Design and synthesis of a novel class of sugar-peptide hybrids: C-linked glyco β-amino acids through a stereoselective "acetate" Mannich reaction as the key strategic element

Article abstract of DOI:10.1021/ja026250s

A new type of sugar-amino acid hybrid, which is comprised of a sugar unit (gluco-, galacto-, or mannopyranose) linked through a C-glycosidic linkage to the β-position of an α-unsubstituted β-amino acid unit, is presented. It is hypothesized that these new compounds, or the oligomeric peptides derived therefrom, might possess the structural features of β-amino acid oligomers and the chemical and enzymatic resistance of C-glycosides to hydrolysis. The synthetic strategy is based on a new Mannich-type reaction between a chiral acetate enolate equivalent and α-amido sulfones derived from the corresponding sugar-C-glycoside aldehydes. While the sugar-C-glycoside aldehyde partner is prepared from well-established transformations on known sugar precursors, the lithium enolate derived from (1 R)-endo-2-acetylisoborneol 3 is employed as the key element. This Mannich approach proceeds with essentially perfect diasteromeric control leading to the new β-amino carbonyl adducts in good yields. Further, cleavage of the camphor auxiliary is smoothly performed by oxidative treatment with ammonium cerium nitrate (CAN). Complementarily, direct peptide-type coupling of the β-amino carbonyl Mannich adducts with an α- or β-amino acid residue and subsequent CAN-promoted detachment of the auxiliary yields dipeptide fragments bearing a sugar-containing aliphatic side chain and is a process that can be iterated. A preliminary conformational study based on the combination of experimental NMR data and molecular mechanics and molecular dynamics (MD) of one particular adduct is also provided.

Full text of DOI:10.1021/ja026250s

Published on Web 06/26/2002
Design and Synthesis of a Novel Class of Sugar-Peptide
Hybrids: C-Linked Glyco â-Amino Acids through a
Stereoselective “Acetate” Mannich Reaction as the Key
Strategic Element
Claudio Palomo,*,† Mikel Oiarbide,† Aitor Landa,† M. Concepcio´n Gonza´lez-Rego,†
Jesu´s M. Garc´ıa,‡ Alberto Gonza´lez,‡ Jose´ M. Odriozola,‡
Manuel Mart´ın-Pastor,§ and Anthony Linden|
Contribution from the Departamento de Qu´ımica Orga´nica I, Facultad de Qu´ımica,
UniVersidad del Pa´ıs Vasco, Apdo. 1072, 20080 San Sebastia´n, Spain;
Departamento de Qu´ımica Aplicada, UniVersidad Pu´blica de NaVarra, Campus de Arrosad´ıa,
31006 Pamplona, Spain; Unidade de Resonancia Magnetica, R.I.A.I.D.T., CACTUS,
UniVersidad de Santiago de Compostela, 15782 A Coruna, Spain; and Organisch-chemisches
Institut der UniVersita¨t Zu¨rich, Winterthurerstrasse 190, CH-8057, Zu¨rich, Switzerland
Received March 20, 2002
Abstract: A new type of sugar-amino acid hybrid, which is comprised of a sugar unit (gluco-, galacto-, or
mannopyranose) linked through a C-glycosidic linkage to the â-position of an R-unsubstituted â-amino
acid unit, is presented. It is hypothesized that these new compounds, or the oligomeric peptides derived
therefrom, might possess the structural features of â-amino acid oligomers and the chemical and enzymatic
resistance of C-glycosides to hydrolysis. The synthetic strategy is based on a new Mannich-type reaction
between a chiral acetate enolate equivalent and R-amido sulfones derived from the corresponding sugar-
C-glycoside aldehydes. While the sugar-C-glycoside aldehyde partner is prepared from well-established
transformations on known sugar precursors, the lithium enolate derived from (1R)-endo-2-acetylisoborneol
3 is employed as the key element. This Mannich approach proceeds with essentially perfect diasteromeric
control leading to the new â-amino carbonyl adducts in good yields. Further, cleavage of the camphor
auxiliary is smoothly performed by oxidative treatment with ammonium cerium nitrate (CAN). Complemen-
tarily, direct peptide-type coupling of the â-amino carbonyl Mannich adducts with an R- or â-amino acid
residue and subsequent CAN-promoted detachment of the auxiliary yields dipeptide fragments bearing a
sugar-containing aliphatic side chain and is a process that can be iterated. A preliminary conformational
study based on the combination of experimental NMR data and molecular mechanics and molecular
dynamics (MD) of one particular adduct is also provided.
Introduction and Objectives
the peptidyl substructures usually have a distinct, often complex
function.2 The limited availability of these structurally complex
compounds is an important obstacle that restrains the study of
their structural features and mode of action, because biosynthetic
approaches to them are lacking, while the chemical synthesis3
is, at present, a time-consuming and encumbered process.4
Remarkably, however, the fact is that less complex, and
therefore easier to access, mimetic compounds may display a
biological activity level similar to that of their native counter-
parts. In the case of glycopeptides, shorter oligomers have been
Carbohydrates and amino acids appear in all cells in some
form or another, and sometimes they merge in the same
molecule as happens in proteoglycans, glycoproteins, and
glycopeptides.1 These types of compounds have been implicated
in fundamental biological processes, in which the glycan and
* To whom correspondence should be addressed. E-mail: qoppanic@
sc.ehu.es.
† Universidad del Pa´ıs Vasco.
‡ Universidad Pu´blica de Navarra.
§ Universidad de Santiago de Compostela (Conformational analysis).
(2) (a) Williams, D. H. Nat. Prod. Rep. 1996, 13, 469-478. (b) Sun, B.; Chen,
Z.; Eggert, U. S.; Shaw, S. J.; LaTour, J. V.; Kahne, D. J. Am. Chem. Soc.
2001, 123, 12722-12723 and references therein.
| Universita¨t Zu¨rich (X-ray analysis).
(1) (a) Glycopetides and Related Compounds; Large, D. G., Warren, C. D.,
Eds.; Marcel Dekker: New York, 1997. Glycobiology and carbohydrate
recognition, see: (b) Kobata, A. Acc. Chem. Res. 1993, 26, 319-324. (c)
Dwek, R. A. Chem. ReV. 1996, 96, 683-720. (d) Davis, A. P.; Wareham,
R. S. Angew. Chem., Int. Ed. 1999, 38, 2978-2996. (e) Chem. ReV. 2002,
102, whole issue No 2. Carbohydrate implications in chemistry and
biology: (f) Nicolaou, K. C.; Mitchel, H. J. Angew. Chem., Int. Ed. 2001,
40, 1576-1624. Glycopeptide and carbohydrate antibiotics: (g) Nicolaou,
K. C.; Boddy, C. N. C.; Bra¨se, S.; Winsingen, N. Angew. Chem., Int. Ed.
1999, 38, 2096-2152. (h) Ritter, T. K.; Wong, C.-H. Angew. Chem., Int.
Ed. 2001, 40, 3508-3533.
(3) Glycopeptide synthesis: (a) Taylor, C. M. Tetrahedron 1998, 54, 11317-
11362. (b) Herzner, H.; Reipen, T.; Schultz, M.; Kunz, H. Chem. ReV.
2000, 100, 4495-4537. Glycoprotein synthesis: (c) Davis, B. G.; Jones,
J. B. Synlett 1999, 1495-1507. (d) Davis, B. G. Chem. ReV. 2002, 102,
579-601. Synthesis and structure of the carbohydrate core of glycopro-
teins: (e) Paulsen, H. Angew. Chem., Int. Ed. Engl. 1990, 29, 823-839.
(4) For a discussion on this subject and new directions in glycopeptide mimetics,
see: (a) Marcaurelle, L. A.; Bertozzi, C. R. Chem.-Eur. J. 1999, 5, 1384-
1390. (b) Hang, H. C.; Bertozzi, C. R. Acc. Chem. Res. 2001, 34, 727-
736.
9
10.1021/ja026250s CCC: $22.00 © 2002 American Chemical Society
J. AM. CHEM. SOC. 2002, 124, 8637-8643
8637

A R T I C L E S
Palomo et al.
Figure 2. Main features of the synthetic strategy.
C-glycosides; (2) the stereochemistry during the process of C-C
bond formation at both the anomeric and the C-â positions must
be fully controlled; and (3) it would be desirable to obtain
monomeric C-glycosyl â-amino acid units that are directly
amenable to a further peptide-type coupling step, while keeping
the functional group protection/deprotection events at a mini-
mum.
Figure 1. Prototypical sugar-peptide links in native (I and II) and mimetic
(III, IV, and V) glycopetides. The new sugar-amino acids pattern VI
reported herein is framed.
To adhere well to the above requirements, we envisaged the
strategy depicted in Figure 2. The key element of the route is
the Mannich reaction of a chiral acetate enolate or equivalent
with the corresponding azomethine function.11 This reaction,
when compared with other approaches to â-amino carbonyls,
inter alia â-amino acids,12 makes it possible to create the C-C
bond and the new stereogenic center(s) in a single synthetic
operation. Despite this synthetic potential11 and the recent
advances in the area of diastereoselective,13 enantioselective,14
and direct15 methods, there are two inherent problems that
hamper the use of the Mannich reaction to increase the pool of
available â-amino acids: the tendency of enolizable aldehyde-
derived azomethines to undergo R-deprotonation rather than
addition,16 and the lack of efficient stereocontrol during the C-C
bond formation process, especially in reactions involving acetate
documented to have significant activity.5 In addition, some
isosteres have demonstrated greater stability6 and bioavailabil-
ity,7 thus increasing the chances for potential therapeutic use.
Of interest are the new glycan-peptide links designed to
substitute the chemically and enzymatically labile O- or
N-glycosyl bond found in native glycopeptides (Figure 1).8
On the other hand, structures comprised of a â-amino acid
oligomeric backbone, that is, â-peptides, have emerged as
promising tools for extending the understanding of protein
structure and stabilization into the realm of folded, nonbiological
polymers.9 Like R-peptides, â-peptides have been shown to fold
into helices, sheets, and turns which are the main structural
elements of proteins, and, in some instances, it has been found
that they possess even higher biological activity than their parent
R-peptides.10 Additionally, â-peptides show greater resistance
to peptidases and proteases, a property that is of interest in
pharmaceutical drug development.9,10 From these precedents,
it could be anticipated that glycoconjugates comprised of a sugar
and a â-peptide unit, both linked through a C-glycosyl bond,
may constitute interesting targets for study. Here we report the
first synthetic approach to this new type of glycoconjugates VI
which relies on a stereoselective Mannich reaction as the key
strategic element.
(11) (a) Mannich, C.; Kro¨sche, W. Arch. Pharm. (Weinheim, Ger.) 1912, 250,
647. For a recent review, see: (b) Arend, M.; Westermann, B.; Risch, N.
Angew. Chem., Int. Ed. 1998, 37, 1044-1070.
(12) For reviews on â-amino acids, see: (a) Cole, D. C. Tetrahedron 1994, 50,
9517-9582. (b) Cardillo, G.; Tomassini, C. Chem. Soc. ReV. 1996, 117-
128. (c) EnantioselectiVe Synthesis of â-Amino Acids; Juaristi, E., Ed.;
Wiley-VCH: Weinheim, New York, 1997. (d) Abdel-Magid, A. F.; Cohen,
J. H.; Maryanoff, C. A. Curr. Med. Chem. 1999, 6, 955-970. (e) Juaristi,
E.; Lo´pez-Ruiz, H. Curr. Med. Chem. 1999, 6, 983-1004.
(13) From trialkylsilyl ketene acetals: (a) Kunz, H.; Burgard, A.; Schanzenbach,
D. Angew. Chem., Int. Ed. Engl. 1997, 36, 386-387. (b) Guenoun, F.; Zair,
T.; Lamaty, F.; Pierrot, M.; Lazaro, R.; Viallefont, P. Tetrahedron Lett.
1997, 38, 1563-1566. (c) Higashiyama, K.; Kyo, H.; Takahashi, H. Synlett
1998, 489-494. (d) Mu¨ller, R.; Ro¨ttele, H.; Henke, H.; Waldmann, H.
Chem.-Eur. J. 2000, 6, 2032-2043. (e) Enders, D.; Oberbu¨rsch, S.; Adam,
J. Synlett 2000, 644-646. (f) Allef, P.; Kunz, H. Tetrahedron: Asymmetry
2000, 11, 375-378. (g) Vicario, J. L.; Bad´ıa, D.; Carrillo, L. Org. Lett.
2001, 3, 773-776. (h) McLaren, A. B.; Sweeney, J. B. Synlett 2000, 1625-
1627. From enolates of N-acyl oxazolidinones and related systems: (i)
Abrahams, I.; Motevalli, M.; Robincon, A. J.; Wyatt, P. B. Tetrahedron
1994, 50, 12755-12772. (j) Roos, G. H. P.; Balasubramaniam, S. Synth.
Commun. 1999, 29, 755-762. (k) Kawakami, T.; Ohtake, H.; Arakawa,
H.; Okachi, T.; Imada, Y.; Murahashi, S.-I. Chem. Lett. 1999, 795-796.
(14) Review: (a) Kobayashi, S.; Ishitani, H. Chem. ReV. 1999, 99, 1069-1094.
Leading examples: (b) Ishitani, H.; Ueno, M.; Kobayashi, S. J. Am. Chem.
Soc. 1997, 119, 7153-7154. (c) Kobayashi, S.; Ishitani, H.; Ueno, M. J.
Am. Chem. Soc. 1998, 120, 431-432. (d) Ishitani, H.; Ueno, M.; Kobayashi,
S. J. Am. Chem. Soc. 2000, 122, 8180-8186. (e) Fujii, A.; Hagiwara, E.;
Sodeoka, M. J. Am. Chem. Soc. 1999, 121, 5450-5458. (f) Hagiwara, E.;
Fujii, A.; Sodeoka, M. J. Am. Chem. Soc. 1998, 120, 2474-2475. (g)
Ferraris, D.; Young, B.; Dudding, T.; Lectka, T. J. Am. Chem. Soc. 1998,
120, 4548-4549. (h) Ferraris, D.; Young, B.; Cox, C.; Drury, W. J., III;
Dudding, T.; Lectka, T. J. Org. Chem. 1998, 63, 6090-6091. (i) Ferraris,
D.; Dudding, T.; Young, B.; Drury, W. J., III; Lectka, T. J. Org. Chem.
1999, 64, 2168-2169. (j) Xue, S.; Yu, S.; Deng, Y.; Wulff, N. D. Angew.
Chem., Int. Ed. 2001, 40, 2271-2274. (k) Juhl, K.; Gathergood, N.;
Jørgensen, K. A. Angew. Chem, Int. Ed. 2001, 40, 2995-2997.
Results and Discussion
Synthetic Plan and Challenges. A satisfactory approach to
C-glyco â-amino acids has at least three requirements: (1) it
must be chemically efficient, starting from readily affordable
(5) For some relevant examples, see: (a) Bertozzi, C. R.; Cook, D. G.; Kobertz,
W. R.; Gonza´lez-Scarano, F.; Bednarski, M. D. J. Am. Chem. Soc. 1992,
114, 10639-10641. (b) Nagy, J. O.; Wang, P.; Gilbert, J. H.; Schaefer, M.
E.; Hill, T. G.; Callstrom, M. R.; Bednarski, M. D. J. Med. Chem. 1992,
35, 4501-4502. (c) Kishi, Y. Pure Appl. Chem. 1993, 65, 771-778. (d)
Marron, G.; Woltering, T. J.; Waitz-Schmidt, G.; Wong, C.-H. Tetrahedron
Lett. 1996, 37, 9037-9040. (e) Wang, L.-X.; Tang, M.; Suzuki, T.;
Kitajima, K.; Inoue, Y.; Inoue, S.; Fan, J.-Q.; Lee, Y. C. J. Am. Chem.
Soc. 1997, 119, 11137-11146. (f) Marcaurelle, L. A.; Rodriguez, E. C.;
Bertozzi, C. R. Tetrahedron Lett. 1998, 39, 8417-8420.
(6) For a comparative study, see: Sparks, M. A.; Williams, K. W.; Whitesides,
G. M. J. Med. Chem. 1993, 36, 778-783.
(7) (a) Fisher, J. F.; Harrison, A. W.; Bundy, G. L.; Wilkinson, K. F.; Rush,
B. D.; Ruwart, M. J. J. Med. Chem. 1991, 34, 3140-3143. (b) Mehta, S.;
Meldal, M.; Duus, J. O.; Bock, K. J. Chem. Soc., Perkin Trans. 1 1999,
1445-1451.
(15) (a) Yamasaki, S.; Iida, T.; Shibasaki, M. Tetrahedron Lett. 1999, 40, 307-
310. (b) Yamasaki, S.; Iida, T.; Shibasaki, M. Tetrahedron 1999, 55, 8857-
8867. (c) List, B. J. Am. Chem. Soc. 2000, 122, 9336-9337. (d) List, B.;
Pojorliev, P.; Biller, W. T.; Martin, H. J. J. Am. Chem. Soc. 2002, 124,
827-833. (e) Notz, W.; Sakthivel, K.; Bui, T.; Zhong, G.; Barbas, C. F.,
III. Tetrahedron Lett. 2001, 42, 199-201. (f) Co´rdova, A.; Notz, W.; Zhong,
G.; Betancort, J. M.; Barbas, C. F., III. J. Am. Chem. Soc. 2002, 124, 1842-
1843. (g) Co´rdova, A.; Watanabe, S.-i.; Tanaka, F.; Notz, W.; Barbas, C.
F., III. J. Am. Chem. Soc. 2002, 124, 1866-1867.
(8) Schweizer, F. Angew. Chem., Int. Ed. 2002, 41, 230-253.
(9) Cheng, R. P.; Gellman, S. H.; DeGrado, W. F. Chem. ReV. 2001, 101,
3219-3232.
(10) For a detailed information on these subjects, see: (a) Seebach, D.; Mathews,
J. L. Chem. Commun. 1997, 2015-2022. (b) Koert, U. Angew. Chem., Int.
Ed. Engl. 1997, 36, 1836-1837. (c) Gellman, S. H. Acc. Chem. Res. 1998,
31, 173-180. (d) Gademann, K.; Hintermann, T.; Schreiber, J. V. Curr.
Med. Chem. 1999, 6, 905-925.
9
8638 J. AM. CHEM. SOC. VOL. 124, NO. 29, 2002

Design of a Novel Class of Sugar-Peptide Hybrids
A R T I C L E S
Scheme 1. Preparation of Reagent 4 from (1R)-(+)-Camphor and
Acetylene as the Elementary Source of the Carbonyl(acetyl) Group
Scheme 2. Asymmetric Mannich Reaction with R-Amido Alkyl
Sulfones
Scheme 3. Preparation of R-Amido Glycoalkyl Sulfones 13/14
from C-Glycosides 10
enolate equivalents.17 Yet, such transformations have tradition-
ally been among the most challenging to accomplish.18
Background. A prior report from these laboratories docu-
mented the efficiency of the lithium enolate of methyl ketone
4, Scheme 1, in successfully addressing the problem of the
insufficient stereoselectivity often encountered in the “acetate”
aldol addition reaction.19 On this basis, it seemed reasonable to
us that the same enolate system may be equally effective, in
terms of stereoselectivity, against azomethine electrophiles.
Importantly, the synthesis of the required methyl ketone 4 from
(1R)-(+)-camphor 1 and acetylene, two commodity chemicals
available in bulk, is straightforward and high yielding. Likewise,
5 is also available from (1S)-(-)-camphor in the same way.
At the outset of our investigation, however, it was not evident
which kind of azomethines would be best suited for this
Mannich reaction.20,21 In addition to the inherent problem
associated with the ease of isomerization of enolizable aldehyde-
derived azomethines, there is the fact that lithium enolates of
acetate are rather inert toward Schiff bases such as N-aryl and
N-trimethylsilylaldimines of benzaldehyde.22 We have addressed
this issue and found that, Scheme 2, R-amido sulfones 6/7,
generated from both enolizable and nonenolizable aldehydes,
sodium p-toluenesulfinate, and benzyl or tert-butyl carbamate,
upon reaction with the lithium enolate of 4, in the presence of
a 1 equiv excess of LDA, provided the Mannich adducts 8/9
with good yields and essentially complete diastereoselectivity.23
Concurrent with our investigation, Petrini and co-workers have
also demonstrated that these R-amido sulfones are capable of
reacting with enolsilanes to give the expected â-amino carbonyl
adducts in their racemic form.24 In a related investigation by
Kise and Ueda, it has been shown that N-alkoxycarbonyl-1-
methoxyamines, upon reaction with metal enolates of chiral
carboxylic acid derivatives, provide â-amino esters, albeit in
moderate levels of diastereoselectivity.25
In the present study, the glucose, galactose, and mannose
derivatives 10a-c,26 Scheme 3, were selected for development,
owing to the ample precedents for their preparation in a highly
stereocontrolled manner.27 The formation of R-amido glycoalkyl
sulfones from these stereochemically defined sugar derivatives
is straightforward. Thus, the hydroboration of the double bond28
in each compound 10a-c, followed by oxidation of the
respective intermediate borane,29 leads to high yields of crystal-
line, easy to purify products 11a-c. Further Swern oxidation
of the primary alcohol to the corresponding aldehyde30 12a-c
and reaction of each one with benzyl carbamate and sodium
p-toluenesulfinate afforded the R-amido glycoalkyl sulfones
13a-c in moderate to good yields. Similarly, compound 14a
(16) For pertinent information on this problem, see: (a) Risch, N.; Arend, M.
In StereoselectiVe Synthesis (Houben-Weyl); Helmchen, G., Hoffmann, R.
W., Mulzer, J., Schaumann, E., Eds.; Thieme: Stuttgart, 1996; Vol. E2/
13, p 1833. (b) Denmark, S. E.; Nicaise, O. J. C. Chem. Commun. 1996,
999-1004.
(17) For exceptions, see: (a) Liebeskind, L. S.; Welker, M. E.; Fengl, R. W. J.
Am. Chem. Soc. 1986, 108, 6328-6343. (b) Saito, S.; Hatanaka, K.;
Yamamoto, H. Org. Lett. 2000, 2, 1891-1894. (c) Saito, S.; Hatanaka,
K.; Yamamoto, H. Tetrahedron 2001, 57, 875-887.
(18) For detailed information on the stereochemical problem of R-unsubstituted
enolates, see: (a) Braun, M. In StereoselectiVe Synthesis (Houben-Weyl);
Helmchen, G., Hoffmann, R. W., Mulzer, J., Schaumann, E., Eds.;
Thieme: Stuttgart, 1996; Vol. E2/13, p 133. (b) Braun, M. Angew. Chem.,
Int. Ed. Engl. 1987, 26, 24-37. (c) Braun, M.; Sacha, H. J. Prakt. Chem.
1993, 335, 653-668.
(19) (a) Palomo, C.; Gonza´lez, A.; Garc´ıa, J. M.; Landa, C.; Oiarbide, M.;
Rodr´ıguez, S.; Linden, A. Angew. Chem., Int. Ed. 1998, 37, 180-182. (b)
Palomo, C.; Oiarbide, M.; Aizpurua, J. M.; Gonza´lez, A.; Garc´ıa, J. M.;
Landa, C.; Odriozola, I.; Linden, A. J. Org. Chem. 1999, 64, 8193-8200.
(20) For reviews, see: (a) Gennari, C. In ComprehensiVe Organic Synthesis;
Trost, B. M., Fleming, I., Eds.; Pergamon: Oxford, 1991; Vol. 2, p 629.
(b) Kleinman, E. F. In ComprehensiVe Organic Synthesis; Trost, B. M.,
Fleming, I., Eds.; Pergamon: Oxford, 1991; Vol. 2, p 893.
(23) (a) Palomo, C.; Oiarbide, M.; Gonza´lez-Rego, M. C.; Sharma, A. K.; Garc´ıa,
J. M.; Gonza´lez, A.; Landa, C.; Linden, A. Angew. Chem., Int. Ed. 2000,
39, 1063-1065. Very recently, Enders has reported the reaction of lithium
enolates of ketones with polymer-bound and carbamate-linked R-amidoalkyl
sulfones: (b) Schunk, S.; Enders, D. Org. Lett. 2001, 3, 3177-3180.
(24) (a) Ballini, R.; Petrini, M. Tetrahedron Lett. 1999, 40, 4449-4452. (b)
Mecozzi, T.; Petrini, M. J. Org. Chem. 1999, 64, 8970-8972. (c) Mecozzi,
T.; Petrini, M. Tetrahedron Lett. 2000, 41, 2709-2712. (d) Giardina, A.;
Mecozzi, T.; Petrini, P. J. Org. Chem. 2000, 65, 8277-8282.
(25) (a) Kise, N.; Ueda, N. J. Org. Chem. 1999, 64, 7511-7514. (b) Kise, N.;
Ueda, N. Org. Lett. 1999, 1, 1803-1805.
(21) For examples on the asymmetric Mannich reaction involving the Eschen-
moser salt and related nonenolizable methaneiminium compounds, see: (a)
Evans, D. A.; Urpi, F.; Somers, T. L.; Clark, J. S.; Bilodeau, M. T. J. Am.
Chem. Soc. 1990, 112, 8215-8216. (b) Page, P. C. B.; Allin, S. M.;
Collington, E. W.; Carr, R. A. E. J. Org. Chem. 1993, 58, 6902-6904. (c)
D’Souza, A. A.; Montevalli, M.; Robinson, A. J.; Wyatt, P. B. J. Chem.
Soc., Perkin Trans. 1 1995, 1-2. (d) Enders, D.; Ward, D.; Adam, J.; Raabe,
G. Angew. Chem., Int. Ed. Engl. 1996, 35, 981-984. (e) Vinkovic, V.;
Sunjic, V. Tetrahedron 1997, 53, 689-696.
(26) Gartzen, D.; Misske, A. M.; Wolbers, P.; Hoffmann, H. M. R. Tetrahedron
Lett. 1999, 40, 6359-6363.
(27) For reviews, see: (a) Postema, M. H. D. C-Glycoside Synthesis; CRC: Boca
Raton, 1995. (b) Du, Y.; Linhardt, R. J. Tetrahedron 1998, 54, 9913-
9959. (c) Dondoni, A.; Marra, A. Chem. ReV. 2000, 100, 4395-4421.
(28) Brown, H. C.; Knights, E. F.; Scouten, C. G. J. Am. Chem. Soc. 1974, 96,
7765-7770.
(22) (a) Gluchowski, C.; Cooper, L.; Bergbreiter, D. E.; Newcomb, M. J. Org.
Chem. 1980, 45, 3413-3416. (b) Ha, D.-C.; Hart, D. J.; Yang, T.-K. J.
Am. Chem. Soc. 1984, 106, 4819-4825.
(29) Lane, C. F. J. Org. Chem. 1974, 39, 1437-1438.
(30) Palomo, C.; Coss´ıo, F. P.; Ontoria, J. M.; Odriozola, J. M. J. Org. Chem.
1991, 56, 5948-5951.
9
J. AM. CHEM. SOC. VOL. 124, NO. 29, 2002 8639

A R T I C L E S
Palomo et al.
Scheme 5. Strategic Combination of the Asymmetric Mannich
Scheme 4. Asymmetric Mannich Reaction for the Synthesis of
C-Glycoalkyl â-Amino Acids
Reaction with a Peptide Coupling Process Leading to Either
â-Peptides or R,â-Peptidesa
a (a) (i) 6 (R1, Cbz; R2, Et) LDA, THF, -78 °C, 15 min, 94%; (ii) H2,
Pd/C (10% w/w), EtAcO, room temperature, 1 h, 98%. (b) (S)-BocPheOH,
EDC, HOBT, CH2Cl2, 97%. (c) TBAF, THF, room temperature, 20 min,
then (NH4)2Ce(NO3)6, CH3CN-H2O, room temperature, 5 min, 90-99%.
(d) 24, EDC, HOBT, CH2Cl2, 95%.
was accessed by reacting 12 with tert-butyl carbamate. In every
case, the R-amido glycoalkyl sulfones were obtained as mixtures
of the corresponding epimers at the newly created stereocenter.
Mannich Reaction. Given the precedent from these labora-
tories, it was expected that the reaction of the lithium enolate
of 4 with these R-amidoglycoalkyl sulfones, Scheme 4, would
generate the corresponding â-amino ketones under an analogous
stereochemically controlled event. Concordant with our expecta-
tions, it was found that the reaction of the lithium enolate of 4
with R-amidoglycoalkyl sulfones 13a-c proceeds to give
Mannich adducts 15a-c in yields of 78, 75, and 60%, respec-
tively. Likewise, the reaction of the lithium enolate of 4 with
14a provided 17a in 70% yield. Most notably, in each case,
essentially only one isomer product was formed, as judged by
both HPLC and 13C NMR analysis of the respective crude
reaction mixture. For stereochemical correlation purposes, the
adduct 15a was selectively N-deprotected31 to give the amino
derivative 16 which was identical to that obtained from the
N-Boc deprotection in 17a.
From the virtually perfect stereocontrol observed for this
Mannich reaction, which involves two chiral reactants, it seemed
of interest to test whether this result corresponds to a matched
combination of the reactants and whether the mismatched
combination would lead to poorer results. With that aim, the
reaction of methyl ketone 5, the latter derived from (-)-
camphor, with 13a was carried out which afforded adduct 20,
again, as the sole isomer. Therefore, the stereochemistry of this
process seems to be fully controlled by the chiral lithium enolate
in both possible combinations, and each one of the four possible
diastereomeric C-glyco â-amino acids (two from the D-sugar
series and two from the L-series) is virtually affordable with
essentially perfect stereocontrol.
Likewise, 20 afforded 21 along with (1S)-(-)-camphor. The
only case in which this oxidative cleavage failed was for
compound 17a. Apparently, the reaction conditions are too harsh
for the N-Boc group34 (vide infra). Separation of each compo-
nent from the mixture was easily effected by basic aqueous
extractions. Alternatively, (1R)-(+)-camphor or its enantiomer
can also be fully extracted from the mixture by washing the
crude products with hexane, while the carboxylic acid, with the
exception of a very small amount in some cases, remains behind.
With the purpose of assessing the stereoisomeric purity of
the resulting C-glyco â-amino acids, all of the compounds 18
and 21 were transformed into their methyl esters 19 and 22,
respectively, and the stereoisomeric composition of each one
was tested by comparison of their HPLC chromatograms with
those of the epimeric samples 23a-c, the latter obtained by
the reaction of the corresponding R-amido glycoalkyl sulfones
13a-c with the lithium enolate of methyl acetate. Such tests
demonstrated diastereomeric purities higher than 98%.
Peptide Synthesis. The excellent diastereoselectivity attained
in the present Mannich reaction is of particular interest in that
it provides, through N-deprotection and subsequent N-acylation,
a concise route to peptide products. To illustrate this latter
aspect, we first elected to use the structurally simple Mannich
adduct 24, Scheme 5, which bears the nonproteinogenic ethyl
side chain. This adduct upon acylation with (L)-Boc-N-phen-
ylalanine under standard peptide conditions afforded the N-
acylated product 25 in 97% yield. The smooth release of the
camphor moiety in adduct 25 by treatment with CAN proceeded
without any byproduct formation to give the R,â-dipeptide 26
in almost quantitative yield. From this result, it was easy to
predict that the same process could be realized in an iterative
fashion. For example, the R,â-dipeptide 26, upon coupling with
24, affords the product 27 in 95% yield, from which the R,â,â-
tripeptide 28 is produced in 90% yield along with the recovery
of camphor in 87% yield. It is worth noting that in these
Concurrently, it was found that the oxidative cleavage of the
acyloin moiety in Mannich adducts 15a-c with ammonium
cerium nitrate (CAN)32 provided â-amino carboxylic acids
18a-c and (1R)-(+)-camphor33 as the only detectable products.
(31) (a) Sajiki, H. Tetrahedron Lett. 1995, 36, 3465-3468. (b) Sajiki, H.; Kuno,
H.; Hirota, K. Tetrahedron Lett. 1998, 39, 7127-7130.
(32) Review: Ho, T.-L. Synthesis 1973, 347-354.
(34) Under certain reaction conditions, deprotection of N-tert-butoxycarbonyl-
amines to the free amine by action of CAN has been reported: Hwu, J. R.;
Jain, M. L.; Tsay, S.-C.; Hakimelahi, G. H. Tetrahedron Lett. 1996, 37,
2035-2038.
25
(33) In every case, the starting (1R)-(+)-camphor (Aldrich [R]D ) +42.2 (c
25
) 1, EtOH)) was easily recovered (recovered material, [R]D ) +41.5/
+42.0 (c ) 1, EtOH)) in yields of 85-90% by simple aqueous workup.
9
8640 J. AM. CHEM. SOC. VOL. 124, NO. 29, 2002

Design of a Novel Class of Sugar-Peptide Hybrids
A R T I C L E S
Scheme 6. Synthesis of C-Glyco â-Peptides
Figure 3. An ORTEP37 drawing of the molecular structure of one of the
symmetry-independent molecules of compound 35 (50% probability el-
lipsoids).
â-amino acid units, and the essentially complete control of the
products configuration are some of the features of the present
model. From this model, not only the neoglycopolymer total
length but also the intersugar distance and the main chain
conformation can virtually be modulated.
Determination of the Configuration of the Products. The
configuration of the adducts was established by a single-crystal
X-ray structure analysis of compound 35 (see Figure 3) and by
assuming a uniform reaction mechanism. To date, a general
survey of azomethines derived from either enolizable or
nonenolizable aldehydes with respect to the generality of this
assumption has been gratifying, and diastereoselective Mannich
reactions with this methyl ketone appear to proceed with
diastereoselection at the 95-98% levels.23a
instances the scission promoted by CAN is tolerant of both the
Boc protecting group and the amide function,35 leading to
chemically almost pure products. Most notably, this realization
implies a dual role for camphor: on one hand, it controls the
stereochemistry of the carbon-carbon bond forming process,
and, on the other hand, it is an efficient carboxylic acid
protecting group (masked in the form of an acyloin moiety). In
this way, concomitant peptide synthesis can be approached with
the required protection/deprotection steps kept to a minimum.
In our and other’s experience in the field of sugar derivatives,
however, it is not always easy to obtain crystalline samples of
sufficient quality for single-crystal X-ray structure analysis.
Therefore, we felt it would be of practical convenience to
establish some alternative methodology for the assignment of
the configuration which would not necessarily depend on any
fanciful property, such as crystallinity. In this respect, the
protocol developed by Riguera for the assignment of the absolute
configuration of chiral amines by NMR spectroscopy38 seemed
attractive. The method relies on the different chemical shift
values corresponding to diastereomeric compounds obtained by
double derivatization of a given substrate with both (R)- and
(S)-methoxyphenylacetic acid chloride. Accordingly, the cor-
responding (R)- and (S)-methoxyphenylacetamides (MPA) of
To gain further insight into the scope of the present model,
the synthesis of peptide chains incorporating C-glyco â-amino
acids was evaluated next. For example, the Mannich adduct 16a,
upon treatment with (L)-N-Boc-phenylalanine under standard
peptide coupling conditions, furnished compound 29 in 85%
yield, Scheme 6. Similarly, the coupling reaction of the adduct
16a with the C-glyco-â-amino acid 18a afforded the product
33 in 82% yield. Subsequent acyloin cleavage in both 29 and
33 provided the R,â-dipeptide 30 and the â,â-dipeptide product
34 in 85 and 92% yields, respectively. In addition, 30 was
transformed into the hydroxyl-free sugar derivative 32 for
preliminary conformational studies. Consequently, this develop-
ment opens the way to further iterative processes en route to
the synthesis of a new family of neoglycopolymers,36 in which
the linear backbone is comprised of a â-peptide chain. The
wealth of peptide synthesis techniques, the ease of modulation
of the peptide chain by incorporating into it new either R- or
1
compound 16a were prepared and analyzed by H NMR. As
shown in Figure 4, a consistent positive ∆δRS value for the
protons at the R site was observed,39 while a negative ∆δRS
value was found for the protons situated at the R′, â′, and γ′
positions. From these observations, and on the basis of the
theoretical and empirical rules set by Riguera, a disposition of
the L1 and L2 portions of the amine can be inferred, which is in
accordance with the assumed S configuration of the adducts.
Preliminary Conformational Studies. In addition to the
observations noted above, it would be instructive to get some
structural insight into this new class of â-amino acids, in
particular with regard to some relevant parameters such as the
(35) Catalytic hydrolysis of R-peptides by the action of CAN at reaction
temperatures of about 50 °C has been reported recently: Takarada, T.;
Yashiro, M.; Komiyama, M. Chem.-Eur. J. 2000, 6, 3906-3913.
(36) For a ring-opening metathesis polymerization strategy to linear neoglyco-
polymers and the biological implications of these compounds, see: (a)
Gordon, E. J.; Sanders, W. J.; Kiessling, L. L. Nature 1998, 392, 30-31.
(b) Kanai, M.; Mortell, K. H.; Kiessling, L. L. J. Am. Chem. Soc. 1997,
119, 9931-9932 and references therein.
(37) Johnson, C. K. ORTEPII, Report ORNL-5138, Oak Ridge National
Laboratory, Oak Ridge, Tennessee, 1976.
(38) (a) Lo´pez, B.; Quin˜oa´, E.; Riguera, R. J. Am. Chem. Soc. 1999, 121, 9724-
9725. (b) Seco, J. M.; Quin˜oa´, E.; Riguera, R. Tetrahedron: Asymmetry
2001, 12, 2915-2925.
(39) The unambiguous assignment of peaks from the spectra to the R, R′, â′,
and γ′ protons was made by means of a HETCOR experiment. See
Supporting Information for details.
9
J. AM. CHEM. SOC. VOL. 124, NO. 29, 2002 8641

A R T I C L E S
Palomo et al.
Figure 5. Glycopeptide 32 showing the atom numbering used and the
definition of torsion angles used for the conformational study.
while the Z-Z and E-Z completely failed to explain those
results.
A cluster analysis of the MD results for the Z-E and E-E
trajectories was performed. In each trajectory, the torsion angles
defining the backbone structure of the glycopeptide in Figure 5
were analyzed, and a total of four and three different cluster
conformers were found independently for the Z-E and E-E
conformations, respectively. The molecular mechanics exhaus-
tive optimization of the geometries of each cluster conformation
showed that within a Z-E or E-E cluster, the different
conformers are close in conformational energy, although the
lowest energy conformer of the E-E cluster is 8.9 kJ mol-1
more stable than the lowest energy conformer in the Z-E
cluster. For both the Z-E and the E-E cluster conformations,
the single conformer NOE simulations gave a similarly good
agreement with the experimental NOESY intensities for all the
nonexchangeable protons, other than the four methylene protons
at C-9 and C-10. For all MD simulations, the glucose moiety
was found in the initially assumed 4C1 pyranose chair conforma-
tion during the complete simulation. A further distinction
between the two sets of Z-E and E-E cluster conformers was
made by comparing the NOEs of the prochiral methylene
protons at C-9 and C-10 (Figure 5), for which overlapping
resonances in the spectrum prevented the stereospecific assign-
ment. Only the set of three cluster structures in the E-E peptidic
bond configuration was found to be compatible with these extra
NOEs, and the values of their torsion angles are given in
Table 1.
Because these three conformers have rather similar confor-
mational energies (Table 1) and several transitions among them
were observed during the MD trajectory, these results suggest
that the glycopeptide 32 has the potential flexibility to adopt a
conformational equilibrium among the three cluster structures
in solution. The NOE simulations also suggest that in the
aforementioned conformational equilibrium, conformer C of
Table 1 could represent a major conformation (Supporting
Information Table S2). A stereoview of conformer C is given
in Figure 6. The analysis of the peptidic torsion angles of the
three conformers (Table 1) shows that the φ/ψ values are in a
region of the Ramachandran plot compatible with a â sheet and
an R-helix secondary structure.
Figure 4. Determination of the absolute configuration of the amine product
by chemical shift correlation on MPA-amide derivatives.
torsion dihedral angles around the peptide bond. To this end,
we adopted the procedure of Seebach and Gusteren40 who have
established that the best understanding of the conformational
behavior of a peptide in solution is obtained by a combined
analysis of experimental (NMR) and modeling (molecular
dynamics, MD) data. Accordingly, we have studied the con-
formational properties in solution of the sugar-based â dipeptide
32 by NOESY NMR experiments assisted by molecular
mechanics and MD calculations. The cluster analysis of the
conformations generated during the different MD trajectories
allowed the selection of those relevant different structures which
populate the trajectory. NOESY intensities were simulated for
both the average of each MD trajectory and the single confor-
mations according to the full relaxation matrix approach,41-43
and plausible models were selected by the ability of the model
to reproduce the NOESY experimental data.
Four starting molecular mechanics minimized models of the
glycopeptide were generated for each of the four possible
combinations, Z-Z, Z-E, E-Z, and E-E, for the two peptidic
bonds at atoms N-1′ and N-4′ (Figure 5) with the glucose
4
pyranose chair of the glycopeptide set in the stable C1 con-
formation. The available conformational space around each
minima was explored by a 4 ns long MD calculation, and from
these simulations only the Z-E and E-E combinations pre-
sented interproton distances and MD average simulated NOEs41-43
compatible quantitatively with the short distance cross-peaks
of the NOESY experiment (Supporting Information Table S1),
Conclusions
New sugar-â-amino acid hybrids8,44 have been designed, and
their synthesis has been approached successfully through a new
diastereoselective Mannich reaction model. This model holds
(40) Daura, X.; Gademann, K.; Scha¨fer, H.; Jaun, B.; Seebach, D.; van
Gunsteren, W. F. J. Am. Chem. Soc. 2001, 123, 2393-2404 and references
therein.
(41) Cumming, D. A.; Carver, J. P. Biochemistry 1987, 26, 6664-6676.
(42) Neuhaus, D.; Williamson, M. P. The Nuclear OVerhauser Effect in
Structural and Conformational Analysis, 2nd ed.; Wiley-VCH: New York,
2000.
(44) After the submission of this manuscript, a paper describing sugar-linked
â-amino acid esters appeared: Sharma, G. V. M.; Reddy, V. G.; Chander,
A. S.; Reddy, K. R. Tetrahedron: Asymmetry 2002, 13, 21-24.
(43) Peter, C.; Daura, X.; van Gunsteren, W. F. J. Biomol. NMR 2001, 20, 297-
310.
9
8642 J. AM. CHEM. SOC. VOL. 124, NO. 29, 2002

Design of a Novel Class of Sugar-Peptide Hybrids
A R T I C L E S
Table 1. Cluster Conformations Found in the 4 ns MD Trajectory of Glycopeptide 32 with E-E Peptidic Bond Conformationa
torsion angles (deg)
cluster
conformer
energy
(kJ mol-1
)
R1
R2
R3
R4
R5
R6
R7
φ
ψ
A
B
C
0.0
2.3
3.9
-60.2
-59.0
39.9
-179.5
-176.3
178.6
-61.2
-58.8
-59.6
-61.8
-56.3
-56.9
106.3
105.3
138.1
-178.9
-180.0
-178.5
178.0
179.3
179.4
-84.8
-141.9
-138.7
-20.2
134.3
52.3
a All torsion angles are defined with respect to the four contiguous heavy atoms involved with the lower possible numbering in Figure 5, except R1, which
is defined as H-1-C1-C10′-C9′.
appropriately fitted into the process. As a result, the synthesis
of sugar-peptide hybrids comprised of a â-amino acid backbone
and a glycoalkyl side chain has been made viable for the first
time. This new class of hybrid compounds, which could be
referred to as C-glyco â-peptides, may be of interest in
pharmaceutical drug design. The synthesis of extended C-glyco
â-peptide oligomers and studies on their structural features are
currently underway. In addition, the potential of these new
compounds as glycopeptide mimetics and as multivalent ligands
in cell surface processes will be evaluated next.
Acknowledgment. We thank The University of the Basque
Country and Ministerio de Ciencia y Tecnologia (Spain) for
financial support. A grant to M.C.G.-R. from Gobierno Vasco
is acknowledged.
Figure 6. Stereoview of the best conformation glycopeptide 32 found
consistent with the NOESY experimental data. This conformation corre-
sponds to conformer C in Table 1.
several interesting features: (i) Mannich adducts of acetate with
very high stereochemical purity, perfectly predictable config-
uration, and a broad range of â-substitution patterns are
accessible; (ii) the chiral controller of the process, (1R)-(+)-
camphor, which is used in stoichiometric quantities, is cheap,
almost fully recoverable, and can be reused without loss of
efficiency; (iii) all carbon atoms employed in the entire process,
including acetylene as the source of acetyl, are integrated into
the final products; and (iv) a peptide coupling event can be
Supporting Information Available: Complete experimental
procedures, determination of stereoisomeric mixtures, analytical
and spectral characterization data including copies of spectra
and chromatograms, detailed methods of the molecular modeling
(PDF), and crystallographic data (CIF file). This material is
available free of charge via the Internet at http://pubs.acs.org.
JA026250S
9
J. AM. CHEM. SOC. VOL. 124, NO. 29, 2002 8643