Amino acid fluoride for glycopeptide synthesis

Article abstract of DOI:10.1016/S0040-4039(99)02228-5

The formation of N-glycosidic linkage between N-acetylglucosamine (GlcNAc) and asparagine (Asn) was effected using aspartic acid γ-fluoride in combination with either glycosyl azide or silyl carbamate, by the action of Lindlar catalyst or Bu₄NF. Further elongation of peptide chain was performed to give pentapeptide. This method was further applied into the synthesis of trisaccharidic asparagine, using p-methoxybenzyl assisted stereoselective β- mannosylation as the key transformation. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4039(99)02228-5

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 41 (2000) 1039–1042
Amino acid fluoride for glycopeptide synthesis†
Yukishige Ito,∗ Manfred Gerz and Yoshiaki Nakahara ‡
RIKEN (The Institute of Physical and Chemical Research) and CREST, Japan Science and Technology Corporation (JST),
2-1 Hirosawa, Wako-shi, Saitama 351-0198, Japan
Received 2 September 1999; revised 4 November 1999; accepted 26 November 1999
Abstract
The formation of N-glycosidic linkage between N-acetylglucosamine (GlcNAc) and asparagine (Asn) was
effected using aspartic acid γ-fluoride in combination with either glycosyl azide or silyl carbamate, by the action
of Lindlar catalyst or Bu4NF. Further elongation of peptide chain was performed to give pentapeptide. This method
was further applied into the synthesis of trisaccharidic asparagine, using p-methoxybenzyl assisted stereoselective
β-mannosylation as the key transformation. © 2000 Elsevier Science Ltd. All rights reserved.
Keywords: azides; carbamates; fluorides; glycopeptides.
Amino acid fluorides are powerful and rapid reacting acylating species, which have proved effective
both in solution phase and solid phase peptide synthesis.1 Excellent reactivity of amino acid fluoride is
well underscored in the synthesis of peptibols, which contain multiple numbers of sterically demanding
aminoisobutylic acid residues.2 The use of amino acid fluorides, especially in the field of glycopeptide
synthesis,3 seems promising, because strong activation of the carboxyl component is likely to be required
in order to incorporate sterically demanding carbohydrate carrying components. We report here the first
demonstration of the utility of acid fluorides for the formation of N-linked carbohydrate containing
peptides.
Common to all types of N-linked glycoproteins is the presence of β-configurated N-glycosidic linkage
between asparagine (Asn) and N-acetylglucosamine (GlcNAc) residues. Here, by using aspartic acid γ-
fluoride in combination with either glycosyl azide or silyl carbamate as latent glycosyl amine (Scheme
1) the formation of this particular type of linkage was effected.
GlcNAc derived azide 14 was reacted with properly protected fluorides 2a–c,5 in the presence of
Lindlar catalyst and Me3SiOMe (to trap HF) under H2 (Method A). After chromatographic purification,
corresponding glyco-amino acid products 3a–c were isolated in 88–93% yield (Table 1). This procedure
is compatible with Fmoc/t-Bu, Cbz/Bn, and Boc/t-Bu protected aspartic acid fluorides (entries 1–3).
Besides Lindlar catalyst, 10% Pd–C deactivated by ethylenediamine [Pd(ene)]6 was also effective for
this transformation (entry 4).
∗
Corresponding author. Tel: +81-48-467-9430; fax: +81-48-462-4680; e-mail: yukito@postman.riken.go.jp (Y. Ito)
Dedicated to Prof. Pierre Sinaÿ, on the occasion of his 62nd birthday.
Alternative address: Department of Industrial Chemistry, Tokai University, Hiratsuka-shi, 259-1292, Japan
†
‡
0040-4039/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(99)02228-5
tetl 16156

1040
Scheme 1. Synthesis of N-glycosylated asparagine. (1) Lindlar cat., Me3SiOMe, H2/THF or i-PrOH; (2) Lindlar cat., H2,
(Boc)2O/EtOH, rt, 20 h, 87%; (3) (i-Pr)3SiOTf (1.5 equiv.), DBMP (1.7 equiv.)/CH2Cl2, rt, 40 min, 100%; (4) Bu4NF (0.2
equiv.)/CH2Cl2
Table 1
Synthesis of glyco-amino acid 3 from 1 (Method A) or 5 (Method B) aBased on 1. bPd(ene) was used
in place of Lindlar catalyst. cCalculated based on 5.
As an alternative precursor of glycosylamine, we envisaged Boc protected 4 assuming that it can be
converted to the corresponding silyl carbamate 5 (Method B).7 Coupling of the latter with acid fluoride
was expected to proceed in the presence of a catalytic amount of Bu4NF, with continuous regeneration of
F−. GlcNAc azide 1 was first converted to crystalline Boc derivative 4 (H2, Lindlar cat., Boc2O/EtOH,
33 h, 87%; m.p. 204–205°C) and then into silyl carbamate 5 under modified Ohfune’s conditions8 [(i-
Pr)3SiOTf, 2,6-di-t-butyl-4-methylpyridine (DBMP)/CH2Cl2, rt, 2 h, quantitative]. Subsequent reaction
with fluorides 2a–d was triggered by Bu4NF (0.2 equiv.) to give 3a–d in 78–87% yield (CH2Cl2, rt, 20
h, Table 1, entries 5–8).
Further elongation of peptide linkage was attempted as follows (Fig. 1). Free acids 3f (1.2 equiv.) and
3g (1.3 equiv.), prepared from 3a (33%TFA/CH2Cl2, rt, 1.5 h) and 3d (H2, Pd/C, 1:1 AcOEt–EtOH, rt,
20 h), respectively, were coupled with tripeptide 6b (prepared from Boc protected 6a by treatment with
15% TFA/CH2Cl2, rt, 20 min), based on Carpino’s in situ acid fluoride protocol [tetramethylfluoroform-
amidinium hexefluorophosphate (TFFH), (i-Pr)2NEt/CH2Cl2, rt, 4 h]9 to afford tetrapeptides 7a and 7b
in 67% and 91% yield, respectively. Compound 7b was converted to silyl carbamate 7c [(i-Pr)3SiOTf,
DBMP/CH2Cl2, rt, 2 h, 95%], which was then reacted with valine fluoride 8 (1.5 equiv.) in the presence
of 0.2 equiv. of Bu4NF in CH2Cl2 (rt, 20 h) to afford 9 in 92% yield.

1041
Fig. 1.
In order to ascertain the extent of epimerization during coupling reactions, enantiomeric fluoride 2e
was prepared from Fmoc-D-Asp-O(t-Bu) (Novabiochem) and converted to 3e and 7d, in a similar manner
as described above. These glycopeptides are diastereomeric to 3a and 7a, with respect to the Asn residue.
1
Stereochemical homogeneity of 3a/3e and 7a/7d was confirmed by 400 MHz H NMR, and it was
concluded that the extent of epimerization was negligible.
The applicability of the method to more functionalized systems was further explored using trisac-
charide azide 13 synthesized by using p-methoxybenzyl-assisted intramolecular aglycon delivery.10 β-
Mannoside containing trisaccharide 12 [13C NMR (CDCl3) δ 99.8, 96.8 and 85.5 (anomeric carbons,
1JC–H=162, 164 and 164 Hz, respectively)] was prepared from mannose donor 1010d and disaccharide 11
and transformed into 13. Coupling with fluoride 2a gave 14 in 70% yield (Scheme 2).
In summary, acid fluorides were successfully used for the synthesis of N-glycosylated asparagine
containing glycopeptides. Using either azide or silyl carbamate as a latent amino component, amide
linkages can be formed under essentially neutral conditions (Lindlar cat. −Me3SiOMe or cat. Bu4NF).
1. Typical experimental procedure: Preparation of 3a
Method A: A mixture of compounds 1 (25.1 mg, 0.067 mmol) and 2a (41.0 mg, 0.103 mmol),
Me3SiOMe (40 µl, 0.29 mmol) and Lindlar catalyst (10 mg) in THF (2 ml) was stirred under H2 at
room temperature for 18 h. The mixture was diluted with CHCl3:MeOH (10:1, 10 ml), treated with
aminopropyl silica gel (Sigma, 1.1 mmol/g, 0.10 g, rt, 0.5 h) and insoluble materials removed by
filtration. The filtrate was evaporated in vacuo and the residue was chromatographed over silica gel
(CH2Cl2:acetone 4:1∼1:1) to afford 44.7 mg (90%) of compound 3a: δH (CDCl3, 400 MHz) 6.09 (d,
1H, J=7.9 Hz, NH), 5.95 (d, 1H, J=8.4 Hz, NH), 5.2–5.0 (m, 3H), 4.6–4.0 (m, 7H), 3.74 (ddd, 1H, J=9.6,
4.0, 2.0 Hz, H–5), 2.85 (dd, 1H, J=16.5, 4.5 Hz, CH2CONH), 2.70 (dd, 1H, J=16.5, 4.0 Hz, CH2CONH),
2.07, 2.05, 2.04 and 1.96 (4s, 3H each, Ac), 1.44 (s, 9H, t-Bu); δC (CDCl3, 100 MHz) 172.1, 171.8,
170.9, 170.5, 169.8, 169.1 and 155.9 (CO), 82.2 (C-1), 80.2 (Me3C), 73.5, 72.9, 67.6, 67.1 (CH2), 64.5,
61.7 (CH2), 53.5, 51.0, 47.2, 38.0 (CH2), 28.0, 23.3, 20.8, 20.7.
Method B: To a solution of compounds 5 (32.9 mg, 0.0625 mmol) and 2a (30 mg, 0.075 mmol)
in CH2Cl2 (1.5 ml), was added Bu4NF (0.15 M, 60 µl, 0.009 mmol) and the mixture stirred at room

1042
Scheme 2. Synthesis of trisaccharidic asparagine. (1) DDQ, MS4A/CH2Cl2, rt, 2.5 h; (2) MeOTf, DBMP, MS4A/Cl(CH2)2Cl,
40°C, 22 h, 78% over two steps; (3) H2N(CH2)2NH2/EtOH, reflux, 24 h; (4) Ac2O/MeOH, rt, 2 h, 75% over two steps; (5)
Bu4NF/THF, rt, 45 min., 91%; (6) 2a (1.9 equiv.), Me3SiOMe (4.8 equiv.), Lindlar cat., H2/THF, rt 26 h, 70%
temperature for 20 h. After processing as described for Method A, 37.7 mg (82%) of compound 3a was
obtained.
Acknowledgements
Financial support from the European Union Science and Technology Fellowship Program (to M.G.)
is acknowledged. We thank Dr. Tomoya Ogawa for his encouragement and Ms. Akemi Takahashi for
technical assistance.
References
1. Carpino, L. A.; Beyermann, M.; Wenschuh, H.; Bienert, M. Acc. Chem. Res. 1996, 29, 268–274.
2. Wenschuh, H.; Beyermann, M.; Krause, E.; Brudel, M.; Winter, R.; Schümann, M.; Carpino, L. A.; Bienert, M. J. Org.
Chem. 1994, 59, 3275–3280.
3. Garg, H. G.; von dem Bruch, K.; Kunz, H. Adv. Carbohydr. Chem. Biochem. 1994, 50, 277–310.
4. Szilágyi, L.; Györgydeák, Z. Carbohydr. Res. 1985, 143, 21–41.
5. Carpino, L. A.; Mansour, M. E. J. Org. Chem. 1992, 57, 6371–6373.
6. Sajiki, H.; Hattori, K.; Hirota, K. J. Org. Chem. 1998, 63, 7990–7992.
7. Conversion of free amine into triisopropylsilyl carbamate was reported recently: Lipshutz, B.; Papa, P.; Keith, J. M. J. Org.
Chem. 1999, 64, 3792–3793.
8. Sakaitani, M.; Ohfune, Y. J. Org. Chem. 1990, 55, 870–876.
9. Carpino, L. A.; El-Faham, A. J. Am. Chem. Soc. 1995, 117, 5401–5402.
10. (a) Ito, Y.; Ogawa, T., Angew. Chem. Int. Ed. Engl. 1994, 33, 1765–1767. (b) Dan, A.; Ito, Y.; Ogawa, T., J. Org. Chem.
1995, 60, 4680–4681. (c) Ito, Y.; Ogawa, T. J. Am. Chem. Soc. 1997, 119, 5562–5566. (d) Ito, Y.; Ohnishi, Y.; Ogawa, T.;
Nakahara, Y. Synlett 1998, 1102–1104.