A sugar aminoacid for the development of multivalent ligands for Escherichia coli 0157 verotoxin

Article abstract of DOI:10.1016/j.tetasy.2009.02.021

Functionalised carbohydrates for incorporation into multivalent ligands for the Escherichia coli 0157 verotoxin are highly desirable. Here, we report the synthesis of a sugar aminoacid based on the known Gal-α-1,4-Gal ligand for verotoxin in eight steps a

Full text of DOI:10.1016/j.tetasy.2009.02.021

Tetrahedron: Asymmetry 20 (2009) 730–732
Contents lists available at ScienceDirect
Tetrahedron: Asymmetry
journal homepage: www.elsevier.com/locate/tetasy
A sugar aminoacid for the development of multivalent ligands
for Escherichia coli 0157 verotoxin
Darren Gibson a, Steven W. Homans b, Robert A. Field c,
*
a School of Chemistry, University of St. Andrews, St. Andrews KY16 9ST, UK
b Institute of Molecular and Cell Biology, University of Leeds, Leeds LS2 9JT, UK
c Department of Biological Chemistry, John Innes Centre, Colney Lane, Norwich NR4 7UH, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
Functionalised carbohydrates for incorporation into multivalent ligands for the Escherichia coli 0157 vero-
Received 4 February 2009
Accepted 10 February 2009
Available online 9 March 2009
toxin are highly desirable. Here, we report the synthesis of a sugar aminoacid based on the known Gal-a-
1,4-Gal ligand for verotoxin in eight steps and in ꢀ10% overall yield.
Ó 2009 Elsevier Ltd. All rights reserved.
Dedicated to Professor George Fleet on the
occasion of his 65th birthday
1. Introduction
reported. Importantly, Kitov and Bundle have shown that, with
respect to the parent disaccharide, substitution in either the 2- or
Escherichia coli O157, which periodically causes outbreaks of
potentially lethal diarrhoea, presents a problem for human health
worldwide.1 This organism produces so-called verotoxins (or shi-
ga-like toxins), which are members of the AB5 class of bacterial
toxins, along with cholera and pertussis toxins.2,3 Verotoxins bind
to cell surface globotriaosylceramide epitopes, resulting in toxin
uptake by the cell, followed by the catalytically active A-subunit
cleaving a single nucleotide base from ribosomal RNA, and hence
shutting down protein synthesis.4 It is now well-established that
the most effective means of neutralising the effect of verotoxin is
by preventing its interaction with host cells with multivalent
carbohydrate-based ligands.5 Building on our work on the develop-
ment of carbohydrate-based sensors and imaging agents for pro-
tein toxins and bacteria,6 and extending our interest in verotoxin
ligands,7 we sought novel ligand components suitable for oligo-
merisation into potential verotoxin inhibitors. Sugar aminoacids
present attractive building blocks in this regard, given their poten-
tial for conjugation to give multivalent constructs and the ease
with which they can be derivatised for ligand optimisation.
6-position of Gal-
a
-1,4-Gal does not significantly reduce affinity
-1,4-
for verotoxin.11 Herein, we report the synthesis of a Gal-
a
Gal aminoacid with amino substitution at C-2 and a carboxymethyl
group at C-6 of the reducing terminal sugar unit (Fig. 1). Our syn-
thetic approach was to first prepare the difunctionalised reducing
terminal sugar unit, and then glycosylate to install the inter-sugar
a-1,4-glycosidic linkage.
2. Results and discussion
Commercial methyl b-D-galactoside, by way of the 3,4,6-O-
acetalated intermediate, was protected as its 3,4-O-acetonide 2
in 93% yield by reaction with 2,2-dimethoxypropane and p-tolu-
enesulfonic acid, followed by careful selective cleavage of the
mixed sugar-(2-methoxy)propyl acetal at C-6 with 50% aq TFA
(Scheme 1). Mono-O-tosylation of the primary alcohol followed
by treatment with sodium azide in DMF gave azide 3.13 Subsequent
alkylation with sodium hydride and methyl bromoacetate then
gave azido-ester 4 in 86% yield. Deprotection of the acetal in 4
was initially attempted with 80% aq TFA, but this gave a mixture
of products, including those resulting from ester hydrolysis. A fur-
ther attempt at acetal removal with 50% aq HBF4 in methanol gave
the desired methyl ester 5 in 90% yield.14
Numerous attempts were made to directly cyclise ester 5 to the
corresponding 2,3-lactone, 6, but this was largely unsuccessful. Es-
ter hydrolysis followed by diimide-based cyclisation again resulted
only in low yields (<20%) or in the desired lactone 6.13 In addition,
attempts to selectively acetylate or benzoylate the equatorial 3-OH
of 3,4-diol 5 typically resulted in a mixture of 3-mono- and 3,4-di-
O-acylated material. The change of oxygen to azide functionality at
a
-C-linked galactose, even in multivalent format, does not give
rise to potent verotoxin ligands.8 In contrast, a Gb3 trisaccharyl C-
linked amino acid, and peptides derived therefrom, shows good
inhibition of verotoxin, but ligand synthesis is lengthy.9 Whilst
verotoxin binding by the Gal-
a-1,4-Gal disaccharide is appreciably
weaker than that by Gb3 trisaccharide,10 it can still serve as an
effective template for inhibitor development.11 In the Gal-
a-
1,4-Gal series, amino-12 and carboxy-derivatives11,12 have been
* Corresponding author. Tel.: +44 1603 450720.
E-mail address: rob.field@bbsrc.ac.uk (R.A. Field).
0957-4166/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2009.02.021

D. Gibson et al. / Tetrahedron: Asymmetry 20 (2009) 730–732
731
HO
HO
OH
O
HO
HO
OH
O
HO
O
NH2
O
HO
O
OH
O
OH
O
HO
OMe
OH
HO
HO
O
O
OH
OH
O
HO
Gal- -1,4-Gal aminoacid 1
Gb3 trisaccharide
α
Figure 1. Natural Gb3 trisaccharide and target sugar aminoacid 1.
HO
HO
O
O
OH
O
OH
O
N3
i
ii, iii
O
74 %
OMe 93 %
O
OMe
O
OMe
OH
OH
OH
2
3
O
HO
BnO
N3
N3
O
HO
HO
N3
v
iv
vi
O
O
O
OMe
OMe
86 %
90 %
51 %
OMe
O
O
O
O
O
O
MeO
BnO
MeO
4
5
7
HO
O
N3
O
OMe
O
6
O
Scheme 1. Reagents and conditions: (i) (a) p-TsOH, (CH3)2C(OCH3)2; (b) (i) 50% TFA (aq), DCM; (ii) p-TsCl, (CH3)2CO, pyridine; (iii) NaN3, DMF; (iv) methyl bromoacetate, NaH,
DMF; (v) 50% HBF4 (aq), MeOH; (vi) (a) MeOH, Bu2SnO; (b) toluene, BnBr, Bu4NBr.
C-6 appears to reduce 3/4-O-acylation selectivity. Tin acetal chem-
istry was therefore considered.15 Reaction of diol 5 with dibutyltin
oxide in refluxing methanol and subsequent addition of 5 mol
equiv of benzyl bromide and 0.5 mol equiv of tetrabutylammo-
nium bromide gave selective 3-O-benzylation resulting in 7 in
moderate yield. A small sample of the benzyl ether was acetylated
BnO
BnO
OBn
O
BnO
BnO
OBn
O
HO
HO
OH
O
HO
BnO
N3
BnO
Cl
O
BnO
ii
8
HO
i
OMe
O
N3
O
NH2
O
O
50 %
70 %
O
BnO
OMe
HO
OMe
O
O
O
9
1
BnO
7
O
O
BnO
HO
BnO
BnO
OBn
O
OBn
OBn
BnO
O
BnO
O
BnO
10
Scheme 2. Reagents and conditions: (i) toluene/diethyl ether (2:1), AgOTf, collidine, 4 Å MS, ꢁ78 C; (ii) EtOH, AcOH, H2, Pd-C.

732
D. Gibson et al. / Tetrahedron: Asymmetry 20 (2009) 730–732
to confirm the regiocontrol of the benzylation [1H NMR data
showed downfield shift for H-4 upon acetylation (3.79–
6. (a) Schofield, C. L.; Haines, A. H.; Field, R. A.; Russell, D. A. Langmuir 2006, 22,
6707–6711; (b) Schofield, C. L.; Field, R. A.; Russell, D. A. Anal. Chem. 2007, 79,
1356–1361; (c) Schofield, C. L.; Mukhopadhyay, B.; Hardy, S. M.; McDonnell, M.
B.; Field, R. A.; Russell, D. A. Analyst 2008, 133, 626–634; (d) Mukhopadhyay, B.;
Martins, M. B.; Karamanska, R.; Russell, D. A.; Field, R. A. Tetrahedron Lett. 2009,
50, 886–889.
7. (a) Shimizu, H.; Brown, J. M.; Homans, S. W.; Field, R. A. Tetrahedron 1998, 54,
9489–9506; (b) Shimizu, H.; Field, R. A.; Homans, S. W.; Donohue-Rolfe, A.
Biochemistry 1998, 31, 11078–11082; (c) Bernlind, C.; Homans, S. W.; Field, R.
A. Tetrahedron Lett. 2009. doi:10.1016/j.tetlet.2009.02.131.
8. Arya, P.; Kutterer, K. M. K.; Qin, H.; Roby, J.; Barnes, M. L.; Lin, S.; Lingwood, C.
A.; Peter, M. G. Bioorg. Med. Chem. 1999, 7, 2823–2833.
9. (a) Debenham, S. D.; Cossrow, J.; Toone, E. J. J. Org. Chem. 1999, 64, 9153–9163;
(b) Lundquist, J. J.; Debenham, S. D.; Toone, E. J. J. Org. Chem. 2000, 65, 8245–
8250.
a
5.36 ppm)]. However, on closer inspection it became apparent that
the 3-O-benzylated product 6 was in fact the benzyl ester, the
methyl ester functionality having been transformed during the
benzylation process. Presumably, residual moisture in conjunction
with the tin reagent contributed to methyl ester hydrolysis in situ,
followed by carboxyl group benzylation, giving 6.
Glycosylation of the sugar aminoacid 7 with 1.5 mol equiv of
galactosyl chloride 816 was effected with silver triflate and collidine
in anhydrous toluene/diethyl ether17 to promote glycosylation
a-selectivity (Scheme 2). The reaction gave the desired a-linked
10. St Hilaire, P. M.; Boyd, M. K.; Toone, E. J. Biochemistry 1994, 33, 14452–
14463.
glycoside, 9, in 70% isolated yield, accompanied by the formation
of 1,10-linked disaccharide 10 (20%),18 presumably arising from
donor quenching and coupling with further donor. Altering the
donor:acceptor stoichiometry did not impact productively on the
yield of the desired glycoside 9. Final reductive deprotection of azi-
do-ester 9 proceeded in moderate yield, reflecting issues associated
with partial removal and/or reduction of benzyl ether-protecting
groups.
11. Kitov, P. I.; Bundle, D. R. J. Chem. Soc., Perkin Trans. 1 2001, 838–853.
12. Hansen, H. C.; Magnusson, G. Carbohydr. Res. 1998, 307, 233–242.
13. All synthetic intermediates gave NMR data and combustion analysis or high-
resolution mass spectrometry data consistent with their proposed structures.
Selected data for azide 3: [a]D = ꢁ16.0 (c 0.95, CHCl3); m
max/cmꢁ1 2097 (N3); dH
(CDCl3, 400 MHz): 1.36 (3H, s, CH3), 1.48 (3H,s, CH3),2.85 (1H, br s, OH), 3.39
(1H, dd, J5,6/6 5, J6,6 13 Hz, H-6/60) 3.49 (1H, dd, J1,2 8.1, J2,3 8.4 Hz, H-2), 3.51
0
0
0
0
0
(3H, s, OMe), 3.69 (1H, dd, J5,6/6 , J6,6 , H-6/6 ), 3.90 (1H, m H-5), 4.01–4.08 (2H,
m, H-3, 4), 4.08 (1H, d, J1,2, H-1); dC (CDCl3, 100 MHz): 26.2, 27.9, 51.1, 57.0,
72.9, 73.5, 73.8, 78.9, 103.3, 110.5.
Selected data for lactone 6: [a]D = ꢁ15.9 (c 0.36, CHCl3); m
max/cmꢁ1 2100 (N3); dH
0 0
(CDCl3, 400 MHz): 3.35 (1H, dd, J5,6/6 2, J6,6 010 Hz, H-6/60), 3.60 (3H, s, OMe),
3. Conclusion
0
3.74 (1H, m, J5,6/6 , H-5), 3.75 (2H, m, H-2, 6/6 ), 4.07 (1H, dd, J3,4 3 Hz, H-4), 4.38
(1H, dd, J2,3 7 Hz, J3,4, H-3), 4.42 (1H, d, J1,2 8 Hz, H-1), 4.43 (1H, d, J 18.0 Hz,
OCH2COOMe) 4.62 (1H, d, J 18.0 Hz, OCH2COOMe); dC (CDCl3, 100 MHz): 50.7,
57.1, 66.2, 67.1, 71.5, 74.0, 79.7, 101.3, 165.8.
In conclusion, the synthesis of di-functionalised, Gal-
a-1,4-Gal
aminoacid 120 from methyl b-
D-galactoside was successfully
14. Pozsgay, V. J. Am. Chem. Soc. 1995, 117, 6673–6681.
achieved in eight steps and in ꢀ10% overall yield. Studies exploring
the use of this reagent in verotoxin inhibitor development will be
reported in due course.
15. David, S.; Hannessian, S. Tetrahedron 1985, 41, 643–663.
16. Iversen, T.; Bundle, D. R. Carbohydr. Res. 1982, 103, 29–40.
17. Demchenko, A.; Stauch, T.; Boons, G. J. Synlett 1997, 818–820.
18. Deprotection of disaccharide 10 using catalytic hydrogenolysis gave the known
unprotected 1,10-linked disaccharide, the
a
/b-stereochemistry of which was
confirmed by 1H NMR spectroscopy. Selected data: [
a]
D
+58 (c 0.22, H2O),
Acknowledgements
(lit.,19 +56); 1H NMR: dH 4.20 (J1a,2a 7.8, H-1a), 5.15 (J1b,2b 3.0, H-1b).
19. Sharp, V. E.; Stacey, M. J. Chem. Soc. 1951, 285–288.
These studies were supported by the BBSRC. We thank the
EPSRC National Mass Spectrometry Service Centre, Swansea, for
invaluable support.
20. Selected data for protected penultimate and final compounds: Methyl 2,3,4,6-
tetra-O-benzyl-
a-D-galactopyranosyl-(1,4)-3-O-benzyl-6-azido-2-benzylacetate-
6-deoxy-b-
D
-galactopyranoside 9: [
a
]
D = +22.7 (c 0.99, CHCl3);
m
0
max/cmꢁ1 1757
(COOBn), 2100 (N3); dH(CDCl3): 3.18 (1H, dd, J5a,6a/6a 4.5, J6a,6a 8.1, H-6a/6a0),
0
0
0
0
3.32 (1H, t, J5b,6b/6b 5.1, J6b,6b 5.1, H-6b/6b ), 3.48 (3H, s, OMe), 3.42–3.47 (2H, m,
H-3a + H-5a), 3.59 (2H, m, H-2a + H-6b/6b0), 3.68 (1H, dd, J5a,6a/6a , J6a,6a , H-6a/
6a0), 3.75 (1H, d, J3a,4a, 3.5, H-4a), 3.99–4.01 (3H, m, H-2b, 3b + 4b), 4.20–4.25
(2H, dd, OCH2Ar), 4.25 (1H, d, J1a,2a, 8.0, H-1a), 4.35 (3H, m, H-5b + OCH2), 4.54–
4.80 (4H, m, 2 ꢂ OCH2Ar), 4.86–4.94 (2H, m, OCH2Ar), 4.94 (1H, d, J1b,2b 3.3,
H-1b), 5.15 (2H, dd, J 9.3 + 12.6, OCH2), 7.04–7.36 (30H, m, 6 ꢂ OCH2Ar);
dC(CDCl3): 50.9, 56.9 (OMe), 66.4, 68.2, 69.8, 70.25, 72.2, 73.0, 73.3, 74.4, 74.7,
74.8, 74.9, 76.4, 77.8, 78.9, 79.6, 80.7, 100.8 (C-1b), 104.3 (C-1a), 127.6, 127.7,
127.8, 127.9, 128.2, 128.2, 128.3, 128.5, 128.7, 135.8, 138.3, 138.5, 138.7, 138.8,
139.1, 170.3.
References
0
0
1. (a) Paton, J. C.; Paton, A. W. Clin. Microbiol. Rev. 1998, 11, 450–479; (b) Karch,
H.; Tarr, P. I.; Blelaszewska, M. Int. J. Med. Microbiol. 2005, 295, 405–418.
2. Merritt, E. A.; Hol, W. G. J. Curr. Opin. Struc. Biol. 1995, 5, 165–171.
3. Lacy, D. B.; Stevens, R. C. Curr. Opin. Struc. Biol. 1998, 8, 778–784.
4. Endo, Y.; Tsurugi, K.; Yutsudo, T.; Takeda, Y.; Ogasawara, K.; Igarashi, K. Eur. J.
Biochem. 1988, 171, 45–50.
5. See: Kitov, P. I.; Mulvey, G. L.; Griener, T. P.; Lipinski, T.; Solomon, D.;
Paszkiewicza, E.; Jacobson, J. M.; Sadowska, J. M.; Suzuki, M.; Yamamura, K. I.;
Armstrong, G. D.; Bundle, D. R. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 16837–
16842. and references cited therein.
Gal-
a-1,4-Gal aminoacid 1: characteristic NMR data—dH (CD3OD, 400 MHz):
3.51 (3H, s, OMe), 4.31 (2H, m, OCH2), 4.35 (1H, d, J1,2 7.6 Hz, H-1), 5.01 (1H, d,
J1 ,2 3.4 Hz, H-10); dC (CD3OD, 100 MHz): 101.7 (C-10), 104.7 (C-1).
0
0