A convenient route to keto-glycosyl phosphates

Article abstract of DOI:10.1016/S0040-4039(99)02020-1

The reagent combination, ruthenium dioxide/sodium periodate/benzyltriethyl ammonium chloride in dichloromethane/aqueous bicarbonate buffer simultaneously oxidises alcohol functions in the sugar ring and glycosyl H-phosphonates to yield keto-glycosyl phosphates. These can be coupled to the respective nucleoside diphosphates to render biosynthetically relevant sugar metabolites and derivatives thereof, useful for further investigation of the polysaccharide biosynthesis in bacteria and plants.

Full text of DOI:10.1016/S0040-4039(99)02020-1

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 41 (2000) 189–192
A convenient route to keto-glycosyl phosphates
Andreas Naundorf,a Stefan Natsch b and Werner Klaffke b,∗
aMeiji Milk Products Co., Ltd, Institute of Health Science, Naruda 540, Odawara 250-0862, Japan
bOrganisch-Chemisches Institut der Westfälischen Wilhelms-Universität, Corrensstrasse 40, D-48149 Münster, Germany
Received 17 August 1999; revised 7 October 1999; accepted 21 October 1999
Abstract
The reagent combination, ruthenium dioxide/sodium periodate/benzyltriethyl ammonium chloride in dichloro-
methane/aqueous bicarbonate buffer simultaneously oxidises alcohol functions in the sugar ring and glycosyl H-
phosphonates to yield keto-glycosyl phosphates. These can be coupled to the respective nucleoside diphosphates
to render biosynthetically relevant sugar metabolites and derivatives thereof, useful for further investigation of the
polysaccharide biosynthesis in bacteria and plants. © 1999 Elsevier Science Ltd. All rights reserved.
Keywords: carbohydrates; oxidation; phosphates; ruthenium.
The pathogenicity of the O-antigen of Gram-negative bacteria is largely influenced by the presence of
deoxy sugars at the non-reducing terminus.1 Furthermore, regulation of hydrophilicity of the aglycone
in commercially applied drugs, and targetting of the glyco-drug towards the endogenous receptor is
attributed to the presence of specifically deoxygenated hexoses.2–4
Cloning techniques made the gene clusters available, which account for the biosynthesis of bacterial
and plant deoxysaccharides5 and much insight has been gained on the respective mechanisms.6,7
However, the ultimate goal of suppressing or modulating the biosynthesis of such a deoxy sugar-
containing structure in order to gain control over the antigeneity of the respective organism, will only
be met if the biosynthetic intermediates and their analogues are at hand for further investigation.
In addition to the reported enzyme-based procedures,8 suitable chemical approaches would enrich the
product spectrum by artificial analogues of the target metabolites for biochemical research.
However, straightforward routes for the chemical generation of keto-glycosyl phosphates are scarce,
if not unavailable and the only method published hitherto has been reported by Schmidt et al.9 An exo-
methylene group masked the keto-function which was then generated upon ozonolysis.
The aim of this contribution is to report on a facile method for the oxidation of glycosyl phosphites
and H-phosphonates, allowing for simultaneous oxidation of the C-3 or C-4 positions of the hexosyl
moiety. Therefore, a series of experiments was set up to investigate: (i) the oxidation of glycosyl H-
phosphonates; (ii) the mild oxidation of hydroxy-groups in the hexose and hence, suppressing imminent
∗
Corresponding author. Fax: 0049 251 836501; e-mail: werner.klaffke@uni-muenster.de (W. Klaffke)
0040-4039/00/$ - see front matter © 1999 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(99)02020-1
tetl 15948

190
rearrangements and eliminations; and (iii) the simultaneous generation of both the keto-group and the
glycosyl phosphate.
Scheme 1 depicts the experiments conducted on the oxidation of glycosyl phosphites to give the
corresponding phosphates. Oxidation of glycosyl phosphates is known to be very sluggish due to the
reported stability of the glycosyl H-phosphonates. Consequently, starting material 3 was converted
according to the reports by van Boom et al.10 to the respective glycosyl diester, either by in situ
esterification to yield the disilyl phosphite or via the glycosyl monobenzyl-H-phosphonate 4. Upon
oxidation with iodine in pyridine of the former, or deprotection by hydrogenation of the latter, both
derivatives were converted in moderate yields to furnish the target phosphate 5. We were prompted
to investigate alternatives to these methods, mainly due to two factors: (i) the facile cleavage under
the conditions of oxidation which accounted for the instability of the disilylester and, therefore, low
yields; and (ii) the extremely inconvenient two-step oxidation/hydrogenolysis sequence along the mono-
benzyl phosphonate/phosphate 4 route. Finally, screening led to the ruthenium tetraoxide reagent in
a two-phase bicarbonate buffered system.11 All hexosyl phosphonates 6, 8 and 10 were oxidised in
nearly quantitative yields. Reactions were complete after addition of 2 equivalents of NaIO4 within
approximately 1 h. Traces of ruthenium ions were removed by either adding EDTA to the aqueous eluents
in SiO2-chromatography or by filtration through cation exchange resins.
Scheme 1. Reagents and conditions: (i) N2H5+OH− (71%), anhyd. DMF; (ii) P(Imi)3, anhydr. CH3CN, then pyridine/H2O
(93%); (iii) HMDSA, then I2/pyridine (20–46%); (iv) RuO2, NaIO4, BnEt3N+Cl−, KHCO3, CH2Cl2, H2O (87%); (v) BnOH,
anhyd. pyr., then I2, pyr./H2O; (vi) H2/Pd–C (75%)
To arrive at the targeted keto-hexosyl phosphates, the efficiency of the RuO4 method was checked
against selectively blocked methyl 6-deoxy-α-D-glucosides 12 and 13 (Scheme 2). Whereas, under
Swern-conditions, a subsequent β-elimination was unavoidable and furnished the methyl 4-keto-hex-
2-enopyranosides 16 and 17, the RuO4-oxidation under the above conditions gave the wanted 4-keto
glycosides 14 and 15 in very good yields. Both compounds could be isolated, purified to produce
excellent NMR and elementary analyses, and stored over longer periods without decomposition. Upon
treatment with either tetrabutylammonium acetate or DBU, the β-elimination to render 16 and 17 was
initiated for dibenzoate 14 in a very clean reaction. The dimethyl ether 15 yielded the ring-contracted
furan 18 via an intramolecular aldoladdition as the main product, the stereochemistry of which was
confirmed by NOE experiments.
Finally, the combined, one-step oxidation was checked for the preparation of dTDP-3-keto- (25) and
dTDP-4-keto-6-deoxy-α-D-glucose (26, Scheme 3). The respective selectively blocked 6-deoxy-glucosyl
phosphonates 19 and 21, which were phosphitylated according to Shibaev et al.12, were deblocked at
O-3 (20) or O-4 (22) by hydrazinolysis. Application of the described oxidation procedure rendered,
after the addition of 3 equivalents of NaIO4, the wanted 6-deoxy-3-keto- and 6-deoxy-4-ketoglucosyl

191
Scheme 2. Reagents and conditions: (i) RuO2, NaIO4, BnEt3N+Cl−, KHCO3, CH2Cl2, H2O (83% for 14; 61% for 15); (ii)
TBA+OAc−, DBU (89% for 16; 3% for 17 and 21% for 18), (iii) DMSO, ClOCCOCl (70% for 17)
phosphates 23 and 24, respectively, as the mixed Na+/K+-salts.13 Coupling, according to previously
described procedures,14 gave the wanted biosynthetic intermediates 25 and 26 in good overall yields.
Scheme 3. Reagents and conditions: (i) N2H5+OH−, anhyd. DMF, rt, 5 h (95% for 20; 91% for 22); (ii) RuO2, NaIO4,
BnEt3N+Cl−, KHCO3, CH2Cl2, H2O (87% for 23; 71% for 24); (iii) dTMP-morpholidate in CH3CN, 3 d then aqueous LiOH,
pH=11, 3 h, then ion exchange chromatography, size exclusion chromatography (47% for 25; 38% for 26)
Acknowledgements
This research has enjoyed financial support from the Deutsche Forschungsgemeinschaft through a
research grant to W.K. (KL631-3).
References
1. Lindberg, B. Adv. Carbohydr. Chem. Biochem. 1990, 48, 279–318. Raetz, C. R. H. In Escherichia coli and Salmonella:
Cellular and Molecular Biology, 2nd ed.; Neidhardt, F. C.; Reznikoff, W. S.; Riley, M.; Schaechter, M.; Umbarger, H. E.,
Eds.; ASM Press: Washington, D.C., 1966; Vol 1, pp. 1035–1063.

192
2. Nicolaou, K. C.; Ajito, K.; Komatsu, H.; Smith, B. M.; Bertinato, P.; Gomez-Paloma, L. Chem. Commun. 1996, 1495–1496;
Nicolaou, K. C.; Ajito, K.; Komatsu, H.; Smith, B. M.; Bertinato, P.; Gomez-Paloma, L., Tor, Y. J. Am. Chem. Soc. 1996,
118, 2303–2304. Liu, C.; Smith, B. M.; Ajito, K.; Komatsu, H.; Gomez-Paloma, L.; Li, T.; Theodorakis, E. A.; Nicolaou,
K. C.; Vogt, P. K. Proc. Natl. Acad. Sci. USA 1996, 93, 940–944.
3. Gao, X.; Patel, D. J. Biochemistry 1989, 28, 751–762.
4. Banville, D. L.; Keniry, M.; Shafer, R. H. Biochemistry 1990, 29, 9294–9304.
5. Lindqvist, L.; Schweda, K. H.; Reeves, P. R.; Lindberg, A. A. Eur. J. Biochem. 1994, 225, 863–872.
6. Liu, H.-W.; Thorson, J. S. Annu. Rev. Microbiol. 1994, 48, 223–256.
7. For recent mechanistic studies in the synthesis of CDP-activated sugars, see: Hallis, T. M.; Liu, H.-W. J. Am. Chem. Soc.
1999, 121, 6765–6766, and references cited therein.
8. A preparative bio-synthetic synthesis is presented in: Stein, A.; Kula, M.-R.; Elling, L.; Verseck, S.; Klaffke, W. Angew.
Chem., Int. Ed. Engl. 1995, 34, 1748–1749.
9. Müller, T.; Schmidt, R. R. Angew. Chem., Int. Ed. Engl. 1995, 34, 1328–1329.
10. Westerduin, P; Veeneman, G. H.; Marugg, J. E.; van der Marel, G. A.; van Boom, J. H. Tetrahedron Lett. 1986, 27,
1211–1214; van Boom, J. H., personal note, 1995.
11. Morris, P. E.; Kiely, D. E. J. Org. Chem. 1987, 52, 1149–1152.
12. Nikolaev, A. V.; Ivanova, I. A.; Shibaev, V. N. Carbohydr. Res. 1993, 242, 91–107.
13. Standard procedure for the oxidation of hexosyl H-phosphonates to ketohexosyl phosphates (for the oxidation of H-
phosphonates to phosphates, the second addition of NaIO4 is omitted): 0.5 mmol of the partially blocked hexosyl H-
phosphonate, KHCO3 (100 mg, 1.0 mmol), RuO2*×H2O (2.6 mg, 0.016 mmol, 3 mol%), BnEt3N+Cl− (2.8 mg, 0.010
mmol, 2 mol%) are dissolved in 10 ml CH2Cl2 and 10 ml dist. water. Upon addition of NaIO4 (260 mg, 1.2 mmol, 2.4
equiv.) the phosphate is formed within 60 min, after which a further aliquot of NaIO4 (130 mg, 0.6 mmol, 1.2 equiv.) and
50 mg KHCO3 is added. Vigorous stirring is maintained for 20 h, and the reaction is quenched with 2 ml 2-propanol and 50
mg KHCO3 for 10 min. The mixture is directly applied to a cellulose column (20×3 cm, EtOH) and filtered. Co-evaporation
of the concentrated appropriate fractions renders the ketohexosyl phosphate, which can be used without further purification.
Selected analytical data for 21: 1H NMR (CD3OD): δ=8.14/8.06 (2d, 4H, o-Bz), 7.14–7.49 (m, 6H, p-Bz, m-Bz), 6.03 (dd,
1H, H-1), 5.75 (ddd, 1H, H-2), 5.33 (dd, 1H, H-4), 4.58 (dq, 1H, H-5), 1.17 (d, 3H, H-6); J1,2=3.9, J1,P=6.7, J2,4=1.0, J4,5=9.8,
J5,6=6.2; negative FAB-MS (2.3 eV): m/z (%)=449.0 (100) [M−Na/K].
14. Liemann, S.; Klaffke, W. Liebigs Ann. Chem. 1995, 1779–1787.