Phosphonic derivatives of carbohydrates: Chemoenzymatic synthesis

Article abstract of DOI:10.1016/S0040-4039(00)00326-9

A series of homologous racemic β-hydroxyaldehydes bearing a phosphonate group in the ω-position were synthesised and allowed to react with dihydroxyacetone phosphate (DHAP) via an enzymatic aldol addition catalysed by fructose 1,6-bisphosphate aldolase (FruA). After enzymatic dephosphorylation, a set of unnatural ω-phosphonic deoxysugars was obtained. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4039(00)00326-9

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 41 (2000) 3181–3185
Phosphonic derivatives of carbohydrates: chemoenzymatic
synthesis
∗
Giuseppe Guanti, Luca Banfi and Maria Teresa Zannetti
Dipartimento di Chimica e Chimica Industriale e C.N.R., Centro di Studio per la Chimica dei Composti Cicloalifatici ed
Aromatici, via Dodecaneso 31, I-16146 Genova, Italy
Received 17 January 2000; accepted 21 February 2000
Abstract
A series of homologous racemic β-hydroxyaldehydes bearing a phosphonate group in the ω-position were
synthesised and allowed to react with dihydroxyacetone phosphate (DHAP) via an enzymatic aldol addition
catalysed by fructose 1,6-bisphosphate aldolase (FruA). After enzymatic dephosphorylation, a set of unnatural
ω-phosphonic deoxysugars was obtained. © 2000 Elsevier Science Ltd. All rights reserved.
Keywords: phosphonic acids and derivatives; enzyme reactions; aldol reactions; carbohydrate mimetics.
Phosphonic acid derivatives, isosteric with natural and synthetic phosphates, in which the P–O bond
of the phosphate is formally replaced with a more stable P–C bond, have attracted considerable attention
due to their stability towards the action of phosphatases and to their potential bioactivity as inhibitors or
regulators of metabolic processes. For example, phosphonate containing molecules have been shown to
1
2
3
4
5,6
be inhibitors of EPSP synthase, HIV protease, renin, farnesyl protein transferase and PTPases.
The synthesis of a large number of isosteric phosphonate analogues of phosphorylated sugars and their
7
interaction with a number of enzymes has been described. Most of the reported syntheses of phosphonic
derivatives of carbohydrates employ natural sugar derivatives as the starting material and the phosphonic
group is introduced in a late step; this strategy suffers from the need of a selective protection/deprotection
scheme and from the limited availability of starting materials.
Enzymatic aldol addition of dihydroxyacetone phosphate (DHAP) has become a useful tool in the syn-
thesis of natural carbohydrates and their analogues. The existence of a set of four stereocomplementary
aldolases and the low substrate specificity of these enzymes, at least for the electrophilic component,
has permitted the synthesis of a large number of carbohydrates and, in general, of polyhydroxylated
8
–10
compounds.
Fructose 1,6-bisphosphate aldolase from rabbit muscle (FruA, EC 4.1.2.13), catalyses
∗
Corresponding author. Tel/fax: +39-010-353-6105; e-mail: guanti@chimica.unige.it (G. Guanti)
0
040-4039/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(00)00326-9
tetl 16606

3
182
the addition of DHAP to a broad range of aldehydes to produce adducts of threo-3S,4R-configuration,
and is the most widely used aldolase for synthetic purposes.
Few phosphonic aldehydes are reported to be substrates for DHAP aldol addition catalysed by
1
1–13
12
FruA
and only in two cases have phosphonic sugars been obtained and characterised. An alter-
native approach to sugar 1-phosphonates, bio-isosteric analogues of sugar 1-phosphates, based on the
use of aldolases, employing a phosphonic analogue of DHAP, has been developed.1
4,15
Here we present
the synthesis of a set of sugar ω-phosphonates via aldol addition of DHAP to the homologous racemic
β-hydroxyaldehydes 3, 6 and 9 catalysed by FruA (Scheme 1).
+
Scheme 1. (a) O
3
, CH
2
Cl
2
/MeOH, −78°C, then Me
2
S; (b) diethylphosphite, KF, rt; (c) Dowex resins (H form), H
2
O, 50°C;
(d) Zn, allyl-Br, H
2
O, THF, rt
Hydroxyaldehyde 3 was synthesised as diethylacetal 2 following a modified Pudovic reaction16
starting from diethylphosphite and 3,3-diethoxypropanal in the presence of KF as catalyst.17 The latter
aldehyde was easily obtained via ozonolysis followed by reductive work-up of 4,4-diethoxy-1-butene.
After deprotection, the pH of the solution was adjusted to 7.0 by addition of 1 M NaOH and the resulting
solution of crude aldehyde 3 was used directly in the enzymatic step.18
1
9
Hydroxyaldehyde 6 was synthesised in its masked form 5 via allylation of the known aldehyde 4
2
0
using allyl bromide and zinc in H O–THF. The allylation in a protic medium gave a cleaner reaction
2
and a better overall yield in comparison with the Grignard addition using allyl magnesium bromide,
probably because of the easy enolisation of 4. Ozonolysis of 5 followed by reductive work-up gave
the hydroxyaldehyde 6. Formaldehyde, the by-product from the cleavage of the double bond, was
quantitatively removed by evaporation in vacuo. The crude aldehyde 6 was taken up in H O, the pH of
2
the solution adjusted to 7.0 by addition of NaOH and the resulting solution used directly in the enzymatic
1
8
step.
Aldehyde 9 was prepared from phosphonate 721 following a similar procedure.18
The racemic mixtures of the aldehydes 3, 6 and 9 in aqueous solution were subjected to the enzymatic
aldol addition catalysed by FruA (Scheme 2).22 DHAP was formed in situ from commercially available
fructose 1,6-bisphosphate (FBP) by a combination of FruA and triose phosphate isomerase (TPI, EC
2
3
5
.3.1.1). In a typical procedure, an aqueous solution of crude hydroxyaldehyde (8 mmol) was treated
with FBP (1 mmol), the pH of the solution was adjusted to 7.0 with NaOH, then FruA (50–100 U) and

3
183
TPI (1000 U) were added. After 3–5 days, the pH of the solution was adjusted to 4.8 with 1 M HCl and
acid phosphatase (50 U) was added. Alternatively the pH of the solution was adjusted to 8.0 with 1 M
NaOH and treated with alkaline phosphatase (200 U). The resulting adducts were purified by silica gel
chromatography.
Scheme 2. (a) FBP, FruA, TPI (pH=7.0); (b) acid phosphatase (pH=4.8) or alkaline phosphatase (pH=8.0)
It is worth noting that, generally, in the FruA catalysed additions of DHAP to racemic β-hydroxy-
aldehydes, the reaction of the enantiomer which produces the more stable adduct is thermodynamically
2
3
favoured. In this case it should be possible to produce exclusively the more stable aldol adduct by the
use of an excess of aldehyde compared to the nucleophilic component. Nevertheless, when an excess of
racemic aldehyde 3 was subjected to the enzymatic sequence of aldol addition/dephosphorylation,24 two
adducts were isolated in 50% global yield and in 1:2 ratio. After a careful chromatographic separation
of the diastereomers, spectroscopic analysis showed that the major adduct corresponds to the monoester
11 (apparently the less stable one), while the minor adduct was the monoester 10 (apparently the more
stable one). In fact, the adduct 11 shows J =1.8 Hz, diagnostic for a diequatorial relationship (trans),
3
/4
3
in a chair conformation C , while the adduct 10 shows J =9.4 Hz, diagnostic for a diaxial relationship
trans) in a chair conformation C , as shown in Scheme 2. This means that, surprisingly, the formation
6
3/4
6
(
3
of the adduct 11 is favoured. Another unexpected result of this reaction is the isolation of a monoethyl
2
5
phosphonate.
When an excess of the racemic aldehydes 6 or 9 were subjected to the enzymatic sequence of aldol
2
4
addition/dephosphorylation, the adducts 12 and 14 were isolated in 12% and 40% yield, respectively.
The spectroscopic analysis in both cases showed that the adduct anticipated to be the more stable was
obtained. As expected, the formation of both 13 and 15 was not observed under these reaction conditions.

3
184
We have described the chemoenzymatic synthesis of a set of new sugar analogues bearing an unnatural
substituent. The construction of complicated structures with three chiral centres in optically active form
was achieved starting from easily accessible racemic phosphonates. Although the overall yields were
modest, we believe that optimisation is possible (i.e. by the use of DHAP instead of FBP and by the use
of more appropriate purification techniques). Further investigations in the chemoenzymatic synthesis of
phosphonic derivatives of carbohydrates are in progress and will be published in due course.
Acknowledgements
We wish to thank C.N.R., University of Genoa, M.U.R.S.T (COFIN 98), CRUI (Programma Vigoni)
for financial assistance and Mrs Chiara Natale for her precious collaboration to this project.
References
1
. Sikorski, J. A.; Miller, M. J.; Braccolino, D. S.; Cleary, D. G.; Corey, S. D.; Font, J. L.; Gruys, K. J.; Han, C. Y.; Lin, K. C.;
Pansegrau, P. D.; Ream, J. E.; Schnur, D.; Shah, A.; Walker, M. C. Phosphorus, Sulfur and Silicon 1993, 76, 375–378.
. Stowasser, B.; Budt, K.-H.; Jian-Qi, L.; Peyman, A.; Ruppert, D. Tetrahedron Lett. 1992, 33, 6625–6628.
. Patel, D.; Rielly-Gauvin, K.; Ryono, D. E. Tetrahedron Lett. 1990, 31, 5587–5594.
2
3
4
5
6
7
8
9
. Pompliano, D. L.; Rands, E.; Schaber, M. D.; Mosser, S. D.; Antony, N. J.; Gibbs, J. B. Biochemistry 1992, 31, 3800–3807.
. Burke, T. R. J.; Li, Z.-H.; Bolen, J. B.; Marquez, V. E. J. Med. Chem. 1991, 34, 1577–1581.
. Burke, T. R. J.; Barchi, J. J. J.; George, C.; Wolf, G.; Shoelson, S. E.; Yan, X. J. Med. Chem. 1995, 38, 1386–1396.
. Engel, R. Chem. Rev. 1977, 77, 349–367.
. Fessner, W.-D.; Walter, C. Topics in Current Chemistry 1996, 184, 97–194.
. Fessner, W.-D. Current Opinion in Molecular Biology 1998, 2, 85–87.
1
0. Wong, C.-H.; Whitesides, G. M. Enzymes in Synthetic Organic Chemistry; Baldwin, J. E.; Magnus, P. D., Eds. Tetrahedron
Organic Chemistry Series; Pergamon: Oxford; 1994.
1
1
1
1
1
1
1
1
1. Gijsen, H. J. M.; Wong, C.-H. Tetrahedron Lett. 1995, 36, 7057–7060.
2. Page, P.; Blonski, C.; Périé, J. Tetrahedron 1996, 52, 1557–1572.
3. Effenberger, F.; Straub, A. Tetrahedron Lett. 1987, 28, 1641–1644.
4. Arth, H. L.; Fessner, W.-D. Carbohydr. Res. 1997, 305, 313–321.
5. Fessner, W.-D.; Sinerius, G. Angew. Chem., Int. Ed. Engl. 1994, 33, 209–212.
6. Pudovic, A. N.; Konovalova, I. V. Synthesis 1979, 81–96.
7. Textier-Boullet, F.; Focaud, A. Synth. Commun. 1982, 165–166.
2
8. An analytical sample of aldehyde was carefully dried and taken up with D O. A satisfactory spectroscopic characterisation
was achieved.
1
2
2
2
9. Nagata, W.; Wakabayashi, T.; Hayase, Y. Org. Synth. Coll. Vol. VI, 448–451.
0. Li, C.-J. Tetrahedron 1996, 52, 5643–5668.
1. Kabak, J.; DeFilippe, L.; Engel, R. J. Med. Chem. 1972, 15, 1074–1075.
2. Although the stereochemistry at the anomeric centre is not fixed, the adducts 10, 11, 12 and 14 are present in aqueous
solution as a single anomer, most likely the α-anomer, as reported in Scheme 2
23. Bednarski, M. D.; Simon, E. S.; Bischofberger, N.; Fessner, W. D.; Kim, M. J.; Lees, W.; Saito, T.; Waldmann, H.;
Whitesides, G. M. J. Am. Chem. Soc. 1989, 111, 627–635.
2
2
4. Approximately the same results (yield and diastereomeric ratio) were obtained by using acid or alkaline phosphatase.
5. A tentative explanation for the unexpected diastereomeric ratio and for the monohydrolysis of the phosphonic diethyl ester
could be the formation of the bicyclic intermediates 10b and 11b, directly after the aldol addition. The formation of 11b from
1
1a is likely to be faster, when compared with the formation of 10b from 10a. The cyclic phosphonates 10b and 11b cannot
undergo a retroaldolic addition, thus withdrawing the adducts from enzymatic equilibration. Then, the acidic or alkaline
conditions used during the enzymatic dephosphorylation and/or the acidity of the silica gel during the chromatography,

3
185
affords ring opening. An alternative explanation could involve enzymatic hydrolysis of the diethylphosphonate catalysed by
2
6
the phosphatase. Although this reaction is reported in the literature, seems to be less likely because we did not observe
this hydrolysis on similar phosphonates and because no differences were detected during the use of acidic and alkaline
phosphatase.24 As we did not isolate the intermediate monophosphate, it is not possible to decide definitively in which step
such hydrolysis takes place.
26. Natchev, I. A. Bull. Chem. Soc. Jpn. 1988, 61, 3699–3715.