Synthesis and evaluation as sialidase inhibitors of xylo-configured cyclohexenephosphonates carrying glycerol side-chain mimics

Article abstract of DOI:10.1016/j.bmcl.2003.11.003

Based on a strategy previously reported by us, we have synthesized D-xylo configured cyclohexenephosphonates designed to mimic the transition state of the sialidase reaction. The double bond orientation corresponds to the benchmark inhibitor Neu5Ac2en and we could selectively introduce hydroxyalkyl substituents in order to simulate the glycerol side-chain of neuraminic acid. The inhibitory activity of a set of compounds towards bacterial sialidases was tested and interesting differences in activity were found.

Full text of DOI:10.1016/j.bmcl.2003.11.003

Bioorganic & Medicinal Chemistry Letters 14 (2004) 361–364
Synthesis and evaluation as sialidase inhibitors of xylo-configured
cyclohexenephosphonates carrying glycerol side-chain mimics
Hansjorg Streicher*
Department of Chemistry, University of Konstanz, D-78457 Konstanz, Germany
Received 22 August 2003; revised 29 October 2003; accepted 3 November 2003
Abstract—Based on a strategy previously reported by us, we have synthesized d-xylo configured cyclohexenephosphonates designed
to mimic the transition state of the sialidase reaction. The double bond orientation corresponds to the benchmark inhibitor
Neu5Ac2en and we could selectively introduce hydroxyalkyl substituents in order to simulate the glycerol side-chain of neuraminic
acid. The inhibitory activity of a set of compounds towards bacterial sialidases was tested and interesting differences in activity were
found.
# 2003 Elsevier Ltd. All rights reserved.
1. Introduction
We have proposed cyclohexenephosphonates such as 1
Sialidases release sialic acids, a family of nine carbon
sugars the most prominent member of which is N-acet-
ylneuraminic acid, from their terminal position on cell
surface glycoproteins or glycolipids.1,2 This is exploited
by various pathogenic microorganisms which utilize the
action of their respective sialidases to promote infec-
tion.3 For instance, influenza virus progeny can only be
efficiently released from infected cells when the density
of sialic acids on the debris of the infected cell surface is
lowered, hence preventing multivalent binding to the
viral surface lectin.4 Another example would be the
cleavage of epithelial cell surface sialic acid moieties by
the sialidase of Vibrio cholerae, thus unmasking the
GM1 pentasaccharide which is the ligand for the bac-
terial toxin responsible for the pathogenesis of cholera
infection.3,5
or 2 (Fig. 1) as a class of sialidase inhibitors which allow
attachment of additional sugar moieties, with or with-
out spacer, at the phosphonate group thus generating a
monoester which retains the negative charge known to
be required for recognition by the enzyme.8ꢀ10 Thus,
bifunctional inhibitors11 could be obtained which bind
both the catalytic centre and the lectin-like binding site
of the enzyme leading to a putative increase in affinity
and selectivity. The feasibility of such a concept has
recently been confirmed by the inhibition of the action
In contrast to the influenza enzyme, for which highly
potent inhibitors such as GS 4071 or GG 167 (Fig. 1)
have been developed, inhibition of bacterial or proto-
zoal sialidases has proven to be far less straightfor-
ward.6 Some of the latter sialidases have been shown to
have a modular structure, consisting of a catalytic
domain and one or more additional carbohydrate-
binding lectin domains, in the case of the V. cholerae
enzyme also referred to as sialidase-wings.7
* Tel.: +49-7531-884403; fax: +49-7531-883135; e-mail: hansjoerg.
streicher@uni-konstanz.de
Figure 1. Prominent influenza sialidase inhibitors and 5-substituted
cyclohexenephosphonates 1 and 2 reported in this contribution.
0960-894X/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2003.11.003

362
H. Streicher / Bioorg. Med. Chem. Lett. 14 (2004) 361–364
Figure 2. Synthetic approach to d-xylo (3) and l-xylo (4) configured cyclohexenephosphonates from d- and l-xylose, respectively.
Figure 3. Reagents and conditions: (i) OsO4, NMO (75%); (ii) Ac2O,
pyridine (qu); (iii) TBAF, THF (qu); (iv) H2S, pridine, H2O; (v) Ac2O,
pyridine (60%); (vi) TMSBr, CHCl3 (qu); (vii) NH4OH, H2O; (viii)
Ac2O, pyridine (55%); (vi) TBAF, THF then Ac2O, pyridine (92%);
Biogel P2, NH4HCO3, H2O.
(vii) TMSBr, CHCl3 (qu); (viii) NH4OH, H2O; (ix) Biogel P2,
Figure 4. Reagents and conditions: (i) NaIO4, EtOH (qu); (ii) NaBH4,
EtOH (95%); (iii) Ac2O, pyridine (qu); (iv) H2S, pyridine, H2O; (v)
NH4HCO3, H2O.
of modular bacterial sialidases on polyvalent substrates
through addition of the lectin-domain ligand, in these
cases galactose.12
give 6 h/l but no attempt to separate them was made
(Fig. 3). Removal of the triisopropylsilyl group with
tetrabutylammonium fluoride followed by reduction of the
azide with hydrogen sulfide and subsequent acetylation led
to 7 h/l.14 Deprotection was achieved by conversion of the
ethyl phosphonate into the corresponding trimethylsilyl
ester, hydrolysis with aqueous ammonia and purifica-
tion of the ammonium phosphonate on Biogel P2 to
give 1 h/l15 in quantitative yield (Fig. 3).
In this paper, we report on the modification of the side
chain of previously synthesized cyclohexene-phospho-
nate 3 (Fig. 2),10 and its impact on the inhibition of
three bacterial sialidases.
2. Synthesis
Starting from 3, we dihydroxylated13 the double bond
of the allyl substituent selectively to afford a 1:1 mixture
of diastereomeric diols 5 h/l which were acetylated to
In order to resemble the glycerol side chain of neura-
minic acid more closely in length, the dihydroxypropyl
substituent had to be converted into a hydroxyethyl

H. Streicher / Bioorg. Med. Chem. Lett. 14 (2004) 361–364
363
Table 1. Inhibition of three bacterial sialidases by 1 h/la and 2 and comparison with previously synthesized compounds A and B. IC50 values are
given in Mb
Compd
Sialidase from
V. cholerae
S. typhimurium
C. perfringens
1.8ꢁ10ꢀ3
1ꢁ10ꢀ3
n.a.
A10
1 (h/l)
2
>4ꢁ10ꢀ3
1.6ꢁ10ꢀ3
>4ꢁ10ꢀ3
1.8ꢁ10ꢀ3
n.a.
n.a.
1.5ꢁ10ꢀ3
<2ꢁ10ꢀ3
n.a.
B10
a A 1:1 mixture of diastereomers was investigated.
b Values are means of three experiments with a deviation of less than 20%; n.a., not active in the concentration range tested (0.01–10 mM).
moiety. This was achieved by glycol cleavage of 5 h/l
with sodium metaperiodate and sodium borohydride
reduction of the resulting aldehyde to give 8 which was
acetylated to afford 9 for the purpose of structure con-
firmation (Fig. 4). Conversion of the azide into the
acetamide led to 10 the triisopropylsilyl group of which
was removed and the hydroxyl group acetylated to yield
the protected target molecule 11.16 Cleavage of all ester
functions as described above followed by a gel filtration
step gave target cyclohexenephosphonate 217 in good
yield (Fig. 4).
Among these three enzymes, the C. perfringens sialidase
seems to be the most selective regarding the double
bond orientation in our cyclohexenephosphonates as
the corresponding GS 4107 analogous molecules do
bind significantly.10 The only exception we found so far
was a monoester,10 this is not surprising as it is known
that the C. perfringens sialidase is inhibited effectively
by C-glycosides of neuraminic acid carrying hydro-
phobic aglycones.21 The combination of a GS 4071-
analogous double bond orientation and hydrophobic
phosphonate monoesters is currently being investigated
in our laboratory.
3. Sialidase inhibition
Acknowledgements
Inhibition by 1 h/l and 2 alongside with previously
reported9,10 molecules A and B of three commercially
available bacterial sialidases (IC50 values) was deter-
mined as described previously (Table 1).10,18
This work was supported by the Deutsche For-
schungsgemeinschaft. We thank Mr. C. Rohrig for his
help in recording the mass spectra.
V. cholerae sialidase: The enzyme was inhibited by
2
hydroxyethyl-substituted cyclohexenephosphonate
and the allyl-derivative B in the same order of magni-
tude as the by the parent compound A. An increase in
side-chain length resulted in decreased binding although
this must be viewed with caution as 1 h/l was investi-
gated as mixture of diastereomers.
References and notes
1. Angata, T.; Varki, A. Chem. Rev. 2002, 102, 439.
2. Schauer, R.; Kamerling, J. P. In Glycoproteins II; Mon-
treuil, J., Vliegenhardt, J. F. G., Schachter, H., Eds.;
Chemistry, Biochemistry and Biology of Sialic Acids;
Elsevier Science: Amsterdam, 1997; p 243.
3. Corfield, T. Glycobiology 1992, 2, 509.
4. Gubareva, L. V.; Kaiser, L.; Hayden, F. G. Lancet 2000,
5, 827.
5. Galen, J. E.; Ketley, J. M.; Fasano, A.; Richardson, S. H.;
Wasserman, S. S.; Kaper, J. B. J. Infect. Immun. 1992, 60,
406.
6. Kiefel, M. J.; von Itzstein, M. Chem. Rev. 2002, 102, 471.
7. Crennel, S.; Garman, E.; Laver, W. G.; Vimr, E.; Taylor,
G. Structure 1994, 2, 535.
Salmonella typhimurium sialidase: The results were
similar to the V. cholerae enzyme and indicated that
there is some tolerance towards side-chain modification,
a fact we try to exploit in our efforts to obtain optimized
inhibitors.19
Clostridium perfringens sialidase: Our previous finding10
that cyclohexenephosphonate A, which has a double bond
orientation comparable to inhibitors DANA20 or GG 167,
does not inhibit the C. perfringens sialidase was confirmed
by the inability of 1 h/l, 2 and B to inhibit the enzyme.
8. Streicher, H. Monatsh. Chem. 2002, 133, 1263.

364
H. Streicher / Bioorg. Med. Chem. Lett. 14 (2004) 361–364
9. Streicher, H.; Meisch, J.; Bohner, C. Tetrahedron 2001,
57, 8851.
16. Selected physical data for 11: Rf=0.25 (ethyl acetate/
methanol, 10:1). [a]2D0=29 (c 0.2, CHCl3). 1H NMR
(600 MHz, CDCl3): d 6.38 (d, 1H, J2,P=21.1 Hz, H-2),
5.59–5.54 (m, 2H, H-3, NH), 4.26, 4.11, 3.78, 3.65 (4m,
4H, H-10a, H-10b, H-20a, H-20b), 4.12–4.09 (m, 4H, 2
CH2CH3), 4.11, (m, 1H, H-4), 3.74 (m, 1H, H-5), 2.80 (m,
1H, H-6a), 2.29 (m, 1H, H-6b), 2.08–1.98 (4s, 12H, 4
COCH3), 1.35–1.32 (m, 6H, 2 CH2CH3). 13C NMR
(150.9 MHz, CDCl3): d 138.5 (C-2), 74.8 (C-5), 71.7 (C-3),
66.9, 63.2 (C-10, C-20), 53.9 (C-4), 30.6 (C-6). 31P NMR
10. Streicher, H.; Bohner, C. Tetrahedron 2002, 58, 7573.
11. (a) Streicher, H.; Meisch, J.; Bohner, C. Book of Abstracts
PA 059; 11th European Carbohydrate Symposium, Lis-
bon, Portugal, Sept. 2–7, 2001. (b) Busse, H.; Streicher, H.
Book of Abstracts PB127; 12th European Carbohydrate
Symposium, Grenoble, France, July 6–11, 2003.
12. Thobani, S.; Ember, B.; Siriwardena, A.; Boons, G.-J. J.
Am. Chem. Soc. 2003, 125, 7154.
13. Streicher, H.; Geyer, A.; Schmidt, R. R. Chem. Eur. J.
1996, 5, 502.
14. Selected physical data for 7 h/l: Rf=0.2 (ethyl acetate/
(242.9 MHz, CDCl3): d 16.90 (s, 1P, P(O)(OEt)2).
C18H30NO9P (M 435.41). MALDI-MS (pos. mode, DHB):
473.8 (M+K)+, 457.8 (M+Na)+, 435.8 (M+H)+.
17. Selected physical data for 2: 1H NMR (600 MHz, D2O): d
5.91 (d, 1H, J2,P=18.1 Hz, H-2), 4.09 (d, 1H, J3,4=9.0
Hz, H-3), 3.73 (dd, 1H, J4,5=9.0 Hz, H-4), 3.60–3.50 (m,
4H, H-10a, H-10b, H-20a, H-20b), 3.47 (m, 1H, H-5), 2.67
(m, 1H, H-6a), 2.09 (m, 1H, H-6b), 1.91 (s, 3H,
NHCOCH3). 13C NMR (150.9 MHz, D2O): d 174.2
(NHCOCH3), 135.4 (d, J1,P=165 Hz, C-1), 131.0 (C-2),
75.9 (C-5), 71.0 (C-3), 70.2, 60.6 (C-10, C-20), 56.8 (C-4),
31.2 (C-6), 21.7 (NHCOCH3). 31P NMR (242.9 MHz,
D2O): d 12.33 (s, 1P, PO32ꢀ). C10H18NO7P (M 295.23)
MALDI-MS (pos. mode, DHB): 318.1 (M+Na)+, 296.1
(M+H)+.
18. (a) Potier, M.; Mameli, L.; Belisle, M.; Dallaire, L.; Mel-
anson, S. B. Anal. Biochem. 1997, 94, 287. (b) Zbiral, E.;
Brandstetter, H. H.; Christian, R.; Schauer, R. Lieb. Ann.
Chem. 1987, 781.
19. Streicher, H.; Bohner, C. Abstracts P#15(47); 17th Inter-
national Symposium on Glycoconjugates, Bangalore,
India, Jan. 12–16, 2003.
1
methanol, 10:1). H NMR (600 MHz, CDCl3): d 6.40 (d,
1H, J2,P=21.3 Hz, H-2), 5.69 (m, 1H, NH), 5.56 (m, 1H,
H-3), 5.12 (m, 1H, H-20), 4.43 and 4.20 (2m, 1H, H-30a),
4.30 and 4.20 (2m, 1H, H-30b), 4.16 (m, 1H, H-4), 4.16–
4.02 (m, 6H, 2 CH2CH3), 3.73 and 3.45 (2m, 1H, H-10a),
3.72 and 3.53 (2m, 1H, H-10b), 3.67 (m, 1H, H-5), 2.82 (m,
1H, H-6a), 2.26 (m, 1H, H-6b), 2.08–1.99 (4s, 12H, 4
COCH3), 1.36–1.32 (m, 6H, 2 CH2CH3). 13C NMR
(150.9 MHz, CDCl3): d 138.5 (C-2), 75.1 (C-5), 71.8 (C-3),
69.9 (C-20), 53.9 (C-4), 30.1 (C-6). 31P NMR (242.9 MHz,
CDCl3): d 16.90 (s, 1P, P(O)(OEt)2). C21H34NO11P (M
507.48). MALDI-MS (pos. mode, DHB): 545.9
(M+K)+, 529.9 (M+Na)+, 507.9 (M+H)+.
15. Selected physical data for 1 h/l: 1H NMR (600MHz, D2O):
d 5.88 (d, 1H, J2,P=17.7 Hz, H-2), 4.09 (m, 1H, H-3), 3.87
(m, 1H, H-20), 3.62, 3.44, 3.00, 2.86 (4m, 4H, H-10a, H-10b,
H-30a, H-30b), 3.72 (m, 1H, H-4), 3.42 (m, 1H, H-5), 2.67 (m,
1H, H-6a), 2.09 (m, 1H, H-6b), 1.91 (s, 3H, NHCOCH3).
13C NMR (150.9 MHz, D2O): d 130.1 (C-2), 76.2 (C-5), 70.3
(C-3), 70.3, 65.7, 41.3 (C-10, C-20, C-30), 56.2 (C-4), 30.4 (C-
6). 31P NMR (242.9MHz, D2O): d 12.34 (s, 1P, PO32ꢀ).
C11H20NO8P (M 325.25) MALDI-MS (pos. mode, DHB):
347.1 (M+Na)+, 325.1 (M+H)+.
20. Meindl, P.; Tuppy, H. Monatsh. Chem. 1969, 100, 1295.
21. Wang, Q.; Wolff, M.; Polat, T.; Du, Y.; Linhardt, R. J.
Bioorg. Med. Chem. Lett. 2000, 10, 941.