Synthesis and evaluation of 4-O-alkylated 2-deoxy-2,3-didehydro-N-acetylneuraminic acid derivatives as inhibitors of human parainfluenza virus type-3 sialidase activity

Article abstract of DOI:10.1016/j.bmcl.2006.12.105

The X-ray crystal structure of the paramyxoviral surface glycoprotein haemagglutinin-neuraminidase (HN) from Newcastle Disease virus was used as a template to design inhibitors of the HN from human parainfluenza virus type-3 (hPIV-3). 4-O-Alkylated derivatives of 2-deoxy-2,3-didehydro-N-acetylneuraminic acid (Neu5Ac2en), accessed from 8,9-O-isopropylidenated-Neu5Ac2en1Me, were found to inhibit the sialidase (neuraminidase) activity of hPIV-3 (strain C243) in the range of 3-30 μM. This is comparable or improved activity compared to the parent 4-hydroxy compound.

Full text of DOI:10.1016/j.bmcl.2006.12.105

Bioorganic & Medicinal Chemistry Letters 17 (2007) 1655–1658
Synthesis and evaluation of 4-O-alkylated 2-deoxy-2,3-didehydro-
N-acetylneuraminic acid derivatives as inhibitors of human
parainfluenza virus type-3 sialidase activity
David J. Tindal,a Jeffrey C. Dyason,a Robin J. Thomson,a Takashi Suzuki,b
Hiroo Ueyama,b Yohta Kuwahara,b Naoyoshi Maki,b
Yasuo Suzukib and Mark von Itzsteina,*
aInstitute for Glycomics, Griffith University (Gold Coast Campus), PMB 50 Gold Coast Mail Centre, Qld 9726, Australia
bDepartment of Biochemistry, School of Pharmaceutical Sciences, University of Shizuoka, Core Research for Evolutional Science
and Technology (CREST), Japan Science and Technology Corporation, and COE Program in the 21st Century, 52-1 Yada,
Shizuoka City, Shizuoka 422-8526, Japan
Received 22 November 2006; revised 21 December 2006; accepted 22 December 2006
Available online 8 January 2007
Abstract—The X-ray crystal structure of the paramyxoviral surface glycoprotein haemagglutinin-neuraminidase (HN) from
Newcastle Disease virus was used as a template to design inhibitors of the HN from human parainfluenza virus type-3 (hPIV-3).
4-O-Alkylated derivatives of 2-deoxy-2,3-didehydro-N-acetylneuraminic acid (Neu5Ac2en), accessed from 8,9-O-isopropylidenat-
ed-Neu5Ac2en1Me, were found to inhibit the sialidase (neuraminidase) activity of hPIV-3 (strain C243) in the range of 3–
30 lM. This is comparable or improved activity compared to the parent 4-hydroxy compound.
Ó 2007 Elsevier Ltd. All rights reserved.
The paramyxoviridae family of viruses includes a num-
ber of respiratory pathogens that cause epidemics of ma-
jor medical and veterinary importance.1 Among these,
the human parainfluenza viruses (hPIV), types 1 to 3,
are leading causes of respiratory disease in infants and
young children, the immunocompromised, the chroni-
cally ill and the elderly (reviewed in Refs. 2 and 32,3).
While a number of agents show in vitro activity against
parainfluenza viruses (reviewed in Ref. 22), there are cur-
rently no antiviral drugs with proven clinical efficacy
against hPIV.2 The wide incidence of infection, and
the severity of disease in immunocompromised pa-
tients,2 suggests that work should be directed towards
the development of chemotherapeutic agents to treat
parainfluenza virus infections.
tinin-neuraminidase (HN).4 HN has three roles in the
pathogenesis of the virus; it mediates the attachment
of the virus to the host cell by interaction with sialic
acids (neuraminic acids) associated with cell-surface
glycoconjugates (e.g., 1);5,6 it promotes fusion activity
enabling viral entry into the cell;7 and it acts through
its sialidase (neuraminidase) function to cleave sialic
acids from the surface of nascent virion particles to
prevent self-agglutination and promote virion elution
from host cells.8 The multifunctional role of HN in the
viral life-cycle makes it an attractive target for the
development of therapeutics to treat hPIV infection. In
particular, inhibition of sialidase activity may potential-
ly reduce degradation of epithelial cell coating thereby
hindering secondary bacterial infection.9,10
Human parainfluenza viruses 1 and 3, and a number of
other paramyxoviridae including Newcastle Disease
virus (NDV) possess two membrane proteins, a fusion
protein and a multifunctional glycoprotein, haemagglu-
The first three-dimensional structure of a haemaggluti-
nin-neuraminidase protein, the HN from Newcastle
Keywords: Parainfluenza virus; Sialidase; Sialic acids; Inhibition;
Computational chemistry; Inhibitor design; Carbohydrates.
*
Corresponding author. Tel.: +61 7 5552 7016; fax: +61 7 5552
9040; e-mail: m.vonitzstein@griffith.edu.au
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.12.105

1656
D. J. Tindal et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1655–1658
Disease virus, was reported in 2000.11 This structure
showed the presence of a single ‘pliable’ active site that
could encompass both the sialic acid receptor-binding
and sialoside hydrolysis functions of the HN, with the
selection of function moderated by a conformational
switch in the protein.11 An X-ray crystal structure of
the HN protein of hPIV-3 reported in 2004 showed sim-
ilar structural flexibility of the active site.12 The location
of both the receptor-binding and sialidase functions in a
single site11,12 suggests that compounds designed to bind
to this site could minimise these functions, and block
both viral infection and virus spread.
via elimination of HCl from the b-chloride derivative 6,
to the unsaturated derivative 7 in ꢀ75% yield over three
steps. De-O-acetylation, followed by isopropylidenation
of the C-8 and C-9 hydroxyl groups, gave derivative 817
in high yield. Methods for the selective alkylation of the
C-4 hydroxyl group of neuraminic acid derivatives have
been reported in the literature.17,18 In this work, selec-
tive 4-O-alkylation of 8 in the presence of sodium hy-
dride gave the products 9a–e in moderate yield.
Alternatively, 4-O-alkylation could be carried out using
silver(I) oxide as catalyst, however the yields for the
straight alkyl derivatives 9c–e were considerably lower
using this method. Subsequent two-step deprotection
of compounds 9a–e gave the target 4-O-alkylated
Neu5Ac2en derivatives 4a–e in good yields.19
In comparison to the sialidases from influenza virus, and
from bacterial sources, there have been relatively few
studies of the inhibition of the paramyxovirus HN sial-
idase activity, and in particular of hPIV-3 HN activity.
In the case of hPIV-3 HN, the unsaturated N-acetylneu-
raminic acid derivative 2-deoxy-2,3-didehydro-N-acetyl-
neuraminic acid (Neu5Ac2en, 2), a naturally occurring
sialidase inhibitor, and the potent influenza virus
sialidase inhibitor 4-deoxy-4-guanidino-Neu5Ac2en
(zanamivir) 3 have been evaluated for their inhibition
of hPIV-3 sialidase activity in vitro,13 with IC50 values
reported in the low millimolar range for both com-
pounds. Interestingly, zanamivir has also been shown
to inhibit the receptor-binding and fusion processes of
hPIV-3,13,14 supporting the premise that an inhibitor
binding to the active site could interfere with both the
receptor-binding and sialidase functions of the HN.
Herein we report the synthesis of 4-O-alkylated Neu5A-
c2en derivatives 4, designed based on the three-dimen-
sional structure of NDV-HN,15 and their evaluation as
inhibitors of hPIV-3 HN sialidase activity.
The 4-O-alkylated Neu5Ac2en derivatives 4a–e were
evaluated for their ability to inhibit the sialidase action
of hPIV-3 (strain C243) in a whole virus assay20,21 using
4-methylumbelliferyl N-acetyl-a-D-neuraminide as the
fluorogenic enzyme substrate. The IC50 values of 4a–e
and the parent compound Neu5Ac2en 2, as well as 4-de-
oxy-4-guanidino-Neu5Ac2en 3 (zanamivir), are given in
Table 1. The 4-O-alkylated Neu5Ac2en derivatives with
the largest hydrophobic groups [4a (benzyl) 4b (2-phen-
yl-benzyl), and 4e (decyl)] showed comparable activity
to the parent 4-hydroxy derivative 2, while derivatives
with the smaller hydrophobic groups [4e (ethyl) and 4d
(hexyl)] showed an order of magnitude improvement
in inhibition over Neu5Ac2en. These data indicate that
hydrophobic groups [even large groups such as the 2-
phenylbenzyl substituent (4b)] introduced onto the C-4
hydroxyl of Neu5Ac2en can indeed bind into the active
site of hPIV-3 HN without detriment to binding affinity.
Analysis of the subsequently published 3D structure of
hPIV-3 HN12 shows that the active site cavity accommo-
dating the C-4 hydroxyl group of Neu5Ac or Neu5A-
c2en is in fact larger than that of NDV HN (used as
the design template in this study), although the environ-
ment of the pocket is slightly less hydrophobic than that
observed in NDV HN.
The design of hPIV inhibitors using the X-ray structure
of NDV HN has also been reported by Alymova et al.22
Two unsaturated neuraminic acid derivatives, C-4
substituted N-(2-methylpropionyl)-2-deoxy-2,3-didehy-
dro-neuraminic acid derivatives BCX-2798 10 and
BCX-2855 11, were reported to inhibit both the haemag-
glutination and sialidase activities of hPIVs 1-3, and the
growth of the viruses in cell culture.22 The IC50
values reported for BCX-2798 10 and BCX-2855 11
against hPIV-3 sialidase activity were 20 and 4.3 lM,
respectively. These values are comparable to those
obtained with the Neu5Ac2en derivatives 4a–e reported
At the beginning of this work, the sole three-dimension-
al structure of a paramyxoviral HN was that from
NDV.11 The crystal structure of NDV HN (1E8V)11
containing Neu5Ac2en (2) shows a large cavity around
the C-4 position of bound Neu5Ac2en, the hydroxyl
group of which makes no interactions with the protein.11
A GRID16 study of the active site showed an area of
hydrophobic interaction within the large C-4 cavity
and directed our molecular modelling studies to the de-
sign of a series of Neu5Ac2en derivatives substituted at
C-4 with hydrophobic groups that were envisaged to
bind well into the large active site. These derivatives of
Neu5Ac2en were anticipated to have enhanced binding
to NDV HN, and potentially to other paramyxoviral
HNs such as those from the human parainfluenza
viruses.
The target 4-O-alkylated Neu5Ac2en derivatives 4a–e
were synthesised using the approach shown in Scheme
1. N-Acetylneuraminic acid (5, Neu5Ac) was converted,

D. J. Tindal et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1655–1658
1657
Scheme 1. Reagents and conditions: (a) MeOH, cat. H+ resin, rt, o/n; (b) AcCl, rt, 2 d; (c) DBU, DCM, rt, 2.5 h (75% over 3 steps); (d) NaOMe,
MeOH, 0 °C, 2 h; (e) 2,2-dimethoxypropane, cat. H+ resin, acetone, rt, o/n (92% over 2 steps); (f) RBr (1.2 equiv), DMF, NaH (1.3 equiv), 0 °C,
0.5–4 h (9a,17 65%; 9b, 61%; 9c, 62%; 9d 61%; 9e, 33%); (g) i—RBr (1.2 equiv), DCM, 3 A mol. sieves, rt, 2 h; ii—Ag2O (3 equiv), n-Bu4NI, rt 2 d (9a,
˚
62%; 9b, 82%; 9c,18 40%; 9d 34%; 9e, 11%); (h) 80% AcOH, 0.5–2 h; (i) 0.1 M KOH, MeOH, rt, 1 h–3 d (ꢀ90% over 2 steps).
ment of more potent inhibitors of infection by viruses of
the Paramyxoviridae subfamily.
Table 1. Inhibition (IC50 lM) of hPIV-3 (strain C243)20 sialidase
activity by C-4 modified Neu5Ac2en derivatives 4a–ea
HO
H
CO2H
O
HO
Acknowledgments
R1
R2
HO
H
R1 = NHAc
We gratefully acknowledge the financial support of the
Australian Research Council (ARC) and the National
Health and Medical Research Council of Australia,
and Griffith University for the award of a Postgraduate
Research Scholarship and the Institute for Glycomics,
Griffith University, for an Institute Postgraduate Award
to D.J.T. M.vI. gratefully acknowledges the support of
the ARC for the award of a Federation Fellowship.
Compound
R2
Sialidase inhibition
(IC50 lM)
4a
4b
OCH2C6H5
OCH2C6H4-o-C6H5
OCH2CH3
10
30
3
4c
4d
OCH2(CH2)4CH3
OCH2(CH2)8CH3
OH
5
4e
2 (Neu5Ac2en)
3 (zanamivir)
10
24
25
NHC(@NH)NH2
a Sialidase activity was measured in
4-methylumbelliferyl N-acetyl-a-D-neuraminide as substrate.20
References and notes
a
fluorometric assay using
1. Mackie, P. L. Paediatric Resp. Rev. 2003, 4, 84.
2. Henrickson, K. J. Clin. Microbiol. Rev. 2003, 16, 242.
3. Glenzen, W. P.; Loda, F. A.; Denny, F. W. Parainfluenza
Viruses. In Viral infection of humans: epidemiology and
control; Evans, A. S., Ed., 3rd ed.; Plenum: New York,
1991; pp 493–503.
4. Lamb, R. A.; Kolakofsky, D. Paramyxoviridae: The
viruses and their replication. In Fundamental virology;
Fields, B. N., Knipe, D. M., Howley, P. M., Eds., 3rd ed.;
Lippincott-Raven: Philadelphia, 1996; pp 577–604.
5. Suzuki, T.; Portner, A.; Scroggs, R. A.; Uchikawa, M.;
Koyama, N.; Matsuo, K.; Suzuki, Y.; Takimoto, T.
J. Virology 2001, 75, 4604.
6. Villar, E.; Barroso, I. M. Glycoconj. J. 2006, 23, 5.
7. Lamb, R. A.; Paterson, R. G.; Jardetzky, T. S. Virology
2006, 344, 30.
8. Huberman, K.; Peluso, R. W.; Moscona, A. Virology
1995, 214, 294.
9. Peltola, V. T.; McCullers, J. A. Pediatr. Infect. Dis.
J. 2004, 23, S87.
10. Alymova, I. V.; Portner, A.; Takimoto, T.; Boyd, K. L.;
Babu, Y. S.; McCullers, J. A. Antimicrob. Agents Chemo-
ther. 2005, 49, 398.
here, that contain significantly different substitutions at
C-4.
The results of both the study described here and that
reported by Alymova et al.22 demonstrate that the crys-
tal structure of NDV HN is suitable as a model for the
design of hPIV-3 HN inhibitors. The X-ray structure of
hPIV-3 HN reported12 subsequent to this work will al-
low further, directed, structure-based design of potential
inhibitors of hPIV-3. With regard to inhibition of other
hPIVs, investigation of the 4-substituted neuraminic
acid derivatives as inhibitors of hPIV-1,22,23 for which
there is as yet no crystal structure, indicates that there
are differences in the C-4 binding pockets between the
HN of hPIV-3 and that of hPIV-1.
In summary, derivatives of Neu5Ac2en with hydropho-
bic substituents on the C-4 oxygen have been shown to
give comparable or improved inhibitory potency against
hPIV-3 sialidase activity compared to the parent com-
pound in vitro. This work sets the stage for the develop-
11. Crennell, S.; Takimoto, T.; Portner, A.; Taylor, G. Nat.
Struct. Biol. 2000, 7, 1068.
12. Lawrence, M. C.; Borg, N. A.; Streltsov, V. A.; Pilling, P.
A.; Epa, V. C.; Varghese, J. N.; McKimm-Breschkin, J. L.;
Colman, P. M. J. Mol. Biol. 2004, 335, 1343.

1658
D. J. Tindal et al. / Bioorg. Med. Chem. Lett. 17 (2007) 1655–1658
13. Greengard, O.; Poltoratskaia, N.; Leikina, E.; Zimmer-
berg, J.; Moscona, A. J. Virology 2000, 74, 11108.
14. Porotto, M.; Greengard, O.; Poltoratskaia, N.; Horga,
M.-A.; Moscona, A. J. Virology 2001, 75, 7481.
15. Tindal, D.J.; Dyason, J.C.; Thomson, R.J.; Suzuki, T.;
Suzuki, Y.; Ryan, C.; Zaitsev, V.; Taylor, G.; von Itzstein,
M. Design, synthesis, and evaluation of inhibitors of
paramyxoviral haemagglutinin-neuraminidase. 13th Euro-
pean Carbohydrate Symposium: Bratislava, Slovakia, 2005;
pp P157.
16. Goodford, P. J. J. Med. Chem. 1985, 28, 849.
17. Castro-Palomino, J. C.; Simon, B.; Speer, O.; Leist, M.;
Schmidt, R. R. Chem. Eur. J. 2001, 7, 2178.
18. Ikeda, K.; Sato, K.; Kitani, S.; Suzuki, T.; Maki, N.;
Suzuki, Y.; Sato, M. Bioorg. Med. Chem. 2006, 14, 7893.
19. All new compounds gave the expected spectroscopic and
analytical data.
20. Sialidase inhibition assay: Sialidase activity was assayed
by a fluorometric assay using 4-methylumbelliferyl N-
acetyl-a-D-neuraminic acid as previously described for
hPIV-1.21 hPIV-3 was isolated from infected LLC-MK2
cells as previously described for hPIV-1.21 Assays were
performed in duplicate and the determined values are
averages and reproducible.
21. Suzuki, T.; Ikeda, K.; Koyama, N.; Hosokawa, C.;
Kogure, T.; Takahashi, T.; Hidari, K. I. P. J.; Miyam-
oto, D.; Tanaka, K.; Suzuki, Y. Glycoconj. J. 2001, 18,
331.
22. Alymova, I. V.; Taylor, G.; Takimoto, T.; Lin, T.-H.;
Chand, P.; Babu, Y. S.; Li, C.; Xiong, X.;
Portner, A. Antimicrob. Agents Chemother. 2004, 48,
1495.
23. Tindal, D.J.; Dyason, J.C.; Thomson, R.J.; Suzuki, T.;
Suzuki, Y.; von Itzstein, M. in preparation.