Galactose-conjugates of the oseltamivir pharmacophore - New tools for the characterization of influenza virus neuraminidases

Article abstract of DOI:10.1039/b903394g

We describe the synthesis of mimetics of the α2-3 and α2-6 sialogalactoside substrates of influenza neuraminidase which include the oseltamivir pharmacophore, and report the sub-nanomolar affinities for these novel neuraminidase inhibitors. The challenge

Full text of DOI:10.1039/b903394g

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Galactose-conjugates of the oseltamivir pharmacophore—new tools for the
characterization of influenza virus neuraminidases†
Benoit Carbain,a Stephen R. Martin,b Patrick J. Collins,c Peter B. Hitchcocka and Hansjo¨rg Streicher*a
Received 17th February 2009, Accepted 20th March 2009
First published as an Advance Article on the web 23rd April 2009
DOI: 10.1039/b903394g
We describe the synthesis of mimetics of the a2–3 and a2–6 sialogalactoside substrates of influenza
neuraminidase which include the oseltamivir pharmacophore, and report the sub-nanomolar affinities
for these novel neuraminidase inhibitors. The challenge of synthesizing a Phospha-Oseltamivir/
Tamiphosphor monoester involving the secondary 3-hydroxy group of galactose required to mimic the
a2–3 sialogalactoside has been overcome by palladium-promoted coupling of the oseltamivir-derived
vinyl iodide with a protected galactose-3-phosphonate. The difference in binding of these two inhibitors
to a given influenza neuraminidase should be a function of its a2–3/a2–6-selectivity, an important, but
not yet fully understood factor in the adaptation of highly pathogenic avian influenza viruses to human
hosts.
Introduction
Influenza continues to be a major health concern for the world
population with epidemics occurring worldwide on a yearly basis.1
Influenza is caused by negative-strand RNA viruses, classified into
influenza types A and B, which are important targets of current
medicinal chemistry. In particular, type A influenza viruses, most
of which are avian viruses, are at the centre of attention as a
possible source of a severe flu pandemic.2,3 They are further
classified as subtypes based on the antigenic behaviour of their two
surface glycoproteins, a haemagglutinin (subtypes H1–H16) and
a neuraminidase (subtypes N1–N9).4 The interplay between the
neuraminidase (NA) and haemagglutinin (HA) crucially affects
both the host range and pathogenicity of influenza virus subtypes.
Consequently, one important aim of medicinal chemists, virolo-
gists, and structural biologists alike is to understand the molecular
basis of how an avian influenza virus, such as the notorious
H5N1 virus, could adapt to the environment of the human
respiratory tract and thus becomes a major threat to human health.
The natural receptors (for HA) and substrates (for NA) for
influenza viruses are the N-acetylneuraminic acids on the host cell
surface, which are a-ketosidically linked to galactose (Fig. 1). A
relatively clear picture has emerged for the trimeric haemagglu-
tinins. Those of human viruses bind more strongly to a2–6-linked
sialylgalactosides (predominant in the human lung) and those of
avian viruses bind more strongly to a2–3-linked sialylgalactosides
(predominant in the intestines of birds) (Fig. 1).5
Fig. 1 Sialoglycoconjugates, influenza neuraminidase inhibitors and
targeted galactose-conjugates.
aDepartment of Chemistry and Biochemistry, University of Sussex, Brighton,
UK BN1 9QG. E-mail: h.streicher@sussex.ac.uk; Fax: +44-1273-876687
bDivision of Physical Biochemistry, National Institute of Medical Research,
Mill Hill, London, UK NW7 1AA
For the tetrameric neuraminidases, which hydrolyze the keto-
sidic linkage in sialosides and thus cleave the neuraminic acid
from the host cell sialoglycoconjugates, the situation is less clear.
Avian influenza neuraminidases have a pronounced preference for
a2–3-linked sialylgalactosides while human influenza neuramini-
dases act on both linkages but, in the majority of studies
cDivision of Virology, National Institute of Medical Research, Mill Hill,
London, UK NW7 1AA
† Electronic supplementary information (ESI) available: 1H-, 13C- and 31P-
NMR spectra. CCDC reference number 720847. For ESI and crystallo-
graphic data in CIF or other electronic format see DOI: 10.1039/b903394g
2570 | Org. Biomol. Chem., 2009, 7, 2570–2575
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
reported, still retain a higher activity with the a2–3-linked
sialylgalactosides.6 These data are generally obtained by analyzing
the reaction kinetics for the hydrolysis of the two substrates by a
given neuraminidase or whole virus.
Unfortunately, compound 4, the mimetic of an a2–3-linked
sialylgalactoside (Fig. 1), proved not to be accessible via
these routes, which so far remain limited to primary hydroxy
groups.
In this contribution, we suggest inhibitor-based tools for
the characterization of influenza neuraminidases. Oseltamivir
carboxylate, marketed as its ester prodrug TamifluTM, is a sub-
nanomolar inhibitor of all influenza virus A neuraminidases
(Fig. 1).7 We recently published an effective synthesis of highly
active phospha-isosteres 1–3 of oseltamivir, where a phosphonate
or phosphonate monoester replaces the carboxylate without
In order to alleviate the steric congestion probably responsible
for the problem, we decided to attempt a direct coupling of the
appropriate vinyl iodide 68,10 (Fig. 2) with the mixed methyl,
3-O-galactosyl phosphonate 10. To obtain 10, the methoxymethyl
protecting group was introduced into selectively protected
methyl b-D-galactopyranoside 711 in 55% yield to give 8, followed
by cleavage of the 3-O-allyl group with PdCl2 in 69% yield to give
9 (Scheme 2).
impairing the potency of the oseltamivir pharmacophore.8
3
represents a chemically stable mimetic of an a2–6-linked sialyl-
galactoside which has been shown to inhibit the neuraminidase
activity of influenza virus A/Norway/1758/07 with an inhibitory
constant in the sub-nanomolar range (Fig. 1).8
Any differential inhibition of an influenza neuraminidase by
3 and the corresponding a2–3 sialylgalactoside mimetic 4 would
necessarily depend on interactions beyond the primary binding
site and hence be determined by linkage recognition. For example,
such data could help to shed light on the adaptation process
undergone by the notorious bird flu virus (H5N1) when it changes
preference from an a2–3 sialylgalactoside receptor environment
in birds to the a2–6 sialylgalactoside receptor environment of the
human lung.
Results and discussion
Synthesis
Cyclohexenephosphonate monoesters of type 2 and 3 have so
far been synthesized by us through alkylation of monomethyl
phosphonates such as 5 with the protected sugar triflate or
condensation under Mitsunobu conditions (Scheme 1).8,9
Fig. 2 Crystal structure of vinyl iodide 6.
The phosphonate was introduced by esterification with dimethyl
chlorophosphonate,12 yielding initially the triester which was
immediately hydrolyzed on silica to give diester 10 in 62%
yield.
Coupling of phosphonate 10 with vinyl iodide 6 using Pd(PPh3)4
proceeded with only moderate 30% yield which was, however, suf-
ficient for our purposes. Interestingly, a vinyl diphenylphosphine
oxide by-product arising from coupling with displaced palladium
ligand was also isolated. Selective phosphonate demethylation
converted the mixture of diastereomers 11 into the monoester
12 in 74% yield. Final deprotection was achieved in aqueous
trifluoroacetic acid, furnishing sialoside mimetic 4 in 61% yield
after purification by gel-permeation chromatography. A spin-echo
difference (SED) spectrum13 of 4 unequivocally confirmed the
linkage of the ‘Phospha-Tamiflu’ moiety to the 3-position of the
galactose (see also the ESI†).
Inhibition of neuraminidase activity
To compare the inhibitory potency of 3, 4 and oseltamivir, a whole-
virus assay using allantoic fluid from infected eggs was employed.
In this standard assay, which is widely used in the influenza field,
Scheme 1 Synthetic routes to galactose conjugates of the oseltamivir
motif (PG = protecting group).
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 2570–2575 | 2571
©

View Article Online
Experimental
General
Reaction solvents were purchased anhydrous and used as received.
Solvents for chromatography were distilled before use. Reactions
were monitored by TLC using precoated silica gel 60 F254
plates. Compounds were detected by UV absorption and/or by
staining with a molybdenum phosphate reagent (20 g ammonium
molybdate and 0.4 g cerium(IV) sulfate in 400 mL of 10% aq.
sulfuric acid) or a basic KMnO4-solution and subsequent heating
at 120 ◦C for 5 min. Silica gel 60 A ‘Davisil’ (particle size 35–
70 mm) from Fisher Scientific, UK was used for flash chromatog-
1
raphy. H NMR, 13C NMR, 31P NMR and all multidimensional
NMR spectra were recorded on Varian VNMRS spectrometers
(600 MHz, 500 MHz or 400 MHz, see compound characterisation
for individual experiments). Chemical shifts in 1H NMR and 13
C
NMR spectra were referenced to the residual proton resonance
of the respective deuterated solvent, CDCl3 (7.24 ppm), D2O
(4.80 ppm), D2O in CD3OD (4.88 ppm). For 31P NMR spectra
H3PO4 was used as external standard (0 ppm). In some cases,
13C chemical shifts were deduced from heteronuclear multiple spin
correlation (HSQC) spectra. The H-6ax and H-6eq assignments
refer to the pseudoaxial and pseudoequatorial protons in the cy-
clohexene systems, respectively, obtained by ROESY spectroscopy.
In pseudo-disaccharidic systems, the cyclohexene ring is indicated
by the suffix ‘a’, the sugar by the suffix ‘b’. Diastereomeric mixtures
of mixed diesters are indicated by the suffix h (higher moving) and
l (lower moving) but no attempts of separation were made. HR-
ESI MS spectra were recorded on a Bruker Daltonics Apex III in
positive mode with MeOH/H2O as solvent.
Fluorescence was measured in a JASCO FP-6300 fluorimeter.
Silica-based MPLC chromatography was carried out on the
Bu¨chi Sepacore system equipped with glass columns packed
with LiChroprep Si 60 (15–25 mm) from Merck, Darmstadt,
Germany. Gel permeation chromatography was carried out in the
1–10 mg scale on a XK 16/70 column (bed volume 130 mL),
from Amersham packed with Sephadex G-10 (particle size 40–
120 mm) and 0.1 M NH4HCO3 as buffer. Detection was achieved
with a differential refractometer from Knauer, Berlin, Germany.
Fine chemicals were purchased from Aldrich-, Sigma- or Acros-
Chemicals and were of the highest purity available.
Scheme 2 Synthesis of an a2–3 sialylgalactoside mimetic.
Table 1 Inhibition constants (Ki [nM]) for target phosphonatesa
3
4
Oseltamivir
0.14 0.02
0.74 0.08
0.29 0.04
a Inhibition of MUNANA hydrolysis catalysed by neuraminidase from
influenza virus A/Norway/1758/07; inhibition constants were determined
as described in the ESI†; KM for MUNANA = 6.4 0.6 mM.
(Methyl-b-D-galactopyranos-3-yl) [(3R,4R,5S)-4-acetamido-3-
amino-5-(1-ethylpropoxy)-1-cyclohexene-1-phosphonate](4). Com-
pound 12 (6 mg, 8.2 mmol) was dissolved in a solution of 1:1
TFA/H2O (2 mL) and the mixture was stirred at room temperature
for 3 hours. The mixture was lyophilised and the residue was
purified by gel permeation chromatography to afford compound
4 (2.5 mg, 5 mmol, 61%) as a white powder after lyophilisation.
the inhibition constant Ki is obtained by measuring the effect of
the inhibitor on the rate of MUNANA hydrolysis.14
When tested for inhibition of the neuraminidase activity of an
H1N1 virus (A/Norway/1758/07), compounds 3 and 4 indeed
displayed a significant difference, with the a2–3 sialylgalactoside
mimetic 4 inhibiting somewhat more strongly (Table 1). In the
absence of detailed structural data, this finding cannot be directly
correlated with the substrate specificity of the virus but it does
demonstrate that the set of the two compounds is useful to
establish a ‘fingerprint’ for a neuraminidase which is, at least
partly, governed by its substrate specificity. In other words, the
ratio of the inhibition constants of 3 and 4 should be to some extent
proportional, or inversely proportional, to the ratio of the binding
constants of the respective a2–3 and a2–6 sialylgalactoside
substrates.
dH (600 MHz, D2O) 0.87 (3 H, t, J
7.4 Hz, pentyl-CH3), 0.92
2,P
(3 H, t, J 7.4 Hz, pentyl-CH3), 1.44–1.52 (1 H, m), 1.54–1.63
(3 H, m), 2.10 (3 H, s, COCH3), 2.37–2.45 (1 H, m, 6a¢-H), 2.73–
2.81 (1 H, m, 6a-H), 3.27 (1 H, s, 5a-H), 3.54 (1 H, tt, J 5.6 Hz,
pentyl-CH), 3.60 (3 H, s, OCH3), 3.64 (1 H, dd, J 9.6, 8.2 Hz,
2b-H), 3.71–3.84 (3 H, m, H-6b, H-6b¢, 5b-H), 3.89 (1 H, dd,
J4a,5a 11.0, J4a,3a 9.4 Hz, 4a-H), 4.02 (1 H, ddd, J3b,2b 8.8, J3b,4b
3.4 Hz, 3b-H), 4.08 (1 H, d, J4b,3b 3.3 Hz, 4b-H), 4.22 (1 H, d,
J
J
3a,4a 8.5 Hz, 3a-H), 4.39 (1 H, d, J1b,2b 8.3 Hz, 1b-H), 6.42 (1 H, d,
2a,P 20.4 Hz, 2a-H); dC (150.9 MHz, D2O) 8.48, 8.51, 25.10, 22.26
2572 | Org. Biomol. Chem., 2009, 7, 2570–2575
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
(NHCOCH3), 25.45, 31.15 (6a-C), 49.49 (5a-C), 57.12 (OCH3),
60.80 (6b-C), 68.35 (4b-C), 69.66 (d, J2b,P 5.6 Hz, 2b-C), 74.68
(5b-C), 76.32 (d, J3a,P 20.4 Hz, 3a-C), 76.61 (d, J3b,P 5.8 Hz, 3b-C),
81.14, 84.08 (pentyl-CH), 103.52 (1b-C), 136.77 (d, J2a,P 7.3 Hz,
2a-C), 174.92 (NHCOCH3); dP (242.9 MHz, D2O) 13.06;
HR-ESI-MS calculated for C19H35N2O10P (M + Na)+ 519.2078,
found 519.2062.
69%). Rf = 0.17 (Tol:EA 1:1). dH (500 MHz, CDCl3) 3.01 (1 H,
bd, J 5.4 Hz, OH), 3.41–3.44 (4 H, m, 5-H, methoxymethyl-CH3),
3.52–3.55 (3 H, m, OCH3), 3.62–3.70 (2 H, m, 3-H, 2-H), 4.05
(1 H, d, J6¢,6 12.5 Hz, 6¢-H), 4.20 (1 H, s, 4-H), 4.25 (1 H, d, J1,2
6.5 Hz, 1-H), 4.32 (1 H, d, J6,6¢ 12.5 Hz, 6-H), 4.76, 4.84 (2 H, 2d,
J 6.5 Hz, methoxymethyl-CH2), 5.53 (1 H, s, CHPh), 7.30–7.37
(3 H, m), 7.46–7.52 (2 H, m); dC (125.8 MHz, CDCl3) 56.03, 57.15
(OCH3), 66.71 (5-C), 69.30 (6-C), 72.48, 75.78 (2-C, 4-C), 77.92
(3-C), 97.74, 101.76 (pentyl-CH), 103.77 (1-C), 126.74, 128.38,
129.36, 137.81; HR-ESI-MS calculated for C16H22O7 (M + Na)+
349.1252960, found 349.1257742.
Ammonium [methyl (3R,4R,5S)-4-acetamido-3-(1,1-dimethyl-
ethyloxycarbonylamino)-5-(1-ethylpropoxy)-1-cyclohexene-1-phos-
phonate] (5). The synthesis of compound 5 has been described
by us before.8
O-Methyl O-(methyl 2-O-methoxymethyl-4,6-O-benzylidene-b-
D-galactopyranos-3-yl) phosphonic acid (10). Under an atmo-
sphere of nitrogen, compound 9 (31 mg, 0.095 mmol) was dissolved
in dry CH2Cl2 (1 mL). N,N-diisopropyl ethyl amine (49 mL,
0.285 mmol) was added to the solution which was then cooled to
(3R,4R,5S)-4-Acetamido-3-(1,1-dimethylethyloxycarbonylamino)-
5-(1-ethylpropoxy)-1-iodocyclohexene (6). Compound
6 was
synthesised from the corresponding azide as described before.8
Crystals were obtained from a solution in CH2Cl2.
◦
0 C. Chloro(dimethyl)phosphine12 (40 mL) was added dropwise,
Methyl 3-O-allyl-4,6-O-benzylidene-b-D-galactopyranoside (7).
Compound 7 was synthesised as described in the literature.15
the reaction mixture was allowed to come to room temperature
and stirred overnight. Solid NaHCO3 (50 mg) was added to the
reaction mixture, followed by MeOH (0.5 mL) and silica (50 mg).
The mixture was evaporated in vacuo to dryness. Purification
by flash chromatography (Tol:EA:MeOH 1:1:0 → 0:10:1) gave
compound 10 (24 mg, 0.059 mmol, 62%). Upper spot on tlc: Rf =
0.18, lower spot on tlc: Rf= 0.16 (EA:MeOH 20:1). NMR data
of the isomer corresponding to the upper spot: dH (600 MHz,
CDCl3) 3.38 (3 H, s, methoxymethyl-CH3), 3.44 (1 H, s, 5-H),
3.54 (3 H, s, OCH3), 3.70 (3 H, d, J 12.1 Hz, POCH3), 3.90 (1
H, dd, J2,3 9.8, J2,1 7.7 Hz, 2-H), 4.05 (1 H, dd, J6¢,6 12.5, J6¢,5
1.6 Hz, 6¢-H), 4.28 (1 H, d, J1,2 7.7 Hz, 1-H), 4.32 (1 H, dd, J6,6¢
12.4, J6,5 1.2 Hz, 6-H), 4.38 (1 H, d, J 3.7 Hz, 4-H), 4.50–4.45
(1 H, m, 3-H), 4.72, 4.87 (2 H, 2d, J 6.3 Hz, methoxymethyl-
CH2), 5.55 (1 H, s, CHPh), 6.93 (1 H, d, JH,P 717.2 Hz, P-H),
7.32–7.36 (3 H, m), 7.48–7.51 (2 H, m); dC (150.9 MHz, CDCl3)
51.90 (d, J 5.8 Hz, POCH3), 56.42 (methoxymethyl-CH3), 57.23
(OCH3), 66.13 (5-C), 69.14 (6-C), 73.98 (d, J2,P 5.7 Hz, 2-C),
75.85 (4-C), 76.25 (d, J3,P 5.9 Hz, 3-C), 97.64 (methoxymethyl-
CH2), 101.43 (benzylidene-CH), 104.17 (1-C), 126.57, 128.40,
129.33, 137.73; dP (161.9 MHz, CDCl3) (8.19), 9.92; HR-ESI-
MS calculated for C17H25O9P (M + Na)+ 427.1143900, found
427.1128400.
Methyl 2-O-methoxymethyl-3-O-allyl-4,6-O-benzylidene-b-D-
galactopyranoside (8). To a solution of compound 7 (0.514 g,
1.59 mmol) in dry THF (4 mL), cooled to 0 ◦C in an atmosphere
of nitrogen, was added NaH (57.4 mg, 2.39 mmol) as a suspension
in dry THF (3mL). The mixture was stirred for 15 min and
then cooled to -20 ◦C. MOMCl (157 mL, 2.07 mmol) was
added dropwise, the mixture was allowed to warm up to room
temperature and stirred overnight. The reaction was quenched by
addition of saturated aqueous NH4HCO3 (5 mL). CH2Cl2 (40 mL)
was added and the mixture was extracted with NH4HCO3 (2 ¥
10 mL), washed with brine (2 ¥ 10 mL), the organic phase was
dried over MgSO4 and evaporated. Purification of the residue
by flash chromatography (Tol:EA 5:1 → 1:5) gave compound 8
(0.323 g, 0.88 mmol, 55%) as a white solid. Rf = 0.53 (Tol:EA
1:1). dH (500 MHz, CDCl3) 3.34 (1 H, s, 5-H), 3.42 (3 H, s,
methoxymethyl-CH3), 3.44 (1 H, dd, J3,2 9.8, J3,4 3.6 Hz, 3-H),
3.53 (3 H, s, OCH3), 3.84 (1 H, dd, J2,3 9.7, J2,1 7.9 Hz, 2-H), 4.03
(1 H, dd, J6¢,6 12.3, J6¢,5 1.6 Hz, 6¢-H), 4.11–4.21 (3 H, m, 4-H, allyl-
OCH2), 4.25 (1 H, d, J1,2 7.8 Hz, 1-H), 4.29 (1 H, dd, J6,6¢ 12.3,
J6,5 1.4 Hz, 6-H), 4.80, 4.82 (2 H, 2d, J 6.2 Hz, methoxymethyl-
CH2), 5.16 (1 H, dd, J 10.4, 1.5 Hz, allyl-CHCH2), 5.29 (1 H, dd,
J 17.2, 1.6 Hz, allyl-CHCH2), 5.50 (1 H, s, CHPh), 5.91 (1 H,
ddd, J 17.3, 10.4, 5.7 Hz, allyl-CHCH2), 7.31–7.36 (3 H, m), 7.50
(2 H, m); dC (125.8 MHz, CDCl3) 56.07 (methoxymethyl-CH3),
56.79 (OCH3), 66.61 (5-C), 69.40 (6-C), 71.02, 73.72 (4-C), 74.46
(2-C), 79.21 (3-C), 97.65 (methoxymethyl-CH2), 101.52 (Ph-CH),
104.22 (1-C), 117.34, 126.72, 128.24, 129.08, 135.17, 138.01; HR-
ESI-MS calculated for C19H26O7 (M + Na)+ 389.1575350, found
389.1570743.
Methyl (methyl 2-O-methoxymethyl-4,6-O-benzylidene-b-D-gal-
actopyranos-3-yl) [(3R,4R,5S)-4-acetamido-3-(1,1-dimethylethyl-
oxycarbonylamino)-5-(1-ethylpropoxy)-1-cyclohexene-1-phospho-
nate] (11). The phosphinic acid 10, tetrakistriphenylphosphine
palladium (7 mg, 6.3 mmol) and vinyl iodide 6 (20 mg, 0.042 mmol)
were mixed under an atmosphere of dry nitrogen. The mixture
was dissolved in anhydrous toluene (3 mL) and triethylamine
(8.8 mL, 0.063 mmol) was added. The mixture was stirred at
85 ◦C for 2.5 hours. After cooling to room temperature, the
reaction was quenched by addition of saturated aqueous NH4Cl
(10 mL). CH2Cl2 (15 mL) was added to the solution which was
then extracted with NH4Cl (5 mL), washed with brine (2 ¥ 5 mL),
the organic phase was dried over MgSO4 and the solvent was
evaporated. Purification by flash chromatography (Tol:EA:MeOH
1:2:0 → 0:10:1) gave a mixture of phosphonate diastereoisomers
11 (9 mg, 0.012 mmol, 30%). Rf = 0.34 (EA:MeOH 10:1).
The product contains a cyclohexenenyl diphenylphosphine oxide
impurity which could be separated and characterised only after
Methyl 2-O-methoxymethyl-4,6-O-benzylidene-b-D-galactopy-
ranoside (9). A mixture of compound 8 (0.316 g, 0.86 mmol),
NaOAc (0.212 g, 2.59 mmol) and PdCl2 (0.168 g, 0.95 mmol)
in AcOH (4 mL) and H2O (0.4 mL) was stirred for 5 hours
at room temperature. The reaction mixture was evaporated in
vacuo. The residue was dissolved in CH2Cl2 (20 mL) and washed
successively with saturated aqueous NaHCO3 (10 mL) and H2O
(10 mL), the organic phase was dried over MgSO4 and evaporated
to dryness. Purification of the residue by flash chromatography
(Tol:EA; 3:1 → 1:3) gave compound 9 (0.194 g, 0.595 mmol,
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 2570–2575 | 2573
©

View Article Online
the following synthetic step (compound 12). dH (500 MHz, CDCl3)
6.53/6.66 (1 H, 2d, J2a,P 22 Hz, 2a-H of the 2 diastereoisomers); dP
(161.9 MHz, CDCl3) 18.07/18.93 (2 diastereoisomers); HR-ESI-
MS calculated for C35H55N2O13P (M + Na)+ 765.3321640, found
765.3333975.
(Methyl 2-O-methoxymethyl-4,6-O-benzylidene-b-D-galactopy-
ranos-3-yl) [(3R,4R,5S)-4-acetamido-3-(1,1-dimethylethyloxy-
carbonylamino)-5-(1-ethylpropoxy)-1-cyclohexene-1-phosphonate]
(12). Under an atmosphere of dry nitrogen, compound 11 (8 mg,
0.011 mmol) was dissolved in anhydrous THF (1 mL). Anhydrous
NEt3 (22 mL, 0.156 mmol) and thiophenol (8 mL, 0.078 mmol) were
added to the solution. After 48 h of stirring at room temperature,
more anhydrous NEt3 (22 mL, 0.156 mmol) and thiophenol (8 mL,
0.078 mmol) were added to the mixture which was stirred for
another 24 h at room temperature. After evaporation of the sol-
vent, purification by flash chromatography (EA:MeOH 10:1→1:5)
afforded compound 12 (6 mg, 8.2 mmol, 74%). dH (600 MHz,
CD3OD) 0.88–0.95 (6 H, m), 1.41 (9 H, s), 1.43–1.60 (4 H, m),
1.97 (3 H, s), 2.35–2.43 (1 H, m, 6a¢-H), 2.70–2.77 (1 H, m, 6a-
H), 3.42–3.45 (1 H, m, pentyl-CH), 3.46 (3 H, s, methoxymethyl-
CH3), 3.57 (4 H, s, 5b-H, OCH3), 3.71–3.77 (1 H, m, 5a-H), 3.81
(1 H, dd, J 8.7 Hz, 2b-H/3b-H), 3.89 (1 H, dd, J 10.0 Hz, 4a-
H), 4.08 (1 H, d, J 6.7 Hz, 3a-H), 4.14–4.21 (2 H, m, 6b¢-H-,
2b-H/3b-H), 4.23 (1 H, d, J6b,6b¢ 12.4 Hz, 6b-H), 4.36–4.41 (2 H,
m, 1b-H, 4b-H), 4.82–4.89 (1 H, 2d, J 5.9 Hz, methoxymethyl-
CH2), 5.65 (1 H, s, CHPh), 6.48 (1 H, d, J2a,P 19.8 Hz, 2a-
H), 7.33–7.38 (3 H, bs), 7.52–7.56 (2 H, bs); dC (150.9 MHz,
CD3OD) 9.84, 10.44, 23.19, 26.91, 27.55, 28.91, 32.96 (d, J6a,P
9.6 Hz, 6a-C), 51.20 (d, J5a,P 13.6 Hz, 5a-C), 56.64 (methoxymethyl-
CH3), 56.94 (d, J4a,P 2.6 Hz, 4a-C), 57.31 (OCH3), 67.86 (5b-
C), 70.19 (6b-C), 74.88 (d, J3b-C/2b-C,P 5.95 Hz, 3b-C/2b-C), 76.05
(d, J3b-C/2b-C,P 5.1 Hz, 3b-C/2b-C), 77.04 (4b-C), 78.10 (d, J3a,P
19.85 Hz, 3a-C), 80.12, 83.39 (pentyl-CH), 98.20 (methoxymethyl-
CH2), 102.27 (benzylidene-CH), 105.40 (1b-C), 127.76, 129.04,
129.85, 137.66 (m, 2a-C), 139.95, 158.09 (NHCOOC(CH3)3),
173.71 (NHCOCH3); dP (161.9 MHz, CD3OD) 10.87; HR-ESI-
MS calculated for C34H53N2O13P (M + Na)+ 775.3140430, found
775.3153421.
The reduced rate of hydrolysis of MUNANA observed in
the presence of inhibitor is predicted by the following equation
(Rameix-Welti et al.)14a
:
where VI is the steady-state rate for MUNANA hydrolysis in the
presence of inhibitor at concentration [I], V0 is the steady-state
rate for MUNANA hydrolysis in the absence of inhibitor, [S] is the
MUNANA concentration, Km is the Michaelis–Menten constant
for hydrolysis of MUNANA, and Ki is the dissociation constant
for the enzyme–inhibitor complex. Because these inhibitors show
a slow approach to the new steady state rate (see Collins et al.14b) it
was necessary to confirm that the new steady-state rate had been
reached; this was done by demonstrating that the first derivative of
the fluorescence change (proportional to the NA activity) became
constant at long times.
Conclusion
Our work provides access to galactose-conjugates of the os-
eltamivir motif mimicking the natural a2–3 and a2–6 sialyl-
galactoside substrates of influenza virus neuraminidases. We
found a difference in binding depending on the position of
the oseltamivir pharmacophore on a b-galactoside, the natural
sialic acid aglycone. We conclude that this difference must be
a function of the preference of a given neuraminidase, and the
respective virus, for a a2–3 or a a2–6 sialylgalactoside substrate.
Consequently, the set of 3 and 4 can be used to characterize
a neuraminidase beyond the primary sialic acid or oseltamivir
binding site, based on inhibition.
Abbreviations
Inhibition of neuraminidase activity
THF, tetrahydrofuran; TBDPS, tert-butyldiphenylsilyl; EA, ethyl
acetate; Tol, toluene; TFA, trifluoroacetic acid; TBAF, tetrabuty-
lammonium fluoride; TBAI, tetrabutylammonium iodide, DMAP,
p-dimethylaminopyridine; BnBr, benzyl bromide; DMF, dimethyl
formamide; MES, 4-morpholinoethanesulfonic acid; MOMCl,
methoxymethyl chloride; MUNANA, 2¢-(4-methylumbelliferyl)-
a-D-N-acetylneuraminic acid.
Neuraminidase (NA) enzymatic activity was studied using
the fluorescent substrate 2¢-(4-methylumbelliferyl)-a-D-N-acetyl-
neuraminic acid (MUNANA). Measurements were made at 37 ◦C
in 32.5 mM MES (pH 6.5) + 4 mM CaCl2 using a JASCO
FP-6300 fluorimeter with excitation at 365 nm and emission
at 450 nm. Michaelis–Menten constants for the enzyme, Km
=
(k2 + k-1)/k1, were determined using standard initial rate measure-
ments with estimated neuraminidase concentrations in the range
0.1 to 0.5 nM and MUNANA concentrations in the range 2 to
200 mM.
Inhibition constants (KI) were determined by measuring the
extent to which different concentrations of inhibitor reduced the
steady-state rate of MUNANA hydrolysis. The data were inter-
preted using the following simple model for competitive inhibition
in which E, S, and I represent neuraminidase, MUNANA, and
inhibitor, respectively:
Acknowledgements
We thank Dr. Iain Day for his help with the NMR experiments and
F. Hoffmann-La Roche Ltd for the donation of valuable starting
material.
References
1 R. B. Belshe, New Eng. J. Med., 2005, 353, 2209–2211.
2 M. D. de Jong and T. T. Hien, J. Clin Virol., 2006, 35, 2–13.
2574 | Org. Biomol. Chem., 2009, 7, 2570–2575
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
3 E. de Clercq and J. Neyts, Trends Pharm. Sci, 2007, 28, 280–285.
4 World Health Organisation, Bull. World Health Organ., 1980, 58, 585–
591.
5 (a) I. A. Wilson, J. J. Skehel and D. C. Wiley, Nature, 1981, 289, 366–
373; (b) Y. Suzuki, T. Ito, T. Suzuki, R. E. Holland, Jr., T. M. Chambers,
M. Kiso, H. Ishida and Y. Kawaoka, J. Virol., 2000, 74, 11825–11831.
6 D. Katinger, L. Mochalova, A. Chinarev, N. Bovin and J. Romanova,
Arch. Virol., 2004, 149, 2131–2140; S. J. Baigent, R. C. Bethell and J. W.
McCauley, Virology, 1999, 263, 323–328.
7 (a) C. U. Kim, W. Lew, M. A. Williams, L. Zhang, X. Chen, P. A.
Escarpe, D. B. Mendel, W. G. Laver and R. C. Stevens, J. Med. Chem.,
1998, 41, 2451–2460; (b) P. J. Collins, L. F. Haire, Y. P. Lin, R. J. Russell,
P. A. Walker, J. J. Skehel, S. R. Martin, A. J. Hay and S. J. Gamblin,
Nature, 2008, 453, 1258–1261.
8 B. Carbain, P. J. Collins, L. Callum, S. R. Martin, A. J. Hay, J. McCauley
and H. Streicher, ChemMedChem, 2009, 4, 335–337.
11 A. Larsson, J. Ohlsson, K. W. Dodson, S. J. Hultgren, U. Nilsson and
J. Kihlberg, Bioorg. Med. Chem., 2003, 11, 2255–2262.
12 Dimethyl chlorophosphonate was synthesised according to the method
of S.-B. Chen, Y.-M. Li, S.-Z. Luo, G. Zhao, B. Tan and Y.-F. Zhao,
Phosphorous Sulfur Silicon, 2000, 164, 277–291.
13 J. S. Cohen, C.-W. Chen and A. Bax, J. Magn. Res., 1984, 59, 181–
184.
14 (a) M. A. Rameix-Welti, F. Agou, P. Buchy, S. Mardy, J. T. Aubin,
M. Veron, S. van der Werf and N. Naffakh, Antimicrobial Agents
Chemother., 2006, 50, 3809–3815; (b) P. J. Collins, L. F. Haire, J. Liu,
R. J. Russel, P. A. Walker, J. J. Skehel, S. R. Martin, A. J. Hay and
S. J. Gamblin, Nature, 2008, 453, 1258–1262; (c) R. G. Duggleby, P. V.
Attwood, J. C. Wallace and B. D. Keech, Biochemistry, 1982, 21, 3364–
3370; (d) W. M. Kati, A. S. Saldivar, F. Mohamadi, H. L. Sham, W. G.
Laver and W. E. Kohlbrenner, Biochem. Biophys. Res. Comm., 1998,
244, 408–413; (e) M. Z. Wang, C. Y. Tai and D. B. Mendel, Antimicrobial
Agents Chemother., 2002, 46, 3809–3816.
9 H. Streicher and H. Busse, Bioorg. Med. Chem., 2006, 14, 1047–1057.
10 J.-J. Shie, J.-M. Fang and C.-H. Wong, Angew. Chem., 2008, 120, 5872–
5875.
15 A. Larsson, J. Ohlsson, K. W. Dodson, S. J. Hultgren, U. Nilsson and
J. Kihlberg, Bioorg. Med. Chem., 2003, 11, 2255–2262.
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 2570–2575 | 2575
©