Novel imino sugar α-glucosidase inhibitors as antiviral compounds

Article abstract of DOI:10.1016/j.bmc.2013.03.014

Deoxynojirimycin (DNJ) based imino sugars display antiviral activity in the tissue culture surrogate model of Hepatitis C (HCV), bovine viral diarrhoea virus (BVDV), mediated by inhibition of ER α-glucosidases. Here, the antiviral activities of neoglycoconjugates derived from deoxynojirimycin, and a novel compound derived from deoxygalactonojirimycin, by click chemistry with functionalised adamantanes are presented. Their antiviral potency, in terms of both viral infectivity and virion secretion, with respect to their effect on α-glucosidase inhibition, are reported. The distinct correlation between the ability of long alkyl chain derivatives to inhibit ER α-glucosidases and their anti-viral effect is demonstrated. Increasing alkyl linker length between DNJ and triazole groups increases α-glucosidase inhibition and reduces the production of viral progeny RNA and the maturation of the envelope polypeptide. Disruption to viral glycoprotein processing, with increased glucosylation on BVDV E2 species, is representative of α-glucosidase inhibition, whilst derivatives with longer alkyl linkers also show a further decrease in infectivity of secreted virions, an effect proposed to be distinct from α-glucosidase inhibition.

Full text of DOI:10.1016/j.bmc.2013.03.014

Bioorganic & Medicinal Chemistry xxx (2013) xxx–xxx
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Novel imino sugar
a-glucosidase inhibitors as antiviral compounds
J. D. Howe a, N. Smith a, M. J.-R. Lee a, N. Ardes-Guisot b, B. Vauzeilles b,c, J. Désiré d, A. Baron c, Y. Blériot d,
M. Sollogoub e, D. S. Alonzi a, T. D. Butters a,
⇑
a Oxford Glycobiology Institute, Department of Biochemistry, University of Oxford, South Parks Road, Oxford OX1 3QU, UK
b CNRS and Univ. Paris Sud, Glycochimie Moléculaire et Macromoléculaire, ICMMO, UMR 8182, Orsay F-91405, France
c Centre de Recherche de Gif, Istitut de Chimie des Substances Naturelles du CNRS, Avenue de la Terrasse, F-91198 Gif-sur-Yvette, France
d Institut de Chimie des Milieux et Matériaux de Poiteirs, UMR 7285, Equipe <Synthése Organique>, 4 rue Michel Brunet, Poitiers, F-8602, France
e UPMC Univ Paris 06, Institut Universitaire de France, Institut Parisien de Chimie Moléculaire, UMR-CNRS 7201, C. 181, 4 place Jussieu, 75005 Paris, France
a r t i c l e i n f o
a b s t r a c t
Article history:
Available online xxxx
Deoxynojirimycin (DNJ) based imino sugars display antiviral activity in the tissue culture surrogate
model of Hepatitis C (HCV), bovine viral diarrhoea virus (BVDV), mediated by inhibition of ER
a-glu-
cosidases. Here, the antiviral activities of neoglycoconjugates derived from deoxynojirimycin, and a
novel compound derived from deoxygalactonojirimycin, by click chemistry with functionalised ada-
mantanes are presented. Their antiviral potency, in terms of both viral infectivity and virion secretion,
Keywords:
Imino sugars
Neoglycoconjugates
Bovine viral diarrhea virus
Endoplasmic reticulum
Glucosidases
with respect to their effect on
the ability of long alkyl chain derivatives to inhibit ER
demonstrated. Increasing alkyl linker length between DNJ and triazole groups increases
a
-glucosidase inhibition, are reported. The distinct correlation between
-glucosidases and their anti-viral effect is
-glucosidase
a
a
inhibition and reduces the production of viral progeny RNA and the maturation of the envelope poly-
peptide. Disruption to viral glycoprotein processing, with increased glucosylation on BVDV E2 species,
is representative of
a-glucosidase inhibition, whilst derivatives with longer alkyl linkers also show a
further decrease in infectivity of secreted virions, an effect proposed to be distinct from
inhibition.
a-glucosidase
Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction
colipid biosynthetic pathway and the N-glycan processing
enzymes of the ER,
a-glucosidases I and II. NB-DNJ has been pro-
Imino sugars are monosaccharide mimics in which the endocy-
clic oxygen has been replaced by nitrogen.1 They have a number of
advantages for compound design and synthesis to define biological
activity. As polyhydroxylated molecules, each chiral centre offers
manipulation to generate isomers with restricted or enhanced
mimicry, and the endocyclic nitrogen atom is readily modified to
gain selectivity, increase potency or improve pharmacodynamics.
The properties of imino sugars have considerable potential for
treating a diverse range of diseases, including lysosomal storage
disorders,2 cystic fibrosis,3 and viral pathogenesis,4,5 with a mech-
anism of action that is dictated by structural modification.6 Two
N-alkylated imino sugars, Miglitol and Miglustat, are approved
medicines for the treatment of type II diabetes and Gaucher
disease, respectively.7
posed as an effective treatment, known as ‘substrate reduction
therapy’ (SRT), for a number of lysosomal storage diseases, in par-
ticular type I Gaucher disease for which clinical trials have been
undertaken8 and NB-DNJ (Miglustat, ‘Zavesca’) is an approved
medicine in USA, Europe and Israel for this disorder.9
The class of DNJ neoglycoconjugates generated through click
chemistry with functionalised adamantanes (Fig. 1) have previ-
ously been studied to gain structure activity relationship insights
into their potential to inhibit both CGT and ER
a-glucosidases I
and II.10 Following on from the initial insight into these cellular
targets this paper attempts to consolidate our knowledge on this
class of compounds by focussing on their potential to be antivi-
ral, using the hepatitis C virus surrogate model, bovine viral
diarrhea virus (BVDV). The mechanism of action behind the anti-
viral mechanism appears to be poorly understood but in the case
of DNJ based imino sugars the inhibition of the ER glucosidases
resulting in the disruption of the N-linked oligosaccharide
processing pathway in the cell is a key factor.11,12 However the
ability of these compounds to inhibit CGT, any viral ion channels
(such as P7 of HCV) plus any other unknown mechanism of
actions provided by these intriguing molecules cannot be
Deoxynojirimycin is the imino sugar mimic of glucose, and its
N-alkylated analogue N-butyl deoxynojirimycin (NB-DNJ) is an
inhibitor of both the known DNJ based imino sugars’ cellular tar-
gets ceramide glucosyl transferase (CGT), a key enzyme in the gly-
⇑
Corresponding author. Tel.: +44 (0) 1865 275725; fax: +44 (0) 1865 275216.
E-mail address: terry.butters@bioch.ox.ac.uk (T.D. Butters).
0968-0896/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmc.2013.03.014
Please cite this article in press as: Howe, J. D.; et al. Bioorg. Med. Chem. (2013), http://dx.doi.org/10.1016/j.bmc.2013.03.014

2
J. D. Howe et al. / Bioorg. Med. Chem. xxx (2013) xxx–xxx
Figure 1. Structures of the adamantane–iminosugar conjugates. Structures of the adamantane–DNJ conjugates, previously synthesised and numbered.10 The adamantane–
DGJ conjugate 6, the synthesis of which is described in this paper, is also shown for comparison.
discounted. Initially analysing the ER glucosidases inhibition by
free oligosaccharide (FOS) analysis it is possible to differentiate
2.2. N-(40-Azidobutyl)-2,3,4,6-tetra-O-benzyl-1-
deoxygalactonojirimycin 3
in a cellular context the relative inhibitions of both
a-glucosi-
dases I and II and compare them to the antiviral capacity against
BVDV to reveal the antiviral potential of these compounds.
2. Materials and methods
2.1. Materials
O
O
1'
3'
6
N3
5
4
N
O
4'
2'
DNJ neoglycoconjugates were synthesized as previously de-
O
1
2
scribed.10 N-(40-(400-(Adamant-1-yl-methoxymethyl)-1H-100-200-300-
3
triazol-100-yl)butyl)-1-deoxygalactonojirimycin hydrochloride
6
was synthesised from 2,3,4,6-tetra-O-benzyl-1-deoxygalactonojiri-
mycin 1, as shown in Scheme 1.
N
N
OBn
BnO
BnO
OBn
NH
OBn
N
OBn
BnO
BnO
N
BnO
BnO
O
i
ii
N3
N
OBn
OBn
1
3
5
iii
Cl
N
OH
N
HO
HO
H
N
N
O
OH
6
Scheme 1. Reagents and conditions: (i) (2), K2CO3, CH3CN (65%); (ii) (4), CuSO4ꢀ5H2O, sodium ascorbate, CH2Cl2/H2O (1:1) (66%); (iii) HCl, Pd/C, H2, EtOH (81%).
Please cite this article in press as: Howe, J. D.; et al. Bioorg. Med. Chem. (2013), http://dx.doi.org/10.1016/j.bmc.2013.03.014

J. D. Howe et al. / Bioorg. Med. Chem. xxx (2013) xxx–xxx
3
To a solution of 2,3,4,6-tetra-O-benzyl-1-deoxygalactonojirimy-
cin (1) (114.6 mg, 0.22 mmol, 1.0 equiv) and 4-azidobutyl 4-meth-
ylbenzenesulfonate 2 (88.4 mg, 0.33 mmol, 1.5 equiv) in CH3CN
(3.0 mL) was added potassium carbonate (63.5 mg, 0.46 mmol,
2.1 equiv). The reaction mixture was refluxed at 85 °C for 24 h un-
der an argon atmosphere then cooled. Most of the solvent was
evaporated under reduced pressure. Water and CH2Cl2 were added
to the residue and the whole mixture was stirred for 10 min, and
then partitioned. The aqueous layer was extracted with CH2Cl2.
The combined organic extracts were dried over sodium sulfate; fil-
trated and concentrated under reduced pressure. The residue was
purified by flash chromatography over silica gel (Pet. ether/EtOAc
9:1 to 7:3) to yield 89.3 mg (0.14 mmol, 65%) of N-(40-azidobu-
tyl)-2,3,4,6-tetra-O-benzyl-1-deoxygalactonojirimycin 3 as a color-
less oil.
solution of copper(II) sulfate (39.0
l
L, 0.2 M, 1.95 mg, 0.008 mmol,
6 mol %). The reaction mixture was purged under argon, the cap
screwed on and the solution then stirred at room temperature
for 20 h. An aqueous saturated solution of sodium bicarbonate
and CHCl3 were added. The mixture was partitioned and the aque-
ous layer was extracted with CHCl3, dried over sodium sulfate, fil-
trated and concentrated under reduced pressure. The residue was
purified by flash chromatography over silica gel (Pet. ether/EtOAc
6:4) to yield 70.6 mg (0.086 mmol, 66%) of N-(40-(400-(adamant-1-
yl-methoxymethyl)-1H-100-200-300-triazol-100-yl)butyl)-2,3,4,6-tetra-
O-benzyl-1-deoxygalactonojirimycin 5 as a colorless oil.
2.5. N-(40-(400-(Adamant-1-yl-methoxymethyl)-1H-100-200-300-
triazol-100-yl)butyl)-2,3,4,6-tetra-O-benzyl-1-
deoxygalactonojirimycin 5
2.3. N-(40-Azidobutyl)-2,3,4,6-tetra-O-benzyl-1-
deoxygalactonojirimycin 3
C
52H64N4O5 Rf 0.23 (Pet. ether/EtOAc 6:4).
1H NMR (500 MHz, CD3CN) d: 7.60 (s, 1H, H-500); 7.38–7.24 (m,
20H, 20 CH aromatic); 4.75 (d, 1H, 3J 11.3 Hz, CHH Bn); 4.71 (d, 1H,
3J 12.0 Hz, CHH Bn); 4.67 (d, 1H, 3J 12.0 Hz, CHH Bn); 4.61 (d, 1H, 3J
11.9 Hz, CHH Bn); 4.58 (d, 1H, 3J 11.9 Hz, CHH Bn); 4.52 (d, 1H, 3J
11.3 Hz, CHH Bn); 4.46 (s, 2H, OCH2Triazole); 4.44 (s, 2H, CH2
Bn); 4.29 (t, 2H, J 7.1 Hz, 2 H-40); 3.99 (dd, 1H, J4,5 3.2, J3,4 2.6 Hz,
C38H44N6O4 Rf 0.33 (Pet. ether/EtOAc 8:2).
1H NMR (500 MHz, CD3CN) d: 7.38–7.23 (m, 20H, 20 CH aro-
matic); 4.76 (d, 1H, 3J 11.3 Hz, CHH Bn); 4.72 (d, 1H, 3J 12.0 Hz,
CHH Bn); 4.68 (d, 1H, 3J 12.0 Hz, CHH Bn); 4.62 (d, 1H, 3J 11.9 Hz,
CHH Bn); 4.59 (d, 1H, 3J 11.9 Hz, CHH Bn); 4.52 (d, 1H, 3J 11.3 Hz,
CHH Bn); 4.46 (s, 2H, CH2 Bn); 4.00 (dd, 1H, J4,5 3.2, J3,4 2.6 Hz, H-
4); 3.74 (ddd, 1H, J1b,2 7.9, J2,3 7.5, J1a,2 3.5 Hz, H-2); 3.72 (dd, 1H,
2
H-4); 3.76–3.72 (m, 1H, H-2); 3.72 (dd, 1H, J6a,6b 10.1, J6a,5
6.1 Hz, H-6a); 3.59 (dd, 1H, J6a,6b 10.1, J6b,5 4.7 Hz, H-6b); 3.55–
3.50 (m, 1H, H-3); 3.03 (s, 2H, OCH2Ad); 2.97–2.89 (br s, 1H, H-
2
2
2
2J6a,6b 9.9, J6a,5 6.0 Hz, H-6a); 3.60 (dd, 1H, J6a,6b 9.9, J6b,5 4.9 Hz,
0
0
0
1a); 2.86–2.76 (br s, 1H, H-5); 2.61 (dt, 1H, J1 a,1 b 13.0, J1 a,2
7.4 Hz, H-10a); 2.54–2.44 (br s, 1H, H-10b); 2.28–2.20 (br s, 1H, H-
H-6b); 3.53 (dd, 1H, J2,3 7.5, J3,4 2.6 Hz, H-3); 3.24 (t, 2H, J 6.4 Hz,
2
0
0
0
0
2 H-40); 2.97 (dd, 1H, J1a,1b 12.2, J1a,2 3.5 Hz, H-1a); 2.88–2.76 (br
1b); 1.93–1.88 (br s, 3H, 3 CH-Ad); 1.78 (tt, 2H, J2 ,3 7.5, J3 ,4
7.1 Hz, 2 H-30); 1.74–1.59 (br m, 6H, 3 CH2-Ad); 1.53–1.47 (br s,
2
s, 1H, H-5); 2.60 (dt, 1H, J1 a,1 b 13.0, J1 a,2 7.1 Hz, H-10a); 2.52–
0
0
0
6H, 3 CH2-Ad); 1.39 (ddt, 2H, J2 ,3 7.5, J1 a,2 7.4, J1 b,2 7.1 Hz, 2 H-20).
13C NMR (125 MHz, CD3CN) d: 146.0 (C-400); 140.4–139.9 (4 Cq
aromatic); 129.4–128.4 (20 CH aromatic); 123.9 (C-500); 82.0
(OCH2Ad); 77.2 (C-3); 76.3 (C-4, C-2); 74.3–72.7 (4 CH2 Bn); 69.3
(C-6); 65.5 (OCH2Triazole); 62.7 (C-5); 53.6 (C-10); 50.7 (C-40);
40.5 (3 CH2-Ad); 38.0 (3 CH2-Ad); 34.8 (Cq-Ad); 29.3 (3 CH-Ad);
29.0 (C-30); 23.8 (C-20).
2
0
0
0
0
0
0
2.40 (m, 1H, H-10b); 2.26 (dd, 1H, J1a,1b 12.2, J1b,2 7.9 Hz, H-1b);
1.54–1.42 (m, 4H, 2 H-20, 2 H-30).
13C NMR (125 MHz, CD3CN) d: 140.4–139.9 (4 Cq aromatic);
129.4–128.4 (20 CH aromatic); 77.2 (C-3); 76.3 (C-4, C-2); 74.5–
72.7 (4 CH2 Bn); 69.4 (C-6); 62.7 (C-5); 53.8 (C-10); 52.2 (C-40);
27.5, 24.1 (C-20, C-30).
MS (ESI+): m/z 621.4 [M+H]+. HRMS (ESI+): calcd for C38H45N4O4
621.3441, found 621.3470.
MS (ESI+): m/z 825.5 [M+H]+. HRMS (ESI+): calcd for C52H65N4O5
825.4955, found 825.4970.
2.4. N-(40-(400-(Adamant-1-yl-methoxymethyl)-1H-100-200-300-
triazol-100-yl)butyl)-2,3,4,6-tetra-O-benzyl-1-
deoxygalactonojirimycin 5
2.6. N-(40-(400-(Adamant-1-yl-methoxymethyl)-1H-100-200-300-
triazol-100-yl)butyl)-1-deoxygalactonojirimycin hydrochloride 6
Cl
N
N
OH
4''
1'
HO
3'
6
H
N
4
5
N
O
5''
4'
2'
N
HO
N
O
1
4''
2
O
3'
3
1'
1
OH
6
N
5
4
N
O
O
5''
2'
4'
O
2
3
A solution of N-(40-(400-(adamant-1-yl-methoxymethyl)-1H-100-
200-300-triazol-100-yl)butyl)-2,3,4,6-tetra-O-benzyl-1-deoxygalacton-
ojirimycin 5 (64.0 mg, 77.6 lmol, 1.0 equiv) in ethanol (5.2 mL)
was acidified to pH ꢁ2 with 1 M aq HCl (10 drops). Argon was
passed through the solution for 5 min, after which palladium
10%wt on carbon (58.8 mg, 0.54 mmol, 7.0 equiv) was added.
Hydrogen was passed through the reaction mixture for 5 min
and the reaction mixture was stirred for 5 days under atmospheric
hydrogen pressure. Pd/C was removed by filtration over CeliteÒ
plug, followed by thorough rinsing with CH3OH. The filtrate was
To
a
solution of N-(40-azidobutyl)-2,3,4,6-tetra-O-benzyl-1-
deoxygalactonojirimycin 3 (80.7 mg, 0.13 mmol, 1.0 equiv) and 1-
((prop-2-yn-1-yloxy)methyl)adamantane 4 (39.8 mg, 0.19 mmol,
1.5 equiv) in CH2Cl2 (560
were added an aqueous solution of sodium ascorbate (97.5
0.2 M, 3.86 mg, 0.019 mmol, 15 mol %), followed by an aqueous
l
L) and H2O (423
lL) in a screw cap tube
lL,
Please cite this article in press as: Howe, J. D.; et al. Bioorg. Med. Chem. (2013), http://dx.doi.org/10.1016/j.bmc.2013.03.014

4
J. D. Howe et al. / Bioorg. Med. Chem. xxx (2013) xxx–xxx
concentrated under reduced pressure. The residue was purified by
flash chromatography over silica gel (CHCl3/CH3OH 7:3) to yield
(Alonzi et al., 2008), and then separated by NP-HPLC (Waters)
using a 4.6 ꢃ 250 mm TSKgel Amide-80 column (Anachem),14 as
described previously.15
31.6 mg (63.1 l
mol, 81%) of N-(40-(400-(adamant-1-yl-methoxy-
methyl)-1H-100-200-300-triazol-100-yl)butyl)-1-deoxygalactonojirimy-
cin hydrochloride 6 as a white powder after freeze-drying.
2.11. Fluorescence focus assay
2.7. N-(40-(400-(Adamant-1-yl-methoxymethyl)-1H-100-200-300-
MDBK cells were seeded in 6-well plates (1.75 ꢃ 106 cells/well)
and infected with BVDV at an MOI of 1, followed by incubation
with different concentrations of imino sugars for 24 h. Cell culture
medium was then harvested and clarified by centrifugation at
5000 rpm for 10 min to remove cell debris. Serial dilutions were
made and used to infect naive MDBK cells seeded in 96-well plates
(7 ꢃ 104 cells/well). Cells were incubated for 24 h, then washed
twice with cold PBS and fixed with 4% (v/v) paraformaldehyde in
PBS for 30 min. Cells were washed again and blocked for two hours
in 5% (w/v) milk in PBS. Cells were then permeabilised with 1% (v/
v) Triton-X-100 in PBS for 20 min, washed with 1% (v/v) Tween20
in PBS and incubated with MAb103/105 (1:500 dilution in 5% milk-
PBS; Animal Health Veterinary Laboratories Agency, Weybridge,
UK) for 1 h. Cells were washed three times with 1% Tween20-PBS
then incubated with anti-mouse FITC-conjugated secondary anti-
body (1:500 dilution in 5% milk-PBS; Sigma). Cells were washed
again, and nuclei stained with 40,6-diamidino-2-phenylindole
(DAPI; Vector Laboratories Inc.). Cells were observed and evaluated
using an inverted Nikon Eclipse TE200-U microscope, equipped
with a Fluor 10ꢃ objective lens, and FITC and UV filter sets. Fluo-
rescent foci were counted, and virus titers (fluorescent focus units
(FFU)/ml) calculated.
triazol-100-yl)butyl)-1-deoxygalactonojirimycin hydrochloride 6
C24H41ClN4O5 Rf 0.17 (CHCl3/CH3OH 8:2).
1H NMR (500 MHz, CD3OD) d: 7.96 (s, 1H, H-500); 4.54 (s, 2H,
OCH2Triazole); 4.44 (t, 2H, J 7.1 Hz, 2 H-40); 3.97 (dd, 1H, J4,5 3.2,
J3,4 1.6 Hz, H-4); 3.84–3.78 (m, 2H, H-6a, H-6b); 3.80 (ddd, 1H,
J1b,2 10.1, J2,3 9.2, J1a,2 5.0 Hz, H-2); 3.23 (dd, 1H, J2,3 9.2, J3,4
2
3.2 Hz, H-3); 3.06 (s, 2H, OCH2Ad); 2.99 (dd, 1H, J1a,1b 11.2, J1a,2
2
5.0 Hz, H-1a); 2.82 (dt, 1H, J1 a,1 b 13.4, J1 a,2 8.0 Hz, H-10a); 2.56
0
0
0
0
2
(dt, 1H, J1 a,1 b 13.4, J1 b,2 7.3 Hz, H-10b); 2.44–2.38 (br s, 1H, H-
0
0
0
0
2
5); 2.10 (dd, 1H, J1a,1b 11.2, J1b,2 10.1 Hz, H-1b); 1.96–1.88 (m,
3H, 3 CH-Ad); 1.89 (tt, 2H, J2 ,3 7.7, J3 ,4 7.1 Hz, 2 H-30); 1.78–1.64
(br m, 6H, 3 CH2-Ad); 1.57–1.52 (br m, 6H, 3 CH2-Ad); 1.52 (ddt,
0
0
0
0
2H, J1 a,2 8.0, J2 ,3 7.7, J1 b,2 7.3 Hz, 2 H-20).
0
0
0
0
0
0
13C NMR (125 MHz, CD3OD) d: 146.6 (C-400); 125.0 (C-500); 82.7
(OCH2Ad); 77.2 (C-3); 72.7 (C-4); 69.0 (C-2); 65.6 (C-5, OCH2Tria-
zole); 62.7 (C-6); 57.8 (C-1); 53.3 (C-10); 51.3 (C-40); 40.9 (3 CH2-
Ad); 38.4 (3 CH2-Ad); 35.2 (Cq-Ad); 29.9 (3 CH-Ad); 29.3 (C-30);
22.7 (C-20).
MS (ESI+): m/z 464.3 [MꢂCl]+. HRMS (ESI+): calcd for
C
24H41N4O5 465.2077, found 465.3080.
2.8. Cell culture
2.12. Real-time RT-PCR analysis
Mardin-Darby Bovine Kidney (MDBK) cells (European Collec-
tion of Animal Cell Cultures, Porton Down, United Kingdom) were
cultured in RPMI-1640 medium (PAA Laboratories, Teddington,
MDBK cells were seeded in 6-well plates (1.75 ꢃ 106 cells/well)
and infected with BVDV at an MOI of 1, followed by incubation
with different concentrations of imino sugars for 24 h. Cell culture
medium was then harvested and clarified by centrifugation at
5000 rpm for 10 min to remove cell debris.
UK) supplemented with 10% fetal calf serum (PAA), 2 mM L-gluta-
mine (PAA) and 1ꢃ penicillin–streptomycin (PAA), at 37 °C in a
humidified atmosphere containing 5% CO2 (v/v).
A QIAamp viral RNA mini kit (Qiagen, Crawley, UK) was used to
purify viral RNA from cell culture medium, according to manufac-
turer’s protocol. The purified RNA samples were treated with
DNase (20 min at 37 °C, followed by 10 min at 75 °C for deactiva-
tion). Real Time RT-PCR was carried out with a Qiagen SYBR green
Quantitect RT-PCR kit, using primers that amplify a 334-bp portion
of the NS2 coding sequence (forward primer, 50-AGA AAA CAC AGA
AAC CCG ACA GAC-30; reverse primer 50-TTC CAC CAC TAT TGT AGC
ATC CG-30). An Opticon 2 real time PCR machine (Bio-Rad) was
used to perform the PCR reaction: 50 °C for 30, 15 min incubation
at 95 °C followed by 35 cycles of 15 s at 95 °C, 1 min at 50 °C, 1 min
at 72 °C and a final extension for 7 min at 72 °C. These data were
calibrated against a standard curve produced using serial dilutions
of a concentrated BVDV stock.
2.9. Cell proliferation assays
A CellTiter 96R aqueous non-radioactive cell proliferation assay
kit (Promega) was used to measure cellular cytotoxicity of com-
pounds according to manufacturer protocol. MDBK cells were
seeded in 96 well plates (5 ꢃ 103/well) and incubated for 24 h.
Media was removed from wells and replaced with fresh media con-
taining different concentrations of imino sugars shown in Figure 1,
then incubated for a further 72 h.
40
l
l
of 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxy-
tetrazolium) (MTS)–phenazine
phenyl)-2-(4-sulfophenyl-2H
methosulfate mixture (20:1) were added to each well and the sam-
ples incubated for 1 h. The absorbance was read at 490 nm with a
UVmax plate reader (Molecular Devices). Samples were analysed in
triplicate and the concentrations at which absorbance at 490 nm
was 50% of untreated controls (CC50) were determined.
2.13. Immunoblotting
MDBK cells were seeded in 6-well plates (1.75 ꢃ 106 cells/well)
and infected with BVDV at an MOI of 1, followed by incubation
with different concentrations of imino sugars for 24 h. Cells were
harvested by scraping in PBS and centrifuged at 2000 rpm for
5 min. The resulting pellet was briefly frozen at ꢂ20 °C then resus-
pended in PBS. The suspension was sonicated on ice using a Bran-
son Sonifier (20 s, 50% duty cycle) to lyse cells. Protein
concentrations were determined using a standard bicinchoninic
acid protein assay. Proteins were then separated by sodium dode-
cyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) under
non-reducing conditions and transferred to Immobilon-P PVDF
membrane (Millipore) by semi-dry electroblotting, then blocked
overnight in 5% (w/v) milk, 0.2% (v/v) Tween20 in PBS. The
2.10. FOS analysis
MDBK cells were seeded in 6-well plates (1 ꢃ 106 cells/well) in
medium containing different concentrations of imino sugars and
incubated at 37 °C, 5% CO2 atmosphere, for 24 h. After incubation,
cells were harvested and free oligosaccharides (FOS) were ex-
tracted and purified as previously described.13 FOS were labelled
with 2-anthranilic acid,14 and purified by chromatography using
Discovery Speed Amide-2 columns.13 The labelled FOS were fur-
ther purified by chromatography using Concanavalin A (ConA)-Se-
pharose 4B columns (100 ml packed resin) as described previously
Please cite this article in press as: Howe, J. D.; et al. Bioorg. Med. Chem. (2013), http://dx.doi.org/10.1016/j.bmc.2013.03.014

J. D. Howe et al. / Bioorg. Med. Chem. xxx (2013) xxx–xxx
5
Table 1
Summary of cytotoxicity and
a
-glucosidase I and II maximal inhibition, as determined by FOS analysis
Compound alkyl linker C atoms 72 h Cytotoxicity CC50
(
lM)
24 h FOS analysis in MDBK cells
Maximal
a-glucosidase II inhibition (
lM) and
Maximal
a-glucosidase I inhibition (
lM) and
relative amount
relative amount
6 C4
>250
ni
ni
ni
ni
7 C1
8 C2
9 C4
10 C4
11 C6
12 C8
13 C10
>250
>250
>250
>250
205 1.1
51.7 4.4
29.2 1.0
100
100
50
100
10
3.0
5.1
5.2
6.0
8.3
5.4
5.8
100
100
100
100
100
25
0.54
0.82
1.44
1.52
11.4
2.0
10
5
10
0.26
Cytotoxicity (CC50 values) is shown for each compound used in MDBK cells. FOS HPLC analysis was used to determine peak areas of mono- or tri-glucosylated Man5GlcNAc1 as
indicators of inhibition of -glucosidase II and I enzymes, respectively. Relative amounts in comparison to a control species Man4GlcNAc1 are shown. ni, not inhibitory.
a
membrane was incubated with MAb214 primary antibody (1:1000
dilution in blocking solution; Animal Health Veterinary Laborato-
ries Agency) for 2 h at room temperature, washed extensively with
PBS–0.1% (v/v) Tween20, then incubated with anti-mouse horse-
radish peroxidase secondary antibody (1:1000; Dako) for 1 h. Anti-
body binding to BVDV E2 protein was detected using Pierce ECL
Western Blotting Substrate (Thermo Scientific, UK) and X-ray film
(GE Healthcare, UK) by following the manufacturer’s instructions.
other compound tested (Table 1). Compounds 9 and 10) have alkyl
linkers of the same length, differing only in the introduction of an
ether bond between the triazole and the adamantyl group in 10.
The two compounds exhibited very similar ER
a-glucosidase
inhibition. Adamantyl–DGJ had no effect on -glucosidase activity,
a
in agreement with our previously published data using
N-alkylated-DGJ compounds.6
3.3. Antiviral activity correlates with ER
inhibition
a-glucosidase
3. Results
Compounds were screened for their ability to inhibit the bio-
genesis and infectivity of bovine viral diarrhea virus (BVDV). Viral
production and release from MDBK cells was determined using RT-
PCR to measure viral RNA copies secreted (Fig. 3A). Compounds
were initially administered at 10 lM and at this concentration,
short alkyl chain compounds 7, 8 and 9 showed no antiviral activ-
3.1. Cellular cytotoxicity increases with N-alkyl chain length
An MTS-based cell proliferation assay was carried out to deter-
mine the maximum concentrations of imino sugar derivatives that
could be used without causing a significant effect to cell viability.
The concentrations compounds 6–13 required to reduce prolifera-
tion in MDBK cells by 50% (CC50) in 72 h are shown in Table 1. No
significant cytotoxic effect was observed for compounds 6 7, 8, 9
and 10. More hydrophobic compounds showed greater anti-prolif-
ity, although 8 and 9 were effective when a higher concentration
was used (50
kyl chain length increased, with 10
RNA fourfold. At 10 M, 10 reduced RNA copies approximately
l
M, Fig. 3A). Reduction in RNA copies increased as al-
l
M 13 treatment reducing viral
l
erative effects; 11 was 205 1.1
lM, 12 was 51.7 4.4 lM and 13
twofold whereas 9 showed no effect, suggesting that the introduc-
tion of the ether bond between the triazole and the adamantyl
group increases antiviral activity.
Infectivity of viral particles secreted from MDBK cells was
determined as focus-forming units/ml (FFU/ml), by inoculating na-
ive MDBK cells with virus secreted from cells administered with
compound, then probing for BVDV NS2/3 protein (Fig. 3B). Com-
pounds were tested at the same concentrations as for RT-PCR anal-
ysis, and reduction in FFU/ml mirrors the reduction in viral RNA
was 29.2 1.0 M. As the length of the alkyl linker between the
l
imino sugar and the triazole ring increased, the CC50 decreased
markedly, a property that has been noted for other N-alkylated
iminosugars.16
3.2. ER
length
a-glucosidase inhibition correlates with N-alkyl chain
Previously, we showed that adamantyl–DNJ conjugates could
inhibit cellular ER-glucosidases.10 Free oligosaccharide (FOS) anal-
ysis of MDBK cells following inhibitor treatment was used to deter-
copies for compounds 7 through 13. Adamantyl-DGJ at 10 lM
however showed no significant reduction in viral RNA and a small
reduction in FFU/ml compared to the untreated control. Increasing
mine the level of in cellulo inhibition of ER
Inhibition of -glucosidase II is known to cause an accumulation of
monoglucosylated FOS, whilst inhibition of -glucosidase I results
a-glucosidases I and II.
the concentration of this compound to 50
lM revealed similar data
a
(results not shown) indicating that inhibition of ER-
a-glucosidases
a
appeared essential to the antiviral activity of the DNJ
neoglycoconjugates.
in triglucosylated FOS accumulation. FOS species were purified
from cells and separated by NP-HPLC for analysis. Man4GlcNAc1,
which remained in constant amount over all concentrations of
inhibitor used in this study, was used as standard and the relative
amounts of mono- and tri-glucosylated Man5GlcNAc1 species were
3.4. ER
a-glucosidase inhibition increases glucosylation of
intracellular BVDV E2 protein
measured as representative markers of
bition respectively (Fig. 2). Concentrations for maximal
dase I and II inhibition for each compound is shown in Table 1.
Maximal -glucosidase II inhibition was achieved for all com-
pounds tested below the maximum concentration used. -Glucosi-
a
-glucosidase II and I inhi-
Inhibition of ER a-glucosidases I and II by N-alkylated DNJ pre-
a
-glucosi-
vents deglucosylation of BVDV E2 protein (Lee et al., manuscript in
preparation). MDBK cells were infected with BVDV (MOI = 1), then
administered with compounds 10, 11, 13 and 6 for 24 h. Com-
pounds were used at concentrations that gave maximal a-glucosi-
dase I inhibition by FOS analysis (Table 1). Proteins from cell lysate
were separated by SDS–PAGE and probed for BVDV E2 protein
(Fig. 4). Treatment with 100 lM 11 caused an increase in apparent
a
a
dase I inhibition occurred at higher concentrations and little
inhibition was observed for compounds with short alkyl linkers.
Treatment with 100
lM 11 showed the highest level of a-glucosi-
dase I inhibition (Fig. 2B), more than three times higher than any
Please cite this article in press as: Howe, J. D.; et al. Bioorg. Med. Chem. (2013), http://dx.doi.org/10.1016/j.bmc.2013.03.014

6
J. D. Howe et al. / Bioorg. Med. Chem. xxx (2013) xxx–xxx
Figure 2. FOS analysis of adamantane–DNJ conjugates in MDBK cell. (A) Representative NP-HPLC profile of 2-AA labelled FOS analysis. (I) untreated MDBK cells and (II)
100 M 11 administered for 24 h. Marked on the chromatograms are the Man4GlcNAc1 (which is constant in amount over all concentrations of inhibitor used in this study)
and the mono- and tri-glucosylated Man5GlcNAc1 species which increase following -glucosidase II and I inhibition respectively. (B) FOS dose response graph showing
l
a
increases in mono- and tri-glucosylated Man5GlcNAc1 species relative to Man4GlcNAc1 following treatment with 11 up to 100 lM for 24 h. Circles, Glc1Man5GlcNAc1; squares,
Glc3Man5GlcNAc1 oligosaccharides. Mean values are displayed, error bars show standard deviation.
molecular weight for all E2 species, due to increased glucosylation,
whilst all other compounds showed little or no bandshift. For E1E2
heterodimer, the main band detected, there are six potential N-
linked glycosylation sites.17 A change in glycan structure from
deglucosylated to triglucosylated could therefore increase the
molecular weight of the heterodimer by up to 3.2 kDa (180.14 g/
mol ꢃ 3 glucose residues ꢃ 6 glycosylation sites), which is in
agreement with the bandshift observed.
Disruption of the folding of envelope glycoproteins E1 and E2,
essential for virus entry,21 by ER
a-glucosidase inhibition can in-
duce misassembly of glycoprotein pre-budding complexes.19
The novel DNJ derivatives discussed here have been shown to
inhibit ER a-glucosidases and to possess anti-viral activity against
BVDV, with a distinct correlation observed between the increasing
length of the alkyl chain linker between the DNJ head group and
the triazole ring and their anti-viral efficacy. Compounds 7 and 8
demonstrate low levels of
a-glucosidase II inhibition and little
inhibition of -glucosidase I even at the highest concentrations
a
4. Discussion
tested. Compounds 9 and 10, which both possess a 4-carbon alkyl
linker, show similar levels of inhibition of both enzymes, suggest-
ing that the introduction of an ether bond in 10 between the tria-
zole and adamantyl group has no effect on glucosidase inhibition.
Longer alkyl chain compounds 11, 12 and 13 efficiently inhibited
DNJ and its derivatives have shown antiviral activity against the
flavivirus BVDV,4 as well as a number of other viruses including
hepatitis B virus,18 and hepatitis C virus pseudoparticles.19 DNJ-
based iminosugars reversibly inhibit ER a-glucosidases I and II by
a
-glucosidase II at low concentrations (5–10
dase I at high concentrations (25–100 M). Compound 11 was able
to attain the greatest level of -glucosidase I inhibition, at similar
concentrations to more hydrophobic compounds 12 and 13 that
lM) and a-glucosi-
competing with glucose substrates for their active site, thereby
perturbing the folding of viral glycoproteins in the N-glycosylation
pathway and impeding viral biogenesis by hindering their
assembly.20
l
a
Please cite this article in press as: Howe, J. D.; et al. Bioorg. Med. Chem. (2013), http://dx.doi.org/10.1016/j.bmc.2013.03.014

J. D. Howe et al. / Bioorg. Med. Chem. xxx (2013) xxx–xxx
7
inhibiting
gar which also contains a six carbon linker and is known to po-
tently inhibit both
-glucosidases,22 11 achieves a comparable
level of -glucosidase I inhibition (Lee et al., manuscript in prepa-
a-glucosidase I. In comparison to NAP-DNJ, an imino su-
a
a
ration). Increasing alkyl chain length in N-alkylated DNJ com-
pounds has previously been shown to increase cellular uptake,23
and an increase in the alkyl chain length linker between DNJ and
phenyl or cyclohexyl groups increases ER
a-glucosidase inhibi-
tion.15 Here, increasing the alkyl chain length of the linker between
DNJ and triazole group similarly increases
inhibition.
a-glucosidase
DNJ-based imino sugars have previously been shown to demon-
strate antiviral activity against BVDV.4 The longer alkyl chain N-
nonyl-DNJ has shown tenfold greater potency than N-butyl-
DNJ,16 with branching and cyclization of side chains able to in-
crease potency further.24 Compound 7, with a single carbon linker
between DNJ and triazole, showed no anti-viral activity at the con-
centrations tested. As the length of the alkyl linker is increased, the
amount of viral RNA in culture medium decreases, with 10 lM 12
reducing viral titre approximately tenfold. Compound 13 reduced
viral titre less than 12, suggesting an upper limit to alkyl linker
length may have been reached. Subsequent infection of naive
MDBK cells with virus secreted from cells administered with com-
pound followed a similar pattern to
a-glucosidase inhibition and
viral secretion levels, although despite showing higher viral titre
levels than 12, less infection was observed following treatment
with 13. These results suggest that the antiviral effects of these
compounds are largely a result of reduced viral biogenesis or secre-
tion, rather than as a result of reduced infectivity, although an
additional reduction is observed with 13. In contrast, treatment
with Adamantyl-DGJ 6, a galactose analogue of Adamantyl–DNJ
10 that does not inhibit a-glucosidases, failed to reduce viral titre
Figure 3. Antiviral results. (A) Quantification of secreted viral RNA levels. MDBK
cells were infected with BVDV at an MOI of 1 (or mock-infected) and 10 M 6, 10,
11, 12 or 13, or 50 M 7, 8 or 9 administered for 24 h. Cell culture supernatant was
but partially reduced infectivity. Long-alkyl-chain DGJ derivatives
have previously been shown to exert their antiviral effect solely
via the production of viral particles with reduced infectivity,16
which has been proposed to be through interaction with p7 ion
channel.25
l
l
harvested and viral RNA purified and subjected to RT-PCR analysis, using primers
against BVDV NS2 gene. (B) Fluorescence focus assay of BVDV-infected MDBK cells.
MDBK cells were infected and administered with drug as described in Figure 3(A).
Cell culture supernatant was harvested and fivefold serial dilutions were used to
infect naive MDBK cells. After 24 h incubation, cells were fixed and probed with
monoclonal antibody against BVDV NS2/3, followed by anti-mouse FITC-conjugated
secondary antibody. Fluorescent focus units were counted and reduction in
infection calculated relative to untreated control. The data were obtained from
three experiments with means and standard deviations shown.
Direct effect on BVDV glycoprotein E2 is observed with Western
blot analysis. Treatment with 11, shown by FOS analysis to be the
most potent
a-glucosidase I inhibitor, resulted in increased glu-
cosylation of intracellular E2 protein. These results taken together
lead to the following proposal for the antiviral effects observed
with compounds 7–13. With the shorter alkyl chain compounds,
antiviral activity is a result of ER
a-glucosidase inhibition, which
prevents efficient and successful processing of viral glycoproteins,
resulting in a reduction in the formation of prebudding complexes
and a subsequent reduction in viral formation and release. In-
creased hydrophobicity of longer alkyl linker derivatives facilitates
their uptake and results in a prolonged period of retention in the
intracellular environment where they can exert inhibition of a-glu-
cosidases in the ER,15 resulting in greater antiviral activity. Sepa-
rately, as alkyl linker length increases, an antiviral effect
unrelated to
a-glucosidase inhibition begins to appear, producing
non- or less infectious virions. Interactions between more hydro-
phobic adamantyl-conjugates and p725 may explain the reduction
in infectivity. However, the improved ER-a-glucosidase inhibitory
Figure 4. Western blot analysis of BVDV E2 protein. Western blot analysis of BVDV-
infected MDBK cells. MDBK cells were mock-infected (Lane 1) or infected with
BVDV at an MOI of 1 (Lanes 2–6) and treated with compounds for 24 h. Lane 2:
activity that appears to correlate with an effective reduction in vir-
al titre and infectivity of these compounds, will allow further eval-
uation of their therapeutic potential.
Untreated control; 3: 100 lM 10; 4: 100 lM 11; 5: 10 lM 13; 6: 100 lM 6. After
treatment, cells were lysed and proteins separated by SDS-10% PAGE, transferred to
Immobilon-P PVDF membrane and probed with MAb214 against BVDV E2 protein.
The main band observed, at 63 kDa in the untreated sample, is BVDV E1E2
heterodimer.
Acknowledgments
The authors thank Professor Raymond Dwek and the Glycobiol-
ogy Institute for support. This work was supported by the ANR pro-
gram ‘Jeunes Chercheuses—Jeunes Chercheurs’, Project Number
ANR-05-JCJC-0199.
were used without affecting cell proliferation limiting their ability.
The 6-carbon linker compound 11 could be administered to cells at
higher concentrations and proved to be even more effective at
Please cite this article in press as: Howe, J. D.; et al. Bioorg. Med. Chem. (2013), http://dx.doi.org/10.1016/j.bmc.2013.03.014

8
J. D. Howe et al. / Bioorg. Med. Chem. xxx (2013) xxx–xxx
12. Norton, A.; Gu, B.; Block, T. M. Iminosugars. From Synthesis to Therapeutic
Applications In Compain, P., Martin, O. R., Eds.; John Wiley Sons Ltd:
Chichester, 2007; p 209.
13. Alonzi, D. S.; Neville, D. C.; Lachmann, R. H.; Dwek, R. A.; Butters, T. D. Biochem.
J. 2008, 409, 571.
14. Neville, D. C.; Coquard, V.; Priestman, D. A.; te Vruchte, D. J.; Sillence, D. J.;
Dwek, R. A.; Platt, F. M.; Butters, T. D. Anal. Biochem. 2004, 331, 275.
15. Alonzi, D. S.; Dwek, R. A.; Butters, T. D. Tetrahedron: Asymmetry 2009, 20, 897.
16. Durantel, D.; Branza-Nichita, N.; Carrouee-Durantel, S.; Butters, T. D.; Dwek, R.
A.; Zitzmann, N. J. Virol. 2001, 75, 8987.
17. Dubuisson, J.; Rice, C. M. J. Virol. 1996, 70, 778.
18. Block, T. M.; Lu, X.; Platt, F. M.; Foster, G. R.; Gerlich, W. H.; Blumberg, B. S.;
Dwek, R. A. Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 2235.
19. Chapel, C.; Garcia, C.; Bartosch, B.; Roingeard, P.; Zitzmann, N.; Cosset, F. L.; Dubuisson,
J.; Dwek, R. A.; Trepo, C.; Zoulim, F.; Durantel, D. J. Gen. Virol. 2007, 88, 1133.
20. Parodi, A. J. Biochem. J. 2000, 348, 1.
21. Ronecker, S.; Zimmer, G.; Herrler, G.; Greiser-Wilke, I.; Grummer, B. J. Gen.
Virol. 2008, 89, 2114.
22. Rawlings, A. J.; Lomas, H.; Pilling, A. W.; Lee, M. J.; Alonzi, D. S.; Rountree, J. S.;
Jenkinson, S. F.; Fleet, G. W.; Dwek, R. A.; Jones, J. H.; Butters, T. D.
ChemBioChem 2009, 10, 1101.
23. Mellor, H. R.; Platt, F. M.; Dwek, R. A.; Butters, T. D. Biochem. J. 2003, 374, 307.
24. Mehta, A.; Ouzounov, S.; Jordan, R.; Simsek, E.; Lu, X.; Moriarty, R. M.; Jacob, G.;
Dwek, R. A.; Block, T. M. Antiviral Chem. Chemother. 2002, 13, 299.
25. Pavlovic, D.; Neville, D. C.; Argaud, O.; Blumberg, B.; Dwek, R. A.; Fischer, W. B.;
Zitzmann, N. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 6104.
References and notes
&
1. Stutz, A. E. Iminosugars as Glycosidase Inhibitors; Wiley-VCH: Weinheim, 1999.
2. Butters, T. D.; Dwek, R. A.; Platt, F. M. Chem. Rev. 2000, 100, 4683.
3. Norez, C.; Noel, S.; Wilke, M.; Bijvelds, M.; Jorna, H.; Melin, P.; DeJonge, H.;
Becq, F. FEBS Lett. 2006, 580, 2081.
4. Zitzmann, N.; Mehta, A. S.; Carrouee, S.; Butters, T. D.; Platt, F. M.; McCauley, J.;
Blumberg, B. S.; Dwek, R. A.; Block, T. M. Proc. Natl. Acad. Sci. U.S.A. 1999, 96,
11878.
5. Chang, J.; Schul, W.; Yip, A.; Xu, X.; Guo, J. T.; Block, T. M. Antiviral Res. 2011, 92,
369.
6. Butters, T. D.; Wormald, M. R.; Dwek, R. A.; Platt, F. M.; Van Den Broek, L. A. G.
M.; Fleet, G. W. J.; Krulle, T. M. Tetrahedron: Asymmetry 2000, 11, 113.
7. Compain, P.; Martin, O. R. Iminosugars. From Synthesis to Therapeutic
Applications; John Wiley & Sons Ltd: Chichester, 2007.
8. Cox, T.; Lachmann, R.; Hollak, C.; Aerts, J.; van Weely, S.; Hrebicek, M.; Platt, F.;
Butters, T.; Dwek, R.; Moyses, C.; Gow, I.; Elstein, D.; Zimran, A. Lancet 2000,
355, 1481.
9. Elstein, D.; Hollak, C.; Aerts, J. M.; van Weely, S.; Maas, M.; Cox, T. M.;
Lachmann, R. H.; Hrebicek, M.; Platt, F. M.; Butters, T. D.; Dwek, R. A.; Zimran, A.
J. Inherit. Metab. Dis. 2004, 27, 757.
10. Ardes-Guisot, N.; Alonzi, D. S.; Reinkensmeier, G.; Butters, T. D.; Norez, C.; Becq,
F.; Shimada, Y.; Nakagawa, S.; Kato, A.; Bleriot, Y.; Sollogoub, M.; Vauzeilles, B.
Org. Biomol. Chem. 2011, 9, 5373.
11. Greimel, P.; Spreitz, J.; Stutz, A. E.; Wrodnigg, T. M. Curr. Top. Med. Chem. 2003,
3, 513.
Please cite this article in press as: Howe, J. D.; et al. Bioorg. Med. Chem. (2013), http://dx.doi.org/10.1016/j.bmc.2013.03.014