N-Alkyl-1,5-dideoxy-1,5-imino-L-fucitols as fucosidase inhibitors: Synthesis, molecular modelling and activity against cancer cell lines

Article abstract of DOI:10.1016/j.bioorg.2018.12.003

1,5-Dideoxy-1,5-imino-L-fucitol (1-deoxyfuconojirimycin, DFJ) is an iminosugar that inhibits fucosidases. Herein, N-alkyl DFJs have been synthesised and tested against the α-fucosidases of T. maritima (bacterial origin) and B. taurus (bovine origin). The

Full text of DOI:10.1016/j.bioorg.2018.12.003

Accepted Manuscript
N-Alkyl-1,5-Dideoxy-1,5-Imino-L-Fucitols as Fucosidase Inhibitors: Synthe-
sis, Molecular Modelling and Activity Against Cancer Cell Lines
Jian Zhou, Arvind Negi, Styliana I. Mirallai, Rolf Warta, Christel Herold-
Mende, Michael P. Carty, Xin-Shan Ye, Paul V. Murphy
PII:
S0045-2068(18)31151-9
DOI:
https://doi.org/10.1016/j.bioorg.2018.12.003
YBIOO 2667
Reference:
Bioorganic Chemistry
To appear in:
Received Date:
Revised Date:
Accepted Date:
9 October 2018
23 November 2018
3 December 2018
Please cite this article as: J. Zhou, A. Negi, S.I. Mirallai, R. Warta, C. Herold-Mende, M.P. Carty, X-S. Ye, P.V.
Murphy, N-Alkyl-1,5-Dideoxy-1,5-Imino-L-Fucitols as Fucosidase Inhibitors: Synthesis, Molecular Modelling and
Activity Against Cancer Cell Lines, Bioorganic Chemistry (2018), doi: https://doi.org/10.1016/j.bioorg.
2018.12.003
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

N-Alkyl-1,5-Dideoxy-1,5-Imino-L-Fucitols as Fucosidase Inhibitors: Synthesis,
Molecular Modelling and Activity Against Cancer Cell Lines
Jian Zhou,a Arvind Negi,b Styliana I. Mirallai,b Rolf Warta,c Christel Herold-Mende,c
Michael P. Carty,d Xin-Shan Ye,e Paul V. Murphy*b
a School of Chemistry and Chemical Biology, University College Dublin, Dublin 4,
Ireland, D04 V1W8
b
School of Chemistry, National University of Ireland Galway, University Road,
Galway, Ireland, H91 TK33
c Division of Experimental Neurosurgery, Department of Neurosurgery, University of
Heidelberg, Im Neuenheimer Feld 400, 69120 Heidelberg, Germany
d
School of Natural Sciences, National University of Ireland Galway, University
Road, H91 TK33 Galway, Ireland, H91 TK33
e
State Key Laboratory of Natural and Biomimetic Drugs and School of
Pharmaceutical Sciences, Peking University, Xue Yuan Road No. 38, Beijing 100191,
China
Abstract
1,5-Dideoxy-1,5-imino-L-fucitol (1-deoxyfuconojirimycin, DFJ) is an iminosugar that
inhibits fucosidases. Herein, N-alkyl DFJs have been synthesised and tested against
the α-fucosidases of T. maritima (bacterial origin) and B. taurus (bovine origin). The
N-alkyl derivatives were inactive against the bacterial fucosidase, while inhibiting the
bovine enzyme. Docking of inhibitors to homology models, generated for the bovine
and human fucosidases, was carried out. N-Decyl-DFJ was toxic to cancer cell lines
and was more potent than the other N-alkyl DFJs studied.
Keywords: 1-deoxyfuconojirimycin, N-alkylation, iminosugar, fucosidase, homology
modelling, anti-cancer activity.
1. Introduction
Iminosugars and their derivatives are inhibitors of enzymes of medicinal interest.
These include glycosidases, glycosyltransferases [1, 2], glycogen phosphorylases [3],
nucleoside-processing enzymes [4],
a
sugar nucleotide mutase [5, 6],
metalloproteinases [7] and others [8]. Iminosugars inhibit glycosidases, due to their
ability to bind at the active sites of these enzymes [9-11]. The range of enzymes
inhibited by iminosugars indicates they have promise as new medicines for diseases
such as diabetes, viral infections or lysosomal storage disorders [12] α-Fucosidase is
involved in the removal of non-reducing terminal L-fucose residues that are connected
to oligosaccharides via α-1,2; α-1,3; α-1,4 or α-1,6-linkages. L-Fucose (1) is found on
glycans that participate in cell-cell interactions and cell migration. These events are
connected to physiological and pathological processes such as fertilization,

embryogenesis, lymphocyte trafficking, immune responses, and cancer metastasis
[13-15]. A variety of physiological and pathological events are associated with
fucose containing glycoconjugates. For instance, an aberrant distribution of α-
fucosidase has been reported as being relevant to inflammation [16], cancer [17], and
cystic fibrosis [18]. These enzymes have been recognized as diagnostic markers for
the early detection of colorectal [19] and hepatocellular cancers, and this is due to the
presence of α-fucosidase in patient serums. α-Fucosidase inhibitors may be used to
study their functions and could form the basis of developing therapeutic agents [20].
Fuconojirimycins 3, 4 and the 1-deoxy analogue 2 (DFJ) are key inhibitors of α-L-
fucosidases. Other compounds assessed for their fucosidase inhibitory properties are
also shown in Fig. 1 with their reported Ki values.
Fig. 1. Various fucosidase inhibitors: L-fuconojirimycin 3 & 4) [21]; natural or semi-
synthetic inhibitors contains 1-deoxyfuconojirimycin (DFJ), β-hydroxymethyl 1-
deoxyfuconojirimycin (β-HFJ), 1-deoxymannojirimycin (DMJ), mannojirimycin (α-
HMJ) [22, 23].
In recent years, N-alkylated iminosuars have shown improved in vivo and in vitro
activities, mainly due to their improved lipophilicity, which may facilitate their
crossing of the plasma membrane into cells [24-26]. In some cases, the appended
alkyl groups are involved in hydrophobic interactions in the hydrophobic pocket of
the target proteins. N-Butyl-1-deoxynojirimycin (zavesca®) has been approved for
Gaucher disease and N-hydroxyethyl-DNJ (glyset®) is used for type-II diabetes
associated complications [27]. Also, N-alkyl DNJs [28, 29] such as zavesca®, N-
nonyl-DNJ and 5a-c [30] (shown in Fig. 2) act as highly potent pharmacological
chaperones for the potential treatment of Gaucher [31] and Pompe [32] diseases by
‘rescuing’ mutant enzymes. Kelly et al [33] and Overkleeft et al [34] have shown that
additional attachment of a large lipophilic substituent (such as the adamantyl group)
(see 5d-e in Fig. 2) increases the interaction with the glucocerebrosidase involved in
lysosomal glycosphingolipid processing.

Fig. 2. Bioactive N-alkylated iminosugars
A number of N-alkyl DFJ derivatives have been synthesised previously. This
includes an N-(aminopropyl) derivative prepared by Hung’s group which had an IC50
of 70 nM as an inhibitor of the fucosidase from T. maritima [35]. Some N-alkyl DFJ
derivatives have been evaluated against fucosyl transferases [36]. N-Methylated-DFJ
showed weak anti-HIV activity and no cytotoxicity in a study where various
glycosidase inhibitors were screened [37]. Conformationally constrained N-alkyl DFJ
derivatives were inhibitors of bovine epididymis -L-fucosidase [38]. An analogue of
castanospermine with the L-fuco configuration and its fucosidase activity was
described by Paulsen and co-workers [39].
Herein, the synthesis of new N-alkylated-1-deoxyfuconojirimycins and their testing
against α-fucosidases and cancer cell lines has been carried out. Molecular modelling
has been used to generate hypotheses about their modes of binding. N-Decyl-1-
deoxyfuconojirimycin (N-decyl-DFJ) was found to be toxic against various cancer
cell lines.
2. Results and Discussion
2.1 Chemistry
The synthetic work investigated herein is summarized in Scheme 1. The attempted
preparation of the N-alkylated DFJ derivatives was investigated initially from the
altrose derivative 7 prepared from 8 as previously described [40, 41]. However,
synthesis from the Fleet bis-mesylate 6 [42] was ultimately the more successful route.

Scheme 1. Summary of the synthesis.
Compound 7, which adopted a furanose structure was a mixture of anomers (α:β =
2:1). A double reductive amination reaction from 7 was initially investigated (Scheme
2) with a view to generating the required benzylated DFJ derivative [43, 44]. N-Butyl
amine was hence reacted with 7 in the presence of NaBH3CN and acetic acid in
methanol at room temperature. The NMR analysis of the product mixture indicated
formation of an inseparable mixture of 1,6-dideoxyaltronojirimycin (DAJ) derivative
9 and DFJ derivative 10 (ratio ~1:1, 79%). Instead, NaBH(OAc)3 at various
temperatures (-60 to -10 °C) was tried but a mixture of stereoisomers was still
obtained [44]. The same approach using N-octylamine had a similar outcome. When
benzyl amine was used, the epimers 9 (R = Bn) and 10 (R = Bn) could be separated
and they were isolated in approximate equal amounts (36% for 9 and 40% for 10). It
was possible to improve the stereoselectivity of the reductive amination reaction by
using a chiral amine [45, 46]. When (R)-1-phenylethanamine was used, a single
isomer 11 was obtained in good yield (85%). However, NMR experiments showed
1
that 11 had an altrose configuration with its ring adopting the C4 conformation; this
1
was supported by coupling constants observed in the H-NMR spectrum (J1a,2 = 8.6
Hz; J1e,2 = 4.5 Hz). The reaction of the enantiomeric amine (S)-1-phenylethanamine
was not stereoselective (Scheme 2).
Scheme 2. Double reductive amination study from 7.

The bis-mesylate 6 (Scheme 3) which Fleet [42] and his group used in a
stereoselective synthesis of 1-deoxyfuconojirimycin was next prepared. The diol 1328
was converted to 6 and without chromatographic purification the freshly generated
bis-mesylate was gently heated in the presence of various primary amines, giving N-
alkyl DFJs. A number of N-substituted DFJs 14a-g were thus obtained in 65-80 %
yield over 2 steps from the diol 13. Adamantyl containing compounds have shown
broad activities, including antiviral activity and for this reason were incorporated.
Terminating the N-alkyl chains with a hydrophobic adamantyl group can lead to
hydrophobic interactions with a target [47]. Thus, the N-ethylamine derivative 14f
was coupled with 1-adamantanecarboxylic acid in the presence of EDC and HOBt to
give the adamantyl derivative 15 in 89% yield.
Scheme 3. The synthesis of 14 and 15.
The benzyl protecting groups were removed from 14a-14e and 15 by catalytic
hydrogenation (Table 1, Scheme 4). The deprotection was first validated by the
reaction of 14e with Pd-C in the presence of hydrogen in THF-H2O-HOAc (4:2:1) and
this gave 16 in 78% yield after ion-exchange chromatography [48]. When 14a was
reacted in the same way then compound 22 was isolated. This is explained by the
formation of DFJ 2 after removal of all the benzyl groups from 14a. The DFJ 2 then
reacted with traces of tetrahydrofuran-2-ol (or 4-hydroxybutanal) which could be
formed from dihydrofuran (DHF) present in THF. The reaction of 2 with 4-
hydroxybutanal would give the iminium ion 21 which would be further reduced with
hydrogen to give 22 [49]. When the solvent was changed to methanol instead of THF
then only 2 was obtained from 14a. The reaction of 14b-e and 15 gave 16-20 (Table
1). The hydrogenation reaction of 9 (R=Bn) and 11 gave 1,6-dideoxyaltronojirimycin
5 (R=H) while reaction of 10 gave 1-deoxyfucojirimycin 2.

Scheme 4. Formation of DFJ 2 and derivatives.
The NMR spectral data for 2 coincided with those reported previously for 1-
deoxyfuconojirimycin [42, 50]. Although the NMR data of compound 5 has not been
reported previously, its specific rotation agreed with the earlier reported value: [α]20
D
0.9 (c 0.55, MeOH) [lit. [α]20 2 (c 1.1, MeOH)] [51, 52].
D
Table 1. Formation of DFJ derivatives by catalytic hydrogenationa
Conditionsa
Isolated Yield (%)b
Reactant
14a
Product
Entry
MeOH, HCl aq.
92
85
2
1
2
THF, H2O, HOAc
(4 :2 :1)
14a
22
MeOH, HCl aq.
90
85
88
78
14b
14c
14d
14e
16
17
18
19
3
4
5
6
MeOH, HCl aq.
MeOH, HCl aq.
THF, H2O, HOAc
(4 :2 :1)
MeOH, HCl aq.
MeOH, HCl aq.
83
93
89
83
15
9
20
5
7
8
MeOH, HCl aq.
MeOH, HCl aq.
10
11
2
9
5
10
a. Pd-C, H2, with various solvents and acids
b. After ion exchange chromatography.

2.2 Inhibition of fucosidases and structure activity relationship
The N-alkyl DFJs were tested as inhibitors of the fucosidase from T. maritima
(bacterial origin) and B. taurus (bovine origin), and the IC50 values obtained are
summarized in Table 2. The T. maritima enzyme shares 38% identity with its human
counterpart and is speculated to have a role in modification of hemicelluloses [48].
The bovine derived fucosidase has been widely used in N-glycan, blood group
oligosaccharide and glycolipid analysis. Only DFJ (2) showed moderate inhibition
(IC50, 8 µM) of the fucosidase from T. maritima. The active site of the fucosidase
from T. maritima did not tolerate the presence of a butyl or hydroxybutyl group on the
piperidine. Examination of the binding pose of L-fucose with the -fucosidase of T.
maritima in a co-crystal structure [53] indicated that the L-fucose is tightly enveloped
by residues in the enzyme. These residues (see in Fig. 3) could prevent binding of N-
substituted iminosugars assuming the fucopyranose mimetic 2 binds in a similar
manner to L-fucopyranose.
Fig. 3. The interactions between α-fucosidase from T. maritima and α-L-
fucopyranose. H-bonds are shown as dashed lines, and van der Waals contacts are
shown as bold dotted lines.
The N-butyl derivative 18 did not inhibit, at concentrations up to 100 M, the
fucosidase B. taurus from bovine kidney (Table 2), whereas the butanol derivative 22
had an IC50 of 30 M. We determined an IC50 of 0.3 µM for 2 which agreed with a
reported IC50 for 2 (0.4 M) for the fucosidase from B. taurus [54]. However,
lengthening the alkyl chain led to improved activity for as seen for N-decyl derivative
16 (IC50 = 2.5 µM). By comparing the aliphatic side chains of 16-18, it can be
concluded that having a longer alkyl group gives rise to improved inhibitory activity.
Compound 20 which had an adamantyl group on the N-side chain was less potent with
an IC50 of 90 µM. The N-ethanol derivative 19 showed IC50 value of 27 µM, similar
to that of the N-butanol derivative The altronojirimycin derivative 5 Table 1) did not
show any inhibition towards the tested α-fucosidases.
In addition to the IC50 determination, some physiochemical parameters (LogP, LipE
and logD) were calculated. In terms of oral administration, LogP values should be
less than 5 according to Lipinski’s rules of five, which most of the N-alkylated

derivatives show [55]. LogP values between 2 and 3 are often considered optimal to
achieve a compromise between permeability and first pass clearance, which is quite
comprehensively shown by compound 16. LipE is another parameter which links the
potency (IC50) and lipophilicity (logD) in an attempt to estimate drug-likeness of a
particular structure. A satisfactory LipE value would indicate selectivity for the target
of interest versus a generic hydrophobic environment [56]. A LipE measurement of 6
represents one-million-fold selectivity for the target versus a generic hydrophobic
environment while a LipE of zero indicates no selectivity. Compounds 2 and 19 had
values higher than 6.
Table 2. IC50 values (µM) and other parameters
Compound
IC50
(fucosidase
from B. taurus)
IC50
(fucosidase
from T.
maritima)
8
LogPd
LogD
For fucosidase
from B. taurus
pIC50
LipEe
2 (DFJ)
17
0.3 (0.4)[54]
-1.71
1.723
2.60
-2.23
1.02
6.52
4.60
5.60
<4
8.75
3.57
3.55
N. A.
6.40
3.47
5.86
25
2.5
>100
27
N. D.b
N. D.
16
2.04
>100
-0.04
-2.13
-0.07
-1.25
-1.71
-1.02
-1.83
1.05
18
N. D.
19
4.56
4.52
4.04
<4
30
>100
22
90
N. D.
20
-1.81
-2.23
>100
>100
5
N. A.
b N. D. = not determined; based on Labute et al method calculated by CCG MOE2018.01;
d
e
LipE = pIC50 - log D [56];
2.3 Molecular modelling
Molecular modelling was used in order to hypothesise how the fuconojirimycin
derivatives may interact with the fucosidases and to provide a basis for the
development of more potent inhibitors. The 3D-coordinates for the fucosidase from T.
maritima (PDB code: 2ZXD) were retrieved from Royal Collaboratory for Structural
Bioinformatics Protein Data Bank (RCSB-PDB) [57]. The fucosidase of bovine origin
was modelled using the α-L-fucosidase of T. maritima as a template. The sequence
for the bovine fucosidase B. taurus was obtained from NCBI (National Center for
Biotechnology Information). A BlastP search was then conducted which showed a
number of possible templates for constructing the homology model (results from the
BlastP search are provided in the supplementary information). Five templates were
found with >35% of residues matching, and the length of each template was ≥ 85% of
that of the B. taurus input sequence and therefore homology models were built for
each of these. These homology models were then evaluated based on their qualitative
structural uniformity. Model-4 (Table 3) developed based on the template of a
fucosyl hydrolase from a strain of T. maritima was selected based on the data
presented in Table 3 for further molecular modelling studies. Three residues were

found to be outliers in this model, and they were later were energy minimised. The
plots obtained before and after minimisation are provided in the supplementary
information. Comparing the pairwise sequence alignment of fucosidase of B. taurus
with that from T. maritima showed the similarity of their structures (Figure 4).
Table 3 Evaluation of the constructed homology models prepared from five available
PDBs with identity ≥ 35% in their amino acid sequence. Qualitative structural
measurements are provided in the form of Ramachandran plot, ERRAT plot, Verify-
3D, ProA, Z-Score and superpose of homology model with their templates
Models Ramachandran Plota
ERRA
T Plot
Verif Superpose ProSa Z-Score
y 3D RMSD (Å)
%
residue
in
favoured
region
% residue
in
additional
allowed
region
% residue
in
generously disallowed
allowed
region
% residue
in
Template
Homolog Templ ΔZ-
y model
ate
score
region
1
2
3
4
5
85.1
84.5
88.2
85.1
69.7
7.3
0.6
1.5
1.5
1.9
3.1
0.9
84.57
86.74
82.20
86.04
79.73
80.65 1HL9_B
(0.739)
-7.26
-7.47
-6.96
-7.69
-6.75
-9.87
-10.29
-9.33
-10.27
-9.81
0.26
0.27
0.25
0.25
0.31
13
0.9
0.9
0.9
1.5
89.25 2ZWY_A
(1.174)
9.3
80.11 1HL9_A
(1.050)
87.37 2WSP_A
(0.926)
86.24 1ODU_A
(1.89)
12.1
25.7
a residues in Ramachandran plot are calculated as ~0.3 = 1
Fig. 4. The sequence alignment of model 4 with its template 2wsp_A showing
evolutionary conserved residues.
The docking of 2 with the crystal structure for the fucosidase from T. maritima was
carried out and it suggested a H-bond acceptor/donor network, where hydroxyl groups
of the iminosugar interacted with the side chain of Glu66 (1.97 Å), Trp67 (2.33 Å)

and Asp224 (2.04 Å), as shown in Fig. 5A. The protonated nitrogen of the iminosugar
had a H-bond donor interaction with the carboxyl group in the side chain of Asp224
(2.05 Å). The docking showed how closely the calculated binding mode of 2
resembles the already reported binding of iso-6FNJ in the crystal structure bound to
the fucosidase from T. maritima (see Fig. 5B). Furthermore, the docking of 2 with the
homology model of the B. taurus fucosidase indicated similar binding interactions to
those seen for T. maritima. The hydroxyl groups of 2 had H-bond acceptor/donor
interactions with Val77 (2.25 Å) and Asn282 (2.31 Å), while the protonated nitrogen
was predicted to have H-bond donor interactions with Cys240 (2.45 Å), see in Fig.
5C. Also, a similar trend was noted after docking of 16 as its C-4 hydroxyl group was
involved in a H-bond donor interaction with Cys240 (1.94 Å), while the protonated
nitrogen had a H-bond donor interaction with Cys240 (1.92 Å). The nonane chain of
16 was placed by docking into a hydrophobic cavity constituted by aromatic residues
(His82, Tyr126, Trp172) on one side while residues 281-283 and Trp248 were found
on the other side of the domain. These are shown in Fig. 5D.
Fig. 5. (A) The docking pose of 2 (orange) with the fucosidase of T. maritima; (B)
superpose of binding pose of 2 (orange) with co-crystallised ligand (brown) of
2WSP_A (model 4, Table 3); (C) docking pose of 2 with fucosidase of B. taurus; (D)
docking pose of 16 (blue) with fucosidase of B. taurus.
In order, to gain confidence regarding the reliability of the docked poses for 2 and 16,
they were selected for 100 ps molecular dynamics simulations (MD). The complex of

2 with the fucosidase of T. maritima showed stability after 10 picoseconds (ps) of the
simulation (RMSD was consistently ~1.9 Å, Fig. 6A). The same protocol was applied
on the docked complex of 2 with the fucosidase of B. taurus; the ligand-protein
complex showed fluctuation in first 10 ps and then the RMSD maintained a stable
value ~ 2.8 Å (Fig. 6B). The complex of 16 to fucosidase of B. taurus complex
showed stability after 20 ps with an RMSD value of ~3.2 Å (Fig. 6C). The MD
simulation demonstrated the docked ligand-protein complexes were stable and
generated confidence that they form the basis of a reasonable hypothesis as to the
nature of inhibitor binding [58].
Figure 6. MD run over 100ps, (A) compound 2 (with fucosidase of T. maritima) with
RMSD 1.88 Å; (A) Compound 2 with B. taurus fucosidase has RMSD 2.76 Å (C)
Compound 16 with B. taurus fucosidase had an RMSD of 3.20 Å.
To further evaluate molecules against human fucosidase, a reverse docking strategy
was performed [59-62]. The consistency and suitability of docking placement

methods (in MOE 2018.01) with respect to the activities of compounds were
evaluated. Hence the 3D structure of human alpha fucosidase was next modelled
using the same template which was used for fucosidase of B. taurus origin (2WSP_A).
The quality of the human homology model generated (Fig. 7A) was evaluated using
Verify 3D (88.81%) and Errat (88.86), Ramachandran plot (residues in most favoured
regions (86.9%), residues in additional allowed regions (10.5%), residues in
generously allowed regions (1.4%) residues in disallowed regions (1.1%,)) and ProSA
Z-score (plots are available in supporting information). The residues Lys52, Phe184,
His396 and Asp432 were found as outliers and are not in the active site (Fig. 7B)
Figure 7. (A) The RMSD value between the template (2WSP, yellow) and the human
fucosidase homology model (green), which are superposed, was 0.71. The black
rectangle outlines the active site. (B) The binding site is shown in grey. The red
spheres are atoms with potential to interact with hydrophobic groups, whereas the
grey spheres correspond to atoms that have potential as H-bonding donors/acceptors.
The IC50/KD values (Table 4) for inhibitors (Fig. 8) of the human fucosidase (HuF)
from the work of Ho et al [63] were used to generate scatter plots (see Fig. 9). Trend
lines were generated and R2 values were determined; these showed an association
between the docking placement method with respect to the inhibitory properties. The
docking scores are shown in Table 4.

1
2
Figure 8. The compounds of Ho et al [35] used in the docking study.
3
4
5
6
Table 4. Comparison of scores attained from docking placement methods against the
IC50/KD values for the compounds against human fucosidase (HuF) and T. maritima
fucosidase ™.
Compd Ki/IC50 Ki/IC50 for Alpha Triangle Triangle Matcher Alpha PMI
for HuF TM (µM) Method
(µM)
Method
HuFa
Method
HuFa TMb
HuFa
TMb
TMb
0.0056c 0.000105c -12.037 -12.114 -11.887 -12.199 -13.011 -12.357
23
24
25
26
27
28
29
30
0.0097c 0.000259c -11.451 -11.78 -11.764 -12.002 -12.882 -12.41
0.0117c 0.00119c
-11.393 -11.491 -11.119 -11.833 -12.50 -12.042
-11.108 -11.542 -10.810 -11.809 -12.619 -11.945
-10.920 -11.304 -10.991 -10.892 -12.451 -11.807
-10.543 -10.928 -10.605 -10.712 -11.521 -11.065
0.018c
0.018d
0.035d
0.092d
0.106d
0.00101c
0.052d
0.064d
0.070d
0.267d
-10.008
-9.332
-11.402
-10.288 -10.42
-10.637
-10.219 -10.419 -9.290 -9.074 -11.477 -10.317
b
c
7
8
9
a Human fucosidase homology model; fucosidase of T. maritima Ki values in µM; d IC50 values in
µM.
10
11
12
13
14
15
16
For this study, three different docking placement methods, available in MOE 2018.01,
were investigated. Of the docking methods, the triangle matcher placement method
performed better than the alpha triangle and alpha PMI placement methods. With this
method there were coefficients of determination of 0.908 and 0.8731, which were
higher than those of the other two methods (see Fig. 9 and 10).

17
18
Figure 9. Plots of docking E-score (y-axis) and Ki/IC50 values (nM) for human
19
20
21
fucosidase (x-axis) for three docking placement methods (See Table 4 for data used).
22
23
Figure 10. Plots of E-score (y-axis) and Ki/IC50 values (x-axis) for inhibitors of the T.
maritima fucosidase for three docking placement methods (See Table 4 for data used)
24
25
26
27
28
29
30
The available enzyme inhibitory data from Ho et al [35] for the two different enzymes
were used to generate the plot in Fig. 11. These gave a trendline with an R2 value of
0.7169. The R2 values observed are sufficiently high to imply that the enzyme
inhibitory properties for compounds are associated with binding to the specified
active site cavity in the fucosidases [64].

31
32
Figure 11. Plots of IC50/KD values for inhibitors of human fucosidase (x-axis) versus
those of T. maritima fucosidase (y-axis, see Table 4 for the data used).
33
34
35
36
37
38
The three docking methods all predicted that the N-propyl amine derivative 29 could
have a different binding location within the active site (E-Score = -10.288) compared
to other DFJ derivatives (Fig. 12).
39
40
Figure 12. (A) The black contour line outlines the active binding site in fucosidase of
T. maritima. Here 29 was predicted to differ in binding position with respect to DFJ
2; (B) The docked conformations of 29, corresponding to the three docking methods,
are shown overlayed in blue (triangle matcher), magenta (alpha triangle), orange
(alpha PMI).
41
42
43
44
45
46
47
48
49
50
In the docking of 18 (E-score = -5.613) and 22 (E-Score = -6.175) high E-scores were
found, which are consistent with the lack of inhibition of the enzymes observed
experimentally.
On the other hand, the N-butyl alcohol derivative 22 did show inhibitory
activity (IC50 = 30 µM) towards the fucosidase of B. taurus and docking indicated this

51
52
53
54
could be reasonably supported by the OH group of the butyl chain being involved in a
hydrogen bonding interaction with the peptide backbone of Lys74 (2.97 Å) of this
fucosidase.
55
56
Figure 13. (A) The binding of 18 (yellow colour) and 22 (cyan colour) in fucosidase
of T. maritima. This showed 18 and 22 outside the binding regions predicted to be
occupied by 29 and DFJ. (B) The docking pose of 22 with respect to the fucosidase
of B. taurus shown with predicted H-bonding interaction (green line) of Lys74.
57
58
59
60
61
62
63
64
65
66
The docking was also performed on human fucosidase (HuF) where binding of both
the N-propylamine derivative 29 (E-score = -9.322) and N-decyl derivative 16 (E-
score = -8.724) is predicted to occur (Fig. 14A). The difference between the binding
score of 16 (E-score = -8.724) to HuF and to the fucosidase of T. maritima (E-score =
-6.901) indicates it may be of interest to evaluate 16 for its inhibition of human
fucosidase in due course.

67
68
69
70
71
72
Figure 14. (A) The predicted binding of 29 (purple) and 16 (green) in the fucosidase
of human origin, along with the clustered conformations. The predicted 2-D
interaction
of
29
(B)
and
16
(C)
with
human
fucosidase.

Cytotoxicity against BT-474, MCF-7 and DU-145 cell lines using the MTT
Assay
The BT-474 (breast cancer), MCF-7 (breast cancer) and DU-145 (prostate cancer)
cell lines, available at the National University of Ireland Galway (NUI Galway),
were used to independently determine whether compounds 5, 16, 18, 20 and 22
were cytotoxic using the MTT assay. Compound 16 was found to be the most toxic
against all cell lines exhibiting a greater potency towards prostate cancer cell line
(DU-145). Toxicity of pleurotin, a positive control, is also given in Table 5 [64,
65]
Table 5 Cytotoxicity evaluation using the MTT colorimetric assay. IC50 values
were obtained after the incubation of cells with the test compounds in DMSO for
72 h.
Compound
Pleurotin (control)
IC50 BT-474 (μΜ) IC50 MCF-7 (μΜ)
IC50 DU-145 (μΜ)
0.43 ± 0.06*
>100
1.94 ± 0.49*
>100
0.28 ± 0.03*
>100
5
57.39 ± 3.38
>100
30 ± 0.04
77 ± 1.53
>100
13.54 ± 0.75
60.99 ± 3.15
>100
16
18
20
22
>100
>100
>100
>100
Inhibition of proliferation of human head and neck carcinoma cell lines
Compound 16 was further tested for its inhibition of proliferation of human head and
neck carcinoma cells (HNO41, HNO97, HNO210), as well that of patient-derived
glioblastoma-initiating cells (NCH644 IDH1-wt GBM) using the CellTiter-Glo assay.
Compound 16 inhibited the proliferation of all these cell lines with IC50 values
ranging from 12 M to 17.6 M (Table 6). The inhibition by pleurotin, a positive
control, is also reported for these various cell types.
Table 6. Inhibition of proliferation of cancer cell lines by 16
IC50
puromycin
[µM]
of
Cell line
IC50 of 16
[µM]
NCH644
HNO210
HNO97
HNO41
17.6
17.4
12.0
13.8
0.27
0.16
0.24
0.24
Conclusions
A series of N-alkylated fuconojirimycin analogues have been prepared and evaluated
as inhibitors of fucosidases of bacterial and bovine origin. The best inhibitory activity
for the new compounds, was observed for N-decyl-DFJ against the bovine fucosidase
(IC50 = 2.5 M). The pairwise sequence alignment for the two enzymes showed their
close resemblance, sharing evolutionary conserved domains. Homology models of

the fucosidases from human and bovine origin were thus both constructed and used in
docking and molecular dynamics. N-Decyl-DFJ was shown to inhibit growth of
various cell lines, including IDH wild type glioblastoma cells [65] and could form the
basis for development of more potent inhibitors of tumour cell growth.
1. Experimental section
1.1 General methods for preparative chemistry
Optical rotations were determined with a Perkin-Elmer 343 model polarimeter at the
sodium D line at 20 °C. Deuterated chloroform (CDCl3), CD3OD or D2O were used as
NMR solvents, unless otherwise stated. NMR spectra were recorded (30 °C) with 400
or 500 MHz spectrometers. Chemical shifts are reported relative to internal Me4Si in
1
CDCl3 ( 0.0) or HOD for D2O ( 4.79) or CD2HOD ( 3.31) for H and Me4Si in
13
1
CDCl3 ( 0.0) or CDCl3 ( 77.0) or CD3OD ( 49.05) for C. H-NMR signals were
assigned with the aid of COSY. 13C-NMR signals were assigned with the aid of
DEPT, gHSQCAD and/or gHMBCAD. Coupling constants are reported in hertz.
Splitting patterns are designated as follows: s, singlet; d, doublet; t, triplet; q, quartet;
m, multiplet and br, broad. Low and high-resolution mass spectra were in positive
and/or negative mode as indicated in each case.
All anhydrous reactions were
performed in flame-dried or oven-dried glassware under a positive pressure of dry
argon or nitrogen. Air or moisture-sensitive reagents and anhydrous solvents were
transferred with oven-dried syringes or cannulae. All flash chromatography was
performed with E. Merck silica gel 60 (230-400 mesh). All solution phase reactions
were monitored using analytical thin layer chromatography (TLC) with 0.2mm pre-
coated silica gel aluminum plates 60 F254. Components were visualized by
illumination with a short-wavelength (254 nm) ultraviolet light and/or staining (ceric
ammonium molybdate, potassium permanganate, or phosphomolybdate stain
solution). Flash chromatography was carried out with silica gel 60 (0.040-0.630 mm)
and using a stepwise solvent polarity gradient correlated with TLC mobility. CH2Cl2,
MeOH, toluene, CH3CN and THF reaction solvents were used as obtained from a
Pure Solv™ Solvent Purification System. Anhydrous DMF, pyridine were used as
purchased. Chromatography solvents were used as obtained from suppliers. All other
reagents were obtained from Sigma-Alrich unless otherwise stated.
1.2. Experimental procedures
1.2.1 N-Benzyl-2,3-di-O-benzyl-1,5,6-trideoxy-1,5-imino-D-altritol 9 and N-
benzyl-2,3-di-O-benzyl-1,5-dideoxy-1,5-imino-L-fucitol 10.
Compound 7 (39
mg, 0.11 mmol) was dissolved in MeOH (2 mL). Benzylamine (20 µL, 0.17 mmol)
in MeOH (1 mL) was mixed with acetic acid (13 µL, 0.22 mmol) to adjust the pH
value to 4-5, and this solution was added to the reaction mixture over a period of 0.5 h
by cooling over an ice-water bath. A portion of NaBH3CN (22 mg, 0.33 mmol) was
then added, and the mixture was stirred for 18 h at room temp. The mixture was

quenched with 0.5 mL 1N HCl aqueous solution. After removal of the solvent, the
residue was dissolved in 10% Na2CO3 (10 mL), and extracted with EtOAc (3 x 7 mL).
The EtOAc layers were combined, dried (Na2SO4), filtered, and the solvent was
removed. Chromatography of the residue on silica gel gave 9 (16 mg, 36 %) and
compound 10 (19 mg, 40%); 9 was eluted with cyclohexane-EtOAc (7:1), while 10
was eluted with cyclohexane-EtOAc (5:1).
1
Analytical data for 9: [α]20 -14.6 (c 1.6, CHCl3); H NMR (500 MHz, CDCl3) δ
D
7.25-7.33(m, 15H, HAr), 4.69 (d, J = 11.7 Hz, 1H, PhCHH), 4.64 (d, J = 11.7 Hz, 1H,
PhCHH), 4.57 (d, J = 11.7 Hz, 1H, PhCHH), 4.53 (d, J = 11.7 Hz, 1H, PhCHH), 3.78
(d, J = 13.0 Hz, 1H, PhCHH), 3.76 (m, 1H, H2), 3.72 (dd, J = 4.6, 3.4 Hz, 1H, H4),
3.62 (dd, J = 7.4, 3.4 Hz, 1H, H3), 3.50 (d, J = 13.0 Hz, 1H, PhCHH), 2.91 (m, 1H,
H5), 2.67 (dd, J = 12.0, 4.5 Hz, 1H, H1-eq), 2.55 (dd, J = 12.0, 8.1 Hz, 1H, H1-ax), 1.69
(s, 1H, OH), 1.09 (d, J = 6.70 Hz, 1H, CH3), 13C NMR (125 MHz, CDCl3) δ 138.82,
138.55, 138.36 (each C), 128.68, 128.38, 128.34, 128.31, 128.15, 127.78, 127.69,
127.67, 127.53, 127.08, 126.95 (each CH), 78.84 (C3), 74.95 (C2), 72.27 (PhCH2),
71.36 (C4), 57.83 (PhCH2), 57.80 (C5), 48.77 (C1), 10.42 (CH3, C6); HRMS: Calcd for
C27H32NO3, [M+H]+, 418.2387; Found, 418.2386.
1
Analytical data for 10: [α]20 16.4 (c 1.0, CHCl3); H NMR (500 MHz, CDCl3) δ
D
7.41 – 7.20 (m, 15H, HAr), 4.74 (dd, J = 14.0, 12.0 Hz, 2H, PhCH2), 4.67 (d, J = 11.6
Hz, 1H, PhCHH), 4.53 (d, J = 11.6 Hz, 1H, PhCHH), 3.95 (d, J = 13.6 Hz, 1H,
PhCHH), 3.82 (dd, J = 3.1, 1.4 Hz, 1H, H4), 3.81 – 3.75 (m, 1H, H2), 3.32 (dd, J =
9.1, 3.2 Hz, 1H, H3), 3.29 (d, J = 13.6 Hz, 1H, PhCHH), 2.99 (dd, J = 11.5, 5.1 Hz,
1H, H1-eq), 2.48 – 2.37 (m, 1H, H5), 1.94 – 1.84 (m, 1H, H1-ax), 1.34 (d, J = 6.5 Hz,
13
3H, CH3). C NMR (125 MHz, CDCl3) δ 138.81, 138.64, 138.34 (each C), 129.06,
128.51, 128.46, 128.41, 127.90, 127.78, 127.76, 127.63, 127.22 (each CH), 83.45
(C3), 75.88 (C2), 73.13 (PhCH2), 72.18 (PhCH2), 72.05 (C4), 58.88 (C5), 56.81
(PhCH2), 54.84 (C1), 16.94 (CH3, C6). HRMS: Calcd for C27H32NO3, [M+H]+,
418.2387; Found, 418.2382.
1.2.2 N-((R)-1-Phenylethyl)-2,3-di-O-benzyl- 1,5,6-trideoxy-1,5-imino-D-altritol
(11)
Compound 7 (29 mg, 0.085 mmol) was dissolved in methanol. (R)-Phenylethanamine
(22 µL, 0.17 mmol) in MeOH (1 mL) was treated with acetic acid to adjust the pH
value to 4-5, and this solution was added to the reaction mixture at room temp. A
portion of NaBH3CN (16 mg, 0.25 mmol) was then added, and the mixture was stirred
at 60 °C for 2 hr then left at room temp overnight. The mixture was quenched with 0.5
mL 1N HCl aqueous solution. After removal of the solvent, the residue was dissolved
in 10% Na2CO3 (8 mL), and extracted with EtOAc (3 x 6 mL). The EtOAc layers
were combined, dried (Na2SO4), filtered, and the solvent was removed.
Chromatography of the residue on silica gel (cyclohexane-EtOAc, 8:1) gave 11 (31
1
mg, 85%) as a white solid; [α]20D -1.8 (c 1.0, CHCl3); H NMR (500 MHz, CDCl3) δ
7.39 – 7.13 (m, 15H, HAr), 4.70 (d, J = 11.9 Hz, 2H, PhCH2,), 4.45 (d, J = 11.7 Hz,
1H, PhCHH), 4.36 (d, J = 11.7 Hz, 1H, PhCHH), 3.82 – 3.72 (m, 2H, 1x NPhCHH &
H4), 3.65 (td, J = 8.3, 4.7 Hz, 1H, H2), 3.60 (dd, J = 7.9, 3.1 Hz, 1H, H3), 3.38 – 3.29

(m, 1H, H5), 2.75 (d, J = 6.6 Hz, 1H, OH), 2.63 (dd, J = 12.1, 4.5 Hz, 1H, H1-eq), 2.29
(dd, J = 12.1, 8.6 Hz, 1H, H1-ax), 1.28 (d, J = 6.7 Hz, 3H, CH3), 1.05 (d, J = 6.8 Hz,
3H, CH3); 13C NMR (125 MHz, CDCl3): δ 145.61 (C), 138.68 (C), 138.64 (C), 128.54
(CH), 128.52 (CH), 128.38 (CH), 127.96 (CH), 127.79 (CH), 127.73 (CH), 127.58
(CH), 127.26 (CH), 127.06 (CH), 79.43 (C3), 75.59 (C2), 72.41 (PhCH2), 72.34
(PhCH2), 71.57 (C4), 59.12 (PhCH(CH3)N), 54.89 (C5), 45.84 (C1), 19.19
(PhCH(CH3)N), 9.78 (CH3, C6). IR (film, CHCl3): νmax 3025, 2920, 2358, 1450, 1369,
1260, 1205, 1092, 799, 738, 698; ES-HRMS: Found 432.2523, C28H34NO3 requires
432.2539 [M+H]+.
1.2.3 N-Benzyl-2,3,4-tri-O-benzyl -1,5 -dideoxy-1,5-imino-L-fucitol (14a)
The diol 13 (95 mg, 0.46 mmol) was dissolved in CH2Cl2 (8 mL), and triethylamine
(0.14 mL, 1.01 mmol) was added and the solution was cooled to 0 °C,
methanesulfonyl chloride (75 µL, 0.96 mmol) was added dropwise and stirred for 4 h.
The solvent was removed under reduced pressure and the residue was used for next
step without further purification. The bis-mesylate was dissolved in THF (1 mL) and
benzylamine (0.2 mL, 1.84 mmol) was added at room temp, and the mixture was
heated at 40 °C for 8 h, after which TLC indicated complete consumption of the bis-
mesylate. The solution was cooled to room temp, and the solvent removed under
reduced pressure. The residue was dissolved in EtOAc (30 mL), and washed with 2 M
NaOH (2 x 20 mL). The organic phase was separated, further washed with brine and
dried with Na2SO4. Removal of the solvent under reduced pressure and flash column
chromatography of the residue on silica gel (cyclohexane-EtOAc, 10:1) gave 14a (80
mg, 77%) as a colourless oil; Rf 0.65 (cyclohexane-EtOAc, 4:3); [α]20 -9.4 (c 0.85,
D
1
CHCl3); H NMR (500 MHz, CDCl3) δ 7.45 – 7.17 (m, 20H, HAr), 4.91 (d, J = 11.8
Hz, 1H, PhCHH), 4.82 – 4.71 (m, 2H, PhCH2), 4.66 (d, J = 11.8 Hz, 1H, PhCHH),
4.62 (d, J = 11.7 Hz, 1H, PhCHH), 4.55 (d, J = 11.7 Hz, 1H, PhCHH), 3.93 (d, J = 3.6
Hz, 1H, H2), 3.85 – 3.69 (m, 2H, PhCHH & H4 ), 3.59 (d, J = 14.0 Hz, 1H, PhCHH),
3.47 (d, J = 6.5 Hz, 1H, H3), 3.00 (dd, J = 11.9, 4.0 Hz, 1H, H1-eq), 2.62 (m, 1H, H5),
13
2.22 – 2.06 (m, 1H, H1-ax), 1.21 (d, J = 6.5 Hz, 3H, CH3); C NMR (125 MHz,
CDCl3): δ 139.21, 138.99, 138.23 (each C), 129.38, 128.43, 128.40, 128.31, 128.28,
128.26, 127.75, 127.55, 127.54, 127.51, 127.03 (each CH ), 83.37 (C3), 79.10 (C4),
75.83 (C2), 74.29 (PhCH2), 73.11 (PhCH2), 72.51 (PhCH2), 57.36 (C5), 57.04
(PhCH2), 52.58 (C1), 14.98 (CH3, C6); ES-HRMS: Found 508.2844, C34H38NO3
[M+H]+ requires 508.2852.
1.2.4 N- Decyl-2,3,4-tri-O-benzyl- 1,5 -dideoxy-1,5-imino-L-fucitol (14b)
Reaction of the bis-mesylate (58 mg, 0.098mmol) as described in the preparation of
14a with n-decylamine gave 14b (45 mg, 83%) as a white solid after column
1
chromatography (cyclohexane-EtOAc, 7:1); [α]20 12.0 (c 1.0, CHCl3); H NMR
D
(500 MHz, CDCl3) δ 7.43–7.21 (m, 15H, HAr), 4.91 (d, J = 12.0, 1H, 1 x PhCHH),
4.73 (m, 5H, 5 x PhCHH), 4.04 – 3.92 (m, 1H, H2), 3.64 (m, 1H, H4), 3.41 (d, J = 7.0,
1H, H3), 3.08 (dd, J = 10.0, 4.5, 1H, H1-eq), 2.63 – 2.45 (m, 2H, CH2), 2.41 (m, 1H,
H5), 2.23 (t, J = 10.0, 1H, H1-ax), 1.47–1.08 (m, 16H, 8 x CH2), 1.00 (d, J = 6.5, 3H,

NCHCH3), 0.88 (t, J = 6.5, 3H, CH2CH3). 13C NMR (125 MHz, CDCl3) δ 139.06 (C),
138.91 (C), 128.74, 128.30, 128.28, 128.05, 127.74, 127.46, 127.44, 127.38 (each
CH), 84.31(C3), 78.54 (C4), 76.32(C2), 74.38 (PhCH2), 73.03 (PhCH2), 72.75
(PhCH2), 57.12 (C5), 54.15 (NCH2Ph), 53.19 (C1), 31.89, 30.32, 29.62, 29.59, 29.55,
29.31, 27.66, 22.66 (each CH2), 15.49 (CH3, C6), 14.09 (CH3). IR (film, CHCl3): νmax
3661, 2924, 2346, 1732, 1455, 1365, 1096, 734 cm-1; HRMS: Calcd for C37H52NO3,
[M+H]+, 558.3947; Found, 558.3975.
1.2.4 N-Octyl-2,3,4-tri-O-benzyl-1,5 -dideoxy-1,5-imino-L-fucitol (14c)
Reaction of the bis-mesylate (44 mg, 0.074 mmol) as described in the preparation of
14a with n-octylamine gave 14c (33 mg, 80%) as a white solid after chromatography
1
(cyclohexane-EtOAc, 8:1); [α]20 11.2 (c 1.5, CHCl3); H-NMR (500 MHz, CDCl3)
D
δ 7.42–7.21 (m, 15H, HAr), 4.91 (d, J = 12.0, 1H, 1 x PhCHH), 4.83–4.62 (m, 5H, 5 x
PhCHH), 3.99 (td, J = 9.0, 4.5, 1H, H2), 3.64 (s, 1H, H4), 3.41 (dd, J = 9.0, 2.5, 1H,
H3), 3.08 (dd, J = 11.5, 4.5, 1H, H1-eq), 2.53 (m, 2H, CH2), 2.41 (m, 1H, H5), 2.23 (t, J
= 10.5, 1H, H1-ax), 1.48–1.35 (m, 1H, 1 x CHHCH2), 1.35 – 1.07 (m, 11H, 11 x
13
CHHCH2), 1.00 (d, J = 6.5, 3H, CHCH3), 0.87 (t, J = 7.0, 3H, CH2CH3); C NMR
(125 MHz, CDCl3) δ 139.07, 138.92 (each C), 128.74, 128.31, 128.29, 128.06,
127.75, 127.47, 127.46, 127.45, 127.39 (each CH), 84.26 (C3), 78.56 (C4), 76.32 (C2),
74.39 (PhCH2), 73.03 (PhCH2), 72.75 (PhCH2), 57.13 (C5), 54.14 (C1), 53.19
(NCH2Ph), 31.82, 29.56, 29.28, 27.67, 23.19, 22.65 (each CH2), 15.70 (CH3-C6),
14.09 (CH3); IR (film, CHCl3): νmax 3389, 3034, 2925, 2347, 1955, 1603, 1455,
1367, 1269, 1096 cm-1. HRMS: Calcd for C35H48NO3, [M+H]+, 530.3634; Found,
530.3654.
1.2.5 N-Butyl -2,3,4-tri-O-benzyl- 1,5 -dideoxy-1,5-imino-L-fucitol (14d)
Compound 14d was prepared from bis-mesylate (81 mg, 0.136 mmol) as described in
the preparation of 14a, yielding 14d (50 mg, 78%) as white solids after column
1
chromatography (cyclohexane-EtOAc 8:1); [α]20D 16.4 (c 1.0, CHCl3); H NMR (500
MHz, CDCl3) δ 7.42 – 7.22 (m, 15H, HAr), 4.92 (d, J = 11.8 Hz, 1H, 1 x PhCHH),
4.80 (d, J = 12.0 Hz, 1H, 1 x PhCHH), 4.71 (m, 4H, 4 x PhCHH), 3.99 (td, J = 9.0,
4.5 Hz, 1H, H2), 3.65 (t, J = 2.3 Hz, 1H, H4), 3.41 (dd, J = 8.8, 2.7 Hz, 1H, H3), 3.08
(dd, J = 11.0, 4.5 Hz, 1H, H1-eq), 2.58 (ddd, J = 13.6, 11.0, 5.0 Hz, 1H, NCHHCH2),
2.50 (ddd, J = 13.6, 11.0, 5.0 Hz, 1H, NCHHCH2), 2.41 (m, 1H, H5), 2.23 (dd, J =
11.0, 9.8 Hz, 1H, H1-ax), 1.48 – 1.16 (m, 4H, 2 x CH2), 1.01 (d, J = 6.5 Hz, 3H,
13
CHCH3), 0.89 (t, J = 7.3 Hz, 3H, CH2CH3); C NMR (125 MHz, CDCl3) δ 139.05
(C), 138.90 (C), 128.73, 128.29, 128.28, 128.05, 127.74, 127.46, 127.43, 127.38 (each
CH), 84.48 (C3), 78.56 (C4), 76.31 (C2), 74.39 (PhCH2), 73.02 (PhCH2), 72.74
(PhCH2), 57.13 (C5), 54.27(C1), 52.86 (NCH2), 25.33 (CH2), 20.81(CH2), 15.36 (CH3-
C6), 14.04 (CH2CH3); HRMS: Calcd for C31H40NO3, [M+H]+, 474.3008; Found,
474.3024.
1.2.6 N-(2-(Benzyloxy)ethyl)-2,3,4-tri-O-benzyl-1,5-dideoxy-1,5-imino-L-fucitol
(14e)

Compound 14e was prepared from bis-mesylate (85 mg, 0.144 mmol) as described in
the preparation of 14a, yielding 14e (60 mg, 75%) as white solids after column
chromatography (cyclohexane-EtOAc4:1); [α]20 2.9 (c 1.25, CHCl3); 1H NMR (500
D
MHz, CDCl3) δ 7.41 – 7.22 (m, 20H, HAr), 4.91 (d, J = 12.0, 1H, 1 x PhCHH), 4.80
(d, J = 12.0, 1H, 1 x PhCHH), 4.77 – 4.66 (m, 3H, 3 x PhCHH), 4.64 (d, J = 12.0, 1H,
1 x PhCHH), 4.47 (s, 2H, 2 x PhCHH), 3.97 (td, J = 9.0, 4.5, 1H, H2), 3.65 (m, 1H,
H4), 3.61 – 3.48 (m, 2H, NCH2CH2), 3.43 (dd, J = 9.0, 2.5, 1H, H3), 3.16 (dd, J =
11.0, 4.5, 1H, H1-eq), 2.93 – 2.78 (m, 2H, CH2OBn), 2.54 (m, 1H, H5), 2.42 (t, J =
13
11.0, 1H, H1-ax), 1.05 (d, J = 6.5, 3H, CH3). C NMR (125 MHz, CDCl3) δ 139.04
(C), 138.86 (C), 138.38 (C), 128.65, 128.33, 128.30, 128.27, 128.07, 127.73, 127.54,
127.50, 127.45, 127.38 (each CH), 84.48 (C3), 78.56 (C4), 76.00 (C2), 74.39 (PhCH2),
73.18 (PhCH2), 73.04 (PhCH2), 72.67 (PhCH2), 67.21(NCH2), 57.72 (C5), 54.38 (C1),
51.94 (CH2), 15.74 (CH3, C6). ES-HRMS: Found 552.3114, C36H42NO4 [M+H]+
requires 552.3122.
1.2.7 N-(2-Aminoethyl)-2,3,4-tri-O-benzyl-1,5-dideoxy-1,5-imino-L-fucitol (14f)
Compound 14f was prepared from bis-mesylate (130 mg, 0.22 mmol) as described in
the preparation of 14a, yielding 14f (82 mg, 81%) as white foam after column
1
chromatography (cyclohexane/EA 2:1); [α]20 -0.8 (c 2.3, CHCl3); H NMR (600
D
MHz, CDCl3) δ 7.38 – 7.25 (m, 15H, HAr), 4.89 (d, J = 11.4, 1H, 1 x PhCHH), 4.62 –
4.81 (m, 5H, 5 x PhCHH), 3.96 (m, 1H, H2), 3.70 (m, 1H, H4), 3.48 (dd, J = 7.2, 1.5,
1H, H3), 3.08 (dd, J = 10.5, 3.5, 1H, H1-eq), 2.80 – 2.66 (m, 3H, NCHHCH2 &
CH2NH2), 2.55 (m, 1H, H5), 2.42 (t, J = 6.0, 1H, NCHHCH2), 2.21 (t, J = 10.5, 1H,
13
H1-ax), 1.09 (d, J = 6.5, 3H, CH3). C NMR (150 MHz, CDCl3) δ 139.03 (C), 138.92
(C), 138.81(C), 128.34, 128.32, 128.26, 128.18, 127.73, 127.53, 127.50, 127.46,
127.42 (each CH), 83.16 (C3), 78.87 (C4), 75.86 (C2), 74.20 (PhCH2), 73.04 (PhCH2),
72.67 (PhCH2), 57.96 (C5), 55.51(CH2), 53.12 (C1), 38.13 (CH2), 14.66 (CH3). IR
(film, CHCl3): νmax 3360, 2915, 1364, 1095 cm-1. ES-HRMS: Found 461.2802,
C29H37N2O3 [M+H]+ requires 461.2804.
1.2.8 N-Allyl- 2,3,4-tri-O-benzyl-1,5-dideoxy-1,5-imino-L-fucitol (14g)
Compound 14g was prepared from bis-mesylate (90 mg, 0.15 mmol) as described in
the preparation of 14a, yielding 14g (59 mg, 85%) as colorless oil after column
1
chromatography (cyclohexane/EA 6:1); Rf 0.52 (cyclohexane-EtOAc, 2:1); H NMR
(500 MHz, CDCl3) δ 7.56 – 7.09 (m, 15H, HAr), 5.97 – 5.77 (m, 1H, NCH2CH=CH2),
5.19 – 5.08 (m, 2H, NCH2CH=CH2), 4.92 (d, J = 11.8 Hz, 1H, 1 x PhCHH), 4.85 –
4.61 (m, 5H, 5 x PhCHH), 4.10 – 3.94 (m, 1H, H2), 3.66 (s, 1H, H4), 3.42 (dd, J = 7.6,
1.0 Hz, 1H, H3), 3.26 (dd, J = 14.3, 6.0 Hz, 1H, NCHHCH=CH2), 3.16 – 3.02 (m, 2H,
H1-eq & NCHHCH=CH2), 2.41 (m, 1H, H5), 2.19 (t, J = 10.4 Hz, 1H, H1-ax), 1.03 (d, J
= 6.4 Hz, 3H, CH3). 13C NMR (125 MHz, CDCl3) δ 139.17, 139.00, 138.98 (each C),
133.64 (CH2CH=CH2), 128.92, 128.46, 128.43, 128.23, 127.89, 127.65, 127.62,
127.60, 127.56 (each CH), 118.38 (CH2CH=CH2), 84.24 (C3), 78.64 (C4), 76.26 (C2),
74.56 (PhCH2), 73.24(PhCH2), 72.87(PhCH2), 57.50 (C5), 56.47(NCH2CH), 54.39

(C1), 15.44 (CH3, C6). ES-HRMS: Found 458.2702, C30H36NO3 [M+H]+ requires
458.2695.
1.2.9 N-((Adamantylcarbonylamino)-ethyl)-2,3,4-tri-O-benzyl-1,5-dideoxy-1,5-
imino-L-fucitol (15)
The 1-adamantanecarboxylic acid (5.6 mg, 0.031mmol) was dissolved in dry CH2Cl2,
cooled to 0 °C and EDC (9.2 mg, 0.048 mmol), HOBt (6.3 mg, 0.048 mmol), and
DMAP (5.7 mg, 0.048 mmol) were added. After stirring for 1hr, the amine 14f (11mg,
0.024 mmol) was added. The reaction was warmed to room temp and stirred for 24 h.
The mixture was then quenched with NH4Cl (sat.) and extracted with CH2Cl2. The
organic fractions were washed with satd. NaHCO3, brine and dried with Na2SO4. The
crude product was purified by flash chromatography on silica gel eluting with
cyclohexane-EtOAc(1:1) to yield 15 (13 mg, 88%); Rf 0.18 (cyclohexane-EA, 1:1);
[α]20 -4.4 (c 1.5, CHCl3); 1H NMR (500 MHz, CDCl3) δ 7.41 – 7.21 (m, 15H, HAr),
D
4.85 (d, J = 11.5 Hz, 1H, 1 x PhCHH), 6.32 (s, 1H, NH), 4.80 – 4.71 (m, 2H, 2 x
PhCHH), 4.65 (m, 3H, 3 x PhCHH), 3.86 (s, 1H, H2), 3.72 (s, 1H, H4), 3.54 (d, J = 5.7
Hz, 1H, H3), 3.26 (d, J = 4.9 Hz, 2H, NCH2CH2), 3.06 (dd, J = 11.8, 3.4 Hz, 1H, H1-
eq), 2.71 (m, 2H, 1 x NCHHCH2 & H5), 2.52 – 2.39 (m, 1H, NCHHCH2), 2.24 (m, 1H,
H1-ax), 1.93 (s, 3H, 3 x CH), 1.77 (d, J = 2.0 Hz, 6H, 3 x CH2), 1.67 (d, J = 12.0 Hz,
3H, 3 x CHH), 1.60 (d, J = 12.0 Hz, 3H, 3 x CHH), 1.12 (d, J = 6.5 Hz, 3H, CH3). 13C
NMR (125 MHz, CDCl3) δ 178.12 (C=O), 139.13, 139.11, 138.81 (each C), 128.50,
128.46, 128.32, 127.98, 127.72, 127.58 (each CH), 84.18 (C3), 78.69 (C4), 75.88 (C2),
73.94 (PhCH2), 73.22 (PhCH2), 72.59 (PhCH2), 57.70 (C5), 51.05 (NCH2CH2), 40.70
(C), 39.48, 39.38, 36.70, 36.62, 35.97, 28.29 (each CH2), 28.25 (CH), 15.40 (CH3).
HRMS: Calcd for C40H50N2O4, [M+H]+, 623.3849; Found, 623.3839.
1.2.10 1,5-Dideoxy-1,5-imino-L-fucitol (2)
A mixture of 14a (23mg, 0.045 mmol) in MeOH (3mL) was added 0.2 mL 6N HCl
and 10% Pd-C (5 mg). The mixture was stirred for 24 h under H2 atmosphere. The
catalyst was then removed by filtration through Celite, and the filtrate was
concentrated. The residue was dissolved in 1.0 M methanolic solution of HCl (1.0
mL) and the solvent was evaporated. The residue, dissolved in H2O, was introduced to
a Dowex-50W-X8-100 (H+) packed ion-exchange column (pre-washed with MeOH
and H2O), washed with water (100 mL) and MeOH (50 mL) and then the product was
eluted with 1.0 M aqueous solution of NH4OH. The combined pure fractions were
evaporated in vacuo to give 2 (6.5 mg, 91%) as a white solid after lyophilization. The
NMR data of the free amine 2 were in excellent agreement to those reported in
literature;28,35 [α]20 -42.4 (c 0.48, MeOH); 1H NMR (500 MHz, D2O) δ 3.70 – 3.51
D
(m, 2H, H4 & H2), 3.30 (d, J = 9.4 Hz, 1H, H3), 2.97 (d, J = 8.4 Hz, 1H, H1-eq), 2.78
(m, 1H, H5), 2.30 (t, J = 11.4 Hz, 1H, H1-ax), 0.94 (d, J = 6.6 Hz, 3H, CH3). 13C NMR
(125 MHz, D2O) δ 75.65 (C3), 72.99 (C4), 68.06 (C2), 54.45 (C5), 49.19 (C1), 16.70
(C6). HRMS: Calcd for C6H14NO3, [M+H]+, 148.0974; Found, 148.0971.
1.2.11 N-(4-Hydroxybutyl)- 1,5-dideoxy-l,5-imino-L-fucitol (22)

A mixture of 14a (23.0 mg, 0.045 mmol) and 10% Pd-C (10.0 mg) in HOAc (0.25
mL), H2O (0.5 mL), and THF (1.0 mL) was stirred for 24 h under H2 atmosphere. The
catalyst was then removed by filtration through Celite, and the filtrate was
concentrated. The residue was dissolved in 1.0 M methanolic solution of HCl (1.0
mL) and the solvent was evaporated. The residue, dissolved in H2O, was introduced to
a Dowex-50W-X8-100 (H+) packed ion-exchange column (pre-washed with MeOH
and H2O), washed with water (100 mL) and MeOH (50 mL) and then the product was
eluted with 1.0 M aqueous solution of NH4OH. The combined pure fractions were
evaporated in vacuo to give 22 (8.5 mg, 85%) as a white powder after lyophilization.
1H NMR (500 MHz, CD3OD) δ 3.88 (td, J = 10.0, 4.9 Hz, 1H, H2), 3.73 (d, J = 1.5
Hz, 1H, H4), 3.70 – 3.57 (m, 2H, CH2OH), 3.33 (dd, J = 9.3, 3.3 Hz, 1H, H3), 3.14
(dd, J = 11.3, 4.9 Hz, 1H, H1-eq), 2.91 – 2.78 (m, 1H, 1x NCHHCH2), 2.70 – 2.53 (m,
2H, 1x NCHHCH2 & H5), 2.25 (t, J = 10.7 Hz, 1H, H1-ax), 1.71 – 1.62 (m, 2H, 1x
NCH2CH2), 1.62 – 1.54 (m, 2H, CH2), 1.30 (d, J = 6.6 Hz, 3H, CH3). 13C NMR (125
MHz, CD3OD) δ 77.06 (C3), 74.59 (C4), 68.73 (C2), 62.65 (CH2OH), 60.20 (C5),
57.33 (C1), 53.79 (NCH2), 31.53 (CH2), 21.81 (CH2), 16.05 (CH3, C6). HRMS: Calcd
for C10H22NO4, [M+H]+, 220.1549; Found, 220.1544.
1.2.12 N-Decyl-1,5-Dideoxy-1,5-imino-L-fucitol (16)
Compound 16 was prepared from 14b as described in the preparation of 2, yielding 16
(90%) as a solid after lyophilization; 1H NMR (500 MHz, CD3OD) δ 3.80 (td, J = 9.4,
4.9 Hz, 1H, H2), 3.65 (dd, J = 3.3, 1.4 Hz, 1H, H4), 3.25 (dd, J = 9.4, 3.3 Hz, 1H, H3),
3.01 (dd, J = 11.2, 4.9 Hz, 1H, H1-eq), 2.77 – 2.63 (m, 1H, NCHH), 2.57 – 2.39 (m,
2H, H5 & NCHH), 2.14 (t, J = 10.8 Hz, 1H, H1-ax), 1.49 (m, 2H, CH2), 1.39 – 1.24 (m,
13
14H, 7 x CH2), 1.21 (d, J = 6.6 Hz, CH3, H6), 0.91 (t, J = 7.0 Hz, 3H, CHCH3). C
NMR (125 MHz, CD3OD) δ 77.31(C3), 74.84 (C4), 69.02 (C2), 59.77 (C5), 57.74 (C1),
53.96 (NCH2CH2), 33.03, 30.70, 30.66, 30.60, 30.42, 28.58, 24.65, 23.70 (each CH2),
16.32(C6), 14.41(CH3). HRMS: Calcd for C16H34NO3, [M+H]+, 288.2539; Found,
288.2539.
1.2.13 N-Octyl-1,5-dideoxy-1,5-imino-L-fucitol (17)
Compound 17 was prepared from 14c as described in the preparation of 2, yielding 17
(85%) as a solid after lyophilization; 1H NMR (500 MHz, CD3OD) δ 3.83 (td, J = 9.9,
4.9 Hz, 1H, H2), 3.69 (d, J = 3.0 Hz,1H, H4), 3.29 (dd, J = 9.3, 3.0 Hz, 1H, H3), 3.07
(dd, J = 11.0, 4.8 Hz, 1H, H1-eq), 2.85 – 2.71 (m, 1H, NCHH), 2.70 – 2.52 (m, 2H, H5
& NCHH), 2.26 (t, J = 10.5 Hz, 1H, H1-ax), 1.53 (m, 2H), 1.32 (m, 10H), 1.24 (d, J =
13
6.6 Hz, 3H, CH3), 0.91 (t, J = 6.9 Hz, 3H, CH3). C NMR (125 MHz, CD3OD) δ
76.92 (C3), 74.42 (C4), 68.62 (C2), 60.11 (C5), 57.21 (C1), 53.97 (NCH2), 32.95,
30.51, 30.35, 28.45, 24.46, 23.68 (each CH2), 15.95 (C6), 14.39 (CH3). HRMS: Calcd
for C14H30NO3, [M+H]+, 260.2226; Found, 260.2232.
1.2.14 N-Butyl-1,5-dideoxy-1,5-imino-L-fucitol (18)
Compound 18 was prepared from 14d as described in the preparation of 2, yielding 18
1
(88 %) as a solid after lyophilization; H NMR (500 MHz, D2O) δ 4.13- 4.18(m, 1H,

H2), 4.11(s, 1H, H4), 3.72 (d, J = 9.5, 1H, H3), 3.52-3.60 (m, 2H, H1-eq & H5), 3.30 (dt,
J = 12.0, 5.5 Hz, 1H, NCHH), 3.20 (dt, J = 12.0, 5.5Hz, 1H, NCHH) 3.03 (t, J = 12.0
Hz, 1H, H1-ax) 1.73-1.81 (m, 2H, CH2), 1.48 (d, J = 7.0, 3H, CH3), 1.42-1.49 (m, 2H,
CH2), 1.02 (t, J = 7.5, 3H, CH3); 13C NMR (125 MHz, D2O) δ 75.48 (C3), 73.71 (C4),
67.40 (C2), 63.15 (C5), 55.90 (C1), 55.14 (NCH2), 26.09 (CH2), 21.86 (CH2), 15.99
(CH3, C6), 15.33 (CH3). HRMS: Calcd for C10H21NO3, [M+H]+, 204.1600; Found,
204.1604.
1.2.15 N-(2-Hydroxyethyl)-1,5-dideoxy-l,5-imino-L-fucitol (19)
Compound 19 was prepared from 14e as described in the preparation of 2, yielding 19
(78 %) as a solid after lyophilization; 1H NMR (500 MHz, D2O) δ 3.85 (m, 4H, H2 &
H4 & CH2OH), 3.51 (d, J = 9.7 Hz, 1H, H3), 3.28 – 3.15 (m, 1H, H1-eq), 3.07 – 2.93
(m, 1H, NCHHCH2), 2.82 – 2.63 (m, 2H, H5 & NCHHCH2), 2.36 (t, J = 10.7 Hz, 1H,
13
H1-ax), 1.13 (d, J = 6.4 Hz, 3H, CH3). C NMR (125 MHz, D2O) δ 74.91(C3), 73.13
(C4), 66.94 (C2), 58.49 (C5), 57.51 (CH2OH), 56.07 (C1), 53.16 (NCH2), 15.24 (C6).
HRMS: Calcd for C8H8NO4, [M+H]+, 192.1236; Found, 192.1228.
1.2.16 N-Adamantylcarbonylaminoethyl-1,5-Dideoxy-1,5-imino-L-fucitol (20)
Compound 20 was prepared from 15 as described in the preparation of 2, yielding 20
1
(83 %) as solids after lyophilization; H NMR (500 MHz, CD3OD) δ 3.98 (ddd, J =
14.2, 9.5, 4.4 Hz, 1H, H2), 3.85 (m, 1H, H4), 3.49 (dd, J = 10.5, 6.0 Hz, 2H, NCH2),
3.45 (dd, J = 9.1, 2.4 Hz, 1H, H3), 3.39 (dd, J = 11.9, 4.9 Hz, 1H, H1-eq), 3.29 – 3.22
(m, 1H, H5), 3.22 – 3.17 (m, 1H, CHHNH), 3.08 – 3.00 (m, 1H, CHHNH), 2.76 (t, J =
11.2 Hz, 1H, H1-ax), 2.04 (s, 3H, 3 x CHCH2), 1.88 (d, J = 2.4 Hz, 6H, 3 x CHCH2),
13
1.78 (m, 6H, 3x CHCH2), 1.39 (d, J = 6.6 Hz, 3H, CH3). C NMR (125 MHz,
CD3OD) δ 180.88 (C=O), 74.11 (C3), 71.60 (C4), 65.74(C2), 60.60 (C5), 55.23(C1),
52.55 (CH2), 40.42(C), 38.85, 38.70, 36.18, 36.11, 34.68 (each CH2), 28.17(CH),
13.43 (C6, CH3); HRMS: Calcd for C19H33N2O4, [M+H]+, 353.2440; Found, 353.2451.
1.2.17 1,5,6-Trideoxy-1,5-imino-D-altritol (5)
Compound 5 was prepared from 9 as described in the preparation of 2, yielding 5
(93%) or prepared from 11 in 83% yield, as solids after lyophilization. [α]20 0.9 (c
D
0.55, MeOH) [lit. [α]20 2 (c 1.1, MeOH)]; 1H NMR (500 MHz, D2O) δ 3.93 – 3.78
D
(m, 2H, H2 & H3), 3.55 (d, J = 8.9 Hz, 1H, H4), 2.93 (m, 2H, H1-eq & H5), 2.76 (dd, J
= 14.0, 2.4 Hz, 1H, H1-ax), 1.35 (d, J = 6.6, 3H, CH3). 13C NMR (125 MHz, D2O) δ
70.82 (C3), 69.96 (C4), 68.67 (C2), 50.49 (C5), 44.23 (C1), 15.86 (C6). HRMS: Calcd
for C6H14NO3, [M+H]+, 148.0974; Found, 148.0967.
1.3 Fucosidase Inhibition Assays
The substrate used for the enzyme assays was 4-nitrophenyl α-L-fucopyranoside.
Substrate solutions were made from a stock solution (10 mM) in 10 mL batches, as
required, by dissolving the appropriate mass of substrate in the correct buffer solution

for the enzyme. These were kept at 4 °C when not in use. Assays were carried out in
triplicate, using water as a blank in place of the inhibitor. The concentrations of the
enzyme were adjusted so that the reading for the final absorbance was in the range of
1.0-2.0 units over a reaction time of 20 min. Linearity over the time course of the
reaction was checked using a series of incubation times. The following were
combined in the well of a flat-bottomed 96-well (300 µL) microtitre plate:10 µL of
enzyme solution; 10 µL inhibitor solution; 80 µL substrate solution. The reaction mix
was incubated at 25 °C for 20 min and was quenched by the addition of 100 µL of 2
M Na2CO3. Absorbance at 405 nm was measured immediately using a microtitre plate
reader.
Percentage inhibition was plotted against the log of the inhibitor
concentration and a trendline was fitted using Microsoft Excel. The IC50 value for
each compound calculated from the value of the log of the inhibitor concentration at
50% inhibition of enzyme activity.
1.4 Molecular modelling
The PDB for fucosidase of T. maritima was available from the RCSB-PDB, while for
the fucosidase of B. taurus, a homology model was constructed. The sequence for the
B. taurus fucosidase (Accession: AAI12589.1 GI: 86827599) was obtained from
NCBI (National Center for Biotechnology Information). Next a BlastP search along
with the BLOSSUM62 substitution matrix, generated putative template PDBs with
the minimum identity of 35% and 87% sequence coverage, provided in Table S1, in
supplementary information. Later, homology modelling on templates having more
than 35 % identity and at least 85% of coverage of the query sequence was conducted.
The software programme MOE 2018.01 implementing the Amber 99 forcefield was
used [66]. After generated homology models were assessed. Initially, the generated
homology model and the template were superimposed giving an RMSD difference,
which was less than 1 Å [67]. See Fig. S1 in the supplementary information. The Z-
score for the templates and homology models, was assessed using ProSA, a protein
structure analysis tool, respectively ([68]. See Fig. S2 in supplementary information.
Next, the Ramachandran plot was generated. In the Ramachandran plot, the core or
the allowed regions are those satisfying the ??/?? angle pairs for residues in a protein.
See Fig. S4 in supplementary information [69, 70]. For those amino acids which did
not comply with the requirements of the Ramachandran plot, further energy
minimization was done. ERRAT plots, which are for the determination of errors in
model building were performed. See Fig. S3 in the supplementary information) [71].
Verify 3D was used to determine the compatibility of the constructed homology
model (3D) with its own amino acid sequence (1D) by assigning a structural class
based on its location and environment (alpha, beta, loop, polar, nonpolar etc); a score
of >80% was considered acceptable (Figure 4) [72]. The active site for docking was
identified with the active site finder tool of MOE and later, the docked pose of the
largest ensembled cluster within a RMSD of 2 Å, was selected [73]. The triangle
matcher placement method and GBVI/WSA dG scoring was used for the docking.
Molecular dynamics (MD) was performed on the selected docked poses of
compounds, 2 and 16 [58]. This analysis was carried out also using MOE 2018. 01
software [66]. Partial charges were calculated and energy minimizations were done.

MD was carried for a duration of 100 ps. The adopted docking procedure was
validated using the co-crystallized structure of iso-6FNJ binding to fucosidase was
obtained from the PDB (2ZXD). After docking iso-6FNJ to the fucosidase, an RMSD
value of 0.391 Å between the co-crystallized structure and the docked structure
indicated that the adopted docking procedure was reliable. [74].
1.5 Cell Culture and Cytotoxicity Evaluation
1.5.1 Materials and Cell Lines
The MCF-7 breast cancer and DU-145 prostate cancer cell lines were obtained
from Dr. Stephen Rea, National University of Ireland, Galway. The BT-474 breast
cancer cell line was obtained from Dr. Helen Dodson, National University of Ireland,
Galway.
MCF-7 were cultured in Dulbecco’s modified Eagle’s medium (DMEM) containing
high glucose (4.5 g/mL) and supplemented with 1% penicillin-streptomycin and 10%
heat-inactivated fetal bovine serum (FBS). DU-145 were cultured in RPMI-1640
medium and supplemented with 1% 2 mM L-glutamine, 1% penicillin-streptomycin
and 10% non heat-inactivated fetal bovine serum (FBS). BT-474 were cultured in
Dulbecco’s modified Eagle’s medium (DMEM)/Nutrient mixture F-12 Ham and
supplemented with 1% penicillin-streptomycin and 10% non heat-inactivated fetal
bovine serum (FBS). All cells grew as adherent cultures. Cell culture reagents were
obtained from Sigma-Aldrich. Disposable sterile plasticware was obtained from
Sarstedt (Numbrecht, Germany).
1.5.2 Cytotoxicity measurements using the MTT assay
The MTT colorimetric assay was used to determine cell viability. Cells were
added to 96-well plates at a cell density of 1000 cells per well (MCF-7, 200 µL per
well), 2000 cells per well (DU-145, 200 µL per well) and 3000 cells per well (BT-
474, 200 µL per well) and allowed to adhere over 24 hours. Compound solutions in
DMSO were added after 24 hours (1% v/v final concentration in well). The control
cells were exposed to the same concentration of the vehicle control alone (DMSO).
All cells were incubated at 37 °C and 5% CO2 (humidified atmosphere) for 72 hours.
MTT (20 µL, 5 mg/mL solution) was added after 72 hours and the cells were
incubated for a further 3 hours. The supernatant was then removed using a multi-
transfer pipette and DMSO (100 µL) added to dissolve the MTT formazan crystals.
The absorbance was determined using a plate reader at 550 nm with a reference at 690
nm. Cell viability is expressed as a percentage of the vehicle-only treated control
(DMSO). Dose-response curves were analyzed by non-linear regression analysis and
IC50 values were determined using GraphPad Prism software, v 8.0 (GraphPad Inc.,
San Diego, CA, USA). The in vitro activity of the drugs towards all cell lines is
expressed as IC50 (i.e. concentration required for the reduction of the mean cell
viability to 50%).
1.6 Tumor cell proliferation assay using HNO41, HNO97, HNO210 & NCH644
cell lines
To further assess the influence of 16 on tumor cell proliferation well-characterized
human head and neck carcinoma cell lines (HNO41, HNO97, HNO210; [75-77] as
well as glioblastoma-initiating cells (NCH644, [63]) were incubated with increasing

concentrations (0.015µM - 45µM) of 16 for 72h. Cells were grown either in media
containing 10% fetal calf serum (HNO41, HNO97, HNO210) or under serum free
conditions as described before [72-74]. Each concentration was measured in four
technical replicates with the CellTiter-Glo® 3D Cell Viability Assay. After adding the
CellTiter-Glo substrate the plate was shaken for 5 min and afterwards incubated for
30 min to get a stable luminescence signal. This was measured as relative light units
by a Tecan Infinite F200 Pro. IC50 was determined by a sigmoidal fit in version 5 of
the GraphPad Prism software.
Acknowledgement
Research presented herein was supported by Science Foundation Ireland
(07/IN.1/B966), the Irish Research Council (GOI PhD Scholarships to JZ, AN) and
the European Commission through COST Action CM1106. The authors thank Guan-
Nan Wang for recording NMR spectra and to Nadine Martinet for organization of the
COST Action CM1106 screening library that led to identification of the activity for
16. The authors thank Dr. Stephen Rea and Dr. Helen Dodson for gifts of tumour
cells.
References
[1] C.H. Wong, D.P. Dumas, Y. Ichikawa, K. Koseki, S.J. Danishefsky, B.W. Weston, J.B. Lowe,
Specificity, Inhibition, and Synthetic Utility of a Recombinant Human Alpha-1,3-Fucosyl-Transferase,
Journal of the American Chemical Society, 114 (1992) 7321-7322.
[2] P. Compain, O.R. Martin, Design, synthesis and biological evaluation of iminosugar-based
glycosyltransferase inhibitors, Current Topics in Medicinal Chemistry, 3 (2003) 541-560.
[3] L. Somsak, V. Nagy, Z. Hadady, T. Docsa, P. Gergely, Glucose analog inhibitors of glycogen
phosphorylases as potential antidiabetic agents: Recent developments, Current Pharmaceutical Design,
9 (2003) 1177-1189.
[4] V.L. Schramm, P.C. Tyler, Imino-sugar-based nucleosides, Current Topics in Medicinal Chemistry,
3 (2003) 525-540.
[5] R.E. Lee, M.D. Smith, L. Pickering, G.W.J. Fleet, An approach to combinatorial library generation
of galactofuranose mimics as potential inhibitors of mycobacterial cell wall biosynthesis: Synthesis of a
peptidomimetic of uridine 5 '-diphosphogalactofuranose (UDP-Galf), Tetrahedron Letters, 40 (1999)
8689-8692.
[6] R.E. Lee, M.D. Smith, R.J. Nash, R.C. Griffiths, M. McNeil, R.K. Grewal, W.X. Yan, G.S. Besra,
P.J. Brennan, G.W.J. Fleet, Inhibition of UDP-Gal mutase and mycobacterial galactan biosynthesis by
pyrrolidine analogues of galactofuranose, Tetrahedron Letters, 38 (1997) 6733-6736.
[7] H. Moriyama, T. Tsukida, Y. Inoue, K. Yokota, K. Yoshino, H. Kondo, N. Miura, S. Nishimura,
Azasugar-based MMP/ADAM inhibitors as antipsoriatic agents, Journal of Medicinal Chemistry, 47
(2004) 1930-1938.
[8] A.E. Stütz, Iminosugars as Glycosidase Inhibitors: Nojirimycin and Beyond, (1999) Wiley-VCH:
Weinheim.
[9] P. Sears, C.H. Wong, Carbohydrate mimetics: A new strategy for tackling the problem of
carbohydrate-mediated biological recognition, Angewandte Chemie-International Edition, 38 (1999)
2301-2324.
[10] T.M. Gloster, G.J. Davies, Glycosidase inhibition: assessing mimicry of the transition state,
Organic & Biomolecular Chemistry, 8 (2010) 305-320.
[11] V.H. Lillelund, H.H. Jensen, X.F. Liang, M. Bols, Recent developments of transition-state
analogue glycosidase inhibitors of non-natural product origin, Chemical Reviews, 102 (2002) 515-553.
[12] O.R. Martin, P. Compain, Iminosugars: Recent insights into their bioactivity and potential as
therapeutic agents - Preface, Current Topics in Medicinal Chemistry, 3 (2003) -.
[13] D.J. Becker, J.B. Lowe, Fucose: biosynthesis and biological function in mammals, Glycobiology,
13 (2003) 41r-53r.