Reprint of "effect of carbohydrate amino group modifications on the cytotoxicity of glycosylated 2-phenyl-benzo[b]thiophenes and 2-phenyl-benzo[b]furans" [Bioorg. Med. Chem. Lett. 21 (2011) 2591-2596]

Article abstract of DOI:10.1016/S0960-894X(11)01095-X

In previous studies, we have identified a family of benzo[b]furan and benzo[b]thiophene derivatives linked to amino sugars (1-6) that are cytotoxic to a range of cancer cell lines. We describe here an exploration of the effect of structural modification o

Full text of DOI:10.1016/S0960-894X(11)01095-X

Bioorganic & Medicinal Chemistry Letters 21 (2011) 5107–5112
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Reprint of ‘‘Effect of carbohydrate amino group modifications on the
cytotoxicity of glycosylated 2-phenyl-benzo[b]thiophenes and
2-phenyl-benzo[b]furans’’ [Bioorg. Med. Chem. Lett. 21 (2011) 2591–2596] q
⇑
Wei Shi, Todd L. Lowary
Alberta Ingenuity Centre for Carbohydrate Science and Department of Chemistry, The University of Alberta, Gunning-Lemieux Chemistry Centre, Edmonton, Canada AB T6G 2G2
a r t i c l e i n f o
a b s t r a c t
Article history:
In previous studies, we have identified a family of benzo[b]furan and benzo[b]thiophene derivatives
linked to amino sugars (1–6) that are cytotoxic to a range of cancer cell lines. We describe here an explo-
ration of the effect of structural modification of the amino group on one of the carbohydrate residues
Received 9 January 2011
Revised 5 February 2011
Accepted 14 February 2011
Available online 17 February 2011
(4-amino-2,3,4,6-tetradeoxy-a-L-threo-hexopyranoside) on in vitro cytotoxicity. It has been found that
maintaining at least one basic functional group around the C-4 position in the carbohydrate moiety is
crucial for cytotoxicity. Furthermore, it appears that modifications around the C-4 position are limited
by suitable hydrophilic/hydrophobic and/or ionic interactions, as well as steric constraints.
Ó 2011 Elsevier Ltd. All rights reserved.
Keywords:
Structure–activity relationship
Amino group
Cytotoxicity
Previous reports from our laboratory have detailed the design,
synthesis and cytotoxicity of novel glycosylated benzo[b]furan
and benzo[b]thiophene derivatives (e.g., 1–6, Chart 1).1–4 These
compounds are cytotoxic to mammalian cells,3 and also possess
weak bacteriocidal activity against gram-positive bacteria.2 The
mode of action of these compounds remains unclear. However,
some of the active compounds have been demonstrated to bind
to DNA1 and also inhibit topoisomerase I and topoisomerase II,3
albeit at relatively high concentration.
Encouraged by low micromolar in vitro cytotoxicity of these
compounds, and their relatively simple structure and ease of syn-
thesis,1–3 we endeavoured to identify key structural features in
these molecules. We envisioned that knowledge of the structural
motifs essential for cytotoxicity would benefit future efforts to-
wards the preparation of more potent analogues, and in turn would
facilitate the discovery of their molecular targets. Our first struc-
ture–activity relationship (SAR) study on analogs containing the
toxicity. In addition, the alkyne linker was only slightly preferred
to the alkane. These studies have also confirmed the crucial role
of the amino group at C-4 for cytotoxicity. Reported here is an
investigation of how structural modifications of the amino group
influence the cytotoxicity of these compounds.
In view of the in vitro anticancer activity of the compounds pre-
viously described,1–3 the benzo[b]furan system was employed in
initial studies. To avoid potential bioactivity loss due to steric
hindrance, small acyl and alkyl groups were first installed for a pre-
liminary study (Scheme 1). To explore the effect of N-acylation, the
acetylated compound 7, and a derivative (9) in which the amine
was N-acylated with L-alanine were prepared from 5 in good yield.
The purpose of preparing 9 was to keep a free amino group in the
molecule while the amine on the carbohydrate ring was masked.
Probing the effect of N-alkylation involved first the synthesis of
the ethylamino and isopropylamino analogues 10 and 11, via
reductive amination with either acetaldehyde or acetone, respec-
tively. In this one-pot process, the condensation of the amino
group with the aldehyde or ketone was followed by in situ reduc-
tion of the intermediate imines with sodium cyanoborohydride
(NaCNBH3). Under these conditions it was possible to generate
the desired products in a yield of 85% for 10 and 82% for 11.
Next, we assayed 5, 7, and 9–11 against three cancer cell lines—
MCF-7 breast cancer, HT29 colon cancer, and HepG2/C3A liver can-
cer (Table 1). These results demonstrated that acylation (7) was
detrimental to cytotoxicity, presumably due to the loss of a basic
nitrogen. This was supported by the observation that 9, which con-
tained both an amide bond and additional basic nitrogen atom, was
cytotoxic at a level comparable to 5. Although 10 and 11 showed
4-amino-2,3,4,6-tetradeoxy-a-L-threo-hexopyranoside (4-N-TDTH)
moiety (e.g., 5 and 6)4 demonstrated that the orientation of the
substituents at C-1 and C-4 appeared not to be important for cyto-
q
A publisher’s error resulted in this article appearing in the wrong issue. The
article is reprinted here for the reader’s convenience and for the continuity of this
special issue. Anyone wishing to cite this article should use the details of the
original publication [Shi, W. Lowary, T.L., Effect of carbohydrate amino group
modifications on the cytotoxicity of glycosylated 2-phenyl-benzo[b]thiophenes and
2-phenyl-benzo[b]furans. Bioorg. Med. Chem. Lett. 21, (2011), 2591-2596, doi:
10.1016/j.bmcl.2011.02.051].
⇑
Corresponding author.
E-mail address: tlowary@ualberta.ca (T.L. Lowary).
doi:10.1016/S0960-894X(11)01095-X

5108
W. Shi, T. L. Lowary / Bioorg. Med. Chem. Lett. 21 (2011) 5107–5112
O
O
O
O
O
O
X
X
X
HO
NH2
HO
NH2
NH2
1 X = O
2 X = S
3 X = O
4 X = S
5 X = O
6 X = S
Chart 1. Structures of glycosylated bezno[b]furan and benzo[b]thiophene analogues 1–6.
O
O
O
a/b
d/e
O
O
O
O
O
O
NHR
NH2
NHR
10 R: ethyl
11 R: isopropyl
5
7 R: acetyl
8 R: Fmoc-L-alanine
9 R: alanine
F
c
F
F
F
O
O
HN
F
Fmoc
12
Scheme 1. Synthesis of 7–11 from 5. Reagents and conditions: (a) acetyl chloride, Et3N, 86%; (b) 12, CH2Cl2; (c) piperidine, CH2Cl2, 55% over two steps; (d) acetaldehyde,
NaCNBH3, CH3OH, 85% for 10; (e) acetone, NaCNBH3, CH3OH, 82% for 11.
Table 1
Cytotoxicity of compounds 5, 7, and 9–11
Compound
IC50
(l
M)
O
O
HepG2a
a
MCF-7
HT29
O
O
O
O
5
7
9
10
11
10.3 0.2
23.4 2.3
8.0 1.0
9.2 1.1
12.6 1.0
6.5 0.5
13.3 1.2
6.8 0.9
5.8 0.8
5.8 0.6
7.9 0.7
18.2 1.9
4.1 0.7
6.0 1.0
13.0 1.2
Aldehydes
NH2
NR1R2
11
Mono: R1 = H, R2 = R
Di: R1 = R = R
2
10
a
HepG2/C3A is simplified as HepG2.
Scheme 2. Monoalkylation and dialkylation of compound 5 via reductive amina-
tion. Reagents and conditions: (a) NaCNBH3, CH3OH or DMF containing 1% acetic
acid. See Supplementary data for the experimental details.
similar cytotoxicity against HT29 cells, the former showed better
cytotoxicity for the other two cancer cell lines. Based on the results
in Table 1, we proceeded with reductive amination of 5 with alde-
hydes to enrich the diversity of our compound library for more
reliable SAR interpretation.
However, not all monoalkylated or dialkylated analogues could
be efficiently made under these conditions. For example, it was dif-
ficult to control monomethylation using formaldehyde. Once the
monomethylated product was formed, it reacted with another
molecule of formaldehyde at a rate faster than the starting mate-
rial. Attempts to quench the reaction before completion failed
and the major component was either 5, or the dimethylated prod-
uct 12. The very similar Rf values, as well as the high polarity, of the
free, monomethylated, and dimethylated amino compounds made
the purification impossible. Use of the stepwise process was also
unsuccessful.
We also found that for most of the aromatic aldehydes, it was
difficult to obtain the dialkylated products, even in DMF containing
1% acetic acid, conditions that favored dialkylation. For some bulky
aromatic aldehydes, monoalkylation of 5 was also difficult. This is
presumably due to the axial orientation of the amino group on the
carbohydrate ring, which would be expected to react slowly with
bulky aldehydes. In cases where dialkylation was successful, the
reaction required substantially longer reaction times. These diffi-
culties, coupled with the fact that the majority of the dialkylated
Reductive amination of aldehydes and ketones is a general
approach for the alkylation of amines.5–8 There are two general
procedures: one-pot (direct)5,7,8 or stepwise (indirect).5,6 The latter
method is usually employed for synthesizing monoalkylated com-
pounds only when the former method fails. A panel of aliphatic and
aromatic aldehydes was chosen to make the N-alkyl amino ana-
logues from 5 (Scheme 2). We found that the one-pot approach,
when carried out in two different solvent systems, can be used to
prepare either monoalkylated or dialkylated analogues, depending
on the aldehyde substrates. In most cases, treatment of 5 with the
aldehyde and sodium cyanoborohydride in methanol led to mono-
alkylation. Under these reaction conditions, dialkylation occurred
mostly with linear or branched aliphatic aldehydes. On the other
hand, by switching the solvent to DMF containing 1% acetic acid,
dialkylated analogues could be prepared from some aromatic or
cyclic aliphatic aldehydes. The alkylated derivatives 12–28 made
under either of the above two conditions are listed in Tables 2
and 3.

W. Shi, T. L. Lowary / Bioorg. Med. Chem. Lett. 21 (2011) 5107–5112
5109
Table 2
Cytotoxicity of the alkylated analogues from compound 5 and aliphatic aldehydesa
O
O
O
NR1R2
a
Compound
R1
H
R2
H
IC50
(lM)
MCF-7
HT29
HepG2b
5
10.3 0.2
10.3 0.8
6.5 1.8
6.1 0.6
7.9 0.7
6.8 0.8
12
10
13
14
15
H
H
9.2 0.9
9.7 1.1
20.8 2.0
>25
5.8 1.0
6.8 0.4
9.8 1.0
>25
6.0 0.4
10.7 1.2
19.0 1.6
>25
16
H
11.3 1.5
6.6 0.5
13.6 1.0
17
18
19
>25
>25
>25
H
H
14.8 1.2
>25
6.0 0.3
10.2 0.5
14.4 0.8
>25
a
All analogues with measured IC50 values showed cell viability of less than 1% at 25
HepG2/C3A is simplified as HepG2.
lM.
b
compounds had very poor water solubility and thus were difficult
to assay led us not to pursue the synthesis of dialkylated com-
pounds from aromatic aldehydes.
SAR analysis. Using the phenyl analog (compound 20) as a refer-
ence, the incorporation of electron-withdrawing substituents into
the carbocyclic aromatics, such as p-nitro (compound 21), lessened
the cytotoxicity, while the inclusion of electron-donating substitu-
ents, such as p-methyl (22), p-methoxy (23), 3,4,5-trimethoxy (24)
and p-hydroxyl (25), raised the cytotoxicity, especially 25. In addi-
tion, by comparing the cytotoxicity of analogues 26–28, which con-
tain a nitrogen-containing heterocyclic substituent, with that of
compound 20, it is clear that the presence of nitrogen-containing
heterocycle (27 and 28) greatly enhanced the cytotoxicity, while
the (2-furanyl)-methyl moiety (35) has little effect. The recovery
of the cytotoxicity for compounds 25, 27 and 28 implies that the
interaction with their biological target(s) depends on the finely de-
fined details of binding, but not the overall size of the substituent
on nitrogen.
Although SAR analysis of the alkylated derivatives containing
the benzo[b]furan system elucidated some of the structural fea-
tures key to bioactivity, we wondered about the generality of these
conclusions. Therefore, some relevant alkylated analogues contain-
ing the benzo[b]thiophene core were synthesized to further vali-
date the trends deduced from the above SAR analysis.
The synthesis of the alkylated benzo[b]thiophene-containing
analogues employed the previously established reductive amina-
tion protocol for the benzo[b]furan system. The structures of the
synthesized analogues (30–40) are listed in Table 4. It is of note
that, unlike the benzo[b]furan system, during the preparation of
the dimethylated analogue 30, it was found that the separation
of the monoalkylated product was possible. Therefore, we explored
conditions to produce a small amount of the pure monomethylated
analogue for SAR analysis. We first performed the reductive amina-
tion in methanol with a large excess of free amino analogue 6, but
After the assembly of this small library consisting of 18 alkyl-
ated analogues of 5, we assayed their cytotoxicity against three
cancer cell lines. The data were summarized in Table 2 for aliphatic
substituents and Table 3 for aromatic substituents, respectively.
Two obvious trends emerge from these data. First, except for small
aliphatic substituents such as methyl (12) and ethyl (13) groups,
the dialkylated analogues had lower cytotoxicity compared to the
monoalkylated compounds (14 vs 15, and 16 vs 17), especially
for HT29 cells. Second, for the monoalkylated analogues, the linear
aliphatic groups (Table 2) usually exhibited a much stronger cyto-
toxic effect than cyclic aliphatic groups (compound 19, Table 2)
and most of the aromatic groups (Table 3). This effect is more pro-
nounced for MCF-7 and HepG2/C3A cells. Both trends could be ex-
plained by two reasons: low aqueous solubility and/or steric
hinderance that may interfere with the interaction with their bio-
logical target(s).
Figure 1 shows the IC50 values for the analogues with linear/
branched aliphatic substituent(s) (excluding inactive compounds
15, 17, and 19). It is of note that there is a spike in the IC50 trends
for monobutylated analogue 14. The increased potency of ana-
logues 16 and 18 might imply that lipophilicity can rescue those
compounds with poor aqueous solubility to some extent, presum-
ably due to increased cell permeability.9
The aromatic substituents were divided into two subgroups:
carbocyclic (left, Fig. 2) and heterocyclic (right, Fig. 2). Because
IC50s could not be obtained for some analogues, such as 20–22,
due to poor aqueous solubility, the cell viability data at 25
lM
(Table S2 in Supplementary data) were employed here for the

5110
W. Shi, T. L. Lowary / Bioorg. Med. Chem. Lett. 21 (2011) 5107–5112
Table 3
Cytotoxicity of the alkylated analogues from compound 5 and aromatic aldehydes
O
O
O
NR1R2
Compound
R1
H
R2
IC50
(
l
M)
MCF-7
HT29
HepG2a
7.9 0.7
5
H
H
10.3 0.2
6.5 1.8
17.9b
20
>25
>25
21
22
23
H
H
H
>25
>25
>25
O2N
>25
11.9 1.0
7.9 0.5
>25
31 (2.2)c
13.1 0.8
MeO
MeO
MeO
MeO
24
H
30 (1.1)c
8.0 0.6
11.3 0.9
25
26
H
H
10.9 0.9
>25
5.8 0.9
13.4 0.7
>25
HO
15.2 0.7
O
27
28
H
H
18.1b
11.1 0.9
10.8 0.6
20.8b
16.3b
N
H
N
14.3 1.3
a
b
c
HepG2/C3A is simplified as HepG2.
Assayed only once.
Cell viability at 25 lM. The numbers in parentheses represent the standard deviation.
under these conditions the major product was still the dialkylated
compound 30, which was formed together with a 2-cyanomethyla-
mino derivative (31) (Scheme 3). Byproduct 31 is presumably
formed as outlined in Scheme 3; a similar method has been
reported for the synthesis of 2-cyano amines from sodium cyanide
and aldehydes.10 When the reaction was carried out in DMF con-
taining 1% acetic acid, the dialkylated analogue 32 was obtained
instead of 30. Apparently, the electron-withdrawing cyano group
prevents the second methylation in methanol, but not in acidified
DMF. Finally, by using solid paraformaldehyde in methanol rather
than aqueous formaldehyde solution, monomethylated analogue
29 (>85% pure based on 1H NMR spectroscopy) was obtained; the
major contaminant was the free amino compound 6.
With alkylated analogues 29–40 in hand, they were subse-
quently subjected to the in vitro cytotoxicity assay (Table 4). The
same cytotoxicity profile was observed for monomethylated
compound 29 as for dimethylated (30) and monoethylated (33)
analogues. All the previously identified trends with the
benzo[b]furan derivatives were confirmed with this series of ana-
logues. For example, except for small groups, monoalkylation is fa-
vored over dialkylation (35 vs 36); aliphatic substituents generally
Figure 1. Cytotoxicity (IC50 value) trends for the aliphatic alkylated analogues.

W. Shi, T. L. Lowary / Bioorg. Med. Chem. Lett. 21 (2011) 5107–5112
5111
provide better bioactivity than aromatic ones; and the lipophilicity
has some influence on the cytotoxicity (37 vs 35). Finally, based on
the assay data for cyano-containing analogues 31 and 32, which
have low cytotoxicity, we can further support the statement made
above that increasing electron density on nitrogen enhances
cytotoxicity.
In conclusion, although amino group modification of the 4-N-
TDTH analogues did not further enhance the cytotoxicity com-
pared to parent compounds 5 and 6, several general trends have
been identified from the above SAR analysis. First, the presence
of at least one basic moiety appears to be essential. This group
can be either directly attached, or tethered through another group,
to the carbohydrate ring. Second, modification of the amine with
linear or branched aliphatic chains is favored over cyclic aliphatic
or aromatic substituents; dialkylation is disfavored compared to
monoalkylation. Third, given the potent cytotoxicity in the case
of the N-(p-hydroxylbenzyl) group, the incorporation of extra func-
tional groups into a linear or branched aliphatic chain may be nec-
essary for stronger interactions between the analogue and its
molecular target(s). Fourth, the proper lipophilicity of an analogue
may facilitate cell membrane permeability, and thus have an influ-
ence on cytotoxicity.
Figure 2. Cytotoxicity (cell viability at 25 lM) trends for the aromatic monoalky-
lated analogues.
Table 4
Cytotoxicity of alkylated analogues from 6 and aldehydes
O
O
S
NR1R2
Compound
R1
H
R2
IC50
(
l
M)
MCF-7
HT29
HepG2a
8.8 0.9
8.8 1.0
6
H
H
7.1 1.2
7.1 1.0
5.9 0.5
5.4 0.2
29b
30
31
32
33
34
35
36
37
8.2 0.9
10.5 1.3
7.7 0.8
5.9 0.5
12.0 0.7
10.2 0.9
>25
3.3 0.4
10.8 1.0
6.2 0.7
3.4 0.4
3.9 0.3
7.0 0.5
>25
10.0 0.9
18.0 1.6
12.3 1.4
6.6 0.7
11.7 0.9
18.6 1.7
>25
H
H
H
NC
NC
H
H
6.8 0.6
5.1 0.4
19.1 1.6
38
39
40
>25
19.0 1.2
6.2 0.5
5.1 0.6
>25
H
H
>25
>25
MeO
HO
12.0 1.0
20.3 0.6
a
HepG2/C3A is simplified as HepG2.
Purity ꢀ85%.
b

5112
W. Shi, T. L. Lowary / Bioorg. Med. Chem. Lett. 21 (2011) 5107–5112
O
O
O
O
O
S
O
S
N
NH
H-B
NC
NH2
31
NC
6
O
O
S
N
NC
29
32
Scheme 3.
Further studies with the daunosamine or acosamine carbohy-
drate moiety in 1–4 are planned to consolidate and enrich the
SAR data. In addition, functionalization of the aromatic domain is
expected to be crucial for pronounced improvement in cytotoxic-
ity. The identification of more potent compounds should facilitate
the elucidation of the molecular target(s) responsible for the cyto-
toxicity for the class of compounds. In this regard we should note
that all of the compounds reported here possess a basic nitrogen
atom and bind to DNA (see Supplementary data) to varying
degrees. However, the affinity is rather weak and thus it is likely
that other biological targets are also involved in the observed
cytotoxicity.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2011.02.051.
References and notes
1. Shi, W.; Coleman, R. S.; Lowary, T. L. Org. Biomol. Chem. 2009, 7, 3709.
2. Shi, W.; Marcus, S. L.; Lowary, T. L. Carbohydr. Res. 2010, 345, 10.
3. Shi, W.; Marcus, S. L.; Lowary, T. L. Bioorg. Med. Chem. 2011, 19, 603.
4. Shi, W.; Lowary, T. L. Bioorg. Med. Chem. 2011, 19, 1779.
5. Abdel-Magid, A. F.; Carson, K. G.; Harris, B. D.; Maryanoff, C. A.; Shah, R. D. J.
Org. Chem. 1996, 61, 3849.
6. Khan, S. N.; Cho, N.-J.; Kim, H.-S. Tetrahedron Lett. 2007, 48, 5189.
7. Skalova, T.; Hasek, J.; Dohnalek, J.; Petrokova, H.; Buchtelova, E.; Duskova, J.;
Soucek, M.; Majer, P.; Uhlikova, T.; Konvalinka, J. J. Med. Chem. 2003, 46, 1636.
8. Wischnat, R.; Martin, R.; Wong, C.-H. J. Org. Chem. 1998, 63, 8361.
9. Camenisch, G.; Folkers, G.; van de Waterbeemd, H. Pharm. Acta Helv. 1996, 71,
309.
Acknowledgements
This work was supported by the University of Alberta, the Nat-
ural Sciences and Engineering Research Council of Canada and the
Alberta Ingenuity Centre for Carbohydrate Science. WS was the re-
cipient of a Studentship from the Alberta Ingenuity Fund.
10. Spaltenstein, A.; Holler, T. P.; Hopkins, P. B. J. Org. Chem. 1987, 52, 2977.