New Disaccharide-Based Ether Lipids as SK3 Ion Channel Inhibitors

Article abstract of DOI:10.1002/cmdc.201600147

The SK3 potassium channel is involved in the development of bone metastasis and in the settlement of cancer cells in Ca²⁺-rich environments. Ohmline, which is a lactose-based glycero-ether lipid, is a lead compound that decreases SK3 channel activity and consequently limits the migration of SK3-expressing cells. Herein we report the synthesis of three new ohmline analogues in which the connection of the disaccharide moieties (1→6 versus 1→4) and the stereochemistry of the glycosyl linkage was studied. Compound 2 [3-(hexadecyloxy)-2-methoxypropyl-6-O-α-d-glucopyranosyl-β-d-galactopyranoside], which possesses an α-glucopyranosyl-(1→6)-β-galactopyranosyl moiety, was found to decrease SK3 current amplitude (70 % inhibition at 10 μm), displace SK3 protein outside caveolae, and decrease constitutive Ca²⁺entry (50 % inhibition at 300 nm) and SK3-dependent cell migration (30 % at 300 nm) at a level close to that of the benchmark compound ohmline. Compound 2, which decreases the activity of SK3 channel (but not SK2 channel), is a new drug candidate to reduce cancer cell migration and to prevent bone metastasis.

Full text of DOI:10.1002/cmdc.201600147

DOI: 10.1002/cmdc.201600147
Full Papers
New Disaccharide-Based Ether Lipids as SK3 Ion Channel
Inhibitors
Wilfried Berthe,[a, c] Charlotte M. Sevrain,[a, c] Aurꢀlie Chantꢁme,[b, c] Ana Maria Bouchet,[b, c]
Maxime Gueguinou,[b, c] Yann Fourbon,[b, c] Marie Potier-Cartereau,[b, c] Jean-Pierre Haelters,[a, c]
Hꢀlꢂne Couthon-Gourvꢂs,[a, c] Christophe Vandier,[b, c] and Paul-Alain Jaffrꢂs*[a, c]
The SK3 potassium channel is involved in the development of
bone metastasis and in the settlement of cancer cells in Ca2+
-rich environments. Ohmline, which is a lactose-based glycero-
ether lipid, is a lead compound that decreases SK3 channel ac-
tivity and consequently limits the migration of SK3-expressing
cells. Herein we report the synthesis of three new ohmline ana-
logues in which the connection of the disaccharide moieties
(1!6 versus 1!4) and the stereochemistry of the glycosyl
linkage was studied. Compound 2 [3-(hexadecyloxy)-2-methox-
ypropyl-6-O-a-d-glucopyranosyl-b-d-galactopyranoside], which
possesses
an
a-glucopyranosyl-(1!6)-b-galactopyranosyl
moiety, was found to decrease SK3 current amplitude (70% in-
hibition at 10 mm), displace SK3 protein outside caveolae, and
decrease constitutive Ca2+ entry (50% inhibition at 300 nm)
and SK3-dependent cell migration (30% at 300 nm) at a level
close to that of the benchmark compound ohmline. Com-
pound 2, which decreases the activity of SK3 channel (but not
SK2 channel), is a new drug candidate to reduce cancer cell mi-
gration and to prevent bone metastasis.
Introduction
The arsenal of drugs available for the treatment of cancer is al-
ready large, but in some cases inefficacy (e.g., toward pancre-
atic cancer) or side effects have inspired chemists, biologists,
and clinicians to investigate new compounds and—more inter-
estingly—new targets. Many anticancer chemotherapies act on
the development of primary tumors by following various
mechanisms of action, including interaction with DNA (interca-
lating agents),[1] the use of antimitotic agents (interaction with
microtubules),[2] or antiangiogenic agents,[3] to mention a few
drug classes. However, one major concern occurs when cancer
cells disseminate in the body, leading to the formation of
metastases distant from the primary tumor. An efficient treat-
ment for limiting the formation of metastases, alongside
action against development of the primary tumor, has yet to
be discovered. Compounds that inhibit the activity of matrix
metalloproteinases (MMPs) could provide a strategy to de-
crease the escape of cancerous cells from primary tumors.[4]
However, despite some promising results, this strategy has
thus far failed,[5] but active research efforts in this field are still
in progress. Another strategy to limit the formation of bone
metastasis is based on the administration of hydroxybisphos-
phonic acid based derivatives.[6] However, if this therapeutic
approach decreases the kinetics of bone resorption associated
with the presence of bone metastasis, this treatment is not
suitable for blocking the invasion process of cancer cells. Final-
ly, denosumab is another treatment for osteoporosis that is
currently in phase III clinical trials for the prevention of bone
metastasis in the context of prostate cancer.[7] In fact, this mon-
oclonal antibody is not directed against cancer cells, but in-
stead targets the RANK ligand, which is a receptor involved in
bone resorption.
More recently, it was demonstrated that the SK3 ion channel
is involved in cancer cell migration. Moreover, this ion channel
is expressed in highly metastatic cancer and invasive cancer
cells (e.g., MDA-MB435s).[8–11] These results led us to assess the
modulation of SK3 activity, which, according to the literature,
can be achieved by certain toxic peptides (e.g., apamin) or by
heterocyclic compounds.[12] We found that amphiphilic ether
lipid compounds can modulate the activity of SK3 potassium
channels when used at a concentration well below (100–
300 nm) their cytotoxic levels (20–50 mm). These original re-
sults, which echo other works relating the modulation of ion
channels with amphiphilic compounds,[12–14] invited us to con-
sider the SK3 channel as a new target for cancer therapy and
to use amphiphilic compounds to modulate its activity.[15]
Edelfosine is another ether lipid that can modulate the activ-
ity of the SK3 channel.[16] However, its toxicity (cytotoxicity at
concentrations up to 3 mm) led us to investigate new structural
[a] W. Berthe, Dr. C. M. Sevrain, Dr. J.-P. Haelters, Dr. H. Couthon-Gourvꢀs,
Prof. P.-A. Jaffrꢀs
Universitꢁ de Brest, CEMCA, CNRS UMR 6521, IBSAM
6 Avenue Le Gorgeu, 29238 Brest (France)
E-mail: pjaffres@univ-brest.fr
[b] Dr. A. Chantꢂme, Dr. A. M. Bouchet, Dr. M. Gueguinou, Y. Fourbon,
Dr. M. Potier-Cartereau, Prof. C. Vandier
INSERM, UMR 1069, Universitꢁ FranÅois Rabelais, Tours, 37032 (France)
[c] W. Berthe, Dr. C. M. Sevrain, Dr. A. Chantꢂme, Dr. A. M. Bouchet,
Dr. M. Gueguinou, Y. Fourbon, Dr. M. Potier-Cartereau, Dr. J.-P. Haelters,
Dr. H. Couthon-Gourvꢀs, Prof. C. Vandier, Prof. P.-A. Jaffrꢀs
Network “Ion Channels and Cancer – Canceropole Grand Ouest”, (IC-CGO),
(France)www.ic-cgo.fr
Supporting information for this article can be found under http://
dx.doi.org/10.1002/cmdc.201600147.
ChemMedChem 2016, 11, 1 – 10
1
ꢃ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
analogues. In particular, we investigated the replacement of
the phosphocholine moiety of edelfosine by a saccharide or
disaccharide group.[17] Some molecules belonging to this com-
pound class (glycosylated antitumor ether lipids) were studied
for their antiproliferative properties at higher concentrations
(IC50 values between 11 and 30 mm).[18] In a study that aimed to
evaluate the action of glyco-ether lipids on SK3-dependent cell
migration, we found ohmline (Figure 1) to be an excellent can-
didate. Indeed, this compound can decrease the migration of
SK3+ cells by up to 50% (no effect on SK3ꢀ cells) and can de-
crease the SK3 current by 70% (patch-clamp experiments). In
an in vivo experiment, these results were confirmed by a reduc-
tion of bone metastases in a murine model of breast cancer.[19]
Ohmline likely interacts with the SK3 channel by insertion into
the plasma membrane; however, this compound is also able to
destabilize an ion channel complex formed through the associ-
ation of an SK3 channel and the Orai1 Ca2+ channel (SK3–
Orai1 complex), which is located only in the lipid rafts of
cancer cells. The SK3 channel is therefore a new target for
acting on the migratory capacity of cancer cells and on their
ability to settle in a calcium-rich environment (e.g., bones).
To better understand the mechanism of action of ohmline
on the SK3 channel, we went further into studying the struc-
ture–activity relationship. First, we found that the introduction
of a phosphate group between the disaccharide and the glyc-
ero-ether lipid fragment also produces compounds that modu-
late the activity of the SK3 channel. However, their biological
activities were not better than that of ohmline, and the selec-
tivity of these compounds toward SKCa channels was de-
creased.[20] We then incorporated the a-galactopyranosyl-(1!
6)-b-galactopyranosyl moiety into the structure of the ether
lipid to produce an ether lipid analogue of digalactosyldiacyl-
glycerol (DGDG; Figure 1).[21] This compound was found to
modulate SK3 channel activity (ꢀ28%) and SK3-dependent cell
migration (ꢀ20%), but these values were not competitive with
respect to ohmline (reduction of SK3 current by 70% and SK3-
dependent cell migration by 50%). In the present work we
report the synthesis of three new ether lipid analogues of
ohmline, but with a distinct structure at the disaccharide polar
head group. The action of these new compounds on SK3 activ-
ity, SK3 localization in lipid rafts, Ca2+ entry, and SK3-depen-
dent cell migration are also reported. For the first time, we
found one of these compounds to possess efficiency similar to
that of ohmline in decreasing SK3 channel activity and, inter-
estingly, with faster kinetics of action; in view of its structure,
this suggests that it could be resistant to lactase activity and
could therefore be administered orally.
Results and Discussion
On the basis of the capacity of ohmline to modulate the activi-
ty of SK3 channels at low concentrations (<1 mm), we focused
on the structure of the disaccharide unit to address two ques-
tions: the influence of the stereochemistry of the glycosyl link
(a versus b) between the two saccharide units, and that of the
saccharide unit bonded to the glycerol backbone. Because we
previously observed that the chirality at the sn-2 position of
the glycerol unit of ohmline had no effect on modulating the
SK3 channel,[17] in this study, we did not control the stereo-
chemistry at this position. Accordingly, all the compounds
were isolated as a mixture of diastereomers. First, we investi-
gated the synthesis of what we call a-ohmline, which is char-
acterized by an a-glycosyl link between the lactose moiety and
the glycerol unit (Scheme 1). For this synthesis we used the tri-
methylsilyl protecting group for construction of the disacchar-
ide moiety. In the present case, lactose was persilylated by
adaptation of the method reported by Uchiyama and Hinds-
gaul.[22] The per-trimethylsilyl lactose 4 was then engaged in
an iodination reaction with trimethylsilyliodide. This reaction
was carried out at 08C for a short period of time to avoid deg-
radation of the glycosyl link between the galactosyl and gluco-
syl moieties as previously reported.[23] The iodo intermediate 5
was directly engaged in a glycosylation reaction with the glyc-
ero-ether lipid 6 to produce the intermediate 7, which features
Figure 1. A) Structures of previously reported disaccharide glycero-ether lipids: lactose-based glycero-ether lipid ohmline and the (1!6) digalactosyl glycero-
ether lipid A. B) New analogues of ohmline reported in this study: compounds 1–3.
&
ChemMedChem 2016, 11, 1 – 10
www.chemmedchem.org
2
ꢃ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
Scheme 1. Synthetic procedures used to prepare a-ohmline 1.
seven trimethylsilyl protecting groups. This intermediate 7 was
mixed with methanol and an acidic resin (Dowex 50X8) to
remove the trimethylsilyl groups. After purification by flash
chromatography on silica gel, the expected a-ohmline 1 was
isolated in 21% yield. The configuration (a) of the glucopyra-
nosyl moiety was confirmed by 13C NMR recorded in CD3OD (a
peak for the anomeric carbon C1 at 100.5 ppm confirmed the
a-glycosylation, whereas the C1’ corresponding to the anome-
ric position between the two saccharide moieties is observed
at 105.1 ppm, as expected for a b-stereoisomer). To confirm
this stereochemistry determined on the basis of the chemical
shift of the anomeric proton, for this compound 1 the 1JHC cou-
pling constants were also determined (NMR in DMSO, see Sup-
porting Information Figure SI1-3). Two doublets were observed
for the C1 anomeric carbon corresponding to two a diastereo-
1 was fully characterized by NMR and mass spectrometry (Fig-
ure SI1).
The preparation of new (1!6) disaccharide-based analogues
of ohmline aimed to complete a previous study that reported
the synthesis of digalactosyl analogues (compound A,
Figure 1), which featured a b-galacopyranosyl unit bonded to
the glycerol unit.[21] We have decided to change the terminal
a-galactopyranosyl unit with an a-glucopyranosyl moiety. In
previous study, we reported the interesting SK3 modulatory ca-
pacity of glucopyranosyl-based disaccharide ether lipids (lac-
tose, melibiose), but this saccharide moiety was never placed
at the terminal position of the disaccharide unit. We therefore
attempted to produce a-glucopyranosyl-(1!6)-b-galactopyra-
nosyl ether lipid 2 (Scheme 2). Its synthesis required a-glycosy-
lation. The methodology previously used for the preparation of
digalactosylglycero-ether lipids[21] was adapted to the synthesis
of 2. The per-trimethylsilylation of glucose was readily ach-
ieved by reaction with chlorotrimethylsilyl in the presence of
mers[24] (1JHC were respectively 166.43 and 169.53 Hz). One dou-
1
blet was observed for the anomeric carbon C1’ with JHC
=
160.44 Hz, as expected for b stereochemistry. Compound
Scheme 2. Synthesis of the a-Glc-(1!6)-b-Gal glycero-ether lipid 2.
ChemMedChem 2016, 11, 1 – 10
www.chemmedchem.org
3
ꢃ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
triethylamine to produce compound 9 (Scheme 2).[22] The
anomeric position of 9 was then iodinated to produce inter-
mediate 10, which was immediately used in the a-glycosyla-
tion with compound 8.[21] The produced disaccharide-based
compound 11 was immediately engaged in a deprotection re-
action to produce compound 2 in 17% yield over three steps.
The a-glycosylation, which was expected,[21,25–27] was confirmed
by NMR (at 4.27 ppm we observed a doublet with a coupling
constant of 3.0 Hz; <5.0 Hz, which confirmed the a-glycosyla-
tion), and the b configuration at the anomeric position of the
galactopyranosyl moiety was also confirmed by NMR (at
4.03 ppm the anomeric proton has a coupling constant of
9.0 Hz, >5.0 Hz, thus confirming the b configuration; see Fig-
ure SI1). With the aim to study the influence of the stereo-
chemistry at the anomeric position of the terminal saccharide,
we assessed the synthesis of the digalactose-based ether lipid
(Scheme 3) that results from a b-glycosylation when the termi-
nal galactopyranosyl moiety is introduced. For this purpose,
we activated the anomeric position with a trichloroacetimidate
group, and we used acetyl protecting groups that are known
to favor the b-glycosylation of galactose derivatives. Accord-
ingly, the b-glycosylation that engaged 8 with 12 followed by
the removal of the trimethylsilyl protecting groups, produced
the disaccharide intermediate 13 in 41% yield. The acetyl-pro-
tected alcohol functions were deprotected with sodium meth-
oxide in methanol followed by treatment with the acidic resin
Amberlite IR-120H to produce compound 3 in 36% yield. The
b configuration of both galactopyranosyl moieties was con-
firmed by NMR (anomeric protons have a coupling constant
larger than 5.0 Hz (9.3 Hz and 9.8 Hz), thus confirming the
b configuration of both anomeric positions; see Figure SI1).
The new disaccharides, glycero-ether lipids 1, 2, and 3 were
first tested for their capacity to modulate the activity of the
SK3 channel. These compounds were tested for their capacity
to decrease SK3 channel activity on HEK293T cells expressing
homotetrameric SK3 channels: HEK293T-rSK3 cells. We tested
the inhibitory effect of these three compounds on whole-cell
SK3 current using the patch-clamp technique (Figure 2). We
previously demonstrated that testing the compounds at 10 mm
Figure 2. Effect of compounds 1–3 on SK3 current and SK3-dependent cell
migration. A) Ohmline (n=4) and compound 2 (n=3) decrease SK3 current
in contrast to 1 (n=4) and 3 (n=4), all assayed at 10 mm. The SK3 current is
the apamin-sensitive current. B) Representative whole-cell SK3 current show-
ing the effect of compound 2 with or without apamin (100 nm) on SK3 cur-
rent. To maintain consistency with conventional representations of current–
voltage curves, current–voltage relationships are represented from ꢀ100 to
+100 mV. C) Time response inhibition of 10 mm ohmline (n=4) and 10 mm 2
(n=3) on SK3 currents (apamin-sensitive current) recorded at 0 mV. D) Histo-
grams showing the effect of 300 nm ohmline and 300 nm compound 2 on
the migration of HEK-rSK3 cells. The normalized cell number corresponds to
the ratio of the total number of migrating cells in the presence of drugs
over the total number of migrating cells in control experiments. Results are
expressed as the mean ꢁSEM; *p=0.05 (N=4, n=12, Kruskal–Wallis and
post hoc tests).
on SK3 current permits a rapid and robust pre-selection of SK3
channel inhibitors that are effective on SK3-dependent cancer
cell migration and at lower concentrations.[17] The effect of
ohmline derivatives on the amplitude of SK3 current was mea-
sured at 0 mV; at this membrane potential the current is only
carried by SK3 channels.
SK3 currents are generated using a ramp protocol from
+100 mV to ꢀ100 mV in the presence or absence of com-
Scheme 3. Synthesis of the b-Gal-(1!6)-b-Gal glycero-ether lipid 3.
&
ChemMedChem 2016, 11, 1 – 10
www.chemmedchem.org
4
ꢃ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
pounds (Figure 2). We first evaluated the importance of the
stereochemistry of the glycosylation link and tested compound
1. Figure 2 shows that at 10 mm, a-ohmline 1 has a slight and
insignificant effect on the amplitude of SK3 current; com-
pound 1 was found to decrease SK3 current by 5% (Figure 2A,
Table 1). Because this compound has no effect of SK3 current
amplitude, we did not test its effect on SK3-dependent cell mi-
gration. These results demonstrated the importance of the
b linkage of the lactose unit to the glycerol moiety present in
ohmline for its capacity to reduce SK3 current amplitude.
We then assessed compound 2, which must be compared
with ohmline as a standard but also to compound A (Figure 1)
due to evident structural similarities. Indeed, the structural dif-
ference between compound 2 and A involves the nature of
to inhibit SK3-dependent cell migration was evaluated. Before
performing these evaluations, the question of toxicity was ad-
dressed in order to establish a concentration that would not
induce cytotoxicity in these evaluations. The toxicity of com-
pound 2 was determined by MTT assays at 300 nm for 24 h (a
maximal concentration used with ohmline to decrease SK3-de-
pendent cell migration). Note that we also evaluated the toxici-
ty of compounds 1 and 3 and found no toxicity at 1 mm for
24 h (data not shown). Compound 2 showed no detectable
levels of toxicity at 300 nm (N=3, n=9, Figure SI2). At 300 nm,
compound 2 was found to significantly inhibit cell migration
of HEK293T-SK3 to values close to those observed with 300 nm
ohmline (Figure 2D). Moreover, the critical micellar concentra-
tion (CMC) values of compound 1 (60 mm), 2 (25 mm), and
ohmline (12 mm) suggest that the effects observed
by patch-clamp and cell migration are not related to
Table 1. Efficiency and comparison of disaccharide-based ether lipids on SK3 channel
a micellar organization of the compound.
activity and cell migration.
We previously demonstrated that the SK3 channel
is localized in caveolae (a region of the plasma mem-
brane rich in cholesterol and where many molecular
transduction platforms are located),[28] and that ohm-
line moves the SK3 channel outside caveolae.[19] Inter-
estingly, compounds 1 and 3 have no effect on the
localization of the SK3 protein into caveolae
(Figure 3, left), whereas compound 2 and ohmline
displace SK3 outside the caveolae (Figure 3, right).
To determine the biological relevance of com-
pound 2 in cancer cells we evaluated the effect of
compound 2 on the breast cancer cell line MDA-
Entry
Disaccharide
unit
Compd
Cell migration
SK3
Ref.
decrease [%] inhibition [%][c]
1
2
3
4
5
6
b-Gal-(1!4)-b-Glc
ohmline
73
5
44, 20
28.2
70
[17]
This study
[17]
[21]
This study
This study
b-Gal-(1!4)-a-Glc a-ohmline 1
a-Gal-(1!6)-b-Glc melibiose-GEL
–
30[a]
19.6
40[b]
0
a-Gal-(1!6)-b-Gal
a-Glc-(1!6)-b-Gal
b-Gal-(1!6)-b-Gal
A
2
3
0
[a] At 300 nm. [b] At 1 mm. [c] Patch-clamp measurement with SK3+ HEK cells; the
potent inhibitors were tested at 10 mm.
the terminal saccharide unit (galactopyranosyl in compound A
and glucopyranosyl in compound 2), whereas the connection
(1!6) and anomeric configuration are identical. Interestingly,
compound 2 reduced the amplitude of SK3 currents by 70%
(Figure 2A,B; Table 1). This value is close to the inhibitory ca-
pacity of ohmline at 10 mm (ꢀ73%).[17] The next experiments
assessed the influence of the configuration of the anomeric
position that links the two saccharide units. For this purpose
compound 3 (featuring a b-Gal(1!6)-b-Gal moiety) was syn-
Figure 3. Localization of SK3 in the caveolae of HEK-rSK3 cells. Representa-
tive immunoblots from two different experiments representing membrane
fractionation on a sucrose gradient of HEK-rSK3 cells treated (or not) with
10 mm compound 1, 10 mm compound 3, 10 mm ohmline, and 10 mm com-
pound 2 for 20 min. b-Adaptin is a protein used as a non-caveolae fraction
marker, and caveolin-1 is a marker of caveolae fractions; NC: non-caveolae
fraction, C: caveolae fraction.
thesized with the aim to compare its capacity to modulate the
SK3 channel with that of compound A (featuring an a-Gal(1!
6)-b-Gal moiety). As shown in Figure 2A compound 3 has no
effect on SK3 channel activity, demonstrating the importance
of the configuration of the anomeric position linking the two
saccharide moieties.
Based on these results indicating that compound 2 is able
to inhibit SK3 channel activity as does ohmline, we further ana-
lyzed the time required for compound 2 to reduce SK3 current
amplitude by 50% at 10 mm (Figure 2C). Regardless of the con-
centration tested, ~40 s was needed to decrease the current
amplitude by 50%. Notably, ~75 s were needed to decrease
the amplitude of the SK3 current by 50% with ohmline. We
next tested the effect of compound 2 on the SK2 channel, an
SKCa channel from the same family as the SK3 channel, using
HEK293T cells expressing homotetrameric SK2 channels. Com-
pound 2 has no effect on SK2 currents (n=3: see Figure SI2),
as also observed with ohmline.[17] The capacity of this molecule
MB435s, which expresses the SK3 channel.[8] With this cell line,
we found that compound 2 decreases cell migration (Fig-
ure 4A, left). In contrast, the same cell line in which the expres-
sion of SK3 protein was silenced (SK3ꢀ) compound 2 has no
effect (Figure 4A right).
These results demonstrate the capacity of compound 2 to
decrease SK3-dependent migration of cancer cells, as observed
earlier with ohmline.[19] We next evaluated the effect of com-
pound 2 on the constitutive Ca2+ entry into MDA-MB435s
cells. Figure 3B showed that at 300 nm, compound 2 decreases
Ca2+ entry by almost 50%. Thus, compound 2 is an inhibitor
ChemMedChem 2016, 11, 1 – 10
www.chemmedchem.org
5
ꢃ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
far, a b configuration is a much better choice. In terms of the
stereochemistry of the anomeric position of the terminal sac-
charide unit, comparison of the modulation of SK3 channel ac-
tivity by compounds A and 3 is an initial indication of the in-
fluence of the a-connection. However, this comparison must
not be too conclusive, because for the benchmark compound
ohmline, which is a 1!4 connected disaccharide, the galac-
tose unit features b stereochemistry. The great interest of the
results reported in this study arises from the excellent efficacy
of compound 2, which features a terminal a-glucopyranosyl
moiety. Despite this compound being slightly less efficient
than ohmline, it could be very promising for in vivo experi-
ments, as we can expect a different kinetics of metabolism.
Indeed, b-galactosidase, which is known to selectively hydro-
lyze b-galactopyranosyl-based di- or polysaccharides such as
the lactose unit, which is present in ohmline, is known to be
inefficient toward a-glucopyranosyl derivatives as substrates.[29]
Figure 4. Effect of compound 2 on SK3-dependent cell migration and on the
constitutive Ca2+ entry of the breast cancer cell line MDA-MB-435s. A) left:
Histograms showing the effect of 100 and 300 nm compound 2 on the mi-
gration of MDA-MB-435s cells that express the SK3 channel (SK3+, N=3,
n=6). A) right: Histograms showing the effect of 100 and 300 nm com-
pound 2 on the migration of MDA-MB-435s cells that do not express the
SK3 channel (SK3ꢀ, N=3, n=6). The normalized cell number corresponds
to the ratio of the total number of migrating cells in the presence of drugs
over the total number of migrating cells in control experiments. Results are
expressed as the mean ꢁSEM, *p=0.05 (Kruskal–Wallis and post hoc tests).
B) left: Time-dependent fluorescence measurement of constitutive Ca2+
entry recorded in MDA-MB-435s cells. B) right: Histograms showing steady-
state relative (to control) fluorescence to Ca2+ entry under control condi-
tions or in the presence of 300 nm compound 2. Data are the mean ꢁSEM;
*p<0.05 (N=2, n=5, Mann Whitney test).
Conclusions
From a synthetic point of view, this work further illustrates the
use of trimethylsilyl protecting groups that allow a-glycosyla-
tion of a pre-formed disaccharide (lactose). It also illustrates
the use of this protecting group to construct a (1!6) disac-
charide unit. In both cases, the number of steps is decreased
relative to the use of classical protecting groups (trityl and
benzyl), and despite the fact that some experimental precau-
tions are required (temperature and reaction time for the prep-
aration of the iodinated intermediates 5 and 10), this proce-
dure is efficient for the production disaccharide glycero-ether
lipids in reasonable yields. This synthetic strategy was used to
produce three new ether lipids, each featuring a polar head
group composed of a disaccharide moiety.
of SK3-dependent cancer cell migration that moves SK3 out-
side caveolae, thus likely dissociating the SK3–Orai1 complex
(it would delocalize SK3 and Orai1 channels from caveolin frac-
tions), and decreasing Ca2+ entry through the Orai1 channel as
previously observed with ohmline.[19] With the aim to compare
the biological effects of the three new compounds reported
herein with other disaccharide-based glycero-ether lipids, we
included the activity of ohmline (Table 1 entry 1), melibiose-
GEL (entry 3), and digalactosyl derivative A (entry 4). These re-
sults indicate that both 1!4 and 1!6 connected disacchar-
ides can produce very efficient modulators of the SK3 channel
(entries 1 and 5), but also poor modulators as illustrated by
compounds a-ohmline 1 (entry 2) and 3 (entry 6). The second
important conclusion is that the stereochemistry of the
anomeric position of the saccharide units has a strong impact
on the capacity of these ether lipids to modulate the SK3
channel. a-Ohmline 1 is the first disaccharide-based ether lipid
tested for the modulation of the SK3 channel that features an
a-glycosyl link between the disaccharide unit and the glycerol
backbone. The consequence of this a-stereochemistry (Table 1
entry 2) is evidenced by comparison of the biological results
with those obtained with ohmline (entry 1). a-Ohmline is inac-
tive, whilst ohmline is the most efficient SK3 modulator report-
ed so far. It is difficult to generalize from one result, but the
stereochemistry of the anomeric position that links the disac-
charide unit to the glycerol unit has a strong influence and, so
From a biological point of view, compound 2 decreased SK3
channel activity and displaced SK3 outside lipid rafts, leading
to a decrease in constitutive Ca2+ entry and finally cancer cell
migration. The level of SK3 inhibition by compound 2 is similar
to that of the gold standard ohmline. The interesting differ-
ence between the two compounds is that 2 features an a-
anomeric junction between the two saccharide units instead of
a b junction in the case of ohmline. This difference should
permit the oral administration of compound 2 with a limited
risk of premature degradation.
Experimental Section
Chemistry
General: Dry solvents (CH2Cl2, DMF) were obtained with a MBraun-
SPS system. 1H NMR (500.13, 400.133, or 300.135 MHz), 13C NMR
(125.773 or 75.480 MHz) were used to characterize the compounds
with Bruker AC300, AvanceDRX 500, AvanceDRX and 400 spectrom-
eters. J corresponds to the coupling constants that are expressed
in Hertz; NMR spectra a described by the following abbreviations:
singlet (s), doublet (d), triplet (t), quadruplet (q), and multiplet (m).
13C heteronuclear HMQC and HMBC were used to further character-
ize the new compounds. Bruker Autoflex MALDI TOF-TOF III
LRF200 CID was used for mass spectrometric analyses. Fourier
&
ChemMedChem 2016, 11, 1 – 10
www.chemmedchem.org
6
ꢃ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
transform infrared (FTIR) spectra were recorded on a Bruker
Vertex 70 FTIR (ATR) spectrometer. Commercial compounds were
used as received, but TMSI was distilled over CaCl2 and DIPEA over
KOH. Optical rotations were measured on a Jasco P-1010 polarime-
ter using a 100 mm cell.
poured into crushed ice. The organic layer was washed with iced
water (3ꢅ50 mL), dried over MgSO4, filtered and concentrated to
give the pure compound as
(400.133 MHz, CDCl3): d=0.14 (m, 45H, SiMe3), 3.30–3.42 (m, 2H,
CH2), 3.60–3.80 (4H, m, H2, H3, H4, H5), 5.00 ppm (1H, d, 3JHH =4,
a
colorless oil (97%). 1H NMR
H1).
Compounds 6,[30,31] 8,[21] 12,[32,33] were prepared according to re-
3-(hexadecyloxy)-2-methoxypropyl-6-O-a-d-glucopyranosyl-b-d-
ported procedures.
galactopyranoside 2: To
a stirred solution of 9 (449 mg,
4-O-[2,3,4,6-tetrakis-O-(trimethylsilyl)-b-d-galactopyranosyl]-
1,2,3,6-tetrakis-O-(trimethylsilyl)-a-d-glucopyranose 4: The pro-
tocol was adapted from the work of Uchiyama and Hindsgaul.[22]
To a stirred solution of d-lactose (650 mg, 1.9 mmol) and Et3N
(1.37 mL, 10.2 mmol) in dry DMF (10 mL) at 08C was added drop-
wise TMSCl (1.29 mL, 10.2 mmol). The mixture was stirred 4 h at
room temperature. n-Pentane (30 mL) was added, and the mixture
was poured into crushed ice. The organic layer was washed with
iced water (10 mL), dried over MgSO4, filtered and concentrated to
give the pure compound as a colorless oil (81%). This compound,
which is an intermediate in this synthesis, was previously report-
0.83 mmol) in dry CH2Cl2 (5 mL) at 08C was added TMSI (130 mL,
0.91 mmol). The mixture was stirred for 20 min at 08C, and toluene
(5 mL) was added. The solution was concentrated at 08C under
vacuum to give the a-iodide. The residue was dissolved in dry
CH2Cl2 (5 mL) and stored under argon. In another reactor a solution
of TBAI (459 mg, 1.24 mmol), DIPEA (206 mL, 1.24 mmol), and com-
pound 8 in dry CH2Cl2 (5 mL) was stirred for 2 h at room tempera-
ture under argon. The solution containing the iodide compound
was added dropwise, and the mixture was stirred for 4 days at
room temperature. The solution was concentrated, and the residue
was dissolved in dry Et2O (20 mL) to precipitate salts, filtered and
concentrated to yield the intermediate 11 as a yellow resin. To
a stirred solution of 11 in methanol (5 mL) was added Dowex
50WX8-200ꢄ ion-exchange resin (300 mg). The mixture was stirred
for 2 h at room temperature, filtered, and concentrated. The resi-
due was purified by three chromatographic separations on silica
gel: first CH2Cl2 with 2 to 30% MeOH, then EtOAc/MeOH/H2O
83:10:7, and finally CH2Cl2 with 6 to 16% MeOH to give the expect-
1
ed.[34] TLC (EtOAc/petroleum ether 1:9): 0.77; H NMR (400.133 MHz,
CDCl3): d=0.06–0.16 (m, 72H, OTMS), 3.14–3.21 (m, 1H, H2), 3.23–
3.32 (m, 3H, H4 +H6’), 3.40–3.82 (m, 7H, H5 +H6 +H2’ +H3’ +H4’ +
3
3
3
H5’), 3.88–3.94 (dd, 1H, JHH =10.8, JHH =3.6, H3), 4.27 (d, 1H, JHH
=
3
7.5, H1’), 4.44 ppm (d, 1H, JHH =7.5, H1).
3-(hexadecyloxy)-2-methoxypropyl-4-O-b-d-galactopyranosyl-a-
d-glucopyranoside 1: To a stirred solution of 4 (1.38 g, 1.50 mmol)
in dry CH2Cl2 (5 mL) at 08C was added TMSI (300 mg, 1.50 mmol).
The mixture was stirred for 30 min at room temperature, and ben-
zene (5 mL) was added. The solution was concentrated under
vacuum to give the a-iodide. The residue was dissolved in dry
CH2Cl2 (5 mL) and stored under an argon atmosphere. In another
reactor a solution of TBAI (828 mg, 2.25 mmol), DIPEA (291 mg,
2.25 mmol), and compound 6 (165 mg, 0.50 mmol) in dry CH2Cl2
(5 mL) was stirred for 2 h at room temperature under argon. The
solution containing the iodide compound was added dropwise,
and the mixture was stirred for 36 h at room temperature. The so-
lution was concentrated, and the residue was dissolved in dry Et2O
(20 mL) to precipitate salts, filtered and concentrated to yield inter-
mediate 7 as a yellow resin. To a stirred solution of 7 in methanol
(5 mL) was added Dowex 50WX8-200ꢄ ion-exchange resin
(300 mg). The mixture was stirred for 4 h at room temperature, fil-
tered and concentrated. The residue was purified by chromatogra-
1
ed product as a white solid (30 mg, 17%). H NMR (500.133 MHz,
CDCl3/CD3OD): d=0.66 (t, 3H, 3JHH =7.0, CH3), 1.05–1.13 (m, 26H,
CH2 fatty chain), 1.53 (m, 2H, CH2 b fatty chain), 3.00–3.85 (m, 21H,
OCH3 +CH2 a fatty chain+CH2 sn-1+CH2 sn-3+CH sn-2+H3 +
H4’ +H5’ +H2 +H4 +H5 +H6 +H2’ +H3’ +H6’), 3.88–3.89 (dd, 1H, 3J=
7.0), 3.96–4.05 (m, 1H, H1), 4.27 ppm (d, 1H, JHH =3.0, H1’); 13C NMR
4
(125.771 MHz, CDCl3/CD3OD): d=13.5 (C40), 14.3 (CH3), 22.3 (C38),
25.6 (C27), 28.9–29.3 (CH2 fatty chain), 31.5 (C38), 54.39 (CH), 57.0–
57.3 (C11), 61.2 (C41), 64.0 (CH2), 65.8 (CH2), 68.1 (CH), 68.3 (C8), 68.7
(CH2), 69.3 (CH2), 69.9 (CH), 69.9 (CH2), 70.8 (CH), 71.5 (CH2), 71.6
(CH), 71.9 (CH), 72.4 (CH), 73.3 (CH), 74.4 (CH), 75.9 (CH), 76.1 (CH),
78.9 (C9), 98.6–98.7 (C17), 103.4–103.5 ppm (C2); MS (MALDI-TOF):
m/z calcd for C32H62O13 [M+Na]: 677.408, found: 677.308.
3-(hexadecyloxy)-2-methoxypropyl-6-O-(2,3,4,6-tetra-O-acetyl-b-
d-galactopyranosyl)-b-d-galactopyranoside 13: Compound
8
(221 mg, 0.31 mmol) and compound 12 (161 mg, 0.33 mmol) were
dissolved in dry CH2Cl2 (3 mL), and the mixture was stirred for 1 h
at room temperature. The mixture was then cooled to ꢀ808C,
TMSOTf (3 mL, 15 mmol) was added, and the mixture was stirred for
2 h at room temperature. The mixture was then filtered and con-
centrated. The residue was dissolved in MeOH (10 mL) and Dowex
50WX8-200 (200 mg) was added and the mixture was stirred for
1 h at room temperature. The mixture was filtered and concentrat-
ed to yield a beige oil which was purified by chromatography on
silica gel (CH2Cl2 with 0 to 10% MeOH) to give the expected prod-
phy on silica gel (CH2Cl2/MeOH 8:2). ½aꢂ20 = +50 (c=0.42, MeOH);
D
Rf (CHCl3/MeOH 8:2): 0.25; IR (ATR): 1031 (CꢀO), 2850–2916 (CꢀH),
3337 cmꢀ1 (OꢀH); 1H NMR (400.133 MHz, CD3OD): d=0.85 (t, 3H,
3JHH =7.5, CH3), 1.18–1.35 (m, 26H, CH2 fatty chain), 1.51 (qt, 2H,
3JHH =7.0, CH2 b fatty chain), 3.35–3.86 (m, 19H, CH2 a fatty chain+
CH2 sn-1+CH2 sn-3+CH sn-2+H2 +H3 +H4 +H5 +H6 +H2’ +H3’ +
H4’ +H5’ +H6’), 3.42 (s, 3H, CH3O), 4.29 (d, 1H, 3JHH =7.5, H1’),
4.72 ppm (m, 1H, H1); 13C NMR (125.771 MHz, CD3OD): d=14.4
(CH3), 23.7 (CH2 fatty chain), 27.2 (CH2 fatty chain), 30.4 (CH2 fatty
chain), 30.6 (CH2 fatty chain), 30.7 (CH2 fatty chain), 33.0 (CH2 fatty
chain), 58.4 (OCH3), 61.8 (C6), 62.5 (C6’), 68.3 (CH2 sn-3), 71.1 (CH2
sn-1), 72.7 (CH2 a fatty chain), 70.3, 72.2, 72.6, 73.4, 73.5, 74.9, 77.1
(C2 +C3 +C5 +C2’ +C3’ +C4’ +C5’), 80.6 (CH sn-2), 80.9 (C4), 100.5 (C1),
105.1 ppm (C1’); MS (MALDI-TOF): m/z calcd for C32H62O13 [M+Na]:
677.408, found: 677.386.
1
3
uct (41%). H NMR (400.133 MHz, CDCl3): d=0.84 (t, 3H, JHH =7.0,
CH3), 1.10–1.40 (m, 26H, CH2 fatty chain), 1.52 (m, 2H, CH2 b fatty
chain), 1.95–2.12 (12H, 4OAc), 3.35–3.70 (m, 12H), 3.70–4.05 (m,
5H), 4,11 (d, 2H, 3JHH =6.0), 4.20 (t, 1H, 3JHH =7.0), 4.56 (dd, 1H,
3JHH =6.4, 3JHH =8.0), 4.92–5.08 (m, 1H), 5.10–5.20 (m, 1H),
3
5.35 ppm (d, 1H, JHH =3.0).
1,2,3,4,6-pentakis-O-(trimethylsilyl)-a-d-glucopyranose 9: To
a stirred solution of glucose (1.8 g, 10.0 mmol) and Et3N (7.6 mL,
55.0 mmol) in dry DMF (60 mL) at 08C was added dropwise TMSCl
(7.0 mL, 55.0 mmol). The mixture was stirred for 4 h at room tem-
perature. n-Pentane (150 mL) was added, and the mixture was
3-(hexadecyloxy)-2-methoxypropyl-6-O-b-d-galactopyranosyl-b-
d-galactopyranoside 3: To a stirred solution of 13 (100 mg,
134 mmol) in dry MeOH was added a small piece of sodium and
the mixture was stirred for 1 h at room temperature. Then 100 mg
of Amberlite IR-120Hꢄ was added, and the mixture was stirred for
ChemMedChem 2016, 11, 1 – 10
www.chemmedchem.org
7
ꢃ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
30 min at room temperature. The mixture was filtered and concen-
trated. The residue was purified by chromatography on silica gel
(CH2Cl2 with 5 to 100% MeOH) to obtain the expected compound
as a white solid (36%). 1H NMR (500.132 MHz, CDCl3/CD3OD): d=
0.84 (t, 3H, 3JHH =7.0, CH3), 1.10–1.40 (m, 26H, CH2 fatty chain),
1.53 (qt, 2H, 3JHH =7.0, CH2 b fatty chain), 3.37–4.08 (m, 22H,
OCH3 +CH2 a fatty chain+CH2 sn-1+CH2 sn-3+CH sn-2+H3 +
counted. Four independent experiments were performed in paral-
lel, and these were done in triplicate.
Constitutive Ca2+ entry measurements: This protocol was previously
validated in MDA-MB-435s cells as described earlier.[19]
Caveolae and non-caveolae membrane fractionation: The membrane
fractions were extracted by using a detergent-free method and
separated by a sucrose gradient performed by ultracentrifugation.
Briefly, cells were lysed with sodium carbonate (pH 11.0), and the
lysate was then adjusted to form a 45% sucrose solution by
mixing with equal volumes of 90% sucrose prepared in MES-buf-
fered saline (MBS: 25 mm MES, pH 6.5, 0.15m NaCl), and placed at
the bottom of an ultracentrifuge tube. A 5–35% discontinuous su-
crose gradient was formed above (1.5 mL of 5% sucrose/1.5 mL of
35% sucrose; both in MBS containing 250 mm sodium carbonate)
and centrifuged at 268000 g for 6 h at 48C in a TLS 55 rotor (Opti-
maTM MAX-XP ultracentrifuge, Beckman Instruments). A total of
nine fractions (500 mL each) were collected from bottom to top of
the tube. Equal volumes of fractions were loaded onto a 10% SDS
polyacrylamide gel, and caveolin 1 (antibody 3228, cell signaling
technology) cells was used as a marker for the identification of cav-
eolae fractions (C), and b-adaptin (antibody sc6425, Santa Cruz Bio-
technology) was used as a marker for non-caveolae fractions (NC).
Then, NCF were pooled as CF, loaded in equal protein quantities
on a 10% SDS polyacrylamide gel and visualized by western blot
as described previously.[19]
3
H4’ +H5’ +H2 +H4 +H5 +H6 +H2’ +H3’ +H6’), 4.19 (dd, 1 H, JHH =16.0,
3JHH =7.0, H1), 4.27 ppm (d, 1 H, 3JHH =7.0, H1’); 13C NMR
(125.771 MHz, CDCl3/CD3OD): d=14.3 (C40/85), 23.2 (C39/84), 26.6 (C27/
72), 29.8–30.2 (CH2 fatty chain), 32.4 (C38/83), 57.9 (C56), 58.1 (C11),
61.7 (C86), 61.9 (C41), 68.2 (C15/60), 68.7 (CH), 69.0 (C8/53), 69.4–69.5
(C3/48), 70.4–70.9 (C23/68), 71.8 (C4/49), 72.3 (C25/70), 73.7 (CH), 74.0–74.1
(C6/51),75.6 (C64), 75.7(C19), 79.8 (C9), 79.9 (C54), 104.2 (C47), 104.3 (C2),
104.4 ppm (C17/62); MS (MALDI-TOF): m/z calcd for C32H62O13 [M+
Na]: 677.408, found: 677.328.
Biological assays
CMC determination: Critical micellar concentration was measured at
228C by mixing increasing concentrations of each alkyl-lipid in PSS
buffer at pH 7.4 with the hydrophobic dye Merocyanine 540 (final
concentration 4 mm). Absorption spectra from the samples were
measured in a spectrophotometer, and CMC was determined by an
increase of ~25 nm in the absorption maximum wavelength (lmax
)
of the dye upon merocyanine entrapping into the ether lipid mi-
celles.
Statistics: Statistical analyses were performed using SigmaStat Soft-
ware (version 3.0.1a, Systat Software, Inc). Unless otherwise indicat-
ed, data are expressed as the mean ꢁSEM (N: number of experi-
ments; n: number of cells from independent experiments). For
comparison between more than two means, we used Kruskal–
Wallis one-way analysis of variance followed by Dunn’s or Dunnet’s
post hoc tests as appropriate. Comparisons between two means
were made using Mann–Whitney. Differences were considered sig-
nificant when p<0.05.
Cell culture: HEK293T cells were obtained from the American Type
Culture Collection (ATCC), maintained in Dulbecco’s modified
Eagle’s medium, supplemented with 10% (v/v) fetal bovine serum
(Lonza, France), 1 mm sodium pyruvate (Sigma–Aldrich, France),
and 1% non-essential amino acids (Sigma–Aldrich, France), and
were grown under a humidified atmosphere at 378C (95% air, 5%
CO2). HEK293T cells were transduced as previously described[9,17] to
obtain a stable expression of homotetrameric SK3 and SK2 chan-
nels.
Patch-clamp experiments: Experiments were performed in the
whole-cell configuration of the patch-clamp technique. The extra-
cellular solution contained 140 mm NaCl, 4 mm KCl, 2 mm CaCl2,
1 mm MgCl2, 10 mm HEPES, 0.33 mm NaH2PO4, and 11.5 mm d-glu-
cose. The solution was adjusted to pH 7.4 with NaOH. The intracel-
lular solution contained 145 mm KCl, 1 mm MgCl2, 1 mm Mg-ATP,
0.7 mm CaCl2, 1 mm EGTA, and 10 mm HEPES (pCa 6). The solution
was adjusted to pH 7.2 with KOH. Patch-clamp experiments and
data analyses were conducted as previously described.[9,17] The ef-
fects of compounds on HEK293T-SK3 and HEK293-SK2 cells were
measured using a ramp protocol from +100 mV to ꢀ100 mV with
a holding potential of 0 mV (500 ms duration, 4 s intervals) to inac-
tive endogenous potassium currents. Potassium SK3 and SK2 cur-
rents were analyzed at 0 mV to minimize chloride currents (EClꢀ:
0 mV). The amplitude of the apamin-sensitive current was obtained
by subtraction of the amplitude of the current before and after ap-
plication of 100 nm apamin (a specific SKCa blocker).
Acknowledgements
We thank Canceropole Grand Ouest, La Nationale Contre le
Cancer, and the network Ion Channel-CGO for financial support.
We thank the University of Brest (Services communs RMN-RPE
and mass spectrometry) for technical assistance. This work was
also funded by the University of Tours, INSERM, the Rꢁgion Centre
of France (LIPIDS project of ARD2020 Biomꢁdicaments), the Ligue
Nationale Contre le Cancer (committees of Charente, Indre et
Loire, Mayenne and Morbihan), the Association “CANCEN”, and
Tours’ Hospital Oncology Association ACORT. M.G. and Y.F. held
fellowships from the Rꢁgion Centre of France. We thank Aurore
Douaud-Lecaille, Morgan Aubry, and Isabelle Domingo for techni-
cal assistance, and Catherine Leroy for secretarial support.
In vitro cell viability and cell migration assay: Cell viability was de-
termined using the tetrazolium salt reduction method (MTT) as
previously described.[16] Cell migration assays were performed in
24-well plates with polyethylene terephthalate membrane cell cul-
ture inserts with a pore size of 8 mm (Becton Dickinson, France) as
previously described.[16] When a compound was evaluated, it was
added into both compartments. After 24 h, stationary cells were re-
moved from the top side, whereas migrated cells in the bottom
side of the inserts were fixed with DAPI (Sigma–Aldrich) and then
Keywords: cell migration · glycolipids · metastases · ohmline ·
potassium channels
[1] R. Martꢆnez, L. Chacꢇn-Garcꢆa, Curr. Med. Chem. 2005, 12, 127–151.
[2] K. E. Gascoigne, S. S. Taylor, J. Cell Sci. 2009, 122, 2579–2585.
[3] S. Bellou, G. Pentheroudakis, C. Murphy, T. Fotsis, Cancer Lett. 2013, 338,
219–228.
[4] M. Whittaker, C. D. Floyd, P. Brown, A. J. Gearing, Chem. Rev. 2001, 101,
2205–2206.
&
ChemMedChem 2016, 11, 1 – 10
www.chemmedchem.org
8
ꢃ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
[5] A. Dufour, C. M. Overall, Trends Pharmacol. Sci. 2013, 34, 233–242.
[6] J. J. Body, S. Casimiro, L. Costa, Nat. Rev. Urol. 2015, 12, 340–356.
[7] M. R. Smith, F. Saad, R. Coleman, N. Shore, K. Fizazi, B. Tombal, K. Miller,
P. Sieber, L. Karsh, R. Dami¼o, T. L. Tammela, B. Egerdie, H. Van Poppel, J.
Chin, J. Morote, F. Gꢇmez-Veiga, T. Borkowski, Z. Ye, A. Kupic, R. Dansey,
C. Goessl, Lancet Lond. Engl. 2012, 379, 39–46.
[8] M. Potier, V. Joulin, S. Roger, P. Besson, M.-L. Jourdan, J.-Y. Leguennec, P.
Bougnoux, C. Vandier, Mol. Cancer Ther. 2006, 5, 2946–2953.
[9] A. Chantome, A. Girault, M. Potier, C. Collin, P. Vaudin, J.-C. Pagꢂs, C.
Vandier, V. Joulin, Exp. Cell Res. 2009, 315, 3620–3630.
[10] M. Potier, T. A. Tran, A. Chantome, A. Girault, V. Joulin, P. Bougnoux, C.
Vandier, F. Pierre, Biochem. Biophys. Res. Commun. 2010, 397, 42–47.
[11] M. Guꢀguinou, A. Chantꢁme, G. Fromont, P. Bougnoux, C. Vandier, M.
Potier-Cartereau, Biochim. Biophys. Acta Mol. Cell Res. 2014, 1843, 2322–
2333.
[12] D. Schmidt, Q.-X. Jiang, R. MacKinnon, Nature 2006, 444, 775–779.
[13] S. B. Hansen, Biochim. Biophys. Acta Mol. Cell Biol. Lipids 2015, 1851,
620–628.
F. Arbion, J.-P. Haelters, P.-A. Jaffrꢂs, M. Pinault, P. Besson, V. Joulin, P.
Bougnoux, C. Vandier, Cancer Res. 2013, 73, 4852–4861.
[20] C. M. Sevrain, J.-P. Haelters, A. Chantꢁme, H. Couthon-Gourvꢂs, A. Gir-
ault, C. Vandier, P.-A. Jaffrꢂs, MedChemComm 2012, 3, 1471–1478.
[21] C. M. Sevrain, J.-P. Haelters, A. Chantꢁme, H. Couthon-Gourvꢂs, M. Gue-
guinou, M. Potier-Cartereau, C. Vandier, P.-A. Jaffrꢂs, Org. Biomol. Chem.
2013, 11, 4479–4487.
[22] T. Uchiyama, O. Hindsgaul, Synlett 1996, 499–501.
[23] S. Honda, T. Ichii, K. Kakehi, Carbohydr. Res. 1983, 115, 95–104.
[24] I. Tvaroska, F. R. Taravel, Carbohydr. Res. 1991, 221, 83–94.
[25] W. Du, S. S. Kulkarni, J. Gervay-Hague, Chem. Commun. 2007, 2336–
2338.
[26] M. Schombs, F. E. Park, W. Du, S. S. Kulkarni, J. Gervay-Hague, J. Org.
Chem. 2010, 75, 4891–4898.
[27] A. S. Bhat, J. Gervay-Hague, Org. Lett. 2001, 3, 2081–2084.
[28] M. Guꢀguinou, A. Gambade, R. Fꢀlix, A. Chantꢁme, Y. Fourbon, P. Boug-
noux, G. Weber, M. Potier-Cartereau, C. Vandier, Biochim. Biophys. Acta
Biomembr. 2015, 1848, 2603–2620.
[14] S. Marra, R. Ferru-Clꢀment, V. Breuil, A. Delaunay, M. Christin, V. Friend,
S. Sebille, C. Cognard, T. Ferreira, C. Roux, L. Euller-Ziegler, J. Noel, E. Lin-
gueglia, E. Deval, EMBO J. 2016, 35, 414–428.
[15] A. Girault, J.-P. Haelters, M. Potier-Cartereau, A. Chantꢁme, P.-A. Jaffrꢀs,
P. Bougnoux, V. Joulin, C. Vandier, Curr. Med. Chem. 2012, 19, 697–713.
[16] M. Potier, A. Chantome, V. Joulin, A. Girault, S. Roger, P. Besson, M.-L.
Jourdan, J.-Y. LeGuennec, P. Bougnoux, C. Vandier, Br. J. Pharmacol.
2011, 162, 464–479.
[29] M. L. Sinnott, Chem. Rev. 1990, 90, 1171–1202.
[30] W. J. Baumann, H. K. Mangold, J. Org. Chem. 1964, 29, 3055–3057.
[31] J. J. Godfroid, C. Broquet, S. Jouquey, M. Lebbar, F. Heymans, C. Re-
deuilh, E. Steiner, E. Michel, E. Coeffier, J. Med. Chem. 1987, 30, 792–
797.
[32] S. Chittaboina, B. Hodges, Q. Wang, Lett. Org. Chem. 2006, 3, 35–38.
[33] J. M. de La Fuente, S. Penadꢀs, Tetrahedron: Asymmetry 2002, 13, 1879–
1888.
[17] A. Girault, J.-P. Haelters, M. Potier-Cartereau, A. Chantome, M. Pinault, S.
Marionneau-Lambot, T. Oullier, G. Simon, H. Couthon-Gourves, P.-A.
Jaffres, B. Corbel, P. Bougnoux, V. Joulin, C. Vandier, Curr. Cancer Drug
Targets 2011, 11, 1111–1125.
[34] N. K. Kochetkov, O. S. Chizhov, N. V. Molodtsov, Tetrahedron 1968, 24,
5587–5593.
[18] G. Arthur, R. Bittman, Anti-Cancer Agents Med. Chem. 2014, 14, 592–
606.
[19] A. Chantꢁme, M. Potier-Cartereau, L. Clarysse, G. Fromont, S. Mario-
nneau-Lambot, M. Guꢀguinou, J.-C. Pagꢂs, C. Collin, T. Oullier, A. Girault,
Received: March 11, 2016
Revised: May 9, 2016
Published online on && &&, 0000
ChemMedChem 2016, 11, 1 – 10
www.chemmedchem.org
9
ꢃ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

FULL PAPERS
W. Berthe, C. M. Sevrain, A. Chantꢂme,
A. M. Bouchet, M. Gueguinou, Y. Fourbon,
M. Potier-Cartereau, J.-P. Haelters,
H. Couthon-Gourvꢀs, C. Vandier,
P.-A. Jaffrꢀs*
Mitigating metastases: Among the
new disaccharide-based ether lipids de-
veloped in this study, compound 2 was
found to exhibit outstanding selective
inhibition of SK3 ion channel activity, in-
ducing a decrease in SK3-dependent
cell migration. This modulation is ex-
plained by a displacement of the SK3
protein outside caveolae.
&& – &&
New Disaccharide-Based Ether Lipids
as SK3 Ion Channel Inhibitors
&
ChemMedChem 2016, 11, 1 – 10
www.chemmedchem.org
10
ꢃ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!