A Novel Phenylchromane Derivative Increases the Rate of Glucose Uptake in L6 Myotubes and Augments Insulin Secretion from Pancreatic Beta-Cells by Activating AMPK

Article abstract of DOI:10.1007/s11095-017-2271-7

Purpose: A series of novel polycyclic aromatic compounds that augment the rate of glucose uptake in L6 myotubes and increase glucose-stimulated insulin secretion from beta-cells were synthesized. Designing these molecules, we have aimed at the two main pathogenic mechanisms of T2D, deficient insulin secretion and diminished glucose clearance. The ultimate purpose of this work was to create a novel antidiabetic drug candidate with bi-functional mode of action. Methods: All presented compounds were synthesized, and characterized in house. INS-1E cells and L6 myoblasts were used for the experiments. The rate of glucose uptake, mechanism of action, level of insulin secretion and the druggability of the lead compound were studied. Results: The lead compound (6-(1,3-dithiepan-2-yl)-2-phenylchromane), dose- and time-dependently at the low μM range increased the rate of glucose uptake in L6 myotubes and insulin secretion in INS-1E cells. The compound exerted its effects through the activation of the LKB1 (Liver Kinase B1)-AMPK pathway. In vitro metabolic parameters of this lead compound exhibited good druggability. Conclusions: We anticipate that bi-functionality (increased rate of glucose uptake and augmented insulin secretion) will allow the lead compound to be a starting point for the development of a novel class of antidiabetic drugs.

Full text of DOI:10.1007/s11095-017-2271-7

Pharm Res
DOI 10.1007/s11095-017-2271-7
RESEARCH PAPER
A Novel Phenylchromane Derivative Increases the Rate
of Glucose Uptake in L6 Myotubes and Augments Insulin
Secretion from Pancreatic Beta-Cells by Activating AMPK
Naomi Rozentul1 & Yosef Avrahami2 & Moran Shubely1 & Laura Levy1 & Anna Munder1
Guy Cohen2 & Erol Cerasi3 & Shlomo Sasson2 & Arie Gruzman1
&
Received: 27 June 2017 /Accepted: 24 September 2017
#
Springer Science+Business Media, LLC 2017
ABSTRACT
μM range increased the rate of glucose uptake in L6 myotubes
and insulin secretion in INS-1E cells. The compound exerted
its effects through the activation of the LKB1 (Liver Kinase
B1)-AMPK pathway. In vitro metabolic parameters of this lead
compound exhibited good druggability.
Conclusions We anticipate that bi-functionality (increased
rate of glucose uptake and augmented insulin secretion) will
allow the lead compound to be a starting point for the devel-
opment of a novel class of antidiabetic drugs.
Purpose A series of novel polycyclic aromatic compounds
that augment the rate of glucose uptake in L6 myotubes and
increase glucose-stimulated insulin secretion from beta-cells
were synthesized. Designing these molecules, we have aimed
at the two main pathogenic mechanisms of T2D, deficient
insulin secretion and diminished glucose clearance. The ulti-
mate purpose of this work was to create a novel antidiabetic
drug candidate with bi-functional mode of action.
Methods All presented compounds were synthesized, and
characterized in house. INS-1E cells and L6 myoblasts were
used for the experiments. The rate of glucose uptake, mecha-
nism of action, level of insulin secretion and the druggability of
the lead compound were studied.
.
.
KEY WORDS AMPK activation glucose uptake insulin
.
.
secretion LKB1 activation phenylchromane scaffold
Results The lead compound (6-(1,3-dithiepan-2-yl)-2-
phenylchromane), dose- and time-dependently at the low
ABBREVIATIONS
[3H]dGlc 1-Tritium-2-deoxyglucose
ACC
Acetyl-CoA carboxylase
AICAR
5-Aminoimidazole-4-carboxamide-1-β-D-
ribofuranoside
Naomi Rozentul and Yosef Avrahami contributed equally to this work.
Akt/PKB Akt/Protein kinase B
Electronic supplementary material The online version of this article
(https://doi.org/10.1007/s11095-017-2271-7) contains supplementary
material, which is available to authorized users.
AMPK
AS160
BDA
AMP kinase
AMPK substrate of phosphorylation 160 kDA
Benzaldehyde dimethyl acetal
Cytochrome P450
* Shlomo Sasson
CYP
shlomo.sasson@mail.huji.ac.il
DCM
DEAD
DMF
Dichloromethane
Diethyl azodicarboxylate
Dimethrylformamide
* Arie Gruzman
gruzmaa@biu.ac.il
GLUT4
GSIS
hERG
HRP
LKB1
mGPD
OPD
Glucose transporter type 4
Glucose-stimulated insulin secretion
BHuman ether-à-go-go^-related gene channels
Horseradish peroxidase
Liver Kinase B1
Mitochondrial glycerophosphate dehydrogenase
O-phenylenediamine
1
Division of Medicinal Chemistry, Department of Chemistry, Faculty of
Exact Sciences, Bar-Ilan University, 52900 Ramat-Gan, Israel
2
Department of Pharmacology, Institute for Drug Research, School of
Pharmacy, Faculty of Medicine, The Hebrew University of
Jerusalem, 91120 Jerusalem, Israel
3
The Endocrinology and Metabolism Service, Department of Medicine
Hadassah-Hebrew University Medical Center, 91120 Jerusalem, Israel

N. Rozentul et al.
PCC
PMSF
QTof
T2D
THF
TMS
Pyridinium Chlorochromate
Phenylmethanesulfonylfluoride
Quadrupole time of-flight mass spectrometer
Type 2 diabetes
Tetrahydrofuran
Tetramethylsilane
in a non-insulin-dependent manner (25) supports the hypoth-
esis that AMPK activation should be a major target for de-
signing novel and potent antidiabetic drugs (14,26,27).
LKB1 controls several key functions also in beta-cells
(28)], which are related to morphological characteristics
such as polarity, cell size and total beta-cell mass in islets
of Langerhans (29). Glucose-stimulated insulin secretion
(GSIS) may also be regulated by this kinase (29,30).
Several studies reported that LKB1 activation downregu-
lated insulin secretion (31–33). Studies on the activation
of AMPK in beta-cells also produced conflicting results
(34,35). Some showed that AMPK activation decreased
insulin secretion and promoted apoptosis (36–38), while
others attested to opposite effects (39–41). For example,
AMPK activators increased GSIS from rat insulinoma
cells and isolated rat islets (42). Moreover, mice lacking
or expressing inactive AMPK forms in beta-cells exhibited
augmented peripheral insulin sensitivity and limited GSIS
capacity that was associated with pathological alterations
in beta-cell morphology (43). The reason for these con-
flicting results is not clear, and can be attributed to dif-
ferent in vitro and in vivo biological models that were used
in these studies..
In our previous studies we showed that D-xylose derivatives
(3-O-methyl- 2,4-benzylidene-D-xylose diethyl dithioacetal, 2,4-
benzylidene-D-xylose diethyl dithioacetal, 2,4:3,5-di-
benzylidene-D-xylose diethyl dithioacetal) augmented glucose
uptake in skeletal muscles in a non-insulin-dependent manner
by activating AMPK (44,45). In addition, using molecular
modelling we also designed and synthesized two AMPK activa-
tors based on the benzothiazole core; 2-(benzo[d]thiazol-2-
y lm e t h y l t h i o) - 6 -e t h o x y b e n z o [ d ] t h i a z o l e a n d
2-(propylthio)benzo-[d]thiazol-6-ol These compounds stimulat-
ed glucose uptake in skeletal muscles and also increased GSIS
from cultured beta-cells and isolated rat islets of Langerhans:
(42,46). However, all these compounds were protected under
different patents. Thus, to accelerate the drug development pro-
cess and to interest pharmaceutical companies, we developed
novel, patentable compounds, which belong to a new class of
bi-functional AMPK activators. In this work we show that mol-
ecules with a phenylchromane core augmented glucose uptake
in myotubes and increased GSIS from beta-cells.
TMSCl
To s C l
Trimethylsilyl chloride
Toluenesulfonyl chloride
INTRODUCTION
Despite recent intensive efforts to find a cure and/or optimal
treatment for type 2 diabetes (T2D), the almost one-century old
metformin remains the first-line oral antidiabetic drug (1).
However, metformin is not an ideal drug: although it is well-
tolerated by patients and exerts efficient blood glucose lowering
effect, normoglycemia is not restored, and it does not prevent all
late complications of diabetes (2). Therefore, the need for novel,
more effective and potent antidiabetic drugs is still intense.
Skeletal muscles are the main consumers of glucose (3).
During exercise or upon insulin stimulation glucose transport-
er type 4 (GLUT4) translocates from the cytoplasm to the
sarcolemma and T-tubules, thus, augmenting the rate of glu-
cose influx, and consequently lowering blood glucose levels (4).
T2D patients suffer from insulin-resistance in skeletal muscles;
however this does not affect contraction-mediated GLUT4
recruitment from intracellular storage compartments to the
plasma membrane (5). This pathway is regulated by the key
energy sensor enzyme AMPK (6–8). LKB1 is a major up-
stream stimulatory kinase that phosphorylates the α-subunit
of AMPK at Thr172 (9,10). Several compounds, such as ber-
berine (11), silibinin (12) and ursolic acid (13) were shown to
have beneficial effects on glucose and lipid homeostasis by
activating LKB1. Yet, other molecules show similar effects
by direct activation of AMPK and not its upstream regulator
LKB1 (14–16). It is suggested that metformin exerts its phar-
macological effects by activating AMPK; yet, other mecha-
nisms were also proposed. For example, inhibition of
gluconeogenic gene expression and allosteric regulation of
key gluconeogenesis enzymes in the liver, direct regulation of
miRNAs, which modulate several downstream genes in met-
abolic or preoncogenic pathways (17,18). Also, direct inhibi-
tion of mGPD (mitochondrial glycerophosphate dehydroge-
nase), resulting in altered cellular redox state (19) was pro-
posed as metformin’s mechanism of action. Direct activators
of AMPK such as A769662, PT-1, 5-aminoimidazole-4-
carboxamide-1-β-D-ribofuranoside (AICAR),
2-(benzo[d]thiazol-2-ylmethylthio)-6-ethoxybenzo[d]thiazole
and others, have been discovered or synthesized, but none
could be developed into antidiabetic drugs (4,14,15,20–24).
The fact that exercise is a potent stimulus for increasing skel-
etal muscle GLUT4 expression, trafficking and glucose uptake
MATERIALS AND METHODS
Organic solvents (HPLC grade) were from Frutarom Ltd.
(Haifa, Israel). Dry THF was obtained using distillation from
a boiled blue color mix with sodium/benzophenone. The
melting points were determined with Fisher-Johns melting
1
point apparatus (Palmerton, PA). The H NMR and 13C
NMR spectra were recorded at room temperature on a
Bruker Advance NMR spectrometer (Vernon Hills, IL)

Bi-functional antidiabetic drug candidate
operating at 200 and 300 MHz and were in accord with the
assigned structures. Chemical shift values were reported rela-
tive to TMS that was used as an internal standard. The sam-
ples were prepared by dissolving the synthesized compounds
in DMSO-d6 (δH = 2.50 ppm, δC = 39.52 ppm) or in CDCl3
(δH = 7.26 ppm, δC = 77.16 ppm) (47). Chemical shifts were
expressed in δ (ppm) and coupling constants (J) in hertz. The
splitting pattern abbreviations are as follows: s, singlet; d, dou-
blet; t, triplet; q, quartet; quint, quintet; m, unresolved multi-
plet due to the field strength of the instrument; dd, doublet of
doublet. A QTof micro spectrometer (Micromass, Milford,
MA) in the positive ion mode was used for mass spectrometry.
Data were processed using massLynX v.4.1 calculation and
deconvolution software (Waters Corporation, Milford, MA).
Column chromatography was performed on Merck Silica gel
60 (230–400 mesh; Merck, Darmstadt, Germany). Analytical
and preparative HPLC (Young Lin Instruments, Anyang,
Korea) were performed on LUNA C18(2) preparative
(10 μm, 100 x 30 mm) or analytical (5 μm, 250 x 4.6 mm)
columns, both from Phenomenex Inc. (Torrance, CA).
Acetonitrile and doubly distilled water were used as an eluent
in different ratios. Analytical thin layer chromatography was
carried out on pre-coated Merck Silica gel 60F254 (Merck)
sheets using UV absorption and iodine physical adsorption for
visualization. The ≥95% purity of the final compounds was
confirmed using HPLC analysis. All analytical data (including
the NMR images) is shown in Supportive Information.
Human insulin (Actrapid) was purchased from Novo
Nordisk (Bagsvaerd, Denmark). AICAR, BSA, 2-deoxy-D-
glucose (dGlc), D-glucose, O-phenylenediamine (OPD), and
the protease inhibitor cocktail were purchased from Sigma-
Aldrich Chemicals (Rehovot, Israel). Glycerol and sodium
fluoride were from Merck (Whitehouse Station, NJ).
Mercaptoethanol, phenylmethanesulfonylfluoride (PMSF),
sodium orthovanadate, sodium-β-glycerophosphate, sodium
pyrophosphate and SDS were purchased from Alfa Aesar
(Ward Hill, MA). American Radiolabeled Chemicals (St.
Louis, MO) supplied [3H]dGlc [2.22 TBq/mmol (60 Ci/
mmol)]. Antibodies against AMPKα, pThr172-AMPK,
LKB1, pSer438-LKB1, pSer437-AKT, ACC, pSer79-ACC
AS-160 and pThr642-AS160 were purchased from Cell
Signaling Technology (Beverly, MD). The α-tubulin was from
Millipore (Billerica, MA). The Anti-c-Myc (A-14) antibody
was from Santa Cruz Biotechnology (Santa Cruz, CA), horse-
radish peroxidase (HRP)-conjugated anti-rabbit IgG and EZ-
ECL chemoluminescence detection kit were from Jackson
ImmunoResearch (West Grove, PA). Goat serum, fetal calf
serum (FCS), L-glutamine, α-minimal essential medium
(MEM) and antibiotics were purchased from Biological
Industries (Beth-Haemek, Israel). P450-Glo™ Assay Systems
was purchased from Promega (Madison, WI, USA).
PredictorTM hERG Fluorescence Polarization Assay kit was
purchased from Invitrogen (Carlsbad, CA, USA).
SYNTHETIC PROCEDURES
3-chloro-1-phenylpropan-1-one (1) was synthesized as de-
scribed (48).
The title compound was obtained starting from 3-
chloropropionyl chloride (yellow oil, 70%): 1H NMR
(200 MHz, CDCl3): δ = 3.45 (t, J = 6.8 Hz, 2H), 3.92 (t,
1
J = 6.8 Hz, 2H), 7.47 (m, 2H), 7.59 (m, H), 7.93 ppm (m,
2H); 13C NMR (50.32 MHz, CDCl3): δ = 38.71, 41.32,
128.08, 128.78, 133.59, 136.40, 196.74 ppm; MS (ESI)
m/z: 169 [M + H].
3-chloro-1-phenylpropan-1-ol (2) was synthesized as de-
scribed (48).
The title compound was obtained starting from 1 (yellow
1
oil, 97%): H NMR (200 MHz, CDCl3): δ = 2.05 (m, 1H),
2.20 (m, 1H), 3.51 (m, 1H), 3.68 (m, 1H), 4.88 (q, J = 4.8 Hz,
1H), 7.32 ppm (m, 5H); 13C NMR (50.32 MHz, CDCl3):
δ = 41.29, 41.58, 71.15, 125.64, 127.76, 128.51,
143.55 ppm; MS (CI) m/z: 171 [M + H].
1-(3-chloro-1-phenylpropoxy)-2, 4-dibromobenzene (3)
was synthesized as described (49).
The title compound was obtained starting from 2 (white
solid, 54%): 1H NMR (300 MHz, CDCl3): δ = 2.24 (m, 1H),
2.53 (m, 1H), 3.64 (m, 1H), 3.89 (m, 1H), 5.39 (q, J = 4.5 Hz,
1H), 6.57 (d, J = 9 Hz, 1H), 7.15 (dd, J = 9 Hz, J = 3 Hz, 1H),
7.34 (m, 5H), 7.64 ppm (d, J = 3 Hz, 1H); 13C NMR
(75.49 MHz, CDCl3): δ = 41.24, 78.22, 113.32, 113.59,
116.23, 125.88, 128.33, 128.99, 131.06, 135.49, 139.67,
153.47 ppm; MS (ESI) m/z: 404 [M].
3, 4-dihydro-2-phenyl-2H-chromene-6-carbaldehyde (4)
was synthesized as described (50).
The title compound was obtained starting from 3 (yellow
solid, 25%): 1H NMR (300 MHz, CDCl3): δ = 2.28 (m, 2H),
2.90 (m, 1H), 3.04 (m, 1H), 5.17 (q, J = 3 Hz, 1H), 7.02 (d,
J = 8.4 Hz, 1H), 7.40 (m, 5H), 7.66 (m, 2H), 9.86 ppm (s, 1H);
13C NMR (75.49 MHz, CDCl3): δ = 41.23, 41.32, 116.29,
126.02, 128.48, 129.20, 130.30, 132.06, 139.95, 163.08,
190.93 ppm; MS (ESI) m/z: 239 [M + H].
General Procedure for the Preparation of (5, 6, 7, 8, 9, 12,
13, 14, 15, 19) (51).
To a well-stirred solution of 10 mmol of 4 or 11 and
20 mmol of corresponding thiol (or 10 mmol dithiol) in chlo-
roform at r. t. 12–15 mmol TMSCl was added slowly over a
period of 20 min. After an hour, the reaction mixture was
washed with a 5% aqueous sodium carbonate solution and
chloroform phase was washed by water. The organic layer
was dried by Na2SO4, filtered and the chloroform was evap-
orated. The residue amorphic solids were purified by HPLC
or by regular silica gel chromatography.
6-(1,3-dithian-2-yl)-2-phenylchroman (5):
The title compound was obtained starting from 4 (white solid,
3%): 1H NMR (300 MHz, CDCl3): δ = 2.02 (m, 4H), 2.85 (m,
6H), 5.16 (d, 1H), 5.64 (s, 1H), 7.33 ppm (m, 8H); 13C NMR

N. Rozentul et al.
(75.49 MHz, CDCl3): δ = 25.06, 25.11, 29.82, 41.37, 57.12,
76.69, 117.06, 121.93, 126.02, 127.19, 127.82, 128.58, 129.05,
132.37, 142.60, 156.06 ppm; MS (EI) m/z: 329 [M + H].
6-(1,3-dithiolan-2-yl)-2-phenylchroman (6):
J = 6 Hz, 2H), 4.96 (d, J = 14.4 Hz, 1H), 5.17 (d,
J = 14.4 Hz, 1H), 6.01 (s, 1H), 6.78 (s, 1H), 7.03 (s, 1H),
7.42 (m, 3H), 7.51 ppm (m, 2H); 13C NMR (75.49 MHz,
CDCl3): δ = 20.66, 61.10, 66.73, 99.0, 124.58, 126.30,
127.89, 128.51, 129.47 ppm; MS (CI) m/z: 256 [M].
General Procedure for the Preparation of (11) (52).
The 10 (300 mg, 1.17 mmol) and PCC (378 mg,
1.75 mmol) in DCM (50 ml) were stirred at r. t. for 2 h.
After the reaction, the mixture was purified by column chro-
matography (silica gel, DCM) to give pure compound 11.
6-methyl-2-phenyl-4H-benzo[d] [1, 3]dioxin-8-
carbaldehyde (11):
The title compound was obtained starting from 4 (colorless
1
oil, 16%): H NMR (300 MHz, CDCl3): δ = 2.17 (m, 2H),
2.85 (m, 2H), 3.43 (m 4H), 5.04 (q, J = 2.7 Hz, 1H), 5.61 (s,
1H), 6.85 (d, J = 8 Hz, 1H), 7.34 ppm (m, 7H); 13C NMR
(75.49 MHz, CDCl3): δ = 25.11, 29.86, 40.28, 56.27, 77.93,
117.05, 121.82, 126.02, 127.19, 127.92, 128.59, 129.06,
131.37, 141.60, 155.05 ppm; MS (EI) m/z: 314 [M].
6-(1,3-dithiepan-2-yl)-2-phenylchroman (7):
The title compound was obtained starting from 4 (white
solid, 13%): 1H NMR (300 MHz, CDCl3): δ = 2.12 (m, 6H),
2.98 (m, 6H), 5.03 (d, J = 2.6 Hz, 1H), 5.26 (s, 1H), 6.84 (d,
J = 8 Hz, 1H), 7.30 ppm (m, 7H); 13C NMR (75.49 MHz,
CDCl3): δ = 25.11, 29.86, 31.71, 32.71, 56.67, 77.90, 117.11,
121.93, 126.01, 126.62, 127.89, 128.49, 128.56, 132.47,
141.63, 154.91 ppm; MS (EI) m/z: 342 [M].
The title compound was obtained starting from 10 (white
solid, 94%): H NMR (300 MHz, CDCl3): δ = 2.32 (s, 3H),
1
5.00 (d, J = 15 Hz, 1H), 5.20 (d, J = 15 Hz, 1H), 6.09 (s, 1H),
7.07 (m, 1H), 7.45 (m, 3H), 7.59 (m, 3H), 10.44 ppm (s, 1H);
13C NMR (75.49 MHz, CDCl3): δ = 20.51, 66.56, 99.39,
121.57, 123.99, 126.33, 126.95, 128.60, 129.69, 130.67,
131.72, 136.45 ppm; MS (ESI) m/z: 277 [M + Na].
8-(1,3-dithian-2-yl)-6-methyl-2-phenyl-4H-benzo[d] [1,
3]dioxin (12):
The title compound was obtained starting from 11 (white
solid, 25%): 1H NMR (300 MHz, CDCl3): δ = 2.03 (m, 2H),
2.28 (s, 3H), 2.98 (m, 4H), 4.92 (d, J = 14.4 Hz, 1H), 5.12 (d,
J = 14.4 Hz, 1H), 5.59 (s, 1H), 6.00 (s, 1H), 6.75 (s, 1H), 7.31 (s,
1H), 7.44 (m, 3H), 7.65 ppm (m, 2H); 13C NMR (75.49 MHz,
CDCl3): δ = 20.64, 25.36, 32.24, 32.32, 43.05, 66.49, 99.16,
120.61, 125.34, 126.60, 126.70, 128.20, 128.44, 129.34,
130.91, 137.23, 146.98 ppm; MS (CI) m/z: 345 [M + H].
8-(1,3-dithiepan-2-yl)-6-methyl-2-phenyl-4H-benzo[d] [1,
3]dioxin (13):
The title compound was obtained starting from 11 (white
solid, 11%): 1H NMR (300 MHz, CDCl3): δ = 2.09 (m, 4H),
2.28 (s, 4H), 2.85 (m, 2H), 3.12 (m, 2H), 4.93 (d, J = 14.4 Hz,
1H), 5.13 (d, J = 14.4 Hz, 1H), 5.70 (s, 1H), 6.01 (s, 1H), 6.73
(s, 1H), 7.33 (s, 1H), 7.44 (m, 3H), 7.62 ppm (m, 2H); 13C
NMR (75.49 MHz, CDCl3): δ = 19.70, 30.61, 30.67, 31.56,
47.52, 65.48, 97.96, 123.81, 125.43, 126.61, 127.36,
128.22 ppm; MS (EI) m/z: 358 [M].
6-(bis(ethylthio)methyl)-2-phenylchroman (8):
The title compound was obtained starting from 4 (colorless
1
oil, 8%): H NMR (300 MHz, CDCl3): δ = 1.23 (t, J = 7.5,
6H), 2.15 (m, 2H), 2.51 (m, 4H), 2.85 (m, 2H), 4.88p (s, 1H),
5.04p (q, J = 2.6 Hz, 1H), 6.85 (d, J = 9 Hz, 1H), 7.16 (m, 2H),
7.35 ppm (m, 5H); 13C NMR (75.49 MHz, CDCl3):
δ = 14.40, 25.21, 26.36, 29.93, 52.07, 77.94, 116.90,
121.91, 126.03, 126.86, 127.92, 128.58, 128.69, 132.18,
141.64, 154.79 ppm; MS (CI) m/z: 344 [M].
6-(bis(propylthio)methyl)-2-phenylchroman (9):
The title compound was obtained starting from 4 (yellow oil,
9%): 1H NMR (300 MHz, CDCl3): δ = 0.97 (t, J = 7.4 Hz, 6H),
1.59 (m, 4H), 2.13 (m, 2H), 2.52 (m, 4H), 2.78 (m, 1H), 2.98 (m,
1H), 4.82 (s, 1H), 5.04 (q, J = 2.4 Hz, 1H), 6.85 (d, J = 9 Hz,
1H), 7.18 (m, 2H), 7.36 ppm (m, 5H); 13C NMR (75.49 MHz,
CDCl3): δ = 13.63, 22.66, 25.20, 29.92, 34.45, 52.77, 77.93,
116.86, 121.88, 126.03, 126.85, 127.91, 128.57, 128.67,
132.35, 141.65, 154.76 ppm; MS (CI) m/z: 297 [M-PrS].
General Procedure for the Preparation of (10) (45).
2, 6-bis (hydroxymethyl)-4-methylphenol (1 g, 5.94 mmol)
was dried by three washes with toluene and evaporation. It
was then dissolved in 10 mL of dry DMF to which 1.4 mL of
BDA, 0.5 g of molecular sieve 3 Å, and 100 μL of dry hydro-
chloric acid 4 M (in dioxane) were added. The reaction mix
was left stirring for 12 h followed by neutralization with 15%
(w/v) sodium bicarbonate. The mixture was then filtered, and
the filtrate was washed with chloroform, dried over MgSO4,
and evaporated. The residue amorphic solids were purified by
HPLC, using a stepwise gradient of water and acetonitrile.
(6-methyl-2-phenyl-4H-benzo[d] [1, 3]dioxin-8-
yl)methanol (10):
8-(1,3-dithiolan-2-yl)-6-methyl-2-phenyl-4H-benzo[d] [1,
3]dioxin (14):
The title compound was obtained starting from 11 (white
1
solid, 23%): H NMR (300 MHz, CDCl3): δ = 2.29 (s, 3H),
3.38 (m, 4H), 4.94 (d, J = 14.4 Hz, 1H), 5.15 (d, J = 14.4 Hz,
1H), 6.01 (s, 1H), 6.74 (s, 1H), 7.43 (m, 4H), 7.63 ppm (m,
2H); 13C NMR (75.49 MHz, CDCl3): δ = 20.91, 39.54,
39.63, 48.37, 66.65, 99.04, 120.49, 124.75, 126.45, 127.25,
128.47, 128.74, 129.37, 130.51, 137.11 ppm; MS (EI) m/z:
330 [M].
8-(bis(ethylthio)methyl)-6-methyl-2-phenyl-4H-benzo[d]
[1, 3]dioxin (15):
The title compound was obtained starting from 2, 6-bis
1
(hydroxymethyl)-4-methylphenol (white solid, 35%): H
The title compound was obtained starting from 11 (white
NMR (300 MHz, CDCl3): δ = 2.29 (s, 3H), 4.68 (d,
solid, 26%): 1H NMR (300 MHz, CDCl3): δ = 1.22 (m, 6H),

Bi-functional antidiabetic drug candidate
2.29 (s, 3H), 2.58 (m, 4H), 4.95 (d, J = 14.4 Hz, 1H), 5.15 (d,
J = 14.4 Hz, 1H), 5.42 (s, 1H), 5.98 (s, 1H), 6.73 (s, 1H), 7.33
(s, 1H), 7.43 (m, 3H), 7.59 ppm (m, 2H); 13C NMR
(75.49 MHz, CDCl3): δ = 14.44, 20.87, 26.46, 26.49, 43.66,
66.72, 99.11, 120.36, 124.62, 126.37, 127.68, 128.47, 129.39,
130.84, 137.21, 147.25 ppm; MS (ESI) m/z: 383 [M + Na],
399 [M + K].
(m, 3H), 7.63 ppm (m, 2H); 13C NMR (75.49 MHz, CDCl3):
δ = 20.72, 23.33, 25.85, 31.22, 43.20, 44.52, 44.58, 66.56,
99.19, 120.64, 122.52, 125.34, 126.65, 128.12, 128.47,
129.38, 130.84, 134.49 ppm; MS (EI) m/z: 372 [M].
Cell Cultures
General Procedure for the preparation of (16) (53).
2,2– dimethyl– 1,3- propanediol (10 g, 0.096 mol) dissolved
in 21 ml of pyridine was added during 30 min to a stirred
solution of p- toluenesulfonyl chloride (55 g, 0.288 mol) in
70 ml of DCM, maintained at 14–18°C. The reaction mixture
was stored at 0°C for 4 days, filtered free from pyridine hy-
drochloride, and washed well with diluted HCl, and then with
diluted NaHCO3 solution. The solvent was removed by dis-
tillation and the viscous residue was warmed briefly under a
vacuum and then poured into vigorously stirred petroleum
ether to precipitate a crystalline product. Recrystallization
from isopropyl alcohol gave pure product.
INS-1E cells were grown as previously described (42). L6
myoblasts were grown and let to differentiate into myotube
as previously described (55,56). Human embryonic kidney
cells (HEK 293) were maintained as described (57).
Isolation of Mononuclear Cells from Human Peripheral Blood
by Density Gradient Centrifugation
Mononuclear cells were isolated from human peripheral
blood by the Ficoll-based protocol developed by Sanderson
et al. (58). Briefly, 20 ml of the whole blood was withdrawn
from a healthy donor under sterile conditions. The blood was
diluted 1:1 with warm PBS.. Ficoll-Paque solution was added
to the diluted blood at a 3:4the ratio. Following centrifugation
at 470xg (~1500 rpm) for 35 min at 20°C in a swinging bucket
rotor without brake, the upper layer containing plasma and
platelets was removed, leaving the mononuclear cell layer un-
disturbed at the interface. This layer was removed and mixed
with cold sterile PBS (30 mland centrifuged at 4°C for 5 min
at 400xg (1000 rpm). The resulting pellet was resuspended in
seeding media (RPMI 1640, 10% FCS, 1% L-glutamine and
seeded (2x106 cells/well in 6-well plates). Cells were incubated
overnight before the common Trypan blue toxicity assay.
2, 2-dimethylpropane-1, 3-diylbis (4-methylbenzenesulfonate)
(16):
The title compound was obtained starting from 2,2– di-
1
methyl– 1,3- propanediol (white solid, 69%): H NMR
(300 MHz, CDCl3): δ = 0.88 (s, 6H), 2.46 (s, 6H), 3.71 (s,
4H), 7.35 (d, J = 8 Hz, 4H), 7.74 ppm (d, J = 8 Hz, 4H);
13C NMR (75.49 MHz, CDCl3): δ = 21.06, 21.70, 35.40,
73.74, 127.92, 129.98, 132.52, 145.03 ppm; MS (ESI) m/z:
413 [M + H], 435 [M + Na], 472 [M + K].
General Procedure for the preparation of (17) (54).
16 (2 g, 4.85 mmol) and KSCN (11.5 g, 118.34 mmol) were
stirred in dry DMF (40 mL) at 125°C for 12 h under nitrogen.
Most of the solvent was removed under reduced pressure on
hot to produce a brownish paste. Water (50 mL) was added to
the paste and the mixture was extracted with ethyl acetate
(4 × 100 mL). The organic layer was dried over Na2SO4
and concentrated under reduced pressure.
Measurement of dGlc Uptake
The rate of [3H]dGlc uptake in myotubes in the absence or
presence of insulin was measured as described [42]. Briefly, L6
myotubes cultures were usually preincubated in α-MEM sup-
plemented with 2% (v/v) FCS and 23.0 mM D-glucose for
24 h, after which they were treated as described in each ex-
periment The cultures were then rinsed 3 times with PBS at
room temperature and further incubated with PBS, pH 7.4,
containing 0.1 mM dGlc and 1.3 μCi/mL [3H]dGlc for 5 min
at room temperature. The uptake was stopped by 3 rapid
washes with ice-cold PBS. The myotubes were then lysed with
0.1% (w/v) SDS in water and taken for liquid scintillation
counting of tritium.
2, 2-dimethyl-1,3-dithiocyanatopropane (17):
The title compound was obtained starting from 16 (yellow
1
oil, 100%): H NMR (300 MHz, CDCl3): δ = 1.24 (s, 6H),
3.09 ppm (s, 4H); 13C NMR (75.49 MHz, CDCl3): δ = 24.81,
37.40, 43.98, 111.86 ppm; MS (CI) m/z: 187 [M + H].
2, 2 -dimethylpropane-1,3-dithiol (18) was synthesized as
described (54).
The title compound was obtained starting from 17 (yellow
oil, 18%): 1H NMR (300 MHz, CDCl3): δ = 0.99 (s, 6H), 2.04
(s, 1H), 2.52 (s, 2H), 2.55 ppm (s, 2H); MS (EI) m/z: 136 [M].
8-(5,5-dimethyl-1,3-dithian-2-yl)-6-methyl-2-phenyl-4H-
benzo[d] [1, 3]dioxin (19) (51).
MTT Assay
The title compound was obtained starting from 11 and 18
(white solid, 44%): 1H NMR (300 MHz, CDCl3): δ = 1.10 (s,
3H), 1.38 (s, 3H), 2.29 (s, 3H), 2.45 (d, J = 14 Hz, 2H), 2.99 (d,
J = 14 Hz, 2H), 4.92 (d, J = 14.4 Hz, 1H), 5.12 (d, J = 14.4 Hz,
1H), 5.44 (s, 1H), 5.99 (s, 1H), 6.75 (s, 1H), 7.39 (s, 1H), 7.44
In order to determine the toxicity of the test compound, L6
myoblasts or INS-1E cells (10,000 cells/well) were plated in
96-well tissue culture plates in 0.1 mL of the culture medium
and incubated overnight. The medium was then replaced
with fresh medium containing a serial dilution of tested

N. Rozentul et al.
compounds starting up to 100 μM. HEK293 cells were incu-
bated with 50 μM of compound 7. Following 24 h incubation
of the plates at 37°C, the cell survival was determined by
MTT assay, as described (59).
AMPK Silencing with shRNA
Subconfluent L6 myoblasts cultures were transfected with
Lentivirus consisting harboring shRNA against rat AMPKα2
(courtesy of Dr. Rachel Hertz (Department of Human
Nutrition and Metabolism, The Hebrew University Faculty
of Medicine, Jerusalem, Israel) and used as described by
Za’tara C., et al. (61). Cells were harvested 72 h after transfec-
tion and treatments, and lysates were prepared and used for
Western blot analyses.
Western Blot Analyses
Whole-cell lysates were prepared in the following manner: the
cells were washed with ice-cold PBS, and 1 mL lysis buffer was
then added and incubated at 4°C for 40 min. The lysis buffer
contained 50 mm Tris·HCl, pH 7.5, 1 mm EDTA, 1 mm
EGTA, 1 mm Na3VO4, 150 mm NaCl, 50 mm
NaF,10 mm sodium glycerophosphate, 5 mm sodium pyro-
phosphate, and 1 mm PMSF, supplemented with 0.1% (v/v)
NP-40, 0.1% (v/v) β-mercaptoethanol, and protease inhibitor
cocktail (1:100 dilution). The resulting cell lysates were centri-
fuged at 8700 g for 30 min at 4°C and the resulting superna-
tant fractions were separated and kept at -80°C until use.
Protein content in the supernatant was determined according
to the Bradford assay, using a BSA standard dissolved in the
same buffer. Aliquots (5–60 μg) were mixed with the sample
buffer [62.5 mM Tris·HCl, pH 6.8, 2% (w/v) SDS, 10% (v/v)
glycerol, 50 mm DTT, and 0.01% (w/v) bromophenol blue]
and heated at 95°C for 5 min. The proteins were then sepa-
rated by 10% SDS-PAGE and Western blot detection were
performed according to suppliers’ protocols for the different
antibodies against AMPKα, pThr172-AMPKα, Akt,
pSer473-AKT, AS160, pThr642-AS160, LKB1, pSer428-
LKB1, ACC pSer79-ACC and GLUT4 (Cell Signaling
Technology, Beverly, MA). Bands were visualized with ECL
plus (GE solution Biosciences, Pittsburgh, PA, USA) and
scanned in BioRad ChemiDoc XRS + Gel Imaging system,
Hercules, CA). The band density was evaluated by the Image
Quant program (GE Healthcare, Chicago, IL).
GSIS
The assays were performed for INS-1E medium and lysates as
previously described using Insulin ELISA kit (Merck
Millipore, Billerica, MA) (62).
Measurement of hERG Function
The experiment was performed using PredictorTM hERG
Fluorescence Polarization Assay kit in accordance with the
manufacturer’s protocol (Protocol PV5365, available online:
http://tools.invitrogen.com/content/sfs/manuals/
Predictor_hERG_FP_Assay_man.pdf). Fluorescence
Polarization (FP) measurement is based on the observation
that when a small fluorescent molecule (tracer) is excited by
the polarized light, the emitted light is largely depolarized
because of the rapid rotation of the molecule in the solution.
The FP measurement was performed by incubating the tracer
and membranes with hERG channel for 2–4 h in the buffer.
The maximal fluorescence polarization was detected with no
interference in the interaction between the tracer and (mini-
mal tracer rotation). However, when test compounds (i.e., 7)
compete with the tracer for the interaction with hERG chan-
nel, the polarization of the emitted light decreases due to the
ability of free unbound tracer to rotate rapidly in the buffer.
The degree of binding between a test compound and hERG is
proportional to the decreasing signal. The reference com-
pound (E-4031, provided by the manufacturer) was used as
a control. To compare the IC50 value of E-4031 in the assay
with the data provided by the manufacturer, dose-response
curve of the E-4031 was generated. Three concentrations of
the test compound were used, 1, 5 μM and 25 μM (n = 3).
Reference inhibitor (E-4031) was used at 30 μM as recom-
mended by the manufacturer.
GLUT4myc Plasma Membrane Content Measurement
in L6 Myotubes
The colorimetric detection of surface GLUT4myc in L6
myotubes was performed as described (60). Briefly, cultured
myotubes were incubated with rabbit anti-c-Myc antibody
(1:200 dilution; Santa Cruz Biotechnology, Santa Cruz,
CA), washed and fixed with 3% formaldehyde, and further
interacted with goat HRP-conjugated anti-rabbit IgG (1:2000
dilution, Santa Cruz Biotechnology, USA). Following the
washes, a solution of OPD (o-phenilendiamine) was added,
and the culture plates were taken for absorbance measure-
ment at 492 nm to estimate the relative abundance of
GLUT4myc on the plasma membrane of the myotubes.
The GLUT4myc L6 cells were the courtesy of Dr. A. Klip,
Hospital for Sick Children (Toronto, ON, Canada).
Measurement of P450 Activity
The experiment was performed using P450-Glo™ Assay
Systems (Promega, Madison, WI, United States) in accor-
dance with the manufacturer’s protocols. The P450-Glo™

Bi-functional antidiabetic drug candidate
Assays provide a luminescent readout-based method for mea-
suring cytochrome P450 activity. A conventional cytochrome
P450 reaction is performed by incubating recombinant cyto-
chrome P450 and a luminogenic cytochrome P450 substrate.
The substrates in the P450-Glo™ assays are derivatives of
beetle luciferin. The derivatives themselves are not substrates
for luciferase but are converted by cytochrome P450 isotypes
to luciferin, which in turn reacts with luciferase to produce
light. The amount of light produced is directly proportional
to the cytochrome P450 activity. All test points were per-
formed in quadruplicates. Control membranes without en-
zymes represent the negative control (baseline). Final
DMSO concentration in the reaction mixtures was 0.25%.
The list of the reference compounds and their concentrations
are listed in Suppl. Table 2. Test concentrations of the refer-
ence compounds correspond to approximately 4-fold report-
ed IC50 values for the corresponding cytochromes P450
(63,64), which is expected to produce nearly 100% inhibition.
RESULTS AND DISCUSSION
Synthesis
We have previously discovered that compounds with a
benzylidene-D-xylose diethyl dithioacetal scaffold (Chart 1)
increased the rate of glucose uptake in skeletal muscles in a
non-insulin-dependent manner (44,45). Presently, we have
aimed at creating more effective and potent novel derivatives
based on this prototype molecule. Scheme 1 describes the
synthesis of such five molecules with a phenylchromane core.
The Friedel–Crafts condensation between a benzene and 3-
chloropropanoyl chloride was used to produce the correspon-
dent ketone (1), which was further reduced to 3-chloro-1-
phenylpropan-1-alcohol by NaBH4 (2). Commercially avail-
able 2,4-dibromophenol was conjugated with 2 using the
Mitsunobu reaction to obtain 2,4-dibromo-1-(3-chloro-1-
phenylpropoxy)benzene (3). This aromatic ether was taken
to intermolecular cyclization and aldehyde addition, which
occurred simultaneously under extra dry conditions, in a one
pot reaction, as described by Hodgetts et al. (50). The outcome
of these reactions was the novel tricycle aldehyde 2-
phenylchromane-6-carbaldehyde (4). This was used as a
starting molecule for the synthesis of five phenylchromane-
based benzylidene-D-xylose diethyl dithioacetal analogs.
Formation and cleavage of thioacetals usually occurs in acid
conditions under kinetic equilibrium (65). Occasionally, it
Statistical Analysis
Results presented as a Mean SEM, statistical significance
*(p < 0.05) was calculated among experimental groups using
the two-tailed Student’s t-test. The QuickCalcs online service
Pad Software: www.graphpad.com/quickcalcs/ttest1.cfm was
used for statistical evaluations.
Dioxin
Phenylchromane
Chart 1 Two groups of synthesized compounds with phenylchromane or dioxin core.

N. Rozentul et al.
Scheme 1 Synthesis of
compounds with phenylchromane
core. Reagents and conditions: (a)
AlCl3, r. t., 2 h.; (b) NaBH4, THF, r.
t., 2 h.; (c) PPh3, DEAD, THF, r. t.,
16 h.; (d) n-BuLi, dry THF, dry DMF,
−50° C to r. t., 2 h.; (e)
corresponding dithiol, CHCl3,
TMSCl, r. t., 2 h.; (f) corresponding
thiol, CHCl3, TMSCl, r. t., 2 h.
may be difficult to define specific acidic conditions for the
catalysis and formation of thioacetals while preserving their
structural integrity. After several unsuccessful attempts to pre-
pare thioacetals using 4 as a starting molecule by applying
canonical methods (e.g.,, DMF with HCl) we introduced a
different method to protect the hydroxyl groups (51)], which
is based on gentle acidic catalysis with TMSCl. The alcohol
moiety, which was not present in the reacting molecule, was
formed during the thioacetal synthesis when the first thiol
group attacked the aldehyde. This hydroxyl reacted with
TMSCl, followed by the release of HCl, which promoted
the formation of the thioacetal moiety, while the second thiol
group attacked the generated cation (51). Of the seven novel
thioacetals five were cyclic: 6-(1,3-dithian-2-yl)-2-
phenylchromane (5), 6-(1,3-dithiolan-2-yl)-2-phenylchromane
(6), 6-(1,3-dithiepan-2-yl)-2-phenylchromane (7); and two
were acyclic: 6-(bis(ethylthio)methyl)-2-phenylchromane (8)
and 6-(bis(propylthio)methyl)-2-phenylchromane (9). All were
synthesized at room temperature using TMSCl as a catalyst
together with the corresponding di- or mono- thiols,.
In addition, a second class of benzylidene-D-xylose diethyl
dithioacetal analogs was synthesized. In this procedure, the
2,4-benzylidene ring was replaced by benzylmethyl domain
to generate 5-methyl-2-phenyl-4H-benzo[d] [1, 3]dioxin core,
as shown in Chart 1. The first intermediate (6-methyl-2-phe-
nyl-4H-benzo[d] [1, 3]dioxin-8-yl)methanol (10) was synthe-
sized under nitrogen atmosphere as described in Scheme 2.
Two starting commercially available molecules, (2-hydroxy-5-
methyl-1,3-phenylene)dimethanol and 1-(dimethoxymethyl)-
4-methylbenzene, were coupled in the presence of a catalytic
amount of HCl. The benzyl alcohol moiety in 10 was oxidized
by PCC to form a novel aromatic aldehyde (11). The yield of
the reaction was very high and close to 100%. Finally, the
dithioacetal moieties were introduced to 11 by an identical
procedure to the above-mentioned phenylchromane analogs
method. Three different compounds that contained 5-methyl-
2-phenyl-4H-benzo[d] [1, 3]dioxin core were synthesized:
8-(1,3-dithian-2-yl)-6-methyl-2-phenyl-4H-benzo[d] [1, 3]di-
oxin (12), 8-(1,3-dithiepan-2-yl)-6-methyl-2-phenyl-4H-
benzo[d] [1, 3]dioxin (13), 8-(1,3-dithiolan-2-yl)-6-methyl-2-
phenyl-4H-benzo[d] [1, 3]dioxin (14). The fourth product,
8-(bis(ethylthio)methyl)-6-methyl-2-phenyl-4H-benzo[d] [1,
3]dioxin (15), was the most structurally closed to the parent
compound due to the presence of the dithioacetal moiety:. To
obtain the fifth compound with a dioxin core, namely, 8-(5, 5-
dimethyl-1, 3-dithian-2-yl)-6-methyl-2-phenyl-4H-benzo[d]
[1, 3] dioxin (19), the corresponding dithiol was first synthe-
sized as shown in Scheme 3. Two hydroxyl groups in the 2,2-
dimethylpropane-1,3-diol were transformed to become good
leaving groups by reactions with tosyl chloride in the presence
of pyridine to obtain 2, 2-dimethylpropane-1,3-diyl bis (4-
methylbenzenesulfonate) (16). Then, potassium thiocyanate
was used as a donor of the thiocyanate groups to replace the
tosyl moieties in 16, resulting in 2, 2-dimethyl-1, 3-
dithiocyanatopropane (17). Reduction cleavage of the thiocy-
anate by lithium aluminum hydride in dry THF produced the
desired 2, 2-dimethylpropane-1, 3-dithiol (18), which was
conjugated with 11 as described above.
Biological Evaluation
All synthesized compounds were first screened for their capacity
to augment glucose uptake in rat L6 myotubes. This rat skeletal
muscle cell-line spontaneously differentiates to multinuclear
myotubes, which are morphologically similar to normal muscle
myotubes (66). The compounds were tested in cultures that were

Bi-functional antidiabetic drug candidate
glucose uptake nearly 3-fold (maximal effect), but at a relative-
ly high concentrations (50 μM). Despite the fact that com-
pound 15 (a dioxin derivative, the structural closest related
molecule to benzylidene-D-xylose diethyl dithioacetals) had
a higher potency than 7, the latter was chosen as the lead
compound and used for the biological investigation due to
its higher activity (three fold stimulatory effect). Dose-
response analysis of the effects of 7 on the rate of glucose
uptake in L6 myotubes was performed, as shown in Fig. 1a.
Already 5 μM of 7 increased significantly the rate of glucose
uptake; at 10 μM 7 nearly doubled the rate of uptake; the
maximal effect of 7 was observed at 50 μM (3.19-fold increase
of rate of glucose uptake compared with the vehicle control).
Noteworthy, at 100 μM 7 was toxic for L6 myotubes (SI Fig.
1A). Time-course analysis (Fig. 1b) shows that 7 augmented
the rate of glucose uptake after a 5-h exposure. This effect
gradually increased and peaked at 24-h of incubation.
However, after 48 h the stimulatory effect of compound 7
decreased from 3-fold (24 h) to 1.8-fold.
Since the parent compound of 7 was an AMPK activator,
we asked whether 7 shared a similar mechanism of action or
mimicked insulin-induced translocation of GLUT4 (69). The
ability of 7 to activate AMPK by Thr172 phosphorylating was
tested at different concentrations of 7 (5, 10 and 25 μM). The
Western blot analyses in Fig. 1c, d indicate that 7 increased
dose-dependently the extent of AMPK phosphorylation.
These analyses show that already within 3 h of incubation 7
(25 μM) increased 2-fold Thr172 phosphorylation.
Comparable levels of Thr172 phosphorylation were also ob-
served after 5 and 24 h of incubation with 25 μM of 7.
Sorbitol, a known hyperosmotic activator of AMPK, was used
as a positive control in these experiments. Importantly, the
time-dependent effects of 7 on glucose uptake (Fig. 1b) corre-
lated well with the activation of AMPK (Fig. 1e, f).
Scheme 2 Synthesis of compounds with dioxin core. Reagents and condi-
tions: (a) dry DMF, molecular sieves 3å, dry HCl in dioxane, r. t., overnight; (b)
PCC, DCM, r. t., 2 h.; (c) corresponding dithiol, CHCl3, TMSCl, r. t., 2 h.; (d)
corresponding thiol, CHCl3, TMSCl, r. t., 2 h.
maintained at high glucose medium (23.5 mM, to mimic hyper-
glycemic conditions). Under such conditions the cellular
autoregulatory mechanism decreases the rate of glucose influx
into the muscle (67,68). Thus, the ability of a test compound to
overcome this physiological protective mechanism and to in-
crease the rate of glucose uptake is considered a good predictor
for potential antihyperglycemic effect in vivo.
Table 1 shows that eight out of the ten tested compounds
increased moderately the rate of glucose uptake in L6
myotubes. In contrast, two compounds (6 and 7), both cyclic
thioether 2-phenylchromane derivatives, increased the rate of
Increased rate of glucose transport in skeletal muscle can
result from translocation of GLUT4 from intracellular compart-
ments to the plasma membrane or from increased intrinsic ac-
tivity of the transporter (6,70–73). To test the possible effect of 7
on plasma membrane GLUT4 content, we used L6 myotubes
stably expressing GLUT4myc. The cell surface expression of
Scheme 3 Synthesis of 19.
Reagents and conditions: (a) TosCl,
pyridine, DCM, 14–18°C; (b)
KSCN, dry DMF, 125°C, 12 h.; (c)
LiAlH4, dry THF, 50° C, overnight;
(d) CHCl3, TMSCl, r. t., 2 h.

N. Rozentul et al.
Tab le 1 Effect of tested com-
pounds on the rate of glucose up-
take in L6 myotubes
Tested compounds
Concentration
Time of maximal effect
% of activation SD Compering to
the rate of glucose uptake in DMSO
treated L6 myotubes (100%)
5
5
24
5
138.14 4.23
279.55 17.12
315.53 21.56
159.39 12.34
156.34 11.24
No effect
6
50
50
50
25
100
40
100
25
25
7
24
24
24
24
5
8
9
12
13
14
15
19
173.35 16.53
No effect
24
5
154.15 11.44
124.76 6.24
5
the maximal effect on glucose uptake is bolded
GLUT4 was measured by detection of the myc epitope in an
extracellular loop of the transporter (60). The total cell content
of GLUT4 in L6 myotubes was not altered by 7 (Fig. 1g). Yet, it
significantly increased GLUT4 abundance in on the cellular
surface in a dose-dependent manner following 24-h incubation
(Fig. 1g). The minimal effective concertation of 7 in increasing
GLUT4 abundance in the plasma membrane was obtained at
5 μM (1.4-fold), and the maximal effect (nearly 2 fold) at 25 μM,
a similar concertation that was required to induce the maximal
effect on the rate of glucose uptake (Fig. 1a).
The effects of 7 on AMPK may involve activation of
LKB1, which is a major upstream kinase that phosphorylates
AMPK α-subunit at Thr172 (74). Therefore, the potential of
7 to activate LKB1 was investigated. Figure 2a, b shows that 7
increased dose-dependently Ser428 phosphorylation in LKB1
in L6 myotubes within 24 h of incubation. Specifically, 10 μM
of 7 increased the phosphorylation nearly by 30%, whereas
25 μM produced maximal phosphorylation (≈80%).
Interestingly, these data also show that insulin induced signif-
icant (2-fold) LKB1 phosphorylation in L6 myotubes; howev-
er, the consequences of this activation are unclear.
Insulin-mediated and AMPK-dependent augmentation of
glucose transport in skeletal muscle requires the activation of
the protein AS160, which is involved in the regulation of the
translocation of GLUT4-rich vesicles to the plasma membrane
of myotubes (75,76). We therefore tested the extent of Thr642
phosphorylation in AS160 in L6 myotubes treated with 7.
Figure 2c, d shows that AS160 was significantly phosphorylated
(≈5-fold) already after 1-h incubation with 25 μM of 7, and this
effect persisted up to 5 h, dropping to nearly 2-fold at 24 h.
Iinterestingly, the effect of 7 on Thr642 phosphorylation in
AS160 was significantly stronger than that of insulin in the same
experimental system. To further test the cellular kinase activity
of AMPK in 7-treated cells we measured its ability to phosphor-
ylate its selective down-stream target Acetyl-CoA carboxylase
(ACC) (77). Figure 2e, f shows that 7 increased Ser79
phosphorylation of ACC in dose-dependent manner. A maxi-
mal 4-fold augmentation of the phosphorylation was observed
with 15 μM of compound 7 after 24 h of incubation. These
findings suggest that 7 augments glucose transport in myotubes
by activating the LKB1-AMPK-AS160 pathway.
Fig. 1 Compound 7 augments glucose uptake in L6 myotubes by activating
„
AMPK. (a) Dose response analysis: L6 myotubes were maintained in αMEM
containing 23.0 mM glucose for 48 h, then washed and incubated with the
same fresh medium supplemented with the indicated concentrations of 7 or
DMSO (0.1%, v/v) for additional 24 h. At the end of incubation the cells were
taken for the standard [3H]dGlc uptake assay. The basal rate of dGlc uptake
with DMSO (0.1%, v/v)-treated L6 myotubes (7.34
0.59 nmol/mg
protein/min) was taken as 100%. (b) Time course analysis: L6 myotubes
were incubated with 25.0 μM of 7 or DMSO for the indicated periods of
time and assayed for [3H]dGlc uptake. The basal rate of [3H]dGlc uptake at
zero time (4.11 0.33 nmol/mg protein/min) was taken as 100%. (c) Dose
response analysis of AMPK activation: L6 myotube cultures were treated with
DMSO or with increasing concentrations of 7 or with 0.25 M sorbitol (S,
positive control). After 24 h whole cell lysates were prepared and Western
blot analyses performed with antibodies against p-172ThrAMPKα and
AMPKα. Representative blots are shown. Level of total AMPKα was used
for the p-172ThrAMPKα band density normalization for each condition. (d)
A graphical representation of the bands density shown in panel C. (e) Time
dependent activation of AMPK: L6 myotube cultures were treated with
DMSO or with 25 μM of 7 or 0.25 M sorbitol and taken for analysis at the
indicated time points. Whole cell lysates were prepared and Western blot
analyses were performed with antibodies against p-172ThrAMPKα and
AMPKα. Representative blots are shown. The level of total AMPK (t-AMPK)
was used for the p-172ThrAMPKα band density normalization per each
treatment. (f) A graphical representation of the bands density shown in the
panel E. (g) Dose response analysis of the effect of 7 on the plasma
membrane abundance of GLUT4myc: Myotubes expressing GLUT4myc
were treated with increasing concentrations of 7, 0.25 M of sorbitol (S) or
DMSO for 24 h. At the end of the incubation the cultures were taken for
immunodetection of surface GLUT4myc as described in the Experimental
Section. The upper panel is shown a reprehensive blot of total level of
GLUT4 in L6 lysates, whereas the lower panel shows plasma membrane-
associated GLUT4myc. The level of GLUT4myc in DMSO- treated
myotubes was taken as 100%. *p < 0.05, in comparison to the respective
controls. n = 5–6.

Bi-functional antidiabetic drug candidate
a
c
d
*
p-AMPK
t-AMPK
*
*
*
10
25
S
DMSO
5
400
7, μM
*
300
200
100
0
DMSO
1
5
10
25
50
*
*
7, μM
b
400
300
200
100
0
*
*
*
*
*
1
5
10
S
DMSO
25
7, μM
g
1
0
2
3
3
5
4
12
5
24
6
48
Time, h
S
DMSO
5
10
25
C
7, 25 µM
7, µM
400
350
300
250
200
150
100
50
e
*
p-AMPK
t-AMPK
*
*
*
C
S
DMSO
3
5
24
300
f
Time, h
*
*
0
1
C
S
2
DMSO
3
5
4
1
5
0
6
25
200
*
7, µM
*
100
0
1
DMSO
C
S
13
5
24
Time, h
To further investigate the role of AMPK in this path-
way, the impact of silencing of the expression of AMPKα2
subunit with shRNA was studied (78). This treatment re-
duced significantly the cellular content of the protein (Fig.
3a). The finding that 7 was unable to produce any glucose
uptake stimulatory effect under this treatment (Fig. 3b)
attests to the critical role of AMPK in mediating the bio-
logical effect of 7 in L6 myotubes.
investigated whether 7 could also employ the insulin transduc-
tion pathway to exert this effect. The phosphorylation of
Ser437 in Akt/PKB, the key protein in the insulin transduc-
tion pathway, was determined (79). Figure 3c, d shows that
insulin, but not 7, induced significant Akt/PKB phosphoryla-
tion in treated myotubes.
Studies on the role of AMPK activators in augmenting insu-
lin secretion led us to investigate whether 7 could dually aug-
ment glucose uptake in skeletal muscle cells and stimulate insulin
secretion from beta cells. The possible effect of 7 on insulin
Since both insulin and 7 increased the plasma membrane
content of GLUT4, albeit with different time-courses, we

N. Rozentul et al.
7, 25 [µM]
c
p-AS160
t-AS160
p-Akt
a
p-LKB1
t-LKB1
t-Akt
1
2.5
INS DMSO
3
5
24
Ins
SF
DMSO
5
10
25
SF
S
7, [µM]
b
d
800
600
400
200
0
400
300
200
100
0
Time, h
*
*
*
*
*
*
*
*
*
*
*
2.5
5
1
S
SF
25
SF
INS
DMSO
3
5
24
INS DMSO
10
Time, h
7, [µM]
e
p-ACC
t-ACC
DMSO
5
10
15
7, [µM]
25
f
500
400
300
200
100
0
*
*
*
DMSO
5
10
15
25
7, [µM]
secretion from INS-1E beta-cells was investigated under static
incubations (65). It exhibited significant dose-dependent stimu-
latory effect on the insulin secretory capacity of the cells after
24 h of incubation (Fig. 4a). Half-maximal and maximal effects
of 7 on GSIS were obtained with 5 and 10 μM, respectively.
Importantly, the stimulatory effects of 7 at 10 and 25 μM were
similar. Noteworthy, the basal rate of insulin secretion (at
3.3 mM glucose) was unaltered in the presence of 7 and the

Bi-functional antidiabetic drug candidate
Fig. 2 Effects of 7 on LKB1, ACC, AS160 and Akt/PKB in L6 myotubes. (a)
it significantly increased AMPK phosphorylation (Thr172) in a
dose-dependent manner; maximal effect was observed with
10 μM (Fig. 4c, d). Figure 4e, f shows the time-dependent effect
of 7 on AMPK phosphorylation: maximal effect (2-fold) was
observed after 12 h of incubation, decreasing slightly at 24 h.
In addition, 7 induced LKB1 phosphorylation (Ser428) in beta-
cells at 24-h incubations in a dose-dependent manner. The min-
imal effective concentration of 7 was 2.5 μM (50% increase) and
at 25 μM led to 3-fold effect (Fig. 5a, b). Compound 7 also
induced ACC phosphorylation (Ser79) within 24 h of incuba-
tion, as shown in Fig. 5c, d; at 25 μM it led to 8-fold augmen-
tation in the phosphorylation level of ACC. Thus, the effect of 7
on beta-cell LKB1/AMPK pathway was similar to that ob-
served in muscle cells, albeit with a somewhat greater efficacy.
Collectively, these data show that 7, increased glucose up-
take in L6 myotubes in a non-insulin-dependent manner and
augmented GSIS from beta-cells through AMPK activation.
This bi-functionality of 7 represents an interesting class of
potential antidiabetic compounds that may reduce blood glu-
cose levels by correcting the two main defects in T2D, the
reduced glucose uptake in skeletal muscles, and the impeded
insulin secretion from beta-cells.
ƒ
Dose response analysis of the effect of 7 on the level of p-428SerLKB1: L6
Myotubes were treated as described in the Fig. 1c. Insulin (100 nM, INS) was
introduced during the last 20 min of incubation. Myotubes which received insulin
were kept 5 h before the insulin addition in free serum medium (SF). Sorbitol
(0.25 M, S) was used as a positive control. Whole cell lysates were prepared and
taken for Western blot analysis. The level of total LKB1 (t-LKB1) was used for the
p-428SerLKB1 band density normalization per each treatment. (b) Graphical
presentation of the bands density of the image shown in A. (c) Time course
analysis of the effect of 7 on the levels of p-Thr642AS160 and p-Ser437AKT/
PKB in L6 myotubes: Myotubes were incubated with 25 μM of 7 or the vehicle
[DMSO, 0.1% (v/v)] for the indicated periods of time. The levels of total As160 (t-
AS160) and total Akt/PKB (t-AKT/PKB) were used for the p-Thr642AS160 and p-
Ser437AKT/PKB band density normalization per each treatment. (d) Graphical
presentation of the bands density of the image shown in C. Band density of the
DMSO-treated cells was taken as 100%. Blue columns represent p-AS160 and
the red represent p-AKT/PKB levels. (e) Dose response analysis of the effect of 7
on the phosphorylation level of ACC: L6 myotubes treated with increased
concentrations of 7 and DMSO (0.1%). Whole cell lysates were prepared and
taken for Western blot analysis of total or phosphorylated p-Ser79ACC and total
ACC. The p-Ser79ACC bands density was normalized to amount of total ACC
(t-ACC) for each treatment. (f) Graphical presentation of the bands density of the
image shown in E. *p < 0.05, in comparison to DMSO treatment, n = 5–7.
total intracellular content of insulin was not affected by 7 (Fig.
4b). The compound was not toxic for INS-1E beta-cells at these
concentrations (SI Fig. 1B).
We then asked whether 7-induced increases of GSIS in INS-
1E also involved AMPK activation like in L6 myotubes. Indeed
The druggability of 7 was then investigated. Two basic
metabolic in vitro pharmacokinetic tests were conducted.
First, possible inhibitory effects of 7 on Bhuman ether-à-
WT
AMPK-/-
a
AMPK
Tubulin
7
C
S
DMSO
7
C
S
DMSO
b
WT
25
20
15
10
5
AMPK-/-
*
*
*
0
5
10
7, [µM]
25
DMSO
5
10
DMSO
25
7, [µM]
Fig. 3 AMPKα2 silencing diminishes the effect of 7 in L6 myotubes. (a) Silencing of AMPK α2: Wild type (WT) cells and shRNA AMPKα2-treated (AMPK−/−)
L6 myotubes were maintained in αMEM containing 23 mM glucose for 24 h, then washed and incubated with the same fresh medium supplemented with 25 μM
of 7 or DMSO (0.1%, v/v) for 24 h. Sorbitol (0.25 M, S) was added for 2 h. Whole cell lysates were prepared and Western blot analyses were performed with
antibodies against AMPKα2 and tubulin (loading control), according to manufacturer’s protocols. Representative blots are shown. (b) Dose response analysis of 7
on glucose uptake in Wild type (WT) cells and shRNA AMPKα2-treated (AMPK−/−) L6 myotubes: The cultures were treated with 7 (5, 10 and 25 μM) and
DMSO as described above. At the end of incubation both types of myotubes were taken for the standard [3H]dGlc uptake assay. The basal rate of dGlc uptake
with DMSO (5.71 0.51 nmol/mg protein/min) was taken as 100%. *p < 0.05, in comparison to the respective controls, n = 3.

N. Rozentul et al.
a
c
8
6
4
2
0
3.3 mM glucose
16.7 mM glucose
p-AMPK
*
*
*
t-AMPK
*
350
300
250
200
150
100
50
d
*
*
Ctrl
C
DMSO
1
2.5
5
10
25
7, µM
*
*
15000
10000
5000
0
b
0
DMSO comp7 DMSO comp7 DMSO comp7
1
2.5
5
10
25
S
DMSO
7, µM
Ctrl DMSO
C
1
2.5
5
10
25
7, µM
3h
12h
24h
e
p-AMPK
t-AMPK
250
200
150
100
50
f
*
*
0
DMSO
+
comp7 DMSO
-
comp7 DMSO comp7
-
DMSO:
-
+
+
-
7:
-
+
-
+
+
go-go^-related gene channels (hERG) and P450 enzymes
were tested. The human hERG potassium channels con-
duct the rapid component of the delayed rectifier potassi-
um current, which is crucial for repolarization of cardiac
action potentials (80). Blocking hERG channels by drugs
and drug candidates can lead to action potential prolon-
gation, lengthening of the QT interval and an increased
risk for Btorsade de pointes^ arrhythmias and sudden
death (80). Thus, the possible inhibition of hERG chan-
nels is one of the negative features of any drug candidate

Bi-functional antidiabetic drug candidate
Fig. 4 Compound 7 augments insulin secretion from INS-1E beta cells. (a)
drugs among them) (82). In humans, approximately 110 genes
encode this battery of P450 enzymes. Interestingly, among all
CYPs the enzymes belonging to group 1, 2 and 3 metabolize
the vast majority of drugs (82). The effect of 7 on five most
relevant P450 isotypes, 2C19, 1A2, 2C9, 2D6 and 3A4 was
tested using the P450-Glo™ Assay Systems for drug metabo-
lism and toxicity (Table 2). The data indicates that 7 exhibited
a relatively high level (≈80%) of inhibition of 2C19 at 5 μM.
Two additional CYPs: 1A2, 2C9 were inhibited by ≈45%.
The activity the 2D6 and 3A4 isotypes was not strongly sup-
pressed by 7. Very strong inhibition of only one CYP out of
five tested is considered encouraging in the process of drug
design and optimization. Interestingly, many approved drugs
such as fluoxetine (83), omeprazole (84), cimetidine (85) and
isoniazid (86) are potent 2C19 inhibitors.
Finally, the possible toxicity of compound 7 in human non-
cancer cells was investigated. Human monocytes and
HEK293 cells were exposed to compound 7 (50 μM) for
24 h. Trypan blue exclusion test or MTT were used to eval-
uate possible toxic effect of 7 on cells viability. In both types of
cells, compound 7 dids not shown any effect on the viability of
the cells (SI Fig. 1C, D).
ƒ
Dose response analysis: The cells were grown and treated with the indicated
concretions of 7 for 24 h. The cells were then washed and taken for the
standard GSIS, as described in the BExperimental Section^. (b) Lack of the
effect of 7 on the total insulin content: Similarly treated cells as in A were
washed and lysed and the whole cells insulin levels were determined as
described under BExperimental Section^. (c) Dose response analysis of
AMPK activation by 7: INS-1E cells were treated with DMSO (0.1%, v/v),
with increasing concentrations of 7 or with 0.25 M of sorbitol (S) for 24 h.
Whole cell lysates were then prepared and Western blot analyses were
performed with antibodies against p-172ThrAMPKα and AMPKα, according
to manufacturer’s protocols. Representative blots are shown. Level of total
AMPK (t-AMPK) was used for the p-172ThrAMPKα band density
normalization for each treatment. The density the band of DMSO treated
cells was taken as 100%. (d) Graphical presentation of the bands density of
the image shown in C. (e) Compound 7 activates AMPK in INS-1E time-
dependently: INS-1E cells were grown and treated with 25 μM of 7 or
DMSO (0.1% v/v). Whole cell lysates were prepared at the indicated times
and taken for Western blot analysis for p-Thr172AMPK and total AMPK t-
AMPK. The density of the total AMPK t-AMPK bands was used for
normalization of the p-Thr172AMPK signal. The density of DMSO treated-
cells at 3 h was taken as 100%. (f) Graphical presentation of the bands density
of the image shown in E. Representative blots are shown * p < 0.05, in
comparison to DMSO treatment, n = 4.
which must be avoided (81). The ability of compound 7 to
bind to hERG was investigated using a fluorescence po-
larization assay. SI Table 1 shows no evidence for 7-
mediated hERG inhibition at the tested concentrations..
Cytochromes P450 (CYPs) form the most important family
of enzymes that catalyze a broad spectrum of oxidation-
derived biotransformation of xenobiotics (and numerous
CONCLUSIONS
This study shows that 6-(1,3-dithiepan-2-yl)-2-
phenylchromane (7), a novel phenylchromane derivative,
a
c
p-LKB1
p-ACC
t-LKB1
t-ACC
b
d
350
1000
300
250
200
150
100
50
*
800
600
400
200
0
*
*
*
*
*
*
*
0
S DMSO
25
DMSO comp7 DMSO comp7 DMSO
2.5
10
DMSO comp7 DMSO comp7 DMSO comp7
5
10
15
25
S
DMSO
7, [µM]
7, [µM]
Fig. 5 Effects of 7 on LKB1 and ACC in INS-1E β-cells. (a) Dose response effect of 7 on the level of p-LKB1 Ser428: The cells were treated as described in the
legend to Fig. 4a. Sorbitol (0.25 M, S) was used as a positive control. Whole cell lysates were used for Western blot analysis. The levels of total LKB1 (t-LKB1) and
p-428SerLKB1 were determined with corresponding antibodies according to manufacturer’s protocols. The level of t-LKB1 was used for the p-428SerLKB1 band
density normalization for each treatment. (b) Graphical presentation of the bands density of the image shown in A: DMSO treated cells were taken as 100% (c)
Dose response analysis of the effect of 7 on the level of Ser79-ACC phosphorylation. The cells were treated as described above and whole cells lysates were used
for Western blot analysis. The levels of total ACC (t-ACC) and p-79SerACC were determined with selective antibodies according to manufacturer’s protocols.
The level of t-ACC was used for the p-79SerACC band density normalization for each treatment. DMSO-treated cells were taken as 100%. (d) Graphical
presentation of the bands density of the image shown in C. Representative blots are shown. *p < 0.05, in comparison to DMSO treatment, n = 4.

N. Rozentul et al.
Tab le 2 Percent of inhi-
bition of P450 enzymes by
compound 7 in
5. Karlssonand HK, Zierath JR. Insulin signaling and glucose trans-
port in insulin resistant human skeletal muscle. Cell Biochem
Biophys. 2007;48:103–13.
CYP
% of inhibition SD
1A2
53.2 1.3
57.1 9.3
78.8 3.3
21.9 2.7
34.7 14.5
concertation of 5 μM
6. Richterand EA, Hargreaves M. Exercise, GLUT4, and skeletal
muscle glucose uptake. Physiol Rev. 2013;93:993–1017.
7. Klip A, Schertzer JD, Bilan PJ, Thong F, Antonescu C. Regulation
of glucose transporter 4 traffic by energy deprivation from mito-
chondrial compromise. Acta Physiol (Oxf). 2009;196:27–35.
8. Ojuka EO. Role of calcium and AMP kinase in the regulation of
mitochondrial biogenesis and GLUT4 levels in muscle. Proc Nutr
Soc. 2004;63:275–8.
9. Hardie DG. AMPK: positive and negative regulation, and its role in
whole-body energy homeostasis. Curr Opin Cell Biol. 2015;33:1–7.
10. Kyriakis JM. At the crossroads: AMP-activated kinase and the
LKB1 tumor suppressor link cell proliferation to metabolic regula-
tion. J Biol. 2003;2:26.
11. Jiang SJ, Dong H, Li JB, Xu LJ, Zou X, Wang KF, et al. Berberine
inhibits hepatic gluconeogenesis via the LKB1-AMPK-TORC2
signaling pathway in streptozotocin-induced diabetic rats. World J
Gastroenterol. 2015;21:7777–85.
12. Xie Z, Ding SQ, Shen YF. Silibinin activates AMP-activated pro-
tein kinase to protect neuronal cells from oxygen and glucose dep-
rivation-re-oxygenation. Biochem Biophys Res Commun.
2014;454:313–9.
2C9
2C19
2D6
3A4
increased the rate of glucose uptake in L6 myotubes under
hyperglycemic conditions at pharmacologically relevant con-
centration. Moreover, 7 augmented glucose-stimulated insu-
lin secretion from beta-cells. Both activities were mediated via
the activation of LKB1-AMPK pathway. Finally, predicted
pharmacokinetic properties investigated using in vitro methods
showed druggable properties, which make it a suitable candi-
date for future development. We anticipate that this bi-
functionality of 7 will make it a lead compound in the devel-
opment of a novel class of antidiabetic drugs.
13. Yang Y, Zhao Z, Liu Y, Kang X, Zhang H, Meng M. Suppression
of oxidative stress and improvement of liver functions in mice by
ursolic acid via LKB1-AMP-activated protein kinase signaling. J
Gastroenterol Hepatol. 2015;30:609–18.
14. Zaks I, Getter T, Gruzman A. Activators of AMPK: not just for
Type II diabetes. Future Med Chem. 2014;6:1325–53.
15. Cameronand KO, Kurumbail RG. Recent progress in the identi-
fication of adenosine monophosphate-activated protein kinase
(AMPK) activators. Bioorg Med Chem Lett. 2016;26:5139–48.
16. Giordanettoand F, Karis D. Direct AMP-activated protein kinase
activators: a review of evidence from the patent literature. Expert
Opin Ther Pat. 2012;22:1467–77.
17. Sliwinskaand A, Drzewoski J. Molecular action of metformin in
hepatocytes: an updated insight. Curr Diabetes Rev. 2015;11:
175–81.
18. Zhou JY, Xu B, Li L, New Role A. for an Old Drug: Metformin
Targets MicroRNAs in Treating Diabetes and Cancer. Drug Dev
Res. 2015;76:263–9.
19. Foretz M, Guigas B, Bertrand L, Pollak M, Viollet B. Metformin:
from mechanisms of action to therapies. Cell Metab. 2014;20:953–66.
20. Cohen G, Riahi Y, Alpert E, Gruzman A, Sasson S. The roles of
hyperglycaemia and oxidative stress in the rise and collapse of the
natural protective mechanism against vascular endothelial cell dys-
function in diabetes. Arch Physiol Biochem. 2007;113:259–67.
21. Cameron KO, Kung DW, Kalgutkar AS, Kurumbail RG, Miller
R, Salatto CT, et al. Discovery and Preclinical Characterization of
6-Chloro-5-[4-(1-hydroxycyclobutyl)phenyl]-1H-indole-3-carbox-
ylic Acid (PF-06409577), a Direct Activator of Adenosine
Monophosphate-activated Protein Kinase (AMPK), for the
Potential Treatment of Diabetic Nephropathy. J Med Chem.
2016;59:8068–81.
22. Hsu MH, Savas U, Lasker JM, Johnson EF. Genistein, resveratrol,
and 5-aminoimidazole-4-carboxamide-1-beta-D-ribofuranoside in-
duce cytochrome P450 4F2 expression through an AMP-activated
protein kinase-dependent pathway. J Pharmacol Exp Ther.
2011;337:125–36.
23. Benziane B, Bjornholm M, Lantier L, Viollet B, Zierath JR,
Chibalin AV. AMP-activated protein kinase activator A-769662
is an inhibitor of the Na(+)-K(+)-ATPase. Am J Physiol Cell
Physiol. 2009;297:C1554–66.
ACKNOWLEDGMENTS AND DISCLOSURES
We thank Dr. Rachel Hertz, Department of Human
Nutrition and Metabolism, The Hebrew University Faculty
of Medicine, Jerusalem, Israel, for providing us the lentivirus
shRNA-AMPK construct. We thank Ms. Duha Fahham for
her technical help. This study was supported partly by a Bar-
Ilan University new faculty grant (A.G), by Diab R&D
(France) for (E.C., A.G. and S.S.) and grants from the
Applied Research Program A of the Hebrew University, and
the Israel Science Foundation of The Israel Academy of
Sciences and Humanities (to S.S). S. Sasson is member of
the David R. Bloom Center for Pharmacy and in Chair Dr.
Adolf and Klara Brettler Center for Research in Molecular
Pharmacology and Therapeutics. S.S. is the Adolf D. and
Horty Storch Chair in Pharmaceutical Sciences, Faculty of
Medicine, The Hebrew University of Jerusalem.
REFERENCES
1. Rojasand LB, Gomes MB. Metformin: an old but still the best
treatment for type 2 diabetes. Diabetol Metab Syndr. 2013;5:6.
2. Consoli A, Gomis R, Halimi S, Home PD, Mehnert H, Strojek K,
et al. Initiating oral glucose-lowering therapy with metformin in type 2
diabetic patients: an evidence-based strategy to reduce the burden of
late-developing diabetes complications. Diabetes Metab. 2004;30:
509–16.
3. Jensenand TE, Richter EA. Regulation of glucose and glycogen
metabolism during and after exercise. J Physiol. 2012;590:1069–76.
4. Rana S, Blowers EC, Natarajan A. Small molecule adenosine 5′-
monophosphate activated protein kinase (AMPK) modulators and
human diseases. J Med Chem. 2015;58:2–29.

Bi-functional antidiabetic drug candidate
24. Mirguet O, Sautet S, Clement CA, Toum J, Donche F, Marques C,
et al. Discovery of Pyridones As Oral AMPK Direct Activators.
ACS Med Chem Lett. 2013;4:632–6.
25. Ojukaand EO, Goyaram V. Mechanisms in exercise-induced in-
crease in glucose disposal in skeletal muscle. Med Sport Sci.
2014;60:71–81.
26. Krishan S, Richardson DR, Sahni S. Adenosine monophosphate-
activated kinase and its key role in catabolism: structure, regulation,
biological activity, and pharmacological activation. Mol
Pharmacol. 2015;87:363–77.
27. Miglianico M, Nicolaes GA, Neumann D. Pharmacological
Targeting of AMP-Activated Protein Kinase and Opportunities for
Computer-Aided Drug Design. J Med Chem. 2016;59:2879–93.
28. Kone M, Pullen TJ, Sun G, Ibberson M, Martinez-Sanchez A,
Sayers S, et al. LKB1 and AMPK differentially regulate pancreatic
beta-cell identity. FASEB J. 2014;28:4972–85.
29. Swisa A, Granot Z, Tamarina N, Sayers S, Bardeesy N, Philipson
L, et al. Loss of Liver Kinase B1 (LKB1) in Beta Cells Enhances
Glucose-stimulated Insulin Secretion Despite Profound
Mitochondrial Defects. J Biol Chem. 2015;290:20934–46.
30. Sun G, Tarasov AI, McGinty JA, French PM, McDonald A,
Leclerc I, et al. LKB1 deletion with the RIP2.Cre transgene mod-
ifies pancreatic beta-cell morphology and enhances insulin secre-
tion in vivo. Am J Physiol Endocrinol Metab. 2010;298:E1261–73.
31. Fu A, Robitaille K, Faubert B, Reeks C, Dai XQ, Hardy AB, et al.
LKB1 couples glucose metabolism to insulin secretion in mice.
Diabetologia. 2015;58:1513–22.
42. Pasternak L, Meltzer-Mats E, Babai-Shani G, Cohen G, Viskind
O, Eckel J, et al. Benzothiazole derivatives augment glucose uptake
in skeletal muscle cells and stimulate insulin secretion from pancre-
atic beta-cells via AMPK activation. Chem Commun (Camb).
2014;50:11222–5.
43. Beall C, Piipari K, Al-Qassab H, Smith MA, Parker N, Carling D,
et al. Loss of AMP-activated protein kinase α2 subunit in mouse β-
cells impairs glucose-stimulated insulin secretion and inhibits their
sensitivity to hypoglycaemia. Biochem J. 2010;429:323–33.
44. Gruzman A, Elgart A, Viskind O, Billauer H, Dotan S, Cohen G,
et al. Antihyperglycaemic activity of 2,4:3,5-dibenzylidene-D-xy-
lose-diethyl dithioacetal in diabetic mice. J Cell Mol Med.
2012;16:594–604.
45. Gruzman A, Shamni O, Ben Yakir M, Sandovski D, Elgart A,
Alpert E, et al. Novel D-Xylose Derivatives Stimulate Muscle
Glucose Uptake by Activating AMP-Activated Protein Kinase α. J
Med Chem. 2008;51:8096–108.
46. Meltzer-Mats E, Babai-Shani G, Pasternak L, Uritsky N, Getter T,
Viskind O, et al. Synthesis and mechanism of hypoglycemic activity
of benzothiazole derivatives. J Med Chem. 2013;56:5335–50.
47. Gottlieb HE, Kotlyar V, Nudelman A. NMR chemical shifts of
common laboratory solvents as trace impurities. J Org Chem.
1997;62:7512–5.
48. La Regina G, Diodata D'Auria F, Tafi A, Piscitelli F, Olla S,
Caporuscio F, et al. 1-[(3-Aryloxy-3-aryl)propyl]-1H-imidazoles,
New Imidazoles with Potent Activity against Candida albicans
and Dermatophytes. Synthesis, Structure-Activity Relationship,
and Molecular Modeling Studies. J Med Chem. 2008;51:3841–55.
49. Hodgetts KJ. A stereocontrolled route to 2-substituted chromans.
Tetrahedron Lett. 2000;41:8655–9.
32. Granot Z, Swisa A, Magenheim J, Stolovich-Rain M, Fujimoto W,
Manduchi E, et al. LKB1 regulates pancreatic beta cell size, polar-
ity, and function. Cell Metab. 2009;10:296–308.
50. Hodgetts KJ. Inter- and intramolecular Mitsunobu reaction based
approaches to 2-substituted chromans and chroman-4-ones.
Tetrahedron. 2005;61:6860–70.
51. Ongand BS, Chan TH. A simple method of dithioacetalization and
dithioketalization. Synth Commun. 1977;7:283–6.
52. Kimura M, Hamakawa T, Hanabusa K, Shirai H, Kobayashi N.
Synthesis of Multicomponent Systems Composed of One
Phthalocyanine and Four Terpyridine Ligands. Inorg Chem.
2001;40:4775–9.
53. Handrickand GR, Atkinson ER. Potential antiradiation drugs. III.
2-Amino-2-alkyl-1,3-propanedithiols and 3-amino-4-mercapto-1-
butanol. J Med Chem. 1966;9:558–62.
54. Zhong W, Tang Y, Zampella G, Wang X, Yang X, Hu B, et al. A
rare bond between a soft metal (FeI) and a relatively hard base
(RO-, R = phenolic moiety). Inorg Chem Commun. 2010;13:
1089–92.
55. Munder A, Moskovitz Y, Redko B, Levy AR, Ruthstein S,
Gellerman G, et al. Antiproliferative Effect of Novel
Aminoacridine-based Compounds. Med Chem. 2015;11:373–82.
56. Shimanovich U, Munder A, Azoia NG, Cavaco-Paulo A, Gruzman
A, Knowles TP, et al. Sonochemically-induced spectral shift as a
probe of green fluorescent protein release from nano capsules.
RSC Adv. 2014;4:10303–9.
57. Suckow AT, Zhang C, Egodage S, Comoletti D, Taylor P, Miller
MT, et al. Transcellular neuroligin-2 interactions enhance insulin
secretion and are integral to pancreatic beta cell function. J Biol
Chem. 2012;287:19816–26.
58. Sanderson RJ, Shepperdson RT, Vatter AE, Talmage DW.
Isolation and enumeration of peripheral blood monocytes. J
Immunol. 1977;118:1409–14.
33. Fu A, Ng AC, Depatie C, Wijesekara N, He Y, Wang GS, et al. Loss
of Lkb1 in adult beta cells increases beta cell mass and enhances
glucose tolerance in mice. Cell Metab. 2009;10:285–95.
34. Fu A, Eberhard CE, Screaton RA. Role of AMPK in pancreatic
beta cell function. Mol Cell Endocrinol. 2013;366:127–34.
35. Langelueddecke C, Jakab M, Ketterl N, Lehner L, Hufnagl C,
Schmidt S, et al. Effect of the AMP-kinase modulators AICAR,
metformin and compound C on insulin secretion of INS-1E rat
insulinoma cells under standard cell culture conditions. Cell
Physiol Biochem. 2012;29:75–86.
36. da Silva Xavier G, Leclerc I, Varadi A, Tsuboi T, Moule SK,
Rutter GA. Role for AMP-activated protein kinase in glucose-
stimulated insulin secretion and preproinsulin gene expression.
Biochem J. 2003;371:761–74.
37. Sun G, Tarasov AI, McGinty J, McDonald A, da Silva Xavier G,
Gorman T, et al. Ablation of AMP-activated protein kinase alpha1
and alpha2 from mouse pancreatic beta cells and RIP2.Cre neurons
suppresses insulin release in vivo. Diabetologia. 2010;53:924–36.
38. Ryu GR, Lee MK, Lee E, Ko SH, Ahn YB, Kim JW, et al.
Activation of AMP-activated protein kinase mediates acute and
severe hypoxic injury to pancreatic beta cells. Biochem Biophys
Res Commun. 2009;386:356–62.
39. Beall C, Watterson KR, McCrimmon RJ, Ashford ML. AMPK
modulates glucose-sensing in insulin-secreting cells by altered
phosphotransfer to KATP channels. J Bioenerg Biomembr.
2013;45:229–41.
40. Matsuda T, Takai T, Suzuki E, Kanno A, Koyanagi-Kimura M,
Asahara S-i, et al. Regulation of Pancreatic β Cell Mass by Cross-
Interaction between CCAAT Enhancer Binding Protein β Induced
by Endoplasmic Reticulum Stress and AMP-Activated Protein
Kinase Activity. PLoS One. 2015;10:e0130757.
41. Wang K, Sun Y, Lin P, Song J, Zhao R, Li W, et al. Liraglutide
Activates AMPK Signaling and Partially Restores Normal
Circadian Rhythm and Insulin Secretion in Pancreatic Islets in
Diabetic Mice. Biol Pharm Bull. 2015;38:1142–9.
59. Eckshtain-Levi M, Lavi R, Yufit DS, Daniel B, Green O, Fleker O,
et al. A versatile water-soluble chelating and radical scavenging plat-
form. Chem Commun (Camb). 2016;52:2350–3.
60. Shapira R, Rudnick S, Daniel B, Viskind O, Aisha V, Richman M,
et al. Multifunctional cyclic D,L-α-peptide architectures stimulate non-

N. Rozentul et al.
insulin dependent glucose uptake in skeletal muscle cells and protect
them against oxidative stress. J Med Chem. 2013;56:6709–18.
61. Zatara G, Hertz R, Shaked M, Mayorek N, Morad E, Grad E, et al.
Suppression of FoxO1 activity by long-chain fatty acyl analogs.
Diabetes. 2011;60:1872–81.
74. Kottakisand F, Bardeesy N. LKB1-AMPK axis revisited. Cell Res.
2012;22:1617–20.
75. Cao S, Li B, Yi X, Chang B, Zhu B, Lian Z, et al. Effects of exercise
on AMPK signaling and downstream components to PI3K in rat
with type 2 diabetes. PLoS One. 2012;7:e51709.
62. Cohen G, Riahi Y, Shamni O, Guichardant M, Chatgilialoglu C,
Ferreri C, et al. Role of lipid peroxidation and PPAR-delta in am-
plifying glucose-stimulated insulin secretion. Diabetes. 2011;60:
2830–42.
76. Deshmukh AS, Hawley JA, Zierath JR. Exercise-induced phospho-
proteins in skeletal muscle. Int J Obes (Lond) 32 Suppl. 2008;4:
S18–23.
77. Hardieand DG, Pan DA. Regulation of fatty acid synthesis and
oxidation by the AMP-activated protein kinase. Biochem Soc
Trans. 2002;30:1064–70.
78. Ross FA, MacKintosh C, Hardie DG. AMP-activated protein ki-
nase: a cellular energy sensor that comes in 12 flavours. FEBS J.
2016;283:2987–3001.
63. Li XQ, Andersson TB, Ahlstrom M, Weidolf L. Comparison of in-
hibitory effects of the proton pump-inhibiting drugs omeprazole,
esomeprazole, lansoprazole, pantoprazole, and rabeprazole on hu-
man cytochrome P450 activities. Drug Metab Dispos. 2004;32:821–7.
64. Niwa T, Shiraga T, Takagi A. Effect of antifungal drugs on cyto-
chrome P450 (CYP) 2C9, CYP2C19, and CYP3A4 activities in
human liver microsomes. Biol Pharm Bull. 2005;28:1805–8.
65. Mori T, Sawada Y, Oku A. Ring-expansion of thioacetal ring via
bicyclosulfonium ylide. Effect Of protic nucleophile on ylide inter-
mediate J Org Chem. 2000;65:3620–5.
66. Cui Z, Chen X, Lu B, Park SK, Xu T, Xie Z, et al. Preliminary
quantitative profile of differential protein expression between rat L6
myoblasts and myotubes by stable isotope labeling with amino acids
in cell culture. Proteomics. 2009;9:1274–92.
67. Wertheimer E, Sasson S, Cerasi E. Regulation of hexose transport
in L8 myocytes by glucose: possible sites of interaction. J Cell
Physiol. 1990;143:330–6.
68. Itani SI, Saha AK, Kurowski TG, Coffin HR, Tornheim K,
Ruderman NB. Glucose autoregulates its uptake in skeletal muscle:
involvement of AMP-activated protein kinase. Diabetes. 2003;52:
1635–40.
69. Ben-Yakir M, Gruzman A, Alpert E, Sasson S. Glucose transport
regulators. Current Medicinal Chemistry-Immunology. Endocr
Metab Agents. 2005;5:519–27.
70. Shamni O, Cohen G, Gruzman A, Zaid H, Klip A, Cerasi E, et al.
Regulation of GLUT4 activity in myotubes by 3-O-methyl-d-glu-
cose. Biochim Biophys Acta. 2017;1859:1900–10.
71. Shamni O, Cohen G, Gruzman A, Zaid H, Klip A, Cerasi E, et al.
Supportive data on the regulation of GLUT4 activity by 3-O-meth-
yl-D-glucose. Data Brief. 2017;14:329–36.
79. Mackenzieand RW, Elliott BT. Akt/PKB activation and insulin
signaling: a novel insulin signaling pathway in the treatment of type
2 diabetes. Diabetes Metab Syndr Obes. 2014;7:55–64.
80. Thomas D, Karle CA, Kiehn J. The cardiac hERG/IKr potassium
channel as pharmacological target: structure, function, regulation,
and clinical applications. Curr Pharm Des. 2006;12:2271–83.
81. Pierson JB, Berridge BR, Brooks MB, Dreher K, Koerner J,
Schultze AE, et al. A public-private consortium advances cardiac
safety evaluation: achievements of the HESI Cardiac Safety
Technical Committee. J Pharmacol Toxicol Methods. 2013;68:7–
12.
82. Zangerand UM, Schwab M. Cytochrome P450 enzymes in drug
metabolism: regulation of gene expression, enzyme activities, and
impact of genetic variation. Pharmacol Ther. 2013;138:103–41.
83. Sager JE, Lutz JD, Foti RS, Davis C, Kunze KL, Isoherranen N.
Fluoxetine- and norfluoxetine-mediated complex drug-drug inter-
actions: in vitro to in vivo correlation of effects on CYP2D6,
CYP2C19, and CYP3A4. Clin Pharmacol Ther. 2014;95:653–62.
84. Quinn DI, Nemunaitis J, Fuloria J, Britten CD, Gabrail N, Yee L,
et al. Effect of the cytochrome P450 2C19 inhibitor omeprazole on
the pharmacokinetics and safety profile of bortezomib in patients
with advanced solid tumours, non-Hodgkin's lymphoma or multi-
ple myeloma. Clin Pharmacokinet. 2009;48:199–209.
85. Furuta S, Kamada E, Suzuki T, Sugimoto T, Kawabata Y,
Shinozaki Y, et al. Inhibition of drug metabolism in human liver
microsomes by nizatidine, cimetidine and omeprazole.
Xenobiotica. 2001;31:1–10.
86. Wen X, Wang J-S, Neuvonen PJ, Backman JT. Isoniazid is a
mechanism-based inhibitor of cytochrome P450 1A2, 2A6, 2C19
and 3A4 isoforms in human liver microsomes. Eur J Clin
Pharmacol. 2002;57:799–804.
72. Furtado LM, Somwar R, Sweeney G, Niu W, Klip A. Activation of
the glucose transporter GLUT4 by insulin. Biochem Cell Biol.
2002;80:569–78.
73. Chiu TT, Jensen TE, Sylow L, Richter EA, Klip A. Rac1 signalling
towards GLUT4/glucose uptake in skeletal muscle. Cell Signal.
2011;23:1546–54.