Abrogation of hyperosmotic impairment of insulin signaling by a novel class of 1,2-dithiole-3-thiones through the inhibition of S6K1 activation

Article abstract of DOI:10.1124/mol.107.044347

A previous study from this laboratory showed that oltipraz and synthetic dithiolethiones prevent tumor necrosis factor-α-induced hepatic insulin resistance via AMP-activated protein kinase-dependent p70S6 kinase (S6K) 1 inhibitory pathway. This study investigated whether oltipraz and a novel class of 1,2-dithiole-3-thiones were capable of preventing insulin resistance induced by hyperosmotic stress, thereby enhancing insulin-dependent signals, and, if so, whether the restoration of insulin signal was mediated with the inhibition of S6K1 activity stimulated by hyperosmotic stress. In HepG2 cells, oltipraz treatment inhibited insulin receptor substrate (IRS) 1 serine phosphorylation, a marker of insulin resistance, induced by sorbitol-, mannitol-, or sodium chloride-induced hyperosmotic stress. Consequently, this allowed cells to restore insulin signals, which was evidenced by decrease in the ratio of serine to tyrosine phosphorylations of IRS1 and increase in the phosphorylations of Akt and glycogen synthase kinase (GSK) 3β. Hyperosmotic stress markedly activated S6K1; S6K1 activation was completely abolished by oltipraz pretreatment. An experiment using dominant-negative S6K1 supports the essential role of S6K1 in the hyperosmolarity-stimulated phosphorylation of IRS1. Transfection of constitutive active mutant S6K1 eliminated the protective effect of oltipraz on GSK3β phosphorylation, indicating that oltipraz restores insulin signaling by inhibiting S6K1 activation. A variety of synthetic 1,2-dithiole-3-thione derivatives also inhibited S6K1 activity and insulin resistance induced by hyperosmotic stress in HepG2 cells. The results of this study demonstrate that a novel class of 1,2-dithiole-3-thiones improve insulin sensitivity under the condition of hyperosmotic stress, which results from the inhibition of S6K1 activation. Copyright

Full text of DOI:10.1124/mol.107.044347

0026-895X/08/7305-1502–1512$20.00
MOLECULAR PHARMACOLOGY
Copyright © 2008 The American Society for Pharmacology and Experimental Therapeutics
Mol Pharmacol 73:1502–1512, 2008
Vol. 73, No. 5
44347/3327893
Printed in U.S.A.
Abrogation of Hyperosmotic Impairment of Insulin Signaling
by a Novel Class of 1,2-Dithiole-3-thiones through the
Inhibition of S6K1 Activation
Eun Ju Bae, Yoon Mee Yang, and Sang Geon Kim
From Innovative Drug Research Center for Metabolic and Inflammatory Diseases, College of Pharmacy and Research Institute
of Pharmaceutical Sciences, Seoul National University, Seoul, Korea
Received December 12, 2007; accepted February 5, 2008
ABSTRACT
A previous study from this laboratory showed that oltipraz and tyrosine phosphorylations of IRS1 and increase in the phos-
synthetic dithiolethiones prevent tumor necrosis factor-␣– phorylations of Akt and glycogen synthase kinase (GSK) 3␤.
induced hepatic insulin resistance via AMP-activated protein Hyperosmotic stress markedly activated S6K1; S6K1 activation
kinase-dependent p70S6 kinase (S6K) 1 inhibitory pathway. was completely abolished by oltipraz pretreatment. An experi-
This study investigated whether oltipraz and a novel class of ment using dominant-negative S6K1 supports the essential role
1,2-dithiole-3-thiones were capable of preventing insulin resis- of S6K1 in the hyperosmolarity-stimulated phosphorylation of
tance induced by hyperosmotic stress, thereby enhancing in- IRS1. Transfection of constitutive active mutant S6K1 eliminated
sulin-dependent signals, and, if so, whether the restoration of the protective effect of oltipraz on GSK3␤ phosphorylation, indi-
insulin signal was mediated with the inhibition of S6K1 activity cating that oltipraz restores insulin signaling by inhibiting S6K1
stimulated by hyperosmotic stress. In HepG2 cells, oltipraz activation. A variety of synthetic 1,2-dithiole-3-thione derivatives
treatment inhibited insulin receptor substrate (IRS) 1 serine also inhibited S6K1 activity and insulin resistance induced by
phosphorylation, a marker of insulin resistance, induced by hyperosmotic stress in HepG2 cells. The results of this study
sorbitol-, mannitol-, or sodium chloride-induced hyperosmotic demonstrate that a novel class of 1,2-dithiole-3-thiones improve
stress. Consequently, this allowed cells to restore insulin sig- insulin sensitivity under the condition of hyperosmotic stress,
nals, which was evidenced by decrease in the ratio of serine to which results from the inhibition of S6K1 activation.
Insulin is an important regulatory hormone that mediates and liver diseases. In the clinical conditions of diabetes or
energy uptake by inhibiting glucose production in liver and
by increasing glucose uptake into muscle and fat (Saltiel and
Kahn, 2001). Insulin resistance is defined as a profound
dysregulation of insulin signaling system and thus repre-
sents a state of impaired ability of peripheral tissues to
respond to the physiological levels of insulin. Insulin resis-
tance may lead to the development of a variety of metabolic
diseases, such as type 2 diabetes, cardiovascular disorders,
dehydration, hyperosmotic stress accounts, at least in part,
for insulin resistance (Bratusch-Marrain and DeFronzo,
1983; Ennis et al., 1994; Gual et al., 2003a,b). Because
inhibition of insulin resistance restores the ability of insu-
lin to suppress hepatic glucose production and to promote
glucose uptake in peripheral tissues, pharmaceutical in-
terventions to prevent insulin resistance are of great ther-
apeutic interest.
Binding of insulin to the insulin receptor (IR) initiates
signaling cascades by activating its receptor tyrosine kinase.
Most signals of IR are transmitted through complexes assem-
bled around insulin receptor substrate (IRS)-1/2, composed of
multiple interaction domains and phosphorylation motifs
(Myers and White, 1996; Paz et al., 1996). Because IRS1 is
This work was supported by the Korea Science and Engineering Foundation
(KOSEF) grant R11-2007-107-01001-0 funded by the Korea government (Min-
istry of Science and Technology).
Article, publication date, and citation information can be found at
http://molpharm.aspetjournals.org.
doi:10.1124/mol.107.044347.
ABBREVIATIONS: IR, insulin receptor; IRS, insulin receptor substrate; mTOR, mammalian target of rapamycin; TNF, tumor necrosis factor; AMPK,
AMP-activated protein kinase; S6K, p70S6 kinase; MAPK, mitogen-activated protein kinase; ERK, extracellular signal-regulated kinase; JNK,
c-Jun N-terminal kinase; CJ11766, 4-ethyl-5-pyrazin-2-yl-1,2-dithiole-3-thione; CJ11788, 5,6-dihydro-4H-cyclopenta-1,2-dithiole-3-thione;
CJ11792, 4,5,6,7-tetrahydrobenzo-1,2-dithiole-3-thione; CJ11840, 5-benzo[b]thiophene-3-yl-1,2-dithiole-3-thione; CJ11842, 4-methyl-5-phenyl-
1,2-dithiole-3-thione; CJ12064, 5-(6-methoxypyrazinyl)-4-methyl-1,2-dithiole-3-thione; CJ12073, 5-(6-ethoxypyrazin-2-yl)-4-methyl-1,2-dithiole-
3-thione; FBS, fetal bovine serum; I␬B, inhibitor of nuclear factor-␬B; ACC, acetyl-CoA carboxylase; DN-S6K1, dominant-negative S6K1;
CA-S6K1, constitutively active mutant S6K1; GSK3␤, glycogen synthase kinase 3␤.
1502

Dithiolethione Effect on Hyperosmotic Stress
1503
centrally located within the insulin-signaling pathway, de- Cruz Biotechnology (Santa Cruz, CA). Anti-phosphotyrosine (4G10) and
anti-p-Ser312-IRS1 (Ser307 in rodent form) antibodies were purchased
from Millipore (Billerica, MA). Antibodies specific for S6, GSK3␤, ERK,
fects of IRS1 function significantly impair downstream re-
sponses of insulin receptor (Yamauchi et al., 1996). In par-
ticular, IRS1 serine phosphorylation leads to decreases in its
tyrosine phosphorylation (Hotamisligil et al., 1996) and in-
JNK, p38 mitogen-activated protein kinase (MAPK), ␤-actin and phos-
pho-specific antibodies directed against p-ACC(Ser79), p-S6K1(Thr389),
p-S6K1(Thr421/Ser424), p-S6(Ser235/236), p-S6(Ser240/244), p-4E-
crease in the proteasome-mediated degradation (Sun et al.,
BP1(Thr37/41), p-Akt(Ser473), p-GSK3␤(Ser9), p-ERK, p-JNK, and
1999). Hence, serine phosphorylations and/or degradation of
p-p38 MAPK were supplied from Cell Signaling Technology (Dan-
IRS1 play a key role in insulin resistance. Other studies
indicated that insulin resistance by hyperosmolarity is also
mediated by IRS1 dysfunction (Gual et al., 2003a).
vers, MA). Horseradish peroxidase-conjugated goat anti-rabbit
and goat anti-mouse IgGs were provided from Zymed Laboratories
(South San Francisco, CA).
Ligand-activated IR and tyrosine-phosphorylated IRS1 re-
lay signal transmission to the phosphoinositide 3-kinase–Akt
pathway, and activation of this pathway then enhances
mammalian target of rapamycin (mTOR)-p70S6 kinase (S6K) 1
activity. Studies showed that insulin resistance induced by cer-
tain pathophysiological situations (e.g., hyperinsulinemia and
obesity) or excess nutrient availability is linked to marked
increases in S6K1 activity (Um et al., 2004; Tremblay et al.,
2005). A S6K1 knockout experiment strengthened the phys-
iological importance of the inhibitory regulation of mTOR-
S6K1 to insulin signaling (Um et al., 2004), which suggests
Chemical Synthesis of 1,2-Dithiol-3-thione Analogs
1,2-Dithiol-3-thione analogs were synthesized at CJ Central Lab-
oratories (Ichon City, Korea) according to the methods described by
Curphey (2002), as described previously (Bae et al., 2007).
4-Methyl-5-(2-pyrazinyl)-1,2-dithiol-3-thione (Oltipraz). Methyl
2-methyl-3-(pyrazin-2-yl)-3-oxopropionate (40 g, 206 mmol) dis-
solved in 100 ml of toluene was added drop-wise to the mixture of 300
ml of toluene, 350 ml of xylene, and 48 g of phosphorus pentasulfide.
Oltipraz crystal obtained from chemical reaction of the mixture was
filtered, washed, and vacuum-dried (6.37 g, 13.6% yield, Ͼ99.5%
purity). NMR (400 MHz, CDCl3): 2.51(s, 3H), 8.70(d, 1H), 8.80(d,
the mTOR-S6K1 pathway as an attractive therapeutic target 1H), 9.21(s, 1H).
4-Ethyl-5-pyrazin-2-yl-1,2-dithiole-3-thione (CJ11766). Phos-
for insulin resistance.
phorus pentasulfide (3.36 g, 7.56 mmol) suspended in 40 ml each of
toluene and xylene was heated to boiling point, and methyl 2-ethyl-
3-oxo-3-pyrazin-2-yl-propionate (1.5 g, 7.20 mmol) was added to the
solution. After reflux for 4 h, the reaction mixture was cooled to
ambient temperature and filtered through Celite (Sigma). Water
(100 ml) and methanol (20 ml) were added to the filtrate, and the
bilayer solution was neutralized with 28% NH4OH. Separation of
organic layer was followed by washing with brine and drying over
anhydrous MgSO4. Evaporation of filtrate and purification by flash
In our previous study, we found that oltipraz and novel
dithiolethiones prevent TNF␣-induced hepatic insulin resis-
tance through AMP-activated protein kinase (AMPK)-depen-
dent S6K1 inhibition (Bae et al., 2007). Hyperosmotic stress
has been shown to activate the mTOR pathway (Gual et al.,
2003a), which causes insulin resistance. However, the kinase
downstream of mTOR leading to hyperosmotic insulin resis-
tance has not been clarified yet. It seemed to us that the
activation of S6K1 is highly likely to be responsible for insu- chromatography on silica gel offered 0.20 g of 4-ethyl-5-pyrazin-2-yl-
1,2-dithiole-3-thione (11.5% yield). NMR (400 MHz, CDCl3): 1.20(t,
3H), 2.90(q, 2H), 8.75(s, 2H), 8.90(s, 1H).
lin resistance induced by hyperosmotic stress. Hyperosmo-
larity also activates AMPK via a mechanism that remains
unclear (Hayashi et al., 1998; Fryer et al., 2000; Hayashi et
al., 2000), which occurs presumably as an adaptive response
to toxic external stress, as indicated by cell shrinkage with
decrement of cell water volume and the dissipation of mito-
chondrial transmembrane potential (Fumarola et al., 2005).
Therefore, the activation of AMPK by hyperosmotic stress
might not be a beneficial signaling pathway for insulin sen-
5,6-Dihydro-4H-cyclopenta-1,2-dithiole-3-thione (CJ11788).
Phosphorus pentasulfide (4.68 g) suspended in 50 ml each of toluene
and xylene was heated to boiling point, and 3 ml of 2-oxo-cyclopen-
tanecarboxylic acid ethylester was added to the solution. After reflux
for 5 h, the reaction mixture was cooled to ambient temperature, and
filtered through Celite. Water (100 ml) and methanol (20 ml) were
added to the filtrate, and the bilayer solution was neutralized with
28% NH4OH. The organic layer was separated, followed by washing
sitivity. In view of the potential important role of mTOR- with brine and drying over anhydrous MgSO4. Evaporation of filtrate
and purification by flash chromatography on silica gel (ethyl acetate/
n-hexane ϭ 1:30) afforded 0.7 g of 5,6-dihydro-4H-cyclopenta-1,2-
dithiole-3-thione (20% yield). NMR (400 MHz, CDCl3): 2.65–2.80(m,
4H), 2.95–3.00(m, 2H).
4,5,6,7-Tetrahydrobenzo-1,2-dithiole-3-thione (CJ11792). To
a mixture of 3 ml each of tetrahydrofuran and dimethylformamide
was added potassium t-butoxide (460 mg) under argon gas. At room
temperature, cyclohexanone (0.17 ml) in 1 ml tetrahydrofuran was
injected, and the mixture was stirred for 15 min. Addition of 0.11 ml
carbon disulfide was followed by stirring for 30 min, injection of 0.52
S6K1 pathway in hyperosmotic insulin resistance, we exam-
ined whether oltipraz and newly synthesized derivatives
were capable of preventing insulin resistance induced by
hyperosmotic stress, thereby enhancing insulin dependent
signals and, if so, whether the restoration of insulin signal
was mediated with the inhibition of S6K1 activity stimulated
by hyperosmotic stress. Here, we report identification of the
compounds comprising 1,2-dithiole-3-thione as a pharma-
cophore that prevent insulin resistance induced by hyperos-
motic stresses as a consequence of the inhibition of S6K1 ml of hexamethyldisilathian and then stirring for 1 h at room tem-
perature. After cooling to 3°C, 1.80 mmol of hexachloroethane in 2 ml
tetrahydrofuran was added, the reaction solution was stirred for 30
min. Quenching with 2 ml of methanol and evaporation of volatile
materials gave sticky red oil, which was purified by flash chroma-
tography on silica gel (ethyl acetate/n-hexane ϭ 1:30). Then, 77 mg of
4,5,6,7-tetrahydrobenzo-1,2-dithiole-3-thione was obtained as a red
activation.
Materials and Methods
Materials
Oltipraz was provided from CJ Corporation (Seoul, Korea). Sorbitol solid (25% yield). NMR (400 MHz, CDCl3): 1.75–7.90(m, 4H), 2.55–
was purchased from Sigma Chemicals (St. Louis, MO). [␥-32P]ATP 2.60(m, 2H), 2.80–2.85(m, 2H).
(3000 mCi/mmol) was supplied from PerkinElmer Life and Analytical
5-Benzo[b]thiophene-3-yl-1,2-dithiole-3-thione (CJ11840). To a
Sciences (Waltham, MA). Antibodies directed against IRS1, IR␤, S6K1, solution of potassium t-butoxide (1.07 g) in 20 ml of tetrahydrofuran
and inhibitor of nuclear factor-␬B (I␬B) ␣ were obtained from Santa was injected 800 mg of 1-benzo[b]thiophen-3-yl-ethanone dissolved in 5

1504
Bae et al.
ml of tetrahydrofuran. After stirring for 15 min, 0.3 ml of carbon 4 days. To assess IRS1 serine phosphorylation, the cells were deprived
disulfide was added, and the mixture was stirred for 30 min followed by of serum for 24 h, pretreated with oltipraz for 1 h, and subsequently
addition of 2.83 mg methyl iodide and stirring at room temperature to exposed to 600 mM sorbitol, 600 mM mannitol, or 150 mM NaCl for the
complete reaction. Work-up was performed as a sequence of dilution time periods (indicated in the figure legends) in the continuing presence
with 50 ml of methylene chloride, neutralization with saturated NH4Cl of oltipraz. To set an insulin resistance model, the cells were incubated
solution, and separation of organic layer. After concentration, purifica- with sorbitol for 50 min and then treated with 10 nM insulin for 10 min.
tion by flash chromatography on silica gel (ethyl acetate/n-hexane ϭ Cell lysates were prepared according to methods published previously
1:10) gave 700 mg of yellow solid as a bis-methylsulfide adduct. Disso- (Kang et al., 2003). In brief, cells were centrifuged at 3000g for 3 min
lution of this solid in 15 ml each of toluene and xylene, followed by and allowed to swell after the addition of lysis buffer. The samples were
addition of 0.58 g of phosphorus pentasulfide, made a suspension to be
refluxed for 3 h. The reaction mixture was cooled to ambient tempera-
ture, and filtered through Celite. Water (50 ml) was added to the
filtrate, and the bilayer solution was neutralized with 28% NH4OH.
Separation of organic layer was followed by washing with brine and
drying over anhydrous MgSO4. Evaporation of filtrate and purification
by flash chromatography on silica gel (ethyl acetate/n-hexane ϭ 1:30)
offered 185 mg of 5,6-dihydro-4H-cyclopenta-1,2-dithiole-3-thione
(15.3% yield). NMR (400 MHz, CDCl3): 7.45–7.55(m, 3H), 7.95(d,
1H), 8.00(s, 1H), 8.10(d, 1H).
centrifuged at 10,000g for 10 min to obtain lysates and stored Ϫ70°C
until use.
Immunoblot Analysis
Immunoblot analysis were performed according to the previously
published procedures. Proteins of interest in lysates were resolved
using 6, 9, or 12% gels and developed using ECL chemiluminescence
system (GE Healthcare, Chalfont St. Giles, Buckinghamshire, UK).
Immunoprecipitation
4-Methyl-5-phenyl-1,2-dithiole-3-thione (CJ11842). To
a
To assess tyrosine-phosphorylated IR␤ or IRS1, cell lysates (250 ␮g
each) were incubated with anti-IR␤ or anti-IRS1 antibody overnight at
4°C. The antigen-antibody complex was immunoprecipitated after in-
cubation for 2 h at 4°C with protein G-agarose. Immune complexes were
solubilized in 2ϫ Laemmli buffer. Protein samples were resolved and
immunoblotted with anti-phosphotyrosine antibody.
mixture of 4 ml of tetrahydrofuran and 2 ml of N,NЈ-dimethylpro-
pyleneurea was added 407 mg of potassium t-butoxide under
argon gas. At room temperature, 0.22 ml of 1-phenyl-propan-1-one
was injected, and the mixture was stirred for 15 min. Addition of
0.11 ml of carbon disulfide was followed by stirring for 15 min,
injection of 0.52 ml of hexamethyldisilathian, and then stirring for
15 min at room temperature. After cooling to 3°C, 391 mg of
hexachloroethane in 2 ml of tetrahydrofuran was added, and the
reaction solution was stirred for 30 min. Quenching with 2 ml of
methanol and evaporation of volatile materials gave sticky red oil,
which was purified by flash chromatography on silica gel (ethyl
acetate/n-hexane ϭ 1:60), and then 78 mg of 4-methyl-5-phenyl-
1,2-dithiole-3-thione was obtained as a red solid (21% yield). NMR
(400 MHz, CDCl3): 2.25(s, 3H), 7.55–7.60(m, 5H).
5-(6-Methoxypyrazinyl)-4-methyl-1,2-dithiole-3-thione
(CJ12064). 5-(6-Chloro-pyrazinyl)-4-methyl-1,2-dithiole-3-thione (200
mg) and 0.32 g of potassium carbonate were dissolved in 10 ml meth-
anol, and the mixture was refluxed for 2 h. Evaporation of solvent was
followed by the addition of saturated NH4Cl aqueous solution and the
extraction with methylene chloride. After concentration, titration with
n-hexane yielded 140 mg of a red solid 5-(6-methoxypyrzin-2-yl)-4-
methyl-1,2-dithiole-3-thione (71.2% yield). NMR (400 MHz, CDCl3):
2.55(s, 3H), 4.40(s, 3H), 8.35(s, 1H), 8.55(s, 1H).
5-(6-Ethoxypyrazin-2-yl)-4-methyl-1,2-dithiole-3-thione
(CJ12073). 5-(6-Chloro-pyrazin-2-yl)-4-methyl-1,2-dithiole-3-thione
(200 mg, 0.77 mmol) and 0.32 g of potassium carbonate were dis-
solved in 10 ml of ethanol, and the mixture was refluxed for 2 h.
Evaporation of solvent was followed by the addition of saturated
NH4Cl aqueous solution and the extraction with methylene chloride.
After concentration, titration with n-hexane yielded red solid 140 mg
of 5-(6-ethoxypyrazin-2-yl)-4-methyl-1,2-dithiole-3-thione (67.5%
yield). NMR (400 ⌴⌯z, CDCl3): 1.45(t,J ϭ 7Hz,3H), 2.55(s,3H),
4.45(q,J ϭ 7Hz,2H), 8.35(s,1H), 8.55(s,1H).
Plasmid Transfection
The S6K1 expression constructs PRK5 myc-tagged E2BQ (domi-
nant-negative, DN-S6K1) and D3E (constitutively active, CA-S6K1)
were supplied from Dr. J. Han (Sungkyunkwan University, Suwon,
Korea), originally provided by Dr. G. Thomas (Friedrich Miescher
Institut, Basel, Switzerland) (Pearson et al., 1995; Jefferies et al.,
1997). HepG2 cells (5ϫ105 cells/well) were replated in six-well plates
overnight, serum-starved for 6 h, and transfected with DN-S6K1 or
CA-S6K1 (1 ␮g) in the presence of Lipofectamine Reagent (Invitro-
gen, Carlsbad, CA). The transfected cells were incubated in Eagle’s
minimum essential medium containing 1% FBS for 24 h and exposed
to sorbitol in the presence or absence of oltipraz.
S6K1 Activity
S6K1 activities were determined by immunocomplex kinase as-
says. S6K1 was immunoprecipitated in lysates of HepG2 cells
treated with 30 ␮M oltipraz or other dithiolethiones for 1 h, and
subjected to kinase reactions. To assess the direct in vitro effects of
oltipraz, S6K1 immunoprecipitates obtained from the cell lysates
were incubated with oltipraz in a kinase reaction mixture. Kinase
reactions were initiated by adding S6 RSK substrate peptide (5 ␮g
per assay; Santa Cruz Biotechnology) and 1 ␮Ci of [␥-32P]ATP (3000
mCi/mmol) to 20 ␮l of kinase buffer containing 25 mM Tris-HCl, pH
7.4, 10 mM MgCl2, 25 mM ␤-glycerophosphate, 1 mM Na3VO4, 2 mM
dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, 1 ␮g/ml leupep-
tin, and 200 ␮M ATP, and continued for 20 min at 30°C. After brief
centrifugation, the supernatants of reaction mixture were spotted
onto p81 Whatman phosphocellulose paper (Whatman Bioscience,
Clifton, NJ). The paper was washed with 0.8% phosphoric acid three
times for 15 min each and subsequently with 90% ethanol for 5 min.
The membrane was dried and the radioactivity of phosphorylated
substrate was measured using a ␤-counter (PerkinElmer Life and
Analytical Sciences, Waltham, MA).
Cell Culture and Drug Treatments
HepG2 hepatocyte, C2C12 myoblast, and 3T3-L1 preadipocyte cell
lines were purchased from American Type Culture Collection (Man-
assas, VA). The cells were maintained in Dulbecco’s modified Eagle’s
medium containing 10% fetal bovine serum (FBS), 50 units/ml pen-
icillin, and 50 ␮g/ml streptomycin (complete medium). C2C12 cells
were differentiated to myotubes by incubating in Dulbecco’s modified
Eagle’s medium containing 2% FBS for 6 to 8 days. 3T3-L1 preadi-
pocytes were differentiated to adipocytes, as described previously
Glucose Uptake
Differentiated myotubes and adipocytes were incubated with each
(Bae and Kim, 2005). In brief, 2 day-postconfluent preadipocytes agent of interest in the presence or absence of 600 mM sorbitol for 50
were treated with 0.5 mM isobutyl-1-methylxanthine, 1 ␮M dexa- min. After the change of culture medium with phosphate-buffered
methasone, and 1 ␮g/ml insulin for 2 days. The cells were then incu- saline, cells were treated with 100 nM insulin for 10 min, and glucose
bated in the complete medium containing insulin for 2 additional days uptake was determined by 2-deoxy-D-[2,6-3H]glucose incorporation
and thereafter exposed to the complete medium without insulin for 2 to into the cells for 20 min. The reaction was terminated by adding

Dithiolethione Effect on Hyperosmotic Stress
1505
ice-cold phosphate-buffered saline. After washing three times, the ment with 30 to 60 ␮M oltipraz effectively blocked the IRS1
cells were dissolved in 0.5 N NaOH containing 0.1% SDS. Radioac-
tivity was measured using a scintillation counter. Specific uptake
was assessed by subtracting nonspecific uptake, which was mea-
sured by incubating the cells with 20 ␮M cytochalasin B. Values
were normalized by protein amounts.
serine phosphorylation (Fig. 1B). Because 30 ␮M oltipraz was
highly active, we chose the concentration in the subsequent
experiments.
Studies have shown that 300 to 600 mM mannitol or 150
mM sodium chloride also induces hyperosmotic stress (Ku¨ltz
et al., 1997; Galvez et al., 2003). Oltipraz similarly inhibited
IRS1 serine phosphorylation induced by hyperosmolar man-
nitol or sodium chloride (Fig. 1C). These results provide
evidence that oltipraz inhibited IRS1 Ser312 phosphorylation
Data Analysis
Scanning densitometry was performed with Image Scan and Anal-
ysis System (Alpha Innotech, San Leandro, CA). Student’s t test was
used to assess significant differences between control and treatment
group and one-way analysis of variance procedures among treatment against hyperosmotic stress.
groups. The criterion for statistical significance was set at p Ͻ 0.05
or 0.01.
Oltipraz Protection of Insulin Signaling. Next, we
measured insulin signaling in the cells treated with sorbitol
or oltiprazϩsorbitol. Exposure of the cells to insulin caused
increases in IR␤/IRS1 tyrosine phosphorylations, and Akt
serine phosphorylation (Fig. 2A, upper). Concomitant sorbi-
tol treatment unchanged the total or tyrosine-phosphory-
lated IR␤ levels in insulin-treated cells, whereas the levels of
IRS1 tyrosine phosphorylation and Akt phosphorylations
were both decreased. We also found that IRS1 serine phos-
phorylation was reciprocal to its tyrosine phosphorylation.
These data suggest that hyperosmotic stress does not impair
the tyrosine kinase activity of insulin receptor, which is con-
sistent with the previous reports (Chen et al., 1999; Gual et
al., 2003a). Oltipraz treatment before sorbitol exposure al-
lowed the cells to recover insulin-dependent IRS1 tyrosine
phosphorylation with a commensurate decrease in the serine
phosphorylation. Therefore, the ratios of IRS1 serine to
tyrosine phosphorylations significantly decreased after olti-
Results
Inhibition of Hyperosmotic Stress-Induced IRS1
Serine Phosphorylation by Oltipraz. Human IRS1 Ser312
phosphorylation, corresponding to Ser307 in the rodent form,
is a representative molecular marker of insulin resistance
(Aguirre et al., 2000, 2002). Because insulin resistance in-
duced by hyperosmolarity involves the IRS1 serine phospho-
rylations (Gual et al., 2003a), we first investigated the effect
of oltipraz on hyperosmotic stress-induced IRS1 Ser312 phos-
phorylation in HepG2 cells. A time-course study showed that
IRS1 serine phosphorylation began to increase as early as 10
min after 600 mM sorbitol treatment, which maximally in-
creased at 1 h and maintained for at least 3 h (Fig. 1A).
Pretreatment of the cells with oltipraz inhibited sorbitol-
induced IRS1 Ser312 phosphorylation at the time points ex-
amined. The dose-response study indicated that pretreat- praz pretreatment (Fig. 2A, lower). Accordingly, oltipraz pre-
Fig. 1. Oltipraz inhibits IRS1 Ser312 phosphorylation of induced by hyperosmotic stress. A, the time responses of hyperosmotic stress-dependent IRS1
Ser312 phosphorylation. Immunoblot analysis was performed in lysates of HepG2 cells treated with 30 ␮M oltipraz for 1 h and subsequently exposed
to 600 mM sorbitol for the indicated time period. The relative levels of IRS1 Ser312 phosphorylation were assessed by scanning densitometry of the
immunoblots. Data represent mean Ϯ S.E. with at least 3 separate experiments (ءء, p Ͻ 0.01, significant compared with vehicle control; significant
compared with respective sorbitol treatment; ##, p Ͻ 0.01; control level ϭ 1). B, the dose-responses of oltipraz treatment on sorbitol-induced IRS1
Ser312 phosphorylation. The cells were treated with 10 to 60 ␮M oltipraz for 1 h and subsequently exposed to 600 mM sorbitol for 1 h. The values
represent mean Ϯ S.E. with three separate experiments (ءء, p Ͻ 0.01, significant compared with vehicle control; ##, p Ͻ 0.01, significant compared
with sorbitol alone). C, oltipraz’s inhibition of IRS1 Ser312 phosphorylation induced by hyperosmolar mannitol or sodium chloride. The cells were
treated with 30 ␮M oltipraz for 1 h and then exposed to 600 mM mannitol or 150 mM sodium chloride for 1 h in the continuing presence of oltipraz.

1506
Bae et al.
treatment enhanced insulin-dependent Akt phosphorylation phosphorylation, which represents cellular S6K1 activity
against hyperosmotic stress. (Fig. 3A, bottom). Next, we determined the effect of oltipraz
Because sorbitol impairs the activity of Akt responsible for on sorbitol-induced S6K1 activation. Treatment of HepG2
the phosphorylation of downstream GSK3␤, insulin-stimu- cells with sorbitol for 10 minϳ1 h caused S6K1 activation,
lated GSK3␤ phosphorylations became defective after sorbi- the extent of which was similar to that induced by insulin (10
tol treatment. Oltipraz treatment allowed the cells to restore nM, 30 min) (Fig. 3B, top). Sorbitol-induced S6K1 activation
insulin-dependent GSK3␤ phosphorylation in a concentra- was significantly inhibited by simultaneous oltipraz treat-
tion-dependent manner (Fig. 2B, top and bottom).
ment throughout the time points examined. Consistent with
The Effects of Oltipraz on Hyperosmotic Stress-In- this result, S6 phosphorylation markedly increased by sorbi-
duced Alterations of Cell Signaling. Because IRS1 Ser tol treatment was almost completely inhibited by oltipraz as
312 phosphorylation is catalyzed by multiple serine/threonine a function of time (Fig. 3B, bottom).
kinases, including S6K1, JNK, ERK, and I␬B kinase complex
Further immunoblot analysis showed that hyperosmotic
(Aguirre et al., 2000, 2002; Engelman et al., 2000; Gao et al., sorbitol promoted S6K1 phosphorylations at Thr389 or
2002; Um et al., 2004), we next determined the activation Thr421/Ser424 (Fig. 3C). Concomitant oltipraz treatment es-
statuses of the kinase signals in the cells treated with sorbi- sentially eliminated phosphorylations of S6K1 induced by
tol or oltipraz ϩ sorbitol. S6K1 immunocomplex kinase as- hyperosmotic stress. 4E-BP1 phosphorylation, which is also
says revealed that oltipraz dose-dependently inhibited the catalyzed by mTOR, was also prevented by oltipraz treat-
basal S6K1 kinase activity. Oltipraz at 3 ␮M began to inhibit ment. These results indicate that oltipraz affects the mTOR
the cellular S6K1 kinase activity, and the inhibition of S6K1 pathway.
activity by 30 ␮M oltipraz was comparable with that caused
Previous observations indicated that hyperosmotic stress
by rapamycin, an mTOR inhibitor (Fig. 3A, top). Incubation activates JNK, ERK, and p38 MAPK (Galcheva-Gargova et
of S6K1 immunoprecipitate, which was prepared from un- al., 1994; Han et al., 1994; Kayali et al., 2000), which
treated HepG2 cell lysates, with oltipraz in vitro resulted in prompted us to determine the effect of oltipraz on these
no change in S6K1 activity (data not shown), confirming our kinases. Hyperosmotic sorbitol treatment increased the
previous observation that oltipraz does not directly inhibit phosphorylations of JNK, ERK, and p38 MAPK, which were
S6K1. Likewise, oltipraz notably decreased the basal S6 unaffected by oltipraz treatment (Fig. 3D). Hyperosmotic
Fig. 2. Oltipraz protects insulin signals against hyperosmotic stress. A, phosphorylations of IR␤, IRS1, and Akt. HepG2 cells were treated with 30 ␮M
oltipraz for 1 h, exposed to 600 mM sorbitol for 50 min, and then continuously incubated with 10 nM insulin for 10 min. IR␤ or IRS1 immunopre-
cipitates were immunoblotted with anti-phosphotyrosine antibody. IRS1 Ser312 phosphorylations or Akt Ser473 phosphorylations were measured in cell
lysates. The bar graphs represent the relative ratios of serine- to tyrosine-phosphorylations of IRS1. Data represent mean Ϯ S.E. with at least three
separate experiments (ءء, p Ͻ 0.01, significant compared with insulin; ##, p Ͻ 0.01, significant compared with sorbitol or sorbitol ϩ insulin). B, GSK3␤
Ser phosphorylations. The relative levels of p-GSK3␤ Ser9 in cells treated, as indicated, were depicted in the bar graph. Data represent mean Ϯ S.E.
with at least three separate experiments (ءء, p Ͻ 0.01, significant compared with insulin; ##, p Ͻ 0.01, significant compared with sorbitol ϩ insulin).

Dithiolethione Effect on Hyperosmotic Stress
1507
stress strongly activates nuclear factor-␬B through proteaso- at multiple serine residues (e.g., Ser312 and Ser636/639 in
mal degradation of I␬B␣ (Eisner et al., 2006). Oltipraz did not human; Ser302, Ser307, and Ser632/635 in rodents)(Harrington
change I␬B␣ degradation, which is consistent with previous et al., 2004; Um et al., 2004; Tremblay et al., 2005). Having
observations (Bae et al., 2007). As an effort to assess the identified oltipraz inhibition of S6K1 activation, we next
effects of oltipraz on AMPK that is also activated by hyper- examined the possible role of S6K1 inhibition in sorbitol-
osmotic stress, we determined the phosphorylations of ACC, induced IRS1 Ser312 phosphorylation. Exposure of HepG2
an AMPK substrate. Sorbitol increased ACC phosphoryla- cells to sorbitol for 1 h resulted in IRS1 serine phosphoryla-
tion, which was not abolished by oltipraz treatment. All of tion, which was repressed by dominant-negative mutant
these results indicate that oltipraz selectively inhibited the S6K1 (DN-S6K1) (Fig. 4A). Inhibition of S6K1 phosphoryla-
activation of S6K1 among the cellular kinases examined.
tion after DN-S6K1 transfection verified transfection effi-
The Role of S6K1 Inhibition in Hyperosmotic Stress- ciency. Rapamycin also inhibited IRS1 serine phosphoryla-
Induced IRS1 Serine Phosphorylation. It has been tion induced by sorbitol (Fig. 4B). These results support
shown that S6K1 activation causes phosphorylation of IRS1 the concept that the inhibition of S6K1 activation leads to
Fig. 3. Oltipraz inhibits the basal and hyperosmolar activation of S6K1. A, inhibition of S6K1 activity by oltipraz. HepG2 cells were treated with 3
to 30 ␮M oltipraz or 3 to 10 nM rapamycin for 1 h. S6K1 kinase activities in cell lysates were determined toward S6 RSK substrate peptide by
immunocomplex kinase assays. Data represent mean Ϯ S.E. with at least three separate experiments (significant compared with vehicle-treated
control, ء, p Ͻ 0.05; ءء, p Ͻ 0.01; control level ϭ 1). Time responses of S6 phosphorylations after oltipraz treatment (bottom). HepG2 cells were
incubated with 30 ␮M oltipraz for the indicated time periods, and the cell lysates were subjected to immunoblotting with antibody specific for p-S6
(Ser235/236). B, inhibition of sorbitol-induced S6K1 activity by oltipraz. HepG2 cells were treated with 600 mM sorbitol for the indicated time period
with or without 30 ␮M oltipraz pretreatment for 1 h. S6K1 activation by insulin (10 nM, 30 min) was included as a positive control. Data represent
mean Ϯ S.E. with at least three separate experiments (ء, p Ͻ 0.05, ءء, p Ͻ 0.01, significant compared with vehicle-treated control; ##, p Ͻ 0.01,
significant compared with respective sorbitol treatment; control level ϭ 1). The phosphorylations of S6 were determined in HepG2 cells that had been
exposed to 600 mM sorbitol with or without oltipraz pretreatment for 1 h (bottom). C, inhibition of sorbitol-induced of S6K1 or 4E-BP1 phosphory-
lations by oltipraz. Immunoblot analyses were conducted in lysates of HepG2 cells treated as described in B. D, the effects of oltipraz treatment on
sorbitol-induced JNK, ERK, p38 MAPK, ACC phosphorylations, and degradation of I␬B␣. Results were confirmed by repeated experiments.

1508
Bae et al.
the inhibition of hyperosmotic stress-induced IRS1 serine treatment allowed the cells to restore AMPK activity inhib-
phosphorylation. ited by TNF␣ (Fig. 6A). The increase in ACC phosphorylation
Abrogation by Constitutively Active Mutant S6K1 of by oltipraz was weaker than that caused by sorbitol. We also
Oltipraz’s Protection of Insulin-Induced GSK3␤ Phos- observed that oltipraz treatment did not further increase
phorylation. To verify the functional role of S6K1 inhibition ACC phosphorylation in the cells treated with hyperosmolar
in oltipraz’s restoration of insulin sensitivity under hyperos- sorbitol. These results suggest that S6K1 activation by TNF␣
motic stress condition, we further determined the effect of was associated with AMPK activity and that the pathway
constitutively active mutant S6K1 (CA-S6K1) on GSK3␤ responsible for mTOR-S6K1 activation by hyperosmolarity
phosphorylation. In this experiment, HepG2 cells were trans- may differ from that by TNF␣. The effects of oltipraz against
fected with the plasmid encoding CA-S6K1 and subsequently hyperosmolarity and TNF␣ challenge were summarized in
treated with sorbitol in combination with insulin. CA-S6K1 Fig. 6B.
transfection virtually eliminated the protective effect of olti-
The Effects of 1,2-Dithiole-3-thione Derivatives on
praz on insulin-dependent GSK3␤ phosphorylation (Fig. 5), Cellular S6K1 Activity and GSK3␤ Phosphorylations.
which indicates that the protection of insulin responses by Because our results indicated that oltipraz improves insulin
oltipraz against hyperosmolar sorbitol is mediated with in- actions under the conditions of hyperosmotic stress through
hibition of S6K1 activation. Immunoblots for c-Myc and phos- the inhibition of S6K1 activation, we tested the effects of
phorylated S6 confirmed transfection efficiency. Our findings 1,2-dithiole-3-thione derivatives to find additional new com-
concerning the inhibition of IRS1 serine phosphorylation and pounds capable of inhibiting S6K1 activation. Among the
of S6K1 activation proved the pharmacologic effect of oltipraz synthetic 1,2-dithiole-3-thione derivatives (Fig. 7A), we iden-
on hyperosmotic stress-induced insulin resistance.
tified several compounds that inhibited the basal S6K1 ac-
The Effect of Oltipraz on ACC Phosphorylation in tivity and would thus be expected to inhibit hyperosmotic
Cells Treated with TNF␣. In an additional experiment, stress-induced insulin resistance. Figure 7B shows that
TNF␣ was compared with hyperosmolar sorbitol for the effect CJ11766 and CJ11788, CJ11792, CJ11842, or CJ12064 sig-
of oltipraz on ACC phosphorylation in HepG2 cells. Oltipraz nificantly inhibited the activity of S6K1. In addition, most of
the compounds were capable of suppressing sorbitol-induced
IRS1 serine phosphorylation (data not shown). To correlate
the inhibition of cellular S6K1 activity by these compounds
with the recovery of insulin action, we monitored insulin-
dependent GSK3␤ phosphorylation in the cells. As expected,
synthetic 1,2-dithiole-3-thione derivatives capable of inhibit-
ing S6K1 activity allowed the cells to recover the insulin
signaling (Fig. 7C).
The Effects of Oltipraz and Other 1,2-Dithiole-3-
Thione Derivatives on Cell Signaling and Glucose Up-
take in C2C12 Myotubes and 3T3-L1 Adipocytes. Next,
we assessed the effects of oltipraz on AMPK and S6K1 in
other types of cells. In differentiated C2C12 myotubes, olti-
Fig. 4. S6K1 induces IRS Ser312 phosphorylation in response to hyper-
osmotic stress. A, reversal of sorbitol-induced IRS Ser phosphorylation by
DN-S6K1. HepG2 cells transfected with pcDNA or a plasmid encoding
DN-S6K1 (1 ␮g) were treated with 600 mM sorbitol for the indicated time
period. Data represent mean Ϯ S.E. with at least three separate experi-
ments (ءء, p Ͻ 0.01, significant compared with vehicle-treated control; ##,
p Ͻ 0.01, significant compared with respective sorbitol treatment). B, the
effect of rapamycin, an mTOR inhibitor, on sorbitol-induced IRS1- Ser312
phosphorylation. HepG2 cells were treated with 10 nM rapamycin for 1 h
and exposed to 600 mM sorbitol for 1 h in the continuing presence of
rapamycin.
Fig. 5. CA-S6K1 abrogates oltipraz’s protection of GSK3␤ phosphoryla-
tion. HepG2 cells transfected with pcDNA or a plasmid encoding CA-
S6K1 (1 ␮g) were treated with 30 ␮M oltipraz for 1 h, exposed to 600 mM
sorbitol for 50 min, and then continuously incubated with 10 nM insulin
for 10 min. Data represent mean Ϯ S.E. with at least three separate
experiments (##, p Ͻ 0.01). Immunoblottings for c-Myc and p-S6 verify
the expression of CA-S6K1.

Dithiolethione Effect on Hyperosmotic Stress
1509
praz increased AMPK activity 4 or 8 h after treatment, as
assessed by ACC phosphorylation (Fig. 8A, left). However,
the extent of S6 phosphorylation was unchanged. In differ-
entiated 3T3-L1 adipocytes that had been incubated in a
medium deprived of serum for 12 h, oltipraz treatment only
slightly enhanced ACC phosphorylation (i.e., 1–4 h) but
failed to decrease S6 phosphorylation (Fig. 8A, right). Be-
cause serum starvation alone increased ACC phosphoryla-
tions in the cell, the increase in ACC phosphorylation was not
distinct 8 h after treatment. These results suggest that olti-
praz might not inhibit S6K1 activity in differentiated myo-
tubes or adipocytes despite AMPK activation.
To additionally assess insulin responses in these cells, we
determined the functional effects of oltipraz on Akt phos-
phorylation and glucose uptake in C2C12 myotubes and
3T3-L1 adipocytes. Incubation of the cells with sorbitol com-
pletely inhibited insulin-dependent Akt phosphorylation,
whereas concomitant oltipraz treatment partly enhanced the
insulin-dependent Akt phosphorylation (Fig. 8B, left). Olti-
praz was also active in restoring insulin-dependent glucose
uptake in sorbitol-treated C2C12 myotubes. In 3T3-L1 adi-
pocytes, sorbitol impairment of insulin-dependent Akt phos-
phorylation and glucose uptake were not notably or sig-
nificantly changed by oltipraz treatment (Fig. 8B, right),
suggesting that oltipraz restores insulin signals in a cell
type-specific manner.
Discussion
In this study, we found that oltipraz treatment effectively
abrogated the IRS1 serine phosphorylation stimulated by
hyperosmotic stress elicited by sorbitol, mannitol, or sodium
chloride. Oltipraz’s inhibition of IRS1 serine phosphorylation
led to the recovery of the impaired insulin signaling cascade
at the levels of IRS1 tyrosine phosphorylations and Akt/
GSK3␤ phosphorylations. These results are in agreement
with our previous findings showing that oltipraz prevented
hepatic insulin resistance in mice challenged with endotoxin
(or TNF␣) or in lepob/ob mice and in mice fed a high-fat diet
(Bae et al., 2007). In an additional experiment, we tried to
establish an in vivo hyperosmotic shock model. When rats
were infused 25% sorbitol solutions for a 10-min period, blood
glucose levels increased from 88 to 293 mg/dl at 6 h after the
infusion (n ϭ 3, data not shown), which was comparable with
the data reported previously (Kaya et al., 2004). However,
the liver tissues obtained 10 h after sorbitol infusion exhib-
ited severe injuries. It seemed that the soaring blood glucose
As a continuing effort to confirm the functional effective-
ness of the 1,2-dithiole-3-thione compounds, we finally
examined glucose uptake into C2C12 myotubes. Among the
compounds, CJ11792 or CJ12073 significantly enhanced
insulin-dependent glucose uptake into C2C12 myotubes un-
der hyperosmotic stress (Fig. 8C). Taken as a whole, these
results demonstrate that treatment with oltipraz or certain
1,2-dithiole-3-thione derivatives allows cells to restore insu-
lin signaling against hyperosmotic stress and thus function-
ally improves insulin response.
Fig. 7. The compounds containing 1,2-dithiole-3-thione as a pharmacoph-
ore inhibit S6K1 activity and thereby protect the ability of insulin to
induce GSK3␤ phosphorylation. A, chemical structures of 1,2-dithiole-3-
thione analogs. B, immunocomplex kinase assays for S6K1. S6K1 activity
was determined in S6K1 immunoprecipitates prepared from HepG2 cells
that had been treated with 30 ␮M of each 1,2-dithiole-3-thione for 1 h.
Data represent mean Ϯ S.E. with at least three separate experiments (ء,
p Ͻ 0.05; ءء, p Ͻ 0.01, significant compared with vehicle-treated control;
control level ϭ 1). C, GSK3␤ phosphorylations. The cells were treated
with 30 ␮M dithiolethione derivative for 1 h, exposed to 600 mM sorbitol
for 50 min, and then continuously incubated with 10 nM insulin for 10
min. Phosphorylated GSK3␤ was immunoblotted in cell lysates.
Fig. 6. The effect of oltipraz on ACC phosphorylation in TNF␣-treated
HepG2 cells. A, immunoblotting for ACC phosphorylation. HepG2 cells
were treated with 10 ng/ml TNF␣ for the indicated time period with or
without 30 ␮M oltipraz pretreatment for 1 h and subjected to immunoblot
analysis. Results were confirmed by repeated experiments. B, a summary
of the effects of oltipraz on HepG2 cells exposed to hyperosmolarity or
TNF␣. Arrows represent increase or decrease in the level of phosphory-
lation of target protein.

1510
Bae et al.
contents may have resulted from strong activation of sympa- S6K1 activity ϳ3-fold in 3T3-L1 adipocytes (Chen et al.,
thetic nervous system and/or severe liver damage. In clinical 1999). However, it was also claimed that hyperosmolarity
situations, hyperosmotic hyperglycemic state is manifested inhibited the basal S6K1 activity in CV1 and Jurkat cells
by serum osmolarity greater than 320 mOsM and blood glu- (Parrott and Templeton, 1999; Fumarola et al., 2005), and
cose levels greater than 600 mg/dl (Kitabchi et al., 2004), insulin-dependent S6K1 activation in 3T3-L1 adipocytes or
with a dramatic increase in the prevalence of type 2 diabetes H4IIE cells (Chen et al., 1999; Lornejad-Scha¨fer et al., 2003).
(Ennis et al., 1994). Although the hyperosmolarity induced Despite the reported cell type-specific controversial effects of
by 600 mM sorbitol in the present cell culture study is not hyperosmolarity on S6K1, our results lend support to the
attainable in a physiological state, our experimental results concept that the mTOR-dependent IRS1 serine phosphoryla-
are worthy of understanding the molecular mechanism of tion results from S6K1 activation.
hyperosmotic insulin resistance.
mTOR, once activated, phosphorylates and activates S6K1,
In the present study, we observed that hyperosmolarity which then increases S6 phosphorylations. As expected, S6
causes S6K1 activation in HepG2 cells, which was evidenced phosphorylations enhanced by hyperosmotic stress were
by increases in both S6K1 immune-complex kinase activity completely abrogated by oltipraz pretreatment. In addition,
and S6 phosphorylation. It has been reported that a high phosphorylation of 4E-BP1, an alternative direct target of
concentration of sorbitol elicits IRS1 dysfunction by an mTOR, could be inhibited by oltipraz. The principal finding of
mTOR-dependent pathway (Gual et al., 2003a). Because this study therefore relates to the inhibitory effect of oltipraz
mTOR targets S6K1, it could be expected that the conditions on hyperosmotic S6K1 activation, and this observation is
of hyperosmotic stress led to S6K1 activation. Chen et al. strengthened by the result of our DN-S6K1 experiment.
have shown that hyperosmotic stress increased the basal These observations lend support to the notion that oltipraz
Fig. 8. Oltipraz or other 1,2-dithiole-3-thione
derivatives functionally improve insulin re-
sponses against hyperosmotic stress in C2C12
myotubes. A, the effect of oltipraz on the phos-
phorylations of ACC and S6. Immunoblot anal-
ysis was performed in cell lysates obtained from
C2C12 myotubes (left) or 3T3-L1 adipocytes
(right) treated with vehicle or 30 ␮M oltipraz for
the indicated time period. B, insulin-dependent
Akt phosphorylation (Ser473) and glucose uptake
in C2C12 myotubes (left) and in 3T3-L1 adipo-
cytes (right). Myotubes or adipocytes were
treated with 30 ␮M oltipraz for 1 h, exposed to
600 mM sorbitol for 50 min, and then continu-
ously incubated with 10 nM insulin for 10 min.
Akt phosphorylations were determined in cell
lysates. Glucose uptake was measured in cells
treated with 100 nM insulin for 10 min after
sorbitol treatment with or without oltipraz pre-
treatment (1 h). Data represent mean Ϯ S.E.
with at least 4 separate experiments (ءء, p Ͻ
0.01, significant compared with vehicle control;
#, p Ͻ 0.05, significant compared with sorbitol ϩ
insulin). C, glucose uptake in C2C12 myotubes
treated with 30 ␮M concentrations of each com-
pound. Metformin (1 mM, 1 h) was used as a
positive control. Data represent mean Ϯ S.E.
with at least four separate experiments (ءء, p Ͻ
0.01, significant compared with vehicle control;
#, p Ͻ 0.05; ##, p Ͻ 0.01, significant compared
with sorbitol ϩ insulin).

Dithiolethione Effect on Hyperosmotic Stress
1511
selectively affects S6K1 activity probably through inhibition thiones (i.e., 4,5,6,7-tetrahydrobenzo-1,2-dithiole-3-thione
of the mTOR pathway. Our current finding that JNK1/2, (CJ11792) and 5-(6-ethoxypyrazin-2-yl)-4-methyl-1,2-dithiole-
ERK1/2, p38 MAPK, and nuclear factor-␬B pathways acti- 3-thione (CJ12073)) significantly increased insulin-dependent
vated by hyperosmotic stress were unchanged by oltipraz glucose uptake impaired by hyperosmotic stress in C2C12
supports the relatively specific inhibition of oltipraz on hy- myotubes. Complete dose response, and pharmacodynamic
perosmotic activation of S6K1. Oltipraz’s inhibition of S6K1 and pharmacokinetic studies of these candidates remain to
led to the recovery of the impaired insulin signaling cascade be established in the future.
at the level of IRS1 tyrosine phosphorylation and Akt and
We have shown that oltipraz and other dithiolethiones
GSK3␤ phosphorylations. Moreover, in this work, CA-S6K1 activate AMPK and that their AMPK-dependent S6K1 inhi-
transfection abrogated the ability of oltipraz to restore bition plays a key role in abolishing TNF␣-induced insulin
GSK3␤ phosphorylation, verifying the functional significance resistance (Bae et al., 2007). Here, in this study, hyperos-
of S6K1 inhibition by oltipraz in improving insulin signals. motic conditions activated both mTOR-S6K1 and AMPK. The
Our data corroborating the selective S6K1 inhibition by olti- results of energy depletion and gene knockout experiments
praz among sorbitol-activated kinases in conjunction with showed that AMPK activation leads to suppression of the
the causal relationship between S6K1 inhibition and im- mTOR pathway via phosphorylation of tuberous sclerosis
proved insulin responses provide strong evidence that olti- complex 2, a negative regulator of mTOR-S6K1 (Krause et
praz’s restoration of insulin sensitivity against hyperosmotic al., 2002; Inoki et al., 2003; Brugarolas et al., 2004; Shaw et
stress results from S6K1 inhibition.
al., 2004). It seems likely that hyperosmotic stress activates
The principal glucose transporter protein that mediates S6K1 irrespectively of its AMPK activation. In the present
glucose uptake into skeletal muscle is GLUT4, which plays a study, we observed that TNF␣ decreased AMPK activity in
key role in regulating whole body glucose homeostasis. In HepG2 cells, which is consistent with the result showing that
response to insulin, Akt and other signaling molecules, such TNF␣ inhibited AMPK via transcriptional up-regulation of
as atypical protein kinase C ␭/␨, AMPK, calcium/calmodulin- PP2C (Steinberg et al., 2006), supporting the concepts that
dependent protein kinase II, and classic protein kinase C S6K1 activation by TNF␣ involves alteration in AMPK activ-
regulate GLUT4 trafficking (Farese et al., 2005; Jessen and ity and that the inhibitory effect of oltipraz on TNF␣-induced
Goodyear, 2005; Rose and Richter, 2005). The beneficial ef- S6K1 activation depends on AMPK activation (Bae et al.,
fects of oltipraz on the functional improvement in insulin 2007). Our results and others indicate that the pathway
actions are supported by the observations demonstrating responsible for mTOR-S6K1 activation by hyperosmolarity
that insulin-dependent glucose uptake inhibited by hyperos- may differ from that by TNF␣. Our observation also showed
motic stress in C2C12 myotubes was significantly restored by that sorbitol treatment markedly increased ACC phosphory-
oltipraz. These results are consistent with the finding that lation without altering cellular ATP content (data not
oltipraz also protected insulin-dependent Akt phosphoryla- shown). In general, hyperosmotic AMPK activation is asso-
tion. However, oltipraz failed to recover Akt phosphorylation ciated with a decrease in cellular generation of ATP, phos-
and glucose uptake in sorbitol-treated 3T3-L1 adipocytes. In phocreatine, and glycogen (Hayashi et al., 1998, 2000), sug-
separate experiments, we found that rapamycin, a potent gesting that the activation of AMPK by hyperosmolarity
inhibitor of mTOR-S6K1 pathway, prevented insulin resis- results from adaptive response to toxic stimuli. Therefore, it
tance in HepG2 or H4IIE cells (evidenced by increases in Akt could be concluded that the signaling pathway of AMPK
phosphorylation), but not in 3T3-L1 adipocytes even at the activation by oltipraz differs from that by hyperosmolarity.
concentration that perfectly inhibited S6 phosphorylations Our observations showing that the prevention of hyperos-
(data not shown). From the above results, it could be specu- motic insulin resistance results from the inhibition of S6K1
lated that pathway(s) alternative to the mTOR-S6K pathway indicate that 1,2-dithiole-3-thiones inhibit S6K1 activation
function(s) in adipocytes exposed to hyperosmolarity.
independently of the hyperosmotic stress-induced AMPK ac-
Furthermore, we demonstrated that a variety of synthetic tivation. This hypothesis is in fact supported by our addi-
1,2-dithiole-3-thione compounds strongly inhibited S6K1 ac- tional data showing that DN-AMPK transfection failed to
tivity, indicating that the active moiety capable of exerting reverse the ability of oltipraz to restore insulin sensitivity in
S6K1 inhibition exists in the 1,2-dithiole-3-thione moiety. cells treated with 600 mM sorbitol for 1 h (data not shown).
S6K1 inhibition by the compounds comprising 1,2-dithiole-3-
In conclusion, the results presented in this study demon-
thione as a pharmacophore contributed to the improvement strate that the compounds comprising 1,2-dithiole-3-thione
of insulin signaling, as evidenced by increases in GSK3␤ as a pharmacophore improved insulin sensitivity under the
phosphorylation against hyperosmotic stress. This concurs condition of hyperosmotic stress, which resulted from the
with our previous finding that the 1,2-dithiole-3-thiones also inhibition of S6K1 activation. Moreover, this study revealed
enhanced insulin responses from TNF␣-induced impairment that S6K1 downstream of mTOR was responsible for hyper-
(Bae et al., 2007). In the present study, the extent of S6K1 osmotic stress-induced insulin resistance. The current study
inhibition by the compounds did not completely match with brings additional insights into the therapeutic effects of 1,2-
that of either GSK3␤ phosphorylation or glucose uptake, dithiole-3-thione compounds on insulin resistance, which may
which might be due to the limit of sensitivity of the kinase be of assistance in developing drugs to treat insulin resistance.
assay or intricate regulatory mechanism underlying glucose
References
uptake into muscle cells. Our contention is supported by the
results showing that the synthetic dithiolethiones consis-
tently decreased S6 phosphorylation in either HepG2 (data
not shown) or H4IIE cells (Bae et al., 2007). Despite the
Aguirre V, Uchida T, Yenush L, Davis R, and White MF (2000) The c-Jun NH2-
terminal kinase promotes insulin resistance during association with insulin re-
ceptor substrate-1 and phosphorylation of Ser(307). J Biol Chem 275:9047–9054.
Aguirre V, Werner ED, Giraud J, Lee YH, Shoelson SE, and White MF (2002)
Phosphorylation of Ser307 in insulin receptor substrate-1 blocks interactions with
variability in their efficacies, some of the 1,2-dithiole-3-
the insulin receptor and inhibits insulin action. J Biol Chem 277:1531–1537.

1512
Bae et al.
Bae EJ and Kim SG (2005) Enhanced CCAAT/enhancer-binding protein beta-liver-
enriched inhibitory protein production by oltipraz, which accompanies CUG re-
peat-binding protein-1 (CUGBP1) RNA-binding protein activation, leads to inhi-
bition of preadipocyte differentiation. Mol Pharmacol 68:660–669.
Bae EJ, Yang YM, Kim JW, and Kim SG (2007) Identification of a novel class of
dithiolethiones that prevent hepatic insulin resistance via adenosine monophos-
phate activated protein kinase-p70S6 kinase 1 pathway. Hepatology 46:730–739.
Bratusch-Marrain PR, and DeFronzo RA (1983) Impairment of insulin-mediated
glucose metabolism by hyperosmolality in man. Diabetes 32:1028–1034.
Brugarolas J, Lei K, Hurley RL, Manning BD, Reiling JH, Hafen E, Witters LA,
Ellisen LW, and Kaelin WG Jr (2004) Regulation of mTOR function in response to
hypoxia by REDD1 and the TSC1/TSC2 tumor suppressor complex. Genes Dev
18:2893–2904.
Chen D, Fucini RV, Olson AL, Hemmings BA, and Pessin JE (1999) Osmotic shock
inhibits insulin signaling by maintaining Akt/protein kinase B in an inactive
dephosphorylated state. Mol Cell Biol 19:4684–4694.
Curphey TJ (2002) Thionation with the reagent combination of phosphorus penta-
sulfide and hexamethyldisiloxane. J Org Chem 67:6461–6473.
Eisner V, Quiroga C, Criollo A, Eltit JM, Chiong M, Parra V, Hidalgo K, Toro B, and
(1997) Rapamycin suppresses 5ЈTOP mRNA translation through inhibition of
p70s6k. EMBO J 16:3693–3704.
Jessen N and Goodyear LJ (2005) Contraction signaling to glucose transport in
skeletal muscle. J Appl Physiol 99:330–337.
Kang KW, Cho IJ, Lee CH, and Kim SG (2003) Essential role of phosphatidylinositol
3-kinase-dependent CCAAT/enhancer binding protein beta activation in the in-
duction of glutathione S-transferase by oltipraz. J Natl Cancer Inst 95:53–66.
Kaya M, Gulturk S, Elmas I, Kalayci R, Arican N, Kocyildiz ZC, Kucuk M, Yorulmaz
H, and Sivas A (2004) The effects of magnesium sulfate on blood-brain barrier
disruption caused by intracarotid injection of hyperosmolar mannitol in rats. Life
Sci 76:201–212.
Kayali AG, Austin DA, and Webster NJ (2000) Stimulation of MAPK cascades by
insulin and osmotic shock: lack of an involvement of p38 mitogen-activated protein
kinase in glucose transport in 3T3–L1 adipocytes. Diabetes 49:1783–1793.
Kitabchi AE, Umpierrez GE, Murphy MB, Barrett EJ, Kreisberg RA, Malone JI, and
Wall BM; American Diabetes Association (2004) Hyperglycemic crises in diabetes.
Diabetes Care 27 (Suppl 1):S94–S102.
Krause U, Bertrand L, and Hue L (2002) Control of p70 ribosomal protein S6 kinase
and acetyl-CoA carboxylase by AMP-activated protein kinase and protein phos-
phatases in isolated hepatocytes. Eur J Biochem 269:3751–3759.
Ku¨ltz D, Garcia-Perez A, Ferraris JD, and Burg MB (1997) Distinct regulation of
osmoprotective genes in yeast and mammals. Aldose reductase osmotic response
element is induced independent of p38 and stress-activated protein kinase/Jun
N-terminal kinase in rabbit kidney cells. J Biol Chem 272:13165–13170.
Lornejad-Scha¨fer MR, Schafer C, Graf D, Haussinger D, and Schliess F (2003)
Osmotic regulation of insulin-induced mitogen-activated protein kinase phospha-
tase (MKP-1) expression in H4IIE rat hepatoma cells. Biochem J 371:609–619.
Myers MG Jr and White MF (1996) Insulin signal transduction and the IRS proteins.
Annu Rev Pharmacol Toxicol 36:615–658.
Parrott LA and Templeton DJ (1999) Osmotic stress inhibits p70/85 S6 kinase
through activation of a protein phosphatase. J Biol Chem 274:24731–24736.
Paz K, Voliovitch H, Hadari YR, Roberts CT Jr., LeRoith D, and Zick Y (1996)
Interaction between the insulin receptor and its downstream effectors. Use of
individually expressed receptor domains for structure/function analysis. J Biol
Chem 271:6998–7003.
D´ıaz-Araya G, Lavandero
S (2006) Hyperosmotic stress activates p65/RelB
NFkappaB in cultured cardiomyocytes with dichotomic actions on caspase activa-
tion and cell death. FEBS Lett 580:3469–3476.
Engelman JA, Berg AH, Lewis RY, Lisanti MP, and Scherer PE (2000) Tumor
necrosis factor alpha-mediated insulin resistance, but not dedifferentiation, is
abrogated by MEK1/2 inhibitors in 3T3–L1 adipocytes. Mol Endocrinol 14:1557–
1569.
Ennis ED, Stahl E, and Kreisberg RA (1994) The hyperosmolar hyperglycemic
syndrome. Diabetes Rev 2:115–126.
Farese RV, Sajan MP, and Standaert ML (2005) Insulin-sensitive protein kinases
(atypical protein kinase C and protein kinase B/Akt): actions and defects in obesity
and type II diabetes. Exp Biol Med 230:593–605.
Fryer LG, Hajduch E, Rencurel F, Salt IP, Hundal HS, Hardie DG, and Carling D
(2000) Activation of glucose transport by AMP-activated protein kinase via stim-
ulation of nitric oxide synthase. Diabetes 49:1978–1985.
Fumarola C, La Monica S, and Guidotti GG (2005) Amino acid signaling through the
mammalian target of rapamycin (mTOR) pathway: role of glutamine and of cell
shrinkage. J Cell Physiol 204:155–165.
Galcheva-Gargova Z, De´rijard B, Wu IH, and Davis RJ (1994) An osmosensing signal
transduction pathway in mammalian cells. Science 265:806–808.
Galvez AS, Ulloa JA, Chiong M, Criollo A, Eisner V, Barros LF, and Lavandero S
(2003) Aldose reductase induced by hyperosmotic stress mediates cardiomyocyte
apoptosis: differential effects of sorbitol and mannitol. J Biol Chem 278:38484–
38494.
Gao Z, Hwang D, Bataille F, Lefevre M, York D, Quon MJ, and Ye J (2002) Serine
phosphorylation of insulin receptor substrate 1 by inhibitor ␬B kinase complex.
J Biol Chem 277:48115–48121.
Pearson RB, Dennis PB, Han JW, Williamson NA, Kozma SC, Wettenhall RE, and
Thomas G (1995) The principal target of rapamycin-induced p70s6k inactivation is
a novel phosphorylation site within a conserved hydrophobic domain. EMBO J
14:5279–5287.
Rose AJ and Richter EA (2005) Skeletal muscle glucose uptake during exercise: how
is it regulated? Physiology 20:260–270.
Saltiel AR and Kahn CR (2001) Insulin signaling and the regulation of glucose and
lipid metabolism. Nature 414:799–806.
Shaw RJ, Bardeesy N, Manning BD, Lopez L, Kosmatka M, DePinho RA, and
Cantley LC (2004) The LKB1 tumor suppressor negatively regulates mTOR sig-
naling. Cancer Cell 6:91–99.
Steinberg GR, Carling D, Hotamisligil GS, Febbraio MA, Kay TW, and Kemp BE
(2006) Tumor necrosis factor alpha-induced skeletal muscle insulin resistance
involves suppression of AMP-kinase signaling. Cell Metab 4:465–474.
Sun XJ, Goldberg JL, Qiao LY, and Mitchell JJ (1999) Insulin-induced insulin
receptor substrate-1 degradation is mediated by the proteasome degradation path-
way. Diabetes 48:1359–1364.
Tremblay F, Krebs M, Dombrowski L, Brehm A, Bernroider E, Roth E, Nowotny P,
Waldhausl W, Marette A, and Roden M (2005) Overactivation of S6 kinase 1 as a
cause of human insulin resistance during increased amino acid availability. Dia-
betes 54:2674–2684.
Um SH, Frigerio F, Watanabe M, Picard F, Joaquin M, Sticker M, Fumagalli S,
Allegrini PR, Kozma SC, Auwerx J, et al. (2004) Absence of S6K1 protects against
age- and diet-induced obesity while enhancing insulin sensitivity. Nature 431:
200–205.
Yamauchi T, Tobe K, Tamemoto H, Ueki K, Kaburagi Y, Yamamoto-Honda R,
Takahashi Y, Yoshizawa F, Aizawa S, Akanuma Y, et al. (1996) Insulin signalling
and insulin actions in the muscles and livers of insulin-resistant, insulin receptor
substrate 1-deficient mice. Mol Cell Biol 16:3074–3084.
Gual P, Gonzalez T, Gremeaux T, Barres R, Le Marchand-Brustel Y, and Tanti JF
(2003a) Hyperosmotic stress inhibits insulin receptor substrate-1 function by
distinct mechanisms in 3T3–L1 adipocytes. J Biol Chem 278:26550–26557.
Gual P, Le Marchand-Brustel Y, and Tanti JF (2003b) Positive and negative regu-
lation of glucose uptake by hyperosmotic stress. Diabetes Metab 29:566–575.
Han J, Lee JD, Bibbs L, and Ulevitch RJ (1994) A MAP kinase targeted by endotoxin
and hyperosmolarity in mammalian cells. Science 265:808–811.
Harrington LS, Findlay GM, Gray A, Tolkacheva T, Wigfield S, Rebholz H, Barnett
J, Leslie NR, Cheng S, Shepherd PR, et al. (2004) The TSC1–2 tumor suppressor
controls insulin-PI3K signaling via regulation of IRS proteins. J Cell Biol 166:
213–223.
Hayashi T, Hirshman MF, Fugii N, Habinowski SA, Witters LA, and Goodyear LJ
(2000) Metabolic stress and altered glucose transport. Activation of AMP-activated
protein kinase as a unifying coupling mechanism. Diabetes 49:527–531.
Hayashi T, Hirshman MF, Kurth EJ, Winder WW, and Goodyear LJ (1998) Evidence
for 5Ј AMP-activated protein kinase mediation of the effect of muscle contraction
on glucose transport. Diabetes 47:1369–1373.
Hotamisligil GS, Peraldi P, Budavari A, Ellis R, White MF, and Spiegelman BM
(1996) IRS-1-mediated inhibition of insulin receptor tyrosine kinase activity in
TNF-alpha- and obesity-induced insulin resistance. Science 271:665–668.
Inoki K, Zhu T, and Guan KL (2003) TSC2 mediates cellular energy response to
control cell growth and survival. Cell 115:577–590.
Address correspondence to: Sang Geon Kim, Ph.D., College of Pharmacy,
Seoul National University, Sillim-dong, Kwanak-gu, Seoul 151-742, Korea.
E-mail: sgk@snu.ac.kr
Jefferies HB, Fumagalli S, Dennis PB, Reinhard C, Pearson RB, and Thomas G