Insulin can block apoptosis by decreasing oxidative stress via phosphatidylinositol 3-kinase- and extracellular signal-regulated protein kinase-dependent signaling pathways in HepG2 cells.

Abstract

Insulin has well-known activities in controlling energy metabolism, cellular proliferation and biosynthesis of functional molecules to maintain a biological homeostasis. Recently, several studies have suggested that insulin may protect cells from apoptosis in different cell lines; however, little is known about the nature of its anti-apoptotic activity. In many clinical disorders, including type 2 diabetes mellitus, oxidative stress and the production of reactive oxygen species (ROS) is increased. With these facts as a background, we examined here whether insulin protects HepG2 cells from apoptosis by decreasing oxidative stress and, if so, which signaling steps are involved in this process.
Intracellular DNA content, the degree of nuclear condensation or poly(ADP-ribose) polymerase hydrolysis was measured to verify the occurrence of apoptotic events. Caspase-3 activity and ROS accumulation within cells were also measured. Western blot analysis was performed to identify signaling molecules activated in response to insulin.
Serum starvation resulted in a marked accumulation of ROS, activation of caspase-3, and subsequent apoptotic cell death which were, in turn, markedly blocked by the addition of insulin. The anti-apoptotic activity of insulin was sensitive to blockade of two different signaling steps, activations of phosphatidylinositol 3-kinase (PI3 kinase) and extracellular signal-regulated protein kinase (ERK).
Insulin exerts an anti-apoptotic activity by suppressing the excessive accumulation of ROS within cells through signaling pathways including stimulation of PI3 kinase and ERK in HepG2 cells.

Full text

Page 1

EXPERIMENTAL STUDY
Insulin can block apoptosis by decreasing oxidative stress
via phosphatidylinositol 3-kinase- and extracellular
signal-regulated protein kinase-dependent signaling
pathways in HepG2 cells
Shinhae Kang1, Jihoon Song1, Heekyoung Kang, Sejae Kim1, Youngki Lee and Deokbae Park
Department of Medicine, College of Medicine, Institute of Basic Science and1Department of Biology, College of Natural Science, Cheju National University,
Ara-1, Cheju, 690-756, South Korea
(Correspondence should be addressed to D Park, Department of Medicine, College of Medicine, Cheju National University, Ara-1, Cheju, 690-756,
South Korea; Email: parkdb@cheju.ac.kr)
Abstract
Objective: Insulin has well-known activities in controlling energy metabolism, cellular proliferation
and biosynthesis of functional molecules to maintain a biological homeostasis. Recently, several
studies have suggested that insulin may protect cells from apoptosis in different cell lines; however,
little is known about the nature of its anti-apoptotic activity. In many clinical disorders, including
type 2 diabetes mellitus, oxidative stress and the production of reactive oxygen species (ROS) is
increased. With these facts as a background, we examined here whether insulin protects HepG2
cells from apoptosis by decreasing oxidative stress and, if so, which signaling steps are involved in
this process.
Methods: Intracellular DNA content, the degree of nuclear condensation or poly(ADP-ribose) polymer-
ase hydrolysis was measured to verify the occurrence of apoptotic events. Caspase-3 activity and ROS
accumulation within cells were also measured. Western blot analysis was performed to identify sig-
naling molecules activated in response to insulin.
Results: Serum starvation resulted in a marked accumulation of ROS, activation of caspase-3, and
subsequent apoptotic cell death which were, in turn, markedly blocked by the addition of insulin.
The anti-apoptotic activity of insulin was sensitive to blockade of two different signaling steps, acti-
vations of phosphatidylinositol 3-kinase (PI3 kinase) and extracellular signal-regulated protein
kinase (ERK).
Conclusion: Insulin exerts an anti-apoptotic activity by suppressing the excessive accumulation of ROS
within cells through signaling pathways including stimulation of PI3 kinase and ERK in HepG2 cells.
European Journal of Endocrinology 148 147–155
Introduction
Insulin has a variety of functions such as glucose
uptake, glycogen biosynthesis, inhibition of lipolysis
and stimulation of cellular proliferation. It can also pro-
tect cells from apoptosis mediated by growth factor
removal in different cell lines (1–3). We have recently
investigated the anti-apoptotic function of insulin and
its related signaling pathways in Chinese hamster
ovary cells expressing wild-type human insulin recep-
tors (CHO-IR) (4, 5). Although a number of molecules
involved in the prevention or induction of apoptosis
have been identified, the mechanisms by which insulin
participates in this process remain largely unknown.
Reactive oxygen species (ROS) have been implicated
as important pathologic mediators in many clinical
disorders (6), including type 2 diabetes mellitus (7, 8).
In addition, oxidative stress disrupts insulin-induced
signaling events such as insulin receptor tyrosine
phosphorylation, insulin receptor substrate-1 (IRS-1)
phosphorylation and activation of phosphatidylinositol
3-kinase (PI3 kinase) (9) as well as redistribution of
IRS-1 and PI3 kinase in adipocytes (10). However, con-
flicting results on the effect of H2O2on insulin signaling
are also available. H2O2mimics the stimulatory effects
of insulin on glucose transport and lipid synthesis in
adipocytes (11, 12) and insulin transiently generates
H2O2via NADPH oxidase, which is integral to activ-
ation of the insulin signaling cascades in adipocytes
(13). Insulin-induced generation of H2O2 reversibly
inhibits protein-tyrosine
enhancing the early insulin action cascades (14). So
phosphatase 1B, thereby
European Journal of Endocrinology (2003) 148 147–155 ISSN 0804-4643
q 2003 Society of the European Journal of Endocrinology Online version via http://www.eje.org

Page 2

far there has been no general agreement on which oxi-
dative stress affects insulin signaling and its physiologi-
cal functions, or on which insulin modulates ROS
generation.
Following insulin binding, the insulin receptors
undergo activation of their intrinsic tyrosine kinase
function and subsequent stimulation of signaling mol-
ecules (15). We have reported that insulin delayed
apoptosis induced by serum starvation (4), by a mech-
anism that is dependent on short-lived farnesylated
proteins in CHO-IR cells (5). However, there is little evi-
dence as to whether insulin really acts as a survival
factor in physiological cell systems expressing intrinsic
insulin receptors. With these facts as a background,
we examined whether insulin exerts an anti-apoptotic
activity and, if so, which signaling steps are important
for completing this process in HepG2 cells expressing
intrinsic insulin receptor molecules.
Materials and methods
Materials
Human recombinant insulin, PD98059, SB202190,
the fluorogenic caspase-3 substrate, Ac-DEVD-AMC,
the caspase inhibitor, z-DEVD-fmk and propidium
iodide, 20,70-dichlorofluorescin diacetate (H2DCFDA)
were obtained from Calbiochem (La Jolla, CA, USA).
Wortmannin, Dulbecco’s modified Eagle’s medium
(DMEM),Dulbecco’s-phosphate-buffered
PBS) and trypsin–EDTA solution were obtained from
Sigma Chemical Corp. (St Louis, MO, USA), and fetal
bovine serum (FBS) from Life Technologies Inc. (Rock-
ville, MD, USA). Monoclonal antibodies against phos-
pho-ERK1/2 (E-4), phospho–C-Jun
kinase (phospho-JNK) (G-7), phospho-p38Map kinase
(D-8), ERK2 (D-2) and polyclonal antibodies against
phospho-Akt1 (ser473) and poly(ADP-ribose) polymer-
ase (PARP) (H-250) were from Santa Cruz Biotechnol-
ogy (Santa Cruz, CA, USA). Electrophoresis reagents,
such as gels, Tris–glycine SDS running buffer, and
poly(vinylidene difluoride) (PVDF) membrane were
from Novex Corp. (San Diego, CA, USA).
saline (D-
NH2
terminal
Cell culture
HepG2, a hepatoblastonema cell line used in this study,
was obtained from Korean Cell Line Bank (Seoul,
Korea) and grown in DMEM containing 100 units/ml
penicillin, 100mg/ml streptomycin and 10% FBS, and
maintained in a humidified atmosphere of 5% CO2in
air at 378C. Two days after plating in 35mm tissue cul-
ture dishes, HepG2 cells were serum starved for 24h
and then treated with different reagents. Cells were
quickly frozen in liquid nitrogen and stored at 2708C
until analysis.
SDS-PAGE and immunoblotting
Unless otherwise indicated, cells were lysed in ice-cold
lysis buffer (50mM Tris–HCl, 1% nonidet P-40,
0.25% sodium deoxycholate, 150mM NaCl, 1mM
sodium orthovanadate, 1mM NaF, 1mM phenyl-
methylsulfonyl fluoride, 1mM aprotinin, 1mM leupep-
tin and 1mM pepstatin A). The same amount of
proteins was separated by SDS-PAGE on 4–20% poly-
acrylamide gels and electrotransferred onto PVDF
membrane. The membrane was incubated in blocking
buffer (5% non-fat dry milk in Tris–buffered saline–
0.1% Tween-20) for 1h at room temperature and
thenprobedwithdifferent
(1:1000–1:5000). After a series of washes, the mem-
brane was further incubated with different horseradish
peroxidase-conjugated secondary antibodies (1:2000–
1:10000). The signal was detected with an enhanced
chemiluminescence detection system (Intron, Seong-
nam, Korea).
primary antibodies
Determination of caspase-3 activity
After treatment with reagents, cells were collected,
lysed inice-cold0.5ml
(50mmol/l Tris–HCl, pH 7.4, 150mM NaCl, 1% non-
idet P-40, 0.25% sodium deoxycholate and 1mmol/l
EGTA) for 15min. After centrifugation at 12000g for
15min at 48C, aliquots of supernatant were incubated
with 10mmol/l Ac-DEVD-AMC for 3h at 378C. The flu-
orescence from the cleaved product was detected with
a Spectrafluor multiwell fluorescence reader (Tecan,
Gro ¨dig, Austria) at 360nm and 465nm wavelengths
for excitation and emission respectively.
caspaseassay buffer
Detection of apoptotic cells with flow
cytometric analysis and H33342 staining
The degree of apoptosis was determined by measuring
the number of cells showing below the G1 DNA content
from flow cytometric analysis after staining of cells with
propidium iodide as originally described by Crissman &
Steinkamp (16). The samples were analyzed with a
Coulter EpicsTM cytometer (Beckman, Fullerton, CA,
USA). Ten thousand events were collected for each
sample. An excitation wavelength of 488nm and a
fluorescence emission of 580nm were used. Otherwise,
cells were stained with a DNA-specific fluorescent dye
(H33342) then observed under a fluorescent micro-
scope equipped with a CoolSNAP-Pro color digital
camera (Media Cybernetics, Meyer, Housten, TX,
USA) to examine the degree of nuclear condensation.
Measurement of ROS
H2DCFDA, a cell-permeable fluorogenic probe that is
useful for the detection of ROS, was used to measure
the degree of ROS accumulation within cells. Cultured
148
S Kang and others
EUROPEAN JOURNAL OF ENDOCRINOLOGY (2003) 148
www.eje.org

Page 3

cells were briefly washed once with D-PBS and further
incubated in D-PBS containing 10mmol/l H2DCFDA
for 10min at 378C. The fluorescence intensity was
measured with a Spectrafluor multiwell fluorescence
reader at 485nm and 535nm wavelengths for exci-
tation and emission respectively, under constant con-
ditions to allow quantitative comparisons of relative
fluorescence intensity from cells with different treat-
ments. Fluorescent cell images were instantly captured
under a fluorescent inverted microscope equipped with
a CoolSNAP-Pro color digital camera.
Statistical analysis
Statistical analysis was performed using an analysis
program, StatView (Abacus Concepts, Berkeley, CA,
USA). The Student’s t-test was used to analyze the
difference between control and experimental groups.
P , 0:05 was considered to be significant.
Results
Insulin delays apoptosis induced by serum
starvation or oxidative stress
Our previous studies (4, 5) have shown that insulin
delayed apoptosis induced by serum starvation in
CHO-IR cells. However, it is not clear whether insulin
can really play a role in inhibiting apoptosis in certain
cellshavingintrinsicinsulinreceptorssuchasadipocytes,
muscle cells or liver cells. HepG2 is a hepatoblastonema
cell line that expresses intrinsic insulin receptors.
HepG2 cells were preincubated in serum-free medium
for 24h and then washed with D-PBS twice, then trea-
ted with 100nmol/l insulin for an additional 48h. The
degree of apoptosis was determined with flow cyto-
metric analysis measuring DNA content of each cell
and counting the number of events below diploid
(,2N) DNA content (Fig. 1A). The percentages of
apoptotic cells were 23:2^3:5% (control, serum-free)
and 4:1^0:7% (100nmol/l insulin) respectively. The
Figure 1 Insulin blocks apoptosis induced by
serum starvation or H2O2addition. (A) Cells
were serum starved (24h), preincubated with
10mmol/l z-DEVD-fmk (30min), and further
incubated with 100nmol/l insulin and 1mmol/l
H2O2for an additional 48h. The degree of
apoptosis is represented as the DNA content
measured by flow cytometric analysis and the
photographs show cells with highly con-
densed nuclei stained with H33342. (B) Pro-
teolytic cleavage of PARP induced by H2O2
and (C) its suppression by insulin after treat-
ment of cells for 48h. Equal amounts of cell
lysates were subjected to electrophoresis and
analyzed by Western blot for PARP, and
ERK as an internal standard.
Signaling pathways underline the anti-apoptotic function of insulin
149
EUROPEAN JOURNAL OF ENDOCRINOLOGY (2003) 148
www.eje.org

Page 4

effect of H2O2(1mmol/l) was also examined in order to
determine whether an additional oxidative stress accel-
erates an apoptotic process induced by serum star-
vation. The addition of H2O2 significantly increased
the percentage of apoptotic cells ð35:3^4:2%Þ com-
pared with control groups ðP , 0:05Þ: Insulin treat-
ment also significantly
P , 0:01Þ the degree of apoptosis compared with
H2O2 alone. To further confirm this finding, cells
were stained with a DNA-specific fluorescent dye,
H33342, to observe the degree of nuclear conden-
sation, which is an apoptotic phenomenon (Fig. 1A).
decreased
ð8:9^2:6%;
A number of cells incubated in serum-free medium or
treated with H2O2for 48h showed highly condensed
nuclei. Insulintreatment
number of cell with condensed nuclei in both groups.
We also investigated whether the activation of cas-
pase-3 is associated with the apoptotic process induced
by serum starvation or oxidative stress. z-DEVD-fmk, a
cell-permeable caspase-3 inhibitor, effectively protected
cells from apoptosis by serum starvation or H2O2treat-
ment from the results of flow cytometric analysis and
H33342 staining. The degree of proteolytic cleavage
of PARP, a target molecule of caspases, was parallel
clearlydecreasedthe
Figure 2 Insulin suppresses caspase-3
stimulation and ROS production induced by
serum starvation. (A) HepG2 cells were
serum starved for 24h and preincubated with
inhibitors for 30min before insulin addition as
described in Results. Caspase-3 activity was
measured at 24h after insulin treatment and
is represented as relative units. Results are
the means^S.E.M. of three separate experi-
ments. *P , 0:05 vs serum free;#P , 0:05
vs insulin; *ND, non-detected. (B) Dichloro-
fluorescein (DCF) fluorescence from cells
imaged with a CoolSNAP-Pro digital camera
attached to the inverted fluorescent micro-
scope. Cells were serum starved for 24h and
further treated with insulin and/or H2O2for an
additional 3h.
150
S Kang and others
EUROPEAN JOURNAL OF ENDOCRINOLOGY (2003) 148
www.eje.org

Page 5

with the dose of H2O2(Fig. 1B). The integrity of PARP
was increased by insulin treatment compared with
serum starvation. H2O2-induced PARP cleavage was
also effectively blocked by insulin (Fig. 1C). These
results suggested that insulin can protect HepG2 cells
from apoptosis induced by serum starvation or oxi-
dative stress.
Insulin suppresses caspase-3 activity and ROS
production
The caspase family of cysteine proteases plays a pivotal
role in mediating apoptosis through the proteolysis of
specific targets that include PARP, the nuclear lamins
and caspase-dependent DNase (17, 18). The function
Figure 3 The anti-apoptotic activity of insulin
is sensitive to ERK inhibition and PI3 kinase
inhibition. Cells were serum starved (24h),
preincubated with inhibitors for 30min and
further incubated with insulin for an additional
48h. The degree of apoptosis is represented
as the DNA content measured by flow cyto-
metric analysis and the photographs show
cells with highly condensed nuclei stained
with H33342.
Signaling pathways underline the anti-apoptotic function of insulin
151
EUROPEAN JOURNAL OF ENDOCRINOLOGY (2003) 148
www.eje.org