(Received for publication, June 20, 1996, and in revised form, September 27, 1996)
From the Departments of Physiology & Biophysics,
¶ Medicine, and
Molecular Pharmacology, Diabetes & Metabolic Diseases Research Program, University Medical Center,
SUNY/Stony Brook, Stony Brook, New York 11794
Differentiation of 3T3-L1 embryonic fibroblasts to adipocytes in response to induction by dexamethasone and isobutylmethylxanthine is blocked by inhibitors of Ca2+-calmodulin-sensitive protein kinase type II, but not by inhibitors of protein kinase A or protein kinase C. CaM kinase II displays a biphasic increase in autonomous activity, rising after an initial transient peak from 1 to 15 h, declining at 24 h, followed by a sustained rise from 24 to 48 h, which is 2.5-fold greater than basal values at induction of adipogenesis. Adipogenesis was blocked effectively by CaM kinase II inhibitors, either KN-62 or KN-93, if the inhibitors are introduced at 6 h and maintained until 12 h of induction of adipogenesis. Equally effective, however, is inhibition of CaM kinase II activity at 24-48 h after induction, during the later phase of autonomous CaM kinase activity. Inhibition of cultures with KN-62 or KN-93 either for 0 to 6 h or for 12 to 24 h failed to influence adipogenesis. Two temporally-distinct phases of CaM kinase II activation, either 6 to 12 h or 24 to 48 h, if inhibited with either KN-62 or KN-93, blocked the conversion to adipocytes. Thus, a biphasic activation of CaM kinase II is obligate for the progression of the embryonic fibroblasts to adipocytes. Inhibition of either phase of CaM kinase activity blocks adipogenesis and expression of several intermediate early gene products.
3T3-L1 embryonic fibroblasts provide a useful model for the study
of cellular differentiation. In response to several inducers (e.g. high concentrations of insulin or dexamethasone + isobutylmethylxanthine (dexamethasone+IBMX),1 in combination),
cultures of 3T3-L1 fibroblasts progress to adipocytes, accumulating
lipid and displaying activation of the number of genes characteristic
of this phenotype (1-4). Recently, a number of key elements to the
expression of the adipogenic phenotype have been reported, including
changes in heterotrimeric G-proteins (5-7), Ras, and Raf-1 activity
(8, 9), mitogen-activated protein kinase (10), CCAAT/enhancer-binding
protein (11-13), protein-tyrosine phosphatase HA2 (14), and peroxisome
proliferator-activation receptor (15). In view of the prominent
role of protein kinases in complex signal pathways, we investigated the
roles of protein kinase A, protein kinase C, and CaM kinase II in the
adipogenic conversion of 3T3-L1 culture in vitro, using
protein kinase inhibitors. The results suggest little role for either
protein kinase A or protein kinase C, but reveal an obligatory role for
CaM kinase II in adipogenic conversion.
KT5720, bisindolylmaleimide, KN-62, A23187, and
ionomycin were purchased from Calbiochem (San Diego, CA). KN-93 was
purchased from Seikagaku America (Jamesville, MD). The CaM kinase
substrate peptide autocamtide-2 (sequence
Lys-Lys-Ala-Leu-Arg-Arg-Gln-Glu-Thr-Val-Asp-Ala-Leu) was synthesized
and purified by the Center for Analysis & Synthesis for Macromolecules
(SUNY/Stony Brook, NY). Fura-2 acetoxymethyl ester was obtained from
Molecular Probes (Eugene, OR). [-32P]ATP was purchased
from DuPont NEN.
The 3T3-L1 cultures were obtained from ATCC, cultured in Dulbecco's modified Eagle's medium plus 10% fetal bovine serum, induced to adipogenesis, and propagated as described elsewhere (5, 6).
Northern Blot AnalysisFor Northern blot RNA hybridization,
cells were harvested in phosphate-buffered saline containing EDTA (1 mM) and collected by low speed centrifugation. Total RNA
was extracted using the RNA STAT-60 reagent (Tel-Test, Friendswood, TX)
following the directions of the commercial supplier. Aliquots (30 µg/lane) of RNA were subjected to separation on 1.5%
agarose-formaldehyde gels and then transferred to nylon membrane in the
presence of 20 × SSC. Membranes were hybridized for 20 h at
42 °C with the cDNA for aP2, FACS, LPL, or
glyceraldehyde-phosphate dehydrogenase, which were gifts from Dr. Nada
Abumrad (Department of Physiology, SUNY, Stony Brook, NY) and labeled
by random priming with [-32P]dCTP. The blots were
hybridized in the presence of 50% formamide and then washed with
1 × SSC three times at 37 °C prior to exposure of film for
autoradiography.
Cells were collected at specific
time points, cooled on ice, and disrupted immediately by sonication at
4 °C in homogenization buffer at 107 cells/ml. The
homogenization buffer consisted of 50 mM Hepes, pH 7.5, 10% glycerol, 1 mM EGTA, 1 mM sodium vanadate,
100 nM okadaic acid, 20 mM benzamidine, 200 µM phenylmethylsulfonyl fluoride, 5 µg/ml aprotinin,
and 5 µg/ml leupeptin. Each homogenate was assayed immediately for
CaM kinase II activity using autocamtide-2 (16). Reactions were carried
out in 25 µl of reaction buffer, containing 50 mM Hepes,
pH 7.5, 10 mM magnesium acetate, 0.14 µM
bisindolylmaleimide, 1 mM sodium vanadate, 100 nM okadaic acid, 25 µM ATP, 0.05 µCi of
[-32P]ATP, and a homogenate from 5 × 104 cells at 30 °C for 2 min. Twenty µl of reaction
mixture was applied onto a piece of phosphocellulose paper (P-81) and
the paper washed with 10 mM phosphoric acid for 15 min and
air-dried. The incorporation of 32P into autocamtide-2 was
determined by liquid scintillation. Phosphorylation of autocamtide-2 in
the absence of Ca2+ relative to its phosphorylation in the
presence of Ca2+ is defined as the "autonomous"
activity (17).
3T3-L1 cells were grown on 25-mm coverslips to confluence and intracellular Ca2+ was measured as described (18). Briefly, coverslip-attached cells were loaded with 2 µM fura-2 acetoxymethyl ester for 45 min at 37 °C in the dark. After washing in fresh Krebs-Henseleit buffer for 15 min at 20 °C, coverslips were inserted into a Sykes-Moore chamber (Bello, Vineland, NJ). Groups of 5-10 fura-2-loaded cells were viewed through a Nikon 40x UV-fluor objective of a Nikon inverted microscope coupled to a Deltascan microspectrofluorometer system (PTI, Princeton, NJ). Excitation at 345 and 380 nm were generated using a 75-watt xenon lamp and a chopper. Emitted light was channeled into the photomultiplier via an ultraviolet dichroic mirror and interference filter of 510 nm (Corion, Holiston, MA). Measurements of fluorescence intensity were performed at a rate of 20 points/s, and the ratio of absorbance 345/380 nm was computed using the PTI software.
The ability of several well known selective inhibitors of protein
kinase activities (19-22) were investigated for their influence on
3T3-L1 cell growth and differentiation in response to the inducers dexamethasone+IBMX (Fig. 1).
Dexamethasone+ IBMX-induced differentiation is >90% of 3T3-L1
embryonic fibroblasts within 6-7 days, as evidenced by staining of
lipid accumulation by oil-red O (23). At concentrations an order of
magnitude greater than their Ki, inhibitors for
protein kinases were added to the cell culture media for 48 h and
the cultures examined for cell growth and differentiation in response
to dexamethasone+IBMX (Fig. 1A). The response was quantified by scoring of the percentage of cells progressing to differentiation (Fig. 1B). The bisindolylmaleimide inhibitor
of protein kinase C (20) failed to alter either cell growth or the
ability of the cells to progress to adipocytes in response to
dexamethasone+IBMX. More than 90% of the cells displayed frank adipogenic conversion in the presence of the protein kinase C inhibitor. Similarly, treating cultures of 3T3-L1 cells with phorbol 13-myristate 12-acetate (10 µM, from 2 to +24 h of
induction in response to dexamethasone+IBMX) in order to down-regulate
protein kinase C did not alter adipogenesis (data not shown).
Short-term activation of protein kinase C with phorbol 13-myristate
12-acetate (250 nM) for 10 min prior to the induction with
dexamethasone+IBMX also was without effect on either the rate or the
extent of adipogenic conversion (data not shown).
CaM kinase II inhibitors (KN-62 and KN-93),
but neither protein kinase C inhibitor (bisindolylmaleimide) nor
protein kinase A inhibitor (KT5720) inhibit dexamethasone+IBMX-induced
3T3-L1 cell differentiation. 3T3-L1 cells grown to confluence on coverslips in six-well plates (day 0) were either untreated
(DM) or induced to differentiation with dexamethasone+IBMX
(+DM). Bisindolylmaleimide (1.4 µM), KT5720 (1 µM), KN-62 or KN-93 (10 µM) were added to the differentiation media for the first 48 h, as indicated.
Panel A, at day 8, cells were fixed, stained with oil-red O,
and photographed using a Zeiss Axiophot system. Panel B,
the number of adipocytes relative to that of total cells on the
coverslip represents the ratio of differentiation. The data are mean
values ± S.E. from at least three separate experiments.
Treatment of cells with KT5720, a selective inhibitor of protein kinase
A (19), did not block adipogenesis, more than 75% of the cells
progressing to adipocytes within 8 days (Fig. 1). By 10 days nearly
100% of the KT5720-treated cells had undergone adipogenesis (not
shown). These data are consistent with earlier data demonstrating the
adipogenic conversion was unaffected by agents like forskolin or
pertussis toxin treatment that elevate intracellular cAMP levels (5).
In addition, suppressing intracellular cAMP levels with
2,5
-dideoxyadenosine (10 µM) was also without effect on
the ability of the cultures to differentiate to adipocytes in response
to dexamethasone+ IBMX (not shown). Thus inhibition of either protein
kinase C or protein kinase A activities fails to influence adipogenic
conversion in 3T3-L1 cells.
In contrast to the lack of effect of inhibitors for either protein
kinase A or C on adipogenic conversion in response to
dexamethasone+IBMX, treating the cells with KN-62, an inhibitor of CaM
kinase II, effectively blocks adipogenesis (Fig. 1, A and
B). KN-62 is highly selective for CaM kinase II, displaying
a Ki for CaM kinase II more than 2 orders of
magnitude lower than those for protein kinase C, protein kinase A, and
myosin light chain kinase (21). KN-93, a second inhibitor of CaM kinase
II with higher water solubility and equal selectivity for CaM kinase II
as compared to protein kinase C, protein kinase A, and myosin light
chain kinase (22), was evaluated also for its ability to effect
adipogenesis. At 10 µM concentrations, both KN-62 and
KN-93 effectively abolish the adipogenic response of 3T3-L1 cells to
induction by dexamethasone+ IBMX. The dose-response relationship
between adipogenic conversion of 3T3-L1 cultures and the inhibitors of
CaM kinase II was explored (Fig. 2). KN-62 and KN-93
both display the ability to block fully the adipogenic response, the
more water-soluble KN-93 compound being less potent in the suppression
of adipogenesis.
The ability of CaM kinase II inhibitors to block the induction of
adipogenesis in 3T3-L1 cells suggested some critical role of CaM kinase
II activity in the differentiation process. Activation of CaM kinase II
activity was investigated in the first 2 h of induction by
dexamethasone+IBMX (Fig. 3A). The autonomous
activity increased by 120 min, more than doubling. Analysis of the
first minute of induction of adipogenesis by dexamethasone+IBMX
revealed a sharp, transient peak of autonomous CaM kinase II activity
(not shown). By 10 s of induction, CaM kinase II activity
increased 5-fold (3.5% as compared to 0.75% at t = 0), the activity declining to control levels with 30-60 s (not shown).
Measured from 2 h to 2 days post-induction with
dexamethasone+IBMX, autonomous CaM kinase II activity increased 4-8
fold (Fig. 3B). Autonomous kinase activity increased from
0.75 to 4.0% at 2 to 18 h (first phase) and then displayed a
sharp decline at 24 h. Following the decline, autonomous CaM
kinase II activity rebounded to 2.5% (second phase), where it remained
up until 48 h post-induction. Thus, after a brief transient
activation, CaM kinase II activity increases from 1 to 18 h of
adipogenesis. This first phase declines by 24 h, followed by a
second phase of autonomous CaM kinase II activity which remains
4-5-fold greater than that in control cells treated only with vehicle.
Treating confluent cultures of 3T3-L1 cells for either 2 or 12 h
with ionophore A23187 (1 µM) or ionomycin (3 µM) in either Dulbecco's modified Eagle's medium or
Dulbecco's modified Eagle's medium supplemented with Ca2+
to 25 mM fails to induce adipogenic conversion of the
embryonic fibroblasts (data not shown). Elevating the concentration of
either ionophore by 10-fold results in cell death (data not shown).
Even in the presence of inducers, CaM kinase II inhibitors do not alter the rate of adipogenesis (data not shown). These data suggest that
activation of CaM kinase activity itself is not sufficient a stimulus
to induce the adipogenic progression.
The transient activation of CaM kinase in 3T3-L1 clones induced by
addition of dexamethasone+IBMX was examined from a second perspective,
measurements of intracellular free Ca2+. In cells
pre-loaded with the calcium-sensitive probe fura-2, measurements of
intracellular free Ca2+ were possible (Fig.
4A). As a control, the ability of the
ionophore ionomycin to stimulate an increase in intracellular free
Ca2+ was demonstrated. Ionomycin produced a rapid and
robust increase in intracellular free Ca2+. Similar
measurements performed in cells stimulated with dexamethasone+IBMX during the first 10 min of induction reveal no significant changes in
intracellular free Ca2+ (Fig. 4B). Measurements
at later time points likewise failed to identify a significant
change in the concentration of intracellular free Ca2+
(data not shown). Thus the temporal nature of a Ca2+
transient, if stimulated by inducers of adipogenesis, may have precluded detection by this approach.
The ability of inhibitors of CaM kinase II to block adipogenesis and
the time course of autonomous CaM kinase II activation during
dexamethasone+IBMX-induced adipogenesis prompted us to investigate the
time constraints on the exposure to the inhibitor with respect to its
ability to block differentiation. As shown by oil-red O staining (Fig.
5A) and quantification of the extent of
adipogenic conversion (Fig. 5B), blockade of adipogenesis by CaM kinase II inhibitor was not effective, if restricted to the first
6 h of the induction protocol. Whereas treatment with a CaM kinase
II inhibitor from 0 to 12 h blocks adipogenesis, treatments from 0 to 6 h, 6 to 24 h, or 12 to 24 h fail to block
adipogenesis. These data suggest that one period of CaM kinase II
activation from 6 to 12 h of induction (Fig. 3A) is
critical for adipogenic conversion to occur, a hypothesis supported by
the ability of a 6-12-h treatment of KN-93 to be fully effective in
blocking adipogenesis. Interestingly, inhibition of CaM kinase II
activity at 6-12 h was not the only window of opportunity for blockade of adipogenesis. Treating cells with KN-93 (or KN-62, not shown) from
0, 6, 12, or 24 h to the end of the 48-h induction period is
equally effective in blocking adipogenesis as was the brief 6-h
exposure to inhibitor from 6 to 12 h. The identification of two
distinct phases for blockade of adipogenesis by CaM kinase II
inhibitors may well reflect the two distinct phases of CaM kinase
activation observed as the cell progress through adipogenesis (Fig.
3).
Inhibition of CaM kinase activities between
6-12 and 24-48 h of induction blocks adipogenesis. Confluent
cultures of 3T3-L1 cells were induced to adipogenesis by incubating
with dexamethasone+IBMX. KN-93 (10 µM) was added into the
culture media for the periods indicated. Panel A, at day 8, cells were fixed, stained with oil-red O, and photographed using a Zeiss Axiophot
system. Panel B, the number of adipocytes relative to that
of total cells on the coverslip represents the ratio of
differentiation. The "+" denotes cultures progressing to 85%
differentiation. The "
" denotes cultures that failed to progress
to adipocytes (
5% differentiation). The data are mean values ± S.E. from at least three separate experiments.
We explored the influence of CaM kinase II inhibition on expression of
marker genes for adipogenic conversion (Fig. 6). Cells were induced to adipogenesis with dexamethasone+ IBMX and KN-93 was
added at either the 6-12-h or 24-48-h periods. aP2, fatty acyl-CoA
synthetase (FACS), and lipoprotein lipase (LPL) are well known markers
of adipogenesis, LPL being an early gene product, aP2 and FACS being
intermediate early gene products (see Ref. 24). The mRNA levels of
glyceraldehyde-phosphate dehydrogenase, which do not change during
differentiation (24), were probed also. Inhibition of CaM kinase II
activity from 24 to 48 h essentially blocked the expression of
aP2, FACS, and LPL mRNAs. Treating the cells with inhibitor for a
period of 6-12 h following induction of adipogenesis severely
attentuated the activation of the marker genes. These data demonstrate
an obligate role of CaM kinase II activation in the early program of
adipogenic conversion, confirming the phenotypic displays of these
cells stained for lipid accumulation with oil-red O at day 8 following
treatment with CaM kinase inhibitors (Fig. 5A).
Activation of CaM kinase II is critical to several aspects of cell
physiology and biology, particularly regulation of the cell cycle (25,
26), control of nuclear envelope breakdown (27), and promotion of
neurite outgrowth and growth cone mobility (28). In the current work we
report the activation of CaM kinase II during adipogenesis. Activation
of CaM kinase has been shown to lead to phosphorylation of C/EBP, a
member of the bZip family of transcription factors (29), providing one
possible paradigm linking CaM kinase activation to gene expression. In
addition, CaM kinase II has been shown to phosphorylate a non-consensus site on ATF-1 and CREB, which inhibits activation of these two transcription factors (30), providing another possible role of CaM
kinase II activation and gene expression. More importantly, blockade of
the CaM kinase II activity, but not that of either protein kinase A or
protein kinase C, prevents the progression of 3T3-L1 embryonic
fibroblasts to adipocytes. The activation of CaM kinase II is dynamic
and has precise temporal boundaries for interruption by selective
inhibitors like KN-62 or KN-93. Identification of substrates for CaM
kinase II critical to adipogenic progression remains an important goal.
In summary, activation of CaM kinase II is shown to be an obligate step
toward induction and progression of adipogenesis.