(Received for publication, October 2, 1996, and in revised form, November 21, 1996)
From the Department of Biochemistry, Previous studies using L6 myotubes have suggested
that glycogen synthase kinase-3 (GSK-3) is phosphorylated and
inactivated in response to insulin by protein kinase B (PKB, also known
as Akt or RAC) (Cross, D. A. E., Alessi, D. R., Cohen, P., Andjelkovic, M., and Hemmings, B. A. (1995) Nature 378, 785-789). In
the present study, marked increases in the activity of PKB have been
shown to occur in insulin-treated rat epididymal fat cells with a time course compatible with the observed decrease in GSK-3 activity. Isoproterenol, acting primarily through
Glycogen synthase kinase-3 (GSK-3)1
was first discovered by virtue of its ability to phosphorylate and
inactivate the regulatory enzyme of glycogen synthesis in mammals,
glycogen synthase (GS) (reviewed in Refs. 1-3). In a number of tissues
GS is activated by insulin, and in skeletal muscle this has been shown
to involve the dephosphorylation of the sites phosphorylated by GSK-3
(4-7). This suggested that GSK-3 might be inactivated by insulin,
although more attention was initially focused on the regulation
(activation) of the relevant phosphatase, a glycogen-bound form of
protein phosphatase-1 (PP-1G) (8). Recently, however, it has become clear that GSK-3 itself is indeed subject to acute regulation, being
inactivated in response to insulin or growth factors (9-16) and also
following T-cell stimulation (17). Where tested, this inactivation was
reversed by treatment with serine/threonine phosphatases indicating
that it was due to increased phosphorylation of GSK-3 on one or more
Ser/Thr residues (11).
It has since been shown that the inactivation of GSK-3 is associated
with increased phosphorylation of a serine residue near the N terminus
(Ser-9 of the In this study we demonstrate that insulin activates PKB in rat fat
cells and that this may underlie the effects of insulin on GSK-3
activity in these cells. In addition, we report that the Male Wistar rats were fed ad libitum
up to the time of killing on a stock laboratory diet (CRM; Bioshore,
Manea, Cambs, UK). Collagenase was purchased from Worthington
Diagnostic Systems (Freehold, NJ). Enhanced chemiluminescence Western
blotting detection kits and [ Adipocytes were isolated from epididymal fat pads,
preincubated, and then incubated (at 150-250 mg cell dry weight/ml) as described previously (13, 24). For fat cells treated with wortmannin,
LY 294002, or rapamycin, the appropriate control incubations contained
0.1% dimethyl sulfoxide.
The activities
of GSK-3 and p70S6k were measured in immunoprecipitates
using peptide substrates as described previously (13). For the
measurement of PKB activity, adipocytes (150-200 mg dry cell weight)
were extracted in 1 ml of 50 mM Hepes (pH 7.6), 0.2 mM EDTA, 2.2 mM EGTA, 1 mM
dithiothreitol, 100 mM KCl, 10% glycerol, 1% Triton
X-100, 1 µM microcystin, and 1 µg/ml each of pepstatin, leupeptin, and antipain. PKB was then immunoprecipitated from 750 µl
of cell extracts with 5 mg of protein A-Sepharose and 5 µl of PKB
antiserum for 2 h at 4 °C. Immunoprecipitates were washed twice
with 1 ml of PKB assay buffer (20 mM MOPS (pH 7.0), 1 mM EDTA, 1 mM EGTA, 0.01% Brij 35, 5%
glycerol) containing 0.5 M NaCl and twice with 1 ml of the
same buffer without added salt. The protein A beads were finally
resuspended in 40 µl of PKB assay buffer containing 0.1%
mercaptoethanol and 2.5 µM cAMP-dependent protein kinase inhibitor peptide (IP20). The activity of
PKB in these immunoprecipitates was measured using either myelin basic protein (MBP, 0.5 mg/ml final concentration) or the synthetic peptide
based on the sequence surrounding the serine phosphorylation site of
GSK-3 (Ser-21 of GSK-3 To separate proteins by anion exchange
chromatography, fat cells were extracted as above and applied to a
Pharmacia SMART system Mono Q column equilibrated with Buffer A (50 mM Tris (pH 7.3), 2 mM EDTA, 2 mM
EGTA, 5% glycerol, 0.03% Brij 35 (w/v), 0.1% mercaptoethanol, and 1 µg/ml pepstatin, leupeptin, and antipain). The column was developed
with a 0-50% gradient of Buffer B (Buffer A plus 1 M
NaCl) at a flow rate of 50 µl/min. Fractions (100 µl) were assayed
for cross-tide kinase activity (10-µl fraction in a final reaction
volume of 25 µl).
Western blotting was performed using immunoprecipitates run on 16-cm
SDS-10% polyacrylamide gels (25) and blotted onto Immobilon-P membrane
(Millipore, Watford, Herts, UK). Blots were incubated with PKB
antiserum at a 1:400 dilution and immunoreacting proteins were
visualized using the enhanced chemiluminescence detection system.
The activity of
glycogen synthase in fat cell extracts was determined as described
elsewhere (13). Lipolysis was assayed by measuring the appearance of
glycerol in the incubation medium (26).
Incubation of adipocytes with isoproterenol led to
a rapid decrease in the activity of GSK-3 (Fig. 1).
Maximum inactivation of GSK-3 in response to isoproterenol was seen
after approximately 20 min and was sustained for at least 40 min. This
time course and the extent of inactivation (approximately 50%) were
broadly similar to the changes in GSK-3 activity seen in response to
insulin (see Fig. 1 and Ref. 13), although the effect of insulin was more rapid in onset. The effects of insulin and isoproterenol on GSK-3
activity were not additive (Fig. 2A). GSK-3
activity in all experiments was measured by following phosphate
incorporation into a peptide substrate corresponding to the GSK-3
phosphorylation site in eukaryotic initiation
factor-2B.2 A similar inactivation of GSK-3
in response to isoproterenol was also observed when the kinase was
assayed using a peptide based on its phosphorylation site in GS (see
Ref. 13) (data not shown).
In order to investigate the mechanism by which isoproterenol
inactivates GSK-3 in fat cells, a number of agents designed to either
mimic or inhibit the effects of The binding of
The effect of various agents on lipolysis and glycogen synthase
activity ratio
3-adrenoreceptors, was found to decrease GSK-3 activity
to a similar extent (approximately 50%) to insulin. However, unlike
the effect of insulin, the inhibition of GSK by isoproterenol was not
found to be sensitive to inhibition by the phosphatidylinositol
3
-kinase inhibitors, wortmannin or LY 294002. The change in GSK-3
activity brought about by isoproterenol could not be mimicked by the
addition of permeant cyclic AMP analogues or forskolin to the cells,
although at the concentrations used, these agents were able to
stimulate lipolysis. Isoproterenol, but again not the cyclic AMP
analogues, was found to increase the activity of PKB, although to a
lesser extent than insulin. While wortmannin abolished the stimulation
of PKB activity by insulin, it was without effect on the activation
seen in response to isoproterenol. The activation of PKB by
isoproterenol was not accompanied by any detectable change in the
electrophoretic mobility of the protein on SDS-polyacrylamide gel
electrophoresis. It would therefore appear that distinct mechanisms
exist for the stimulation of PKB by insulin and isoproterenol in rat
fat cells.
-isoform of mammalian GSK-3 which corresponds to
Ser-21 of the
-isoenzyme) (11, 14, 16). This is of special interest
since several insulin/growth factor-activated protein kinases have been
shown to phosphorylate these residues in GSK-3 in vitro.
These include the p90 ribosomal S6 protein kinase (p90rsk),
which is downstream of the mitogen-activated protein kinase (MAP
kinase) cascade, and the p70 ribosomal S6 kinase (p70S6k)
(18, 19). Studies employing rapamycin, which selectively blocks
activation of p70S6k, appear to rule out a role for this
enzyme in regulating GSK-3 in the cell types thus far studied (10, 12,
13). The evidence for a role for the MAP kinase pathway is mixed, with
transfection approaches (either using p90rsk itself (14) or
dominant negative mutants of an upstream kinase (15)) providing
evidence for a role for this pathway and recent data from fat cells
(13) and L6 myotubes (16) giving, respectively, correlative and strong
evidence against a role for the MAP kinase pathway in regulating GSK-3
activity in response to insulin. The latter studies (16) provided
evidence that protein kinase B (PKB, also known as RAC or Akt) might be
responsible for the control of GSK-3 in L6 myotubes. PKB phosphorylates
Ser-9 and Ser-21 in the
- and
-isoforms of GSK-3, respectively,
is activated by insulin (16, 20, 21), and may lie downstream of
phosphatidylinositol 3
-kinase (PI 3
-kinase) (16, 22). PI 3
-kinase is
implicated in the control of GSK-3 in several cell types by the
observation that wortmannin, a selective but not absolutely specific
inhibitor of PI 3
-kinase, blocks the inactivation of GSK-3 by insulin
and other agents (10, 12, 13, 16). LY 294002, a structurally unrelated
inhibitor of PI 3
-kinase (23), has recently been shown to block the
insulin-induced inactivation of GSK-3 in L6 myotubes (16).
-adrenergic
agonist, isoproterenol, also causes the stimulation of PKB and
inactivation of GSK-3 in fat cells. The effects of isoproterenol on
these two kinases do not appear to be mediated by increases in the
intracellular level of cAMP. A number of lines of evidence suggest that
distinct mechanisms are involved in the activation of PKB by insulin
and isoproterenol.
Materials
-32P]ATP were from
Amersham Int. (Amersham, Bucks, UK), and pepstatin, leupeptin, and
antipain were from Cambridge Research Biochemicals (Harston, Cambs,
UK). Microcystin, dithiothreitol, and rapamycin were obtained from
Calbiochem (Nottingham, UK). LY 294002 was purchased from the Alexis
Corp. (Nottingham, UK). All other chemicals and biochemicals were from
Sigma or BDH (both of Poole, Dorset, UK). The
3-agonist,
BRL 37344, was a gift from Dr. P. Young of SmithKline Beecham
Pharmaceuticals (Welwyn, Herts, UK). The PKB antiserum was raised in
rabbits immunized with a synthetic peptide corresponding to residues
465-480 of the human PKB sequence and was used without further
purification. Anti-GSK-3 serum was a gift from Dr. J. Vandenheede
(Katholieke Universiteit, Leuven, Belgium). All synthetic peptides,
including those used for the assay of GSK-3, p70S6k, and
PKB were synthesized by Dr. G. Bloomberg (Dept. of Biochemistry, University of Bristol, UK). Wortmannin and rapamycin were dissolved in
dimethyl sulfoxide and stored at
20 °C.
and Ser-9 of GSK-3
), "cross-tide" (16)
(100 µM), for 15 min at 30 °C.
Inactivation of GSK-3 in Fat Cells in Response to
Isoproterenol
Fig. 1.
Time course for the effect of insulin and
isoproterenol of GSK-3. Adipocytes were incubated with insulin (83 nM, ) or isoproterenol (1 µM,
) for the
times indicated prior to extraction. Results are expressed as a
percentage of the control value at time 0 (39.5 ± 4.9 pmol of
phosphate incorporated into substrate peptide/min/g dry cells,
n = 3) and are means ± S.E. for three separate
cell preparations. In the absence of hormones the activity of GSK-3 did
not alter significantly with time and was 39.3 ± 3.2 pmol of
phosphate/min/g dry cells (n = 3) after the 40-min incubation.
[View Larger Version of this Image (15K GIF file)]
Fig. 2.
The effect of -adrenergic agonists and
agents that increase intracellular cAMP levels on GSK-3 activity.
A, adipocytes were incubated for 10 min with 83 nM insulin (Ins), 1 µM
isoproterenol (Iso), 5 nM BRL37344
(BRL), 2 mM cpt-cAMP (cpt), 5 mM db-cAMP (db), or 5 nM forskolin
(For) prior to extraction. Where indicated 10 µM propranolol (Pro) was added 10 min before
addition of isoproterenol. Results are expressed as a percentage of the
control value (no additions) for GSK-3 activity (45.6 ± 3.0 pmol
of phosphate incorporated into substrate peptide/min/g dry cells,
n = nine separate cell preparations) and are means ± S.E. for the number of separate cell preparations shown in
parentheses. Significance, as assessed by the Student's
t test, is indicated as follows: **, p < 0.0001, *, p < 0.001 versus appropriate
control. B, dose response for the effect of BRL37344 on
GSK-3 activity. Adipocytes were incubated for 10 min with varying
concentrations of BRL 37344. Results are expressed as a percentage of
the control value in the absence of BRL37344 (see A) and are
means ± S.E. for two separate cell preparations.
[View Larger Version of this Image (19K GIF file)]
-adrenergic agents was tested. The
effects of isoproterenol on fat cells are mediated through its actions
on
-adrenergic receptors. However, the effects of isoproterenol on
GSK-3 in this system were not blocked by the classical
1-/
2-antagonist propranolol (Fig.
2A). This finding suggested that isoproterenol may be acting
primarily via the
3-receptor which is known to be
insensitive to propranolol (see Ref. 28). The
3-receptor
is thought to be the primary adrenergic receptor mediating the effects
of
-adrenergic agents on lipolysis in rat white adipose tissue (29).
Like isoproterenol, the
3-receptor-specific agonist, BRL
37344 (30, 31), caused a significant decrease in the activity of GSK-3
(Fig. 2A). The effects of BRL 37344 on GSK-3 activity were
half-maximal at approximately 5 nM (Fig. 2B). This is consistent with reported EC50 values for the
actions of BRL 37344 on the stimulation of lipolysis in this cell type
(32-34).
-adrenergic agonists to their receptors activates
adenylate cyclase and thereby increases cytoplasmic cAMP levels. In
order to address the possibility that the inactivation of GSK-3 was a
consequence of elevated cAMP, we adopted two approaches. In the first,
cells were treated with the cell-permeant cAMP analogues, db-cAMP and
cpt-cAMP. This had no effect on the activity of GSK-3 (Fig.
2A) although they did stimulate lipolysis over 8-fold (see Table I). In the second, we made use of forskolin, which
activates adenylate cyclase. Treatment of cells with forskolin also
failed to cause inactivation of GSK-3 (Fig. 2A) but did
increase the rate of lipolysis in these cells to an extent comparable
with isoproterenol (Table I). It therefore seems unlikely that
isoproterenol acts via cAMP to inactivate GSK-3 in these cells. Table I
also shows the effects of various agents on the activity ratio
((
/+)-glucose 6-phosphate) of GS in fat cells. Despite decreasing the
activity GSK-3, neither isoproterenol nor BRL 37344 on their own had
any effect on the activity ratio of GS. The pronounced stimulation of
GS seen with insulin was approximately halved when isoproterenol was
included in the incubation.
Condition
Rate of lipolysis
(nmol glycerol released/10 min/g dry cells)
GS activity ratio
((
/+)-glucose 6-phosphate)
Control
207
± 39 (6)
0.048 ± 0.009 (6)
Insulin (83 nM)
225 ± 47 (6)
0.238 ± 0.048 (6)*
Isoproterenol (1 µM)
2495
± 340 (6)*
0.059 ± 0.008 (6)
Insulin + isoproterenol
2833 (2)
0.113 (2)
BRL37344 (5 nM)
1393 (2)
0.030 (2)
cpt-cAMP (2 mM)
1648 (2)
0.063 (2)
db-cAMPa (5 mM)
2512 (2)
0.055 (2)
Forskolin (5 nM)
2425 (2)
0.057 (2)
Since the inactivation
of GSK-3 by isoproterenol could not be mimicked by raising
intracellular cAMP levels, we sought to examine other possible
mechanisms by which this effect might occur. In rat fat cells, the
inactivation of GSK-3 by insulin is blocked by wortmannin, an inhibitor
of PI 3-kinase (13). In contrast, wortmannin had no effect on the
inactivation of GSK-3 by isoproterenol or by BRL 37344. (Fig.
3). Since wortmannin does block the regulation of GSK-3
by other stimuli such as epidermal growth factor and insulin-like
growth factor-1 in other cell types (10, 12, 16), the mechanism by
which adrenergic agonists decrease GSK-3 activity appears to be quite
distinct from that employed by agents acting through receptor tyrosine
kinases. To further test for the possible involvement of PI 3
-kinase
in the effect of isoproterenol on GSK-3, the fat cells were pretreated
with the unrelated PI 3
-kinase inhibitor LY 294002. As is the case for
wortmannin, although LY 294002 abolished the effect of insulin on
GSK-3, it was unable to prevent the inhibition seen in response to
isoproterenol (Fig. 3).
As the inactivation of GSK-3 induced by isoproterenol was not affected by wortmannin, which completely blocks activation of p70S6k by insulin in fat cells (13), it is very unlikely that the inactivation of GSK-3 is mediated by this signaling pathway. This is supported by the observation that rapamycin (which specifically blocks the activation of p70S6k without affecting MAP kinase) had no effect upon the inactivation of GSK-3 induced by isoproterenol or BRL 37344 (Fig. 3). Rapamycin also has no effect upon the control of GSK-3 in other cell types tested (10, 12, 13, 16). At the dose used here (20 nM), rapamycin completely blocked the activation of p70S6k induced by insulin (see Ref. 13).
Insulin and Isoproterenol Activate PKB in Rat Epididymal Fat CellsCross et al. (16) have recently shown that PKB
can phosphorylate GSK-3 in vitro and cause its inactivation.
PKB phosphorylates the same N-terminal serine residue phosphorylated by
p70S6k or p90rsk (18, 19). Since PKB is thought to
be downstream of PI 3-kinase (20, 22), this would provide an
explanation of the ability of wortmannin to block completely the
regulation of GSK-3 by insulin and other stimuli. Thus, insulin by, for
example, activating PI 3
-kinase would also bring about the activation
of PKB and therefore the phosphorylation and inactivation of GSK-3.
We have therefore examined the regulation of PKB activity by insulin
and isoproterenol in fat cells. Extracts from adipocytes treated with
insulin or isoproterenol were separated by Mono Q chromatography
essentially as described by Cross et al. (16). Insulin
increased the kinase activity toward cross-tide in two broad peaks (I
and II) (Fig. 4). Peak I (eluting between 150-250 mM NaCl) was found to coincide with the elution position of
PKB as shown by immunoblotting (Fig. 4, inset). The
cross-tide kinase activity eluting at this position was thus due, at
least in part, to the activity of PKB. Immunoreactive PKB could not be
detected in peak II, and thus the activity toward cross-tide in these
fractions is likely to be due to some other unidentified kinase.
Isoproterenol increased the kinase activity toward cross-tide in peak I
but not peak II. The effect of isoproterenol on the kinase activity in
this peak was less than the stimulation seen in response to insulin.
The stimulation of PKB by both insulin and isoproterenol in whole cell extracts was confirmed by measuring kinase activity in anti-PKB immunoprecipitates using either cross-tide or myelin basic protein as a substrate. Table II shows that the both hormones increase the activity of PKB, with the effects of insulin being greater than those of isoproterenol. The fold effects observed were similar whether cross-tide or MBP was used to measure kinase activity. To further verify that the kinase activity measured in immunoprecipitates was actually PKB, we preincubated the protein A-Sepharose beads with anti-PKB antibody plus the peptide to which the antiserum was raised (see "Experimental Procedures") prior to the addition of beads to fat cell extracts. This peptide was able to compete with PKB for binding to the antiserum, as shown by its ability to abolish kinase activity toward MBP in immunoprecipitates from control, insulin-, or isoproterenol-treated cell extracts (data not shown).
|
The time courses for activation of PKB by insulin and isoproterenol are
shown in Fig. 5. Insulin caused an extremely rapid stimulation of PKB activity which declined slowly over 40 min. In
contrast, the activation in response to isoproterenol was slower and
more transient. A comparison of the time courses for GSK-3 inactivation
(Fig. 1) and PKB stimulation (Fig. 5) shows a number of discrepancies
between the two. First, complete activation of PKB, even with
isoproterenol, was seen after 5 min, while maximal inhibition of GSK-3
occurred only after 20 min. Second, inactivation of GSK-3 was sustained
even after PKB activities had returned to basal levels.
The stimulation of PKB by insulin has also been reported in a number of
other cell types (16, 20, 21) and is accompanied by a change in the
mobility of the protein on SDS-PAGE. This band-shift in response to
insulin is thought to be a result of changes in the phosphorylation of
PKB (21, 22, 35) and may be seen on stimulation of fat cells with the
hormone (see Fig. 6 inset and Ref. 21).
Interestingly the increase in PKB activity on treatment with
isoproterenol is not associated with any detectable change in the
electrophoretic mobility of the protein. The effects of insulin and
isoproterenol on PKB activity were not additive (Fig. 6). The primary
effect of isoproterenol again appears to be via the
3-adrenergic receptor because BRL 37344 is also able to
stimulate PKB. The lack of effect of either cpt-cAMP or db-cAMP indicates that the increase of PKB activity in response to
-adrenergic agonists does not occur as a result of increased
cytoplasmic cAMP levels (Fig. 6).
The data in Fig. 7 show that while the effect of insulin
was abolished by preincubation of the adipocytes with wortmannin, the
inhibitor had no significant effect on the stimulation of PKB by
isoproterenol. In agreement with others (21), wortmannin also reversed
the effect of insulin on the band-shift seen with PKB on SDS-PAGE (data
not shown). Rapamycin was without effect on changes in the activity of
PKB in response to either hormone (data not shown).
Given that PKB is reported to lie upstream of p70S6k (22),
we determined whether isoproterenol activated this kinase.
Isoproterenol gave a significant activation, which was blocked by
rapamycin but was modest in comparison to that induced by insulin (Fig. 8). In contrast to the effect of isoproterenol on PKB
activation and GSK-3 inhibition, the increase in p70S6k
activity was largely eliminated by wortmannin and rapamycin.
The data presented in this paper show that in
freshly isolated rat adipocytes treated with isoproterenol, there is a
marked inhibition of GSK-3 activity. Isoproterenol was also found to increase the activity of the kinase proposed to be upstream of GSK-3,
namely PKB. We have been unable to find any change in the activity
ratio of GS ((/+)-glucose 6-phosphate) in response to isoproterenol;
indeed, isoproterenol partially reverses the activation of GS seen in
response to insulin (see also Ref. 36). Since decreases in GSK-3
activity have been proposed to be responsible, at least in part, for
activation of GS, these findings were of interest. Our results seem to
indicate that GSK-3 is not the primary regulator of GS activity in rat
fat cells and suggest that PP1-G probably plays a more important role
in the regulation of glycogen synthesis in vivo. It has been
suggested that PP1-G plays the dominant role in the regulation of rat
muscle GS (see Ref. 37). The mechanisms involved in the regulation of
glycogen metabolism in different tissues may be distinct. In rat muscle
(37) and in 3T3-L1 adipocytes (38) the stimulation of GS by insulin is partially sensitive to inhibition by rapamycin, whereas in primary rat
adipocytes it is not (13).
As PKB has been implicated as the direct upstream regulator of GSK-3 in L6 myotubes (16), it was important to determine its activity in adipocytes. Here we report the first direct measurement of PKB activity in rat fat cells and have shown substantial and rapid stimulation of the kinase in response to insulin. Isoproterenol was also found to activate PKB in these cells. Overall, the congruence between the actions of all effectors on GSK-3 inactivation and PKB activation supports the view that PKB is the upstream regulator of GSK-3 in fat cells, as concluded by Cross et al. (16) from their studies in L6 myotubes. In further agreement with this view, the activation of PKB in the present study is considerably more rapid than the inactivation of GSK-3 (Figs. 1 and 5). However, in the absence of a specific inhibitor of PKB it is not possible to obtain conclusive proof that PKB is indeed upstream of GSK-3.
The finding that isoproterenol also causes a small stimulation of p70S6k is interesting in light of the proposal that this kinase is downstream of PKB (22). The physiological relevance for the activation of this S6 kinase under these conditions remains to be explained.
Our results also have important implications for understanding the
mechanisms by which PKB is regulated in fat cells. The exact means by
which PKB activity is controlled remains obscure, with a number of
different factors being implicated, acting either independently or in
conjunction with each other. In other cell types there is strong
evidence that the activation of PKB is mediated via PI 3-kinase (20,
22), although the direct involvement of PI 3
-kinase products remains
controversial (20, 39). In fat cells, the differing effects of
wortmannin on PKB activation by insulin and isoproterenol, along with
the failure of isoproterenol to produce a band-shift in PKB, would seem
to indicate that different mechanisms are involved in the stimulation
of the kinase by these hormones. A role for PI 3
-kinase in the
activation of PKB by insulin in fat cells seems likely as the
stimulation is abolished by preincubation with wortmannin. Changes in
PKB activity in response to isoproterenol do not appear to involve
PI 3
-kinase activation as this inhibitor was without effect.
Isoproterenol has no effect on adipocyte PI 3
-kinase activity when
measured in anti-PI 3
-kinase p85 immunoprecipitates (40). It is,
however, possible that
-adrenergic agonists increase the amount of
PI 3
-kinase products in fat cells via activation of another, as yet
uncharacterized, isoform of PI 3
-kinase. Indeed, a
G
-sensitive isoform of PI 3
-kinase has been
described in myeloid-derived cells and in platelets (41, 42).
In agreement with Kohn et al. (21), the activation of PKB in fat cells in response to insulin was accompanied by a change in the mobility of the kinase on SDS-PAGE. In this respect, the mechanism by which isoproterenol activates PKB again clearly differs from that involved in insulin action, as it is not associated with a change in the migration of the protein on gel electrophoresis (Fig. 6, inset). The activation of the kinase and SDS-PAGE band-shift in response to platelet-derived growth factor in Rat-1 fibroblasts is thought to be a direct result of changes in the phosphorylation of PKB, as both are reversed by treatment with phosphatase (22). It is likely that activation of PKB by insulin in fat cells involves a similar phosphorylation step. Activation in response to isoproterenol, however, appears to occur independently of the phosphorylation event responsible for band-shift of the protein. It is possible that isoproterenol induces changes in the phosphorylation of PKB which cause changes in kinase activity but are not manifested as changes in its electrophoretic mobility.
Finally, it has been proposed that activation of PKB can occur as a
result of protein-protein interactions mediated by its pleckstrin
homology domain (43, 44). The pleckstrin homology domains from the
three known PKB isoforms (,
, and
) have recently been shown
to interact with PKC-
, -
, and -
subtypes and with G protein
subunits (G
) in vitro (43), but the
effects of these interactions on PKB activity have yet to be
determined. It is possible that the effect of isoproterenol is mediated
by the binding of G
subunits to the pleckstrin
homology domain, in a manner analogous to the activation of
-adrenergic receptor kinase on binding of G
(see
Ref. 27). Further work clearly needs to be done to clarify the question
of how
-adrenergic agonists stimulate PKB.