Arsenite stimulated glucose transport in 3T3-L1 adipocytes involves both Glut4 translocation and p38 MAPK activity Merlijn Bazuine, D.. These data show that in 3T3-L1 adipocytes both ars
Trang 1Arsenite stimulated glucose transport in 3T3-L1 adipocytes involves both Glut4 translocation and p38 MAPK activity
Merlijn Bazuine, D Margriet Ouwens, Daan S Gomes de Mesquita and J Antonie Maassen
Department of Molecular Cell Biology, Leiden University Medical Centre, Leiden, the Netherlands
The protein-modifying agent arsenite stimulates glucose
uptake in 3T3-L1 adipocytes In the current study we have
analysed the signalling pathways that contribute to this
response By subcellular fractionation we observed that
arsenite, like insulin, induces translocation of the GLUT1
and GLUT4 glucose transporters from the low-density
membrane fraction to the plasma membrane Arsenite did
not activate early steps of the insulin receptor (IR)-signalling
pathway and the response was insensitive to inhibition of
phosphatidylinositol-3¢-kinase (PI-3¢) kinase by
wortman-nin These findings indicate that the ÔclassicalÕ
IR–IR substrate–PI-3¢ kinase pathway, that is essential for
insulin-induced GLUT4 translocation, is not activated by
arsenite However, arsenite-treatment did induce
tyrosine-phosphorylation of c-Cbl Furthermore, treatment of the
cells with the tyrosine kinase inhibitor, tyrphostin A25,
abolished arsenite-induced glucose uptake, suggesting that
the induction of a tyrosine kinase by arsenite is essential for
glucose uptake Both arsenite and insulin-induced glucose
uptake were inhibited partially by the p38 MAP kinase inhibitor, SB203580 This compound had no effect on the magnitude of translocation of glucose transporters indica-ting that the level of glucose transport is determined by additional factors Arsenite- and insulin-induced glucose uptake responded in a remarkably similar dose-dependent fashion to a range of pharmacological- and peptide-inhibi-tors for atypical PKC-k, a downstream target of PI-3¢ kinase signalling in insulin-induced glucose uptake These data show that in 3T3-L1 adipocytes both arsenite- and insulin-induced signalling pathways project towards a similar cel-lular response, namely GLUT1 and GLUT4 translocation and glucose uptake This response to arsenite is not func-tionally linked to early steps of the IR–IRS–PI-3¢ kinase pathway, but does coincide with c-Cbl phosphorylation, basal levels of PKC-k activity and p38 MAPK activation Keywords: PKC-k; PKB; PI-3¢ kinase; insulin; Cbl
Insulin induces multiple responses in target tissues such as
adipocytes and muscle through the intracellular activation
of several signal transduction pathways These responses
include a pronounced anabolic action on protein and lipid
metabolism, an antiapoptotic response, an increase in
glucose uptake, and stimulation of glycogen synthesis
[1,2] Insulin-stimulated glucose uptake occurs primarily
via translocation of the GLUT4 glucose transporter to the
plasma membrane [3,4] This process is initiated by the
activation of the insulin receptor (IR) tyrosine kinase
followed by receptor autophosphorylation and tyrosine
phosphorylation of downstream effectors like insulin
receptor substrate-1 (IRS-1), IRS-2 and related proteins
Tyrosine phosphorylated IRS proteins provide docking
sites for class I phosphatidylinositol-3¢ (PI-3¢) kinase that becomes activated upon binding to these proteins [5,6] Numerous studies have shown that PI-3¢ kinase activation provides an essential signal for the stimulation of glucose uptake by insulin [7,8] Downstream targets of PI-3¢ kinase
in 3T3-L1 adipocytes that have been implicated in signalling towards GLUT4 translocation are the AGC kinase family members PDK1, PKB and the atypical PKC-k/-f [9–11], of which 3T3-L1 adipocytes only express the k-isoform [12] Recent data also demonstrate the involvement of an additional, nonPI-3¢ kinase dependent pathway involving c-Cbl which becomes tyrosine-phosphorylated upon APS (adapter protein with a PH and SH2 domain)-mediated association with the activated insulin receptor [13] Sub-sequently, tyrosine-phosphorylated c-Cbl translocates towards the caveolae and induces the activation of the small GTP-binding protein, TC10 [14], ultimately signalling towards the exocyst complex (Exo70) involved in GLUT4 translocation [15]
Apart from insulin, some other stimuli, like muscle contraction, H2O2 and hyperosmotic shock, have been shown to stimulate GLUT4-mediated glucose uptake in adipocytes and muscle Most studies show that these stimuli are not sensitive to inhibition by wortmannin, indicating PI-3¢ kinase is not involved in glucose uptake mediated by these agents [16–18]
Sodium arsenite is known for its atherogenic, carcino-genic and genotoxic effects Recently, arsenite has also
Correspondence to J A Maassen, Department of Molecular Cell
Biology, Leiden University Medical Centre, Wassenaarseweg 72,
PO Box 9503, 2333 AL, Leiden, the Netherlands.
Fax: + 31 71 5276437, Tel.: + 31 71 5276127,
E-mail: J.A.Maassen@lumc.nl
Abbreviations: IR(s), insulin receptor (substrates); IBMX,
1-methyl-3-isobutylxanthine; PI-3¢, phosphatidylinositol-3¢-kinase;
2-DOG, 2-deoxy- D [ 14 C]glucose; BIM I, bisindolylmaleimide I;
LDM, low density microsome; PM, plasma membrane;
TPA, 12-O-tetradecanoylphorbol 13-acetate.
(Received 18 December 2002, revised 24 July 2003,
accepted 28 July 2003)
Trang 2been used effectively as a chemotherapeutic drug in the
treatment of acute promyelocytic leukaemia patients
[19,20] At the protein level, arsenite exerts its biological
effects through modification of vicinal sulfhydryl groups
in specific target proteins For instance, arsenite
specific-ally inactivates the E2 subunit of branched-chain
alpha-keto acid dehydrogenase (but not the other subunits) [21]
and activates heat shock protein 70 [22] Arsenite is also a
potent activator of the stress kinases, JNK and p38, by
modulating the activity of an unidentified target protein
[23] Furthermore, arsenite has been shown to induce
glucose uptake in 3T3-L1 adipocytes, baby hamster
kid-ney cells and L6 muscle cells [24–26] As the action of
arsenite involves the modification of a limited number of
arsenite-sensitive target proteins, we hypothesized that
elucidation of the mechanism of arsenite-induced glucose
uptake may contribute to a better understanding of
insulin-induced glucose uptake
In this study, we observed that arsenite displays
insulin-like effects on GLUT4-mediated glucose transport in
adipocytes To explore the underlying mechanism we
analysed the signalling pathways that are activated by
arsenite in comparison to insulin and that contribute to
stimulation of glucose uptake
Experimental procedures
Materials Dulbecco’s modified Eagle’s medium (DMEM) was pur-chased from Life Technologies, Inc.; foetal bovine serum was from Brunschwig, Amsterdam; bovine insulin, 1-methyl-3-isobutylxanthine (IBMX), dexamethasone, 12-O-tetradecanoylphorbol 13-acetate (TPA) and 2-deoxy-glucose were obtained from Sigma 2-deoxy-D-[14C]glucose was purchased from NEN-Dupont Tyrphostin A25, SB203580, chelerythrine chloride, bisindolylmaleimide I, Go¨ 6976, Ro 31-8220 and Ro 32-0432 were from Calbio-chem LY-294002, microcystin LR and wortmannin were obtained from Alexis Myristoylated pseudosubstrate pep-tide inhibitors for PKC-a/-b and PKC-k/-f were purchased from Biomol For an overview of the characteristics of the pharmacological inhibitors applied in this study, see Table 1
Antibodies Polyclonal antisera recognizing IRS-1, and the regulatory subunit of PI-3¢ kinase were described previously [27] Table 1 Characteristics of pharmacological inhibitors applied in this study.
Pharmacological inhibitor Concentration applied Described target IC 50 Reference
MAPK activated protein-kinase 1b 50 n M [52]
Mitogen and stress activated kinase-1 <1 l M [53]
4 l M in vivo
[11] Glycogen synthase kinase 3b 38 n M
2.8 n M
[53] [50] MAPK activated protein-kinase 1b 3 n M [52]
Mitogen and stress activated protein kinase 8 n M [53]
a
Disputed by Davies et al [53].
Trang 3A polyclonal antiserum recognizing IRS-2 is described in
a recent paper from our group [28] HRP-conjugated
mouse monoclonal anti-(phosphotyrosine Ig) pY-20,
mouse monoclonal pY-20, rabbit polyclonal antibody
against PKC-k/-f (C-20) and polyclonal against Cbl
(C-15), goat polyclonal antibodies recognizing the
cata-lytic (p110a) subunit of PI-3¢ kinase (C-17) and GLUT4
(C-20) were obtained from Santa Cruz Biotechnology,
Inc Rabbit polyclonal antibody recognizing IRb-chain
and mouse monoclonal integrin b1, Cbl and PKC-k were
purchased from Transduction Laboratories The
phospho-specific antibodies recognizing PKC-k (T403), caveolin-1
(Y14), PKB (T308 and S473), MAPKAP-K2 (T334), p38
(T180/Y182) and ERK-1/)2 (T202/Y204) were obtained
from Cell Signalling Technology Sheep polyclonal
anti-body recognizing PKB was purchased from Upstate
Biotechnology The rabbit polyclonal antibodies against
GLUT-1 and -4 (used in the PM-lawn assay) have been
described [29] The appropriate HRP- and
FITC-conjugated secondary antibodies were obtained from
Promega
Cell culture
3T3-L1 Fibroblasts, obtained from American Type
Cul-ture Collection (Manassas, VA, USA), were culCul-tured and
differentiated to adipocytes as described previously [30]
Cells were routinely used 7 days after completion of the
differentiation process, with only cultures in which >95%
of cells displayed adipocyte morphology being used Prior to
use, adipocytes were serum starved for 16 h with DMEM
supplemented with 0.5% foetal bovine serum
Membrane isolation assay
3T3-L1 adipocytes were stimulated as indicated in the figure
legends Subsequently cells were washed twice in ice-cold
HES buffer (20 mM Hepes pH 7.4, 1 mM EDTA and
250 mMsucrose) on ice and scraped in HES buffer in the
presence of protease inhibitors (complete protease inhibitor
cocktail, Boehringer Mannheim) Samples were
homogen-ized by nine times three strokes in a glass potter
homo-genizer after which low density microsome (LDM) and
plasma membrane (PM) fractions were isolated by
differ-ential centrifugation as described by Simpson et al [31]
Equal amounts of protein as determined with BCA
protein assay reagent (Pierce) were subjected to immunoblot
analysis using various antibodies
Plasma membrane-lawn assay
The plasma membrane-lawn assay was performed as
described previously [32] Digital fluorescence imaging was
performed using a Leica DM-RXA epifluorescence
micro-scope (Leica, Germany) equipped with a 100-W mercury
lamp and the appropriate filters
Assay of 2-deoxyglucose uptake
3T3-L1 adipocytes, grown in 12-well plates (Costar), were
subjected to an assay of 2-deoxy-D-[14C]glucose (0.075 lCi
per well) uptake as described previously [33]
Immunoprecipitations and Western blotting Dishes (9 cm) of 3T3-L1 adipocytes were stimulated with agonists Immunoprecipitation and immunoblotting pro-cedures were as described previously [27] For c-Cbl immunoprecipitation 9-cm dishes of 3T3-L1 adipocytes were stimulated with agonists and scraped in lysis buffer (1 mM Na3VO4, 1 mM EGTA, 1 mM EDTA, 50 mM
Tris/HCl pH 7.4, 1% NP-40, 0.5% sodium deoxycholate,
150 mM NaCl, 5 mM NaF in the presence of protease inhibitors) Cell lysates were tumbled for half an hour at
4C, cell lysate was cleared from cellular debris by spinning
at 14 000 g, for 10 min at 4C in a table-top centrifuge About 1 mg of cell lysate was subjected to immunopreci-pitation using 5 lg of anti-Cbl mouse monoclonal 7G10 (UBI) for 1.5 h at 4C Immunocomplexes were harvested
by incubating with ProtGbeads for 1.5 h at 4C Beads were washed in lysis buffer and dissolved subsequently in sample buffer Phosphotyrosine was demonstrated by immunoblotting using anti-pY20 followed by anti-(mouse HRP) secondary Ig Immunoblots were quantified using
LUMIANALYST software on a LumiImager (Boehringer-Mannheim)
PI-3¢ kinase activity assay Dishes (9 cm) of 3T3-L1 adipocytes were stimulated with agonists and immunoprecipitated using time, concentration and antibodies as indicated in the figure legends Cells were lysed and IRS-1 and p85 immunoprecipitates collected on protein A–Sepharose beads were analysed for the copreci-pitation of in vitro PI-3¢ kinase activity using 5 lCi c-32 P-labelled ATP per reaction as described by Burgering et al [34] Incorporated radioactivity was quantified on a Molecular Dynamics phosphorimager
PKC-k kinase assay Dishes (9 cm) of 3T3-L1 adipocytes were stimulated with
100 nMof insulin for 10 min or 0.5 mMarsenite for 30 min Cells were lysed in an NP-40 based lysis buffer (see above) in the presence of 1 lM microcystin LR and immunopreci-pitated with 5 lg mouse monoclonal PKC-k for 1.5 h at
4C Subsequently, Prot-Gbeads were added and com-plexes were harvested after another 1.5 h The precipitate was washed three times with lysis buffer and two times with kinase assay buffer (100 mMHepes pH 7.4, 10 mMMgCl2,
1 mM dithiothreitol) PtdSer (4 lg per sample) was dried under N2(g)and dissolved in 25 lL kinase buffer per sample Subsequently, PtdSer was waterbath-sonicated three times for 5 min and 25 lL sample kinase buffer, ATP (40 lM, final concentration), 5 lCi c-32P-labelled ATP per reaction, dithiothreitol (1.5 mM), PKI (1 mM), the indicated concen-trations of Ro 31-8220, and PKCe-substrate (40 lM) were added As a control, 10 lM of a peptide identical to the PKC-k pseudosubstrate domain was added to determine the specificity of the assay Kinase reactions were allowed
to proceed for 10 min at 37C under gentle agitation Twenty microlitres of each reaction was spotted on p81 paper and washed three times for 5 min with 0.85% (v/v) phosphoric acid, and once for 5 min with acetone P81 papers were air dried and analysed in a scintillation counter
Trang 4Statistical analyses
Data were analysed with an independent-samples t-test
usingSPSS10.0 Curves represent fits to data by nonlinear
regression analysis usingGRAPHPAD PRISM2.01
Results
Arsenite induces 2-DOG uptake, GLUT1 and GLUT4
translocation in 3T3-L1 adipocytes
Incubation of 3T3-L1 adipocytes with arsenite stimulated
the uptake of hexose in a time- and dose-dependent manner
Stimulation of 2-deoxy-D[14C]glucose (2-DOG) uptake was
maximal at a 30-min preincubation period with 0.5 mM
arsenite (Fig 1A,B) On average, maximal stimulation of
glucose uptake was approximately sevenfold for insulin and
threefold for arsenite When insulin was added to adipocytes
during the final 15 minutes of a 30-min incubation with
arsenite, no significant additive response was seen on
arsenite-induced glucose uptake (Fig 1C) Under these
conditions, insulin did induce tyrosine phosphorylation of
IRb, IRS-1 and IRS-2, indicating that arsenite does not
interfere with the insulin-induced activation of this pathway
(data not shown)
Insulin-stimulated glucose transport predominantly
involves GLUT4 translocation from an intracellular
microsomal compartment to the plasma membrane of
adipocytes along with some induction of GLUT1
trans-location To determine whether arsenite stimulates
GLUT1 and GLUT4 translocation, we fractionated
adipocytes into membrane (PM) and microsomal vesicle
(LDM) fractions Equal amounts of protein ( 10 lg)
were subjected to immunoblot analysis (Fig 2A,C)
Plasma membrane fractions were identified using
antibod-ies against the IR b-chain and integrin-b1 (data not
shown) As can be seen in Fig 2A, when probing the
fractions with an antibody against GLUT4, both
insulin-and arsenite-treatment resulted in a shift of GLUT4 from
the microsomal fractions towards the plasma membrane
fractions Figure 2C shows that insulin induces some
translocation of GLUT1 towards the plasma membrane,
as did arsenite, albeit at lower levels than insulin The
amounts of GLUT1 and GLUT4 in each fraction were
quantified and expressed as a relative amount of total
GLUT protein in these fractions Arsenite significantly
increases the amount of GLUT4 in the PM (Fig 2B),
albeit at a lower level than insulin With respect to
GLUT1, although a consistent increase in the amount of
GLUT1 translocating towards the plasma membrane was
observed this did not reach significant levels compared to
basal levels of GLUT1 in the PM (Fig 2D) Furthermore,
arsenite did not change the total amount of GLUT4 or
GLUT1 in 3T3-L1 adipocytes (data not shown) It should
be noted that both GLUT1 and GLUT4 are heavily
glycosylated and show heterogeneous mobility
An alternative method to investigate GLUT protein
translocation is the plasma-lawn assay In this analysis,
sonicated cells are probed with an antibody recognizing
GLUT1 or GLUT4 and subjected to immunofluorescnce
microscopy As can be seen in Fig 2E, both GLUT1 and
GLUT4 are present in PM-lawns at higher quantities
after arsenite treatment as compared to the unstimulated situation
Combined, these data indicate that arsenite-stimulated glucose uptake involves translocation of the GLUT1 and the insulin-responsive GLUT4 glucose transporter
The effect of arsenite on early events in insulin receptor signalling
To elucidate the signalling-pathways that contribute to arsenite-induced GLUT4 translocation, we examined
Fig 1 Arsenite induces glucose uptake in 3T3-L1 adipocytes in a dose-and time-dependent manner (A) 3T3-L1 adipocytes were stimulated with the indicated concentrations of arsenite for 30 min and assayed for 2-deoxy- D [14C]glucose (2-DOG) uptake (B) 3T3-L1 adipocytes were stimulated with 0.5 m M arsenite for the indicated times and assayed for 2-DOGuptake (C) 3T3-L1 adipocytes were stimulated as indicated with 100 n M insulin for 15 min (ins), 0.5 m M arsenite for
30 min (as), or 0.5 m M arsenite for 30 min combined with 100 n M
insulin added after 15 min (as/ins) and assayed for 2-DOGuptake Incorporated radioactivity was determined by liquid scintillation counting Values are mean ± SEM of at least four determinations;
*P < 0.05 compared to basal and P < 0.05 for as/ins compared
to ins.
Trang 5whether arsenite activates intermediates of the insulin
signalling-pathway Following stimulation of 3T3-L1
adi-pocytes with either insulin or arsenite, IR, IRS-1 and IRS-2
were immunoprecipitated and assessed for tyrosine
phos-phorylation by Western blotting As shown in Fig 3A–C,
no significant increase in tyrosine phosphorylation of either
IRb, IRS-1 or IRS-2 could be detected in response to
arsenite Under these conditions, stimulation with insulin
led to a pronounced tyrosine phosphorylation of these
proteins In agreement with the lack of IRS-tyrosine
phosphorylation in response to arsenite, no association of
the p85 or p110a subunits of PI-3¢ kinase with the IRS
proteins was detected, again in contrast to the situation seen
after insulin stimulation (Fig 3B) Treatment with arsenite
did not lead to phosphorylation of PKB on either Ser473 or
Thr308, nor of phosphorylation of PKC-k onT403 This
observation agrees with the absence of PI-3¢ kinase
activa-tion by arsenite in vivo (Fig 3D) Consistent with these
observations, no increase of in vitro PI-3¢ kinase activity
was observed in IRS-1 immunoprecipitates after arsenite
treatment (Fig 4A), nor did we find any arsenite-induced
stimulation of in vitro PI-3¢ kinase activity in
immuno-precipitates of PI-3¢ kinase (Fig 4B)
To investigate the possibility that basal PI-3¢ kinase activity in combination with arsenite-induced signals is needed to stimulate glucose uptake, we examined the effect
of the PI-3¢ kinase inhibitor, wortmannin, on arsenite-induced glucose uptake Using concentrations that fully inhibited insulin-induced glucose uptake, arsenite-induced glucose uptake was unaffected by wortmannin (Fig 4C) Similar data were obtained using LY-294002 (data not shown) These observations suggest that arsenite-induced glucose uptake occurs without the need for PI-3¢ kinase activity, a situation that is in marked contrast to insulin induced glucose uptake Another early target of insulin action, the activation of ERK-1, -2 was also not activated appreciably in response to arsenite-treatment either (Fig 3D)
The effect of arsenite on Cbl and caveolin-1 tyrosine phosphorylation
A recently described PI-3¢ kinase independent pathway involved in insulin-induced glucose uptake in 3T3-L1 adipocytes involves Tyr phosphorylation of c-Cbl and caveolin-1 mediated by the IR Remarkably, we found
Fig 2 Arsenite-treatment induces GLUT1 and GLUT4 translocation to the plasma membrane in 3T3-L1 adipocytes 3T3-L1 adipocytes were mock-treated (basal), stimulated for 15 min with 100 n M insulin (insulin/ins) or for 30 min with 0.5 m M arsenite (arsenite/as) (A) Adipocytes were fractionated and equal amounts of protein from both microsomal (LDM) and plasma membranes (PM) were analysed by immunoblot with anti-GLUT4 Igs (B) anti-GLUT4 levels in each fraction subjected to immunoblot analysis as in A were quantified using a LumiImager and expressed as a fraction of GLUT4 residing in either LDM or PM (C) Subcellular fractions as in (A) were also subjected to immunoblot analysis using antibodies against GLUT1 (D) GLUT1 levels in each fraction were quantified as in (B) and expressed as a fraction of GLUT1 residing in either LDM or PM The total amount of GLUT1 or GLUT4 in the HDM fraction did not alter during either arsenite- or insulin-treatment Data are expressed as mean ± SEM of at least three independent observations, *P < 0.05 compared to basal (E) 3T3-L1 adipocytes were mock-treated (none), stimulated for 15 min with 100 n M insulin (ins) or for 30 min with 0.5 m M arsenite (as) Adipocytes were subjected to PM-lawn analysis using antibodies against either GLUT1 or GLUT4 Data shown are a representative example of five independent observations for each condition.
Trang 6that arsenite did induce tyrosine-phosphorylation of both
c-Cbl (Fig 5B) and caveolin-1 (Fig 5A), in spite of the
absence of IR activation This observation suggests that
arsenite does induce an as of yet unidentified tyrosine
kinase activity in 3T3-L1 adipocytes To evaluate whether
this tyrosine kinase activity is required for arsenite-induced
glucose uptake, we applied the tyrosine kinase inhibitor,
tyrphostin A25 Insulin-induced IRS and caveolin-1
tyro-sine phosphorylation are attenuated, but still present after
pretreatment with tyrphostin A25 (Fig 5A)
Insulin-induced Cbl phosphorylation is even enhanced by
tyr-phostin (Fig 5B) In contrast, arsenite-induced Cbl
and caveolin-1 phosphorylation are strongly inhibited
(Fig 5A,B) In a glucose uptake assay, tyrphostin A25
attenuated insulin-induced glucose uptake, but completely
blocked arsenite-induced glucose uptake (Fig 5C) These
data illustrate the distinct nature of the insulin- (namely
the IR) and arsenite-induced tyrosine kinases, and
fur-thermore they suggest that tyrosine kinase activity is a
requirement for arsenite-mediated induction of glucose
uptake in 3T3-L1 adipocytes
The effect of the p38 MAPK inhibitor, SB203580,
on arsenite-induced glucose uptake
SB203580 is a pharmacological inhibitor of the MAP kinase
family member p38 and has been shown to inhibit insulin
induced glucose uptake by 3T3-L1 adipocytes and L6
muscle cells [35,36] Arsenite-treatment induced
p38-phos-phorylation on Thr180 and Tyr182, and phosp38-phos-phorylation of
MAPKAP-K2 on Thr334 (a direct target site of p38 MAP kinase activity [37]) Treatment with 10 lM SB203580 significantly inhibited arsenite-induced glucose uptake
as well as MAPKAP-K2 and p38-phosphorylation (Fig 6A,B)
In insulin-signalling, p38 MAPK has been implied in enhancing the intrinsic activity of the GLUT4 glucose transporter Thus, SB203580 has been shown to reduce induced glucose uptake without an effect on insulin-induced GLUT4 translocation [35] SB203580 had a similar
Fig 4 Arsenite-induced in vitro PI-3¢ kinase activity 3T3-L1 adipo-cytes were stimulated as indicated with 100 n M insulin for 5 min or 0.5 m M arsenite for 30 min Cell lysates were incubated for 3 h with polyclonal antiserum against IRS-1 (A), or against the 85-kDa regu-latory subunit of PI-3¢ kinase (B) Immunoprecipitates were washed and subsequently subjected to an in vitro PI-3¢ kinase assay Copre-cipitating PI-3¢ kinase activity was determined on a phosphorimager as the relative stimulation of [c- 32 P]ATP incorporation into phosphati-dylinositol standardized against untreated cells Data are expressed as the mean ± SEM of three observations, statistically significant com-pared to basal (*P < 0.05) (C) 3T3-L1 adipocytes were pretreated for
15 min with 100 n M wortmannin Subsequently, adipocytes were mock-treated (basal) or stimulated as indicated with 100 n M insulin for
15 min or with 0.5 m M arsenite for 30 min in the continued presence of the pharmacological inhibitor and assayed for 2-DOGuptake Data are expressed as mean ± SEM of at least six observations Statistically significant data when compared to the samples not treated with wortmannin are indicated (*P < 0.05).
Fig 3 The effect of arsenite on the activation status of early steps in
insulin-responsive signal-transduction pathways 3T3-L1 adipocytes
were stimulated as indicated with 100 n M insulin for 5 min or 0.5 m M
arsenite for 30 min Cell lysates were immunoprecipitated with anti-IR
(A), anti-IRS-1 (B), anti-IRS-2 (C) followed by immunoblot analysis
with anti-phosphotyrosine (ap-Tyr), anti-PI-3¢ kinase regulatory
sub-unit (ap85) or anti-PI-3¢ kinase catalytic subsub-unit (ap110a) as indicated.
Equal loading was confirmed using the respective antibodies used for
immunoprecipitation (D) Total cell lysate of adipocytes (10 lg)
sti-mulated as described above was analysed by immunoblot using
phosphospecific antibodies against T202/Y204 of ERK-1/2 (apERK),
T308 and S473 of PKB (apThr308 and apSer473), T403 of PKC-k
(apPKCk) or PKB (aPKB) as indicated.
Trang 7effect on arsenite-induced glucose uptake (Fig 6; compare
A with C,D) i.e a reduction in glucose uptake without a
reduction in GLUT4 translocation
The effect of pharmacological inhibitors
of PKC-isoforms on arsenite-induced glucose uptake Atypical PKC isoforms (PKC lambda and zeta) have been implicated in insulin induced glucose uptake in adipocytes [11,12] We compared the effect of a number of pharmaco-logical inhibitors for various PKC isoforms on arsenite- and insulin-induced glucose uptake in 3T3-L1 adipocytes Ro
31-8220 inhibited insulin- and arsenite-induced glucose uptake with similar dose–response relations More precisely, an IC50 value of approximately 5 lMwas found for the inhibition
of insulin- and arsenite-induced glucose (Fig 7A) When PKC-k was purified by immunoprecipitation and subjected
to an in vitro kinase assay, a similar dose dependency for the inhibition of PKC-k (an IC50 of 5 lM) was observed (Fig 7B) Co-incubation with 50 lMof a peptide resembling the PKC-k pseudosubstrate domain reduced32P incorpor-ation by 90%, demonstrating the specificity of the assay In support of the data presented in Fig 3D, arsenite did not induce PKC-k activation over basal levels Hence, arsenite appears to require basal levels of PKC-k in conjunction with other signals to induce glucose uptake
In addition the PKC-inhibitors, chelerythrine chloride,
Ro 32-0432, bisindolylmaleimide I (BIM I) and Go¨ 6976 were studied at concentrations well above the IC50values for their respective conventional and novel PKC target proteins (Table 1) All inhibitors reduced TPA-induced ERK phosphorylation in 3T3-L1 adipocytes, demonstra-ting their functional interference with PKC (Fig 7C) Their effects on arsenite- and insulin-induced glucose uptake were minimal (Fig 7D) (although BIM I had a small but significant inhibitory effect), demonstrating that neither conventional nor novel PKC isoforms are involved in arsenite- or insulin-induced glucose uptake
The effect of myristoylated PKC-k/-f and PKC-a/-b pseudosubstrate peptides on arsenite-induced glucose uptake
To substantiate the observations made using Ro 31-8220 we investigated the effect of myristoylated peptide-inhibitors for PKC [11] on insulin- and arsenite-induced glucose uptake As can be seen in Fig 8B, a myristoylated peptide with a sequence similar to the pseudosubstrate domain of the atypical PKCs was capable of inhibiting insulin- as well
as arsenite-induced glucose uptake A peptide resembling the pseudosubstrate domain of conventional PKC-a/-b had
no significant inhibitory effect on either insulin- or arsenite induced glucose uptake (Fig 8B), whereas it did block TPA induced ERK phosphorylation demonstrating its functionality (Fig 8A) These observations corroborate the observations made with Ro 31-8220
Discussion
Insulin-induced glucose uptake by adipocytes is determined
by multiple factors, including: the translocation of glucose transporters from intracellular sites to the plasma mem-brane, expression levels of individual members of the glucose transporter family and by modulating the intrinsic activity (or, degree of occlusion) of glucose transporters (Fig 9)
Fig 5 The involvement of tyrosine kinase activity in arsenite-induced
glucose uptake 3T3-L1 adipocytes were pretreated with for 15 min
with 25 l M tyrphostin A25 as indicated Subsequently, adipocytes
were mock-treated (-), stimulated for 5 min with 100 n M insulin (INS)
or for 30 min with 0.5 m M arsenite (As) in the continued presence of
the pharmacological inhibitor (A) Total cell lysate (10 lg) was
sub-jected to immunoblot analysis using antibodies against phosphoY14 of
caveolin-1 (ap-Cav1), phosphotyrosine (ap-Tyr) (shown are the
IRS-bands at 180 kDa) and IRS-1 (aIRS) for equal loading (B) 3T3-L1
adipocytes treated as described above were immunoprecipitated using
antibodies against c-Cbl followed by immunoblot analysis using
antibodies against phosphotyrosine (ap-Tyr) or Cbl (ac-Cbl) (C)
3T3-L1 adipocytes were pretreated for 15 min with 25 l M tyrphostin A25.
Subsequently, adipocytes were mock-treated (basal), stimulated for
15 min with 100 n M insulin or for 30 min with 0.5 m M arsenite in the
continued presence of the pharmacological inhibitor and assayed for
2-DOGuptake Data are expressed as mean ± SEM of at least six
observations Statistically significant (*P < 0.05) when compared to
the samples not treated with tyrphostin Statistically significant
(P < 0.05) when compared to the basal (or arsenite) samples treated
with tyrphostin A25.
Trang 8Arsenite is a protein modifying agent known to react with
sulfhydryl groups in a discrete number of proteins As a
result these proteins are modified in their function and this
can couple back to altered activity of signal transduction
pathways
In this report we demonstrate that arsenite displays
insulin-mimicking effects in 3T3-L1 adipocytes: thus,
arse-nite stimulates 2-DOGuptake and induces translocation of
the insulin-responsive GLUT4 glucose transporter from the
low-density microsomal fraction towards the plasma
mem-brane Comparable to other stress-inducing agents [38,39]
arsenite acutely blocks insulin-induced glucose uptake
(Fig 1C), however, and in contrast to oxidative and
osmotic stress, arsenite did not interfere with early events
in insulin-induced signalling A normal level of
phosphory-lation of IRS-1,2 and PKB was observed by
insulin-induction after incubation with arsenite (data not shown)
Whereas PI-3¢ kinase activity is pivotal for
insulin-induced GLUT4 translocation, arsenite increases the uptake
of 2-DOGwithout the need for PI-3¢ kinase activity as
judged from the absence of an effect of arsenite on
PI-3¢ kinase activation and the lack of inhibition by either
wortmannin or LY 294002 Furthermore, arsenite does not
activate other signalling steps normally activated in response
to insulin, such as IR tyrosine kinase, IRS-1 and IRS-2
tyrosine phosphorylation, phosphorylation of PKC-k on T403 or phosphorylation of PKB on either Ser473 or Thr308 These observations suggest that the target of arsenite action resides downstream of PI-3¢ kinase or in a separate pathway
A PI-3¢ kinase-independent pathway in insulin-induced GLUT4 translocation has recently been identified and involves tyrosine phosphorylation on several residues of the proto-oncogene c-Cbl [13,40] and caveolin-1 [41] by the activated insulin receptor Arsenite also induces c-Cbl and caveolin-1 tyrosine phosphorylation in 3T3-L1 adipo-cytes, however, given that arsenite does not activate the insulin receptor (Fig 3A) the two processes are mechan-istically different The distinct nature of the insulin- and arsenite-induced tyrosine kinase activities is illustrated by the effects of tyrphostin A25 Whereas insulin-induced tyrosine kinase activity was attenuated (and Cbl-tyrosine phosphorylation levels even potentiated), all arsenite-induced tyrosine phosphorylation was strongly reduced The effects of tyrphostin A25 on insulin- and arsenite-induced glucose uptake mirrored these observations, i.e., insulin-induced glucose uptake was attenuated whereas arsenite-induced glucose uptake was lost These data also demonstrate that a tyrosine kinase activity is apparently required for the induction of glucose uptake by arsenite,
Fig 6 The effect of the p38 MAP kinase inhibitor, SB203580, on arsenite-induced glucose uptake and translocation of GLUT4 3T3-L1 adipocytes were pretreated for 30 min with 10 l M SB203580 (A–D) Subsequently adipocytes were mock-treated (basal) or stimulated as indicated with
100 n M insulin for 15 min (insulin) or with 0.5 m M arsenite for 30 min (arsenite) in the continued presence of the pharmacological inhibitor and assayed for 2-DOGuptake (A) Data are expressed as the mean value ± SEM of at least six observations Statistically significant (*P < 0.05) when compared to samples without SB203580 (B) 3T3-L1 adipocytes were lysed and subjected to immunoblot analysis using antibodies against p38 MAPK (p38), phospho-specific antibodies against p38 (p-p38) and MAPKAP-K2 (p-MAPKAP-K2) (C) 3T3-L1 adipocytes treated as described above were subjected to cell fractionation and the effect of SB203580 on GLUT4 translocation was determined by immunoblotting followed by quantification in a lumni-imager as described for Fig 2B Samples pretreated with SB203580 are indicated with SB Data are expressed as the mean ± SEM of three independent experiments (D) Representative immunoblot probed with anti-GLUT4 Igs, used to obtain the data described in C.
Trang 9as is the case for insulin The identity of the
tyrosine-kinase activity activated in response to arsenite remains to
be resolved
The pharmacological p38 MAPK inhibitor, SB203580,
affects insulin-induced glucose transport by affecting
intrinsic GLUT4 activity [35,36] Our results with this
inhibitor confirmed this observation Furthermore, we
demonstrate a similar effect on arsenite-induced glucose
uptake Pre-treatment with 10 lM SB203580 had no
effect on GLUT4 (or GLUT1) translocation but did
reduce arsenite-induced glucose uptake by 30% Thus,
our data show a similar contribution of p38-MAPK
activity in combination with GLUT4 translocation in
insulin- and arsenite-induced glucose uptake at the level
of modulating the GLUT4 mediated transport activity
Though we cannot fully exclude a similar effect of
SB203580 on GLUT1 as well, this seems unlikely given
that SB203850 had no effect whatsoever on arsenite- or
insulin-induced glucose uptake levels in 3T3-L1
preadipo-cytes (expressing GLUT1 and no GLUT4) (data not
shown)
The term coined for the modulatory effect of SB203850
is Ôintrinsic activityÕ [35], possibly by altering the speed of transition between ÔoutwardÕ and ÔinwardÕ conformations
of the transporter [42] The term ÔocclusionÕ has been used
to describe a state in which the GLUT4 transporter is fully inserted to the plasma membrane, but incapable of binding and/or transporting glucose yet [43] This could
be due to associating proteins blocking glucose transport, differences in LDM-derived and PM-membrane compo-sition and/or a conformational change in the GLUT4-transporter that is required for its activation If SB203580 hampers the progress through these stages, similar consequences are expected (i.e., GLUT4 being present
in the plasma membrane, but less glucose being taken up) Arsenite-induced glucose uptake, which demonstrates
a similar sensitivity to SB203580, may provide an additional tool for future research addressing these models
Cellular stresses like hypoxia [44] and hyperosmolarity [45] increase glucose uptake through an upregulation of the amount of GLUT1 Arsenite in contrast, does not increase
Fig 7 The effect of PKC-inhibitors on arsenite-induced glucose uptake 3T3-L1 adipocytes were incubated with the indicated concentrations of
Ro 31-8220 for 30 min prior to stimulation Subsequently, adipocytes were mock-treated (basal), stimulated with 100 n M insulin for 15 min (insulin) or 0.5 m M arsenite for 30 min (arsenite) in the continued presence of Ro 31-8220 (A) 2-DOGuptake was assayed and data are expressed
as mean ± SEM of two independent experiments each performed in triplicate (B) In vitro kinase assay performed in the continued presence of
Ro 31-8220 Incorporated counts (in k c.p.m.) are expressed as mean ± SEM of two independent experiments each performed in duplicate (C,D) 3T3-L1 adipocytes were pretreated for 30 min with 0.1 l M Go¨ 6976, 5 l M bisindolylmaleimide I (BIM I), 10 l M chelerythrine chloride (Chel-erythrine), or 10 l M Ro 32–0432 as indicated Subsequently, 3T3-L1 adipocytes were stimulated with 100 n M TPA for 15 min and analysed for ERK-1/2 phosphorylation (C) 2-DOGuptake was tested in a separate experiment (D) After pretreatment with the indicated pharmacological inhibitors, adipocytes were mock-treated (basal), stimulated with 100 n M insulin for 15 min (insulin), or 0.5 m M arsenite for 30 min (arsenite) in the continued presence of the inhibitors 2-DOGuptake was assayed and data are expressed as mean ± SEM of at least two independent experiments each performed in triplicate, statistically significant compared to the uninhibited samples (*P < 0.05).
Trang 10the amount of GLUT1 Indeed, treatment of the adipocytes
with the protein synthesis inhibitors cycloheximide or
emetine had no effect on arsenite-induced glucose uptake
(data not shown) Furthermore, the time-course of
arsenite-induced glucose uptake, being maximal after 30 min and
declining thereafter (Fig 1B) already seems to argue against
de novo synthesis of GLUT-transporters in mediating
arsenite-induced glucose uptake
Although most GLUT1 is already localized in the plasma
membrane of an unstimulated adipocyte [46], some GLUT1
is known to cotranslocate with GLUT4 [47] and Fig 2
C,D,E Moreover, treatment of 3T3-L1 adipocytes with
TPA induced the specific translocation of GLUT1 and not
GLUT4 towards the plasma membrane of 3T3-L1
adipo-cytes [48] Arsenite in contrast, induces the translocation of
GLUT1 at about half the levels obtained with insulin
Though failing to reach statistical significance (Fig 2D) this
effect was consistently reproducible (e.g Fig 2C,E) Thus,
clearly, arsenite differs from other types of cellular stress in
that it projects towards a more insulin-like response (i.e.,
translocation of GLUT1 and 4) albeit at a lower level of
efficiency
When applying multiple pharmacological- and
peptide-inhibitors for several PKC isoforms we observed a common
pattern of inhibition for insulin- and arsenite-induced
glucose uptake and similar concentration dependencies for the various agents Most notably, Ro 31-8220 inhibits both insulin- and arsenite-induced glucose uptake with an IC50of
5 lMsuggesting an involvement of atypical PKCs The IC50
of atypical PKC for BIM I is 5.8 lM, hence the significant reduction in glucose uptake measured using 5 lM(Fig 7D) fits with the observations made with Ro 31-8220 Inhibitors against conventional or novel PKCs such as Go¨ 6976, chelerythrine chloride and Ro 32-0432 (a compound struc-turally related to Ro 31-8220) (Table 1) remained without effect, or acted even in a slightly potentiating manner Furthermore, both insulin- and arsenite induced glucose uptake was inhibited by treatment with a myristoylated PKC-k/f pseudosubstrate peptide, but not significantly sensitive to treatment with a PKC-a/b pseudosubstrate A formal exclusion of other intracellular targets with similar sensitivities to the inhibitors mentioned cannot, however, be excluded
In the case of arsenite-induced glucose uptake, the myristoylated PKC-k/f pseudosubstrate peptide inhibited this response by approximately 50% (Fig 8B) Remark-ably, when the effect of Ro 31-8220 on arsenite-induced GLUT4 translocation was determined a similar reduction in GLUT4 translocation was observed (data not shown) This
is in contrast to the situation in response to insulin, where the inhibition is complete
Another observation was that in contrast to insulin, arsenite did not induce T-loop phosphorylation of PKC-k (Fig 3D), nor did we observe an increase in the amount of incorporated radiolabelled phosphate in immunoprecipi-tated PKC-k (data not shown) Indeed, when analysing PKC-k activity in an in vitro kinase assay, no induction of PKC-k activity in response to arsenite was observed (Fig 7B) Thus, taken together, these data suggest that arsenite does not activate PKC-k, but does require the basal
Fig 8 Arsenite-induced glucose uptake is inhibited by a myristoylated
PKC-k/-f, but not by PKC-a/-b pseudosubstrate peptide 3T3-L1
adipocytes were incubated with either myristoylated PKC-a/-b
pseudosubstrate (myrPKC-a/b ps), or myristoylated PKC-k/–f
pseudosubstrate (myrPKC-k/f ps) at the indicated concentrations for
1 h prior to stimulation (A) 3T3-L1 adipocytes were treated with
100 n M TPA for 15 min and subjected to immunoblotting as described
in the legend of Fig 7C (B) 2-DOGuptake was tested in a separate
experiment 3T3-L1 adipocytes were mock-treated (basal), stimulated
for 15 min with 100 n M insulin (insulin) or for 30 min with 0.5 m M
arsenite (arsenite) Data are expressed as mean ± SEM of at least two
independent experiments each performed in triplicate, statistically
significant compared to the uninhibited samples (*P < 0.05).
Fig 9 A model highlighting the insulin- and arsenite-induced pathways
to glucose uptake in 3T3-L1 adipocytes Our data suggest some com-mon steps in both arsenite- and insulin-induced glucose uptake: acti-vation of p38 MAPK and tyrosine-phosphorylation of c-Cbl In contrast to insulin, arsenite does not activate PI-3¢ kinase (and con-sequently does not activate PKC-k) However, the data suggests that basal levels of PKC-k activity are needed for arsenite-induced glucose uptake, as is indicated by the dashed arrow.