AUTHOR COPY ONLY. Glycogen synthase kinase 3b mediates high glucose-induced ubiquitination and proteasome degradation of insulin receptor substrate 1

Transcription

1 Glyogen synthse kinse 3 medites high gluose-indued uiquitintion nd protesome degrdtion of insulin reeptor sustrte Snhu Leng, Wenshuo Zhng, Ynin Zheng, Ziv Liermn 1, Christopher J Rhodes, Hgit Eldr-Finkelmn 1 nd Xio Jin Sun The Setion of Endorinology, The University of Chigo, KCBD Room 8122, 9 Est 57th Street, Chigo, Illinois 6637, USA 1 The Deprtment of Humn Moleulr Genetis nd Biohemistry, Skler Shool of Mediine, Tel Aviv University, Tel Aviv 69978, Isrel (Correspondene should e ddressed to X J Sun; Emil: xsun@mediine.sd.uhigo.edu) (S Leng is now t New York College of Helth Professions, Syosset, New York 11791, USA) Astrt High gluose (HG) hs een shown to indue insulin resistne in oth type 1 nd type 2 dietes. However, the moleulr mehnism ehind this phenomenon is unknown. Insulin reeptor sustrte (IRS) proteins re the key signling moleules tht medite insulin s intrellulr tions. Geneti nd iologil studies hve shown tht redutions in nd/or IRS2 protein levels re ssoited with insulin resistne. In this study we hve shown tht protesome degrdtion of, ut not of IRS2, is involved in HG-indued insulin resistne in Chinese hmster ovry (CHO) ells s well s in primry heptoytes. To further investigte the moleulr mehnism y whih HG indues insulin resistne, we exmined vrious moleulr ndidtes with respet to their involvement in the redution in protein levels. In ontrst to the insulin-indued degrdtion of, HG-indued degrdtion of did not require IR signling or phosphtidylinositol 3-kinse/Akt tivity. We hve identified glyogen synthse kinse 3 (GSK3 or GSK3B s listed in the MGI Dtse) s kinse required for HG-indued serine 332 phosphoryltion, uiquitintion, nd degrdtion of. Overexpression of with muttion of serine 332 to lnine prtilly prevents HG-indued degrdtion. Furthermore, overexpression of onstitutively tive GSK3 ws suffiient to indue degrdtion. Our dt revel the moleulr mehnism of HG-indued insulin resistne, nd support the notion tht tivtion of GSK3 ontriutes to the indution of insulin resistne vi phosphoryltion of, triggering the uiquitintion nd degrdtion of. Introdution Insulin resistne is mjor pthophysiologil prolem in oth type 1 nd type 2 dietes mellitus, nd is key omponent of the metoli syndrome (DeFronzo et l. 1982, Fourlnos et l. 24). Insulin resistne is hrterized y progressive redution in the response of insulin-sensitive tissues to norml level of insulin (Khn 1994). The moleulr events tht medite insulin s tions in its trget tissues hve een well detiled in reent yers due to extensive geneti nd moleulr studies (Virkmki et l. 1999, Lee & White 24, Tniguhi et l. 26). In omprison, the moleulr sis of insulin resistne is onsiderly less ler. Insulin reeptor sustrte (IRS) proteins re mjor doking proteins tht medite insulin/insulin-like growth ftor 1 (IGF1) signling in vrious insulin-sensitive tissues (White 26)., like other IRS proteins, ontins mny tyrosyl phosphoryltion sites for IR tyrosine kinse (Sun et l. 1993). Tyrosine phosphoryltion of retes inding sites for proteins ontining the Sr homology-2 domin inluding phosphtidylinositol 3-kinse (PI3K), the GRB2/SOS omplex, SHP-2 (or PTPN11 s listed in the MGI Dtse), -fyn, nd NCK (Sun et l. 1993, White 26). These intertions led to the tivtion of multiple signling pthwys required for insulin s pleiotropi tions (White 26). In ddition, lso ontins multiple serine/ threonine phosphoryltion sites (Sun et l. 1991). It is sujet to phosphoryltion y wide rry of protein kinses tht re involved in diverse signling pthwys inluding those stimulted y insulin, inflmmtion, nutrients, nd ellulr stress (Sun & Liu 29). Mny serine/threonine kinses tht phosphorylte hve een identified, inluding glyogen synthse kinse 3 (GSK3 or GSK3B), p7 S6K, protein kinse Cq (PKCq), ERK, JNK, MTOR, nd AMPK mong others (Sun & Liu 29). Most serine/threonine phosphoryltion events ttenute insulin signling, nd re thought to indue insulin resistne. Studies in humns, nimls (Sd et l. 1992, Rondinone etl.1997),ndulturedells(leeetl.2,zhndeetl.22) DOI: /JOE /1/ q 21 Soiety for Endorinology Printed in Gret Britin Online version vi

2 172 S LENG nd others. Protesome degrdtion of hve reported n inverse orreltion etween protein levels nd insulin resistne. Mie defiient in re insulin resistnt nd growth retrded (Arki et l. 1994, Lee et l. 2); dietes develops when the loss of is omined with heterozygous null IR muttion (Bruning et l. 1997). Chroni insulin tretment, well-estlished method for induing insulin resistne (Del Prto et l. 1994, Koopmns et l. 1997), indues uiquitin-medited protesome degrdtion of, leding to ellulr insulin resistne (Sun et l. 1999, Lee et l. 2, Rui et l. 22, Zhnde et l. 22). Reently, n in vivo sirna-sed study demonstrted the essentil role of in the regultion of gluoneogeni gene expression in the liver, suh tht knokdown of Irs1 resulted in the upregultion of gluoneogeni genes (Tniguhi et l. 25). These studies emphsize the importne of ellulr levels of in insulin sensitivity. It hs een reported tht hroni hyperglyemi n indue insulin resistne in oth humns nd nimls (Vuorinen- Mrkkol et l. 1992, Buren et l. 23). In ontrst to tht indued y hyperinsulinemi, hyperglyemi-tlyzed resistne is less strightforwrd. It is unler whether high gluose (HG) inhiits insulin signling diretly through metolite effets, seondrily y stimultion of exessive insulin seretion, vi omintion of the two forementioned mehnisms, or through some yet undisovered mehnism. From ell-sed studies nd reserh in type 1 dietes, it n e onluded tht hyperglyemi ertinly does not require seondry hyperinsulinemi to ntgonize insulin funtion. However, even in insulin-free systems, the mehnism of hyperglyemi-indued insulin resistne is unler. Ativtion of PKC, degrdtion of IRS2, nd genertion of retive oxygen speies hve ll een proposed, with no onlusive model to dte (Berti et l. 1994, Ide et l. 1994, Pilly et l. 1996, Nkjim et l. 2). GSK3 is serine kinse negtively regulted y insulin tht ws originlly identified s key regultor of glyogen synthesis. Insulin tivtes AKT/protein kinse B through well-defined mehnism medited y the /PI3K pthwy. This leds to the phosphoryltion of GSK3 t serine 9, resulting in its intivtion (Cross et l. 1995, Shw et l. 1997, Frme & Cohen 21). In ddition to glyogen synthesis, GSK3 plys roles in mny other iologil funtions inluding Wnt signling (ruil in development) nd NFkB (or NFKB) signling ( key meditor of the inflmmtory response; Frme & Cohen 21, Lee & Kim 27). GSK3 is lso thought to e involved in insulin resistne. In type 2 dietes, GSK3 tivity is elevted twofold in oth sl nd insulin-stimulted musle (Nikoulin et l. 2). Additionlly, inhiition of GSK3 improves insulin s tion nd gluose metolism in humn smooth musle (Nikoulin et l. 22). GSK3 hs een shown to diretly phosphorylte in vitro nd in vivo t serine 332 nd impir insulin signling (Eldr-Finkelmn & Kres 1997, Liermn & Eldr-Finkelmn 25). Thus, GSK3 my ply ustive role in the development of insulin resistne. In this study, we hve investigted the moleulr mehnism of HG-indued insulin resistne in oth Chinese hmster ovry (CHO) ells nd primry mouse heptoytes. We report tht HG promotes protesome-medited degrdtion of. Unlike insulin, HG-indued degrdtion is independent of IR signling nd PI3K tivity. We hve identified GSK3 s key kinse tht medites HG-indued degrdtion of. Mterils nd Methods Cells nd trnsfetion CHO ells nd ells (CHO ells overexpressing the IR) were grown in F12 medium supplemented with 1% fetl ovine serum (FBS). They were ultured t 37 8C in humidified inutor mintined t 5% CO 2. ells overexpressing His-tgged ( His6 ), wild-type, with muttion of serine 332 to lnine, or with muttion of serine 332/336 to lnine were otined y trnsfetion of ells with pcmv/his/ His6, pcmv/his/, pcmv/ His/ S332A, or pcmv/his/ S332/336A onstruts respetively using lipofetmine (Invitrogen; Zhnde et l. 22, Liermn & Eldr-Finkelmn 25). Trnsfeted ells were seleted for resistne to 1 mm histidinol for 3 weeks efore eing used in the experiments. Adenovirl infetion For the expression of GSK3 vrints, reominnt denoviruses were generted nd used s previously desried (Finly et l. 24). In totl, 4% onfluent ells were inuted with 2 MOI of the following denoviruses: AdV-Flu (virus ontrol), AdV-GSK3-KD (kinse-defiient mutnt), or AdV-GSK3 S9A (onstitutively tive mutnt). Infetions were performed in F12 medium ontining 2% FBS for 2 h. The medium ws then removed, nd the ells were wshed twie with PBS. F12 medium ontining 1% FBS ws dded gin, nd the ells were llowed to reover for t lest 2 h. The denovirus-infeted ells were then used in the experiments s desried. Primry heptoyte isoltion nd ulture Primry mouse heptoytes (C57/BJ6) were isolted y modified nonirulting ollgense perfusion method s desried (Berry & Friend 1969, Zhnde et l. 26). Prolonged HG exposure For gluose exposure in CHO ells, the ells were ultured to 8 85% onflueny, nd were then inuted with low gluose (5. 5 mm or 1 g/l) or HG ( mm or 4. 5 g/l) DMEM ontining. % BSA for vrious time points. The medium ws refreshed every 24 h if inution ws done for longer

3 Protesome degrdtion of. S LENG nd others 173 thn 24 h. For gluose exposure in primry mouse heptoytes, the ells were ultured in low gluose (5. 5 mm) or HG ( mm) DMEM with 1% FBS. In some ses, insulin (5 or 1 nm) or inhiitors, inluding yloheximide (. 5mM), MG132 (mm), LY2942 (5 mm), LiCl (2 mm), rpmyin (2 nm), nd PD9859 (. 1 mm), were lso inluded in the medium. At the end of the inution, ells were left unstimulted or stimulted with 1 nm insulin for 1 min efore eing lysed in n pproprite uffer. His-tgged protein purifition ells overexpressing His6-tgged ( His6 ; Zhnde et l. 22) were wshed one with ie-old PBS ontining 1 mm N 3 VO 4 nd 1 mm phenylmethylsulfonyl fluoride (PMSF), nd lysed in 1 ml lysis uffer (5 mm NH 2 PO 4, ph 8., 3 mm NCl, 1 mm imidzole, 1% Tween 2, 1 mm PMSF, 1 mm N 3 PO 4, 5 mg/ml protinin, 5 mg/ml leupeptin, nd 5 mg/ml miroystin) for 1 min. Lystes were entrifuged t 13 g for 15 min, nd the superntnts were mixed with Ni-NTA eds (Qigen; 2 ml per 1 m plte of ells). The eds with His6 were wshed with wsh uffer (5 mm NH 2 PO 4,pH8., 3 mm NCl, 2 mm imidzole, 1% Tween 2, 1 mm PMSF, 1 mm N 3 PO 4, 5 mg/ml protinin, nd 5 mg/ml leupeptin) three times, nd were then dentured in Lemmli smple uffer ontining. 1M dithiothreitol (DTT) nd oiled for 5 min. Whole ell lyste preprtion CHO ells, ells, or primry mouse heptoytes were lysed diretly in Lemmli smple uffer ontining. 1 M DTT, sonited, nd oiled for 5 min. Immunolotting nlysis Smples from Ni-NTA purifition or whole ell lystes were seprted y SDS-7. 5% or 1% PAGE, nd trnsferred onto nitroellulose memrnes. Memrnes were riefly wshed with tris uffered sline (TBS; 2 mm Tris HCl, ph 8., nd. 15 M NCl) nd loked (5% skim milk in TBST (TBS ontining. 5% Tween 2)) for 2 h t room temperture or overnight t 4 8C, followed y inution with the orresponding ntiodies in the pproprite loking solution (5% skim milk TBSTor 5% BSA TBST): PY (PY2), IR (C-19) (Snt Cruz Biotehnology, Snt Cruz, CA, USA), PS 9 -GSK3-P44/42 ERK1/2, ERK1/2, PT 24 FoxO1, GSK3 (Cell Signling Tehnology, In.), uiquitin (FK2, Biomol Interntionl, LP, Plymouth Meeting, PA, USA),, IRS2, nd PS 332. Memrnes were then wshed three times with TBST nd proed with HRP-onjugted protein A (Snt Cruz) for 3 min, or HRP-onjugted rt nti-mouse ( Jkson ImmunoReserh Lortories, In., West Grove, PA, USA) or HRP-onjugted nti-rit ntiody ( Jkson ImmunoReserh Lortories, In.) for 1 h. After inution with seondry ntiodies, memrnes were wshed three times with TBSTnd one with TBS. Speifi proteins were visulized using ECL (GE Helthre, Pittsurg, PA, USA; Zhnde et l. 22). Results HG indues degrdtion of, ut not of IRS2 We hve shown previously tht prolonged insulin tretment n indue degrdtion of, ut not of IRS2, in CHO ells overexpressing the humn IR (; Sun et l. 1999, Zhnde et l. 22). To investigte if HG ould led to the sme effet, ells were exposed to norml gluose (5. 5 mm) or HG ( mm) for vrious lengths of time. Redutions in protein levels were oserved fter 72 h of exposure (Fig. 1A, lnes d nd f versus lne e). This ws not due to hnges in ell numer sine nonspeifi nds () nd GSK3 nds were onstnt t ll time points. To ensure tht this effet ws not limited to ells, we performed similr experiments using primry mouse heptoytes. We otined omprle dt, lthough the redution (5%) in protein levels ws oserved fter 22 h of exposure to HG (Fig. 1B, pnel 1). Interestingly, this effet ws speifi to the moleule, sine HG exposure hd no effet on IRS2 protein levels under the sme onditions (Fig. 1B, pnel 2). Intrellulr protein levels re determined y the lne etween protein synthesis nd degrdtion. To evlute the ontriution of impired synthesis versus enhned degrdtion to HG-indued redution of protein levels, ells were oinuted with HG nd yloheximide, protein trnsltion inhiitor, for 14 h. Without yloheximide, the redution in protein levels y HG ws not detetle fter 14 h of exposure (Fig. 1C, lne versus lne ). When yloheximide ws present, signifint redution in protein levels y HG ws oserved (Fig. 1C, lne d versus lne ). This suggests tht the effet of HG on is not medited y suppression of gene trnsription or mrna trnsltion, ut rther tht HG ts y ugmenting protein degrdtion. Indeed, MG132, reltively speifi protesome inhiitor, signifintly loked HG-indued degrdtion of (Fig. 1C, lnes e f versus lnes d), implying the involvement of the uiquitinmedited protesome degrdtion pthwy. To test if the redution in protein levels ould impt insulin signling, we exmined insulin-stimulted tyrosyl phosphoryltion of IRS proteins nd downstrem signling events in primry heptoytes. HG impired insulin signling, inluding insulin-indued tyrosine phosphoryltion of IRS proteins (3%), threonine 24 phosphoryltion of FoxO1 (2%), serine 9 phosphoryltion of GSK3 threonine 22 /tyrosine 24 phosphoryltion of ERK (35%), nd serine 473 phosphoryltion of AKT (65%; Fig. 2, lne versus lne ).

4 174 S LENG nd others. Protesome degrdtion of A 72 h 48 h 24 h h Gluose 5 5 mm Gluose mm B h C MG132 (/GSK3) GSK d e f ells 14 h d e f P < 1 d e f ells IRS2 (IRS/) P > 5 P < Heptoytes / IRS2/ Figure 1 Effet of high gluose on protein levels. (A) ells were inuted in DMEM. % BSA ontining 5. 5 mm (solid lines) or mm gluose (dshed lines) for the indited time periods. The medium ws refreshed every 24 h. (B) Primry mouse heptoytes were inuted in DMEM 1% FBS ontining 5. 5 or mm gluose for 22 h. (C) ells were inuted in DMEM. % BSA ontining 5. 5 or mm gluose in the presene or sene of yloheximide (. 5 mm) nd MG132 ( mm) for 14 h. After tretment(s), ll ells were lysed in Lemmli smple uffer ontining. 1 M DTT. nd IRS2 proteins in the ell lystes were deteted y immunolotting nlysis using nti- nd nti-irs2 ntiodies respetively s desried in Mterils nd methods. Results re representtive of t lest three seprte experiments. The densities for western lot nds were quntified y Imge J softwre. The rtio of density for nd GSK3 proteins ws lulted for eh lne, nd the vlue for norml (5. 5 mm) ws onsidered s 1 nd ompred with other tretments for eh individul experiment. Dt re expressed s mensgs.e.m., nz3. GSK3 is required for HG-indued degrdtion To investigte the signling pthwy tht medites HG-indued degrdtion of, we used severl speifi inhiitors for vrious serine/threonine kinses tht re involved in the modultion of insulin sensitivity. These inlude LiCl (GSK3), rpmyin (MTOR), PD9859 (MEK), nd LY2942 (PI3K). Among these tested inhiitors, LiCl ws the only inhiitor tht ws ple of ompletely loking HG-indued degrdtion of (Fig. 3A, pnel 1, lne d versus lnes, f, nd h; Fig. 4C, pnels 1 nd 3, lne versus lne d), suggesting the involvement of GSK3 in HG-indued degrdtion of. Sine phosphoryltion of GSK3 t serine 9 intivtes the kinse, we next heked the phosphoryltion level of GSK3. Indeed, HG exposure led to signifint redution in GSK3 serine 9 phosphoryltion (tivtion; Fig. 3A, pnel 2, lne versus lne ), nd LiCl enhned the phosphoryltion of GSK3 (intivtion; Fig. 3A, pnel 2, lnes d versus lnes ). The effet of LiCl on GSK3 phosphoryltion nd HG-indued degrdtion of, ut not of IRS2, ws lso oserved in primry heptoytes (Fig. 3B, pnels 1 3, lnes d versus lnes ). PD9859, rpmyin, nd LY2942 were unle to lok the redution of GSK3 phosphoryltion or degrdtion of indued y HG (Fig. 3A, pnels 1 2, lnes e h versus lnes ; Fig. 4C, pnels 1 4, lnes versus lnes d). These dt strongly suggest speifi role for GSK3 in the medition HG-indued degrdtion of. It is well known tht mny kinse inhiitors re not s speifi s ommerilly dvertised (Bin et l. 27). To verify if tivtion of GSK3 is suffiient to indue degrdtion, we overexpressed GSK3 kinse-defiient

5 IRS2 PY-IRS PT 24 -FOXO1 PS 9 - P-ERK42/44 ERK44 PS-AKT473 AKT (P-/) Protesome degrdtion of. S LENG nd others 175 Insulin P < 5 P < 5 5 (P-ERKs/ERK44) 3 1 (PY/) (PT-FOXO1/) (PS-AKT/AKT) mutnt s well s onstitutively tive GSK3 (serine 9 mutted to lnine) y denovirl infetion of ells. Overexpression of the onstitutively tive GSK3 mutnt signifintly redued protein levels (Fig. 3C, lne versus lne ), wheres overexpression of the kinse-defiient mutnt slightly inresed the level of protein (Fig. 3C, lne versus lne ). Sine high levels of endogenous GSK3 re present in ells, the kinse-defiient mutnt presumly served s dominnt negtive mutnt to lok endogenous GSK3 tivity towrds. This result demonstrtes tht tivtion of GSK3 is suffiient to prompt the degrdtion of. HG nd insulin indue degrdtion of through distint signling pthwys Insulin-indued protesome degrdtion depends on the tivity of the IR tyrosine kinse nd PI3K (Sun et l. 1999, Zhng et l. 2). We first tested whether the IR ws lso required for HG-indued degrdtion y ompring ells to prentl CHO ells. Although ells expressed signifint quntities of IR when ompred with CHO ells, omprle degree of degrdtion of ws oserved in oth ell lines fter 72 h of HG exposure (Fig. 4A, lnes d nd versus lnes nd ). This pttern ws mintined even when protein synthesis ws loked with yloheximide (Fig. 4B, lnes d nd h versus lnes nd g). Thus, IR is not required for HG-indued degrdtion of P < 5 P < 5 P < 5 Figure 2 Effet of high gluose on insulin signling. Primry mouse heptoytes were inuted in DMEM 1% FBS ontining 5. 5or mm gluose for 22 h. Cells were hllenged with insulin (1 nm) for 1 min efore eing lysed in Lemmli smple uffer ontining. 1 M DTT. Signl proteins in the ell lystes were deteted y immunolotting nlysis using the orresponding ntiodies. Quntifition of western lot ws done in the sme wy s desried in Fig. 1. Dt re expressed s mensgs.e.m., nz3. Next, we tested whether PI3K tivity ws required for HG-indued degrdtion of y stimulting ells with insulin, or loking PI3K tivity with LY2942, potent PI3K inhiitor. CHO nd ells were inuted with or without insulin (1 nm) in the presene of norml gluose (5. 5 mm) or HG ( mm) for 16 h in the presene of yloheximide (Fig. 4C). As expeted, IR ws required for insulin-indued degrdtion under norml gluose onditions, whih ws evidened y the oservtion tht insulin indued degrdtion only in ells, ut not in CHO ells (Fig. 4C, pnels 1 nd 3, lne e versus lne ). LY2942 ompletely loked insulin-indued degrdtion of (Fig. 4C, pnel 3, lne g versus lnes e nd ), suggesting tht PI3K medites the insulin-indued degrdtion of. In ontrst, HG lone redued in oth CHO nd ell lines even when PI3K tivity ws loked (Fig. 4C, pnels 1 nd 3, lnes nd d versus lnes nd ), inditing tht IR nd PI3K re not required for HG-indued degrdtion of. The ft tht HG nd insulin imposed opposite effets on GSK3 tivity (HG tivtes GSK3 while insulin inhiits GSK3; Fig. 4C, pnels 2 nd 4) prompted us to exmine whether insulin signling hd ny effet on HG-indued degrdtion of. In ells, lthough 1 nm insulin indued the inhiition of GSK3 (inresed phosphoryltion; Fig. 4C, pnel 4, lnes e f), it indued the degrdtion of under oth norml nd HG onditions (Fig. 4C, pnel 3, lnes e f), suggesting mehnism independent of GSK3 tivity. However, loking PI3K tivity y LY2942 only prevented insulin-indued degrdtion under norml gluose onditions, ut not under HG onditions (Fig. 4C, pnel 3, lnes g h versus lnes e f), whih is onsistent with the ft tht HG-indued degrdtion of is independent of PI3K tivity. Thus, HG nd insulin indue degrdtion of vi distint pthwys. Although 1 nm insulin filed to prevent the HG-indued degrdtion of in ells (sensitive to insulin), it did prevent HG-indued degrdtion to modest extent in CHO ells (less sensitive to insulin; Fig. 4C, pnels 1 nd 3, lnes e f versus lnes d). There is possiility for the effet of insulin on HG-indued degrdtion of eing minimized y the insulin-indued degrdtion of during overnight exposure of ells to high-dose insulin (1 nm). To investigte if short-term exposure to low dose of insulin ould revel ny effet of insulin on HG-indued tivtion of GSK3 nd degrdtion of, ells were treted with norml nd HG in onjuntion with yloheximide in the presene or sene of low dose of insulin (5 nm) for vrious time periods. Fivennomolr insulin in norml gluose indued degrdtion of t slow rte. In omprison, HG indued degrdtion t muh fster rte (Fig 4D, pnel 1, lnes d versus lnes e g, nd Fig 4E). Interestingly, HG-indued degrdtion of ws gretly redued in the presene of 5 nm insulin (Fig. 4D, lnes h j versus lnes d nd e g; Fig. 4E). The dynmi hnges in HG-indued degrdtion

6 176 S LENG nd others. Protesome degrdtion of A PD Rpmyin + LiCl PS 9 - B IRS2 PS 9 - None LiCl C -CA denovirus -DN denovirus Control denovirus d e f g h (IRS/) P > 5 P > 5 P < 5 P < Heptoytes d / IRS2/ Figure 3 Involvement of GSK3 in HG-indued degrdtion. (A) ells were inuted in DMEM. % BSA with yloheximide (. 5 mm) nd 5. 5 or mm gluose ontining vrious inhiitors: LiCl (2 mm), rpmyin (2 nm), or PD9859 (. 1 mm) for 16 h. (B) Primry mouse heptoytes were inuted in DMEM 1% FBS with 5. 5ormM gluose for 22 h. In lnes d, LiCl (2 mm) ws dded to the medium nd inuted for the lst 8 h. (C) ells were infeted with lnk denovirus or denovirus ontining the kinse-defiient GSK3 mutnt or the onstitutively tive GSK3 mutnt (2 MOI) for 2 h. After tretment(s), ll ells were lysed in Lemmli smple uffer ontining. 1 M DTT. Lystes were proessed for immunolotting nlysis s desried in Mterils nd Methods. Results re representtive of t lest two seprte experiments. Quntifition of western lots (B) ws done in the sme wy s desried in Fig. 1. Dt re expressed s mens GS.E.M., nz3. of y insulin were onsistent with hnges in GSK3 tivity (Fig. 4D, pnel 2), further supporting the involvement of GSK3 in HG-indued degrdtion of. HG promotes the phosphoryltion of serine 332 in GSK3 hs een shown to negtively regulte insulin signling y phosphorylting t serine 332 fter phosphoryltion of the priming site serine 336 y PKCII (Eldr-Finkelmn & Kres 1997, Liermn & Eldr-Finkelmn 25, Liermn et l. 28). To investigte if is phosphorylted t this speifi site during HG exposure, we used phospho-speifi ntiody ginst the site to monitor the speifi phosphoryltion of. Indeed, serine 332 phosphoryltion ws inresed in HG-treted ells, nd ws ompletely loked y LiCl (Fig. 5A, lnes nd versus lne ), suggesting tht GSK3 is the potentil kinse tht medites the phosphoryltion of y HG. To further investigte the role of serine 332 in the HG-indued degrdtion of, two nonphosphoryltle serine/lnine mutnts (Ser/ Al 332 nd Ser/Al 332/336 ) were overexpressed in ells (Liermn & Eldr-Finkelmn 25), s well s in wildtype ontrol. Trnsfeted ells were sujeted to HG tretment nd exmined for degrdtion. Both ser/l mutnts prevented HG-indued degrdtion of to signifint extent (Fig. 5B, lnes d nd f versus lne ). Considering the high rtio of endogenous versus overexpressed proteins, the degree to whih ws proteted from degrdtion y the mutnts is potentilly greter thn tht suggested y the immunolots. Thus, serine 332 phosphoryltion in y GSK3 is required for nd ppers to e the priniple site with respet to HG-indued degrdtion of. GSK3 tivity is required for HG-indued uiquitintion of As shown in Fig. 1C, HG-indued degrdtion ws loked y MG132, 26S protese inhiitor, inditing the involvement of uiquitin-medited protesome degrdtion. To onfirm the uiquitintion of during HG exposure, uiquitintion of ws mesured in the presene of MG132 (to preserve the uiquitinted ). Indeed, HG inresed uiquitintion (Fig. 6, lne versus lne ). Consistent with our dt, LiCl inhiited GSK3 tivity (Fig. 3A), redued phosphoryltion of t serine 332 (Fig. 5A), nd redued uiquitintion (Fig. 6, lne versus lnes nd ). These dt support the onlusion tht HG-indued uiquitintion nd degrdtion of re medited y GSK3.

7 Protesome degrdtion of. S LENG nd others 177 A CHO B CHO IRβ IRβ C LY2942 PS 9 - None Insulin d d e f CHO g h PS 9 - d e f g h D Gluose Insulin (5 nm) Time (h) PS mm (NG) mm (HG) d e f g h i j E / () HG Ins/NG Ins/HG Time (hours) Figure 4 Independene of insulin signling nd PI3K tivity for HG-indued degrdtion of. (A) CHO nd ells were inuted in DMEM. % BSA ontining 5. 5or mm gluose for 72 h. (B) CHO nd ells were inuted in DMEM. % BSA ontining 5. 5 or mm gluose in the presene or sene of yloheximide (. 5 mm) for 16 h. (C) CHO nd ells were inuted in DMEM. % BSA ontining yloheximide (. 5 mm) nd 5. 5 or mm gluose in the presene or sene of LY2942 (5 mm) nd/or insulin (1 nm) for 16 h. (D) ells were inuted in DMEM. % BSA ontining yloheximide (. 5 mm) nd 5. 5 or mm gluose in the presene or sene of insulin (5 nm) for 6, 12, or 24 h. After tretment(s), ll ells were lysed in Lemmli smple uffer ontining. 1 M DTT. Lystes were proessed for immunolotting nlysis s desried in Mterils nd methods. Results re representtive of t lest two seprte experiments. (E) Densities of nd GSK3 in (D) were quntified y Imge J softwre. The rtio of the :GSK3 density ws plotted ginst time. NG, norml gluose (5. 5 mm) nd HG, high gluose ( mm). Disussion is one of the mjor IRSs in the insulin signling pthwy, nd is required for norml insulin signling (Lee & White 24, Tniguhi et l. 26). protein levels re ontrolled y mny ftors inluding insulin/igf1, gluoortioids, oxidtive stress, nd ertin nutrients (free ftty ids, mino ids, nd gluose). The uiquitin-medited protesome degrdtion of is n importnt ellulr ontrol mehnism for insulin sensitivity (Sun et l. 1999, Rui et l. 22, Zhnde et l. 22). Here, we hve shown tht long-term HG exposure (mimiking hroni hyperglyemi) lso indues uiquitin-medited protesome degrdtion of. Interestingly, HG-indued degrdtion of is medited y mehnism distint from tht of insulin, in tht it is IR independent nd does not require PI3K tivity. More importntly, we hve shown tht the tivtion of GSK3 is required for HG-indued degrdtion of. HG tivtes GSK3, whih phosphoryltes t serine 332, leding to the uiquitintion nd protesome-medited degrdtion of. Thus, our study, for the first time, provides diret evidene tht phosphoryltion of is linked to its uiquitin-medited protesome degrdtion, nd revels moleulr mehnism for HG-indued degrdtion of. We nd others hve reported previously tht insulin/igf1 n indue degrdtion through uiquitin-medited protesome degrdtion pthwy (Sun et l. 1999, Lee et l. 2, Rui et l. 22, Zhnde et l. 22). In this study, we hve lerly shown tht lthough oth insulin nd HG tivte the uiquitin-medited protesome pthwy to

8 178 S LENG nd others A LiCl PS B pcmv/ IRβ (PS-/IRβ) 1 5. Protesome degrdtion of WT 5 5 / S332A 5 5 P < 1 d e degrde, there exist t lest two distint upstrem signling hins tht ultimtely led to uiquitintion of (Fig. 7). Contrry to the insulin-indued degrdtion of, whih requires oth the IR tyrosine kinse nd PI3K tivities (Sun et l. 1999, Lee et l. 2, Zhng et l. 2, Rui et l. 22), HG-indued protesome degrdtion is n IR- nd PI3K-independent proess. This onlusion is supported y the ft tht overexpression of the IR nd/or inhiition of PI3K did not signifintly ffet HG-inited degrdtion (Fig. 4). Although high onentrtions of insulin nd HG individully led to the degrdtion of, insulin t low onentrtions, for short periods of time, seems to inhiit HG-indued degrdtion of. This n e lerly explined y our finding tht insulin nd HG hve opposing effets on GSK3 phosphoryltion. Insulin, through d S332A/S336A P < Figure 5 Serine 332 phosphoryltion of in HG-indued degrdtion. (A) / ells were inuted in DMEM. % BSA ontining yloheximide (. 5 mm) nd MG132 ( mm) nd 5. 5 or mm gluose in the presene or sene of LiCl (2 mm) for 24 h. (B) ells overexpressing wild-type, S332A, or S332/336A mutnts were inuted in DMEM. % BSA ontining yloheximide (. 5 mm) nd 5. 5or mm gluose for 16 h. After tretment(s), ll ells were lysed in Lemmli smple uffer ontining. 1 M DTT. S332 phosphoryltion of, fprotein, nd the -suunit of insulin reeptor in the lystes were deteted y immunolotting nlysis using the orresponding ntiodies s desried in Mterils nd Methods. Results re representtive of t lest three seprte experiments. Quntifition of western lot (B) ws done in the sme wy s desried in Fig. 1. Dt re expressed s mensgs.e.m., nz3. e f f tivtion of PI3K/Akt, phosphoryltes serine 9 on GSK3, inhiiting kinse tivity. In ontrst, HG tivtes GSK3 through unknown mehnism y reduing serine 9 phosphoryltion. Sine insulin-indued degrdtion of is GSK3-independent proess (LiCl nnot lok insulinindued degrdtion, dt not shown), long-term insulin exposure even t lower onentrtions tht initilly ounter HG s effets on GSK3 would eventully trigger insulin-indued protesome degrdtion (Fig. 4D, 6- nd 12-h versus 24-h insulin o-tretment with HG). The entrl finding in this study is the identifition of GSK3 s the key omponent in HG-indued degrdtion of. There re few onfliting reports tht desrie the effet of HG on GSK3 tivity. In ner ells, inresing gluose uptke nd high rtes of gluose utiliztion indued the phosphoryltion of GSK3/ (inhiition) in PKCmedited proess, leding to ell survivl (Zho et l. 27). In glomerulr mesngil ells, HG-stimulted GSK3 tivity ugmented poptosis y ntgonizing Wnt signling, n effet preventle y LiCl tretment (Lin et l. 26). GSK3 tivity ws found to e higher in STZ-treted hyperglyemi rts due to deresed AKT tivity (Gurusmy et l. 26). HG hs lso een shown to downregulte oth sl nd insulin-stimulted glyogen synthse in rt-1 firolsts seondry to the stimultion of GSK3 (Singh & Crook 2). Our urrent study lerly shows tht HG redues GSK3 serine 9 phosphoryltion without hnging GSK3 protein levels. More importntly, we hve demonstrted in severl experiments tht GSK3 is neessry nd suffiient for the medition of HG-indued degrdtion of. First, n inhiitor of GSK3 loks HG-indued protesome degrdtion of ; seondly, HG redues GSK3 phosphoryltion t serine 9, whih is inditive of tivtion; LiCl Poly uiquitinted His6 His6 MG / His6 Figure 6 Uiquitintion of indued y HG. / His6 ells were inuted in DMEM. % BSA ontining MG132 ( mm) nd 5. 5 or mm gluose in the presene or sene of LiCl (2 mm) for 24 h. After tretments, His6-tgged ( His6 ) ws purified using Ni-NTA eds nd deteted y immunolotting with uiquitin ntiody (FK2) nd ntiody respetively. Results re representtive of two seprte experiments.

9 LiCl Protesome degrdtion of. S LENG nd others 179 High gluose PS332 Insulin nd finlly, overexpression of onstitutively tive GSK3 inreses degrdtion, wheres overexpression of kinse-defiient GSK3 slightly loks sl degrdtion. The mehnism y whih GSK3 medites degrdtion is unknown. Our results provide possile mehnism explining how GSK3 medites the uiquitintion of. Serine/threonine phosphoryltion is known to e n importnt signl for the uiquitintion nd protesome degrdtion of ellulr proteins (Nsh et l. 21). GSK3 phosphoryltes t serine 332 nd ttenutes insulin s tion (Liermn & Eldr-Finkelmn 25). Our results lerly show tht inhiition of GSK3 not only loks the protesome degrdtion of, ut lso prevents serine 332 phosphoryltion nd its susequent uiquitintion. Furthermore, muttion of serine 332 to lnine prevents degrdtion indued y HG. Tken together, the dt support model where HG tivtes GSK3 promoting serine 332 phosphoryltion of. This phosphoryltion tgs for uiquitintion, leding to protesome-medited destrution. Gluosmine, gluose metolite, hs lso een shown to indue insulin resistne in different ells nd nimls (DeFronzo et l. 1982, Rossetti et l. 1995, Ptti 1999, Hert et l. 2). We were unle to detet the effet of gluosmine on uiquitintion nd protesome degrdtion in our ell systems (dt not shown), inditing tht gluose nd gluosmine might indue insulin resistne vi different mehnisms s suggested in the literture (DeFronzo et l. 1982, Nelson et l. 2, 22, Hn et l. 23, Dun et l. 25). Extly how gluose tivtes GSK3 remins to e reveled in future study. IR PI3K U MG132 LY2942 Degrdtion Figure 7 Model of HG- nd insulin-indued protesome degrdtion of. Insulin-indued degrdtion of is medited y PI3K-dependent pthwy; wheres high gluoseindued degrdtion of depends on GSK3 tivity. Although oth high gluose nd insulin indue the degrdtion of under ertin onditions (low insulin onentrtion nd short tretment times), insulin n lok the high gluose-indued degrdtion of through the inhiition of GSK3 tivity. Muh evidene in the literture suggests tht GSK3 is involved in insulin resistne. GSK3 is onstitutively tive in unstimulted ells, nd is intivted y vriety of extrellulr stimuli, notly insulin (Woodgett 199). Elevted GSK3 tivity nd expression hve een oserved in oese nd dieti humns nd rodents (Eldr-Finkelmn et l. 1999, Nikoulin et l. 2, Cirldi et l. 22) nd in the dipose tissue of high-ft-fed mie (Eldr-Finkelmn et l. 1999). Overexpression of humn GSK3 (fivefold) in the skeletl musle of mie results in hyperglyemi, hyperinsulinemi, hyperlipidemi, redution in glyogen levels, nd signifint derese in protein levels (Pere et l. 24). Homozygous GSK3 knokout is emryonilly lethl, suggesting tht it plys n importnt role in erly development (Hoeflih et l. 2). However, GSK3 knokout nimls re vile, nd demonstrte enhned gluose nd insulin sensitivity (MAuly et l. 27). Insulin-stimulted Akt nd GSK3 phosphoryltion is higher in GSK3 KO livers thn in the wild-type littermtes. Interestingly, hepti expression is mrkedly inresed in these nimls (MAuly et l. 27). Although our experiments did not test GSK3, it is very likely tht GSK3 is lso involved in HG-indued degrdtion. Thus, our studies re onsistent with these reports, nd further provide the moleulr mehnism through whih GSK3 is involved in insulin resistne. Delrtion of interest The uthors delre tht there is no onflit of interest tht ould e pereived s prejudiing the imprtility of the reserh reported. Funding This work ws supported y grnt from NIH Grnt R1 DK6128 to XJS. Referenes Arki E, Lipes MA, Ptti ME, Bruning JC, Hg B III, Johnson RS & Khn CR 1994 Alterntive pthwy of insulin signlling in mie with trgeted disruption of the IRS-1 gene. Nture Bin J, Plter L, Elliott M, Shpiro N, Hstie CJ, MLuhln H, Kleverni I, Arthur JS, Alessi DR & Cohen P 27 The seletivity of protein kinse inhiitors: further updte. Biohemil Journl Berry MN & Friend DS 1969 High-yield preprtion of isolted rt liver prenhyml ells: iohemil nd fine struturl study. Journl of Cell Biology Berti L, Mosthf L, Kroder G, Kellerer M, Tippmer S, Mushk J, Seffer E, Seedorf K & Hring H 1994 Gluose-indued trnslotion of protein kinse C isoforms in rt-1 firolsts is prlleled y inhiition of the insulin reeptor tyrosine kinse. Journl of Biologil Chemistry Bruning JC, Winny J, Bonner-Weir S, Tylor SI, Aili D & Khn CR 1997 Development of novel polygeni model of NIDDM in mie heterozygous for IR nd IRS-1 null lleles. Cell Buren J, Liu HX, Luritz J & Eriksson JW 23 High gluose nd insulin in omintion use insulin reeptor sustrte-1 nd -2 depletion nd protein

10 18 S LENG nd others. Protesome degrdtion of kinse B desensitistion in primry ultured rt dipoytes: possile implitions for insulin resistne in type 2 dietes. Europen Journl of Endorinology Cirldi TP, Nikoulin SE & Henry RR 22 Role of glyogen synthse kinse-3 in skeletl musle insulin resistne in type 2 dietes. Journl of Dietes nd its Complitions Cross DA, Alessi DR, Cohen P, Andjelkovih M & Hemmings BA 1995 Inhiition of glyogen synthse kinse-3 y insulin medited y protein kinse B. Nture DeFronzo RA, Hendler R & Simonson D 1982 Insulin resistne is prominent feture of insulin-dependent dietes. Dietes Del Prto S, Leonetti F, Simonson DC, Sheehn P, Mtsud M & DeFronzo RA 1994 Effet of sustined physiologi hyperinsulinemi nd hyperglyemi on insulin seretion nd insulin sensitivity in mn. Dietologi Dun W, Pk L & Pillrisetti S 25 Distint effets of gluose nd gluosmine on vsulr endothelil nd smooth musle ells: evidene for protetive role for gluosmine in theroslerosis. Crdiovsulr Dietology Eldr-Finkelmn H & Kres EG 1997 Phosphoryltion of insulin reeptor sustrte 1 y glyogen synthse kinse 3 impirs insulin tion. PNAS Eldr-Finkelmn H, Shreyer SA, Shinohr MM, LeBoeuf RC & Kres EG 1999 Inresed glyogen synthse kinse-3 tivity in dietes- nd oesityprone C57BL/6J mie. Dietes Finly D, Ptel S, Dikson LM, Shpiro N, Mrquez R, Rhodes CJ & Sutherlnd C 24 Glyogen synthse kinse-3 regultes IGFBP-1 gene trnsription through the thymine-rih insulin response element. BMC Moleulr Biology Fourlnos S, Nrendrn P, Byrnes GB, Colmn PG & Hrrison LC 24 Insulin resistne is risk ftor for progression to type 1 dietes. Dietologi Frme S & Cohen P 21 GSK3 tkes entre stge more thn 2 yers fter its disovery. Biohemil Journl Gurusmy N, Wtne K, M M, Prksh P, Hiryshi K, Zhng S, Muslin AJ, Kodm M & Aizw Y 26 Glyogen synthse kinse 3 together with protein regultes dieti rdiomyopthy: effet of losrtn nd tempol. FEBS Letters Hn DH, Chen MM & Holloszy JO 23 Gluosmine nd gluose indue insulin resistne y different mehnisms in rt skeletl musle. Amerin Journl of Physiology. Endorinology nd Metolism 285 E1267 E1272. Hert E, Choi WS & Sung CK 2 Gluosmine-indued insulin resistne in 3T3-L1 dipoytes. Amerin Journl of Physiology. Endorinology nd Metolism 278 E13 E112. Hoeflih KP, Luo J, Ruie EA, Tso MS, Jin O & Woodgett JR 2 Requirement for glyogen synthse kinse-3 in ell survivl nd NF-kB tivtion. Nture Ide R, Megw H, Kikkw R, Shiget Y & Kshiwgi A 1994 High gluose ondition tivtes protein tyrosine phosphtses nd detivtes insulin reeptor funtion in insulin-sensitive rt 1 firolsts. Biohemil nd Biophysil Reserh Communitions Khn CR 1994 Insulin tion, dietogenes, nd the use of type II dietes (Bnting Leture). Dietes Koopmns SJ, Ohmn L, Hywood JR, Mndrino LJ & DeFronzo RA 1997 Seven dys of euglyemi hyperinsulinemi indues insulin resistne for gluose metolism ut not hypertension, elevted teholmine levels, or inresed sodium retention in onsious norml rts. Dietes Lee J & Kim MS 27 The role of GSK3 in gluose homeostsis nd the development of insulin resistne. Dietes Reserh nd Clinil Prtie 77 S49 S57. Lee YH & White MF 24 Insulin reeptor sustrte proteins nd dietes. Arhives of Phrmologil Reserh Lee AV, Gooh JL, Oesterreih S, Guler RL & Yee D 2 Insulin-like growth ftor I-indued degrdtion of insulin reeptor sustrte 1 is medited y the 26S protesome nd loked y phosphtidylinositol 3 -kinse inhiition. Moleulr nd Cellulr Biology Liermn Z & Eldr-Finkelmn H 25 Serine 332 phosphoryltion of insulin reeptor sustrte-1 y glyogen synthse kinse-3 ttenutes insulin signling. Journl of Biologil Chemistry Liermn Z, Plotkin B, Tennenum T & Eldr-Finkelmn H 28 Coordinted phosphoryltion of insulin reeptor sustrte-1 y glyogen synthse kinse-3 nd protein kinse CII in the dieti ft tissue. Amerin Journl of Physiology. Endorinology nd Metolism 294 E1169 E1177. Lin CL, Wng JY, Hung YT, Kuo YH, Surendrn K & Wng FS 26 Wnt/et-tenin signling modultes survivl of high gluose-stressed mesngil ells. Journl of the Amerin Soiety of Nephrology MAuly K, Dole BW, Ptel S, Hnsoti T, Sinlir EM, Druker DJ, Ngy A & Woodgett JR 27 Glyogen synthse kinse 3-speifi regultion of murine hepti glyogen metolism. Cell Metolism Nkjim K, Ymuhi K, Shigemtsu S, Ikeo S, Komtsu M, Aizw T & Hshizume K 2 Seletive ttenution of metoli rnh of insulin reeptor down-signling y high gluose in heptom ell line, HepG2 ells. Journl of Biologil Chemistry Nsh P, Tng X, Orliky S, Chen Q, Gertler FB, Mendenhll MD, Siheri F, Pwson T & Tyers M 21 Multisite phosphoryltion of CDK inhiitor sets threshold for the onset of DNA replition. Nture Nelson BA, Roinson KA & Buse MG 2 High gluose nd gluosmine indue insulin resistne vi different mehnisms in 3T3-L1 dipoytes. Dietes Nelson BA, Roinson KA & Buse MG 22 Defetive Akt tivtion is ssoited with gluose- ut not gluosmine-indued insulin resistne. Amerin Journl of Physiology. Endorinology nd Metolism 282 E497 E56. Nikoulin SE, Cirldi TP, Mudlir S, Mohideen P, Crter L & Henry RR 2 Potentil role of glyogen synthse kinse-3 in skeletl musle insulin resistne of type 2 dietes. Dietes Nikoulin SE, Cirldi TP, Mudlir S, Crter L, Johnson K & Henry RR 22 Inhiition of glyogen synthse kinse 3 improves insulin tion nd gluose metolism in humn skeletl musle. Dietes Ptti ME 1999 Nutrient modultion of ellulr insulin tion. Annls of the New York Ademy of Sienes Pere NJ, Arh JR, Clphm JC, Coghln MP, Cororn SL, Lister CA, Llno A, Moore GB, Murphy GJ, Smith SA et l. 24 Development of gluose intolerne in mle trnsgeni mie overexpressing humn glyogen synthse kinse-3 on musle-speifi promoter. Metolism Pilly TS, Xio S & Olefsky JM 1996 Gluose-indued phosphoryltion of the insulin reeptor. Funtionl effets nd hrteriztion of phosphoryltion sites. Journl of Clinil Investigtion Rondinone CM, Wng LM, Lonnroth P, Wesslu C, Piere JH & Smith U 1997 Insulin reeptor sustrte (IRS) 1 is redued nd IRS-2 is the min doking protein for phosphtidylinositol 3-kinse in dipoytes from sujets with non-insulin-dependent dietes mellitus. PNAS Rossetti L, Hwkins M, Chen W, Gindi J & Brzili N 1995 In vivo gluosmine infusion indues insulin resistne in normoglyemi ut not in hyperglyemi onsious rts. Journl of Clinil Investigtion Rui L, Yun M, Frntz D, Shoelson S & White MF 22 SOCS-1 nd SOCS-3 lok insulin signling y uiquitin-medited degrdtion of nd IRS2. Journl of Biologil Chemistry Sd MJ, Arki E, Mirlpeix M, Rothenerg PL, White MF & Khn CR 1992 Regultion of insulin reeptor sustrte-1 in liver nd musle of niml models of insulin resistne. Journl of Clinil Investigtion Shw M, Cohen P & Alessi DR 1997 Further evidene tht the inhiition of glyogen synthse kinse-3 y IGF-1 is medited y PDK1/PKB-indued phosphoryltion of Ser-9 nd not y dephosphoryltion of Tyr-216. FEBS Letters Singh LP & Crook ED 2 The effets of gluose nd the hexosmine iosynthesis pthwy on glyogen synthse kinse-3 nd other protein kinses tht regulte glyogen synthse tivity. Journl of Investigtive Mediine Sun XJ & Liu F 29 Phosphoryltion of IRS proteins Yin-Yng regultion of insulin signling. Vitmins nd Hormones

11 Protesome degrdtion of. S LENG nd others 181 Sun XJ, Rothenerg PL, Khn CR, Bker JM, Arki E, Wilden PA, Chill DA, Goldstein BJ & White MF 1991 The struture of the insulin reeptor sustrte IRS-1 defines unique signl trnsdution protein. Nture Sun XJ, Crimmins DL, Myers MG Jr, Mirlpeix M & White MF 1993 Pleiotropi insulin signls re engged y multisite phosphoryltion of IRS-1. Moleulr nd Cellulr Biology Sun XJ, Golderg JL, Qio LY & Mithell JJ 1999 Insulin-indued insulin reeptor sustrte-1 degrdtion is medited y the protesome degrdtion pthwy. Dietes Tniguhi CM, Ueki K & Khn R 25 Complementry roles of IRS-1 nd IRS-2 in the hepti regultion of metolism. Journl of Clinil Investigtion Tniguhi CM, Emnuelli B & Khn CR 26 Critil nodes in signlling pthwys: insights into insulin tion. Nture Reviews. Moleulr Cell Biology Virkmki A, Ueki K & Khn CR 1999 Protein protein intertion in insulin signling nd the moleulr mehnisms of insulin resistne. Journl of Clinil Investigtion Vuorinen-Mrkkol H, Koivisto VA & Yki-Jrvinen H 1992 Mehnisms of hyperglyemi-indued insulin resistne in whole ody nd skeletl musle of type I dieti ptients. Dietes White MF 26 Regulting insulin signling nd -ell funtion through IRS proteins. Cndin Journl of Physiology nd Phrmology Woodgett JR 199 Moleulr loning nd expression of glyogen synthse kinse-3/ftor A. EMBO Journl Zhnde R, Mithell JJ, Wu J & Sun XJ 22 Moleulr mehnism of insulinindued degrdtion of insulin reeptor sustrte 1. Moleulr nd Cellulr Biology Zhnde R, Zhng W, Zheng Y, Pendleton E, Li Y, Polkiewiz RD & Sun XJ 26 Dephosphoryltion y defult, potentil mehnism for regultion of insulin reeptor sustrte-1/2, Akt, nd ERK1/2. Journl of Biologil Chemistry Zhng H, Hoff H & Sell C 2 Insulin-like growth ftor I-medited degrdtion of insulin reeptor sustrte-1 is inhiited y epiderml growth ftor in prostte epithelil ells. Journl of Biologil Chemistry Zho Y, Altmn BJ, Coloff JL, Hermn CE, Jos SR, Wiemn HL, Wofford JA, Dimsio LN, Ilkyev O, Kelekr A et l. 27 Glyogen synthse kinse 3 nd 3 medite gluose-sensitive ntipoptoti signling pthwy to stilize Ml-1. Moleulr nd Cellulr Biology Reeived in finl form 5 My 21 Aepted 12 My 21 Mde ville online s n Aepted Preprint 12 My 21