DIABETES, VOL. 49, FEBRUARY 2000 263 Potential Role of Glycogen Synthase Kinase-3 in Skeletal Muscle Insulin Resistance of Type 2 Diabetes Svetlana E. Nikoulina, Theodore P. Ciaraldi, Sunder Mudaliar, Pharis Mohideen, Leslie Carter, and Robert R. Henry Glycogen synthase ( GS) activity is reduced in skeletal muscle of type 2 diabetes, despite normal protein expression, consistent with altered GS regulation. Glycogen synthase kinase-3 (GSK-3) is involved in reg- ulation (phosphorylation and deactivation) of GS. To access the potential role of GSK-3 in insulin resistance and reduced GS activity in type 2 diabetes, the expres- sion and activity of GSK-3 were studied in biopsies of vastus lateralis from type 2 and nondiabetic subjects before and after 3-h hyperinsulinemic (300 mU · m – 2 · min – 1 )-euglycemic clamps. The specific activity of GSK-3 did not differ between nondiabetic and dia- betic muscle and was decreased similarly after 3-h insulin infusion. However, protein levels of both and isoforms of GSK-3 were elevated (~30%) in diabetic muscle compared with lean ( P < 0.01) and weight- matched obese nondiabetic subjects ( P < 0.05) and were unchanged by insulin infusion. Thus, both basal and insulin-stimulated total GSK-3 activities were ele- vated by approximately twofold in diabetic muscle. GSK-3 expression was related to in vivo insulin action, as GSK-3 protein was negatively correlated with maxi- mal insulin-stimulated glucose disposal rates. In sum- mary, GSK-3 protein levels and total activities are 1) elevated in type 2 diabetic muscle independent of obesity and 2) inversely correlated with both GS activ- ity and maximally insulin-stimulated glucose disposal. We conclude that increased GSK-3 expression in dia- betic muscle may contribute to the impaired GS activ- ity and skeletal muscle insulin resistance present in t ype 2 diabetes. Diabetes 4 9 :2 6 3–271, 2000 G lycogen synthesis represents a major pathway o f glucose disposal in skeletal muscle after insulin stimulation (1). The rate o f glycogen synthesis is impaired in type 2 diabetes (2). Several labora- tories have previously shown that the activity of glycogen synthase (GS), the rate-limiting enzyme for glycogen syn- thesis, is reduced in type 2 diabetes (3–5). This impairment in GS activity occurs in the presence of normal enzyme pro- tein levels in diabetic muscle (6), although there is evidence that GS mRNA expression may be lower in muscle from dia- betic subjects (7). GS is activated by insulin, mainly through dephosphoryla- tion of specific serine residues at sites designated 3a, b, and c, although dephosphorylation of additional sites may also play a role (8). Pho spho rylatio n at sites 3a, b, and c, with sub- sequent deactivation of the enzyme, is catalyzed by glycogen synthase kinase-3 ( GSK-3) (9–12) , whereas dephosphorylation is catalyzed by PP1g, the glycogen-bound form of protein phosphatase 1 (13,14). Previous studies of possible defects in the regulation of GS have focused on the role of the PP1g in diabetic muscle. The limited evidence suggests that PP1g activity is not impaired in diabetic skeletal muscle, even when GS activity (fractional velocity) is decreased (15). Recently, it has become clear that GSK-3 is also subject to acute regula- tion. Insulin has been demonstrated to cause inactivation of GSK-3 in vivo (16) and in several c ell types (17–20). This inac- tivation appears to be regulated by a protein kinase B, also known as Akt (17,21), which is a downstream target of phos- phatidylinositol (PI) 3-kinase (22). Blo ckage of insulin stimu- lation of PI 3-kinase also eliminates the activation of GS in human myo blasts (23). Studies in c ells over-expressing GSK-3 have shown that twofold activation of GSK-3 is sufficient to inhibit GS (12) and provided additional evidence supporting a physiolo gical role fo r GSK-3 in the regulation of GS. Two isoforms of human GSK-3, with molecular weights of 51 kDa ( ) and 46 kDa ( ), have been identified (24). The - and -isoforms share 85% homology at the amino acid level and are expressed ubiquitously, but functional differences between them are still unclear (25). Insulin administration results in phosphorylation of Ser-9 and Ser-21 in the - and -isoforms of GSK-3, respectively, which inhibits their activ- ity (26). In intac t cells, GSK-3 is also phospho rylated o n tyro - sine-216 (27); this modification activates GSK-3 (19,26,28). H o w ever, it is no t clear if tyro sine phosphorylation is respon- sive to insulin or if it mo dulates GSK-3 activity in vivo (17,19). The aim of the current study was to evaluate whether impairment of whole body glucose uptake or skeletal mus- cle GS activity in type 2 diabetes is related to alterations in the expression or activity of GSK-3. To determine this, we measured levels of protein expression of the two isoforms ( and ) of GSK-3 and investigated the effects of insulin infusion on GSK-3 phosphorylation and activity in skeletal muscle of lean and obese nondiabetic and obese type 2 dia- betic patients. Department of Medicine, University o f Califo rnia, San Diego, La Jolla, and Veterans Affairs Health Care System, San Diego, California. Address correspondence and reprint requests to R. R. Henry, VA Health Care System, San Diego (9111G), 3350 La Jolla Village Drive, San Diego, CA 92161. E-mail: rrhenry@ vapop.ucsd.edu. Received for publication 12 May 1999 and accepted in revised form 25 October 1999. DTT, dithiothreitol; FV, fractional velocity; GDR, glucose disposal rate; GS, glycogen synthase; GSK-3, glycogen synthase kinase-3; IRS-1, insulin receptor substrate-1; OD, optical density; PI, phosphatidylinositol; PMSF, phenylmethylsulfonyl fluoride; PP1g, glycogen-bound form of protein phosphatase 1; R g l , muscle-specific glycogen-associated phosphatase.