Frick W, Bauer A, Bauer J, Wied S, Müller G
Hoechst Marion Roussel Deutschland GmbH, Chemical and Metabolic Diseases Research Frankfurt, Frankfurt am Main, Germany.
Biochemistry. 1998 Sep 22;37(38):13421-36. doi: 10.1021/bi9806201.
Phosphoinositolglycan (PIG) molecules have been implicated to stimulate glucose and lipid metabolism in insulin-sensitive cells and tissues in vitro and in vivo. The structural requirements for this partial insulin-mimetic activity remained unclear so far. For establishment of a first structure-activity relationship, a number of PIG compounds were synthesized consisting of the complete or shortened/mutated glycan moiety derived from the structure of the glycosylphosphatidylinositol (GPI) anchor of the GPI-anchored protein, Gce1p, from yeast. The PIG compounds were divided into four classes according to their insulin-mimetic activity in vitro with the typical representatives: compound 41, HO-SO2-O-6Manalpha1(Manalpha1-2)-2Manalpha1 (6-HSO3)- -6Manalpha1-4GluNb eta1-6(D)inositol-1,2-(cyclic)-phosphate; compound 37, HO-PO(H)O-6Manalpha1(Manalpha1-2)-2Manalpha1-6Manal pha1-4GluNbeta1-6( D)inositol-1,2-(cyclic)-phosphate; compound 7, HO-PO(H)O-6Manalpha1-4GluN(1-6(L)inositol-1,2-(cyclic)-ph osp hate; and compound 1, HO-PO(H)O-6Manalpha1-4GluN(1-6(L)inositol. Compounds 41 and 37 stimulated lipogenesis up to 90% (at 20 microM) of the maximal insulin response but with differing concentrations required for 50% activation (EC50 values 2.5 +/- 0.9 vs 4.9 +/- 1.7 microM) as well as glycogen synthase (4.7 +/- 1 vs 9.5 +/- 1.5 microM) and glycerol-3-phosphate acyltransferase (3.5 +/- 0.8 vs 8.0 +/- 1.1 microM). Compound 7 was clearly less potent (20% of the maximal insulin response at 100 microM), whereas compound 1 was almost inactive. This relative ranking in the insulin-mimetic potency between members of the PIG classes (e.g., 41 > 37 >> 7 > 1) was also observed for the (i) activation of glucose transport and glucose transporter isoform 4 translocation in isolated normal and insulin-resistant adipocytes, (ii) inhibition of lipolysis in adipocytes, (iii) stimulation of glucose transport and glycogen synthesis in isolated normal and insulin-resistant diaphragms, and (iv) induction of tyrosine phosphorylation of insulin receptor substrate-1 (IRS-1) in diaphragms. The complete glycan core structure (Man3-GluN) of typical GPI anchors including a mannose side chain and the inositolphosphate moiety was required for maximal insulin-mimetic activity of the PIG compounds with some variations possible with respect to the type of residues coupled to the terminal mannose/inositol as well as the type of linkages involved. These data argue for the potency and specificity of the interaction of PIG molecules with putative signaling component(s) (presumably at the level of the IRS proteins) in adipose and muscle cells which finally lead to insulin-mimetic metabolic activity even in insulin-resistant states.
磷酸肌醇聚糖(PIG)分子已被证实,在体外和体内均可刺激胰岛素敏感细胞和组织中的葡萄糖及脂质代谢。迄今为止,这种部分胰岛素模拟活性的结构要求仍不明确。为了建立首个构效关系,我们合成了许多PIG化合物,它们由来自酵母GPI锚定蛋白Gce1p的糖基磷脂酰肌醇(GPI)锚结构的完整或缩短/突变的聚糖部分组成。根据其体外胰岛素模拟活性,PIG化合物被分为四类,典型代表如下:化合物41,HO-SO2-O-6Manα1(Manα1-2)-2Manα1 (6-HSO3)- -6Manα1-4GluNβ1-6(D)肌醇-1,2-(环状)-磷酸酯;化合物37,HO-PO(H)O-6Manα1(Manα1-2)-2Manα1-6Manα1-4GluNβ1-6( D)肌醇-1,2-(环状)-磷酸酯;化合物7,HO-PO(H)O-6Manα1-4GluN(1-6(L)肌醇-1,2-(环状)-磷酸酯;以及化合物1,HO-PO(H)O-6Manα1-4GluN(1-6(L)肌醇。化合物41和37刺激脂肪生成的效果可达最大胰岛素反应的90%(在20 microM时),但50%激活所需的浓度不同(EC50值分别为2.5 +/- 0.9和4.9 +/- 1.7 microM),糖原合酶(分别为4.7 +/- 1和9.5 +/- 1.5 microM)以及甘油-3-磷酸酰基转移酶(分别为3.5 +/- 0.8和8.0 +/- 1.1 microM)也是如此。化合物7的效力明显较低(在100 microM时为最大胰岛素反应的20%),而化合物1几乎无活性。在(i)分离的正常和胰岛素抵抗脂肪细胞中葡萄糖转运和葡萄糖转运蛋白4亚型转位的激活、(ii)脂肪细胞中脂解的抑制、(iii)分离的正常和胰岛素抵抗膈肌中葡萄糖转运和糖原合成的刺激以及(iv)膈肌中胰岛素受体底物-1(IRS-1)酪氨酸磷酸化的诱导方面,也观察到了PIG类成员之间这种胰岛素模拟效力的相对排名(例如,41 > 37 >> 7 > 1)。典型GPI锚的完整聚糖核心结构(Man3-GluN),包括一个甘露糖侧链和肌醇磷酸部分,是PIG化合物最大胰岛素模拟活性所必需的,对于连接到末端甘露糖/肌醇的残基类型以及所涉及的连接类型可能存在一些变化。这些数据表明,PIG分子与假定的信号成分(可能在IRS蛋白水平)在脂肪和肌肉细胞中的相互作用具有效力和特异性,最终导致即使在胰岛素抵抗状态下也具有胰岛素模拟代谢活性。