Intracellular signal transduction: The role of endosomes.

Polypeptide hormones, growth factors, and other biologically significant molecules are specifically internalized by target cells. Exposure of cells to these ligands results in the formation of ligand-receptor complexes on the cell surface and subsequent internalization of these complexes into the endosomal apparatus (endosomes, or ENs). The study of ENs has identified several important functions for this unique cellular organelle. These include the dissociation of ligand from receptor and receptor recycling to the cell surface and the degradation of some internalized ligands, as well as the delivery of others to lysosomes. More recently, it has become apparent that ENs fulfill another critical role, that of signal transduction. In this article, we review the evidence substantiating this role for ENs and propose three models by which ENs participate in cell signaling.

A role for tyrosine phosphorylation in both activation and inhibition of the insulin receptor tyrosine kinase in vivo.

Upon insulin binding, a conformational change in the insulin receptor (IR) leads to IR beta-subunit autophosphorylation, an increase in IR beta-subunit exogenous tyrosine kinase activity, and the rapid endocytosis of the ligand-receptor complex into endosomes. Previous work has shown that upon internalization, rat hepatic endosomal IRs manifest increased autophosphorylating and exogenous tyrosine kinase activity compared to IRs located at the plasma membrane. As this period of enhanced activity is associated with reduced endosomal IR beta-subunit phosphotyrosine content, it has been proposed that partial dephosphorylation of the internalized IR beta-subunit by an endosomally located phosphotyrosine phosphatase(s) [PTPase(s)] mediates this effect. To test whether endosomal PTPase activity was required for internalization-dependent augmentation of IR tyrosine kinase activity, the present study used the peroxovanadium PTPase inhibitor, bisperoxo(1,10-phenanthroline)oxovanadate anion [bpV(phen)], to block IR dephosphorylation within this subcellular compartment. Rats were pretreated with bpV(phen) before receiving insulin (1.5 micrograms/100 g BW). bpV(phen) inhibited the dephosphorylation of 32P-labeled hepatic endosomal IR by approximately 97% at 15 min post-bpV(phen) injection and prevented a decrease in IR beta-subunit phosphotyrosine content after IR internalization. Fifteen-minute bpV(phen) pretreatment produced a significant reduction (75%; P < 0.001) in maximal insulin-stimulated endosomal IR exogenous kinase activity and decreased IR autophosphorylating activity by 4.3-fold in this subcellular fraction. In conclusion, these findings suggest that an hepatic endosomal PTPase(s) regulates internalization-dependent increases in IR exogenous tyrosine kinase activity.

Ragaglitazar: the pharmacokinetics, pharmacodynamics, and tolerability of a novel dual PPAR alpha and gamma agonist in healthy subjects and patients with type 2 diabetes.

Ragaglitazar is a novel dual peroxisome proliferator-activated receptor (PPAR) alpha and gamma agonist intended to restore insulin sensitivity and correct diabetic dyslipidemia. These studies assessed single-dose pharmacokinetics and tolerability of ragaglitazar in healthy subjects, as well as multiple-dose pharmacokinetics, pharmacodynamics, and tolerability of ragaglitazar in healthy subjects and in patients with type 2 diabetes. Healthy subjects received a single oral dose (1-120 mg), and healthy subjects and type 2 diabetic patients received a loading dose and thereafter once-daily doses (0.5-16 mg) of ragaglitazar for 6 and 20 days, respectively. Ragaglitazar was rapidly absorbed (tmax: 1.5-1.7 h), with mean AUC0-24 h and Cmax proportional to dose after single and multiple dosing; t1/2 was 80 hours following a single dose and 104 hours in healthy subjects and 122 hours in patients after multiple dosing. Administration of 4 mg ragaglitazar to patients (n = 4) for 21 days resulted in mean decreases from baseline in fasting levels of plasma glucose (18%), C-peptide (18%), fructosamine (6%), triglycerides (36%), free fatty acids (49%), total cholesterol (11%), low-density lipoprotein (LDL) cholesterol (21%), and very low-density lipoprotein (VLDL) cholesterol (15%), as well as an increase in high-density lipoprotein (HDL) cholesterol (33%). Overall, ragaglitazar was well tolerated; with multiple dosing, there was a higher incidence of adverse events for patients that, at the highest dose level (16 mg), included peripheral edema and anemia.

Protein phosphatase type-1 mRNA levels in response to starvation-refeeding and streptozotocin-diabetes.

Levels of the mRNA encoding the catalytic subunit of protein phosphatase type-1 (PP-1cat) were reduced in skeletal muscle but not liver in response to short-term (2h) chow refeeding after prolonged (40h) starvation in the rat. This reduction did not appear to be mediated by insulin per se since streptozotocin-induced diabetes was associated with a reduction in PP-1cat levels in skeletal muscle. It is suggested that glucose levels may be one factor that modulates skeletal muscle PP-1cat mRNA levels. Despite the changes in PP-1cat mRNA levels in skeletal muscle, total protein phosphatase-1 catalytic activity was not altered by either chow refeeding or streptozotocin-diabetes. By contrast, although total hepatic PP-1cat mRNA levels were not altered in response to chow refeeding, there was a marked reduction in glycogen phosphorylase phosphatase activity in the cytosol but not in the glycogen/microsomal fraction.

Insulin receptor-associated protein tyrosine phosphatase(s): role in insulin action.

Protein tyrosine phosphatases (PTPs) play a critical role in regulating insulin action in part through dephosphorylation of the active (autophosphorylated) form of the insulin receptor (IRK) and attenuation of its tyrosine kinase activity. Following insulin binding the activated IRK is rapidly internalized into the endosomal apparatus, a major site at which the IRK is dephosphorylated in vivo. Studies in rat liver suggest a complex regulatory process whereby PTPs may act, via selective IRK tyrosine dephosphorylation, to modulate IRK activity in both a positive and negative manner. Use of peroxovanadium (pV) compounds, shown to be powerful PTP inhibitors, has been critical in delineating a close relationship between the IRK and its associated PTP(s) in vivo. Indeed the in vivo administration of pV compounds effected activation of IRK in parallel with an inhibition of IRK-associated PTP activity. This process was accompanied by a lowering of blood glucose levels in both normal and diabetic rats thus implicating the IRK-associated PTP(s) as a suitable target for defining a novel class of insulin mimetic agents. Identification of the physiologically relevant IRK-associated PTP(s) should facilitate the development of drugs suitable for managing diabetes mellitus.

Peroxovanadium compounds: biological actions and mechanism of insulin-mimesis.

When used alone, both vanadate and hydrogen peroxide (H2O2) are weakly insulin-mimetic, while in combination they are strongly synergistic due to the formation of aqueous peroxovanadium species pV(aq). Administration of these pV(aq) species leads to activation of the insulin receptor tyrosine kinase (IRK), autophosphorylation at tyrosine residues and inhibition of phosphotyrosine phosphatases (PTPs). We therefore undertook to synthesize a series of peroxovanadium (pV) compounds containing one or two peroxo anions, an oxo anion and an ancillary ligand in the inner co-ordination sphere of vanadium, whose properties and insulin-mimetic potencies could be assessed. These pV compounds were shown to be the most potent inhibitors of PTPs yet described. Their PTP inhibitory potency correlated with their capacity to stimulate IRK activity. Some pV compounds showed much greater potency as inhibitors of insulin receptor (IR) dephosphorylation than epidermal growth factor receptor (EGFR) dephosphorylation, implying relative specificity as PTP inhibitors. Replacement of vanadium with either molybdenum or tungsten resulted in equally potent inhibition of IR dephosphorylation. However IRK activation was reduced by greater than 80% suggesting that these compounds did not access intracellular PTPs. The insulin-like activity of these pV compounds were demonstrable in vivo. Intra venous (i.v.) administration of bpV(pic) and bpV(phen) resulted in the lowering of plasma glucose concentrations in normal rats in a dose dependent manner. The greater potency of bpV(pic) compared to bpV(phen) was explicable, in part, by the capacity of the former but not the latter to act on skeletal muscle as well as liver. Finally administration of bpV(phen) and insulin led to a synergism, where tyrosine phosphorylation of the IR beta-subunit increased by 20-fold and led to the appearance of four insulin-dependent in vivo substrates. The insulin-mimetic properties of the pV compounds raises the possibility for their use as insulin replacements in the management of diabetes mellitus.

Association of phosphatidylinositol 3-kinase with the insulin receptor: compartmentation in rat liver.

Phosphatidylinositol 3-kinase (PI 3-kinase) plays an important role in a variety of hormone and growth factor-mediated intracellular signaling cascades and has been implicated in the regulation of a number of metabolic effects of insulin, including glucose transport and glycogen synthase activation. In the present study we have examined 1) the association of PI 3-kinase with the insulin receptor kinase (IRK) in rat liver and 2) the subcellular distribution of PI 3-kinase-IRK interaction. Insulin treatment promoted a rapid and pronounced recruitment of PI 3-kinase to IRKs located at the plasma membrane, whereas no increase in association with endosomal IRKs was observed. In contrast to IRS-1-associated PI 3-kinase activity, association of PI 3-kinase with the plasma membrane IRK did not augment the specific activity of the lipid kinase. With use of the selective PI 3-kinase inhibitor wortmannin, our data suggest that the cell surface IRK beta-subunit is not a substrate for the serine kinase activity of PI 3-kinase. The functional significance for the insulin-stimulated selective recruitment of PI 3-kinase to cell surface IRKs remains to be elucidated.

Selective activation of the rat hepatic endosomal insulin receptor kinase. Role for the endosome in insulin signaling.

Insulin administration activates the insulin receptor kinase (IRK) in both plasma membrane (PM) and endosomes (ENs) raising the possibility of transmembrane signaling occurring in the endosomal compartment. Peroxovanadium compounds activate the IRK by inhibiting IR-associated phosphotyrosine phosphatase(s). Following the administration of the phosphotyrosine phosphatase inhibitor bisperoxo(1,10-phenanthroline)-oxovanadate (V) anion (bpV(phen)) activation of the hepatic IRK in ENs preceded that in PM by 5 min. When colchicine treatment preceded bpV(phen) administration IRK activation in ENs was unaffected but was totally abrogated in PM. Insulin receptor substrate-1 tyrosine phosphorylation followed the kinetics of IRK activation in ENs not PM and a hypoglycemic response similar to that achieved with a pharmacological dose of insulin ensued. These studies demonstrate that ENs constitute a site for IR-mediated signal transduction.

Insulin receptor internalization and signalling.

The insulin receptor kinase (IRK) is a tyrosine kinase whose activation, subsequent to insulin binding, is essential for insulin-signalling in target tissues. Insulin binding to its cell surface receptor is rapidly followed by internalization of insulin-IRK complexes into the endosomal apparatus (EN) of the cell. Internalization of insulin into target organs, especially liver, is implicated in effecting insulin clearance from the circulation. Internalization mediates IRK downregulation and hence attenuation of insulin sensitivity although most internalized IRKs readily recycle to the plasma membrane at physiological levels of insulin. A role for internalization in insulin signalling is indicated by the accumulation of activated IRKs in ENs. Furthermore, the maximal level of IRK activation has been shown to exceed that attained at the cell surface. Using an in vivo rat liver model in which endosomal IRKs are exclusively activated has revealed that IRKs at this intracellular locus are able by themselves to promote IRS-1 tyrosine phosphorylation and induce hypoglycemia. Furthermore, studies with isolated rat adipocytes reveal the EN to be the principle site of insulin-stimulated IRS-1 tyrosine phosphorylation and associated PI3K activation. Key steps in the termination of the insulin signal are also operative in ENs. Thus, an endosomal acidic insulinase has been identified which limits the extent of IRK activation. Furthermore, IRK dephosphorylation is effected in ENs by an intimately associated phosphotyrosine phosphatase(s) which, in rat liver, appears to regulate IRK activity in both a positive and negative fashion. Thus, insulin-mediated internalization of IRKs into ENs plays a crucial role in effecting and regulating signal transduction in addition to modulating the levels of circulating insulin and the cellular concentration of IRK in target tissues.

In vivo insulin mimetic effects of pV compounds: role for tissue targeting in determining potency.

Peroxovanadium (pV) compounds activate the insulin receptor kinase in hepatocytes and inhibit the dephosphorylation of insulin receptors in hepatic endosomes with highly correlated potencies (Posner, B. I., R. Faure, J. W. Burgess, A. P. Bevan, D. Lachance, G. Zhang-Sun, J. B. Ng, D. A. Hall, B. S. Lum, and A. Shaver J. Biol. Chem. 269: 4596-4604, 1994). After intravenous administration, K2[VO(O2)2(picolinato)].2H2O [bpV(pic)], VO(O2) (picolinato) (H2O)2 [mpV(pic)], K[VO(O2)2(picolinato)].3H2O [bpV(phen)], and K[VO(O2)2(4,7-dimethyl-1,10-phenanthroline)].1/2H2O [bpV(Me2phen)] produced 50% of their maximal hypoglycemic effect at doses of 0.04, 0.04, 0.32, and 0.65 mumol/100 g body wt, respectively. In contrast, their potencies as inhibitors of dephosphorylation were bpV(pic) = bpV(phen) > mpV(pic) = bpV(Me2phen). bpV(pic) stimulated [14C]glucose incorporation into rat diaphragm glycogen in vivo, and its effect was dose dependent, synergistic with insulin, and evident in other skeletal muscles. In contrast, bpV(phen) displayed no effect on glycogen synthesis in skeletal muscle. mpV(pic) stimulated and bpV(Me2phen) had no effect on glycogen synthesis in the diaphragm. bpV(pic) augmented rat diaphragm insulin receptor kinase 2.2-fold with a time-integrated response 70% that of insulin. In contrast, the effect of bpV(phen) was delayed and much reduced. Thus, the in vivo potencies of pV compounds reflect differing capacities to act on skeletal muscle. The ancillary ligand within the pV complex may target one tissue in preference to another.