PI-3 kinase-independent pathways for stimulation of glucose transport
The CAP/Cbl/TC10 pathway
Substantial evidence confirms the presence of PI-3 kinase independent pathways for stimulation of glucose transport. It had long been clear that other growth factor receptors (i.e., the platelet-derived growth factor (PDGF) receptor, cytokine receptors such as IL-4, and certain integrins) activate PI-3 kinase to the same extent as the insulin receptor with generation of PI(3,4,5)P3, yet still do not stimulate glucose transport [103,104]. Since stimulation of transport was blocked by PI-3 kinase inhibitors (e.g. wortmannin), investigators concluded that PI-3 kinase is necessary but not sufficient for this key metabolic effect of insulin . The additional factor that endows the insulin receptor with the specific capacity to stimulate glucose transport could include another parallel signaling pathway, compartmentalization of the PI-3 kinase response, or the utilization of specific signal protein isoforms in insulin target cells.
Multiple observations support a second parallel complementary pathway for insulin-stimulated glucose transport [105,106]. For example, adenovirus-mediated hyperexpression of dominantly interfering PH and PTB domains blocks insulin receptor/IRS interaction and inhibits some responses involving PI-3 kinase, such as membrane ruffling and DNA synthesis, without any effect on GLUT4 translocation. Similarly, blockade of PI-3 kinase activation, via prolonged pretreatment with PDGF or overexpression of GSK3β, inhibits association of PI-3 kinase with IRS proteins without affecting stimulation of glucose transport or GLUT4 translocation. In addition, specific mutations in the insulin receptor have been identified that exhibit a near normal ability to activate PI-3 kinase but completely fail to induce GLUT4 translocation . Finally, introduction of a cell permeable analogue of PI(3,4,5)P3 does not have any effect to stimulate basal glucose transport, but in the presence of wortmannin plus insulin does elicit a maximal glucose transport response . These data indicate that insulin signaling requires a separate wortmannin-insensitive pathway, which must operate in parallel with the PI-3 kinase pathway, for stimulation of glucose transport.
Evidence has accumulated strongly supporting the CAP/ Cbl/TC10 pathway as the complementary pathway for insulin- stimulated glucose transport [105,106], as illustrated in Figure 12.4. This pathway was elucidated as a linear sequence of interacting factors using technologies such as the yeast two-hybrid system and co-immunoprecipitation of interacting factors. The pathway diverges from the PI3-kinase pathway at the level of the insulin receptor kinase, which mediates tyrosine phosphorylation of Cbl proto-oncogene through a process that does not involve IRSs. c-Cbl phosphorylation does require another adapter protein, APS, which binds the insulin receptor via SH2 domain-phosphotyrosine interaction and couples the receptor to Cbl . c-Cbl is also complexed with another adapter protein, the c-Cbl associating protein (CAP), which helps facilitate recruitment of CAP/Cbl to the microdomain of the insulin receptor . Tyrosine phosphorylation of Cbl results in disengagement of the CAP/Cbl complex from the insulin receptor.
CAP contains three tandem SH3 domains in the COOH- terminus that binds to a proline-rich domain of Cbl, and, in the NH2-terminal region, also contains a sorbin homology (SoHo) domain. After disengagement of CAP/Cbl from the insulin receptor, the CAP SoHo domain binds to flotillin in caveolin-containing lipid raft domains of the plasma membrane . Lipid rafts are plasma membrane domains, enriched in cholesterol, glycolipids, and sphingolipids, which coordinate signaling events by accumulating specific protein constituents. One class of lipid rafts contains caveolin proteins in invaginated structures referred to as caveolae (“little caves”). Flotillin is anchored in lipid raft domains via its interaction with caveolin proteins. Thus, CAP is responsible for relocation of the CAP/Cbl complex to lipid raft domains. Once bound to flotillin, tyrosine phosphorylated Cbl presents a recognition site for recruitment of the CrkII-C3G complex to the lipid raft. CrkII is a small adapter protein that contains an SH2 domain that binds phosphorylated Cbl, and SH3 domains that interact with a proline-rich region of C3G. C3G is a guanyl-nucleotide exchange factor for TC10 and other small molecular weight GTP binding proteins . TC10 is a member of the Rho family of GTPases, and is also targeted to lipid raft domains as a result of its capacity to undergo posttranslational modification by farnesylation and palmitoylation . Insulin-mediated recruitment of C3G to the lipid raft results in the activation of TC10 via GTP exchange for GDCIP4/2P. TC10 in turn interacts with and recruits Cdc42-interacting protein 4/2 (CIP4/2) to the lipid raft . The signaling events between TC10-CIP4/2 and translocation of GLUT4-containing vesicles have not yet been elucidated. It is tempting to hypothesize that TC10, as a Rho family member, participates in the regulation of the actin cytoskeleton, given that investigators have demonstrated a functional requirement for filamentous actin in GLUT4 translocation .
Experiments involving transgene expression in adipocytes have provided strong support for the contribution of the CAP/Cbl/TC10 pathway. Deletion of CAP SH3 domains creates a dominant-interfering mutant that blocks binding to Cbl and recruitment of Cbl to lipid rafts, and prevents GLUT4 translocation and stimulation of glucose transport without loss of PI-3 kinase or MAP kinase activation . SoHo deletion mutations in CAP, which prevent binding of the CAP/Cbl complex to flotillin in lipid rafts, also block GLUT4 translocation and glucose transport stimulation . Hyperexpression of C3G potentiates insulin action and shifts the insulin dose–response curve to the left . Overexpression of TC10 (but not other Rho family members) results in mistargeting to nonlipid raft regions of the plasma membrane, and inhibits recruitment of GLUT4 translocation associated with disruption of cortical actin in adipocytes . Mutant forms of CIP4/2 also inhibit GLUT4 translocation. To illustrate the dependence on caveolin-containing lipid rafts, disruption of rafts by cholesterol-extracting drugs (β-cyclodextrin) or by expression of mutant caveolins blocks insulin stimulation of TC10 and glucose transport [117,118]. These experiments demonstrate that the CAP/Cbl/TC10 pathway is necessary for insulin stimulation of glucose transport. However, it is important to emphasize that these findings are well established in the adipocyte system, whereas, in muscle cells, a definitive role for the CAP/Cbl/TC10 pathway has yet to be firmly established.
In future research, it will be important to elucidate the mechanisms by which activation of Akt/PKB, PKC ζ/λ, and TC10 interact with insulin’s metabolic effector systems and, in particular, with the trafficking and translocation of GLUT4 vesicles (see later). Detailed study of these processes will be necessary to better define defects causing insulin resistance in human disease.
Acute exercise or muscle contraction results in GLUT4 translocation and stimulates muscle glucose transport activity . This effect occurs without any change in serum insulin concentrations and does not involve activation of PI-3 kinase or Akt/PKB. The mobilization of GLUT4 to the cell surface by acute exercise may involve a different intracellular pool of GLUT4 than that recruited by insulin, and the effects of acute exercise and insulin are partially additive . These observations indicate that signaling systems mediating glucose transport stimulation are different in response to acute exercise versus insulin . This is underscored by the finding that, in muscle from insulin resistant humans and rodents, GLUT4 translocation is impaired in response to insulin but normal in response to acute exercise.
While signal transduction mechanisms are not fully understood, this response to acute exercise appears to be at least partially dependent on increments in intracellular 5′ AMP induced by acute exercise, and subsequent activation of 5′ AMP-activated protein kinase (AMPK) [120–122]. Hypoxia and mitochondrial uncoupling agents also increase 5′ AMP levels and increase cellular glucose transport rates. Investigators in this field have made frequent use of 5-aminoimidazole-4-carboxamide-1-β-D-riboside (AICAR), which acts as a 5′AMP analog by being phosphorylated after cellular uptake and then activating AMPK. AICAR administration mimics acute exercise by stimulating both GLUT4 translocation and glucose transport in a PI-3 kinase independent manner [121,122]. AMPK is an enzyme consisting of one catalytic subunit (α) and two regulatory subunits (β and γ), and exhibits a complex mechanism of regulation that “senses” fuel status in the cell. When ATP levels are low and 5′ AMP is elevated, AMPK activates pathways for ATP regeneration and limits further ATP utilization by modifying activity of multiple metabolic enzymes, including acetyl-CoA carboxylase, hydroxymethylglutaryl-CoA reductase, creatine kinase, and hormone-sensitive lipase . Subsequent studies using AMPK-inactive mouse models confirmed the importance of this pathway in vivo .
Downstream mechanisms leading to translocation of a distinct intracellular GLUT4 pool following muscle contraction have not been fully identified. However, an Akt substrate of 160 kDa (AS160) was identified as part of the signaling network for insulin and Akt leading to GLUT4 translocation [124–126]. Insulin-stimulated phosphorylation of AS160 was demonstrated, and this was sensitive to PI-3 kinase inhibition, placing AS160 in the pathway of insulin-stimulated GLUT4 translocation . In addition, AICAR-stimulated phosphorylation of AS160 was ablated in AMPK-inactive mice . Together, these studies identified AS160 as a point of convergence in insulin-mediated and contraction-mediated GLUT4 translocation. Subsequent investigation revealed that AS160 has a complex pattern of serine/threonine phosphorylation that may define its specificity [128–130].
Subsequently, AS160, designated as gene TBC1D4, was identified as a Rab GTPase activating protein  that serves to retain GLUT4 intracellularly, and its activation by either insulin or contraction-mediated signaling releases GLUT4 and allows it to be translocated to the plasma membrane. Rabs are key elements in vesicle trafficking pathways [132–134] that toggle between active and inactive forms through GTP loading and hydrolysis, respectively. The finding that AS160 is a Rab GTPase places it squarely in the machinery of vesicle trafficking. More recently, there is evidence that AS160 plays two roles; one, in which unphosphorylated AS160 restrains GLUT4 translocation, and a second, where phosphorylation of AS160 can facilitate fusion of GLUT4 vesicles with the plasma membrane .