The activation of protein kinase B (or Akt) plays a central role in the stimulation of glucose uptake by insulin. Currently, however, numerous questions remain unanswered regarding the role of this kinase in bringing about this effect. For example, we do not know precisely where in the GLUT4 trafficking pathway this kinase acts. Nor do we know which protein substrates are responsible for mediating the effects of protein kinase B, although two recently identified proteins (AS160 and PIKfyve) may play a role. This paper addresses these important questions by reviewing recent progress in the field.
- AS 160
- glucose uptake
- protein kinase B
Insulin stimulates glucose uptake into muscle and adipose tissues by promoting a rapid translocation of the insulin-responsive isoform of the glucose transporter GLUT4 from intracellular vesicles to the plasma membrane . In the basal state, most of GLUT4 is excluded from the plasma membrane and resides both in the endosomal system, together with GLUT1 and transferrin receptors, and in a specialized intracellular pool termed ‘GLUT4 storage vesicles’ (GSVs), which excludes endosomal proteins [2,3]. Under these conditions, any GLUT4 found at the plasma membrane is rapidly endocytosed, returning to the GSV pool through a pathway supposed to involve trafficking through the endosomal system and TGN (trans-Golgi network). The addition of insulin results in a 10–20-fold increase in GLUT4 levels at the plasma membrane. The primary effect of insulin on the translocation of GLUT4 to the plasma membrane is to release GLUT4 from the GSV pool (i.e. by stimulating step 1 in Figure 1), although insulin does also increase the translocation of GLUT4 from the recycling endosomal system by stimulating step 2 and decreasing the rate of GLUT4 endocytosis (step 3; Figure 1) [3,4]. Insulin has also been reported to increase the rate at which GLUT4 is sorted from internalizing endosomes away from the transferrin receptor (step 4), and this may reflect an increase in the flux of GLUT4 through the TGN and back into the GSV pool, probably to replenish these insulin-sensitive pools [5,6].
Even though the signalling mechanism by which insulin can mobilize GLUT4 to the plasma membrane has been extensively studied, many aspects are yet to be defined fully. It is known that insulin, through the activation of its receptor tyrosine kinase, leads to the stimulation of the class 1a PI3Ks (phosphoinositide 3-kinases). This results in the production of PtdIns(3,4,5)P3 in the plasma membrane, which leads to the activation of PKB (protein kinase B; also known as Akt) and the atypical PKC (protein kinase C) isoforms PKCλ and PKCζ. PKB and PKCλ/PKCζ have both been implicated in insulin-stimulated GLUT4 translocation .
The PI3K pathway is known to be essential for the action of insulin, since the promotion of GLUT4 translocation and the stimulation of glucose transport are both blocked by the PI3K inhibitor wortmannin and by dominant-negative mutants of this enzyme (reviewed in ). It has been proposed that a second PI3K-independent signalling pathway, involving the activation of the GTP-binding protein TC10 through association of CAP, APS (adapter protein containing a PH and an SH2 domains) and Cbl adaptor proteins with the activated insulin receptor, is involved in the stimulation of glucose transport by insulin [7–9]. However, recent work has questioned the importance of this second pathway, since insulin stimulation of GLUT4 translocation and glucose uptake are not inhibited under conditions where components of the pathways are ablated (namely, in adipocytes in which members of this pathway have been subject to siRNA (small interfering RNA)-mediated depletion or in adipocytes from mice in which Cbl or APS was knocked out [10–13]).
In contrast, there is a great deal of evidence that PKB is a crucial mediator of the effect of insulin on glucose transport. For example, constitutively active PKB mutants have been shown to induce GLUT4 translocation in the absence of insulin. Furthermore, dominant-negative PKB mutants and the ablation of PKB using siRNAs decrease insulin-stimulated glucose uptake. Finally, PKBβ knockout mice are diabetic, showing decreased glucose uptake into muscle and adipose cells (reviewed in [3,12]).
While insulin stimulates the flux of GLUT4 in step 1, step 2 and possibly step 4 and decreases flux in step 3, little is known about which steps are subjected to regulation by PKB. Some time ago, we utilized two toxins from Clostridium botulinium, namely BoNT/B and BoNT/E, to attempt to answer this question . These toxins block GLUT4 vesicle docking and/or fusion with the plasma membrane by cleaving a t-SNARE and a v-SNARE (where SNARE stands for soluble N-ethylmaleimide-sensitive fusion protein attachment protein receptor) present in the plasma membrane and GLUT4 vesicle respectively. BoNT/B and BoNT/E inhibited insulin-stimulated translocation by approx. 65%. In contrast, they completely blocked the translocation induced by a constitutively active PKB. Interestingly, the insulin-stimulated translocation of GLUT1 and transferrin receptors to the plasma membrane was unaffected by these toxins. Taken together, the results suggest that PKB acts by stimulating the translocation of GLUT4 from the GSV pool to the plasma membrane by increasing the flux through step 1. It has no obvious effect on the flux of GLUT4 from the endosomal system through step 2. This is further supported by the recent observation that a constitutively active form of PKB stimulated the translocation of GLUT4, but not GLUT1 or transferrin receptors, to the plasma membrane .
If constitutively active PKB and insulin act by stimulating the flux through step 1, then where, precisely, do they act? Much of the GLUT4 found in basal cells is stored at some distance from the plasma membrane, either in the perinuclear region or in tubulovesicular structures distributed throughout the cytoplasm. So, do PKB and insulin initiate the release of GLUT4 from the stored intracellular pool or do they accelerate the final fusion step between the GLUT4 vesicle and the plasma membrane? Time-lapse imaging of GFP (green fluorescent protein)-tagged GLUT4 provides some clues, at least for insulin. In the basal state, GLUT4 vesicles show little apparent movement other than very short linear movements and rapid vibrations around a central point as if the vesicles are tethered to an intracellular structure. When insulin is added, we rarely if ever observe the direct translocation of any GFP–GLUT4 vesicle to the plasma membrane [15,16]. In contrast, we observed a gradual dimming of GLUT4 vesicles as they deliver their GFP–GLUT4 content to the plasma membrane through some so far ill-defined mechanism.
This strongly suggests that the insulin signal that begins at the plasma membrane must propagate into the cell towards the intracellular GLUT4 vesicles, initiating the delivery of their contents into the plasma membrane. In support of this idea, it has been reported that PKB associates with intracellular GLUT4 vesicles in an insulin-dependent manner [17,18]. We studied the functional consequence of this translocation by artificially targeting constitutively active and kinase-dead mutants of PKB to intracellular GLUT4 vesicles by fusing them with the N-terminus of GLUT4 itself. We examined the effect of these mutants on the insulin-dependent translocation of the IRAP (insulin-responsive amino peptidase; a bona fide GLUT4-vesicle-resident protein). We found that a kinase-dead PKB targeted to GLUT4 vesicles (PKB[KD]GLUT4) was an extremely effective dominant-negative inhibitor of insulin-stimulated translocation of IRAP to the plasma membrane. Furthermore, a kinase-dead PKB containing a myristoylation signal sequence (Myr-PKB[KD]), which was expressed both at the plasma membrane and within intracellular vesicles containing endogenous IRAP, was also a potent dominant-negative inhibitor of IRAP–GFP translocation. In contrast, a kinase-dead PKB expressed in the cytoplasm did not exhibit any dominant-negative effect. This mutant does not appear to act by blocking the activation of endogenous PKB by insulin, since the PKB[KD]GLUT4 chimaera had no effect on the ability of insulin to promote the translocation of the transcription factor FKHR (forkhead in rhabdosarcoma) out of the nucleus . These results suggest that PKB has an important functional role at, or in the vicinity of, GLUT4 vesicles. We propose that this step is the initiation stage of translocation. However, we cannot exclude the possibility that PKB also modulates the final fusion step between the GLUT4 vesicle and the plasma membrane.
In summary, therefore, PKB must phosphorylate a key substrate (or substrates) localized close to the GLUT4 vesicle that plays a central role in translocation. Clearly, the identification of these substrates represents the next most important stage in understanding the role of PKB in glucose uptake. More than 30 PKB substrates have been identified to date (Table 1) [20,21]. At present, there is only evidence for an involvement of two of these in insulin-stimulated glucose transport. The first is AS160, a Rab-GAP (where GAP stands for GTPase-activating protein) that contains five sites (Ser318, Ser570, Ser588, Thr642 and Thr751), which conform to the PKB substrate's consensus sequence (RXRXX[pS/pT]), and whose phosphorylation has been shown to be increased by insulin. Rab GTPases are key players in membrane-trafficking events and have been shown to have critical roles in vesicle formation, fusion and movement. The Rab protein regulated by AS160 is yet to be identified; however, it has been demonstrated that a mutant AS160 lacking the PKB phosphorylation sites blocks the ability of insulin to stimulate the exocytosis of GLUT4 at a step preceding the docking and fusion of GLUT4 vesicles with the plasma membrane. As such, this protein may regulate the flux of GLUT4 through step 1 (Figure 1). This mutant has no effect on the inhibition of GLUT4 internalization by insulin (step 3) [4,22,23].
We have recently identified a PtdIns3P 5-kinase (PIKfyve) as another PKB substrate that plays a role in insulin-regulated GLUT4 trafficking. To do this, we used a proteomic approach utilizing a commercially available antibody (PAS) raised against the minimal PKB consensus phosphorylation site found on almost all of its known substrates, RXRXX(pS/pT). A similar approach was used by Gus Lienhard and co-workers  to identify AS160. Other insulin-stimulated adipocyte phosphoproteins identified using this method include three proteins of unknown function (PRAS40, pp47 and pp105) and pp250, a protein with a predicted GAP domain for Rheb and/or Rap at its C-terminus [24,25]. We have also used this technique to show that PKB phosphorylates ATP-citrate lyase in vitro on Ser454, a previously reported insulin- and isoprenaline-stimulated phosphorylation site .
PIKfyve contains a PtdIns3P-binding FYVE domain, a DEP (dishevelled, Egl-10 and pleckstrin homology) domain of unknown function, a chaperonin-like region and a PtdIns 5-kinase catalytic domain. PIKfyve is the mammalian homologue of yeast Fab1p that has been shown to be important for cargo sorting from late endosomes to lysosome/multivesicular body . Indeed, PIKfyve has previously been suggested to have a role in insulin-stimulated GLUT4 translocation , although this was on the basis of the dominant-negative effect of a kinase-dead PIKfyve, which is known to promote vacuolation of intracellular membranes, an observation which may complicate the interpretation of this experiment.
PIKfyve has a highly stringent PKB phosphorylation site at Ser318 and we have shown that this site is phosphorylated by PKB in vitro and in intact cells, both in response to insulin and in a PI3K-dependent manner, and by constitutively active mutants of PKB or PI3K. Although, at first sight, using laser scanning confocal microscopy, there appears to be little obvious co-localization between GLUT4 and PIKfyve, rapid time-lapse imaging showed that PIKfyve co-localizes with a highly motile subpopulation of GLUT4 vesicles, which appeared to traffic from the cell periphery to the perinuclear region. Furthermore, expression of a PIKfyve mutant [S318A (Ser318→Ala)] in 3T3-L1 adipocytes enhances insulin-stimulated GLUT4 translocation . These results suggest that PIKfyve and its phosphorylation by PKB may have a role in insulin-regulated GLUT4 trafficking, potentially being involved in increasing the rate at which GLUT4 is sorted from the internalizing endosomes through the TGN and into the GSV pool. This may represent the acceleration of step 4 (Figure 1), which has been observed in the previous studies in response to insulin and constitutively active PKB mutants [5,6].
Whether AS160 and PIKfyve turn out to be the key controllers of GLUT4 trafficking and translocation to the plasma membrane requires further testing. One of the main problems in studying insulin-regulated GLUT4 trafficking is the fact that GLUT4 is in dynamic equilibrium with a number of intracellular compartments, including GSVs and endosomes (which are insulin-sensitive compartments) as well as the endoplasmic reticulum, TGN, Golgi and other compartments (which are most probably insulin-insensitive). Such manipulations could perturb any one of these trafficking steps that could alter the extent to which GLUT4 is capable of entering, or being retained, in the insulin-responsive compartment(s). This could indirectly decrease or enhance insulin-stimulated GLUT4 translocation. Such a protein may not, therefore, be directly involved in the translocation step. This could conceivably be the case for either AS160, which controls members of the Rab family of GTP-binding proteins that are involved in numerous intracellular trafficking events, or PIKfyve. To address this issue, we must, therefore, put some effort into developing methods that truly look at the key regulatory steps in insulin-stimulated GLUT4 trafficking and take into account non-specific effects on non-regulated steps.
We thank the Medical Research Council and Diabetes U.K. for financial support.
Kinases in Diabetes: Focused Meeting held at Hulme Hall, University of Manchester, U.K., 11 November 2004. Organized and Edited by D. Smith, E. Kilgour and M. Coghlan (AstraZeneca UK). Sponsored by AstraZeneca UK.
Abbreviations: GAP, GTPase-activating protein; GFP, green fluorescent protein; GSV, GLUT4 storage vesicle; IRAP, insulin-regulated aminopeptidase; PI3K, phosphoinositide 3-kinase; PKB, protein kinase B; siRNA, small interfering RNA; TGN, trans-Golgi network
- © 2005 The Biochemical Society