|Home | About | Journals | Submit | Contact Us | Français|
Protein kinase B/Akt has been implicated in the insulin-dependent exocytosis of GLUT4-containing vesicles, and, more recently, insulin secretion. To determine if Akt also regulates insulin-independent exocytosis, we used adrenal chromaffin cells, a popular neuronal model. Akt1 was the predominant isoform expressed in chromaffin cells, although lower levels of Akt2 and Akt3 were also found. Secretory stimuli in both intact and permeabilized cells induced Akt phosphorylation on serine-473, and the time course of Ca2+-induced Akt phosphorylation was similar to that of exocytosis in permeabilized cells. To determine if Akt modulated exocytosis, we transfected chromaffin cells with Akt constructs and monitored catecholamine release by amperometry. Wild-type Akt had no effect on the overall number of exocytotic events, but slowed the kinetics of catecholamine release from individual vesicles, resulting in an increased quantal size. This effect was due to phosphorylation by Akt, as it was not seen in cells transfected with kinase-dead mutant Akt. As overexpression of cysteine string protein (CSP) results in a similar alteration in release kinetics and quantal size, we determined if CSP was an Akt substrate. In vitro 32P-phosphorylation studies revealed that Akt phosphorylates CSP on serine-10. Using phospho-serine10-specific antisera, we found that both transfected and endogenous cellular CSP is phosphorylated by Akt on this residue. Taken together, these findings reveal a novel role for Akt phosphorylation in regulating the late stages of exocytosis and suggest that this is achieved via the phosphorylation of CSP on serine-10.
Exocytosis is the fusion of secretory vesicles with the plasma membrane. Constitutive exocytosis, where fusion is apparently unregulated, is used by all cells to deliver integral membrane proteins to the plasma membrane, and for the secretion of various substances. In contrast, regulated exocytosis, where fusion is triggered by an intracellular signal, is characteristic of ‘professional’ secretory cells that release material only on demand, such as neurons, endocrine and exocrine cells (1). Regulated exocytosis is not always used for secretion, however, as it is also an important mechanism for the stimulus-dependent insertion of cell surface receptors and transporters. In the great majority of cell types, the intracellular signal that triggers regulated exocytosis is an increase in the cytoplasmic free Ca2+ concentration. While Ca2+ can be thought of as a near universal trigger for exocytosis, protein phosphorylation can be considered as an equally widespread modulator of regulated exocytosis (2). Indeed, many studies over the past 20 years have shown that Ca2+-stimulated exocytosis is controlled by protein kinases (PKs) and/or phosphatases in almost all cell types, including neurons (3-6). Although a variety of kinases have been implicated from these studies, to date only PKA and PKC are candidates for general modulators of regulated exocytosis across a wide range of cell types. For example, activation of PKC has been shown to enhance exocytosis in exocrine pancreatic acinar cells (7), endocrine adrenal chromaffin cells (8) and in various neuronal systems, including neuromuscular junctions (9), synaptosomes (10,11) and the calyx of Held (12). Likewise, activation of PKA increases exocytosis in pancreatic acinar cells (7), adrenal chromaffin cells (13) and in neuronal preparations ranging from the squid giant synapse (14) to the mammalian hippocampus (15) and cerebellum (16). Abundant evidence suggests that these effects of PKA and PKC are due to phosphorylation of components of the exocytotic machinery. Although the molecular details are not entirely clear, good candidates for such PKA substrates are cysteine string protein (CSP) (17,18), Snapin (19), Rim1 (20) and SNAP-25 (21).
PKA and PKC may not be the only kinases with a general function in modulating exocytosis, however. Recent studies have hinted that Akt/PKB may also be an important kinase in the control of regulated exocytosis. Akt is an evolutionarily conserved serine/threonine kinase, three isoforms of which have been identified in mammals (Akt 1, 2, 3; PKB a, ß, ?), which has important functions in the regulation of metabolism and cell fate (22). A role for Akt in regulated exocytosis was first discovered in the insulin-stimulated exocytosis of glucose transporter 4 (GLUT4) containing vesicles. Expression of a constitutively active Akt construct stimulated GLUT4 translocation, whereas microinjection of an Akt substrate peptide or an antibody to Akt inhibited translocation in adipocyte cell lines (23,24). Similarly, in transfected skeletal muscle myoblast cell lines, overexpression of constitutively active Akt1 was seen to increase GLUT4 translocation whereas a dominant negative Akt1 construct inhibited translocation (25,26). Studies of Akt2 knockout mice have revealed defects in glucose disposal due to an impairment of GLUT4 translocation in adipocytes, thus clearly demonstrating a physiological role for this Akt isoform in exocytosis (27,28). Most recently, it has been shown that insulin secretion is inhibited in transgenic mice expressing a kinase-dead mutant Akt construct in pancreatic ß cells (29). The molecular mechanism(s) by which Akt regulates exocytosis is unknown, however. Here we report a novel role for Akt in regulating exocytotic release kinetics and quantal size in adrenal chromaffin cells. We also identify Akt as a cellular CSP kinase that phosphorylates CSP on serine10, a residue that is essential for the alteration of exocytosis release kinetics and quantal size by CSP. Taken together, these findings suggest a mechanism whereby Akt phosphorylation of CSP on serine-10 regulates the late stages of exocytosis.
Plasmids encoding wild type and mutant constructs of CSP (pQE30-CSP1, pQE30-CSP1 (S10A), pcDNA3-myc-CSP1, pcDNA3-myc-CSP1(S10A)) and Akt (pLNCX1-HA-AKT1, pLNCX1-HA-myr-AKT1, pcDNA3-HA-AKT AAA) have been previously described (17,18,23,26,30). Anti Akt2 and Akt3 antisera have been previously described (31). Anti-Akt1 antibody and purified, recombinant Akt isoforms were obtained from Upstate (Dundee, UK). Anti-pan-Akt, and anti-phospho-serine-473-specific antisera were from Cell Signalling Technology (Beverly, MA, USA). Anti-rabbit biotinylated antibody, 32P-ATP and Hyperfilm ECL were obtained from Amersham (Amersham, Buckinghamshire, UK). Streptavidin-Alexa 488, anti-mouse-Alexa 594 and anti-sheep-Alexa 350 were from Molecular Probes (Eugene, OR). Synthetic peptides were generated by MWG Biotech (Milton-Keynes, UK). Lipofectamine was from Invitrogen (Paisley, UK) and protease inhibitor cocktail was from Roche (Welwyn, UK). All other reagents were from Sigma (Poole, Dorset, UK).
Recombinant His6-CSP was expressed and purified as previously described (32). In vitro phosphorylation of His6-CSP was performed in 50 mM Tris, pH 7.5, 20 mM MgCl2, 1 mM EGTA, 15 mM DTT, 0.25 mM NaVO43-, 0.03% Brij-35 for Akt; and in 50 mM MES, pH 6.9, 10 mM MgCl2, 0.5 mM EDTA, 1 mM DTT for PKA. In both cases, 5 μg/ml purified kinase was added and the reaction initiated by the addition of 2 μCi [γ-32P]ATP and unlabelled ATP to a final concentration of 100 μM. Samples were solubilised in Laemmli buffer, boiled and run on SDS-PAGE. 32P incorporation was visualised using a phosphoimager (Molecular Dynamics). For peptide phosphorylation assays, Csp(4-14) peptides or Crosstide were used at 0.1-30 μM and incubated for 5 min with 0.5 μg/ml purified Akt. Under these conditions, the incorporation of phosphate was linear with respect to time and enzyme concentration. Reactions were terminated by spotting onto Whatman P81 phosphocellulose paper followed by extensive washing in 5 mM orthophosphoric acid and determination of incorporated 32P was by liquid scintillation counting. Kinetic parameters were calculated by linear regression of S/V v. S plots (S=substrate concentration and V=initial rate of phosphorylation)
Rabbit polyclonal anti-serum to CSP has been previously described (30). Sheep polyclonal antiserum to CSP was generated by ProSci (Poway, CA, USA) by immunisation with recombinant purified His6-CSP protein. The generation of rabbit polyclonal phospho-CSP-specific antiserum is described elsewhere (33). Briefly, the phosphopeptide CQRQRSLpSTSGE, containing amino acid residues 4-14 of CSP with a phosphorylated serine at position 10, was used as an immunogen in rabbits. The resulting antiserum was affinity purified using the P-CSP phosphopeptide coupled to an immobilised matrix. Specificity of detection in immunoblotting and immunofluorescence was demonstrated by pre-incubating antisera with recombinant purified His6-CSP protein, the P-CSP phosphopeptide, or an unphosphorylated CSP(4-14) peptide (CQRQRSLSTSGE).
Human embryo kidney 293 cells were grown in DME medium with 10% FBS. Cells were transfected with CSP, CSP(S10A), myr-Akt and AAA-Akt plasmids using lipofectamine according to the manufacturer’s instructions. 36 hours after transfection, the cells were serum starved for 8 hours and then treated with or without 172 nM insulin for 20 min. The cells were chilled at 4 degrees, washed with PBS three times and collected in lysis buffer; 20mM Tris, pH=7.5, 150mM NaCl, 1mM EDTA, 1mM EGTA, 1% Triton with protease inhibitors and phosphatase inhibitors. The lysates were centrifuged at 10,000g for 5min and the supernatants were used for immunoprecipitation by myc antibody. SDS-PAGE and immunoblots were performed using standard procedures.
Bovine adrenal chromaffin cells were isolated as previously described (17). For transfection, cells were plated onto non-tissue culture treated 10 cm Petri dishes and left overnight at 37°C. Non-attached cells were resuspended in fresh growth medium at a density of 1×107 cells/ml. Plasmid constructs were added to the chromaffin cell suspension at 20 μg/ml. For amperometry experiments, plasmids were co-transfected with a vector encoding EGFP in order to identify co-transfected cells. Cells were electroporated using a Bio-Rad Gene Pulser II (Bio-Rad, Hercules, CA), immediately diluted to 1×106 cells/ml and maintained in culture for 3-5 days. For secretion assays and western blotting, cells were plated directly on to 24-well plates at a density of 1 million cells per well and maintained in culture for 3-6 days before use.
Chromaffin cells grown on collagen-coated cover slips were fixed in 4 % formaldehyde for 20 min and then permeabilized and blocked for non-specific antibody binding in PBS, 0.1% Triton X100 and 1 % bovine serum albumin. Primary antibodies (mouse anti-HA-tag, sheep anti-CSP, or rabbit anti-phospho-CSP) were applied for 2 h at room temperature and then the cells were washed 3 times in PBS. For triple labelling studies, anti-phospho-CSP antisera was incubated first, followed by anti-CSP in order to prevent epitope masking by the anti-CSP antiserum, which was raised against the full-length recombinant protein. Secondary antibodies were incubated for 1 hr followed by washing 3 times in PBS; then streptavidin-fluorophore was added for 30 min. Coverslips were washed, air dried and mounted on slides in anti-fade glycerol (90 % glycerol, 0.25 % triethylenediamine, 0.002 % p-phenylenediamine). For conventional fluorescence microscopy, staining was visualised with a NIKON TE300 inverted microscope and images were acquired using a NIKON Coolpix 995 digital camera. For confocal microscopy, staining was visualised using a Leica AOBS TCS SP2 laser scanning confocal microscope using a 63x oil immersion objective with a numerical aperture of 1.2. The following fluorophores and parameters were used: 416 nm excitation and 506 nm light collection for Alexa-350; 476 nm excitation and 500-581 nm light collection for Alexa-488; and 606 nm excitation and 757 nm light collection for Alexa 594. For quantification of fluorescence, the region of interest was defined and the CSP (blue) and P-CSP (green) signals quantified using the Leica confocal software. The ratio of P-CSP:CSP was then calculated after subtraction of background fluorescence. At least 10 cells were imaged and quantified for each condition. Statistical significance was assessed using a student’s t test.
Electrophysiological recording conditions were as described previously (34). Briefly, cells were incubated in bath buffer (139 mM potassium glutamate, 0.2 mM EGTA, 20 mM PIPES, 2 mM ATP and 2 mM MgCl2, pH 6.5) and a 5 μm-diameter carbon fibre was positioned in contact with the target cell. Exocytosis was stimulated with a permeabilization/stimulation buffer (139 mM potassium glutamate, 20 mM PIPES, 5 mM EGTA, 2 mM ATP, 2 mM MgCl2, 20 μM digitonin and 10 μM free Ca2+, pH 6.5) pressure-ejected from a glass pipette on the opposite side of the cell. Amperometric responses were monitored with a VA-10 amplifier (NPI Electronic, Tamm, Germany) and saved to computer using Axoscope 8 (Axon Instruments). Experiments were carried out in parallel on control (untransfected cells) and transfected cells from the same batch of cells in the same cell culture dishes. Transfected cells were identified by expression of EGFP. Previous studies have established that 95% of cells co-express proteins from both plasmids in the transfection (35). Recordings of both control and transfected cells used the same carbon fibres to eliminate any potential effects of inter-fibre variability. Amperometric data were exported from Axoscope and analyzed using Origin (Microcal Software, Northampton, MA). Amperometric spikes were selected for analysis provided that the spike amplitude exceeded 40 pA, in order to remove any confounding effects of diffusion by selecting those fusion events not occurring directly beneath the carbon fibre end. Individual spikes were analyzed for total charge released (measured by the integral of the spike), amplitude (the height from baseline to peak), rise time (time from spike onset to peak) and fall time (time from spike peak to return to baseline). Spike frequency (spikes per cell) was calculated as the number of exocytotic events within the 210 seconds of recording time. All data presented are shown as mean ± S.E.M. Statistical differences were assessed with nonparametric Mann-Whitney tests.
Assays of bulk exocytosis from populations of cultured cells were performed as previously described (13). For intact cells, cells were incubated in Krebs buffer in the presence or absence of nicotine; for permeabilized cells, cells were incubated in permeabilization/stimulation buffer or in the same buffer with no added Ca2+. Catecholamine release was assayed using a fluorimetric assay and expressed as the percentage of total cellular catecholamine.
To determine which Akt isoforms are expressed in bovine adrenal chromaffin cells, cell lysates were run on SDS-PAGE alongside recombinant Akt protein standards and western blotted with isoform-specific Akt antibodies (Fig. 1A). Anti-Akt1 gave a readily detectable band in chromaffin cells, equivalent to between 5 and 10 ng of recombinant Akt1. This antibody was not entirely specific, however, as bands of lesser intensity were also detected using 10 ng of Akt2 and Akt3 (Fig. 1A, upper panel). In contrast, the Akt2 and Akt3 antibodies were absolutely specific for their corresponding recombinant protein. The Akt2 and Akt3 signals in chromaffin cells were much lower than those seen with 5 ng of recombinant Akt2/3. Therefore, the partial cross-reactivity of anti-Akt1 with Akt2/3 can at most account for only a small proportion of the staining intensity seen with anti-Akt1. We therefore concluded that Akt1 is the predominant isoform expressed by bovine chromaffin cells, although lower levels of Akt 2 and 3 are also present.
Exocytosis is triggered by Ca2+ in chromaffin cells, so we investigated whether secretory stimuli resulted in activation of cellular Akt. To do this we treated cells with nicotine, which activates acetylcholine receptors and triggers Ca2+ entry through channels to elicit exocytosis. Akt activity was assayed by western blotting of cell lysates with a phospho-serine473-specific antibody. Nicotine treatment resulted in a clear increase in phospho-Akt staining, despite no change in total Akt levels (Fig. 1B, left panel). To determine if Ca2+ directly drives Akt activation, we used digitonin-permeabilized cells. Application of 10 μM Ca2+, the optimal concentration for triggering exocytosis, again resulted in increased Akt phosphorylation (Fig. 1B, right panel). To determine the time course of Akt activation relative to secretion, we simultaneously monitored catecholamine release and Akt phosphorylation (Fig. 1C, D). Both processes followed broadly similar time courses, with Ca2+-induced Akt phosphorylation being detectable within 2 minutes of stimulation.
To determine if Akt regulates exocytosis in adrenal chromaffin cells, we transfected cell cultures with HA-tagged Akt1 constructs and analysed individual exocytotic release events using carbon fibre amperometry. Transfected cells were visualised by the endogenous fluorescence of an EGFP reporter plasmid that was co-transfected as a marker. After stimulation of regulated exocytosis with digitonin and calcium, transient spikes of catecholamine oxidation were recorded, indicative of individual chromaffin granule fusion events (Fig. 2A). Transfection of wild type Akt had no significant effect on the frequency of exocytotic fusion events elicited (Fig. 2B). In contrast, analysis of individual exocytotic events revealed a 43% increase in the amount of catecholamine released per fusion event (Fig. 2C). This increased quantal size was due to a slowing of catecholamine release kinetics, as evidenced by an increase in both the rise- and fall-time of amperometric spikes (Fig. 2D-E). Transfection of a myristoylated, constitutively active form of Akt produced a similar increase in quantal size of individual fusion events and likewise had no effect on overall frequency of release events (data not shown). To confirm that the Akt constructs were active upon transfection into chromaffin cells, double label immunofluorescence was used. Total transfected Akt was visualised using a HA-tag antibody (red) while active Akt was visualised using a phospho-serine473-specific antibody (green). A clear increase in cellular activated Akt (as defined by phospho-serine473 immunoreactivity) was seen in all HA-positive cells transfected with wt-Akt (Fig. 2F) or constitutively active Akt (data not shown).
In order to establish whether these effects of Akt on exocytosis were due to Akt phosphorylation of target proteins, or alternatively reflected a kinase-independent function, we transfected chromaffin cells with a mutant Akt construct, AAA-Akt. This construct encodes Akt1 containing three point mutations: K179A, T308A and S473A. These mutations result in a kinase-dead, phosphorylation-deficient form of Akt that has been shown to act as a dominant negative mutant in some cell types (26). Stimulation of AAA-Akt transfected cells by digitonin and calcium again resulted in transient spikes of catecholamine release due to exocytosis (Fig. 3A). As with wild type Akt, there was no statistically significant effect on the overall frequency of exocytotic fusion events (Fig. 3B). In pointed contrast to wt-Akt, however, no significant changes in charge, rise-time or fall-time were observed upon transfection of AAA-Akt (Fig. 3C-E). These data strongly suggest that the effect of Akt on quantal size and the kinetics of exocytotic release requires Akt kinase activity. In order to rule out the possibility that the AAA mutant was not stably expressed in the transfected cells, transfected Akt was visualised using an HA-tag antibody (Fig. 3F, red). Similar transfection efficiency and fluorescence intensity was seen for both AAA-Akt and wt-Akt, thus ruling out this trivial explanation. In contrast to wt-Akt, however, no increase in phospho-serine473 immunoreactivity (Fig. 3F, green) was apparent in AAA-Akt transfected cells, as predicted for this mutant protein.
Transfection of CSP in chromaffin cells has two effects on exocytosis as measured by amperometry: a reduction in the frequency of exocytotic fusion events, and an increased quantal size as a result of a slowing of release kinetics (36). This latter effect on quantal size (but not the effect on frequency) is abolished in a mutant CSP(S10A) construct that cannot be phosphorylated on serine-10 (17) (Table 1). This raised the possibility that the similar effect of Akt and CSP transfection in amperometry was due to direct Akt phosphorylation of CSP on this residue. A comparison of the coding sequence of CSP with the minimal Akt consensus phosphorylation motif revealed a potential Akt phosphorylation site at serine-10 (Fig. 4A), the same residue that is phosphorylated in vitro by PKA (17). An in vitro kinase assay using activated Akt and recombinant purified CSP in the presence of 32P-labelled ATP was then set up to test this experimentally. As can be seen in Fig. 4B, 32P was readily incorporated into wild type CSP, indicating phosphorylation by Akt. However, when the assay was performed using S10A mutant recombinant CSP, which has a non-phosphorylatable alanine in place of serine-10, 32P incorporation was barely detectable despite equal levels of recombinant protein being present in both cases (Fig. 4B). To determine the efficiency of Akt phosphorylation, we analysed the kinetics of 32P incorporation under initial rate conditions using synthetic peptides. The peptides used corresponded to amino acids 4-14 of CSP, the same peptide but with an S10A substitution, or the optimal Akt substrate peptide, Crosstide. This revealed the N-terminal CSP peptide to be a comparable Akt substrate to Crosstide, whereas the S10A peptide exhibited no detectable 32P incorporation under the same conditions (Fig. 4C,D).
Although these data clearly showed that CSP was an efficient in vitro Akt substrate, a different approach was required to assess whether Akt was a cellular CSP kinase. To address this issue we raised an antibody (P-CSP) against a synthetic CSP4-14 peptide containing a phosphorylated serine 10 residue (33). As can be seen in Fig. 5A, this P-CSP antibody specifically detected CSP that had been phosphorylated in vitro by PKA, but displayed no observable binding to mock-phosphorylated CSP or to serine-phosphorylated recombinant Munc18-1. The phospho-specificity of the antibody was further confirmed by the abolition of binding by the phospho-CSP4-14 peptide used for immunisation and the lack of effect of a non-phosphorylated version of the same peptide (Fig. 5A). When in vitro phosphorylation reactions were blotted and probed with the P-CSP antibody, a strong signal was observed with both Akt- and PKA-phosphorylated wild type CSP, but not with S10A mutant CSP (Fig. 5B), confirming that CSP is efficiently phosphorylated by Akt on serine-10 in vitro. To determine if this also occurred within living cells, we co-transfected HEK293 cells with plasmids encoding CSP and Akt constructs and monitored CSP phosphorylation using the P-CSP antibody (Fig. 5C). CSP-transfected cells co-transfected with constitutively active Akt (myr-Akt) gave a much stronger P-CSP signal than those co-transfected with AAA-Akt, despite similar levels of total expressed CSP. In contrast, CSP phosphorylation was barely detectable in S10A mutant CSP-transfected cells co-transfected with either Akt construct.
Finally, we sought to determine if Akt phosphorylates CSP in the adrenal chromaffin cells used for our functional exocytosis experiments. Unlike HEK cells, chromaffin cells express endogenous CSP and have a very low transfection efficiency, thus precluding the use of the western blotting approach. We therefore used a triple-labelling immunofluorescence approach on chromaffin cells transfected with CSP and Akt plasmids. For this purpose, we raised a new sheep polyclonal antibody to recombinant CSP protein to enable simultaneous detection of total CSP and phospho-CSP in the same cells. Characterisation of this sheep antiserum on recombinant protein, transfected cells and native tissue revealed a very similar specificity to the previously described rabbit antiserum (30) (data not shown). Double CSP/Akt-co-transfected cells were identified via the HA-tag antibody (red) marker for recombinant Akt and in all cases a large increase in total CSP expression (blue) was observed. Despite these similar levels of total CSP, however, there was a striking increase in the phospho-CSP (green) signal in wt-Akt transfected cells compared to AAA-Akt transfected cells (Fig. 6A). Quantification of the fluorescence intensity of these double transfected cells using confocal microscopy revealed an approximately 3-fold higher CSP phosphorylation in wt-Akt-transfected cells compared to AAA-Akt transfected cells (Table 2). Transfection of CSP alone resulted in a similar level of phosphorylation to that seen with AAA-Akt, further demonstrating the specificity of recombinant CSP phosphorylation by Akt in chromaffin cells (Fig. 6A and Table 2). In cells transfected with Akt constructs alone, a significant increase in phosphorylation of endogenous CSP was also observed with wt-Akt in comparison to AAA-Akt and non-transfected controls (Table 2). Therefore, Akt is a chromaffin cell CSP kinase capable of phosphorylating both transfected and native CSP on serine-10. To determine if Ca2+-induced activation of Akt would further increase CSP phosphorylation, we permeabilized untransfected chromaffin cells and stimulated them with Ca2+ (Fig. 6B). Tonic Akt activity was evident in the absence of Ca2+, and this was increased by Ca2+ application. With CSP, a high level of basal phosphorylation on serine-10 was observed, but no further increase in this signal occurred in response to Ca2+. It is likely that this high level of tonic CSP phosphorylation would prevent detection of small Ca2+- induced acute changes in phosphorylation using our phospho-specific antibody, however. Indeed, using a 32P-metabolic labelling approach, in which the high ‘noise’ of basal phosphorylation is reduced, it has previously been shown that nicotine increases CSP phosphorylation (18).
The existing evidence linking Akt to the regulation of exocytosis is mainly based on studies of insulin-stimulated GLUT4 translocation. Although not without controversy (37,38), the available evidence supports a physiological role for Akt in the exocytosis of GLUT4 vesicles in both muscle and fat cells (24-28,39). Very recently, it was reported that insulin secretion is inhibited in transgenic mice expressing a kinase-dead mutant Akt construct in pancreatic ß cells (29), suggesting that Akt’s role in exocytosis is not restricted to GLUT4 translocation. The data presented here indicates that Akt modulates Ca2+-triggered exocytosis in adrenal chromaffin cells. These cells are derived from the same precursor cells that give rise to sympathetic neurons and are commonly used as neuronal models. This demonstration of a role for Akt in neurosecretion suggests that Akt, like PKA and PKC, may be a general regulator of exocytosis across radically different cell types. It is conceivable that this novel role for Akt in chromaffin cell exocytosis underlies previous observations that insulin-like growth factor 1 enhances catecholamine secretion from these cells (40). Although phosphatidylinositol 3-kinase inhibitors, which would be predicted to indirectly inhibit Akt, have been shown to inhibit catecholamine release from chromaffin cells when used at high doses, it has been assumed that this is due to effects on other targets, such as myosin light chain kinase and Ca2+ channels (41,42). However, it has very recently been shown that phosphatidylinositol 3-kinase C2 alpha is essential for exocytosis in chromaffin cells (43). As this isoform is only inhibited by very high concentrations of phosphatidylinositol 3-kinase inhibitors, this may, in part, explain the relative insensitivity of chromaffin cell secretion to these drugs noted in the earlier studies. Indeed, we found that application of wortmannin at sub-micromolar doses only partially inhibited Akt Ser473 phosphorylation (Fig. 6C), even when incubated for prolonged periods, consistent with an important role for phosphatidylinositol 3-kinase C2 alpha in chromaffin cells.
We found that the effect of Akt in chromaffin cells is restricted to a very late stage in the exocytotic process, manifesting as an alteration in the rate and extent of release from individual vesicles. The precise mechanism that underlies this Akt-induced slowing of release kinetics and increased quantal size is unclear. However, similar effects are caused by overexpression of CSP (36) and mutant versions of syntaxin (44) and Munc18 (45), and these effects have been suggested to represent a shift away from transient ‘kiss and run’ exocytosis and towards full fusion exocytosis. Other interpretations are also possible, however, and further work is needed to shed light on the mechanism by which Akt affects the late stages of exocytosis. Interestingly, Akt had no effect on the early stages of vesicle recruitment in chromaffin cells. This lack of effect of Akt on the frequency of exocytotic fusion events is confirmed by experiments in PC12 cells, where both AAA-Akt (46) and wt-Akt constructs (our unpublished data) have no effect on bulk assays of growth hormone exocytosis. This contrasts with the situation in GLUT4 exocytosis, where an increase in the overall extent of GLUT4 translocation and glucose uptake is associated with Akt activity. Akt appears to regulate multiple stages of the GLUT4 translocation process. For example, it has been shown to accelerate the transit of GLUT4 through endosomes (47), to regulate a pre-fusion recruitment stage (48), and to regulate a late, GLUT4 externalisation stage (49). One possible explanation for the ability of Akt to regulate both early and late stages of exocytosis is that Akt may phosphorylate distinct target proteins involved in each stage. For example, in chromaffin cells, PKC regulates both the early phase of vesicle recruitment into the releasable pool and the late stages of membrane fusion via phosphorylation of SNAP -25 and Munc18-1, respectively (34,50). Alternatively, it may be that a single Akt substrate performs distinct early and late functions in these different cell types. Although the recently reported inhibitory effect of kinase-dead Akt on insulin secretion appears to be downstream of the Ca2+ trigger for exocytosis, the stage in the exocytotic process affected was not determined and indirect effects on the biogenesis of releasable granules cannot be ruled out in such transgenic mice. In contrast, our study assays release of pre-formed granules in permeabilized cells, thus demonstrating a direct effect of Akt phosphorylation on the exocytotic machinery. In view of our findings, it would be interesting to assess whether the inhibition of insulin secretion by kinase-dead Akt is a consequence of reduced insulin release from individual granules.
Despite the intensive study of the role of Akt in GLUT4 translocation, the molecular mechanism(s) by which Akt regulates exocytosis remains unclear. Potentially important Akt substrates include AS160, a Rab GTPase activating protein, and PIKfyve, a phosphatidylinositol-3-phosphate 5-kinase, both of which are Akt targets affecting GLUT4 translocation (48,51). Our data establish CSP as a novel Akt substrate whose phosphorylation on serine 10 may underlie the effect of Akt on exocytotic release kinetics and quantal size in chromaffin cells. CSP is a synaptic vesicle-localised member of the DnaJ family of co-chaperones, and several studies support a role for CSP as a chaperone in the synapse (32,52-56). CSP is not restricted to neurons, however, and is expressed in a wide range of cell types, including adipocytes (57), adrenal chromaffin cells (58) and pancreatic ß cells (59). Furthermore, CSP has been shown to modulate exocytosis in pancreatic ß cells (59-61) and chromaffin cells (36). Transfection of CSP in chromaffin cells has two distinct effects on exocytosis: a reduction in the frequency of exocytotic fusion events, and an increased quantal size as a result of a slowing of release kinetics (36). Phosphorylation of CSP on serine-10 has been implicated specifically in the latter effect on quantal size, as this is abolished in a mutant CSP(S10A) construct that cannot be phosphorylated on serine-10 (17). Furthermore, in vitro phosphorylation of CSP on serine-10 by PKA reduces the binding affinity of CSP for the key exocytotic proteins syntaxin and synaptotagmin, suggesting a potential mechanism for the observed effects on the late stages of exocytosis. Intriguingly, treatment of chromaffin cells with pharmacological activators of PKA (62) produces remarkably similar effects on release kinetics and quantal size to that seen in response to Akt and CSP overexpression. This suggests a simple explanation whereby both kinases act via a single effector, CSP, on a common phosphorylation site, serine-10. This notion is supported by our observation here that both PKA and Akt phosphorylate serine-10 in vitro. Furthermore, Akt-transfected chromaffin cells exhibit both increased phosphorylation of endogenous CSP on serine-10 and increased quantal size, suggesting that Akt regulates exocytotic release via CSP phosphorylation. It may be that some redundancy may exist between Akt and PKA in regulating CSP phosphorylation and exocytosis, but that distinct extracellular signals may favour one pathway over another. Nevertheless, the high level of constitutive CSP phosphorylation evident under basal conditions is likely to be due to the tonic Akt activity observed here. CSP is expressed in all cell types in which Akt has been shown to modulate exocytosis, so it is tempting to speculate that CSP may be a general Akt effector in these and possibly other cell types, including neurons. Changes in neurotransmitter release kinetics and/or quantal size have been suggested to underlie forms of synaptic plasticity such as long-term potentiation and long-term depression (63-65). Akt is highly expressed in brain and in subcellular fractionation a substantial proportion of neuronal Akt co-migrates with synaptic marker proteins (66). Furthermore, CSP phosphorylation on serine-10 occurs in a variety of central synapses (33). It is possible that the regulation of presynaptic Akt activity may act via CSP phosphorylation to fine-tune the kinetics of neurotransmitter release and so impact on synaptic plasticity.
We thank Nick Dolman and Alexei Tepikin for expert assistance with confocal microscopy, Tim Craig for PKC-phosphorylated Munc18-1, and Miles Houslay for helpful suggestions. This work was funded by research grants from the UK Medical Research Council to AM, by grant R01 DK56886 from NIH to MJB and from the Wellcome Trust to RDB. GRP is supported by a Wellcome Trust Prize Studentship.