Stimulation of numerous G protein-coupled receptors leads to the elevation of intracellular concentrations of cAMP, which subsequently activates the PKA pathway. Specificity of the PKA signaling module is determined by a sophisticated subcellular targeting network that directs the spatiotemporal activation of the kinase. This specific compartmentalization mechanism occurs through high-affinity interactions of PKA with A-kinase anchoring proteins (AKAPs), the role of which is to target the kinase to discrete subcellular microdomains. Recently, a peptide designated “AKAPis” has been proposed to competitively inhibit PKA-AKAP interactions in vitro. We therefore sought to characterize a cell-permeable construct of the AKAPis inhibitor and use it as a tool to characterize the impact of PKA compartmentalization by AKAPs. Using insulin-secreting pancreatic β-cells (INS-1 cells), we showed that TAT-AKAPis (at a micromolar range) dose dependently disrupted a significant fraction of endogenous PKA-AKAP interactions. Immunoflurescent analysis also indicated that TAT-AKAPis significantly affected PKA subcellular localization. Furthermore, TAT-AKAPis markedly attenuated glucagon-induced phosphorylations of p44/p42 MAPKs and cAMP response element binding protein, which are downstream effectors of PKA. In parallel, TAT-AKAPis dose dependently inhibited the glucagon-induced potentiation of insulin release. Therefore, AKAP-mediated subcellular compartmentalization of PKA represents a key mechanism for PKA-dependent phosphorylation events and potentiation of insulin secretion in intact pancreatic β-cells. More interestingly, our data highlight the effectiveness of the cell-permeable peptide-mediated approach to monitoring in cellulo PKA-AKAP interactions and delineating PKA-dependent phosphorylation events underlying specific cellular responses.
- subcellular compartmentalization
- signal transduction
- pancreatic β-cells
- insulin exocytosis
- protein kinase A
- A-kinase achoring protein
cellular responses to a wide variety of extracellular stimuli involve signal transduction mechanisms, which invariably rely on the interplay between intracellular phosphatases and kinases, the latter being regulated by fluctuations in the levels of second messengers (10, 51). The PKA signaling module, acting downstream from G protein-coupled receptors via the activation of adenylyl cyclases, is one of the best-characterized kinase systems. Indeed, this signaling pathway served during the last 50 years as the paradigm for signal transduction research (4). However, a key advance in the field was the discovery of a fundamental aspect in the mode of action of the PKA module consisting in the tight regulation of PKA functions via subcellular localization of the kinase through physical interactions with A-kinase anchoring proteins (AKAPs) (9, 47). According to this compartmentalization model, it is well established that AKAPs tether PKA holoenzyme pools (mainly type II PKA tetramers) and localize them at discrete subcellular microdomains in close proximity to kinase-specific downstream substrates (9, 44, 52). AKAPs, which now represent a large family of functionally related proteins, orchestrate the spatial organization pattern of PKA signaling within the cell, thereby ensuring a precise coordination of PKA reversible phosphorylation events within defined subcellular compartments. To this end, each AKAP isoform contains within its sequence a targeting signal that serves to localize PKA-AKAP complexes to discrete and specific subcellular compartments: membrane, mitochondrial, nuclear, and several other subcellular locations (11). In addition to this targeting signal, and despite their diversity, all AKAPs share a conserved amphipathic helix motif that binds to a hydrophobic groove formed by the dimerization domain of regulatory subunit (R) dimers of the PKA holoenzyme (9, 40). This interaction provides the basis for high-affinity binding of AKAPs to PKA in situ. Peptides mimicking the amphipathic helix domain of AKAPs, such as the Ht31 inhibitor peptide, which is derived from the PKA-binding domain of a human thyroid AKAP (AKAP-Lbc), were originally shown to competitively disrupt PKA binding to AKAPs in vitro (8, 9). Microinjection of such anchoring inhibitor peptides into neuronal or skeletal muscle cells alters, respectively, PKA potentiation of glutamate receptor channels (43) and voltage-gated Ca2+ channels (30) by disrupting PKA pools from their AKAP-directed locations. Development of a modified Ht31 peptide, covalently linked to a lipid carrier, has significantly helped to understand the cellular roles of AKAP-mediated PKA targeting and associated molecular mechanisms (54). Interestingly, a recent analysis of the RII binding domain of AKAPs, using bioinformatics and peptide arrays, led to the design of AKAPis, a peptide with high-affinity binding and specificity for RII PKA and a higher dissociation effect on PKA-AKAP anchoring than Ht31 (1). Owing to this potency, development of a cell-permeable peptide (CPP) analog of AKAPis is also of great interest, not only in terms of effective specificity for type II PKA but also to further explore AKAP-mediated signaling mechanisms on endogenous substrates and effectors in intact cells (33). To this end, the human immunodefiency virus (HIV)-1-derived TAT peptide has been extensively used in the protein transduction area since the discovery of its transmembrane transport properties (17, 21). Indeed, like some other CPPs, when covalently linked to virtually any compound (peptides, nucleic acids, proteins, or even small iron beads), the TAT peptide can efficiently transport them into almost all cell types in a receptor- and transporter-independent manner (45, 46, 55). Here, we designed a cell-permeable analog for AKAPis, named TAT-AKAPis, and used it to analyze AKAPs functional targeting of endogenous RII PKA in intact cells. We focused on a pancreatic β-cell line (INS-1) and insulin secretion since both represent appropriate models to investigate cAMP/PKA-dependent potentiation mechanisms. We showed that the TAT-AKAPis construct, when delivered into cells, is an effective means for monitoring PKA-AKAP disruption and for analyzing the resulting effects on cAMP/PKA-dependent amplification of signal transduction.
MATERIALS AND METHODS
Chemistry and peptide synthesis.
TAT (GRKKRRQRRR) and AKAPis (QIEYLAKQIVDNAIQQA) peptides were synthesized manually using Boc-HF solid-phase chemistry as previously described (35), with slight modifications including 4-methylbenzhydrylamine hydrochloride salt resin as the solid support (0.59 meq/g, Novabiochem). In both peptides, a cysteine residue was added for subsequent disulfide bonding. The TAT sequence was initiated at its COOH terminus with a cysteine residue (TAT-Cys). In AKAPis, the extra cysteine was added at the NH2 terminus (Cys-AKAPis). Upon completion of the stepwise elongation and classical workup, TAT-Cys and Cys-AKAPis were purified by preparative reverse-phase HPLC. TAT-Cys was then dissolved in dimethylacetamide and treated with 2,2′-dithiobis(5-nitropyridine), an activating reagent for the thiol function of cysteine, to facilitate heterodimeric peptide conjugation by the introduction of the 5-nitro-2-pyridinesufenyl (Npys) group (41). The resulting activated peptide [TAT-Cys(Npys)] was HPLC purified, lyophilized, and then allowed to react with Cys-AKAPis. TAT-Cys(Npys) (2 mg, 120 nmol) and Cys-AKAPis (3 mg, 145 nmol, 1.2 excess) were dissolved in CH3CN/H2O [50:50 (vol/vol)] under mild stirring, and the reaction was monitored by analytical HPLC. Complete disappearance of TAT-Cys(Npys) was observed after 6 h. The expected heterodimeric peptide TAT-(S-S)-AKAPis was finally obtained after HPLC purification (purity > 99%) and lyophilization (yield 65%). The TAT-AKAPis peptide was dissolved in DMSO-water [10% (vol/vol)] before use in the subsequent experiments. In some control experiments, where an inactive form of TAT-AKAPis (DTT-inactivated TAT-AKAPis) was used, the TAT-AKAPis peptide preparation was first reduced in 10 mM DTT for 1 h at 37°C before it was added to INS-1 cells. All peptides were analyzed, and their integrity was confirmed by electrospray ionization mass spectrometry.
The insulin-secreting INS-1 cells used in this study (a kind gift from Prof. C.B. Wollheim) were cultured as previously described (3). INS-1 cells were regularly checked for their physiological glucose responsiveness by measuring insulin release in response to different glucose concentrations (2.8, 5.6, 8.3, and 11.2 mM; data not shown). Only cell cultures retaining normal glucose responsiveness were used in functional experiments.
Immunoanalysis of the TAT-AKAPis disrupting effect on endogenous PKA-AKAP complexes.
INS-1 cells grown to subconfluency in 35-mm culture dishes were treated with or without TAT-AKAPis peptide, lysed, and then examined for protein-protein interactions between PKA and AKAPs endogenously expressed by INS-1 cells. Briefly, cells were first starved for 2 h in glucose-free HEPES-balanced Krebs-Ringer bicarbonate (KRB) buffer [which contained (in mM) 119 NaCl, 4 KCl, 1.2 KH2PO4, 1.2 MgSO4, 2.5 CaCl2, and 20 HEPES (pH 7.2), with 0.1% BSA]. TAT-AKAPis peptide solutions were directly added to the incubation medium after the first hour of the starvation period at the indicated concentrations. Cells were starved for an additional hour to allow optimal internalization of the TAT-AKAPis construct. In “vehicle” control experiments, and to check for any artifactual effect due to DMSO (used as a solvent in the TAT-AKAPis peptide solutions), control cells were treated with KRB buffer containing 0.2% DMSO (corresponding to the amount of DMSO delivered with the highest concentration of TAT-AKAPis used). At the end of the starvation/peptide incubation step, cells were harvested, and whole cell lysates were prepared.
To evaluate the effect of TAT-AKAPis in disrupting endogenous PKA-AKAP complexes, INS-1 protein lysates were subjected to immunoprecipitation of AKAP95. Precipitated complexes were then subjected to Western blot analysis to quantify the amount of the PKA RIIα subunit captured with AKAP95 immunocomplexes. Immunoprecipitations were carried out on INS-1 cell whole protein lysates using 2 μg of anti-AKAP95 antibody (Upstate Biotechnology, Lake Placid, NY) on a rotating device at 4°C for 4 h. This step was followed by the addition of protein A/G PLUS-agarose beads (1:25, Santa Cruz Biotechnology, Santa Cruz, CA) for an additional hour. The resulting immunocomplexes were sedimented by centrifugation and washed three times with ice-cold PBS. Samples were then dissolved with 20 μl SDS extraction buffer, separated by SDS-PAGE, transferred to nitrocellulose membranes, and analyzed by immunoblot analysis using a rabbit anti-RIIα PKA antibody (Chemicon, Temecula, CA). To correct for any variations due to sample preparations, fractions from the same INS-1 protein lysates used for immunoprecipitation (50 μg) were resolved by SDS-PAGE and analyzed for total RIIα PKA content using the rabbit anti-RIIα PKA antibody. Blot membranes were then incubated with a horseradish peroxidase-linked secondary antibody followed by enhanced chemiluminescence detection. Quantification of protein bands was performed using a Kodak imaging-analysis station 2000 System and software (Kodak, Rochester, NY). The extent of interaction between PKA and AKAP95 was expressed as percentage ratios of RIIα detected on AKAP95-immunoprecipitate blots relative to total RIIα detected in input protein lysates.
Immunofluorescence analysis of PKA localization.
INS-1 cells were seeded on a poly-l-lysine (Sigma Aldrich)-coated Lab-Tek Chamber Slide System. At subconfluency, cells were starved in glucose-free medium for 2 h. During the second hour of starvation, cells were treated with or without 10 μM TAT-AKAPis peptide (either intact or inactivated by DTT treatment). After fixation with 2% paraformaldehyde for 20 min, cells were permeabilized for 5 min in 0.1% Triton X-100. Subsequently, and after being blocked in 2% BSA, cells were incubated with rabbit polyclonal antibodies to either RIIα PKA (1:100, Santa Cruz Biotechnology) or AKAP95 (1:50, Upstate Biotechnology, Lake Placid, NY). After several washes followed by nuclear staining with 4′,6-diamidino-2-phenylindole (DAPI; Sigma Aldrich), slides were then incubated for 1 h with a Texas red-conjugated anti-rabbit antibody (1:100, Vector Laboratories, Burlingame, CA). After three additional washes, cells were mounted in Citifluor (Citifluor, Canterbury, UK) and observed using a Bio-Rad MRC 1024 confocal microscope. Confocal images were taken using Image software. The specificity of the immune staining was tested by incubating cells only with the secondary Texas red-conjugated anti-rabbit antibody, and no background signal was observed (not shown).
Analysis of cAMP-dependent phosphorylation of p44/p42 MAPKs and cAMP response element binding protein in INS-1 cells.
To evaluate the impact of disruption of PKA-AKAP interactions on the phosphorylation of PKA downstream targets, glucagon-stimulated INS-1 cells were analyzed for the phosphorylation of p44/p42 MAPKs and cAMP response element binding protein (CREB) in the presence or absence of TAT-AKAPis.
Time courses were first determined for p44/p42 MAPK and CREB phosphorylations after cAMP stimulation elicited by glucagon. Maximal stimulation time points obtained for p44/p42 MAPKs and CREB, respectively, were thereafter considered to evaluate the impact of PKA-AKAP complex disruption by the TAT-AKAPis peptide. In these experiments, cells were starved for 2 h in glucose- and serum-free KRB medium, and TAT-AKAPis peptide was then added during the second hour of the starvation period at the indicated concentrations where appropriate. INS-1 cells were subsequently washed and stimulated with glucagon (10 nM) for the indicated time periods. After glucagon stimulation, INS-1 cells were harvested, and whole cell lysates (40–60 μg) were subjected to SDS-PAGE and Western blot analyses of p44/p42 MAPK and CREB phosphorylation as previously described (12). To correct for protein loading/transfer variations, blots were stripped and reblotted with either the anti-RIIα PKA antibody for p44/p42 MAPK blots or an anti-total CREB antibody for phospho-CREB blots.
Analysis of TAT-AKAPis effects on glucagon-dependent potentiation of glucose-stimulated insulin secretion.
Insulin secretion assays were performed in INS-1 cells to evaluate the effect of disruption of endogenous PKA-AKAP interactions by TAT-AKAPis peptide. Briefly, INS-1 cells were seeded on 24-well plates (105 cells/well). On the day of the experiment, INS-1 cells were starved in glucose-free KRB buffer containing 0.1% (wt/vol) BSA for 1 h, at the end of which TAT-AKAPis was added to the starvation medium at the indicated concentrations. Cells were then starved for an additional hour. After this 2-h starvation step, cells were washed twice with glucose-free KRB buffer and then incubated in KRB medium containing either nonstimulating (2.8 mM) or stimulating (8.3 mM) glucose concentrations with or without glucagon (10 nM) as indicated. To evaluate the impact of PKA anchoring by AKAPs on insulin secretion, INS-1 cells were stimulated for 1 h with glucose alone or combined with glucagon in the absence or presence of increasing concentrations of TAT-AKAPis peptide as indicated. At the end of the stimulation, incubation media were collected and spun at 1,000 rpm for 10 min (to discard any detached cells), and the resulting supernatants were stored at −20°C until insulin was quantified. Total cellular insulin contents were also extracted from each individual culture well using acid-ethanol buffer overnight at 4°C. Insulin concentrations in the samples were determined using a homogeneous time-resolved fluorescence insulin kit (CisBio, Marcoule, France). Immunoreactive insulin was then counted on a Rubystar microtiter plate reader (BMG Labtech, Offenburg, Germany). Insulin release was expressed as the percent ratio of secreted insulin relative to total insulin content.
Results are expressed as means ± SE. Differences between groups were compared using the unpaired Student's t-test included in the SPSS software package (version 11.5, SPSS, Chicago, IL).
TAT-AKAPis peptide construct synthesis.
As shown in Fig. 1, TAT and AKAPis peptides were synthesized separately. We anticipated a system in which the TAT carrier peptide delivers the AKAPis inhibitor peptide into the cell, where AKAPis could antagonize endogenous PKA-AKAP interactions. Therefore, we chemically conjugated the TAT peptide (the cell-permeable carrier) to the AKAPis peptide (the PKA-AKAP anchoring inhibitor) through a disulfide link (Fig. 1). In this way, whatever the mechanism of TAT peptide entry into the cell, it was expected that intracellular glutathione would reduce the disulfide bond (22), subsequently releasing free AKAPis molecules within the cell. This reducible linkage was preferred and adopted for various reasons. First, it allows the independent synthesis and purification of both peptides before quantitative heterodimer formation with highly purified peptides (∼99%). Second, it allows the intracellular delivery of free AKAPis peptide molecules, thus avoiding any problem of inappropriate localization of the TAT-AKAPis chimera and/or possible steric hindrance on the dissociation of the PKA-AKAP complex due to the presence of the TAT peptide. Despite the high hydrophilic nature of the TAT peptide, the TAT-AKAPis peptide construct was found to be highly hydrophobic and was efficiently solubilized in 10% (vol/vol) DMSO-water.
Before functional experiments, and to verify whether INS-1 cells could be transduced with the TAT peptide, INS-1 cells were incubated with fluorescein-labeled TAT (range: 1–20 μM). We observed efficient and time-dependent uptake of the peptide into the cells, reaching maximal intracellular accumulation within ∼60 min (E. Vives and E. H. Hani, unpublished observations). Furthermore, the entire cell population was found to be transduced, as it was unambiguously demonstrated by flow cytometry on different cell types, showing the homogeneity of the transduction process mediated by the TAT peptide (15, 39, 49).
Disruption of endogenous PKA-AKAP complexes by TAT-AKAPis.
To check whether the chimeric TAT-AKAPis peptide was effective in disrupting PKA-AKAP complexes inside the cell, we treated INS-1 cells with increasing concentrations of TAT-AKAPis, lysed them to immunoprecipitate PKA-AKAP complexes (using an anti-AKAP directed antibody), and then quantified the relative amounts of RIIα PKA cocaptured with the immunoprecipitates.
Several AKAPs are expressed by rat INS-1 pancreatic β-cells, including AKAP95, AKAP150, AKAP220, and the γ-isoform of AKAP18 as well as MAP2B (E. H. Hani, unpublished observations). We focused on AKAP95 as a model to study the disrupting effect of TAT-AKAPis on the physical interaction between PKA and INS-1 endogenously expressed AKAPs since AKAP95 is a typical high-affinity PKA-binding AKAP, with a 10-fold higher affinity for RIIα than RIIβ PKA subunits and no detectable binding to RI PKA isoforms (24). In coimmunoprecipitation assays, we observed a significant interaction between endogenous RIIα PKA and AKAP95 proteins in the absence of TAT-AKAPis disruptor peptide (Fig. 2A). Preincubation of INS-1 cells with increasing concentrations (0.1–20 μM) of TAT-AKAPis resulted in a dose-dependent disruption of INS-1 PKA-AKAP95 interactions (Fig. 2, A and B). At the highest TAT-AKAPis concentration tested (20 μM), nearly 60% inhibition of the interaction between RIIα PKA and AKAP95 was observed (Fig. 2B). In agreement with this finding, a similar dose-dependent disanchoring effect of TAT-AKAPis on PKA-AKAP79 complexes was observed in intact MIN6 cells, another pancreatic β-cell line of mouse origin (O. M. Faruque and E. H. Hani, unpublished observations).
We noted that the 10 μM concentration point of TAT-AKAPis did not perfectly fit the overall dose-dependent effect observed in our experiments on RIIα PKA-AKAP95 interactions (Fig. 2B). In this regard, some concentration effects have been suggested, since, e.g., high CPP concentrations (>10 μM) can lead to an energy-independent internalization (15, 53). Therefore, we cannot exclude that at some particular concentration of CPP, different transport mechanisms could be active and could lead to a more or less marked cellular response.
To check for any undesirable effect due to DMSO used as a vehicle for TAT-AKAPis peptide solutions, INS-1 cells were treated with an amount of DMSO equivalent to the one delivered with 20 μM TAT-AKAPis (the highest concentration used in the disruption experiments). We found that DMSO alone had no significant effect on the endogenous RIIα PKA-AKAP95 interaction (Fig. 2, A and B).
As a result, AKAPis, when rendered cell permeable using TAT conjugation, is able to dose dependently inhibit intracellular PKA-AKAP molecular interactions.
Immunofluorescence analysis of the TAT-AKAPis effect on PKA subcellular localization.
To confirm the observed disrupting effect of TAT-AKAPis on PKA-AKAP complexes at the subcellular level, we performed immunofluorescent analyses of INS-1 cells pretreated with or without TAT-AKAPis and monitored RIIα PKA distribution in these cells using a polyclonal anti-RIIα PKA-directed antibody. At the basal state (i.e., no cAMP or glucose stimulation) and in the absence of the TAT-AKAPis inhibitor peptide, RIIα PKA was found to be distributed in the form of marked clusters, mainly concentrated around the nucleus of INS-1 cells (Fig. 3A), with apparently no detectable nuclear staining. In contrast, when cells were treated with 10 μM TAT-AKAPis, RIIα PKA staining was found to be significantly altered. Indeed, in these cells, distribution of RIIα PKA was found to be more diffuse, with less marked clusters. In addition, some nuclear staining, not seen in control cells, was also observed in TAT-AKAPis-treated cells (Fig. 3B). As a control, DTT-inactivated TAT-AKAPis (10 μM) treatment of INS-1 cells had no effect on RIIα PKA localization (Fig. 3C), which displayed a similar staining pattern as untreated cells. DAPI nuclear labeling of cells stained with RIIα PKA antibody is also shown in Fig. 3, D–F. As an internal control, we also performed immunofluorescence analysis of AKAP95, which is known to display a nuclear localization (Fig. 3, G–L). In the same experimental conditions as for PKA, no changes in nuclear staining of AKAP95 could be observed in TAT-AKAPis peptide-treated cells (Fig. 3H) compared with untreated or DTT-inactivated TAT-AKAPis control cells (Fig. 3, G and I). DAPI nuclear labeling of cells stained with AKAP95 antibody is also shown in Fig. 3, J–L.
Therefore, and in agreement with the PKA-AKAP molecular cointeraction data drawn from coimmunopecipitations assays, TAT-AKAPis seems to be efficient in displacing PKA from its subcellular locations.
Analysis of cAMP-dependent phosphorylation of INS-1 endogenous p44/p42 MAPKs and CREB.
We have previously shown that, under glucagon stimulation, cAMP-dependent phosphorylation and activation of p44/p42 MAPKs and CREB in pancreatic β-cells are PKA-dependent events (12). To assess the impact of PKA dissociation from AKAPs on the phosphorylation of these effectors, we first characterized the kinetics of p44/p42 MAPK and CREB phosphorylation after glucagon stimulation of INS-1 cells to identify the maximal stimulation time points for their respective phosphorylations (Figs. 4, A and B, and 5, A and B). Next, we performed inhibition assays using TAT-AKAPis to evaluate whether cAMP-dependent phosphorylation of both effectors after glucagon stimulation involves AKAP-mediated compartmentalization of PKA pools (Figs. 4 and 5).
In the absence of glucagon, a low (basal) phosphorylation level of p44/p42 MAPKs was observed in INS-1 cells extracts (Fig. 4, A and B). In contrast, we found that glucagon stimulation of INS-1 cells resulted in a rapid and transient phosphorylation of both p44 and p42 MAPKs (Fig. 4, A and B), reaching maximal levels 2 min after glucagon administration (3- and 1.5-fold over basal levels for p44 and p42 MAPKs, respectively). The 2-min time point was subsequently considered to evaluate the effect of TAT-AKAPis on p44/p42 MAPK phosphorylation. We therefore investigated whether the phosphorylation of p44/p42 MAPKs after glucagon stimulation involved PKA anchoring by AKAPs. We found that TAT-AKAPis at concentrations of 5 and 10 μM attenuated glucagon-stimulated phosphorylation of p44 MAPK by nearly 40% and 50%, respectively (Fig. 4, C and D). Similarly, we observed that 5 and 10 μM TAT-AKAPis decreased glucagon-stimulated phosphorylation of p42 MAPK by ∼55% and 80%, respectively (Fig. 4, C and D).
Next, we studied in glucagon-stimulated INS-1 cells the phosphorylation pattern of CREB, a typical PKA phosphorylation target, and evaluated, using TAT-AKAPis peptide, the possible implication of PKA anchoring by AKAPs in CREB phosphorylation. As for p44/p42 MAPKs, INS-1 cells exhibited a relatively low (basal) phosphorylation level of CREB in the absence of glucagon stimulation (Fig. 5, A and B). In contrast, treatment of INS-1 cells with glucagon elicited a transient increase in the phosphorylation level of CREB at Ser133, which peaked to a 2.5-fold increase over the basal level at 5 min after stimulation and returned to the basal level by 20 min (Fig. 5, A and B). The 5-min time point was thereafter considered to evaluate the effect of TAT-AKAPis on CREB phosphorylation. We found that TAT-AKAPis peptide resulted in a dose-dependent reduction of CREB phosphorylation at Ser133 in INS-1 cells: CREB phosphorylation was reduced by 35% and 45% when 5 and 10 μM of TAT-AKAPis peptide were used, respectively (Fig. 5, C and D).
Taken together, these results indicate that cAMP-stimulated phosphorylations of p44/p42 MAPKs and CREB after glucagon stimulation of INS-1 cells are likely to depend on AKAP-mediated tethering of PKA pools.
Analysis of cAMP-dependent potentiation of glucose-induced insulin secretion after glucagon stimulation.
Insulin secretion from pancreatic β-cells represents a prototypical cellular model for cAMP-dependent potentiation. Indeed, glucose is the main physiological signal that triggers insulin release, whereas hormones stimulating cAMP and PKA act as potentiating factors with crucial functions in the glycemic control (23, 32). We therefore focused on insulin secretion potentiation to evaluate whether TAT-AKAPis might be useful to delineate this cAMP-dependent physiological response and assess the role of PKA anchoring in the regulation of insulin secretion. As shown in Fig. 6A, a low glucose concentration (2.8 mM) induced only a minor insulin secretory response from INS-1 cells, corresponding to the basal insulin release (∼1% of the total insulin content). In contrast, a stimulatory glucose level (8.3 mM) resulted in a nearly fourfold increase in insulin secretion over the basal level. This secretory profile recapitulates the physiological stimulation of insulin release upon glucose stimulation. When glucagon (10 nM) was combined with a stimulatory glucose concentration, insulin secretion was increased to >1.5-fold over the 8.3 mM glucose stimulation (Fig. 6A). As expected, glucagon elicited a clear potentiation of insulin release in INS-1 cells. When increasing amounts of TAT-AKAPis were included in the assays, we observed a dose-dependent inhibition of the glucagon-induced potentiation of insulin release (in the range of 5–20 μM of TAT-AKAPis competitor peptide; Fig. 6A). In the presence of 20 μM TAT-AKAPis, glucagon potentiation of the glucose-induced insulin release was lowered by ∼80% (Fig. 6A). To exclude any unwanted effect attributable to the TAT peptide itself in disrupting PKA from INS-1-expressed AKAPs, and the consequences of such interference on insulin secretion, we carried out control experiments using inactivated TAT-AKAPis. Accordingly, when TAT-AKAPis was inactivated using DTT before its application to INS-1 cells, no significant inhibitory effect was noted on the glucagon-induced potentiation of insulin secretion (Fig. 6B). It is noteworthy that TAT-AKAPis (20 μM) had no effect on the insulin secretion response stimulated by glucose (8.3 mM) alone.
Accordingly, AKAP-mediated anchoring of PKA is necessary for the potentiation of insulin secretion from pancreatic β-cells by cAMP.
In a variety of cell types, numerous cellular cAMP-dependent processes are closely controlled through the interaction of intracellular PKA pools with distinct functional AKAPs (for reviews, see Refs. 52 and 56). In this regard and in addition to the essential role of PKA in potentiating pancreatic β-cell responses (23, 32), PKA-AKAP interactions have been documented to represent a prerequisite for cAMP-mediated insulin exocytosis (18, 36, 37). One way to approach these interactions and to elucidate their physiological relevance is the use of disrupting competitors of endogenous PKA-AKAP complexes, such as the Ht31 peptide, which has served now for several years to study PKA subcellular compartmentalization (8, 9, 54).
Here, we have successfully characterized an original hybrid peptide construct consisting of the recently described and highly potent PKA-AKAP disrupting peptide, AKAPis (1), combined with the well-known CPP, TAT peptide, derived from the HIV-1 TAT protein (17, 21).
We focused on pancreatic β-cells, which are, like several other cell types, highly regulated by cAMP and PKA. These regulations by the cAMP/PKA pathway occur at various subcellular and molecular sites, a number of which are essential for pancreatic β-cell physiology (23, 48). These include K+ and Ca2+ channels (5, 57) and components of the insulin granule exocytosis pathway (16) as well as the transcriptional machinery, among which CREB represents a major regulator of pancreatic β-cell plasticity (28). Using this cell model, we provide evidence that TAT-AKAPis is a potent disruptor of anchored type IIα PKA pools in situ and show that, through endogenous displacement of PKA from AKAPs by the TAT-AKAPis disruptor, the cellular implications of type II PKA signaling can be delineated at different levels. Indeed, the application of CPP TAT-AKAPis to pancreatic β-cells dramatically reduced in situ physical interactions of PKA to AKAPs in a concentration-dependent manner, as ascertained by specific dissociation of a high-affinity AKAP (namely, AKAP95) from PKA pools in INS-1 cells. Since AKAP95 is one of the PKA high-affinity binding AKAPs (24), it is expected that other expressed AKAPs (with similar or lower affinity to PKA) might be displaced from kinase holoenzyme pools as well. In addition, the localization of PKA in β-cells after treatment with TAT-AKAPis peptide was significantly altered. Indeed, the typically perinuclear and clustered localization of PKA in control cells, in agreement with another previous report (37), is rendered diffuse throughout the cytoplasm upon TAT-AKAPis peptide treatment. Moreover, and probably consequently, TAT-AKAPis significantly reduced glucagon-induced phosphorylation of some endogenous PKA effectors, and this allowed us to demonstrate a crucial role of PKA subcellular targeting in the cAMP-mediated potentiation of insulin secretion. This is in full agreement with the hypothesis that anchoring inhibitors can disrupt AKAP targeting of PKA holoenzyme pools at specific subcellular locations (44, 52, 56), thereby altering normal cAMP-dependent cellular responses. Use of CPP-based approaches may help to specifically identify endogenous AKAP-mediated phosphorylation events and the associated cellular responses in a given cell type. These phosphorylation events are exemplified in the present work by p44/p42 MAPKs and CREB, for which the respective activation mechanisms are known to depend on PKA activity (12, 20, 29).
The affinity of AKAPis for PKA is described as being five times greater than Ht31 peptide and highly more specific for the RII isoforms of PKA than for RI isoforms (1). Thus, the TAT-AKAPis construct has definite advantages in the context of in vivo molecular dissection of cAMP signaling events that rely on AKAP-directed subcellular compartmentalization. On the other hand, it is well established that AKAPs play, through their diversified structural domains, essential roles in organizing multimolecular complexes and, subsequently, essential functions in integrating multiple signal transduction inputs and outputs (13, 14, 58). Therefore, it is likely that CPP versions of anchoring inhibitors such as TAT-AKAPis will be of importance in investigating endogenous molecular assemblies involving AKAP-mediated compartmentalization.
Accordingly, we show here that the CPP analog of PKA-AKAP can efficiently translocate and act inside cells and may represent an attractive alternative approach over existing gene-based transfection or microinjection techniques (31, 33) to study in intact cells PKA-AKAP molecular assemblies of specific relevance for a given cellular response. To date, and in this regard, different methods have been used to introduce PKA-AKAP disruptor peptides into living cells, with successful results in demonstrating PKA-AKAP cellular roles. Early studies have successfully used patch pipette-based intracellular microinjection of active PKA-AKAP disruptor peptide (30, 43). Alternative delivery methods, consisting of either plasmid-driven overexpression or direct delivery of PKA-AKAP disruptor peptide using the lipofectamine reagent, have also proven valuable in demonstrating the role of AKAP-mediated PKA anchoring (37). Others have used a modification of the PKA-AKAP disruptor peptide itself by the addition of a stearyl lipid group to the active peptide sequence (54). This method is also known to facilitate the delivery of diverse peptides into intracellular compartments and thus helped in the development of the well-known stearated Ht31, a reagent that has been used to disrupt intracellular PKA-AKAP interactions in a large number of studies. With advantages such as relative lower toxicity, simplicity (in contrast with the case of cells that are difficult to transfect), time considerations (when other methods, such as vector expression, may be laborious) and direct analysis on intact cells, TAT peptide conjugation to PKA-AKAP disruptor peptides represents an alternative and complementary method to other previously described approaches.
Such an approach can be extended to other modulators (inhibitors or activators) of cellular kinases and thus should prove an efficient tool to directly delineate the in vivo functions of expressed kinases and their interplay with other molecular partners. In this regard, a large majority of the AKAPs found in nature are RII specific, whereas others exhibit dual specificity for both RI and RII PKA (25, 26, 42) and a few AKAPs seem to be selective for RI (2, 34, 38). Interestingly, diverse approaches have recently led to the development of more potent and highly isoform-specific anchoring disruptors (6, 7, 19, 27, 50), allowing in the future a greater potential for monitoring discrete subpools of PKA and their anchoring mechanisms with respect to the associated molecular events and cellular responses. The design of cell-permeable peptide versions for these novel anchoring disruptors should greatly enhance the power of future studies on the mechanisms and functions of anchored type I and II PKA pools in vivo (33, 52) and certainly allow a better knowledge of regulatory mechanisms of signal transduction biology.
This work was funded in part by the Centre National de la Recherche Scientifique, the University of Montpellier, and the Association de Recherche sur le Diabète via the Institut National de la Santé et de la Recherche Médicale/Programme National de la Recherche sur le Diabète Program. We also thank the French Ministry of Foreign Affairs (through the Egide Foundation) for the unrestricted support to O. M. Faruque during his PhD training. We are especially grateful to the European Foundation for the Study of Diabetes/European Association for the Study of Diabetes for the appreciable support at the initiation of this research program, through an Albert Renold Career Development Award (to E. H. Hani).
The authors thank Dr. René Gross for critical reading of the manuscript, Dr. Sharon Lynn Salhi for presubmission editorial assistance, and Dr. Stéphane Dalle for helpful discussions.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2009 the American Physiological Society