|Home | About | Journals | Submit | Contact Us | Français|
Protein kinase A anchoring proteins (AKAPs) provide the backbone for targeted multimolecular signaling complexes that serve to localize the activities of cAMP. Evidence is accumulating of direct associations between AKAPs and specific adenylyl cyclase (AC) isoforms to facilitate the actions of protein kinase A on cAMP production. It happens that some of the AC isoforms (AC1 and AC5/6) that bind specific AKAPs are regulated by submicromolar shifts in intracellular Ca2+. However, whether AKAPs play a role in the control of AC activity by Ca2+ is unknown. Using a combination of co-immunoprecipitation and high resolution live cell imaging techniques, we reveal an association of the Ca2+-stimulable AC8 with AKAP79/150 that limits the sensitivity of AC8 to intracellular Ca2+ events. This functional interaction between AKAP79/150 and AC8 was observed in HEK293 cells overexpressing the two signaling molecules. Similar findings were made in pancreatic insulin-secreting cells and cultured hippocampal neurons that endogenously express AKAP79/150 and AC8, which suggests important physiological implications for this protein-protein interaction with respect to Ca2+-stimulated cAMP production.
Multimolecular signaling complexes organized around protein kinase A anchoring proteins (AKAPs)2 provide a focal point for the activities of cAMP-dependent protein kinase (PKA) and other signaling molecules to produce spatially and temporally discrete cAMP-dependent events. Such localized action of cAMP is believed to account for the diverse physiological effects of this ubiquitous messenger within the cell (1, 2). The AKAPs form a large family of scaffold proteins that preferentially target PKA to discrete regions of the cell (e.g. plasma membrane, Golgi, or mitochondria) (1, 3). In addition, AKAPs associate with a range of other signaling proteins, including protein kinase C, Epac (exchange protein directly activated by cAMP), phosphatases (e.g. PP2B), and cAMP-dependent phosphodiesterases (e.g. PDE4) as well as a wide range of specific ion channel and receptor subtypes (1, 4).
Recent studies have revealed that specific adenylyl cyclase (AC) isoforms can also form an integral part of AKAP-based signaling complexes. In the first demonstration of this kind, AKAP79 was shown to directly interact with AC5 and AC6 to facilitate PKA-mediated inhibition of cAMP production (5). More recently, the plasma membrane-targeted AKAP, Yotiao (a splice variant of AKAP9), was found to directly interact with AC1 and AC2 (6). Interactions with Yotiao inhibited AC2 activity, but the functional consequence of the association between Yotiao and AC1 was not apparent. In a separate study, the muscle-specific AKAP, mAKAPβ, was found to interact with and inhibit the activity of AC5 to provide an additional role for mAKAPβ in the regulation of hypertrophy (7).
Further compartmentalization of cAMP signals within the cell is achieved by the Ca2+-dependent regulation of specific AC isoforms. Activities of four of the nine membrane-bound AC isoforms are sensitive to submicromolar changes in local [Ca2+]. AC1 and AC8 are stimulated by Ca2+/calmodulin and, in vivo, are highly selective for Ca2+ increases arising from store-mediated Ca2+ entry or voltage-gated Ca2+ channel activity. AC5 and AC6 display similar selectivity for the source of Ca2+-rise, but they are directly inhibited by Ca2+ (8). Previous work from our laboratory has shown that the largest, most dynamic changes in intracellular [cAMP] arise in subcellular domains close to sites of Ca2+ entry (9,–11). Furthermore, cAMP levels can oscillate as a function of agonist-evoked intracellular Ca2+ transients (12,–14). Interestingly, a number of the Ca2+-regulated ACs are among those isoforms that are reported to interact with AKAPs (e.g. the interaction of Ca2+-inhibitable AC5 with AKAP79 (5) or mAKAPβ (7) and interaction of Ca2+-stimulated AC1 with Yotiao (6)). Nevertheless, the functional role of AKAP association with respect to Ca2+ regulation of the ACs has not been examined. Because these enzymes are predominantly regulated by local Ca2+ changes and a number of AKAP species have been shown to interact with Ca2+ regulatory proteins (e.g. L-type Ca2+ channels or inositol 1,4,5-trisphosphate receptor (15, 16)), an additional function of the AKAPs to localize specific AC isoforms, not only with their effector molecules but also with sites of Ca2+ entry, could be envisaged.
In the present study, we have investigated whether AC8, a well characterized Ca2+-stimulable AC, interacts with an AKAP and if such an interaction might affect Ca2+-dependent AC8 activity. Our initial screen using GST pull-down assays identified AKAP79 and its rodent orthologue AKAP150 as candidate binding partners for AC8. Using a combination of co-immunoprecipitation, FRET analysis of protein-protein interaction, live cell imaging with a high resolution cAMP biosensor, and selected knockdown of AKAP79/150 using shRNA, we were able to establish a defined role for AKAP79/150-AC8 interaction and the regulation of Ca2+-stimulated cAMP production. The functional interaction was observed in HEK293 (human embryonic kidney) cells overexpressing AC8 and AKAP79/150 but also in pancreatic MIN6 (mouse pancreatic β-cell line) insulin-secreting cells and cultured hippocampal neurons that endogenously express AC8 and AKAP79/150. These data provide the first evidence that AC8 interacts with an AKAP and demonstrate a role for AKAP79/150 in the attenuation of Ca2+-dependent AC8 activity in vivo.
AKAP18α-Myc-His and GFP-AKAP149 were gifts from Dr. Enno Klussmann (LMP, Berlin, Germany); Myc-Yotiao was a gift from Dr. Carmen Dessauer (University of Texas, Houston, TX); FLAG-sAKAP84 and FLAG-AKAP-Lbc were gifts from Dr. Hiroshi Ariga (Hokkaido University, Japan); HA-gravin was a gift from Dr. Craig Malbon (SUNY, Buffalo, NY); AKAP150, pSilencer vector, and shRNA AKAP79 were gifts from Prof. John Scott (University of Seattle); WAVE-2 was a gift from Dr. Marc Kirschner (Harvard University, Boston, MA); AKAP79-YFP and AKAP79-CFP were gifts from Dr. Marc Dell'Acqua (University of Colorado, Denver, CO), GST-Ezrin was a gift from Dr. Kjetil Tasken (University of Oslo, Norway); and Epac2-camps was a gift from Dr. Martin Lohse (Würzburg, Germany).
Untagged versions of AKAP150 and WAVE were C-terminally HA-tagged using donated plasmid DNAs as template and subcloned into pcDNA3.1. DNA fragments encoding C-terminally HA-tagged AKAP proteins were generated by PCR. Primer sequences for AKAP150-HA were 5′-GGCGGATCCACCATGAAAGAGTGCAGTGTCAAAATG (forward) and 5′-GGCTCTAGATCAAGCGTAATCTGGTACGTCGTATGGGTACTGGAACAGCGTATTTATTTGATTA (reverse), and primer sequences for WAVE-HA were 5′-GGCGGTACCACCATGCCGTTAGTAACGAGGAAC (forward) and 5′-GGCGGTACCACCATGCCGTTAGTAACGAGGAAC (reverse). AKAP79-HA was generated from AKAP79-YFP. Primer sequences were 5′-GGCGGATCCACCATGGAAACCACAATTTCAGAAATTC (forward) and 5′-GGCGAATTCAAGCGTAATCTGGTACGTCGTATGGGTACTGTAGAAGATTGTTTATTTTATTATC (reverse). pcDNA3.1 AKAP149-HA was generated using the primer sequences 5′-GGCGGATCCACCATGGCAATCCAGTTCCGTTCG (forward) and 5′-GGCGAATTCAAGCGTAATCTGGTACGTCGTATGGGTAAAGGCTTGTGTAGTAGCTGTCTA (reverse). Ezrin-HA was generated from GST-Ezrin using the primer sequences 5′-GGCAAGCTTCCACCATGCCGAAACCAATCAATGTCC (forward) and 5′-GGCGAATTCAAGCGTAATCTGGTACGTCGTATGGGTACAGGGCCTCGAACTCGTCG (reverse).
The generation of DNA fragments encoding GST-AC8 1–179 and GST-AC8 1106–1248 was described previously (17). cDNAs encoding the AC8 fragments 1–77, 73–179, and 582–703 were subcloned into pGEX4T-1 using the EcoRI and SalI (residues 1–77 and 582–703) or the EcoRI and XhoI (residues 73–179) restriction sites to produce GST-tagged proteins. For AC8 1–77, primers were 5′-CGGAATTCATGGAACTCTCGGATGTGCAC (forward) and 5′-ACGCGTCGACTTACGCGTGGTGGTTTGGTCC (reverse). For AC8 73–179, primers were 5′-CGCGAATTCCCAAACCACCACGCGCCG (forward) and 5′-GCGCTCGAGCTACTCCGATTTGCGCCTCTGG (reverse). For AC8 582–703, primers were 5′-GGAATTCGAGACCTATTTGATTAAGCAGC (forward) and 5′-ACGCGTCGACAGTGATAAACAGAAGAACG (reverse). AC8M1(Δ1–106) with a C-terminal HA tag was inserted into pcDNA3 using the KpnI and XbaI restriction sites. Primer sequences were 5′-GGGGTACCGCCACCATGCCGGAACGCAGCGGGAGCGGC (forward) and 5′-GCTCTAGATTATGGCAAATCGGATTTG (reverse).
The original Epac2-camps sensor containing the FRET pair CFP and YFP (18) was modified to generate Ci/C-Epac2-camps, in which the YFP was replaced with citrine. The DNA fragment encoding citrine was generated by PCR using plasmid 12150 DNA as template (19). Primer sequences were 5′-GGCAAGCTTCCACCATGGTGAGCAAGGGCGAGG (forward) and 5′-CGAATTCCTTGTACAGCTCGTCCATGC (reverse). The PCR fragment was then cloned into pcDNA Epac2-camps (YFP/CFP) and digested with HindIII and EcoRI to remove the original YFP sequence.
HEK293 cells (European Collection of Cell Cultures, Porton Down, UK) were grown in minimum essential medium supplemented with 10% (v/v) fetal bovine serum, 100 units/ml penicillin, 100 μg/ml streptomycin, and 2 mm l-glutamine. Mouse insulin-secreting MIN6 β-cells (a gift from Dr. Anders Tengholm (Uppsala, Sweden)) were cultured in Dulbecco's modified Eagle's medium containing 4500 mg/ml glucose, supplemented with 15% fetal bovine serum, 100 units/ml penicillin, 100 μg/ml streptomycin, 2 mm l-glutamine, and 50 μm 2-mercaptoethanol. All cells were maintained at 37 °C in a humidified atmosphere of 5% CO2, 95% air.
To produce stably expressing HEK-AC8 cells, wild-type HEK293 cells were plated on 100-mm dishes at ~60% confluence 1 day prior to transfection with 2 μg of rat AC8, AC8-HA, or 8M1-HA cDNA using the Lipofectamine 2000 transfection method. Two days later, the culture medium was replaced with fresh medium containing 800 μg/ml G-418 disulfate (Formedium Ltd., Hunstanton, UK) to select transfected cells. After selection, cells were maintained in medium containing 400 μg/ml G-418. Stable AC8-expressing HEK293 cells were established from ~2 weeks following transfection.
For single cell cAMP measurements and micro-FRET experiments, MIN6, untransfected HEK293, or stable AC8 (HEK-AC8) cells were plated onto 25-mm poly-l-lysine-coated coverslips at ~60% confluence 24 h prior to transient transfection with 1 μg of total cDNA (0.5 μg each if transfecting two constructs (e.g. Epac2-camps and AKAP)), using the Lipofectamine 2000 method. For in vitro FRET measurements, 10 μg of Epac2-camps or Ci/C-Epac2-camps sensor was transfected into HEK293 cells in 150-mm dishes using Lipofectamine 2000. For GST pull-down and co-immunoprecipitation assays, cells were plated onto 100-mm dishes at ~40% confluence 1 day prior to the transient transfection with 2 μg of AKAP cDNA according to the calcium phosphate method as described previously (20). All experiments were carried out ~48 h post-transfection.
Hippocampal neurons were cultured from newborn Wistar rats as described previously (21) except that the growth medium used was Neurobasal medium (Invitrogen) containing 2% B27 (Invitrogen), 5% fetal calf serum (PAA Laboratories, GmbH, Pasching, Austria), 1 mm l-glutamine, 35 mm glucose (Sigma), 100 units/ml penicillin, and 0.1 mg/ml streptomycin (Invitrogen). 2.4 μm cytosine arabinoside (Sigma) was added to cultures 2–4 days after plating to inhibit proliferation of non-neuronal cells. All cultures were plated on poly-d-lysine-coated 18-mm glass coverslips and maintained at 37 °C in a humidified atmosphere of 95% air and 5% CO2. At 10–12 days in vitro, neurons were transfected with a Ci/C-Epac2-camps sensor using the Lipofectamine 2000 method of transfection. Transfected neurons were imaged 48 h later.
For knockdown of endogenously expressed AKAP150 in primary cultured hippocampal neurons, shRNA AKAP150 lentiviral particles (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) were added directly to cultured cells 72 h prior to experiments. For introduction of shRNA AKAP150 to cultured MIN6 cells, the culture medium was replaced with fresh complete Dulbecco's modified Eagle's medium containing Polybrene (Santa Cruz Biotechnology, Inc.) at a final concentration of 5 μg/ml. shRNA AKAP150 lentiviral particles were added for overnight incubation before replacing the medium once again with complete Dulbecco's modified Eagle's medium. Control scrambled shRNA lentiviral particles were used as a negative control (Santa Cruz Biotechnology, Inc.).
Fusion proteins of GST and AC8 fragments were expressed in BL21 cells at 30 °C following induction with 0.1 mm isopropyl 1-thio-β-d-galactopyranoside. Cells were lysed by sonication (three times for 30 s at 7–10 watts) in lysis buffer (50 mm Tris, pH 8, 20% (w/v) sucrose, 10% (v/v) glycerol, 2 mm MgCl2, 200 μm Na2S2O5, 1 mm dithiothreitol, and protease inhibitors). Homogenates were centrifuged (27,000 × g, 4 °C, 15 min), and supernatant was passed through a glutathione-Sepharose 4B resin by chromatography and washed until no protein remained in the eluate (assessed by measurement of absorbance at 280 nm). An equal volume of phosphate-buffered saline to resin with 0.02% NaN3 was added to create a 50% slurry.
Confluent 100-mm dishes of HEK293 cells transiently expressing the AKAP of interest were lysed in GST-Fish buffer (10% (v/v) glycerol, 100 mm NaCl, 50 mm Tris, pH 7.4; supplemented with 0.5% Tween 20, 100 μm EGTA, 2 mm dithiothreitol, 1 mm phenylmethylsulfonyl fluoride, 1 mm benzamidine, protease inhibitors, 10 mm β-glycerophosphate, and 2 mm sodium orthovanadate) by rotating for 30 min at 4 °C, before centrifugation (16,000 × g, 4 °C, 15 min). The cell lysate was incubated with the appropriate GST-fragment for 4 h at 4 °C with rotation. GST-beads were washed twice in GST-Fish buffer, and bound proteins were eluted by adding an equivalent volume of 2× Laemmli buffer and boiled for 5 min prior to Western blot analysis.
HEK293 cells expressing AC8-HA or co-expressing AC8-HA and CFP or AKAP79-CFP or pancreatic MIN6 cells were washed with phosphate-buffered saline and lysed in C12E9 solubilization buffer (50 mm Tris, pH 7.4, 1 mm EDTA, 1 mm MgCl2, 150 mm NaCl, 0.5% (v/v) C12E9, and protease inhibitors) by repeatedly passing the cell suspension through a 21-gauge needle. The crude cell lysate was then centrifuged (200 × g, 4 °C, 5 min) to remove cellular debris. Lysates were precleared with 20 μl of prewashed protein A/G Plus-agarose beads (50% slurry; Santa Cruz Biotechnology, Inc.) for 1 h at 4 °C, and beads were precipitated by brief centrifugation. The appropriate immunoprecipitating antibody (anti-AKAP150 antibody (Santa Cruz Biotechnology, Inc.), anti-AKAP79 antibody (Millipore, Watford, UK), or anti-HA antibody (Sigma)) or control preimmune serum (normal goat, rabbit, or mouse IgG, respectively) was added to the supernatant and rotated for 1 h at 4 °C. Prewashed protein A/G Plus-agarose bead slurry (60 μl) was then added to each tube, and samples were rotated for 2 h at 4 °C. Beads were washed three times in wash buffer containing 50 mm Tris, pH 7.4, 1 mm EDTA, 1 mm MgCl2, 150 mm NaCl, 0.05% (v/v) C12E9, and protease inhibitors and once with 50 mm Tris, pH 7.4. Bound proteins were eluted by adding an equivalent volume of 2× Laemmli buffer and incubated at 37 °C for 30 min prior to Western blot analysis.
For immunoprecipitation using anti-HA affinity-agarose beads (Roche Applied Science), lysates from HEK293 cells expressing AC8-HA or 8M1-HA were prepared as described above in Nonidet P-40 solubilization buffer (50 mm Tris, pH 7.4, 1 mm EDTA, 150 mm NaCl, 0.3% (v/v) Nonidet P-40, and protease inhibitors). The cell lysates were rotated with 100 μl of prewashed bead slurry (50%) for 4 h at 4 °C. Beads were washed five times in Nonidet P-40 solubilization buffer, and proteins were eluted with Nonidet P-40 solubilization buffer supplemented with 1% (w/v) SDS. Laemmli buffer was added to the elution and incubated at 37 °C for 30 min prior to Western blot analysis.
Proteins were resolved using 8, 10, or 12% SDS-polyacrylamide gels. Separated proteins were then transferred to a supported nitrocellulose membrane. Nitrocellulose membranes were blocked in TBS (20 mm Tris, pH 7.5, 150 mm NaCl) containing 5% (w/v) skimmed milk, for 30 min, followed by three 10-min washes in TBS supplemented with 0.05% (v/v) Tween 20 (TTBS). Membranes were incubated overnight at 4 °C with anti-AC8 antibody (1:1,000; gift from Dr. James Cali (Madison, WI)); anti-HA antibody (1:5,000; Sigma), anti-GFP antibody (1:5,000; Sigma), anti-AKAP79 antibody (1:5,000; Millipore); anti-AKAP149 (1:2,000; BD Biosciences); anti-AKAP150 (1:800; Santa Cruz Biotechnology, Inc.); anti-Myc antibody (1:5,000; Santa Cruz Biotechnology, Inc.), anti-FLAG antibody (1:5,000; Stratagene), or anti-GST antibody (1:40,000; Sigma) in TTBS containing 1% (w/v) skimmed milk. Membranes were washed (3 times for 10 min each) in TTBS and then incubated with goat anti-rabbit IgG conjugated to horseradish peroxidase (1:20,000; for anti-AC8 and anti-AKAP79 antibodies), donkey anti-goat IgG conjugated to horseradish peroxidase (1:5,000; for anti-AKAP150), or goat anti-mouse IgG conjugated to horseradish peroxidase (1:20,000; for anti-HA, anti-FLAG, anti-Myc, anti-GST, and anti-GFP antibodies) in TTBS containing 5% (w/v) skimmed milk for 1 h. Finally, the membrane was washed three times in TTBS (10 min) and once in TBS (10 min), visualized with ECL Plus reagent (GE Healthcare) according to the manufacturer's protocol, and exposed to film. Immunoreactive bands were quantified by densitometry using Image J.
HEK-AC8 cells, MIN6 cells, or cultured hippocampal neurons (previously plated onto 25- or 18-mm coverslips) were loaded with 4 μm Fura-2/AM and 0.02% Pluronic F-127 (Invitrogen) for 35 min at room temperature. For HEK-AC8 cell experiments, extracellular buffered saline contained 140 mm NaCl, 4 mm KCl, 1 mm CaCl2, 0.2 mm MgCl2, 11 mm d-glucose, 10 mm HEPES, pH 7.4. For MIN6 cell experiments, extracellular buffered saline contained 125 mm NaCl, 4.8 mm KCl, 1.3 mm CaCl2, 1.2 mm MgCl2, 3 mm d-glucose, and 25 mm HEPES, pH 7.4. For hippocampal experiments, extracellular buffered saline contained 140 mm NaCl, 5 mm KCl, 5 mm NaHCO3, 2 mm CaCl2, 1 mm MgCl2, 5.5 mm d-glucose, 20 mm HEPES, 1 mm glycine, at pH 7.4. After loading, cells were washed several times and then imaged using an Andor Ixon+ EMCCD camera (Andor, Belfast, UK) and monochromator system (Cairn Research, Kent, UK) attached to a Nikon Eclipse TE2000-S microscope (×40 objective). Emission images (ET510/80M) at 340 and 380 nm excitation were collected at 1 Hz using MetaFluor software (Molecular Devices). For zero calcium buffers, the constituents were the same as for extracellular buffer, but CaCl2 was omitted and replaced by 0.1 mm EGTA.
Fluorescent imaging of Epac2-camps or Ci/C-Epac2-camps expressing HEK293 cells, MIN6 cells, or cultured hippocampal neurons was performed using an Andor Ixon+ EMCCD camera and an Optosplit (505DC) to separate CFP (470 nm) and YFP/citrine (535 nm) emission images (Cairn Research, Kent, UK). For dual emission ratio imaging, cells were excited at 435 nm using a monochromator (Cairn Research) and 51017 filter set (Chroma Technology Corp.) attached to a Nikon eclipse TE2000-S microscope (×40 oil immersion objective). Emission images at 470 and 535 nm were collected every 3 s (250 ms integration time) and analyzed using Metamorph imaging software (Molecular Devices). Cells in which the CFP and YFP/citrine fluorescence intensity was less than twice the background signal were excluded, as were cells with excessive expression of the fluorescent probe. Single cell FRET data were plotted as changes in background-subtracted 470 nm versus 535 nm (CFP/YFP or CFP/citrine) emission ratio relative to maximum FRET ratio change seen with saturating cAMP concentrations.
All FRET images were collected with YFP- and CFP-tagged constructs using an Andor Ixon+ EMCCD camera and an Optosplit (505DC) to separate CFP (470 nm) and YFP (535 nm) emission images. Images were acquired and analyzed according to the three-cube micro-FRET method (22). In brief, a series of live cell images were first collected from cells expressing tagged or untagged CFP or YFP alone (300 ms integration time, 2 × 2 binning) to calculate values for CFP bleed-through into the YFP channel (63%) and YFP cross-excitation of CFP (5.7%) for our imaging system. To calculate FRET in cells co-expressing CFP- and YFP-tagged constructs of interest, three separate live cell images were obtained: (i) 435 nm excitation/470 nm emission (CFP image), (ii) 500 nm excitation/535 nm emission (YFP image), and (iii) 435 nm excitation/535 nm emission (uncorrected CFP/YFP FRET image). All images were then background-subtracted. To correct for CFP bleed-through and cross-excitation of YFP, 63% of the CFP image and 5.7% of the YFP image were subtracted from the CFP/YFP FRET image to produce a corrected pseudocolor FRET image, FRETc. As well as examination of the AC8 and AKAP tagged constructs and the PKA-RII and AKAP tagged constructs, negative controls were performed in HEK293 cells expressing untagged CFP and YFP alone. In these cells, no FRETc signal was seen following fractional subtraction of the CFP and YFP images from the uncorrected FRET signal.
To quantify the degree of sensitized FRET seen between different pairings of CFP- and YFP-tagged constructs, average intensities of the FRETc images were divided by the product of the CFP and YFP image intensities to produce normalized FRETNc values (22).
HEK293 cells transiently expressing Epac2-camps and Ci/C-Epac2-camps were washed twice with ice-cold phosphate-buffered saline, scraped from the dishes, and centrifuged at 200 × g for 5 min (room temperature). Cell pellets were resuspended in 5 mm Tris-HCl, 2 mm EDTA (pH 7.3) and lysed by repeatedly passing through a 21-gauge needle. Lysates were centrifuged at 20,000 × g, 4 °C for 1 h, and subsequently the fluorescence emission spectra of the supernatants were measured (excitation at 436 ± 8 nm, emission from 450 to 550 nm) in a PerkinElmer Life Sciences LS50B spectrofluorometer. Concentrations of cAMP were spectrofluorometrically established at λ259 nm, and sigmoidal dose-response curves were obtained using GraphPad Prism version 4 (GraphPad Software Inc., La Jolla, CA). Spectra were also obtained from Epac2-camps or Ci/C-Epac2-camps lysate at pH 6.7 to examine the effects of acidification on FRET signal.
Unless stated otherwise, data were analyzed by one-way analysis of variance followed by Newman-Keuls multiple comparison tests using GraphPad Prism software. Data are presented as means ± S.E., with significance set at p < 0.05.
To probe for potential interactions between the Ca2+-stimulated AC8 and AKAPs, a number of HA-, Myc-, or FLAG-tagged AKAPs were transiently overexpressed in HEK293 cells and screened by GST pull-down assays (Fig. 1). Three cytosolic regions of AC8 were chosen as possible binding sites because these regions are highly divergent in length and sequence between ACs and could be expected to provide isoform-specific interactions with other proteins. Comparisons were made for the interaction of each AKAP with GST alone (lane 1) or with GST fused to the N terminus of AC8 (comprising residues 1–179; lane 2), the cytosolic C1b domain of AC8 (residues 582–703; lane 3), and the C terminus of AC8 (residues 1106–1248; lane 4). Immunoreactivities were then assessed using primary antibodies raised against the relevant AKAP fusion tag (HA, Myc, or FLAG). These Western blot analyses provided evidence of association between the N terminus of AC8 and both human AKAP79-HA and its rat orthologue AKAP150-HA (Fig. 1A, lane 2). A band was also present in lane 2 of the Ezrin-HA blot; however, this appeared to be due to nonspecific interactions with GST alone (see Fig. 1A, lane 1). Interestingly, AKAP150-HA was also shown to interact with the C terminus of AC8 (Fig. 1A, lane 4), although there was no evidence of AKAP79-HA interaction with this domain, suggesting differences in the specific interactions between these two AKAP orthologues and AC8. Full-length AKAP79 and AKAP150 orthologues contain 427 amino acids and 1417 amino acids, respectively (23, 24). These two AKAPs share similar primary sequence and function; however, AKAP150 also contains multiple octapeptide sequence repeats of unknown function (23). It is possible that the octapeptide repeats form an additional site of interaction with AC8 when overexpressed in HEK293 cells. AKAP18α-Myc-His, Myc-Yotiao, HA-gravin, FLAG-AKAP-Lbc, WAVE-HA, AKAP149-HA, and FLAG-sAKAP84 showed no sign of interaction with any of the GST-tagged regions of AC8 (Fig. 1A).
Having identified AKAP79 and AKAP150 as potential binding partners for the N terminus of AC8, further experiments were performed to narrow down the site of interaction between the AC and scaffold proteins (Fig. 1, B and C). GST pull-down assays were performed on whole cell lysates from HEK293 cells expressing AKAP79-HA or AKAP150-HA to compare the interaction of the two AKAP orthologues with the first half of the N terminus of AC8 (amino acids 1–77 of AC8 fused to GST) and the second half of the N-terminal region (amino acids 73–179 of AC8 fused to GST) (Fig. 1B). The GST pull-downs revealed that interaction between either AKAP79 or AKAP150 and AC8 occurred within the first 77 amino acids of the N terminus of AC8 (Fig. 1B) with little or no AKAP association with GST-AC8 73–179. Quantitation of the densitometries of HA-specific bands relative to GST signals revealed a 3.1-fold greater interaction by GST-AC8 1–77 compared with GST-AC8 73–179 with AKAP79 (Fig. 1C). For AKAP150, GST-AC8 1–77 bound with 1.6-fold greater efficiency than GST-AC8 73–179. This region of AC8 has previously been found to contain a helical calmodulin binding domain of AC8 (between residues 34 and 51) (25) and a binding site for protein phosphatase 2A that overlaps with the calmodulin binding domain (17).
We next examined whether AKAP79 could interact with full-length, functional AC8 or a truncated form of AC8 (8M1), lacking the apparently critical first 106 residues, when co-expressed in HEK293 cells. AC8-HA immune complexes immunoprecipitated from HEK293 cell lysates were tested for interaction with AKAP79-CFP by Western blot using mouse anti-GFP antibody (Fig. 2A, top blot). Co-purification of AKAP79-CFP with AC8-HA was confirmed by the presence of a band at ~110kDa in lane 6, with no band detected in the IgG control (lane 7). A reverse co-immunoprecipitation experiment provided supporting evidence for the interaction of AC8-HA and AKAP79-CFP in HEK293 cell lysate (Fig. 2A, bottom blot). AKAP79-CFP immune complexes were immunoprecipitated using an AKAP79-specific antibody and probed for interaction with AC8-HA. The presence of an HA-specific band of ~170 kDa in lane 6 confirmed co-purification of AC8-HA with AKAP79-CFP. The higher molecular weight band seen near the top of the blot corresponds to AC8 dimers (26, 27) and suggests that AKAP79 may also interact with higher orders of AC8 organization. Evidence of co-immunoprecipitation of AC8-HA with AKAP79 immune complexes in cells expressing AC8-HA alone (lane 4) was consistent with the interaction of AC8-HA with endogenous AKAP79 in HEK293 cells. Previous work from our laboratory has shown that AKAP79 is one of the major membrane-bound AKAP isoforms endogenously expressed in HEK293 cells (28).
To determine whether the N terminus of AC8 is indeed required for the interaction with AKAP79, similar co-immunoprecipitations were performed using lysate of HEK293 cells expressing either AC8-HA or the N-terminally truncated AC8 mutant, 8M1-HA (Δ1–106), with anti-HA affinity-agarose beads (Fig. 2B). The top blot confirmed the presence of AC8-HA (~170 kDa, lane 4) and 8M1-HA (~150 kDa, lane 5) in bound fractions. Blotting of the immune complexes with an AKAP79-specific antibody revealed interaction of AC8-HA with endogenously expressed AKAP79 (middle blot, lane 4). In contrast, 8M1-HA did not co-immunoprecipitate with AKAP79 (middle blot, lane 5). These results were in good agreement with the findings from Fig. 1, which suggested that the N terminus of AC8 (absent in 8M1) provides the site for interaction with AKAP79. Consistent with GST pull-down data (Fig. 1A), no band corresponding to the predicted molecular mass of AKAP149 (~150 kDa) could be seen in the AC8-HA-bound fraction (Fig. 2B, bottom blot, lane 4), showing no nonspecific binding of AKAPs to the beads and confirming the specific interaction between AKAP79 and AC8.
The 10 AKAP isoforms used in our initial GST pull-downs to identify potential binding partners for AC8 (Fig. 1A) were examined in parallel studies to see if their expression evoked any functional effects on AC8 activity. Capacitative Ca2+ entry (CCE) was induced in intact HEK293 cells to isolate AC8-dependent cAMP production from that mediated by endogenously expressed (Ca2+-insensitive) ACs. Previous studies have shown that Ca2+-dependent stimulation of AC8 activity is highly selective for store-operated, or capacitative, Ca2+ entry over other modes of Ca2+ rise in non-excitable cells (10). Thus, HEK-AC8 cells expressing the Epac2-camps FRET-based biosensor for cAMP (18) were pretreated with the sarco-/endoplasmic reticulum Ca2+-ATPase pump inhibitor, thapsigargin (Tg; 200 nm) in the absence of external Ca2+. This enabled passive depletion of the endoplasmic reticulum Ca2+ stores, priming the system for sustained store-operated Ca2+ entry upon the addition of 2 mm external Ca2+ to the bath solution (as measured using Fura-2; Fig. 3B).
Measurements from HEK-AC8 cells expressing Epac2-camps were used to quantitate the effects of transient AKAP expression on CCE-induced AC8 activity (Fig. 3, A and C). Pseudocolor images presented in Fig. 3D represent typical FRET ratio changes evoked by CCE in control cells expressing the cAMP biosensor. The switch to warmer colors at 400 s (during CCE) compared with the FRET ratio at rest (200 s) indicated an increase in cAMP levels as a consequence of enhanced AC8 activity. Expression of AKAP79-HA attenuated the peak response of AC8 to CCE by around 20% (p < 0.01 compared with control) and suggested that the interaction between the N termini of AKAP79 and AC8 (Fig. 1) played an important role in the regulation of AC8 by Ca2+. Interestingly, expression of AKAP150-HA potentiated the degree of Ca2+-stimulated AC8 activity (p < 0.05). At present, the reason for opposing effects of the human AKAP79 and the rat AKAP150 orthologues with respect to AC8 activity in HEK293 cells is unclear. However, the additional association of the C terminus of AC8 with AKAP150, not seen with AKAP79, could be linked to different functional effects of the two AKAPs (Fig. 1A, lane 4). Despite showing no signs of direct interaction, based on our GST pull-down data, three other AKAPs appeared to significantly up- or down-regulate AC8 activity. Expression of plasma membrane-targeted AKAP18α-Myc-His (also known as AKAP15) reduced AC8 activity but to a lesser extent than that seen with AKAP79 expression (p < 0.05). Furthermore, an enhancement of CCE-mediated AC8 activity was seen following expression of the membrane-associated AKAP Myc-Yotiao (p < 0.05), and the mitochondrially targeted AKAP149-HA produced a more robust potentiation of AC8 activity (p < 0.01) (Fig. 3, A and C). In summary, the most potent effects on CCE-induced AC8 activity were seen with AKAP79-HA, AKAP150-HA, and AKAP149-HA. Because AKAP149-HA and the other two AKAPs, Myc-Yotiao and AKAP18α-Myc-His, did not appear to interact directly with AC8 (Figs. 1A and and22B), their effects on AC8 activity were not investigated further.
We considered the possibility that the maximal effects of AKAP79-HA expression on AC8 activity were curtailed due to high levels of endogenous AKAP79 in HEK293 cells (supplemental Fig. 1A) (28). To further examine the potential of AKAP79 and AC8 interaction, HEK-AC8 cells were treated with an shRNA directed against AKAP79. This shRNA depleted endogenous AKAP79 levels in HEK293 cells by more than 80% in our hands (supplemental Fig. 1A). Around 72 h following shRNA AKAP79 plasmid transfection, Ca2+-stimulated AC activity in HEK-AC8 cells was significantly increased, by around 30% (p < 0.001; Fig. 4, A and D). The peak rate of signal rise was also enhanced to 2.5 times that seen under control conditions (p < 0.001; Fig. 4D). Fura-2 experiments were carried out to examine whether the effects of AKAP79 knockdown (Fig. 4, A and D) or the effects of overexpression of AKAP79-HA (Fig. 3C) or AKAP150-HA (Fig. 3C) on AC8 activity were an indirect consequence of altered CCE under these conditions. Measurements of intracellular Ca2+ changes showed comparable Ca2+ entry upon the addition of 2 mm extracellular Ca2+ to store-depleted cells under all four transfection conditions (Fig. 4B), indicating more direct effects of the AKAPs on Ca2+-regulated AC8 activity.
To further validate a specific role for AKAPs in the control of Ca2+-stimulated AC8 activity, the effects of the AKAP disruptor peptide, St-Ht31, were examined. Pretreatment of HEK-AC8 cells with 10 μm St-Ht31 for 40 min significantly enhanced CCE-mediated AC8 activity (p < 0.001; Fig. 4, C and D), and this enhancement was comparable with the effects of AKAP79 knockdown (Fig. 4A). The negative control peptide, St-Ht31P, was without effect (Fig. 4, C and D).
Biochemical studies using GST pull-downs and co-immunoprecipitation in whole cell lysate have provided evidence for an association between the N termini of AKAP79 and AC8 (Figs. 1 and and2).2). Here, we have used “micro-FRET” to examine this protein-protein interaction at the single cell level in HEK293 cells expressing YFP- and CFP-tagged AC8 and AKAP79 constructs (Fig. 5). This technique takes into account variable CFP and YFP expression in transient transfections to provide a high resolution measurement of sensitized FRET, which is corrected for any potential bleed-through in fluorescence signals in the YFP and CFP emission images (22). Using micro-FRET, we were able to demonstrate a protein-protein interaction between YFP-AC8 and AKAP79-CFP. The degree of FRET seen at the plasma membrane was enhanced when we co-expressed 8Tm1/YFP/Tm2 with AKAP79-CFP. 8Tm1/YFP/Tm2 is a truncated version of AC8 in which YFP has been inserted between residues 1–397 and residues 654–956 (26). Importantly, this AC8 construct contains the full-length N terminus of AC8 but lacks most of the cytosolic C1 and C2 domains. Quantification of the FRET signal was achieved by normalization of the corrected FRET image (FRETNc; see “Experimental Procedures”) and indicated 70% greater FRETNC when 8Tm1/YFP/Tm2 was used instead of YFP-AC8 (Fig. 5, bar chart). The average FRETNc value of 10.8 ± 2.1 for 8Tm1/YFP/Tm2 and AKAP79-CFP interaction was comparable with our positive control, which used AKAP79-YFP and PKA-RIIα-CFP binding to provide a FRETNc value of 13.4 ± 3.4. This latter value is consistent with previous micro-FRET studies examining the interaction between PKA and its scaffold protein (29). The ability of 8Tm1/YFP/Tm2 to bind AKAP79 further supports the importance to the interaction of the N terminus of AC8 and the insignificance of the C1 and C2 domains. Negative controls in cells that co-expressed both AKAP79-YFP and AKAP79-CFP or untagged CFP and YFP demonstrated good co-localization but, as expected, no FRET signal.
Having proven that AC8 and AKAP79/150 can interact in HEK293 cells overexpressing tagged versions of AC8 and AKAP79/150, we sought to establish whether a similar interaction could be detected in a more physiological system in which both proteins are expressed endogenously. A recent study by our laboratory demonstrated the presence of Ca2+-stimulable AC8 activity in a mouse insulin-secreting pancreatic β-cell line, MIN6 (30). Here, the endogenous expression of the rodent AKAP79 orthologue, AKAP150, in MIN6 cells was confirmed by Western blot analysis (Fig. 6A, left). This AKAP150-specific band was absent in HEK293 cells stably expressing AC8. Using an AC8-specific antibody, we also confirmed the presence of AC8 (band ~170 kDa) but at far lower levels than those seen when AC8 is overexpressed in HEK293 cells (Fig. 6A, right). Using the AKAP150-specific antibody to immunoprecipitate endogenous AKAP150 from MIN6 cell lysates, we could demonstrate the interaction with endogenously expressed AC8 (Fig. 6B, lane 2) but not with the IgG control (lane 1).
To establish whether the AC8-AKAP150 signaling complex plays an important functional role in the regulation of AC8 in MIN6 cells, we assessed the consequences of AKAP150 knockdown and overexpression on CCE-induced AC8 activity (Fig. 7). Because MIN6 cells express both AC8 and the Ca2+-inhibitable AC6, conditions were optimized for stimulation rather than inhibition of AC activity by using 2 mm extracellular Ca2+ to trigger CCE. Higher Ca2+ concentrations have been shown to preferentially influence Ca2+-inhibitable AC6 activity (30). Store-mediated Ca2+ entry was induced by prior store depletion in Ca2+-free conditions using 1 μm Tg treatment for 3 min followed by the subsequent addition of 2 mm external Ca2+ (Fig. 7A). Around 60% of this Ca2+ entry is inhibited by 100 μm 2-aminoethoxydiphenyl borate and can be attributed to CCE (see bar chart inset). Parallel experiments were performed in MIN6 cells transiently expressing the Epac2-camps FRET sensor (Fig. 7B). For cAMP measurements, the AC activator, forskolin (FSK; 20 nm), and non-selective phosphodiesterase inhibitor, 3-isobutyl-1-methylxanthine (IBMX; 100 μm), were also present to potentiate cAMP production. All data are plotted as the percentage of FRET signal change relative to the maximum FRET signal that is attainable with saturating [cAMP]. Under these conditions, CCE induced an initial decrease in cAMP production (due to a reduction in AC6 activity) that was followed by a steady rise in cAMP levels. In MIN6 cells transiently transfected with AKAP150-HA, the amplitude of Ca2+-stimulated cAMP production was reduced (Fig. 7, B and C). This effect was comparable with the effects of AKAP79 overexpression in HEK-AC8 cells (Fig. 3). Lentiviral shRNA directed against AKAP150 was found to reduce endogenous AKAP150 levels by around 55% (supplemental Fig. 1B). This knockdown of endogenous AKAP150 expression levels produced a marked potentiation of cAMP production compared with scrambled shRNA controls following the initial decrease in AC activity (Fig. 7B, lower traces) (p < 0.05). These findings further support an AKAP79/150 and AC8 interaction attenuating the stimulatory effects of Ca2+ on AC8 activity.
Previous studies on the distribution patterns for AC8 expression have shown that the enzyme is highly expressed within the hippocampus (31, 32), where it is believed to play an important role in memory formation (33, 34). Interestingly, the hippocampus is also a major source of AKAP79/150 expression (35,–37), where it associates with a range of membrane proteins, including N-methyl-d-aspartate receptors (38) and L-type Ca2+ channels (15). Using primary cultured rat hippocampal neurons, we investigated whether a change in the expression levels of AKAP79/150 altered Ca2+-stimulated AC activity. Because activity-induced changes in neuronal [Ca2+] are typically accompanied by substantial local acid shifts (39), we generated a modified Epac2-camps sensor for our hippocampal experiments in which the pH-sensitive YFP (pKa ~6.9) was replaced with citrine (pKa ~5.8) (Fig. 8A), thus minimizing the potential for pH artifacts in the measured FRET signal. This new sensor, Ci/C-Epac2-camps, expressed well in ~2% of our cultured hippocampal neurons following transient transfection at 10–12 days in vitro. Of the neurons expressing the cAMP sensor, only pyramidal neurons (identified by cell morphology) were selected for experimental use.
In vitro calibration of the Ci/C-Epac2-camps used cell lysate from HEK293 cells transiently transfected with the new sensor and showed good sensitivity to cAMP with an estimated EC50 value of 210 ± 11 nm cAMP. This value fell slightly below our estimated EC50 of 349 ± 21 nm for the parent YFP-containing Epac2-cAMP sensor (Fig. 8B). Spectral analysis of the Ci/C-Epac2-camps and original Epac2-camps sensors revealed reduced pH sensitivity of the citrine compared with YFP at around 530 nm (peak emission for YFP and citrine), with the citrine showing no obvious decline in signal when switching from pH 7.3 to pH 6.7 (Fig. 8B, lower plots).
To directly examine the effects of AKAP150-HA expression on endogenous Ca2+-stimulated AC activity in cultured hippocampal neurons, AC activity was compared in the absence and the presence of 2 mm external Ca2+. Under control conditions, switching from 2 mm Ca2+ containing saline to Ca2+-free saline resulted in a cessation of spontaneous Ca2+ transients associated with voltage-gated Ca2+-channel activity (data not shown) and a notable decrease in Ci/C-Epac2-camps emission ratio, consistent with a drop in basal AC activity (Fig. 8C). Upon the addition of 1 μm FSK in Ca2+-free conditions, cAMP levels began to rise slowly. This rate of cAMP increase was markedly potentiated upon the readdition of external Ca2+ at 180 s, due to stimulation of endogenous AC8 (and/or AC1) activity (Fig. 8, C and D). Maximum FRET ratio change was obtained by the subsequent addition of 10 μm FSK, 10 μm isoproterenol, and 100 μm IBMX. In parallel experiments, the transient overexpression of AKAP150-HA significantly delayed the Ca2+-dependent rise in AC activity (Fig. 8, C and D). This effect of AKAP150 overexpression is again consistent with the attenuating effects of AKAP79 on AC8 activity in HEK293 cells and of AKAP150 on AC8 activity in MIN6 cells, which suggests that AKAP79/150 dampens Ca2+-dependent AC8 activity as a consequence of its interaction with the enzyme. Experiments attempting to assess the effects of knockdown of endogenous AKAP150 levels proved difficult to interpret because 72–96-h lentiviral transfection with shRNA targeted against AKAP150 caused a pronounced enhancement of Ca2+-insensitive AC activity in the cultured neurons (Fig. 8C). This enhanced rate of cAMP production was not increased further upon the addition of 2 mm Ca2+ to the bath solution (Fig. 8, C and D).
Despite mounting evidence for the direct association of AKAPs with Ca2+-regulated ACs (5, 6), our understanding of the role of such interactions is limited. In the case of AC5/6, AKAP-targeted PKA serves to facilitate inhibition of cAMP production (5, 6); however, nothing is known regarding the effects of AKAP interaction on the highly sensitive and potentially dynamic regulation of the ACs by changes in intracellular [Ca2+]. This issue is of particular interest because both the Ca2+-regulated ACs and a number of specific AKAPs reside in close apposition to or direct association with sites of Ca2+ entry (9, 11, 15, 40, 41). Here we provide the first evidence for interaction of Ca2+-stimulated AC8 with an AKAP and begin to address the function of such AKAP-AC interactions with respect to Ca2+-regulated cAMP production in a range of cell types.
Direct binding between the N terminus of AC8 and AKAP79/150 was illustrated by GST pull-downs and confirmed by co-immunoprecipitation of tagged AC8 and AKAP79 constructs when overexpressed in HEK293 cells. In addition, AC8-HA associated with endogenous human AKAP79 in HEK293 cells. A protein-protein interaction was also observed between AC8 and AKAP150 (the rodent orthologue of AKAP79) in pancreatic cells endogenously expressing both protein species.
Using truncated N-terminal regions of AC8, we could narrow down the site of interaction with AKAP79/150 to residues within the first 77 amino acids of AC8. This was supported by the lack of interaction of AKAP79 with the N-terminally truncated AC8 mutant, 8M1 (Δ1–106). Our data were consistent with the hypothesis that the non-conserved N termini of ACs provide a general site of interaction during the assembly of AKAP-AC complexes (7) because the N termini of both AC2 and AC5 provided the site for interaction with Yotiao and mAKAPβ, respectively (6, 7). However, common contact sites for AKAP-AC interaction seem unlikely because mAKAPβ and AKAP79 have been shown to interact with different sites within AC5 (7). The first 77 amino acids of AC8 contain a helical calmodulin binding domain, located between residues 34 and 51 (25, 42). An overlapping region of amino acid residues is also reported to bind the protein phosphatase, protein phosphatase 2A (17). The precise residues within the N terminus of AC8 that were responsible for binding to AKAP79/150 are yet to be identified. Studies of AKAP79/150 binding to the L-type Ca2+ channel, Cav1.2, have identified a modified leucine zipper motif within the distal C terminus of the channel subunit that is responsible for the interaction with AKAP79/150 (15, 41, 43). We considered the possibility of a similar means of interaction of AKAP79/150 with AC8. However, sequence analysis of AC8 suggested that no such leucine zipper-like domains were present within the first 77 amino acids. Interestingly, the AKAP150 orthologue also interacted with the C2 domain of AC8 (amino acids 1106–1248) when expressed in HEK293 cells. Similar additional sites of AKAP interaction were reported for the binding of AC5 to mAKAPβ, which was found to involve all three cytosolic domains of AC5 (N terminus and C1 and C2 domains) (7).
In addition to our biochemical data, direct evidence of binding between AKAP79 and AC8 was observed using micro-FRET to confirm protein-protein interactions between AKAP79-CFP and YFP-AC8 in live cells. Switching the location of the YFP from the N terminus of AC8 to the C1 domain, as in 8Tm1/YFP/Tm2, enhanced the degree of FRET between tagged AC8 and AKAP79 constructs. This improved FRET efficiency is likely to be mediated by subtle differences in the distance and orientation between the two fluorophores when YFP is inserted at the C1 domain of AC8. The degree of FRET observed between 8Tm1/YFP/Tm2 and AKAP79-CFP was comparable with that observed following co-expression of AKAP79-CFP with its ubiquitous binding partner PKA-RIIα (tagged to CFP).
To determine a functional role for the proposed AKAP79/150-AC8 signaling complex, Ca2+-dependent AC8 activity was monitored using the FRET-based cAMP biosensor, Epac2-camps (18). In an overexpression system and when expressed endogenously, the AKAP79/150 interaction with AC8 mediated an inhibitory effect on AC8 activity. In HEK-AC8 cells, AKAP79-HA expression decreased CCE-mediated cAMP production by ~20% compared with controls, independently of any indirect effect on the amplitude of the Ca2+ entry signal. Because AKAP79 is endogenously expressed in HEK293 cells at relatively high levels (supplemental Fig. 1A), more potent modulation of the effect of AKAP79 on AC8 activity was seen when shRNA was used to selectively knock down endogenous expression of the scaffold protein. Comparable effects were also seen following disruption of AKAP-PKARII interaction (using St-Ht31). Thus, targeting of PKA to AC8 via AKAP79 might underlie the regulatory effects of AKAP79 with respect to AC8 activity. This would provide the first evidence of a regulatory effect of PKA on AC8 activity and, more specifically, an inhibitory effect of PKA on Ca2+-dependent AC8 activity. We can only speculate on the precise mechanism of action for PKA. The simplest scenario would be a direct phosphorylation of AC8 by PKA. The amino acid sequence of AC8 reveals a number of consensus PKA phosphorylation sites in the intracellular domains. An alternative mode of regulation via PKA could arise from the ability of PKA to phosphorylate and activate certain protein phosphatase 2A subunits (44). This latter option is of particular interest because protein phosphatase 2A has been shown to directly associate with the N terminus of AC8 (17). In addition, further modulatory actions of other AKAP79-associated proteins, such as protein kinase C or PP2B, cannot be discounted.
The overexpression of AKAP150-HA had opposing effects on Ca2+-stimulated AC8 activity compared with AKAP79-HA, resulting in enhanced cAMP production during CCE. One possible explanation for the different effects of these two AKAP orthologues when overexpressed in HEK293 cells is the ability of the longer AKAP150 to bind to the C2 domain of AC8, in addition to the N terminus, potentially leading to competitive inhibition of any interaction of endogenous AKAP79 with AC8 in HEK293 cells. Furthermore, it is possible that the correct association of rodent AKAP150 with its typical array of signaling molecules is limited in a human cell line endogenously expressing AKAP79. Both AKAP79 and AKAP150 behaved similarly when studied in their native environments.
When we examined the role of the AKAP-AC8 interaction in physiological systems in which AKAP150 and AC8 were expressed endogenously, we saw further evidence of inhibitory regulation of AC8 activity that was dependent upon its association with AKAP150. In insulin-secreting mouse pancreatic β-cells (MIN6 cells), overexpression of AKAP150 decreased CCE-mediated AC8 activity. This effect of AKAP150 was comparable with the effects of AKAP79 in HEK-AC8 cells. However, the role of AKAP150 with respect to AC8 activity was more clearly seen when the cultured pancreatic cells were treated with a shRNA lentivirus designed to suppress endogenous expression of the AKAP. Knockdown of AKAP150 in MIN6 cells resulted in a significant increase in Ca2+-stimulated AC8 activity (Fig. 7). A modest enhancement of the transient Ca2+-dependent inhibition of cAMP levels was also observed. This latter effect was probably due to loss of the association of endogenous AC6 with AKAP150 that has been shown to mediate PKA inhibition of AC5/6 activity (5).
In primary cultured hippocampal pyramidal neurons, the inhibitory effects of AKAP150 on Ca2+-stimulated AC activity displayed different characteristics. Overexpression of AKAP150-HA in hippocampal neurons significantly delayed Ca2+-stimulated cAMP production mediated by endogenously expressed AC8 (rather than inhibiting the overall degree of AC8 activity). Bidirectional regulation of L-type Ca2+ channels has been reported in hippocampal neurons due to interaction of the channels with AKAP79/150 and the opposing actions of PKA and calcineurin (15). Thus, it was important to determine if the effects of AKAP150-HA expression on AC8 activity might be explained by reduced Ca2+ entry. Control Fura-2 experiments revealed a modest increase in basal Ca2+ levels and the amplitude of spontaneous Ca2+ transients following AKAP150 expression, but no significant difference in the amplitude or rate of Ca2+ entry was seen upon the readdition of external Ca2+ (data not shown). We therefore conclude that the inhibitory effects of AKAP150 expression on Ca2+-stimulated cAMP production in hippocampal neurons were a direct consequence of an AKAP150-AC8 interaction. Previous studies have implicated essential roles for both AC8 and AKAP150 in hippocampal memory formation (33, 34, 36, 41, 45). Our evidence for a direct association between these two proteins, with a clear bifunctional consequence, has identified a potentially important site of communication between two key signaling components linked to memory formation. This is further supported by the selective targeting of AKAP79/150 and AC8 to dendritic spines (31, 35, 46, 47).
To conclude, we propose that AKAP79/150 interaction with AC8 mediates inhibition of Ca2+-dependent cAMP production. Such modulated activity of AC8 is likely to have physiological significance in the pancreas and in the hippocampus in which AKAP79/150 and AC8 are co-expressed. These tissues are subject to transient fluxes in local Ca2+ levels that have the potential to generate localized, dynamic cAMP signals linked to important events, such as insulin secretion (12, 14) and memory formation (48, 49). The evidence presented here of the functional association of AKAP79/150 with AC8, combined with the positioning of AKAP79/150 close to sites of Ca2+ entry (15), could form the basis of a self-regulated multimolecular signaling complex in neurons, where the local cAMP events have the potential to regulate the activity of other AKAP79/150-associated proteins, such as PKA or L-type Ca2+ channels. The influence of AKAP-binding on Ca2+-driven cAMP signals generated by other AC isoforms is yet to be investigated, but such a device could add a sophisticated component in tuning the interplay between Ca2+ and cAMP signaling.
2The abbreviations used are: