AKAP79/150 Signal Complexes in G-Protein Modulation of Neuronal Ion Channels

Abstract

Voltage-gated M-type (KCNQ) K⁺ channels play critical roles in regulation of neuronal excitability. Previous work showed A-kinase-anchoring protein (AKAP)79/150-mediated protein kinase C (PKC) phosphorylation of M channels to be involved in M current (I_M) suppression by muscarinic M₁, but not bradykinin B₂, receptors. In this study, we first explored whether purinergic and angiotensin suppression of I_M in superior cervical ganglion (SCG) sympathetic neurons involves AKAP79/150. Transfection into rat SCG neurons of ΔA-AKAP79, which lacks the A domain necessary for PKC binding, or the absence of AKAP150 in AKAP150^−/− mice, did not affect I_M suppression by purinergic agonist or by bradykinin, but reduced I_M suppression by muscarinic agonist and angiotensin II. Transfection of AKAP79, but not ΔA-AKAP79 or AKAP15, rescued suppression of I_M by muscarinic receptors in AKAP150^−/− neurons. We also tested association of AKAP79 with M₁, B₂, P2Y₆, and AT₁ receptors, and KCNQ2 and KCNQ3 channels, via Förster resonance energy transfer (FRET) on Chinese hamster ovary cells under total internal refection fluorescence microscopy, which revealed substantial FRET between AKAP79 and M₁ or AT₁ receptors, and with the channels, but only weak FRET with P2Y₆ or B₂ receptors. The involvement of AKAP79/150 in G_q/11-coupled muscarinic regulation of N- and L-type Ca²⁺ channels and by cAMP/protein kinase A was also studied. We found AKAP79/150 to not play a role in the former, but to be necessary for forskolin-induced upregulation of L-current. Thus, AKAP79/150 action correlates with the PIP₂ (phosphatidylinositol 4,5-bisphosphate)-depletion mode of I_M suppression, but does not generalize to G_q/11-mediated inhibition of N- or L-type Ca²⁺ channels.

Introduction

To ensure specificity of intracellular events, signaling complexes at specific subcellular locations are formed by scaffold proteins, such as A-kinase anchoring proteins (AKAPs), which anchor receptors and signaling proteins to physiological substrates (Bauman et al., 2006). AKAP79/150 is a member of this family, with three orthologs, human AKAP79, rodent AKAP150, and bovine AKAP75. AKAP79/150 recruits protein kinase A (PKA), protein kinase C (PKC), calmodulin (CaM), and calcineurin (CaN) to ion channels, such as glutamate receptors (Tavalin et al., 2002; Dell'Acqua et al., 2006), Kir 2.1 (Dart and Leyland, 2001), L-type Ca²⁺ channels (Oliveria et al., 2007), and GABA receptors (Brandon et al., 2003).

AKAP79/150 is also involved in the modulation of M-type (KCNQ, Kv7) K⁺ channels. These channels have properties that enable them to control neuronal excitability and action potential firing (Brown et al., 2007). M channels are inhibited by several G_q/11-coupled receptors using two intracellular pathways that arise from the need of M channels for membrane phosphatidylinositol 4,5-bisphosphate (PIP₂) (Suh and Hille, 2002; Ford et al., 2003; Zhang et al., 2003; Li et al., 2005; Suh et al., 2006). The first pathway, used by muscarinic M₁ and angiotensin II (AngII) AT₁ receptors in sympathetic neurons, acts by depletion of PIP₂ (Horowitz et al., 2005; Li et al., 2005; Winks et al., 2005; Suh et al., 2006). Another mechanism, used by bradykinin B₂ and purinergic P2Y receptors, does not involve depletion of PIP₂ but, rather, intracellular Ca²⁺ signals that act in concert with CaM (Gamper and Shapiro, 2003). This mechanism probably involves alterations in channel affinity for PIP₂ (Delmas and Brown, 2005; Zaika et al., 2007; Hernandez et al., 2008). Previous work showed AKAP79 to coprecipitate with KCNQ2 channels (Cooper et al., 2000), and to facilitate PKC phosphorylation of KCNQ2 channels (Hoshi et al., 2003). In sympathetic neurons expressing a dominant-negative AKAP150 that cannot bind PKC, or those with AKAP150 knocked down using siRNA, there was reduced suppression of M current (I_M) by M₁, but not B₂, receptor stimulation (Hoshi et al., 2003, 2005) Studies on AKAP150^−/− mice suggest a role of AKAP79/150 in epilepsy, with AKAP150^−/− mice demonstrating increased seizure resistance (Tunquist et al., 2008). Recent work reveals AKAP79/150 to interact with KCNQ2-5, an interaction that is inhibited by CaM (Bal et al., 2010).

Here, we studied whether AKAP79/150 is also involved in the P2Y and AT₁ receptor modulation of I_M. We used patch-clamp electrophysiology to explore the functional actions of AKAP79/150 in sympathetic neurons isolated from rat, or wild-type or AKAP150^−/− mice, and Förster resonance energy transfer (FRET) on tissue-culture cells to show receptor association. The involvement of AKAP79/150 in similar G_q/11-coupled pathways regulating activity of N- and L-type Ca²⁺ channels was also studied. We show that the AKAP79/150 action is highly receptor specific, being used in angiotensin, but not in purinergic suppression of I_M. The AKAP79/150 action does not generalize to G_q/11-mediated modulation of N- or L-type type Ca²⁺ channels, but is involved in L-channel modulation by PKA.

Materials and Methods

Superior cervical ganglion neuron culture and transfection.

Sympathetic neurons were isolated from the superior cervical ganglion (SCG) of 7- to 14-d-old rats (Sprague Dawley) or mice (C57BL/6) and cultured for 2–4 d. Rats or mice were anesthetized with halothane and decapitated. Neurons were dissociated using established methods (Bernheim et al., 1991), plated on 4 × 4 mm glass coverslips (coated with poly-l-lysine), and incubated at 37°C (5% CO₂). Fresh culture medium containing nerve growth factor (NGF) (50 ng/ml) was added to the cells 3 h after plating. In some experiments, pertussis toxin (PTX) (100 ng/ml) was added to the culture medium. For exogenous expression of cDNA constructs in SCG neurons, we used the PDS-1000/He biolistic particle delivery system (Bio-Rad), as described previously (Gamper and Shapiro, 2006). Transfection efficiency was assumed to be determined by the random distribution of fired gold particles, and was up to 10% of cultured neurons.

AKAP150 knock-out mice.

We maintain colonies of AKAP150^+/+ and AKAP150^−/− mice (C57BL/6 background) at the University of Texas Health Science Center at San Antonio (originally supplied to us by Dr. G. Stanley McKnight, University of Washington, Seattle, WA). Mice were housed in groups of five and maintained under a 12 h light/dark cycle with food and water provided ad libitum. Mice were backcrossed (+/+ × −/− to yield +/−) every sixth generation. Genotype comparisons were made among littermates and mice were bred by intercross to produce all three genotypes within a litter. A small piece of tissue taken from the end of the tail (tail clip) was used to genotype the mice by real-time (quantitative) (qRT)-PCR on an Applied Biosystems PRISM 7000 Real-Time PCR sequence detection system. The probe and detailed protocol for performing these analyses were also kindly provided by Drs. G. Stanley McKnight and Johannes Hell (University of California, Davis, CA). The primers used for the qRT-PCR for AKAP150 were 5′-GGCCTTGTGACACACAGGAA-3′ and 5′-CAGGCGGCTTCTGCTTCTT-3′, and the fluorescent probe was VIC-AGGTCCGAGCCTGC-MGBNFQ; for Neo, the primers were 5′-ATGGCCGCTTTTCTGGATT-3′ and 5′-GCCAAGCTCTTCAGCAATATCA-3′, and the fluorescent probe was VIC-CGGACCGCTATCAGG-MGBNFQ. Samples were run through 40 cycles and each sample was run in duplicate for both AKAP150 and Neo. To minimize user bias, cycle thresholds (C_t) were calculated automatically using the Applied Biosystems software. As a control against Neo contamination, all of our littermate genotyping analyses included tail clips from two AKAP150^+/+, two AKAP150^+/−, and two AKAP150^−/− mice as control. Genotyping results were only accepted if the C_t values for both AKAP150 and applying a 500 ms hyperpolarizing pulse and Neo were not significantly different from controls.

Perforated-patch electrophysiology.

Pipettes were pulled from borosilicate glass capillaries (1B150F-4, World Precision Instruments) using a Flaming/Brown micropipette puller P-97 (Sutter Instruments), and had resistances of 1–4 MΩ when filled with internal solution and measured in standard bath solution. Membrane current was measured with pipette and membrane capacitance cancellation, and was sampled at 5 ms or 200 μs and filtered at 1 or 2.5 kHz [for I_M and I_Ca (Ca²⁺ current), respectively] by an EPC-9 amplifier and PULSE software (HEKA/Instrutech). In all experiments, the perforated-patch method of recording was used with amphotericin B (600 ng/ml) in the pipette (Rae et al., 1991). Amphotericin was prepared as a stock solution as 60 mg/ml in DMSO. In these experiments, the access resistance was typically 5–10 MΩ 5–10 min after seal formation. Cells were placed in a 500 μl perfusion chamber through which solution flowed at 1–2 ml/min. Inflow to the chamber was by gravity from several reservoirs, selectable by activation of solenoid valves (Warner Scientific). Bath solution exchange was essentially complete by <30 s. Experiments were performed at room temperature.

M currents in SCG cells were studied by holding the membrane potential at −25 mV and applying a 500 ms hyperpolarizing pulse to −60 mV every 5 s. M-current amplitude was measured at −60 mV from the decaying time course of the deactivating current sensitive to the M channel-specific blocker XE991 (Zaczek et al., 1998). UTP, bradykinin (BK), angiotensin II, and XE991 were used at concentrations of 2 μm, 100 nm, 500 nm, and 10 μm, respectively. Oxotremorine methiodide (oxo-M) was used at concentrations of 1 μm and 0.3 μm for rat and mouse SCG neurons, respectively. To evaluate the amplitude of I_Ca, cells were held at −80 mV, and 15 ms depolarizing steps to 5 mV were applied every 5 s. The amplitude of I_Ca was usually defined as the inward current sensitive to Cd²⁺ (100 μm). For experiments in which the long-lasting tail current was measured, cells were held at −90 mV, and 15 ms depolarizing steps to 10 mV and 40 ms steps to −40 mV were applied every 5 s, and the tail current amplitude was measured ∼15 ms after the start of the −40 mV pulse.

The external solution used to record M currents contained the following (in mm): 150 NaCl, 5 KCl, 2 CaCl₂, 1 MgCl₂, 10 glucose, 10 HEPES, 500 nm tetrodotoxin (TTX), pH 7.4, with NaOH. For I_Ca recordings, KCl was reduced to 2.5 mm, and CaCl₂ was increased to 5 mm. The pipette solution for voltage-clamp experiments contained the following (in mm): 150 KCl/CsCl, 5 MgCl₂, and 10 HEPES to record I_M/I_Ca. Data are presented as mean ± SEM. Statistical tests were performed using ANOVA, paired t test, or unpaired t test, where appropriate.

cDNA constructs.

Human KCNQ2-4 constructs (GenBank accession numbers AF110020, AF071478, and AF105202, respectively) were kindly given to us by David McKinnon (State University of New York, Stony Brook, NY) [KCNQ2 (isoform d)] and Thomas Jentsch (Zentrum fur Molekulare Neurobiologie, Hamburg, Germany) [KCNQ3, KCNQ4 (isoform a)]. Plasmids were subcloned into pECFP-N1 or PEYFP-N1 vectors (Clontech) using standard techniques. The membrane-localized cyan fluorescent protein (CFP)-yellow fluorescent protein (YFP) tandem construct (Rho-pYC) was kindly given to us by Paul Slesinger (Salk Institute, La Jolla, CA). It consists of the C-terminal prenylation site of Rho (RQKKRRGCLLL) appended to the C terminus of a YFP-CFP fusion protein (Fowler et al., 2007).

Cell culture and cDNA transfections.

Chinese hamster ovary (CHO) cells were grown in 100 mm tissue culture dishes (Bymaster et al., 1999) in DMEM with 10% heat-inactivated fetal bovine serum plus 0.1% penicillin and streptomycin in a humidified incubator at 37°C (5% CO₂), and passaged approximately every 4 d. Cells were discarded after ∼30 passages. For the total internal reflection fluorescent (TIRF)/FRET experiments, cells were first passaged onto 35 mm plastic tissue culture dishes and transfected 24 h later with Polyfect reagent (QIAGEN), according to the manufacturer's instructions and as previously described (Gamper et al., 2005). The next day, cells were plated onto poly-l-lysine-coated glass-bottom 35 mm tissue culture dishes (MatTek) or cover glass chips, and experiments were performed over the following 1–2 d.

TIRF microscopy.

Fluorescence emission from enhanced CFP-tagged or enhanced YFP-tagged proteins were collected at room temperature using TIRF (also called evanescent field) microscopy. Total internal reflection fluorescence generates an evanescent field that declines exponentially with increasing distance from the interface between the cover glass and the cytoplasm, illuminating only a thin section (300 nm) of the cell very near the cover glass, including the plasma membrane (Axelrod, 2003). All TIRF experiments were performed in the TIRF microscopy core facility housed within the Department of Physiology. Images were collected (200–600 ms exposure time) immediately before and after photobleaching. Images were not binned or filtered, with pixel size corresponding to a square of 122 × 122 nm.

FRET.

We used the “acceptor photobleaching” (donor dequenching) method of evaluating FRET efficiency as previously described (Bal et al., 2010). The emission of the donor fluorophore was compared before and after photobleaching of the acceptor (Centonze et al., 2003). The medium in the glass-bottom dishes was exchanged with Ringer's solution that contained (in mm): 160 NaCl, 5 KCl, 1 MgCl₂, 2 mm CaCl₂, 10 HEPES, pH 7.4, with NaOH. Cells were first examined using the mercury lamp and standard CFP or YFP filter cubes to find a suitable cell robustly expressing both CFP- and YFP-tagged proteins. Under TIRF illumination, the focal plane was adjusted if necessary, immediately before each image acquisition to obtain a sharp TIRF image. The focusing and cell-centering protocol resulted in CFP photobleaching of <1%. TIRF images using 442 and 514 laser lines were acquired before and after photobleaching of the YFP fluorophores. FRET efficiency was calculated as the percentage increase in CFP emission after YFP photobleaching by using the formula: %FRET = ((CFP_post − CFP_pre)/CFP_pre)*100, where CFP_post is CFP emission after YFP photobleaching and CFP_pre is CFP emission before YFP photobleaching. The %FRET was calculated by drawing regions of interest around the entire area of the cell and subtracting the background in a cell-free region for each image.

Immunostaining.

Cells grown on poly-l-lysine-coated coverslips were fixed in 4% paraformaldehyde, washed twice with 100 mm sodium phosphate (PB), pH 7.4 and three times with PB plus 150 mm NaCl (PBS), and blocked with 5% goat serum and 0.1% saponin in PBS (PBS + GS). The cells were incubated for 3 h at room temperature with primary affinity-purified goat anti-AKAP150 (diluted 1:1500 in PBS + GS, Santa Cruz Biotechnology) and mouse anti-tyrosine hydroxylase (1:5000) antibodies. In the blocking controls, the anti-AKAP150 antibody was preadsorbed with a tenfold molar excess of the peptide used to raise the antibody. Cells were washed six times with PBS and then incubated with goat Rhodamine Red-conjugated anti-rabbit and FITC-conjugated anti-mouse secondary antibodies (1:500, Jackson ImmunoResearch) in PBS + GS for 1 h. Cells were then washed three times with PBS, twice with PB, and three times with water. Air-dried slides were mounted on a drop of Vectashield (Vector Laboratories) and sealed with nail polish. Stained cells were viewed using the Olympus FV-1000 confocal microscope located in our institutional confocal imaging core facility.

Reagents.

The following reagents were used: UTP, TTX, NGF, bradykinin, oxo-M, angiotensin II, and collagenase type I (Sigma); DMEM, fetal bovine serum, penicillin/streptomycin (Invitrogen); amphotericin B and PTX (Calbiochem); and FPL 64176 and XE991 (Tocris Bioscience).

Results

We studied AKAP79/150-mediated regulation of the M-type (KCNQ) K⁺ current, N-type Ca²⁺ current (N-current), and L-type Ca²⁺ current (L-current) in sympathetic neurons of rodent SCG. Our work using Förster resonance energy transfer and perforated-patch voltage clamp of CHO cells has shown AKAP79/150 to interact with KCNQ2-5 subunits, and AKAP79/150-mediated PKC phosphorylation to sensitize KCNQ2/3 heteromers and KCNQ2-5 homomers to inhibition by stimulation of muscarinic receptors (Bal et al., 2010). In this study, we used patch-clamp electrophysiology of cultured SCG neurons. We chose to use these well studied neurons because (1) they express robust I_M that can be isolated without a complex “cocktail” of blockers normally needed for M-current study in central neurons, (2) N-current constitutes most of the I_Ca in rat SCG neurons, and mouse SCG neurons express a large and easily isolated L-current, and (3) they express a robust repertoire of G_q/11-coupled receptors that reliably modulate M-, N-, and L-current activity. Previous studies in these neurons using dominant-negative (DN) constructs, RNAi, and AKAP150 knock-out (KO) mice have shown that AKAP79/150-mediated PKC phosphorylation of M channels is involved in I_M suppression by G_q/11-coupled muscarinic M₁, but not bradykinin B₂ receptors (Hoshi et al., 2003, 2005; Tunquist et al., 2008).

Overexpression of dominant-negative AKAP79 suppresses M-current inhibition by muscarinic and angiotensin II receptors

Our first question was to ask whether purinergic and angiotensin suppression of I_M in SCG neurons involves AKAP79/150, since purinergic P2Y receptors depress I_M in SCG neurons via a mechanism similar to that of bradykinin, involving IP₃-mediated Ca²⁺ signals (Delmas and Brown, 2005; Zaika et al., 2007; Hernandez et al., 2008), whereas angiotensin AT₁ receptors depress I_M via a mechanism similar to that of M₁ receptors, by depletion of PIP₂ (Horowitz et al., 2005; Li et al., 2005; Winks et al., 2005; Suh et al., 2006). We wished to overexpress, as a DN, ΔA-AKAP79 (Fig. 1A), which lacks the A domain necessary for PKC binding to AKAP79/150 (Klauck et al., 1996), in rat SCG neurons to answer this question. Overexpression of ΔA-AKAP79 is thought to compete with, and to disrupt, any endogenous AKAP150 bound to M channels (Hoshi et al., 2003), but we first decided to verify this notion. Thus, before functional study, we tested whether ΔA-AKAP79 interacts with KCNQ channels heterologously expressed in CHO cells by FRET under TIRF microscopy, which selectively excites fluorophores within 300 nm of the plasma membrane (Axelrod, 2003). In these experiments, we used enhanced CFP and enhanced YFP as the donor and acceptor, respectively, as in our previous experiments (Bal et al., 2010). FRET was measured under TIRF illumination by the donor dequenching method, in which CFP emission is compared before and after photobleaching of YFP. The FRET efficiency for our positive control, the membrane-targeted tandem CFP-YFP construct (Rho-pYC), was 26 ± 1% (n = 15). Our negative control was coexpression of YFP-tagged AKAP79 with a membrane-targeted enhanced CFP (ECFP-M), which yielded a FRET value of only 2 ± 2% (n = 13). CHO cells were then transfected with CFP-tagged KCNQ2-4 channels and YFP-tagged ΔA-AKAP79. TIRF images of CFP and YFP emission, before and after YFP photobleaching, are shown in Figure 1B, in which CFP and YFP images are displayed in “rainbow” and yellow pseudocolor, respectively. The CFP emission was significantly greater than that for our negative control after YFP photobleach for all KCNQ2-4 channels. These data are summarized in Figure 1C. The FRET efficiency for KCNQ2-4 with ΔA-AKAP79 was 17 ± 2% (n = 12), 13 ± 3% (n = 9), and 16 ± 2% (n = 13), respectively. Thus, ΔA-AKAP79 retains intimate association with KCNQ2-4 channels similar to wild-type AKAP79 (Bal et al., 2010), confirming that it should act as a DN.

Figure 1.

ΔA-AKAP79 retains undiminished interaction with KCNQ2-4 channels. A, Structure of ΔA-AKAP79, with the important domains labeled. B, Shown are images of CHO cells under TIRF illumination expressing CFP-tagged KCNQ2-4 and YFP-tagged ΔA-AKAP79, using 442 or 514 nm laser lines. Images of CFP (left, in rainbow pseudocolor) and YFP (right, in yellow pseudocolor) emissions are shown before or after YFP photobleach, as labeled. C, Bars show the percentage increase in CFP emission after YFP photobleach for the groups in B, as well as for the membrane-targeted CFP-YFP tandem (Rho-pYC) and for ECFP-M + AKAP79-YFP that serve as our positive and negative FRET controls, respectively.

We then performed voltage-clamp experiments to study the effect of overexpression of ΔA-AKAP79 in sympathetic neurons. In these experiments, rat SCG neurons were transfected with either enhanced GFP (EGFP)-tagged ΔA-AKAP79, or EGFP only as a control, by the biolistic gene delivery system, and studied under perforated-patch voltage clamp. The suppression of I_M was measured by application of the M₁ muscarinic, B₂ bradykinin, P2Y purinergic, or AT₁ angiotensin II receptor agonists oxo-M, BK, UTP, or AngII, respectively. Since AKAP79 “sensitizes” KCNQ2/3 and KCNQ2-5 channels to muscarinic inhibition, manifested as shifts in the dose–response relations to lower concentrations, we used concentrations of these agonists that yield half-maximal inhibition of the current (i.e., the IC₅₀), at which concentrations the responses should best detect changes in the dose–response relations, and most accurately indicate effects of AKAP79/150. Normalized I_M amplitudes, quantified as the amplitude of the time-dependent relaxation during the step to −60 mV, are plotted from a rat SCG neuron transfected with EGFP only (Fig. 2A,B), or with EGFP-tagged ΔA-AKAP79 (Fig. 2C,D). The insets show examples of M-current traces before and after application of agonists. In the neuron expressing ΔA-AKAP79 (Fig. 2C,D), the suppressions of I_M by oxo-M (1 μm) and AngII (500 nm) were severely reduced, whereas the responses to BK (100 nm) and UTP (2 μm) were unaltered, compared with the control neuron (Fig. 2A,B). Such data are summarized in Figure 2E. For neurons transfected with EGFP only, the suppressions of current amplitudes by UTP, BK, oxo-M, or AngII were 45 ± 3% (n = 10), 37 ± 3% (n = 5), 50 ± 1% (n = 9), and 45 ± 3% (n = 6), respectively. In ΔA-AKAP79-transfected neurons, the suppressions were 43 ± 6% (n = 7), 36 ± 3% (n = 5), 26 ± 2% (n = 14, p < 0.001), and 25 ± 2% (n = 4, p < 0.01), respectively. These data suggest that AKAP79/150 action is involved in muscarinic and angiotensin, but not in bradykinin or purinergic suppression of neuronal M current.

Figure 2.

Overexpression of dominant-negative AKAP79 suppresses M current inhibition by muscarinic and angiotensin II receptors. A–D, Plotted are normalized M-currents during the experiment from rat SCG neurons transfected with EGFP only (A, B) or with EGFP-tagged ΔA-AKAP79 (C, D), while oxo-M (1 μm), BK (100 nm), UTP (2 μm), angiotensin II (500 nm), or XE991 (10 μm) was bath-applied, during the periods shown by the bars. Representative current traces are shown in the insets, using the indicated voltage protocol. E, Bars are summarized data for the groups of cells as in A–D. **p < 0.01, ***p < 0.001.

AKAP150^−/− neurons display suppressed M-current inhibition induced by muscarinic and angiotensin II receptors

The ΔA-AKAP79 experiments on rat SCG neurons presented so far indicate marked receptor specificity in AKAP79/150 actions, with the scaffold protein being used by M₁ and AT₁, but not by B₂ or P2Y, receptors. To probe the role of AKAP79/150 in M-channel modulation in a more direct way, we used colonies of AKAP150 KO mice. Previous work showed neurons from AKAP150 KO mice to have reduced suppression of I_M by stimulation of M₁, but not B₂, receptors, and AKAP150^−/− mice to demonstrate increased resistance to pilocarpine-induced seizures (Hoshi et al., 2003, 2005, 2010; Tunquist et al., 2008). In our work, AKAP150^+/− heterozygotes were bred, and SCG neurons were prepared from AKAP150^+/+ (WT) and AKAP150^−/− (KO) littermates identified by tail-snip genotyping, to compare the modulation of I_M in neurons cultured from such WT and KO mice. We first confirmed the lack of AKAP150 expression in SCG neurons isolated from AKAP150^−/− mice by immunostaining and confocal microscopy. Cultured neurons from WT or KO mice were fixed and immunolabeled with anti-AKAP150 antibodies and an antibody against tyrosine hydroxylase as a sympathetic neuron marker. Figure 3A shows confocal images of SCG neurons from AKAP150^+/+ mice showing robust expression of AKAP150 protein. As controls, there was no significant labeling when the anti-AKAP150 antibody was preadsorbed with a tenfold molar excess of the immunizing peptide used to raise the antibody or when the primary antibody was omitted. In contrast, SCG neurons from AKAP150^−/− mice did not show any AKAP150 labeling.

Figure 3.

AKAP150^−/− neurons display suppressed M-current inhibition induced by muscarinic and angiotensin II receptors. A, SCG neurons isolated from AKAP150^−/− mice lack AKAP150 expression. Shown are confocal fluorescence images of fixed SCG neurons isolated from AKAP150^+/+ or AKAP150^−/− mice immunostained for AKAP150 and tyrosine hydroxylase (TH), visualized using Rhodamine Red- and FITC-conjugated secondary antibodies, respectively. Cells were stained with either the anti-AKAP150 antibody (1° ab), the antibody preadsorbed with a tenfold excess of the peptide used to raise the antibody (1° ab + BP), or the secondary antibody only (2° ab only). B–E, Plotted are normalized M-currents during the experiment from SCG neurons isolated from AKAP150^+/+ mice (B, C) or from AKAP150^−/− mice (D, E), while oxo-M (0.3 μm), BK (100 nm), UTP (2 μm), angiotensin II (500 nm), or XE991 (10 μm) was bath-applied, as indicated by the bars. Representative current traces are shown in the insets, using the voltage protocol as in Figure 2. F, Bars are summarized data for the groups of cells as in B–E. **p < 0.01, ***p < 0.001.

With confidence in the accuracy of the genotyping of our AKAP150 WT and KO mouse colonies, we used SCG neurons cultured from these mice to probe the receptor specificity in AKAP150 action. Figure 3B–E shows examples of perforated-patch voltage-clamp experiments on neurons from AKAP150^+/+ and AKAP150^−/− mice. Shown are normalized M-current amplitudes from a SCG neuron from a WT mouse (Fig. 3B,C) or from a KO mouse (Fig. 3D,E). In the insets are representative M-current traces, using the “classic” M-current voltage protocol. Again, we tested the suppression of M-current by receptor agonists at concentrations near their respective IC₅₀ values, so as to most sensitively assay AKAP150 action. Consistent with our results in rat SCG neurons using the DN approach, the suppression of I_M by oxo-M (0.3 μm) and AngII (500 nm) was significantly reduced in neurons isolated from AKAP150^−/− mice, compared with those isolated from AKAP150^+/+ mice. On the other hand, the responses to BK (100 nm) and UTP (2 μm) were not significantly different in the two populations of neurons. For neurons isolated from AKAP150^+/+ mice, suppressions of I_M by oxo-M, BK, UTP or AngII, were 63 ± 3% (n = 19), 52 ± 4% (n = 11), 47 ± 4% (n = 8), and 65 ± 6% (n = 6), respectively. For neurons isolated from AKAP150^−/− mice, the suppressions were 37 ± 3% (n = 13, p < 0.001), 53 ± 3% (n = 9), 54 ± 5% (n = 7), and 40 ± 3% (n = 7, p < 0.01), respectively. Since M₁ and AT₁ receptors depress I_M in SCG similarly, by depletion of PIP₂ in the membrane (Suh and Hille, 2002; Ford et al., 2003; Zhang et al., 2003), whereas B₂ and P2Y receptors depress I_M not via depletion of PIP₂, but rather by intracellular Ca²⁺ signals in concert with CaM (Zaika et al., 2007; Hernandez et al., 2008), our data suggest that AKAP79/150 is only involved in I_M modulation by receptors that deplete PIP₂, not those acting via Ca²⁺/CaM.

Transfection of AKAP79 in AKAP150^−/− neurons “rescues” muscarinic suppression of M current

The patch-clamp data presented above indicate that lack of functional AKAP150 in the neurons reduces muscarinic and angiotensin, but not bradykinin or purinergic, suppression of I_M. If these effects are truly the result of the lack of AKAP150, heterologous expression of functional AKAP79 or AKAP150 in the neurons from AKAP150^−/− mice should rescue the responses to stimulation of M₁ and AT₁ receptors, without affecting the responses to stimulation of B₂ or P2Y receptors. To test this, we chose to examine the suppression of I_M by M₁ and B₂ receptors, to represent receptors for which the lack of functional AKAP79/150 did, or did not, affect suppression of I_M, respectively.

In these experiments, SCG neurons isolated from AKAP150^−/− mice were transfected with either EGFP-tagged AKAP79 or two controls, by the biolistic gene delivery system. The first control was EGFP-tagged ΔA-AKAP79, which cannot recruit PKC and so should not rescue AKAP150 action. The other control was a different AKAP, AKAP15, which has been shown to facilitate PKC phosphorylation of L-channels (Hulme et al., 2002, 2003), but has not been tested against M channels. Such neurons were studied under perforated-patch voltage clamp (Fig. 4). Normalized I_M amplitudes are plotted from SCG neurons isolated from AKAP150^−/− mice transfected with EGFP-tagged AKAP79 (Fig. 4A), EGFP-tagged ΔA-AKAP79 (Fig. 4B), or AKAP15 + EGFP (Fig. 4C). In the insets are representative examples of M-current traces, before and after application of agonists (oxo-M, 0.3 μm; BK, 100 nm). Transfection of AKAP79, but not ΔA-AKAP79 or AKAP15, in the neurons from AKAP150^−/− mice restored the muscarinic suppression of I_M. However, none of these transfections had any effect on the BK suppression of I_M. These data are summarized in Figure 4D. In neurons from AKAP150^−/− mice transfected with AKAP79, the suppressions of I_M by oxo-M or BK were 73 ± 4% (n = 10) and 56 ± 4% (n = 8), respectively. In those transfected with ΔA-AKAP79, the suppressions of I_M by oxo-M or BK were 37 ± 2% (n = 7, p < 0.001) and 59 ± 3% (n = 8), respectively. In neurons from AKAP150^−/− mice transfected with AKAP15, the suppressions of I_M by oxo-M or BK were 34 ± 2% (n = 15, p < 0.001) and 61 ± 3% (n = 15), respectively. Notice that the rescued muscarinic suppression of I_M in AKAP79-transfected AKAP150^−/− neurons was significantly larger than those in neurons isolated from AKAP150^+/+ mice (p < 0.05), probably because AKAP79 was overexpressed in these neurons. However, such AKAP79 overexpression did not have any effect on the BK suppression of I_M, further confirming that AKAP79/150 is only involved in M₁, but not in B₂ receptor modulation of M channels. Finally, AKAP15 cannot mediate PKC phosphorylation of M channels, since expression of AKAP15 did not rescue the lack of AKAP150 in the neurons.

Figure 4.

Transfection of AKAP79 in AKAP150^−/− neurons rescues the suppression of M-current by muscarinic receptors. A–C, Plotted are normalized M-currents during the experiment from SCG neurons isolated from AKAP150^−/− mice transfected with AKAP79 (A), EGFP-tagged ΔA-AKAP79 (B), or AKAP15 (C), while oxo-M (0.3 μm), BK (100 nm), or XE991 (10 μm) was bath-applied, as indicated by the bars. Representative current traces are shown in the insets, using the voltage protocol as in Figure 2. D, Bars are summarized data for the groups of cells as in A–C. ***p < 0.001.

AKAP79 and KCNQ channels are closely associated with muscarinic and angiotensin II receptors

The role of AKAPs in orchestrating intracellular pathways typically involves inclusion of the receptor that initiates the signal into the signaling complex. For instance, AKAP79/150 assembles the β-adrenergic receptor with the L-type Ca²⁺ channel that is the target of recruited PKA (Bauman et al., 2004; Wong and Scott, 2004). For actions on KCNQ channels, the M₁, but not the B₂, receptor was shown to physically associate with AKAP150, correlating with the involvement of AKAP150 in muscarinic, but not bradykinin, modulation of I_M (Hoshi et al., 2005). The patch-clamp experiments presented here so far indicate functional involvement of AKAP79/150 in the actions of M₁ and AT₁ receptors, but not B₂ or P2Y receptors. To probe the physical association between AKAP79/150 and M₁, B₂, P2Y, and AT₁ receptors, we performed further TIRF/FRET experiments on CHO cells transfected with CFP-tagged receptors and YFP-tagged AKAP79_1-153. The latter contains the proximal 153 residues of AKAP79, which includes the A, B, and C domains (Fig. 1A) sufficient for physical association with KCNQ channels, CaM and PIP₂ (Dell'Acqua et al., 1998; Hoshi et al., 2003; Bal et al., 2010), and was sufficient to rescue AKAP150 activity in RNAi experiments (Hoshi et al., 2005).

We again used the donor dequenching paradigm on individual living cells, in which CFP emission is compared before and after selective photobleaching of YFP. Figure 5A shows representative images of CFP and YFP before and after YFP photobleach. As a P2Y receptor to test, we chose the P2Y₆ subtype, which has been shown to predominate in rat SCG (Calvert et al., 2004). In the cells cotransfected with YFP-tagged AKAP79_1-153 and CFP-tagged M₁ or AT₁ receptors, the CFP emission was significantly stronger after YFP photobleach, indicating robust FRET. However, in cells cotransfected with CFP-tagged B₂ or P2Y₆ receptors, the FRET was much less. Such data are summarized in Figure 5B. The FRET efficiency for M₁, AT₁, B₂, and P2Y₆ receptors with AKAP79_1-153 was 17 ± 2% (n = 23), 21 ± 4% (n = 9), 6 ± 3% (n = 11, p < 0.01), and 7 ± 2% (n = 12, p < 0.01), respectively. Thus, in parallel with the receptor specificity seen in neurons, the interaction between AKAP79 and P2Y₆ or B₂ receptors was much less than that for M₁ and AT₁ receptors.

Figure 5.

AKAP79 and KCNQ channels are associated with muscarinic and angiotensin II receptors. A, Shown are images of CHO cells under TIRF illumination expressing CFP-tagged M₁, B₂, AT₁, or P2Y₆ receptors and YFP-tagged AKAP79_1-153, using 442 or 514 nm laser lines. Images of the CFP (left, in rainbow pseudocolor) and YFP (right, in yellow pseudocolor) emissions are shown before or after YFP photobleach, as labeled. Note the brighter CFP emission (warmer colors) after YFP photobleach for M₁ and AT₁, but not for B₂ or P2Y₆ receptors. B, Bars show the percentage increase in CFP emission after YFP photobleach for the groups in A, as well as for cells cotransfected with WT or DN CaM. C, Bars show the percentage increase in CFP emission after YFP photobleach for cells expressing CFP-tagged M₁, B₂, P2Y₆, or AT₁ receptors and YFP-tagged KCNQ2 or KCNQ3 channels, *p < 0.1,**p < 0.01, ***p < 0.001.

We previously found calcified CaM, but not Ca²⁺-free CaM, to disrupt the interactions between KCNQ2-5 channels and AKAP79 (Bal et al., 2010). Hence, it was of interest to know whether CaM might similarly influence the interactions between AKAP79 and the receptors. Thus, we performed parallel TIRF/FRET measurements on cells cotransfected with either WT or DN CaM, in which all four Ca²⁺-binding sites have been mutated as in previous work (Bal et al., 2010), and compared those data with those from cells without overexpressed WT or DN CaM. However, we did not observe any effect of cotransfection of either WT or DN CaM. For cells cotransfected with YFP-tagged AKAP79_1-153, WT CaM, and CFP-tagged M₁, AT₁, B₂, or P2Y₆ receptors, the FRET efficiencies were 18 ± 1% (n = 16), 20 ± 3% (n = 11), 0 ± 4% (n = 12), and 9 ± 2% (n = 11). For cells cotransfected with YFP-tagged AKAP79_1-153, DN CaM, and CFP-tagged M₁, AT₁, B₂, or P2Y₆ receptors, the FRET efficiencies were 15 ± 2% (n = 14), 18 ± 2% (n = 17), 4 ± 2% (n = 15), and 1 ± 3% (n = 15). Since CaM is intimately associated with KCNQ channels in the membrane (Bal et al., 2010), we also performed a FRET experiment on CHO cells coexpressing KCNQ2 together with WT CaM, YFP-tagged AKAP79_1-153, and CFP-tagged M₁ receptors. However, we still did not observe any effect of CaM on the interaction of AKAP79/150 with M₁ receptors. For those cells, the FRET efficiency was 16 ± 2% (n = 15). Thus, CaM does not seem to play a regulatory role in the association of AKAP79/150 with its cognate G_q/11-coupled receptors.

Our data presented so far show that AKAP79/150 is physically and functionally associated with M₁ and AT₁ receptors. Considering our previous work indicating the physical association between AKAP79/150 and KCNQ2-5 channels (Bal et al., 2010), a prediction would be that KCNQ channels are also physically associated with the same G_q/11-coupled receptors that associate with AKAP79; i.e., that AKAP79 receptors and KCNQ channels are components of a macromolecular complex. Thus, TIRF/FRET experiments were performed on CHO cells transfected with YFP-tagged KCNQ2 or KCNQ3 and CFP-tagged M₁, AT₁, B₂, or P2Y₆ receptors. Although we did not cotransfect AKAP79 or AKAP150 as well in these experiments, CHO cells do express modest levels of endogenous AKAP150 (data not shown), as found previously (Hoshi et al., 2003). For both KCNQ2 and KCNQ3, there was substantial FRET with M₁ or AT₁ receptors, but almost none with B₂ or P2Y₆ receptors. These data are summarized in Figure 5C. For cells cotransfected with YFP-tagged KCNQ2, and CFP-tagged M₁, AT₁, B₂, or P2Y₆ receptors, the FRET efficiencies were 20 ± 2% (n = 16), 19 ± 2% (n = 16), 2 ± 1% (n = 13, p < 0.001), and 4 ± 2% (n = 15, p < 0.001). For cells cotransfected with YFP-tagged KCNQ3 and CFP-tagged M₁, AT₁, B₂, or P2Y₆ receptors, the FRET efficiencies were 20 ± 2% (n = 16), 20 ± 3% (n = 14), 2 ± 1% (n = 17, p < 0.001), and 0 ± 2% (n = 15, p < 0.001). Our data suggest that AKAP79/150 assembles KCNQ channels and M₁/AT₁, but not B₂/ P2Y₆, receptors, to form a signaling complex. It is important to note, however, that these data do not demonstrate physical binding of M₁ or AT₁ receptors to the channels, only that they are localized close enough together to exhibit FRET. Indeed, this FRET may rather be the result of common binding of both receptors and channels to the same AKAP150 molecules.

AKAP79/150 is not involved in the muscarinic modulation of N-type Ca²⁺ channels

G_q/11-coupled M₁ receptors also inhibit N- and L-type Ca²⁺ channels by a similar voltage-independent, PTX-insensitive “slow pathway” (Hille, 1994), which is thought also to be largely the result of PIP₂ depletion (Wu et al., 2002; Gamper et al., 2004; Suh et al., 2010). Although arachidonic acid, which is also produced downstream of PIP₂ breakdown by PLC, also seems to play a role in muscarinic suppression of both N- and L-type Ca²⁺ channels (Roberts-Crowley et al., 2009), involvement of PKC (as the downstream molecule of DAG) in G_q/11-mediated regulation of these channels has not been identified. To ask whether AKAP79/150 might play a role in M₁ receptor action on N- an L-type Ca²⁺ channels, we performed perforated-patch experiments on SCG neurons similar to those performed for M channels. In these experiments, we used oxo-M at 2 and 10 μm, concentrations that induce half-maximal and supramaximal inhibitions, respectively, of N- and L-type Ca²⁺ current (Bernheim et al., 1992). Thus, responses at these two concentrations of oxo-M should be sensitive to any AKAP79/150 effect on the dose–response relationship or maximal suppression, respectively, of muscarinic suppression.

We first tested the involvement of AKAP79/150 in the muscarinic modulation of N-current. Although our AKAP150^+/+ and AKAP150^−/− mouse colonies are convenient, we chose to use the well studied rat SCG neurons for these experiments because their N-type Ca²⁺ current constitutes most of the I_Ca (>90%) (Plummer et al., 1989; Regan et al., 1991) that can be isolated without a complex cocktail of blockers of L- and P/Q-type Ca²⁺ currents. In these experiments, rat SCG neurons were preincubated with pertussis toxin (100 ng/ml) overnight to inactivate membrane-delimited modulation of N-current by G_o/i-coupled M₄ receptors (Beech et al., 1992). Neurons were transfected with either EGFP-tagged ΔA-AKAP79 or EGFP only as a control, by the biolistic gene delivery system, and I_Ca recorded by perforated patch clamp. Figure 6 shows representative experiments. The neuron transfected with ΔA-AKAP79 did not display reduced suppression of N-current by either 2 or 10 μm oxo-M (Fig. 6B), compared with the neuron transfected with EGFP only (Fig. 6A). Such data are summarized in Figure 6C. For ΔA-AKAP79-transfected neurons, the suppression of I_Ca by 2 or 10 μm oxo-M was 22 ± 3% (n = 7) and 47 ± 6% (n = 7), respectively. For neurons transfected with EGFP only, the suppressions were 24 ± 3% (n = 8) and 54 ± 6% (n = 8), respectively. Therefore, AKAP79/150 is not involved in G_q/11-coupled muscarinic modulation of N-type Ca²⁺ channels.

Figure 6.

AKAP79/150 is not involved in the muscarinic modulation of N-type Ca²⁺ channels. A, B, Plotted are normalized I_Ca during the experiment from rat SCG neurons transfected with EGFP only (A) or EGFP-tagged ΔA-AKAP79 (B), while half-maximal or supramaximal concentrations of oxo-M (2 or 10 μm), or Cd²⁺ (100 μm) were bath-applied, as indicated by the bars. Representative current traces are shown in the insets, using the indicated voltage protocol. C, Bars are summarized data for the groups of cells as in A and B.

AKAP79/150 mediates modulation of L-type Ca²⁺ channels by PKA, but not by muscarinic receptors

Modulation of L-type Ca²⁺ channels by cAMP/PKA has been intensively studied in the four decades since its discovery in the heart (Reuter, 1967; Tsien et al., 1972). L-type channels in the hippocampus, whose predominant isoform is Ca_V1.2 (Hell et al., 1993), are also modulated by PKA action (Kavalali et al., 1997; Hoogland and Saggau, 2004), which is orchestrated by AKAP79/150 (Oliveria et al., 2007). Thus, we thought it likely that L-current in SCG neurons is also modulated by PKA, and AKAP79/150 is likely to be involved. We used our AKAP150^+/+ and AKAP150^−/− mouse colonies to study this question, because the fraction of I_Ca in mouse SCG neurons that is L-type is much larger (∼20% of I_Ca) (Shapiro et al., 1999) than in rat (<5%) (Plummer et al., 1989). L-type I_Ca was isolated by adding the L-type Ca²⁺ channel agonist FPL 64176 to the bath solution to elicit a long-lasting tail-current comprised entirely of L-type current (Liu et al., 2006, 2008).

Our first use of our mouse/FPL system was to test whether AKAP79/150 mediates PKA phosphorylation of the L-current in mouse SCG neurons (Fig. 7A–C). Under perforated-patch voltage clamp, a brief depolarization to 10 mV elicits a large inward current, followed by a uniformly fast tail-current when the membrane is repolarized to −40 mV, both from multiple types of I_Ca. Application of FPL 64176 (1 μm) induces a large, slow component of the tail-current, whose amplitude we monitored as an assay of L-type I_Ca. As seen previously (Liu et al., 2006, 2008), we observed FPL 64176 to induce slowed current inactivation during the test pulse, but we did not explore this phenomenon further. Normalized tail-current amplitudes measured at a point 15 ms into the step to −40 mV are plotted from a neuron isolated from an AKAP150^+/+ mouse (Fig. 7A) or from an AKAP150^−/− mouse (Fig. 7B). The diterpene forskolin (FSK, 20 μm) was used to activate PKA by elevating cAMP levels (Hedin and Rosberg, 1983). In the control neuron, the increase in the slowed tail-current by FSK was large, but in the neuron from the AKAP150^−/− mouse, the response was small. Such data are summarized in Figure 7C. For neurons isolated from AKAP150^+/+ or AKAP150^−/− mice, the FSK-induced augmentation of the L-current was 67 ± 9% (n = 5) and 16 ± 4% (n = 8, p < 0.001), respectively. We note that the absence of AKAP150 in the KO neurons did not totally abolish FSK-induced augmentation of the L-current, which could be due either to coincidental spatial proximity of signaling components, or to the actions of AKAP15/18, which has been shown to also organize PKA phosphorylation of Ca_V1.2 channels (Hulme et al., 2002, 2006).

Figure 7.

AKAP79/150 mediates modulation of L-type Ca²⁺ channels by PKA, but not by muscarinic receptors. A, B, D, E, Plotted are the normalized amplitudes of the long-lasting tail-current reflecting L-type Ca²⁺ currents, during experiments on mouse SCG neurons isolated from AKAP150^+/+ (A, D) or AKAP150^−/− mice (B, E), while FSK (20 μm) or half-maximal or supramaximal concentrations of oxo-M (2 or 10 μm) or Cd²⁺ (100 μm) were bath-applied, as indicated by the bars. Representative current traces are shown in the insets, using the indicated voltage protocol. The dotted line shows the time of measurement of the tail current. C, Bars are summarized data for the groups of cells as in A and B. ***p < 0.001. F, Bars are summarized data for the groups of cells as in D and E.

We then tested the involvement of AKAP79/150 in the M₁ receptor-mediated modulation of L-current in mouse SCG neurons (Fig. 7D–F). Again, neurons were pretreated overnight with PTX to block G_o/i-mediated actions. We again used FPL 64176 to isolate the L-current as a long-lasting tail-current, and used half-maximal and supramaximal concentrations of oxo-M (2 and 10 μm) to stimulate M₁ receptors. Examples of experiments on a neuron from an AKAP150^+/+ mouse (Fig. 7D) or from an AKAP150^−/− mouse (Fig. 7E) are shown. We found that the absence of AKAP150 did not observably affect the response to either 2 or 10 μm oxo-M. For neurons isolated from AKAP150^+/+ mice, the suppressions of L-current amplitudes by 2 or 10 μm oxo-M were 35 ± 3% (n = 7), and 48 ± 5% (n = 7), respectively. For neurons isolated from AKAP150^−/− mice, the suppressions were 34 ± 4% (n = 8) and 47 ± 5% (n = 8), respectively. These data suggest that AKAP79/150 is not involved in the muscarinic modulation of L-type Ca²⁺ channels in sympathetic ganglia.

Discussion

In this study, we demonstrate high receptor specificity in AKAP79/150 actions on I_M modulation in sympathetic neurons. Occlusion or genetic deletion of AKAP150 reduced the efficacy of G_q/11-coupled M₁ and AT₁, but not B₂ or P2Y, receptors. The deficit was fully rescued by functional AKAP79 expression in the neurons, but not by AKAP15. The TIRF/FRET data showing receptor-specific physical association with AKAP79 are in accord with the functional measurements. These results are consistent with previous studies showing involvement of AKAP79/150-mediated PKC phosphorylation in I_M modulation by stimulation of M₁, but not B₂, receptors in SCG neurons (Hoshi et al., 2003, 2005, 2010). Interestingly, this AKAP79/150 action is not involved in the G_q/11-mediated muscarinic modulation of N- or L-type Ca²⁺ channels, although they share similar mechanisms of muscarinic suppression as M channels, via PLC activation and PIP₂, break down in the membrane. As in hippocampal pyramidal neurons, FSK-induced upregulation of L-current in mouse SCG neurons heavily depends on the presence of AKAP150. Figure 8 summarizes our results showing involvement of AKAP79/150 in G_q/11-coupled muscarinic modulation of M current, but not N- or L-current. It also shows our model for receptor specificity in Ca²⁺ signals and AKAP79/150 action in SCG neurons, with the two microdomains regulating M channel activity proposed in this discussion.

Figure 8.

Model for receptor specificity of Ca²⁺ signals and AKAP79/150 actions in SCG neurons. M channel, and L-type and N-type Ca²⁺ channels are all regulated by G_q/11-coupled muscarinic M₁ receptors, which activate PLCβ, hydrolyzing PIP₂ into IP₃ and DAG. Depletion of PIP₂ is primarily responsible for the suppression of M-, N-, and L-current. M channels are also regulated by PLC-linked bradykinin B₂ and purinergic P2Y receptors, which induce significant cytoplasmic Ca²⁺ signals as a result of their spatial colocalization with IP₃ receptors. The Ca²⁺ signals activate CaM to suppress M-channel activity. Similar to M₁ receptors, angiotensin II AT₁ receptors suppress I_M by PIP₂ depletion. Another mechanism that inhibits I_M activity is AKAP79/150-mediated PKC phosphorylation activated by DAG, which is only involved in M₁ and AT₁ receptor modulation. The data in this paper suggest two distinct microdomains regulating M-channel activity, as shown here. In addition to the already identified microdomain that contains B₂/P2Y receptors in the plasma membrane and IP₃ receptors on the ER (light aqua), our data suggest a novel microdomain including KCNQ2/3 channels, AKAP79/150, PKC, and M₁/AT₁ receptors (light yellow). As for Ca_V1.2, AKAP79/150 is also involved in the PKA regulation of Ca_V1.3 channels, by anchoring PKA and CaN to the channels, but is not involved in G_q/11 action (light purple).

The receptor specificity suggested by previous work (M₁ vs B₂) (Hoshi et al., 2005) made us inquire more deeply into the mechanism of this phenomenon. Indeed, our results showing parallel specificity in AT₁ versus P2Y receptor actions strongly suggest that AKAP79/150 action correlates with receptor mechanism. Thus, the “mode 1” receptors (M₁ and AT₁ types) suppressing I_M by PIP₂ depletion use AKAP79/150, whereas the “mode 2” receptors (B₂ and P₂Y types), which induce intracellular Ca²⁺ (Ca²⁺_i) signals and depress I_M via Ca²⁺/CaM binding (Hernandez et al., 2008), do not. Our previous study using signaling components reconstituted in cells showed Ca²⁺/CaM to disrupt AKAP79/KCNQ interactions, and to prevent AKAP79/150-mediated sensitization of KCNQ2/3 channels to muscarinic suppression (Bal et al., 2010). Based on these observations and the overlapping binding sites of AKAP150 and CaM in the KCNQ channel C terminus (Wen and Levitan, 2002; Yus-Najera et al., 2002; Gamper and Shapiro, 2003; Haitin and Attali, 2008), we have hypothesized a mechanism underlying the receptor specificity of AKAP79/150 actions in sympathetic neurons: Ca²⁺/CaM, produced by B₂ and P2Y, but not M₁ or AT₁, receptor activation competes with AKAP79/150 binding to KCNQ channels, thus preventing subsequent PKC phosphorylation. This hypothesis predicted AKAP79/150 to sensitize I_M suppression by AT₁, but not by P2Y receptors, which has been confirmed here on both rat and mouse SCG. In accord with this idea, it was recently shown that phospho-mimicking mutants, or in vitro phosphorylation, of KCNQ2 no longer retained CaM in the channel complex (N. Hoshi, Society for Neuroscience 2009 annual meeting). Thus, both biochemical and functional studies support our competition hypothesis of CaM-mediated disruption between AKAP79/150 and KCNQ channels, either via mass action competition between Ca²⁺/CaM and AKAP70/150 molecules, or via CaM-mediated reduction of channel affinity with AKAP79/150, or both. Certainly, it will require direct biochemical analysis of affinities between KCNQ channels, AKAP79/150 and CaM, to test whether functional CaM induces changes in channel affinity for AKAP79/150.

In SCG neurons, stimulation of B₂ and P2Y, but not M₁ or AT₁, receptors induces IP₃-mediated Ca²⁺_i signals (Beech et al., 1991; Cruzblanca et al., 1998; Delmas and Brown, 2002; Calvert et al., 2004; Zaika et al., 2007). This receptor specificity has been postulated because of clustering of plasma membrane B₂ and P₂Y, but not M₁ or AT₁, receptors with ER membrane IP₃ receptors; thus, only IP₃ produced near the IP₃ receptors can cause Ca²⁺ release (Delmas et al., 2002, 2004, Hernandez et al., 2008; Zaika et al., 2011). Since AKAP79/150 assembles another microdomain containing PKC and KCNQ channels, there appears to be two distinct signal complexes regulating M-channel activity by mode 1 or mode 2 receptors, defining two local events. The first event involves PKC phosphorylation of KCNQ channels (in the same complex with AKAP79/150), whereas the other determines IP₃-mediated Ca²⁺ release. Since M₁/AT₁ receptors, but not B₂/P2Y receptors, are physically associated with AKAP79/150 and KCNQ2/3 channels, this suggests that mode 1 receptors, but not mode 2 receptors, are in the same microdomain with KCNQ channels, although our results do not demonstrate a direct physical interaction. This notion is supported by the TIRF/FRET data in Figure 5C. Since mode 2 receptors inhibit I_M via intracellular Ca²⁺ and CaM, another conclusion is that the Ca²⁺/CaM signal must be “global” to be able to reach, and act on, M channels, whereas the PKC phosphorylation of the channels must be “local.” Thus, we suggest the existence of two distinct microdomains containing the components of separate, yet synergistic, intracellular signals acting on M channels (Fig. 8).

We do not know whether PLC is part of the AKAP79/150-orchestrated signaling complex. If it is, might this mean that the PIP₂ depletion induced by mode 1 receptors is local to M-channels? The issue of whether PIP₂ depletion is local or global in native cells is very controversial. The Ho laboratory has suggested localized depletion of PIP₂, because of its suggested low membrane mobility, to be a mechanism underlying receptor-specific modulation of channels in cardiomyocytes (Cho et al., 2005a,b, 2006). However, other laboratories have reached the opposite conclusion by cell-attached patch-recording and by analysis of the mobility of exogenous fluorescently tagged and endogenous PIP₂ molecules in the membrane (Selyanko et al., 1992; Marrion, 1993; Yaradanakul and Hilgemann, 2007). In addition, our laboratory and the Hille laboratory have successfully modeled muscarinic suppression of KCNQ channels in tissue culture cells to result from global PIP₂ depletion (Suh et al., 2004; Hernandez et al., 2009; Falkenburger et al., 2010a,b).

The signaling complex acting on hippocampal L-current mediated by AKAP79/150 includes β-adrenergic receptors, PKA, and CaN (Davare et al., 2001; Oliveria et al., 2007). Might there be a relevant phosphatase in this KCNQ-AKAP79/150 complex as well; and if so, what phosphatase might it be? Unfortunately, we lack this information, but we believe it likely, since it seems necessary to have a mechanism for dephosphorylating M channels once the receptor-mediated signal is turned off. A reasonable candidate is CaN, since it mediates dephosphorylation of L-current, as well as of Ca²⁺/CaN-sensitive NFAT (nuclear factor of activated T-cells) family transcription factors expressed in SCG neurons (Hernández-Ochoa et al., 2007).

Although PIP₂ depletion is also heavily involved in G_q/11-mediated muscarinic depression of N- and L-current (Wu et al., 2002; Gamper et al., 2004; Suh et al., 2010), AKAP79/150 seems not participate in this slow pathway regulation of N- and L-current (Bernheim et al., 1991). Although N-type Ca²⁺ channels can be phosphorylated by PKC in vitro at positions 422, 774, and 898 (Dai et al., 2009), and PKC phosphorylation has been shown to increase N-current in frog SCG and cells heterologously expressing the cloned channels (Yang and Tsien, 1993; Stea et al., 1995), our results are in accord with the previous study that PKC activation by phorbol esters failed to attenuate muscarinic suppression of N-current in rat SCG (Shapiro et al., 1996). Indeed, the upregulation of N-current by PKC has to do with antagonizing inhibition of N-current by Gβγ (Zamponi et al., 1997; Dolphin, 1998), acting via the membrane-delimited, voltage-dependent, and PTX-sensitive pathway (Bean, 1989; Hille, 1994; Herlitze et al., 1996; Ikeda, 1996), but not the voltage-independent, PTX-insensitive, slow pathway we study here. AKAP150 is reported to coimmunoprecipitate with Ca_V1.2 from brain (Hall et al., 2007; Oliveria et al., 2007), and the coimmunoprecipitation of PKA with Ca_V1.2 is dramatically reduced in AKAP150^−/− mice (Hall et al., 2007). The L-current in rodent SCG has been shown to be predominantly caused by the Ca_V1.3 isoform (Lin et al., 1996). Here, the absence of AKAP150 caused a dramatic decrease of FSK-induced augmentation of L-current in AKAP150^−/− mice, suggesting a major role of AKAP79/150 in recruiting PKA to Ca_V1.3 channels as well. PKC activation induces an upregulation, a downregulation, or a biphasic effect on Ca_V1.2-mediated cardiac L-current (Catterall, 2000; Kamp and Hell, 2000), and the upregulation of Ca_V1.2 currents is AKAP150 dependent in arterial smooth muscle (Navedo et al., 2008). Ca_V1.2 contains three PKC-phosphorylatable sites in vitro and has been shown to coimmunoprecipitate with PKCα, PKCβ, and PKCγ from brain (Dai et al., 2009). Although our results suggest AKAP79/150 action is not involved in the muscarinic modulation of L-current, further investigation of the role of these phosphorylations is required. In addition, possible interplay between arachidonic acid and PKC could also occur, which has not been studied.