Scaffolding by A-Kinase Anchoring Protein Enhances Functional Coupling between Adenylyl Cyclase and TRPV1 Channel

Background: AKAP79 scaffolds TRPV1 channel and PKA to block channel desensitization.

Results: Adenylyl cyclase (AC) anchoring to TRPV1-AKAP79-PKA complex sensitizes the channel to forskolin and is required for PGE₂ sensitization of TRPV1 in dorsal root ganglia (DRG).

Conclusion: AC scaffolding enhances functional coupling between G_s-coupled agonist and effector.

Significance: AKAP complexes containing AC and PKA facilitate responses to inflammatory mediators.

Abstract

Scaffolding proteins often bring kinases together with their substrates to facilitate cell signaling. This arrangement is critical for the phosphorylation and regulation of the transient receptor potential vanilloid 1 (TRPV1) channel, a key target of inflammatory mediators such as prostaglandins. The protein kinase A anchoring protein AKAP79/150 organizes a multiprotein complex to position protein kinase A (PKA) and protein kinase C (PKC) in the immediate proximity of TRPV1 channels to enhance phosphorylation efficiency. This arrangement suggests that regulators upstream of the kinases must also be present in the signalosome. Here, we show that AKAP79/150 facilitates a complex containing TPRV1 and adenylyl cyclase (AC). The anchoring of AC to this complex generates local pools of cAMP, shifting the concentration of forskolin required to attenuate capsaicin-dependent TRPV1 desensitization by ∼100-fold. Anchoring of AC to the complex also sensitizes the channel to activation by β-adrenergic receptor agonists. Significant AC activity is found associated with TRPV1 in dorsal root ganglia. The dissociation of AC from an AKAP150-TRPV1 complex in dorsal root ganglia neurons abolishes sensitization of TRPV1 induced by forskolin and prostaglandin E₂. Thus, the direct anchoring of both PKA and AC to TRPV1 by AKAP79/150 facilitates the response to inflammatory mediators and may be critical in the pathogenesis of thermal hyperalgesia.

Introduction

A-kinase anchoring proteins (AKAPs)³ are well known for their ability to couple regulatory proteins to downstream effectors. For example, AKAP79 (also known as AKAP5 or mouse AKAP150) scaffolds PKA, PKC, and PP2B (calcineurin) to a large number of different channel effectors (1, 2). In doing so, AKAP79 can facilitate both temporal and spatial signaling. This is due to the concentration of signaling molecules and the enhanced rate of phosphorylation that occurs upon tethering a kinase to its substrate (2, 3).

We have previously shown that AKAPs can also facilitate the formation of complexes containing both PKA and its upstream activator, adenylyl cyclase (AC) (3–6). Additionally, AKAPs facilitate AC-effector complex formation (7, 8); however, the functional significance of such a complex on a bound effector is unknown. To directly address this, we have examined the role of AKAP79-anchored AC on the regulation of the transient receptor potential vanilloid type 1 (TRPV1) channel in dorsal root ganglia (DRG) neurons.

TRPV1 is a calcium-permeable ion channel that responds to chemical stimuli (e.g. capsaicin, the pungent component of hot chili peppers) or thermal changes (>42 °C) for nociceptive signaling. TRPV1 is essential for the development of inflammatory thermal hyperalgesia (9, 10). Like many channels and receptors, TRPV1 is desensitized upon continuous activation or in response to repeated exposures of capsaicin, known as tachyphylaxis (11). Although the exact mechanisms for desensitization are still unclear, one common thread is the balance between phosphorylation and dephosphorylation of the channel. Inflammatory mediators (such as prostaglandins) act to block desensitization, a process often referred to as sensitization. For example, in DRG neurons prostaglandin E₂ (PGE₂) activates the Gα_s-coupled EP4 receptor to increase adenylyl cyclase production of cAMP and the subsequent activation of the cAMP-dependent kinase, PKA (12). PKA directly phosphorylates TRPV1 on Ser-116, blocking TRPV1 desensitization and enhancing the TRPV1 response to capsaicin and other activators (reviewed in Refs. 13 and 14). AKAP anchored PKA is required in this process as the Ht31 peptide (a competitor of PKA-AKAP association) blocks forskolin- and PGE₂-induced TRPV1 sensitization in sensory neurons (15, 16). AKAP79/150 is expressed in ∼80% of TRPV1 positive DRG neurons and directly binds to TRPV1 (17–19). AKAP79/150 is required in the reduction of TRPV1 desensitization by forskolin or PGE₂ in DRG and trigeminal ganglia (17, 18, 20). PKA must be anchored to AKAP150 as knock-in mice lacking the PKA binding site (AKAP150Δ36) are strongly impaired in PGE₂ regulation of TRPV1. Finally, TRPV1 regulation by AKAP150-anchored PKA has important physiological consequences as disruption of this pathway reduces PGE₂-induced thermal hyperalgesia (17).

We demonstrate herein that AKAP79/150 facilitates assembly of an AC5-TRPV1 complex. If AC generates locally elevated pools of cAMP, it should sensitize effectors to lower concentrations of G protein-coupled receptor agonists or direct activators of adenylyl cyclases. However, this had not been directly proven for an AC-AKAP-effector complex. We observe an approximate 100-fold shift in the concentration of forskolin required to block capsaicin-evoked TRPV1 desensitization. Anchoring of AC to the complex also sensitizes the channel to activation by β-adrenergic receptor agonists. In addition, significant AC activity is found associated with TRPV1 in mouse DRG preparations. Dissociation of AC from the AKAP150 complex in isolated DRG neurons blocks the effect of forskolin and PGE₂ on TRPV1 desensitization. Thus, the direct anchoring of both PKA and AC to TRPV1 facilitates the response of the channel to inflammatory mediators, which may be critical for the development of inflammatory pain and thermal hyperalgesia. Anchoring of AC to effector complexes may also serve as a basic paradigm to more efficient and/or spatially confined receptor-effector coupling.

EXPERIMENTAL PROCEDURES

Plasmids

Plasmids AC5-pcDNA3, AKAP79WT-pECFP, AKAP79ΔPKA-pECFP, AKAP79ΔAC-pECFP, and purified recombinant H6-S-tag-AKAP79 polypeptide fragments (77–153) and (109–290) were previously described (7, 21). Mouse TRPV1-pcDNA3 was described in Ref. 22.

Antibodies

Antibodies used were mouse α-AKAP79 (BD Transduction Laboratories), goat α-AKAP150 (Santa Cruz Biotechnology, Santa Cruz, CA), and rabbit α-TRPV1 used for IP (EMD Biosciences, Inc., Billerica, MA) and Western blotting (Alomone Labs, Jerusalem, Israel).

Immunoprecipitation of Adenylyl Cyclase Activity

Immunoprecipitations using α-TRPV1 antibody followed by measurement of associated AC activity (IP-AC assays) were performed as described (4–6). Immunoprecipitants were immediately assayed for adenylyl cyclase activity upon stimulation with 50 μm forskolin and 50 nm Gα_s. IP-AC assays from mouse DRGs were performed using 14–16 whole DRGs. Acetylated cAMP was detected by enzyme immunoassay (ELISA kit from Enzo Life Sciences).

Calcium Imaging

HEK293 cells were transfected using Lipofectamine 2000 and plated on 18-mm coverslips. After 26–36 h, cells were loaded with 3 μm fura-2 AM in extracellular solution for 40 min at room temperature in the dark. extracellular solution contained (in mm): 140 NaCl, 5 KCl, 2 CaCl₂, 1 MgCl₂, 10 HEPES, 10 glucose (pH adjusted to 7.4 with NaOH and osmolarity to 310 mosm with d-mannitol). After washing twice, coverslips were placed on a microscope stage in a 0.5-ml recording chamber and constantly perfused with extracellular solution at a flow rate of 10 ml/min. Time-lapse imaging was acquired (excitations at 340 and 380 nm; emission at 510 nm) using a TE2000 Nikon microscope with a DG4 xenon light source and two CoolSNAP cameras (Roper Scientific) controlled by SlideBook 5 software. After 30–60 s of base line recording, 300 nm capsaicin was applied twice for 10 s separated by a 3-min extracellular solution wash. Where indicated, forskolin or isoproterenol was added at the start of the recording and was present for the duration of the experiment. Only cells showing an initial Ca²⁺ increase with capsaicin were included in analysis. The response ratio (amplitude peak 2/amplitude peak 1) was quantified using at least 20 different cells for each condition from three separate experiments.

Isolation of DRG Neurons

Neurons were isolated from 3–5-week-old C57BL/6 mice as described (23). Neurons were cultured with Ham's nutrient mixture F12 supplemented with 10% fetal bovine serum (Invitrogen), 1% penicillin-streptomycin, and 50 ng/ml nerve growth factor (NGF). Neurons were plated on 12-mm glass coverslips coated with poly-d-ornithine and used 24 h after isolation for patch clamp experiments.

Electrophysiology

Whole-cell patch clamp recordings were performed using an EPC9 amplifier and PatchMaster v2.50 software for data acquisition (HEKA Elektronik). Patch pipettes were pulled from borosilicate glass capillaries (World Precision Instruments) using a horizontal P-97 micropipette puller (Sutter instrument) and then fire-polished using an MF-830 microforge (Narishige Group). The patch pipette resistance ranged from 3 to 6 megaohms when filled with the intracellular solution and placed in extracellular solution. Intracellular solution contained (in mm): 140 CsCl₂, 1 MgCl₂, 3 NaATP, 0.5 NaGTP, 0.2 EGTA, 10 HEPES (pH adjusted to 7.2 with CsOH and osmolarity to 290 mosm with d-mannitol). Where indicated, 10 μm of control or disrupting protein was added to the intracellular solution. TRPV1 current was monitored by applying every 3 s a voltage clamp protocol consisting of a first step of 100 ms at −100 mV followed by an ascending ramp (rate 0.4 mV/ms) to +100 mV and a final step at +100 mV for 100 ms. Between episodes, the membrane potential was clamped at 0 mV.

Data Analysis

For patch clamp experiments, data were analyzed with PatchMaster v2.50 and Origin 5.0 (MicroCal Software, Inc.). Statistical significance between different groups was determined by unpaired Student's t test or analysis of variance analysis using InStat software (version 3.06, GraphPad Software, Inc.). For biochemical and calcium imaging assays, data were analyzed among groups using analysis of variance analysis from an average of at least 3 experiments (means ± S.E. are shown). Comparison between different experimental groups was determined with the unpaired Student's t test using SigmaStat software. p < 0.05 (*), p < 0.01 (**), and p < 0.001 (***) is indicated for all figures.

RESULTS

TRPV1 Associates with AC5 in an AKAP79-dependent Manner

TRPV1 and AC5 have been independently shown to directly bind AKAP79/150 (4, 7, 17, 18, 20), but it was unclear whether a macromolecular complex of all three proteins could be formed. To measure complex formation, HEK293 cells were transfected with TRPV1 and AC5, in the absence or presence of AKAP79 (Fig. 1A). TRPV1 was immunoprecipitated from cell lysates followed by measurement of associated AC activity in each immune complex (referred to as an IP-AC assay) for detection of direct or indirect binding of AC. Expression of AKAP79 and TRPV1 is shown by Western blotting (Fig. 1A), whereas AC5 expression was confirmed by AC activity assays of cell lysates (data not shown). Antibodies against TRPV1 pulled down a 4-fold increase in AC activity as compared with IgG. This interaction was dependent on expression of AKAP79 (Fig. 1A). AC5 binds the second polybasic domain of AKAP79 (7). A fragment of AKAP79 consisting of residues 77–153 serves as a specific disruptor of AC5 or AC6 interactions with AKAP79 (7). The control fragment AKAP79^109–290 had no effect on AC5 binding. The addition of the AC5/6-AKAP79-disrupting polypeptide fragment (AKAP79^77–153) significantly reduced TRPV1-associated AC activity, such that the remaining activity was about equivalent to that associated with control IgG (Fig. 1A). Similarly, deletion of the AC5 binding domain (Δ77–108) but not the PKA binding domain of AKAP79 resulted in loss of all TRPV1-associated AC activity (Fig. 1B). This was not due to loss of TRPV1 interactions with AKAP79 as deletion of the AC binding domains had no effect on the direct binding of TRPV1 to AKAP79 (Fig. 1C). Thus, we conclude that TRPV1 forms a complex with AC5 that is dependent on their respective interactions with AKAP79.

FIGURE 1.

TRPV1 forms a complex with AC5 via interactions with AKAP79/150. A, HEK293 cells were transfected with AC5 and TRPV1 ± AKAP79. Samples were immunoprecipitated with anti-TRPV1 or control IgG and assayed for AC activity with 50 nm Gα_s and 100 μm forskolin. Where indicated, control (AKAP79^109–290) or AC-AKAP79-disrupting polypeptide (AKAP79^77–153; 5 μm) was included during homogenization. Data were normalized to control (TRPV1 IP without AKAP79) and are presented as the mean ± S.E., n = 4–5. AKAP79 and TRPV1 expression was confirmed by Western blotting. B, HEK293 cells were transfected with AC5 and TRPV1 ± AKAP79 or AKAP79 with a deletion of the AC (ΔAC) or PKA binding site (ΔPKA). IP-AC assays were performed with anti-TRPV1 antibody as in A (n = 3). C, AKAP79ΔAC does not interfere with TRPV1 binding as anti-TRPV1 pulls down both AKAP79 and AKAP79ΔAC when co-expressed in HEK293 cells (n = 3). D, IP-AC assays were performed with AKAP150 as described in A. Data were normalized to control (IP using IgG) and are presented as the mean ± S.E., n = 3. E, HEK293 cells were transfected with AC5 and TRPV1 ± AKAP150. Immunoprecipitations using anti-AKAP150 were subjected to Western blotting using the indicated antibodies (n = 3). Expression levels (inputs) for each protein are shown in the lower panel. *, p < 0.05, **, p < 0.01, and ***, p < 0.001 relative to the indicated controls.

The mouse ortholog of AKAP79, AKAP150, also supports a complex between TRPV1 and AC5, which can be disrupted with AKAP79^77–153 (Fig. 1D). As expected, endogenous PKA regulatory subunits (RIIβ) and PKA catalytic subunits are also associated with the AKAP150-TRPV1-AC5 complex (Fig. 1E). However, we did not detect association of PP2B (calcineurin) with this complex. This is consistent with previous studies that show that PP2B regulation of TRPV1 is independent of AKAP79/150 association (24)

Forskolin Stimulation of AC Abolishes TRPV1 Desensitization in HEK293 Cells

Although AC has been shown to associate with several ion channels (7, 8, 25), it is unknown whether anchoring AC to a channel complex has any direct effect on channel function. Therefore, we examined TRPV1 activity by calcium imaging in HEK293 cells transfected with TRPV1, AC5, and AKAP79. As shown previously (26–28), activation of TRPV1 by capsaicin generates a large and transient intracellular Ca²⁺ signal. This is followed by desensitization, such that the response to a subsequent application of capsaicin given 3 min later becomes greatly diminished (Fig. 2A). Desensitization is quantified as the response ratio of the amplitude of the second capsaicin application divided by the amplitude of the initial application (28–30). Exposure to cAMP-elevating agents such as forskolin (300 nm) abolishes the desensitization, allowing the subsequent stimulation to yield a similar response as the initial one (Fig. 2A). This regulation is due to PKA-dependent phosphorylation of TRPV1 that requires AKAP79 anchoring of the channel and PKA (17, 18, 20).

FIGURE 2.

Formation of a TRPV1-AKAP79-AC5 complex increases TRPV1 sensitivity to forskolin. HEK293 cells were transfected with TRPV1, AC5, and AKAP79. A, intracellular Ca²⁺ transients were evoked by two applications of capsaicin (300 nm for 10 s; indicated by arrowheads) separated by 3 min. Cells were continually perfused under control conditions (top) or in the presence of 300 nm forskolin (bottom). B, plot of the response ratio (amplitude peak 2/amplitude peak 1) as a function of forskolin concentrations. Inactive dideoxyforskolin (ddFsk) was used as a control. Data are presented as mean ± S.E. and were obtained from 11–26 cells for each concentration point. C, HEK293 cells were transfected with AC5 and TRPV1 ± AKAP79 or AKAP79 with deletion of the AC or PKA binding site. Intracellular Ca²⁺ transients were measured in the presence of 1,9-dideoxyforskolin or forskolin (30 nm), and the response ratio was plotted. D, cAMP accumulation was measured in HEK293 cells expressing TRPV1 and AKAP79 ± AC5 with increasing concentrations of forskolin. E, expression of AC5 and AKAP79 shifts the sensitivity of TRPV1 to forskolin. The response ratio for intracellular Ca²⁺ transients in HEK293 cells expressing TRPV1 alone or with AC5 ± AKAP79 was measured at the indicated forskolin concentrations (mean ± S.E., n = 3–4 experiments). *, p < 0.05, **, p < 0.01, and ***, p < 0.001 relative to the indicated controls.

Association of AC with the TRPV1 complex should sensitize the channel to the effects of cAMP and facilitate PKA phosphorylation. Therefore, we performed an initial forskolin dose-response curve with the TRPV1-AKAP79-AC5 complex. With as little as 30 nm forskolin, we observed a significant block of TRPV1 desensitization (Fig. 2B). A subsaturating forskolin concentration (300 nm) nearly completely abolished the TRPV1 desensitization (Fig. 2A), whereas exposure of cells to the inactive form of forskolin (1,9-dideoxyforskolin) did not produce any effect on capsaicin-dependent desensitization of the channel (Fig. 2, B and C). The sensitivity of the channel to low forskolin concentrations (30 nm) was dependent on the presence of both PKA and AC5 as deletion of either binding site negated the forskolin effect (Fig. 2C).

To ensure that the overexpression of AC5 did not produce a saturating level of cAMP at low forskolin concentrations, we measured cAMP accumulation in response to forskolin with and without AC5 expression (Fig. 2D). Stimulation of AC5 with low forskolin (<100 nm) is shown to produce cellular cAMP levels at or below maximal levels as compared with endogenous AC activity in HEK293 cells. Finally, to demonstrate that it is locally anchored, rather than the global expression of AC, which is important for TRPV1 regulation, we measured TRPV1 desensitization in response to forskolin with and without AC5 and AKAP79 expression (Fig. 2E). Expression of AC5 in the absence of AKAP79 shifted the forskolin dose-response curve necessary to abolish the TRPV1 desensitization effect by ∼10-fold (Fig. 2E; from EC₅₀ of 6 μm to 0.4 μm), similar to the increase in cAMP produced by AC5 expression (Fig. 2D). However, co-expression of AC5 and AKAP79 gave rise to a shift in the forskolin dose-response curve by 2 orders of magnitude, greatly increasing the sensitivity of the channel to the effects of cAMP. This measured effect of anchored AC5 likely represents an underestimation because HEK293 cells express low levels of AKAP79 and AC6 (4, 31).

Scaffolding of AC5 to TRPV1 by AKAP79 also affects the desensitization of TRPV1 current (I_TRPV1). A desensitization protocol similar to the calcium measurements was used, where two successive capsaicin applications (300 nm for 30 s) were separated by a 3-min interval during whole-cell recordings. Currents at +100 mV were used to assess relative TRPV1 activities. Inclusion of 30 nm forskolin had little effect on the desensitization of TRPV1 expressed alone (Fig. 3, A and E). In the absence of forskolin, coexpression of AC5 and AKAP79 did not affect I_TRPV1 desensitization (Fig. 3, B and E); however, with 30 nm forskolin, the desensitization was completely removed (Fig. 3, C and E). Anchoring of AC5 was once again required as forskolin action on I_TRPV1 desensitization was compromised in cells co-expressing AKAP79ΔAC instead of the wild-type AKAP79 (Fig. 3, D and E). Please note that overexpression of AC5 and AKAP79 did not significantly alter the maximal TRPV1 currents (Fig. 3F).

FIGURE 3.

Scaffolding of AC5 to TRPV1 by AKAP79 reduces I_TRPV1 desensitization. A–D, representative time courses of I_TRPV1 recorded at +100 mV evoked by two successive applications of capsaicin (300 nm, 30 s) separated by a 3-min interval in HEK293 cells transfected with TRPV1 alone or in combination with AC5 and AKAP79 or AKAP79ΔAC in the presence or absence of forskolin (Fsk) (30 nm). Data are normalized by the peak amplitude of the first capsaicin-evoked response. E, quantification of I_TRPV1 desensitization in HEK-293 cells transfected with the indicated plasmids in the presence or absence of forskolin (30 nm) (mean ± S.E., n = 5–6 in each group). *, p < 0.05 and **, p < 0.01 relative to the indicated controls (Ctrl). F, quantification of I_TRPV1 densities in HEK293 cells transfected with TRPV1 alone or in combination with AC5 and AKAP79 and treated in the presence or absence of 300 nm forskolin (mean ± S.E., n = 5–6 for each group). pF, picofarads.

Isoproterenol Stimulation of AC Abolishes TRPV1 Desensitization in HEK293 Cells

Although AC anchoring is clearly important for TRPV1 regulation by forskolin, a direct activator of AC, it was not clear whether it was also required for regulation by activation of Gs-coupled receptors. In HEK293 cells expressing TRPV1, AC5, and AKAP79, the capsaicin-induced desensitization of I_TRPV1 was completely blocked by isoproterenol, an agonist of endogenously expressed G_s-coupled β-adrenergic receptors (Fig. 4). This is similar to the effect produced by forskolin. Moreover, the effects of isoproterenol on I_TRPV1 required anchoring of AC5 to AKAP79 as expression of AKAP79ΔAC was unable to support the isoproterenol-stimulated block in desensitization of I_TRPV1 (Fig. 4).

FIGURE 4.

Stimulation of G_s-coupled β-adrenergic receptors abolishes TRPV1 desensitization in HEK293 cells. A, time courses of I_TRPV1 recorded at +100 mV evoked by two successive applications of capsaicin (300 nm, 30 s) separated by a 3-min interval in HEK-293 cells transfected with TRPV1, AC5, and AKAP79 or AKAP79ΔAC in the presence or absence of isoproterenol (ISO, 1 μm). Data are normalized by the peak amplitude of the first capsaicin-evoked response. B, quantification of I_TRPV1 desensitization. Cells expressed TRPV1 and AC5 plus AKAP79 (A79) or AKAP79ΔAC (A79ΔAC) as indicated (mean ± S.E., n = 3–5 in each group). *, p < 0.05, relative to the indicated controls.

AC-TRPV1-AKAP150 Complex Is Important for TRPV1 Desensitization in Mouse DRG Neurons

Formation of an AC-TRPV1 complex also occurs in vivo as IP-AC assay from dissected mouse DRGs showed a 4-fold higher level of AC activity associated with TRPV1 as compared with the IgG control (Fig. 5A). To measure the effect of anchored AC on TRPV1 regulation, capsaicin-evoked I_TRPV1 was recorded from isolated DRG neurons by whole-cell patch clamping. Currents at −70 mV were used to assess relative TRPV1 activities. The vast majority of TRPV1-expressing DRG neurons (∼91%) also express AKAP150 (18). In the absence of forskolin, DRG neurons displayed a clear desensitization of I_TRPV1 (Fig. 5, B and C), as observed in TRPV1-expressing HEK293 cells. The addition of 1 μm forskolin to the extracellular bath solution blocked the desensitization of I_TRPV1 (Fig. 5, B and C), whereas inclusion of an AC-AKAP-disrupting polypeptide (AKAP79^77–153, 10 μm) in the intracellular pipette solution completely blocked the effect of forskolin on I_TRPV1 desensitization. By contrast, inclusion of 10 μm control polypeptide in the pipette solution did not alter the effect of forskolin (Fig. 5, B and C). Therefore, in native DRG neurons, AC anchored to the AKAP150-TRPV1 complex is required for functional coupling of AC stimulation to TRPV1 channel regulation.

FIGURE 5.

Anchoring AC5 to AKAP150 is required for forskolin and PGE₂-mediated blockade of TRPV1 desensitization in mouse DRG neurons. A, AC activity is associated with TRPV1 in DRGs. Mouse DRG extracts were immunoprecipitated with IgG (control) or anti-TRPV1 antibody. Immunoprecipitates were stimulated with Gα_s and forskolin, and the associated AC activity was measured (n = 3). B, representative whole-cell currents at −70 mV evoked by two successive applications of capsaicin (1 μm, 30 s) separated by a 3-min interval ± forskolin (Fsk) (1 μm). DRG neurons were dialyzed with control intracellular solution or intracellular solution containing 10 μm control peptide (AKAP79^109–290) or disrupting peptide (AKAP79^77–153). pF, picofarads. C, quantification of panel B (mean ± S.E., n = 3–5 experiments). D, similar to C, but DRG neurons were treated with PGE₂ (1 μm). E, quantification of panel D (mean ± S.E.; n = 4–5 experiments). *, p < 0.05, **, p < 0.01, and ***, p < 0.001 relative to the indicated controls.

Finally, we examined the regulation of native TRPV1 activity in DRG neurons by stimulation of an endogenous upstream regulator, namely PGE₂. The PGE₂ receptor is coupled to G_s, which sensitizes TRPV1 thermal activation and is often associated with thermal hyperalgesia (13, 14, 32–34). The addition of 1 μm PGE₂ to DRG neurons abolished desensitization of I_TRPV1 (Fig. 5, D and E). The PKA-disrupting peptide Ht31 blocked the effects of PGE₂, as shown previously (15, 17, 20).⁴ Supplementation of the intracellular pipette solution with the AC-disrupting peptide AKAP79^77–153 eliminated the effects of PGE₂ (Fig. 5, D and E), underscoring the necessity of an intact AC-TRPV1-AKAP150-PKA complex in the facilitation of TRPV1 function by the endogenous inflammatory mediator.

DISCUSSION

The experiments reported herein show that anchoring of AC by AKAP79/150 is essential in sensitizing TRPV1 to cAMP/PKA-dependent modulation in mouse sensory neurons. Several lines of evidence support this conclusion. First, AC activity is found associated with TRPV1 in an AKAP79-dependent manner when coexpressed in HEK293 cells. This interaction is dependent on the AC5/6 binding site on AKAP79 and independent of PKA binding. Second, the formation of an AC-AKAP79-TRPV1 complex sensitizes the channel to cAMP/PKA signaling, attenuating the desensitization of TRPV1 in response to capsaicin activation. Specifically, anchoring of AC5 to the TRPV1 complex shifts the dose response to forskolin by ∼100-fold. This shift to lower concentrations of forskolin requires bound PKA and AC5. Similar results are obtained using both calcium imaging and electrophysiological methods to measure TRPV1 channel activity. Third, anchoring of AC to TRPV1 also sensitizes the channel to stimulation by the G protein-coupled β-adrenergic receptors. Fourth, the association of TRPV1 and AC is also found in DRGs. Finally, disruption of AC anchoring in isolated DRG neurons blocks the sensitization of TRPV1 by either forskolin or PGE₂.

TRPV1 is a homotetrameric, nonselective cation channel and serves as a polymodal sensor for a wide variety of stimuli, including heat, changes in pH, endogenous lipids such as anandamide, and exogenous molecules such as capsaicin, the spicy ingredient in hot chili peppers (35). Activation of PKA, PKC, and the calcium-dependent phosphatase PP2B (calcineurin) controls the phosphorylation of several TRPV1 residues and ultimately the functional state of TRPV1 (13). All three of these regulators have been shown to directly bind AKAP79/150 (36–38). In addition, activation of the tyrosine kinase Src by NGF sensitizes TRPV1 by phosphorylating the channel and increasing the trafficking of TRPV1 to the cell surface (18, 39). However, only PKA and PKC require AKAP79/150 association to mediate their effects, whereas PP2B and Src act independent of AKAP79/150 expression (17, 18, 20, 24). This is consistent with our finding that PP2B is not associated with the AC5-TRPV1-AKAP150 complex.

Prostaglandins exert their effect on TRPV1 via activation of the G_s-coupled receptors EP3C and EP4 in sensory neurons to increase cAMP/PKA signaling (12, 17, 40, 41). However, the requirement for anchored AC raises several important questions about the upstream components of the system. For example, AKAP79/150 scaffolds the β-adrenergic receptor and the relaxin receptor (RXFP1) either directly or indirectly (42–46). Does the PGE₂ receptor bind similarly to the TRPV1 complex? Is the G_s heterotrimer recruited indirectly by the receptor or through interactions with AC5 (47)? Future studies will be required to address the exact nature of the TRPV1 complex in DRGs.

The identity of the AC isoform(s) coupled to PGE₂ sensitization of TRPV1 in DRGs is also unclear. RT-PCR analysis and opioid inhibition of cAMP in DRGs suggest the presence of all AC isoforms except AC2 and AC7 (48). Knock-out studies of AC isoforms implicate both AC1 and AC5 in pain pathways. Deletion of AC1 has significantly reduced behavioral nociceptive responses in acute muscle pain and chronic muscle inflammatory pain (49). Deletion of AC5 strongly attenuates acute thermal pain, thermal allodynia, and chronic inflammatory pain (50) and attenuates the behavioral responses to morphine (51). AC8 does not have a role in formalin-mediated inflammatory pain (49); roles for additional AC isoforms have not been examined. However, AC isoforms 1, 4, and 7 do not bind to AKAP79/150. In addition, the AC-AKAP79-disrupting polypeptide is effective only with AC5/6, but not AC2 or -9. Therefore, it is highly likely that AC5 or possibly AC6 couples to AKAP79/150 to regulate TRPV1 in DRG neurons.

AKAP79/150 also facilitates the PKA phosphorylation of anchored AC5/6 (4). The activity of AC5/6 is inhibited by AKAP79 expression due to PKA phosphorylation of the anchored enzymes (4, 7). This raises the question of how anchored AC can facilitate TRPV1 phosphorylation if AC activity is simultaneously inhibited. However, expression of AKAP79 does not decrease the initial amplitude of isoproterenol-stimulated cAMP accumulation or PKA activation. Rather, AKAP79 assembles a negative feedback loop to facilitate rapid termination of cAMP synthesis (4).

Herein we have concentrated mainly on the functional consequences of AC association with TRPV1, but there is significant biochemical evidence for additional AC-effector complexes. For example, AKAP79/150 also mediates interactions between AC5/6 and the AMPA-glutamate receptors in brain (7). In heart, AKAP79/150 scaffolds AC5/6 and L-type calcium channels, whereas the AKAP Yotiao mediates interactions between AC9 and KCNQ1 K⁺ channels (8, 25). The functional consequences of AKAP-anchored AC for the activation of these channels remain to be explored; however, the scaffolding of AC9 to the anchoring protein Yotiao facilitates PKA phosphorylation of KCNQ1 (8). The sensitization of channel effectors to lower agonist concentrations may be a general consequence of recruiting AC to these complexes. Whether or not this is a regulated process remains to be determined.

The phosphorylation of TRPV1 elicited by inflammatory mediators such as bradykinin, ATP, and prostaglandins has many physiological consequences. The reduction in TRPV1 desensitization by PKA and/or PKC phosphorylation can facilitate channel function when continuously activated by protons during periods of inflammation and may be responsible for chronic pain associated with a number of disease states, including but not limited to, pancreatitis, irritable bowel syndrome, and cancer (11, 13, 14, 33). This in turn also leads to increased calcium influx through TRPV1 to enhance the release of substance P or calcitonin gene-related peptide from nociceptive terminals to further enhance the inflammatory state. In addition, TRPV1 phosphorylation reduces the temperature threshold for heat activation of TRPV1, thus contributing to thermal hyperalgesia arising from inflammation, peripheral nerve injury, diabetes, or herpes simplex (33). AKAP79/150-anchored PKA is indeed at least partially required for the PGE₂-induced decrease in temperature threshold in DRG neurons, consistent with the diminished PGE₂-induced thermal hyperalgesia observed in mice lacking the PKA binding site within AKAP150 (17). The disruption of AC anchoring to TRPV1 may therefore prove to be an attractive target for ameliorating the effects of chronic TRPV1 activation under inflammatory conditions.