A-Kinase Anchoring Protein 79/150 Coordinates Metabotropic Glutamate Receptor Sensitization of Peripheral Sensory Neurons

Abstract

Glutamate serves as the primary excitatory neurotransmitter in the nervous system. Previous studies have identified a role for glutamate and group I metabotropic receptors as targets for study in peripheral inflammatory pain. However, the coordination of signaling events that transpire from receptor activation to afferent neuronal sensitization has not been explored. Herein, we identify that scaffolding protein A-Kinase Anchoring Protein 79/150 (AKAP150) coordinates increased peripheral thermal sensitivity following group I metabotropic receptor (mGluR5) activation. In both acute and persistent models of thermal somatosensory behavior, we report that mGluR5 sensitization requires AKAP150 expression. Furthermore, electrophysiological approaches designed to record afferent neuronal activity reveal that mGluR5 sensitization also requires functional AKAP150 expression. In dissociated primary afferent neurons, mGluR5 activation increases TRPV1 responses in an AKAP dependent manner through a mechanism that induces AKAP association with TRPV1. Experimental results presented herein identify a mechanism of receptor-driven scaffolding association with ion channel targets. Importantly, this mechanism could prove significant in the search for therapeutic targets that repress episodes of acute pain from becoming chronic in nature.

Introduction

Chronic pain is a debilitating symptom of many disease states that is now being studied as its own separate and unique pathological event. Although central nervous system changes have been shown to contribute to this pathology [46; 53], increasingly more studies highlight the contributions of peripheral neurons [7; 21]. Despite this newfound focus, few have identified signaling mechanisms that peripherally support the perpetual nature of persistent somatosensitivity. This represents an opportunity to identify molecular coordinators of signaling events that may support chronic pain.

Glutamate serves as the primary excitatory neurotransmitter throughout the body, both in central and peripheral nervous systems. Growing evidence indicates that glutamate plays a significant role in afferent neuronal sensitization following injury and/or inflammation. Indeed, electrical stimulation of peripheral afferents, peripheral application of capsaicin, and circumstances of chemical and arthritic inflammation stimulate glutamate accumulation in the periphery [18; 37; 40]. This accumulated glutamate is capable of activating both ionotropic and metabotropic receptors throughout the body, although only certain isoforms are expressed by peripheral sensory neurons [17]. Of these isoforms, those belonging to group I that include mGluR1 and mGluR5 display significant physiological relevance in peripheral pain sensitization.

Previous studies have identified signaling pathways correlated with peripheral somatosensitivities following Group I mGluR activation. Indeed, the mGluR5 G-protein coupled receptor signals through a classic Gαq/11-coupled pathway to stimulate phospholipase C (PLC) [5]. mGluR5 activation stimulates protein kinases A (PKA) and C (PKC) to reduce the activation threshold for transient receptor potential type V1 (TRPV1) [16; 29] contributing to thermal sensitivity and peripheral pain behaviors. TRPV1, a non-selective cation permeable, ligand gated ion channel activated by heat (>43°C), protons (pH< 5.9), endogenous eicosanoids, and exogenous vanilloids, is dynamically modified via scaffolding functions coordinated by A-Kinase Anchoring Protein 79/150 (AKAP) [30; 31; 56]. Further, PLC hydrolysis of AKAP150 anchorage to the plasma membrane stimulates AKAP150 association with TRPV1 [32]. Without this PLC-induced association, TRPV1 modification by PKA/PKC would not be possible, thereby preventing receptor sensitization from occurring. Therefore, we hypothesize that receptors coupled to Gαq-signaling pathways should stimulate AKAP150 association with TRPV1 to coordinate peripheral glutamate-induced somatosensitization through mGluR5.

Herein, we investigated the role of AKAP150 in glutamate-induced peripheral sensitization. Experimental findings indicate that AKAP serves an integral role as signal coordinator between mGluR activation and somatosensitization through TRPV1. Importantly, this biochemical coordination of peripheral signaling events may support an inflammatory positive feedback system to maintain sensory afferent sensitivity to sub-threshold stimulation.

Materials and Methods

Reagents

(S-3,5-Dihydroxyphenylglycine (DHPG), 3-((2-Methyl-1,3-thiazol-4-yl)ethynyl)pyridine hydrochloride (MTEP), and fenobam were purchased from Tocris/R&D Systems (Minneapolis, MN). Capsaicin (CAP) and all other chemicals were purchased from Sigma Aldrich, unless otherwise noted.

Animals

Procedures using animals were approved by either the Institutional Animal Care and Use Committee of The University of Texas Health Science Center at San Antonio (UTHSCSA) or The University of Texas Medical Branch at Galveston (UTMB) and were conducted in accordance with policies for the ethical treatment of animals established by the National Institutes of Health and International Association for the Study of Pain. Male Sprague-Dawley rats 175 – 200g in weight (Charles River Laboratories, Wilmington, MA) and AKAP WT/KO mice with a C57/Bl6 background [49] were used for behavior analyses, and for dorsal root ganglia (DRG) and trigeminal ganglia (TG) dissection. DRG (L4-6) and TG were removed bilaterally from male rodents, and dissociated by collagenase treatment (30 min, Worthington, Lakewood, NJ), followed by trypsin treatment (15 min, Sigma, St.Louis, MO), Cells were centrifuged and re-suspended between each treatment with Pasteur pipettes. Cells were centrifuged, aspirated, and re-suspended in DMEM (Gibco, Grand Island, NY) with 10% FBS (Gibco), 250 ng/ml NGF (Harlan, Indianapolis, IN), 1% 5-fluoro deoxyuridine (Sigma), 1% penicillin/streptomycin (Gibco), and 1% L-glutamine (Sigma), and then placed onto plates coated with poly-D lysine, or coverslips coated with poly-D lysine and laminin. Cultures were maintained at 37°C, 5% CO₂, and grown in 10 cm plates for 5 – 7 days for phosphorylation experiments. Rat TG cultures used for co-immunoprecipitation were maintained at 37°C for 5-7 days prior to the experiment. Rat tissue was preferred over mouse TG tissue since rat TGs are larger and contains more neurons, thereby requiring significantly fewer animals, following guidelines released by the NIH Office of Laboratory Animal Welfare. Mouse DRG cultures for calcium imaging were grown 18-36 h. AKAP150 wild-type (WT) and knock-out (KO) mice were originally created and characterized in the laboratory of John D. Scott [49], and maintained at UTHSCSA. Genotyping was performed as previously described [32], with males aged 4-9 weeks old used for experiments. All WT and KO animals used for experiments were littermates.

Calcium Imaging

DRG neurons were isolated from male AKAP150 WT and KO mice. Prior to measurements, cells were incubated for 30 min at 37°, 5% CO₂, with Fura2-AM (2μM, Invitrogen Life Technologies, Grand Island, NY) in the presence of 0.05% Pluronic (Invitrogen Life Technologies). Experiments were performed with an inverted Nikon Eclipse TE-2000-U microscope, equipped with 20X/0.8 NA Fluor objective. Cultured neurons were excited at the 340, 380 nm wavelength from a Lambda LS Fluorescent light source (Sutter Instruments), while a Hamatasu digital CCD Camera detected emissions at 510 nm. Collected data were analyzed with MetaFluor Software (Molecular Devices). Ratiometric 340/380 nm values were converted into nM of Ca²⁺ according to the formula [Ca²⁺] = 199 × (R-0.46)/(1.45-R); where R = 340/380 ratio and all the other values are specific to the experimental setup and were determined with use of a Fura2-AM calibration kit (Invitrogen Life Technologies).

Experiments were performed at room temperature (RT) with use of a gravity feed perfusion system, which allowed for continuous exchange of bath solution. Recordings were performed in standard extracellular solution (SES) that contained the following (in mM): 140 NaCl, 5 KCl, 2 MgCl₂, 2CaCl_2, 10 HEPES, 10 Glucose, pH 7.4 [NaOH]. Coverslips of cultured DRG neurons were treated with DHPG (100μM, [29]) via bath application, and capsaicin (CAP, 100nM) was applied for 30 sec via pipet application. SES also served as vehicle for capsaicin (100nM, Sigma Aldrich) and DHPG (100μM, Tocris). The micropipette, which delivered drug-containing solutions, was precisely placed in the optic field that includes our cells of interests. Both bath and pipet solution were under automatically controlled pressurized system (PM2000 Cell Microinjector, MicroData Instruments Inc.) that, in pairing with pneumatic pump, supported precise, fast and consistent fluid flow and exchange. Data were analyzed by one-way ANOVA, with Bonferroni post hoc correction as needed.

Electrophysiological Recordings: In Vitro Skin-Nerve Preparation

The preparation used to record from nociceptors has been previously described in detail [12; 27]. Briefly, male C57/Bl6 WT and KO mice (10-13 weeks of age) were sacrificed with an overdose of CO₂. The glabrous skin was dissected from each hindpaw starting at the ankle and moving to the tips of the toes. The medial and lateral plantar nerves were also dissected free and kept intact with the glabrous skin. The preparation was placed corium side up in a 2-chambered tissue bath. The nerves were drawn into one chamber and the skin laid out and pinned to the gel floor of the other chamber.

In the recording chamber the nerves were desheathed and carefully teased apart so that small nerve bundles could be obtained. These were laid across the gold wire recording electrode. A glass rod was pressed into the glabrous skin to find receptive fields of units in the nerve bundles. Once the receptive field of a unit was located, the conduction velocity (CV) was measured and a sequence of testing was done as described below.

Thermal and Chemical Testing Procedures

For thermal stimulation, radiant heat was applied to each receptive field by a feedback-controlled lamp (made in house) beneath the organ bath. The beam was focused through the bottom of the bath onto the epidermal surface of the skin. A thermocouple in the corium above the light beam measured intracutaneous temperature. A standard heat ramp starting from an adapting temperature of 34°C and rising to 51°C in 10 s was applied to each unit from the epidermal side.

To investigate unit responses to chemicals, a small plastic ring (4.5 mm diameter) was placed over the receptive field of each unit. The synthetic interstitial fluid (SIF) in the ring was replaced with a drug dissolved in SIF (buffered to pH 7.40 + 0.05). As with any in vivo intraplantar injections, drugs must diffuse through the plantar connective tissue, dermis and epidermis to reach the nociceptor nerve terminals, thus drug concentrations were considerably diluted by the time they reach their target sites. Drug concentrations for DHPG (1 and 3mM) and CAP (1μM) were similar to those previously used for activating primary sensory neurons [3; 5; 57].

To determine the percent of units responding to a drug application, only those units whose discharge rate increased more than the mean + 2 SD's (calculated using the background discharge rate for the population) were considered “responders”.

Behavior

Paw withdrawal latency to a thermal stimulus was measured with a plantar test apparatus (IITC, Woodland Hills, CA) as previously described using the Hargreave's method [26]. Briefly, animals were placed in plastic boxes on a warm glass surface. After a 30 min habituation period, the plantar surface of the hindpaw was exposed to a beam of light that gradually heated the glass floor. The latency to withdraw the paw from the glass surface was measured and the intensity of the light and basal temperature of the glass floor were adjusted so that baseline withdrawal latencies for rats and mice were close to 10 sec. Cutoff was set at 25 sec to avoid tissue damage from repeated testing. Measurements were taken in duplicate at least 30 sec apart and the average was used for statistical analysis. Data are presented as mean ± SEM paw withdrawal latencies or as mean ± SEM of the individual changes from individual pre-injection baseline values. Observers were blinded to the treatment conditions and genotypes of the animals.

All injections were given intraplantarly in 50 μl (rat) or 10μl (mouse) volumes via a 28-gauge needle inserted through the lateral footpad just under the skin to minimize tissue damage. Drug doses were taken from previous work by Aley et al [1] and Bhave et al (for MTEP [5]). Drug stocks were dissolved in Phosphate Buffered Saline (PBS). For rat priming experiments, rats were tested for baseline responses prior to injections, and then injected with carrageenan (50 μl of a 0.1% solution in PBS) or vehicle (PBS) on day 1. On days 2, 3, and 4, rats were injected with MTEP (400 μg) or vehicle (PBS). Rats were tested again on day 5 for baseline responses, and then injected with PGE₂ (100 ng) followed by repeated thermal testing at 15 min, 30 min, 2 h, and 4 h post-PGE₂ injection. For mouse priming experiments, WT and AKAP150 KO mice were tested for baseline responses prior to injections, and then injected with carrageenan (10 μl of a 0.5% solution in PBS) on day 1. Thermal withdrawal latency was measured again on day 5 (Post-Carageenan) prior to injection with PGE₂ (100 ng) and repeated thermal testing at 15 min, 30 min, 1 h, 2 h, and 4 h post-PGE₂ injection.

Co-immunoprecipitation and Western Blotting

TG cultures in 10-cm plates were treated and prepared for homogenization and isolation of crude plasma membrane fractions, as described previously [30]. Crude plasma membrane homogenates are quantified for protein concentration by Bradford analysis [9]. Equal samples underwent co-immunoprecipitation (100μg total protein) or gross homogenate WB analysis (10 μg total protein). Samples isolated for co-immunoprecipitation were diluted to 500 μl with homogenization buffer [30] and incubated with antibodies specific to AKAP150 (1 μg, Santa Cruz Biotechnology, Santa Cruz, CA [30]), pulled down with protein agarose-A (SigmaAldrich, St. Louis, MO), and resolved by Sodium Dodecyl Sulfate -Polyacrylamide Gel Electrophoresis (SDS-PAGE). Gels were transferred to polyvinyldifluoride (PVDF, EMD Millipore, Billerica, MA), blocked in 5% nonfat milk in Tris buffered saline with 0.1% Tween-20 non-ionic detergent, and incubated 18 h with primary antibody anti-TRPV1 (SantaCruz Biotechnology [8; 55]) or anti-AKAP150. Anti-rabbit secondary antibodies (GE Healthcare Life Sciences, Piscataway, NJ) were applied to rinsed blots, incubated at RT for 1 h, and then blots were rinsed again. Blots were incubated with enhanced chemiluminescence solution (GE Healthcare Life Sciences), exposed to X-ray film and developed for analysis. Films were scanned and immunoreactive bands were quantified using NIH Image 1.62 shareware.

Results

mGluR1/5 and AKAP150 support acute pain

In Figure 1, AKAP150 WT and KO littermate male mice were intraplantarly (i.p.) injected in the hindpaw with DHPG prior to stimulating the same hindpaw with radiant heat using the Hargreaves’ method [26] to monitor timed latency to remove the injected hindpaw from a thermal stimulus. Mice were injected with vehicle or 50nmol DHPG (10μl injection volume), and tested for paw withdrawal latency to a thermal stimulus 30 min post-injection, as performed previously [5]. In Figure 1, AKAP150 WT mice experienced a significant increase in sensitivity to a thermal stimulus following DHPG treatment, as paw withdrawal latency was dramatically reduced. Importantly, no change was observed in AKAP150 KO mice following DHPG injection, indicating that AKAP150 plays a physiological role in the sensitized thermal response following mGluR5 agonist administration.

Figure 1. DHPG sensitization of thermal sensitivity in AKAP WT and KO mice.

AKAP WT and KO mice were monitored for baseline paw withdrawal latency responses to a thermal stimulus prior to experimentation. AKAP WT and KO mice were injected intraplantarly (ipl) with DHPG (50nmol in 10μl PBS) or vehicle (Veh) as indicated, and tested 10min later for paw withdrawal latency to a thermal stimulus. n=6 mice/treatment paradigm, **p<0.01, *p<0.05, NS = not significant, two-way ANOVA with Bonferroni correction.

Persistent thermal sensitivity requires mGluR1/5 and AKAP150

Next, we evaluated whether acute thermal behavior observed in AKAP150 KO animals would translate to persistent pain models. Studies published by the Levine group identify a behavioral protocol by which acute pain events become persistent in nature [7; 41; 42]. In this protocol, the peripheral injection of an inflammatory algogen, such as carrageenan, several days prior to another inflammatory event led to a significant temporal extension of mechanical sensitivity and nociceptive behavior. However, rodents that underwent this treatment protocol were not studied for thermal sensitivity. Given the importance of TRPV1 to chronic pain [28; 34; 39; 44; 51], and the inclusive importance of IB4(+) neurons to hyperalgesic priming [33], we sought to determine whether the hyperalgesic priming protocol could affect behavioral measures of TRPV1 responsiveness, including thermal sensitivity. We employed the Hargreaves’ method for measuring sensitivity to a thermal stimulus, following the carrageenan pre-injection, PGE2-timed injection protocol [42]. As demonstrated in Figure 2A, we observed a significant difference in paw withdrawal latency between animals receiving vehicle vs. carrageenan pre-injection in the hind paw 2 h following PGE2 injection. Further, antagonism of mGluR5 with MTEP reversed priming effects following PGE2 injection, indicating an important role for mGluR5 in persistent pain.

Figure 2. Carageenan priming of PGE2-induced thermal allodynia requires mGluR5 and AKAP.

A. Rats were injected intraplantarly (ipl) with vehicle (Veh) or carrageenan (Cg, 5μl of 1% solution in 50μl PBS final) on day 1 (see Materials and Methods for full protocol). Rats were then injected with Veh or MTEP (400μg in 50 μl PBS) on days 2, 3, and 4. Rats were tested for baseline paw withdrawal latency day 5, then injected with PGE2 (indicated by arrow, 100ng in 50μl PBS, ipl) and tested again for paw withdrawal latency to a thermal stimulus at 15 min, 30 min, 2 h, and 4 h. Paw withdrawal latency displayed as change from baseline, n=6 rats/treatment paradigm, **p<0.01, *p<0.05, two-way ANOVA with Bonferroni correction. B. Mice were tested for baseline paw withdrawal latency to a thermal stimulus before intraplantar (ipl) injections (Pre-Cg). Mice were then injected with Cg (5μl of 1% solution in 10μl PBS final, ipl) on day 1. Mice were tested for baseline PWL on the Day 5 (Post-Cg), then injected with PGE2 (indicated by arrow, 50ng in 10μl, ipl) and tested for PWL at 15 min, 30 min, 2 h, and 4 h. Paw withdrawal latency displayed as change from baseline illustrated, n=6 mice/group, ***p<0.005, two-way ANOVA with Bonferroni correction.

Next, AKAP150 WT and KO littermates underwent a similar protocol, receiving carrageenan-priming injections prior to PGE2 injections. In Figure 2B, AKAP150 WT mice displayed persistent hypersensitive behavior in response to a thermal stimulus as rats previously did (Fig 2A). However, AKAP150 KO mice demonstrated little persistent response to carrageenan priming, returning to baseline readings 2 h after an initial PGE2 response. Taken together, these data indicate that functional AKAP150 expression is necessary to maintain persistent thermal hypersensitivity.

mGluR1/5 sensitizes peripheral TRPV1 responses in AKAP-dependent manner

Next, we employed an electrophysiological approach to monitor agonist-activated mGluR1/5 sensitization of TRPV1 in peripheral, primary afferent neurons responsible for innervating the tissues tested in previous behavior experiments. We monitored action potential generation of peripheral nerves using a skin-nerve preparation approach following previously published protocols [12; 27]. In Figures 3 and 4, glabrous skin preparations were generated from AKAP150 WT and KO littermate mice for in vitro measurements of peripheral neuron depolarization by TRPV1 agonists CAP (1μM) and heat (51°C). A total of 170 nociceptors were recorded from WT mice and 119 from KO. For WT, units consisted of 129 C fibers with a median conduction velocity (CV) of 0.68 (range = 0.38 - 1.18 m/s) and 41 Aδ fibers with a median CV of 2.64 (range = 1.21 – 8.43 m/s). For KO, units consisted of 82 C fibers with a median CV of 0.62 (range = 0.45 – 1.10 m/s) and 37 Aδ fibers with a median CV of 2.15 (range = 1.23 – 9.43 m/s). All units had their receptive fields on the glabrous skin of the hindpaw. The mean background discharge rate for units from each strain was not different (WT = 0.10 ± 0.04; KO = 0.13 ± 0.03 imp/sec). However, the deletion of AKAP150 resulted in a change in the response of units to a noxious heat stimulus. The thermal threshold of activation was significantly lower in nociceptors from KO compared to WT (40.61 ± 0.59 vs 43.01 ± 0.43°C, Fig 3B). Furthermore, the discharge rate was elevated in KO (3.19 ± 0.64 imp/s) compared to WT (1.66 ± 0.18 imp/s); however, this difference did not reach significance (p = 0.33, Fig 3A).

Figure 3. mGluR Sensitization of Heat Response in Nociceptors from AKAP WT and KO Mice.

A. Baseline activity in nociceptors was measured for 2 min prior to any manipulation. Mean discharge rate is higher in KO compared to WT, however, the increase is not significant. B. A 10 sec heat ramp, ascending from 34 to 51°C, demonstrates that the mean threshold to heat activation is significantly lower in KO compared to WT mice. *p < 0.05, Mann Whitney U test. C. Traces of unfiltered raw data show unit responses a few seconds before and after a 10 sec heat stimulus is applied to the receptive field of a nociceptor. In the KO mice there is no change in heat-induced activity following a 2 min treatment with 1 or 3 mM DHPG. D. In contrast, a unit from WT shows an enhanced discharge rate to heat following 3mM DHPG. The population data are summarized in E, F showing that a 2 min application of 1 or 3 mM DHPG to nociceptors in KO has no effect on heat-induced discharge rate (E) or threshold (F) but both are modified in WT. *p < 0.05, Kruskal-Wallis followed by Dunn's post hoc test.

Figure 4. mGluR Sensitization of Capsaicin Response in Nociceptors from AKAP WT and KO Mice.

A, B Traces of unfiltered raw data demonstrate that repeated 2-min exposures of receptive fields to 10 μM DHPG has no effect on discharge rate in WT and KO. C. The response to CAP alone is similar in KO vs WT mice. However, 10μM DHPG greatly enhances 1 μM CAP-induced activity in WT (A,C) compared to KO (B,C) *p < 0.05, Kruskal-Wallis followed by Dunn's posthoc test. The percent of CAP responders is similar in KO compared to WT, however, deletion of AKAP significantly reduces the percent of nociceptors enhanced by DHPG-CAP (D) *p < 0.05, Fischer exact test.

DHPG applied at 1 or 3 mM prior to heat stimulation produced no change from pre drug levels in discharge rate or threshold to activation in units from KO mice (Fig. 3C-F). The 1 mM dose also had no effect on heat responses in WT. In contrast, 3 mM DHPG sensitized WT units, significantly elevating the discharge rate (150 ± 26%) and lowering the mean threshold for activation compared to pre drug levels (94 ± 1%, p < 0.05).

Changing from a natural stimulus (heat) to a chemical stimulus (CAP) to activate TRPV1, it was clear that DHPG enhanced the TRPV1 activation in WT mice compared to KO mice (Fig. 4A and B). Neither DHPG (Fig 4A, C) nor CAP (Fig 4C) had any effect on discharge rate in KO or WT when applied alone, but when combined with CAP, DHPG sensitized WT units produced significantly enhanced discharge rates when compared to KO (0.90 ± 0.1 vs 0.69 ± 0.13 imp/s, Fig 4C). To ensure that deletion of AKAP150 did not change the number of CAP-sensitive units in the KO mice, the percentage of responders was calculated and there was no significant difference in the number of units in KO and WT responding to CAP (57 vs 62%, Fig 4D, Chi Square); however, the percent of units responding to DHGP + CAP was significantly higher in WT compared to KO (69 vs 50%, Fig 4D, Chi Square). Taken together, data presented in Figures 3 and 4 demonstrate that mGluR activation increases CAP- and heat-induced neuronal excitability in an AKAP150-dependent manner.

mGluR5 sensitization of TRPV1 is AKAP150- and PLC-dependent

Next, we sought to identify AKAP150 as a functional mediator of mGluR1/5 modulation of TRPV1 sensitivity to agonist activation. DRG neurons from AKAP150 WT and KO mice were prepared for real-time, single-cell calcium imaging to analyze mGluR5 effects on CAP-responses of TRPV1. For this work, we followed a previously published protocol detailing the effects of mGluR5 agonists on pharmacological desensitization of TRPV1 in sensory neurons [29]. We report similar findings in Figure 5, as pre-treatment with DHPG (2 min, 100μM) prior to the second application of CAP significantly reduces pharmacological desensitization of TRPV1. Importantly, this relationship was lost in AKAP150 KO neurons, indicating an important role for the scaffolding protein in downstream signal transduction pathway(s) following mGluR5 activation.

Figure 5. AKAP plays a role in DHPG enhanced Ca²⁺ responses following sequential capsaicin treatment.

Dorsal root ganglia neurons were isolated from WT and AKAP KO mice. Changes in intracellular Ca²⁺ levels were assessed by measurements with Fura2 calcium indicator. A. Representative tracings of single cell measurements. Starting with establishment of baseline, cells were treated with CAP (100 nM) for 30 s, washed with constant perfusion of SES buffer for 5 min, followed by administration of the second CAP treatment. In DHPG groups, cells were washed with SES buffer for 3 min and DHPG (100 μM) was administered for 2 min prior to second CAP treatment. B. The net change of intracellular Ca²⁺ concentration was calculated by subtraction of established basal concentration from peak concentration following CAP administration, n=28-76 neurons per group. In case of the second peak, concentration at the end of wash was taken as base value. C. The percentage of desensitization was determined by comparison of responses after first and second CAP application. The first Ca²⁺ increase was normalized to 100 %, and the second response was calculated as its percentage and the difference between the two illustrated the rate of desensitization for each group. ***p<0.001, NS = no significance, as determined by one-way ANOVA, with Bonferroni correction.

Since mGluR1/5 activation drives Gαq-coupled PLC activity [47], we next sought to determine whether PLC inhibition would reverse the DHPG-induced reduction in TRPV1 pharmacological desensitization. In Figure 6, DRG neurons were prepared for real-time calcium imaging as in Figure 5, but co-treated between successive CAP-applications with DHPG (2 min, 100μM) and either U73122 (PLC inhibitor, 1μM) or U73343 (negative control, inactive analog for U73122, 1μM) [32]. Following PLC inhibition, DHPG treatment is no longer able to reduce TRPV1 pharmacological desensitization, indicating that the Gαq-coupled signaling pathway is primarily responsible for the mGluR5 affects on TRPV1 responses in sensory neurons.

Figure 6. DHPG reversal of TRPV1 pharmacological desensitization is reversed by inhibition of PLC.

Dorsal root ganglia neurons were isolated from WT mice. Changes in intracellular Ca²⁺ levels were assessed by measurements with Fura2 calcium indicator. A. Representative tracings of single cell measurements. Starting with establishment of baseline, cells were treated with CAP (100 nM) for 30 s, washed with constant perfusion of SES buffer for 5 min, followed by administration of the second CAP treatment. In DHPG treatment groups, cells were washed with SES buffer for 3 min and vehicle, DHPG (100 μM) and vehicle, DHPG and U73122 (1μM), or DHPG and U73343 (1μM) was administered for 2 min prior to second CAP treatment. B. The net change of intracellular Ca²⁺ concentration was calculated by subtraction of established basal concentration from peak concentration following CAP administration, n=77-107 neurons per group. In case of the second peak, concentration at the end of wash was taken as base value. C. The percentage of desensitization was determined by comparison of responses after first and second CAP application. The first Ca²⁺ increase was normalized to 100 %, and the second response was calculated as its percentage and the difference between the two illustrated the rate of desensitization for each group. ***p<0.001, NS = no significance, as determined by two-way ANOVA, with Bonferroni correction.

mGluR1/5 Activation increases AKAP150/TRPV1 Association

We previously demonstrated that direct PLC activation increases TRPV1 association with AKAP150 [32]. This is important because AKAP150 scaffolds PKA and PKC to coordinate post-translational modification and subsequent sensitization of TRPV1. However, no one has demonstrated this phenomenon through receptor-mediated activation of PLC, as occurs via Gαq-coupled metabotropic receptors such as mGluR5. In Figure 7, cultured sensory neurons were treated with either vehicle and DHPG (100μM, [29]) or fenobam (1μM, [38]) and DHPG for 5 min each, and TRPV1 association with AKAP150 was determined by co-immunoprecipitation (Co-IP). As shown in Figure 7A-B, DHPG treatment induces a two-fold increase in AKAP association with TRPV1 in a manner sensitive to mGluR5 antagonism. In order to confirm the PLC-dependent nature of this receptor-activated association, we pre-treated cultures with the PLC inhibitor U73122 (1μM, 5 min) or its negative control U73343 (1μM, 5 min) prior to administering DHPG (5 min) as in panel A. Co-IP results in Figure 7C-D confirm that PLC inhibition summarily prevents DHPG-stimulated AKAP association with TRPV1. Furthermore, U73343 has no effect, thus identifying AKAP as an important coordinating scaffolding protein that manages mGluR5 sensitization of TRPV1.

Figure 7. Figure 1. mGluR1/5 activation stimulates AKAP association with TRP Receptors.

A. Cultured rat TG neurons were treated with DHPG (100μM, 5 min) and AKAP:TRPV1 association was determined by co-immunoprecipitation (Co-IP). Increased Co-IP is inhibited by pre-treatment with the mGluR5 negative allosteric modulator, fenobam (1μM, 5 min prior to DHPG). B. Quantified Co-IP results from A, normalized to total AKAP IP. Results representative of 3 independent trials, *p<0.05, **p<0.01, two-way ANOVA with Bonferroni correction. C. Cultured rat TG neurons were pre-treated with PLC inhibitor U73122 (1μM, 5 min) or negative control U73433 (1μM, 5 min) prior to DHPG (100μM, 5 min). AKAP:TRPV1 association was determined by Co-IP. D. Quantified Co-IP results from C, normalized to total AKAP IP. Results representative of 3 independent trials, *p<0.05, **p<0.01, two-way ANOVA with Bonferroni correction. For A and C, molecular weights of immunoreactive proteins in kilo Daltons (kDa) are indicated to the left of representative Western blots.

Discussion

Glutamate serves as the most abundant excitatory neurotransmitter in the human body. In addition to this, it also plays an important role in peripheral inflammation [18; 37; 40], and participates in afferent sensitization to somatosensation [5; 50]. Previous work has identified that glutamate increases peripheral thermal sensitivity through both PKC [36] and PKA activities [29], yet it is unclear how these signaling processes are coordinated. This is especially important given that both PKA and PKC increase peripheral thermal sensitivity via post-translational modifications of the ligand-gated ion channel TRPV1 [4; 6]. Here, we identify that AKAP150 (AKAP), a scaffolding protein responsible for organizing PKA- and PKC-mediated phosphorylation of TRPV1 [30; 31; 56], coordinates mGluR5 sensitization of TRPV1 in peripheral sensory neurons. AKAP scaffolding supports mGluR5-sensitive acute and persistent pain models. Also, AKAP is required for peripheral mGluR5-sensitization of sensory neurons by either thermal or chemical stimuli. In addition, mGluR5 activation reduces TRPV1 pharmacological desensitization in a manner that requires AKAP expression, but is sensitive to PLC inhibition. Furthermore, mGLuR5 stimulation of AKAP association with TRPV1 in sensory neurons is sensitive to PLC inhibition. Taken together, these results indicate that mGluR5-stimulated AKAP association with TRPV1 coordinates inflammatory thermal sensitization on an acute time-scale that can become more persistent in nature.

mGluR5 belongs to the group I family of metabotropic glutamate receptors, of which both subtypes (mGluR1 and mGluR5) are expressed by peripheral sensory neurons [5]. Both receptor subtypes couple to Gαq-mediated PLC signaling pathways in multiple models, including central nervous system neurons [25] and transfected cells [43]. Additional work in peripheral sensory neurons indicates that PLC mediates mGluR5-induced sensitization of TRPV1 responses [29]. Herein, we similarly demonstrate a significant importance for PLC in mGluR5 modulation of TRPV1 responses in cultured sensory neurons (Figure 6). Co-immunoprecipitation experiments reveal that mGluR5 activation stimulates AKAP association with TRPV1 in a PLC-dependent manner (Figure 7). Although this result confirms previous biochemical findings [19; 32], it also serves to identify that activation of peripherally-expressed, Gαq-coupled receptors can stimulate scaffolded coordination of phosphorylation of certain TRP channels, including TRPV1. Taken together, these data identify AKAP as a target for drug-development that may reduce TRPV1 sensitization linked to inflammatory thermal hyperalgesia [13].

Thermal hyperalgesia can be measured using multiple behavioral models. In this study, we employed the acute Hargreave's thermal sensitivity behavioral model (Figure 1, [26]), and the Levine hyperalgesic priming behavioral model (Figure 2, [1]). Following the acute behavioral paradigm, we found AKAP expression to be of significant importance to mGluR5-sensitization of thermal behavior. However, we modified the Levine model for mechanosensitivity to measure for persistent thermal sensitivity, and found that carageenan-induced hyperalgesic priming was sensitive to mGluR5 inhibition by MTEP. Importantly, use of AKAP KO animals in the same model revealed that AKAP expression is required for persistent thermal sensitivity. Together, these data identify an important role for peripheral glutamate in persistent pain that may involve AKAP.

A prominent feature of mechanical hyperalgesic priming in the Levine behavioral model is the extended length of time that the behavior persists (>24 h, [1]). Using the same model, we were only able to observe significant priming effects for thermal sensitivity at 2 h in rats (Fig 2A) and at 0.5 h and 2 h in WT mice (Fig 2B). This difference could be explained by a number of factors. First, several reports have identified that PGE2 effects on thermal sensitivity are short lived, losing significance less than 2 h post-exposure [2; 45]. Second, peripheral glutamate release following afferent depolarization [18] might affect receptors/channels sensitive to thermal activation differently than those activated by mechanical stimulation. Third, central sensitization could account for extended mechanical somatosensitivities [52] not observed here. However, an autocrine feed-forward loop [22] could also explain the findings presented here, and may significantly contribute to persistent peripheral somatosensitivities.

Autocrine feed-forward plasticity often arises in neuronal signaling paradigms when depolarized neurons release neurotransmitters capable of acting on receptors expressed by the depolarized neuron. In this vein, glutamate released from depolarized peripheral afferents could act on metabotropic glutamate receptors, including receptors belonging to the group I family, thereby resulting in TRPV1 phosphorylation and sensitization. This sensitized ligand-gated ion channel then has a reduced threshold for activation [4], increasing the likelihood for neuronal depolarization, and subsequent repeated release of glutamate. Indeed, this feed-forward scenario has been shown to be sensitive to peptidergic inhibition of AKAP/TRPV1 association [10; 23], identifying a unique protein-protein interaction that could support persistent thermal hyperalgesia. In conclusion, behavioral, functional, and biochemical data presented here demonstrate that mGluR5-stimulates AKAP150 association with TRPV1 in a model of thermal hyperalgesic priming, suggesting an innovative direction for future therapeutic interventions.