Nicotinic acetylcholine receptors in dorsal root ganglion neurons include the α6β4* subtype

Abstract

The α6-containing nicotinic acetylcholine receptors (nAChRs) have recently been implicated in diseases of the central nervous system (CNS), including Parkinson's disease and substance abuse. In contrast, little is known about the role of α6* nAChRs in the peripheral nervous system (where the asterisk denotes the possible presence of additional subunits). Dorsal root ganglia (DRG) neurons are known to express nAChRs with a pharmacology consistent with an α7, α3β4*, and α4β2* composition. Here we present evidence that DRG neurons also express α6* nAChRs. We used RT-PCR to show the presence of α6 subunit transcripts and patch-clamp electrophysiology together with subtype-selective α-conotoxins to pharmacologically characterize the nAChRs in rat DRG neurons. α-Conotoxin BuIA (500 nM) blocked acetylcholine-gated currents (I_ACh) by 90.3 ± 3.0%; the recovery from blockade was very slow, indicating a predominance of α_xβ4* nAChRs. Perfusion with either 300 nM BuIA[T5A;P6O] or 200 nM MII[E11A], α-conotoxins that target the α6β4* subtype, blocked I_ACh by 49.3 ± 5 and 46.7 ± 8%, respectively. In these neurons, I_ACh was relatively insensitive to 200 nM ArIB[V11L;V16D] (9.4±2.0% blockade) or 500 nM PnIA (23.0±4% blockade), α-conotoxins that target α7 and α3β2*/α6β2* nAChRs, respectively. We conclude that α6β4* nAChRs are among the subtypes expressed by DRG, and to our knowledge, this is the first demonstration of α6β4* in neurons outside the CNS.—Hone, A. J., Meyer, E. L., McIntyre, M., McIntosh, J. M. Nicotinic acetylcholine receptors in dorsal root ganglion neurons include the α6β4* subtype.

Dorsal root ganglia (DRG) contain neurons of the peripheral nervous system whose axons convey somatosensory information to the central nervous system (CNS). Attenuation of nociceptive information through DRG is used in the treatment of pain states (1–3). DRG neurons can be divided into various categories based on their different sensory modalities and size and by their expression of specific receptors and ion channels, including nicotinic acetylcholine receptors (nAChRs; refs. 4–8). nAChRs are cation channels composed of 5 homologous subunits that arrange in various combinations that impart each individual subtype with different biophysical and pharmacological properties. In mammals, there are 16 of these subunits, which are designated α1–α7, α9, α10, β1–β4, δ, ε, and γ based on sequence homology. Members of a subset of these subunits, α2–α10 and β2–β4, are expressed by neurons and have therefore been called neuronal nAChR subunits.

Functional studies indicate that DRG neurons express several nAChR subtypes, but the composition of these receptors is only partially known. Pharmacological evidence supports the presence of receptors with an α7-, α3β4*-, and α4β2*-like composition (where the asterisk denotes the possible presence of additional subunits; refs. 4, 8). However, RT-PCR analysis indicates that these neurons may express all of the neuronal nAChR subunits, including α6 (4, 7). α6* nAChRs in the CNS play a prominent role in modulating the release of catecholamines (9–12). These receptors are restricted to a limited number of discrete brain regions (13, 14). Nevertheless, recent evidence suggests that these nAChRs may play a central role in nicotine reinforcement and addiction (10, 15–17) and movement disorders such as Parkinson's disease (18, 19) Thus, there is intense interest in further defining the role of α6* nAChRs in the nervous system.

We therefore investigated whether DRG neurons functionally express nAChRs that contain the α6 subunit. Previous characterization of α6* nAChRs has been hampered due to the relative lack of ligands that discriminate between α6* receptors and other nAChR subtypes. In this study, we used whole cell voltage-clamp electrophysiology in combination with nAChR subtype-selective α-conotoxin (α-Ctx) analogs to probe for the expression of α6* receptors. Our results demonstrate that DRG neurons express a variety of nAChRs subtypes that include the α6β4* subtype.

MATERIALS AND METHODS

Male Sprague-Dawley rats were obtained from Charles River Laboratories (Wilmington, MA, USA). HBSS, HEPES, 2.5% trypsin, DMEM, B27 supplement, Glutamax, d-glucose, and penicillin/streptomycin were purchased from Invitrogen (Carlsbad, CA, USA). Nerve growth factor-7S (NGF), DNase (II) type-I, collagenase-A, poly-d-lysine hydrobromide, and acetylcholine chloride (ACh) were purchased from Sigma Aldrich (St. Louis, MO, USA). Heat-inactivated FBS was from HyClone (Logan, UT, USA). Dihydro-β-erythroidine hydrochloride was from Tocris Biosciences (Ellisville, MO, USA). All α-conotoxins were synthesized as described previously (20).

Glass coverslips (5 mm; cat. no. 64-0700) were purchased from Warner Instruments (Hamden, CT, USA), and thin-walled borosilicate glass capillaries (1B100F-4) were purchased from World Precision Instruments (Sarasota, FL, USA). Primers used in endpoint RT-PCR experiments were synthesized by the University of Utah Genome Core Facility, and the TaqMan riboprobes and all other reagents used in real-time RT-PCR experiments were purchased from Applied Biosystems (Foster City, CA, USA).

DRG neuron isolation and culture

PCR and electrophysiology experiments were carried out using male Sprague-Dawley rats, ages 25–45 d. For PCR experiments, rats were given an intraperiotoneal injection of 65 mg of sodium pentobarbitol in 1 ml of 0.9% normal saline and then perfused transcardially with 200 ml of ice-cold PBS until blood free. For electrophysiology experiments, rats were killed with 100% CO₂. For both PCR and electrophysiology experiments, the L1-L6 DRG were removed and placed in ice-cold HBSS buffered with 10 mM HEPES, pH 7.2 (dissection buffer). After the nerve roots were trimmed, the ganglia were bisected or quartered, depending on the size, and transferred to a 15 ml conical tube with 2 ml of dissection buffer containing 0.25% (v/v) trypsin and 0.1% (w/v) collagenase-A. The ganglia were then incubated at 37°C for 60 min, rinsed once with dissection solution, and mechanically dissociated in the presence of 5 mM MgCl₂ and 10 μg/ml DNase (II) type I using a Pasteur pipette. The cells were then suspended in 10 ml of dissection solution and passed through a 70-μm cell strainer to remove large pieces of tissue and then centrifuged at 200 g for 2 min. For electrophysiology experiments, the supernatant was aspirated off, and the cells were resuspended in 1.5 ml of DMEM containing 10% (v/v) FBS, 2% (v/v) B27, 2 mM glutamax, 100 U/ml penicillin, 100 μg/ml streptomycin, and 100 ng/ml NGF. The cells were plated on poly-d-lysine coated 5-mm coverslips in a single well of a 12-well Falcon culture plate and cultured at 37°C in 95% air/5% CO₂. All recordings were made the following day after plating.

PCR

Total RNA was isolated using the Ambion RNAqueous-4PCR kit (AM1914; Ambion, Austin, TX, USA) following the manufacturer's instructions. The samples were treated with DNase type I to digest contaminating DNA, and the RNA was quantified with a spectrophotometer. Total RNA was used to generate cDNA with the high-capacity cDNA reverse transcriptase kit from Applied Biosystems (4368814). Endpoint RT-PCR and agarose gel electrophoresis were performed using the DRG cDNA samples at the University of Utah Genomics Core Facility. Primers used to generate Fig. 1A have been published elsewhere (21). At least 1 primer per pair spanned an exon/exon boundary to eliminate amplification of genomic DNA, or both primers matched sequences in different exons, so that any genomic DNA amplification would be obvious due to band size. An additional primer set described in Genzen et al. (4) was also tested and produced similar results (data not shown). Real-time RT-PCR was performed in 96-well plates using the Applied Biosystems 7900HT system and the Applied Biosystems TaqMan Gene Expression Master Mix (4369016). TaqMan primer probe sets were purchase from Applied Biosystems: α2, Rn00591542_m1; α3, Rn00583820_m1; α4, Rn00577436_m1; α5, Rn00567155_m1; α6, Rn00589325_m1; α7, Rn00563223_m1; α9, Rn01413370_m1 and Rn00575997_m1; α10, Rn00575309_m1; rat β-actin, Rn00667869_m1. At least one primer of each pair spanned an exon/exon boundary to eliminate the possibility of amplifying genomic DNA.

Figure 1.

RT-PCR analysis of nAChR subunit transcripts. A) Endpoint RT-PCR products were detected for nAChR subunits α3–α7, α10, and β2–β4 and visualized using ethidium bromide fluorescence. Reactions for each primer set were performed in the absence of template (H₂O only) as controls and were negative (data not shown). B) Real-time RT-PCR analysis confirms the presence of α3–α7 and α10 nAChR subunit transcripts; transcripts for α2 and α9 were not detected or were below the threshold level of detection at cycle number 40 (data not shown).

Whole-cell patch-clamp electrophysiology

To initiate whole-cell voltage-clamp experiments, a 5-mm coverslip with neurons was placed in a diamond-shaped electrophysiology chamber (64-0280; Warner Instruments) and continuously gravity perfused at a rate of 2.5 ml/min with extracellular solution composed of 145 mM NaCl, 5 mM KCl, 2 mM CaCl₂, 1 mM MgCl₂, and 10 mM HEPES. The solution was adjusted to pH 7.4 with NaOH and to 320 mosmolar with d-glucose. Patch electrodes were pulled using a Flamming Brown P97 micropipette puller (Sutter Instruments, Novato, CA, USA). The resistances of the electrodes were 2–4 MΩ when filled with an internal solution containing 145 mM K-gluconate, 10 mM KCl, 1 mM EGTA, 10 mM d-glucose, 10 mM HEPES, 4 mM Na-ATP, and 0.5 mM Na-GTP. The pH of the internal solution was adjusted to 7.3 with KOH; osmolarity was 322 mosmolar. Recordings were made exclusively from neurons having diameters between 20 and 40 μm measured with a ×40 objective and a Nikon DS-Qi1 camera (Nikon Instruments, Melville, NY, USA). Gigaohm seal formation and whole-cell access were obtained using an ez-gSEAL 100A pressure controller (Neobiosystems, San Jose, CA, USA). The neurons were clamped at a holding potential of −80 mV using a Multiclamp 700B amplifier (Molecular Devices, Sunnyvale, CA, USA) and stimulated once every 2 min by pressure ejection (6 psi) of 1 mM ACh through a pipette, identical to that used for patch recordings, using a Picospritzer (General Valve Corp., Pine Brook, NJ, USA). The position of the pipette was always 50 μm upstream from the neuron under study. The duration of the ACh pulse (100 ms) and the 2-min interval between pulses avoided long-term desensitization. The ACh-gated currents (I_ACh) were filtered at 1 kHz through an 8-pole Bessel filter and digitized at 5 kHz, using a Digidata 1440A and pClamp 10.2 software (Molecular Devices), and stored on a personal computer for later analysis with Clampfit 10.2 analysis software (Molecular Devices) or GraphPad Prism (GraphPad Software, La Jolla, CA, USA). Access resistances (R_a) were generally <10 MΩ and were routinely compensated up to 80%. Pipettes with values of R_a > 15 MΩ after break-in were not used.

To assess the pharmacology of nAChR subtypes, the neurons were exposed to toxins by manually switching among extracellular solutions containing various toxins by means of a 3-way valve system and a 6-port manifold (64-0210; Warner Instruments). To determine the level of block produced by the toxins, 3 steady-state responses in the absence of toxins were averaged to determine the baseline response. Toxins were then applied sequentially and cumulatively, i.e., after application of toxin A, toxin B was applied in the continued presence of toxin A, etc. The amount of block from application of toxin A was determined by taking the response amplitude in the presence of the toxin divided by the baseline response. Block by each successive toxin was calculated as a percentage blockade of the response measured just before the application of that toxin. Unless stated otherwise, the percentage recovery from blockade was calculated as the difference between the responses in the presence of the toxin and those after washout. All recordings were made from only one cell for a given coverslip.

Two-electrode voltage-clamp electrophysiology of Xenopus laevis oocytes

Detailed methods for conducting electrophysiological experiments of nAChRs heterologously expressed in X. laevis oocytes have been previously described in detail (22). Briefly, stage IV–V oocytes were injected with equal ratios of cRNA encoding clones of rat nAChR subunits α3, α6/α3, β2, β3, and β4. The α6/α3 subunit chimera has been previously described and consists of an α3 subunit where the first 237 aa are replaced with the corresponding α6 amino acids (23). This chimera was used because of poor expression of the nonchimeric form of the α6 construct.

RESULTS

DRG neurons express multiple nAChR subunits including α6

Endpoint RT-PCR and real-time RT-PCR analyses of nAChR transcripts were performed on acutely dissociated lumbar L1–L6 DRG neurons from rats. We used the endpoint RT-PCR primer sets for nAChR subunits α2–α7, α9, α10, and β2–β4 described by Mikulski et al. (21) and a second set described by Genzen et al. (4). Qualitatively, the results were the same for all subunits using both primer sets. The results obtained using the primer set from Mikulski et al. (21) are shown in Fig. 1A. We were unable to detect transcripts for the α2 or α9 subunits and therefore used real-time RT-PCR that monitors the amount of product produced after each PCR cycle and may be more efficient at detecting transcripts of low abundance. As shown in Fig. 1B, the real-time RT-PCR results confirmed the presence of α6 and of other nAChR subunits. These results were also similar in comparison to the end point RT-PCR results, and in all cases signals for α2 and α9 subunits were either not detected or were below the threshold level of detection (data not shown).

α-Ctxs ArIB[V11L;V16D] and BuIA identify 2 types of responses to ACh

Previous pharmacological analysis suggests that DRG neurons express ≥3 classes of nAChRs (4). These receptors include α7 homopentamers and heteropentamers that contain either β2 or β4 subunits. As an initial step in identifying responses attributable to these classes, we initiated our studies using α-Ctx ArIB[V11L;V16D], a modified analog of a peptide isolated from Conus arenatus, and α-Ctx BuIA from Conus bullatus. ArIB[V11L;V16D] is a highly selective antagonist with nanomolar potency (IC₅₀=1.0 nM) on native and heterologously expressed α7 nAChRs (24, 25), whereas BuIA blocks most neuronal nAChRs with picomolar to nanomolar potency except for α4β2 (26). We observed 2 distinct types of responses to ACh in DRG neurons. The first type (category I) was characterized by rapid activation and desensitization kinetics with a 10% to 90% rise time of 31.6 ± 6 ms and a 90 to 10% decay time of 600.0 ± 129 ms (n=8; Fig. 2A). The second type of response (category II) was characterized by slower activation and desensitization kinetics. The 10 to 90% rise time for this response type was 160.6 ± 27 ms with a 90 to 10% decay time of 2486.0 ± 250 ms (n=17; Fig. 2B). Category I responses were rapidly blocked (95.5±1%, n=8; Fig. 3A) by 200 nM ArIB[V11L;V16D] and recovered slowly after 12 min toxin washout (20.1±2%; Fig. 3B). The kinetics of toxin blockade and recovery are consistent with those reported in Xenopus oocytes expressing homomeric rat α7 nAChRs (24). As an additional confirmation, we assessed the potency and off-rate kinetics of α-cobratoxin (α-Cbtx) from Naja naja kaouthia that would be expected to potently and essentially irreversibly block α7 nAChRs (27). Perfusion with 100 nM α-Cbtx abolished the response to ACh (96.4±1%, n=4; Fig. 3C); there was little recovery even after a 20-min wash with extracellular solution (8.7±6%, n=4; Fig. 3D). In contrast, category II responses were relatively insensitive to 200 nM ArIB[V11L;V16D] (12.6±4% blockade, n=6; Fig. 3E); however, substantial blockade was achieved using 500 nM BuIA (90.3±3%, n=6; Fig. 3E). The recovery from blockade by BuIA was slow and recovered by 8.0 ± 1% (n=6) from the initial level of blockade after a 12 min wash.

Figure 2.

Whole-cell voltage-clamp electrophysiology of DRG neurons. Neurons were held at −80 mV and stimulated with 1 mM ACh for 100 ms every 2 min. Under these conditions, 2 broad types of responses were observed. A) Category I responses were characterized by rapid activation and desensitization kinetics. B) Category II responses were characterized by slower kinetics relative to category I responses.

Figure 3.

Pharmacological evaluation of category I and II response types. A, B) Category I responses are blocked by α7-selective nAChR antagonists (A); blockade was observed using ArIB[V11L;V16D] (200 nM), and the response slowly recovered during washout (B). C, D) α-Cbtx (100 nM) also blocked category I responses (C); there was very little recovery from blockade even after 20 min of wash (D). E) Category II responses are mainly mediated by heteromeric nAChRs. These responses were relatively insensitive to ArIB[V11L;V16D] (200 nM) but were substantially blocked by BuIA (500 nM). F) In the continued presence of ArIB[V11L;V16D] (200 nM), recovery from blockade by BuIA was very slow; there was little recovery after a 12 min wash. c, control response before application of toxins.

Category II responses are mainly attributable to nAChRs containing the β4 subunit

During our initial investigation of category II responses, we observed that the kinetics of blockade and recovery from block by BuIA were remarkably slow. In Xenopus oocytes, BuIA distinguishes between heteromeric receptors that contain β2 vs. β4 subunits based on kinetics of blockade and unblock. Both the on- and off-rate kinetics of BuIA are markedly slower for β4- vs. β2-containing nAChRs (26). We used this distinguishing characteristic as a first approximation to estimate the relative contribution of heteromeric β2- vs. β4-containing nAChRs that mediate these slower responses. After blocking α7 nAChRs with ArIB[V11L;V16D], we perfused the neurons with 500 nM BuIA and monitored the inhibition of the responses. The rate of blockade using BuIA was slow (Fig. 4A, B) and consistent with the slow rate of blockade observed for rat β4-containing receptors heterologously expressed in Xenopus oocytes. Similarly, the recovery rate was also very slow. We assessed recovery rates by simultaneously perfusing 200 nM ArIB[V11L;V16D] together with 5 μM BuIA, and then we monitored the rate of recovery in the continued presence of ArIB[V11L;V16D] (to keep any α7 nAChRs blocked). Under these conditions, the 2 toxins together produced a 91.8 ± 3% (n=8) blockade of the response after a 6-min exposure (Fig. 4C). The responses were then monitored for recovery during washout of BuIA. In most cases, there was very little recovery of the responses even after a 20-min wash (3.7±2%, n=8; Fig. 4D). As shown in Fig. 4A–C, we observed a residual response of ∼10%. BuIA is least potent at blocking α4β2, IC₅₀ > 10 μM (26), and therefore we presumed that this residual response was attributable to this nAChR subtype, which has been previously reported in DRG neurons (4). We applied 200 nM ArIB[V11L;V16D], 500 nM BuIA, and 1 μM DHβE (to block α4β2* nAChRs) simultaneously and observed nearly complete blockade of the response (96.9±1%; n=9; Fig. 4E). Taken together, these results led us to hypothesize that the predominant nAChR subtypes expressed by DRG neurons with category II responses was a heteromeric nAChR with an α_xβ4* composition.

Figure 4.

Slow kinetics of blockade and recovery from blockade category II responses by BuIA indicate a predominance of α_xβ4 nAChRs. A) Superimposed trace recordings of ACh (100 ms pulse every 2 min) responses from a neuron before and during a 20-min perfusion with BuIA (500 nM). All solutions contained ArIB[V11L;V16D] (200 nM) to inhibit any α7 nAChRs. B) Single exponential fit of data showing the observed rate of blockade by BuIA (500 nM); t_1/2 = 3.9 min (95% CI: 2.9–5.9). Error bars = se for 6 neurons. C, D) To determine recovery from BuIA blockade, neurons were simultaneously perfused with ArIB[V11L;V16D] (200 nM) and BuIA (5 μM) for 6 min (C), and ACh responses were monitored for recovery during washout of BuIA (in the continued presence of ArIB[V11L;V16D] to keep α7 nAChRs inhibited); there was little recovery after a 20 min wash (D). E) Blockade of α4β2* by DHβE. Near complete blockade of the response was observed by simultaneous perfusion with a combination of BuIA (500 nM), ArIB[V11L;V16D] (200 nM), and DHβE (1 μM). Additional blockade by DHβE is consistent with the presence of α4β2* nAChRs. For comparison, see panels A–C and Fig. 3E. c, control response before perfusion with toxin solution. Panels A, C, E show results from single neurons; see Results for averages of multiple neurons.

α-Ctx PnIA blocks α6/α3β2β3 and α3β2 but neither α6/α3β4 nor α3β4 nAChRs

Some α-Ctxs and their analogs show remarkable specificity for nAChR subtypes. We sought to exploit this specificity to further identify the nAChRs mediating category II responses. α-Ctx PnIA from Conus pennaceous is a potent, IC₅₀ = 9.6 nM (28), blocker of α3β2 nAChRs but also blocks the α6/α3β2β3 subtype in Xenopus oocytes (29). Since the potency on the α6/α3β2β3 subtype was not previously quantitated, we expressed the receptor in Xenopus oocytes and determined the IC₅₀ value to be 10.9 nM (Fig. 5A). Next, we assessed PnIA for its ability to discriminate between different α3* and α6* subtypes. At a concentration of 500 nM, PnIA blocked >95% of responses mediated by α6/α3β2β3 (Fig. 5B; 98.1±0.4%; n=5) and α3β2 (Fig. 5C; 95.4±0.2%, n=5) but had little effect on α6/α3β4 or α3β4 nAChRs. The average response to ACh in the presence of 500 nM PnIA was 92.3 ± 4% (Fig. 5D; n=4) for α6/α3β4 and 98.5 ± 3% (Fig. 5E; n=4) for α3β4 of control responses, respectively. Thus, in a heterogeneous population of receptors, PnIA would block any α3β2* or α6β2* nAChRs without substantially blocking α6β4* or α3β4* nAChRs.

Figure 5.

α-Ctx PnIA blocks α3β2 and α6/α3β2β3 but not α3β4 or α6/α3β4 nAChRs in Xenopus oocytes. A) Blockade of α6/α3β2β3 was determined by perfusing the oocyte with solutions containing increasing concentrations of PnIA and yielded a value of IC₅₀ = 10.9 nM (95% CI: 7.9–14.9), n_H = 1.0 (95% CI: 0.7–1.4). Error bars = se for 5 individual experiments. B) Blockade of α6/α3β2β3 by PnIA (500 nM). C–E) PnIA (500 nM) also blocks α3β2 nAChRs (C) but neither α6/α3β4 nAChRs (D) nor α3β4 nAChRs (E). Panels B-E show responses in a single oocyte;see Results for averages of multiple oocytes. /–/ indicates a 5-min exposure to PnIA.

Subtype-selective α-Ctxs identify α6β4* nAChRs in DRG neurons

Several subtype-selective α-Ctxs were chosen from a panel of ligands, shown in Table 1, to develop a cocktail of antagonists that would allow us to further isolate α_xβ4* nAChRs. First, we assessed whether category II responses were sensitive to PnIA. Since PnIA also blocks α7 nAChRs, IC₅₀ = 252 nM (28), we perfused the cells first with 200 nM ArIB[V11L;V16D] (to eliminate any α7 current) followed by 500 nM PnIA. Similar to the results shown in Fig. 3E, ArIB[V11L;V16D] produced minimal blockade (7.2±2%; n=9; Fig. 6A); PnIA blocked 23.0 ± 4% (n=9; Fig. 6A) of the residual response, indicating that approximately one-fifth of the category II response was mediated by α3β2* and/ or α6β2* nAChRs. With these results in mind, we combined 200 nM ArIB[V11L;V16D], 500 nM PnIA, and 1 μM DHβE in a cocktail to block α7, α3β2*, α4β2*, and α6β2* nAChRs, and to effectively isolate α_xβ4* subtypes. When tested on DRG neurons with category II responses, this cocktail blocked 35.9 ± 6% (Fig. 6B; n=4). Next, we designed a series of experiments using the α-Ctxs shown in Table 1 to assess for the presence of α6β4* nAChRs. The criteria used for the selection of α-Ctxs included the potency, the subtype selectivity, the kinetics of blockade and the reversibility of the blockade. For example, in the presence of the cocktail, substantial and reversible blockade by 300 nM BuIA[T5A;P6O] would indicate the presence of α6β4* receptors (30). Indeed, we observed a 51.9 ± 11% (n=3; Fig. 7A) block by a 10 min exposure to 300 nM BuIA[T5A;P6O]; the responses recovered to 97.0 ± 8% (n=3) of baseline after a 12-min wash. Since the predominant subtype expressed in DRG neurons contained the β4 subunit, and partial inhibition was observed using the α6β4-selective BuIA[T5A;P6O], the residual current was hypothesized to be mediated by α3β4* receptors. Following the sequential and cumulative application of the cocktail and then 300 nM BuIA[T5A;P6O], 10 μM α-Ctx AuIB, from Conus aulicus (31), was applied. This procedure produced 7.4 ± 6, 47.4 ± 4, and 77.8 ± 4% blockade (n=4), respectively, for each of the toxin conditions. The substantial blockade by AuIB confirms the presence of α3β4* nAChRs.

Table 1.

IC₅₀ values of antagonists obtained in Xenopus oocytes

Antagonist	nAChR subtype
	α3β2	α3β4	α4β2	α4β4	α6/α3β2β3	α6β4	α7
ArIB[V11L;V16D]	>20 μM (24)	>20 μM (24)	>20 μM (24)	>20 μM (24)	>20 μM (24)	>20 μM (24)	1.1 nM (24)
AuIB	>100 μM (31)	750 nM (31)	>100 μM (31)	>100 μM(31)	>100 μM (31)	ND	∼10 μM (31)
BuIA	5.7 nM (26)	27.7 nM (26)	>10 μM (26)	69.9 nM (26)	258 pM (26)	1.5 nM (26)	272 nM (26)
BuIA[T5A;P6O]	>10 μM (30)	1.2 μM (30)	>10 μM (30)	>10 μM (30)	>10 μM (30)	58.1 nM (30)	>10 μM (30)
DHβE	2 μM (32)	25 μM (33)	100 nM (32)	190 nM (32)	1.1 μM (32)	ND	8.0 μM (32)
MII[E11A]	8.7 nM (23)	2.1 μM (23)	ND	>10 μM (23)	154 pM (23)	6.4 nM (23)	1.1 μM (23)
PnIA	9.6 nM (28)	>1 μM (29)	>1 μM (29)	ND	10.9 nMa	>500 nMa	252 nM (28)

ND, not determined.

^aPresent study.

Figure 6.

Effect of PnIA on category II responses in DRG neurons. PnIA blocks a portion of category II responses. A) Neurons were perfused with ArIB[V11L;V16D] (200 nM) followed by PnIA (500 nM). B) A cocktail of antagonists consisting of ArIB[V11L;V16D] (200 nM), PnIA (500 nM), and DHβE (1 μM), to block α7, α3β2*/α6β2*, and α4β2* nAChRs, respectively, blocked a portion of the response similar to that observed in panel A. c, control response before perfusion with toxins.

Figure 7.

Category II responses are blocked by α-Ctxs that target α6β4* nAChRs. A panel of nAChR subtype-selective α-Ctxs, listed in Table 1, was used to isolate and block α6β4* nAChRs in neurons with category II responses. A) After perfusion with the cocktail, the α6β4 nAChR antagonist BuIA[T5A;P6O] (300 nM) reversibly blocked a substantial portion of the remaining response. B) Sequential and cumulative application of the cocktail, followed by BuIA[T5A;P6O] (300 nM), then AuIB (10 μM) to block α3β4* nAChRs. C) Omission of DHβE from the cocktail, followed by perfusion with 100 nM, then 300 nM BuIA[T5A;P6O]. D) Following the same protocol as in panel B, perfusion of the cocktail, followed by MII[E11A] (200 nM) to block α6β4* nAChRs, and then AuIB (10 μM). Bottom graphs show time course of blockade by the toxins; bars labeled a–c indicate perfusion duration for each toxin solution and correlate with same labels in trace recordings (top panels). Results are from single neurons; see Results for averages of multiple neurons. c = control response before application of toxin.

Although BuIA[T5A;P6O] is most potent at blocking α6β4* vs. α3β4*, some blockade of α3β4* nAChRs might nevertheless occur using a concentration of 300 nM. This consideration, coupled with concerns about the activity of DHβE on β4-containing nAChRs (see Table 1 and refs. 32, 33), prompted us to conduct the following experiment. We omitted DHβE from the cocktail solution and then employed a lower concentration of BuIA[T5A;P6O] to block α6β4* specifically. Under these conditions, 100 nM BuIA[T5A;P6O] blocked 40.5 ± 9%, and 300 nM blocked an additional 35.0 ± 3% (n=4) of the remaining response (Fig. 7C). As an additional confirmation of the presence of α6β4*, we used α-Ctx MII[E11A], a modified analog of α-CTx MII from Conus magus (23), which has a selectivity ratio > 300 for α6β4* vs. α3β4* (23). Thus, in the presence of the cocktail, any α6β2* or α3β2* nAChRs are blocked, and therefore MII[E11A] would selectively block α6β4* nAChRs. Following the application protocol used in Fig. 7B, perfusion with the cocktail blocked 14.2 ± 7% (n=4), and application of 200 nM MII[E11A] further blocked the response by 46.7 ± 8% (n=4), providing additional evidence for the expression of an α6β4*-containing nAChR. Lastly, 10 μM AuIB was applied, which blocked 75.3 ± 6% (n=4) of the remaining response (Fig. 7D), consistent with the additional presence of α3β4* nAChRs.

DISCUSSION

In this study, we used PCR analysis and subtype-selective α-Ctxs to demonstrate for the first time that α6β4* nAChRs are among the nAChR subtypes expressed by rat DRG neurons. Transcripts for the α6 subunit were detected using both endpoint RT-PCR and real-time RT-PCR in addition to transcripts for other nAChR subunits (Fig. 1A, B). Despite using 2 primer sets for the α2 subunit, we were unable to detect transcripts in endpoint RT-PCR, nor were we able detect transcripts for the α9 subunit as previously reported (7). Signals for both subunits were occasionally observed with real-time RT-PCR but were either not detectable or barely above the threshold level for detection (data not shown).

Functional evaluation of α6β4* nAChR expression was accomplished using whole-cell voltage-clamp electrophysiology to record ACh-evoked responses and a panel of subtype-selective α-Ctxs. In general, DRG responses to ACh could be classified into 2 categories based on activation and desensitization kinetics. Category I responses displayed kinetics characteristic of α7 nAChRs and were sensitive to blockade by the α7-selective antagonists α-Ctx ArIB[V11L;V16D] and α-Cbtx (Fig. 3A–D). In contrast, category II responses displayed markedly slower kinetics and were relatively insensitive to ArIB[V11L;V16D] (Figs. 3E and 6A). These responses were slowly, and in some cases irreversibly, blocked by α-Ctx BuIA (Figs. 3F and 4D), suggesting that category II responses were predominantly mediated by nAChRs with an α_xβ4 composition. Nevertheless, the ∼20% block by α-Ctx PnIA suggests that DRG neurons also express α3β2* and/or α6β2* nAChRs (Fig. 6). At the present time, we are unable to investigate this finding further, because, to our knowledge, there are no ligands available that can distinguish between α3β2* and α6β2* without also blocking α6β4* nAChRs.

A cocktail of α-Ctxs in combination with DHβE was used to isolate nAChRs containing the β4 subunit. After the selective blockade of α3β2*, α4β2*, α6β2*, and α7 nAChRs with this cocktail, perfusion with either BuIA[T5A;P6O] or MII[E11A] produced substantial blockade of the remaining response, providing pharmacological evidence for α6β4* nAChRs (Fig. 7). In addition, the remaining response was sensitive to AuIB, indicating the presence of α3β4* nAChRs. We note that these experiments do not rule out the possible presence of the α5 subunit, as suggested by Rau et al. (8), nor do they rule out other multisubunit combinations, such as α3α6β4 or those proposed by Grinevich et al. (34), including α6β4β3 and α6β4β3α5. Nevertheless, the simplest interpretation of category II responses is that neurons that display this response type express a heterogeneous population of nAChRs that minimally consist of α3β2* and/or α6β2*, α3β4*, α6β4*, α7, and probably α4β2* nAChRs. Thus, it would be highly desirable to conduct immunoprecipitation studies on DRG neurons that may reveal which subunits assemble together with α6 (35–37).

While we broadly classified DRG neurons into 2 categories based on activation and desensitization kinetics, this classification was done for simplification purposes and should not be taken to imply that there are only 2 ACh-sensitive populations of DRG neurons. There may, in fact, be many subpopulations, and the nAChRs subtypes may be distributed heterogeneously across the different subpopulations. For example, α7 nAChRs appear to coexpress with heteromeric nAChRs in category II responses, but their contribution to the total I_ACh is generally small (∼10%). This is consistent with studies that used selective agonists and positive allosteric modulators to demonstrate the expression of α7 nAChRs; in these studies, the contribution of α7 was small relative to the overall size of the response (6).

α6 nAChR mRNA was first reported by Lamar et al. (38) in 1990. Since then, numerous studies have demonstrated a functional role for α6* nAChRs in dopaminergic neurons. More recent studies have also reported their presence in noradrenergic terminals in the hippocampaus (39), presynaptic GABAergic boutons (40), and human adrenal chromaffin cells (41). α6* nAChRs have a relatively limited tissue distribution pattern and have been implicated in drug reward and dependence (16, 42) and in Parkinson's disease (43–45). Consequently, significant efforts have been devoted to the study of α6* nAChRs. As demonstrated in the present study, α6 nAChRs are also expressed by sensory neurons; therefore, they may also play a role in disorders related to sensory perception.

Compounds that act on nAChRs are analgesic and are currently being investigated for their application in the treatment of chronic and neuropathic pain syndromes (46–49). For example, epibatidine, isolated from the South American frog Epipedobates tricolor, is a potent analgesic. However, severe adverse side effects due to lack of nAChR subtype specificity limit the use of this toxin as therapeutic compound. Novel nAChR agonists have substantially improved selectivity and have implicated α4* nAChRs in the analgesic response (50, 51). Notably, however, there are no reported agonists that discriminate well between α6* and α4* nAChRs. The results of the present study extend our knowledge of the tissue distribution of α6* nAChR subtypes in general and in particular in DRG. This information may help enable a more defined target-based search for novel therapeutics.