Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The pelvic ganglia provide autonomic innervation to the lower bowel and various urogenital organs, including the urinary bladder, prostate, and penis (for review, see Keast, 1999). Physiologically, the ganglia play important roles in various autonomic reflexes, including micturition and penile erection (de Groat and Booth, 1993a; de Groat et al., 1993). A distinctive feature of the pelvic ganglia that differentiate them from other autonomic ganglia is the colocalization of both sympathetic and parasympathetic postganglionic neurons within the same ganglion capsule (Keast, 1999). Anatomical structures of the pelvic ganglia show variability among different species and genders. Because of their relatively simple anatomy and thus, ease of isolation, manipulation and quantification (Keast, 1999), male rat pelvic ganglia, termed the major pelvic ganglia (MPG), have been used as a model system for studying physiological and pathophysiological aspects of the neural control of pelvic viscera.

MPG neurons are known to express various putative neurotransmitters, including neuropeptide Y (NPY), vasoactive inhibitory peptide, and nitric oxide in addition to classical neurotransmitters such as norepinephrine and acetylcholine (Keast and de Groat, 1989; Keast, 1995; Zhu et al., 1995). Electrophysiological studies have shown that these neurotransmitters act as modulators of voltage-activated Ca²⁺ channels, which play important roles in synaptic transmission and neuronal excitability (Zhu et al., 1995; Zhu and Yakel, 1997). Indeed, MPG seem to function as a potential integration site where both adrenergic and cholinergic synaptic transmission toward effector organs would be edited (Theobald and de Groat, 1989; Zoubek et al., 1993; Warren and Lavidis, 1996; Félix et al., 1998; Keast, 1999).

Considerable evidence has accumulated suggesting that purines such as ATP and adenosine act as neurotransmitters/modulators in the pelvic plexus (De Groat and Booth, 1993b). Generally, ATP exerts excitatory effects via P₂ receptors on the urogenital smooth muscles, including the vas deferens, urinary bladder, and urethra (Fujii, 1988). A recent study has shown that ionotropic P_2X receptors are also present in rat MPG neurons although their roles in excitatory ganglionic transmission are unclear (Zhong et al., 1998). In comparison with ATP, adenosine is known to produce inhibitory effects on ganglionic and neuromuscular transmission via adenosine (P₁) receptors presumably located on postganglionic pelvic neurons and nerve terminals (Akasu et al., 1984; Theobald and de Groat, 1989). To date, however, the nature of the adenosine receptor subtype and the cellular mechanisms underlying the inhibitory actions of adenosine remain unclear. In preliminary experiments with rat MPG neurons, we observed that adenosine, but not ATP, was capable of inhibiting voltage-activated Ca²⁺ currents. In the present study, thus, we identified the subtype of adenosine receptor and the signaling pathway responsible for the adenosine-induced Ca²⁺ current inhibition in rat MPG neurons by using patch-clamp and RT-PCR techniques. Our data suggest that adenosine inhibits N-type Ca²⁺ currents by activation of A₁ receptors via a voltage-dependent and pertussis toxin (PTX)-sensitive pathway in rat MPG neurons.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Preparation of Pelvic Ganglion Neurons. MPG neurons were enzymatically dissociated with some modifications of the method described previously (Zhu et al., 1995). Briefly, adult (200-300 g) male Sprague-Dawley rats were anesthetized with pentobarbital sodium (50 mg/kg i.p.). MPG clusters were dissected out from the lateral surface of the prostate gland and placed in cold Hanks' balanced salt solution. The ganglia were then desheathed, cut into small pieces, and incubated in Earle's balanced salt solution containing 0.7 mg/ml collagenase type D (Roche Molecular Biochemicals, Indianapolis, IN), 0.1 mg/ml trypsin type I (Roche Molecular Biochemicals), and 0.1 mg/ml DNase type I (Sigma Chemical, St. Louis, MO) at 35°C for 1 h in a shaking water bath. After incubation, ganglia were dispersed into single neurons by vigorous shaking of the culture flask containing the ganglia. After centrifugation at 50g, the neurons were resuspended in RPMI 1640 containing 10% fetal calf serum and 1% penicillin-streptomycin (all from Invitrogen, Carlsbad, CA). Neurons were then plated onto culture dishes (35-mm) coated with poly-L-ornithine and maintained in a humidified 95% air, 5% CO₂ incubator at 37°C. Neurons were used within 24 h after plating. As appropriate, neurons were incubated overnight (14-18 h) with 500 ng/ml PTX.

RT-PCR Analysis. Total RNA from dissociated MPG neurons was prepared using a modified guanidinium thiocyanate-phenol-chloroform extraction method (Chomczynski and Sacchi, 1987). Synthesis of the first strand of cDNA was performed in an RT-PCR buffer containing 2 µg of total RNA, 25 of nmol of dNTP, 0.5 µg of random hexamer, 20 U of RNase inhibitor, and 200 U of murine leukemia virus reverse transcriptase (all from Promega, Madison, WI) in a final volume of 25 µl at 37°C for 60 min. Specific sense and antisense primer pairs were designed based on the known cloned rat adenosine receptor sequences deposited in GenBank (Table 1). Single-stranded cDNA products were denatured at 94°C for 5 min and then subjected to PCR amplification (32 cycles). Each PCR cycle consisted of denaturing at 94°C for 30 s, annealing at 56°C for 30 s, and extension at 72°C for 1 min in a GeneAmp thermocycler (PerkinElmer Instruments, Norwalk, CT). The PCR buffer (50 µl) contained the transcribed cDNA, 10 pmol of primers, 10 nmol of dNTP, and 1.25 U of Taq polymerase (PerkinElmer Instruments). As a PCR control, amplification of rat 28S RNA was performed for 24 cycles under the same temperature conditions. The resultant PCR products were separated and visualized on a 1.1% agarose gel containing ethidium bromide.

Electrophysiology. Ca²⁺ channel currents were recorded using the whole-cell variant of the patch-clamp technique. Patch electrodes were fabricated from a borosilicate glass capillary (Corning 7052; Garner Glass Co., Claremont, CA) by using a P-97 Flaming Brown micropipette puller (Sutter Instrument Co., San Rafael, CA). The patch electrodes were fire polished on a microforge (Narishige, Tokyo, Japan) and had resistances of 1 to 3 M when filled with the internal solution described below. An Ag/AgCl wire was used to ground the bath. The cell membrane capacitance and series resistance were compensated (>80%) electronically using the patch-clamp amplifier (Axopatch 1D; Axon Instruments, Foster City, CA). Voltage protocol generation and data acquisition were performed using pClamp 6.03 software on an IBM computer equipped with an analog-to-digital converter (Digidata 1200; Axon Instruments). Current traces were filtered at 2 to 5 kHz by using the four-pole Bessel filter in the clamp amplifier and stored on the computer hard drive for later analysis.

Solution and Drugs. Ca²⁺ currents were isolated using patch electrodes filled with an internal solution containing 120 mM N-methyl-D-glucamine-methanesulfonate (MS), 20 mM tetraethylammonium-MS, 20 mM HCl, 11 mM EGTA, 1 mM CaCl₂, 10 mM HEPES, 4 mM Mg-ATP, 0.3 mM Na₂-GTP, 14 mM creatine phosphate, pH 7.2. External recording solution contained 145 mM tetraethylammonium-MS, 10 mM HEPES, 10 mM CaCl₂, 15 mM glucose, 0.0003 mM tetrodotoxin, pH 7.4. Drugs were applied to single neurons via a gravity-fed fused silica capillary tube connected to an array of seven polyethylene tubes. The outlet of the perfusion system was located within 100 µm of the cell. The bath superfusion rate was approximately 1 to 2 ml/min. All experiments were performed at room temperature (20-24°C). Drugs used in experiments were obtained as follows: -conotoxin GVIA (-CgTx GVIA) from Peninsula Laboratories (Belmont, CA); guanosine-5'-thiodiphosphate (GDPS), N⁶-cyclopentyladenosine (CPA), CGS 21680, and 8-cyclopentyl-1,3-dipropylxanthine (DPCPX) from Sigma/RBI (Natick, MA); and adenosine, tetrodotoxin, PTX, and nimodipine from Sigma Chemical. For stock solutions (1-100 mM), all drugs were dissolved in distilled water except CPA and CGS 21680, which were dissolved in dimethyl sulfoxide.

Data Analysis. Amplitudes of step currents were usually determined isochronally 10 ms after the onset of a test pulse, normalized to membrane capacitance, and expressed as pA/pF. Membrane capacitance (C_m) was measured by applying a 20-ms, 10-mV hyperpolarizing step from a holding potential of 80 mV and calculated according to the following equation (Bénitah et al., 1993): C_m =  · I_o/V_m (1  I/I_o), where  is the time constant of the capacitive current, I_o is the maximum capacitive current value, V_m is the amplitude of a voltage step, and I is the amplitude of the steady-state current. Concentration-response curves and IC₅₀ values for half-maximal Ca²⁺ current inhibition were obtained from fitting to a single-site binding isotherm with least-squares nonlinear regression. Data were presented as means ± S.E.M. Statistical significance was determined using Student's t test, and p < 0.05 was considered significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Characteristics of Ca²⁺ Currents in MPG Neurons. Figure 1A illustrates inward Ca²⁺ currents in a MPG neuron elicited by a voltage ramp to +80 mV from a holding potential of 80 mV. Based on the absence or presence of low voltage-activated (LVA) T-type Ca²⁺ currents, two different subpopulations of MPG neurons could be distinguished. According to a previous report (Zhu et al., 1995), adrenergic MPG neurons express T-type Ca²⁺ channels, whereas cholinergic MPG neurons do not. In a subpopulation of the tested neurons, there were prominent voltage humps between 50 and 20 mV on the current-voltage (I-V) curves, indicating the presence of LVA Ca²⁺ currents (Fig. 1A). The average C_m of neurons expressing T-type Ca²⁺ channels was 82 ± 3 pF (n = 50), compared with 47 ± 3 pF (n = 54) for neurons lacking T-type Ca²⁺ channels (p < 0.01) (Fig. 1B). The correlation between C_m (an indirect measurement of membrane surface area) and expression of T-type channels was consistent with previous findings in rat MPG neurons (Zhu et al., 1995). HVA Ca²⁺ current density was not significantly different for neurons with or without T-type Ca²⁺ channels (44 ± 4 versus 49 ± 5 pA/pF, respectively; p > 0.05) (Fig. 1C).

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Whole-cell Ca2+ currents in male rat MPG neurons. A, representative traces of inward Ca2+ currents in adrenergic and cholinergic neurons. The Ca2+ currents were elicited by voltage ramps from 80 to +80 mV in neurons held at 80 mV. Adrenergic neurons were distinguished from cholinergic neurons by the presence of low voltage-activated T-type Ca2+ currents. B and C, summary of average Cm and current densities (pA/PF) in adrenergic (n = 50) and cholinergic (n = 54) neurons, respectively. The peak Ca2+ currents were elicited by test pulses to +10 mV from a holding potential of 80 mV. Current amplitudes were determined isochronally 10 ms after onset of test pulses and normalized to membrane Cm. Data represent the mean ± S.E.M. ***p < 0.001.

Inhibition of Ca²⁺ Currents by Adenosine. We tested whether adenosine modulates Ca²⁺ currents elicited by the ramp protocol in MPG neurons. As shown in Fig. 2A, application of 30 µM adenosine significantly inhibited HVA, but not LVA Ca²⁺ currents recorded from adrenergic neurons. The I-V relationships for the Ca²⁺ currents in the absence or presence of adenosine are shown in Fig. 2B. On average, adenosine inhibited the peak Ca²⁺ currents by 36 ± 3% (n = 9). We also determined whether Ca²⁺ currents were differentially modulated by adenosine in adrenergic and cholinergic neurons. The degree of Ca²⁺ current inhibition by adenosine at concentrations ranging between 0.01 and 30 µM was not different for the two subpopulations (Fig. 2C). This is in contrast to the previous findings for ₂-adrenergic receptor-mediated Ca²⁺ current inhibition in rat MPG neurons (Zhu and Yakel, 1997). Accordingly, we did not discriminate between the two subpopulations in the following experiments.

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Adenosine-induced Ca2+ current inhibition in rat MPG neurons. A, representative traces of Ca2+ currents in absence (solid line) and presence (dotted line) of 30 µM adenosine. The ramp currents were elicited by the protocol described in Fig. 1 in an adrenergic neuron. Note that T-type Ca2+ currents were not modulated by adenosine. B, averaged I-V relationship in the absence () and presence () of adenosine for adrenergic neurons (n = 9). C, concentration-response curves for adenosine-induced Ca2+ current inhibition in adrenergic (n = 9; ) and cholinergic (n = 8; ) neurons. Inhibition (%) was determined as (1  Idrug/Icontrol) × 100%. The smooth curves were obtained by fitting data to a single-site binding isotherm with a nonlinear least-squares regression program. In B and C, data represent the mean ± S.E.M.

Identification of Adenosine Receptor Subtype Involved in Ca²⁺ Current Inhibition. To identify subtypes of adenosine receptors expressed in MPG neurons, we performed RT-PCR with four pairs of primers specific to adenosine receptor isoforms (A₁, A_2a, A_2b, and A₃) (Table 1). We confirmed that all the primers did properly work in control RT-PCR experiments by using the brain (A₁, A_2a, and A_2b) and the testis (A₁, A_2b, and A₃) (data not shown). In addition, the possibility of genomic DNA contamination was eliminated by PCR experiments without prior reverse transcription (data not shown). As shown in Fig. 3, A₁ and A_2a receptor mRNAs, predicted as 205- and 371-bp products, respectively, were detected after the RT-PCR reaction in MPG neurons. In contrast, A_2b and A₃ receptor mRNAs were not detected in MPG neurons.

View larger version (49K): [in this window] [in a new window]   Fig. 3.   RT-PCR analysis of mRNAs encoding adenosine receptors expressed in rat MPG neurons. Total RNA isolated from MPG neurons was reverse transcribed, and amplified by PCR with four pairs of primers specific to adenosine receptor isoforms (A1, A2a, A2b, and A3) (Table 1). The resultant PCR products were separated and visualized on a 1.1% agarose gel containing ethidium bromide. As an internal control, rat 28S RNA was amplified (103 bp). M, DNA size marker.

View larger version (16K): [in this window] [in a new window]   Fig. 4.   Identification of the receptor subtype involved in adenosine-induced Ca2+ current inhibition. A, concentration-response curves for Ca2+ current inhibition induced by adenosine and adenosine analogs: adenosine (ADO, ), CPA (A1-selective agonist, ), and CGS 21680 (A2-selective agonist, ). The currents were elicited and amplitudes were determined as in Fig. 1. The inhibition (%) was normalized to that induced by each agonist at a maximal concentration. The curve was obtained from fitting to a single-site binding isotherm with least-squares nonlinear regression. Parentheses represent the EC50 values. B, summary of Ca2+ current inhibition induced by different agonists in absence () and presence () of 0.1 µM DPCPX, an A1-selective antagonist: NE (10 µM; n = 4), ADO (30 µM; n = 7), CPA (10 µM; n = 4), and ATP (100 µM; n = 4). DPCPX was pretreated for 10 min. Data represent the mean ± S.E.M. ***p < 0.01.

Effects of GDPS and PTX on Adenosine-Induced Ca²⁺ Current Inhibition. Dialysis of neurons with GDPS, a hydrolysis-resistant GDP analog, has been shown to abolish the G protein-mediated effects of agonists by acting as a competitive inhibitor of GTP binding to the G subunits (Holz et al., 1986; Jeong and Wurster, 1997). As summarized in Fig. 5A, GDPS significantly decreased the Ca²⁺ current inhibition produced by adenosine (from 33 ± 4%, n = 5 to 8 ± 1%, n = 12) and NE (from 47 ± 4%, n = 9 to 6 ± 1%, n = 11). In general, A₁ adenosine receptors are coupled to PTX-sensitive Go/i proteins (Fredholm et al., 1994; Ralevic and Burnstock, 1998). To identify the nature of G protein coupling between the adenosine receptors and Ca²⁺ channels, MPG neurons were incubated overnight in a medium containing PTX (500 ng/ml). The PTX treatment significantly attenuated the adenosine-induced Ca²⁺ current inhibition from 29 ± 2% (n = 9) to 9 ± 1% (n = 7). In control experiments, NE-induced inhibition was also reduced by PTX treatment from 40 ± 4% (n = 7) to 10 ± 1% (n = 11) (Fig. 5B). Taken together, these data suggest that adenosine-induced Ca²⁺ current inhibition is primarily mediated by PTX-sensitive Go/i proteins.

View larger version (15K): [in this window] [in a new window]   Fig. 5.   Involvement of G proteins in adenosine-induced Ca2+ current inhibition. A, summary of inhibition (%) of Ca2+ currents by adenosine in neurons dialyzed with GTP (0.3 mM, control) and GDPS (2 mM). The currents were elicited after 5-min dialysis with GDPS as in Fig. 1. B, summary of effects of PTX on adenosine-induced Ca2+ current inhibition. Neurons were incubated 14 to 18 h with 500 ng/ml PTX. As a positive control, NE (10 µM) was also tested. Data (acquired from 5 to 12 neurons) represent the mean ± S.E.M. ***p < 0.01.

Voltage-Dependence of Adenosine-Induced Ca²⁺ Current Inhibition. As determined from the I-V curves in Fig. 2B, the relationship between the adenosine-induced Ca²⁺ current inhibition and test potentials displayed a "bell-shaped" profile (Fig. 6A). The voltage-dependent inhibition of Ca²⁺ currents by adenosine was also demonstrated using a double pulse protocol consisting of two identical test pulses to +10 mV separated by a large depolarizing conditioning pulse to +80 mV (Fig. 6B, bottom). The time course of the adenosine-induced Ca²⁺ current inhibition is shown in Fig. 6B (top). The adenosine-induced Ca²⁺ current inhibition displayed the hallmarks of voltage-dependent inhibition, i.e., kinetic slowing and relief of current inhibition by the conditioning pulses (Elmslie et al. 1990) (Fig. 6B). Facilitation, defined as the ratio of the postpulse to prepulse current amplitude, increased from 1.09 to 1.55 after adenosine application.

View larger version (14K): [in this window] [in a new window]   Fig. 6.   Voltage-dependent inhibition of Ca2+ currents by adenosine. A, bell-shaped relationship between inhibition (%) and test potential. The profile was derived from the data shown in Fig. 2B (n = 16). B, time course of current inhibition (top) and superimposed current traces (bottom) in the absence and presence of 30 µM adenosine. The currents were elicited by a double pulse protocol consisting of two identical test pulses to +10 mV separated by a large depolarizing conditioning pulse to +80 mV. The facilitation was defined as the ratio of the postpulse () to prepulse () current amplitude (post/pre). Data represent as the means ± S.E.M.

Modulation of -Conotoxin GVIA-Sensitive N-Type Ca²⁺ Currents by Adenosine. As established previously (Zhu et al., 1995; Zhu and Yakel, 1997), -CgTx GVIA-sensitive N-type Ca²⁺ channels contribute to the majority (60 ± 4%; n = 9) of HVA Ca²⁺ channel currents in MPG neurons (Fig. 7B). Thus, we tested whether N-type Ca²⁺ channels were modulated by adenosine with a toxin occlusion experiment. Figure 7A illustrates the effects of 30 µM adenosine on the Ca²⁺ currents before and after application of 1 µM -CgTx GVIA. Adenosine-induced Ca²⁺ current inhibition was almost completely occluded by -CgTx GVIA. On average, adenosine-induced Ca²⁺ current inhibition was 33 ± 1% (n = 5) and 3 ± 1% (n = 5) before and after -CgTx GVIA, respectively. These results suggest that activation of adenosine receptors modulates mainly -CgTx GVIA-sensitive N-type Ca²⁺ channels in MPG neurons.

View larger version (21K): [in this window] [in a new window]   Fig. 7.   Effects of adenosine on -CgTx GVIA-sensitive Ca2+ currents. A, time course of the effects of the consecutive application of 30 µM adenosine (ADO), 1 µM -CgTx GVIA, and 10 µM nimodipine on Ca2+ currents. The currents were elicited and amplitudes were determined as in Fig. 1. B, contribution of N-type (-CgTx sensitive), L-type (nimodipine sensitive), and non-N/L-type (-CgTx and nimodipine-resistant, respectively) currents to the total Ca2+ current. C, summary of inhibition (%) of Ca2+ current by adenosine in the absence and presence of -CgTx GVIA. In B and C, data (acquired from 5 to 9 neurons) represent the mean ± S.E.M. ***p < 0.01.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Identical Modulation by Adenosine in Adrenergic and Cholinergic MPG Neurons. The present study describes identification of an adenosine receptor subtype and signal transduction pathway responsible for the adenosine-induced HVA Ca²⁺ current inhibition in rat MPG neurons. As described previously, the MPG contains distinct populations of adrenergic and cholinergic neurons providing motor inputs to pelvic effectors (Keast, 1999). Accordingly, it is possible that the expression level of a certain receptor and the resultant modulation of Ca²⁺ currents are phenotype-specific. For example, NE and NPY have been found to produce much larger Ca²⁺ current inhibition in adrenergic MPG neurons compared with cholinergic MPG neurons (Zhu et al., 1995; S. K. Cha, K. S. Park, and S. W. Jeong, unpublished observations). This phenotype-dependent differential modulation might be explained by the fact that both NE and NPY are colocalized in adrenergic nerve terminals in MPG neurons (Keast and de Groat, 1989; Keast, 1991; Keast, 1999). However, this is not the case for adenosine-induced Ca²⁺ current inhibition because the potency and efficacy of adenosine were almost identical in the two types of MPG neurons (Fig. 2C). This suggests that adenosine can act as an inhibitory modulator for both adrenergic and cholinergic MPG neurons (see below).

Identification of Adenosine Receptor Subtype Involved in Ca²⁺ Current Inhibition. Based on amino acid sequence and pharmacology, adenosine/P1 receptors are subdivided into four subtypes, A₁, A_2a, A_2b, and A₃ (Fredholm et al., 1994; Ralevic and Burnstock, 1998). A comprehensive study has revealed that all subtypes of adenosine receptor are expressed in rat brain tissues (Dixon et al., 1996). However, RT-PCR analysis delineated only A₁ and A_2a in rat MPG neurons, suggesting tissue-specific distribution of adenosine receptors. Several lines of evidence indicate that adenosine-induced Ca²⁺ current inhibition is mainly mediated via A₁ receptors in MPG neurons. First, CPA, an A₁-selective agonist, potently mimicked adenosine-induced Ca²⁺ current inhibition. In contrast, CGS 21680, an A₂-selective agonist was an order of magnitude less potent than CPA. The rank order of potency for adenosine and adenosine analogs (CPA > adenosine CGS 21680) in MPG neurons is reminiscent of that in superior cervical ganglion (SCG) neurons where activation of A₁ receptors inhibits Ca²⁺ currents (Zhu and Ikeda, 1993). Second, pretreatment of DPCPX, the most widely used A₁-selective antagonist, completely blocked the adenosine-induced Ca²⁺ current inhibition. Finally, the pharmacological data described above are consistent with the experiments with PTX. It is well established that A₁ receptors are coupled to PTX-sensitive Go/i, whereas A₂ receptors are coupled to cholera toxin-sensitive Gs (Ralevic and Burnstock, 1998). Recently, Jeong and Ikeda (2000) have demonstrated that A₁ receptors can couple to Gi2, GoA, and GoB to inhibit Ca²⁺ currents in rat SCG neurons. Indeed, blockade of adenosine-induced Ca²⁺ current inhibition by PTX supports the involvement of A₁ receptors. It should be noted, however, that A₁ receptors in supraoptic neurons are coupled to a PTX-insensitive G protein, probably Gz (a member of Gi family) (Noguchi and Yamashita, 2000; but see Jeong and Ikeda, 1998). Although cholera toxin-sensitive Gs has been shown to couple vasoactive inhibitory peptide receptors to Ca²⁺ current inhibition (Zhu and Ikeda, 1994), most studies have reported that activation of Gs-coupled A₂ receptors potentiates Ca²⁺ currents (mostly P-type) to facilitate synaptic transmission via a cyclic AMP/protein kinase A-dependent pathway (Mogul et al., 1993; Umemiya and Berger, 1994; Gubitz et al., 1996). Because there appear to be few, if any, P/Q-type Ca²⁺ channels in MPG neurons (Zhu et al., 1995; Zhu and Yakel, 1997), it is unlikely that stimulation of A₂ receptor modulates Ca²⁺ currents in MPG neurons.

Voltage-Dependent Inhibition of N-Type Ca²⁺ Currents by Adenosine. In most neurons, PTX-sensitive G proteins transduce the voltage-dependent and membrane-delimited inhibition of Ca²⁺ currents (Hille, 1994). Likewise, the voltage-dependence of adenosine-induced Ca²⁺ current inhibition was also evident by 1) the bell-shaped relationship between the current inhibition and test potentials, 2) slowing of the activation kinetics, and 3) greatly increased prepulse facilitation (Elmslie et al., 1990). The latter two characteristics can be explained by interconversion between "willing" and "reluctant" channels (Bean, 1989; Zhu and Ikeda, 1993) that requires direct binding of G subunits released from Go/i to Ca²⁺ channels (for review, see Ikeda and Dunlap, 1999). Ca²⁺ current inhibition mediated by A₁ receptors has also been described in other neurons, including those of dorsal root ganglia (Dolphin et al., 1986), hippocampus (Scholz and Miller, 1991; Mogul et al., 1993; Wu and Saggau, 1994), brain stem (Umemiya and Berger, 1994), spinal cord (Mynlieff and Beam, 1994), SCG (Zhu and Ikeda, 1993), and supraoptic nucleus (Noguchi and Yamashita, 2000). In all cases, N-type Ca²⁺ currents were inhibited by adenosine. Previous studies have shown that MPG neurons express at least three different HVA Ca²⁺ channels, i.e., N-, L-, and non-N/L-types (Zhu et al., 1995; Zhu and Yakel, 1997). Consistent with previous studies, the major target of A₁ receptor activation in this study was the -CgTx GVIA-sensitive N-type Ca²⁺ channels, which underlie the majority (60%) of whole-cell Ca²⁺ currents in MPG neurons (Fig. 7).

Functional Relevance of Adenosine-Induced Ca²⁺ Current Inhibition. It is well known that ATP can be released from both adrenergic and cholinergic nerve terminals in pelvic viscera (Fujii, 1988; Hoyle, 1992). Thus, ATP mediates neurally elicited contraction of effectors via P2 receptors, whereas adenosine, catabolized from ATP by ectonucleotidases at an extracellular side, exerts inhibitory effects on neuromuscular transmission via adenosine receptors located in presynaptic terminals (Theobald and de Groat, 1989). On the other hand, a study has proposed that adenosine is released by stimulation of preganglionic nerves and acts as an inhibitory neurotransmitter/modulator (Akasu et al., 1984). In addition, one can consider the soma of postganglionic neurons as a potential source of endogenous adenosine. Regardless of the source, on-going neuronal activity seems to influence the local concentration of adenosine, which should in turn provide an activity-dependent autoregulatory mechanism to control excessive excitation in the pelvic plexus. In this scenario, a strong stimulation of pre- and postganglionic nerves facilitates corelease of ATP with NE or acetylcholine. The subsequent breakdown of ATP by ectoenzymes results in a high concentration of adenosine at synapses that reduces the release of neurotransmitters from synaptic terminals by inhibiting Ca²⁺ currents via activation of adenosine receptors. Previously, adenosine has been shown to produce slow hyperpolarizing potentials in cat vesical parasympathetic ganglia by activating a K⁺ conductance (Akasu et al., 1984; de Groat and Booth, 1993b). Likewise, we observed that adenosine augments A-type K⁺ currents via activation of A₁ receptors, which may reduce neuronal firings in rat MPG (S. K. Cha and I. D. Kong, unpublished observations). In addition to the modulation of K⁺ conductance, thus, the present study introduces another ionic mechanism by which adenosine might play an inhibitory role in the rat pelvic plexus.