Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Previous studies have demonstrated that activation of prejunctional ₂-adrenoceptors reduces, while activation of prejunctional ₂-adrenoceptors facilitates, the nerve stimulation-evoked release of NE from sympathetic nerves at a variety of vascular and nonvascular smooth muscle neuroeffector junctions. Consequently, it has been proposed that the amount of NE available to act at the postjunctional receptors, and thus the amplitude of response of the target cells, is controlled not only by the degree of activation of sympathetic nerves but also by auto-feedback mechanisms that involve prejunctional ₂- and ₂-adrenoceptors (Langer, 1981; Misu and Kubo, 1986; Nedergaard and Abrahamsen, 1990).

It is now appreciated that the sympathetic nerves release other chemical mediators in addition to NE, e.g., the sympathetic nerves innervating the rodent vas deferens use ATP and NE as motor neurotransmitters (Fedan et al., 1981; Sneddon and Westfall, 1984; Burnstock, 1999). The question arises then as to whether prejunctional receptors that modify the release of NE do likewise to its cotransmitter ATP.

It appears that this is not the case with prejunctional ₂-adrenoceptors. For example it has been shown that activation of prejunctional ₂-adrenoceptors by endogenous NE causes a profound reduction of its own release but exerts little influence on the release of ATP (Driessen et al., 1993; Todorov et al., 1996). This observation and the evidence for temporal dissociation of neurogenic release of ATP and NE (Todorov et al., 1994, 1996) indicate that the cotransmitters are stored separately and released by different mechanisms rather than being stored and released from the same synaptic vesicle (Stjarne, 1989).

The literature is not clear in regard to regulation of cotransmitter release by prejunctional -adrenoceptors. Goncalves et al. (1996) and Driessen et al. (1996) have reported that activation of prejunctional ₂-adrenoceptors by exogenously applied agonists enhances the nerve stimulation-evoked release of [³H]NE but decreases the release of ATP, whereas, studies by Brock et al. (1997) with sympathetically innervated rat tail artery have demonstrated that the nerve-stimulated release of ATP (as measured by the amplitude of excitatory junction potentials) and the release of NE (as measured by the amplitude of oxidation currents of a carbon-fiber electrode) were increased equivalently by agonists of -adrenoceptors.

One of the difficulties of these types of studies is the variety of methods used to measure transmitter overflow, sometimes, in fact, using the postjunctional response to ATP as an index of release. We thought that some of the conflicting results might be reconciled if the effects of -adrenoceptor activation on the release of the endogenous cotransmitters were measured along with the postjunctional action of the cotransmitters under similar experimental conditions. Therefore, in this study we examine, in the guinea pig vas deferens, the effects of -adrenoceptor activation on the nerve stimulation-evoked release of ATP and NE under experimental conditions that allow quantification and direct comparison of the amplitude as well as time course of release of each cotransmitter (Todorov et al., 1994, 1996). We also tested the effects of activation of postjunctional -adrenoceptors on the contractile effects of endogenously released (neurogenic contractile response) and exogenously applied ATP. Using selective agonists and antagonists we determined the subtype of -adrenoceptors involved in modulation of the sympathetic cotransmission of the guinea pig vas deferens.

Together, the results from overflow and contraction experiments indicate that, in this tissue, there are ₂-adrenoceptors present both prejunctionally and postjunctionally. Activation of the prejunctional ₂-adrenoceptors causes an increase of the nerve stimulation-evoked release of ATP as well as of NE. The time courses of release of ATP and NE, however, remain dissociated. These data indicate that prejunctional ₂-adrenoceptors modulate equally well the two otherwise separate mechanisms for release of ATP and NE.

Activation of postjunctional ₂-adrenoceptors causes a concentration-dependent reduction of contractions evoked by exogenously applied, and endogenously released ATP but has no effect on the amplitude of contractions evoked by NE. These results suggest that the ₂-adrenoceptor-mediated inhibition of the postjunctional actions of neurotransmitter ATP prevails over the prejunctional facilitation of release of neurotransmitters. Thus, the functional outcome of activation of ₂-adrenoceptors in the guinea pig vas deferens is a reduction of sympathetic neuroeffector transmission.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Tissue Preparation. Male albino guinea pigs (350-400 g) were killed by decapitation. The vasa deferentia were removed from the body, cleaned of connective tissue, and the lumen exposed by section along the longitudinal axis.

Recording of Contractile Responses Evoked by Exogenously Applied ATP. The prostatic half of the vas deferens tissue preparation was fixed at one end to a holder and suspended in an organ chamber filled with 5 ml of Krebs' solution (37°C) of the following composition: 150 mM NaCl, 4.6 mM KCl, 1.2 mM MgCl₂, 2.5 mM CaCl₂, 24.8 mM NaHCO₃, 1.2 mM KH₂PO₄, and 5.6 mM glucose, which was constantly bubbled with 95% O₂ and 5% CO₂. The other end of the tissue was attached by silk surgical suture to a force-displacement transducer (Grass FTO3, Grass Instruments, Quincy, MA). The tissue preparation was allowed to equilibrate for 20 min before and 40 min after an initial tension of 0.5 g was applied. Bolus injections (50 µl) of a 10 mM solution of ATP were delivered at 20-min intervals and the contractile responses of the tissue were recorded on paper by the means of a Grass (RPS 7C 8B) polygraph. After each injection of ATP the tissue was washed three times with fresh Krebs' solution. At the beginning of each experiment three consecutive control contractile responses to ATP (100 µM) were obtained and their average amplitude was calculated. To test the effects of activation of postjunctional -adrenoceptors on the ATP-induced contractions of the guinea pig vas deferens -adrenoceptor agonists, in increasing concentrations, were injected in the organ bath (0.1 nM-100 µM) 5 min before injection of ATP. When used, adrenoceptor antagonists were added to the Krebs' solution and their concentration was maintained constant throughout the experiment. The amplitudes of contractile responses of the tissue preparation were calculated as percentages from the average amplitude of the three control responses and presented graphically (Prizm v 3; GraphPad Software, San Diego, CA).

Recording of Neurogenic Contractile Responses. In some experiments the prostatic portion of a vas deferens was placed in a horizontally oriented organ chamber equipped with two platinum ring electrodes for producing electrical field stimulation (EFS). One end of the tissue was fixed in place and the other attached by silk surgical suture to a Grass force-displacement transducer (FTO3) connected via a bridge amplifier to a PowerLab recording system (ADInstruments, Castle Hill, Australia). The tissue was superfused with oxygenated Krebs' solution (37°C) by the means of a Rabbit-Plus peristaltic pump (Rainin, Woburn, MA) at a rate of 2 ml/min. The tissue was allowed to equilibrate for 20 min before and 40 min after an initial tension of 0.5 g was applied. At the beginning of each experiment three trains (20 s) of electrical impulses (8 Hz, 0.1-ms duration) were delivered to the tissue preparation at 20-min intervals. The average amplitudes of the first, twitch-like phase and the second, tonic phase of the resulting contractile responses were calculated. Isoproterenol (0.01, 0.1, and 1.0 µM) was added to the superfusing Krebs' solution for 10 min prior to nerve stimulation. The contractile responses were recorded by the means of Chart v 4.01 software (ADInstruments) and the digital information stored on an Intel-equipped personal computer. The maximal amplitudes of the first and the second phase of the neurogenic contractions of each tissue were calculated as percentages of the average amplitudes of the corresponding control contractile responses and statistically evaluated (Prizm v 3; GraphPad Software).

Overflow Experiments. The prostatic halves of three tissues, each from a different animal, were loaded in a "Brandel" superfusion chamber with a volume of 200 µl. Whatman 541 filters were cut to fit both ends of the chamber. The chamber was then inserted vertically into a thermostatic block (36°C) and closed with platinum "screen" electrodes at each end. The tissues were superfused from bottom to top with Krebs' solution by the means of a Rabbit-Plus peristaltic pump (Rainin) at a rate of 2 ml/min. When used, drugs were added to the superfusing solution. The sympathetic nerves were stimulated once for 60 s by EFS at 8 Hz, pulse duration of 0.1 ms, and supramaximal voltage. Samples of superfusate (320 µl) were collected in ice-cold test tubes at 10-s intervals before, during and after sympathetic nerve stimulation. The sympathetic nerves of the guinea pig vas deferens release specific metabolic enzymes that rapidly degrade the neurotransmitter ATP to ADP, AMP, and ADO (Todorov et al., 1996, 1997; Mihaylova-Todorova et al., 2001). Therefore, the sum total of adenine nucleotides and ADO that are present in the superfusate collected during nerve stimulation of the tissue preparations approximates the amount of ATP that was released initially. For these experiments we routinely add desipramine (0.3 µM) to the superfusion solution to block the neuronal uptake of NE. Neuronal uptake is recognized as the most important mechanism involved in the clearance of NE from the synapse (Graefe and Bonisch, 1988). Thus, the sum total of the adenine nucleotides and ADO (referred to as "purines" in the figures) and NE, with uptake blocked, provides an approximation of the molecular ratios at which ATP and NE were coreleased from the sympathetic nerves.

HPLC Analysis of Endogenously Released ATP. The amount of ATP released during sympathetic nerve stimulation was estimated from the sum total of ATP and its products of degradation (ADP, AMP, and ADO) present in the superfusate (Todorov et al., 1996). Briefly, 200-µl aliquots taken from each 10-s collection of superfusate were acidified with 90 µl of citric phosphate buffer, pH 4, and incubated for 40 min at 80°C in a dry heating block in the presence of 10 µl of 2-chloroacetaldehyde. During the incubation, the adenine nucleotides and ADO present in the sample were transformed to their respective fluorescent derivatives 1N⁶-etheno ATP, 1N⁶-etheno ADP, 1N⁶-etheno AMP, and 1N⁶-etheno ADO. Two hundred microliters of the mixture, containing the etheno-adenine nucleotides and etheno-adenosine, were injected by the means of Waters 715 Ultra WISP sample processor (Waters, Milford, MA) onto a 4-µm, 8- × 10-mm Nova-Pak phenyl cartridge (Waters) connected to Waters 510 HPLC pumps. Bound nucleotides were eluted during a gradient change from buffer A (0.1 M KH₂PO₄, adjusted to pH 6.0 with NaOH) to buffer B (25% methanol in buffer A) according to Waters gradient profile 7 at flow rate of 2 ml/min. Fluorescence of 1N⁶-etheno ATP, 1N⁶-etheno ADP, 1N⁶-etheno AMP, and 1N⁶-etheno ADO, at retention times 4.5, 5.5, 8 and 12 min, respectively, was quantified with an excitation wavelength of 230 nm and an emission wavelength of 420 nm by Shimadzu RF 535 fluorescent monitor (Columbia, MD).

HPLC Analysis of Endogenously Released NE. To measure the overflow of NE, 90-µl aliquots from each 10-s collection of superfusate were acidified with 10 µl of 1 M perchloric acid and filtered through a 0.22-µm Cameo 3N syringe filter (Westborough, MA) into limited volume inserts (Waters) by centrifugation at 1000g for 1 min. The acidified sample (30 µl) was injected by Waters 715 Ultra WISP sample processor (Waters) onto a catecholamine HR-80 column (ESA, Chelmsford, MA), connected to a Waters 510 HPLC pump. Catecholamines were eluted under isocratic conditions using a mobile phase of 50 mM Na₂PO₄, 0.2 mM EDTA, 3 mM 1-heptanesulfonic acid, 3% v/v methanol in deionized water, pH 2.6, adjusted with phosphoric acid, at a flow rate of 1.5 ml/min. NE (retention time 2.2 min) was quantified using Coulochem II electrochemical detector (ESA).

Chemicals Used. The following chemicals were purchased from Sigma (St. Louis, MO): adenosine 5'-triphosphate (disodium salt), adenosine 5'-diphosphate (sodium salt), adenosine 5'-monophosphate (sodium salt), adenosine (hemisulfate salt), atenolol, chloroacetal, clenbuterol, isoproterenol, norepinephrine, phentolamine, prazosin, and propranolol. BRL 37 344 and ICI 118, 551 were purchased from Sigma/RBI (Natick, MA) and xamoterol hemifumarate from Tocris Cookson (Ballwin, MO). Methanol was purchased from B&J (Muskegon, MI) and 1-heptanesulfonic acid from Fisher Scientific (Pittsburgh, PA). 2-Chloroacetaldehyde was prepared in the laboratory, as described previously (Todorov et al., 1996).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effects of -Adrenoceptor Activation on Amplitude and Time Course of Nerve Stimulation-Evoked Overflow of Cotransmitters ATP and NE from Guinea Pig Vas Deferens. There were no detectable amounts of NE, ATP, ADP, and AMP but there was a small amount of ADO in samples collected before the onset of the EFS. Stimulation of the sympathetic nerves of the guinea pig vas deferens resulted in an overflow of ATP and its metabolites ADP, AMP, ADO (presented as their sum total in Fig. 1, A, C, E, and G) and in the overflow of NE (control in B, D, F, and H). Under control conditions, the EFS-evoked overflow of purines reached a peak by 20 s and then quickly declined to prestimulation levels even though the stimulation continued for 60 s. The overflow of NE on the other hand increased steadily until the end of the neurogenic stimulation. Superfusion of tissue preparations with Krebs' solution containing 1 µM isoproterenol for 5 min before the onset of EFS resulted in 2-fold increase of the nerve stimulation-evoked overflow of purines (Fig. 1A). The overflow of NE was also doubled in amplitude under these conditions (Fig. 1B). However, the dissociated temporal patterns of overflow of ATP and NE were not changed in the presence of isoproterenol. Treatment with clenbuterol (1 µM), a selective ₂-adrenoceptor agonist also produced a 2-fold increase in the amplitude of overflow of purines (Fig. 1E) and NE (Fig. 1F), while having no effect on the temporal pattern of release of cotransmitters. Two other -adrenoceptor agonists, the ₁-subtype-selective agonist xamoterol (Fig. 1, C and D), and the ₃-subtype-selective agonist BRL 37 344 (Fig. 1, G and H) had no effect on the amplitude or the time course of overflow of either cotransmitter.

View larger version (27K): [in this window] [in a new window]   Fig. 1.   Effects of -adrenoceptor agonists on the amplitude and time course of nerve stimulation-evoked overflow of ATP and NE. Superfused guinea pig vas deferens tissue preparations were stimulated with EFS at 8 Hz for 60 s and samples of the superfusing solutions were collected at 10-s intervals before (0), during (1-6), and after (7) the onset of EFS. When plotted against time the sum total of ATP, ADP, AMP, ADO (purines in A, C, E, and G) and NE (NE in B, D, F, and H) present in each of the samples reflects the time course of overflow of concomitantly released sympathetic cotransmitters ATP and NE. The data from control experiments (, n = 7) show that the overflow of purines occurs only at the beginning of EFS, while the overflow of NE is maintained throughout the stimulation period. Perfusion with isoproterenol (1 µM) leads to a 2-fold increase (statistically significant at P < 0.05) in amplitude of overflow of both purines (A) and NE (B) (n = 5). Virtually the same facilitatory effect (P < 0.05) is produced by the 2-adrenoceptor-selective agonist clenbuterol (1 µM) (E and F) (n = 4). Neither the 1-adrenoceptor-selective agonist xamoterol (1 µM) (C and D) (n = 4) nor the 3-adrenoceptor-selective agonist BRL 37 344 (1 µM) (G and H) (n = 4) has any effect on the EFS-evoked overflow of ATP and NE. Note that the time course of overflow of the two cotransmitters remains dissociated in the presence of any of the -adrenoceptor agonists used in this study. In all graphs the vertical bars represent ± S.E.M.

View larger version (30K): [in this window] [in a new window]   Fig. 2.   Effects of subtype-selective -adrenoceptor antagonists on the facilitation of nerve stimulation-evoked overflow of ATP and NE produced by isoproterenol and clenbuterol. The increase in amplitude of nerve stimulation-evoked overflow of purines and NE by isoproterenol (1 µM) (, A and B) (n = 5) or clenbuterol (1 µM) (, C and D) (n = 4) is not affected by the 1-adrenoceptor-selective antagonist atenolol (1 µM) () (n = 4). Pretreatment with the 2-adrenoceptor-selective antagonist ICI 118, 551 (1 µM) () (n = 5) prevents (P < 0.05) the facilitation of cotransmitter release by both isoproterenol and clenbuterol.

Effects of Isoproterenol on Neurogenic Contractile Response of Guinea Pig Vas Deferens. Stimulation of the sympathetic nerves of the guinea pig vas deferens with a train of electrical impulses delivered at frequency of 8 Hz results in biphasic contraction (Fig. 3A, control). The first phase of this neurogenic contraction is a twitch, which is mediated by neuronally released ATP (Sneddon and Westfall, 1984; Todorov et al., 1999). NE mediates the second tonic phase of the neurogenic contraction (Todorov et al., 1996). Addition of isoproterenol (0.01-1.0 µM) caused a concentration-dependent decrease of the amplitude of the purinergic twitch contraction (Fig. 3A, isoproterenol). In the presence of 0.01 µM isoproterenol the amplitude of the neurogenic twitch contraction was 81.6 ± 3.42% of the control (p = 0.012, n = 4). The amplitude of the neurogenic twitch contraction was further decreased to 58.7 ± 3.62 (p = 0.0015, n = 4) in the presence of 0.1 µM isoproterenol, and to 37.76 ± 6.04 (p = 0.0019, n = 4) in the presence of 1 µM isoproterenol (Fig. 3B). There was a slight but not statistically significant increase in amplitude of the second, adrenergic phase of neurogenic contraction of the guinea pig vas deferens in the presence of isoproterenol (Fig. 3B). When applied in higher concentrations (10 and 100 µM) isoproterenol produced a concentration-dependent increase of both purinergic and adrenergic phases of the neurogenic contractions of the guinea pig vas deferens (data not shown).

View larger version (23K): [in this window] [in a new window]   Fig. 3.   Effect of isoproterenol on the EFS-induced contractile responses of the guinea pig vas deferens. A, electrical field stimulation at 8 Hz for 20 s evokes a biphasic contraction of the guinea pig vas deferens consisting of a fast twitch-like phase and a delayed, tonic phase (control). Pretreatment with isoproterenol (0.01-1.0 µM) inhibits only the first, twitch-like phase of the neurogenic contraction. The statistically evaluated (n = 4) amplitudes of the first and the second phases are presented in B. Superfusion with isoproterenol in increasing concentrations causes a statistically significant (P < 0.05) decrease in amplitude of the first phase and statistically not significant increase of the second phase of the neurogenic contractions of the guinea pig vas deferens.

Effects of Isoproterenol on Contractile Response of Guinea Pig Vas Deferens Evoked by Exogenously Applied ATP. The -adrenoceptor agonists used in this study did not produce contractile responses of the in vitro guinea pig vas deferens (data not shown). However, the contractile responses evoked by bolus injections of ATP (100 µM) were progressively decreased in amplitude in the presence of low concentrations (0.0001-1 µM) of the nonselective -adrenoceptor agonist isoproterenol (Fig. 4A). A maximal 64.4 ± 4.1% inhibition of ATP-induced contractions was observed with isoproterenol in a concentration of 0.1 µM. In concentrations higher than 1 µM, isoproterenol produced a concentration-dependent facilitation of the ATP-induced contractions (Fig. 4B). In the presence of isoproterenol (100 µM) the amplitude of contractions evoked by ATP was increased by 359.6 ± 46.34%. Pretreatment of tissue preparations with the nonselective -adrenoceptor antagonist propranolol prevented the reduction of the ATP-induced contractions by low concentrations of isoproterenol (Fig. 4A). However, the facilitation of ATP-induced contractions by isoproterenol at higher concentrations (10, 50, and 100 µM) was not affected by propranolol (Fig. 4B). The opposite was found in experiments with the -adrenoceptor antagonists phentolamine and prazosin. Pretreatment with either -adrenoceptor antagonist abolished the facilitation of ATP-induced contractions by high concentrations of isoproterenol (Fig. 4, D and F), whereas the inhibitory effects of low concentrations of isoproterenol remained unaffected (Fig. 4, C and E).

View larger version (28K): [in this window] [in a new window]   Fig. 4.   Isoproterenol-evoked inhibition and facilitation of ATP-induced contractions of the guinea pig vas deferens: effects of propranolol, phentolamine, and prazosin. The amplitudes of contractions of the guinea pig vas deferens produced by ATP (100 µM) in the presence of increasing concentrations of isoproterenol (0.0001-100 µM) are presented as a percentage of the average amplitude of three control ATP-induced contractile responses (n = 8). Isoproterenol in low concentrations inhibits (P < 0.05), while in high concentrations (above 1 µM) increases (P < 0.05) the amplitude of the ATP-induced contractions (isoproterenol control). Pretreatment with the -adrenoceptor antagonist propranolol (1 µM) prevents the inhibition of ATP-induced contractions of the guinea pig vas deferens by low concentrations of isoproterenol (A) but has no effect on the facilitatory effects of isoproterenol in high concentrations (B) (n = 4). The -adrenoceptor antagonist phentolamine (n = 4) and the 1-subtype-selective antagonist prazosin (n = 4) (1 µM each) do not affect the isoproterenol-induced inhibition of ATP-induced contractions (C and E), while they abolish its facilitatory effects (D and F).

View larger version (26K): [in this window] [in a new window]   Fig. 5.   Effects of -adrenoceptor agonists on the amplitude of ATP-induced contractile responses of guinea pig vas deferens. The 2-adrenoceptor-agonist clenbuterol (C) (n = 5) produces inhibition (P < 0.05) of the ATP-induced contractile responses of the guinea pig vas deferens that closely resembles the inhibitory effects of isoproterenol in low concentrations. Unlike isoproterenol, clenbuterol in high concentrations produces inhibition instead of facilitation (D). Neither the 1-adrenoceptor agonist xamoterol (A and B) (n = 4) nor the 3-adrenoceptor agonist BRL 37 344 (E and F) (n = 4) has inhibitory (A and E) or facilitatory (B and F) effects on the ATP-induced contractions of the guinea pig vas deferens.

View larger version (27K): [in this window] [in a new window]   Fig. 6.   Effects of subtype-selective antagonists of -adrenoceptors on the inhibition of ATP-induced contractions of the guinea pig vas deferens produced by isoproterenol and clenbuterol. The 1-adrenoceptor antagonist atenolol (1 µM) has no effect on the inhibition (A) or facilitation (B) of ATP-induced contractions produced by isoproterenol (n = 6). The 2-adrenoceptor-antagonist ICI 118, 551 (1 µM) prevents the inhibitory effects of isoproterenol in low concentrations (C) but does not influence its facilitatory effects in high concentrations (D) (n = 4). The inhibitory effects of clenbuterol both in low (E) and high (F) concentrations are reversed in the presence of ICI 118, 551 (n = 5).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

It has been known for some time that activation of -adrenoceptors can lead to an increase in the amount of NE released upon sympathetic nerve stimulation (Adler-Graschinsky and Langer, 1975; Stjarne, 1975). It would seem logical therefore, that the response of the target cells would be increased under such circumstances (Misu and Kubo, 1986; Nedergaard and Abrahamsen, 1990). However, there are a number of examples where this does not occur. The excitatory junction potentials in the guinea pig vas deferens (Sjostrand, 1973) and rabbit mesenteric arteries (Kuriyama and Makita, 1984), two sympathetically innervated tissues, are reduced in the presence of the -adrenoceptor agonist isoproterenol. Isoproterenol is also known to decrease the nerve stimulation-induced phasic contraction of the guinea pig and rat vas deferens (Goncalves et al., 1996; Huang et al., 1998). Perhaps these discrepancies between -adrenoceptor activation of NE release and postjunctional reduction in response are due to the fact that cotransmitters are involved, e.g., the excitatory junction potentials and phasic component of neurogenic response in vas deferens and certain blood vessels are not mediated by NE but rather by ATP that is also released from the nerves (Sneddon and Burnstock, 1984; Sneddon and Westfall, 1984). This could mean that while the nerve stimulation-induced release of NE is increased by activation of -adrenoceptors, the release of ATP is decreased. However, reports in the literature to date are inconsistent. For example, Goncalves et al. (1996) and Driessen et al. (1996) have reported that the nerve stimulation-evoked release of ATP from the guinea pig vas deferens is decreased by activation of -adrenoceptors, while Brock et al. (1997) found that activation of -adrenoceptors in the rat tail artery increases ATP release.

This is indeed a complicated situation. There are at least two transmitters involved and furthermore, the release of ATP and NE may be differentially modulated (Todorov et al., 1996, 1999). Another complication is that -adrenoceptors may influence the final neurogenic response not only by influencing release of the transmitters but by influencing the action of the transmitters postjunctionally as well. It is in this context that we carried out the experiments reported here.

Our results showed that both release of NE and release of ATP were increased in the presence of isoproterenol. Furthermore, this effect of the nonselective agonist isoproterenol is apparently mediated by ₂-adrenoceptors. Clenbuterol, a ₂-adrenoceptor-selective agonist, produced effects that were similar to isoproterenol, whereas xamoterol, a ₁-adrenoceptor-selective agonist and BRL 37 344, a ₃-adrenoceptor-selective agonist did not. The actions of isoproterenol and clenbuterol were antagonized by ICI 118, 551, a ₂-adrenoceptor-selective antagonist, but not by atenolol, a ₁-adrenoceptor-selective antagonist. These findings show that prejunctional ₂-adrenoceptors facilitate in parallel the release of ATP and NE not only in the rat-tail artery (Brock et al., 1997) but also in the guinea pig vas deferens.

Parallel modulation of release of cotransmitters would be the expected result if two neurotransmitters were stored together in the same synaptic vesicle and therefore released simultaneously. However, the fact that stimulation of prejunctional ₂-adrenoceptors increases the amplitude of release of the sympathetic cotransmitters ATP and NE to a similar degree does not necessarily indicate that these cotransmitters are stored and released together as suggested by Brock et al. (1997). It is equally possible that activation of facilitatory ₂-adrenoceptors coupled to each of otherwise independent mechanisms for release of ATP and NE may very well yield a parallel increase of the amplitude of their release.

Our experimental approach allows us to compare not only the amplitude but also the time course of nerve stimulation-evoked release of ATP and NE. The results from overflow experiments (Figs. 1 and 2) demonstrate that purines are released for a short time only at the beginning of nerve stimulation, while the release of NE continues until the end of the stimulation period. As discussed previously elsewhere (Todorov et al., 1996) this dissociated temporal pattern of release of ATP and NE is not consistent with the idea that the two cotransmitters originate from the same synaptic vesicle. It supports the alternative hypothesis that the sympathetic nerves store separately ATP and NE and release each cotransmitter via independent mechanism. The fact that activation of ₂-adrenoceptors increases the amplitude of release of ATP and NE to a similar degree, but does not affect their time courses, which remain vastly dissociated, suggests that prejunctional ₂-adrenoceptors modulate equally well the mechanism for release of ATP and the mechanism for release of NE.

In spite of the fact that isoproterenol increases the nerve stimulation-evoked release of both cotransmitters, we found, as have others (Goncalves et al., 1996; Huang et al., 1998), that the neurogenic contractile response of the tissue is decreased in the presence of isoproterenol. This reduction in the neurogenic response is obviously independent of an effect on transmitter release, since transmitter release was increased, not decreased.

To determine whether the reduction in the size of the neurogenic contraction by isoproterenol was due to a postjunctional action, we examined the effect of isoproterenol on contractions produced by exogenously administered ATP. We concentrated on ATP because it was the purinergic component of the neurogenic contraction that was preferentially affected (Fig. 3). As shown in Figs. 4 to 6, isoproterenol in concentrations of 0.001 to 1 µM decreased the size of the ATP-induced contractions with a maximal reduction of approximately 70%. At higher concentrations (>10 µM) isoproterenol actually increased the size of the ATP-induced contraction. The potentiation of the size of contractions produced by ATP is an odd finding in that it occurred with isoproterenol but not with other -adrenoceptor agonists. The fact that this effect was blocked by the -adrenoceptor antagonists phentolamine and prazosin but not by the -adrenoceptor antagonists propranolol, atenolol, and ICI 118, 551 suggests that isoproterenol-evoked facilitation of ATP-induced contractions is mediated by -adrenoceptors. While isoproterenol is generally regarded as a "pure" -adrenoceptor agonist, it has been known for a number of years that at high concentrations isoproterenol can stimulate -receptors (Furchgott, 1967). This ability of isoproterenol to activate not only -adrenoceptors but also -adrenoceptors may account for the failure of earlier studies (Goncalves et al., 1996; Huang et al., 1998) to detect the -adrenoceptor-mediated inhibition of the ATP-induced contractile responses in rodent vas deferens.

The postjunctional inhibition of the ATP-induced contractile response appears to be also mediated via ₂-adrenoceptors. Clenbuterol, a ₂-adrenoceptor-selective agonist, mimics the inhibitory effects of isoproterenol, whereas xamoterol and BRL 37 344, selective ₁- and ₃-adrenoceptor agonists, respectively, do not. Furthermore, the inhibition of the ATP-induced contractions by isoproterenol and clenbuterol is reversed by ICI 118, 551, a ₂-adrenoceptor-selective antagonist but not by the ₁-adrenoceptor-selective antagonist atenolol.

The endogenous agonist and the physiological role of prejunctional ₂-adrenoceptors that increase neurotransmitter release from the sympathetic nerves remain elusive. It appears that in virtually all of the tissues studied, neuronally released NE does not activate prejunctional -adrenoceptors (Stjarne and Brundin, 1976; May et al., 1986; Kazanietz and Enero, 1989; Apparsundaram and Eikenburg, 1995). In this study ICI 118, 551, a potent inhibitor of ₂-adrenoceptors also failed to decrease the release of ATP or the release of NE evoked by electrical nerve stimulation, suggesting that facilitatory ₂-adrenoceptors are not activated by endogenously released NE. Moreover, it has been suggested that feedback inhibition mediated via prejunctional ₂-adrenoceptors would buffer the -adrenoceptor-mediated increase in neurotransmitter release (Majewski and Rand, 1981; Costa and Majewski, 1988; Kahan et al., 1988). In addition, the evidence provided in the current study demonstrates that the ₂-adrenoceptor-mediated postjunctional inhibition would override the effect of facilitation of neurotransmitter release, which is also mediated via ₂-adrenoceptors, i.e., the neurogenic response is decreased. It seems, therefore, that even though present prejunctionally, the ₂-adrenoceptors are not involved in a functional feedback regulation of release of cotransmitters from the sympathetic nerves of the guinea pig vas deferens. Our results lead us to conclude that, even if there were situations where an endogenous agonist (i.e., epinephrine) might selectively activate ₂-adrenoceptors, the result would be an inhibition rather than facilitation of neuroeffector transmission.