Address correspondence and reprint requests to Elizabeth D. Abercrombie, PhD, 197 University Avenue, Newark, NJ 07102, USA. E-mail: email@example.com
Nigrostriatal dopaminergic neurons release dopamine from dendrites in substantia nigra and axon terminals in striatum. The cellular mechanisms for somatodendritic and axonal dopamine release are similar, but somatodendritic and nerve terminal dopamine release may not always occur in parallel. The current studies used in vivo microdialysis to simultaneously measure changes in dendritic and nerve terminal dopamine efflux in substantia nigra and ipsilateral striatum respectively, following intranigral application of various drugs by reverse dialysis through the nigral probe. The serotonin releasers (+/–)-fenfluramine (100 µm) and (+)-fenfluramine (100 µm) significantly increased dendritic dopamine efflux without affecting extracellular dopamine in striatum. The non-selective serotonin receptor agonist 1-(m-chlorophenyl)-piperazine (100 µm) elicited a similar pattern of dopamine release in substantia nigra and striatum. NMDA (33 µm) produced an increase in nigral dopamine of a similar magnitude to mCPP or either fenfluramine drug. However, NMDA also induced a concurrent increase in striatal dopamine. The D2 agonist quinpirole (100 µm) had a parallel inhibitory effect on dopamine release from dendritic and terminal sites as well. Taken together, these data suggest that serotonergic afferents to substantia nigra may evoke dendritic dopamine release through a mechanism that is uncoupled from the impulse-dependent control of nerve terminal dopamine release.
The dependence of nerve terminal neurotransmitter release on impulse activity has been well established (for a review see Cooper et al. 1991). Upon reaching threshold, action potentials generated at the axon hillock propagate down the axon to the terminal region and stimulate neurotransmitter release by activating voltage-sensitive calcium channels. In DA neurons, action potentials can actively propagate back into the dendrites with only modest attenuation owing to the existence of sodium channels within the dendrites (Hausser et al. 1995). Somatodendritic DA release is sensitive to activation of voltage-gated calcium channels (Elverfors et al. 1997; Bergquist et al. 1998), so propagation of sodium-dependent action potentials within the dendrites may provide a mechanism for influx of calcium necessary to evoke somatodendritic DA efflux. In this scenario, somatodendritic and nerve terminal DA release will co-vary in an impulse-dependent manner (Westerink et al. 1992a,b)
Alternatively, several lines of evidence indicate that the unique electrophysiological properties of DA neurons may make possible the uncoupling of local DA release within the dendrites and DA efflux from axon terminals. DA neuron action potentials are comprised of an initial segment spike and a distinctive somatodendritic spike (Grace and Bunney 1983a,b; Kita et al. 1986; Grace and Onn 1989; Grace 1990). This may be caused, in part, by the emergence of the axon from a primary dendrite a significant distance from the soma (Grace and Bunney 1983b; Tepper et al. 1987; Hausser et al. 1995; Richards et al. 1997). Based on this anatomical arrangement, the somatodendritic component potentially could be generated independently from the initial segment spike in an isolated ‘compartment’ within the dendrites (Grace 1990). The somatodendritic spike is pharmacologically distinct from the initial segment spike in that it is mediated by a calcium current (Grace 1990). DA neurons display low-threshold and high-threshold calcium spikes that can be activated independently of fast sodium channels and are expressed within the dendrites (Llinas et al. 1984; Harris et al. 1989; Hounsgaard et al. 1992; Nedergaard and Greenfield 1992). Entry of calcium into the dendrites occuring independently of sodium spikes may therefore provide an additional mechanism for somatodendritic DA release such that action potential generation may not be compulsory for somatodendritic DA release under all conditions. In this way, modulatory afferent inputs may conceivably activate dendritic DA release in the absence of back-propagating sodium-dependent action by eliciting dendritic calcium spikes through activation of dendritically localized conductances. For instance, tetrahydrocannabinol stimulates local dendritic DA release within the ventral tegmental area without affecting terminal release in nucleus accumbens (Chen et al. 1993). Moreover, stimulation of dorsal raphe nucleus attenuates antidromic activation of DA neurons without reducing the probability of neuronal firing (Trent and Tepper 1991). These authors propose that serotonin (5-HT)-evoked dendritic DA release occurs locally in SN and inhibits neighboring dendritic regions through inhibitory autoreceptor mechanisms, without affecting overall neuronal excitability. 5-HT application has been shown to evoke somatodendritic DA release from nigral slices and 5-HT selectively activates a tetrodotoxin-insensitive calcium conductance in DA neurons (Williams and Davies 1983; Nedergaard et al. 1988). Taken together, these data suggest that certain afferent inputs, including 5-HT, may evoke dendritic DA efflux independent of changes in release of DA from axon terminals in striatum.
To evaluate the extent to which 5-HT input to nigral DA neurons might selectively evoke local dendritic DA efflux in a manner uncoupled from nerve terminal DA in striatum, microdialysis probes were implanted in the SN and ipsilateral striatum of conscious, freely moving animals. Various pharmacological agents were applied by reverse dialysis through the nigral probe in order to compare the relative changes in extracellular DA efflux simultaneously measured in both SN and striatum.
Sodium pentobarbital, (+/–)-fenfluramine hydrochloride, (+)-fenfluramine hydrochloride, NMDA, mCPP and quinpirole were purchased from Sigma (St Louis, MO, USA). All other reagents and chemicals were of the highest purity commercially available (Fisher Scientific, Springfield, NJ, USA).
Adult male Sprague–Dawley rats (Zivic-Miller Laboratories, Pittsburgh, PA, USA) were used in all experiments. Before the experiment, animals were housed individually in plastic shoebox cages with free access to food and water. During the microdialysis experiments, animals were housed in plexiglass cylinders (Abercrombie and Finlay 1991), also with food and water ad libitum. Animals were kept under conditions of constant temperature (21°C) and humidity (40%), and maintained on a 12-h light/dark cycle. Rats weighed between 275 and 325 g at the time of each microdialysis experiment. All efforts were made to minimize animal suffering and limit the number of animals utilized for these experiments. Animal procedures were conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals (NIH Publications no. 80–23 revised 1996.).
Microdialysis probe implantation
Microdialysis probes were of the vertical concentric design, as described in detail previously (Abercrombie and Finlay 1991; Cobb and Abercrombie 2002). Probes implanted into SN and striatum had an active area of 1.5 and 2 mm respectively. They were perfused continuously with artificial CSF (147 mm NaCl, 2.5 mm KCl, 1.3 mm CaCl2, 0.9 mm MgCl2, pH ∼ 7.4) at a rate of 1.5 µL/min with a syringe pump (Harvard Apparatus, Holliston, MA, USA). Animals were anesthetized with sodium pentobarbital (60 mg/kg, i.p.) and mounted into a stereotaxic device (David Kopf Instruments, Tujunga, CA, USA). For implantation into SN, probes were set at a 30° lateral angle and lowered into SN at the following coordinates: AP − 5.2 mm, ML ± 5.6 mm relative to bregma, and DV − 8.0 mm below dura (Paxinos and Watson 1986). The probe assembly was fixed to the skull with fast-curing dental cement and two set screws. A second microdialysis probe was implanted into the ipsilateral striatum at the following coordinates: AP + 0.5 mm, ML ± 2.5 mm relative to bregma, and DV − 6.5 mm below dura (Paxinos and Watson 1986). The probe assembly was fixed to the skull with fast-curing dental cement and an additional set screw. Experiments were conducted at least 18 h after probe implantation under conscious conditions.
Analysis of dialysate
Dialysis samples (20 µL) were collected from both SN and striatum every 15 min throughout the experiment. Subsequently, each sample was analyzed for DA content by HPLC coupled with electrochemical detection. In brief, each system utilized a Velosep RP-18 column (100 × 3.2 mm; Applied Biosystems, Inc., Foster City, CA, USA) and the mobile phase was composed of 0.1 m sodium acetate buffer, pH 4.1, 0.1 mm EDTA, 1.2 mm sodium octyl sulfate and 8.5% (v/v) methanol. An electrochemical detector (Waters Model 460; Millipore Corporation, Bedford, MA, USA) with an amperometric electrode set at an applied potential of + 0.6 V was used. A solvent delivery pump (Model LC-10AD; Shimadzu Corporation, Columbia, MD, USA) delivered the mobile phase at a flow rate of 0.7 mL/min. Each system was calibrated daily with 20 µL of 10 nm standard solution in 0.1 m perchloric acid. Retention time was used to identify DA, which was quantified based on peak height. Samples were analyzed on two separate HPLC systems; the limit of detection for DA in these analyses in SN and striatum was ∼ 0.3 and 2.0 pg respectively.
For each experiment, a syringe containing drug solution was prepared by dissolving (+/–)-fenfluramine (100 µm), (+)-fenfluramine (100 µm), 1-(m-chlorophenyl)-piperazine (mCPP), (100 µm), NMDA (33 µm) or quinpirole (100 µm) in artificial CSF to the desired concentration for delivery into SN. The concentrations of (+/–)-fenfluramine, (+)-fenfluramine and mCPP were determined empirically in pilot studies based on the ability to elicit increases of similar magnitude in dendritic DA efflux. After three successive stable baseline samples had been collected from each probe (variability < 10%), the syringe perfusing the probe in SN was switched from one containing artificial CSF only to a syringe containing drug solution. After allowing 15 min for the flow rate to normalize, dialysis sample collection resumed. Extracellular DA content in SN and striatum was analyzed for an additional 3 h while drug was continually delivered through the nigral probe. The time delay before the drug effects were observed can be attributed to the amount of time required for the drug to travel through the tubing of the inlet of the probe to the active area.
Upon completion of each experiment, a lethal dose of sodium pentobarbital was administered (120 mg/kg, i.p.) and animals were perfused transcardially with 10% buffered formalin. The brain was then removed and stored in formalin for subsequent histological processing. Coronal slices of 50 µm thickness were stained with cresyl violet to verify placement of microdialysis probes in SN or striatum. Only data from animals with probe placement within the pars reticulata of SN (Cobb and Abercrombie 2002) and striatum were included in the data analysis.
Data are presented as mean ± SEM picograms of DA per 20 microlitres of microdialysis sample. To examine the effect of drug treatments on extracellular DA in SN and striatum, within-group effects were analyzed using a one-way anova with repeated measures over time coupled with Dunn's post hoc test (p < 0.05).
Effect of (+/–)-fenfluramine application in SN on extracellular DA in SN and striatum
Reverse dialysis application of (+/–)-fenfluramine (100 µm) through the nigral probe significantly increased DA measured in SN. Nigral (+/–)-fenfluramine application elicited a rise in extracellular DA measured in SN from 1.1 ± 0.2 pg/sample immediately before drug application to an overall absolute peak level of 2.4 ± 0.4 pg/sample in the presence of drug perfusion (F14,84 = 8.7; p < 0.01, n = 7; Fig. 1a). The increase was observed in the third post-drug sample and remained for the duration of the experiment.
Reverse dialysis application of (+/–)-fenfluramine (100 µm) through the nigral probe had no effect on DA efflux in striatum. No significant increase or decrease in DA was observed while (+/–)-fenfluramine was applied within SN compared with the value immediately before drug application (F14,84 = 0.5; p = 0.94, n = 7; Fig. 1b).
Effect of application of (+)-fenfluramine in SN on extracellular DA in SN and striatum
Reverse dialysis application of the more active enantiomer (+)-fenfluramine (100 µm) through the nigral probe significantly increased extracellular DA measured in SN from 1.0 ± 0.1 pg/sample immediately before drug application to a peak level of 2.4 ± 0.3 pg/sample during drug administration (F14,56 = 8.7; p < 0.01, n = 5; Fig. 2a). The increase was observed in the third post-drug sample and remained present for the duration of the experiment.
Reverse dialysis application of (+)-fenfluramine (100 µm) through the nigral probe had no effect on spontaneous DA release in striatum (F14,48 = 1.0; p = 0.42, n = 5; Fig. 2b).
Effect of mCPP application in SN on extracellular DA in SN and striatum
Reverse dialysis application of the non-selective 5HT1B and 5-HT2B/2C serotonin receptor agonist mCPP (100 µm) through the nigral probe significantly increased DA measured in SN. Nigral mCPP application increased extracellular DA measured in SN from 1.0 ± 0.1 pg/sample immediately before drug application to a peak level of 2.4 ± 0.3 pg/sample during drug administration (F14,56 = 11.4; p < 0.01, n = 4; Fig. 3a). The increase was observed in the third post-drug sample and remained for the duration of the experiment.
Reverse dialysis application of mCPP (100 µm) through the nigral probe had no effect on spontaneous DA release in striatum (F14,42 = 1.1; p = 0.41, n = 4; Fig. 3b).
Effect of NMDA application in SN on extracellular DA in SN and striatum
Local application of NMDA (33 µm) by reverse dialysis through the nigral probe significantly increased DA measured in SN from a pre-drug value of 0.7 ± 0.1 pg/sample to a maximal level of 2.6 ± 0.5 pg/sample during drug administration (F14,84 = 20.3; p < 0.01, n = 7; Fig. 4a). The increase was observed in the third post-drug sample and remained for the duration of the experiment.
A significant increase in DA measured in striatum was also observed during NMDA (33 µm) application into SN, from 8.3 ± 2.6 pg/sample immediately before drug application to a maximal level of 10.5 ± 2.6 pg/sample during reverse dialysis administration of NMDA into SN (F14,84 = 16.8; p < 0.01, n = 7; Fig. 4b).
NMDA also produced a dose-dependent response as greater concentrations of NMDA produced more robust increases in nigral DA release. Intranigral NMDA (50 µm) increased nigral DA efflux from a pre-drug value of 1.0 ± 0.1 pg/sample to a maximal value of 4.8 ± 1.4 pg/sample (n = 4; data not shown) and NMDA at a concentration of 100 µm increased extracellular DA in SN from 1.0 ± 0.3 pg/sample to a peak value of 6.7 ± 0.8 pg/sample (n = 2; data not shown).
Effect of quinpirole application in SN on extracellular DA in SN and striatum
Reverse dialysis application of the D2 DA receptor agonist quinpirole (100 µm) through the nigral probe significantly decreased extracellular DA measured in SN from a pre-drug value of 1.0 ± 0.1 pg/sample to a minimum level of 0.5 ± 0.1 pg/sample during drug administration (F14,70 = 10.9; p < 0.01, n = 6; Fig. 5a).
A significant decrease in striatal DA was also observed during quinpirole (100 µm) application into SN. Local quinpirole administration decreased striatal DA from 9.2 ± 1.0 pg/sample immediately before drug application to a minimum level of 5.8 ± 0.3 pg/sample during reverse dialysis administration in SN (F14,70 = 7.3; p < 0.01, n = 6; Fig. 5b).
In the current set of experiments, we examined the effects of serotonergic, glutamatergic and dopaminergic inputs to SN on DA efflux in pars reticulata of SN and in striatum. To achieve this, animals were implanted with microdialysis probes in SN and the ipsilateral striatum. Drugs were delivered via reverse dialysis through the nigral probe. The release of DA in striatum is fundamentally dependent upon action potential generation in the axon hillock of the midbrain DA neurons (Kuhr et al. 1987; Keefe et al. 1992), so release of DA in striatum provides a relative index for the level of nerve impulse activity. The concurrent comparison of nigral and striatal DA efflux in these experiments provides additional information relevant to consideration of the relationship between nerve impulse activity and somatodendritic DA release.
We observed that nigral application of (+/–)-fenfluramine via the microdialysis probe significantly increased dendritic DA efflux in SN without affecting release from axon terminals in striatum (Fig. 1). Fenfluramine increases extracellular 5-HT by blocking 5-HT uptake and evoking 5-HT release from 5-HT terminals (Gobbi et al. 1992; Crespi et al. 1997). SN is richly innervated by serotonergic terminals arising from the dorsal raphe nucleus (van der Kooy and Hattori 1980; Wirtshafter et al. 1987; Lavoie and Parent 1990; Corvaja et al. 1993), so application of (+/–)-fenfluramine via the nigral probe was used to increase serotonergic tone within SN. The enantiomer (+/–)-fenfluramine may affect DA outflow directly at DA uptake sites (Bettini et al. 1987). Therefore we also administered the more active enantiomer (+)-fenfluramine alone; a similar pattern of DA release in SN and striatum as that induced by the mixed enantiomer (+/–)-fenfluramine was obtained (Fig. 2). In an additional group of animals, application of the non-selective 5-HT receptor agonist mCPP evoked an equivalent increase in nigral DA efflux in the absence of any effect on extracellular DA in striatum (Fig. 3). Intravenous administration of mCPP has been shown to have a minimal inhibitory effect on DA neuron firing rate (Di Giovanni et al. 2000). In this same study, DA release in striatum was unaltered following intraperitoneal application of mCPP as measured by microdialysis. These findings, together with the present data, indicate that stimulation of nigral 5-HT receptors can evoke local somatodendritic DA release selectively without affecting the release of DA from axon terminals in striatum.
5-HT has a weak inhibitory influence over DA neuron firing rate (Dray et al. 1976; Kelland et al. 1990, 1993), yet our data reveal that 5-HT has a facilitory effect on dendritic DA. Trent and Tepper (1991) hypothesized that the apparent inhibitory effect of dorsal raphe stimulation on DA neuron dendritic excitability that they observed in the absence of an effect on overall neuronal excitability was the result of 5-HT-evoked somatodendritic DA release and a subsequent increase in the binding of D2 inhibitory autoreceptors localized to the soma and dendrites (Groves et al. 1975; Paden et al. 1976; Trent and Tepper 1991; Pucak and Grace 1994). 5-HT has been shown to have excitatory effects on DA neurons in vitro, including activation of a high-threshold calcium spike in the dendritic region of DA neurons (Nedergaard et al. 1988, 1991). Moreover, exogenous 5-HT application stimulates DA release from nigral slices (Williams and Davies 1983). Whereas DA release from axon terminals relies exclusively upon action potential generation for initiation, the results from the current study indicate that additional mechanisms may have the ability to evoke somatodendritic DA release such that dendritic DA efflux may not be coupled to neuronal impulse activity under all conditions. As a result, afferent input may excite local dendritic regions and stimulate somatodendritic DA release without changing the overall firing rate or firing pattern of the neuron. In this scenario, select excitatory inputs that are incapable of generating action potentials may powerfully regulate pars reticulata output neurons of SN and the output of the basal ganglia via stimulation of local somatodendritic DA release in SN (Cameron and Williams 1993; Timmerman and Abercrombie 1996; Abercrombie and DeBoer 1997; Radnikow and Misgeld 1998).
Our data reveal that stimulation of nigral 5-HT receptors evokes somatodendritic DA release, but it is unclear if this is a direct effect through activation of 5-HT receptors localized to DA neurons or whether this action is mediated through stimulation of 5-HT receptors localized to afferent terminals or non-dopaminergic neurons in SN. SN contains among the densest serotonergic innervation in the brain and serotonergic terminals directly contact TH-positive dendrites predominantly in SN pars reticulata but also within SN pars compacta (Nedergaard et al. 1988; Corvaja et al. 1993; Moukhles et al. 1997). Both 5-HT1 and 5-HT2 serotonin receptor classes are expressed within SN (Pazos and Palacios 1985; Waeber et al. 1989; Pompeiano et al. 1994; Wright et al. 1995; Eberle-Wang et al. 1997; Di Matteo et al. 2001; Varnas et al. 2001), but the specific profile of 5-HT receptor expression by nigrostriatal DA neurons compared with afferent terminals or other neurons in SN has yet to be defined fully. Therefore, serotonergic interactions within SN will require further investigation to identify the exact mechanisms of serotonergic action in SN. An indirect effect of 5-HT on NMDA-mediated stimulation of dendritic DA efflux in SN appears unlikely based on the present data because nigral application of NMDA was associated with increased DA output in striatum, whereas the pharmacological 5-HT manipulations were devoid of this action.
It is possible that reverse dialysis application of either mCPP or fenfluramine increases impulse activity in nigrostriatal DA neurons, but the net effect is too small for the temporal and spatial resolution of the microdialysis probe within the striatum to allow for detectable changes. To investigate this possibility further we applied NMDA or quinpirole via reverse dialysis and measured extracellular DA in SN and striatum. NMDA application elicits excitatory post-synaptic currents in DA neurons, increases DA neuron impulse activity and induces burst firing (Mereu et al. 1991; Overton and Clark 1992; Zhang et al. 1994; Christoffersen and Meltzer 1995; Meltzer et al. 1997). Moreover, NMDA applied directly into SN has also been shown to evoke somatodendritic and terminal DA release in the presence of nomifensine (Westerink et al. 1992b). Conversely, nigral application of quinpirole suppresses DA neuron discharge rate via activation of inhibitory autoreceptors in SN and decreases DA efflux in SN and striatum (Santiago and Westerink 1991; Westerink et al. 1992a; Shi et al. 1997). In the current studies, we replicated these findings in the absence of DA uptake inhibition. Importantly, we found that the increase in extracellular DA in SN produced by nigral application of NMDA was of similar amplitude to the increase observed following administration of mCPP or fenfluramine (Fig. 4). However, the increase in nigral DA efflux evoked by intranigral NMDA administration induced a simultaneous increase in extracellular DA measured within striatum, which was absent during nigral application of mCPP or fenfluramine despite the equivalent effects of these compounds on nigral DA efflux. Consequently, our data suggest that 5-HT or glutamate inputs to SN may differentially regulate DA neuron physiology, such that an increase of similar magnitude of somatodendritic DA efflux may or may not be coupled with a concurrent increase in DA release from nerve terminals in striatum. Higher concentrations of NMDA (50 µm, n = 4; 100 µm, n = 2) evoked larger increases in extracellular DA in SN and striatum (data not shown) so that this disparity cannot be explained as a ceiling effect. Furthermore, quinpirole mediated a parallel inhibitory effect on DA efflux in SN and striatum (Fig. 5).
Taken together, the present data have demonstrated that, under conditions whereby stimulation of nigral NMDA receptors and stimulation of nigral 5-HT receptors have equivalent effects in stimulating dendritic DA efflux, differential effects on nerve terminal DA release in striatum are observed. Based on these observations, it is suggested that action potential generation at the axon hillock of the DA neuron is not compulsory for initiation of somatodendritic DA release under all conditions, and that somatodendritic and nerve terminal DA release therefore does not necessarily parallel under all conditions. Interestingly, a TTX-insensitive mechanism for the initiation of dendritic DA release by 5-HT is predicted by this model (Cheramy et al. 1981). The physiological condition(s) under which such a mechanism might contribute to the regulation of basal ganglia function remain to be determined.
The authors would like to thank James Zackheim for his contribution to the manuscript and Mary Russo for her technical assistance. This research was supported by United States Public Health Service grants NS 19608 and MH 12968.