|Home | About | Journals | Submit | Contact Us | Français|
The role and contribution of the dense noradrenergic innervation in the ventral bed nucleus of the stria terminalis (vBNST) and anteroventral thalamic nucleus (AV) to biological function and animal behaviors is poorly understood due to the small size of these nuclei. The aim of this study was to compare norepinephrine release and uptake in the vBNST with that in the AV of anesthetized rats. Measurements were made in vivo with fast-scan cyclic voltammetry following electrical stimulation of noradrenergic projection pathways, either the dorsal noradrenergic bundle (DNB) or the ventral noradrenergic bundle (VNB). The substance detected was identified as norepinephrine based upon voltammetric, anatomical, neurochemical, and pharmacological evidence. Fast-scan cyclic voltammetry enables the selective monitoring of local norepinephrine overflow in the vBNST evoked by the stimulation of either the DNB or VNB while norepinephrine in the AV was only evoked by DNB stimulation. The α2-adrenoceptor antagonist, yohimbine, and the norepinephrine uptake inhibitor, desipramine, increased norepinephrine overflow and slowed its disappearance in both regions. However, control of extracellular norepinephrine by both autoreceptors and uptake was greater in the AV. The greater control exerted by autoreceptors and uptake in the AV resulted in reduced extracellular concentration compared to the vBNST when large numbers of stimulation pulses were employed. The differences in noradrenergic transmission observed in the terminal fields of the vBNST and the AV may differentially regulate activity in these two regions that both contain high densities of norepinephrine terminals.
Norepinephrine is involved in the regulation of a number of behaviors including learning and memory, arousal and stress reactions to challenging environments, and drug addiction (Onaka & Yagi, 1998; Berridge & Waterhouse, 2003; Forray & Gysling, 2004; Aston-Jones & Cohen, 2005). Despite its important roles in the regulation of behavior, neuronal regulation of its release and uptake has been studied far less than for dopamine. In part, this is because the distribution of norepinephrine within the brain is diffuse (Mitchell et al., 1994) and high sensitivity is required for detection of its low physiological concentrations. Brain regions that have appreciable amounts of norepinephrine are often only a few hundred microns across (Kilts & Anderson, 1986; Palij & Stamford, 1992), further complicating detection.
Two important noradrenergic pathways in the brain are the dorsal noradrenergic bundle (DNB) and ventral noradrenergic bundle (VNB). The DNB primarily originates in the locus coeruleus (LC; A6) and then merges into the medial forebrain bundle with projections to targets that include the thalamus, hypothalamus, and cortex (Paxinos, 1995; Berridge & Waterhouse, 2003). The VNB has its origin in the nucleus tractus solitarius (NTS; A2) and other noradrenergic cell groups (A1, A5 and A7). It courses between the substantia nigra and ventral tegmental area (Ungerstedt, 1971; Johnston et al., 1987) and then also merges with the medial forebrain bundle before projecting to target regions such as the septum and hypothalamus (Paxinos, 1995). Until they merge into the medial forebrain bundle, the VNB and DNB are distinct (Miyahara & Oomura, 1982; Saphier, 1993), facilitating independent excitation of each pathway by electrical stimulation. Electrically evoked norepinephrine can be detected with fast-scan cyclic voltammetry at carbon-fiber microelectrodes (Palij & Stamford, 1992), a technique demonstrated to be useful for monitoring the dynamics of extracellular dopamine in the brain (Millar et al., 1985; Garris & Wightman, 1994; Owesson-White et al., 2008). The high time resolution of this approach has enabled critical differences in the dynamics of extracellular dopamine to be characterized in specific brain regions such as the caudate-putamen and amygdala (Garris & Wightman, 1994). The small size (μm range) of the chemical sensor allows a local view of the dynamics of catecholamine release and uptake.
Here we report on norepinephrine release evoked by stimulation in the DNB and VNB. Release was monitored in the anteroventral thalamic nucleus (AV), a region characterized in previous work (Brun et al., 1993; Dugast et al., 2002) and compared to norepinephrine release in the ventral bed nucleus of the stria terminalis (vBNST), a region that receives a dense norepinephrine innervation (Oke et al., 1983). The AV plays an important role in learning and memory while the vBNST regulates autonomic and behavioral responses related to stress, anxiety and noxious stimuli (Onaka & Yagi, 1998; Delfs et al., 2000; Van Groen et al., 2002). Comparison of noradrenergic transmission in the terminal fields in the vBNST and the AV provides new insights into the regulation of extracellular norepinephrine in these two regions that regulate very different aspects of animal behavior.
Adult male Sprague-Dawley rats (320-400g) were purchased from Charles Rivers (Wilmington, MA) and housed in temperature and humidity controlled rooms with a 12 hour on/off light cycle. Food and water were available ad libitum. All procedures for handling and caring for the laboratory animals were in accordance with the Guide for Care and Use of Laboratory Animals and were approved by the Institutional Animal Care and Use Committee of the University of North Carolina at Chapel Hill.
Male Sprague-Dawley rats were anesthetized with urethane (1.5 mg/kg) and immobilized in a stereotaxic frame (David Kopf Instruments, Tujunga, CA). Temperature was maintained at 37 °C with a heating pad (Harvard Apparatus, Holliston, MA). The dorsal skull surface was exposed, and small holes were drilled in the skull for reference (Ag/AgCl) and stimulating electrodes as well as the carbon-fiber microelectrodes. Anteroposterior (AP), mediolateral (ML) and dorsoventral (DV) positions were referenced from bregma. The dura mater was punctured, carefully removed, and a carbon-fiber microelectrode was lowered into the BNST (AP 0.0 mm, ML +1.2 mm, DV from -6.0 to -8.0 mm), the AV (AP -2.1 mm, ML +1.5 mm, DV from -5.0 to -6.0 mm) or the CPu (AP +1.3 mm, ML +2.0 mm, DV from -4.5 to -6.5 mm; coordinates from a stereotaxic atlas (Paxinos & Watson, 2007). An Ag/AgCl reference electrode placed in the contralateral cortex was employed for all measurements and it was secured on the skull with dental cement.
Electrical stimulation was accomplished with a bipolar, stainless-steel electrode (0.2 mm diameter for each tip, the tips are separated by ~1.0 mm, Plastics One, Roanoke, VA). The stimulating electrode was insulated to the tip. It was placed into either the DNB (AP -5.2 mm, ML +1.2 mm, DV from -5.0 to -6.0 mm) or the VNB that is at the same AP and ML coordinates but is ~2 mm lower. Because the tips of the bipolar electrode were separated by ~1.0 mm, the VNB placement also stimulated the VTA and SN (Miyahara & Oomura, 1982; Saphier, 1993). Supplemental Figure 1 is a schematic of the ascending norepinephrine and dopamine pathways and the sites for their respective terminal areas (Ungerstedt, 1971).
Stimulus pulses were computer-generated (NL 800A, Neurolog, Medical Systems Corp., Great Neck, NY, U.S.A.) and were electrically isolated from the electrochemical system. The electrical stimulation consisted of biphasic square wave pulses (300 μA and 2 ms each phase) with stimulation frequencies between 10 and 60 Hz. The number of stimulus pulses was normally held constant at 60 for the vBNST and the AV and 24 for the CPu. Stimulations were repeated every 4 or 5 min. With this time interval, stimulation resulted in reproducible responses for both norepinephrine and dopamine release. Shorter interval (< 2 min) resulted in diminished release.
Glass-encased cylindrical carbon-fiber microelectrodes and Ag/AgCl reference electrodes were prepared as described earlier (Cahill & Wightman, 1995). T-650 untreated carbon fibers (Thornel, Amoco Corp., Greenville, SC) with an exposed length of 75 - 100 μm and 6 μm in nominal diameter were used. fast-scan cyclic voltammetry was computer-controlled and has been described in detail previously (Heien et al., 2003). A triangular scan (-0.4 to +1.3 V, 400V/s) was repeated every 100 ms. Data was digitized and stored on a computer using software written in LABVIEW (National Instruments). Background-subtracted cyclic voltammograms were obtained by digitally subtracting voltammograms collected during stimulation from those collected during baseline recording. Temporal responses were determined by monitoring the current at the peak oxidation potential for norepinephrine and dopamine in successive voltammograms. The current was converted to concentration based on calibration curves of each carbon-fiber microelectrode obtained after the in vivo experiment with known concentrations of dopamine and norepinephrine in vitro.
Rats were anesthetized with urethane and decapitated. The brain was rapidly removed and placed on ice. Coronal brain slices (500 μm thick) containing the ventral bed nucleus of the stria terminalis (vBNST) were prepared in ice cold artificial cerebral spinal fluid (aCSF) using a Lancer Vibratome (World Precision Instruments, Sarasota, Fl.). The aCSF contained (in mM): 20 HEPES, 2.4 CaCl2, 1.2 MgCl2, 1.2 NaH2PO4, 2.45 KCl, 126 NaCl, 11 glucose, and 25 NaHCO3. The pH was adjusted to pH 7.4 and the aCSF was saturated with 95% O2 / 5% CO2. Tissue samples were dissected from the brain slices containing the area of interest (500 μm thickness). The tissue was wet weighted in a pre-tared volume of 200 μL 0.1 N HClO4 spiked with 1 μM hydroquinone (HQ). The tissue was homogenized with a sonic dismembrator (Fisher Sci., Model 60, Pittsburgh, PA, USA). The homogenized tissue was centrifuged at 6000 rpm for 10 min and the supernatant was removed and filtered using a 0.2-μm syringe microfilter (Millex-LG). Injections (10 μL) were made onto a reverse phase column (C-18, 5 μm, 4.6 × 250 mm, Waters symmetry 300). The mobile phase (prepared in HPLC grade water) contained 0.1 M citric acid, 1 mM hexyl sodium sulfate, 0.1 mm EDTA, and 10% methanol (pH 3.5) at a flow rate of 1 mL/min. Catecholamines were detected with a thin-layer radial electrochemical cell (BASi, West Lafayette, IN, USA) at a potential of 700 mV versus a Ag/AgCl reference electrode. Catecholamine standards were prepared from 10 mM stock solutions in 0.1 N HClO4. A Labview stripchart recorder program (Jorgenson Lab, UNC-CH) was used for data collection through home built electronic equipment. The peak area of the analyte was ratioed to the peak area of the internal standard, HQ. After correction for differential detector response, the amount of analyte in the brain tissue was calculated.
Lidocaine was administered via an infusion cannula (33 gauge) that was combined with the bipolar stimulating electrodes and inserted into the implanted guide cannula (26 gauge, Plastics One Inc., VA, USA). Microinfusions (0.5 μL) were made over 60 s with a syringe pump (Kent Scientific Corporation, CT, USA). The regions targeted were normally the noradrenergic pathways, DNB and VNB. In one set of experiments, the noradrenergic cell bodies in the LC (AP -9.8 mm, ML +1.4 mm, DV 6.5-7.5 mm) and the NTS (AP -12.0 mm, ML +1.8 mm, DV 7.6-8.6 mm) were targeted.
At the end of experiments, electrode placements were verified by stereotaxically lowering a tungsten electrode to the original position of the carbon-fiber microelectrode. A lesion was made at the recording site by applying continuous current (20 μA for 10 s) to the tungsten electrode. Some lesions (> 20 rats) of the vBNST and the AV were made directly with the carbon-fiber microelectrodes, but this precluded postcalibration of the electrode sensitivity. These experiments verified the accuracy of the repositioning of tungsten electrodes in recording sites. Brains were removed from the skull, stored in 10 % formaldehyde for at least 3 days, and coronally sectioned into 40 μm slices with a cryostat. The sections were mounted on slides, stained with 0.2 % thionin, and coverslipped before viewing under a light microscope.
All chemicals and drugs were reagent-quality or better and were used without additional purification. Drugs were obtained from Sigma-Aldrich (St. Louis, MO, USA) unless otherwise indicated. In vitro postcalibration of carbon-fiber microelectrodes was performed in a Tris buffer solution at a pH 7.4 containing 15 mM Tris, 140 mM NaCl, 3.25 mM KCl, 1.2 mM CaCl2, 1.25 mM NaH2PO4, 1.2 mM MgCl2, and 2.0 mM Na2SO4 in double distilled water (Mega Pure System, Corning Glasswork, Corning, NY). Lidocaine (100 μg/0.5 μL) was dissolved in saline and sterile filtered with the pH adjusted to 6 with NaOH. Desipramine-HCl, raclopride-HCl, clonidine-HCl and yohimbine-HCl were dissolved in saline. GBR 12909-HCl was dissolved in double distilled water and then diluted with saline. Injected volumes were 0.6 ml/kg and were given intraperitoneally (i.p.).
Clampfit 8.1 as part of pCLAMP 8.1 software package (Axon Instruments, Foster City, CA) was used to analyze half-life and signal amplitude values according to procedures described in the literature (Park et al. 2005). The half life of catecholamine uptake, t1/2, was taken as the time to descend from its maximum value to half of that value. Since uptake follows Michaelis-Menten kinetics, this procedure is an oversimplification. Nevertheless, we used it to obtain figures that give an approximate measure of the changes in data induced by the various drugs. Data are represented as mean ± S.E.M. and ‘n’ values indicating the number of rats. Mean values were compared by using the two-tailed Student's t-test (GraphPad Software version 4.0, San Diego, CA, USA) to calculate the level of significance. P < 0.05 was regarded as statistically significant.
Two regions of vBNST were dissected from fresh brains and their monoamine (norepinephrine, dopamine and serotonin) content determined by HPLC (Table 1). The ratio of norepinephrine to dopamine content for the anteromedial vBNST (AP +0.5 to 0.0 mm from bregma) was ca. 2 to 1 whereas in the medioposterior vBNST region (AP 0.0 to -0.5 mm from bregma) it was ca. 10 to 1 consistent with a previous report (Kilts & Anderson, 1986). Serotonin was not detected in either subregion. The norepinephrine content in the medioposterior vBNST (3.25 ± 1.89 μg/g of tissue) is similar to the 2.1 μg/g of tissue previously found in the AV (Mitchell et al., 1994), a region that also contains negligible dopamine (Oke et al., 1983). The relatively large standard deviation for the neurochemical values in the vBNST is atypical of liquid chromatographic measurements and may well reflect the heterogeneity of this region and our failure to dissect exactly the same nuclei each time (the site in the vBNST that supports norepinephrine release is ~300 μm thick). However, the norepinephrine content in both of these regions is considerably less than the content of dopamine (9 μg/g of tissue) in the CPu, a region that contains little norepinephrine (Garris & Wightman, 1994).
In the first set of in vivo experiments, the carbon-fiber microelectrode was placed in the vBNST or AV while the stimulating electrode was lowered from 3.5 mm below the skull surface. Based on catecholamine content of vBNST, we lowered the carbon-fiber microelectrode in the medioposterior vBNST to monitor extracellular norepinephrine as shown in supplementary Figure 2.
Catecholamine overflow was monitored while the electrical stimulation was delivered at different depths. Stimulations (60 Hz, 60 pulses) were delivered at 200-300 μm intervals. Figure 1A shows the coronal plane (AP -5.2 mm) and track (dotted lines, centered on 1.2 mm ML) of the tips of the bipolar stimulating electrode. The DNB is located ~2 mm above the VTA/SN while the VNB passes within the VTA/SN (Ungerstedt, 1971; Miyahara & Oomura, 1982; Saphier, 1993). The maximum current in the cyclic voltammograms at the potential where catecholamines are oxidized (0.6 V vs Ag/AgCl) was recorded and plotted as a function of the position of the stimulating electrode (Figure 1B).
When the carbon-fiber microelectrode was in the vBNST responses were not observed until the stimulating electrode was below a depth of 5 mm. Evoked release reached an initial maximum at depths of 5.5-7.0 mm, the location of the DNB (Figure 1B) and a second maximal release was evoked between 8.2 and 8.8 mm beneath the skull surface, the location of the VNB. When at the level of the VNB, the stimulating electrode tips are in close proximity to the VTA/SN. Indeed, the dopamine cell bodies are also activated with VNB stimulation since catecholamine release was evoked in the CPu, a region containing little norepinephrine. DNB stimulation does not evoke catecholamine release in the CPu (Garris et al., 1993). When the carbon-fiber microelectrode was in the AV, maximal release was also observed with the stimulating electrode in the DNB (between 5.5 and 7.0 mm). However, catecholamine release in the AV was not detected when the stimulating electrode was at the level of the VNB. Individual traces and background-subtracted cyclic voltammograms recorded in the vBNST and AV along with their respective cyclic voltammograms are shown in Figure 1C.
The cyclic voltammograms recorded in different terminal regions are compared to authentic dopamine and norepinephrine in Figure 2. Although slight differences exist, within experimental error dopamine and norepinephrine cannot be distinguished based on their voltammograms. However, the cyclic voltammograms for the catecholamines are distinct from those for all other substances establishing that the evoked release is a catecholamine (Heien et al., 2003). The signal for norepinephrine at a specific concentration is only 60 % of that for dopamine because it adsorbs to the electrode to a lesser extent (Heien et al., 2004).
In the next series of experiments the stimulating electrode was fixed at a location in the DNB or the VNB while the carbon-fiber microelectrode was lowered through terminal regions. Figure 3A shows the coronal plane (AP 0.0 mm) that was used for investigating catecholamine release along a track that includes the vBNST. The carbon-fiber microelectrode was lowered from 4.2 to 8.5 mm below the skull surface (track shown by a dotted line in Figure 3A (left), 1.2 mm ML), and cyclic voltammograms were recorded at different vertical positions during electrical stimulations (60 Hz, 60 pulses). When the VNB was stimulated, catecholamine release was seen in two regions of the BNST as the carbon-fiber microelectrode was lowered (Figure 3A, right). Catecholamine overflow was maximal at ~6.0 mm, the level of the dorsal BNST (dBNST), disappeared at the level of the anterior commissure, and then reached a maximal response at ~7.3 mm, the level of the vBNST. When the stimulating electrode was in the DNB, only low release was evoked in the dBNST, none was seen at the level of the anterior commissure, and catecholamine release was maximal at the level of the vBNST (Figure 3A right). The responses in the vBNST were lower with the carbon fiber further posterior (AP - 0.4 to - 0.7 mm, data not shown), consistent with previous work (Forray & Gysling, 2004).
Figure 3B shows the coronal section containing the AV (AP -2.2 mm) in which the carbon-fiber microelectrode was lowered (ML 1.5 mm) from 4.2 to 7.2 mm from the skull surface while stimulated release was evoked as described above. The AV is at 4.9 to 5.7 mm along this tract. When the stimulating electrode was placed in the DNB, maximal catecholamine release was observed at 5.5 mm below the skull surface, coincident with the AV (Figure 3B, right). No release was seen below 6.2 mm. When the stimulating electrode was in the VNB, there was no detectable signal in the AV with the exception of a small signal in one out of six rats. This result is consistent with previous studies that catecholamine input into the AV is primarily through the DNB (Kobayashi et al., 1974; Brun et al., 1993).
To confirm the origin of the signals in the AV, we tested the effects of drugs known to act at noradrenergic neurons and compared their responses to drugs with known actions at dopaminergic neurons. α2-adrenoceptors down-regulate norepinephrine release, a process that is blocked by the antagonist yohimbine (Ghasemzadeh et al., 1993). D2 receptors play a similar role on dopamine neurons where they act as autoreceptors and inhibit release. Thus, raclopride, a D2 antagonist, increases stimulated release from dopamine neurons (Seeman & Van Tol, 1993), but does not affect release from norepinephrine neurons (Gobert et al., 1998). Desipramine selectively inhibits the norepinephrine transporter (NET), reducing the clearance rate of norepinephrine (Dugast et al., 2002), but it does not affect the dopamine transporter (DAT) (Kula et al., 1999). In contrast, GBR12909 blocks the DAT while it has no affect on the NET (Palij & Stamford, 1992; Dugast et al., 2002).
The effects of noradrenergic drugs on stimulated release were evaluated in the AV (60 pulses, 60 Hz) and compared to dopaminergic responses in the CPu (24 pulses at 60 Hz, representative examples in Figure 4). Before drug administration the clearance half-life (t1/2, the time required for the concentration to decay to 50 % of the maximum) in the AV was 1.28 ± 0.11 s (n = 15) in close agreement with that reported previously (Mitchell et al., 1994). The half-life for catecholamine in the CPu was 0.83 ± 0.05 sec (n = 7) was almost half of that in the AV (t = 2.70, d.f. = 20, P = 0.01). Overflow in the AV was significantly increased 30 min after yohimbine (5 mg/kg). These changes were quantified by recording the maximal evoked release [CA]max and t1/2. These are reported as a percent of predrug values in Table 2. Both parameters significantly increased after yohimbine in the AV. Injection of desipramine (15 mg/kg) 30 min after yohimbine further increased [CA]max and t1/2. The effects of the drugs were apparent within 10 min of administration and reached a maximum value at ~30 min. Desipramine (15 mg/kg) alone, also significantly increased both [CA]max and t1/2 in the AV. Note that the noradrenergic drugs at the same doses did not significantly change dopamine overflow in the CPu elicited by VTA/SN stimulations (Figure 4A, right, Table 2).
When dopaminergic drugs were applied, changes in stimulated release were only observed in the CPu with VTA/SN stimulation, but not in the AV with DNB stimulation (Table 2). The evoked response in the AV was not significantly changed by raclopride (2 mg/kg) or GBR 12909 (15 mg/kg)(Figure 4B, left). This work is consistent with prior results showing that electrically stimulated release in the AV is sensitive to drugs specific for the noradrenergic system but insensitive to drugs that affect dopaminergic neurons (Stamford et al., 1988; Brun et al., 1993). In contrast, both [CA]max and t1/2 in the CPu were significantly increased after the raclopride (Figure 4B, right). Administration of GBR 12909, 30 min after raclopride, dramatically increased stimulated catecholamine overflow in the CPu. Evoked catecholamine overflow during control experiments with saline injection was stable and reproducible for at least 2 h (data not shown).
While evoked release in the vBNST has been pharmacologically characterized in brain slices (Palij & Stamford, 1992; 1993), it has not in vivo. Therefore, evoked catecholamine release in the vBNST was also evaluated with noradrenergic specific (yohimbine, desipramine and the norepinephrine α2-adrenoceptor agonist, clonidine) and dopamine specific (raclopride and GBR 12909) drugs using both DNB and VNB stimulations (60 Hz, 60 pulses). Before drug administration, norepinephrine clearance in the vBNST had t1/2 of 2.45 ± 0.19 sec (n = 8) for VNB and 2.19 ± 0.16 sec (n = 9) for DNB stimulations, values that are not significantly different from each other but that are significantly greater than in the AV (t = 5.73, d.f. = 21, P < 0.0001 for VNB and t = 4.83, d.f. = 22, P < 0.0001 for DNB). The half lives for catecholamine clearance before drugs in the vBNST with stimulation of either noradrenergic pathway were almost 3 times longer than in the CPu (t = 7.74, d.f. = 13, P < 0.0001 for VNB and t = 7.23, d.f. = 14, P < 0.0001 for DNB). Evoked release was significantly decreased 30 min after clonidine injection (0.2 mg/kg) with either DNB or VNB stimulations, a decrease that was reversed by yohimbine administration (5.0 mg/kg, Figure 5A). When yohimbine was administered alone, evoked release was significantly increased 30 min later (Figure 5B) with stimulations of both pathways. These responses were further increased with injection of desipramine 30 min later. The average results from all animals (Table 2) show that both [CA]max and t1/2 were significantly increased by the noradrenergic drugs. However, the effects of desipramine and yohimbine on [CA]max and t1/2 in the vBNST were significantly less than in the AV (Table 2). Catecholamine overflow evoked by DNB or VNB stimulation in the vBNST was not changed by raclopride (2 mg/kg) or GBR 12909 (15 mg/kg, Figure 5C, Table 2).
After the experiments with yohimbine and desipramine in the vBNST, catecholamine overflow evoked by the stimulation of the DNB and the VNB was recorded for 20 min at 5 min intervals and, then, lidocaine (350 nmol) was microinfused onto the electrical stimulation tips. Lidocaine is a voltage-gated Na+ channel blocker that prevents action potential generation (Boehnke & Rasmusson, 2001). It eliminated stimulated release within 5 min after its microinfusion (Figure 6) and stimulated release was partially restored 30 min later. Identical results were obtained when the stimulating electrode and microinfusion were placed in the noradrenergic cell bodies, either the LC or the NTS. Thus, catecholamine overflow in the vBNST is due to the neuronal responses originating from both the NTS and the LC.
The responses in each noradrenergic region to different stimulation parameters were investigated. Figure 7A shows that the amplitude increased in the AV and the vBNST with increasing stimulation frequency (10, 20, 40, and 60 Hz, 60 pulses). The evoked concentration change during a 20 Hz stimulation in the AV approached a plateau while the concentration change in the vBNST continued to increase during the stimulation. However, at higher frequencies in both regions the concentration increased during the stimulation. Figure 7B shows the responses in the vBNST and the AV with different pulse numbers (10, 20, 40 and 80 and 120 pulses at 60 Hz). Measurable responses (signal to noise (S/N) ≥ 5) were normally observed with at least 10 pulses at 60 Hz with our present instrumentation. The extracellular catecholamine concentration evoked by both DNB and VNB stimulation in the vBNST increased with pulse number but the stimulated release in the AV did not increase when more than 40 pulses were used. Pooled maximal responses are shown in Figure 7C and D. The maximum evoked catecholamine concentration ([CA]max) at 60 Hz, 40 pulses in the CPu with stimulation at the level of the VNB was ~11 times greater than in the vBNST and AV in Table 3.
This study provides novel neurochemical information for two brain regions, the AV and vBNST, that receive high densities of noradrenergic input. The AV receives major input from the hippocampal formation via the mammillary bodies of the hypothalamus and has reciprocal connections with limbic circuits. While the functional significance of norepinephrine signaling in the AV has not been extensively studied, it is thought to play an important role in behavioral learning and memory and the conscious perception of emotion (Gabriel et al., 1980; Sparenborg & Gabriel, 1992; Oda et al., 2003). Conversely, the BNST is highly connected to the ventral subiculum of the hippocampus, neocortex, the paraventricular and supraoptic nucleus of the hypothalamus, the ventral tegmental area and other limbic structures such as the amygdala (Dong & Swanson, 2004; Forray & Gysling, 2004; Massi et al., 2008). The norepinephrine input to vBNST plays a key role in stress and drug related behaviors (Delfs et al., 2000; Fendt et al., 2005). However, investigation of noradrenergic transmission within the AV and vBNST is extremely difficult because of the small size of these nuclei. The present study provides a foundation for studies of norepinephrine communication in animal behaviors such as learning and fear in these brain regions. The results clearly show that norepinephrine neurotransmission has regionally specific characteristics in discrete regions of the brain that can be monitored with carbon-fiber microelectrodes and fast-scan cyclic voltammetry. The select positioning of stimulating electrodes in either noradrenergic pathways (Figure 1) or cell body regions (Figure 6) allows characterization of noradrenergic afferents that innervate small brain regions such as the AV and vBNST.
These two regions were selected for this initial study because their norepinephrine content is high relative to dopamine. Moreover, measurements in brain slices have shown that norepinephrine release in the vBNST can be evoked with local stimulation and have identified the pharmacological control points in these regions (Palij & Stamford, 1992; 1993; Miles et al., 2002). Our brain mapping and pharmacological investigations in the intact animal expand upon that work. A major finding is that the noradrenergic inputs into the vBNST from both the DNB and VNB provide have quite similar capacity for release. We also found that control by autoreceptors and NET is quite similar for the neurons of both pathways into the vBNST. In contrast, the regulation of release and uptake was strikingly different in the AV, a region only innervated by the DNB.
The background-subtracted cyclic voltammograms provide a “fingerprint” that identifies the released substance(s) as a catecholamine (Figure 2). These cyclic voltammograms enable the catecholamines to be distinguished from other substances such as serotonin, ascorbic acid, and catecholamine metabolites (Palij & Stamford, 1992; Heien et al., 2003). Another approach that has been used to monitor norepinephrine is differential pulse amperometry (Suaud-Chagny et al., 1990; Brun et al., 1993). This method can resolve catecholamines from ascorbate but not from their metabolites and suffers from a slow response time. Continuous amperometry has also been used (Dugast et al., 2002; Yavich et al., 2005). This approach has little chemical selectivity but can be used to establish that the half-life of evoked norepinephrine and dopamine is on the time scale of seconds.
Unfortunately, like all other electrochemical techniques, background-subtracted cyclic voltammograms cannot distinguish between norepinephrine and dopamine although the cyclic voltammograms enable the catecholamines to be distinguished from the other substances as mentioned above. Epinephrine also has similar electrochemical properties to dopamine and norepinephrine but the epinephrine tissue content is sufficiently low in the brain regions investigated that it is unlikely to be measured (Van der Gugten et al., 1976; Kilts & Anderson, 1986). Our assignment of the evoked signals measured in the AV and vBNST to norepinephrine thus relies heavily on the neurochemical results (Table 1) that show that dopamine is a minority catecholamine in each region (Oke et al., 1983; Kilts & Anderson, 1986; Mitchell et al., 1994). Furthermore, immunofluorescence and autoradiographic tracing have shown that the AV and the vBNST receive few dopaminergic terminals (Segal et al., 1973; Swanson & Hartman, 1975; Versteeg et al., 1976; Freedman & Cassell, 1994; Kozicz, 2001). However, we are hesitant to assign the species detected in regions where the tissue content of dopamine and norepinephrine is similar or has not been established even when the cyclic voltammograms have the characteristic shape of a catecholamine. For example, release seen in the dBNST with VNB stimulation (Figure 3A), a region that contains similar amounts of dopamine and norepinephrine (Kilts & Anderson, 1986), likely contains a dopamine component because the VNB stimulation also excites dopamine cell bodies in the VTA/SN.
The small size of the carbon-fiber microelectrode enables maps to be generated of sites that support catecholamine release with up to 100 μm resolution, and enable in vivo tracking of neuronal pathways. While such maps have been generated previously for dopamine pathways (Garris et al., 1994; Garris & Wightman, 1994), their use to explore norepinephrine (Suaud-Chagny et al., 1986) innervation is less extensive. When the carbon-fiber microelectrode was implanted either within the vBNST or the AV, electrical stimulation of the DNB, the pathway of the noradrenergic neurons from the LC and other nuclei (Kobayashi et al., 1974; Moore & Bloom, 1979), evoked catecholamine release in both brain regions (Figure 3). Prior work has shown that this stimulation location does not support release in the CPu (Garris et al., 1993). When the stimulating electrode was lowered to the VNB, robust catecholamine release was observed in the vBNST, but no release was observed in the AV. This is consistent with the lack of VNB noradrenergic projections to the AV (Moore & Card, 1984). The VNB stimulation also evoked release in the CPu because the cell bodies for dopamine, the predominant catecholamine in this region, are in the VTA/SN which is also stimulated at the coordinates we used for the VNB. Thus, the absence of a response in the AV also confirms that it lacks dopaminergic inputs.
The VNB noradrenergic input into the vBNST has been demonstrated by retrograde labeling (Terenzi & Ingram, 1995; Delfs et al., 2000; Forray et al., 2000). This assignment is supported by our pharmacological, voltammetric, and neurochemical measurements that show that the catecholamine detected is norepinephrine and not dopamine even though we were directly activating dopamine neurons with the site we selected for VNB stimulation. More surprising is the robust norepinephrine release obtained with DNB stimulation. These findings are consistent with anatomical evidence of a projection to the vBNST from the LC (Lindvall & Stenevi, 1978; McNaughton & Mason, 1980) via the DNB (Moore, 1978; Phelix et al., 1992), a concept suggested by Adams and coworkers (Mitchell et al., 1994). In support of the dual noradrenergic inputs to the vBNST, norepinephrine release was evoked by both the direct stimulation of the LC and the NTS (Figure 6), the origins of the two pathways. Despite similar amounts available for release in the vBNST from the two pathways, it appears that they are activated by different conditions during animal behaviors. For example, during opiate withdrawal, the noradrenergic neurons of the VNB are stimulated to a greater degree than those of the DNB (Delfs et al., 2000)
The various stimulation parameters and pharmacological agents that we employed revealed significant differences in the characteristics of release and uptake in these two norepinephrine rich regions. The rapid time resolution of the recordings at carbon-fiber microelectrodes enabled the dynamics of release and uptake to be characterized with subsecond temporal resolution, revealing these differences. The pharmacological studies reveal that release and uptake in both regions is affected by noradrenergic drugs providing additional evidence that the observed release is norepinephrine. Furthermore, release in both regions is unaffected by dopaminergic drugs at doses that are highly effective in the modulation of dopamine release in the striatum. Indeed, the high selectivity of the drugs used in this work for the dopaminergic and noradrenergic pathways indicates that they can be used to distinguish dopamine and norepinephrine release as long as autoreceptors and the NET or DAT are exerting exclusive control on extracellular catecholamine levels.
The half-life measurements for norepinephrine disappearance after evoked release suggest that uptake is approximately twice as fast in the AV as in the vBNST. The sensitivity of the clearance to desipramine in both regions, coupled with their insensitivity to GBR 12909, clearly shows that clearance is due to uptake mediated by NET. While the absolute values of t1/2 given in the Results section are not corrected for the response time of the electrode, they do show similar trends to those reported previously with continuous amperometry. In accord with our findings, norepinephrine uptake in the AV was reported to be significantly slower than for dopamine uptake in the striatum (Dugast et al., 1994; Dugast et al., 2002). Slower uptake allows extrasynaptic neurotransmitters to participate in volume transmission because they can diffuse greater distances before removal from the extracellular fluid by the transporter. Interestingly, the measured half-life for clearance measured in the vBNST was independent of the noradrenergic pathway stimulated. Thus, it appears that the same uptake sites are controlling uptake, irrespective of the pathway from which release originates.
The inhibition of the α2-adrenoceptor increased release evoked by VNB stimulation in the vBNST and by DNB stimulation in both the AV and vBNST. However, the increase in release was significantly greater in the AV than with stimulation of either pathway innervating the vBNST (Table 2), indicating stronger control by noradrenergic autoreceptors in the AV. Thus, release and uptake by autoreceptors and NET are both more tightly regulated in the AV compared to the vBNST.
In addition to the greater control of extracellular norepinephrine in the AV evidenced by pharmacological studies, there are notable differences in the release evoked in each region. DNB stimulation causes similar amounts of release in each region (Table 3). Furthermore, VNB stimulation, which does not evoke release in the AV, causes release of a similar concentration in the vBNST. In the vBNST the frequency dependence of release is quite similar to that reported in dopamine regions (Garris & Wightman, 1994). During low frequency stimulations (10-20 Hz), which are in the physiological range, norepinephrine concentrations approach a new, steady-state level during the stimulation (Figure 7). This behavior is that expected for neurotransmitter systems in which extracellular levels are maintained by the balance between uptake and release. At higher frequencies release takes more pulses to overwhelm uptake, and evoked concentrations continue to rise during the stimulation in the vBNST. However, in the AV, the stronger uptake coupled with the greater inhibition by autoreceptors, results in a greater control of norepinephrine concentrations during long stimulations leading to a lack of sensitivity to pulse number even with frequencies as high as 60 Hz stimulations (Figure 7). The importance of uptake control in the AV is clearly apparent with the dramatic increase in release amplitude following NET inhibition (Table 2). Thus, norepinephrine response in the AV resembles that at the rat tail artery where norepinephrine shows strong control by uptake (Msghina et al., 1999), with additional control of extracellular levels by the noradrenergic autoreceptor.
In the present study, we showed that in vivo fast-scan cyclic voltammetry coupled with carbon-fiber microelectrodes enable monitoring of rapid changes in norepinephrine concentration in the vBNST and the AV of anesthetized rats despite the small size of these regions and the low amounts of norepinephrine compared to that of dopamine in the striatum. This study also demonstrates that norepinephrine release in the vBNST was evoked by the stimulation of either DNB or VNB whereas only DNB stimulation evoked norepinephrine release in the AV. In both regions, released norepinephrine was regulated by autoreceptors and restored to prestimulus levels by uptake. However, control by autoreceptors and uptake was much stronger in the AV than in the vBNST.
Supplemental Figure 1. Schematic drawing shows the noradrenergic (left side) and dopaminergic (right side) nuclei and their ascending pathways, adapted from Ungerstedt (1971). Abbreviations used: LC (A6), locus coeruleus; NTS (A2), nucleus tractus solitarius; DNB, dorsal noradrenergic bundle; VNB, ventral noradrenergic bundle; VTA (A10), ventral tegmental area; SN (A9), substantia nigra; n, nucleus (reprinted with permission).
Supplemental Figure 2. The representative placements of the carbon-fiber microelectrode visualized by the electrolytic lesion (A) in the vBNST and (B) in the AV are shown in coronal sections. The black circles are based on a lesion made at the end of the experiments. The coronal section was taken from the atlas of Paxinos and Watson (2007).
We thank Andy Seipel, Leslie Sombers and Brandon Aragona for providing technical assistance. The authors are also indebted to Khristy Fontillas, Brandon Aragona and Nii Addy for helpful comments regarding the manuscript. This work was supported by NIH (NS 15841).