In Vivo Electrochemical Evidence for Simultaneous 5-HT and Histamine Release in the Rat Substantia Nigra pars Reticulata Following Medial Forebrain Bundle Stimulation

Abstract

Exploring the mechanisms of serotonin (5-hydoxytryptophan (5-HT)) in the brain requires an in vivo method that combines fast temporal resolution with chemical selectivity. Fast-scan cyclic voltammetry (FSCV) is a technique with sufficient temporal and chemical resolution for probing dynamic 5-HT neurotransmission events; however, traditionally it has not been possible to probe in vivo 5-HT mechanisms. Recently, we optimized FSCV for measuring 5-HT release and uptake in vivo in the substantia nigra pars reticulata (SNR) with electrical stimulation of the dorsal raphe nucleus (DRN) in the rat brain. Here, we address technical challenges associated with rat DRN surgery by electrically stimulating 5-HT projections in the medial forebrain bundle (MFB), a more accessible anatomical location. MFB stimulation elicits 5-HT in the SNR; furthermore, we find simultaneous release of an additional species. We use electrochemical and pharmacological methods and describe physiological, anatomical and independent chemical analyses to identify this species as histamine. We also show pharmacologically that increasing the lifetime of extracellular histamine significantly decreases 5-HT release, most likely due to increased activation of histamine H-3 receptors that inhibit 5-HT release. Despite this, under physiological conditions, we find by kinetic comparisons of DRN and MFB stimulations that the simultaneous release of histamine does not interfere with the quantitative 5-HT concentration profile. We therefore present a novel and robust electrical stimulation of the MFB that is technically less challenging than DRN stimulation to study 5-HT and histamine release in the SNR.

Introduction

5-HT is an important neuromodulator and dysfunctions of the 5-HT system are particularly well-documented in neurological disorders such as anxiety and depression (Petty et al., 1996). In vivo neurochemical measurements of 5-HT are necessary for furthering our understanding of the mechanisms that govern these disorders, and will improve their diagnosis and treatment. Microdialysis studies have correlated in vivo basal level 5-HT changes to behavioral and pharmacological manipulations (Rueter et al., 1997; Barnes and Sharp, 1999). Basal 5-HT levels are determined by numerous individual neurotransmission events that are averaged in one microdialysis reading (Robinson and Wightman, 2007) and to fully understand the mechanisms that underlie these slow changes requires quantitative, sub-second endogenous 5-HT detection. This has traditionally been accomplished with fast-scan cyclic voltammetry (FSCV) in tissue slice preparations (O'Connor and Kruk, 1991; Rice et al., 1994; Bunin and Wightman, 1998; John et al., 2006). These studies have provided a solid understanding of 5-HT release and uptake kinetics (Bunin et al., 1998), 5-HT receptor pharmacology (Davidson and Stamford, 1996; Threlfell et al., 2010), and 5-HT metabolism by monoamine oxidase (John and Jones, 2007). Recently, FSCV has been used to perform similar characterizations in D. melanogaster (Borue et al., 2009; Borue et al., 2010). While these types of studies are essential to describe basic 5-HT mechanisms, in vivo measurements in a mammalian brain are necessary to understand how 5-HT dynamics are modulated in the complex entity of the intact nervous system. Recently we described a FSCV technique in which carbon-fiber microelectrodes were coated with Nafion to monitor 5-HT in vivo (Hashemi et al., 2009). The Nafion modification reduces electrode fouling while increasing its sensitivity to 5-HT. This is the first technique capable of monitoring endogenous, in vivo 5-HT release and uptake on a sub-second time scale. In this previous work, we established a physiological model where we evoked 5-HT release with a bipolar stimulating electrode in the dorsal raphe nucleus (DRN), the location of 5-HT cell bodies, and recorded terminal 5-HT release and uptake in the substantia nigra, pars reticulata (SNR) (Hashemi et al., 2009).

In this paper, we explore an alternative method of evoking 5-HT release in the SNR. Using the well-documented efferent circuitry of the DRN (Azmitia and Segal, 1978; Moore et al., 1978; Parent et al., 1981; Imai et al., 1986), we exploit a branched 5-HT projection to both the SNR and the striatum (van der Kooy and Hattori, 1980; Imai et al., 1986) that is located in the medial forebrain bundle (MFB). We previously showed that electrical stimulation of the MFB released an unidentified substance in the red nucleus of the rat, that we suspected was 5-HT (Kita et al., 2009). This suggested that MFB stimulation could evoke 5-HT release in targets posterior to the stimulating electrode. Here, we confirm this expectation by demonstrating that branched 5-HT fibers can be retrogradely activated, allowing the experimenter to avoid targeting the DRN, an anatomically challenging surgical technique. However, we find that MFB stimulation elicits release of an additional species in the SNR. With electrochemical and pharmacological data, in addition to literature well documenting histamine's physiological, anatomical and independent chemical verification in the SNR (Brown et al., 2001), we identify this substrate as histamine. Threllfel et al. have previously shown that pharmacological activation of H-3 receptors inhibits 5-HT release (Threlfell et al., 2004). In accord with this, we find significant decreases in 5-HT release upon administration of SKF 91488, an agent that prolongs histamine lifetime in the synapse. With no drug present, we compare the 5-HT concentration release and uptake profiles between DRN and MFB stimulations and find that histamine release does not interfere with the quantification of the 5-HT signal. We thereby present a novel and robust method for studying 5-HT and histamine neurotransmission in the SNR.

Materials and Methods

Animals

Male Sprague–Dawley rats, 8–12 weeks old, weighing 250–350 g, were purchased from Charles River Labs (Raleigh, NC, USA). Rats were housed under 12h/12h light cycles with controlled temperature and humidity. Food and water were available ad libitum. All animal care was in accordance with the Guide for the Care and Use of Laboratory Animals and was approved by the Institutional Animal Care and Use Committees of the University of North Carolina.

Surgery

Rats were anesthetized with urethane (1.5 g kg⁻¹ rat weight) and positioned into a stereotaxic frame (David Kopf Instruments, Tujunga, CA, USA). Holes were drilled in the skull according to stereotaxic coordinates referenced from bregma and taken from Paxinos and Watson's Rat Brain Atlas (Paxinos and Watson, 2007). Nafion-modified carbon-fiber microelectrodes were implanted in the SNR (stereotaxic coordinates AP −4.8 to −5.2; ML +2.0; DV −8.5). A bipolar stainless steel stimulating electrode, insulated to the tip (0.2 mm diameter, Plastics One, Roanoke, VA, USA) was implanted into the MFB (AP −2.5 to −2.8; ML 1.7; DV −8.0) or DRN as described previously (Hashemi et al., 2009). An Ag/Ag Cl wire serving as a reference electrode was implanted into the contralateral hemisphere. Computer-generated biphasic pulse trains were applied through constant current stimulators (NL 800A, Neurolog, Medical Systems Corp., Great Neck, NY, USA), 2 ms in width and 350 μA each phase (unless otherwise noted), at 60 Hz for 2 s to evoke 5-HT release.

Voltammetric Procedures

Cylindrical carbon-fiber microelectrodes were constructed by aspiration of a single 2.5-μm radius carbon fiber (T-650, Thornel, Amoco Co.) into a glass capillary of 0.6 mm external diameter and 0.4 mm internal diameter (A-M Systems, Inc., Sequim, WA). A micropipette puller (Narishige, Tokyo, Japan) was used to taper the glass and form a carbon-glass seal. The exposed carbon fiber was cut to approximately 100 μm in length and was soaked for 30 minutes in isopropyl alcohol (IPA) to clean the surface. The procedure for electrodeposition of Nafion was described previously (Hashemi et al., 2009). Dopamine and 5-HT specific electrochemical detection waveforms were used as described elsewhere (Jackson et al., 1995; Heien et al., 2003). A customized version of TH-1 software (ESA, Chelmsford, MA) written in LABVIEW (National Instruments, Austin, TX) was used for waveform generation and data acquisition. A custom-built UEI potentiostat (University of North Carolina at Chapel Hill, Department of Chemistry Electronics Facility) was employed. All potentials are reported versus an Ag/Ag Cl reference electrode. Signal processing (background subtraction, signal averaging, and digital filtering (4-pole Bessel Filter, 5 kHz)) was also done in TH-1 software. Voltammetric data are visualized as color plots that show multiple background-subtracted cyclic voltammograms that were consecutively collected. The abscissa is time and the current is encoded as false color (Michael et al., 1998). For most of the work in this paper, the ordinate is the voltage axis that peaks at 1.0 V, scans back to −0.1 V and returns to 0.2 V, the rest potential. The sensitivity with this waveform on Nafion modified microelectrodes to 5-HT is 49.5 nA μM⁻¹ and 0.4 nA μM⁻¹ to histamine.

Histology

To verify the spatial placement of the electrodes in vivo, the carbon-fiber microelectrode that acquired data was used to create a small, specific lesion in the recording site by a applying constant voltage (20 V for 10 s) (Park et al., 2009). Following the experiment, rats were sacrificed and perfused with 10% formalin solution. Brains were then removed from the skull and stored in 10% formalin. After at least 3 days, the brains were flash-frozen, sectioned into 40 μm slices in a cryostat, mounted on glass slides, and stained with 0.2% thionine. The brains were visualized and photographed with an optical microscope.

Drugs and Reagents

Serotonin hydrochloride, histamine dihydrochloride and thioperamide maleate were obtained from Sigma-Aldrich (St. Louis, MO, USA) at reagent quality and used without purification. SKF 91488 dihydrochloride was obtained from Tocris Biosience (Ellisville, MO, USA) and was delivered at a high dose to ensure permeation across blood brain barrier and cause robust in vivo effects (50 mg kg⁻¹). Drugs were dissolved in saline and were injected intraperitoneally at a volume of 0.6 ml kg⁻¹.

Data Analysis

Kinetic characterization of 5-HT release and uptake was adapted from techniques previously used to describe kinetics of release and uptake of the dopamine system (Garris and Wightman, 1994). Release was described as [5-HT]_p*f, where [5-HT]_p is the amount of 5-HT released per stimulation pulse, and f is the frequency of stimulation pulses. The rate of change during stimulation is:

$$\frac{d[5-\mathit{HT}]}{\mathit{dt}}=\left({[5-\mathit{HT}]}_p\ast f\right)-{\left\{\frac{d[5-\mathit{HT}]}{\mathit{dt}}\right\}}_{\mathit{uptake}}$$ (1)

in which the duration of the release term is determined by the number of pulses in the stimulation. Only the uptake term dominates after the stimulation terminates. The uptake rate of 5-HT from the extra-cellular space (v) following electrically stimulated release was assumed to follow the Michaelis-Menten equation:

$$v={\left\{\frac{d[5-\mathit{HT}]}{\mathit{dt}}\right\}}_{\mathit{uptake}}=\frac{V_{\mathit{max}}}{\frac{K_m}{[5-\mathit{HT}]}+1}$$ (2)

V_max is the maximal rate of uptake, and K_m is the Michaelis-Menten constant that describes the affinity of the 5-HT transporter for the 5-HT molecule. It was taken to be 170 nM, a value found in rat brain synaptosomes (Shaskan and Snyder, 1970; Mosko et al., 1977).

In all simulations, analyte diffusion through a thin layer (300 nm) of Nafion was accounted for (Kristensen et al., 1987). The amount of 5-HT released per stimulation pulse ([5-HT]_p) and K_m and V_max were determined by fitting the model to the experimental data. In experiments involving transport inhibition, V_max was fixed to values determined in pre-drug models.

Student's t-tests were performed on paired data sets, p<0.05 was taken as significant.

Flow Injection Analysis

For experiments characterizing histamine and 5-HT in vitro, flow injection analysis was used (Kristensen et al., 1986). The carbon-fiber microelectrode was placed in the output of a six-port HPLC loop injector mounted on a two-position actuator (Rheodyne model 7010 valve and 5701 actuator), operated by a 12 V DC solenoid valve kit (Rheodyne, Rohnert Park, CA). The apparatus enabled the introduction of a rectangular pulse of analyte to the microelectrode surface using a syringe infusion pump (Harvard Apparatus model 940, Hollison, MA) at a flow rate of 2 ml min⁻¹.

For in vivo experiments, the recording electrode was used to make a lesion in the tissue at the end of all experiments to verify its placement histologically. The high voltage across the working electrode used to achieve this necessarily over-oxidizes the carbon surface altering its sensitivity. Therefore, post-calibrations would not be a reliable measure of the electrode response. Rather, pre-calibrations were used to obtain a calibration curve, as described previously (Hashemi et al., 2009).

Results

Comparison of 5-HT following DRN or MFB stimulation

Histology verified the location of the stimulating and carbon-fiber microelectrodes in coronal slices of brains used in in vivo experiments (Figure 1). The stimulating electrode was in the MFB and the carbon-microelectrode was in the SNR as indicated by the small lesions.

Figure 1. Histology of Stimulating and Carbon-Fiber Electrode Placements in the MFB and SNR.

Left hemisphere is a diagram showing intended placement of (i) stimulation electrode in the MFB or (ii) carbon fiber microelectrode in the SNR at coordinates described in the methods section. Right hemisphere shows actual placement in a representative brain.

We have previously shown that electrical stimulation of the DRN releases 5-HT in the SNR (Hashemi et al., 2009). In Figure 2, we compare this type of electrically stimulated 5-HT release (Figure 2A) to 5-HT release elicited via MFB stimulation (Figure 2B). The horizontal dashed lines in panel Figure 2A(iii) are at the peak potential for 5-HT oxidation (0.65 V) (1) and oxidation of an additional substrate (0.85 V) (2). The currents at these potentials were converted to concentrations and are plotted directly above this ((i) 5-HT, (ii) additional substrate). In (i) the current begins to rise at stimulus initiation and peaks within 0.5 seconds of the stimulation termination. In Figure 2A(ii), there is a small increase in current in response to the stimulation. The vertical dashed line in the color plot (iii) at the end of the stimulation and the current at this time were used to construct the cyclic voltammogram (inset). This cyclic voltammogram is identical to those obtained in vitro for 5-HT where the presence of the reverse wave is characteristic of 5-HT. Similar results were obtained with MFB stimulation (Figure 2B). In Figure 2B(i), the 5-HT current profile closely matches the 5-HT current profile in Figure 2A(i), however the current obtained upon stimulation in Figure 2B(ii) is 3-fold greater with MFB stimulation that with DRN stimulation. The cyclic voltammogram (inset α) is taken at the vertical dashed line at the end of the stimulation. An additional vertical dashed line (β) is taken two seconds after the stimulus termination, the current at this time was used to construct the cyclic voltammogram in inset β. Panel C shows the averaged maximal release amplitudes of (1) 5-HT and (2) the additional substrate as the stimulating electrode is lowered down the dorsal/ventral tract above the MFB (n=6 ± SEM). For each line, the data are normalized to the maximal signal along the stimulation track. When two points are statistically different, the p value is added.

Figure 2. 5-HT Color Plot, CV and Time Profile of Response in SNR with DRN and MFB Stimulation.

(A) and (B) - Comparison of signals obtained in SNR obtained following DRN and MFB electrical stimulation respectively. Concentrations vs. time taken at the horizontal white dashed lines are shown above the color plots in panels (i) and (ii), and representative cyclic voltammograms (vs. Ag/ Ag Cl) taken at the vertical white dashed lines are inset. Bipolar stimulation onset at 5 seconds is 350 μA, 60 Hz and 120 pulses, represented by blue bar. (C) shows averaged, normalized maximal release amplitude of 5-HT (1) and the additional substrate (2) at 0.5 mm dorsal/ventral intervals of the stimulating electrode (n=6 ± SEM). Where the two values are statistically different, the p value is added.

The concentration changes extracted from the current at 0.65 V exhibit a time-course in the SNR very similar to the DRN stimulation. DRN and MFB averaged maximal concentrations in the SNR are statistically identical: 12.7 ± 1.6 nM (n=6 ± SEM) 5-HT with DRN stimulation (Hashemi et al., 2009) and 12.8 ± 1.0 nM (n=6 ± SEM) with MFB (p=0.29). Similarly, the time for 5-HT to be cleared to one half of its maximal value (t_1/2) is 1.7 ± 0.3 s with DRN stimulation (n=6 ± SEM) and 1.8 ± 0.2 s with MFB stimulation (n=6 ± SEM) (p=0.45).

Examination of the color plot reveals additional features around the time of the stimulation. An oxidation process at 0.85 V on the reverse scan accompanied by a reduction process at 0.1 V are simultaneous with the 5-HT signal. Because these processes disappear rapidly after the stimulation, a cyclic voltammogram collected at 2 s after the stimulation (inset β) has the characteristic 5-HT cyclic voltammogram. However, when it is recorded at the end of the stimulation (inset α), it has additional features. Because 5-HT is oxidized at a potential that is reached before the additional species, the presence of the second species does not distort the time-course of 5-HT detection.

Characterization of Histamine Cyclic Voltammetry

The additional oxidation and reduction occur on the backwards oxidation scan and forwards reduction scan. This indicates a kinetically limited reaction that is dependent on the initial sweep that regenerates the electrode's carbon surface (Takmakov et al., 2010). The SNR is populated with histaminergic terminals (Panula et al., 1989) and histamine displays this type of kinetically limited electrochemistry as demonstrated by following the direction of the potential sweep with the arrows in Figure 3. This figure compares in vitro responses at a carbon-fiber microelectrode, using the 5-HT waveform (see above) of (A) 5-HT (500 nM), (B) histamine (20 μM), (C) 5-HT + histamine (500 nM + 20 μM, respectively) to (D) the in vivo response in the SNR during MFB stimulation. Panel (i) shows the color plots and panel (ii) shows the cyclic voltammograms taken at the white dashed vertical lines. The microelectrode's response to histamine (20 μM) injected alone is 7.9 ± 0.9 nA (n=4 ± SEM) and to histamine (20 μM) injected with 5-HT is 7.5 ± 0.7 nA (n=4 ± SEM)(p=0.6). The features of the in vivo response (D) are clearly electrochemically reproduced via a 5-HT + histamine mixture (C).

Figure 3. Comparison of In Vivo Signal to a Mixed In Vitro 5-HT and Histamine Signal.

In vitro injections of (A) 5-HT (500 nM), (B) histamine (20 μM), or (C) 5-HT (500 nM) + Histamine (20 μM), and (D) in vivo response shown in (i) color plot format and (ii) cyclic voltammograms. Current is represented in false color and is a dual scale (left for in vitro and right for in vivo). Representative cyclic voltammograms were taken at the white dashed lines; in (D), the cyclic voltammogram was taken at 0.5 s after stimulation duration, indicated by the white dashed line. This position best captures both substrates.

Figure 4 shows how the histamine signal changes with the applied waveform. Figure 4A shows a log-log trace of the current response (n=5 ± SEM) in vitro to flow injections of histamine (20 μM) as a function of increasing the scan rate of the detection waveform. It can be seen that the histamine response is proportional to the scan rate; the slope of the log-log plot is 0.56 indicating diffusion driven electrochemistry. In Figure 4B, the peak amplitude for histamine oxidation is shown as a function of increasing the positive limit of the applied voltage. Increasing the potential window of the detection waveform causes an increase in the histamine response from 6.6 ± 2.2 nA (n=5 ± SEM) at 1.2 V to 13.4 ± 3.7 nA (n=5 ± SEM) at 1.3 V (p=0.01). This behavior has been seen previously for dopamine (Hafizi et al., 1990). At the same time, the peak on the reverse scan occurs at more positive potentials as the scan limit is increased, shown by the inset cyclic voltammograms.

Figure 4. Effect of Varying Scan Rate and Upper Potential Limit on Histamine Response In Vitro.

(A) Log-log current response to in vitro injections of histamine (20 μM) as the waveform scan rate is increased from 100 V s⁻¹ to 2000 V s⁻¹ with conventional 5-HT waveform (−0.2 – 1.0 V) (n=5 ± SEM), where the line of best fit is a linear regression. (B) Current response to in vitro injections of histamine (20 μM) as the upper potential window is increased from 0.9 V to 1.4 V (n=5 ± SEM) vs. Ag / Ag Cl. Representative cyclic voltammograms are inset at 1.0, 1.3 and 1.4 V.

Responses to Inhibition of Histamine N-methyltransferase

Unlike most biogenic amine neurotransmitters, histamine is not thought to be inactivated by a specific transporter (Brown et al., 2001). Instead, it is primarily inactivated by methylation, a process that is catalyzed by histamine N-methyltransferase. To pharmacologically verify the species detected at more positive potentials as histamine, an inhibitor of histamine N-methyltransferase, SKF 91488 (50 mg kg⁻¹) was administered. SKF 91488 is a potent, non-competitive inhibitor of histamine N-methyl transferase that is inactive at histamine receptors (Beaven and Shaff, 1979). Panel (i) of Figure 5 displays the effects of this agent in one rat. Figure 5A shows the average color plot of 5 control stimulations, taken 10 minutes apart. Figure 5B shows the average color plot of 5 stimulations one hour after drug administration, taken 10 minutes apart. Horizontal white dashed lines α and β were used to construct concentration vs. time plots of histamine (α) and 5-HT (β) which were then averaged for 5 animals and displayed in panel (ii) (n=5 ± SEM). The control responses are shown in blue and the drug responses are shown in black. Electrically stimulated histamine release was 2.8 ± 1.3 μM (n=5 ± SEM). After drug administration, the t_½ for histamine was significantly increased from 4.1 ± 0.9 s to 10.4 ± 2.1 s (p=0.03) with no change in amplitude. The maximal evoked concentration of 5-HT decreased from 9.3 ± 1.0 nM to 4.6 ± 0.4 nM (p=0.01) without a change in its t_½ value. Figure 5 (iii) shows the effects of SKF 91488 on 5-HT release amplitude in control rats (A) (n=5 ± SEM) and in rats pre-treated with thioperamide (10 mg kg⁻¹) (n=6 ± SEM), a selective H-3 receptor antagonist (B). In rats pre-treated with thioperamide, 5-HT release is not significantly affected after SKF 91488 administration; the control response was 10.2 ± 1.3 nM and 10.4 ± 1.5 nM after SKF 91488 administration (p=0.76).

Figure 5. Pharmacological Characterization of Histamine with SKF 91488.

(i) Color plots showing electrically stimulated response in the SNR in (A) control and (B) 1 hour after intra-peritoneal SKF 91488 (50 mg kg⁻¹) administration. (ii)(α) Averaged concentration vs. time responses taken at oxidation potential α in control (blue) and 1 hour after intra-peritoneal SKF 91488 (50 mg kg⁻¹) administration (black) (n=5 ± SEM). (ii)(β) Averaged concentration vs. time responses taken at oxidation potential β in control (blue) and 1 hour after intra-peritoneal SKF 91488 (50 mg kg⁻¹) administration (black) (n=5 ± SEM). The blue bars represent the durations of the stimulus (350 μA, 60 Hz and 120 pulses) and error bars are SEM. (iii) Averaged maximal 5-HT release concentraion in control rats (A) and rats pre-treated with thioperamide (10mg kg⁻¹) (B). The bars are averages of the maximal 5-HT release in control stimulations (blue) and after SKF 91488 (50 mg kg⁻¹) administration (n=6 ± SEM).

Modeling 5-HT Release and Uptake

Figure 6 compares experimental and modeled 5-HT data between the two stimulations. The experimental data (blue dots) are superimposed on the model (black line). The model is a composite of linear release and uptake governed by Michaelis-Menten kinetics (Wightman et al., 1988). The concentration of 5-HT released per stimulation pulse, [5-HT]_p, and the maximal uptake velocity, V_max, were adjusted for the best fit while the value of the K_m for uptake was fixed at 170 nM. During DRN stimulation, the model yields [5-HT]_p of 1.5 nM and a V_max of 0.63 μM s-1, and during MFB stimulation, the model yields [5-HT]_p of 1.1 nM, with a V_max of 0.67 μM s-1.

Figure 6. A Comparison of 5-HT Experimental Data and Kinetic Analysis Between DRN and MFB Stimulation.

Averaged, normalized experimental data evoked by a 350 μA, 60 Hz and 120 pulse stimulation of the DRN (A) and the MFB (B) (blue dots, n=6 ± SEM). Kinetically modeled data (black line) is superimposed and kinetic parameters V_max and [5-HT]_p are inset when K_m is fixed at 170 nM. Blue box indicates duration of stimulation and error bars are SEM.

Discussion

Electrical MFB stimulation evokes 5-HT Release in the SNR

The efferent circuitry of the DRN to the SNR is well-documented (Dray et al., 1976; Fibiger and Miller, 1977; Wirtshafter et al., 1987; Corvaja et al., 1993), as are its projections to the striatum (Miller et al., 1975; Steinbusch et al., 1980; Steinbusch et al., 1981). Some studies, including a retrograde labeling study by Van der Kooy et al., show that the same 5-HT projection branches to both the SNR and the striatum (van der Kooy and Hattori, 1980; Imai et al., 1986). Striatal bound 5-HT fibers travel within the MFB (Miller et al., 1975). If axonal stimulation in the MFB could be used in place of DRN stimulation, certain experimental challenges associated with DRN stimulation can be avoided. These challenges arise primarily from the small size, inaccessible location and complicated topography of the DRN (Whishaw et al., 1977; Dib, 1994). However, before this work, it had not been established whether electrical stimulation of the MFB at a location anterior to the terminal site (SNR) would release 5-HT. We had previously reported a similar phenomenon: the release of an unidentified substance in the red nucleus terminals upon anterior stimulation of the MFB (Kita et al., 2009). We suspected this substance was 5-HT, but since we had not yet optimized our electrochemical detection method with Nafion, we were unable to electrochemically verify this. Due to our recent technological advances for in vivo 5-HT detection (Hashemi et al., 2009), we are able to study whether this phenomenon applies to 5-HT release in the SNR with MFB stimulation.

In Figure 2, we compare the terminal output in the SNR upon DRN and MFB stimulation. Comparison of the inset cyclic voltammograms electrochemically confirms that 5-HT is released upon MFB stimulation. However another substrate is present in the MFB cyclic voltammogram in Figure 2 (B). The main reductive process of 5-HT is delayed with both DRN and MFB stimulation and indeed in in vitro (inset β). However, the reduction associated with the other substrate occurs simultaneously with the oxidation (inset α) showing this substrate is electrochemically independent from 5-HT. While the second substrate overlaps with the voltammograms for 5-HT, it does not interfere with the concentration of 5-HT obtained at 0.65 V because its electrolysis peaks occur later in the cyclic voltammogram. We have previously found somatodendritic dopamine release in the VTA upon MFB stimulation (Kita et al., 2009), however due to the low levels of dopamine in the SNR (Heeringa and Abercrombie, 1995) and the low sensitivity of the 5-HT specific waveform to dopamine (Hashemi et al., 2009), it is unlikely that there is a significant contribution from dopamine to this signal. Moreover, we had previously shown that the signal in the SNR is unresponsive to pharmacological manipulations by GBR 12909 when we stimulate the DRN (Hashemi et al., 2009). We also confirmed this to be true for MFB stimulation (data not shown).

Histamine is Simultaneously Released with 5-HT in the SNR during MFB Stimulation

The oxidation and reduction peaks of the additional substrate occur on the backward sweep of the oxidation scan and the forward sweep of the reduction scan. Such features have been previously seen in fast-scan cyclic voltammetry (Pihel et al., 1995), and they are indicative of an electrochemical process that requires electrochemical cleaning of the electrode prior to its oxidation. Possible candidates based on these electrochemical characteristics are histamine, adenosine, or hydrogen peroxide (Pihel et al., 1998; Cechova and Venton, 2008; Sanford et al., 2010). All three of these compounds are oxidized at very positive potentials and could be released by electrical stimulation of the medial forebrain bundle. We have previously shown that 5-HT itself can have distorted peaks when the electrode surface is fouled, but Nafion electrodeposition employed in our technique minimized this effect (Hashemi et al., 2009). Additionally, some groups have found `switching' artifacts that occur at the switching potential of the waveform (Bull et al., 1990). The small current obtained at 0.85V in Figure 2A (ii) may have contributions by such a switching artifact at around 1 V, however, the equivalent MFB stimulation causes a much larger current at 0.85 V (Figure 2B(ii)).

Collectively, the evidence presented in this paper indicates that histamine is the substance detected along with 5-HT. Histaminergic projections originate from the tuberomammillary nucleus (TM), located immediately ventral to the MFB, and traverse the MFB (Garbarg et al., 1974; Garbarg et al., 1976; Kohler et al., 1985; Panula et al., 1989). One of the target regions is the SNR where dense populations of histamine containing nerve terminals and histaminergic receptors are found (Panula et al., 1989; Schwartz et al., 1991). High amounts of histamine and histidine decarboxylase, the enzyme responsible for histamine synthesis from histidine, have been chemically verified in the SNR using microenzymatic methods (Pollard et al., 1978). Furthermore, the H-3 receptors present in the SNR (Cumming et al., 1991; Pollard et al., 1993; Anichtchik et al., 2000) have been shown to modulate GABA and 5-HT dynamics (Garcia et al., 1997; Korotkova et al., 2002; Threlfell et al., 2004). Thus, histamine release in the SNR following MFB stimulation is an anticipated result. Given the close locality of these two locations (MFB and TM), it is not possible to selectively stimulate the TM without activating the MFB. Thus, to gain more insight into the nature of the stimulation, we recorded maximal 5-HT and histamine release in the SNR while we lowered the stimulating electrode down the dorsal/ventral tract above our MFB coordinates in 0.5mm intervals from 5.5 – 10mm. Figure 2C shows that the maximal response for both 5-HT and histamine are at D/V 8.5 mm (n=6 ± SEM) which is not surprising since this is the dorsal/ventral location of the MFB. However, the histamine response significantly peaks earlier than the 5-HT response dorsal to this location. This implies that the histaminergic fibers have a broader and independent fiber distribution through the MFB than the 5-HT fibers. In contrast to the MFB, the DRN has very little histaminergic innervation (Panula et al., 1989), explaining why stimulation of this region causes 5-HT release in the SNR with little or no histamine release.

Further evidence that the additional species is histamine comes from our electrochemical and pharmacological experiments. First, as shown in Figure 3, the cyclic voltammetry of the unknown species closely resembles that for histamine in pH 7.4 solution. Second, administration of the highly selective ligand, SKF 91488, which potently inhibits histamine N-methyltransferase (Beaven and Shaff, 1979), significantly increased the clearance time of the signal in comparison to the control with no effect on the signal amplitude (Figure 5 (ii)(α)). This drug also caused a significant decrease in the release amplitude of 5-HT after drug administration (β). Prior work has shown that activation of H-3 receptors inhibits 5-HT release (Threlfell et al., 2004). Since SKF 91488 prolongs histamine's life-time in the extracellular space, this would result in increased activation of H-3 receptors and suppression of 5-HT release. We confirm this by pre-treating rats with thioperamide (10 mg kg⁻¹), a highly selective H-3 receptor antagonist (Arrang et al., 1987) prior to SKF 91488 administration. Figure 5 (iii) shows that thioperamide pre-treatment significantly arrests the SKF 91488 mediated inhibition of 5-HT release (n=6 ± SEM).

Thus, on the basis of electrochemical, anatomical, physiological, and pharmacological evidence, we assign the second signal simultaneously released upon MFB stimulation to histamine.

The Dynamic Rate of Electrically Stimulated 5-HT Release is Independent of DRN or MFB Stimulation

To confirm that MFB and DRN stimulations evoke release from a similar population of terminals in the SNR, we compared 5-HT signals obtained with each stimulation and compared kinetic parameters based on a model previously used to characterize 5-HT release in brain slices containing the SNR (Bunin and Wightman, 1998). Figure 6 illustrates that there are no significant differences in [5-HT]_p or V_max in the SNR with stimulation of the DRN or MFB. Furthermore, the calculated values of V_max in the SNR with both types of stimulation (0.63 and 0.67 μM s⁻¹, respectively) agree with ones previously established in tissue slice preparations (0.57 μM s⁻¹) (Bunin et al., 1998). Since V_max is a function of the number of active transporters, this term is expected to be the same in slices and in vivo. Thus, our finding of similar maximal uptake rates for 5-HT is consistent with removal of 5-HT by a transporter in each preparation.

However, these quantitative modeling studies reveal that 5-HT release evoked in vivo is much smaller than evoked in tissue slice preparations ([5-HT]_p = 55 nM in slices containing the SNR (Bunin et al., 1998) but 1.5 nM in vivo with MFB stimulation). This large difference in release amplitude may have contributed to the previous physical and analytical challenges of in vivo 5-HT voltammetric detection. It would appear that the local stimulations that were employed in the tissue slice preparation, bypassed the regulatory mechanisms that tightly control 5-HT release in vivo. This suggests that intact in vivo physiological mechanisms, combining both cell body and terminal feedback, may profoundly regulate 5-HT release. This regulation may be at the level of 5-HT autoreceptors (Blier et al., 1998; Daws et al., 2000), release of different vesicular pools (Pellegrino de Iraldi, 1992), or activation of other inhibitory mechanisms (Invernizzi et al., 2007). These regulatory mechanisms, in combination with the fact that there is no known transporter for histamine, may also explain why the magnitude of histamine release is far greater than 5-HT release. These mechanisms are the focus of our ongoing studies.

In conclusion, this work shows that 5-HT release can be evoked by stimulating a collateral 5-HT projection in the MFB that branches to both the SNR and the striatum. We verified electrochemically, anatomically, physiologically and chemically that the additional species present upon MFB is histamine. We confirmed that MFB stimulation evokes 5-HT release that is quantitatively equivalent to that evoked by DRN stimulation; however, in both models, release is profoundly less than in tissue slice culture experiments. This highlights the importance of in vivo methods in exploring the mechanisms of 5-HT.