Astrocytes in the Ventral Tegmental Area Are Involved in Cotinine Self-Administration in Male Wistar Rats

TL;DR

This study shows that astrocytes in the ventral tegmental area may contribute to the reinforcing effects of cotinine, a nicotine metabolite, by influencing dopamine and glutamate levels.

Contribution

The study identifies a novel role for VTA astrocytes in cotinine reinforcement, which was previously unexplored.

Findings

Cotinine self-administration increases GFAP protein levels specifically in the ventral tegmental area.

Fluorocitrate in the VTA reduces cotinine self-administration and lowers extracellular dopamine and glutamate levels.

These findings suggest that VTA astrocytes may regulate cotinine reinforcement through extracellular neurotransmitter transmission.

Abstract

What are the main findings?
Cotinine self-administration elevates GFAP protein levels in the VTA.Fluorocitrate in the VTA attenuates cotinine self-administration and reduces extracellular glutamate and dopamine levels.

Cotinine self-administration elevates GFAP protein levels in the VTA.

Fluorocitrate in the VTA attenuates cotinine self-administration and reduces extracellular glutamate and dopamine levels.

What are the implications of the main findings?
These results suggest astrocytes may play an important role in cotinine self-administration.These findings may enhance our understanding of nicotine reinforcement.

These results suggest astrocytes may play an important role in cotinine self-administration.

These findings may enhance our understanding of nicotine reinforcement.

Background: Our recent studies indicate that astrocytes in a key mesocorticolimbic region play an important role in nicotine reinforcement. Nicotine self-administration elevated the astrocyte marker glial fibrillary acidic protein (GFAP) in the nucleus accumbens (NAc) core. Metabolic inhibition of astrocytes in the NAc core with fluorocitrate attenuated nicotine self-administration and disrupted local extracellular glutamate and dopamine transmission. Cotinine is the major neuroactive metabolite of nicotine, demonstrating its own reinforcing effects and contributing to the development of nicotine reinforcement. Mechanisms underlying cotinine reinforcement remain underexplored. The objective of this study was to investigate the potential involvement of astrocytes in cotinine reinforcement. Methods: GFAP protein expression was measured in key mesocorticolimbic regions with a Western blot following chronic cotinine self-administration. The effects of fluorocitrate on cotinine self-administration and extracellular glutamate and dopamine levels were determined. Results: GFAP protein levels were higher in rats undergoing chronic cotinine self-administration than in those with saline self-administration within the ventral tegmental area (VTA) but not the nucleus accumbens or the medial prefrontal cortex. Intra-VTA microinjection of fluorocitrate inhibited the maintenance of cotinine self-administration. Perfusion of fluorocitrate in the VTA reduced local extracellular levels of glutamate and dopamine. Conclusions: These results indicate that cotinine self-administration augmented GFAP expression in the VTA and that metabolic inhibition of VTA astrocytes attenuated cotinine self-administration and impaired extracellular dopamine and glutamate transmission. Overall, these findings suggest that astrocytes in the VTA may play an important role in cotinine reinforcement, potentially through regulation of local extracellular glutamate and dopamine transmission.

Figures

Full text

1. Introduction

Tobacco use disorder remains a prevalent psychiatric disorder and is the leading cause of preventable morbidity and mortality [1,2]. Nicotine is the primary addictive component in tobacco [3,4]. After absorption, nicotine undergoes extensive metabolism, and 70–80% of it is converted to cotinine, its main metabolite [5]. Research has shown that cotinine produces its own pharmacological and behavioral effects and may contribute to various effects of nicotine [6,7,8,9]. Therefore, elucidating mechanisms underlying cotinine’s actions may enhance understanding of nicotine’s effects.

Our recent studies suggest that cotinine may play an important role in nicotine reinforcement. Cotinine supported intravenous self-administration (IVSA) in rats, and reliable cotinine IVSA produced physiologically relevant plasma cotinine levels comparable to typical levels in smokers [10]. Rats with a history of cotinine IVSA displayed relapse-like cotinine-seeking behaviors [11]. Methoxsalen, a potent inhibitor of nicotine metabolism, reduced plasma cotinine levels and inhibited acquisition of nicotine IVSA; enhancement of plasma cotinine levels attenuated the inhibitory effects of methoxsalen and restored acquisition of nicotine IVSA [12].

Cotinine is a weak agonist of nicotinic acetylcholine receptors (nAChRs) and is several orders of magnitude less potent than nicotine [6,13,14]. Our studies indicate that the nonselective nAChR antagonist mecamylamine and the partial α4β2* nAChR agonist varenicline reduced nicotine, but not cotinine, IVSA in rats [10]. Bupropion, a first-line smoking cessation medication, functions as a monoamine reuptake inhibitor and a nAChR antagonist and has been shown to reduce nicotine IVSA in rodents [15,16,17]. However, bupropion did not significantly alter cotinine IVSA in rats [18]. These results suggest that nAChRs may be differentially involved in nicotine and cotinine reinforcement. On the other hand, our studies indicate that the reinforcing effects of nicotine and cotinine may involve a similar cellular mechanism. Cotinine activated the mesolimbic dopamine pathway and increased extracellular dopamine levels in the nucleus accumbens (NAc) [18,19]. Chronic cotinine exposure induced neuroadaptations within the mesolimbic dopamine system [19]. These effects are qualitatively similar to those induced by nicotine [20]. Taken together, mechanisms underlying cotinine reinforcement are complex and warrant more research.

Our recent study indicates that astrocytes may play an important role in nicotine reinforcement. Astrocytes are one of the most abundant glial cells in the brain and provide both anatomical and functional support to neuronal networks [21]. Our study demonstrates that nicotine IVSA enhanced the protein levels of glial fibrillary acidic protein (GFAP), an important astrocyte marker, in the NAc core but not in other mesocorticolimbic regions, including the NAc shell, medial prefrontal cortex, or ventral tegmental area (VTA). In addition, metabolic inhibition of astrocytes in the NAc core with fluorocitrate, an astrocyte toxin, inhibited nicotine IVSA and cue-induced reinstatement of nicotine seeking [22]. Fluorocitrate enhanced extracellular dopamine levels but reduced extracellular glutamate levels in the NAc core, suggesting that dopamine and glutamate transmissions may contribute to the behavioral effects of fluorocitrate on nicotine reinforcement [22]. Given the role of cotinine in nicotine reinforcement, it is possible that cotinine’s effects may also involve astrocytes.

The objective of the current study was to investigate a potential role of astrocytes in cotinine reinforcement. Experiments were performed to determine (a) effects of cotinine IVSA on GFAP expression in key mesocorticolimbic regions, (b) impact of fluorocitrate on cotinine IVSA, and (c) effects of fluorocitrate on extracellular dopamine and glutamate transmission. The hypothesis to be tested was that astrocytes in key mesocorticolimbic regions would play an important role in cotinine reinforcement.

2. Materials and Methods

2.1. Animals

A total of 39 young adult male Wistar rats were acquired at the age of ~8 weeks from Inotiv and housed in a room maintained on a reversed light–dark cycle with light on at 9:30 p.m. and off at 9:30 a.m. Rats were housed in groups of 2–4 rats per cage upon arrival. After acclimation for about one week, the rats received surgery and were housed individually afterwards. Cages were enriched with nestlets and a polycarbonate play tunnel. Food and water were available ad libitum except during the IVSA period when light food restriction was applied to maintain rats at ~85–90% of free-feeding weight. Protocols used were approved by the Institutional Animal Care and Use Committee at Pennsylvania State University College of Medicine. All experiments were performed in accordance with the principles outlined in the Guide for the Care and Use of Laboratory Animals [23].

2.2. Chemical Agents

Fluorocitrate, (-)-cotinine, dopamine, glutamate, ascorbic acid, sodium chloride, potassium chloride, calcium chloride, magnesium chloride, sodium sulfite, and ethylenediaminetetraacetic acid disodium salt (EDTA) were obtained from MilliporeSigma (St. Louis, MO, USA). Citric acid monohydrate was from Alfa Aesar (Ward Hill, MA, USA). Acetonitrile was from Ricca Chemical Company (Arlington, TX, USA). Methanol was from IBI Scientific (Peosta, IA, USA). Phosphoric acid was from Spectrum Chemical MFG Corp (New Brunswick, NJ, USA). O-phthalaldehyde was obtained from Tokyo Chemical Industry Co. LTD (Tokyo, Japan). Bromophenol blue and 1-octanesulfonic acid (OSA) were purchased from Acros Organics (GEEL, Belgium). Bupivacaine was purchased from Hospira, Inc. (Lake Forest, IL, USA). Carprofen was acquired from Zoetis Inc. (Kalamazoo, MI, USA). Heparin and gentamicin were purchased from Fresenius Kabi (Lake Zurich, IL, USA). Methohexital sodium was from Par Pharmaceutical (Chestnut Ridge, NY, USA). All chemicals were dissolved in distilled water or saline to the desired concentrations.

2.3. Catheterization Surgery and IVSA

Intravenous catheterization surgery was conducted following procedures previously described [10,11,12,19]. Briefly, rats were anesthetized under 2–3% isoflurane inhalation. The right jugular vein was exposed, and an incision was made in the vein. Then, catheters constructed with polyurethane tubing (inner diameter (I.D.) × outer diameter (O.D.) = 0.6 × 1.0 mm; Instech Laboratories, Inc., Plymouth Meeting, PA, USA) were inserted into the vein. The remaining portion of the catheter coursed via a subcutaneous tunnel over the shoulder blade and exited the back of the rat via a 22-gauge cannula. Bupivacaine (0.5%) and carprofen (5 mg/kg) were applied as analgesics. Catheters were flushed daily with ~0.5 mL of saline containing 20 IU/mL heparin and 0.13 mg/mL gentamicin sulfate. Catheter patency was checked once a week with intravenous administration of ~0.1 mL of 10 mg/mL methohexital sodium, a fast-acting barbiturate. Rats were excluded from experiments when the catheters were clogged or failed the patency test.

After 5–7 days of recovery, IVSA started in standard rat self-administration chambers equipped with two levers, one active and the other inactive, counterbalanced (Med Associates Inc., St. Albans, VT, USA). After meeting the pre-set reinforcement schedules, responses on the active lever led to intravenous infusion of either saline or cotinine. These schedules included fixed ratio 1 and 2 (FR1 and FR2) and progressive ratio (PR) during which a step increase of 3 active responses was required for each additional infusion. Cotinine was self-administered at 0.03 mg base/kg/infusion because it was shown to induce optimal IVSA and produce clinically relevant blood cotinine levels in rats [10]. Each infusion was delivered in a volume of 55 µL over 3 s via a syringe pump (PHM-100, Med Associates Inc., St. Albans, VT, USA); then a 17 s timeout followed. During each infusion, the house light was turned off and the cue light above active lever was turned on. During timeout periods, both the cue light and the house light were turned off. Lever presses during the infusion and timeout periods were recorded with no further infusion. Responses on the inactive lever were recorded with no programmed consequences. FR sessions were 2 h and PR sessions were 4 h in duration. Sessions were conducted daily during weekdays. A total of 4 rats were excluded due to loss of catheter patency (2 rats) and failure in acquiring cotinine self-administration (~7 infusions/session) across the last 3 sessions.

2.4. Western Blot

Protein levels of GFAP were examined following the procedures previously described [19,22]. Briefly, rat brains were harvested immediately after the last IVSA session. Frozen tissue was collected via micro-punch from the prelimbic (PL) and infralimbic (IL) cortices, the NAc shell and core, and the VTA. Total protein was extracted with a NucleoSpin^®^ RNA/Protein purification kit (MACHEREY-NAGEL GmbH & Co., Duren, Germany) followed by quantitation of protein content with the Qubit^®^ Protein Assay on a Qubit 4 fluorometer (ThermoFisher Scientific, Waltham, MA, USA). Western blotting was carried out on a ProteinSimple Wes automated Western blot platform (Bio-techne, Minneapolis, MN, USA). A 12–230 kDa separation microplate kit was used with 0.6 µg protein loaded onto the plate. The primary antibody was mouse anti-GFAP antibody MAB360 (1:200; MilliporeSigma, Burlington, MA, USA). Total protein was used for loading control. Densitometric analysis was performed with the ProteinSimple Compass software ver. 5.0.1.

2.5. Stereotaxic Surgery

Guide cannulae were implanted into the brain following the procedures previously described [19,22]. Briefly, rats were anesthetized under 2–3% isoflurane inhalation, and guide cannulae (P1 Technologies, Roanoke, VA, USA) were implanted aimed at the target region. Cannulae at 22-gauge and 18-gauge were used for microinjection and microdialysis, respectively. Stylets were inserted into the guide cannulae. Bupivacaine and carprofen were applied as analgesics during surgery. Following surgery, rats recovered for 3–5 days during which the rats were handled on a daily basis.

2.6. Microinjection

Fluorocitrate was injected following the procedures previously described [22,24]. Briefly, on the day of microinjection, two 28-gauge microinjectors (P1 Technologies, Roanoke, VA, USA) were inserted into the target region. Each injector was connected via PE50 tubing to a 25 µL Hamilton syringe mounted on an infusion pump (Harvard Apparatus, Holliston, MA, USA). A ringer’s solution (147 mM NaCl, 3 mM KCl, 1.2 mM CaCl_2_, 1.2 mM MgCl_2_) or fluorocitrate (0.25 or 1.0 nmol in ringer’s solution) was infused at a volume of 0.5 µL/side over 2 min. After microinjection, the injectors remained in place for an additional two minutes before removal. Microinjection was conducted ~30 min prior to the start of the IVSA sessions. Treatments followed a within-subject design in a random order. Treatment-free sessions were included between treatments to allow responses to return to baseline levels before next treatment. Each concentration of fluorocitrate and vehicle was administered once in each rat.

2.7. Probe Insertion and Microdialysis

Microdialysis probes were inserted as previously described [19,22]. Briefly, rats were anesthetized with 2–3% isoflurane, and microdialysis probes containing a 1.5 mm active membrane (I.D. × O.D. = 200 μm × 216 μm, molecular weight cut-off: 13 kDa, Spectrum Laboratories, Inc., Rancho Dominguez, CA, USA) were slowly inserted into the target region. Microdialysis was conducted ~20 h later. On the day of microdialysis, rats were placed into Plexiglas chambers, and probes were connected via PE20 tubing to a syringe mounted on a Harvard infusion pump. A ringer’s solution (147 mM NaCl, 3 mM KCl, 1.2 mM CaCl_2_, 1.2 mM MgCl_2_, and 0.2 mM ascorbic acid (for dopamine only), pH 7.2–7.4) was perfused through probes at a flow rate of 1.0 µL/min. Baseline samples were collected after a 90 min washout period. Then, fluorocitrate was perfused through probes at 100 µM followed by 250 µM. Each concentration was perfused for 50 min. Following fluorocitrate, the perfusate was switched back to ringer’s solution. Samples were collected at 10 min intervals from one side for dopamine and from the other side for glutamate and were stored at −80 °C for later analysis. A total of 2 rats were excluded due to microdialysis probe blockage during microdialysis.

2.8. Dopamine and Glutamate Analysis

Microdialysis samples were analyzed with an ALEXYS^TM^ Neurotransmitter Analyzer coupled with electrochemical detection (Antec Scientific USA, Boston, MA, USA) as previously described [19,22]. Dopamine and glutamate were analyzed separately. An AS 110 UHPLC autosampler was used for sample loading. For dopamine, samples were loaded onto an Acquity UPLC^®^ BEH C18 column (50 mm × 1.0 mm, 1.7 µm, Waters Corporation, MA, USA) with a mobile phase containing 100 mM phosphoric acid, 0.1 mM EDTA, 100 mM citric acid, 600 mg/L OSA, and 3.5–4.0% acetonitrile at pH 6.0. For glutamate, samples were first processed for in-needle derivatization with o-phthalaldehyde and sulfite via the autosampler. Then samples were loaded onto an Acquity UPLC^®^ HSS T3 column (50 mm × 1.0 mm, 1.8 µm, Waters Corporation, MA, USA) with mobile phase A (50 mM phosphoric acid, 50 mM citric acid, 0.1 mM EDTA, 2% acetonitrile, pH 3.5) for separation followed by mobile phase B (50 mM phosphoric acid, 50 mM citric acid, 0.1 mM EDTA, 50% acetonitrile, pH 3.5) for post-separation elution. A SenCell with a 2 mm glassy-carbon working electrode was used for sample detection. Oxidation potentials were set at +460 mV for dopamine and +850 mV for glutamate. Signals were analyzed with the DataApex Clarity software ver. 8.1.

2.9. Histology

Placements for microinjection sites and microdialysis probes were verified as previously described [19,22]. At the end of experiments, rats were euthanized with CO_2_ overdose, and bromophenol blue dye was injected into the microinjection sites or perfused through microdialysis probes. Brains were quickly removed, frozen, and sliced on a cryostat microtome to 40 µm coronal sections. Placements of microinjection sites and microdialysis probes were determined by locating the bromophenol blue dye with reference to the rat brain atlas of Paxinos and Watson [25].

2.10. Statistical Analysis

Statistics were conducted in SPSS ver. 31.0.0.0 as previously described [12,19]. Time course data were analyzed with linear mixed-effects modeling for repeated measures. These data include lever responses and number of infusions across IVSA sessions and extracellular neurotransmitter levels over time. Significant main effects were followed by multiple comparisons with Bonferroni correction. For Western blotting, densitometric data were first normalized against total protein loading control and then against averages from the saline group. Data normality was analyzed with the Shapiro–Wilk test. Student t tests and Mann–Whitney U tests were performed on data with normal and non-normal distributions, respectively. The significance level was set at p < 0.05.

3. Results

3.1. Effects of Cotinine IVSA on GFAP Expression

GFAP protein levels were examined in rats with a history of IVSA of saline (n = 10) or cotinine (n = 9) at 0.03 mg/kg/infusion. The detailed procedure and IVSA data from these rats have been published elsewhere [10]. Briefly, these rats were trained for IVSA under an FR1 schedule for 15 sessions, an FR2 schedule for five sessions, and a PR schedule for five sessions. Number of infusions per session across different reinforcement schedules are summarized in Figure 1A. Cotinine induced significantly more infusions than saline during self-administration (linear mixed-effects modeling: session: F_24, 408_ = 5.09, p < 0.001; drug: F_1, 17_ = 12.72, p = 0.002; session × drug: F_24, 408_ = 3.751, p < 0.001). GFAP levels are summarized in Figure 1B. Cotinine IVSA significantly increased GFAP expression within the VTA (Figure 1B; t15 = 2.753, p = 0.015) compared with saline IVSA. No significant difference was observed in the NAC shell or core or PL or IL cortices between cotinine and saline IVSA (all p values > 0.05).

3.2. Effects of Intra-VTA Microinjection of Fluorocitrate on IVSA of Cotinine

During IVSA training, rats (n = 6) differentiated the active lever from the inactive lever and elicited more active responses than inactive responses (Figure 2A; linear mixed-effect modeling: session: F_14, 150_ = 2.131, p = 0.013; lever: F_1, 150_ = 54.717, p < 0.001; interaction: F_14, 150_ = 1.081, p = 0.379). The number of cotinine infusions increased at the beginning of the second week and maintained at that high level thereafter (Figure 2A; linear mixed-effect modeling: session: F_14, 70_ = 3.453, p < 0.001).

Intra-VTA microinjection of fluorocitrate at both concentrations reduced the number of cotinine infusions (Figure 2B left panel; linear mixed-effect modeling: F_2, 15_ = 25.027, p < 0.001) and active-lever responses (Figure 2B center panel; linear mixed-effect modeling: F_2, 15_ = 20.344, p < 0.001). Responses on the inactive lever were not altered by fluorocitrate (Figure 2B right panel; linear mixed-effect modeling: F_2,_ 15 = 0.051, p = 0.961).

3.3. Effects of Intra-VTA Fluorocitrate Perfusion on Local Extracellular Levels of Dopamine and Glutamate

For glutamate (Figure 3A; n = 8), fluorocitrate perfusion decreased extracellular glutamate levels, with significantly lower glutamate levels observed during both 100 and 250 µM fluorocitrate perfusion compared with baseline levels (Figure 3A; linear mixed-effect modeling: time: F_17, 119_ = 7.039, p < 0.001). For dopamine levels (Figure 3B; n = 7), fluorocitrate perfusion reduced extracellular dopamine levels (Figure 3B; linear mixed-effect modeling: time: F_17, 102_ = 12.75, p < 0.001). Significantly lower dopamine levels were seen during fluorocitrate perfusion at both concentrations compared with baseline levels. Following fluorocitrate perfusion, both dopamine and glutamate levels returned to baseline levels, indicating a reversible effect of fluorocitrate on both dopamine and glutamate.

4. Discussion

Our studies demonstrate that chronic cotinine IVSA elevated GFAP protein levels in the VTA, but not in the NAc core and shell or PL and IL cortices, compared with saline IVSA, suggesting that cotinine’s effect is region-selective. Microinjection of fluorocitrate into the VTA attenuated cotinine IVSA, suggesting that metabolic inhibition of VTA astrocytes may inhibit cotinine reinforcement. Furthermore, local perfusion of VTA with fluorocitrate significantly decreased extracellular levels of glutamate and dopamine, suggesting that metabolic inhibition of VTA astrocytes may impair local extracellular glutamate and dopamine transmission, which may contribute to inhibitory effects of fluorocitrate on cotinine IVSA. Taken together, these results suggest that astrocytes in the VTA may play an important role in cotinine reinforcement.

Our study indicates that cotinine IVSA enhanced GFAP expression in the VTA (Figure 1). These results agree with previous findings that passive exposure to cotinine increased GFAP levels in the frontal cortex and hippocampus in mice [26,27,28]. It is noted that nicotine and cotinine IVSA altered GFAP expression differently; GFAP increased in the VTA with cotinine (Figure 1) but in the NAc core with nicotine [22]. Given that nicotine is mostly metabolized to cotinine, one would expect that nicotine IVSA may also lead to GFAP increase in the VTA. This lack of such an effect remains unexplained. It is noted that blood cotinine levels were ~230 and ~450 mg/mL in rats following similar nicotine and cotinine IVSA, respectively [10]. It is possible that the lower cotinine level after nicotine IVSA may be below the threshold level to increase GFAP expression. Future studies may test this possibility.

Our results suggest that cotinine IVSA induced region-selective alterations in GFAP expression. These region-specific effects are consistent with astrocyte heterogeneity across brain regions. Astrocytes exhibit neural circuit-specific properties including but not limited to composition, morphology, transcriptomic and proteomic signatures, basal metabolism, electrophysiological characteristics, Ca^2+^ signaling, and astrocyte–synapse proximity [29,30]. In addition, an early study reported that brain uptake of cotinine was region dependent in mice with cotinine enriched in the cerebral cortex and basal ganglia compared with the hippocampus and cerebellum [31]. Although these studies are not specific to the regions examined in the current study, it is possible that astrocyte diversity and differential distribution of cotinine may exist within the mesocorticolimbic system, resulting in a region-dependent GFAP response to cotinine IVSA.

The mechanisms underlying cotinine-enhanced GFAP expression remain unknown. A recent study demonstrated that the cotinine-induced increase in GFAP expression in the frontal cortex and hippocampus was attenuated by DhβE, an α4β2* nAChR antagonist, suggesting an involvement of α4β2* nAChRs [28]. Another study suggests that α7 nAChRs may be involved in cotinine’s effects. Repeated cotinine exposure downregulated α7 nAChR binding in the hippocampus [9]. These results suggest that both α4β2* and α7 nAChRs may be involved in cotinine’s effects on GFAP expression. Astrocytes express nAChRs, e.g., α7 nAChRs [32,33], which may mediate cotinine’s effects on GFAP expression. GFAP is one of the most widely examined astrocyte markers, and upregulation of GFAP is frequently associated with an adaptive response to pathological stimuli [34]. The cotinine-induced increase in GFAP in the VTA suggests that cotinine IVSA may alter astrocyte properties, which may be important to the development of cotinine reinforcement.

Fluorocitrate is one of a handful of pharmacological tools that display relative specificity to astrocytes [35]. Fluorocitrate is preferentially taken into astrocytes via astroglia-specific acetate transporters and functions as a reversable inhibitor of aconitase to suppress the tricarboxylic acid cycle, leading to the blockade of ATP production and impairment of astrocyte function [36,37]. Our study demonstrates that intra-VTA administration of fluorocitrate attenuated the maintenance of cotinine IVSA, suggesting that inhibition of astrocyte function may impair cotinine reinforcement. These results are consistent with our recent findings that fluorocitrate inhibited nicotine IVSA [22]. These findings add to the growing literature implicating astrocytes in drug addiction [38,39,40]. Together, these findings enhance our understanding of astrocyte mechanisms underlying drug addiction.

Local perfusion of the VTA with fluorocitrate lowered extracellular glutamate levels in the VTA (Figure 3A). Our previous study shows that fluorocitrate perfusion in the NAc core reduced local extracellular glutamate levels [22]. These results are consistent with both in vivo and in vitro studies demonstrating that fluorocitrate inhibits glutamate overflow in various brain regions and from cultured astrocytes [41,42,43]. It remains unknown whether cotinine alters extracellular glutamate levels in the brain. However, since the mesolimbic glutamate system plays an important role in drug addiction, including nicotine, ethanol, and psychostimulants [44,45], it is possible that inhibition of glutamate transmission in the VTA may contribute to the inhibitory effects of fluorocitrate on cotinine IVSA.

Several mechanisms may contribute to the inhibitory effects of fluorocitrate on extracellular glutamate levels. Astrocytes are the source of the glutamate precursor, glutamine, which is synthesized in astrocytes in an ATP-dependent manner [21]. Fluorocitrate has been shown to inhibit glutamine synthesis and release [41,42]. In addition, astrocytes may release glutamate via putative gliotransmission, and this process is also energy dependent [46]. Therefore, fluorocitrate may impair both glutamate synthesis and gliotransmission to reduce extracellular glutamate levels.

Fluorocitrate also reduced extracellular dopamine levels in the VTA (Figure 3B). Extracellular dopamine in the VTA originates mainly from somatodendritic dopamine release involving a depolarization-dependent process [47]. This process is regulated by various neurotransmitters including glutamate, the main excitatory neurotransmitter. Glutamate innervation contributes to the excitation and depolarization of the VTA dopamine neurons, leading to enhanced somatodendritic dopamine release [47]. Our study shows that fluorocitrate reduced extracellular glutamate levels in the VTA (Figure 3A). Therefore, it is possible that the decrease in extracellular dopamine in the VTA may be secondary to fluorocitrate inhibition of glutamate, leading to a decrease in glutamate innervation of the VTA dopamine neurons and somatodendritic dopamine release. Interestingly, previous studies demonstrated that local perfusion with fluorocitrate elevated extracellular dopamine levels in the NAc core [22] and dorsolateral striatum [48]. Moreover, fluorocitrate was found to reduce extracellular glutamate levels in the NAc core [22]. It remains unknown why fluorocitrate induced opposite effects on extracellular dopamine levels in the VTA and the NAc under the same background in which fluorocitrate reduced extracellular glutamate transmission in both regions. These region-dependent effects on dopamine transmission suggest that some unknown factors other than glutamate may come into play to alter fluorocitrate’s effects in the NAc core. Future studies are warranted to identify these potential factors.

Cotinine has been shown to activate the mesolimbic dopamine system to support IVSA [19]. In addition, microinjection of cotinine into the VTA increased dopamine release in the NAc, suggesting cotinine stimulation of the VTA dopamine neurons [18]. On the other hand, potential effects of cotinine on glutamate transmission remain unknown. Given these, it is possible that fluorocitrate inhibition of somatodendritic dopamine activity in the VTA may attenuate cotinine stimulation of the VTA dopamine neurons, resulting in a reduction in cotinine IVSA.

Fluorocitrate preferentially inhibits astrocytes. Microinjection of fluorocitrate at 1 nmol into the striatum selectively inhibited the astrocytic tricarboxylic acid cycle and disrupted astrocyte ultrastructure without impacting neurons over 4 h [41,49,50]. Perfusion of fluorocitrate via a microdialysis probe into the brain at 1 mM over 4 h caused white and swollen and empty appearance of astrocytes but spared neurons [51]. Fluorocitrate administered in our microinjection and microdialysis studies (1 nmol and 100–250 µM, respectively) is within the range of those used in previous studies, suggesting that fluorocitrate may also selectively alter astrocytes in the VTA. On the other hand, a recent study showed that fluorocitrate at concentrations greater than 250 µM could directly affect neuronal metabolism and synaptic plasticity [52]. Given these, a direct contribution from neurons may not be excluded, especially at higher doses and concentrations of fluorocitrate.

One limitation of the current study is that only male rats were included. The lack of inclusion of female rats may prevent the generalization of current findings to female rats. Nonetheless, our studies indicate that cotinine IVSA enhanced GFAP expression in the VTA and that metabolic inhibition of VTA astrocytes inhibited cotinine IVSA and impaired local extracellular glutamate and dopamine transmission in male rats. These findings suggest that the VTA astrocytes may play an important role in cotinine reinforcement via the potential regulation of extracellular dopamine and glutamate transmission, at least in male rats.