Glucagon-like peptide-1 receptor regulation of basal dopamine transporter activity is species-dependent

1. Introduction

Natural reward and drug reward converge on a common neural pathway, the mesolimbic dopamine system. Since drugs activate the same reward system that underlies food reward¹ it is perhaps no surprise that appetite-regulating peptides, besides governing energy homeostasis, also target brain areas associated with reward and addiction². Glucagon-like peptide-1 (GLP-1) is a peptide that acts both as an incretin hormone that regulates blood sugar and as a neuropeptide in the brain regulating satiety³. Its anorexic and glucoregulatory effects are well-established and GLP-1 receptor (GLP-1R) agonists (GLP-1RA) are approved for clinical use in type 2 diabetes and obesity⁴. Recent years of research have broadened the understanding of the GLP-1 system especially by linking it to the rewarding properties of food^5–7. In the brain, GLP-1Rs are found in areas associated with reward, reinforcement and addiction including the ventral tegmental area (VTA) and striatum^8–10. The VTA and its dopaminergic projections to the nucleus accumbens (NAc) orchestrate motivated behavior to obtain natural reward (such as food and sex) but are prone to be hijacked by artificial reward substances (such as drugs of abuse and alcohol) ultimately leading to addiction¹¹. Importantly, many appetite-regulating hormones including GLP-1 exert a direct influence on the VTA and NAc¹². Taken together these data have led to the hypothesis that the central GLP-1R system modulates the reward system. Indeed, systemic or central administration of GLP-1RAs in rodents and non-human primates attenuate addiction-related effects of alcohol, central stimulants and nicotine^13–20. In addition, gene variants of GLP-1R in humans have been associated with the prevalence of alcohol use disorder (AUD)²¹.

Although GLP-1R activation attenuates the rewarding properties of food, drugs and alcohol, the underlying mechanisms remain largely unknown. We recently suggested a link between GLP-1R activation and the dopamine transporter (DAT) that regulates dopamine (DA) homeostasis in the lateral septum (LS) of the brain¹⁶, and elsewhere. GLP-1Rs are highly expressed in LS, which is also associated with reward²⁵. Brain DA plays a pivotal role in drug addiction and since plasma membrane DAT is essential for terminating DA neurotransmission, regulation of DAT may or may not underlie the observed effects of GLP-1R on reward. Since the striatum serves as a central interface in the reward system²⁷ we wanted to investigate if the GLP-1R system modulates DAT and consequently, levels of synaptic DA.

Here we report data using native GLP-1 (rodents) and GLP-1RA (humans) as tools in three different studies; two studies in rodents and one study in healthy humans. Our hypothesis was supported by the present rat striatal data, but not by data from mice or humans.

2. Materials and Methods

2.1. Experiment 1 - Assessing ex vivo and in vivo effects of GLP-1R on DA homeostasis in striatum in rats

2.1.1. Animals and striatal slice preparation

Experimentally naive male Sprague-Dawley rats (275–300g Charles River) were used. Corticostriatal hemislices (300 pm; hereafter “striatal slices”) were prepared with a vibratome (Leica VT1000S) in an ice cold oxygenated (95% O2 / 5% CO2) sucrose cutting solution consisting of (in mM): 210 sucrose, 20 NaCl, 2.5 KCl, 1 MgCl2, 1.2 NaH2PO4, 10 glucose, 26 NaHCO3. Slices were then transferred to oxygenated artificial cerebrospinal fluid (aCSF) at 28 °C for a minimum of 1 h. The aCSF consisted of (in mM): 125 NaCl, 2.5 KCl, 1 MgCl2, 2 CaCl2, 1.2 NaH2PO4, 10 glucose, 26 NaHCO3. All chemicals were purchased from Sigma unless otherwise indicated. GLP-1 (7–36)-amide was purchased from Phoenix Pharmaceuticals. Drug treatments were performed in oxygenated aCSF at 28 °C unless otherwise indicated. Slices were anatomically matched before assignment for drug treatments.

All experiments were conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and Vanderbilt University’s and University of Texas Health Science Center at San Antonio Institutional Animal Care and Use Committee guidelines.

2.1.2. Dopamine uptake

Striatal slices were treated at 28 °C for 10 min with control aCSF solution or 100 nM GLP-1 (7–36)-amide. Slices were then transferred to a 37 °C chamber with the same drugs plus 50 nM [³H]DA for 10 min. At the end of the [³H]DA treatment, the slices were washed with cold aCSF and the striatum dissected. Tissue samples were then homogenized in 200 μL of 1% Triton lysis buffer consisting of 150 mM NaCl, 25 mM HEPES, 2 mM sodium orthovanadate, 2 mM sodium fluoride, 1 μg/ml leupeptin, 1 μg/ml aprotinin, 1 μg/ml pepstatin, 0.5 mM phenylmethylsulfonyl fluoride, 1% Triton-X100. The homogenate was centrifuged at 13,000 g at 4 °C for 30 min and the supernatant added to 500 μL of 0.1% Triton buffer consisting of 150 mM NaCl, 25 mM HEPES, 2 mM sodium orthovanadate, 2 mM sodium fluoride, 1 μg/ml leupeptin, 1 μg/ml aprotinin, 1 μg/ml pepstatin, 0.5 mM phenylmethylsulfonyl fluoride, 0.1% Triton-100. The protein concentration of each sample was measured and 500 μL of each sample was added to scintillation vials to count [³H]DA. Counts were expressed as a ratio to protein content and normalized to the mean value for the control condition within each experiment.

2.1.3. DAT cell surface expression

Slices were treated at 28 °C for 20 min with control aCSF solution or 100 nM GLP-1 (7–36)-amide and were then washed with cold aCSF to terminate the drug treatment. Biotinylation was performed as previously described²⁸. In brief, slices were incubated with 1 mg/mL Sulfo-NHS-SS-Biotin (Pierce) in oxygenated aCSF on ice for 45 min, then washed and incubated in cold aCSF for 10 min twice. The biotin reaction was quenched by washing twice for 20 min each with cold oxygenated aCSF with glycine. Slices were washed with cold aCSF and the striatum dissected and immediately frozen on dry ice. Tissue samples were then homogenized in 200 μL of 1% Triton lysis buffer and the homogenate was centrifuged at 13,000 g at 4 °C for 30 min and the supernatant added to 500 μL of 0.1% Triton buffer. Protein concentration of each sample was measured and equal amounts of protein for each sample were incubated with ImmunoPure immobilized streptavidin beads (Pierce) overnight at 4 °C under gentle rotation. Beads were washed three times with 0.1% Triton buffer and biotinylated proteins were then eluted in 50 μL of Laemmli sample loading buffer at 95 °C for 5 min and then cooled to room temperature. Biotinylated and total lysate samples were analyzed by SDS- PAGE and western blotted with a DAT primary mouse monoclonal antibody (1:2000 Dr. Roxanne Vaughan, University of North Dakota).

2.1.4. Dopamine clearance

Measurement of DA clearance in vivo was performed using the Fast Analytical Sensing Technology system (Quanteon) as previously described^29,30. Briefly, a carbon fiber electrode (30 pm diameter, 150 pm in length; Specialty Materials) was coated with Nafion 3–4 times baked at 200 °C for 5 min per coat. Electrodes displayed linear amperometric responses to DA during in vitro calibration in phosphate-buffered saline and were ≥ 1000 fold selective for DA over the DA metabolite dihydroxyphenylacetic acid (DOPAC). The electrode was coupled to a glass multi-barrel micropipette at a center-to-center distance of 300 μm. Animals were anesthetized with injections of urethane (850 mg/kg, i.p.) and α-chloralose (85 mg/kg, i.p.), fitted with an endotracheal tube to facilitate breathing, and placed into a stereotaxic frame (David Kopf Instruments). The electrode/micropipette assembly was lowered into the striatum at the following coordinates (in mm from bregma): A/P +1.5; M/L, +/−2.2; D/V, −3.5 to −5.5 and DA pressure-ejected at 5 min intervals. Pressure application of solutions was accomplished using a Picospritzer II (General Valve Corporation) in an ejection volume of 100–150 nl (5–25 psi for 0.25–3 s). DA concentration at the recording electrode was measured by high-speed chronoamperometry. Voltage steps, consisting of 100 ms square pulses of 550 mV each separated by a 900 ms interval during which the resting potential was maintained at 0 mV, were applied with respect to an Ag/AgCl reference electrode implanted into the contralateral superficial cortex. Oxidation signals were converted to DA concentration based on in vitro calibration with DA standards. For each recording session, DA was identified by its reduction/oxidation signal ratio: 0.55–0.80. Electrode placement was confirmed by making an electrolytic lesion at the recording site at the conclusion of the experiment. Following establishment of a stable baseline DA signal, either vehicle or GLP-1 (7–36)-amide (0.25 pmol) was locally microinjected, and two min later DA was applied again. DA clearance rate was quantified by the slope of the linear portion of the current decay curve (20–60% decay amplitude from the maximal signal amplitude). For each experiment, DA clearance rate following vehicle or GLP-1 (7–36)-amide injection was expressed as a percentage of the pre-injection baseline DA clearance rate^31,30.

2.2. Experiment 2 - Assessing the effect of exenatide- and of GLP-1R knockout on DA uptake in striatum in mice

2.2.1. Animals

In Experiment 2, DA uptake was assessed in global GLP-1R knockout (KO) mice, as well as mice treated acutely with the GLP-1RA, exenatide (purchased from Tocris Bioscience). Therefore, two different strains of mice were used for the DA uptake experiments. For the effect of acutely administered exenatide on DA uptake, we used naive male C57BL/6NRj mice acquired at 6 weeks of age (Janvier Labs, France). Mice were allowed to acclimate to the facility for one week prior of initiation of the experiment. For the assessment of the effect of global knockout of endogenous GLP- 1R on DA uptake we used Glp1rflox/flox nestin-Cre+/− KO mice with a neuronal-specific deletion of the GLP-1R, generated as previously described⁷ and bred at the University of Copenhagen, Panum facility. Glp1rflox/flox nestin- Cre−/− littermate mice were used as wild-type (WT) controls. Genotype was determined by polymerase chain reaction on DNA extracted from ear clip samples.

All mice were group-housed with ad libitum water and rodent chow (Altromin 1310, Brogaarden, Denmark), in a temperature- and humidity-controlled room. Lights were on from 07:00 to 19:00, and assessments were performed during the light phase. Study procedures were conducted at the University of Copenhagen Department of Experimental Medicine, Panum facility, in accordance with the EU directive 2010/63/EU and were approved by the Animal Experiments Inspectorate under the Danish Ministry of Food, Agriculture, and Fisheries.

2.2.2. Dopamine uptake

We assessed the role of the GLP-1R on striatal DA uptake in mice by two different experimental approaches. The acute effects were assessed in C57BL/6NRj mice (n = 3 per group) receiving either 3.2 μg/kg exenatide at an injection volume of 10 μL/g or 10 ml/kg 0.9% saline injected intraperitoneally exactly 60 min prior of decapitation of the animal. Striatal tissue was dissected on ice from coronal sections using a brain matrix and a puncher. For assessment of the effect of global knockout of GLP-1R, GLP-1R nestin-Cre KO and WT mice (n = 5) received no treatment before decapitation. In both experiments, synaptosomal DA uptake was assessed as previously described^32,33. In brief, striata were weighed and homogenized in ice-cold HEPES buffer (4 mM HEPES, 0.32 M sucrise, pH 7.4) followed by homogenizations. Pellet was resuspended and 10 μL incubated at 37 °C in 440 μL uptake buffer (25 mM HEPES pH 7.4, 120 mM NaCl, 5 mM KCl, 1.2 mM CaCl2, 1.2 mM MgSO4, 1 mM L-ascorbic acid, 5 mM D-glucose, 1 μM pargyline, 100 nM desipramine and 10 nM cathechol-O-methyl-transferase inhibitor (RO-41–0960, R108 Sigma Aldrich, Germany)) with or without the presence of 500 μM cocaine (used to assess non-specific uptake). Six different concentrations of 2, 5, 6-[3H]-DA (91.1 Ci/mmol, Perkin Elmer Life Sciences, Boston, MA, USA) was added for creation a concentration curve of 5 min DA uptake. A standard BCA™ Protein Assay kit (Thermo Scientific Pierce, Rockford, IL, USA) was used to determine protein concentrations of the synaptosomes for adjustments from counts per min to fmol/min/μg for analysis. Background-subtracted fmol/min/μg protein counts were fitted by Michaelis–Menten kinetics using the median of triplicate determinations for each mouse.

2.3. Experiment 3 - Assessing the effect of exenatide on DAT availability in striatum in healthy volunteers

2.3.1. Participants

Participants were screened within 7 days prior to the experiment. Inclusion criteria were 18–40 years of age, healthy and not taking any medication on a daily basis (except contraceptives), alcohol consumption less than the limit recommended by Danish Health Authority (14/21 units/week w/m) and no heavy drinking days (defined as > 5 units of alcohol in one occasion) for the last week prior to the date of inclusion. None of the participants had any history of physical, neurological, or psychiatric disorders, nor history of or current treatment of any addiction or any active substance abuse/dependence for the past six months assessed by Copenhagen Stimulant Screening Questionnaire. All had an Alcohol Use Disorder Identification Test score < 8 (mean score was 5.8) and were non-smokers. Neurological examination, routine blood and urine tests including plasma glucose, glycaeted haemoglobin A1c (HbA1c) and urine drug test were normal in all participants. The females were excluded if they were pregnant, breast-feeding, had intentions of becoming pregnant, or if they did not use adequate contraceptives. Finally, the participants were excluded if they had been exposed to excessive radioactivity during the last 12 months or had received any investigational drug within the last 3 months.

The participants were recruited through advertising on www.forsoegsperson.dk and they all gave informed written consent. The study was performed in accordance with the ethical standards of the Declaration of Helsinki and was approved by the ethical committee of Copenhagen Capital Region (Protocol number: H-1 2010–109 and Protocol H-B- 2008–024).

2.3.2. Design

The study was an open-label, placebo-controlled, repeated-measures experiment.

2.3.3. Radioligand administration

We used ¹²³I-labeled-N-(3-iodoprop-2E-enyl)-2-β-carbomethoxy-3β -(4-methylphenyl) nortropane (¹²³I-PE2I) (MAP Medical Technologies) which is a highly selective DAT ligand with fast kinetics and high target-to-background ratio in striatum³⁴. Cannulas in each cubital vein and one in the back of the hand were established for tracer administration, blood sampling, and exenatide infusion, respectively. To block thyroid uptake of free radioiodine, all participants received 200 mg potassium perchloride intravenously 10 min before ¹²³I-PE2I injection. A bolus/infusion protocol of 2.7 (i.e. the bolus is worth 2.7 h of infusion) was used to reach a steady state condition³⁵. An average intravenous bolus of 81.8 MBq (range, 71.4–88.6 MBq) of ¹²³I-PE2I was administered, immediately followed by a constant infusion until the end of experiment (3 h and 40 min; mean 111.3 MBq, range, 92.8–124.7 MBq).

2.3.4. GLP-1RA infusion

All participants underwent a 100-min Single-photon emission computed tomography (SPECT) scan. A baseline scan was obtained during the first 40 min during which saline was infused (0.5 ml/min). Without interruption, the scan was continued for 60 min during which exenatide was infused at two different rates to achieve steady state³⁶: 0.2 pmol/kg/min for 30 min followed by 0.1 pmol/kg/min until the end of the scan.

2.3.5. Blood sampling

Venous blood was collected at different time points for measuring plasma radioactivity (10, 20, 30, 60, 80 and 100 min) and exenatide concentration (40, 70, 85 and 100 min). For plasma radioactivity, the blood sample was centrifuged at 2800 rpm 4 °C for 10 min, plasma was collected, and octanol was added to derive the lipophilic phase containing the radioligand parent compound³⁷. The sample was shaken and centrifuged, the octanol phase was pipetted into a counting vial, and radioactivity was measured in a γ -counter (Cobra II; Packard Instrument Co). As a safety measure, we also measured blood glucose levels by finger stick at different time points during the infusion of exenatide (40, 70 and 100 min).

2.3.6. SPECT acquisition and image reconstruction

SPECT imaging was performed with a triple-head IRIX camera (Philips Medical, Cleveland, USA) fitted with low-energy, general all-purpose parallel-holed collimators (spatial resolution 8.5 mm at 10 cm). Mean radius of rotation was 15.5 cm. Each camera covered 120° of the circular orbit. Scans were performed in continuous mode. Reconstruction of the images was performed with a MATLAB 6.5 (MathWorks) based program in 128×128 matrices (2.33 mm pixels and identical slice thickness) using standard filtered back projection with a low pass fourth-order Butterworth filter at 0.3 Nyquist (= 0.64 c-1). The imaging energy window was positioned at 143–175 keV. High-energy photons of ¹²³I penetrated through the lead of the collimator, and Compton scatter in the scintillation crystal caused erroneous counts in the imaging energy window. A second energy window positioned at 184–216 keV was used to correct for these down- scattered photons in the imaging window. Before reconstruction, the projection images of the second energy window were subtracted from the imaging energy window with a weight of 1.1.

2.3.7. Binding Potentials

¹²³I-PE2I binding potential relative to plasma (BP_P) was used as a measure of DAT availability. ¹²³I-PE2I BP_P was calculated as the ratio at steady-state of the concentration of intracerebrally non-displaceable and specifically bound ¹²³I-PE2I, i.e., the concentration of ¹²³I-PE2I in a region of interest, denoted C_T, minus the concentration of intracerebrally ¹²³I-PE2I in a reference region devoid of DAT, denoted C_ref, to the concentration of ¹²³I-PE2I in plasma, denoted C_P:

$$B{P}_P=\frac{C_T-C_{\mathit{ref}}}{C_P}$$

Tracer steady-state was attained 120 min after ¹²³I-PE2I injection and cerebellum was used as a reference devoid of DAT ³⁵. For all participants, BP_P values were calculated in striatum using an algorithm (DATquan) developed in-house. DATquan offers a fast, accurate, and highly reproducible method for semiautomatic ROI delineation and ¹²³I-PE2I quantifications of C_T and C_ref using a template-based approach³⁸.

For statistical comparison we calculated three different BP_P values: a baseline value comprising the first 40 min of scan (saline), and two BP_P values during exenatide infusion comprising 20 min of scan acquired at time intervals 50–70 min and at 70–90 min. Time-corresponding C_P values were used in the calculations, i.e., for the baseline BP_P we used a mean C_P from timepoints 10, 20 and 30 min and for the BP_P’s during exenatide infusion, we used mid-time C_P values at 60 and 80 min.

3. Data analysis

In Experiment 1, DA uptake data were analyzed by one-way ANOVA followed by Bonferroni-Holm post-hoc test. DAT data were analyzed by one tailed paired t-test and DA clearance data were analyzed by Student’s t-test. In Experiment 2, DA uptake data were analyzed by two-way ANOVA with exenatide treatment or genotype as between- subjects factor and DA concentration as repeated-measures factor. In Experiment 3, data were analyzed by one-way ANOVA with BP_P as repeated-measures factor. All data are presented as means ± standard error of the mean (s.e.m) unless otherwise stated and the level of significance was set at p < 0.05. Statistical tests were conducted using Prism software (Experiment 1), Stata software (Experiment 2) and SPSS software (Experiment 3).

4. Results

4.1. Experiment 1 - GLP-1 (7–36)-amide increased DA uptake, DAT expression and DA clearance in rats

Incubating slices with 100 nM GLP-1 (7–36)-amide significantly increased DA uptake relative to aCSF control (figure 1A). Cell surface DAT expression, labeled by biotinylation, was also significantly increased in GLP-1 (7–36)-amide - treated slices compared to vehicle (figure 1B). Using high-speed chronoamperometry (HSCA), we measured the effect of locally microinjected GLP-1 (7–36)-amide in vivo on DA clearance in the striatum. GLP-1(7–36)-amide produced a transient increase in the rate of DA clearance that was maximal 10 min after injection (figure 2A, representative traces). In contrast, intrastriatal injection of vehicle had no effect on DA clearance (figure 2B, representative traces). DA clearance rates were significantly increase by GLP-1 (7–36)-amide (figure 2C).

Figure 1 – Experiment 1: increased DA uptake (A) and DAT expression (B) after treatment with GLP-1 (7–36)-amide.

CAPTION: A, DA uptake into striatal synaptosomes as % of aCSF control in striatal slices incubated with aCSF (Control) or 100 nM GLP-1 (7–36)-amide (n = 20 slices - one-way ANOVA followed by Bonferroni-Holm post-hoc test). B, ratio of surface to total DAT protein as % of aCSF control in striatal slices incubated with aCSF (control) or 100 nM GLP-1 (7–36)-amide (n = 7 slices - one tailed paired t-test).*p < 0.05 vs. aCSF control.

Figure 2 – Experiment 1: accelerated DA clearance after treatment with GLP-1 (7–36)-amide.

CAPTION for figure 2AB: Representative DA oxidation currents, converted to concentration using DA standards, produced by intrastriatal pressure pulse application of exogenous DA. A, shows the effect of local microinjection of GLP-1 (7–36)-amide (0.25 pmol), blue trace, or B, vehicle, red trace, on DA clearance as a function of time, compared to baseline (black dashed traces). Traces are superimposed for ease of comparison.

CAPTION for figure 2C: DA clearance rate was quantified by the slope of the linear portion of the DA concentration time course, from 20 to 60% decay from the maximal DA concentration amplitude. For each treatment (vehicle or GLP- 1 (7–36)-amide), DA clearance rate was expressed as a percentage of baseline DA clearance rate. GLP-1 (7–36)-amide (blue bar) microinjection significantly increases DA clearance rate compared with vehicle (red bar) injection (*p < 0.05 Student’s t-test; n = 6).

4.2. Experiment 2 - No effect of exenatide on DA uptake and no change in DA uptake in GLP-1R KO in mice

Synaptosomal DA uptake did not differ between striatum taken from mice treated with exenatide in vivo relative to saline-treated mice (figure 3A). DA uptake was related to DA concentration [F(5, 20) = 163, p < 0.0001] with no significant effect of exenatide or interaction. Exenatide treatment did not change the calculated maximal DA uptake capacity Vmax (saline: 62.7 ± 0.8, exenatide: 62.1 ± 7.0 fmol/min/μg) or apparent affinity K_M (0.36 ± 0.11, 0.38 ± 0.10 μM). DA uptake in striatum was also comparable between GLP-1R nestin-Cre KO mice and WT mice (figure 3B); effect of DA concentration [F(5, 40) = 15.4, p < 0.0001] ). WT and KO mice had comparable Vmax (WT: 35.96 ± 3.26, KO: 31.93 ± 3.72 fmol/min/μg) and K_M (WT: 0.053 ± 0.019, KO: 0.037 ± 0.019 μM).

Figure 3 – Experiment 2: no changes in DA uptake after treatment with exenatide in C57BL/6NRj mice (A) or in GLP-1R KO mice vs WT (B).

CAPTION: DA uptake as a function of DA concentration in striatum of C57BL/6NRj mice treated with saline (n = 3) or exenatide (n = 3) in vivo (A), and in striatum of KO mice lacking neuronal GLP-1Rs and WT littermates (n = 5) (B). Abbreviations: Sal; saline, Ex4; exendin-4, WT; wild type, KO; knockout.

4.3. Experiment 3 - No effect of exenatide on DAT availability in striatum in healthy individuals

Ten healthy participants (5 males) were included. See figure 4 for flowchart. Each participant arrived in the morning, not fasting, and underwent the same procedure followed by an hour of observation with a light meal. See figure 5 for overview of the experiment. The mean age was 24.0 years (SD ± 3.5 years; range, 18–29 years), the mean body weight was 73.1 kg (SD ± 13.8 kg; range, 52.0–98.5 kg). One participant failed to complete the experimental day due to vasovagal hypotension when establishing intravenous cannulas. In one participant, we had to extrapolate values of plasma exenatide and plasma radioactivity due to an intravenous access that became faulty towards the end of scan (i.e. the last three blood samples are missing). No adverse reactions were observed. None of the participants became hypoglycemic (defined as < 55 mg/dL i.e. < 3.0mmol/L)³⁹ at any time during the experiment (mean blood glucose at t100 min was 3.47 ± 0.31 mmol/L).

Figure 4 – Flowchart of Experiment 3.

Figure 5 – Overview of Experiment 3.

4.3.2. Plasma exenatide

Mean exenatide dose was 2.8 (± 0.17) μg. As intended, plasma exenatide increased during the exenatide infusion for each of the participants. Infusion began after 40 min of scan and plasma exenatide concentrations increased significantly within the following 30 min and remained elevated throughout the rest of the scan (effect of time [F(3, 24) = 25.1, p < 0.001]). Mean concentrations were 102.3 ± 7.4, 112.5 ± 14.2, and 131.0 ± 24.1 pmol/L at 70, 85, and 100 min, respectively.

4.3.1. BP_P during exenatide infusion

Figure 6 depicts the striatal DAT binding potentials for each participant over time. We found no statistically significant effect of exenatide on DAT BP_P when comparing with baseline [F(1.9, 17.2) = 0.84, p = 0.44]. Furthermore, there was no correlation between BP_P and plasma exenatide concentrations (r2 = 0.0007). Assessing changes in BP_P for caudate and putamen separately did not change the result (not depicted).

Figure 6 – Experiment 3: no change in DAT availability in striatum during exenatide infusion in healthy individuals.

CAPTION: Each line represents a subject (n = 10). No significant difference comparing BP_P at baseline (40 min) with BP_P values during exenatide infusion (60 min and 80 min), p = 0.44.

5. Discussion

5.1. Preclinical results

A regulatory role of GLP-1Rs for reward modulating DAT and DA homeostasis seems likely since it has been shown that GLP-1Rs are widely expressed in the mesolimbic dopamine system including striatum⁹ and several preclinical studies have demonstrated an attenuating effect on feeding as well as alcohol- and drug-related behavior^2,6. We have previously demonstrated that the GLP-1RA exenatide attenuates cocaine-induced DA release in LS¹⁶. Cocaine alters DA homeostasis by targeting DAT and we found that exenatide exerted its effect via an increase of DAT expression¹⁶. We and others have also shown that GLP-1RAs blunt cocaine’s action i.e. the ability to increase DA levels in striatum^14,15,40. Accordingly, in striatal slices of rats (Experiment 1) we found increased DA uptake after GLP-1 (7–36)- amide application. DA homeostasis is primarily regulated by DAT function⁴¹. DAT is exclusively found on DA neurons⁴² and densely expressed in striatum⁴³. DAT acts as the primary mechanism responsible for terminating DA neurotransmission by translocating DA from the extracellular space back into neurons⁴¹. Changes in DAT membrane trafficking can rapidly regulate dopaminergic tone by altering the number of transporters present at the cell surface⁴¹. Overall transport capacity of DAT depends on the number expressed on the cell surface, hence increased DA uptake could be explained by increased cell surface expression of DAT. Indeed, we found support for this hypothesis in biotinylated rat striatal preparations (Experiment 1), where GLP-1 (7–36)-amide increased DAT cell surface expression. These findings are in line with our previous results in LS¹⁶. In further support, we demonstrated increased striatal DA clearance in an in vivo assay after locally microinjecting GLP-1 (7–36)-amide into striatum of rats. We found that the increase in DA clearance happened within 10 min, suggesting that GLP-1 (7–36)-amide evoked rapid trafficking of DATs to the cell surface rather than de novo synthesis. Collectively, these findings suggest involvement of GLP-1R in DAT function in rat striatum.

In Experiment 2 we turned to mice to take advantage of GLP-1R KOs in parsing out the role of endogenous GLP-1R signaling on DA uptake. We found no differences in DA uptake into striatal synaptosomes between GLP-1R KO mice and WT controls. Because compensation in uptake mechanisms in GLP-1R KO mice might account for this, we examined the effect of exenatide on DA uptake in C57BL/6NRj mice. Unlike our findings in rat, DA uptake into striatal synaptosomes prepared from C57BL/6NRj mice was not affected by systemic administration of exenatide. For this reason, we did not assess DAT cell surface expression. We used a behaviorally relevant dose of 3.2 μg/kg exenatide, which is within a range that produces significant effects on behavioral and dopamine striatal microdialysis measures in mice^14,15. Based on these reports, and our results from Experiment 1, we expected DA uptake into striatal synaptosomes to be increased following exenatide treatment, which was not the case. Collectively, mouse data suggest that GLP-1R signaling does not regulate DA uptake in striatum, which is not in accord with rat data (Experiment 1). The reason for this discrepancy may be due to differences in species since differences in striatal GLP-1R expression have been reported^44,8,9,26. Indeed, mice show very sparse GLP-1R expression in dorsal striatum⁴⁵, suggesting that behavioral effects of exenatide previously reported in mice are not mediated by its actions on striatal GLP-1Rs. Mixed results may also be due to use of different ligands, GLP-1 (7–36)-amide (in rats) vs. exenatide (in mice), as different GLP-1R ligands are known to produce divergent effects on some endpoints, likely due to different signaling bias for second messenger pathways^46–49. Differences in tissue collection methods or route of administration (bath applied vs intraperitoneal injection) might also play a role.

5.2. Clinical results

In Experiment 3 we investigated whether an acute infusion of the GLP-1RA exenatide would affect DAT availability in humans. To our knowledge, we are the first to assess the role of the GLP-1 system on DAT availability in healthy individuals. We found that intravenously infused exenatide did not result in any acute changes in DAT availability.

Few studies have investigated the central effect of GLP-1R activation in humans. Two fMRI studies investigated the central GLP-1 system and reported that exenatide infusion vs. placebo decreased food-related brain responses in patients with type 2 diabetes and obese individuals in appetite and reward-related brain areas including putamen⁵⁰. Furthermore, they found that exenatide also decreased anticipatory food reward⁵¹. These effects were blocked by prior GLP-1R antagonism using the GLP-1R antagonist exendin 9–39. Notably, one study has investigated the effect of long-term treatment (60 weeks) with exenatide, i.e., a 2 mg once weekly prolonged-release formulation, and DAT availability in patients with Parkinson’s disease (PD)⁵². PD is characterized by loss of DA neurons in striatum which can be diagnosed with DAT-SPECT-imaging. The study found a beneficial effect of exenatide on PD symptoms, but no significant change in DAT availability compared to placebo⁵².

Our participants received a clinically relevant dose with a mean infusion of 2.8 μg exenatide within 60 min. Patients diagnosed with type 2 diabetes receive an initial dose of 5 μg s.c, which corresponds to 3.25 μg since bioavailability is approximately 65% after subcutaneous injection⁵³. Exenatide was administered by intravenous infusion in the periphery and is thought to cross the blood-brain barrier⁵⁴. The measured exenatide plasma concentrations were comparable to other studies^55,56. Notably, participants in the mentioned fMRI study⁵⁰ had a mean concentration in the range of 35–43 pmol/L whereas we found a mean of 131 pmol/L (at 1100 min) in our participants. Importantly, in three participants with a very high exenatide concentration > 180 pmol/L did not show any changes in DAT availability either and we found no correlation between plasma exenatide concentrations and DAT availability.

Our results suggest that exenatide does not acutely alter DAT availability in healthy individuals. It should be noted that the sample size was small i.e. n = 5 per sex and there is a risk of type II error due to the low statistical power. It is, however, possible that GLP-1R activation affect DAT differently with longer duration of exposure and/or in people with abnormal DAT availability and DA. Decreased DAT availability has been reported in patients diagnosed with AUD⁵⁷ and we are currently investigating the effects of long-term (26 weeks) treatment with 2 mg exenatide once weekly on both alcohol consumption and DAT availability in AUD patients²³.

Our study includes different experimental designs and methodologies, which perhaps in part can explain the somewhat diverse results presented. We are not confident that our hypothesis that GLP-1R modulates DAT and consequently synaptic levels of DA, should be disregarded entirely, but other possible central mechanisms should be considered. For example, one study has investigated GLP-1R modulation of dynamic DA release during cocaine administration and suggest that instead of regulating DAT, GLP-1R activation possibly suppresses the excitability of DA neurons¹⁷. Other studies suggests that GLP-1R activation regulates reward in a non-dopaminergic fashion via GABAergic inhibitory medium spiny neurons^26,58. Lastly, it has been suggested that activation of GLP-1R reduces excitatory synaptic strength of VTA projections to the striatum in mice⁵⁹.

6. Conclusion

A solid body of preclinical evidence shows that GLP-1R agonists attenuate the effects of substance use disorder-related behaviors. The mechanisms underlying these effects remain elusive. In the present study, we hypothesized that GLP-1R activation modulates DAT and thus DA homeostasis in striatum. In rats and using GLP-1 (7–36)-amide, we found support for this hypothesis corresponding with our previous findings in the LS¹⁶ However, we could not replicate the present rat striatal data in mice and humans using exenatide. This may be due to interspecies- and methodological differences. Another explanation could be that striatal DAT is not of pivotal importance with respect to the GLP-1R mediated effects on addiction-related behaviors and that other reward-related brain areas are of more importance. The present study does not rule this out.

Highlights.

• GLP-1R activation increases rat striatal DA uptake in vitro. No effects of GLP-1R activation on striatal DA uptake ex vivo were observed.

• No acute effect of exenatide infusion on striatal DAT availability in healthy humans was found.