Reserpine-induced rat model for depression: Behavioral, physiological and PET-based dopamine receptor availability validation

baseline

were transient.The data suggest that RES administration can induce a rodent model for depression with mild face validity.

Introduction
Major Depressive Disorder (MDD) is a psychiatric condition marked by persistent feelings of disinterest, anhedonia, changes in appetite, sadness, anxiety, reduced social engagement, and occasionally, anger, disrupting daily activities and affecting mood, behavior, and emotional stability (American Psychiatric Association, 2013;Kennedy, 2008).Worldwide 280 million people are affected by the disorder and diagnosis is on the rise (WHO, 2017), there is an urgency to improve the understanding of the disease and treatment strategies.However, the complex etiology of MDD poses the challenge to create valid animal models that resemble the clinical pathophysiology and symptomatology of MDD (Czéh et al., 2016;Overstreet, 2012;Planchez et al., 2019;Wang et al., 2017).
RES, an indole alkaloid plant extract substance, was originally used as a drug for treating hypertension but it led to depressive episodes and parkinsonism in patients (Cheung and Parmar, 2023;O'Neil and Moore, 2003;Rijntjes and Meyer, 2019).The irreversible inhibition of the vesicular monoamine transporter (VMAT 2 ) by RES at the presynaptic terminals was shown to cause catecholamines to be metabolized resulting in the reduced availability of dopamine, noradrenaline, or serotonin for release into the synaptic cleft.The initial experimental findings from studies involving RES during the 1960s were instrumental in shaping the formulation of the monoamine hypothesis of depression (Kaakkola and Teräväinen, 1990).The drug has been used for >40 years in preclinical studies to evoke parkinsonism and depression-like symptoms in rodents and other species (Leão et al., 2015;Pierzchała-Koziec et al., 1999).Despite this, there is an absence of understanding of the RES-induced pathophysiology evoking depressive symptoms.Moreover, there is a lack of consensus on dosages or protocols most appropriate for experimental models (Acheson et al., 1961;Ikram and Haleem, 2017;Sethy and Hodges Jr., 1985).
The present study aimed: i.) to validate the chronic effects of different doses of RES and ii.) to investigate the temporal robustness of the depression-like phenotype following a wash-out period.We hypothesized that an enduring, dose dependent depression-like phenotype will emerge, accompanied by changes in dopamine-related markers, including a decrease in dopamine receptor availability (increased BP ND ), and that this will persist up to two weeks after the final drug administration.To monitor the outcome, broad-spectrum behavioral phenotyping was conducted focusing on anxiety and depression-like behaviors, and in-vivo dopamine D 2/3 receptor (D 2/3 R) PET with the [ 18 F]DMFP tracer (D 2/3 R antagonist) was carried out to assess D 2/3 R availability.Post-mortem changes in mRNA expression were measured for biomarkers associated with depression pathology.

Animals
Male Sprague-Dawley (SDs) rats (Janvier®, Germany) weighing 250-300 g at the start of the study were housed at a constant temperature (22 ± 1 • C) and light-controlled conditions (12 h light-dark cycle, lights on 07:00 h) with food/ water ad libitum.Rats (3-4/cage) were handled before the start of the experiments to habituate them to the experimenter.All experiments were approved by the local Animal Welfare and Ethical Committee (Regierungspräsidium Freiburg, TierSchG; TVA: G-19/123) and the study conducted according to the German Animal Welfare Act and European legislation.

Drug treatment and experimental design
Fig. 1 A-B summarizes the experimental design of Experiments 1 and 2. Experiment 1 consisted of three phases (Fig. 1A): i) Phase 1: Baseline behavior testing was performed and animals were divided into matched experimental groups based on the Forced Swim Test scoring (Control, Low RES or High RES).Health monitoring including weight and scoring of feces/orifices, body position/muscle tone or lacrimation/eye openness, among others, were noted every three days (scoring was 0, 5 or 10; 0, normal healthy state; 5, mild variation from normal (e.g.softer feces); 10, larger deviations from normal (e.g.diarrhea); ii) Phase 2: Over a day predefined timetable, rats received intraperitoneal (i.p.) injections of either RES solution (Low RES: 0.2 mg/kg / 0.8 ml/kg or High RES: 0.8 mg/kg / 0.73 ml/kg) or vehicle (Control: 0.0 mg/kg / 0.8 ml/kg).RES powder (Sigma-Aldrich® (Germany) was dissolved in 0.3% glacial acetic acid following dilution in distilled water to the required doses.Vehicle injection was 0.3% glacial acetic acid diluted in distilled water.Rats were injected at 17 h00 on days 1, 2, 3, 4, 6, 8, 12, 16 and 20.Depression-like phenotype was re-assessed at the end of the Phase 2. Daily health monitoring and rectal temperature assessment were conducted at noon; iii) Phase 3: Following a 10 days post-injection wash-out period, the depression-like phenotype was re-assessed for a third time.
In Experiment 2, rats were subjected to the same experimental design and timetable as described above, but without any behavioral testing (Fig. 1B).PET scans (Control, n = 4; Low RES, n = 4; High RES, n = 4) were conducted at the end of each of the three phases to investigate the chronic in vivo effects of RES administration on the dopaminergic system.In both Experiments 1 and 2, following Phase 3 (16 days after lastinjection), fresh brain tissue was collected, micro-dissected and prepared for molecular analysis.

Locomotor activity
Locomotor activity was measured continuously over 24 h using individual cages (40 cm × 40 cm × 40 cm) equipped with two pairs of opposite infrared motion sensors.Horizontal activity was determined by counting the number of times the animal crossed/interrupted the beams and the data expressed as the total beam-breaks during 24 h.

Sucrose consumption test (SCT)
Anhedonic behavior was assessed by measuring the consumption of a condensed milk solution.On Day 1, the rats were habituated to individual testing cages (40 cm × 40 cm × 40 cm) during 4 h with free access to sweet condensed milk solution (Nestlé®, 1:3 dilution in water).Next, 15 h food deprivation was followed by a 15 min test period the morning of Day 2. The volume of the consumed condensed milk solution was calculated from the difference in the weight of the bottle before and after the test.Locomotor activity was monitored during this period and expressed as total beam breaks/ 15 min.

Social interaction test (SIT)
The three-chamber SIT measures sociability and preference for social novelty (Thiele et al., 2020).The transparent apparatus (100 cm × cm × 45 cm) consisted of three partially separated zones: Zone 1 with cylinder Cage 1; Middle zone; and Zone 2 with cylinder Cage 2 (See Fig. 2 SIT).The cylindrical cages (diameter: 20 x height: 25 cm) had transparent bars with gaps conducive to low-level physical contact and permissive to odors.Prior to testing, animals were habituated to the chamber for 5 min.The paradigm consisted of two trials.In trial 1 (T1) the rat was placed into the Middle zone with open access to both Zones L.M. Telega et al. and 2. In T1, Cage 1 contained an unknown rat ("Stranger 1" that was to become the "Familiar" rat) and Cage 2 was left empty.The rat had 10 min to explore the apparatus and to interact with Stranger 1, before returning to its home cage for a 5-min inter-trial interval.Following the inter-trial-interval, in trial 2 (T2), the rat was placed again in the Middle zone for 10 min where Cage 1 contained a novel rat ("Stranger 2") and Cage 2, the now Familiar rat (previously Stranger 1).Activity was recorded with a digital camera (Panasonic®, Japan) and the movement tracked (Biobserve, GmbH Germany).The results are expressed as the percentage of time spent in novel zone vs. familiar zone, i.e.Cage 1/ Cage2 in T1 and Cage 2/Cage1 in T2.

Open field test (OFT)
Exploratory behavior and fear/avoidance to open spaces was measured as described by Thiele and colleagues (Thiele et al., 2020).Spontaneous movement was analyzed (Biobserve GmbH, Germany) over 30 min in an open top dark gray PVC box (75 cm × 75 cm × 30 cm).The four corners (18.75 cm × 18.75 cm) and the Centersquared region (37.5 cm × 37.5 cm) centered in the middle of the arena were marked as areas of interest (See Fig. 1 OFT) and the distance covered and time spent in these zones measured.The box was disinfected with 70% ethanol after each animal.

Elevated plus maze (EPM)
Anxiety-associated spontaneous behaviors were probed as described previously (Thiele et al., 2020).The EPM testing apparatus was made of dark gray PVC consisting of two pairs of opposite arms (50 cm × 12 cm) in the form of a "plus", elevated 1 m above ground on a central stand.In the middle of the maze, at the meeting point of the four arms, was a 12 cm × 12 cm center zone.The opposite arms were either "Open", i.e., no walls other than a 0.5 cm ledge; or "Closed", with 50 cm high wall on three sides.At the beginning of the trial, the rat was released in the middle, facing open arm, and the movement tracked over 5 min.All trials were carried out under low light conditions (30 lx white light in the center).The proportion of the time spent in the Open and Closed arms were measured (Biobserve GmbH, Germany), and the maze was disinfected with 70% ethanol to remove odors following each animal.

Forced swim test (FST)
The FST is used as an index of depression-like phenotype by measuring change in activity as the animal switches from active to passive behavior in the face of an acute stressor (Molendijk and de Kloet, 2022).The protocol described by Thiele and colleagues (Thiele et al., 2020) was employed.On Day 1, rats were placed in a plastic cylinder (40 cm high, 20 cm in diameter) with 30 cm of water (21-23 • C) for 15 min.On Day 2, their immobility score was assessed with a 7-min trial.Activity was recorded by a digital video camera (Sony Corporation, Japan) and scored by a blinded assessor."Immobility" was defined as floating behavior, absence of struggling, and the absence of movement in three out of four paws.

RT-qPCR
The day following the last behavioral test/PET scan in Phase 3, rats were anesthetized (100 mg/kg ketamine (10%, Medistar Arzneimittelvertrieb GmbH, Germany) mixed with 10 mg/kg xylazine (Rom-pun® 2%, Bayer, Germany)) and decapitated.Selected brain targets were excised bilaterally, and were dissected on an ice-cold plate.Initially, the area of interest was block-sliced in the antero-posterior direction, followed by the precise manual dissection of the volume of Fig. 1.A-B.The experimental design for Experiments 1 and 2. A) Experiment 1.The protocol comprised of three phases.Phase 1 (baseline): acclimatization, handling and baseline behavior (immobility score from FST was used to assign the animals to matched experimental groups (Control, low RES or high RES).Phase 2 (injection): Following a predetermined schedule, rats were administered i.p. injection of reserpine (RES) (low dose: 0.2 mg/kg or high dose: 0.8 mg/kg) or vehicle.Depression-like phenotype was assessed at the end of each phase.Phase 3 (wash-out): 10 days post-injection, maintenance of depression-like phenotype was behaviorally re-assessed.Post mortem brains were microdissected for mRNA expression analysis.B) Experiment 2. Additional animals underwent the same protocol as in Experiment 1, but without any behavioral assessment.On three occasions immediately following each phase, dopamine D 2/3 receptor binding potential was assessed using microPET.interest using a 2 mm cutting edge circular knife (Fine Science Tools, CA, USA).Samples were stored at − 80 • C until further use for mRNA expression analysis.The cerebellum was used as the control region for mRNA expression levels as it was also the reference region for PET image analysis.

RNA isolation -cDNA preparation
RNA isolation (Holz et al., 2019) was conducted on micro-dissected brain tissue from medial prefrontal cortex (mPFC), nucleus accumbens (NAc) and dorsal striatum (Str) for Experiment 1 and Str and cerebellum (Cer) for Experiment 2. RNA concentrations were determined (Nano-Drop™ 2000, Thermo Fisher Scientific Inc., USA).For the reverse transcription procedure in Experiment 1, 1 μg of RNA from mPFC and NAc, each respectively, and 0.25 μg from Str were used.For Experiment 2, 0.05 μg for Str and 1 μg for Cer were used.SuperScript® III First-Strand Synthesis System (Thermo Fisher Scientific Inc., USA) was used for the reverse transcription procedure following the preparation instructions from the manufacturer.
Samples were run in duplicates and after verification of primer specificity, relative quantification of the target genes with respect to both housekeeping genes was calculated using the LightCycler® 480 software.All runs were adjusted to their respective efficiencies.Samples that did not result in single amplicons were excluded from analysis.

Radiochemistry and PET acquisition
[ 18 F]DMFP was prepared using a cassette-based fully automated module Trasis AllInOne with an integrated HPLC system (Trasis, Belgium).The disposable cassettes and materials supplied by Trasis were modified and adjusted to the production of [ 18 F]DMFP.The Sep-Pak® Alumina N Plus Light Cartridge in position 13 and tC18 Plus long cartridges in position 32 (Waters, Eschborn, Germany) were preconditioned with 10 ml, 10 ml ethanol and 20 ml water respectively prior to start the synthesis.[ 18 F]Fluoride was trapped on a Sep-Pak® Light Accell™ Plus QMA-Cartridge (Waters, Eschborn, Germany) in position 5 and eluted with a solution of 9 mg tetrabutyl ammonium tosylate in 0.7 ml methanol.Methanol was evaporated under vacuum and nitrogen flow at 80-100 • C for 8 min.4 mg of the precursor Tosyl-Desmethoxyfallypride (ABX GmbH, Radeberg, Germany) in 0.8 ml DMSO was added to the dried residue and heated at 150 • C for 19 min.After cooling down, the reaction mixture was neutralized with 10% phosphoric acid and passed through Alumina N Plus Light Cartridge into the HPLC injection loop.The crude product was purified by HPLC as previously described (Gründer et al., 2003) using a semi-preparative column (MultoHigh® 100 RP8-3 μ, 10 mm × 250 mm) and a mixture of 30% acetonitrile in 250 mM ammonium acetate containing 2 ml acetic acid as a mobile phase (flow rate 5 ml/min).The product was collected into a 50-ml vial (position 35) and diluted with Water (37 ml).The mixture was passed through a second Sep-Pak® tC18 Plus long Cartridge (position 32) and rinsed with water (40 ml).The product was eluted with ethanol (1 ml) and diluted with saline (10 ml).The final solution was passed through a Cathivex GV 0.22 μm sterile filter (Merck, Darmstadt, Germany) into a 15-ml sterile vial.The quality control procedure included pH, pyrogenicity tests, identification of radiochemical purity (≥97) by radio-high-performance liquid chromatography and radio-thin-layer chromatography.The residual solvents were determined by gas chromatography.The molar activity ranged from 47 to GBq/μmol.The image acquisitions were performed using a small animal PET/MRI BioSpec 3 T (Bruker BioSpin MRI GmbH, Ettlingen, Germany) at the end of every phase.[18F]DMFP was injected into a tail vein under anesthesia (1-2.5% isoflurane in O 2 ) as a slow bolus injection of 30.0 ± 2.2 MBq.Dynamic list mode data acquisition started directly after tracer injection and lasted 60 min under anesthesia.Body temperature and respiration rates were monitored.Data were reconstructed using maximum likelihood expectation maximization algorithm (MLEM, iterations, voxel size 0.5 × 0.5 × 0.5 mm 3 ) into a sequence of 21 frames increasing systematically from 20 s to 10 min.Scatter, random and decay corrections were applied to each dataset.The final group size due to dropouts (injection failures and deceased animals) for each of the phases was n = 3/4/3 for low RES, n = 4/4/4 for high RES, and n = 4/2/ 2 for Control, for Phase 1, 2, and 3, respectively.

PET image analysis
Image analysis was performed using the PMOD software (V.3.7, PMOD Technologies, Zurich, Switzerland).D 2 R availability was expressed by the non-displaceable binding potential (BP ND ) (Innis et al., 2007) which is the ratio of the distribution volumes of specifically bound to non-displaceable radioligand at equilibrium.For this, an in-house [ 18 F]DMFP PET template was developed in the same space as the SD rat atlas (Schiffer et al., 2006) ([ 18 F]FDG/glucose metabolism PET atlas) provided by PMOD to extract striatal and cerebellar time-activity curves (TAC) using pre-defined volumes of interest (VOI).First, a perfusionphase template was created by co-registering and spatially normalizing perfusion-phase [ 18 F]DMFP PET images (0-5 min post-injection (p.i.)) of Phase 1 to the SD rat atlas 21 .The resulting co-registered images were averaged, and the process was iterated using the average image as a new perfusion template.At each iteration (3 times), visual quality control was performed to ensure a good fit of the data to the atlas space.Second, individual perfusion-phase images of Phase 1 were co-registered and spatially normalized to the final perfusion-phase template.The individually derived transformation matrix was then applied to an early uptake-phase (2-15 min p.i) image of Phase 1 and these were averaged to obtain the final in-house [ 18 F]DMFP PET template.
Individual early uptake-phase [ 18 F]DMFP PET images were coregistered and spatially normalized to the derived [ 18 F]DMFP PET template to define the individual spatial transformation parameters, which were then applied to the entire respective dynamic datasets (0-60 min p.i.).Anatomical definitions of the striatum and cerebellar white matter (WM) were adopted from the Sprague Dawley rat atlas (Schiffer et al., 2006).The cerebellar WM VOI was eroded isotropically by 0.4 mm in order to minimize the impact of blood and bone spillover.Static [ 18 F]DMFP PET images (0-60 min) were scaled to the average uptake of cerebellar WM (i.e., uptake ratio image) and averaged for each group and phase for illustration purposes.BP ND was estimated for each animal at every phase using the Simplified Reference Tissue Model (SRTM) (Lammertsma and Hume, 1996).For validation purposes, BP ND was also quantified using the Logan Reference Tissue Model (LRTM) (Logan et al., 1996) (time of linearization, t* = 25 min) and estimated using a simple area under the curve (AUC) ratio with trapezoid integration (Meyer et al., 2005).

Statistical analysis
Statistical analysis were done using GraphPad Prism 9.5.0.®Version 2019b (GraphPad Software, Inc., La Jolla, CA, United States) and R version 4.2.1 (R Core Team, 2023).All results are presented as mean ± SEM if not otherwise indicated.Data were verified for normality using Shapiro-Wilk test and for equal variances using Lavene's test, when required.Two-way Repeated Measures (rm) ANOVA (main effects for Phase, Dose and their interaction) following post hoc Bonferroni corrections, when applied, were used to analyze the behavioral outcomes.In the cases where sphericity was not fulfilled in the Mauchly's test of sphericity, Greenhouse-Geisser corrections were applied.Data from the qPCR were analyzed using one-way ANOVA test when normality and equal variances were fulfilled; otherwise, non-parametric Kruskal-Wallis ANOVA was employed.Given the limited PET data (see above for group sizes), we restricted the statistical analyses to a) one-way ANOVA for comparison of baseline BP ND across groups and b) two-way repeatedmeasures ANOVA to assess the effect of the factors Dose (only low and high RES) and Phase (and their interaction) on BP ND .For Phases 2 and 3, the effect sizes between BP ND of high RES and low RES compared to Phase 1 were calculated for paired observations using Cohen's d (Hedge's corrected for small sample size) (Hedges and Olkin, 1985).Associations between BP ND values calculated with different kinetic analyses were assessed with Pearson's correlation analysis.Statistical effects were considered significant if p < 0.05.Number of subjects are given in the figure captures and each section of results.
In Phase 3, 48 h after the last injection, animals started to recover their weight.However, the RES treated rats remained significantly lighter compared to Controls (Experiment 1: Control vs. low RES p < 0.0001, low RES vs. high RES p = 0.002; Experiment 2: Control vs. low RES p = 0.0003, Control vs. high RES p < 0.0001, low RES vs. high RES p < 0.0001).
In Experiment 1, one-way rm.ANOVA revealed a significant main effect of Dose on the rectal temperature (F 2,32 = 17.84, p < 0.0001; Suppl.Table 1).The post hoc analysis revealed those effects to be present between Control versus High RES (p < 0.0001) and low RES versus high RES (p < 0.0001) groups.The minimum average temperature measured was 34.7 • C in the high RES group, 1 • C below Control group and 0.5 • C below low RES group; while the maximum was 36.8 • C for the Control group, 0.3 • C above the maximum registered in the high RES group and 0.1 • C below the maximum registered in the low RES group.

Experiment 1
Results from Experiment 1 are summarized in Suppl.Table 1.

Locomotor activity
Fig. 2B-C shows locomotor activity scores.Initially, all groups had similar baseline activity levels during the 15-min SCT and 24-h spontaneous activity (p > 0.05).In Phase 2, low and high RES groups had reduced, and Controls increased activity levels.Statistical analysis indicated significant Phase and Phase x Dose interaction effects for both measurements (SCT: F 2,14 = 4.99, p < 0.05 and F 2,14 = 7.75, p < 0.01; and F 4,28 = 3.03, p < 0.05, respectively; 24 h: Phase F 2,14 = 8.19, p < 0.01; Dose F 2,13 = 6.07, p < 0.05).Post hoc analysis revealed significant differences in Phase 2 between Control and high RES group for the 15min measurement (p = 0.02, See Fig. 2B) and significant effects for both low and high RES groups in the 24-h measurement (p < 0.01 and p < 0.05, respectively, See Fig. 2C).After the washout period, activity increased in all groups in both measurements, except for the Controls over 24 h.

Sucrose consumption test
Baseline sweet condensed milk consumption was similar for the three experimental groups (Fig. 2D and Suppl.Table 1).Main significant differences were found across the three phases (Phase, two-way rm-ANOVA: F 2,8 = 65.45,p < 0.0001).During Phase 2, all groups showed a reduction in consumption in comparison to Phase 1, albeit significance was detected only for the low RES and high RES group (p < 0.0001 and p = 0.007, respectively).In Phase 3, consumption returned to baseline level in all groups; this represented a significant increase for the low RES group (Phase 2 vs. Phase 3 low RES, p = 0.01; see Fig. 2D, Suppl.Table 1).

Social interaction test
Fig. 2E shows the percent of time spent in the novel zone versus the familiar zone for all experimental groups in the SIT.In T1, from Phase 1 to Phase 2, all groups experimented reductions in the percent time spent in the novel zone versus the familiar zone, which remained decreased for all groups when compared to Phase 3. In these last 2 phases, treated groups had an increased tendency to spend more time in novel vs. familiar zone with respect to Controls.Two-way rm-ANOVA revealed significant main effects in Phase (F 2,14 = 5.97, p < 0.05) but not in Dose nor their interaction (F 2,14 = 1.70, p = 0.22 and F 4,28 = 0.76, p = 0.56, respectively).In T2, treated groups had also very similar mean values between Phase 1 and 2 while Control group spent more time in the novel versus familiar zone, with similar tendency kept in Phase 3 compared to the treated groups.On the contrary, there were opposing trends from Phase 2 to Phase 3 among the treated groups: low RES showed a decrease, whereas high RES showed an increase.Two-way rm-ANOVA did not reveal any significant main effects (Phase: F 2,14 = 0.20, p = 0.82; Dose: F 2,14 = 0.67, p = 0.53; interaction: F 4,28 = 0.34, p = 0.85)).Mean values for all groups in all phases can be found in Suppl.Table 1.

Open field test
The performance of each of the groups in the OFT is presented in Fig. 2F and in Suppl.Table 1.Two way rm-ANOVA revealed significant effects across Phases (Phase, F 2,14 = 4.04, p = 0.04) and the interaction Dose x Phase (Dose v Phase, F 4,28 = 13.23,p = 0.0045) but not in Dose (Dose, F 2,14 = 2.71, p = 0.10).In Phase 1, all groups spent a similar amount of time in the Corners (p > 0.05).In Phase 2, the treated groups increased their time in the Corners compared to Controls, with the low RES group showing the most significant increase (p = 0.03).Phase 3 performance in all three groups returned to Phase 1 levels.Time spent in the Center was similar, with a slightly lower mean for the high RES group.During the Phase 2, Controls and high RES group increased their time in the Center, while low RES decreased it.In Phase 3, this trend reversed: Controls and high RES group decreased Center time, while low RES increased.The interaction between Phase and Dose was statistically significant for Center time (F 4,28 = 3.03, p < 0.05), but Dose and Phase alone did not show main significant effects (F 2,14 = 0.53, p = 0.60 and F 2,14 = 1.83, p = 0.20, respectively).

Elevated plus maze
Fig. 2G-H summarizes the EPM results.Baseline % time spent in the different arms were similar for Controls and low RES groups.The high RES group spent higher average fraction of time in the Closed arms and lower mean fraction of time in the Open arms, albeit neither measures were significant (p > 0.05, See Suppl.Table 1).During the injection phase (Phase 2), Control and low RES groups maintained similar mean values in both arms with respect to Phase 1, while the high RES group decreased the time spent in the Closed arms, matching the other groups.Following the wash-out period (Phase 3), an mild decrease in the percentage of time spent in the Closed arms was noted in Control and low RES groups, with − 2% and − 4%, respectively (p > 0.05).Conversely, the high RES group increased the time spent in the Closed arms by 10%.Concerning the Open arms, only minor changes were observed across Phases 2 and 3, with a decrease of 3% and 2% for low and high RES groups, respectively.Neither Dose, nor Phase or their interaction were found to be statistically significant in the % time spent in Open arms (Phase: F 2,14 = 1.64, p = 0.23; Dose: F 2,14 = 1.62, p = 0.23; Phase x Dose: F 2,14 = 0.42, p = 0.67) nor in Closed arms (Phase: F 2,14 = 0.68, p = 0.52; Dose: F 1,8 = 1.52, p = 0.26; Phase x Dose: F 2,14 = 0.94, p = 0.41).

Forced swim test
The mean immobility times obtained for all groups are contained in Suppl.Table 1.During Phase 1, immobility scores were similar as the groups were matched on this criteria (p > 0.05).During Phase 2, mean immobility time increased for both RES injected groups with respect to Phase 1 (21.0 ± 4.9 and 19.6 ± 3.7 s, for low RES and high RES, respectively), while it decreased in the Control group (p > 0.05).Despite the opposite trends in the RES injected and the Controls, the immobility scores were not significantly different (two way rm-ANOVA Phase: F 2,12 = 0.97, p = 0.41; Dose: F 2,12 = 1.27, p = 0.32; and Phase x Dose: F 4,24 = 0.90, p = 0.48).After RES withdrawal (Phase 3), mean immobility time increased for all groups.

mRNA-expression levels
Suppl.Table 2 summarizes Experiment 1 qPCR results and statistical analysis.No statistical significant changes were found across groups for any of the markers and regions assessed (p > 0.05), with the exception of the decrease of accumbal D 2 R mRNA expression in the low RES group compared to the Controls (p = 0.048).

Experiment 2 3.4.1. mRNA-expression levels
The effects of RES administration on the gene-expression of selected makers were varied across regions with no statistical significant maineffects (p > 0.05, See Suppl.Table 3 for statistical results).

D 2 R PET
Across the different phases, the BP ND in Controls varied between 1.36 and 1.64 and did not obviously change (mean ± standard deviation of change from Phase 1 to 2 and 3 in two animals scanned three times: 5 ± 4%; see Fig. 4 and Suppl.Table 4).There was no significant difference in BP ND between the three groups at Phase 1 (one-way ANOVA, F 2,8 = 0.09, p = 0.91).In contrast, the low RES and high RES groups showed a substantial increase in BP ND (or D 2 R availability, in line with a decrease of synaptic dopamine) during the RES injection and wash-out phase (change from Phase 1 to 2 and 3: 50 ± 40%; see Fig. 4 and Suppl.Table 4).These findings are well appreciable by simple visual inspection of the uptake ratio images (Fig. 4A).At Phase 2, the low RES group showed an increase in BP ND of 54 ± 28% (Cohen's d = 3.6, n = 3) and the high RES of 65 ± 57% (d = 1.9, n = 4) compared to Phase 1.At washout (Phase 3), the BP ND increase compared to Phase 1 was slightly smaller both in the low RES (21 ± 6%, d = 1.7, n = 2) and the high RES group (47 ± 43%, d = 1.4,n = 4).Aforementioned data and visual inspection (Fig. 4) suggests a dose effect (i.e., higher BP ND increase in high compared to low RES), but in contrast to the Phase effect, the Dose could not be confirmed by preliminary statistical analysis (two-way rm-ANOVA, Dose: F 1,4 = 0.57 p = 0.49; Phase: F 2,8 = 8.12, p = 0.012; Dose x Phase: F 2,8 = 0.22, p = 0.81).
Validation analyses indicate that BP ND estimates from the SRTM were highly correlated with those from LRTM (r = 0.96, p < 0.001) and the AUC ratio method (r = 0.93, p < 0.001) (Suppl.Fig. 1).BP ND values for all three methods are provided in Suppl.Table 4.

Discussion
The complex etiology and varied symptom presentation in Major Depressive Disorder (MDD) makes it a critical challenge to create or select an appropriate experimental model to investigate this disease.Rodent models cannot replicate the spectrum of behavioral and physiological pathologies observed in MDD, but can focus on key clinical symptoms such as reduced motivation and anhedonia or physiological parameters.The present study explored the consequences and capacity of long-term RES administration to generate a chronic rodent model for MDD.The wide range of behavioral tests allowed for the assessment of the face validity of this model.Furthermore, non-invasive imaging and the molecular evaluations allowed for a deeper insight about the chronic impact of RES's monoamine depleting actions.In addition, we investigated RES's effects even after a prolonged washout period, a parameter not previously assessed in other studies.

Reserpine effects on physiology and well-being
The results confirmed previous findings of RES administration to rodents.The effect on body weight loss is well established and can serve as a predictor of the degree of "reserpinization" (Halaris and Freedman, 1975).Along with the current study, chronic RES administration under different protocols have reported induced body weight loss in rats (Kuzay et al., 2022;Sethy and Hodges Jr., 1985).However, contradictory outcomes have also been shown (Ikram and Haleem, 2017).The disparity of effects can be a consequence of the dose and solution preparation.RES has poor solubility in Phosphate Bovine Serum (PBS), thus, it is probable that the defined doses could have been lower once injected.On the other hand, from the clinical perspective, weight change does not represent a robust physiological marker for the pathology since changes in both directions have been noted in depressive patients (Martel et al., 1962;Simmons et al., 2016).Moreover, the secondary effects derived from the action of the drug (diarrhea, sedation, lacrimation, decreased spontaneous activity, reduced response to stimuli) are consistent with other RES studies (Acheson et al., 1961;Martel et al., 1962).
The average decrease in rectal temperature for the RES administered groups have been well documented (Cremades et al., 1982;Goodrich, 1982;Rohte and Müntzing, 1973;Taylor and Fregly, 1962).Overall, decrease or impaired body temperature regulation have been linked to decreased heat production rather than increased heat loss, as a consequence of lower thyroid gland response or impaired vasoconstriction mechanisms in the skin (Alper et al., 1963).Monoaminergic thermoregulatory mechanisms in hypothalamic neurons have been reported.Specifically, serotonin and noradrenalin may act as antagonists of input signals from thermosensors, regulating the effects mediated by other transmitters (Brück and Zeisberger, 1987).Serotonin might also play a role in thermoregulation through the VTA (Ishiwata et al., 2017), in particular in heat loss processes.Thus, systemic lowering of these transmitters might hinder an effective control within these networks and lead to a body temperature downregulation.In mice, RES (0.6 mg/kg intravenous (i.v.)) has shown to decrease body temperature after 3 h post-injection (Rohte and Müntzing, 1973), while others reported only decreases in young but not adult mice after a single dose of 1 mg/kg (Goodrich, 1982).In rats, other studies have shown hypothermic tendencies only when the doses were above 1 mg/kg (Taylor and Fregly, 1962) or shortly after a hyperthermic phase appearing directly after the drug administration (Cremades et al., 1982).In our study, temperature was measured daily at noon (19 h post-injection/5 h pre-injection), making a direct comparison with previous data difficult due to protocol differences and other factors that could contribute to changes in body temperature.

Reserpine effects on behavior
The main RES induced behaviors reported here are hypo-locomotion, mild anxiety-like and anhedonic-like symptoms and reduced exploratory behavior.The mild behavior changes were predominantly transient, and did not persist following the wash-out period.The dose- dependent decreased motoric symptoms have also been previously reported and studied in more detail under different drug administration schemes and linked to molecular outcomes too (Heslop and Curzon, 1999;Ikram and Haleem, 2017;O'Neil and Moore, 2003;Skalisz et al., 2002).However, the opposite observationthe induction of hyperlocomotion -have also been reported (Neisewander et al., 1991) which can be explained by different treatment schedules, doses and other experimental factors.Nonetheless, hypo-locomotion is akin to decreased mobility, lethargy and loss of energy typically presented in depressed humans providing face validity to the model.These findings together with the decreased striatal dopamine availability indicated by the PET results concords with the dopamine-associated locomotive effects found in other studies using dopaminergic agonists (Oliveira de Almeida et al., 2014).
Anhedonia, reduced motivation and socialization are seminal symptoms of clinical depression.The SCT data shows a tendency of the drug-administered groups for reduced sweet condensed milk consumption, and similar results in SCT and Sucrose Preference Tests have been reported by others (Ozerov et al., 2016;Skalisz et al., 2002).This behavior, interpreted as reduced motivation and manifestation of anhedonic phenotype, could be induced by the decreased dopaminergic neurotransmission due to the VMAT 2 blockade by RES (Belujon and Grace, 2017;Ikram and Haleem, 2017).Interestingly, in the washout phase, the differences were no longer present between treated and nontreated groups, suggesting the reversible/ temporal nature of this phenotype following drug withdrawal.Nevertheless, the lowered sweet condensed milk consumption in the higher dose group could be potentially associated with the minimally decreased mRNA VMAT 2 levels found in that group, aligning with results reported by Isingrini and colleagues (2016) in the SCT.
Clinical depression is associated with increased anxiety, but this was not reflected in the EPM data among the RES treated groups.In the OFT, drug treatment lead to an increased time spent in the Corner zones in Phase 2 and a decrease in exploratory behavior, indices of anxiety-like phenotype.In the current paper, these phenotypes can be ascribed to an overall reduction in mobility/catalepsy (measured semi-qualitatively through visual scoring).Earlier studieswith different injection protocolshave reported an increase in anxiety-like behaviors (Qian et al., 2023;Santos et al., 2013).
Altered socialization was investigated using the SIT, but the current study did not replicate this phenotype.Finally, the FST immobility scoresreflecting reduced perseverance and adaptation to stressremained similar across the three experimental groups following drug administration and after the washout period.Overall, the RES administration reduced mobility and exploration, and induced mild anhedonia.These phenotypes were observed following the drugadministration, but were absent following the washout period.

Reserpine effects on mRNA expression levels
RES treatment has previously shown to modify Tyrosine Hydroxylase (TH) mRNA expressionthe rate limiting enzyme in catecholamine biosynthesis -across diverse regions like the locus coeruleus (LC) and adrenal medulla of rats (Cubells et al., 1995;Pasinetti et al., 1990).In the PFC region, Santos reported reduction in TH+ immunoreactivity after administration of 0.1 mg/kg RES (Santos et al., 2013), similar to our low RES group (0.2 mg/kg), although we report a mild increase in mRNA TH expression.The differences are likely due to differences in the rat strains used, injection routes (subcutaneous (s.c.) vs. i.p.), injection protocols employed, assessment criteria (TH+ immunoreactivity vs. TH mRNA) or time at which post mortem samples were acquired.
The study did not find statistically significant alterations in the mRNA levels of the biomarkers tested (TH, D 1 R, BDNF, VMAT 2 ) across the regions or doses, with the exception of a decrease in NAc D 2 R expression levels following low-RES treatment.This effect is in accord with findings reported in other models of depression, like Chronic Unpredictable Stress (Qiao et al., 2020) and has been related to altered locomotion induced by RES (Sugita et al., 1989) or in addiction studies (Chen and Xu, 2010;Karasinska et al., 2005;Trifilieff and Martinez, 2014).The current findings suggest either: i.) that the wash-out period permitted a regional and dose-dependent restoration of the biomarker mRNAs that might have been altered during the RES injection phase and was accompanied by the decrease in available DA pool; or ii.) the synaptic adaptation mechanisms following the washout period are balanced through other mechanisms, for example, via reduced DAT expression (Neumeister et al., 2001).Studies suggest that these synaptic effects, consequence of a reduced dopaminergic transmission, might be modulated via the upregulation of pre-and post-synaptic D 1 and D 2 receptors (Butkerait and Friedman, 1993;Jaber et al., 1992;Leão et al., 2015).However, in the studies cited, brain tissue were collected a day after the last injection, and did not include a wash out period.This key divergence in the study designleading to different time dependent adaptive mechanisms post drug administrationis a probable reason behind the contrasting findings.

D 2 R availability following chronic reserpine treatment: [ 18 F]DMFP PET
PET imaging in animal models for depression permits is a powerful tool to complement the assessment of longitudinal behavioral changes with the monitoring of changes in brain physiology (Vazquez-Matias et al., 2023).In the current study, the PET tracer [ 18 F]DMFP was used to evaluate D 2 R availability chronically before, during and after RES administration.We found a marked increase in striatal BP ND of [ 18 F] DMFP in both the low and the high RES groups, at both the end of the 3 weeks of drug administration and also after the 2 weeks of wash-out period.Assuming a competition of endogenous dopamine with the [ 18 F]DMFP for D 2 R binding (Döbrössy et al., 2012;Rominger et al., 2010), the emerging PET data from the current study would suggestas predicted -a reduction in synaptic dopamine in our model induced by both the low and the high RES treatments.However, following the final PET measurement, we report a significant reduction in the mRNA levels coding for D 2 R in the NAc.Yamamoto and colleagues have recently described a region-wise correlation between the mRNA expression levels of dopamine D 2 receptors in human post mortem tissue and their availability as measured by in vivo PET (Yamamoto et al., 2023).Still, the relation between mRNA and receptor levels can vary, as there are diverse parameters such as temporal/ regional dynamics, ligand availability, and post-translation regulatory events (Jongen-Rêlo et al., 1994;Komorowski et al., 2020;Wigestrand et al., 2011).In addition, in PET, there are imaging processing factors, e.g.partial volume effects, which can impact the measures (Vazquez-Matias et al., 2023).Furthermore, due to limited spatial resolution, the increase in the BP ND value observed in the current study refers to the entire striatum -not only the NAcand which are congruent with the human (Peciña et al., 2017) and the nonhuman primate findings (Felger et al., 2013).On the other hand, the decrease in the D 2 R mRNA was found only in the nucleus accumbens, but not in the striatum (that included the dorsal striatal areas in the microdissection).Our data conforms to the interpretation that the RES induced depletion of synaptic dopamine (indicated by the increase in striatal BP ND of [ 18 F]DMFP) was followed by a loss of D 2 R receptors (as reflected by the reduction in accumbal D 2 R mRNA).Interestingly, the effect of RES treatment on striatal BP ND lasted markedly longer (at least 2 weeks) than the actual drug administration, revealing a chronic pathophysiological impact.

Study limitations
The study used only male rats, which limits the generalizability of the findings.Additional well-powered studies are needed to take into account i.) inter-individual differences in the reaction to the RES administration, and ii.) to explore the mRNA expression and BP ND relation.Longer washout phases would help to elucidate the effect of dosage and timing on D 2 R availability under RES treatment.Furthermore, including an additional control group without injections would strengthen the impact of our findings.

Conclusion
Chronic administration of 0.2 and 0.8 mg/kg of RES induced mild and transient changes in measures of mobility, anxiety-like, anhedonialike, and exploratory behaviors.The behavioral differences across the groups were nearly absent following the wash-out period.The RES administration resulted in dose-dependent weight loss that started to recover during the wash-out period.Repetitive PET measures indirectly indicated the reduction of synaptic dopamine following the drug administration and persisting up to 2 weeks after the last injection, accompanied by a decrease in accumbal D 2 R mRNA message.In summary, following the chronic RES administration -contrary to our hypothesis -the emerging behavioral phenotype was transient and mild.However, the physiological consequences on biomarkers of dopamine transmission were more long-term and moderate.On the grounds of the transient phenotype, but a more consistent physiological response, we consider this model to have mild face validity as a rodent model of depression.The discrepancy between the transient behavior changes and the longer lasting physiological read-outs will require additional investigation.

Fig. 2 .
Fig. 2. A-I.Physiological and behavioral results from Experiment 1 probing depression-like phenotype.A) Longitudinal weight monitoring during Phases 1-3 (g); B) 15 min locomotor activity during SCT (beam breaks); C) 24 h spontaneous locomotor activity (beam breaks); D) Sucrose Consumption Test (ml); E) Social Interaction Test (% time in novel versus familiar zone); F) Open Field Test (time spent in Corners); G) Elevated Plus Maze (% time in Closed arm); H) EPM (% time in Open arm); I) FST (total immobility time).Results expressed as mean ± SEM; *p < 0.05 as compared with two-way rm.ANOVA.Graphical designs created with BioRender.com.

F:
AGG TCG GTG TGA ACG GAT TTG R: AGC CTT GAC TGT GCC GTT GAA CTT 62 • C ARBP F: CCT GCA CAC TCG CTT CTT AGA G R: CAA CAG TCG GGT AGC CAA TCT G 62 • C TH F: CTT TGA CCC AGA CAC AGC A R: TGG ATA CGA GAG GCA TAG TTC 58 • C D 1 R F: GGA GGA CAC CGA GGA TGA R: ATG AGG GAC GAT GAA ATG G 58 • C D 2 R F: CAC TCA GAT GCT TGC CAT TGT TC R: GTG GGA TGT TGC AAT CAC AGT GTA 62 • C BDNF F: GCC ACT GAA ATG CGA CTG AAT G R: CTG CTC TGC CAG GAA ATA GTA TGT C 60 • C VMAT 2 TGC TGG AGA AGG CAA ACA TAA CTG R: TCC AGC TCC TCA CTA ACC CAT TCA 65 • C L.M. Telega et al.

Fig. 4 .
Fig. 4. A-B.[ 18 F]DMFP PET results.A) Group-averaged parametric uptake ratio images in each group and phase (selected axial slices at the level of striatum, reference: cerebellar white matter, WM); B) BP ND estimates (box plots, brown square indicating the mean value) provided by the Simplified Reference Tissue Model (group sizes for the Phases 1/2/3 are as follows: n = 3/4/3 for low RES, n = 4/4/4 for high RES, and n = 4/2/2 rats for Control.