Ventral striatal islands of Calleja neurons bidirectionally mediate depression-like behaviors in mice

Introduction

Depression, one of the most common mood disorders, is associated with dysfunction of brain reward systems [1]. In both rodents [2, 3] and humans [4–7], depressive phenotypes are linked to dysfunction of the ventral striatum, a key reward center that integrates brain-wide inputs including from midbrain dopamine neurons and sends inhibitory outputs to downstream structures [8, 9]. The ventral striatum contains two major divisions, the nucleus accumbens (NAc) and the olfactory tubercle (OT; also called tubular striatum [10]). Similar to the NAc, the OT is also involved in motivational and reward-related behaviors in rodent models [11–18]. In contrast to extensive research on the role of NAc in depression [19–27], whether and how the OT circuitry contributes to depression is largely unknown.

Like the rest of the striatum, the OT contains spiny projection neurons (SPNs) and several types of interneurons [28, 29]. SPNs are GABAergic principal neurons that express either D1- or D2-type dopamine receptor [8, 9]. Functional and anatomical changes in NAc D1 and D2-SPNs or interneurons have been linked to depressive behaviors in rodents [19, 26, 30]. The ventral striatum also contains neurons expressing D3-type dopamine receptors, the majority of which are granule cells concentrated in the islands of Calleja of the OT [30]. The drd3 gene encoding the D3 receptor [30] as well as the islands of Calleja [31, 32] have been linked to the pathophysiology of neuropsychiatric diseases. Moreover, cariprazine, an atypical antipsychotic drug prescribed to treat several neuropsychiatric disorders including bipolar depression, predominantly binds to islands of Calleja neurons in the mouse brain [33], supporting a potential role of these neurons in regulating emotion.

We recently discovered that OT D3 neuron activity bidirectionally regulates self-grooming in mice: optogenetic activation of these neurons robustly initiates orofacial grooming while optogenetic inhibition halts ongoing grooming [34]. Self-grooming, an evolutionally-conserved, repetitive behavior, serves important functions including de-arousal and stress reduction [35, 36]. Notably, altered levels of grooming are considered a behavioral biomarker for a number of neurological and neuropsychiatric disorders including depression [35, 36]. Since self-grooming increases dopamine release in the NAc [37], it is possible that the activity of OT D3 neurons and self-grooming are intimately linked to the reward system and emotion, and that these neurons therefore may be integral to the pathophysiology of depressive states.

In this study, by combining optogenetic and chemogenetic manipulations, genetic ablation, ex vivo electrophysiology and mouse behavioral assays, we asked 1) what is the relationship between OT D3 neuron activity and depressive behaviors, 2) does activation of these neurons have a rewarding effect and whether it depends on the elicited self-grooming, and 3) which underlying neural pathway transmits the activity of OT D3 neurons to influence dopamine release? To answer these questions, we used the chronic restraint stress (CRS) model [38–41] to induce robust depression-like behaviors (increased inactivity and anhedonia) in mice and found that CRS significantly decreased excitability of OT D3 neurons without changing that of neighboring D1/D2-SPNs. Loss-of-function of OT D3 neurons led to depression (particularly increased inactivity but not anhedonia), whereas activation of these neurons ameliorated CRS-induced depression-like behaviors. Furthermore, optogenetic activation of OT D3 neurons produced conditioned place preference, indicating a rewarding effect, which diminished when self-grooming was blocked. Finally, we provide ex vivo electrophysiological evidence to support a model in which OT D3 neurons influence dopamine release into the NAc via synaptic connections with OT SPNs that subsequently project to dopamine neurons in the ventral tegmental area (VTA). Our study reveals a critical role of OT D3 neurons in regulating emotional responses, suggesting a new target for treatment of depression.

Results

CRS induces anxiety- and depression-like behaviors and decreases excitability of OT D3 neurons

To investigate how chronic exposure to stress may influence function of the OT, we subjected double transgenic D1-tdTomato/D2-EGFP mice and transgenic D3-Cre/tdTomato mice (see Materials and Methods for details) to CRS for 14 consecutive days (2 h per day). Twenty-four hours after the last CRS session, the mice underwent multiple behavioral tests (sequentially with stress level from low to high) to assess anxiety- and depression-like behaviors (Fig. 1A; see Materials and Methods for more details). Since CRS caused similar behavioral changes in the two mouse lines used, we pooled the data for statistical analysis in Fig. 1 (see Supplemental Figs. S1 and S2 for data and statistical analysis separated by each mouse line and sex). For anxiety-like behaviors, we used the open field test (OFT), light-dark box transition test (LDT) and elevated zero maze (EZM) test. In the OFT, CRS did not have significant effects on the total distance travelled nor the time spent in the center zone (Fig. 1B), indicating that CRS did not affect general locomotion. In the LDT, compared to the controls, CRS mice exhibited a similar latency for the first entry into the dark area but spent longer time in the dark area (Fig. 1C). In the EZM test, CRS mice showed an increased latency to the first entry into the open sections while shortened the duration stayed in the open sections compared to the controls (Fig. 1D). For depression-like behaviors, we performed the forced swimming test (FST) and tail suspension test (TST). Compared to the controls, CRS mice spent more time in immobility in both tests (Fig. 1E, 1F), inferred as enhanced “helplessness” and “despair” (see Materials and Methods for details). We then conducted the sucrose preference test to investigate CRS-induced anhedonia, another core behavioral symptom of depression. To avoid potential influence from prior behavioral tests, we used another cohort of mice to conduct this test (same for the following sucrose preference tests). CRS decreased the sucrose preference index to about 73% of the controls (Fig. 1G), indicating anhedonia. Overall, CRS induced some anxiety-like phenotypes and robust depression-like behaviors in mice, in general agreement with previous reports [42, 43]. Since stress affects grooming behavior in rodents [37, 44–47], we also quantified the time spent in orofacial grooming and found that CRS mice groomed more than controls (Fig. 1H; see Discussion).

To examine potential effects of CRS on the electrophysiological properties of OT neurons, we performed whole-cell patch-clamp recordings on D1-tdTomato/D2-EGFP SPNs and D3-Cre/tdTomato neurons (hereafter referred to as D1/D2 SPNs or D3 neurons, respectively) in acute brain slices from control and CRS mice (Fig. 2A, 2E). Representative traces of firing patterns of D1/D2 SPNs and D3 neurons are shown in Fig. 2B and Fig. 2F, respectively. Since the data from D1 and D2 SPNs showed very similar trends, they were pooled together. Without altering the input resistance of these neurons (Fig. 2C, 2G), CRS significantly lowered the firing frequencies of D3 neurons but not D1/D2 SPNs upon current injections (Fig. 2D, 2H), indicating that CRS specifically reduces excitability of D3 neurons in the OT.

Ablation Or Inhibition Of Ot D3 Neurons Induces Robust Depression-like Behaviors

We next explored the potential involvement of OT D3 neurons in anxiety- and depression-like behaviors in physiological conditions (without CRS treatment) by genetically ablating these neurons. We bilaterally injected the Cre-dependent DTA virus or control virus into the OT of D3-Cre/ChR2 mice (ChR2-EYFP as a marker for D3 neurons) (Fig. 3A). We previously showed that four weeks later, this approach efficiently ablated OT D3 neurons revealed by reduced EYFP signals [34]. We performed the same behavioral tests on DTA and control virus mice four weeks post injection (Fig. 3B). Ablation of D3 neurons had little effect on anxiety-like behaviors (except for the longer time spent in the center zone in the OFT, reflecting a mild anxiolytic effect) (Fig. 3C-E). By contrast, ablation of these neurons induced robust depression-like behaviors, leading to significantly longer immobility time in both the FST and TST (Fig. 3F, 3G), similar to CRS effects. Ablation of D3 neurons did not change the sucrose preference index (Fig. 3H), suggesting that these neurons are dispensable for this hedonic behavior.

In addition, we tested the potential role of OT D3 neurons in mediating anxiety- and depression-like behaviors by chemogenetic manipulations. We bilaterally injected a mixture of Cre-dependent excitatory DREADD hM3D(Gq) and inhibitory DREADD KORD (Gi coupled DREADD based on the kappa-opioid receptor template) viruses into the OT of D3-Cre mice (Fig. 4A; Supplemental Fig. S3) to achieve bidirectional manipulations of D3 neuronal activity in the same mice [48] (the results of excitatory DREADD are described later). Ex vivo patch-clamp recordings confirmed that KORD-expressing D3 neurons were inhibited by its ligand salvinorin B (SALB), reflected by significantly decreased firing frequencies compared to control condition (Fig. 4B). Three weeks after viral injection, we conducted the same behavioral tests with subcutaneous injection of either DMSO (as control) or SALB (Fig. 4A). Unlike the mild anxiolytic phenotypes resulting from ablation of D3 neurons, inhibition of these neurons consistently and robustly produced anxiety-like behaviors, characterized by the decreased time spent in the center zone in the OFT (Fig. 4C), shorter latency to the first entry into the dark box and longer time spent in the dark box in the LDT (Fig. 4D), as well as the increased latency to the first entry into the open sections and less time spent in the open sections in the EZM test (Fig. 4E). Similar to ablation of D3 neurons, chemogenetic inhibition of these neurons also induced robust depression-like behaviors. SALB injected mice spent longer time in immobility in both the FST and TST compared to DMSO control conditions (Fig. 4F, 4G). The altered immobility duration was not due to the impaired motor ability since inhibition of D3 neurons did not affect general locomotion in the OFT (Fig. 4C). In addition, inhibition of D3 neurons did not affect the sucrose preference index (Fig. 4H), consistent with the ablation experiment (c.f. Figure 3H). To exclude the potential non-specific effect of SALB, we bilaterally injected the Cre-dependent AAV8-DIO-mCherry virus into the OT of D3-Cre mice as controls. Application of DMSO or SALB did not change behavior in these control mice (Supplemental Fig. S4), supporting that inhibitory DREADD-induced effects result from the interaction between SALB and KORD. Taken together, these results indicate that loss-of-function of OT D3 neurons via both genetic ablation and chemogenetic inhibition reliably induces depression-like behaviors.

Activation Of D3 Neurons Normalizes Crs-induced Depression-like Behaviors

In order to evaluate whether activation of OT D3 neurons can mitigate CRS-induced anxiety- and depression-like behaviors, we used both optogenetic and chemogenetic approaches. In D3-Cre/ChR2 mice, the islands of Calleja D3 neurons in the OT were visualized by EYFP signals (Fig. 5A, top). These D3-Cre/ChR2 neurons reliably fired action potentials upon blue light stimulation at 20 Hz (473 nm; 10 ms pulse) [34] (Fig. 5A, bottom), and these parameters were applied in the following behavioral tests. Immediately after each CRS session, blue (for activating ChR2-expressing D3 neurons) or green light (less efficiency in activating ChR2 as comparison) was applied for 15 min in the home cage, and all behavioral tests were conducted during blue or green light stimulation, except for the sucrose preference test in which photostimulation was applied immediately before the test (Fig. 5B). Compared to green light, activation of D3 neurons by blue light alleviated some CRS-induced anxiety-like phenotypes (increased latency to the first entry into the dark area in the LDT and shortened latency to the first entry into the open sections in the EZM test) but not others (Fig. 5C-E). By contrast, optogenetic activation of D3 neurons ameliorated CRS-induced depression-like behaviors in both the FST and TST. Compared to green light, activation of D3 neurons by blue light decreased the immobility time by 57% (FST) and 63% (TST) (Fig. 5F, 5G). Furthermore, optogenetic activation of D3 neurons (15 min/day) was not sufficient to weaken CRS-induced anhedonia in the sucrose preference test (Fig. 5H), consistent with the results from the ablation and inhibition experiments (c.f., Figs. 3H and 4H). Since blue light activation of OT D3 neurons induces orofacial grooming [34, 49], it may lead to underestimation of the total distance travelled in the OFT as well as the immobility time in the FST and TST. We therefore rectified potential grooming-related deviations under light stimulation (see Materials and Methods for details), and similar conclusions could be drawn from these behavioral tests (Supplemental Fig. S5A-E).

We next tested the effects of chemogenetic activation of D3 neurons on CRS-induced anxiety- and depression-like behaviors. We performed same behavioral assays on CRS, excitatory DREADD hM3D(Gq) mice with intraperitoneal injection of either saline (as control) or CNO (activating OT D3 neurons) (Fig. 6A). Ex vivo patch-clamp recordings confirmed that hM3D(Gq)-expressing D3 neurons were activated by its ligand CNO, reflected by significantly increased firing frequencies compared to control condition (Fig. 6B). Chemogenetic activation of D3 neurons did not mitigate any CRS-induced anxiety phenotypes (Fig. 6C-E). By contrast, activation of these neurons normalized CRS-induced depression-like behaviors, characterized by less immobility time in both the FST and TST (Fig. 6F, 6G). Similar to optogenetic manipulations, chemogenetic activation of D3 neurons also did not improve CRS-induced anhedonia in the sucrose preference test (Fig. 6H). To exclude the potential non-specific effect of CNO, we bilaterally injected the Cre-dependent AAV8-DIO-mCherry virus into the OT of D3-Cre mice as controls. Intraperitoneal injection of saline or CNO did not influence behavior of these control mice (Supplemental Fig. S6), supporting that excitatory DREADD-induced effects are ascribed to the interaction between CNO and hM3D(Gq).

Optogenetic Activation Of Ot D3 Neurons Induces Conditioned Place Preference

Since optogenetic or chemogenetic activation of OT D3 neurons has antidepressant effects, we then asked whether activation of these neurons has any inherently rewarding effects (or associated positive valence). We used the conditioned place preference (CPP) assay, a Pavlovian conditioned paradigm, which contained four sessions in four days (see Materials and Methods for details). In the pre-conditioning session, a mouse was allowed to freely explore the three-compartment arena to determine its most and least preferred side chamber (Fig. 7A). In the following two conditioning sessions, the mouse was stimulated by blue light in the least preferred chamber. In the last session, the post-conditioning session, the mouse was again allowed to freely explore the arena, the time it spent in each compartment was measured and the CPP difference score (the time difference between the post- and pre-conditioning session in a chamber) was calculated (Fig. 7B). Unlike the D3-Cre control mice, which maintained their initial bias for the most preferred chamber, D3-Cre/ChR2 mice spent more time in the laser-paired chamber in the post-conditioning session (Fig. 7C), and thus had a significantly higher CPP score (Fig. 7A, 7B). These results suggest that optogenetic activation of OT D3 neurons has a rewarding effect.

We previously reported that optogenetic activation of OT D3 neurons reliably induced self-grooming while inactivation of these neurons halted ongoing grooming [34]. Similarly, chemogenetic manipulations of these neurons also bidirectionally mediated grooming (Fig. 4I and Fig. 6I). We then asked whether grooming induced by D3 neuron activation is required for the ability of these neurons to drive CPP. We performed the same CPP test using D3-Cre/ChR2 mice with a collar around the neck to block grooming by preventing the forepaws from contacting the face and head (Fig. 7D, inset). Ablating self-directed orofacial grooming eliminated CPP caused by OT D3 neuron stimulation (Fig. 7D, 7E), suggesting that grooming elicited by activation of D3 neurons is necessary for the rewarding effect.

Ot D3 Neurons Inhibit Ot Spns Which Directly Inhibit Nac-projecting Vta Dopamine Neurons

We propose a neural circuit model that may explain bidirectional regulation of grooming and depression-like behaviors via the activity of OT D3 neurons (Fig. 8A). OT D3 neurons directly inhibit OT SPNs which in turn inhibit the ventral tegmental area (VTA) dopamine neurons that mediate dopamine release into the NAc. This model is based on several lines of evidence in the literature: 1) self-grooming (both spontaneous and stress-elicited) induces transient dopamine release into the NAc [37], 2) the VTA◊NAc pathway is implicated in regulating depression-like behaviors [30, 50, 51], 3) OT D3 neurons are local interneurons and provide direct inhibition onto OT D1/D2 SPNs [34], and 4) OT SPNs project directly to VTA [52].

To provide direct evidence to support this model, we examined whether OT SPNs make monosynaptic connections onto NAc-projecting VTA neurons. Since the VTA receives denser innervation from OT D1-SPNs than D2-SPNs [52], we tested functional connections of the OT SPNs◊VTA◊NAc pathway in D1-Cre mice. We bilaterally injected Cre-dependent AAV1-EF1a-DIO-ChR2-EYFP virus and cholera toxin subunit B-555 (CTB) into the OT and NAc, respectively (Fig. 8B). The ChR2-EYFP⁺ OT D1-SPNs, CTB⁺ NAc neurons, and retrogradely labeled CTB⁺ VTA neurons surrounded with dense D1-SPNs axonal fibers were confirmed post mortem (Fig. 8C, 8D). We performed whole-cell patch-clamp recordings on CTB⁺ VTA neurons in acute brain slices and recorded blue light evoked inhibitory postsynaptic currents (IPSCs, which were inward currents due to high intrapipette [Cl⁻]; see Materials and Methods for details). In ~ 72% (43 out of 60) of CTB⁺ VTA neurons, repeated light pulses evoked IPSCs, which had short latency (~ 5 ms) and little jitter (< 1 ms) (Fig. 8E). These currents were blocked by GABA_A receptor antagonist bicuculline but not changed by glutamate receptor antagonists, (2R)-amino-5-phosphonovaleric acid (AP5) and cyanquixaline (6-cyano-7-nitroquinoxaline-2,3-dione) (CNQX) (Fig. 8F), supporting the existence of GABA_A mediated monosynaptic connections from OT D1-SPNs onto these NAc-projecting VTA neurons. Among synaptically connected CTB⁺ VTA neurons, we classified them into several types based on electrophysiological properties [53, 54]. The majority (~ 67% or 29 out of 43) displayed characteristics of dopamine neurons with low firing frequency (< 5 Hz) upon current injection, apparent spike frequency adaptation and voltage “sag”. Another 23% were deemed as GABAergic neurons as they had higher firing frequency (≥ 5 Hz) with little spike frequency adaption or voltage “sag”, while 9% belonged to either class (Fig. 8G). Further post-hoc immunostaining with a tyrosine hydroxylase (TH) antibody confirmed that at least some of CTB⁺ VTA neurons were also TH⁺ (Fig. 8E), supporting their identity as dopamine neurons. Taken together, these results support that OT SPNs make direct inhibitory synaptic connections onto NAc-projecting VTA neurons.

Discussion

By using cell-type-specific manipulations, ex vivo electrophysiology and behavioral assays, we reveal that OT D3 neurons (mostly in the islands of Calleja) bidirectionally mediate depression-like behaviors in mice. Loss-of-function of OT D3 neurons leads to depression-like behaviors (specifically increased inactivity), gain-of-function of these neurons alleviates CRS-induced depression-related symptoms. We propose a model which links the activity of OT D3 neurons and self-grooming with the reward system and stress induced responses.

We first examined the effects of CRS treatment using multiple behavioral assays to assess anxiety- and depression-like phenotypes in mice. CRS induced robust depressive phenotypes, (i.e., increased immobility in both forced swimming and tail suspension tests) and anhedonia (decreased preference for sucrose) (Fig. 1E-G), supporting that CRS effectively causes depression-like behaviors in rodents as previously reported [38–41]. In addition to depressive phenotypes, CRS mice also exhibited anxiety-like behaviors. Unlike the consistent performance in depression-related assays, CRS mice exhibited divergent states in different anxiety assays. Anxiety-like behaviors were observed in light-dark box and elevated zero maze tests, but not in the open field test (Fig. 1B-D), suggesting that these behavioral assays may have different sensitivities or test different aspects of anxiety. Given that there is no single ideal animal model for anxiety and that each existing test has its advantages [55], a combination of different behavioral tests produces a better understanding in anxiety-related processes [56]. Our findings further support the necessity of using multiple behavioral assays for testing anxiety-like behaviors.

CRS specifically decreases excitability of OT D3 neurons but not OT D1/D2 SPNs (Fig. 2), which is in sharp contrast to results from NAc circuits. Stress produces distinct changes (e.g., morphology, excitability, synaptic transmission) in NAc D1 and D2 SPNs, which have opposing roles in depression-like behaviors [2, 30]. Specifically, chronic stress induces hyperexcitability of NAc D1 SPNs [20, 21]. Here we report that OT D3 neurons are vulnerable but OT SPNs are resilient to stressful and depressive states, suggesting chronic stress exerts differential influences on the OT and NAc circuitry. The role of dopamine receptor-expressing neurons in the pathophysiology of depression-related symptoms is also implicated in other brain areas. For example, dysfunctions of p11 in dopamine D2 receptor-expressing neurons in the lateral habenula and prelimbic cortex contribute to depression-like behaviors [40, 41]. Our results suggest a causal relationship between activity of OT D3 neurons and depressive phenotypes, which is supported by both loss-of-function and gain-of-function experiments. Ablation or chemogenetic inhibition of OT D3 neurons caused depressive behaviors (Figs. 3, 4), especially increased inactivity, whereas optogenetic or chemogenetic activation of these neurons alleviates CRS-induced depression-like behaviors or makes the mice more resilient (Figs. 5, 6). These findings add OT to the existing brain areas that are causally linked to depression-related symptoms.

In contrast to the striking effects on the immobility time in forced swimming and tail suspension tests, optogenetic or chemogenectic activation of OT D3 neurons is insufficient to ameliorate anhedonia (reduced sucrose preference) in CRS mice (Figs. 5, 6). One possibility is that different aspects of CRS-induced depressive phenotypes are mediated by distinct neuronal types and/or brain regions, which is supported by the finding that distinct ventral pallidal neurons mediate separate depressive symptoms [57]. Activation of OT D3 neurons may specifically improve “the decreased motivation for activity” in CRS mice, while anhedonic phenotypes might be mediated by other neuronal subtypes such as cholinergic neurons [25, 58] and/or other brain regions. There is evidence to support that NAc D1-SPNs are critical for the expression of anhedonia [2, 26, 30]. The other possibility is that OT D3 neurons are involved in mediating CRS-induced anhedonia but our relatively short activation of OT D3 neurons (15 min/day in optogenetic manipulation in Fig. 5 and approximately 2 h chemogenetic manipulation in Fig. 6) is not sufficient to improve the reduced sucrose preference. This hypothesis is, to some extent, supported by our finding that activation of OT D3 neurons had a rewarding effect in a Pavlovian conditioned paradigm (Fig. 7). A recent report suggests that acute and chronic optogenetic stimulations on certain neural pathways could compensate each other in their effects on normalizing depressive phenotypes [59]. It would be interesting to separate the effects of chronic and repetitive optogenetic activation of OT D3 neurons during CRS treatment versus acute optogenetic activation during the anxiety- and depression-related assays.

In this study, the two loss-of-function approaches produced different effects on anxiety-like behaviors. Chemogenetic inhibition of OT D3 neurons induced anxiety-like behaviors in all three tests (open filed, light-dark box transition and elevated zero maze) (Fig. 4). However, ablation of these neurons produced a minor anxiolytic effect highlighted by the increased time in the center zone in the open field test, but no significant changes in the other two tests (Fig. 3). Unlike transient inhibition via chemogenetic manipulations, ablation of OT D3 neurons permanently alters the OT circuitry. For instance, it may induce compensatory changes in the neuronal activity of neighboring OT D1/D2 SPNs, which in turn affects both local and long-range circuits (e.g., the VTA◊lateral septum pathway) that are involved in regulating anxiety-related behaviors [60].

Consistent with our previous finding that optogenetic activation or inactivation of OT D3 neurons initiates or halts self-grooming [34], here we demonstrate that chemogenetic manipulations also bidirectionally mediate the total grooming time (Figs. 4, 6). Interestingly, CRS treatment decreased excitability of OT D3 neurons (Fig. 2), which should act to reduce the grooming drive from these neurons. However, CRS mice exhibited more grooming than controls (Fig. 1H), suggesting that other grooming centers may be more active after CRS treatment to ensure this adaptive behavior in stressed situation [35, 37, 44, 47]. The potential role of D3 receptor in grooming and depression-like behaviors remains elusive. D3 receptor knockout mice display high basal level of grooming [61], consistent with the inhibitory action of this receptor in OT D3 neurons. Whereas some studies suggest that mice lacking D3 receptor are more resistant to stressful conditions and display normal emotional behaviors [62, 63], others show that D3 receptor deficiency results in depression-like behaviors [64, 65]. Since D3 receptor is expressed in multiple brain regions, specific manipulation of D3 receptor expression in distinct subpopulations of D3 neurons would be required to dissect out its role.

As OT D3 neurons are local GABAergic interneurons [34], we propose a model which links these neurons and self-grooming with the NAc reward system through OT SPNs and VTA dopamine neurons (Fig. 8). Decreased OT D3 neuron activity would disinhibit monosynaptically connected SPNs, leading to enhanced inhibition onto NAc-projecting VTA dopamine neurons. This would attenuate dopamine release into the NAc and ultimately induce depressive phenotypes in these mice. On the other hand, activation of OT D3 neurons eventually leads to more dopamine release into the NAc, manifested as increased motivation for activity (characteristic of resilience from depressive phenotypes) in CRS mice. This model is supported by the rewarding effect of activation of OT D3 neurons, and interestingly, this effect diminishes when grooming is physically blocked (Fig. 7). Furthermore, we provide ex vivo electrophysiological evidence to support that OT D1 SPNs directly inhibit NAc-projecting VTA dopamine neurons (Fig. 8). However, we do not rule out other potential pathways that link OT D3 neurons and grooming with NAc dopamine signaling.

Taken together, we discovered a novel role of OT D3 neurons in bidirectionally mediating depression-like behaviors. The findings that activation of OT D3 neurons has a rewarding effect and efficiently alleviates CRS-induced depressive phenotypes by increasing the motivation for activity suggest that ventral striatal OT/islands of Calleja D3 neurons are an attractive target for the intervention and treatment of depression.

Materials And Methods

Animals

The D1-tdTomato [66] and D2-EGFP (Tg(Drd2-EGFP)S118Gsat) [67] mice were crossed to obtain double-transgenic mice in which dopamine D1 and D2 receptor-expressing neurons are labeled with red and green fluorescence, respectively. Bacterial artificial chromosome (BAC) transgenic D3-Cre (Tg(Drd3-cre)KI198Gsat) was obtained from Mutant Mouse Resource & Research Centers (MMRRC) and crossed with the Cre-dependent tdTomato reporter line (JAX Stock No: 007909 or Ai9 line: Rosa26-floxed-tdTomato) or Cre-dependent channel rhodopsin 2 (ChR2) line (JAX Stock No: 024109 or Ai32 line: Rosa26-floxed-ChR2) [68] to generate D3-Cre/tdTomato or D3-Cre/ChR2 mice, respectively. Drd1-Cre (D1-Cre for simplicity; MMRRC Strain# 37156-Jax) mice were obtained from MMRRC [69]. Both male and female mice (8–12 weeks old) were used in all experiments and the data were pooled. Mice were housed on a 12 h light/dark cycle with food and water available ad libitum. Mice were group-housed until the surgery of viral injection and/or intra-cranial optical fiber implantation and singly-housed afterwards. All experimental procedures were performed in accordance with the guidelines of the National Institutes of Health and were approved by the Institutional Animal Care and Use Committees at the University of Pennsylvania.

Chronic Restraint Stress Model

Chronic restraint stress (CRS) treatment was performed as previously documented [39–41]. Briefly, mice were individually placed head first into a well-ventilated 50 ml polypropylene conical tube, which was then plugged with a 4.5-cm-long middle tube, and finally tied with the cap of the 50 ml tube. The restraint stress lasted for 2 hours per day (approximately from 9 to 11 am) for consecutive 14 days. After each session of the restraint stress, mice were returned to their home cage, where they were housed in pairs with food and water available ad libitum. Twenty-four hours after the last restraint session, mice were subjected to behavioral assays or electrophysiological recordings.

Virus/toxin Injection And Optical Fiber Implantation

Mice were anesthetized with isoflurane (~ 3% in oxygen) and secured in a stereotaxic system (Model 940, David Kopf Instruments). Isoflurane levels were maintained at 1.5-2% for the remainder of the surgery. Body temperature was maintained at 37 ℃ with a heating pad connected to a temperature control system (TC-1000, CWE Inc.). Local anesthetic (bupivacaine, 2 mg/kg, s.c.) was applied before skin incision and hole drilling on the dorsal skull. To target the islands of Calleja in the OT, we used two sets of coordinates from bregma: anteroposterior (AP) 1.2 (or 1.54) mm; mediolateral (ML), ± 1.1 (or ± 1.15) mm; dorsoventral (DV), -5.5 (or -5.0) mm. For the NAc, in order to retrogradely label more NAc-projecting VTA neurons, we injected CTB (recombinant, Alexa Fluor™ 555 conjugate, Invitrogen #C34776) into both the core and shell subregions: core, AP, 1.6 mm; ML, ± 0.8 mm; DV, -4.1 mm; medial shell, AP, 1.5 mm; ML, ± 0.55 mm; DV, -4.7 mm; lateral shell, AP, 0.98 mm; ML, ± 1.8 mm; DV, -4.92 mm. For viral injection, as appropriate, AAV8-CMV-TurboRFP-WPRE-rBG (2.94x10¹⁰ vg/ml), AAV8-EF1α-mCherry-FlEX-DTA (3.3x10⁹ viral units/ml) (University of North Carolina Viral Vector Core, Chapel Hill, NC) (800 nl each side), AAV1-DIO-ChR2-EYFP (C-34777, Life Technologies) (300–500 nl each side), AAV8-hSyn-DIO-hM3D(Gq)-mCherry (≥ 4×10¹² vg/ml; Addgene, cat.# 44361), AAV8-hSyn-dF-HA-KORD-IRES-mCitrine (≥ 7×10¹² vg/ml; Addgene, cat.# 65417) or AAV8-hSyn-DIO-mCherry (≥ 1×10¹³ vg/ml; Addgene, cat.# 50459) was bilaterally (300–500 nl each side) injected into the OT via a Hamilton syringe (5 µl) with a flow rate of 50 nl/min controlled by an Ultra Micro Pump III (UMP3) with a SYS-micro4 controller attachment (World Precision, Sarasota, USA). In D1-Cre mice, cholera toxin subunit B (300–500 nl) was also bilaterally injected into NAc. The tip of the syringe was left for 10–15 min after the injection. For optical fiber implantation, a cannula (CFMC14L10-Fiber Optic Cannula, Ø2.5 mm Ceramic Ferrule, Ø400 µm Core, 0.39 NA; Thorlabs, Newton, NJ), customized to 6 mm length, was unilaterally placed in the OT at the coordinates aforementioned and fixed on the skull with dental cement in D3-Cre/ChR2 mice. Mice were returned to home cage for recovery for one week before behavioral tests and mice with viral injection had at least three week waiting period before tests. All optical fiber implantation locations were post-hoc verified and only mice with the intended targeted site were included in data analysis.

Ex vivo electrophysiological recording

Whole-cell patch-clamp recordings were performed as described previously [30]. Briefly, mice were deeply anesthetized (ketamine-xylazine; 200 and 20 mg/kg body weight, respectively) and quickly decapitated. The dissected brain was immediately placed in ice-cold cutting solution containing (in mM) 92 N-Methyl D-glucamine, 2.5 KCl, 1.2 NaH₂PO₄, 30 NaHCO₃, 20 HEPES, 25 glucose, 5 Sodium L-ascorbate, 2 Thiourea, 3 Sodium Pyruvate, 10 MgSO₄, and 0.5 CaCl₂; osmolality ~ 300 mOsm and pH ~ 7.3, bubbled with 95% O₂-5% CO₂. Coronal sections (250 µm thick) containing the OT were cut using a Leica VT 1200S vibratome. Brain slices were incubated in oxygenated artificial cerebrospinal fluid (ACSF in mM: 124 NaCl, 3 KCl, 1.3 MgSO₄, 2 CaCl₂, 26 NaHCO₃, 1.25 NaH₂PO₄, 5.5 glucose, and 4.47 sucrose; osmolality ~ 305 mOsm and pH ~ 7.3, bubbled with 95% O₂-5% CO₂) for ~ 30 min at 31ºC and at least 30 minutes at room temperature before use. For recordings, slices were transferred to a recording chamber and continuously perfused with oxygenated ACSF. Fluorescent cells were visualized through a 40X water-immersion objective on an Olympus BX61WI upright microscope equipped with epifluorescence. Whole-cell patch-clamp recordings were made under both current and voltage clamp mode. Recording pipettes were made from borosilicate glass with a Flaming-Brown puller (P-97, Sutter Instruments; tip resistance 5–10 MΩ). The pipette solution contained (in mM) 120 K-gluconate, 10 NaCl, 1 CaCl₂, 10 EGTA, 10 HEPES, 5 Mg-ATP, 0.5 Na-GTP, and 10 phosphocreatine. Light stimulation was delivered through the same objective via pulses of blue laser (473 nm, FTEC2473-V65YF0, Blue Sky Research, Milpitas, USA) with 10 ms light pulse at 20 Hz. For light-evoked inhibitory postsynaptic currents (IPSCs), a high Cl⁻ intrapipette solution (120 mM KCl instead of K-gluconate) was used so that the reversal potential of [Cl⁻] was at ~ 0 mV and GABA_A receptor-mediated currents were inward at a holding potential of -60 mV. Light stimulation was delivered through the same objective via 10 ms pulses of blue laser (473 nm, FTEC2473-V65YF0, Blue Sky Research, Milpitas, USA). Viral infection in the OT was confirmed in brain slices during recordings. Pharmacological drugs CNO and SALB (dissolved in DMSO) were bath perfused.

Immunohistochemistry

After patch clamp recordings, acute brain slices prepared form D1-Cre mice with AAV1-DIO-ChR2-EYFP virus and cholera toxin subunit B bilaterally injected into the OT and NAc, respectively, were immediately incubated in 4% paraformaldehyde (PFA) for 10–15 min, and then transferred into 1X phosphate buffered solution (PBS) overnight. Slices were washed with PBS three times (20 min each), and then incubated with 0.5% Triton X-100 in PBS for 10 min, followed by incubation with 0.5% Triton X-100 and 0.5% Tween-20 in PBS for 10 min. Slices were incubated with a primary antibody against tyrosine hydroxylase (TH) (1:500; Millipore, cat# AB152, host: rabbit) for 24 h at 4°C. After three PBS washes (20 min each), the slices were incubated with a secondary antibody (1:1000; Life Technology, cat# A31573, host: donkey) and incubated for 24 h at 4°C. Slices were washed three more times in 0.5% Triton X-100 and 0.5% Tween-20 in PBS (20 min each) and treated with glycerol in PBS (volume ratio 1:1) for 30 min followed by glycerol in PBS (volume ratio 7:3) for 30 min before being mounted onto superfrost slides for confocal imaging.

Behavioral Tests

All behavioral procedures were performed during the light cycle (9:00 am -12:00 pm) except for the sucrose preference test which was conducted during the dark cycle (6:00 pm-6:00 am).

All mice tested were transferred to the testing room 1 h before the test for habituation. The following behavioral tests were performed sequentially (stress level from low to high): open field test (OFT), light-dark box transition test (LDT), elevated-zero maze (EZM) test, forced swimming test (FST) and tail suspension test (TST) with an interval of 24 h between two individual behavioral assays. All apparatuses were cleaned with 70% ethanol before and between trials. Mice in all behavioral tests were videotaped using a webcam at 30 frames/sec, and behaviors were scored using the ANY-maze software (Stoelting Co.) or manually by those who were blinded to the experimental conditions.

Optogenetic and chemogenetic manipulations For optogenetic experiments, blue light (473 nm, 10–15 mW/mm², 20 Hz, 10 ms pulse, 10 s stimulation every 30 s) was applied to activate ChR2-expressing neurons after each daily CRS treatment for 15 min and during behavioral tests. Green light (532 nm), less efficient in activating these neurons, was applied with the same parameters as control. D3-Cre mice with AAV8-hSyn-DIO-hM3D(Gq)-mCherry (excitatory DREADD) and AAV8-hSyn-dF-HA-KORD-IRES-mCitrine (inhibitory DREADD) or AAV8-hSyn-DIO-mCherry (control virus) in the OT were intraperitoneally injected with saline or CNO (5 mg/kg) (for excitatory DREADD), or subcutaneously injected with DMSO or SALB (10 mg/kg) (for inhibitory DREADD), 30 min before behavioral tests.

Open field test (OFT) The mice were placed in an open field arena (40 cm x 40 cm) in a room with dim light and allowed to freely explore the apparatus for 5 min. The total distance travelled and the total time in the central zone (20 cm x 20 cm) were calculated.

Light-dark box transition test (LDT) The light-dark box (46 cm x 28 cm x 30 cm) was composed of two compartments. Two-thirds of the box was the light compartment and the remaining part was the dark compartment. Mice were placed in the center of the light box with the head oppositely facing the dark compartment, and were allowed to freely explore the two compartments for 10 min. The latency to the first entry into the dark compartment and the total time spent in the dark compartment were calculated.

Elevated zero maze (EZM) test The zero-maze, composed of two open and two closed sections, was elevated 80 cm above the floor. The test mice were placed at the interface of an open and a closed section with the head facing the closed section. The latency to the first entry into the open section and the total time spent in the open sections were calculated.

Forced swimming test (FST) The FST, commonly used to assess antidepressant activity [70, 71], was also used for testing depression-like behaviors (or “learned helplessness”) in rodents [72, 73]. Briefly, the mouse was placed in a cylindrical plexiglass tank (40 cm high × 12 cm in diameter) with water (25 ± 2°C) in a depth of 10 cm. Mice were submitted to the swimming test for 6 min on the test day. The time spent in immobility was calculated post-hoc. Immediately after the FST, each mouse was removed from the water, towel-dried, and returned to its home cage. The water was changed and the cylinder was cleaned for each mouse tested.

Tail suspension test (TST) The TST is widely used for testing “behavioral despair”, another depression-related symptom [74]. Briefly, the mouse, held by the tail, was suspended 50 cm from the floor for 6 min. The time spent in immobility was recorded. The mice were considered immobile only when they hung down passively and were completely motionless.

Sucrose preference test Prior to the test, mice were habituated to the presence of two drinking bottles (one containing 2% sucrose and the other plain water) for 2 days in their home cage. Following this acclimation, mice had the free choice of either drinking the 2% sucrose solution or plain water for a period of 4 days. Water and sucrose solution intake was measured daily, and the positions of the two bottles were switched daily to reduce any confounding side bias. Sucrose preference was calculated as a percentage of the weight of sucrose intake over the total weight of fluid intake and averaged over the 4 days of testing. Some D3-Cre/ChR2 mice with optical fiber implantation in the OT were photostimulated for 15 min after each daily CRS treatment and during the tests with the same parameters as aforementioned. For chemogenetic manipulations, D3-Cre mice with AAV8-hSyn-DIO-hM3D(Gq)-mCherry and AAV8-hSyn-dF-HA-KORD-IRES-mCitrine in the OT were intraperitoneally (I.P.) injected with saline or CNO (5 mg/kg) (after two-week’s CRS treatment), or subcutaneously injected with DMSO or SALB (10 mg/kg) (without CRS treatment), respectively, one time each day with only one drug in a counterbalanced way during the four days test session.

Conditioned place preference A custom-built conditioned place preference (CPP) apparatus consisted of a rectangular cage with three compartments: a left black chamber (35 cm × 20 cm) with a metal wire mesh floor, a connecting zone (35 cm × 10 cm) with a smooth gray floor, and a right white chamber (35 cm × 20 cm) with a soft floor. The CCP test was conducted as previously described [75]. Briefly, the CPP test consisted of 4 days. Day 1 (pre-conditioning, 15 min): a preconditioning test was performed to obtain a baseline preference for the apparatus of each mouse tested. The side chamber a mouse spent the most time was assigned as the most preferred side, and the other side chamber as the least preferred side. On days 2 and 3 (conditioning): the mouse was firstly kept in either the most or least preferred side (counterbalanced across mice) for 15 min, then transferred to the other side for 15 min. Mice in the least-preferred side were paired with blue light stimulation (same parameters as aforementioned), and no photostimulations for mice in the most preferred side. On Day 4 (post-conditioning, 15 min): 24 h after the conditioning session on Day 3, the mice were placed back into the arena with all three compartments accessible, to evaluate preference for the stimulation and non-stimulation paired chambers. The CPP difference score was calculated by the time spent in the post-conditioning session minus that in the pre-conditioning session in the corresponding chamber.

Rectification of potential grooming-related effects in behavioral tests Since blue light activation of OT D3 neurons induces orofacial grooming, it may deviate some of the measurements of the behavioral tests; e.g., underestimation of the total distance travelled in the OFT and overestimation of the activity time in the FST and TST. We therefore rectified potential grooming-related deviations under light stimulation as follows. We first quantified the total duration of grooming induced by blue light activation of OT D3 neurons in the 5 min OFT - approximately 60 s grooming out of 100 s light stimulation. In the OFT, the rectified total distance travelled = (the actual distance x 300s/240s) as mice did not travel during the 60 s of blue light-induced grooming. For potentially overestimated parameters including the time and latency in anxiety tests (OFT, LDT and EZM) and the activity time in depression tests (FST and TST), we subtracted the presumptive grooming time proportionally to the blue light duration for each measurement.

Confocal Imaging

Mice were perfused transcardially with 4% paraformaldehyde (PFA) in fresh phosphate buffered saline (PBS). The brain was post fixed in 4% PFA overnight at 4 ºC and then transferred into PBS. Coronal slices (100 µm thick) were prepared using a Leica VT 1200S vibratome. The slices were treated with glycerol in PBS (volume ratio 1:1) for 30 min followed by glycerol in PBS (volume ratio 7:3) for 30 min before being mounted onto superfrost slides for imaging. Confocal imaging was performed by sequential scanning of slices at 10 x and 40 x in a SP5/Leica confocal microscope.

Statistical analysis

Shapiro-Wilk tests were used to verify normal distribution of datasets. For normally distributed datasets, parametric statistical tests (student’s t test or two-way ANOVA test) were used; otherwise, non-parametric tests (Mann-Whitney or Wilcoxon matched-pairs signed rank test) were applied. Statistical analysis was performed in GraphPad Prism and figures were assembled in Adobe Photoshop.

References

1. Fox, M.E. and M.K. Lobo, The molecular and cellular mechanisms of depression: a focus on reward circuitry. Mol Psychiatry, 2019. 24(12): p. 1798-1815.

2. Francis, T.C. and M.K. Lobo, Emerging role for nucleus accumbens medium spiny neuron subtypes in depression. Biological psychiatry, 2017. 81(8): p. 645-653.

3. Heshmati, M. and S.J. Russo, Anhedonia and the brain reward circuitry in depression. Current behavioral neuroscience reports, 2015. 2(3): p. 146-153.

4. Malone Jr, D.A., et al., Deep brain stimulation of the ventral capsule/ventral striatum for treatment-resistant depression. Biological psychiatry, 2009. 65(4): p. 267-275.

5. Hou, Z., et al., Increased temporal variability of striatum region facilitating the early antidepressant response in patients with major depressive disorder. Progress in Neuro-Psychopharmacology and Biological Psychiatry, 2018. 85: p. 39-45.

6. Kumar, P., et al., Impaired reward prediction error encoding and striatal-midbrain connectivity in depression. Neuropsychopharmacology, 2018. 43(7): p. 1581.

7. Morgan, J.K., et al., Maternal response to positive affect moderates the impact of familial risk for depression on ventral striatal response to winning reward in 6-to 8-year-old children. Biological Psychiatry: Cognitive Neuroscience and Neuroimaging, 2022.

8. Burke, D.A., H.G. Rotstein, and V.A. Alvarez, Striatal local circuitry: a new framework for lateral inhibition. Neuron, 2017. 96(2): p. 267-284.

9. Kreitzer, A.C. and R.C. Malenka, Striatal plasticity and basal ganglia circuit function. Neuron, 2008. 60(4): p. 543-554.

10. Wesson, D.W., The tubular striatum. Journal of Neuroscience, 2020. 40(39): p. 7379-7386.

11. Murata, K., et al., Opposing roles of dopamine receptor D1-and D2-expressing neurons in the anteromedial olfactory tubercle in acquisition of place preference in mice. Frontiers in behavioral neuroscience, 2019. 13: p. 50.

12. Zhang, Z., et al., Activation of the dopaminergic pathway from VTA to the medial olfactory tubercle generates odor-preference and reward. Elife, 2017. 6: p. e25423.

13. Gadziola, M.A. and D.W. Wesson, The neural representation of goal-directed actions and outcomes in the ventral striatum's olfactory tubercle. Journal of Neuroscience, 2016. 36(2): p. 548-560.

14. Murata, K., et al., Mapping of learned odor-induced motivated behaviors in the mouse olfactory tubercle. Journal of Neuroscience, 2015. 35(29): p. 10581-10599.

15. Gadziola, M.A., et al., The olfactory tubercle encodes odor valence in behaving mice. Journal of Neuroscience, 2015. 35(11): p. 4515-4527.

16. DiBenedictis, B.T., et al., DREADD-induced silencing of the medial olfactory tubercle disrupts the preference of female mice for opposite-sex chemosignals. Eneuro, 2015. 2(5).

17. Agustín-Pavón, C., F. Martínez-García, and E. Lanuza, Focal lesions within the ventral striato-pallidum abolish attraction for male chemosignals in female mice. Behavioural brain research, 2014. 259: p. 292-296.

18. Ikemoto, S., Involvement of the olfactory tubercle in cocaine reward: intracranial self-administration studies. Journal of Neuroscience, 2003. 23(28): p. 9305-9311.

19. Cheng, J., et al., HCN2 channels in cholinergic interneurons of nucleus accumbens shell regulate depressive behaviors. Neuron, 2019. 101(4): p. 662-672. e5.

20. Francis, T.C., et al., Nucleus accumbens medium spiny neuron subtypes mediate depression-related outcomes to social defeat stress. Biological psychiatry, 2015. 77(3): p. 212-222.

21. Francis, T.C., et al., The Selective RhoA Inhibitor Rhosin Promotes Stress Resiliency Through Enhancing D1-Medium Spiny Neuron Plasticity and Reducing Hyperexcitability. Biological psychiatry, 2019. 85(12): p. 1001-1010.

22. Lim, B.K., et al., Anhedonia requires MC4R-mediated synaptic adaptations in nucleus accumbens. Nature, 2012. 487(7406): p. 183.

23. Wang, J., et al., Dopamine D3 receptor in the nucleus accumbens alleviates neuroinflammation in a mouse model of depressive-like behavior. Brain, behavior, and immunity, 2022. 101: p. 165-179.

24. Miyamoto, Y., et al., Striatal N-acetylaspartate synthetase Shati/Nat8l regulates depression-like behaviors via mGluR3-mediated serotonergic suppression in mice. International Journal of Neuropsychopharmacology, 2017. 20(12): p. 1027-1035.

25. Warner-Schmidt, J.L., et al., Cholinergic interneurons in the nucleus accumbens regulate depression-like behavior. Proceedings of the National Academy of Sciences, 2012. 109(28): p. 11360-11365.

26. Fox, M.E., et al., Dendritic remodeling of D1 neurons by RhoA/Rho-kinase mediates depression-like behavior. Molecular psychiatry, 2018: p. 1.

27. LeGates, T.A., et al., Reward behaviour is regulated by the strength of hippocampus–nucleus accumbens synapses. Nature, 2018. 564(7735): p. 258-262.

28. Martin-Lopez, E., et al., Embryonic and postnatal development of mouse olfactory tubercle. Molecular and Cellular Neuroscience, 2019. 98: p. 82-96.

29. Millhouse, O. and L. Heimer, Cell configurations in the olfactory tubercle of the rat. Journal of Comparative Neurology, 1984. 228(4): p. 571-597.

30. !!! INVALID CITATION !!!

31. Inta, D., et al., Adult neurogenesis in the human striatum: possible implications for psychiatric disorders. Mol. Psychiatry, 2016. 21: p. 446-447.

32. Inta, D., A. Meyer-Lindenberg, and P. Gass, Alterations in postnatal neurogenesis and dopamine dysregulation in schizophrenia: a hypothesis. Schizophrenia bulletin, 2011. 37(4): p. 674-680.

33. Prokop, S., et al., PharmacoSTORM nanoscale pharmacology reveals cariprazine binding on Islands of Calleja granule cells. Nature communications, 2021. 12(1): p. 1-19.

34. Zhang, Y.-F., et al., Ventral striatal islands of Calleja neurons control grooming in mice. Nature neuroscience, 2021. 24(12): p. 1699-1710.

35. Kalueff, A.V., et al., Neurobiology of rodent self-grooming and its value for translational neuroscience. Nature Reviews Neuroscience, 2016. 17(1): p. 45-59.

36. Smolinsky, A.N., et al., Analysis of grooming behavior and its utility in studying animal stress, anxiety, and depression, in Mood and anxiety related phenotypes in mice. 2009, Springer. p. 21-36.

37. Sun, J., et al., Excitatory SST neurons in the medial paralemniscal nucleus control repetitive self-grooming and encode reward. Neuron, 2022.

38. Willner, P., The chronic mild stress (CMS) model of depression: History, evaluation and usage. Neurobiology of stress, 2017. 6: p. 78-93.

39. Oh, Y.-S., et al., SMARCA3, a chromatin-remodeling factor, is required for p11-dependent antidepressant action. Cell, 2013. 152(4): p. 831-843.

40. Seo, J.-S., et al., Cellular and molecular basis for stress-induced depression. Molecular psychiatry, 2016.

41. Seo, J., et al., Elevation of p11 in lateral habenula mediates depression-like behavior. Molecular Psychiatry, 2017.

42. Chiba, S., et al., Chronic restraint stress causes anxiety-and depression-like behaviors, downregulates glucocorticoid receptor expression, and attenuates glutamate release induced by brain-derived neurotrophic factor in the prefrontal cortex. Progress in Neuro-Psychopharmacology and Biological Psychiatry, 2012. 39(1): p. 112-119.

43. Joo, Y., et al., Chronic immobilization stress induces anxiety-and depression-like behaviors and decreases transthyretin in the mouse cortex. Neuroscience letters, 2009. 461(2): p. 121-125.

44. Mu, M.-D., et al., A limbic circuitry involved in emotional stress-induced grooming. Nature communications, 2020. 11(1): p. 1-16.

45. Song, C., K.C. Berridge, and A.V. Kalueff, 'Stressing'rodent self-grooming for neuroscience research. Nature Reviews Neuroscience, 2016. 17(9): p. 591-591.

46. Fernández-Teruel, A. and C. Estanislau, Meanings of self-grooming depend on an inverted U-shaped function with aversiveness. Nature Reviews Neuroscience, 2016. 17(9): p. 591-591.

47. Xu, Y., et al., Identification of a neurocircuit underlying regulation of feeding by stress-related emotional responses. Nature communications, 2019. 10(1): p. 1-16.

48. Vardy, E., et al., A new DREADD facilitates the multiplexed chemogenetic interrogation of behavior. Neuron, 2015. 86(4): p. 936-946.

49. Zhang, Y.-F., et al., Self-directed orofacial grooming promotes social attraction in mice via chemosensory communication. iScience, 2022: p. 104284.

50. Tye, K.M., et al., Dopamine neurons modulate neural encoding and expression of depression-related behaviour. Nature, 2013. 493(7433): p. 537-541.

51. Yadid, G., D.H. Overstreet, and A. Zangen, Limbic dopaminergic adaptation to a stressful stimulus in a rat model of depression. Brain research, 2001. 896(1-2): p. 43-47.

52. Zhang, Z., et al., Whole-brain mapping of the inputs and outputs of the medial part of the olfactory tubercle. Frontiers in Neural Circuits, 2017. 11: p. 52.

53. Koyama, S., et al., Obesity decreases excitability of putative ventral tegmental area GABA ergic neurons. Physiological Reports, 2013. 1(5).

54. Zhang, T.A., A.N. Placzek, and J.A. Dani, In vitro identification and electrophysiological characterization of dopamine neurons in the ventral tegmental area. Neuropharmacology, 2010. 59(6): p. 431-436.

55. Ramos, A., Animal models of anxiety: do I need multiple tests? Trends in pharmacological sciences, 2008. 29(10): p. 493-498.

56. Bouwknecht, J.A. and R. Paylor, Behavioral and physiological mouse assays for anxiety: a survey in nine mouse strains. Behavioural brain research, 2002. 136(2): p. 489-501.

57. Knowland, D., et al., Distinct ventral pallidal neural populations mediate separate symptoms of depression. Cell, 2017. 170(2): p. 284-297. e18.

58. Virk, M.S., et al., Opposing roles for serotonin in cholinergic neurons of the ventral and dorsal striatum. Proceedings of the National Academy of Sciences, 2016. 113(3): p. 734-739.

59. An, K., et al., A circadian rhythm-gated subcortical pathway for nighttime-light-induced depressive-like behaviors in mice. Nature Neuroscience, 2020: p. 1-12.

60. Tong, Q., et al., D1 receptor-expressing neurons in ventral tegmental area alleviate mouse anxiety-like behaviors via glutamatergic projection to lateral septum. Molecular Psychiatry, 2022: p. 1-14.

61. Le Foll, B., J. Diaz, and P. Sokoloff, Neuroadaptations to hyperdopaminergia in dopamine D3 receptor-deficient mice. Life sciences, 2005. 76(11): p. 1281-1296.

62. Chourbaji, S., et al., Dopamine receptor 3 (D3) knockout mice show regular emotional behaviour. Pharmacological research, 2008. 58(5-6): p. 302-307.

63. Xing, B., et al., Effects of immobilization stress on emotional behaviors in dopamine D3 receptor knockout mice. Behavioural brain research, 2013. 243: p. 261-266.

64. Leggio, G.M., et al., Dopamine D3 receptor as a new pharmacological target for the treatment of depression. European journal of pharmacology, 2013. 719(1-3): p. 25-33.

65. Moraga-Amaro, R., et al., Dopamine receptor D3 deficiency results in chronic depression and anxiety. Behavioural brain research, 2014. 274: p. 186-193.

66. Shuen, J.A., et al., Drd1a-tdTomato BAC transgenic mice for simultaneous visualization of medium spiny neurons in the direct and indirect pathways of the basal ganglia. Journal of Neuroscience, 2008. 28(11): p. 2681-2685.

67. Gong, S., et al., A gene expression atlas of the central nervous system based on bacterial artificial chromosomes. Nature, 2003. 425(6961): p. 917.

68. Madisen, L., et al., A robust and high-throughput Cre reporting and characterization system for the whole mouse brain. Nature neuroscience, 2010. 13(1): p. 133.

69. Zhang, J., et al., c-Fos facilitates the acquisition and extinction of cocaine-induced persistent changes. Journal of Neuroscience, 2006. 26(51): p. 13287-13296.

70. Petit-Demouliere, B., F. Chenu, and M. Bourin, Forced swimming test in mice: a review of antidepressant activity. Psychopharmacology (Berl), 2005. 177(3): p. 245-55.

71. Slattery, D.A. and J.F. Cryan, Using the rat forced swim test to assess antidepressant-like activity in rodents. Nature Protocols, 2012. 7(6): p. 1009-1014.

72. Gonzalez, M.M. and G. Aston-Jones, Light deprivation damages monoamine neurons and produces a depressive behavioral phenotype in rats. Proc Natl Acad Sci U S A, 2008. 105(12): p. 4898-903.

73. Zhang, Y.-F., et al., Alterations of motor cortical microcircuit in a depressive-like mouse model produced by light deprivation. Neuroscience, 2016.

74. Steru, L., et al., The tail suspension test: a new method for screening antidepressants in mice. Psychopharmacology (Berl), 1985. 85(3): p. 367-70.

75. Stamatakis, A.M. and G.D. Stuber, Activation of lateral habenula inputs to the ventral midbrain promotes behavioral avoidance. Nature neuroscience, 2012. 15(8): p. 1105.