BNST PACAP signaling in the dorsal medial habenula mediates depression-related behavioral responses after chronic stress

Abstract

BACKGROUND:

Hallmark behaviors associated with depression-related disorders include reduction in reward responsivity and increased despair. However, the neurocircuits and mechanisms underlying reward responding and despair-like behavior have remained unclear. Our current studies show that pituitary adenylate cyclase activating polypeptide (PACAP, ADCYAP1) connections from the bed nucleus of the stria terminalis (BNST) to the dorsal medial habenula (dMHb) may contribute to depression-like behaviors.

METHODS:

Viral tracing of BNST afferents was examined in PACAP-Cre mice; MHb immunocytochemical staining for PACAP and the PAC1 receptor was also performed. Reward responsiveness was measured using the sucrose preference test (SPT) following MHb-targeted infusions of PACAP, PAC1 receptor antagonist (PACAP(6–38); N-terminally truncated PACAP peptide), or vasoactive intestinal peptide (VIP) in rats. Despair-like behavior was measured in rats using the forced swim test following PACAP infusions. The ability PACAP(6–38) to blunt chronic stress-induced depression-like responses was also evaluated. Finally, the BNST → dMHb PACAPergic pathway was chemogenetically activated and the effect on the SPT was measured.

RESULTS:

PACAP and PAC1 receptors are both expressed in the MHb, and BNST PACAP neurons have afferent projections to the dMHb. PACAP signaling in the MHb is necessary and sufficient for depression-like behaviors in rats and chemogenetic activation of the BNST → dMHb PACAPergic pathway reduces sucrose preference score in mice.

CONCLUSIONS:

We identify a novel PACAPergic BNST → dMHb pathway and characterize its involvement in depression-like responses. These results further our understanding of the neurocircuitry underlying depression-like behaviors, particularly the behavioral roles of BNST PACAP → dMHb signaling.

Introduction

Psychopathologies associated with stress-responding and disruptions in emotional systems represent a significant health concern worldwide, particularly as they relate to disorders of mood. A significant body of literature has now shown that chronic or repeated activation of stress response systems can lead to functional changes in emotion-related neurocircuits that produce long-lasting and sometimes maladaptive changes in behavioral responding. One manifestation of chronic or severe stress is major depressive disorder (MDD) which represents one of the most common mental disorders in the United States with 8.3% of adults having at least one major depressive episode within the past year (1). Further, following the Covid-19 pandemic, the prevalence of depression world-wide spiked several-fold (2). Some of the behavioral hallmarks of MDD include anhedonia or the inability to derive pleasure, pain sensitivity, social avoidance, depressed motivation, lethargy, suppression of active defensive responding and behavioral despair. The effects of MDD can be debilitating and as the neurocircuits and mechanisms underlying stress-induced depressive disorders remain very much unclear, the development of targeted therapeutics has significant challenges.

The bed nucleus of the stria terminalis (BNST) is a forebrain nucleus and represents a nexus of principal stress-associated behavioral neurocircuits (3). The BNST has been implicated as key to threat responding and has been argued to be recruited by stimuli that predict future threat - stimuli that elicit sustained threat responding and/or temporally unpredictable threatening stimuli (4, 5). Accordingly, BNST activation has been shown to drive anxiety-related responses, facilitate downstream fear/panic circuits and upon chronic stimulation may recruit or activate pathways that promote depression (6–8).

Many studies have now shown critical roles for pituitary adenylate cyclase activating polypeptide (PACAP, ADCYAP1) in stress responding at every level of stress response systems (9). BNST PACAP expression is elevated following a chronic variate stress procedure (CVS) and BNST PACAP/PAC1 receptor signaling is both necessary and sufficient for many of the consequences of CVS in rodents (9–11). In addition to endogenous BNST PACAP production, the BNST also receives significant PACAP afferents from the lateral parabrachial nucleus (LPBn) (12, 13), and activation of these LPBn PACAP afferents in the BNST is anxiogenic (14). Translationally, circulating PACAP levels and a single nucleotide polymorphism (SNP) in the PAC1 receptor gene (ADCYAP1R1) are associated with PTSD symptoms in women, and methylation of the PAC1 receptor gene is associated with PTSD symptoms in both men and women (15). SNPs in the PAC1 gene are also associated with BNST-dependent dark-enhanced startle task (15, 16); further, PACAP genes have been associated with MDD (17). In aggregate, these studies demonstrate that BNST PACAP activation is critical for many consequences of chronic stress.

In our ongoing investigations of the downstream targets of BNST PACAP neurons, we employed Cre-dependent reporters in PACAP-Cre mice for neurocircuit tracings and observed a novel robust BNST PACAP neuronal projection to the medial habenula (MHb), an epithalamic structure that has been increasingly implicated in a number of behavioral processes including reward, depression (18–20) and nociception/pain responding (21, 22). Accordingly, we have further examined this projection pathway to determine whether PACAP/PAC1 receptor signaling in the MHb participates in behavioral responses in rodents that have associated with anhedonia as well as passive coping behavior.

Methods and Materials

Detailed methods and extended data are available in Supplementary Information.

Animals

Sprague-Dawley rats were from Charles River Canada and PACAP-IRES-Cre breeding mice were a gift from Michael Krashes and Bradford Lowell (Beth Israel Deaconess Hospital, Harvard Medical School). All the procedures were approved by the University of Vermont Institutional Animal Care and Use Committee.

Stereotaxic surgeries and treatments

In neurocircuit tracings, 0.5 μl AAV2-EF1A-DIO-mCherry vector was infused into the BNST of PACAP-Cre mice using coordinates from bregma in mm: AP = +0.1, ML = ±1.0, DV = −4.6. For rat MHb peptide infusions (PACAP, PACAP(6–38) or VIP; 0.5 – 1 μg/0.5 μl), the cannulae were lowered at a 15° angle above the MHb with coordinates from bregma in mm; AP = −3.4, ML = ± 1.8 and DV = −3.8. The chemogenetic studies were performed in PACAP-Cre mice by first injecting AAV2-hSyn-DIO-hM3D(Gq)-mCherry into the BNST using coordinates in mm: AP = +0.3, ML = ±2.6, DV = −4.3, at a 20° angle. Simultaneously, cannulae targeting the MHb were placed using coordinates AP = −1.3, ML = ±1.2, DV = −3.7, 20° angle, for subsequent clozapine-N-oxide (CNO; 1 mM/0.3 μl) at the PACAP BNST → MHb projections.

Tissue Immunocytochemistry

Mouse brain sections were prepared and immunocytochemically processed as described previously (11–13, 23, 24).

Behavioral tests

Chronic stress in rats were performed as described previously but over 14 days (10, 25). In sucrose preference tests, rats were first presented with one bottle containing tap water and another bottle with 2% sucrose. Following 24 h food/water withdrawal, the rats received bilateral MHb peptide or vehicle infusions before presentation of the same water and 2% sucrose bottles for unrestricted consumption over 60 min. For forced swim tests, rats with MHb peptide infusions were placed into a room temperature water-filled container for 5 min for immobility measurements.

Statistics

Statistics were analyzed using Student’s t test, one-way ANOVA with Tukey’s HSD for the two-week CVS antagonist test, or 2×2 factorial ANOVA. Data analysis was done using GraphPad Prism and IBM SPSS Statistics.

Results

Cre-dependent mCherry reporter expression in PACAP-Cre mice demonstrates BNST PACAP neurons project to the dorsal medial habenula (dMHb)

The BNST represents a nexus of stress circuits, receiving and relaying stress signals that can integrate and coordinate physiological and behavioral responses. The BNST not only receives PACAP afferents from stress-associated nuclei, notably such as the lateral parabrachial nucleus (LPBn) (12, 13, 26), but expresses PACAP endogenously that appears to be regulated by chronic stress (10). The endogenous BNST PACAP neurons may have activities intrinsically within the BNST and may have long axonal projections to more distal regulatory nuclei. To better appreciate the roles of these intrinsic BNST PACAP neurons, we infused a Cre-dependent mCherry reporter vector in the BNST of PACAP-Cre mice to establish the BNST PACAP neuronal fiber tracts.

As anticipated, one prominent BNST PACAP projection appeared within the stria terminalis which was followed into amygdala nuclei (Supplemental Figure 1). However, slightly posterior to these amygdala-projecting neurons, another PACAP fiber projection was readily apparent in the stria medullaris that could be followed in successive tissue sections to terminate as large axonal boutons in the medial habenula (Figure 1). A population of PACAP-immunoreactive neurons could be identified in the dorsolateral part of the medial habenula as described previously (36) whereas PACAP-immunoreactive nerve fibers were found in the remaining and more medial aspects of the MHb (Figure 2A, 2C, 2D). PAC1 receptor immunoreactivity was also found in the medial habenula in good agreement with previous transcript work (Figure 2B – 2F) (27). At high magnification, the PACAP nerve fibers were found in close apposition to PAC1 receptor-expressing neurons most likely representing synaptic contacts (Figure 2F, arrows). The medial habenula can be further compartmentalized into dorsal and ventral subnuclei with different phenotypic identities (28, 29). The dorsal medial habenula (dMHb), distinguished by a population of tachykinin (Tac1, substance P (SP)) neurons and the ventral medial habenula (vMHb), identified by choline acetyltransferase (ChAT) expression, can exhibit differential functions. When tissue sections from the PACAP-Cre tracing studies were processed for ChAT immunoreactivity, the BNST PACAP axonal boutons were clearly in the dMHb and segregated from vMHb ChAT neurons (Figure 2G – 2I). These observations implicated BNST PACAP distant signaling onto dMHb tachykinin neurons.

Figure 1.

BNST PACAP neurons have axonal projections to the medial habenula. A) The BNST of PACAP-Cre mice were infused with AAV2-hSyn-DIO-mCherry vector for reporter expression in PACAP neurons. The bulk of the BNST PACAP neuronal projections traveled in stria terminalis (see Supplemental Figure 1); a significant fraction of the fibers was also found in the stria medullaris. The stria medullaris PACAP fibers could be traced in successive cryosections to terminate in the medial habenula (MHb, panels B – E) in large synaptic bouton clusters. 3V, third ventricle; Scale bars, 100 μm.

Figure 2.

Medial habenula expresses PACAP and PAC1 receptor immunocytochemistry. Mouse brain tissue sections were immunocytochemically processed for dual PACAP (A, green) and PAC1 receptor (B, red) expression and colocalization (C, yellow; merge). PACAP immunoreactivity was apparent in both the medial (MHb) and lateral (LHb) habenula; there was heavy PACAP immunoreactivity in the paraventricular (PV) thalamic nucleus. PAC1 receptor immunoreactivity appeared dense in the MHb, LHb and PV. Panels D – F, increasingly high magnification of the MHb show PACAP fiber immunoreactivity (green) in apparent contact with PAC1 receptors (red) that circumscribe MHb neurons. Boxed insets show enlarged areas in succeeding panels. Magenta arrows in Panel F illustrate apparent PACAP fiber terminal coincidence with PAC1 receptors (white puncta). Panels G - I, cryosection shown in Figure 1D with MHb mCherry reporter expression in BNST PACAP fiber terminals (D, red) was also immunocytochemically processed for choline acetyltransferase (ChAT; E, green). As described previously, ChAT immunoreactivity was restricted to the LHb; the BNST PACAP terminals were found in the dMHb and did not colocalize with ChAT (F, merge). 3V, third ventricle; dMHb, dorsal medial habenula; vMHb, ventral medial habenula; DG, dentate gyrus. Scale bars: Panels A – C, 200 μm; panel D, 100 μm; panel E, 10 μm; panel F, 7 μm; panels G – I, 10 μm.

MHb PACAP signaling increases depression-associated behaviors in unstressed male rats.

The MHb has been well studied with respect to aversive states and mood regulation; the dMHb in particular has been associated with mood, stress and reward processing that contribute to aversion, anxiety, depression and anhedonia-related behaviors. These initial behavior studies were evaluated in rats. First, we infused PACAP into the MHb of unstressed male or female rats for sucrose preference tests as described in Methods. Following acclimation in their home cages with bottles of water and 2% sucrose, the rats were deprived of food and water for 3 h before subsequent MHb PACAP or vehicle infusions and sucrose water consumption measurements over 60 min. From these studies, MHb PACAP infusions into male rats facilitated anhedonia and decreased sucrose preference scores in a dose-dependent manner (Figure 3A). Whereas MHb infusions with 0.5 μg PACAP peptide significantly decreased the sucrose preference score approximately 9% (PACAP, M = 63.86, SD = 7.00) compared to control groups (vehicle, M = 69.56, SD = 8.17; t(29) = 2.08, p = 0.047), similar infusions with 1.0 μg peptide decreased sucrose preference 30% (PACAP, M = 48.53, SD = 13.23) against vehicle-treated animals (vehicle, M = 67.63, SD = 8.93; t(11) = 2.99, p = 0.012). Interestingly, these PACAP responses were observed only in male and not female rats (Figure 3C). Despite similar total consumption, there were no apparent differences in sucrose preference between MHb PACAP (1.0 μg) (PACAP, M = 60.98, SD = 14.37) and vehicle infusions in female groups (vehicle, M = 67.08, SD = 10.54; t(12) = 0.92, p = 0.37).

Figure 3.

MHb PACAP infusions produced anhedonia- and depression-like responses. Male rats were infused with PACAP for sucrose preference or forced swim tests as described in Methods. The anhedonia-like responses in sucrose preference tests appeared markedly greater with higher PACAP concentrations (panel A, compare 0.5 vs 1.0 μg PACAP infusions). MHb VIP infusions in male rats had no apparent effects (panel B); MHb PACAP infusions also had no apparent effects in female rats (panel C). MHb PACAP infusions into male rats also increased immobility times in forced swim tests (panel D) without affecting latency to immobility (panel E).

PACAP binds with high affinity not only to its selective cognate PAC1 receptor but also to the vasoactive intestinal peptide (VIP) VPAC1 and VPAC2 receptors (30). To distinguish whether the MHb PACAP-induced anhedonia-like responses were mediated by PAC1 or VPAC receptors, the exact same experimental sucrose preference paradigm was performed using VIP peptide. However, unlike the PACAP-mediated anhedonia-related responses, male rats that received VIP infusions (VIP, M = 70.22, SD = 21.00) did not differ from the vehicle control group (vehicle, M = 70.84, SD = 16.53; t(11) = 0.059, p = 0.95), suggesting that the MHb PACAP effects were PAC1 receptor dependent (Figure 3B).

To assess MHb PACAP effects on passive coping, PACAP was infused into the BNST of unstressed male rats prior to the forced swim test. Immobility times for PACAP infused animals were greater than those treated with vehicle (t(20) = 2.159, p = 0.0432; Figure 3D), although the latency to immobility did not differ (t(20) = 0.8783, p = 0.3902) (Figure 3E). Together these results suggest that MHb PACAP infusion increases behaviors related to depression (both anhedonia and passive coping) via PAC1 receptors in previously unstressed rats.

PAC1 receptor antagonism blocks chronic stress-induced anhedonia in male rats

Chronic stress can cause a variety of maladaptive behavioral responses including diminished reward responding and anhedonia. To evaluate whether MHb PACAP/PAC1 receptor signaling participates in chronic stress-induced anhedonia-like behavior, we examined the ability of a MHb PACAP receptor antagonist to block the chronic stress-induced decreases in sucrose preference in rats. As shown, a two-week chronic variate stress paradigm (CVS) was sufficient to reduce sucrose consumption in male rats (Figure 4). To assess the roles of MHb PACAP signaling in these responses, the rats were exposed to the same 2 week CVS paradigm and the PAC1 receptor antagonist PACAP(6–38) was infused into the MHb before administering the sucrose preference test (Figure 4). A 2 × 2 factorial ANOVA revealed a significant main effect of infusion (F(1,43) = 8.76; p = 0.005) and a significant interaction between stress and infusion (PACAP(6–38) or vehicle, F(1,43) = 6.13; p = 0.017). Post-hoc testing with Tukey’s HSD revealed significant pair-wise differences between the stress/vehicle group (M = 49.81, SD = 10.58) and the stress/PACAP(6–38) (M = 68.59, SD = 8.53), no stress/vehicle (M = 65.31, SD = 13.03), and no stress/PACAP(6–38) (M = 65.31, SD = 13.03) groups. As in previous work, these observations suggested that PACAP/PAC1 receptor signaling is sustained in mediating the effects of chronic stress and that PAC1 receptor antagonism during stress or even after stressor termination can ameliorate the maladaptive effects. In the current studies, MHb PAC1 receptor antagonism with PACAP(6–38) was capable of alleviating the stress-induced anhedonia-like decrease in sucrose preference. These studies in aggregate suggest that the BNST PACAP → MHb projection may be a contributory neurocircuit in mediating the panoply of maladaptive behavioral consequences of chronic stress.

Figure 4.

Chronic variate stress-induced anhedonia-like responses were blocked with MHb infusions with PAC1 receptor antagonist. Male rats with MHb cannula implants were exposed to 14 day chronic variate stress (CVS) as described in Methods. Chronic stress diminished sucrose preference compared to control unstressed animals receiving either vehicle or PAC1 receptor antagonist PACAP(6–38) (P(6–38)) alone. By contrast, stressed animals infused with PACAP(6–38) were relieved of the anhedonia-like behavior effects of stress.

BNST PACAP neuronal signaling to the dMHb increases anhedonia-related behaviors in unstressed male PACAP-Cre mice.

While the studies above demonstrated the abilities for MHb PACAP/PAC1 receptor signaling to elicit depression-like anhedonia responses, the studies did not evaluate directly whether the effects were consequences of BNST PACAP → MHb neurocircuit activation. Hence to probe this question, a chemogenetic approach was used to express an excitatory Gq-coupled DREADD in BNST PACAP neurons using male PACAP - Cre mice as described in Methods. The synthesis and subsequent translocation of these DREADDs to axonal terminals allowed the specific interrogation of the BNST PACAP → dMHb circuit following direct CNO infusions into the MHb to stimulate terminal PACAP release (Figure 5; Supplemental Figure 2, representative cannulation). Further, the specificity of the DREADD-mediated responses could be verified upon MHb pretreatments with the PAC1 receptor antagonist. A two-way ANOVA revealed a significant interaction effect (F(1,24) = 5.065, p = 0.0338), but no significant main effect of antagonist (F(1,24) = 1.675, p = 0.208) nor virus (F(1,24) = 0.5794, p = 0.454). Furthermore, post-hoc comparisons revealed that MHb CNO-mediated activation of the BNST PACAP DREADD projection terminals was sufficient to depress sucrose preference compared to control reporter vector expression (p = 0.037) and importantly, MHb pretreatment with the PAC1 receptor antagonist PACAP(6–38) was able to completely block the anhedonia-like effects of DREADD stimulated PACAP signaling to vehicle control levels (p = 0.59) (Figure 5). These results demonstrate that long distance BNST PACAP neuronal projections can be consequential in affecting a multitude of stress-related behavioral responses and that those traveling within the stria medullaris to the dMHb can participate in reward and anhedonia responses that may be contributory to depression-like behaviors.

Figure 5.

Chemogenetic activation of BNST PACAP → MHb circuit decreased sucrose preference which could be blocked with PAC1 receptor antagonist. A) Schematic of BNST PACAP chemogenetic activation. Gq DREADD was infused into the BNST of PACAP-Cre mice; vector reporter expression in the BNST (B) could be followed into the MHb (C) consistent with data in Figure 1. Following expression, CNO infusion into the MHb to stimulate BNST PACAP terminal peptide release decreased sucrose preference (D, column 3) compared to reporter vector controls (D, column 4). By contrast, MHb PAC1 receptor antagonist infusions prior to CNO administration blocked the anhedonia-like effects of BNST PACAP signaling (D, column 1); PAC1 receptor antagonist administration alone had no effects in control treated animals (D, column 2). Scale bar, 100 μm.

Discussion

In the current work, we have identified a novel PACAP pathway from the BNST to the MHb critical for depression-related responses. In addition to anatomically describing this novel PACAP projection from the BNST to the dorsal aspects of the MHb associated with SP/Tac1 neurons, we demonstrated that MHb PACAP (but not VIP) infusion produced an anhedonic response in the sucrose-preference test and also increased passive coping. Moreover, we found that MHb PACAP receptor antagonism abrogated the anhedonic effects of chronic stress as well as the chemogenetic activation of BNST PACAP afferents in the MHb.

The stria medullaris is a septohabenular tract relaying information from forebrain structures to the habenular (Hb) nuclei in the dorsal epithalamus. The two divisions of the Hb, the lateral habenula (LHb) and MHb are predominantly glutaminergic and associated with the regulation of mood states. Accordingly, the stria medullaris and septohabenular tract carries diverse forebrain signals critical to the regulation of moods and emotions including anxiety, fear, motivation and reward, aversion and depression-like behaviors (31–35). Our current tracing studies demonstrate that BNST PACAP neurons have projections in the stria medullaris to specifically target the dorsal medial habenula (dMHb), and importantly do not terminate in nearby regions such as the LHb and ventral pallidum (PV). These observations were fully supported by complementary immunocytochemical studies demonstrating PACAP and PAC1 receptor expression in the MHb and consistent with data from the Allen Brain Atlas and previous in situ hybridization results (27, 36). The MHb demonstrated high PAC1 receptor expression that was associated with BNST PACAP terminals in the dMHb. That MHb PACAP expression was lower than PAC1 expression suggests that MHb PAC1 activity may be regulated by PACAP afferents from other brain regions, since VIP is unlikely to activate PAC1 receptors at physiological levels. The MHb can be further divided into dMHb and vMHb subnuclei distinguished by the presence of SP/Tac1 and cholinergic neurons, respectively (28, 29), and additional characterizations demonstrated that the BNST PACAP terminals were localized preferentially in the dMHb suggesting regulatory activities on glutaminergic/ SP/Tac1 neurons.

The selectivity of BNST PACAP afferents in targeting the MHb (Figure 1), and the observation that selectively activating these afferents reduces sucrose preference (Figure 5) provides the strongest evidence for a novel BNST→ MHb PACAP circuit in the regulation of depression-related anhedonia. However, as lower levels of PAC1 receptor expression were also observed in nearby LHb and PV, areas that have also been associated with motivation, these data do not rule out the possibility of complementary LHb PACAP circuits in depression-related behaviors, nor the possibility that other PACAP afferents other than the BNST regulate MHb activities and subsequent behavioral changes.

PACAP signaling has often been associated with maladaptive stress-related responses and in coherence, exogenous PACAP applications or BNST → MHb circuit activation of BNST PACAP afferents in the dMHb resulted in reductions in sucrose preference. Among PACAP receptor subtypes, the effects were mediated specifically by the PAC1 receptor and notably, the depression-like responses were apparent only in male rats suggesting that the PACAP effects may be sexually dimorphic. This apparent sexual dimorphism should also be interpreted with care as it is possible that a behavioral effect in females would have been evident with greater statistical power. However, the current results differ from other PACAP-related work when sex differences are reported (15, 37–39). The apparent preferential BNST PACAP → dMHb circuit impact on males suggests that for reasons still unclear, the PACAP-mediated effects may differ among behavioral tasks in a sex-dependent manner. Importantly, the anhedonia observed in chronically stressed rats could be reversed upon MHb administration of the PAC1 receptor antagonist PACAP(6–38), further demonstrating that this is a functionally significant circuit associated with depression disorders.

Within the last decade, there has been increasing interest in the roles of the Hb in mood disorders. The LHb has been described to encode anti-reward, negative motivational and maladaptive stress states (31, 32); the MHb has been associated variably with fear, anxiety, anhedonia, mood regulation and nicotine addiction (28, 31–35). Hence, the two nuclei appear to possess overlapping and contributory activities to facilitate depression-like behaviors. However, exactly how each Hb nuclei or subnuclei contributes to the depression-like responses has also been complex and sometimes inconsistent. Studies using dMHb^CKO mice in which the dMHb neurons were genetically ablated, for example, showed depression-related decreases in sucrose preference, no effects in forced swim tests for depression, and reduced immobility in tail-suspension tests, which could be interpreted as anti-depressant effects (28, 33). Optogenetic stimulation of the dMHb or dMHb neurokinin neurons were reinforcing in intracranial self-stimulation tests but had no apparent effects in real-time place preference paradigms (28). There may be several reasons for these apparent disparate results. Firstly, the genetic ablation of a complete nuclei may have disrupted normal wiring networks and resulted in circuit alterations during developmental plasticity to contribute to the variable results. The optogenetic stimulation of dMHb neurons may have affected mixed populations; further, if the circuit or dMHb were sexually dimorphic, then from mixed sex animal usage some of the changes may have been obscured. The observations reported here were consistent with some previous work on MHb roles in anhedonia and depression (40), and in general agreement of the contributory roles of maladaptive PACAP signaling in psychopathologies.

Impaired function in central reward circuits have been implicated in anhedonia and depression (62). Hence, ablation of dopaminergic (DA) neurons in the ventral tegmental area (VTA) significantly reduces sucrose preference (63), and the hedonic value of sugar intake is associated with enhanced dopamine release in the ventral striatum, whereas the nutritional value is associated with dorsal striatum dopamine (64). Interestingly, antagonism of α3β4 nicotinic acetylcholine receptors in either the MHb or downstream IPN prevented the increased sensitized dopamine response to chronic morphine treatment in the nucleus accumbens (65). There are several possible pathways by which the regulation of MHb activity might impact DA reward circuits. Intriguingly, the MHb substance P neurons targeted by BNST PACAP afferents may project through the IPN and terminate in the VTA (66), and may also indirectly regulate DA activity via one-way projections through the LHb and or other brain regions downstream from the IPN involved in motivation (see 67 for review).

While substantial data have suggested roles for central PACAP systems in producing anxiety and panic, some of the investigations on the PACAPergic system and depression have been variable. In coherence with current work, intracerebroventricular (ICV) administration of PACAP raised intracranial self-stimulation threshold, diminished sweet solution preference and reduced novel animal interactions, which were consistent with depression-associated behaviors (49). Because the relationship of the forced swim test to depression is controversial, and its primary utility may be as a test of antidepressants, we focused most of the investigation reported here on sucrose preference and anhedonia. It is unknown whether the activation of BNST PACAP afferents in the MHb would decrease coping behaviors in the forced swim test, although interestingly, ICV PACAP did not increase immobility in forced swim tests even as other depression-related behaviors were increased (49). While our work supports a BNST → MHb PACAP circuit in anhedonia, further work is necessary to similarly implicate this circuit in depression-related changes in coping behavior. Moreover, PACAP deficient mice had lower depression responses (50, 51) and stressed PACAP knockout mice demonstrated marked attenuations in forced swim and social interaction depression-like behaviors, which correlated with reduced stress-induced serum corticosterone levels and hypothalamic transcriptional ΔFosB expression (52). We further demonstrate here that MHb PACAP receptor antagonism abrogated decreased sucrose preference following repeated variate stress, as well as the decreased sucrose preference observed following the chemogenetic activation of BNST PACAP afferents in the MHb (suggesting that the anhedonia-related effects of that stimulation were mediated by PACAP receptor activation). It is currently unknown whether the inhibition of BNST PACAP afferents in the MHb would prevent the depression-related behavioral consequences of repeated variate stress; however, because BNST afferents in the MHb may co-express and release other transmitters with PACAP, the results of that inhibition may be complex. Interestingly, other studies have also suggested that PACAP deficient mice have inherent increased depression-related behavior in similar tests (50, 53). The reasons for the discordant results are not quite clear but may be related to PACAP knockout backcrossings onto different genetic backgrounds (i.e., CD1 vs C57BL/6N). Nevertheless, the preponderance of data suggests that PACAP facilitates depression-like behaviors. Translationally, BNST PACAP levels appeared increased in postmortem tissues from patients with major depressive and bipolar disorders (54). The PACAP gene is found within a locus linked to schizophrenia and bipolar disorders (55–57), and polymorphism of the PACAP gene has been associated with MDD (17) suggesting that PACAP may be a risk factor leading to the pathogenesis of mental disorders.

The identities of the dMHb neurons receiving the BNST PACAP projections have not been established. Approximately 67% of dMHb neurons express the tachykinin 1 (Tac1) gene encoding substance P (58) and SP signaling has been associated with depression-related responses. Stress paradigms can increase SP levels in forebrain and hippocampal structures; blood SP levels are elevated in MDD and notably, neurokinin 1 (NK1) receptor antagonists have anti-depressant effects (59–61). The dMHb neurons project specifically to the lateral segments of the IPN and the increase in SP-stimulated IPN excitatory activities has been implicated in mediating anhedonia- and depression-like behaviors (40). The stimulatory effects of BNST PACAP signaling on dMHb glutaminergic and/or SP neurons may be consistent with these circuit mechanisms.

Nevertheless, the abilities for PACAP signaling to affect depression-like behaviors appear coherent with its roles in chronic states to facilitate or amplify the aversive maladaptive effects of stress on behavior. As described above, PACAP signaling has been associated with a number of psychopathologies including anxiety disorders, fear, and panic; this work and others suggest that depression may also fall within the compendium. PACAP expression and function in the BNST and LPBn appear to be central participants in these intersectional hubs mediating these responses; the activation and dysregulation of specific pathways from these hubs to their distal targets such as the amygdala, hypothalamus, hippocampus, dMHb and DR among others may be drivers of the particular behavioral pathologies.

Supplementary Material

Supplement Methods and Figures S1–S2

KEY RESOURCES TABLE

Resource Type	Specific Reagent or Resource	Source or Reference	Identifiers	Additional Information
Add additional rows as needed for each resource type	Include species and sex when applicable.	Include name of manufacturer, company, repository, individual, or research lab. Include PMID or DOI for references; use “this paper” if new.	Include catalog numbers, stock numbers, database IDs or accession numbers, and/or RRIDs. RRIDs are highly encouraged; search for RRIDs at https://scicrunch.org/resources.	Include any additional information or notes if necessary.
Antibody	Anti-ChAT	Millipore	AB144P
	Anti-PACAP	Hannibal	523C
	Anti-PAC1	Hannibal	35J8
Bacterial or Viral Strain	AAV2-EF1A-DIO-mCherry	Addgene	50462-AAV2
	AAV2-hSyn-DIO-hM3D(Gq)-mCherry	Addgene	44361-AAV2
	AAV2-hSyn-DIO-mCherry reporter	Addgene	50459-AAV2
Chemical Compound or Drug	PACAP38	Bachem	4031157
	PACAP6–38	Bachem	4031158
Organism/Strain	PACAP-Cre mouse (male and female)	Bred in house