Abstract
Cocaine- and amphetamine-regulated transcript (CART) peptide has been implicated in stress-related behaviors that are regulated by central serotonergic (5-HT) systems in the dorsal raphe nucleus (DRN). Here, we aimed to investigate the interaction between CART and DRN 5-HTergic systems after initially observing CART axonal terminals in the DRN. We found that microinfusion of CART peptide (55–102) into the DRN-induced anxiogenic effects in male C57BL/6J mice, while central administration of CART reduced c-Fos in 5-HTDRN neurons. This inhibitory effect of exogenous CART on 5-HTDRN activity and local 5-HT release was also demonstrated via in vivo fiber photometry coupled with calcium and 5-HT biosensors. CART inputs to the DRN were observed in various subcortical nuclei, but only those in the centrally projecting Edinger–Westphal nucleus (EWcp) were highly responsive to stress. Chemogenetic activation of these DRN-projecting CARTEWcp neurons recapitulated the effects of intra-DRN CART infusion on anxiety-like behavior in males, but not in females, suggesting a sex-specific role for this pathway. Interestingly, CARTEWcp projections to the DRN made direct synaptic contact primarily with non-5-HT neurons, which were also found to express putative CART receptors. Furthermore, chemogenetic stimulation of this CARTEWcp→DRN pathway inhibited 5-HT neurons while increasing activity in local GABAergic neurons. In summary, this study establishes for the first time a neuromodulatory role for CARTEWcp neurons in 5-HTDRN neurotransmission and suggests that CART may drive anxiety-like behavior by promoting feedforward inhibition of 5-HT neurons.
Significance Statement
Cocaine- and amphetamine-regulated transcript (CART) peptide is widely expressed in brain areas that are implicated in anxiety, stress, reward, and addiction. Genetic mutations in the CART gene have been linked to a heightened risk of developing anxiety/stress in adolescents, and particularly in the centrally projecting Edinger–Westphal nucleus (EWcp), upregulation of CART has been linked to depression in suicide victims. Here, we report that CARTEWcp signaling in the dorsal raphe nucleus (DRN) promotes anxiety-like behavior and inhibits serotonin (5-HT) activity in a sex-specific manner.
Introduction
Approximately 280 million individuals globally are affected with anxiety and depression, a number that has surged by 25% since the COVID-19 pandemic (Brunier, 2022). Neuropeptides are neuromodulators that are frequently co-released with neurotransmitters and have been implicated in anxiety and affective disorders (Kupcova et al., 2022). Emerging evidence indicates the potential role of neuropeptides as therapeutic agents for treating neurological diseases (Yeo et al., 2022). Cocaine- and amphetamine-regulated transcript (CART) peptide, a key player in regulating emotion, has been associated with a heightened risk of developing anxiety/depression in an adolescent population with a missense mutation in the CART gene (Miraglia Del Giudice et al., 2006). Moreover, increased CART expression in the centrally projecting Edinger–Westphal (EWcp) nucleus has been observed among suicide victims (Bloem et al., 2012).
CART neurons implicated in the stress response are predominantly localized in the arcuate nucleus (ARC), nucleus accumbens (NAc), EWcp, ventral tegmental area and locus ceruleus and sparsely within the amygdala and hippocampus (HP; Ahmadian-Moghadam et al., 2018; Priest et al., 2023). Administration of CART (55–102) peptide into either the lateral ventricle (LV), NAc, or amygdala produces anxiogenic effects (Chaki et al., 2003; Stanek, 2006; H. S. Yoon et al., 2014), whereas heightened CART levels in the EWcp, HP, ARC, and paraventricular hypothalamic nucleus (PVN) were observed in stressed rats (Kong et al., 2003; Gozen et al., 2007; Hunter et al., 2007; Okere et al., 2010; Xu et al., 2014). Furthermore, central administration of CART increased c-Fos expression in corticotropin-releasing factor (CRF) neurons of the PVN, thereby modulating CRF secretion and underscoring CART's crucial role in mediating the stress response (Smith et al., 2004).
It is estimated that over 30% of individuals with depression are refractory to current medications targeting monoamine transporters in the central nervous system (Rana et al., 2022), which speaks to the diverse etiology of mood-related disorders and emphasizes the pressing need for new therapeutic strategies addressing this unmet clinical need. The neuromodulatory potential of small neuropeptides represents a promising alternative avenue for the development of novel therapeutics for depression and mood disorders. While the regulatory role of CART in anxiety, depression, and stress within the mesolimbic circuitry has been characterized, its physiological function in the brainstem nuclei, particularly the EWcp and dorsal raphe nucleus (DRN), remains largely unexplored. Given the substantial contribution of DRN 5-HT to mood and stress/anxiety regulation (Andrews et al., 2015; Żmudzka et al., 2018; Zangrossi et al., 2020), we aim to investigate the interplay between the CARTergic and 5-HTergic systems. To ascertain the functional role of CART in the DRN, we investigated the impact of exogenous CART and endogenous CART originating in EWcp on anxiety-like phenotypes. Employing state-of-the-art techniques, we explored the physiological and functional aspects of CARTEWcp neurons and their role in modulating 5-HTDRN neuronal activity and anxiogenic behavior.
Materials and Methods
Animals
Adult (2–3 months old) male C57BL/6J (000664), Fos2A-iCreER (030323), Sert-Cre (014554), and Ai14 (007908) mice and adult male and female Cart-IRES2-Cre-D (CART-cre; 028533) mice from the Jackson Laboratory were used in these experiments. All mice were kept in a temperature- and humidity-controlled vivarium with ad libitum access to food and water under a 12 h light/dark cycle in accordance with AAALAC guidelines. All experimental procedures were approved by the IACUC at the University of Iowa.
Immunofluorescence
Immunofluorescence (IF) was performed to localize CART projections and GPR-160 (CART-putative receptor) onto 5-HT neurons in the DRN of C57BL/6J, as described previously (Balasubramanian et al., 2023; Khan et al., 2023a). Briefly, 4–5 DRN sections (25 µm) were used across the rostrocaudal axis. Sections were permeabilized in 0.5% Triton X-100/PBS for 30 min, followed by blocking in 10% NDS–0.1% Triton X-100–PBS solution, and then incubated with primary and secondary antibodies (Table 1). Images were acquired on an Olympus FV-3000 confocal microscope with z-stacks (0.5 μm/step) at 20× or 40× with an optical resolution factor of 1×, 2×, or 3×. Multiple tiles were imaged for every region of interest, and mosaic stitching was performed using FV3000 Multi-Area Time-Lapse (MATL) Software. After imaging, confocal stacks were converted to maximum projection images using ImageJ software.
List of antibodies used in immunofluorescence
Stereotaxic surgeries and viral constructs
Stereotaxic surgeries were performed for single/bilateral/fiber cannula implantations (Table 2) and intracranial AAV infusions (Table 3). Surgeries were performed using an Angle Two stereotaxic frame (Leica Biosystems). A retrograde tracing experiment was performed by injecting 250 nl of AAVrg-hsyn-DIO-EGFP (Addgene #50457) into the DRN (ML = ±0.0, AP = −4.65, DV = −3.30, θ = 23.58°) of CART-cre mice with a Quintessential Stereotaxic Injector (Stoelting) at a rate of 100 nl/min. Anterograde tracing experiments were performed by injecting 150 nl/site cre-dependent synaptophysin tracer (AAV8.2-ef1α-DIO-synaptophysin-eYFP) and a transsynaptic tracer (AAV2/9-CAG-DIO-WGA-ZsGreen) into two sites of the EWcp (rostral, ML = ±0.0, AP = −3.16, DV = −3.9; caudal, ML = ±0.0, AP = −3.6, DV = −3.30) of CART-cre mice.
List of the cannula used in the surgery
List of AAV viruses used in the study
For the CART overexpression study, 150 nl/site of AAV2/8-ef1α-DIO-nlstdTomato-WPRE (control) or AAV2/8-ef1α-DIO-Cartpt-2A-nlstdTomato-WPRE was injected into the EWcp of male CART-cre mice. For the chemogenetics study, 300 nl of AAVrg-EF1a-DIO-FLPo-WPRE-hGHpA was injected into DRN, and 150 nl/site of AAV8-ef1a-fDIO-mCherry-WPRE (control) or AAVDJ-hsyn-fDIO-hM3Dq-mCherry was injected into the EWcp of both male and female CART-cre mice.
For fiber photometry experiments, a fiber cannula (OmFC) was implanted in the DRN, and AAV was infused at the rate of 100 nl/min through OmFC guide cannula into the DRN (ML = ±0.0, AP = −4.65, DV = −3.2, θ = 23.58°) using a fluid injector (with 0.4 mm projection). The OmFC cannula was implanted in such a way that the cannula was just placed above the DRN and the fiber was close to the DRN (Fig. 3D,E, DRN implant diagram). Diluted (1:2) 500 nl of AAV1-CAG-Flex-GCaMP6s-WPRE-SV40 (Addgene #100842) and AAV9-syn-5HT3.5 (WZ Biosciences) was infused into the DRN of male Sert-cre and C57BL/6J mice, respectively.
In a separate cohort, a single cannula (P1 Technologies) was implanted into the DRN of male C57BL/6J for CART microinjection. In addition, bilateral cannulae were implanted in the LV (ML = ±1.0, AP = −0.30, DV = −2.20) of male Fos2A-iCreER mice for experiments involving targeted recombination in activated populations (fos-TRAP). The internal cannula used for both single and bilateral injections had a projection of 0.4 mm (Table 2).
Drugs
CART peptide infusion experiment
The lyophilized CART peptide (55–102) (003-62, Phoenix Pharmaceuticals) was reconstituted in 0.2 M PBS to obtain a stock concentration of 1 µg/µl. The final working CART peptide solution was freshly made in aCSF contaning the following (in mM): 124 NaCl, 4.4 KCl, 2 CaCl2, 1.2 MgSO4, 1 NaH2PO4, 10.0 glucose, and 26.0 NaHCO3 (osmolality, 300–310 mOsm/kg).
Targeted recombination in TRAP experiment
4-Hydroxytamoxifen (4-OHT; H6278; Millipore Sigma) was made as described previously (DeNardo et al., 2019). Briefly, 20 mg/ml 4-OHT stock solution was prepared in ethanol by incubating at 40°C for 2–3 min. A working concentration of 5 mg/ml was made consisting of a 1:4 mixture of castor and sunflower seed oil. Ethanol from the working solution was evaporated by heating at 56°C. Mice were intraperitoneally injected at a volume of 10 ml/kg for a final dose of 50 mg/kg.
Chemogenetic experiments
Clozapine-N-oxide dihydrochloride (CNO; Hello Bio) was dissolved in 0.9% sterile saline to a concentration of 0.30 mg/ml and intraperitoneally injected at a volume of 10 ml/kg, 30 min before the beginning of behavioral experiments for a dose of 3 mg/kg. For c-fos studies, CNO was injected systemically 2 h before perfusion and brain extraction.
Behavioral tests
All the behavioral tests were performed in C57BL/6J (male) and CART-cre (male and female) mice between 7.30 A.M. to 12.00 P.M. In the C57BL/6J mice (n = 8–9/group), the behaviors were recorded 15 min after the infusion of either CART or aCSF in a between-subject design (Bakhtazad et al., 2020; Somalwar et al., 2020). Before the test day, the C57BL/6J mice were habituated to infusion to reduce handling stress. For infusion, the mouse was gently scruffed, and the CART peptide (5, 50, and 100 ng) or aCSF was infused through the guide cannula using the internal cannula and the tubing attached to a 1 µl syringe (7101, Hamilton Company). Because CART (62–102) ICV administration did not affect anxiogenic behavior, in the present study, we used CART (55–102) to decipher its effect on DRN and anxiety (Chaki et al., 2003). For the chemogenetics stimulation and anxiety assessment experiment, the CART-cre mice (males, n = 7/group; females, n = 5–6/group) were habituated to intraperitoneal injection 2 d before the test day.
Elevated plus maze test
Anxiety was assessed in the elevated plus maze (EPM) as reported previously (Wise et al., 2019). The mice were placed in the center of an elevated plus-shaped maze with white open and dark/black closed arms and allowed to explore for 5 min. Sessions were recorded with a Basler GenICam and Media Recorder software (Noldus). The % time spent in the open and closed arms as well as a neutral zone, the number of entries to each arm, the distance moved, and the mean velocity were analyzed using EthoVision XT15.
Light–dark box exploration test
The light–dark box exploration (LDB) test was conducted as reported previously (Bourin, 2015). The apparatus consists of a light (390 lux) and a dark (5 lux) compartment with a small opening between the compartments. The mice were placed in the center of the light compartment, and the behavior was recorded for 10 min with a camera and Media Recorder software (Noldus). The % time spent in each compartment, number of transitions to the light–dark box, distance moved, and mean velocity were analyzed with EthoVision XT15.
Social interaction test
Social deficits were measured using a social interaction test (SIT) as described previously (Khan et al., 2023a,b). Briefly, the test was conducted in a three-chambered plexiglass apparatus with a small opening between the chambers. The test mouse was habituated to the entire arena for 10 min. At the end of the habituation phase, the test mouse was confined to the center chamber while a conspecific stranger mouse was placed inside a metal holding cage in one of the outermost chambers, and an empty holding cage was placed in the opposite chamber. The social interaction test began by removing the doors between the chambers so that the test mouse could move freely throughout the arena for another 10 min. The behavior was recorded, and the total time spent by the test mouse interacting with the conspecific stranger mouse cage (SC) or empty cage (EC) and % social interaction was scored by a trained observer blinded to the experimental treatment.
Targeted recombination in transiently active populations
To selectively label neurons activated by CART microinjection, we employed the TRAP method using Fos-icreER x Ai14 male mice as described previously (DeNardo et al., 2019). Before the test day, the Fos-icreER x Ai14 mice (n = 5/group) were habituated to intraperitoneal injections and infusions to minimize handling stress. On the test day, either aCSF or CART (500 ng/mouse) was bilaterally infused in the LV of Fos-icreER x Ai14 mice. These mice express tdTomato in a Cre-dependent manner after induction of c-fos. 4-OHT (50 mg/kg, i.p.) was administered 30 min after infusion of aCSF or CART, and the mice were left undisturbed for several hours. After 2 weeks, the brains were extracted, and immunofluorescence was performed on DRN slices (rostral, mid, and caudal) with TPH2 antibodies. The images were captured with a confocal microscope, and multiple tiles were stitched using FV3000 Multi-Area Time-Lapse Software Module. The DRN images were analyzed for colocalization of tdTomato and TPH2 signals using QuPath software (Version 0.4.2).
Liquid chromatography mass spectrometry (LCMS)
Neurotransmitter levels were measured using LCMS in the DRN tissue of aCSF/CART-microinjected male C57BL/6J mice (n = 8–9). The brains were extracted 15 min after aCSF/CART microinjection into the DRN and snap-frozen immediately. DRN tissue was micropunched, and LCMS was performed at the University of Iowa Metabolomics Core. The samples were processed as described previously (Cantor et al., 2017). Briefly, the tissue samples were first lyophilized and transferred to ceramic bead tubes, and 18-fold (w/v) extraction solvent (with nine heavy internal standards) was added to each sample. The sample was homogenized and incubated in the rotor at −20°C for 1 h and then centrifuged at 21,000 × g for 10 min to collect the supernatant. The supernatant (300 nl) was dried in a SpeedVac, further reconstituted in 30 ul acetonitrile and water (1:1 v/v), and kept overnight at −20°C. The sample was then centrifuged, and the supernatant was transferred to LCMS autosampler vials for analysis. LCMS data were acquired on a Thermo Q Exactive hybrid quadrupole Orbitrap mass spectrometer with a Vanquish Flex UHPLC system or Vanquish Horizon UHPLC system (Thermo Fisher Scientific). The LC column used was a Millipore SeQuant ZIC-pHILIC (2.1 × 150 mm, 5 µm particle size) with a ZIC-pHILIC guard column (20 × 2.1 mm; MilliporeSigma). Two microliters of the sample were run on a mobile phase containing Solvent A, 20 mM ammonium carbonate [(NH4)2CO3] and 0.1% ammonium hydroxide (v/v) [NH4OH], and Solvent B, acetonitrile, at the flow rate of 0.150 ml/min.
MS data acquisition was performed in the range of m/z 70–1,000 with the resolution set at 70,000, the AGC target at 1 × 106, and the maximum injection time at 200 ms (Cantor et al., 2017). LCMS data were processed by Thermo Scientific TraceFinder 4.1 software, and metabolites were identified based on an in-house library. NOREVA software was used for signal drift correction (B. Li et al., 2017). Data were normalized to the sum of all the measured metabolite ions in that sample.
Fiber photometry
Home cage fiber photometry experiments were conducted to investigate the impact of CART on 5-HT neuronal activity and release using a within-subject design. Mice were habituated to patch cord tethering and aCSF infusion for 3 consecutive days. On test day 1, a 10 min baseline recording was conducted, after infusing 200 nl of aCSF through an OmFC cannula into DRN. Subsequent 15 min post-infusion signals were recorded. On test day 2, instead of aCSF, CART was infused after baseline recordings, and 15 min post-infusion recordings were obtained. Data were recorded using a fiber photometry system (Neurophotometrics).
GCaMP6s biosensor in 5-HTDRN neurons
The effect of CART on 5-HT neuronal activity was measured in Sert-cre male mice (n = 7) injected with AAV1-CAG-Flex-GCaMP6s-WPRE-SV40. Signals were recorded from alternating pulses of 470 nm (calcium-dependent) and 415 nm (calcium-independent, isosbestic) with an excitation power set at 60 µW. Fiber photometry data were acquired with Bonsai software (Open Ephys) and processed via MATLAB R2021b software (MathWorks) as described previously by our group (Wang et al., 2023). For all recordings, 470 and 415 signals were deinterleaved and processed to create a linear regression model using the MATLAB fitlm function to fit the raw data and correct motion artifacts. Next, each fluorescent intensity value was normalized to the corresponding value predicted by the linear regression model. The fiber photometry system generated data every 3 s, and in the analysis, 20 normalized values from 1 min were downsized to one value. Finally, we computed peri-event percent fluorescent changes (% of ΔF/F) between the last 5 min pre-infusion event (baseline) and 12 min post-infusion event and compared the area under the curve (AUC) in the peri-event plot for the CART versus aCSF groups.
g5-HT2h (GRAB5-HT3.5) biosensor in DRN
The effect of CART on 5-HT release was measured in C57BL/6J male mice (n = 8) injected with a new generation of g5-HT2h (AAV9-hsyn-5HT3.5) biosensor that can detect changes in extracellular 5-HT levels sensitively with less signal-to-noise ratio. GRAB5-HT3.5 biosensor inherited the pharmacological specificity of the parent 5-HT4R receptor and shows high specificity to 5-HT and not any other 5-HT precursor or metabolites of 5-HT (Deng et al., 2024). For g5-HT2h (5HT3.5) recordings, 415 isosbestic signals were not used in the analysis as it is not an appropriate wavelength for motion control for this sensor (Hon et al., 2022). Peri-event fluorescent plots were generated by normalizing traces to the first 20 s of each trace.
Retrograde tracing
DRN-projecting CART neurons were identified by a cre-dependent retrograde tracing approach as reported previously (Green et al., 2020). AAVrg-hsyn-DIO-EGFP was injected into the DRN of CART-cre x Ai14 reporter male mice. These mice express tdTomato in a cre-dependent manner in CART neurons across the brain. After 3 weeks of viral expression, brains were extracted for analysis. CART-expressing brain regions, especially those involved in stress and anxiety-related responses, were included in this study (Ahmadian-Moghadam et al., 2018). NAc, EWcp, medial amygdala (MeA), hypothalamic nuclei, ventromedial hypothalamus (VMH), tubular lateral hypothalamus (TuLH), ARC, PVN, and HP are the key targets we chose for retrograde tracing. Three to four sections from each level of the rostrocaudal axis of each brain region were selected for the retrograde tracing analysis. The images were captured on a confocal microscope and analyzed for colocalization of tdTomato (CART neurons) and EGFP (DRN-projecting CART neurons) using QuPath software (Version 0.4.2). The mean values for a given sub-region along the rostro-caudal axis were collapsed to one value for each brain region in each mouse.
Anterograde tracing
A cre-dependent anterograde tracer (AAV8.2-ef1a-DIO-synaptophysin-eYFP) was injected into the EWcp of CART-cre mice to label efferent projections of CARTEWcp neurons to the DRN. This anterograde tracing strategy has been described previously by our group (Wang et al., 2023). AAV2/9-CAG-DIO-WGA-ZsGreen, a transsynaptic anterograde tracer, was also injected into the EWcp of a separate cohort of CART-cre mice to label post-synaptic cellular targets of CARTEWcp neurons, which is a technique we also described previously (Wang et al., 2023). To allow for anterograde trafficking of virally expressed proteins, brains were extracted after 8 weeks for immunofluorescence and imaging. After verifying cell type–specific expression of eYFP/ZsGreen in CARTEWcp neurons by immunofluorescence with a CART antibody, four sections from each level of the rostro-caudal axis from DRN were stained using a TPH2 antibodies to label 5-HT neurons and map the synaptophysin and ZsGreen tracers within the DRN. The images were captured on a confocal microscope, and stacks were converted to maximum projection images using ImageJ software. The number of ZsGreen neurons in the DRN and the percentage of ZsGreen-immunoreactive area within 5-HT and non-5-HT neurons were measured using QuPath software (Version 0.5.1).
Ex vivo electrophysiology
Three weeks before the electrophysiological recordings, CART-Cre x Ai14 male mice (n = 6) were injected with AAVrg-hSyn-DIO-eGFP in the DRN to label DRN-projecting CART neurons. Acute restraint stress (30 min) was applied to the CART-Cre x Ai14 mice to induce anxiety; undisturbed mice served as controls. The ex vivo electrophysiological recording method has been described in detail previously (Khan et al., 2023a; Pierson et al., 2024). Briefly, acute 300 µm brain slices containing the PVN, EWcp, VMH, and NAc were prepared with a vibratome (VT1200S; Leica Biosystems) after mice were transcardially perfused with carbogenated modified artificial cerebrospinal fluid (aCSF) containing the following (in mM): 110 choline-Cl, 2.5 KCl, 7 MgSO4, 0.5 CaCl2, 1.25 NaH2PO4, 26.2 NaHCO3, 25 glucose, 11.6 Na-ascorbate, 2 thiourea, and 3.1 Na-pyruvate (pH, 7.3–7.4; osmolality, 300–310 mOsmol/kg). Brain slices were then recovered in modified aCSF containing the following (in mM): 92 NaCl, 2.5 KCl, 2 MgSO4, 2 CaCl2, 1.25 NaH2PO4, 30 NaHCO3, 20 HEPES, 25 glucose, 5 Na-ascorbate, 2 thiourea, and 3 Na-pyruvate (pH, 7.3–7.4; osmolality, 300–310 mOsmol/kg), continuously bubbled with 95% O2/5% CO2. During recordings, slices were perfused with standard aCSF containing the following (in mM): 124 NaCl, 4 KCl, 1.2 MgSO4, 2 CaCl2, 1 NaH2PO4, 26 NaHCO3, and 11 glucose (osmolality, 300–310 mOsmol/kg; saturated with 95% O2/5% CO2 and maintained at 30 ± 1°C). Fluorescent filters were utilized to screen cells expressing both tdTomato and EGFP, i.e., DRN-projecting CART neurons. Patch electrodes (3–5 MΩ) were filled with a solution containing the following (in mM): 135 K-gluconate, 5 NaCl, 2 MgCl2, 10 HEPES, 0.6 EGTA, 4 Na2-ATP, and 0.4 Na2-GTP. The current-clamp mode was employed to assess intrinsic excitability, and cells were kept at −70 mV to offset the variation of resting membrane potentials. Rheobase was measured by the minimum depolarizing current needed to evoke action potentials. Input resistance was examined by the decrease in membrane potential after the injection of −100 pA hyperpolarizing current. Numbers of spikes (i.e., evoked action potentials) were recorded upon the injection of depolarizing currents for 250 ms at 10 pA incremental steps (10–200 pA).
For the CART overexpression experiments, AAV2/8-ef1α-DIO-Cartpt-2A-nlstdTomato-WPRE (or the control virus, AAV2/8-ef1α-DIO-nlstdTomato-WPRE) was injected in the EWcp of male CART-cre mice (n = 6/group), and electrophysiological recordings in DRN 5-HT neurons were performed 6 weeks later. The procedures were the same as those described above, except that no acute stress was exerted and no fluorescent filters were used during recordings, but biocytin (2 mg/ml, HB5035, Hello Bio) was added to the intracellular solution for post hoc identification of 5-HT neurons. After recordings, the slices were post-fixed with 4% paraformaldehyde for 1 h and then stained with the primary antibody, rabbit anti-TPH2 (NOVUS, NB100-74555, 1:1,000), and secondary antibodies, Alexa Fluor 488 donkey anti-rabbit (Invitrogen, A21206, 1: 1,000) and Alexa Fluor 647 streptavidin (to bind with biocytin, Jackson ImmunoResearch, 016-600-084, 1:1,000). Confocal imaging was performed to identify recorded 5-HT neurons that co-expressed biocytin and TPH2.
Fluorescence in situ hybridization (FISH) or RNAscope
Multiplexed FISH was performed using the ACDBio RNAscope V2 reagents and protocols (catalog #323100) on 20 µm fixed EWcp or DRN slices as reported previously (Wang et al., 2023). Briefly, slices were mounted on HistoBond Plus slides and baked for 30 min. After baking, the slices were post-fixed in 10% chilled neutral buffered formalin (MilliporeSigma) at 4°C for 15 min and dehydrated serially in ethanol (50, 70, 100%). The slides were re-baked for 30 min and then treated with hydrogen peroxide (10 min). Target retrieval was performed for 9 min using the reagent provided by the manufacturer. Next, the proteins were digested using protease III for 30 min at 40°C in a HybEZ Hybridization Oven. After the pretreatment steps, respective mRNA probes for Cartpt, tdtomato (Cart overexpression experiment) and Gad1, Tph2, and cfos (chemogenetics c-fos experiment) were applied (probe details in Table 4). A series of amplification and Opal detection reagents were applied to the tissue sections according to the manufacturer's instructions. Sections were counterstained using DAPI provided in the detection reagent kits and mounted in ProLong Gold mounting media (P36930, Thermo Fisher Scientific). The slides were imaged using an Olympus SLIDEVIEW VS200 digital slide scanner, and the images were analyzed in QuPath (Version 0.5.1), including cell detection and RNAscope puncta quantification. For the Cartpt overexpression experiment, the pixel classifier was set for Cartpt mRNA, and the % positive area was calculated relative to the total ROI area. For the chemogenetic c-fos experiment, an object classifier was set for c-fos-Tph2, c-fos-Gad1, or c-fos-Gad1-Tph2 mRNA-expressing neurons. The percentage of total Tph2 neurons, total Gad1 neurons, and total Tph2-Gad1 neurons that also expressed c-fos were then calculated.
List of probes used in RNAscope
Data and statistical analysis
Statistical analysis was performed using GraphPad Prism 10. Student’s t tests were used for comparisons between two groups, and a one-way or two-way ANOVA was applied for comparisons between more than two groups with one or more independent variables, respectively. Post hoc Tukey, Dunnett's, or Šidák tests were employed for pairwise comparisons after ANOVA. A p-value of <0.05 was considered significant for all the analyses. Data are expressed as mean ± SEM.
Results
Microinfusion of CART peptide in the DRN induces anxiogenic behavior
CART axonal terminals were initially detected in DRN by immunofluorescence, and a subset of these was found to be proximal to 5-HT cell bodies (Fig. 1A). We then aimed to determine whether CART peptide could elicit anxiogenic effects in the DRN of male C57BL/6J mice using a between-subject design. Various CART doses (5, 50, and 100 ng) were microinjected into the DRN, and anxiety-like behaviors were assessed in the EPM, LDB, and SIT (Fig. 1B). Mice receiving a high CART dose (100 ng) exhibited heightened anxiety in the EPM, spending more time in the closed arm (one-way ANOVA, F3,30 = 4.847, p = 0.0072; followed by post hoc Dunnett's tests showing aCSF vs CART-100 ng, p < 0.0117) versus the open arm (one-way ANOVA, F3,30 = 4.847, p = 0.0072; followed by post hoc Dunnett's tests showing aCSF vs CART-100 ng, p = 0.0117; Fig. 1C,D). Moreover, the mice that received 100 ng CART spent less time in the neutral zone as compared with the aCSF mice (one-way ANOVA, F3,30 = 4.419, p = 0.0109; followed by post hoc Dunnett's tests showing aCSF vs CART-100 ng, p = 0.0033; Fig. 1E), thus suggesting heightened anxiety and reduced exploration or risk-taking behavior. While measuring the anxiogenic behavior in the EPM test, distance moved (Fig. 1F) and mean velocity (data not shown) parameters were used to assess any locomotory dysfunction in the test mice. No significant locomotor changes were observed in the CART-injected mice compared with aCSF. The heat map visualization of the EPM test illustrates a CART dose-dependent increase in anxiety in comparison with aCSF controls (Fig. 1G–J).
CART peptide infusion into the DRN increases anxiogenic behavior. A, Representative confocal images of immunofluorescence (IF) showing a single IF image of CART peptide (left) and 5-HT (middle) and their colocalization (right) in DRN of C57BL/6J mice. Scale bar, 50 µm. B, Schematic representation of the experiments showing the surgery and behavior schedule. Parameters of elevated plus maze (EPM) test showing % time spent in (C) open arm, (D) closed arm, and (E) neutral zone and (F) distance moved after infusion of aCSF or different doses of CART (5, 50, 100 ng) into DRN of C57BL/6J mice. G–J, Heat map visualization of time spent in the EPM arms that received either aCSF or CART (5, 50, 100 ng). Light dark box (LDB) exploration test showing % time spent in (K) light box and (L) dark box and (M) distance moved and (N) mean velocity in aCSF and CART microinfused mice. O, Heat map visualization of time spent in the light box of LDB apparatus. Social interaction test showing (P) % social interaction with strangers and (Q) interaction time (s) with empty (EC) and stranger cages (SC) in aCSF and CART-100 ng microinfused mice. Values (n = 7–10/group) are represented as means (±SEM), and the data were analyzed by one-way or two-way ANOVA (*p < 0.05, **p < 0.01, vs aCSF). Part of the figure was made in BioRender.
The LDB test exhibited a similar dose-dependent anxiogenic effect, with the highest dose of 100 ng CART showing the strongest anxiogenic effect compared with lower doses and the vehicle controls. Mice receiving 100 ng of CART spent more time in the dark box (one-way ANOVA, F3,29 = 3.796, p = 0.0207; followed by post hoc Dunnett's tests showing aCSF vs CART-100 ng, p = 0.0107) versus the light box (one-way ANOVA, F3,29 = 3.796, p = 0.0207; followed by post hoc Dunnett's tests showing aCSF vs CART-100 ng, p = 0.0107; Fig. 1K,L). Similar to the EPM test, no significant effect of CART was observed in locomotor functions assessed by distance moved and mean velocity parameters (Fig. 1M,N). The heat map visualization of the LDB test illustrates a CART dose-dependent increase in anxiety in comparison with aCSF controls (Fig. 1O).
Next, we measured social deficits in the SIT and observed a decrease in the percentage of time spent in social interaction in the CART-100 ng group as compared with aCSF controls (one-way ANOVA, F3,30 = 3.129, p = 0.0402; followed by post hoc Dunnett's tests showing aCSF vs 100 ng, p = 0.0268). Interestingly, the time spent interacting with the empty cage (EC) increased, while interaction with a stranger cage (SC) decreased in the CART-100 ng group compared with the aCSF group (two-way ANOVA: interaction effect between aCSF/CART and EC/SC, F1,28 = 12.07, p = 0.0017, with post hoc Sidak's tests showing increased EC interaction, p = 0.0404, and decreased SC interaction, p = 0.0406, in mice infused with CART vs aCSF; Fig. 1P,Q). Collectively, these experiments indicate that intracranial infusion of CART peptide into the DRN has an anxiogenic effect in mice.
ICV administration of CART peptide reduced fos activity in 5-HTDRN neurons
Subsequently, we sought to investigate whether the anxiogenic effects mediated by CART were indicative of altered activity in DRN neurons. To address this, we conducted a fos-TRAP experiment utilizing Fos2A-iCre x Ai14 (tdtomato reporter mice) to assess CART-mediated 5-HT neuronal activation (Fig. 2A). Microinjection of CART into the lateral ventricle (ICV injection) resulted into a significant decrease in the number of tdTomato+ TPH2 neurons in the CART-500 ng compared with the aCSF control group (t9 = 2.943, p = 0.018; Fig. 2B). Analysis of the proportion of tdTomato+ TPH2 neurons within the TPH2 neuron population also indicated an overall decrease in c-Fos activity in TPH2 neurons (t9 = 3.343, p = 0.01; Fig. 2C). We then analyzed c-Fos activity in non-TPH2 neurons and found that there was no significant change in the overall tdTomato signal in the CART-500 ng group as compared with aCSF controls (Fig. 2D). Furthermore, a specific examination of c-Fos activation in the rostral to caudal axis revealed a non-significant decrease in rostral (t9 = 2.044, p = 0.07; Fig. 2E,K, top panel) and mid-subregion (t9 = 1.943, p = 0.087; Fig. 2E,K, middle panel), upon CART infusion. It was recently reported that 5-HTDRN neurons differentially modulate stress-related behavior depending on their relative position across the dorsoventral axis, with neurons in the dorsal DRN activating subcortical stress pathways and neurons in the ventral DRN activating cortical anti-stress systems (Ren et al., 2018). Therefore, we performed a separate analysis of the dorsal and ventral regions of the DRN and observed a significant decrease in the tdTomato signal in ventral TPH2 neurons (t9 = 4.146, p = 0.01; Fig. 2F). Analysis of the proportion of tdTomato+ TPH2 neurons within the TPH2 population indicated a significant reduction observed in mid (t9 = 2.869, p = 0.02; Fig. 2G,K, middle panel), caudal (t9 = 3.477, p = 0.008; Fig. 2G,K, lower panel), and ventral TPH2DRN neurons (t9 = 3.183, p = 0.01; Fig. 2H). Moreover, the number of tdTomato+ non-TPH2 neurons was only significantly reduced in the rostral DRN (t9 = 2.430, p = 0.041; Fig. 2I,J). Together, these findings suggest that CART may promote anxiety-like behavior by inhibiting cortical-projecting ventral 5-HTDRN neurons that activate anti-stress pathways in the brain.
CART peptide decreases tdTomato (c-Fos) expression in the DRN-TPH2 neurons. A, Schematic representation of targeted recombination of active populations (fos-TRAP) experiment after aCSF/CART infusion in the lateral ventricle (LV) of Fos-iCreER x Ai14 mice. Violin plot showing the (B) number of tdTomato-TPH2 double-positive neurons and (C) percentage of TPH2 neurons positive for tdTomato signal and (D) number of tdTomato non-TPH2 double-positive neurons in the DRN of aCSF and CART microinfused (ICV) Fos-iCreER x Ai14 mice. Violin plot showing the number of total tdTomato and TPH2 double-positive neurons in (E) rostral, mid, and caudal DRN and (F) dorsal and ventral DRN. Violin plot showing the percentage of total TPH2 neurons positive for tdTomato signal in (G) specific to rostral, mid, and caudal DRN and (H) dorsal and ventral DRN. Violin plot showing the number of tdTomato non-TPH2-positive neurons in (I) rostral, mid, and caudal regions and (J) dorsal and ventral regions of the DRN. K, Confocal images showing tdTomato andTPH2 double-positive neurons in the rostral (top), mid (middle), and caudal (bottom) DRN of aCSF and CART microinfused Fos-iCreER x Ai14 mice. The dorsal and ventral parts of the DRN are shown by dashed lines. Scale bar, 200 µm. Values (n = 5/group) are represented as means (±SEM), and the data were analyzed by single and unpaired t test (*p < 0.05, **p < 0.01 vs aCSF). Part of the figure was made in BioRender.
Microinfusion of CART peptide in the DRN reduces 5-HT neuronal activity and transmitter release in real time
We next sought to elucidate the impact of CART on neurotransmitter levels in the DRN. Micropunched DRN tissues from mice that received CART/aCSF were processed for measurement of neurotransmitter levels via LCMS. Interestingly, LCMS data revealed a significant decrease in 5-HT levels in the CART-100 ng versus aCSF group (one-way ANOVA, F3,29 = 2.497, p = 0.0430; followed by post hoc Dunnett's tests showing aCSF vs 100 ng, p = 0.0252; Fig. 3A). The chromatograph illustrating 5-HT peak intensity indicated a decrease in 5-HT levels upon CART infusion in the DRN (Fig. 3B). However, the 5-HT metabolite, 5-hydroxyindoleacetic acid (5HIAA), exhibited no significant change (Fig. 3A). Furthermore, GABA, glutamate, and homovanillic acid (HVA, a dopamine metabolite) showed no apparent alterations following CART infusion (Fig. 3C).
CART peptide inhibits 5-HT neuronal activity and transmitter release. A, Line graph showing 5-HT and 5-hydroxy indole-acetic acid (5HIAA) levels in the DRN and (B) LCMS chromatograph depicting a reduction in the 5-HT peak at 16–20 retention time (RT) after aCSF or CART (5, 50, 100 ng) infusion. C, Graph showing levels of GABA (purple), glutamate (yellow), and homovanillic acid (HVA; green) in LCMS after CART/aCSF infusion. Values (n = 8–10/group) are represented as means (±SEM), and the data were analyzed by one-way ANOVA and (*p < 0.05, vs aCSF). D, E, Schematic representation of the timeline of the in vivo fiber photometry experiment using GCaMP6s and 5-HT biosensors in Sert-Cre and C57BL/6J mice, respectively. F, Confocal image showing the expression of AAV1-CAG-Flex-GCaMP6s-WPRE-SV40 in DRN of Sert-Cre mice. Scale bar, 200 µm. G, (% ΔF/F) showing decreased GCaMP signal as a measure of 5-HT activity in the CART versus aCSF group. H, I, Violin plot showing decreased area under the curve (AUC) GCaMP values in the CART versus aCSF mice. J, Confocal image showing the expression of AAV9-syn-5HT3.5 in DRN of C57BL/6J mice. Scale bar, 200 µm. K, (% ΔF/F) showing a decrease in g5-HT2h (5HT3.5) signal as a measure of 5-HT release in the CART versus aCSF group. L, M, Violin plot showing decreased AUC values for g5-HT2h (5HT3.5) in the CART versus aCSF mice. Values (n = 8/group) are represented as means (±SEM), and the data were analyzed by unpaired t test (*p < 0.05, vs aCSF). Part of the figure was made in BioRender.
We further monitored the impact of CART microinjection on 5-HTDRN neuronal activity in real time in awake, behaving mice. For this experiment, Sert-cre mice were microinjected with a cre-dependent GCaMP6s calcium biosensor in the DRN and implanted with a specialized OmFC mount comprising an optical fiber for activity measurement and a cannula for aCSF/CART delivery (Fig. 3D,F). We observed a decrease in the GCaMP signal (% ΔF/F) post–CART-100 ng microinfusion into the DRN (Fig. 3G). This effect persisted for 15 min post-infusion, with significantly lower AUC values in the CART-100 ng group compared with the aCSF group (t13 = 2.400, p = 0.0062; Fig. 3H,I). Real-time recording of 5-HT release in C57BL/6J mice using the g5-HT2h (5HT3.5) biosensor corroborated the GCaMP results (Fig. 3E,J). The g5-HT2h (5HT3.5) signal (% of ΔF/F) decreased after CART microinfusion, with concurrently reduced AUC values (t15 = 2.297, p = 0.0375; Fig. 3K–M). Collectively, these findings suggest that CART diminishes 5-HTDRN neuronal activity and transmitter release, ultimately culminating in reduced 5-HT concentrations in the DRN. This complements our initial findings that CART reduces c-Fos activity in 5-HT neurons in the ventral DRN, which were previously reported to promote stress-coping behaviors (Ren et al., 2018).
Retrograde tracing of CART projection neurons to the DRN
So far, the results suggest that CART signaling in the DRN may play a role in orchestrating anxiety-like behavior via modulation of 5-HT activity. We next sought to map the anatomical origin of CART inputs to the DRN by injecting a Cre-inducible retrograde tracer, AAVrg-hsyn-DIO-EGFP, into the DRN of CART-cre x Ai14 (tdTomato reporter mice) (Fig. 4A). Of the regions analyzed, the highest density of retrogradely labeled CART neurons was observed in the EWcp (Fig. 4B,F) and the NAc (Fig. 4B,E), and the least was observed in the HP (Fig. 4B) and MEA (Fig. 4B,G). Various hypothalamic nuclei also contained labeled CART neurons that project to DRN. This includes the ARC (Fig. 4B,H), VMH (Fig. 4B,I), TuLH (Fig. 4B,J), and PVN (Fig.4B,K). The proportion of DRN-projecting CART neurons relative to the total CART population within these regions is depicted in Figure 4C, and the percentage of DRN-projecting CART neurons within these regions is depicted in Figure 4D.
Retrograde tracing of DRN-projecting CART neurons. A, Diagram showing a cre-dependent retrograde tracer (AAVrg-hsyn-DIO-EGFP) injected into DRN of CART-Cre x Ai14 mice and a representative confocal image showing the expression of EGFP in the DRN after 3 weeks. Scale bar, 200 µm. B, Number of DRN-projecting CART neurons per square millimeter in the nucleus accumbens (NAc), Edinger–Westphal nucleus (EW), medial amygdala (MEA), arcuate nucleus (ARC), ventromedial hypothalamus (VMH), tubular hypothalamus (TuLH), paraventricular hypothalamus (PVN), and hippocampus (HP). C, The proportion of CART input neurons to total CART neurons in NAc, EWcp, MEA, ARC, VMH, TuLH, PVN, and HP per square millimeter. D, Percentage of DRN-projecting CART neurons in NAc, EWcp, MEA, ARC, VMH, TuLH, PVN, and HP. Representative confocal image showing CART-positive retrograde labeled neurons in (E) NAc, (F) EWcp, (G) MEA, (H) ARC, (I) VMH, (J) TuLH, and (K) PVN. Scale bars: 200 µm (NAc) and 100 µm (other regions). Part of the figure was made in BioRender.
Acute stress increased the excitability of DRN-projecting CART neurons located in the EWcp
We first performed ex vivo electrophysiological recordings to compare the intrinsic excitability of DRN-projecting CART neurons located in the PVN, EWcp, VMH, and NAc, four brain regions with large populations of such neurons. Interestingly, DRN-projecting CART neurons in the PVN were the most excitable, followed by those in the EWcp, as assessed by rheobase (i.e., the minimum depolarizing current required to evoke action potentials), input resistance, and evoked spike frequency. Nevertheless, PVN CART neurons appeared to easily reach the state of depolarization block (i.e., being unable to generate more action potentials in response to further depolarization). In comparison, NAc CART neurons were quiescent, while VMH CART neurons appeared to be a heterogeneous neuronal population with diverse intrinsic electrophysiological properties. First, based on comparisons of rheobase, it was the easiest to evoke action potentials in PVN and EWcp CART neurons, and the hardest in NAc neurons, followed by VMH neurons (one-way ANOVA, F3,38 = 17.94, p < 0.0001, with post hoc Tukey tests showing multiple significant pairwise comparisons; sample sizes: PVN, n = 11; EWcp, n = 8; VMH, n = 14; and NAc, n = 9; Fig. 5A,B). In line with the rheobase results, DRN-projecting CART neurons in the PVN had the highest input resistance, followed by those in the EWcp and VMH, and the NAc CART neurons had the lowest input resistance (one-way ANOVA, F3,39 = 23.04, p < 0.0001, with post hoc Tukey tests showing multiple significant pairwise comparisons; sample sizes: PVN, n = 11; EW, n = 8; VMH, n = 15; and NAc, n = 9; Fig. 5C,D). Moreover, injecting incremental depolarizing currents (at 20 steps, 10 pA/step until 200 pA) evoked more action potentials in PVN neurons than in EWcp or NAc neurons (one-way ANOVA for the area under the curves/AUCs between the spike frequency curves and the x-axis: F3,39 = 6.626, p = 0.001, with post hoc Tukey tests showing PVN vs EWcp, p = 0.0113, and PVN vs NAc, p = 0.0012; sample sizes: PVN, n = 11; EWcp, n = 8; VMH, n = 15; and NAc, n = 9; Fig. 5E–G). In addition, hyperpolarization-activated cation current (Ih) appeared in every EWcp neuron we recorded, as shown by a prominent “sag” in the voltage response to a hyperpolarizing current injection (Fig. 5C), which is consistent with a previous report (Topilko et al., 2022). Furthermore, in PVN CART neurons, depolarization block was often observed when injected currents exceeded 110 pA (Fig. 5E,F), which could serve as a mechanism to protect PVN neurons from excessive firing activity (Bianchi et al., 2012).
Acute stress increased the excitability of DRN-projecting CART neurons in the EWcp of CART-Cre x Ai14. A, Representative traces of a current ramp used to determine rheobase by incrementing the magnitude of depolarizing currents until action potentials were evoked. B, Violin plot of rheobase in DRN-projecting CART neurons in the PVN, EWcp, VMH, and NAc. C, Representative traces for the input resistance experiments: assessing changes in membrane potential upon the injection of −100 pA hyperpolarizing currents. D, Violin plot of input resistance in DRN-projecting neurons in the PVN, EWcp, VMH, and NAc. E, Representative traces for assessing evoked spike (action potential, AP) frequency upon the injection of incremental depolarizing currents for 250 ms at 20 steps (from 10 to 200 pA, with 10 pA/step). F, Curves depicting the frequency of spikes (APs) evoked by incremental depolarizing currents in DRN-projecting CART neurons in the PVN, EWcp, VMH, and NAc. G, Violin plot of area under the curve (AUC) shown in F. H, L, P, T, Coronal sections that contain the brain regions targeted in electrophysiological recordings. I, M, Q, U, Confocal images of representative DRN-projecting CART neurons in the PVN, EWcp, VMH, and NAc, respectively. Scale bar, 20 µm. J, N, R, V, Representative traces obtained at the 200 pA current step from the PVN, EWcp, VMH, and NAc, respectively. K, O, S, W, Curves depicting frequency of spikes (APs) evoked by incremental depolarizing currents in mice with and without exposure to 30 min restraint stress from the PVN, EWcp, VMH, and NAc, respectively. Acute stress led to increased spike frequency in EWcp neurons. All the current-clamp recordings were performed when the membrane potential was held at −70 mV. Data were analyzed using one-way or two-way ANOVA followed by post hoc Tukey or Šidák pairwise comparisons. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001; ns, nonsignificant.
To assess whether DRN-projecting CART neurons in the PVN, EWcp, VMH, and NAc were associated with stress response and anxiety, we performed ex vivo electrophysiological recordings right after the mice were exposed to acute (30 min) restraint stress. Our data showed that acute stress led to increased excitability of CART neurons in the EWcp, but not in the VMH or NAc (for evoked spike frequency, in the EWcp, two-way ANOVA: interaction effect between current and stress, F19,323 = 3.339, p < 0.0001, with a post hoc Šidák test showing that at the 200 pA current step, control vs stress: p = 0.0486, ncontrol = 8 and nstress = 11, Fig. 5L–O; no significant interaction effects or main effects of stress in the VMH, ncontrol = 15 and nstress = 16, Fig. 5P–S, or in the NAc, ncontrol = 9, and nstress = 5, Fig. 5T–W). Interestingly, visual inspection indicates that acute stress might have delayed the onset of depolarization block in PVN CART neurons, and a two-way ANOVA indicated that there was a significant interaction between current and stress. However, pairwise comparisons did not detect significant differences in spike frequency between the control and stress groups in response to any magnitude of current injection (two-way ANOVA: interaction effect between current and stress, F19,437 = 3.391, p < 0.0001; post hoc Šidák tests showing no significant pairwise comparisons, ncontrol = 11 and nstress = 14; Fig. 5H–K).
CART neurons in the EWcp form synaptic connections with DRN neurons
Restraint stress electrophysiology experiments confirmed that EWcp CART neurons are robustly modulated by stress, which led us to speculate that these neurons may regulate stress-related behaviors via efferent projections to the DRN 5-HT system. We next sought to map EWcp CART projections to the DRN by injecting the anterograde tracer AAV8.2-ef1α-DIO-synaptophysin into the EWcp of CART-cre mice (Fig. 6A), which revealed a prominent CART projection to the rostral DRN (Fig. 6B–D) compared with the mid-DRN (data not shown). Notably, we detected a few eYFP-labeled presynaptic CART terminals in close proximity to TPH2 neurons (Fig. 6D). To determine whether DRN 5-HT neurons receive direct synaptic input from EWcp CART neurons, we injected CART-cre mice with the anterograde transsynaptic tracer AAV9-DIO-WGA-ZsGreen (Fig. 6E) in the EWcp (Han et al., 2021). This resulted in strong ZsGreen labeling in the rostral DRN (Fig. 6F–H) with 26.53 and 73.46% of the ZsGreen area co-labeled with TPH2 neurons and non-TPH2 neurons, respectively (Fig. 6J, left pie chart). The mid-DRN received fewer CARTEWcp projections as compared with rostral DRN, showing 11.49 and 88.5% of the ZsGreen area co-labeled with TPH2 neurons and non-TPH2 neurons, respectively (Fig. 6J, right pie chart). In both experiments, the caudal DRN neurons received fewer to no projections from CARTEWcp neurons. Both synaptophysin and WGA-ZsGreen revealed that CARTEWcp projections are predominant in non-TPH2 neurons as compared with TPH2 neurons in DRN. Recently, orphan GPR-160 GPCR has been identified as a putative receptor for CART (Yosten et al., 2020). Immunofluorescence analysis of GPR-160 in the DRN demonstrates a higher expression of GPR-160 within non-5-HT neurons as compared with 5-HT neurons (Rostral-DRN, 44.26% in 5-HT and 55.73% in non-5-HT neurons in the rostral DRN; Mid-DRN, 25.84% in 5-HT and 74.15% in non-5-HT neurons; Fig. 6K–N). The proportion of GPR-160 expression aligns with the proportion of CARTEWcp anterograde transsynaptic tracer (ZsGreen) in DRN neurons. These results suggest that CARTEWcp neurons project predominantly to non-5-HT DRN neurons and may activate microcircuits in the DRN that ultimately regulate 5-HT neuron activity via feedforward inhibition.
Anterograde tracing of CARTEWcp→DRN neurons. A, A confocal image of the Edinger–Westphal nucleus of CART-cre mice that received an anterograde tracer AAV8.2-hef1a-DIO-synaptophysin-EYFP. The viral construct expresses synaptophysin in a cre-dependent manner in CART neurons. Scale bar, 100 µm. Confocal image of rostral DRN showing (B) synaptophysin processes, (C) TPH2 and synaptophysin, and (D) an inset of the merged image. Scale bar, 200 µm. E, A confocal image of the Edinger–Westphal nucleus of CART-cre mice that received an anterograde transsynaptic tracer AAV2/9-CAG-DIO-WGA-ZsGreen. The viral construct expresses WGA-ZsGreen in a cre-dependent manner in post-synaptic neurons. Scale bar, 100 µm. Confocal image of rostral DRN showing (F) ZsGreen-labeled neurons, (G) TPH2 and ZsGreen, and (H) an inset of the merged image. Scale bar, 200 µm. I, Number of ZsGreen-positive cells in the DRN (rostral, mid, and caudal). J, Percentage of ZsGreen area in TPH2 and non-TPH2 neurons in rostral (right) and mid-DRN (left). Confocal image of rostral DRN showing (K) GPR-160, (L) GPR-160 and 5-HT, and M an inset of the merged image. Scale bar, 200 µm. N, Percentage of GPR-160 area in 5-HT and non-5-HT neurons in rostral (right) and mid-DRN (left). Part of the figure was made in BioRender.
CART overexpression in CART-producing neurons of the EWcp decreases 5-HTDRN activity
To assess whether the CARTEWcp affects DRN 5-HT neuronal excitability, we employed a Cre-dependent AAV to overexpress CART in the EWcp of CART-Cre mice (Funayama et al., 2023) and recorded from 5-HTDRN neurons 6 weeks later using ex vivo whole-cell patch-clamp electrophysiology (Fig. 7A). We validated the AAV2/8-ef1α-DIO-Cartpt-2A-nlstdTomato-WPRE construct using RNAscope. The pixel puncta analysis revealed that the CART mRNA levels were higher in the overexpressing mice as compared with the control AAV: AAV2/8-ef1α-DIO-nlstdTomato-WPRE (t7 = 2.831, p = 0.0299; Fig. 7B,C). Results showed that CART overexpression in the EWcp led to a reduction in the excitability of DRN 5-HT neurons. First, the minimum depolarizing current (i.e., rheobase) needed to evoke action potentials was higher after EWcp CART overexpression, although no change in input resistance was observed (for rheobase, unpaired t test: t28 = 3.351, p = 0.0023, n = 15 in each group, Fig. 7E,H; for input resistance, unpaired t test: t28 = 0.8369, p = 0.4098, n = 15 in each group, Fig. 7F,I). Second, injecting incremental depolarizing currents (20 steps, 10 pA/step until 200 pA) evoked fewer action potentials after EWcp CART overexpression (two-way ANOVA showing an interaction effect between current and CART overexpression: F19, 513 = 3.941, p < 0.0001; post hoc Šidák tests producing significant group differences when injected currents were 130, 150, 160, and 190 pA, respectively, Fig. 7G,J; an unpaired t test showing decreased AUCs between firing frequency curves and the x-axis: t27 = 2.507, p = 0.0185, n = 14 and 15 in the control and CART overexpression groups, respectively, Fig. 7K).
CART overexpression in the EWcp decreased the excitability of DRN 5-HT neurons. A, Diagram showing injection of an AAV overexpressing CART (AAV2/8EF1a-DIO-Cartpt-2A-nlstdTomato-WPRE) and tdTomato (AAV2/8EF1a-DIO-nlstdTomato-WPRE) in a cre-dependent in Edinger–Westphal (EW) nucleus. B, Representative RNAscope images showing mRNA expression of CART and tdTomato in EWcp nucleus. C, Quantification of percent CART mRNA-positive area in the EWcp of mice that received control and CART overexpression AAV. D, Representative confocal image of dorsal raphe nucleus (DRN) showing biocytin-labeled TPH2 neurons in electrophysiological recordings. E, Representative traces for accessing rheobase (i.e., minimum depolarizing current to evoke action potentials). F, Representative traces for accessing input resistance upon the injection of −100 pA hyperpolarizing currents. G, Representative traces for spikes (action potentials, APs) evoked within 250 ms upon the injection of 200 pA depolarizing currents. H, Violin plot showing that EWcp CART overexpression led to increased rheobase in DRN 5-HT neurons. I, Violin plot showing that EWcp CART overexpression did not alter input resistance in DRN 5-HT neurons. J, Curves depicting spike (AP) frequency corresponding to the magnitude of depolarizing current. K, Violin plots of area under the curves (AUCs) between the curves in (J) and the x-axis, showing that CARTEW overexpression decreased evoked spike frequency in DRN 5-HT neurons. All the current-clamp recordings were performed when the membrane potential was held at −70 mV. Data were analyzed using unpaired t tests or two-way ANOVA followed by post hoc Šidák pairwise comparisons. *p < 0.05; **p < 0.01; ns, nonsignificant. Part of the figure was made in BioRender.
Chemogenetic activation of the CARTEWcp→DRN pathway induces anxiety
Our data so far suggested that stress activates DRN-projecting CART neurons, which may in turn inhibit 5-HT neurons in the DRN. This implies a potential role for the CARTEWcp→DRN circuit in contributing to anxiety. To investigate this possibility further, we used a chemogenetic approach to activate this circuit and examine its effects on anxiety-like behavior in male and female mice. We targeted Gq-coupled DREADDs to DRN-projecting CARTEWcp neurons by injecting CART-cre mice with a retrograde AAV expressing Cre-dependent FLP recombinase (AAVrg-ef1α-DIO-FLPo-WPRE-hGHpA) in the DRN and a FLP-dependent excitatory, Gq-coupled DREADD (AAVDJ-hsyn-fDIO-hM3Dq-mCherry) or control (AAV8-ef1a-fDIO-mCherry-WPRE) in the EWcp (Fig. 8A). This viral targeting strategy restricted expression of the Gq-coupled DREADDs to pathway-specific CARTEWcp neurons that project to the DRN. Immunofluorescence images confirmed selective expression of the mCherry tag in control and hM3Dq mice to a subset CARTEWcp neurons (Fig. 8B). Chemogenetic activation of the CARTEWcp→DRN circuit following systemic CNO administration induced anxiety-like behavior in male mice, as evidenced by reduced time spent (%) in the open arms of the EPM (t13 = 2.685, p = 0.0199) and increased time spent (%) in the closed arms (t13 = 2.685, p = 0.0199) compared with mCherry controls (Fig. 8C,D). Additionally, hM3Dq males spent significantly less time (%) in the neutral zone (t13 = 4.480, p = 0.0008), suggesting increased anxiety (Fig. 8E). Importantly, no significant locomotor dysfunction was observed in the hM3Dq males (Fig. 8F). Heatmaps from the EPM clearly showed less time spent in the open arms and more time in the closed arms by hM3Dq mice (Fig. 8G), indicating the involvement of this circuit in anxiety-like behavior in males. In the LDB test, chemogenetic activation of this circuit also increased the percentage of time spent in the dark box (t13 = 3.495, p = 0.0044) and decreased the time spent in the light box (t13 = 3.495, p = 0.0044) in males (Fig. 8H,I). No locomotor dysfunction was detected, as measured by distance moved and mean velocity parameters (Fig. 8J,K). Heatmaps from the LDB test further confirmed the reduced time spent in the light box (Fig. 8L). We also assessed social deficits in these mice. The hM3Dq males spent less time interacting with a stranger mouse compared with mCherry controls, indicating social deficits (t13 = 2.638, p = 0.0217; Fig. 8M). Additionally, the time spent interacting with a SC was significantly reduced compared with an EC, with a significant interaction effect between virus and cage type (F1,24 = 5.275, p = 0.0307) and post hoc tests showing decreased SC interaction (p = 0.0294; Fig. 8N).
Chemogenetic activation of CARTEWcp→DRN circuit induces anxiogenic behavior in males. A, Diagram of pathway-specific Gq-coupled (hM3Dq) designer receptors activated only by designer drug (DREADD) manipulation of CARTEWcp→DRN circuit. A Cre-dependent FLP (AAVrg-EF1a-DIO-FLPo-WPRE-hGHpA) was injected into the DRN and an FLP-dependent hM3Dq (AAVDJ-hsyn-fDIO-hM3Dq-mCherry) or control (AAV8-ef1a-fDIO-mCherry-WPRE) in the EWcp of CART-cre male mice. This will activate the CART neurons in the EWcp that project to DRN. Schematic representation of the experiments showing the surgery and behavior schedule. B, Representative confocal image showing mCherry expression in CART neurons in EWcp of CART-cre mice that received a control AAV overexpressing mCherry (right) AAV overexpressing hM3Dq-mCherry (left). Parameters of elevated plus maze (EPM) test showing % time spent in (C) open arm, (D) closed arm, and (E) neutral zone and (F) distance moved in mCherry and hM3Dq male mice. G, Heat map visualization of time spent in the EPM arms by mCherry and hM3Dq male mice. Light dark box (LDB) exploration test showing % time spent in (H) dark box and (I) light box and (J) distance moved and (K) mean velocity in mCherry and hM3Dq male mice. L, Heat map visualization of time spent in the light box of LDB apparatus by mCherry and hM3Dq male mice. Social interaction test showing (M) % social interaction with strangers and (N) interaction time (s) with empty and stranger cages in mCherry and hM3Dq male mice. O, Diagram showing pathway-specific DREADD manipulation of CARTEWcp→DRN circuit in the females. Parameters of elevated plus maze (EPM) test showing % time spent in (P) open arm, (Q) closed arm, and (R) neutral zone. S, Distance moved in mCherry and hM3Dq female mice. Light dark box (LDB) exploration test showing % time spent in (T) dark box and (U) light box. V, Distance moved in mCherry and hM3Dq female mice. Social interaction test showing (W) % social interaction with strangers and (X) interaction time (s) with empty and stranger cages in mCherry and hM3Dq female mice. Values (n = 5–7/group) are represented as means (±SEM), and the data were analyzed by unpaired t tests or two-way ANOVA followed by post hoc Bonferroni's pairwise comparisons (*p < 0.05, **p < 0.01, ***p < 0.001 vs mCherry). Part of the figure was made in BioRender.
Next, we explored the anxiogenic effects of CARTEWcp→DRN circuit activation in female CART-cre mice (Fig. 8O). Unlike the males, activation of this circuit did not induce anxiety-like behavior in females across any of the measures tested. In the EPM, there was no significant change in the percentage of time spent in the open, closed, or neutral zones (Fig. 8P–R). No locomotor dysfunction was observed in females (Fig. 8S). Similarly, in the LDB test, no changes were observed in the percentage of time spent in the light or dark box (Fig. 8T,U). Locomotor assessments showed no change in activity level in hM3Dq females (Fig. 8V). Additionally, no social deficits were observed in female mice following activation of the CARTEWcp→DRN circuit (Fig. 8W,X).
In summary, chemogenetic activation of the CARTEWcp→DRN circuit induces anxiogenic behavior and social deficits in male CART-cre mice, but not in females, suggesting a sex-specific role for this circuit in anxiety.
Chemogenetic activation of the CARTEWcp→DRN circuit activates GABADRN neurons and inhibits 5-HTDRN neurons
To investigate the effects of CARTEWcp neurons on DRN neurotransmission, we used c-fos as a marker of neuronal activity following chemogenetic activation of the CARTEWcp→DRN pathway. Following systemic CNO injection, brains were collected after 2 h to detect c-fos expression in a subset of DRN neurons. RNAscope was performed on DRN slices from both hM3Dq and mCherry control groups using mRNA probes specific to Gad1, Tph2, and c-fos. We analyzed c-fos expression in Tph2 and/or Gad1 neurons. Our analysis revealed a lower proportion of c-fos-positive neurons within the Tph2 population in hM3Dq males compared with mCherry controls, indicating reduced activity in Tph2 neurons (t12 = 3.131, p = 0.0096; Fig. 9A). Conversely, the proportion of c-fos-positive neurons within the Gad1 population was higher in hM3Dq males, suggesting increased activity in Gad1 neurons (t12 = 3.014, p = 0.0118; Fig. 9B). We conducted separate analyses on rostral, mid, and caudal DRN neurons, as EWcp CART projections are robust to rostral DRN. In rostral DRN, the proportion of c-fos-positive Tph2 neurons was significantly reduced (t12 = 2.837, p = 0.0162; Fig. 9D), while the percentage of c-fos-positive Gad1 neurons increased (t12 = 2.942, p = 0.0134; Fig. 9E). Similar changes were observed in mid-DRN, with decreased c-fos-positive Tph2 neurons (t12 = 2.396, p = 0.0355; Fig. 9G) and increased c-fos-positive Gad1 neurons (t12 = 2.399, p = 0.0353; Fig. 9H). No changes were observed in the caudal DRN neurons following chemogenetic activation of the circuit (Fig. 9J,K), suggesting that CARTEWcp neurons primarily influence rostral and mid-DRN Tph2-Gad1 neurons. These findings are consistent with the anterograde tracing data, which showed strong projections of CARTEWcp neurons to the rostral and mid-DRN, but not to the caudal DRN. Furthermore, no changes in c-fos activation were detected in Tph2-Gad1 coexpressing neurons following circuit activation (Fig. 9F,I,L). Overall, the RNAscope data suggest that chemogenetic activation of the CARTEWcp→DRN circuit increases c-fos activity in GABAergic neurons and decreases c-fos activity in serotonergic neurons (Fig. 9M–T). These results imply that CARTEWcp neurons could modulate serotonergic neuronal activity, and by extension anxiety-like behavior, indirectly via activation of a local inhibitory microcircuit in the DRN.
Chemogenetic activation of CARTEWcp→DRN circuit activates GABA neurons and inhibits TPH2 neurons in the males. Violin plot showing (A) % of c-fos-positive Tph2 neurons, (B) % of c-fos-positive Gad1 neurons, and (C) % of c-fos-positive Tph2-Gad1 co-expressing neurons in the DRN of mCherry and hM3Dq group. Violin plot showing (D) % of c-fos-positive Tph2 neurons, (E) % of c-fos-positive Gad1 neurons, and (F) % of c-fos-positive Tph2-Gad1 co-expressing neurons in the rostral DRN of mCherry and hM3Dq group. Violin plot showing (G) % of c-fos-positive Tph2 neurons, (H) % of c-fos-positive Gad1 neurons, and (I) % of c-fos-positive Tph2-Gad1 co-expressing neurons in the mid-DRN of mCherry and hM3Dq group. Violin plot showing (J) % of c-fos-positive Tph2 neurons, (K) % of c-fos-positive Gad1 neurons, and (L) % of c-fos-positive Tph2-Gad1 coexpressing neurons in the caudal DRN of mCherry and hM3Dq group. Representative images showing the c-fos mRNA puncta within Tph2 and/or Gad1 mRNA-positive neurons in the (M) mCherry and (N) hM3Dq group. Single-channel images showing (O, P) c-fos mRNA, (Q, R) Gad1 mRNA, and (S, T) Tph2 mRNA in the mCherry and hM3Dq males. Scale bar, 50 µm. Values (n = 7/group) are represented as means (±SEM), and the data were analyzed by unpaired t tests (*p < 0.05, **p < 0.01, vs mCherry).
Discussion
We investigated the interactions between EWcp-CARTergic and DRN 5-HTergic systems and their role in mediating anxiety states by employing state-of-the-art tools for manipulating and monitoring neural circuits. Previous studies have shown that central or site-directed administration of CART in the amygdala or NAc induces anxiogenic behaviors in rats (Chaki et al., 2003; Stanek, 2006; H. S. Yoon et al., 2014), but the role of DRN serotonergic circuits in CART-mediated behaviors had not been explored. We demonstrate that CART infusion into the DRN invokes anxiety-like behavior and social aversion in mice, which may be mediated by inhibition of 5-HT signaling as indicated by LCMS and fiber photometry data. This model is congruent with previous reports suggesting that optogenetic stimulation of 5-HTDRN neurons rescues social deficits in Shank3 KO mice (Luo et al., 2017). The Fos-TRAP findings further suggest that CART inhibits 5-HT neurons in the ventral DRN, which was previously implicated in stress-coping behaviors via efferent projections to the cortex (Ren et al., 2018). Overall, these findings strongly suggest that CART signaling promotes anxiety-like behavior in the DRN via inhibition of a serotonergic subpopulation that buffers against the aversive effects of stress.
This study also pinpoints the EWcp as the likely source of CART inputs to the DRN that orchestrates stress-related behaviors. While CART cell bodies were not directly observed in the DRN, there was robust labeling of CART-containing axons which likely represent afferent inputs from other regions. This was confirmed by retrograding tracing studies confirming the presence of DRN-projecting CART neurons in the EWcp, ARC, VMH, LH, PVN, MEA, HP, neocortex, and NAc, some of which have been indicated by previous studies (Y. S. Yoon and Lee, 2013; Lee and Lee, 2014). These regions are implicated to some extent in anxiety and stress (Hunter et al., 2007; Dandekar et al., 2008; H. S. Yoon et al., 2014; Ahmadian-Moghadam et al., 2018; Walker et al., 2021), so we characterized CART neurons in these input regions and their responsiveness to stress using ex vivo slice electrophysiology. DRN-projecting CART neurons in the EWcp, VMH, PVN, and NAc exhibit varying levels of intrinsic excitability at baseline, with CARTEWcp→DRN and CARTPVN→DRN neurons being the most excitable. PVN neurons quickly reached a depolarization block, suggesting a protective mechanism against excessive firing activity (Bianchi et al., 2012). Importantly, stress selectively increased excitability in CARTEWcp→DRN input neurons, but not in other regions examined. Previous studies have similarly demonstrated heightened CARTEWcp c-Fos activity and elevated CART mRNA levels following restraint stress (Okere et al., 2010; Xu et al., 2014). Increased CART expression in the EWcp has also been observed in suicide victims, suggesting an involvement of these neurons in anxiety and depression (Bloem et al., 2012) and underscoring the need for further investigation into the CARTEWcp→DRN circuit in psychiatric conditions.
Surprisingly, the impact of stress on the excitability of CART input neurons in the hypothalamus was less robust despite its well-delineated role in stress response. CART input neurons in the VMH displayed high variability in basal excitability and stress responsiveness, suggesting a heterogeneous population that may co-express other neurotransmitters that are crucial in appetitive or reward regulation (Ahmadian-Moghadam et al., 2018; Sagarkar et al., 2021). Thus, all CARTVMH→DRN neurons may not be critically involved in stress response, which bears further examination. The PVN also abundantly expresses CART and has been linked to stress in traumatic brain injury (TBI), where CART expression was reduced in male rats with TBI (J. Liu et al., 2017). In females, however, CART was increased in the PVN under swim stress conditions (Gozen et al., 2007). We also observed a significant interaction effect of stress on firing frequency in the subset of CARTPVN neurons that project to the DRN, suggesting that this subpopulation may also be involved in stress physiology. Specifically, stress appeared to delay depolarization block in CARTPVN→DRN neurons, potentially increasing their intrinsic excitability. Future studies are warranted to better understand these nuanced responses of CARTPVN→DRN neurons to stress.
Interestingly, DRN-projecting CART neurons in the NAc were the least excitable relative to those located in the EWcp, PVN, and VMH and were not modulated by stress. The function of the CART peptide in the NAc is not entirely clear, but it may be a positive or negative homeostatic regulator of dopamine-mediated activity (Job and Kuhar, 2017). This is supported by another study showing that stimulation of 5-HT4Rs in the NAc regulated anorexia by increasing CART mRNA levels (Jean et al., 2007). Thus, there may be reciprocal projections between CARTNAc and 5-HTDRN neurons, but DRN-projecting CARTNAc neurons may not be the primary mediators of stress or anxiety. Instead, they may play a significant role in regulating reward and feeding behavior (Nakamura, 2013).
Based on these electrophysiological studies, we focused on the stress-sensitive CARTEWcp→DRN pathway as a potential source of CART input to the DRN that plays a physiological role in orchestrating anxiety and aversive behavior. CARTEWcp projections to the DRN have been reported previously (Topilko et al., 2022; Priest et al., 2023), so we wanted to assess their regional distribution and post-synaptic targets using anterograde and trans-synaptic tracers. EYFP-synaptophysin was used to label CART fibers in the DRN, while WGA-ZsGreen–labeled post-synaptic target neurons. EYFP-labeled CART axons were observed in the rostral and mid-DRN, with more robust labeling in the rostral portion. ZsGreen was more abundant in the rostral DRN and primarily targeted non-5-HT neurons, suggesting that CARTEWcp inputs may modulate 5-HT neuronal activity indirectly via a non-5-HT inhibitory microcircuit. This is supported by additional data showing that chemogenetic activation of the CARTEWcp→DRN pathway increases c-fos activity in local GABAergic neurons while decreasing activity in 5-HT neurons.
Given that CARTEWcp neurons coexpress other neuropeptides that have been implicated in anxiety [e.g., nesfatin, pituitary adenylate cyclase-activating polypeptide (PACAP), cholecystokinin (CCK); Merali et al., 2008; Bowers et al., 2012; Hammack and May, 2015], we wanted to specify the role of CART peptide in driving these effects. To this end, we showed that CART overexpression within CARTEWcp neurons reduces the excitability of 5-HTDRN neurons. This provides converging evidence that stress induction of CART peptide in CARTEWcp neurons elevates anxiety by modulating serotonergic circuitry in the DRN.
In addition to its effects on local 5-HT circuits, chemogenetic stimulation of the CARTEWcp→DRN pathway also increased anxiety and social deficits in a sex-specific manner. Prior research has shown that the CART and ER-β colocalize in EWcp neurons and that the CART-immunoreactive area is 19% lower in females, suggesting a discrepancy in CART biosynthesis within EWcp neurons that may influence downstream signaling related to anxiety and sociability (Cano et al., 2021). Moreover, c-Fos expression in the EWcp of rats subjected to chronic mild stressors was selectively increased in males (Xu et al., 2012). Taken together, the evidence suggests that the CARTEWcp→DRN circuit plays a critical role in stress, anxiety, or affective behavior in males, but not in females. However, another study recently reported that chemogenetic manipulation of CARTEWcp neurons modulates anxiety-like behavior in both sexes (Priest et al., 2023), which contrasts with our sex-specific effects. This discrepancy may be attributed to the fact that this prior study targeted all CARTEWcp neurons, whereas we isolated the DRN-projecting population. In our retrograde tracing study, we found that approximately one-third of CARTEWcp neurons were non-DRN projecting, so this population may have contributed to the effects observed in females.
Given that CARTEWcp neurons project predominantly to non-5-HT neurons in the DRN, we speculate that CART may exert its effects on 5-HT function and anxiety by modulating non-5-HT neurons, which consist primarily of GABA- and glutamate-producing neurons (Bang et al., 2012). Trans-synaptic tracing studies revealed that DRN-GABA interneurons densely project to and monosynaptically inhibit 5-HT neurons (R. Liu et al., 2000; Sparta and Stuber, 2014; Weissbourd et al., 2014). Our data also confirm that CARTEWcp neurons activate GABAergic (Gad1) neurons in the DRN while inhibiting 5-HT (Tph2) neurons. This aligns with the distribution of the putative CART receptor, GPR-160, in non-5-HT neurons relative to 5-HT neurons. Previous studies have also demonstrated the involvement of GABA-5-HT DRN microcircuits in response to aversive or reward stimuli. During aversive stimuli, GABAergic neurons are activated with reduced 5-HT neuron firing, while positively valenced stimuli inhibit GABAergic activity (Y. Li et al., 2016). In the social defeat model, DRN-GABAergic sensitization and 5-HT inhibition persisted in susceptible mice but not in resilient mice. This highlights the critical role of the GABA-5-HT DRN microcircuit in regulating socio-affective behaviors (Challis et al., 2013). Similarly, our study demonstrates that CARTEWcp neurons may regulate this DRN-GABA microcircuit, contributing to anxiety and social deficits in male mice. This CARTEWcp→DRN circuit likely influences other subcortical or cortical structures that modulate stress-related behaviors (Garcia-Garcia et al., 2018; Ren et al., 2018).
In summary, this study establishes the neuromodulatory role of CARTEWcp neurons in 5-HTDRN neurotransmission, highlighting the CARTEWcp→DRN circuit as a key driver of sex-specific anxiety-like behavior in males. It further suggests that CART may modulate 5-HT transmission by activating inhibitory microcircuits via the putative CART receptor GPR-160. Future studies are warranted to explore the CART regulation of this DRN microcircuit and further elucidate the intricate role of the CARTEWcp→DRN circuit in stress-related behavior.
Footnotes
We thank Dr. Jeffrey Friedman (The Rockefeller University, New York, NY, USA) for donating the Cart-IRES2-Cre-D mice, Dr. Uchida Shusaku (Kyoto University, Kyoto, Japan) for providing the CART overexpression virus, and Dr. Yulong Li (Peking University, Beijing, China) for offering the g5-HT2h (GRAB5-HT3.5) biosensor. We acknowledge the University of Iowa Metabolomics Core Facility for performing LCMS analysis and the Viral Vector Core Facility for production of AAV particles. We thank Yu Xu for maintaining mouse colonies and Thomas D. James for technical assistance in fiber photometry. This work was supported by funds from the Iowa Neuroscience Institute and the Department of Neuroscience and Pharmacology at the University of Iowa to C.A.M.
The authors declare no competing financial interests.
- Correspondence should be addressed to Catherine A. Marcinkiewcz at catherine-marcinkiewcz{at}uiowa.edu.