Abstract
The lateral habenula (LHb) has emerged as a pivotal brain region implicated in depression, displaying hyperactivity in human and animal models of depression. While the role of LHb efferents in depressive disorders has been acknowledged, the specific synaptic alterations remain elusive. Here, employing optogenetics, retrograde tracing, and ex vivo whole-cell patch-clamp techniques, we investigated synaptic transmission in male mice subjected to chronic social defeat stress (CSDS) at three major LHb neuronal outputs: the dorsal raphe nucleus (DRN), the ventral tegmental area (VTA), and the rostromedial tegmental nucleus (RMTg). Our findings uncovered distinct synaptic adaptations in LHb efferent circuits in response to CSDS. Specifically, CSDS induced in susceptible mice postsynaptic potentiation and postsynaptic depression at the DRN and VTA neurons, respectively, receiving excitatory inputs from the LHb, while CSDS altered presynaptic transmission at the LHb terminals in RMTg in both susceptible and resilient mice. Moreover, whole-cell recordings at projection-defined LHb neurons indicate decreased spontaneous activity in VTA-projecting LHb neurons, accompanied by an imbalance in excitatory–inhibitory inputs at the RMTg-projecting LHb neurons. Collectively, these novel findings underscore the circuit-specific alterations in LHb efferents following chronic social stress, shedding light on potential synaptic adaptations underlying stress-induced depressive-like states.
- chronic social defeat stress
- dorsal raphe nucleus
- lateral habenula
- RMTg
- synaptic adaptations
- ventral tegmental area
Significance Statement
The lateral habenula (LHb) is a brain region responsible for encoding negative signals and tends to be overactive in both depressed individuals and animal models of depression. Distinct groups of neurons within the LHb connect with the dorsal raphe nucleus, the ventral tegmental area, and the rostromedial tegmental area, implying that they serve distinct functions. Our study demonstrates that chronic social defeat stress, a widely used animal model of clinical depression, leads to specific adaptations in synaptic transmission and neuronal activity along these pathways. These findings suggest that the outputs of LHb neurons play distinct roles in the onset and progression of depressive symptoms commonly observed in major depression.
Introduction
Major depressive disorders are among the most prevalent mental illnesses affecting ∼16% of the global population (Kessler et al., 2003). Symptoms associated with depressive disorders include an inability to experience positive emotions (anhedonia), difficulty concentrating, memory impairment, loss of motivation, exaggerated feelings of helplessness and hopelessness, and the presence of suicidal thoughts. Underlying these different symptoms are alterations in different brain structures of the reward circuit involved in the integration and processing of emotions (Russo and Nestler, 2013). However, our understanding of the specific brain nuclei implicated, along with the synaptic changes, in their microcircuitry responsible for the diverse impairments observed in major depression is still deficient (Krishnan and Nestler, 2008).
One of the most important factors precipitating depressive episodes is emotional and psychological stress (L. Yang et al., 2015). Particularly, the persons subjected to social intimidation and bullying suffer from the consequences of chronic physical and emotional stress and consequently have greater chances of developing depressive disorders (Moore et al., 2017). The victims of social intimidation have impaired social skills, are submissive, and show signs of helplessness (Kaltiala-Heino and Fröjd, 2011). Chronic social defeat stress (CSDS), which mimics social bullying in humans, is a well-validated animal model that has been extensively used to understand the pathophysiology of depressive disorders (Hollis and Kabbaj, 2014). In this model, mice exposed to 10 d of social defeat stress exhibit several depressives-like behaviors, including helpless behavior, anhedonia, and social avoidance (Golden et al., 2011), which can be reversed by antidepressants (Berton et al., 2006; Heshmati et al., 2020). The CSDS has particularly been a valuable model of clinical depression to delineate the maladaptive synaptic adaptations within the brain reward circuit that are believed to mediate the behavioral and cognitive deficits induced by chronic social stress (Christoffel et al., 2011; Golden et al., 2013; Knowland et al., 2017).
The lateral habenula (LHb) has emerged as a key nucleus in depressive disorders (Proulx et al., 2014; Hu et al., 2020). The LHb receives signals from the basal ganglia involved in the selection of actions (Bromberg-Martin et al., 2010; Shabel et al., 2012) and from various nuclei of the limbic system that play roles in the processing of various cognitive and affective signals and in the control of emotions (Lecourtier and Kelly, 2007; Proulx et al., 2014). In turn, the LHb sends excitatory efferents to the dopaminergic VTA (Lammel et al., 2012; Beier et al., 2019), the serotoninergic dorsal raphe nucleus (DRN; Sego et al., 2014; Weissbourd et al., 2014; Zhou et al., 2017), and the GABAergic rostromedial tegmental nucleus (RMTg), which inhibits dopaminergic centers (Jhou et al., 2009; Hong et al., 2011).
The LHb encodes negative value signals: its neurons activate when an animal receives a punishment or a reward that is less than expected or anticipates punishment (Matsumoto and Hikosaka, 2007). Notably, studies in humans (Morris et al., 1999; Sartorius et al., 2010), and in animal models of depression (Shumake et al., 2003; B. Li et al., 2011; K. Li et al., 2013), suggest that maladaptive neuronal activity in the LHb contributes to major depression. In support of this hypothesis, neurons in the LHb (K. Li et al., 2013; Tchenio et al., 2017; Cui et al., 2018; Y. Yang et al., 2018) and afferences synapsing onto LHb neurons (Shabel et al., 2014; Knowland et al., 2017; Tchenio et al., 2017) are hyperactive in animal models of depression. However, how chronic stress impacts neuronal transmission at LHb neuronal outputs is still largely limited. The LHb inputs to the dorsal raphe nucleus (DRN), VTA, and RMTg arise from distinct LHb neuronal populations (Gonçalves et al., 2012; Cerniauskas et al., 2019), suggesting that these outputs have segregated functional roles in the transmission of aversive signals and behavior control. Consequently, maladaptive transmission induced by chronic emotional stress at these LHb neuronal outputs is likely to have distinct roles in major depression (Bernard and Veh, 2012; Gonçalves et al., 2012; Petzel et al., 2017).
Here, we have combined whole-cell electrophysiological recordings, optogenetic, and neuronal tracing approaches to assess whether chronic social stress impacts intrinsic activity and synaptic transmission at the LHb neurons innervating major downstream aminergic centers. Our results show that CSDS promotes distinct and specific synaptic adaptations at the LHb neuronal outputs projecting to the DRN, VTA, and RMTg. We also found that chronic stress differently altered intrinsic properties at LHb projecting neurons. These results support that signal transmission at the LHb neuronal outputs is likely to play distinct roles in the etiology and development of depressive deficits observed in major depression.
Materials and Methods
Animal
Male C57BL/6 mice of 6–7 weeks used in this study were purchased from Charles River Laboratories and allowed 1 week of acclimatation before experimentations. Sexually experienced retired male CD1 breeders (∼40 g) of at least 4 months of age (Charles River Laboratories) were used as aggressors (AGG). C57BL/6 mice were housed in four per cage under standard conditions at their arrival (12 h light/dark cycle at 22–23°C and free access to food and water). Mice were singly housed following CSDS. All the experiments were performed in accordance with the Canadian Guide for the Care and Use of Laboratory Animals guidelines and were approved by the Université Laval Animal Protection Committee.
Stereotaxic injections
Stereotaxic injections were performed under general isoflurane (5%) anesthesia as described before (Martianova et al., 2023). Viral titers ranged from 1012 to 1013 genome copies per milliliter, and 100 nl of AAV2/9-hSyn-ChR2(H134R)-mCherry was injected bilaterally into the lateral habenula [LHb; AP = −1.65, ML = ±0.45, DV = −2.85]. Injections were performed using a glass pipette mounted on a stereotaxic table. The AAVs were infused at a rate of 1 nl/s. At the end of the injection, the pipet was left in situ for 5 min to allow the virus to diffuse into the surrounding tissue. For retrograde labeling, mice were injected unilaterally with fluorescent retrobeads (200 nl; Lumafluor) in the DRN (AP = −4.65, ML = 0, DV = −3.45), VTA (AP = −3.15, ML = ±0.5, DV = −4.4), and RMTg (AP = −4.0, ML = ±0.5, DV = −4.2) using a 1 ml Hamilton syringe (Hamilton).
The animals were kept on a heating pad until they recovered from anesthesia. Experiments were performed 4 weeks (for AAVs) or 1 week (for retrobeads) after stereotaxic injections. Injection sites were confirmed in all animals by preparing coronal sections (100 mm) of injection sites.
Chronic social defeat stress (CSDS)
Chronic social defeat stress (CSDS) was conducted as previously described (Golden et al., 2011). CD1 mice were screened for aggressive behavior during intermale social interactions for 3 consecutive days based on previously described criteria (Golden et al., 2011) and housed in the social defeat cage (26.7 cm width × 48.3 cm depth × 15.2 cm height, Allentown Inc.) for 24 h before the start of defeats on one side of a clear perforated Plexiglass divider. C57BL/6 mice (intruder) were introduced to unknown retired aggressive CD1 mice in their home cages for a period of 5 min during which the C57BL/6 intruder were attacked by the resident CD1. Following this 5 min of physical aggression, C57BL/6 mice were housed on the opposite side of cages divided in half by the perforated wall where it remained in continuous visual and olfactory contact with the CD1 mice without physical harm for the next 24 h. The same procedure was repeated for the following 10 d with each C57BL/6 meeting a new and unknown CD1 mouse every day. The social interaction (SI) test was used to characterize the susceptibility of the defeated mice.
Social interaction
Social avoidance behavior was evaluated with the social interaction (SI) test 24 h after the end of the CSDS protocol as previously described under red-light conditions (Golden et al., 2011). The SI test consisted of two phases. First, mice were placed in a Plexiglas open-field arena (42 cm × 42 cm × 42 cm) with a small wire animal cage placed at one end. Movements were monitored and recorded automatically with a video tracking system (ANY-maze 6.1, Stoelting Co) for 2.5 min to determine baseline exploratory behavior and locomotion in the absence of a social target (no social target). After this initial 2.5 min, mice were removed, and the arena was cleaned. During the second phase, mice were returned to the arena but this time in the presence of an unfamiliar CD1 mouse placed in the wired enclosure (social target). Time spent in the different zones of the arena was determined automatically by the video tracking system. The social interaction (SI) ratio for the interaction zone was calculated by dividing the time spent in the interaction zone with the CD1 mouse present by the time spent in the interaction zone in the absence of the CD1 mouse. Time spent in the corners in the absence and presence of the CD1 was always calculated. Stressed mice were categorized as susceptible if their SI ratio fell below 1 or if their time spent in the interaction zone with the social target present was <25 s. Conversely, mice were categorized as resilient if their SI ratio was above 1 or if their time spent in the interaction zone was above 90 s.
Ex vivo slice electrophysiology
Slice preparation
Mice were deeply anesthetized with isoflurane and intracardially perfused with ice-cold NMDG artificial cerebrospinal fluid (aCSF) containing the following (in mM): 1.25 NaH2PO4-H2O, 2.5 KCl, 10 MgCl2-6H2O, 20 HEPES, 0.5 CaCl2-2H2O, 28 NaHCO3, 8 glucose, 5 Na-ascorbate, 3 Na-pyruvate, 2 thiourea, and 93 NMDG [osmolarity adjusted with sucrose to 300–310 mOsmol/L and pH adjusted to 7.3 with HCl 10N (oxygenated with 95%O2/5%CO2); Ting et al., 2018]. Kynurenic acid (2 mM) was added to the perfusion solution on the day of the experiment. Coronal slices (250 µm) were then cut with a vibratome (VT2000; Leica) to obtain sections containing LHb, DRN, RMTg, or VTA. Slices were placed in a 32°C oxygenated NMDG-aCSF solution for 10 min before incubation for 1 h at room temperature (RT) in HEPES-aCSF solution containing the following (in mM): 1.25 NaH2PO4-H2O, 2.5 KCl, 10 MgCl2-6H2O, 20 HEPES, 0.5 CaCl2-2H2O, 28 NaHCO3, 2.5 D-glucose, 5 Na-ascorbate, 1 Na-pyruvate, 2 thiourea, 92 NaCl, and 20 sucrose (osmolarity adjusted to 300–310 mOsmol/L at pH 7.4).
Whole-cell patch-clamp electrophysiology
All data were acquired by using an Axopatch 200B amplifier, digitized with a Digidata 1500A, and analyzed with pCLAMP 10.6 software (Molecular Devices). Slices were transferred into the recording chamber of an upright microscope (Zeiss) and perfused at a rate of 3–4 ml/min with artificial cerebrospinal solution (aCSF) containing the following (in mM): 120 NaCl, 5 HEPES, 2.5 KCl, 1.2 NaH2PO4, 2 MgCl2, 2 CaCl2, 2.5 glucose, 25 NaHCO3, and 7.5 sucrose. The aCSF in the perfusion chamber was kept at 32°C. For the recording of excitatory postsynaptic currents (EPSCs), gabazine (10 microM) was added to block inhibitory currents mediated by GABAA receptors. Cells were visualized with a 60× objective on an upright fluorescent microscope with a video camera (Zeiss). Borosilicate glass patch pipettes (3–6 MΩ) were filled with internal solution containing the following (in mM): 115 cesium methanesulfonate, 20 cesium chloride, 10 HEPES, 2.5 MgCl2, 4 Na2-ATP, 0.4 Na-GTP, 10 Na-phosphocreatine, 0.6 EGTA, 5 QX314, and 0.2% biocytin (pH 7.35). Signals were filtered at 5 kHz. Pipettes and cell capacitances were fully compensated. Optically evoked postsynaptic responses were obtained delivering 5 ms pulses of 473 nm light through the light path of the microscope using a Colibri 7 LED light source (Zeiss) under computer control. Neurons were voltage-clamped at −60 mV to record AMPAr EPSCs and at +40 mV to record dual component EPSCs containing AMPAr and NMDAr EPSCs. At +40 mV voltage, EPSC from AMPAr and NMDAr are kinetically different. Consequently, to obtain the AMPAr/NMDAr ratio, the peak of the AMPAr EPSC was divided by the magnitude of the NMDAr EPSC measured 50 ms after stimulation, at this point AMPAr-mediated current is negligible (Béïque et al., 2006). This approach allowed us to sample a much larger cell population from an individual mouse. Paired-pulse ratios (PPR) were recorded by giving two 5 ms blue light pulses at −60 mV with a 100 ms interval and calculated by dividing the amplitude of the second peak by the amplitude of the first peak (averaged responses from 12 sweeps).
Spontaneous excitatory postsynaptic currents (sEPSCs) were recorded at −60 mV, and spontaneous inhibitory postsynaptic currents (sIPSCs) at 0 mV in gap-free mode for 5 min in retrogradely identified LHb neurons. sEPSC and IPSC frequency and amplitude were analyzed with Clampfit 11. The E/I ratio was calculated by dividing the frequency of EPSCs by the frequency of IPSCs from the same cell. Pipettes were filled with an intracellular patch solution containing the following (in mM): 127 cesium methanesulfonate, 8 cesium chloride, 10 HEPES, 2.5 MgCl2, 4 Na2-ATP, 0.4 Na-GTP, 10 Na-phosphocreatine, 0.6 EGTA, 5 QX314, and 0.2% biocytin (pH 7.35).
For recordings of action potential characteristics, we obtained whole-cell recordings from retrogradely identified LHb neurons in the current-clamp mode. Pipettes were filled with an intracellular patch solution containing the following (in mM): 130 K-gluconate, 5 KCl, 10 HEPES, 2.5 MgCl2, 4 Na2-ATP, 0.4 Na3-GTP, 10 Na-phosphocreatine, 0.6 EGTA, and 0.2% biocytin (pH 7.35). Signals were filtered at 5 kHz. Pipettes and cell capacitances were fully compensated. One to 2 min after obtaining whole-cell configuration, the resting membrane potential (RMP) was recorded in the current-clamp mode right after whole-cell configuration had been obtained. To examine evoked firing properties, depolarizing current steps (−20 to +100 pA, 20 pA increments, and 300 ms duration) were applied to the cells. Action potentials (APs) generated during this period were counted, and we obtained the number of spikes and frequency of firing. Also, some action potential (AP) properties were analyzed: the first AP half-width, the first AP decay slope, and the first AP amplitude (Pachenari et al., 2019).
Immunohistochemistry
Following recordings, brain slices were incubated in 4% w/v paraformaldehyde (PFA) for 1 h at room temperature (RT), rinsed in 0.1 M phosphate buffer solution (PB, pH 7.4), and stored in PB at 4°C until used for immunohistochemistry. Free-floating coronal sections were first blocked in PB containing 3% normal serum (NS), 0.4% Triton X-100, and 0.3% bovine serum albumin (BSA) for 1.5 h at RT. They were then incubated overnight at 4°C with primary antibodies diluted in the blocking buffer. Primary antibodies were rat anti-mCherry (1:1,000, Thermo Fisher Scientific, M11217), sheep anti-TH (1:500, MilliporeSigma AB1542) for VTA slices, sheep anti-TPH (1:500, MilliporeSigma, AB1541) for DRN slices, and rabbit anti-FoxP1 (1:20,000, Abcam) for RMTg slices. Slices were then rinsed 3 × 10 min in PB and were incubated for 4 h at RT with secondary antibodies diluted in PB containing 3% BSA and 0.4% Triton X-100. Secondary antibodies were goat anti-rat Alexa Fluor 568 (1:1,000, Thermo Fisher Scientific, A11077), donkey anti-sheep Alexa Fluor 647 (1:1,000, MilliporeSigma, SAB4600178), and donkey anti-rabbit Alexa Fluor 647 (1:1,000, Thermo Fisher Scientific, A31573). DAPI was added to stain the nuclei and Streptavidin Alexa Fluor 488 conjugate (1:1,000, Thermo Fisher Scientific, S32354) to specifically label the neurons that were patched and filled with biocytin. After rinsing, sections were mounted on slides using Fluoromount Aqueous Mounting Medium (MilliporeSigma, F4680). For retro beads experiments, free-floating LHb slices were incubated for 4 h in PB with 3% BSA and 0.4% Triton X-100 with Streptavidin Alexa Fluor 568 conjugate (1:1,000, Thermo Fisher Scientific, S11226A) to label patched neurons. DAPI was added to label cell nuclei. Fluorescent images were taken using a Zeiss LSM700 confocal microscope.
Statistics
For most of the data, a nested two-level one-way ANOVA was used specifying the experimental groups (control, resilient, susceptible) as the main factor, and mice were nested within these groups. To do so, and thus to consider the lack of independence of the measurements from the same mouse in our model, we added mouse (subgroups) as a random effect. Post hoc analysis was conducted using Dunnett's test to identify significant pair-wise differences between these groups. Nested analyses were performed in the R environment with the aov function from the stats native package and the DunnetTest function from the DescTools package. Two experimental groups were compared using an unpaired t test (nondirectional) or Mann–Whitney U test using GraphPad Prism (GraphPad Software). Values are reported as mean ± standard error of the mean (SEM). Data distributions were tested using the Shapiro–Wilk normality test. Parametric or nonparametric tests were chosen depending on the number of observations and the distribution.
Results
CSDS induces resilient and susceptible phenotypes
To examine synaptic adaptations to chronic social stress at the LHb circuits, we subjected mice to chronic social defeat stress (CSDS), a well-validated model of major depression, which mimics the impact of social bullying in humans (Golden et al., 2011). C57BL/6 mice were exposed to 5 min physical aggression by an aggressive CD1 mouse for 10 consecutive days. Twenty-four hours after the last defeat, mice were evaluated in the social interaction (SI) test to determine their resilient (RES) or susceptible (SUS) phenotype for social avoidance (Fig. 1A,B). CSDS induced the expression of stress susceptibility and resilience with a main effect of phenotype for the time spent in the interaction zone, the time in the corners when the social target is present, and the social interaction (SI) ratio (Fig. 1C–F). No significant effect was measured for the time spent in the interaction zone and the time in the corners when the target was not present, nor for the locomotion (Fig. 1C–F). Of all the mice that were subjected to CSDS in this study, 39.9% were found resilient in social interaction. Taken together, these results are consistent with previous publications and validate the impact of CSDS on social avoidance.
CSDS induces resilient and susceptible phenotypes. A, Schematic of the CSDS protocol. B, Representative heatmap of control (CTRL), resilient (RES), and susceptible mice evaluated on the interaction test in the presence of a social target. C, CTRL, RES, and SUS mice do not show a significant difference in the time spent in the interaction zone when the social target is absent (F(2,251) = 1.89, p > 0.05), but there is an effect of CSDS when the social target is present (F(2,251) = 267, p < 0.0001). D, There is a significant effect of phenotype for the SI ratio (F(2,251) = 172, p < 0.0001). E, CTRL, RES, and SUS mice do not show a significant difference in the time spent in the corners when the social target is absent (F(2,251) = 1.69, p > 0.05), but there is an effect of CSDS when the social target is present (F(2,251) = 82.3, p < 0.0001). There is no significant effect of phenotype for the locomotion (F(2,251) = 1.33, p > 0.05). ***p < 0.001 between SUS and CTRL.
CSDS induces synaptic potentiation in DRN neurons receiving inputs by the LHb in susceptible mice
To determine whether CSDS elicits synaptic changes at the LHb terminals in the DRN, mice were first injected with an adeno-associated virus (AAV) encoding the opsin channelrhodopsin-2 (ChR2) fused with the fluorescent protein mCherry (AAV-ChR2-mCherry). Three weeks later, they were placed in the CSDS for 10 d, and their phenotype was determined in the SI test. Twenty-four hours later, acute brain slices encompassing the DRN were prepared for whole-cell recordings (Fig. 2A). Figure 2B shows robust expression of ChR2-mCherry in LHb axon terminals in the DRN, interspersed between serotonin neurons, identified by their expression for tryptophan hydroxylase (TPH), a serotoninergic marker. To investigate synaptic transmission at LHb-DRN synapses, optically evoked postsynaptic currents (oEPSCs) were obtained with pairs of light pulses at −60 mV to calculate the paired-pulse ratio (PPR: P2/P1), a marker of presynaptic plasticity, and at +40 mV to calculate the ratio of a-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor (AMPAr)-mediated EPSC to N-methyl-D-aspartate receptor (NMDAr)–mediated EPSC (AMPAr/NMDAr), to assess postsynaptic changes of synaptic transmission (Fig. 2C,D). Our results show that CSDS had no significant impact on the PPR (Fig. 2C). However, CSDS had a significant effect on the AMPAr/NMDAr ratio with a significant increase in SUS mice suggesting postsynaptic potentiation. No significant difference was observed in the absolute amplitude of oEPSC measured at the first peak obtained at −60 mV holding potential [CTRL 113 pA ± 13.5, RES 190 pA ± 30.1, SUS 185 pA ± 33.2 (F(2,28) = 2.19, p > 0.05)]. During whole-cell recordings, DRN neurons were filled with biocytin and processed for post hoc staining for TPH to determine their phenotype as serotoninergic or nonserotoninergic. From 159 DRN neurons, 128 could be phenotyped as TPH+ or TPH−, and from these, 74.2% (95 out of 128) were identified as serotoninergic (TPH+). When restricting the analysis to the TPH+ neurons (Fig. 2C,D), CSDS had no significant effect on the PPR (F(2,26) = 0.09, p > 0.05), but we observed a trend toward a significant effect on the AMPAr/NMDAr ratio (F(2,26) = 3.02, p = 0.07), with significant increase for SUS mice compared with CTRL. No significant effects were found for both the PPR (F(2,17) = 0.44, p > 0.05) and for the AMPAr/NMDAr ratio (F(2,17) = 1.38, p > 0.05) for the TPH− neurons. The significant difference in the AMPAr/NMDAr ratio in SUS mice was due to a fraction of DRN neurons with large (>10 AMPAr/NMDAr ratio) values (Fig. 2D, shaded area). This prompted us to examine the distribution of these neurons in the DRN (Fig. 2E). Interestingly, DRN neurons with large AMPAr/NMDAr were largely clustered in the medial section of the more rostral section of the DRN. Together, these results show that CSDS induces synaptic potentiation in a specific subpopulation of the DRN innervated by the LHb.
CSDS induces synaptic potentiation at the LHb terminals in the DRN in susceptible mice. A, Schematic representation of the experimental timeline. B, Confocal image of the DRN expressing ChR2-mCherry in LHb terminals, DRN neurons positive for the serotoninergic marker TPH, and merged image. C, Representative traces of paired pulses for CTRL, RES, and SUS (top) and plots for paired-pulse ratio for all DRN neurons included (bottom left panel; F(2,28) = 0.48, p > 0.05), TPH+ neurons (bottom middle panel; F(2,26) = 0.09, p > 0.05), and TPH− neurons (bottom right panel; F(2,17) = 0.44, p > 0.05). D, Representative traces and plots for AMPAr/NMDAr ratio for all DRN neurons (left panel; F2,28) = 4.75, p < 0.05), TPH+ neurons (middle panel; F(2,26) = 3.02, p > 0.05), or TPH− neurons (right panel; F(2,17) = 1.38, p > 0.05; number of mice: 11, 14 and 5 for CTRL, SUS, and RES respectively). E, Schematics of the distribution of DRN evaluated, color coded for their phenotype and their ratio of AMPAr/NMDAr >10. The vertical and horizontal scale bars are 50 pA and 100 ms, respectively. *p < 0.05, **p < 0.01 between RES or SUS and CTRL. Raw data are provided in Extended Data Figure 1-1.
Figure 1-1
Tables for raw data presented in Figures 1 and 2. Download Figure 1-1, XLSX file.
CSDS induces distinct adaptation at LHb synapses in the VTA and RMTg
Using the same approach, we next examined whether CSDS had a similar impact on synaptic transmission in the dopaminergic VTA and the GABAergic RMTg, two major LHb neuronal outputs directly and indirectly controlling dopaminergic transmission (Fig. 3A,D) (Lammel et al., 2011; Jhou, 2021). LHb terminals positive for ChR2-mCherry can be found in the VTA, intermingled with neurons positive for the dopaminergic marker TH (Fig. 3B), and in the RMTg, localized for its expression of the protein FoxP1 (Fig. 3E). At the LHb-VTA pathway, presynaptic function was not changed by CSDS with no significant difference in the PPR. However, CSDS significantly decreased the AMPAr/NMDAr ratio in SUS mice compared with the CTRL group (Fig. 3C). From the 73 recorded VTA neurons, 21 were positive for the dopaminergic marker tyrosine hydroxylase (TH+), and 12 were negative for TH (TH−). No significant impact of CSDS was observed for the TH + VTA neuron for either the PPR (F(2,8) = 0.15, p > 0.05) or the AMPAr/NMDAr ratio (F(2.8) = 0.50, p > 0.05). No significant difference was observed in the absolute amplitude of oEPSC measured at the first peak obtained at −60 mV holding potential [CTRL 286 pA ± 72.2, RES 119 pA ± 19.9, SUS 147 pA ± 56.2 (F(2,69) = 2.52, p > 0.05)]. At the LHb-RMTg pathway, while CSDS did not change the AMPAr/NMDAr ratio, we observed a significant decrease in the PPR, suggesting that presynaptic function was altered. This presynaptic adaptation was observed for both the RES and SUS groups when compared with CTRL (Fig. 3F). No significant difference was observed in the absolute amplitude of oEPSC measured at the first peak obtained at −60 mV holding potential [CTRL 217 pA ± 36.1, RES 403 pA ± 85.6, SUS 253 pA ± 35.1 (F(2,75) = 3.05, p > 0.05)]. Taken together, our results show that CSDS induced specific and distinct synaptic adaptations at LHb terminals in three major LHb neuronal outputs.
CSDS induces synaptic depression in the VTA and presynaptic adaptation at the LHb terminals in the RMTg. A, Schematic representation of the experimental timeline for the LHb-VTA pathway. B, Confocal image of the VTA expressing ChR2-mCherry in LHb terminals and VTA neurons positive for the dopaminergic marker TH. C, Representative traces and plots for the PPR (F(2,17) = 0.84, p > 0.05) and AMPAr/NMDAr ratio (F(2,17) = 3.98, p < 0.05; number of mice: 8, 8, and 6 for CTRL, SUS, and RES, respectively). D, Schematic representation of the experimental timeline for the LHb-RMTg pathway. E, Confocal image of the RMTg expressing ChR2-mCherry in LHb terminals and neurons positive for FoxP1. F, Representative traces and plots for the PPR (F(2,19) = 6.20, p < 0.01) and AMPAr/NMDAr ratio (F(2,18) = 0.70, p > 0.05; number of mice: 6, 9, and 5 for CTRL, SUS, and RES, respectively). The vertical and horizontal scale bars are 50 pA and 100 ms, respectively. **p < 0.01, ***p < 0.001 between RES or SUS and CTRL. Raw data are provided in Extended Data Figure 1-1.
CSDS alters the intrinsic properties of projection-defined LHb neurons
We next examined whether CSDS induced changes in the intrinsic properties of LHb neurons defined by their projection outputs. Fluorescent retrograde beads were injected either in the DRN, VTA, or RMTg, and whole-cell patch-clamp recordings were performed from retrogradely labeled LHb neurons (Fig. 4A,I,Q). At the DRN-projecting LHb neurons, CSDS significantly decreased resting membrane potential in SUS mice (Fig. 4B). To examine the intrinsic excitability of DRN-projecting LHb neurons, we have measured the neuronal firing frequency induced by depolarizing current steps (Fig. 4C). No difference was found in the firing frequency. We then analyzed the electrophysiological properties of the first action potential evoked by each depolarizing current step (Fig. 4D–H). We observed a significant increase in rise slope and decrease in post train after-hyperpolarization potential (AHP) in SUS mice and increased half-width in RES mice. At the VTA-projecting LHb neurons, we observed a significant difference in membrane resting potential between SUS and RES (Fig. 4J). Current injection in VTA-projecting LHb neurons significantly decreased the number of elicited action potentials induced by current injection in SUS mice and significantly reduced the action potential amplitude in RES mice (Fig. 4K,M). Finally, specifically recordings from RMTg-projecting LHb neurons revealed a significant reduction in the amplitude of the first action potential in SUS mice while all other parameters were unaltered by CSDS (Fig. 4R–X). Taken together, these results show that CSDS induced specific alterations in the intrinsic properties of LHb neurons based on their projection specificity.
CSDS alters the intrinsic properties of projection-defined LHb neurons. A, Schematic representation of the experimental timeline for the DRN-projecting LHb neurons. B, Resting membrane potential for CTRL, RES, and SUS groups at the DRN-projecting LHb neurons (F(2,40) = 4.62, p = 0.015). C, Representative traces of action potential responses for CTRL, RES, and SUS during current injection and the number of action potentials elicited by different current steps (F(2,41) = 1.22, p = 0.31). D, Example traces of the first action potential for CTRL (gray), RES (blue), and SUS (red). Scale bars: 5 ms and 20 mV. First action potential amplitude (E; F(2,37) = 0.07, p = 0.93), half-width (F; F(2,36) = 5.21, p = 0.01), rise slope (G; F(2,37) = 3.476, p = 0.04), and the post train after-hyperpolarization (AHP; H; F(2,35) = 4.358, p = 0.02; number of mice: 15, 15, and 13 for CTRL, SUS, and RES respectively). I–P, Same convention as for A–H but for the VTA-projecting LHb neurons. J, Resting membrane potential (F(2,56) = 3.274, p = 0.045). K, Representative traces of action potential responses for CTRL, RES, and SUS during current injection and the number of action potential elicited by different current steps (F(2,53) = 6.425, p = 0.003). L, Example traces of the first action potential for CTRL (gray), RES (blue), and SUS (red). Scale bars: 5 ms and 20 mV. First action potential amplitude (M; F(2,51) = 4.696, p = 0.0134), half-width (N; F(2,51) = 0.939, p = 0.398), rise slope (O; F(2,51) = 2.203, p = 0.121), and the post train after-hyperpolarization (AHP; P; F(2,56) = 0.716, p = 0.493; number of mice: 18, 22, and 16 for CTRL, SUS, and RES, respectively). Q–X, Same convention as for A–H, but for RMTg-projecting LHb neurons. R, Resting membrane potential (F(2,61) = 0.708, p = 0.497). S, Representative traces of action potential responses for CTRL, RES, and SUS during current injection, and the number of action potential elicited by different current steps (F(2,59) = 0.218, p = 0.805). T, Example traces of the first action potential for CTRL (gray), RES (blue), and SUS (red). Scale bars: 5 ms and 20 mV. First action potential amplitude (U; F(2,60) = 6.591, p = 0.003), half-width (V; F(2,60) = 0.539, p = 0.586), rise slope (W; F(2,60) = 2.236, p = 0.106), and the post train after-hyperpolarization (AHP; X; F(2,56) = 0.385, p = 0.683; number of mice: 23, 15, and 26 for CTRL, SUS, and RES, respectively). *p < 0.05, **p < 0.01, ***p < 0.001 between RES or SUS and CTRL, #p < 0.05, ##p < 0.01 between RES and SUS. Raw data are provided in Extended Data Figures 2-1–2-3.
Figure 2-1
Tables for raw data presented in Figures 4 for the DRN-projecting LHb neurons. Download Figure 2-1, XLSX file.
Figure 2-2
Tables for raw data presented in Figures 4 for the VTA-projecting LHb neurons. Download Figure 2-2, XLSX file.
Figure 2-3
Tables for raw data presented in Figures 4 for the RMTg-projecting LHb neurons. Download Figure 2-3, XLSX file.
Presynaptic transmission at the projection-defined LHb neurons
Finally, we have characterized whether CSDS had an impact on presynaptic excitatory and inhibitory transmission at projection-defined LHb neurons. LHb neurons were retrogradely labeled as detailed in Figure 4, and spontaneous excitatory postsynaptic current (sEPSC) and inhibitory postsynaptic current (sIPSC) were measured on LHb neurons specifically projecting to the DRN, VTA, or RMTg (Fig. 5A,G,M). No significant effects of CSDS were observed at DRN-projecting LHb for frequency and amplitude of sEPSC (Fig. 5B,C) and sIPSC (Fig. 5D,E) nor any changes in the excitatory–inhibitory balance when looking at the sEPSC/sIPSC frequency ratio and amplitude ratio (Fig. 5F). Similarly, no significant effects were noticed at the VTA-projecting LHb neurons, except observing an almost significant increase in the excitatory balance in the frequency sEPSC/sIPSC ratio between CTRL and SUS mice (Fig. 5H–L). Given that VTA-projecting LHb neurons have previously been shown to exhibit increased excitatory transmission in the learned helplessness model (B. Li et al., 2011), we have recorded spontaneous activity at VTA-projecting LHb neurons in control mice and mice subjected to 60 min of inescapable footshock stress (IS). Consistent with previous findings, we have found that inescapable foot shocks significantly increased the ratio of sEPSC/sIPSC frequency (CTRL 1.145 ± 0.19, IS 2.485 ± 0.38 p < 0.01, unpaired t test), with no change for the sEPSC/sIPSC amplitude (CTRL 0.886 ± 0.08, IS 1.205 ± 0.25 p > 0.05, unpaired t test). At the RMTg-projecting LHb neurons, while CSDS had no significant effect on sEPSC amplitude, it had a significant effect on sEPSC frequency. However, post hoc analysis did not reveal significant changes between CTRL and SUS or RES mice (Fig. 5N,O). Finally, CSDS had a significant effect on the amplitude of sIPSC with a significant increase between CTRL and SUS mice, which translated into a significant decrease in the amplitude of excitatory–inhibitory balance for SUS compared with CTRL mice (Fig. 5Q–R). As observed before, CSDS differently impacted presynaptic transmission at the projection-defined LHb neurons.
Excitatory and inhibitory inputs to projection-defined LHb neurons are altered by CSDS. A, Characterization of excitatory and inhibitory synaptic inputs onto LHb neurons projecting to the DRN. Representative traces for sEPSC (B) and sIPSC (D) for CTRL, RES, and SUS mice and average plots for sEPSC [C; frequency (F(2,17) = 0.46, p > 0.05) and amplitude (F(2,17) = 0.65, p > 0.05)] and sIPSC [E; frequency (F(2,15) = 0.30, p > 0.05) and amplitude (F(2,15) = 0.01, p > 0.05)]. F, Plots for the excitatory–inhibitory ratio for frequency (F(2,14) = 0.04, p > 0.05) and amplitude (F(2,14) = 0.09, p > 0.05; number of mice: 5, 8, and 7 for CTRL, SUS, and RES, respectively). G–L, Same convention as for A–F for VTA-projecting LHb neurons. Representative traces for sEPSC (H) and sIPSC (J) for CTRL, RES, and SUS mice and average plots for sEPSC [I; frequency (F(2,8) = 0.75, p > 0.05) and amplitude (F(2,8) = 0.26, p > 0.05)] and sIPSC [K; frequency (F(2,8) = 0.77, p > 0.05) and amplitude (F(2,8) = 3.24, p > 0.05)]. L, Plots for the excitatory–inhibitory ratio for frequency (F(2,8) = 3.06, p > 0.05) and amplitude (F(2,8) = 0.01, p > 0.05; number of mice: 4, 5, and 3 for CTRL, SUS, and RES, respectively). M–R, Same convention as for A–F for RMTg-projecting LHb neurons. Representative traces for sEPSC (N) and sIPSC (P) for CTRL, RES, and SUS mice and average plots for sEPSC [O; frequency (F(2,12) = 5.67, p < 0.05) and amplitude (F(2,12) = 0.002, p > 0.05)] and sIPSC [Q; frequency (F(2,12) = 0.04, p > 0.05) and amplitude (F(2,12) = 7.45, p < 0.01)]. R, Plots for the excitatory–inhibitory ratio for frequency (F(2,12) = 2.71, p > 0.05) and amplitude (F(2,12) = 5.25, p < 0.05; number of mice: 5, 6, and 4 for CTRL, SUS, and RES, respectively). The vertical and horizontal bars are 15 pA and 1 s respectively. *p < 0.05. Raw data are provided in Extended Data Figures 3-1–3-3.
Figure 3-1
Tables for raw data presented in Figures 4 for the DRN-projecting LHb neurons. Download Figure 3-1, XLSX file.
Figure 3-2
Tables for raw data presented in Figures 4 for the VTA-projecting LHb neurons. Download Figure 3-2, XLSX file.
Figure 3-3
Tables for raw data presented in Figures 4 for the RMTg-projecting LHb neurons. Download Figure 3-3, XLSX file.
Discussion
The LHb is a brain nucleus that plays a central role in the processing of aversive signals, also known to be hyperactive in depressive disorders (Proulx et al., 2014). In recent years, multiple studies have revealed how LHb neurons adapt to chronic stress and how synaptic inputs onto LHb neurons become hyperactive in animal models of depression. Still, limited data exist on the consequences of stress on transmission at the LHb neuronal outputs, which directly and indirectly control serotonin and dopamine centers. Here, we show that CSDS, an important animal model of clinical depression, induces distinct adaptations at the LHb neuronal outputs innervating the DRN, VTA, and RMTg, three major LHb targets. Since DRN, VTA, and RMTg are innervated by distinct LHb neuronal subpopulations (Gonçalves et al., 2012; Cerniauskas et al., 2019), these results suggest that stress-induced adaptations at these major LHb neuronal outputs are likely to mediate distinct functions, either in the etiology or the maintenance of depressive and cognitive dysfunctions found in depression. However, since SUS and RES mice were categorized solely based on their social interaction profile, future studies will be required to directly address how the synaptic adaptations revealed here may impact other depressive-like behaviors.
The DRN receives direct excitatory inputs from the LHb (Zhou et al., 2017). Optogenetic stimulation of LHb terminals in the DRN suggests that aversive signals are encoded at the LHb-DRN pathway (Cerniauskas et al., 2019). Aversive events are strong activators of DRN5HT neurons, and at least some of the aversive signals originate from LHb transmission. Particularly, uncontrollable stress activates DRN5HT neurons leading to the release of 5-HT in the DRN and projection regions such as the amygdala, which is required to induce the behavioral deficits induced by uncontrollable stress (Amat et al., 2001; Dolzani et al., 2016). Importantly, lesion or inhibition of the habenula is sufficient to abolish the activation of DRN5HT neurons and to prevent the behavioral consequences of uncontrollable stress (Amat et al., 2001; Dolzani et al., 2016). These findings show that uncontrollable stress induces an exaggerated activation of the LHb-DRN pathway, which is an important underlying mechanism in the etiology of depressive deficits found in the learned helplessness. We have found that in SUS mice, CSDS induces a synaptic potentiation in a subpopulation of synapses at the LHb-DRN pathway. It is then possible that excessive activity at the LHb-DRN pathway during social defeat in SUS mice promotes synaptic adaptation at these synapses. The fact that these synapses are found in a specific subregion of the DRN may further suggest that increased transmission occurs in DRN neurons projecting to a specific target region such as the amygdala. In support of this hypothesis, it has been shown that uncontrollable stress induces an intense activation of DRN5HT causing a sensitization of these neurons and leading to exaggerated responses to subsequent stressors, which is reflected by increased release of 5-HT in projection regions involved in stress-induced behavioral deficits (Rozeske et al., 2011). This could partially be mediated by a facilitated transmission at the LHb-DRN synapses.
In parallel, the LHb also makes synaptic contact with dopaminergic and GABAergic neurons in the VTA (Ogawa et al., 2014; Faget et al., 2016). Optogenetic activation of VTA-projecting LHb neurons also encodes an aversive signal, which could be mediated by the release of dopamine in the PFC (Lammel et al., 2012). Here, we have found that CSDS depressed synaptic transmission at the LHb-VTA pathway in SUS mice. Moreover, VTA-projecting LHb neurons show reduced excitability in SUS mice, reflected by a significantly reduced firing rate induced by current injection. Altogether, these results suggest that signal transmission at the LHb-VTA pathway is less efficient in SUS mice following CSDS. Future experiments will be required to understand the behavioral consequences of a hypofunctional transmission at the LHb-VTA pathway. The VTA and the LHb play important roles in associative learning and defensive behaviors (Brown et al., 2012; Lecca et al., 2017; Trusel et al., 2019; Barbano et al., 2020; Root et al., 2020). It is then possible that altered transmission at the LHb-VTA pathway may impair associative learning and the expression of defensive behaviors observed after chronic stress (Giovanniello et al., 2023).
The RMTg-projecting LHb neurons encode negative reward prediction errors: they are excited by unexpected nonrewarding or unpleasant events and are inhibited by unexpected rewarding events (Matsumoto and Hikosaka, 2007, 2009). Transmission from the LHb to the GABAergic RMTg leads to a robust inhibition of dopaminergic neurons in the VTA (Hong et al., 2011). Through this mechanism, the LHb neurons encode motivational values and prediction errors in an opposite manner compared with DA neurons in the VTA (Matsumoto and Hikosaka, 2007), which would control motivated responses where unrewarded or aversive events discourage action while rewarded or appetitive events reinforce action (Matsumoto and Hikosaka, 2009, 2011; Baker et al., 2016). In line with this hypothesis, we have recently reported that activity at the LHb-RMTg pathway controls the effort to either escape an aversive context or to acquire rewards, supporting its role in gating motivation to exert effort (Proulx et al., 2018). Consequently, alteration in this pathway may play a role in amotivational states observed in depression (Liu et al., 2017). Here, we are showing that CSDS significantly impacted presynaptic transmission at the LHb terminals synapsing onto RMTg neurons, both in RES and SUS mice. Decreased PPR can be an indication of an increased release probability, which could be reflected by a facilitated or increased spontaneous activity at the LHb-RMTg, leading to an exaggerated inhibition of dopaminergic activity and reduced motivation (Sun et al., 2020). RMTg-projecting LHb receives inputs from the globus pallidus internal segment (GPh) of the basal ganglia (Hong et al., 2011). Our previous findings have shown that these inputs transmit aversive signal (Shabel et al., 2012) and are hyperactive in animal models of depression, which could be reversed by antidepressants (Shabel et al., 2014). Altogether, these findings suggest that the circuit GPh-LHb-RMTg-VTA is altered by chronic stress resulting in an exaggerated inhibition of dopamine activity, which could reduce motivation (Liu et al., 2017).
RMTg neurons receiving inputs from the LHb also send inhibitory projections to the DRN. Hyperactivity at the LHb-RMTg could also result in constitutive inhibition of DRN5HT activity. Consistent with this hypothesis is the finding that LHb lesion, in chronically stressed rats, was sufficient to increase serotonin level in the DRN, this was correlated with a decreased immobility in the forced swim test, a reflection of increased motivation to swim in an aversive context (L-M. Yang et al., 2007; Proulx et al., 2018).
Using a chronic variable stress model (CMS), Cerniauskas et al. have examined synaptic adaptation at LHb neurons projecting to the VTA and the DRN. CMS increased excitability specifically at the VTA-projecting LHb neurons. CMS increased current-evoked action potentials at these neurons. They have also observed an increased release probability at the GPh [entopeduncular nucleus (EP) in rodents] terminals synapsing onto VTA-projecting LHb neurons. We suspect that some of these VTA-projecting LHb neurons are projecting to the RMTg and would support the hypothesis, together with our results, that chronic stress increased transmission at the GPh-LHb-RMTg, which reduced motivation to exert effort (Proulx et al., 2018; Cerniauskas et al., 2019). Surprisingly, CMS failed to induce significant synaptic adaptation at the DRN-projecting LHb neurons. The circuit adaptation induced by the CMS (Cerniauskas et al., 2019) and CSDS (this study) at the LHb circuits illustrates the importance of examining how different chronic stress paradigms can impact functional outputs at specific pathways and circuits, which may underly specific dysfunctions found in major depressions.
The present study also demonstrates that LHb neurons display pathway-specific changes in their electrophysiological properties induced by CSDS. In DRN-projecting LHb neurons, CSDS significantly hyperpolarized the resting membrane potential in SUS mice. Previous studies have demonstrated that increased bursting activity in LHb neurons contributes to depression-like behaviors (Y. Yang et al., 2018; Cerniauskas et al., 2019). This increased bursting is associated with a hyperpolarized resting membrane potential. However, it remains to be directly explored whether CSDS induces heightened burst activity specifically in LHb neurons projecting to the DRN. The excitability of DRN-projecting LHb neurons measured by the number of action potentials elicited by current injection was not changed, but we observed significant differences in rise slope and AHP amplitude in susceptible mice and increased half-width in RES mice. In VTA-projecting LHb neurons, a significant decrease in membrane resting potential in SUS mice compared with RES mice was observed, which could explain the significant reduction in the action potential amplitude. Notably, the excitability of VTA-projecting LHb neurons was significantly reduced in SUS mice. Together with a significant decrease in the AMPAr/NMDAr ratio observed at the LHb-VTA synapses, these results suggest that CSDS induces an overall decreased transmission in this specific LHb pathway. Finally, RMTg-projecting LHb neurons showed a significant decrease in the AP amplitude in SUS mice while other parameters were unaltered.
These results suggest that CSDS induces distinct changes in the expression or functionalities of ion channels responsible for the electrophysiological properties of LHb neurons resulting in altered neuronal functions within the LHb in a pathway-specific manner. Voltage-gated sodium and potassium channels as well as calcium-activated BK channels play major roles in neuronal AP initiation and repolarization (Zhang et al., 2003; Kress and Mennerick, 2009; Kim and Nimigean, 2016) influencing action potential rise slope, amplitude, and AP duration in mammalian central nervous system. Resting membrane potential is primarily dependent on potassium conductance (Wright, 2004). Transcriptional analysis of gene expression in projection-defined LHb neuronal outputs previously revealed altered gene expression induced by stress (Cerniauskas et al., 2019; Levinstein et al., 2020). Particularly, a change in gene expression for potassium channels (Kcnc1 and Kcnc2; Cerniauskas et al., 2019) might explain the differences in AP amplitude and half-width. Understanding the precise mechanisms and consequences of ion channel dysregulations in the LHb following CSDS may offer insights into potential therapeutic interventions for stress-related psychiatric disorders. Intersectional viral approaches and single-cell RNA sequencing will be required to reveal the exact mechanisms underlying the changes in firing properties induced by CSDS (Bittar et al., 2021). Future experiments will be required to directly assess the physiological consequences of these changes in depressive-like behaviors.
Our results finally showed an imbalance in the excitatory–inhibitory inputs to the RMTg-projecting LHb neurons in SUS mice suggesting that CSDS may also alter presynaptic transmission at LHb neurons projecting to RMTg. While CSDS did not induce significant changes in presynaptic transmission at the VTA-projecting LHb neurons, inescapable footshock stress significantly increased the excitatory–inhibitory balance, suggesting that this pathway might be sensitive to the type of stressors. LHb neurons projecting to DRN, VTA, and RMTg receive inputs from many brain regions including the ventral pallidum, the lateral hypothalamus, the prefrontal cortex, the VTA, and the entopeduncular nucleus. EPN terminals preferentially synapse onto RMTg-projecting LHb neurons while DRN-projecting LHb neurons receive significantly more inputs from the VTA (Cerniauskas et al., 2019; H. Li et al., 2019). The roles of synaptic inputs to the LHb from the lateral hypothalamus and the ventral pallidum in promoting behavioral deficits found in CSDS have previously been shown (Knowland et al., 2017; Wang et al., 2021). Future studies will be necessary to fully understand synaptic adaptations at the LHb circuitry in depressive-like behaviors.
Altogether, we have shown that chronic social stress–induced maladaptive changes at the main LHb neuronal outputs. Considering the important role of the LHb in depressive disorders, how these stress-induced adaptations may impact cognitive and behavioral dysfunctions in depression will be important to open new avenues for more selective and efficient therapeutical approaches.
Inclusion and Diversity
We support inclusive, diverse, and equitable conduct of research.
Footnotes
We thank Dr. Joaquin Piriz for his critical review of the manuscript and Maryse Pinel for her technical assistance. We also thank the CERVO Canadian Optogenetics and Vectorology Foundry Core Facility for producing the viral vectors. C.D.P. is supported by the Canadian Institutes of Health Research Grant PJT169117 and the Natural Science and Engineering Research Council of Canada Grant RGPIN-2017-06131 and received Fonds de Recherche en Santé du Québec (FRQS) Junior-2 salary support. J.C.H.S. was supported by a merit scholarship program for foreign students from the FRQS and a doctoral scholarship from CONACYT.
The authors declare no competing financial interests.
- Correspondence should be addressed to Christophe D. Proulx at christophe.proulx{at}fmed.ulaval.ca.
