Hypoactive thalamic Crh+ cells in a female mouse model of alcohol drinking after social trauma

Abstract

Background:

Comorbid stress-induced mood and alcohol use disorders are increasingly prevalent among female patients. Stress exposure can disrupt salience processing and goal-directed decision making, contributing to persistent maladaptive behavioral patterns; these and other stress-sensitive cognitive and behavioral processes rely on dynamic and coordinated signaling by midline and intralaminar thalamic nuclei. Considering the role of social trauma in the trajectory of these debilitating psychopathologies, identifying vulnerable thalamic cells may provide guidance for targeting persistent stress-induced symptoms.

Methods:

A novel behavioral protocol traced the progression from social trauma to the development of social defensiveness and chronically escalated alcohol consumption in female mice. Recent cell activation – measured as cFos - was quantified in thalamic cells after safe social interactions, revealing stress-sensitive corticotropin releasing hormone-expressing (Crh+) anterior central medial thalamic (aCMT) cells. These cells were optogenetically stimulated during stress-induced social defensiveness and abstinence-escalated binge drinking.

Results:

Crh+ aCMT neurons exhibited substantial activation after social interactions in stress-naïve but not stressed female mice. Photoactivating Crh+ aCMT cells dampened stress-induced social deficits whereas inhibiting these cells increased social defensiveness in stress-naïve mice. Optogenetically activating Crh+ aCMT cells diminished abstinence-escalated binge alcohol drinking in female mice, regardless of stress history.

Conclusions:

This work uncovers a role for Crh+ aCMT neurons in maladaptive stress-induced social interactions and in binge drinking after forced abstinence in female mice. This molecularly-defined thalamic cell population may serve as a critical stress-sensitive hub for social deficits caused by exposure to social trauma and for patterns of excessive alcohol drinking in female populations.

Introduction

The debilitating repercussions of stress can manifest in distinct psychiatric disease trajectories, often according to gender; while mood disorders are more prevalent among women, alcohol use disorders (AUD) are more frequent in men (1–20). Yet, the gender gap in AUD prevalence is steadily closing as diagnostic rates increase in women (9, 21–24). Probing the etiology of this rise in AUDs reveals that mood disorders commonly precede the initiation of alcohol abuse (6, 7, 25). With a growing population of women diagnosed with comorbid mood and alcohol use disorders (26), translational animal models can serve as an important tool for examining the neural and behavioral maladaptations that contribute to stress-induced defensiveness, hypervigilance, and increased drinking in female subjects (27, 28).

Experiencing social trauma can result in the development of posttraumatic stress disorder, characterized by impaired fear inhibition in a safe context (29, 30). To model social trauma, translational mouse protocols employ chronic social defeat stress (CSDS) to forge a learned association between social cues and a pending attack. Repeated experiences with aggressive, dominant conspecifics lead to the rapid development of distinct defensive tactics accompanied by molecular and physiological disruptions that mirror some of the consequences of social trauma in humans (31–46). In female mice, CSDS yields a prominent defensive behavioral profile that persists for weeks after stress exposure, even during safe social interactions with a non-aggressive partner in a familiar non-threatening environment (46). Here, CSDS yields social defensiveness and increases alcohol consumption in female mice, modeling several cardinal symptoms of mood and alcohol use disorders in women who experience social trauma.

Considering its prominent involvement in hypothalamic and extrahypothalamic stress response mechanisms, corticotropin-releasing hormone (CRH; 47–52) has been examined extensively in translational preclinical experiments that model stress-induced psychiatric symptoms (42, 44, 45, 53–59). Cells that transcribe Crh - the gene encoding CRF - have been identified in the paraventricular thalamus (PVT) and the rostral intralaminar nuclei (rILN), which include the anterior central medial (aCMT), paracentral (PC), and central lateral thalamus (CL; 60–62). The PVT and rILN receive brainstem inputs from the ascending reticular activating system (63, 64) and provide both direct inputs to the central and basolateral amygdala (BLA), medial prefrontal cortex (mPFC), and nucleus accumbens (NAc) as well as disynaptic accumbal inputs via the BLA or mPFC; many of these pathways are bidirectional, forming information processing loops (65–79). As such, PVT and rILN cells receiving arousal-related signals may functionally communicate salience to limbic targets that process emotional and reward-related information (80–87). Growing evidence implicates PVT/rILN cell populations and extended amygdala CRH signaling in reward processing and alcohol drinking (88); however, the functional role of thalamic Crh+ neurons remains unclear. Because midline and intralaminar thalamic nuclei are instrumental in maintaining adaptive cognitive and behavioral processes that are derailed by stress-induced mood and drug use disorders (89–100), we hypothesize that thalamic Crh+ neurons contribute to the development of persistent stress-induced social deficits and dysregulated alcohol drinking in female mice.

Here, a translational model of female CSDS (46) is used to identify a novel population of aCMT neurons with suppressed cFos activation during safe social interactions in mice with a stress history. We hypothesized that a subpopulation of cells in the aCMT – those expressing Crh – may be specifically sensitive to social defeat stress and may be responsible for some of the persistent maladaptive behavioral repercussions of social defeat stress in female mice. After finding that Crh+ aCMT neurons are highly stress-sensitive, we optogenetically targeted these cells to reveal that: 1.) activation of Crh+ aCMT cells reverses stress-induced social avoidance and defensiveness, 2.) inhibition of these cells promotes social defensiveness in stress-naïve mice, 3.) prior social stress exposure increases chronic alcohol consumption in female mice, and 4.) that optogenetic activation of Crh+ aCMT neurons blunts forced abstinence-escalated alcohol consumption in control and stressed animals. In sum, the present work describes a novel stress-sensitive population of Crh+ aCMT neurons that was identified and characterized using an experimental approach that closely models the disease trajectory from social trauma to dysregulated sociability and alcohol drinking.

Methods and Materials

Animals:

Twelve-week-old Swiss Webster (CFW) female mice (Charles River Laboratories, Wilmington, MA, USA) were pair-housed with castrated CFW males (n=78 pairs). Six-to-eight-week-old C57BL/6J (B6; Jackson Laboratories, Bar Harbor, ME) females were group-housed (n=10/26×48×15 cm cage) to serve as stimuli for either sociability or aggression tests (n=100). Experimental females included 10-12-week-old B6, CRH-ires-Cre, or CRH-ires-Cre/Ai9 reporter mice (100, 101). Animals were cared for according to the NIH Guide for the Care and Use of Laboratory Animals and procedures were approved by the Institutional Animal Care and Use Committee of Tufts University.

Home cage sociability testing:

Single-housed experimental mice were tested repeatedly for sociability toward non-aggressive B6 female mice during 1.5-minute home cage sociability tests (46). Social interaction time was quantified as the duration of social contact initiated by the experimental female. Defensiveness was quantified as total time spent escaping, kicking, avoiding, jumping, and displaying hypervigilance or social risk assessment, operationalized as time spent orienting toward but physically avoiding the social stimulus (46, 102, 103). Baseline social interaction times were used to assign experimental mice into counterbalanced defeat and control conditions. All social interactions were videotaped under dim red light and social and defensive behaviors were scored manually with Observer XT software (Noldus, Leesburg, VA, USA).

Female chronic social defeat stress:

As described previously (46), CFW females were tested for aggression toward unfamiliar group-housed B6 female intruder mice daily during 2-minute encounters. CFW females that bit >15 times during >3 consecutive tests were used as residents for 10-day CSDS. Twenty-four-hours pre-CSDS, aggressive CFW resident females and their male cage mates were housed in a cage (26x48x15 cm) divided in half by a perforated plastic partition (35, 46). Daily, CFW males were moved to holding cages before an unfamiliar experimental female was introduced into the territory of the aggressive resident female. Following the 5-minute defeat, the experimental female was housed adjacent to the resident aggressor, protected from further attack by the partition. Male-female CFW pairs were reunited between daily defeats to preserve female-directed aggression in resident mice. Daily for 10 days, each defeated experimental mouse was attacked by an unfamiliar CFW female and housed adjacent to that resident until the subsequent defeat. Non-defeated experimental control mice were housed adjacent to an unfamiliar CFW pair daily but were never defeated. Control and defeated mice were individually housed twenty minutes after the final defeat.

Continuous Access to Alcohol:

Mice received access to 20% ethyl alcohol (EtOH) and water as described previously (42, 43, 45). Alcohol solutions (w/v) were made weekly by diluting 95% EtOH in tap water (PHARMCO-AAPER, Brookfield, CT, USA), and presented daily on alternating sides of the cage lid. Bottles were weighed three hours into the dark photoperiod (1030h); fluid evaporation and spillage were controlled for by recording EtOH and water bottle weights from an empty cage and subtracting these values from daily experimental measurements. Body weights were recorded weekly to calculate grams of EtOH consumed per kilogram of body weight. Compared to intermittent access, continuous alcohol access yields moderate consumption (43), thereby allowing detection of bidirectional changes in intake.

Experiment 1: Alcohol consumption in control and defeated female mice:

Prior studies identify social stress as a powerful tool for modeling stress-escalated alcohol consumption in male mice (42–45). We employed CSDS in female mice to develop a novel model of social stress-escalated alcohol consumption. Beginning one week after their arrival, singly housed B6 females were habituated to vaginal lavage (104) for daily estrous cycle phase determination throughout the experiment. Ten days post-CSDS, mice received continuous access to alcohol for four weeks. The effects of forced alcohol abstinence were examined after the fourth week by providing mice with only water for 24-hours prior to alcohol reintroduction. After two hours of alcohol access, blood was collected from the submandibular vein and plasma was extracted to quantify blood ethanol concentration (BEC). Plasma samples (5 μL) were run in duplicate (AM1 Analyzer; Analox Instruments Ltd., Stourbridge, UK).

Experiment 2: aCMT cFos after social interactions in control or defeated female mice:

Female B6 mice were individually housed for one week. Sociability testing occurred twenty-four hours before and after the CSDS protocol. Mice were deeply anesthetized 1-hour after the second sociability test and transcardially perfused with 4% paraformaldehyde in 1xphosphate-buffered saline (PBS). Brain sliced containing the aCMT and BNST were selected for cFos immunohistochemistry (see Supplement).

Experiment 3: cFos in Crh+ aCMT cells after social interactions in control or defeated female mice

To examine if effects of CSDS on cFos activation were specific to Crh+ aCMT cells, we replicated and extended Experiment 2 in female CRH-ires-Cre/Ai9 mice that expressed endogenous tdTomato in Crh+ neurons.

Experiment 4: Optogenetic interrogation of Crh+ aCMT cells during pre- and post-CSDS social interactions and abstinence-induced binge alcohol drinking:

Female CRH-ires-Cre mice were infected with of adeno-associated virus (1-8x10¹² vg/mL; UNC Vector Core, Chapel Hill, NC, USA) for Credependent expression of excitatory channelrhodopsin (ChR2; AAV2-EF1a-DIO-hChR2(H134R)-mCherry-WPRE-pA) (105) or inhibitory halorhodopsin (NpHR; AAV2-EF1a-DIO-eNpHR3.0-mCherry) (106) in Crh+ aCMT cells (AP:−0.95; ML:0; DV:3.60). Optic fibers were implanted over the aCMT (16°; AP:−0.95, ML:1.00; DV:3.75).

Sociability during optogenetic stimulation of Crh+ aCMT neurons:

To allow mice a full range of motion during sociability tests, implanted optic fiber ferrules were attached to an optic fiber patch cord that connected to a rotary joint. This rotary joint was fitted to an overhead gimbal holder and a second patch cord connected the joint to the laser. Light pulses were controlled by a pulse generator (A.M.P.I, Jerusalem, Israel), and were administered for 2 minutes beginning 15 seconds prior to the 1.5-minute sociability tests. For pre-CSDS ChR2 stimulations, 50, 10, 5 or 1 ms 473 nm pulses were delivered at 2, 10, 20, or 40 Hz, respectively (107, 108). Pre-CSDS inhibitory 561 nm pulses were delivered to NpHR-expressing mice at frequencies and pulse widths of 2 Hz/50 ms, 2 Hz/125 ms or 20 Hz/5 ms (109, 110). Pre-CSDS laser control trials were conducted to determine if behavioral effects were induced by laser light alone; specifically, ChR2 mice received 561 nm pulses (40 Hz, 1 ms) while NpHR mice received 473 nm pulses (20 Hz, 5 ms). To mitigate wavelength-specific heat effects, power intensity was maintained at 0.5-2 mW and stimulations were restricted to <5 minutes (111). At the conclusion of Experiment 4, cFos was quantified after active or control laser pulse deliveries to examine the extent to which opsins might be affected by the control wavelength.

Pre-CSDS effects of laser pulse deliveries guided the selection of post-CSDS stimulation conditions; NpHR mice received 20 Hz stimulations whereas ChR2 mice received 40 Hz stimulations during post-CSDS sociability tests and alcohol drinking sessions after forced abstinence. Pre-CSDS sociability tests were conducted twice weekly for three weeks; post-CSDS tests occurred in the week following the final social defeat. All tests were conducted at least 48 hours apart and >24 hours after cages were cleaned.

Optogenetic stimulation of Crh+ aCMT cells during alcohol drinking:

Three days after their final sociability test, mice received continuous two-bottle choice access to water and 20% ethyl alcohol (w/v; EtOH). In the third week of alcohol access, home cages were fitted with custom-made stainless-steel lickometer panels, allowing alcohol and water sipper tubes to be presented through holes in the panel. Stainless steel mesh flooring was secured to the bottom of each panel to form a raised platform within each home cage. To drink, mice stood on the mesh platform and made tongue contact with the metal sipper tube. Each contact completed a circuit which registered as an event relayed via the lickometer controller to an interface (Med Associates, St. Albans, VT, USA) (112). Licks were recorded each time a closed circuit was detected with a minimum interlick interval sensitivity of 1-ms.

Mice received two days of continuous alcohol and water access during habituation to the lickometer panel (lickometer days 1 and 2). Three hours into the dark photoperiod on day 3, alcohol was replaced with a second water bottle and was subsequently reintroduced 24-hours later (day 4) to induce abstinence-induced alcohol binge drinking. Concurrent with alcohol reintroduction on day 4, NpHR-expressing females received 20 Hz 561 nm light pulses and ChR2-expressing mice received 40 Hz 473 nm pulses. Laser stimulations occurred over 2-hour sessions with 8×15-minute trials consisting of 5-minute laser ON, 10-minute laser OFF. This stimulation schedule was designed to detect: 1.) the effect of optogenetic stimulations on deprivation-induced binge drinking during the first five minutes after alcohol reintroduction, and 2.) potential changes in intake coinciding with laser offset. From day 5-11, animals received continuous access to alcohol and water and were subsequently alcohol-deprived on day 12. Alcohol was reintroduced on day 13 to quantify post-deprivation licking without optogenetic stimulations.

Effects of laser light pulses on aCMT cFos:

ChR2- or NpHR-expressing mice received stimulations in the home cage with or without concurrent sociability tests, respectively. One hour later, mice were transcardially perfused and tissue was collected to determine if ChR2 activation (473 nm, 40 Hz) induced cFos expression compared to control laser pulses (561 nm, 40 Hz) and to test if NpHR stimulation (561 nm, 20 Hz) reduced social cFos activation compared to control laser stimulations (473 nm, 20 Hz).

Statistical analyses:

The D’Agostino-Pearson omnibus K2 test for normality was conducted on raw or – for data sets containing zeroes - on y=(y+1) transformed data. When necessary, a logarithmic or square root transformation was applied to achieve normality before performing parametric statistical analyses; figures depict back-transformed data. The Geisser-Greenhouse correction was applied to repeated-measures (RM) and mixed analyses of variance (ANOVA). F tests of homoscedasticity preceded unpaired t tests of between-subject data sets with a single-level factor. For within-subject and mixed designs with multi-level factors, RM one-way, RM two-way ANOVA or mixed two-way ANOVA identified significant interactions and main effects. Significant omnibus tests were followed by Dunnett’s post-hoc comparisons between baseline and factor levels. Statistical analyses were conducted in GraphPad Prism v.8 (GraphPad Software, San Diego, CA, USA).

See Supplement for additional details and experimental group numbers (Table S1).

Results

Experiment 1: Social defeat stress-increased chronic alcohol consumption and binge drinking after forced abstinence in female mice:

Chronically socially defeated wild-type B6 females consumed more alcohol compared to non-defeated controls (Fig. 1). Tracking estrous cycling through vaginal cytology revealed that neither stress nor alcohol intake resulted in protracted acyclicity (Fig. S1). Control and defeated mice achieved BECs >80 mg/dL in the two hours of alcohol access following 24-hour forced abstinence (Fig. 1), modeling a pattern of alcohol intake associated with high-risk binge drinking in women (113, 114).

Figure 1. Social stress and forced abstinence increase alcohol drinking in female mice.

(A) Wild-type C57BL/6J female mice were exposure to 10-day chronic social defeat stress (CSDS) or the non-defeated control condition (n=10/condition). Ten days after the final day of CSDS, mice received four weeks of continuous access to 20% ethyl alcohol (EtOH (w/v)) and water in the home cage. Alcohol was removed for 24-hours and reintroduced for 2-hours; submandibular blood was collected to determine blood EtOH concentration (BEC). Estrous cycle phase was tracked throughout (Fig. S1). (B) Defeated female mice consistently consumed more alcohol (g/kg/24hr) than control mice in the four weeks of alcohol access [F(1,18)=8.29, p=0.01]; data depicted as two-day group Means ± SEM. (C) Average daily alcohol intake was greater in defeated mice compared to controls [t(18)=3.22, p=0.0048] whereas (D) water intake was similar across conditions. (E) Defeated mice consumed more alcohol during two-hour access after forced abstinence [t(18)=3.065, p=0.0067], (F) producing a trend of increased BEC in defeated vs. control mice (p=0.08). Average 2-hour BECs were > 80 mg/dL (dotted line), indicative of a binge pattern of alcohol intake in control and defeated mice. (C, D) Data points depict four-week averages for individual mice; bars represent four-week group Mean ± SEM. (E, F) Data points are individual mice; bars represent the group Mean ± SEM. * p <0.05, ** p<0.01 control vs. defeated

Experiments 2 and 3: CSDS reduced cFos in Crh+ aCMT cells after safe social interactions:

In B6 female mice, CSDS exposure reduced time spent socially interacting, increased defensive behaviors, and reduced the number of cFos+ cells in the anterior central medial thalamus (aCMT) compared to stress-naive control mice (Fig. S2). Colocalization of cFos and the endogenous fluorescent Crh-reporting protein - tdTomato - (Fig. S3, S4) was quantified in control and defeated CRH-ires-Cre/Ai9 mice after safe social interactions. As in wild-type mice (Fig. S2), CSDS reduced social interactions (Fig. 2A–C) and increased defensive behaviors (Fig. 2D–E) in female CRH-ires-Cre/Ai9 mice. Tissue collected from these animals revealed fewer aCMT cells with colocalized cFos and Crh-reporter and a reduction in the percentage of activated Crh+ aCMT cells in defeated compared to control mice (Fig. 2F–I). CSDS had no effect on cFos activation in Crh+ neurons within the bed nucleus of the stria terminalis (BNST), suggesting a regionally selective effect of CSDS on Crh+ cell populations (Fig. S5).

Figure 2. Chronic social defeat stress blunts social cFos activation in Crh+ aCMT cells.

(A) Female CRH-ires-Cre/Ai9 mice were tested for baseline sociability toward a safe social partner, then exposed to either 10-day chronic social defeat stress (CSDS; n=5 Defeated) or the 10-day control condition (n=5 Control. Sociability was retested the day after 10-day CSDS/control protocol; tissue was collected one hour later to examine colocalization of cFos with the endogenous Crh reporter protein, tdTomato, in anterior central medial thalamic (aCMT) cells. (B, C) CSDS reduced sociability, measured as cumulative nasonasal and anogenital contact time [F(1,8)=5.37, p=0.049] and (D, E) increased defensiveness, measured as total time spent kicking, flinching, and maintaining a vigilant-like pose [F(1,8)=5.56, p=0.046]. (B, D) Yellow asterisks indicate CRH-ires-Cre/Ai9 mice as they engage in representative social and defensive behaviors. (F) Colocalization of cFos and tdTomato in the aCMT was suppressed by CSDS [t(8)=2.87, p=0.021]. (G) A greater percentage of Crh+ aCMT cells were activated by safe social interactions in (H) control mice compared to (I) defeated individuals [t(8)=3.19, p=0.013]; (H, I) representative images taken at 10x magnification of (left panels) cFos (green) and (right panels) cFos colocalization with tdTomato (magenta) and DAPI counterstaining (blue); 200 μm scale bar. Data shown as individual values and as the group Mean ± SEM; *p<0.05 control vs. defeated

Experiment 4: Crh+ aCMT cell activation recovered sociability in chronically defeated female mice:

Female CRH-ires-Cre mice expressing ChR2 (Fig. 3A–B) received optogenetic stimulation to activate Crh+ aCMT neurons (Fig. S6). Activating these cells during sociability tests produced a frequency-dependent increase in defensiveness without affecting social interaction time, burrowing or walking compared to no-laser baseline (BL); control 561 nm light pulses had no effect on social interaction or walking durations (Fig. 3C–E; Fig. S8A). Defeat stress reduced BL social interactions and increased defensiveness compared to controls; activating Crh+ aCMT cells recovered sociability and reduced defensiveness in stressed mice without affecting burrowing (Fig. 3F–H; Video 1).

Figure 3. Optogenetic activation of Crh+ aCMT neurons reverses chronic social defeat stress-induced social deficits whereas optogenetic inhibition produces a defeat-like phenotype in stress-naïve female mice.

(A) Female CRH-ires-Cre mice received intra-aCMT AAV to express Cre-dependent channelrhodopsin (ChR2) or halorhodopsin (NpHR) in Crh+ neurons and were implanted with optic fibers targeting the aCMT. (B) Laser pulses were delivered during home cage safe social interactions to activate (ChR2) or inhibit (NpHR) Crh+ aCMT cells before and after 10-day chronic social defeat stress (CSDS) or the 10-day control procedure (ChR2: n=8 Control, n=10 Defeated; NpHR: n=8 Control, n=8 Defeated). (C) Before 10-day CSDS, neither 473 nm laser pulses nor control 561 nm laser pulses affected social interaction times in ChR2-expressing mice as compared with no-laser baseline (BL) tests. (D) Optogenetic activation of Crh+ aCMT cells produced a frequency-dependent increase in social defensiveness [F(4,68)=6.25, p<0.001] without affecting (E) burrowing. (F-H; BL) Compared with control mice, CSDS suppressed sociability [t(16)=4.35, p<0.001] and increased defensiveness [t(16)=3.70, p=0.002] without affecting burrowing. In defeated mice, Crh+ aCMT cell activation (F) recovered sociability [F(1,16)=11.63, p=0.004], and (G) diminished defensiveness [F(1,16)=9.00, p=0.009] without influencing (H) burrowing. (I) Compared to BL tests, inhibiting Crh+ aCMT cells with 561 nm light pulses reduced the duration of social interactions [F(3,45)=7.61, p<0.001], (J) increased defensive behaviors [F(3,45)=12.24, p<0.001], and (K) increased time spent burrowing [F(3,45)=4.07, p=0.012] whereas 473 nm pulses had no effect. (L-N) Post-CSDS, defeated mice interacted less with a social partner compared to control mice [t(14)=2.61, p=0.021] and were more defensive [t(14)=2.94, p=0.011]. Optogenetic inhibition of Crh+ aCMT cells produced a defeat-like phenotype in control mice by (F) increasing defensiveness [F(1,14)=5.35, p=0.036] while also (G) increasing time spent burrowing [F(1,14)=7.25, p=0.018]. (C-N) Data shown as the Mean ± SEM; Defeat vs. control: *p<0.05, **p<0.01; no stimulation (BL) vs. stimulation: +p<0.05, ++p<0.01, +++p<0.001. (A) Coronal brain images adapted from the Allen Institute Mouse Brain Atlas

Optogenetically inhibiting Crh+ aCMT cells produced a defeat-like phenotype in stress-naïve mice:

CRH-ires-Cre females expressing NpHR received optogenetic stimulations to inhibit Crh+ aCMT cells (Fig. S7). In stress-naïve control mice, inhibiting Crh+ aCMT cells produced a defeat-like phenotype by suppressing social interactions, escalating defensiveness, increasing burrowing, and decreasing walking compared to BL; control 473 nm laser pulses had no effect on sociability or walking (Fig. 3I–K; Fig. S8B; Video 2). CSDS suppressed BL social interactions and increased defensiveness (Fig. 3L, M); as observed prior to CSDS, inhibiting Crh+ aCMT cells increased defensiveness and burrowing and produced a trend toward social avoidance in controls (Fig. 3L–N).

Optogenetically activating Crh+ aCMT cells reduced deprivation-escalated alcohol drinking:

CSDS increased daily home cage continuous access alcohol intake in CRH-ires-Cre mice (Fig. 4A). Drinking was subsequently measured in a lickometer setup (Fig. 4B) during drinking sessions preceded by unrestricted alcohol access (No Dep, No Stim; light grey traces) or beginning with alcohol reintroduction after 24-hour deprivation (Dep, No Stim; dark grey traces). Mice engaged in a pattern of abstinence-induced binge alcohol drinking during the initial five minutes after alcohol reintroduction (Fig. 4C–H; left panels). The effect of deprivation on alcohol intake was evident for up to two hours, particularly in socially defeated female mice (Fig. S9C, Fig. S10B).

Figure 4. Optogenetic activation of Crh+ aCMT neurons diminishes abstinence-induced binge alcohol drinking in female mice.

(A) Chronically socially defeated female CRH-ires-Cre mice drank more alcohol (g/kg/24hr) than stress-naïve controls [one-tailed: t(33)=1.99, p=0.027] when given (B) continuous access to water (blue bottle) and 20% ethyl alcohol ((w/v); red bottle) in their home cage. Cages were fitted with contact lickometer panels to count alcohol and water bottle licks during unrestricted access (No Dep). Mice were alcohol-deprived (Dep) for 24 hours prior to alcohol reintroduction and optogenetic activation (ChR2) or inhibition (NpHR) of Crh+ aCMT neurons (Dep, Stim) or during sessions without laser stimulations (Dep, No Stim). (C, D, F, G) Average alcohol licks/minute over the first 10 minutes of each drinking session; data shown as the group Mean ± SEM (line ± shaded areas) with light grey lines indicating No Dep, No Stim sessions, dark grey lines indicating Dep, No Stim sessions, blue or green lines denoting Dep, Stim of ChR2 or NpHR sessions, respectively. Blue or green bars beneath x-axes indicate 5-minute stimulation of ChR2 or NpHR, respectively. Vertical dotted lines indicate laser stimulation offset or the matching time point during No Stim sessions. (C, D; left panels) Control (n=8) and defeated (n=10) ChR2-expressing female mice demonstrated deprivation-escalated alcohol drinking during the first five minutes of drinking during Dep vs. No Dep sessions [F(1,16)=17.37, p<0.001]; (C, D; right panels) optogenetic activation of Crh+ aCMT neurons blunted deprivation-induced binge drinking in both control and defeated mice [F(1,16) = 5.72, p=0.029]. (E) Summarized effects of deprivation and Crh+ aCMT cell activation on the first five minutes of alcohol intake; data shown as Mean ± SEM collapsed over control and defeat conditions. (F, G; left panels) Control (n=9) and defeated (n=8) NpHR-expressing mice exhibited deprivation-escalated alcohol drinking during the first five minutes of Dep vs. No Dep sessions [F(1,15) =11.33, p=0.004]. (F, G, right panels) Optogenetic inhibition of Crh+ aCMT neurons had no effect on deprivation-escalated intake. (H) Summary of effects of alcohol deprivation and optogenetic inhibition on alcohol licking during the first five minutes of each session type in NpHR-expressing mice; data depicted as Mean ± SEM collapsed over control and defeat groups. **p<0.01, ***p<0.001 deprivation effect; # p<0.05 stimulation effect.

To examine the effect of Crh+ aCMT cell stimulation on deprivation-escalated drinking, the onset of laser pulse deliveries coincided with alcohol reintroduction (Dep, Stim; blue and green traces; Fig. 4C, D, F, G; right panels). Stimulations occurred for two hours with 8×15-minute trials, each consisting of 5-min laser ON, 10-min laser OFF (Fig. S11). Optogenetically activating Crh+ aCMT cells blunted binge drinking during the first five minutes of alcohol access (Fig. 4C, D; right panels; Fig. S9B). Alcohol intake was suppressed throughout the two-hour stimulation protocol (Fig. S9D) and recovered after cessation of laser pulse deliveries (Fig. S9F, H). Optogenetic activation of Crh+ aCMT cells had no effect on water consumption (Fig. S9B–H). In NpHR mice, optogenetic inhibition of Crh+ aCMT cells had no effect on deprivation-escalated alcohol consumption or concurrent water intake (Fig. 4F–H; Fig. S10C, D).

cFos was quantified as a readout for laser stimulation effects. ChR2 mice exhibited increased cFos activation in response to 473 nm, 40 Hz stimulations compared to 561 nm pulses (Fig. S6). NpHR mice exhibited reduced cFos during social interactions paired with 561 nm light pulses compared to 473 nm stimulations (Fig. S7). Examination of viral expression revealed Crh+ projections from the midline and intralaminar thalamus (Fig. S12) to the mPFC, NAc, lateral septum, BNST, and BLA (Fig. S13).

Discussion

Traumatic social stress can lead to the development of mood disorders and dysregulated alcohol drinking, particularly among female patient populations (6, 7). Here, a translational model of this disease trajectory revealed a population of Crh+ anterior central medial thalamic (aCMT) cells that are rendered hypoactive – as measured by cFos, a correlate of recent neuronal activity - by social defeat stress exposure in female mice. Optogenetically activating Crh+ aCMT neurons recovered sociability in chronically socially defeated female mice and blunted alcohol abstinence-induced binge drinking. Interestingly, this high-frequency activation of Crh+ aCMT neurons increased defensiveness in stress-naïve mice. These opposing outcomes of high-frequency stimulations could reflect a stress-induced deviation from adaptive Crh+ aCMT cell firing patterns that ultimately shifts the stimulation-effect curve for defensive behaviors. Longitudinal in vivo calcium imaging and electrophysiology studies in both male and female mice will help to explore such predictions and provide greater insight into the real-time effects of ongoing stress exposure and the long-term consequences of social trauma within this cell population.

In stress-naïve female mice, optogenetically inhibiting Crh+ aCMT neurons elicited defensiveness to mimic defeat-like phenotype. The same stimulation conditions reduced exploratory locomotion during social interactions without suppressing other measures of motor activity (e.g., burrowing, defensive behaviors, licking). While locomotor suppression may contribute to the observed reduction in thalamic cFos in defeated mice, it is unlikely that photoinhibition blunts motor activity to produce a defeat-like phenotype. Rather, the present optogenetics studies and examinations of locomotion in stressed females during non-social probes (46) would indicate that diminished locomotor activity is caused by the presence of a social partner. This evidence supports a specific role for stress-sensitive Crh+ aCMT neurons in mediating adaptive sociability in female mice.

Female mice that were subjected to social defeat stress consumed more alcohol compared to stress-naïve individuals, as reported previously in male mouse models (42–45, 115–118). Mice with a history of chronic voluntary alcohol consumption also engaged in intense binge drinking during the initial five minutes of alcohol access after a period of forced abstinence. Optogenetic activation of Crh+ aCMT neurons reduced abstinence-induced binge drinking without affecting concurrent water intake; this effect was only observed during the two-hour optogenetic stimulation protocol and not in the hours of drinking following cessation of cell activation. Importantly, compensatory drinking bouts did not emerge during the ten-minute durations that separated five-minute epochs of optogenetic activation, suggesting that activation of Crh+ aCMT cells may selectively and persistently lessen the negative affect that motivates binge drinking caused by acute cessation of chronic alcohol consumption. Suppression of alcohol intake during optogenetic activation of Crh+ aCMT cells was observed in both control and socially defeated mice, indicating that this cell population may be sensitive to a broad range of stressors that yield persistent, maladaptive affective states, including historical social trauma and involuntary abstinence from chronic alcohol consumption. Additional studies are required to further our understanding of relevant stress-sensitive thalamic nuclei; in particular, it will be crucial to determine if thalamic cell populations can be manipulated to counter the negative affective symptoms of alcohol withdrawal that contribute to relapse in patients with alcohol use disorders.

Social avoidance was apparent when Crh+ aCMT neurons were optogenetically inhibited. Easing inhibitory constraints on thalamic cells within thalamo-corticothalamic loops may disinhibit cortical interneurons and blunt corticothalamic afferents. Interestingly, a distinct population of thalamocortical cells - classified by the absence of dopamine D2 receptor gene expression (Drd2−) - is transiently hypoactive in response to acute stressors and social interactions (119). Determining the relationship between thalamic Drd2− and Crh+ cell populations will deepen our understanding of thalamic cell types and the functional and transcriptional characteristics of inherently dynamic, stress-sensitive neural ensembles with specific projection targets (120). Midline and intralaminar thalamic Crh+ cells projected to the mPFC, NAc, lateral septum, BNST, and BLA. Together, these target regions comprise corticolimbic networks that regulate reward and threat processing (121–128). Growing support implicates midline and rILN thalamic nuclei as hubs in a broader stress-sensitive network (e.g., mPFC, NAc, BNST, BLA; 59–79). As such, gene expression profiling could reveal targetable stress-sensitive transcription factors in molecularly defined thalamic cell populations that promote or maintain network-level gene expression changes and downstream behavioral repercussions of trauma (129–131).

The present work identifies a novel social stress-sensitive population of Crh+ aCMT neurons using a highly translational model of social trauma in female mice (46). Targeting this cell population with optogenetics revealed that: 1.) stress-naïve female mice become socially defensive when Crh+ aCMT neurons are inhibited, 2.) sociability is recovered in stress-exposed females that receive laser pulses activating Crh+ aCMT cells, 3.) social defeat stress promotes chronic alcohol consumption in female mice, 4.) acute forced abstinence from chronic alcohol access induces binge drinking when mice regain access to alcohol, and 5.) optogenetic activation of Crh+ aCMT neurons blunts abstinence-induced binge alcohol drinking. Together, the present studies implicate Crh+ aCMT neurons in social avoidance and defensiveness after stress exposure and in binge drinking induced by involuntary abstinence after chronic alcohol drinking. In a broader context, these findings point to Crh+ cells in the aCMT as a novel neuronal population that demands further consideration for its role in maintaining the persistent maladaptive affective states and behavioral repercussions that can result from social trauma and chronic alcohol use.

Key Resource Table

Resource Type	Specific Reagent or Resource	Source or Reference	Identifiers	Additional Information
Add additional rows as needed for each resource type	Include species and sex when applicable.	Include name of manufacturer, company, repository, individual, or research lab. Include PMID or DOI for references; use “this paper” if new.	Include catalog numbers, stock numbers, database IDs or accession numbers, and/or RRIDs. RRIDs are highly encouraged; search for RRIDs at https://scicrunch.org/resources.	Include any additional information or notes if necessary.
Antibody
	Polyclonal rabbit anti-cFos	Millipore Sigma	F7799
	Goat Anti-Rabbit IgG AlexaFluor-488-conjugated	Jackson ImmunoResearch Laboratories Inc.	111-545-144
Bacterial or Viral Strain
	AAV2-EF1a-DIO-hChR2(H134R)-mCherry-WPRE-pA	University of North Carolina Vector Core	N/A
	AAV2-EF1a-DIO-NpHR3.0-mCherry	University of North Carolina Vector Core	N/A
Chemical Compound or Drug
	95% Ethyl Alcohol (ACS/USP Grade)	PHARMCO-AAPER	111000190
	Ketamine	VedCo	CAS: 6740-88-1	KETAVED (KETAMINE HCL 100MG)
	Xylazine	Patterson Veterinary	CAS: 7361-61-7	AnaSed
	Carprofen	Patterson Veterinary	CAS: 53716-49-7	Rimadyl Injectable
	Saline	Hospira	00409-4888-12
	4% Paraformaldehyde in 1xPBS	Alfa Aesar	CAS: 30525-89-4
	Phosphate buffered saline (10x)	Fisher Scientific	BP3991
	Triton X-100	Millipore Sigma	CAS: 9002-93-1
	DAPI	Millipore Sigma	CAS: 28718-90-3
	Krystalon	Millipore Sigma	64969
	Cresyl violet	Millipore Sigma	C5042 CAS: 10510-54-0
Commercial Assay Or Kit
	Alcohol analyzer	Analox	AM1
	Alcohol analyzer reagent	Analox	GMRD-113
	Alcohol analyzer standard	Analox	GMRD-110(100)
Organism/Strain
	Swiss webster (CFW) mice (male and female)	Charles River Laboratories	024	Males castrated by Charles River Laboratories
	C57BL/6J mice (female)	The Jackson Laboratory	000664
	B6.Cg-Gt(ROSA)26Sor tm9(CAG-tdTomato)Hze/J (Ai9) mice	The Jackson Laboratory	007909	Bred to CRH-ires-Cre mice in-house
	B6(Cg)-Crh tm1(cre)Zjh/J (CRH-ires-Cre) mice	The Jackson Laboratory	012704	Experimental mice bred in-house
Software; Algorithm
	GraphPad Prism	GraphPad Software, Inc	v. 8.0 and v. 9.0
	The Observer XT	Noldus Information Technology	v. 12.0
	FIJI ImageJ	Schindelin J, Arganda-Carreras I, Frise E et al. (2012): Fiji: an open-source platform for biological-image analysis. Nat Methods 9: 676-682, PMID 22743772, doi:10.1038/nmeth.2019	v. 1.53e	https://imagej.net/Fiji
	ZEN Lite	Zeiss	N/A	https://www.zeiss.com/microscopy/us/products/microscope-software/zen-lite.html
	MATLAB	Mathworks	R2019b
	Med-PC	Med Associates Inc	IV
	Biorender	Biorender	N/A	Biorender.com
Other
	Optic fiber implants	Doric lenses	MFC_200/240-0.22_5mm_ZF1.25_FLT
	Mating sleeve	Doric lenses	SLEEVE_ZR_1.25_BK
	Patch cords	Doric lenses	MFP FC-FC and FC-ZF1.25
	Rotary joint	Doric lenses	FRJ_1×1_FC
	Gimbal	Doric lenses	GH_FRJ
	Stereotaxic frame	Kopf	Model 902
	Stereotaxic syringe pump	World Precision Instruments	UMP3T-2
	Sub-microliter syringe	World Precision Instruments	NANOFIL (10)
	Microinfusion needles (beveled)	World Precision Instruments	NF35BV-2
	Submandibular mouse lancet	MEDIpoint Inc.	Goldenrod 3 mm
	Dental cement	Stoelting Co.	51459
	Lasers	CrystaLaser	473 nm, 561 nm
	Pulse generator	A.M.P.I.	Master-8
	Lickometer insert	Med Associates	ENV-350RM
	Lickometer controller	Med Associates	ENV-250B
	Interface module, connection panel	Med Associates	DIG-716B, SG-716
	Home cage lickometer panel	Machined in-house	N/A
	Video camera	Logitech	C920
	Microscope	Zeiss	AxioImager Z1
	Microscope camera	Zeiss	AxioCam Mrm v.3.0
	Cryostat	Leica	CM3050