Alcohol Drinking Alters Stress Response to Predator Odor via Extended Amygdala Kappa Opioid Receptor Signaling in Male Mice

Maladaptive responses to stress are a hallmark of alcohol use disorder, but the mechanisms that underlie this are not well characterized. Here we show that kappa opioid receptor signaling in the bed nucleus of the stria terminalis (BNST) is a critical molecular substrate underlying abnormal stress responses to predator odor following heavy alcohol drinking. Exposure to predator odor during protracted withdrawal from intermittent alcohol drinking resulted in enhanced prefrontal cortex (PFC)-driven excitation of prodynorphin-containing neurons in the BNST compared to drinking or stress alone. Furthermore, deletion of prodynorphin in the BNST and chemogenetic inhibition of the PFC-BNST pathway restored abnormal responses to predator odor in alcohol-exposed mice. These findings suggest that increased corticolimbic drive may promote abnormal stress behavioral responses to predator odor during protracted withdrawal from heavy drinking. Various nodes of this PFC-BNST dynorphin-related circuit may serve as potential targets for potential therapeutic mediation as well as biomarkers of negative responses to stress following heavy alcohol drinking.Heavy alcohol drinking primes dynorphin / kappa opioid systems in the bed nucleus of the stria terminalis to alter stress responses in mice.


Introduction
Alcohol abuse exacts a tremendous toll on society, and long term drinking can dysregulate stress systems in the brain. Prolonged alcohol drinking and withdrawal experiences result in enhanced responsiveness and behavioral sensitivity to stress during protracted abstinence (Heilig et al. 2010). Reciprocally, clinical studies show that negative stress coping is 5 predictive of higher levels of drinking in alcoholics (Noone et al. 1999). As blunted responses to stress have been identified in alcohol-dependent people (Sinha et al. 2011), it is essential to consider mechanisms by which alcohol drinking impacts stress responses during protracted abstinence. While many studies have utilized animal models to investigate how stress drives increased alcohol drinking behaviors (Becker et al. 2011, Gilpin andWeiner 2017), few have 10 explored the effects of alcohol drinking on subsequent stress responsivity.
Chronic alcohol exposure engages brain stress signaling systems that influence drinking behaviors in a dynamic and complex manner (Koob and Kreek 2007). One such stress system is the neuropeptide prodynorphin (Pdyn) and its receptor, the kappa opioid receptor (KOR), which has been studied in the contexts of both mood and alcohol use disorders (Lutz and Kieffer 2013). 15 Limbic structures implicated in alcohol and stress behaviors, such as the bed nucleus of the stria terminalis (BNST), are rich in Pdyn and KOR (Le Merrer et al. 2009). The BNST is an integrative hub that may mediate the negative affective state associated with chronic alcohol use and withdrawal (Koob 2009, Kash 2012. KORs throughout the extended amygdala and the BNST alter anxiety-like behavior in mice (Bruchas et al. 2009;Crowley et al. 2016) and mediate 20 stress-induced reinstatement for alcohol reinforcement (Lê et al. 2018).
In this study, we tested whether BNST KOR/Pdyn signaling regulates abnormal stress responses after long-term alcohol drinking. We employed the ethologically relevant predator BNST PDYN alcohol-induced stress behavior 4 odor trimethylthiazoline (TMT) as a stressor, which is a compound isolated from fox feces. In rats and C57BL/6J mice, TMT activates specific brain regions involved in stress, anxiety, and fear, including the BNST (Day et al. 2004, Asok et al. 2013, Janitzky et al. 2015, and inactivation of the BNST blocks TMT-induced freezing (Fendt et al. 2003). Recent work suggests that distinct neuropeptide circuits in the BNST may drive opposing emotional states 5 (Giardino et al. 2018), which may be dependent on inputs from cortical sites to affect stress coping behaviors (Johnson et al. 2019). The current series of experiments investigate whether Pdyn neurons and KOR signaling in the BNST can modulate behavioral responses to stress in alcohol-exposed animals. We show that dysregulation of cortical inputs to BNST Pdyn neurons and BNST Pdyn neurons themselves underlie lasting behavioral changes to stressors that emerge 10 after chronic drinking. This is a critical area of study, as mitigating stress responses can contribute to improved alcohol relapse outcomes.

Results
Male C57BL/6J mice were given six weeks of intermittent access to alcohol (EtOH), a 15 protocol known to induce heavy voluntary drinking (Hwa et al. 2011), before behavioral testing during protracted (7-10 days) abstinence [ Fig 1A]. Mice consumed high amounts of EtOH [ Fig   1B] and increased their EtOH preference over time [ Fig 1C]. Further, mice achieved greater than 80 mg/dl blood EtOH concentrations, indicative of intoxication, which correlated with drinking behavior [Fig 1D;R 2 =0.59,p=0.0036]. To test stress responsivity during protracted abstinence 20 from EtOH, mice were exposed to the predator odor TMT in the home cage ).
Both water (H2O)-drinking controls and EtOH drinking mice showed a TMT-induced increase in plasma corticosterone [Fig 1E;TMT main effect: F1,10=26.79,p=0.0004, H2O BL vs TMT BNST PDYN alcohol-induced stress behavior 5 t10=3.32, p=0.0154, EtOH BL vs TMT t10=3.99, p=0.005]. We tracked the location of the mouse relative to the TMT and measured the time spent contacting the TMT and in the far corners [ Fig   1F]. EtOH-drinking mice displayed reduced avoidance of the TMT compared to the water (H2O)-drinking controls during protracted abstinence . As an initial screen to identify altered behavior separate from avoidance, we examined stress-related and exploratory behavior 5 in three mice per condition on a second-by-second basis [Fig 1 Suppl 1]. Since the primary difference among stress-related activities was burying, we focused our further analyses on this typical behavior in response to noxious stimuli . Specifically, EtOH drinkers demonstrated reduced burying behavior compared to controls [Fig 1I]. Previous studies have shown that activation of the Pdyn/KOR system can modulate stress-induced EtOH seeking (Lê et al. 2018 protracted time point, we next focused on identifying the mechanism for this long-lasting adaptation in the brain's dynorphin system. The BNST is a brain site known for its involvement in stress, anxiety, and addiction, and is regulated by the Pdyn/KOR system (Crowley et al. 2016). Previous studies in rats have shown that TMT increases BNST activity using c-Fos as a marker for active neuronal populations (Day 10 et al. 2004, Asok et al. 2013 (Fendt et al. 2003); however, the role of KOR signaling in this process has not been explored.
Thus, we next tested whether microinfusions of norBNI directly into the BNST would alter behavioral responses to TMT during protracted abstinence [ Fig 3A- 15 We then tested if dynorphin produced in the BNST played a role in behavioral changes following EtOH and TMT, as we have previously shown that BNST Pdyn can modulate synaptic transmission in the BNST (Crowley et al. 2016). Pdyn was deleted from the BNST using the Pdyn lox/lox mouse line (Bloodgood et al. 2020) via AAV Cre-GFP microinfusions [ Fig 5A- Given that we have previously reported increased glutamatergic transmission in the mPFC following acute TMT exposure ) and recent reports from the Radley lab 10 indicated a key role in PFC inputs to the BNST in stress regulation, we next wanted to investigate if EtOH and TMT together may strengthen the functional connection between mPFC and BNST Pdyn neurons. To do this, we injected an AAV encoding channelrhodopsin (ChR2) into the mPFC of Pdyn-GFP mice [ Fig 6A] and measured BNST cell responses to photostimulation of this pathway using slice electrophysiology [ Fig 6B]. A large proportion of 15 BNST PDYN neurons were light responsive after TMT in both H2O and EtOH mice, whereas nonstressed H2O mice had mostly non-responsive cells [ Fig 6C; 15 Here, we have identified a causal role for the mPFC-BNST PDYN pathway in mediating alcohol-induced alterations in TMT predator odor-evoked stress responses. First, we identified BNST PDYN as a stress-and alcohol-sensitive population using immunohistochemistry and in situ hybridization. With whole cell patch clamp electrophysiology, we found that enhanced synaptic drive in BNST PDYN cells was reduced by KOR antagonism in stressed mice with a history of 20 alcohol drinking. Finally, experiments with ex vivo optogenetics indicated that EtOH-drinking stressed mice had increased prefrontal cortical synaptic connectivity onto BNST PDYN cells compared to stressed H2O drinkers and unstressed EtOH drinkers. We were able to manipulate BNST PDYN alcohol-induced stress behavior 12 EtOH-induced alterations in TMT stress reactions using BNST KOR antagonism, BNST Pdyn deletion, and PFC-BNST chemogenetic inhibition. Altogether, our findings indicate that engagement of Pdyn/KOR signaling in the BNST promotes an allostatic shift in stress-responses following EtOH drinking. 5 Previous articles from our laboratory have shown that wild-type and transgenic mice exhibit relatively modest intermittent EtOH drinking and preference (Bloodgood et al. 2020) compared to those reported in Hwa et al. (2011) publication, which was likely a results of varying vivarium conditions. However, mice in this study still achieved intoxicating blood EtOH concentrations, and this intermittent schedule may be favorable over drinking levels in 10 continuous two-bottle choice access (Yu et al. 2019). In our hands, six weeks of intermittent access to EtOH affected behavioral responses to TMT predator odor. We interpret the EtOHinduced lack of avoidance of the predator odor as a maladaptive reaction to an innately stressful stimulus. While control mice displayed an array of stress behaviors in response to TMT (i.e. freezing, grooming, stretch-attend, etc.), a lack of burying was a prominent behavioral feature of 15 EtOH mice. Burying in response to an immediate threat is commonly interpreted as an innate, active coping behavior in rodents (De Boer and Koolhaas, 2003). While EtOH mice also showed increased anxiety-like behavior in the elevated plus maze during protracted withdrawal, this group difference was eliminated following TMT exposure, as seen in our control drug/virus experiments after BNST norBNI, Pdyn deletion, and mPFC-BNST inhibition, suggesting long- 20 lasting impact of TMT on performance in the elevated plus maze.

BNST KOR/Pdyn gates stress reactions after EtOH
Using converging approaches of intra-BNST norBNI infusions and genetic deletion of BNST PDYN using a floxed mouse line, we show that reducing Pdyn/KOR signaling at the pre-or BNST PDYN alcohol-induced stress behavior 13 post-synaptic level, respectively, normalizes alcohol-induced impairments in TMT behavioral responses during protracted abstinence. These results are in line with literature showing KOR antagonists can block anxiety-like behaviors precipitated by acute withdrawal from alcohol vapor (Valdez andHarshberger 2012, Rose et al. 2016) and suppress alcohol self-administration in post-dependent rats (Walker and Koob 2008, Schank et al. 2012, Kissler et al. 2014. KORs in engagement of BNST PDYN neurons during TMT exposure using fiber photometry, as specific subpopulation of BNST neurons are known to exhibit TMT-elicited calcium transients (Giardino et al. 2018).

Glutamatergic contribution to stress-enhanced signaling in BNST PDYN neurons
After assessing population activity of BNST PDYN/KOR after EtOH and stress and the 5 contributions of this population to drinking-induced alterations in behavior, we performed synaptic transmission experiments on BNST PDYN neurons during protracted abstinence from EtOH. In addition, non-stressed intermittent EtOH mice displayed modestly increased sIPSC frequency in BNST PDYN cells. While some studies from our laboratory have reported increased sIPSC frequency in the BNST 24 hr after drinking in monkeys (Pleil et al. 2015), others have 10 found increased sEPSC/sIPSC ratios in C57BL/6J mice 48 hr after ethanol vapor (Pleil et al. 2016) These differences are likely the result of cell-type specific population targeting, variations in drinking/exposure protocols, and withdrawal time points. During protracted abstinence, there were no apparent synaptic transmission differences between withdrawn mice and controls, although previous reports have found increased sEPSC frequency at this time point in female 15 drinkers in a BNST CRF population (Centanni et al. 2019). Rather, we found that TMT exposure increased glutamatergic transmission in BNST PDYN neurons after EtOH and TMT, suggesting enhanced glutamatergic activity across the region. While BNST PDYN synaptic drive did not differ between stressed H2O and EtOH mice, differences in transmission were revealed during KOR blockade with norBNI pretreatment. Altogether, while other studies have found that chronic 20 EtOH exposure and withdrawal can impact BNST spontaneous glutamatergic and NMDAR function (Kash et al. 2009, Wills et al. 2012, McElligott and Winder 2009, our findings are the first to highlight plastic shifts in response to stressors and identify pathway-specific alterations in neuropeptide signaling.

EtOH and stress interact revealing synaptic plasticity from cortical input
The mPFC, among other brain regions, is a known source of increased glutamatergic signaling onto BNST neurons. Previous work in the lab found that both central amygdala (CeA) 5 and basolateral amygdala (BLA) inputs to the BNST are KOR-sensitive, but mPFC inputs are KOR-insensitive (Crowley et al. 2016). Notably, photostimulation of BLA inputs promotes anxiolysis in alcohol-naïve mice (Crowley et al. 2016). Taken together, these findings suggest a model in which increased activity of BNST PDYN neurons promote release of Pdyn, which in turn inhibit amygdala inputs to the BNST to promote increased engagement of mPFC glutamate 10 signaling. We have previously identified PL layer 2/3 neurons as a population engaged in response to acute TMT using a combination of slice physiology and immunohistochemical approaches , providing converging data for engagement of this pathway by this specific aversive stimulus. However, while Pdyn/KOR signaling in the CeA appears to promote EtOH consumption (Bloodgood et al. 2020), Pdyn/KOR signaling in the BNST did not affect 15 EtOH drinking in our study.
Our study provides valuable insight into how the synaptic strength of the mPFC to BNST PDYN pathway may be altered by combined exposure to EtOH and stress. Indeed, we found that a higher proportion of BNST PDYN neurons were light-responsive following stress or EtOH compared to H2O controls, suggesting that these stimuli increase connectivity between the mPFC 20 and BNST DYN neurons. Taking into account the observed increases in oEPSC, AMPA, and NMDA amplitudes and AMPA/NMDA ratio after the combination of EtOH and TMT, it appears that EtOH exposure primes the synapse for aberrant responses to stressors under the control of a BNST PDYN alcohol-induced stress behavior 16 glutamatergic, mPFC-driven mechanism. Further, with repeated stimulation pulse trains, the EtOH TMT BNST PDYN cells show reduced short-term depression, suggesting increased fidelity and short-term plasticity in this circuit. Again, this is in line with known chronic EtOH-induced glutamate plasticity in BNST cells (Wills et al. 2012). In contrast to the mPFC-BNST pathway, we observed no differences in the strength of BLA inputs to BNST PDYN neurons after EtOH and 5 stress. Since there was a stress-EtOH interaction observed from the cortical projection, we wanted to examine how inhibition of this pathway could alter behavior. In an mPFC-BNST DREADD experiment, we found that chemogenetic inhibition also improved burying behavior in EtOH mice. This was specific to TMT behavior, as the manipulation did not affect EtOH drinking or anxiety-like behavior in the elevated plus maze. These combined strategies of testing 10 synaptic strength in slice and pathway-specific manipulation of behavior provide a mechanism for how long-term drinking and stress interact to dysregulate prefrontal inputs to BNST PDYN neurons.

BNST circuitry control of stress behavior
A recent paper from the Radley lab explored the role of the prelimbic (PL) mPFC to 15 BNST pathway in stress-related behaviors in rats using optogenetics (Johnson et al. 2019).
Activation of the PL to BNST circuit negatively correlated with freezing behavior, a measure of passive coping, in response to a shock prod, while photoinhibition increased freezing and decreased burying, a measure of active coping. Notably, they found that these behavioral effects were related to downstream control of the periaqueductal gray. An important future direction will 20 be to assess the role of specific downstream projection targets of BNST Pdyn neurons, including the periaqueductal gray, in EtOH-induced alterations in TMT behavioral responses. It is also possible that excitatory local microcircuity in the BNST activates GABA neurons that inhibit ventral tegmental area GABA output signaling reward, which leads to anhedonia-like behavior and reduced stress responding. Our results illustrating aberrant responses to stress during protracted withdrawal from alcohol complement established research characterizing a more general role for the BNST in anxiety-related behaviors employing chemo-and optogenetics in mice (Kim et al. 2013, Jennings et al. 2013, Marcinkiewcz et al. 2016, Mazzone et al. 2018, 5 Crowley et al. 2016.

Conclusions
Maladaptive responses to stress are a hallmark of alcohol use disorder, but the mechanisms that underlie this effect are not well characterized. Here we show that Pdyn/KOR signaling in the BNST is a critical molecular substrate disrupting stress-related behavioral 10 responses following heavy alcohol drinking. Further, our findings suggest that increased corticolimbic connectivity may underlie this phenomenon; thus, altered mPFC-BNST connectivity could serve as a potential biomarker of negative outcomes in alcohol use disorder.
Disentangling this imbalance of corticolimbic-driven stress neuropeptide signaling may lead to the development of novel therapeutics to enhance stress coping in persons with alcohol use 15 disorder.

Materials and Methods
Animals. Eight-week old male C57BL/6J mice (Jackson Laboratories, Bar Harbor, ME) were used for behavioral pharmacology experiments. To visualize Pdyn-expressing neurons, we generated a Pdyn-GFP reporter line by crossing preprodynorphin-IRES-Cre mice (Crowley et al. 2016, Bloodgood et al. 2020, Al-Hasani et al. 2015) (Pdyn-Cre) and Rosa26-flox-stop-L10-GFP 5 reporter mice. For conditional knockout of BNST Pdyn, we used the Pdyn lox/lox mouse line (Bloodgood et al. 2020). These mice were bred in the UNC facilities. All mice were grouphoused for at least 3 days before being singly housed in polycarbonate cages (GM500, Tecniplast, Italy) with a 12:12-h reversed dark-light cycle with lights off at 7:00am. Mice had Intermittent EtOH Drinking. Mice were given 24 hr access to a 20% (w/v) EtOH solution and water on an intermittent schedule (Hwa et al. 2011). Two bottles were held in modified drinking 15 spouts of plastic cage tops and weighed before and after daily EtOH access. A dummy cage without an animal was used to simulate fluid lost while positioning the bottles, so average fluid drip was subtracted from each mouse's daily drinking. Mice were tested for stress reactions to TMT during 7-10 day protracted abstinence after 6 weeks of intermittent drinking. Blood EtOH concentrations were measured in a subset of mice. Tail blood was collected after two hours of 20 intermittent EtOH drinking, and then centrifuged at 3000 rpm at 4°C. Separated blood plasma was stored at -20°C before analysis using the AM1 Analox Analyzer (Analox Intstruments Ltd., Lunenberg, MA).

Behavioral Assays after EtOH Drinking
TMT Predator Odor Exposure. Exposure to fox-derived synthetic predator odor, trimethylthiazoline (TMT), was performed in the home cage as previously described to elicit stress reactions in mice . Animals were moved to a separate experimental room for testing that included a fume hood and a large fan. Mice were tested one at a time for odor Drugs and Viral Vectors. 5 mg/kg norBNI (Cat no. 0347, Tocris) or saline was administered i.p., 1ml/100g, 16 hr before testing to both EtOH and H2O mice. For intra-BNST norBNI microinfusions, 5 ug/ul norBNI was injected with 50 nl AAV5-CamKII-eGFP to mark the injection site. Intra-BNST norBNI or PBS occurred 7 days prior to TMT testing. Both 16 hour 15 and 7 day drug pretreatment minimized handling stress prior to the predator odor exposure and allowed for KOR antagonism instead of non-specific mu opioid antagonism that occurs initially post-injection. Pdyn lox/lox mice received 300 nl AAV5-CamKII-Cre-eGFP (UNC Vector Core, Lot 6450) and control AAV5-CamKII-eGFP (Lot 4621B) in the BNST. 300 nl AAV5-CamKIIa-hChR2(H134R)-mCherry-WPRE-hGH (Addgene, Lot CS1096) was injected into the mPFC or 20 BLA of Pdyn-GFP mice for synaptic connectivity experiments in the BNST. C57BL/6J mice were injected with 300 nl AAV8-hSyn-DIO-hM4D(Gi)-mCherry (Addgene, Lot 6048) into the mPFC and AAV2retro-SL1-CAG-Cre into the BNST for mPFC-BNST inhibition. All BNST PDYN alcohol-induced stress behavior 21 intracranial injections were bilateral. 3 mg/kg clozapine N-oxide (CNO; Hello Bio, Princeton, NJ) was dissolved in saline before i.p. administration, 1ml/100g, 20 min before testing.
c-Fos Immunohistochemistry, Histology, and Microscopy. For c-Fos and histological verification, mice were deeply anesthetized with Avertin before transcardial perfusion with phosphate buffered saline and 4% paraformaldehyde. Brains were extracted 90 min following 5 TMT exposure for c-Fos, cryoprotected, and then sliced at 45 µm on a Leica 1200S vibratome.  Technologies), and imaged on a Zeiss 800 laser scanning confocal microscope. ImageJ was used to count mean intensity of fluorescence for Pdyn (550 channel) and Oprk (647 channel), and the cell counter plug-in was used to hand count Pdyn-positive neurons. After verification with DAPI, cells labeled with at least 3 puncta were counted as containing Pdyn mRNA. Hand counts were tabulated by a blind observer. 5 Slice Electrophysiology. Ninety minutes following TMT, during 7-10 day protracted withdrawal, mice were sacrificed via deep isoflurane anesthesia, and coronal brain slices containing the BNST were collected according to standard laboratory protocols , Crowley et al. 2016, Bloodgood et al. 2020

BNST PDYN alcohol-induced stress behavior
For ex-vivo optogenetic experiments, tissue was evaluated for light-evoked action potentials in the mPFC. Brains were discarded and not used for further experimentation if injection sites were missed or if action potentials were not present. A blue LED (470 nm, CoolLed) was used to optically stimulate release from channelrhodopsin (ChR2)-containing fibers (Crowley et al. 2016). Picrotoxin (25 uM), tetrodotoxin (500 nM), and 4-AP (200 uM) 5 were added to the aCSF to isolate monosynaptic oEPSCs with cells held at -70 mV. The intracellular solution was cesium gluconate (117 mM D-gluconic acid and cesium hydroxide, 20 mM HEPES, 0.4 mM EGTA, 5 mM tetraethyl ammonium chloride, 2 mM MgCl26H2O, 4 mM Na2ATP, 0.4 Na2GTP, pH 7.3, 287-292 mOsmol). oEPSC amplitude (pA) was the first peak of the paired pulse ratio with a 50 ms interstimulus interval. Paired pulse ratio was calculated as the 10 second peak amplitude divided by the first peak amplitude. Cells were held at -70 mV to isolate AMPAR-mediated current and at +40 mV for NMDAR-mediated current. In separate slices, ten 1, 5, and 10 Hz pulse trains were performed at -55 mV voltage clamp without the presence of ionotropic inhibitors with the cesium methanesulfonate internal solution. The nine subsequent amplitudes in the pulse train were normalized to the first peak. 15 For DREADD validation in slice, mPFC cell bodies were identified with mCherry expression. 10 µM CNO (Hello Bio, Princeton, NJ) was bath applied for 10 min, and the resting membrane potential was monitored in voltage clamp. Action potential firing was assessed before and after CNO application using an increasing ramp protocol in current clamp.
Statistics. Time spent in contact with the TMT, far corners, and burying behavior in seconds (s), 20 and distance traveled (cm) were analyzed with two-way ANOVA with drug/virus and EtOH as factors. Post-hoc paired and unpaired t-tests were two-tailed and used where appropriate. In experiments where virus was injected before EtOH, cumulative 6-week alcohol intake and BNST PDYN alcohol-induced stress behavior 25 average ethanol preference were compared via t-test. BNST c-Fos, Pdyn-containing, and c-Fos and Pdyn colocalization were analyzed with two-way ANOVA with TMT exposure and EtOH as factors. With norBNI physiology, saline-and norBNI-injected stressed mice were compared in separate two-way ANOVA with drug and EtOH as factors. To compare proportion of lightresponsive cells per condition, a Χ 2 test was performed. Furthermore, optically-evoked 5 experiments (e.g. oEPSC amplitude) were analyzed with two-way ANOVAs comparing TMT and EtOH exposure. Pulse trains were analyzed with repeated measures two-way ANOVA across stimulus time and condition. Alpha was set to 0.05. Biological replicates throughout behavioral, immunohistochemical, and electrophysiological studies were combined. Statistical tests were analyzed with GraphPad Prism 8 (La Jolla, CA, USA).        Figure 7. DREADD-mediated inhibition of mPFC-BNST pathway and assessment of EtOH drinking and TMT-related behaviors. A, Time course of mPFC-BNST chemogenetic strategy in C57BL/6J mice before EtOH and TMT. B, Images of AAV-hM4Di-mCherry expression in mPFC cell bodies, left, and BNST terminals, right (H 2 O mCherry n=7, H 2 O hM4Di-mCherry n=7. EtOH mCherry n=8, EtOH hM4Di-mCherry n=7). mCherry was enhanced with a GFP immunostain. Slice physiology validation of the DREADD strategy, in mPFC neurons, as represented by C, hyperpolarization of resting membrane potential after CNO bath application. Inset scale bar indicates 2 mV height and 30 sec time. D, Latency to action potential threshold before and after CNO with 100 pA current ramp steps. Scale bar indicates 20 pA height and 100 ms time. E, EtOH drinking (g/kg/24hr) across 6 wks with mPFC-BNST hM4Di (red) or mCherry (light red). EtOH intake (g/kg) across F, 1 hr, G, 4 hr, and H, 24 hr after i.p. saline (circles) or 3 mg/kg CNO (diamonds). I, Sample TMT heatmaps of EtOH mPFC-BNST mCherry (left) and EtOH mPFC-BNST hM4Di (right) mice. In the TMT test, J, distance traveled (cm), K, TMT contact, L, far corners (sec), and M, burying (sec). In the elevated plus maze, duration in the N, open arms (sec), and O, closed arms (sec). *p<0.05.