Potentiated GABAergic neuronal activities in the basolateral amygdala alleviate stress‐induced depressive behaviors

Abstract Aims Major depressive disorder is a severe psychiatric disorder that afflicts ~17% of the world population. Neuroimaging investigations of depressed patients have consistently reported the dysfunction of the basolateral amygdala in the pathophysiology of depression. However, how the BLA and related circuits are implicated in the pathogenesis of depression is poorly understood. Methods Here, we combined fiber photometry, immediate early gene expression (c‐fos), optogenetics, chemogenetics, behavioral analysis, and viral tracing techniques to provide multiple lines of evidence of how the BLA neurons mediate depressive‐like behavior. Results We demonstrated that the aversive stimuli elevated the neuronal activity of the excitatory BLA neurons (BLACAMKII neurons). Optogenetic activation of CAMKII neurons facilitates the induction of depressive‐like behavior while inhibition of these neurons alleviates the depressive‐like behavior. Next, we found that the chemogenetic inhibition of GABAergic neurons in the BLA (BLAGABA) increased the firing frequency of CAMKII neurons and mediates the depressive‐like phenotypes. Finally, through fiber photometry recording and chemogenetic manipulation, we proved that the activation of BLAGABA neurons inhibits BLACAMKII neuronal activity and alleviates depressive‐like behavior in the mice. Conclusion Thus, through evaluating BLAGABA and BLACAMKII neurons by distinct interaction, the BLA regulates depressive‐like behavior.

brain regions.The BLA plays a crucial role in orienting and processing threats. 10BLA reactivity may emphasize mood disorders, heightened anxiety, biased emotional processing, and preferential representation of negative information typical of depression. 11[14][15] The BLA consists of a majority (80%-85%) of spiny glutamatergic neurons and a minority (up to 20%) of GABAergic neurons. 16,178][19] Hyperactivity of BLA during major depression is one of the consistent findings in clinical neuroscience. 20,21Therefore, enhancing the inhibition in the BLA could be a potential treatment for depression.We recently found that the high-frequency activation of GABAergic neurons in the auditory cortex enhances inhibition and reduces glutamatergic neuronal activity in response to auditory stimuli. 22Regardless of proof from human experiments accusing changes in volume, 23 metabolism, 24,25 and valence response 26 of the BLA in depressed patients, how BLA individual neurons (BLA GABA and BLA CAMKII ) are engaged in the pathogenesis of depression remains largely unexplored.
Here, we adopted fiber photometry to investigate how aversive stimuli affect the BLA neuronal activity.Furthermore, we investigated how optogenetic stimulation affects real-time behavior.Next, we explored whether the inhibitory GABAergic neurons in the BLA present major functional input to BLA CAMKII neurons at the neural activity by using fiber photometry and behavioral level using chemogenetics and optogenetic methods.Finally, we used retrograde labeling to find the input sources to BLA.

| Animals
All experiments were certified by the Animal Advisory Committee at the City University of Hong Kong Guidelines for the Care and Use of Laboratory Animals.Male C57BL/6 J (8-12 weeks old) of normal appearance and weight were used for all behavioral tests, immunohistochemistry, fiber photometry, optogenetics, and chemogenetics experiments.Ai14 mice were used for retrograde labeling.All male mice were socially housed, five per cage, and kept under typical housing environments on corn cob litter in a temperature (23 ± 1°C)   animal room on a 12 h light/dark cycle (lights were on from 8:00 to 20:00 every day) with food and water ad libitum until surgery.All the behavioral experiments were performed in the night-time (Table S1).

| Virus injection and optical fiber implantation
Mice were anesthetized with sodium pentobarbital (50 mg/kg, i.p. injection) for bilateral stereotaxic injection of viruses into the BLA (AP: −1.60; ML: ±3.40; DV: −4.00; mm AP and ML are corresponding to bregma and DV from dura matter; AP, ML, and DV refer to the anteroposterior, mediolateral, and dorsoventral distance from the bregma, respectively).The coordinates were determined from the bregma corresponding to the mouse brain atlas.We infused 300 nL of the virus into each side at a rate of 50 nL/min.After each injection, the needle was kept in place for another 10 min and then gradually withdrawn.
To record the fiber photometry GcaMP signal mice were implanted with an optical fiber cannula (length 6 mm, NA 0.

| Fiber photometry-based calcium measurements
Mice were attached with a fiber patch cord and allowed to habituate for a minimum of 10 min prior to each testing session.To record fluorescence signals, the fiber-photometry system (Doric Lenses) used two continuous sinusoidally modulated LEDs (DC4100, ThorLabs) at 473 nm (220 Hz) and 405 nm (330 Hz) as light sources to excite GCaMP6s and an isosbestic autofluorescence signal, respectively.Both lights were connected to a large-core (200 μm), high-NA (0.37) optical fiber patch cord, which was coupled with a corresponding brain implant in each mouse.The light intensity at the interface between the fiber tip was set to 15 μW.GCaMP6s and autofluorescence signals were gathered by the same fiber and driven onto two separate photoreceivers (2151, Newport Corporation).An RZ5P acquisition system (Tucker-Davis Technologies; TDT), fitted out with a real-time signal processor, operated the LEDs and also autonomously demodulated the fluorescence brightness due to 473-nm and 405-nm excitations.Behavioral activities were recorded by the same system via TTL input.
The acute tail suspension protocol for fiber photometry was adopted from a previous study. 27Briefly, animals were connected to an optical patch cord and permitted to habituate for 10 min.
After that, animals were suspended with their tails by a single experimenter for a total of three trials, each lasting 30 s every 2 min.
Changes in baseline fluorescence during the acute tail suspension test (TST) were calculated by evaluating the 30 s of recording data during tail suspension to the 5 s before the onset of the TST.For the social interaction task, mice were permitted to habituate for 10 min with an optical patch cord.After that, animals were placed into the social interaction chamber by a single experimenter for a total of 5 min.For the social interaction test (SIT), stimulated changes in fluorescence were calculated by comparing the recording data 5 s before social interaction and 10 s after mice entered the CD-1 aggressor mouse interaction zone.In a separate group of animals, foot shock-induced neural activity was analyzed by comparing the 5 s before foot shock and 10 s after foot shock.Data were analyzed by pMAT an open-source software suite for the analysis of fiber photometry data. 28The resulting fitted 405-nm signal was then used to normalize the 473-nm signal as follows: ΔF/F0 = (473-nm signal − fitted 405-nm signal)/fitted 405-nm signal.

| Optogenetic alleviation of depressive-like behavior
Mice injected with either eArchT3.0 or EYFP (control) in the BLA of C57 mice.For optogenetic inhibition of BLA CAMKII neurons mice were attached to an optical patch cord and placed on opposite sides of CD-1 aggressor mice with the divider.Mice received yellow light inhibition for 4 s on and 1 s off for 10 min for three consecutive days post chronic social defeat stress.

| Chemogenetic enhancement of depressive-like behavior
Mice injected with either hM4Di or mcherry (control) virus in the BLA of C57 mice.For chemogenetic facilitation of depressive-like behavior with subthreshold social defeat stress.Mice received 5 mg/ kg CNO immediately after the first trial of two trials of social defeat and were kept on opposite sides of the CD-1 aggressor for 15 min before the second trial of physical defeat.After the second trial mice were kept in their home cage and behavior test were conducted the following days to evaluate depressive-like behavior.

| Immunohistochemistry
Mice were anesthetized by an overdose of pentobarbital sodium 60 min after sensory stimuli and transcardially perfused with 30 mL PBS, followed by 30 mL of 4% paraformaldehyde (PFA) in PBS.Brains were removed and fixed with 4% PFA overnight.
Coronal sections 50 μm in thickness were cut using a vibratome.
Brain sections were first washed with PBS and then blocked with 5% goat serum in PBST (0.3% Triton X-100 in PBS) for 2 h.Then slices were incubated with primary antibody anti-c-fos rabbit (ab190289) with a concentration of (1:1000) for 24 h at 4°C.Slices were washed 3 times with PBS and incubated with secondary antibodies either Alexa fluor 488 (1:500) or Alexa fluor 647 (1:500) at room temperature for 2 h, followed by washing 3 times with PBS and then incubated with DAPI (1:5000) for 5 min and washed again with PBS for 3 times.Slices were mounted on slides and images were taken with a Nikon confocal microscope.Data were analyzed with Image J software (Table S1).Real-Time Place Aversion (RTPA) was performed 3 weeks after surgery, both ChrimsonR-expressing mice and their controls were bilaterally implanted with optical fibers above the BLA.Mice were placed into the behavioral arena (32 cm × 32 cm) containing two chambers.

|
Mice were habituated to the chamber for 10 min before starting a testing session.We designated one counterbalanced side as the stimulation site for the following 10-min test.At the beginning of the test session, the mouse was placed on the non-stimulated side of the chamber, and every time the mouse crossed to the stimulation site, 40 Hz laser stimulation was given until the mouse crossed back into the non-stimulation side.The movements of mice were tracked by a video camera located above the chamber and analyzed using Smart 3 software.The avoidance score was determined as (Time on Laser ON Side ÷ Total Time) × 100.

| Chronic social defeat stress
Chronic social defeat stress was conducted as defined previously. 29iefly, before the experiment, retired male breeder CD1 mice were assessed on three consecutive days to validate their aggressive traits.Subsequently, each experimental male C57BL6/JL mouse (intruder) was placed into the home cage of a novel aggressive CD1 mouse (resident) for 5-10 min for physical defeat for consecutive 10 days.
Following the physical defeat, experimental mice were maintained in sensory contact for 24 h by being placed on the opposite side of CD 1 aggressors using a perforated Plexiglass partition.After CSDS, mice were housed individually, and a SIT was conducted 24 h later.
Control mice were kept in pairs on opposite sides of the perforated Plexiglass partition in the same cage for 10 days without any physical defeat.

| Social interaction
Social avoidance behavior was assessed corresponding to a two-trail SIT.In the first trial, mice were introduced in an open-field chamber (44 × 44 × 44 cm 3 ) including an empty cage made of mesh wire and plastic sheets (10 × 6 × 8 cm 3 ).During behavioral experiments, the mice were tracked with Debut video-tracking software.Time spent in the social interaction zone (8 cm around the cage) was quantified individually.In the first trial, the time spent in the interaction zone was named "No target."Mice were then put back in the home cage for 1 min.In the second trial, a novel, CD1 aggressor mouse was put in the cage and was named (Target).The social interaction ratio was evaluated by (time spent in interaction zone with Target/time spent in the interaction zone with No target).Animals were regarded as susceptible when this ratio was <1 and resilient when the ratio was >1.

| Two-trial subthreshold social defeat stress
A two-trial subthreshold social defeat (SSDS) stress experiment was carried out as defined previously. 30The experimental mouse was placed into the home cage of an aggressive CD1 for 10 min for physical defeat.Following 10 min of physical stress, the experimental mice underwent 15 min of sensory stimuli opposite side of CD1 by using a perforated Plexiglass wall.Then the mice underwent a second trial of physical defeat with a novel CD1 mouse and then returned to the home cage, and behavior tests were conducted the following days to evaluate the depressive-like behavior.

| Sucrose preference test
Following the SIT, mice were habituated for two consecutive days to 50 mL tubes with rubber stoppers with ballpoint steel sipper tubes (two-bottle choice).Subsequently, one bottle was changed with 1% sucrose to determine the sucrose preference.Both bottles including water and sucrose were weighed at several time points, 09:00, 15:00, and 18:00.The bottle's position was changed each time after weight measurements (left to right, right to left) to confirm that the mice do not develop the side preference.Sucrose preference was determined as the percentage (amount of sucrose consumed × 100 (bottle a)/total volume consumed (bottles a + b)).
Total sucrose intake over the 24 h was quantified and used to find sucrose preference.

| Tail-suspension test
Briefly, mice were hanged by their tails with adhesive tape (17 cm in length, 1 cm from the tail tip) and approximately 50 cm higher than the surface so no contact could be made.Plastic tubes were positioned on the tails to guarantee mice could neither climb nor hang on to their tail.The mice were with a video camera for 6-min and the immobile duration of over 6-min was used to determine the despairlike behavior.S1).

| Aversive stimuli induced BLA neuronal activities
To investigate how the aversive stimuli modulate the BLA neuronal activities, we adopted a well-established TST 27,31 and chronic social defeat stress (CSDS) paradigm-induced social avoidance task. 32rst, we found that the CSDS decreased the social interaction time in the presence of the target (Figure S1A,B

| Effect of laser activation of BLA CAMKII or BLA GABA on real-time behavior
We have shown above that the aversive stimuli led to the hyperactivity of BLA via increased BLA CAMKII neuronal activity.The next experiment examined the effect of laser stimulation of BLA CAMKII or BLA GABA on real-time behavior. 6,34We injected ChrimsonR with a specific promoter, CAMKII for glutamatergic neurons or Dlx for GABAergic neurons into BLA and implanted optic fiber over the BLA to specifically activate either BLA CAMKII or BLA GABA neurons soma in awake, freely moving mice (Figure 2A).Laser stimulation of BLA CAMKII

| BLA CAMKII neurons mediate depressive-like behaviors
It is known that CSDS induced depressive-like behaviors, 35,36 but SSDS could not induce depressive-like behavior unless it is combined with chemogenetic/optogenetic manipulation of stress-mediating neurons. 30,36Consistent with previous studies, 30 we found that the SSDS could not induce depressive-like behavior but CSDS induced the depressive-like behavior which is indicated by decreased social interaction ratio, decreased sucrose preference, and center time in the open field along with increased immobile time in the TST (Fig- ure S2A-E).The following experiment examined whether optogenetic stimulation of BLA CAMKII neurons with subthreshold social defeat stress (SSDS) could induce depressive-like behavior (decreased social interaction, decreased sucrose preference, increased immobile time), in the mice (Figure 3A).We applied the laser stimulation during a 10-min sensory period following a physical defeat.
Chronic aversive stimuli experiencing is a fundamental risk factor that leads to the development of depressive-like states.Therefore, we next adopted a CSDS model to further explore whether BLA CAMKII neurons mediate stress-induced depressive-like behavior, we silenced these neurons by expressing inhibitory opsin eArch (Figure 3F).[39][40] Chronic therapy could be the best way to treat depression.Thus, we adopted the three consecutive days of optogenetic inhibition of BLA CAMKII manipulation to treat the depressive behaviors in the mice to achieve the best outcome.We found that optogenetic inhibition of eArch expressing neurons reduced stress susceptibility in the SIT (Figure 3G,H; EYFP before laser 0.68 ± 0.05 vs. EYFP after laser 0.58 ± 0.10; p < 0.96; eArch before laser 0.53 ± 0.09 vs. eArch after laser 1.12 ± 0.12; p < 0.0012), increased sucrose intake in SPT

| BLA GABA modulates the BLA CAMKII neuronal activities and depressive-like behaviors
Then, we explored how the BLA GABA neurons control the BLA CAMKII neurons' activities.We injected mDlx-hM4Di and CAMKII-GCAMP6s viruses into BLA and performed fiber photometry recording after recovery (Figure 4A).CNO injection increased the firing of BLA CAMKII neurons as indicated by the increased number of spikes (Figure 4B) and increased firing frequency of BLA CAMKII neurons (Figure 4C; Baseline 0.13 ± 0.05 vs. Saline 0.16 ± 0.06 vs. CNO 0.56 ± 0.04; p < 0.009) as compared with baseline and saline recording duration.In a separate group of mice, we injected mDlx-hM3Dq and CAMKII-GCAMP6s viruses into BLA to examine whether the chemogenetic activation of BLA GABA could inhibit the BLA CAMKII neuronal activity.The saline was injected 30 min before fiber photometry recording and looked at the foot shock-induced (0.2 mA) BLA CAMKII neuronal activity then CNO was injected and 30 min later again fiber photometry recording was performed (Figure 4D).It has been found that CNO injection decreased the foot shock-induced BLA CAMKII neuronal activity as shown by decreased calcium responses after CNO application (Figure 4E,F; Saline 8.46 ± 2.25 vs. CNO 2.88 ± 1.09; p < 0.002).
Afterward, we investigated the role of BLA GABA neurons in depression at the behavior level using chemogenetic inhibition.We injected the Dlx-hM4Di or Dlx-mCherry virus into the BLA of C57 mice.Mice underwent a two-trail sub-threshold social defeat paradigm.Immediately after the first trial of physical defeat, mice received a single injection of 5 mg/kg CNO and were kept in a cage opposite to (facing) the CD-1 mouse for 15-min sensory stimulation, followed by a second trial of physical defeat (Figure 5A).Chemogenetic inhibition of BLA GABA neurons facilitated the induction of depressive-like behavior with SSDS in mice.We found that chemogenetic inhibition of BLA GABA neurons induced stress susceptibility in the SIT (Figure 5B,C; mCherry 1.23 ± 0.14 vs. hM4Di 0.44 ± 0.11; p < 0.0001), decreased performance in the SPT (Figure 5D Chemogenetic inhibition of BLA GABA neurons during SSDS not only facilitated the induction of depressive-like states but also increased the c-fos activity in the BLA (Figure S4D,E; mCherry 8.18 ± 1.12 vs. hM4Di 20.67 ± 2.21; p < 0.0001).
If hyperactivity of the BLA is characteristic of susceptibility to stress, we hypothesized that chemogenetic activation of BLA GABA could promote resilience to social stress.To test this, we selectively expressed either hM3Dq-EGFP or only EGFP in BLA GABA neurons (Figure 5F).Chronically chemogenetic activation of the BLA GABA neurons over three consecutive days in the defeated susceptible mice expressing hM3Dq-EGFP showed an increase in social interaction (Figure 5G,H; Before CNO hM3Dq 0.82 ± 0.09 vs.After CNO hM3Dq 1.57 ± 0.20; p < 0.003), as well as a dramatic improvement in the SPT (Figure 5I p < 0.009).However, the same 3-day chemo manipulation in the EGFP group had no effect.These results suggest an antidepressant effect caused by 3-day chronic chemo manipulation of the BLA GABA neurons in susceptible mice.Additionally, we observed that the chemogenetic stimulation of BLA GABA reduced the c-fos activity in the BLA (Figure S4J,K; EGFP 17.29 ± 2.26 vs. hM3Dq 8.81 ± 1.20; p < 0.003).In summary, BLA GABA modulates depressive-like behavior.detected in several regions, in the particular inferior limbic cortex (IL), prelimbic cortex (PL), anterior cingulate cortex (ACC), auditory cortex (Auc), anterior insular cortex (AIC), and lateral entorhinal cortex (LEnt) (Figure S5A-C).In general, these input sources are consistent with previous anatomical data, 41 These structures have been implicated in stress responses, and anxiety, 18,[42][43][44][45] supporting that BLA may integrate aversive events information from multiple sources.1][52][53] Hyperactivity of the basolateral amygdala during stressful event re-experience and in depression is one of the most reliable findings of clinical 20,21 and pre-clinical neuroscience. 54,55though BLA has been implicated in the regulation of emotional behavior, the functional role of BLA CAMKII neurons is still unclear.

| Input sources to the basolateral amygdala
Consistent with previous studies, we found that aversive stimuli increased the activity of BLA.Of note, the tail suspension aversive stimuli and chronic social defeat-induced SIT aversive stimuli dramatically magnified the activity of BLA CAMKII neurons.Activation of these neurons triggers real-time place aversion and despair-like behavior, while suppressing these neurons alleviates the expression of depressive-like behavior after exposure to chronic social stress.
The increased activity of CAMKII neurons after exposure to stress accounts for the persistence of the induced depressive state and suppressing these neuron activities is sufficient to reduce the basallevel depressive state.Together, our results manifest that BLA CAMKII neurons regulate the induction, expression, and maintenance of depressive states in general.
The regulation of negative emotional states including fear, anxiety, and depression, depends on the processing of negative stimuli.
The activation of BLA CAMKII neurons appears to result in an extremely strong negative emotion as compared with the previously studied structures such as the bed nucleus of the stria terminalis (BNST) 56 and paraventricular nucleus (PVN). 57Notably, the percentage of time spent in the photo-stimulation chamber is comparatively less than when other structures are activated. 56,57These results offer strong proof that the activity of BLA CAMKII neurons can encode extremely negative valence, and may thus lead to relentless fear and a depressive state.This property of BLA CAMKII neurons manifests that they might be a unique therapeutic target for treating severe depressive disorders.
Opposite to the CAMKII neurons, our experiments demonstrate that the BLA GABA neurons encode positive valence and promote antidepressant-like effects.Interestingly, by comparing CAMKII and GABAergic neurons, our current experiments reveal that the antidepressant-like effect caused by the activation of the GABAergic neurons is much stronger than suppressing the CAM-KII neurons.In addition to antidepressant-like effects, activation of GABAergic neurons is also anxiolytic which is consistent with previous findings showing the impaired decreased GABAergic neuron activity associated with anxiety state. 58These results suggest that the activity of CAMKII neurons cannot encode a full scale of valence values and positive effects are primarily coded by GABAergic neurons.
Our results indicated that the inhibition of BLA GABA increased the firing frequency of BLA CAMKII neurons.Besides, we found that the foot shock increased the activity of BLA CAMKII , whereas chemogenetic stimulation of BLA GABA neurons reduced the activity of BLA CAMKII neurons.These results supported that BLA CAMKII neuron activity is tightly controlled by local GABAergic interneurons. 19,59,60Chemogenetic stimulation of BLA GABA neurons may enhance the inhibition by high-frequency long-term activation of BLA GABAergic neurons.This is supported by a study published in the SSRN elibrary (preprint) showing that chemogenetic inhibition decreases the number of seizures in the KA-induced epilepsy mice model by enhancing GABAergic inhibition. 61High-frequency stimulation of GABA neurons may facilitate the release of inhibitory CCK which strengthen the inhibitory synapses and enhanced the inhibition in the amygdala.Indeed, a recent study showed that the somatodendritic release of CCK enhanced the GABAergic transmission onto VTA dopamine neurons. 62Recently, we showed that the high-frequency stimulation of GABAergic neurons induced iLTP in the auditory cortex, and this effect was mediated by a novel CCK receptor. 22Furthermore, our previous studies also showed 37) held in a ceramic ferrule (Inper company) over the BLA (AP: −1.60; ML: ±3.40; DV: −3.85; mm AP and ML are corresponding to bregma and DV from dura matter.For optogenetic stimulation, mice were implanted with a bilaterally optical fiber cannula (length 6 mm, NA 0.37) held in a ceramic ferrule (Inper company) over the BLA (AP: −1.60; ML: ±3.40; DV: −3.70; mm AP and ML are corresponding to bregma and DV from dura matter.

2. 9 . 3 |
Open-field test Mice were separately presented to the central zone of an open-field (44 × 44 cm 2 ) arena with dim light in the dark phase for 10 min.Their movements were recorded by video camera and analyzed with smart 3.0 software.Anxiety-like behavior was determined by the time spent in the central zone.

2. 10 |
Quantification and statistical analyses All the data are shown as means ± s.e.m.Statistical evaluations were achieved using GraphPad Prism 8.0.1.software with suitable inferential methods.Normally distributed data were assessed by paired, and unpaired t-tests for two-group comparisons, one-and two-way analysis of variance (ANOVA), and multiple comparisons were performed by the Bonferroni test.Statistical significance was set at *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 (Table photometry recording was performed during SIT Post-CSDS (Figure 1A).We observed the tail suspension aversive stimuli robustly activated BLA CAMKII neurons as indicated by increased calcium responses from 0.57 ± 0.37 to 5.34 ± 1.3 during TST (Figure 1B-E; p < 0.003).Footshock (FS) is another aversive stimuli that also evoked strong calcium responses from −0.37 ± 0.42 to 10.45 ± 1.71 (Figure S1E,F; p < 0.0003).Moreover, we found that the BLA CAMKII neurons showed increased activity during the aversive social stimuli as exhibited by elevated calcium responses from 0.22 ± 0.21 to 5.1 ± 0.73 during the SIT (Figure 1F-I; p < 0.0013).

F I G U R E 2
Effect of laser activation of BLA CAMKII or BLA GABA on real-time behavior.(A) Left: Experimental design, Day 1 virus injection, Day 21 RTPP/RTPA test, and Day 22 TST.Right: representative image showing the location of fiber tips and virus expression in the BLA.(B) Heat map showing the time spent in light on the paired chamber and light off chamber between mcherry and CAMKII-chrimsonR group.(C) % Time spent on light ON side between mcherry and CAMKII-chrimsonR group, ***p < 0.0003, t = 4.985, df = 12, N = 7 mice, two-tailed unpaired t-test.(D) Heat map showing the time spent in light on the paired chamber and light off chamber between mcherry and Dlx-chrimsonR group.(E) % Time spent on light ON side between mcherry and Dlx-chrimsonR group, p < 0.2, t = 1.335, df = 14, N = 8 mice, two-tailed unpaired t-test.(F) Experimental setup for laser stimulation and TST.(G) % Immobile time during tail suspension test mcherry and CAMKII-chrimsonR, *p < 0.0381, t = 2.330, df = 12, N = 7 mice, two-tailed unpaired t-test.(H) % Immobile time during tail suspension test mcherry and Dlx-chrimsonR, *p < 0.0497, t = 1.765, df = 14, N = 8 mice, one-tailed unpaired t-test.All data are shown as mean ± s.e.m. (Figure 3I; EYFP 63.11 ± 4.82 vs. eArch 84.55 ± 2.18; p < 0.0004), reduced immobility in the TST (Figure S3F; EYFP 47.78 ± 3.94 vs. eArch 24.63 ± 2.54; p < 0.0001), however, there was no improvement in center time in OFT (Figure S3G-I; EYFP 6.21 ± 0.77 vs. eArch 5.78 ± 1.03; p < 0.74).Optogenetic inhibition of BLA CAMKII neurons not only inhibited the depressive-like states but also reduced the cfos activity in the BLA (Figure S3J,K; EYFP 24.91 ± 2.20 vs. eArch 13.88 ± 1.30; p < 0.0004).Together these results of activation and inactivation experiments suggest that the BLA CAMKII neurons mediate the expression of physical stress-induced depressive-like behavior.

F I G U R E 4 FIGURE 5
FIGURE 5 Legend on next page