Arc-mediated plasticity in the paraventricular thalamic nucleus promotes habituation to stress

Abstract

BACKGROUND

Habituation is defined as a progressive decline in response to repeated exposure to a familiar and predictable stimulus and is highly conserved across species. Disrupted habituation is a signature of post-traumatic stress disorder (PTSD). In rodents, habituation is observed in neural, neuroendocrine and behavioral responses to repeated exposure to the predictable and moderately intense stress or restraint. We previously demonstrated that lesioning the posterior paraventricular thalamic nucleus (pPVT) impairs habituation. However, the underlying molecular mechanisms and specific neural connections among the pPVT and other brain regions that underlie habituation are unknown.

METHODS

Behavioral and neuroendocrine habituation was assessed in adult male Sprague-Dawley rats using the repeated restraint paradigm. Pan-neuronal and Cre-dependent Designer Receptors Exclusively Activated by Designer Drugs (DREADDs) were used to chemogenetically inhibit the pPVT and the subpopulation of pPVT neurons that project to the medial prefrontal cortex (mPFC), respectively. Activity-regulated cytoskeleton-associated protein (Arc) expression was knocked down in the pPVT using siRNA. Structural plasticity of pPVT neurons was assessed using Golgi staining. Local field potential recordings were used to assess coherent neural activity between the pPVT and mPFC. The attentional set-shifting task was used to assess mPFC-dependent behavior.

RESULTS

Here, we show that Arc promotes habituation by increasing stress-induced spinogenesis in the pPVT, increasing coherent neural activity with the mPFC, and improving mPFC-mediated cognitive flexibility.

CONCLUSION

Our results demonstrate that Arc induction in the pPVT regulates habituation and mPFC function. Therapies that improve synaptic plasticity during PTSD therapy may enhance habituation and the efficacy of PTSD treatment.

One Sentence Summary:

We demonstrate that Arc in the posterior division of the paraventricular thalamic nucleus promotes habituation to repeated stress by increasing dendritic spines.

INTRODUCTION

Habituation is defined as progressively reduced responsiveness to repeated exposure to the same predictable, mild to moderately intense stressor (1–3). Habituation is displayed by species ranging from rodents to humans and is characterized in rodents by progressive reductions in behaviors and hypothalamic-pituitary-adrenal (HPA) axis responses to repeated experience with predictable and familiar homotypic stressors (1, 2, 4–6). Habituation is disrupted in individuals with post-traumatic stress disorder (PTSD) and is a key contributor to hyperarousal and re-experiencing symptom clusters (7–10). Additionally, habituation is critical for the success of behavioral therapies for PTSD (8, 11, 12). Therefore, understanding the mechanisms underlying habituation may improve the efficacy of PTSD therapy. Habituation to stress is considered adaptive because it allows humans and animals to filter out irrelevant stimuli and focus selectively on important stimuli (13). However, a clear understanding of the molecular substrates underlying habituation to repeated stress has not emerged.

We have previously demonstrated that activity in the posterior division of the paraventricular thalamic nucleus (pPVT) is necessary for habituation (14). The PVT is an extensive midline thalamic nucleus divided into anterior, middle and posterior divisions based on segregation of afferent and efferent projections (15–20). Efferents of the anterior and medial divisions of the PVT are widespread (17, 21–23). In contrast, efferent projections of the pPVT are limited to: the central, basomedial, basolateral amygdala, nucleus accumbens, anterior olfactory nucleus, bed nucleus of the stria terminalis (BNST), peri-posterior paraventricular nucleus of the hypothalamus (peri-PVN), but not the PVN, and the medial prefrontal cortex (mPFC) (17, 18, 20, 24–27). The specific molecular substrates within the pPVT underlying habituation are not clear.

Activity-regulated cytoskeleton-associated protein (Arc/Arg3.1) is an immediate early gene that couples changes in neuronal activity with synaptic plasticity (28–30). Arc expression is increased within 30min of an activity-inducing event and returns to baseline within 4 hours (31). Arc is necessary for late-phase long-term potentiation (LTP) in the hippocampus (30, 32, 33). This is attributed to Arc’s role in interacting with proteins that stabilize and promote the branching of actin filaments (29, 32, 34, 35), thereby increasing dendritic spine numbers (36, 37). Indeed, Arc is required for hippocampus-dependent memory (30). Therefore, Arc-mediated mechanisms that have been identified in the hippocampus may also contribute to habituation to repeated stress by promoting neuronal plasticity in the pPVT.

Here, we show that Arc in the pPVT is induced following restraint, and that this induction is necessary for habituation to repeated restraint. We demonstrate that Arc promotes habituation through the formation of new spines on pPVT dendrites. Further, Arc in the pPVT regulates the coherence of oscillatory activity between the pPVT and the mPFC and mediates PFC-mediated cognitive flexibility in habituated animals. Together, these findings suggest that Arc-mediated synaptic plasticity in the pPVT and its regulation of projections to the mPFC underlie a phylogenetically-conserved adaptation to stress. Treatments that promote synaptic plasticity during behavioral therapy for PTSD may enhance habituation and improve therapeutic efficacy.

METHODS AND MATERIALS

Animals

Adult male Sprague–Dawley rats (225–250 g) were obtained from Charles River Laboratories and singly housed. Rats were euthanized by rapid decapitation and their brains were immediately snap-frozen in 2-methylbutane or prepared for Golgi staining (see supplemental information).

Restraint

Rats were restrained daily in clear, plexiglass tubes. On day 1 or day 5, they were euthanized 60min following restraint onset (4, 38, 39). Tail blood was collected to assess plasma ACTH and corticosterone (14, 38, 40) (MP Biomedicals; Orangeburg, NY) (see supplemental information).

Immunohistochemistry

The following primary antibodies were used: rabbit anti-Arc (156003 1:200, Synaptic Systems), rabbit anti-mCherry (ab167453, 1:200, Abcam), and rabbit anti-HA (C29F4, 1:250, Cell Signaling). See supplemental information for all antibodies.

Stereotaxic Injections and drug administration

The following siRNA constructs were used: iAAV-scramble (iAAV01508, serotype 8) and iAAV-Arc (iAAV06494008, serotype 8). Non-Cre-dependent AAV8-hSyn-hM4D-HA-IRES-mCherry and Cre-dependent AAV8-hSyn-DIO-hM4D-HA-mCherry were purchased from the University of North Carolina Vector Core. CAV2-Cre was purchased from Institut de Génétique Moléculaire de Montpellier. Arc knockdown surgeries were performed 7 days prior to restraint. Cre and hM4D overexpression studies were performed 28 days prior to restraint. For DREADDs experiments, rats were intraperitoneally administered clozapine-N-oxide (CNO, 2 mg/kg) or vehicle (4% DMSO in saline) 60min prior to each restraint. For pharmacological inhibition of spinogenesis via MK-8931 injections, cannulae (Plastic One, C313I/SPC) were placed to target the pPVT. 1 µL of either vehicle (10% DMSO in saline) or MK-8931 (Verubecestat, 100 ng/1µL, S8564, Selleckchem) was injected in awake rats (see supplemental information).

Local field potential recordings

Recording electrodes were placed in the mPFC and within 0.5mm of the AAV injection site in the pPVT (see supplemental information).

Golgi Staining

Golgi staining was performed using the Rapid GolgiStain™ kit (FD Neurotechnologies, PK401) per manufacturer instructions. Neurons were traced at 100x using Neurolucida. In the pPVT and aPVT, typically 2 neurons in 4 sections were traced for each rat. Group sizes indicate the number of rats in each group (see supplemental information).

Attentional Set Shifting Task

Following 5 consecutive days of restraint, iAAV-scramble and iAAV-Arc rats underwent the attentional set shifting task (AST) as in (40) to assess the role of pPVT Arc knockdown on PFC function following restraint. Rats were tested in side discrimination, side reversal, and light discrimination phases. The number of trials required to reach criterion (8 consecutive correct trials), time required to reach criterion, number of omissions, and number of errors during the test phase were recorded and analyzed using MATLAB (see supplemental information).

Statistical Analyses

Student’s t-tests, one-way ANOVAs, ordinary two-way ANOVAs, and repeated measures two-way ANOVAs were performed in GraphPad Prism 7 (see supplemental information).

RESULTS

Restraint induces Arc expression in the pPVT but not in the anterior (aPVT).

Consistent with previous results (4, 14, 40), we confirmed that plasma ACTH and corticosterone concentrations were reduced during the 5^th day of restraint (Supplemental Figure S1A-D). By the 3^rd restraint, struggle duration was significantly lower than on day 1 (Figure S1E). The density of Arc-immunoreactive (IR) neurons was increased in the pPVT (Figure 1A,B), but not the aPVT (Figure 1C), of rats restrained for 1 and 5 days compared to non-restrained controls. The density of c-Fos-IR cells is also increased in the pPVT, but not the aPVT, of rats restrained for 1 or 5 days compared to non-restrained controls (Supplemental Figure 1F-H). Together, these results confirm that restraint induces neuronal activity in the pPVT.

Figure 1.

Chemogenetic inhibition of the pPVT impairs habituation to repeated restraint. A) pPVT Images of and quantification of Arc-IR neurons in the B) pPVT and C) aPVT (no restraint n = 8, 1day n = 6, 5day n = 8). D) Timeline for chemogenetic inhibition of the pPVT. E) Image of HA tag in hM4D-expressing pPVT neurons. Plasma ACTH in vehicle- (n = 6) and CNO-treated rats (n = 8) during restraint on F) day1 and G) day5. H) Integrated plasma ACTH in vehicle- (n = 6) and CNO-treated rats (n = 8). Plasma corticosterone in vehicle- (n = 6) and CNO-treated rats (n = 8) during restraint on I) day1 and J) day5. K) Integrated plasma corticosterone in vehicle- (n = 6) and CNO-treated rats (n = 8). L) Struggle duration in vehicle- and CNO-treated rats (n = 7/group). M) Density of pPVT Arc-IR neurons in vehicle- and CNO-treated rats (no restraint vehicle n = 7, no restraint CNO n = 6, 1day vehicle n = 7, 1day CNO n = 5, 5day n = 8/treatment). Bars represent mean ± SEM. B) *p < 0.05 compared to non-restrained controls, Tukey’s multiple comparisons, one-way ANOVA. F-L, *p<0.05, **p<0.01, ***p<0.001, #p<0.10, Sidak’s multiple comparisons, repeated measures two-way ANOVA. M, ***different from all groups except each other, Tukey’s multiple comparisons, ordinary two-way ANOVA.

Chemogenetic inhibition of the pPVT impairs habituation.

In otherwise naïve rats, CNO alone did not affect plasma concentrations of ACTH or corticosterone at any timepoint during days 1 or 5 of restraint (Supplemental Figure 2A-F) nor did CNO affect struggle duration (Supplemental Figure 2G). In rats expressing hM4D in the pPVT (Figure 1D,E), CNO treatment increased plasma ACTH concentrations at the 60min timepoint on day 1 (Figure 1F) and the 30- and 60-min timepoints on day5 (Figure 1G) compared to vehicle-treated controls. Integrated ACTH was also increased on days 1 and 5 in CNO-treated rats (Figure 1H). CNO treatment did not affect plasma corticosterone concentrations during day1 of restraint (Figure 1I). However, compared to vehicle-treated controls, CNO-treated rats displayed increased plasma corticosterone concentrations 60min following restraint onset on day5 of restraint (Figure 1J). Vehicle-treated, but not CNO-treated, rats displayed reduced integrated corticosterone on day5 compared to day1. CNO-treated rats displayed increased integrated plasma corticosterone concentrations compared to vehicle-treated controls on day5 (Figure 1K). Vehicle-treated, but not CNO-treated, rats displayed reduced struggle duration on day5 compared to day1, resulting in increased struggle duration in CNO-treated rats on day5 (Figure 1L). Indeed, CNO treatment in hM4D-expressing rats chemogenetically inhibited the pPVT as vehicle-, but not CNO-treated, rats displayed increased densities of Arc+ pPVT neurons 60min following the onset of 1 or 5 restraints (Figure 1M). Together, our findings indicate that chemogenetic inhibition of the pPVT impairs neuroendocrine and behavioral stress habituation.

Stress-induced increases in Arc expression in the pPVT are necessary for habituation.

We investigated whether Arc in the pPVT regulates habituation to repeated stress by using siRNA directed toward Arc (iAAV-Arc) to knockdown Arc expression (Figure 2A). Arc-IR neuron density was significantly reduced in iAAV-Arc rats compared to iAAV-scramble controls following 5 consecutive daily restraints (Figure 2B,C), thereby confirming siRNA efficacy. In iAAV-scramble controls, ACTH concentrations were significantly reduced 15- and 30-min following restraint onset on day 5 compared to day 1 (Figure 2D). In iAAV-Arc rats, ACTH concentrations were similar at all timepoints on days 1 and 5 of restraint (Figure 2E). iAAV-scramble, but not iAAV-Arc, rats displayed reduced integrated ACTH on day 5 compared to day 1 of restraint (Figure 2F). In iAAV-scramble controls, corticosterone concentrations were significantly reduced 60min following restraint onset on day5 compared to day1 (Figure 2G). In iAAV-Arc rats, corticosterone concentrations were similar at all timepoints on days 1 and 5 of restraint (Figure 2H). iAAV-scramble rats displayed reduced integrated corticosterone on day5 compared to day1 of restraint but this decrease was not observed in iAAV-Arc rats (Figure 2I). iAAV-scramble rats displayed reduced struggle duration on day5 of restraint compared to day1 but iAAV-Arc rats did not show this reduction (Figure 2J). Together, these results indicate that restraint-induced Arc in the pPVT is necessary for behavioral and neuroendocrine habituation to repeated restraint stress.

Figure 2.

Arc in the pPVT regulates habituation to repeated restraint. A) Timeline for investigating habituation in iAAV-scramble and iAAV-Arc rats. B) Images of Arc in the pPVT of iAAV-scramble and iAAV-Arc rats. C) Density of Arc-IR neurons in the pPVT of iAAV-scramble (n = 10) and iAAV-Arc rats (n = 8) following 5 days of restraint. Plasma ACTH concentrations at 0, 15, 30, and 60min following the onset of 30min restraint on days 1 and 5 in D) iAAV-scramble (n = 10) and E) iAAV-Arc rats (n = 8). F) Integrated plasma ACTH concentrations in iAAV-scramble (n = 10) and iAAV-Arc rats (n = 8) from 0–60min following restraint onset on days 1 and 5. Plasma corticosterone concentrations at 0, 15, 30, and 60min following the onset of 30min restraint on days 1 and 5 in G) iAAV-scramble (n = 10) and H) iAAV-Arc rats (n = 8). I) Integrated plasma corticosterone concentrations in iAAV-scramble (n = 10) and iAAV-Arc (n = 8) rats from 0–60min following restraint onset on days 1 and 5. J) Struggle duration during the first 15min of restraint in iAAV-scramble and iAAV-Arc rats on days 1 and 5 (n = 8/group). Bars represent mean ± SEM. For C, ****p < 0.0001, Student’s unpaired two-tailed t-test. For D-J, *p < 0.05, **p < 0.01, ***p < 0.001, Sidak’s multiple comparisons following repeated measures two-way ANOVA. Horizontal bars indicate differences between groups.

Arc knockdown in the pPVT prevents stress-induced spinogenesis.

We investigated whether Arc regulates structural plasticity in the pPVT (Figure 3A). Arc knockdown had no effect on spine density in non-restrained rats (Figure 3B,C). 24 hours following a 5^th consecutive restraint, iAAV-Arc rats displayed reduced dendritic spine densities in the pPVT 80 and 100 µm distal from the soma (Figure 3D). Collapsing across distance from soma, mean pPVT spine density was increased by repeated restraint in iAAV-scramble, but not in iAAV-Arc rats, resulting in lower spine densities in iAAV-Arc rats compared to scramble controls on day5 (Figure 3E). Mean dendrite length in the pPVT was not affected by restraint or Arc knockdown (Figure S3A-C), indicating that effects on spine density cannot be solely attributed to changes in dendritic length. Together, these data indicate that Arc is necessary for restraint-induced increases in pPVT spine density.

Figure 3.

Arc knockdown prevents stress-induced spine densities and dendritic remodeling in the pPVT. A) Timeline for investigating structural plasticity of pPVT neurons in iAAV-scramble and iAAV-rats. B) Images of dendritic stubby spines in pPVT neurons. Mean spine density at increasing 10 µm radial increments in the pPVT of C) non-restrained controls (iAAV-scramble n = 8, iAAV-Arc n = 6) and D) 24 h following a 5^th daily restraint (n = 7/group). E) Mean pPVT spine density over all distances from the soma. F) Images of dendritic complexity in the pPVT. Mean number of intersections at increasing 10 µm radial increments from the soma in the pPVT of G) non-restrained rats (iAAV-scramble n = 7, iAAV-Arc n = 6) and H) 24 h following a 5^th daily restraint (n = 7/group). I) Mean dendritic intersections averaged over all distances from the soma (no restraint iAAV-scramble n = 7, no restraint iAAV-Arc n = 6, 5day restraint n = 7/group).J) Timeline for investigating habituation in vehicle- and MK-8931-treated rats. K) Mean spine density of pPVT neurons averaged over all distances from the soma in vehicle- (n = 8) and MK-8931-treated rats (n = 7). L) Integrated plasma ACTH concentrations in vehicle- and MK-8931-treated rats from 0–60min following restraint onset (n = 7/group). M) Struggle duration during the first 15min of restraint in vehicle- (n = 8) and MK-8931-treated rats (n = 7). Bars represent mean ± SEM. *p < 0.05, **p < 0.01, #p < 0.10. C, D, G, H, L, and M, Sidak’s multiple comparisons, repeated measures two-way ANOVA. E and I, Tukey’s multiple comparisons, ordinary two-way ANOVA. K, Student’s unpaired t-test.

Arc knockdown in the pPVT prevents reductions in dendritic complexity caused by stress.

Arc knockdown did not affect dendritic complexity in pPVT neurons in the absence of stress as non-restrained iAAV-scramble and iAAV-Arc rats displayed similar numbers of dendritic intersections at all distances from the soma (Figure 3F,G). Following restraint, iAAV-Arc rats display trends for increased dendrite intersections 60 and 70 µm from the soma in pPVT neurons compared to iAAV-scramble controls (Figure 3H). Collapsing across distance from the soma, restraint reduced dendritic intersections in the pPVT of iAAV-scramble rats but not in iAAV-Arc rats (Figure 3I). Taken together, these findings indicate that stress reduces dendritic complexity in pPVT neurons and that this reduction is dependent on Arc. These structural effects are specific to pPVT neurons as aPVT neurons exhibited no restraint- or Arc-mediated effects on dendritic spine density (Figure S3D-F), dendrite intersections (Figure S3G-I), or dendrite length (Figure S3J-L).

Inhibition of spine formation in the pPVT impairs habituation.

We next investigated whether the formation of new spines in the pPVT is necessary for habituation (Figure 3J). We used the β-site amyloid precursor protein cleavage enzyme 1 (BACE1) inhibitor MK-8931, which inhibits the formation of new dendritic spines but not the stability of existing ones (41). The use of more common pharmacological inhibitors of spinogenesis, such as latrunculin A, could not be used as they inhibit actin dynamics (42) and therefore actin-mediated localization of Arc to the synapse (43, 44). This would be problematic as it would be impossible to determine whether spinogenesis or other Arc-mediated processes regulate habituation. Following restraint, dendritic spine density in the pPVT was reduced in MK-8931-treated rats compared to vehicle-treated controls (Figure 3K, Figure S3M). MK-8931 treatment did not affect dendritic branching (Supplemental Figure 3N,O) or dendrite length (Figure S3P,Q) in the pPVT. MK-8931 treatment impaired habituation as vehicle-treated, but not MK-8931-treated, rats exhibited reduced plasma ACTH concentrations (Figure 3L, Figure S3R,S) and struggle duration (Figure 3M) on day 5 of restraint compared to day 1 although plasma corticosterone was not impacted (Figure S3T-V). These findings indicated that restraint-induced spinogenesis in the pPVT is necessary for habituation.

Arc knockdown in the pPVT reduces network activity in the pPVT and reduces coherent activity between the pPVT and mPFC in repeatedly stressed rats

The mPFC inhibits HPA activity (45, 46) and is the primary cortical target of the pPVT (26, 47–49). We hypothesized that Arc in the pPVT regulates habituation, at least in part, via projections to the mPFC. Power spectral density (PSD) was analyzed in iAAV-scramble and iAAV-Arc rats to assess network activity in the delta (2–4 Hz), theta (4–9 Hz), alpha (9–12 Hz), beta (12–20 Hz), and gamma (20–40 Hz) frequency bands before, during, and after restraint (Figure 4A). Delta is associated with quiet wakefulness whereas theta and alpha frequency ranges are associated with vigilance and information processing required for attention (50–57). Compared to iAAV-scramble rats, iAAV-Arc rats displayed reduced delta power in the pPVT during baseline and during the 5–10min bin following restraint on day5 (Figure 4B, Figure S4D). Compared to iAAV-scramble controls, iAAV-Arc rats also displayed reduced alpha power during the 5–10min bin of restraint on day5 (Figure S4E). Arc knockdown did not alter power in the pPVT in the theta, beta, or gamma frequency ranges at any timepoints (Figure S4F-H). Arc knockdown in the pPVT did not alter power in any frequency ranges in the mPFC at any time points (Figure S4I-N). Significant virus effects were observed for pPVT-mPFC coherence in the delta, theta, and alpha frequency ranges, indicating that iAAV-Arc rats displayed reduced overall coherence across all timepoints in these ranges compared to iAAV-scramble controls. Arc knockdown had no effect on pPVT-mPFC coherence during the day 1 baseline recording. During the day 5 baseline recording, iAAV-Arc rats displayed reduced pPVT-mPFC coherence in the delta, theta, and alpha (Figure 4C-F) frequency ranges compared to iAAV-scramble controls. During the 0–5min bin following day1 restraint, iAAV-Arc rats displayed reduced pPVT-mPFC coherence in the delta range (Figure 4D) and a trend for reduced coherence in the theta range (Figure 4E) compared to iAAV-scramble controls. Following day 5 restraint, iAAV-Arc rats displayed reduced pPVT-mPFC coherence in the theta range 0–5min following restraint a trend for reduced theta coherence 5–10min following restraint. Arc knockdown in the pPVT did not affect pPVT-mPFC coherence at any time point in the beta or gamma frequency ranges (Figure S4O,P). Together, these findings indicate that Arc expression in the pPVT regulates population activity in the pPVT important for delta and alpha frequency bands. Arc in the pPVT also promotes coherent neural activity between the pPVT and the mPFC during habituation.

Figure 4.

Arc knockdown in the pPVT disrupts pPVT-mPFC coherence. A) Timeline for investigating LFPs in the pPVT and mPFC in iAAV-scramble and iAAV-Arc rats. B) Delta PSD percentages in the pPVT of iAAV-scramble and iAAV-Arc rats during baseline, restraint, and recovery on days 1 and 5 of restraint (iAAV-scramble n = 6, iAAV-Arc n = 8). C) Images of mean pPVT-mPFC coherence value heat maps from 0–50 Hz in the pPVT of iAAV-scramble and iAAV-Arc rats during a 5 min baseline, three 5-min restraint bins, and two 5-min recovery bins on days 1 and 5 of restraint. Warmer colors represent higher coherence. Coherence values in the D) delta, E) theta, and F) alpha frequency ranges in the pPVT of iAAV-scramble and iAAV-Arc rats during baseline, restraint, and recovery on days 1 and 5 of restraint (iAAV-scramble n = 6, iAAV-Arc n = 8). Bars represent mean ± SEM. For B and D-F, *p < 0.05, **p < 0.01, ***p < 0.001, # p < 0.10, Fisher’s least significant difference following two-way ANOVA. Horizontal bars indicate differences between groups.

Chemogenetic inhibition of mPFC-projecting pPVT neurons impairs behavioral but not neuroendocrine habituation.

We next determined whether mPFC-projecting pPVT neurons impair habituation. CAV2-Cre, which transduces axon terminals to retrogradely express Cre recombinase, was injected into the mPFC. The pPVT was injected with AAV8-hSyn-DIO-hM4D-HA-mCherry, which expresses the inhibitory hM4D DREADD in a Cre-dependent manner (17) (Figure 5A). No differences in the density of mCherry+ pPVT neurons were observed in vehicle- and CNO-treated groups (Figure S5A). Sixty min following the onset of the 5^th restraint, the numbers of Arc+/mCherry+ pPVT neurons (Figure 5B,C), percentage of Arc+/mCherry+ pPVT neurons (Figure S5B), and number of c-Fos+/mCherry+ pPVT neurons (Figure S5C,D) were reduced in CNO-treated rats compared to vehicle-treated controls. Vehicle-treated rats displayed reduced struggle duration during the 5^th restraint compared to the 1^st restraint but CNO-treated rats did not show this reduction (Figure 5D). CNO treatment did not affect plasma ACTH or corticosterone concentrations (Figure S5E-J). These results indicate that mPFC-projecting pPVT neurons regulate behavioral, but not neuroendocrine, habitation.

Figure 5.

The pPVT projections to the mPFC regulate struggle habituation and mPFC function. A) Timeline investigating chemogenetic inhibition of mPFC-projecting pPVT neurons on habituation. B) Images of mCherry, Arc, and merged images from the pPVT of rats injected with CAV2-Cre in the mPFC and with AAV8-hSyn-DIO-hM4D-HA-mCherry in the pPVT after 5 restraints. C) The number of Arc+/mCherry+ neurons in the pPVT (n = 8/group). D) Struggle duration on days 1 and 5 of restraint vehicle- and CNO-treated rats (n = 8/group). E) Timeline for investigating AST performance in iAAV-scramble and iAAV-Arc rats. F) Descriptive graphic of AST methods. The G) number of trials and H) time required to reach criterion (8 consecutive correct trials) during the side discrimination phase of the AST (n = 6/group). The I) number of trials and J) time required to reach criterion during the side reversal phase of the AST (n = 6/group). The K) number of trials and L) time required to reach criterion during the light discrimination phase of the AST (n = 6/group). M) Figure summarizing findings. Stress increases pPVT activity, which induces Arc. Arc-mediated increases in dendritic spine density facilitate habituation by sensitizing stress-activated input to pPVT neurons, particularly those projecting to the mPFC. pPVT Arc promotes pPVT-mPFC coherence and regulates behavioral habituation and cognitive flexibility. pPVT neurons that do not project to the mPFC may negatively regulate the HPA axis. Bars represent mean ± SEM. For C and I-L, *p < 0.05, ***p < 0.001, ****p < 0.0001, #p < 0.10, unpaired, two-tailed Student’s t-test. For D, *p < 0.05, Sidak’s multiple comparisons, repeated measures two-way ANOVA. Horizontal bars indicate differences between groups.

Restraint-induced Arc in the pPVT promotes mPFC-dependent cognitive flexibility.

We hypothesized that pPVT Arc knockdown would impair performance in the attentional set shifting task (AST) following restraint (Figure 5E,F), particularly in the mPFC-dependent light discrimination phase (58), as Arc knockdown disrupted coherent activity with the mPFC. We confirmed that Arc knockdown in the pPVT impaired habituation of struggle duration (Figure S5K). Arc knockdown did not affect the number of trials (Figure 5G) or time required to reach criterion (Figure 5H) in the side discrimination phase. In the side reversal phase, iAAV-Arc rats displayed an increased number of trials (Figure 5I) and a trend towards increased time (Figure 5J) required to reach criterion compared to iAAV-scramble rats. In the light discrimination phase, iAAV-Arc rats displayed an increased number of trials (Figure 5K) and increased time (Figure 5L) required to reach criterion compared to iAAV-scramble rats. Arc knockdown in the pPVT did not affect the percentage of omissions (Figure S5L-N) or number of errors (Figure S5O-Q) in the AST. Behavioral habituation impairments, as assessed by the ratio of time spent struggling on day5 compared to time spent struggling on day1, did not correlate with AST impairments as assessed by trials to reach criterion in the in the side discrimination (Figure S5R) and side reversal (Figure S5S) phases. However, behavioral habituation impairments correlated with AST impairments in the light discrimination phase (Figure S5T). These findings indicate that as male rats habituate to repeated restraint stress, induction of Arc in the pPVT contributes to improvements in subsequent mPFC-mediated cognitive flexibility. In sum, our findings indicate that restraint-induced Arc regulates habituation by increasing spine density in the pPVT. Increased output of mPFC-projecting pPVT neurons regulates behavioral habituation and cognitive flexibility, whereas pPVT projections to other brain regions regulate neuroendocrine habituation (Figure 5M).

Discussion

Here, we demonstrate that neuronal activity in the pPVT produced by repeated stress induces the expression of Arc, which promotes habituation by increasing dendritic spine densities in the pPVT. Knockdown studies demonstrate that pPVT Arc Is necessary for habituation, stress-induced increases in spine density, and reduced dendritic complexity in pPVT neurons. Thus, Arc mediates stress-induced reformatting of pPVT neuron inputs. Stress-induced spinogenesis contributes to habituation as pharmacological inhibition of spine formation in the pPVT using MK-8931 was sufficient to impair habituation. Increased spine density in the pPVT may sensitize pPVT neurons to be more responsive to specific afferents activated by restraint. Arc-mediated increases in pPVT spine densities may increase activity of projections to the mPFC as pPVT Arc knockdown reduced pPVT-mPFC coherence. We hypothesize that pPVT-mPFC coherence is an important mechanism underlying habituation as pPVT Arc knockdown impaired PFC-dependent cognitive flexibility and chemogenetic inhibition of mPFC-projecting pPVT neurons impaired behavioral habituation. These findings are important as they are, to best of our knowledge, the first to identify 1) a mechanism of neuronal plasticity in the pPVT that promotes adaptation to repeated stress and 2) that restraint-induced pPVT Arc induction is important for cortically-dependent cognitive flexibility.

While restraint-induced expression of other immediate early genes in the pPVT may also contribute to habituation, we focused on Arc because of its well-established role in regulating structural plasticity (29, 59, 60), a property of neurons that is also tightly regulated by stress (61, 62). We found that Arc is necessary for restraint-induced spinogenesis in the pPVT. We provide evidence that spinogenesis is necessary for habituation as inhibiting spinogenesis with MK-8931 impaired habituation, although other effects caused by BACE1 inhibition cannot be completely ruled out. In addition to increasing spine density, restraint-induced Arc also mediated reductions in dendritic branching of pPVT neurons. Stress-mediated reductions in dendritic branching have been described in the pPVT, mPFC, and hippocampus (61, 63, 64). In the hippocampus, NMDA receptors are necessary for reductions in the dendritic complexity caused by stress (65). Arc is a key link between NMDA receptor activity and structural remodeling. Therefore, Arc contributes to stress-mediated reductions in dendritic complexity in the pPVT and perhaps in other regions as well.

We report that in the pPVT of iAAV-Arc rats, power in the delta frequency range is reduced during the baseline recording and during recovery following restraint on day 5 compared to iAAV-scramble controls. Reduced delta power may reflect a decrease in quiet wakefulness (50–52, 57) resulting from anticipatory activity prior to restraint and impaired return to quiet wakefulness following restraint, respectively. Power in the alpha range was also reduced during restraint on day 5 in iAAV-Arc rats. Alpha and theta frequencies are associated with vigilance, information processing and the filtering of sensory stimuli required for optimal attention (53–56). Attention is essential for establishing restraint stress as not being physically threatening, a key aspect that makes habituation to these stressors adaptive (2). Together, reduced delta and alpha power in the pPVT of iAAV-Arc rats suggests a heightened state of arousal that may impair the ability to attend to restraint stress.

We hypothesized that pPVT projections to the mPFC regulate habituation. The mPFC is the primary cortical target of the pPVT (23, 26, 48) and inhibits activity in key stress-related brain regions including the amygdala (66, 67) and HPA axis (45, 46). pPVT-mPFC coherence was reduced in the delta, theta, and alpha frequency ranges during certain baseline and/or recovery bins in iAAV-Arc rats compared to iAAV-scramble controls. These reductions in coherence suggest that functional connectivity between the pPVT and mPFC may be impaired as the PVT directly projects to the mPFC (26, 47–49). It is worth noting that other structures, such as the locus coeruleus, may modulate both pPVT (68) and mPFC activity in parallel (69). Regardless, reductions in synchronous oscillations between the pPVT and mPFC may impair the ability of the pPVT to influence mPFC activity. To directly examine whether pPVT-mPFC projections regulate habituation, we demonstrated that inhibition of mPFC-projecting pPVT neurons impaired behavioral habituation. It will be important in future studies to investigate the effects of repeated chemogenetic inhibition of pPVT neurons in non-stressed rats as repeated pPVT inhibition might increase baseline plasma ACTH and corticosterone or have other relevant effects in the absence of stress. Together, these findings demonstrate that pPVT Arc promotes pPVT-mPFC coherence and that mPFC-projecting pPVT neurons regulate behavioral habituation during repeated restraint.

We then examined whether Arc in the pPVT regulates mPFC-mediated behaviors. In previous work, we showed improved cognitive flexibility in the AST, a PFC-dependent task, in male rats that have habituated compared to non-stressed controls (40). Based on our electrophysiology studies, we hypothesized that pPVT Arc may contribute to mPFC-dependent behavior in the AST. We observed that restrained iAAV-Arc rats displayed impairments in the side reversal and light discrimination phases, which are regulated by the OFC and the mPFC, respectively (58). Impairments in the side reversal phase suggest that some pPVT neurons may project to the OFC or to OFC-projecting mPFC neurons, and/or that the side reversal phase of the AST is not exclusively dependent on the OFC. To the best of our knowledge, our results are the first to identify a role of the pPVT in contributing to PFC-dependent cognitive flexibility.

Arc induction by stress and Arc actions are specific to the posterior division of the PVT. In the aPVT, restraint did not alter Arc expression or affect structural plasticity. This specificity is important because aPVT projections are widespread (21–23) whereas projections from the pPVT are more exclusive and, in general, target stress-related brain regions like the mPFC and amygdala (17, 18, 20, 26). The differences in efferent projections from aPVT and pPVT neurons confer different functions and uniquely position the posterior division to regulate habituation to repeated stress, consistent with previous findings (14).

The findings presented here are the first to identify Arc as a key mediator of structural plasticity in the pPVT and for how individuals adapt to repeated stress. Much less is known about the function of Arc in the thalamus compared to other brain regions, like the hippocampus. Our results may provide a foundation for understanding other plasticity-dependent processes regulated by thalamic circuits. Chemogenetic inhibition of the subpopulation of pPVT neurons projecting to the mPFC attenuated behavioral, but not neuroendocrine, habituation. This dichotomy suggests that pPVT projections to brain regions other than the mPFC (e.g. the BNST) may be responsible for pPVT regulation of the HPA axis. Together, our findings provide novel insights into the molecular and network mechanisms underlying habituation to repeated stress, a phylogenetically conserved property of the stress response.

KEY RESOURCES 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	Rabbit anti-Arc (polyclonal)	Synaptic Systems, PMID:21456007	156003, RRID:AB_887694
Antibody	Rabbit anti-cFos (monoclonal)	Santa Cruz, PMID:28467927	sc-166940, RRID:AB_10609634
Antibody	Mouse anti-Arc (monoclonal)	Santa Cruz, PMID:26875623	sc-17839, RRID:AB_626696
Antibody	Goat anti-cFos (polyclonal)	Abcam	ab156802, RRID:AB_2747692
Antibody	Rabbit anti-mCherry (polyclonal)	Abcam, PMID:27477019	ab167453, RRID:AB_2571870
Antibody	Rabbit anti-HA (monoclonal)	Cell Signaling, PMID:28193319	C29F4, RRID:AB_10693385
Antibody	donkey anti-rabbit Alexa Fluor ® 647 (polyclonal)	Abcam, PMID:30726741, PMID:30830967	ab150075, RRID:AB_2752244
Antibody	donkey anti-mouse Alexa Fluor ® 647 (polyclonal)	Abcam	ab150107, RRID:AB_2890037
Antibody	donkey anti-mouse Alexa Fluor ® 594 (polyclonal)	Abcam, PMID:29803967, PMID:31968248	ab150108, RRID:AB_2732073
Antibody	donkey anti-rabbit Alexa Fluor ® 594 (polyclonal)	Abcam, PMID:30926749, PMID:32801156	ab150076, RRID:AB_2782993
Chemical Compound or Drug	MK-8931 (Verubecestat)	Selleckchem	S8564
Commercial Assay Or Kit	ACTH double antibody radioimmunoassay kit	MP Biomedicals	SKU:0710610-CF
Commercial Assay Or Kit	corticosterone double antibody radioimmunoassay kit	MP Biomedicals	SKU:0712010-CF
Commercial Assay Or Kit	Rapid GolgiStain™ kit	FD Neurotechnologies	PK401
Organism/Strain	Male Sprague-Dawley rats	Charles River laboratories	RRID:MGI:5651135
Recombinant DNA	iAAV8-Arc	Applied Biological Materials	iAAV06494008, serotype 8, rat
Recombinant DNA	iAAV8-scramble control	Applied Biological Materials	iAAV01508, serotype 8, rat
Recombinant DNA	AAV8-hSyn-hM4D-HA-IRES-mCherry	University of North Carolina Vector Core	N/A
Recombinant DNA	AAV8-hSyn-DIO-hM4D-HA-mCherry	University of North Carolina Vector Core	N/A
Recombinant DNA	CAV2-Cre	Institut de Génétique Moléculaire de Montpellier, University of Montpellier	N/A
Other