Skip to main content
NCBI home page
Neuroscience Bulletin logoLink to Neuroscience Bulletin
. 2020 Nov 12;37(2):229–241. doi: 10.1007/s12264-020-00603-6

Projections from Infralimbic Cortex to Paraventricular Thalamus Mediate Fear Extinction Retrieval

Yan Tao 1,#, Cheng-Yun Cai 1,#, Jia-Yun Xian 1,#, Xiao-Lin Kou 1, Yu-Hui Lin 1, Cheng Qin 1, Hai-Yin Wu 1, Lei Chang 1, Chun-Xia Luo 1, Dong-Ya Zhu 1,2,3,✉
PMCID: PMC7870747  PMID: 33180308

Abstract

The paraventricular nucleus of the thalamus (PVT), which serves as a hub, receives dense projections from the medial prefrontal cortex (mPFC) and projects to the lateral division of central amygdala (CeL). The infralimbic (IL) cortex plays a crucial role in encoding and recalling fear extinction memory. Here, we found that neurons in the PVT and IL were strongly activated during fear extinction retrieval. Silencing PVT neurons inhibited extinction retrieval at recent time point (24 h after extinction), while activating them promoted extinction retrieval at remote time point (7 d after extinction), suggesting a critical role of the PVT in extinction retrieval. In the mPFC-PVT circuit, projections from IL rather than prelimbic cortex to the PVT were dominant, and disrupting the IL-PVT projection suppressed extinction retrieval. Moreover, the axons of PVT neurons preferentially projected to the CeL. Silencing the PVT-CeL circuit also suppressed extinction retrieval. Together, our findings reveal a new neural circuit for fear extinction retrieval outside the classical IL-amygdala circuit.

Electronic supplementary material

The online version of this article (10.1007/s12264-020-00603-6) contains supplementary material, which is available to authorized users.

Keywords: Paraventricular thalamus, Infralimbic cortex, Medial prefrontal cortex, Amygdala, Fear extinction retrieval, Neural circuit, Post-traumatic stress disorder

Introduction

Fear can be elicited by perceived threats. Proper behavioral responses to threats enable all animals, including humans, to reduce the risk of danger and increase the chance of survival [1]. When individuals are unable to overcome acquired fear, they may develop fear-related disorders [2]. Post-traumatic stress disorder (PTSD) is a highly prevalent fear-related disorder, costing affected individuals a lot [3]. Fear extinction is the first-line treatment for PTSD patients under the term “exposure therapy” [4]. Prolonged Exposure (PE) and Virtual Reality Exposure (VRE) have been highly effective in recent years [5]. Besides, pharmacological interventions, such as paroxetine, venlafaxine and fluoxetine, are also used to treat PTSD but their effects are subject to many limitations [6]. However, learned extinction memories are often fragile and extinguished fear relapses under a number of circumstances [7, 8]. Impaired retrieval of extinction memory may contribute to the return of this fear. Thus, a better understanding of the circuit mechanism underlying extinction retrieval will benefit the treatment of PTSD.

Classical experimental extinction is based on Pavlovian fear conditioning, in which a conditioned stimulus (CS, such as a tone or context) is paired with a naturally aversive unconditioned stimulus (US, such as an electric footshock). Presentation of the CS will evoke a conditioned response (CR) after fear conditioning. Fear extinction learning refers to the gradual, within-session decrements of learned fear with repeated exposure to a stimulus or place, when the CS no longer predicts the US [9]. Extinction consists of three phases: acquisition, consolidation, and retrieval. Extinction retrieval refers to the recall of the learned extinction memory after a delay. Presentation of the extinguished CS may elicit a low CR in normal retrieval. Poor retrieval of extinction results from spontaneous recovery (SR), renewal, reinstatement, and pathology [10]. SR refers to the reappearance of a strong CR at a long interval after extinction training. The degree of CR recovery depends on the length of the retention interval [11].

Neural circuits underlying fear extinction have been well studied over the years. The infralimbic (IL) subregion of the medial prefrontal cortex (mPFC) is responsible for fear extinction. IL neurons are strongly activated during extinction training and retrieval [12]. Evidence has shown that the IL mediates fear extinction by inhibiting the medial division of the central amygdala (CeM) via intercalated cells [13]. The CeM mediates the output of the fear response. The periaqueductal grey (PAG), a target of the CeM in the midbrain, is responsible for fear expression [1]. But, increasing evidence suggests that extinction is not only mediated by the IL-amygdala pathway; other circuits could also be implicated, such as prefrontal projections to the thalamic nucleus reuniens [14].

The paraventricular thalamus (PVT), an important part of thalamus below the third ventricle, is composed of glutamatergic neurons and lacks GABAergic interneurons [15]. It serves as a hub because it not only relays information from many areas such as the mPFC and brainstem [16], but also sends dense projections to the nucleus accumbens, bed nucleus of the stria terminalis, and central nucleus of the amygdala (CeA) [17]. Recent experimental evidence has shown that the PVT plays a critical role in wakefulness [18], food intake [19] and associative learning [20]. However, little is known about its role in extinction memory and the relationship between the PVT and the IL-amygdala circuit. Here, we found that the activity of PVT neurons was necessary for fear extinction retrieval. Silencing the activity of the PVT suppressed normal extinction retrieval in the short term, while stimulating PVT activity facilitated retrieval in the long term. Moreover, pharmacogenetically silencing IL-PVT and PVT-lateral division of central amygdala (CeL) circuits also inhibited normal extinction retrieval. Taken together, we reveal an IL-PVT-CeL circuit for mediating extinction retrieval and provide new evidence for the existence of prefrontal-thalamic circuits for fear extinction.

Materials and Methods

Animals

Young male adult C57BL/6 mice (6-7 weeks old) were purchased from Mode Animal Research Center of Nanjing University (Nanjing, China). The mice were maintained at a controlled temperature (20 ± 2°C) and group housed (12-h light/dark cycle) with access to food and water ad libitum. Every effort was made to minimize the number of animals used and their suffering. All experiment were approved and conducted in accordance with the guidelines and regulation designed by the Institutional Animal Care and Use Committee of Nanjing Medical University.

Viruses and Drugs

AVV2/9-CaMKIIα-hM3D(Gq)-mCherry-WPRE-pA (PT-0049), AVV2/9-CaMKIIα-hM4D(Gi)-mCherry-WPRE-pA (PT-0017), and rAAV2/R-hsyn-EGFP-WPRE-pA (PT-0241) were from Brain VTA (Wuhan, China). AAV-CMV-EGFP, AAV9-CaMKIIα-EYFP (AVV00042), and AAV9-CaMKIIα-eNpHR3.0-EYFP (AVV00013) were from Genechem (Shanghai, China). Clozapine-n-oxide (CNO) (#4936) was from Tocris Bioscience (Missouri, USA).

Stereotaxic Microinjections and Implantation of Cannulae

Mice were anesthetized with isoflurane and fixed into a stereotaxic frame (KOPF instruments, Tujunga, CA, USA). Viral solution was microinjected into brain areas using microinjection pump (WPI, Sarasota, FL, USA) with a glass pipette at 2 nL/s, and then the pipette was left for 15 min to ensure restriction of the virus to the required area. We infused 0.3 μL of virus solution unilaterally into the PVT and bilaterally into the IL. Four to six weeks later, mice were used in experiments. A 26-guage guide cannula (RWD Life Science, Shenzhen, China) was used for implantation into the PVT and CeL. Immediately after insertion into the target region, we fixed cannula by using dental cement. After implantation surgery, mice were raised alone for one week. Cannulae were implanted bilaterally except for the PVT. After virus microinjection and cannulae implantation surgery, mice were placed on the warm blanket until they were awake. All coordinates for the virus microinjection and cannulae implantation were: AP + 1.9 mm, ML ± 0.3 mm, DV -3.0 mm for the IL; AP -1.3 mm, ML 0.0 mm, DV -3.1 mm for the PVT; and AP -1.2 mm, ML ±2.6 mm, DV -4.5 mm for CeL.

Drug Delivery

We dissolved CNO in sterile saline. Mice received an intraperitoneal injection of CNO (2 mg/kg) at 0.2 mg/mL or saline. CNO at 0.5 mmol/L was used for brain microinjection [21]. The infusion rate (6 nL/s) was controlled by a microinjection pump (WPI, Sarasota, FL, USA) and the volume of solution was 0.6 μL. All mice were handled for 3 min before drug delivery. Then they were restricted gently by hand and a thin 34-guage metal needle (RWD Life Science, Shenzhen, China) was inserted into the guide cannula. The needle was left for another 5 min after microinjection to promote diffusion of the drug.

Fear Conditioning/Extinction Paradigm

We used an auditory fear conditioning/extinction paradigm consisting of auditory fear conditioning, fear memory recall, massed extinction training, and a retrieval test. Mice were habituated to handling for three days before the paradigm, randomly assigned to experimental groups, and each training or testing cohort was counterbalanced to equally represent each group in the cohort. Freezing behavior (as a percentage of a block of time or trial[s]) served as the dependent measure of fear in all training and test sessions. On day 1, mice were placed in the conditioning context (Context A, 30 cm × 30 cm × 30 cm) for 2 min to explore freely in order to assess baseline behaviors. After habituation, mice were exposed to three tones (CS; 30 s, 80 dB, 4 kHz) at 30-s interstimulus intervals (ISIs), and 2 s before the end of each tone, a mild electric footshock (US; 0.75 mA, 2 s) was paired and co-terminated with the tone. After fear conditioning, each mouse was transferred to its home cage. Conditioning chambers were cleaned with 75% ethanol before each test. On day 2, mice were placed in a novel context (Context B, cylinder-shaped chambers) with a different floor texture instead of the electrifiable grid floor. Three minutes after exploring freely, two tones were presented at a 120-s ISI in Context B for the fear memory recall test without shock. On day 3, mice underwent fear extinction in Context B where they freely explored for 2 min, followed by12 tone-alone presentations (10-s ISIs). On day 4, as on day 3, mice underwent fear extinction. On day 5, mice had a retrieval test to assess the acquired extinction memory in Context B as on day 2. On day 11, mice had the same extinction retrieval to assess the spontaneous recovery of fear. All chambers were cleaned with 0.5% acetic acid before tests except day 1. Experiments were performed with computerized software (Freezescan, Clever Systems, Reston Inc., VA), which quantified the freezing level. For optogenetic manipulation, an optical fiber (OD, 200 μm; NA, 0.22) was implanted 200 μm above the PVT. During fear extinction retrieval, the implanted fiber was connected to a laser stimulator (Plexon) and the power at the tip of the fiber was adjusted to 5 mW for the yellow laser (550 nm) with 20 Hz, 25-ms pulse-width. The yellow laser was triggered 10 s before tone onset and persisted throughout the 30-s tone.

Immunohistochemistry

Mice were anesthetized with isoflurane and decapitated after behavioral tests. Each mouse was transcardially perfused with PBS, followed by 4% PFA and the brain was removed into an EP tube containing 4% PFA and kept overnight at 4°C. Coronal sections (40 μm) were cut on a vibratome (Leica) and stored in PBS until use. For c-Fos immunocytochemistry, sections were washed in PBS (3 × 10 min) and then blocked in 10% donkey serum containing 0.1% Triton-100 for permeation at room temperature. Two hours later, the sections were incubated overnight at 4°C with primary antibody solution containing c-Fos (rabbit 1:200, Synaptic system)/0.1% Triton-100/5% donkey serum in PBS. Next day, the sections were washed in PBS (3 × 10 min) and then incubated with the fluorescent secondary antibody solution containing cy3 (donkey anti-rabbit 1:200, Jackson Immuno Research Labs) at room temperature for 2 h, followed by washes in PBS (3 × 10 min). Then, sections were stained with Hoechst (1:400, sigma) for 10 min to label nuclei, followed by washes in PBS (3 × 10 min). Finally, the sections were mounted on slides with Fluoromount-G (Southern Biotech). Immunofluorescence images were acquired using a confocal microscope (LSM500, Zeiss). We selected the first of every four sections through the targeted areas for counting. The number of c-Fos-positive cells was counted by an experimenter who was blind to the group design.

Data Analysis

GraphPad Prism software was used to make graphs. Data were analyzed by SPSS software (power analysis and sample size, version 19). Two-way ANOVA followed by the Bonferroni post hoc test was used for extinction retrieval tests, and one-way ANOVA followed by Scheffe’s post hoc test for comparison among multiple groups. Comparisons between two groups in c-Fos test were analyzed by two-tailed Student’s t-test. All data are presented as means ± SEM. P < 0.05 was set as statistical significance.

Results

PVT and IL are Involved in Fear Extinction Retrieval

To test whether PVT and IL neurons are involved in fear extinction retrieval, we used c-Fos immunohistochemistry to assess their activity during extinction retrieval. We trained mice in the auditory fear conditioning/extinction paradigm, after which the mice were randomly divided into an extinction retrieval group and a non-retrieval group (Fig. 1A). We found that extinction retrieval significantly increased the number of c-Fos+ neurons in the PVT, compared to controls (two-tailed Student’s t-test; n = 5 per group, t2/8 = 4.381, P = 0.006; Fig. 1B, C). Moreover, significantly more c-Fos+ neurons were found in the IL of mice that experienced extinction retrieval than in controls (two-tailed Student’s t-test; n = 5 per group, t2/8 = 4.190, P = 0.003; Fig. 1D, E). Thus, the PVT and IL may be involved in fear extinction retrieval.

Fig. 1.

Fig. 1

Open in a new tab

Extinction retrieval increases c-Fos expression in the PVT and IL. A Schematic of the extinction retrieval protocol for B–E. B Representative c-Fos immunofluorescence images of the PVT (left, scale bars 200 μm) and high-magnification images (right, scale bars 20 μm) from selected areas in the left images. C Number of c-Fos-positive neurons in the PVT (two-tailed Student’s t-test; n = 5, t2/8 = 4.381, **P = 0.006). D Representative c-Fos immunofluorescence images of the IL (left, scale bars 200 μm) and high-magnification images (right, scale bars 20 μm) from selected areas in the left images. E Number of c-fos-positive neurons in the IL (two-tailed Student’s t-test; n = 5, t2/8 = 4.190, **P = 0.003). D3V: dorsal 3rd ventricle; MHb: medial habenular nucleus; PVT: paraventricular thalamus; Cg1: cingulate cortex, area 1; PrL: prelimbic cortex; IL: infralimbic cortex; DP: dorsal peduncular cortex; Ret: retrieval; NO-Ret: no retrieval.

Activity of the PVT is Critical for Extinction Retrieval

To test the role of the PVT in regulating extinction retrieval, we used pharmacogenetics to manipulate PVT neurons. We infused a CaMKIIα promoter-driven adeno-associated virus (AAV) expressing Gi-coupled hM4Di labelled with mCherry (AVV2/9-CaMKIIα-hM4Di-mCherry), a recombinant virus vector that silences neurotransmission in the presence of the designer drug agonist clozepine-N-oxide (CNO), into the PVT (Fig. 2A). Four weeks after microinjection of the vector, hM4Di-mCherry, identified as red fluorescence, was abundantly expressed in the PVT (Fig. 2B). We trained mice in the auditory fear conditioning/extinction paradigm, then at 24 h after training, the freezing levels assessed and CNO was intraperitoneally injected at 1 h before retrieval (Fig. 2A). In the saline-treated mice, no significantly different freezing level was found at 24 h after training (Ext vs Retr, P > 0.05), suggesting successful extinction retrieval in the short term. In the CNO-treated mice, however, a significantly higher freezing level was found at 24 h after training (Ext vs Retr, P < 0.001), and moreover, CNO-treated mice displayed more freezing than controls (Saline vs CNO, P < 0.001) (Fig. 2C), indicating that silencing PVT neurons inhibits extinction retrieval in the short term.

Fig. 2.

Fig. 2

Open in a new tab

Silencing the activity of PVT inhibits fear extinction retrieval behavior in the short term. A Schematic of the experimental protocol for B, C. B Representative images of hM4Di-mCherry expression in the PVT (upper left, scale bar 200 μm) and high-magnification images from selected area in the upper left image (upper right and lower, scale bars 20 μm). C Effect of specific inactivation of PVT neurons on extinction memory retrieval (two-way ANOVA, stage, F3/80 = 129.401, P < 0.001; treatment, F1/80 = 9.991, P = 0.002; interaction, F3/80 = 6.503; P < 0.001; ***P < 0.001, nsP > 0.05). Freezing level of mice measured during the 2-min baseline (Bas), acquired freezing (Acq, average freezing of 2 tone periods), extinction (Ext, percentage of freezing during last extinction block, each block represents average freezing of 3 extinction trials) and extinction retrieval (Retr, acquired extinction, average freezing of 2 tone periods). Saline, n = 10; CNO, n = 12. D3V: dorsal 3rd ventricle; MHb: medial habenular nucleus; PVT: paraventricular thalamus.

Next, we infused a CaMKIIα promoter-driven adeno-associated virus (AAV) expressing Gq-coupled hM3Dq labelled with mCherry into the PVT to stimulate its activity (Fig. 3A). Four weeks after the microinjection of virus vector, hM3Dq-mCherry, identified as red fluorescence, was abundantly expressed in the PVT (Fig. 3B). We trained mice in the auditory fear conditioning/extinction paradigm as above. At 7 days after extinction training, the freezing levels were assessed and CNO was intraperitoneally injected at 1 h before retrieval (Fig. 3A). In the saline-treated mice, a significantly higher freezing level was found 7 days after extinction training (Ext vs Retr, P < 0.001), suggesting failed extinction retrieval in the long term. In the CNO-treated mice, no significant change in freezing level was found 7 days after extinction training (Ext vs Retr, P > 0.05), and moreover, CNO-treated mice displayed less freezing than controls (Saline vs CNO, P < 0.001; Fig. 3C), suggesting that activating PVT facilitates extinction retrieval. Together, PVT activity is necessary and sufficient for the recall of extinction memory.

Fig. 3.

Fig. 3

Open in a new tab

Stimulating PVT activity promotes fear extinction retrieval in the long term. A Schematic of the experimental protocol for B, C. B Representative image of hM3Dq-mCherry expression in the PVT (upper left, scale bar 200 μm) and high-magnification images from selected area in the upper left image (upper right and lower, scale bars 20 μm). C Effect of specific activation of PVT neurons on extinction memory retrieval (two-way ANOVA, stage, F3/88 = 151.581, P < 0.001; treatment, F1/88 = 9.864, P = 0.002; interaction, F3/88 = 7.898, P < 0.001; ***P < 0.001, nsP > 0.05). Saline, n = 13; CNO, n = 11. D3V: dorsal 3rd ventricle; MHb: medial habenular nucleus; PVT: paraventricular thalamus.

IL-PVT Projections Regulate Fear Extinction Retrieval

Considerable evidence has demonstrated that the consolidation and retrieval of fear extinction memories require the IL [1]. It is known that the prefrontal cortex sends glutamatergic projections to the PVT [22] and some studies have shown that projections from the prelimbic (PL) cortex to the PVT mediate fear acquisition [23]. However, whether the IL sends enough projections to the PVT and whether these projections evoke extinction retrieval need to be elucidated. To test the relationship between PVT and mPFC, retrograde tracing adeno-associated virus (rAAV-hSyn-EGFP) was infused into the PVT by microinjection. Six weeks later, infected neurons and surrounding nerve terminals in the PVT were revealed by EGFP expression (Fig. 4A). Interestingly, when the retrograde tracing AAV was infused into the PVT, there were many EGFP+ neurons in the IL, but only a few EGFP+ neurons in the PL (Fig. 4B), suggesting that, in the mPFC-PVT circuit, projections from the IL rather than the PL are dominant.

Fig. 4.

Fig. 4

Open in a new tab

The IL-PVT circuit regulates fear extinction retrieval. A Schematic showing the rAVV microinjection site, and representative images of hsyn-eGFP expression in the PVT (upper, scale bar 200 μm) and high-magnification images from selected area in the upper image (lower, scale bars 20 μm). B Representative images of hsyn-eGFP expression in the mPFC (left, scale bar 200 μm) and high-magnification images from selected area in the left image (right, scale bars 20 μm) after microinjection of rAVV-hsyn-eGFP into the PVT. C Schematic of the experimental protocol for D-F. D Representative images of hM4Di-mCherry expression in the IL (left, scale bar 200 μm) and high-magnification images from selected area in the left image (right, scale bars 20 μm). E Schematic showing the AVV microinjection site and cannula implantation for CNO microinjection. F Effect of specific inactivation of IL-PVT projections on extinction memory retrieval (two-way ANOVA, stage, F3/68 = 119.766, P < 0.001; treatment, F1/68 = 3.387, P = 0.070; interaction, F3/68 = 6.042; P = 0.001; ***P < 0.001, nsP > 0.05). Saline, n = 10; CNO, n = 9. D3V: dorsal 3rd ventricle; MHb: medial habenular nucleus; PVT: paraventricular thalamus; M2: secondary motor cortex; Cg1: cingulate cortex, area 1; PrL: prelimbic cortex; IL: infralimbic cortex; DP: dorsal peduncular cortex.

To determine whether the projections from IL to PVT are necessary for extinction retrieval, we transfected neurons in the bilateral IL with AAV2/9-CaMKIIα-hM4Di-mCherry, and at the same time, implanted cannulas into the PVT for microinjection of CNO (Fig. 4C, E). Previous studies have demonstrated that the activity of hM4Di-expressing axon terminals is reduced after CNO application in the mPFC- basolateral nucleus of amygdala (BLA) circuit [24]. At 28 days after microinjection of the AAV into the IL, hM4Di-mCherry, identified as red fluorescence, was abundantly expressed in the IL (Fig. 4D). At 24 h after extinction training, the freezing levels were assessed and CNO was infused via cannulas in the PVT at 30 min before retrieval to silence the activity of IL-PVT projections (Fig. 4C). In the saline-treated mice, no significant change in freezing level was found 24 h after extinction training (Ext vs Retr, P > 0.05), while in the CNO-treated mice, a significantly higher freezing level was found at 24 h after extinction training (Ext vs Retr, P < 0.001), and moreover, CNO-treated mice displayed more freezing than controls (Saline vs CNO, P < 0.001; Fig. 4F). Thus, the IL-PVT circuit is necessary for extinction retrieval.

PVT-CeL Projections Contribute to Fear Extinction Retrieval

The CeA is composed of CeL and CeM subregions. The CeM serves as the fear output region and is modulated by parallel circuits that drive fear expression and inhibit conditioned fear responses [25]. The CeL consists of GABAergic neurons that project to the CeM [26]. A recent study showed that CeL is also important for fear extinction retrieval [27]. So, we investigated whether IL delivers extinction retrieval information to the CeL via the PVT-CeL circuit.

The axons of PVT neurons project into diverse regions and elicit different behaviors. To determine whether the PVT has preferred projections to different subregions of the amygdala, we infused anterograde tracing cytomegalovirus promoter-driven adeno-associated virus expressing enhanced green fluorescent protein (AAV-CMV-eGFP) into the PVT. Four weeks after infusion, eGFP-labeled neurons were revealed in the PVT (Fig. 5A) and a large number of eGFP-labeled projection fibers were observed in the CeL (Fig. 5B), consistent with the previous report [28]. Interestingly, we found only sparse eGFP-labeled projection fibers in other subregions of the amygdala (Fig. 5B), indicating that PVT neurons preferentially innervate neurons in the CeL.

Fig. 5.

Fig. 5

Open in a new tab

PVT-CeL projections contribute to fear extinction retrieval. A Representative image of CMV-eGFP expression in the PVT (upper, scale bar 200 μm) and high-magnification images from selected area in the upper image (lower, scale bars 20 μm). B Representative image of CMV-eGFP expression in the amygdala (upper left, scale bar 200 μm) and high-magnification images from selected area in the upper left image (lower left and right, scale bars 20 μm) after microinjection of AVV-CMV-eGFP into the PVT. C Schematic of the experimental protocol for D and E. D Schematic showing the AVV microinjection site and cannula implantation for CNO microinjection. E Effect of specific inactivation of PVT-CeL projections on extinction memory retrieval (two-way ANOVA, stage, F3/68 = 56.237, P < 0.001; treatment, F1/68 = 2.894, P = 0.093; interaction, F3/68 = 2.959; P = 0.038; **P < 0.01, ***P < 0.001, nsP > 0.05). Saline, n = 10, CNO, n = 9. D3V: dorsal 3rd ventricle; MHb: medial habenular nucleus; PVT: paraventricular thalamus; BLA: basolateral amygdala; CeM: centromedial amygdala; CeL: centrolateral amygdala.

Next, we examined whether silencing the activity of PVT-CeL projections affects extinction retrieval. We infused AVV2/9-CaMKIIα-hM4Di-mCherry into the PVT and implanted cannulas bilaterally into the CeL for CNO microinjection (Fig. 5D). Four weeks after AAV infection, we trained mice in the auditory fear conditioning/extinction paradigm. At 24 h after extinction training, the freezing levels were assessed and CNO was infused via cannulas in the CeL 30 min before retrieval to silence the PVT-CeL circuit (Fig. 5C). In the saline-treated mice, no significant change in freezing level was found at 24 h after extinction training (Ext vs Retr, P > 0.05), while in the CNO-treated mice, a significantly higher freezing level was found 24 h after extinction training (Ext vs Retr, P < 0.001), and moreover, CNO-treated mice displayed more freezing than controls (Saline vs CNO, P = 0.002) (Fig. 5E). Thus, the PVT-CeL circuit also contributes to fear extinction retrieval.

The IL-PVT-CeL Circuit Regulates Fear Extinction Retrieval

Although pharmacogenetically silencing the IL-PVT or the PVT-CeL circuit inhibited the extinction retrieval (Figs. 4 and 5), whether a IL-PVT-CeL circuit mediates extinction retrieval remained unclear. We thus investigated whether disrupting the IL-PVT circuit prevents the extinction retrieval induced by stimulating the PVT-CeL circuit, by combining pharmacogenetics with optogenetic manipulation.

We delivered AAV9-CaMKIIα-eNpHR3.0-EYFP into the IL and AVV2/9-CaMKIIα-hM3Dq-mCherry into the PVT by microinjection (Fig. 6A–C). Four weeks after the microinjections, we trained mice in the auditory fear conditioning/extinction paradigm (Fig. 6D). Cannulas were implanted bilaterally into the CeL for CNO microinjection and an optical fiber was implanted above the PVT for illumination (Fig. 6E). CNO or saline was infused into the CeL 30 min before yellow light delivery. We found that the extinction retrieval induced by activating the PVT-CeL circuit was diminished by silencing the IL-PVT circuit (EYFP+CNO vs EYFP+saline, P < 0.001; eNpHR3.0+CNO vs EYFP+CNO, P < 0.001) (Fig. 6F), suggesting that activity in the IL-PVT circuit is necessary for extinction retrieval.

Fig. 6.

Fig. 6

Open in a new tab

The IL-PVT-CeL circuit regulates fear extinction retrieval. A Schematic showing the AVV microinjection site. B Representative image of AVV2/9-CaMKIIα-hM3Dq-mCherry expression in the PVT and eNpHR3.0-expressing axon terminals from the IL (left, scale bar 200 μm; right, scale bar 20 μm). C Representative image of AAV9-CaMKIIα-eNpHR3.0-EYFP expression in the IL (left, scale bar 200 μm; right, scale bars 20 μm). D Schematic of the experimental protocol for A, E, and F. E Optical fibers implanted in the PVT for yellow light delivery and cannulae implanted in the CeL for CNO or saline microinjection. F Effect of inactivation of IL-PVT projections with stimulation of the PVT-CeL circuit on extinction memory retrieval (two-way ANOVA, Light, F1/54=24.979, P < 0.001; treatment, F1/54=18.215, P < 0.001; interaction, F1/54 = 1.351; P = 0.25; ***P < 0.001, nsP > 0.05). eNpHR3.0 + CNO, n = 15; eNpHR3.0 + saline, n = 13; EYFP + CNO, n = 16; EYFP + saline, n = 14. D3V: dorsal 3rd ventricle; MHb: medial habenular nucleus; PVT: paraventricular thalamus; M2: secondary motor cortex; Cg1: cingulate cortex, area 1; PrL: prelimbic cortex; IL: infralimbic cortex; DP: dorsal peduncular cortex.

Discussion

PVT, a part of medial thalamic nuclei, is involved in extensive circuits that control various behaviors. We found that silencing PVT activity suppressed the retrieval of fear extinction in the short term, while stimulating PVT activity facilitated the retrieval of fear extinction in the long term. Therefore, PVT plays a critical role in regulating the recall of fear extinction. It consists of excitatory neurons without GABAergic interneurons [29] and acts as a comprehensive transmission hub to receive projections from many regions and send projections to downstream regions. How does the PVT mediate fear extinction retrieval and which upstream region controls its activity? The mPFC sends glutamatergic projections to the PVT [22]. While the IL plays a crucial role in modulating the consolidation and retrieval of extinction memory [1], the PL, located above the IL, is mainly responsible for fear conditioning and expression [30]. We found that neurons in the PVT and IL were strongly activated during fear extinction retrieval. Retrograde tracing showed that projections to the PVT from the IL rather than the PL were dominant in the mPFC-PVT circuit. More importantly, silencing the glutamatergic projections from IL to PVT inhibited extinction retrieval. Thus, the IL controls PVT activity, and the IL-PVT circuit mediates extinction retrieval.

How does activation of the IL-PVT circuit realize the retrieval of extinction memory? The CeL may be its downstream target. The CeA has emerged as a critical site in the early formation of fear memories and it can be subdivided into CeL and CeM, containing GABAergic medium spiny-like neurons [28]. Somatostatin-expressing (SOM+) neurons in the CeL are responsible for fear memory formation and storage, while corticotropin releasing factor-expressing (CRF+) neurons in the CeL are important for flight when mammals respond to fear [28, 31, 32]. Evidence has shown that SOM+ and CRF+ neurons mutually inhibit each other in the CeL [31]. The competitive interactions are associated with the dynamic remodeling of neural circuits regulating the acquisition and extinction of conditioned fear. The acquisition of learned fear responses remodels the BLA-CeL circuit bias toward SOM+ neuron-mediated conditioned fear, while extinction training remodels the BLA-CeL circuit bias toward CRF+ neuron-mediated fear extinction. Moreover, recent studies have also revealed that CeL-CRF+ neurons play a critical role in the subsequent retrieval of extinction memory, rather than extinction training [27]. Previous evidence demonstrated that fear conditioning and extinction have opposite effects on dendritic spine remodeling [33]. Moreover, protein kinase c-δ-expressing (PKC-δ+) neurons in the CeL appear to be a counterpart of SOM+ neurons. It has previously been shown that PKC-δ+ neurons directly inhibit CeM output neurons, thus playing an important role in fear extinction learning [34–37]. We found that inhibition of the PVT-CeL circuit suppressed extinction retrieval. Therefore, it is possible that activation of the IL-PVT circuit after extinction training shifts its downstream projections from PVT-CeL SOM+ neurons to PVT-CeL CRF+/PKC-δ+ neurons, thereby facilitating extinction retrieval. Whether CRF+ directly or indirectly inhibits CeM output neurons remains to be established. Of course, IL can also regulate fear extinction through the direct IL-amygdala circuit. Previous studies have shown that extinction learning-induced IL activation inhibits CeM activity via the intercalated neurons, resulting in suppression of fear responses, and at the same time, synaptic strength in the circuit from the mPFC to the BLA is decreased after fear extinction training [13].

Of course, PVT does not project to the CeL of the amygdala only. It has been reported that there are glutamatergic projections from the PVT to the BLA. However, the PVT-BLA circuit is too weak to change the firing of BLA neurons [38]. Moreover, we found that PVT neurons preferentially innervated neurons in the CeL but not other subregions of the amygdala. Thus, the CeL is a dominant downstream target of the IL-PVT circuit. In addition, the PVT-CeL circuit not only modulates extinction retrieval but also fear acquisition, as Chen and Bi reported that the optogenetically-induced long-term depression (LTD) of the PVT-CeL circuit decreases fear expression [38]. Which type of neuron is activated may determine whether the PVT-CeL circuit mediates fear acquisition or extinction retrieval, since Chen and Bi showed that the percentage of relatively high excitability (RHE) neurons is decreased and the percentage of relatively low excitability (RLE) neurons is increased in the CeL by LTD induction [38]. Whether tone exposure during extinction training activates RHE or RLE neurons in the CeL remains to be studied.

Based on these, we proposed a model in which PVT-mediated fear expression and fear extinction: fear conditioning remodels the PVT-CeL circuit bias toward SOM+ neuron-mediated fear expression (Fig. S1A), while extinction training remodels the PVT-CeL circuit bias toward CRF+/PKC-δ+-mediated fear extinction (Fig. S1B). And moreover, IL-mediated extinction retrieval may have two pathways: one is the direct IL-amygdala circuit, from IL to CeM output neurons via ITC neurons, and the other is the indirect IL-PVT-amygdala circuit, from IL to PVT, and then to CeM output neurons via CRF+/PKC-δ+ neurons in the CeL (Fig. S1B). Besides, recent work has reported that the nucleus reuniens (NR) of the midline thalamus is responsible for both encoding and retrieving extinction memories, and projections from the mPFC to the NR also mediate extinction memory [14]. Our findings provide evidence for the critical role of prefrontal-thalamic circuits in fear extinction.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary material 1 (PDF 162 kb) (162.7KB, pdf)

Acknowledgments

This work was supported by the National Natural Science Foundation of China (31530091 and 81870912), the National Key Research and Development Program of China (2016YFC1306703), the Science and Technology Program of Guangdong Province, China (2018B030334001), and the Collaborative Innovation Center for Cardiovascular Disease Translational Medicine, China.

Conflict of interest

The authors declare no competing financial interests.

Footnotes

Yan Tao, Cheng-Yun Cai and Jia-Yun Xian contributed equally to this work.

References

  • 1.Tovote P, Fadok JP, Lüthi A. Neuronal circuits for fear and anxiety. Nat Rev Neurosci. 2015;16:317–331. doi: 10.1038/nrn3945. [DOI] [PubMed] [Google Scholar]
  • 2.Brunello N, Davidson JR, Deahl M, Kessler RC, Mendlewicz J, Racagni G, et al. Posttraumatic stress disorder: diagnosis and epidemiology, comorbidity and social consequences, biology and treatment. Neuropsychobiology. 2001;43:150–162. doi: 10.1159/000054884. [DOI] [PubMed] [Google Scholar]
  • 3.Ferry F, Bunting B, Murphy S, O’Neill S, Stein D, Koenen K. Traumatic events and their related PTSD burden in Northern Ireland: a consideration of the impact of the ‘Troubles’. Soc Psychiatry Psychiatr Epidemiol. 2014;49:435–446. doi: 10.1007/s00127-013-0757-0. [DOI] [PubMed] [Google Scholar]
  • 4.Mataix-Cols D, Fernández de la Cruz L, Monzani B, Rosenfield D, Andersson E, Pérez-Vigil A, et al. D-cycloserine augmentation of exposure-based cognitive behavior therapy for anxiety, obsessive-compulsive, and posttraumatic stress disorders: a systematic review and meta-analysis of individual participant data. JAMA Psychiatry 2017, 74: 501–510. [DOI] [PubMed]
  • 5.Rauch SAM, Koola C, Post L, Yasinski C, Norrholm SD, Black K, et al. In session extinction and outcome in Virtual Reality Exposure Therapy for PTSD. Behav Res Ther. 2018;109:1–9. doi: 10.1016/j.brat.2018.07.003. [DOI] [PubMed] [Google Scholar]
  • 6.Hoskins M, Pearce J, Bethell A, Dankova L, Barbui C, Tol WA, et al. Pharmacotherapy for post-traumatic stress disorder: systematic review and meta-analysis. Br J Psychiatry. 2015;206:93–100. doi: 10.1192/bjp.bp.114.148551. [DOI] [PubMed] [Google Scholar]
  • 7.Vervliet B, Craske MG, Hermans D. Fear extinction and relapse: state of the art. Annu Rev Clin Psychol. 2013;9:215–248. doi: 10.1146/annurev-clinpsy-050212-185542. [DOI] [PubMed] [Google Scholar]
  • 8.Goode TD, Maren S. Animal models of fear relapse. ILAR J. 2014;55:246–258. doi: 10.1093/ilar/ilu008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Dunsmoor JE, Niv Y, Daw N, Phelps EA. Rethinking extinction. Neuron. 2015;88:47–63. doi: 10.1016/j.neuron.2015.09.028. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Quirk GJ, Mueller D. Neural mechanisms of extinction learning and retrieval. Neuropsychopharmacology. 2008;33:56–72. doi: 10.1038/sj.npp.1301555. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Myers KM, Davis M. Mechanisms of fear extinction. Mol Psychiatry. 2007;12:120–150. doi: 10.1038/sj.mp.4001939. [DOI] [PubMed] [Google Scholar]
  • 12.Laurent V, Westbrook RF. Inactivation of the infralimbic but not the prelimbic cortex impairs consolidation and retrieval of fear extinction. Learn Mem. 2009;16:520–529. doi: 10.1101/lm.1474609. [DOI] [PubMed] [Google Scholar]
  • 13.Cho JH, Deisseroth K, Bolshakov VY. Synaptic encoding of fear extinction in mPFC-amygdala circuits. Neuron. 2013;80:1491–1507. doi: 10.1016/j.neuron.2013.09.025. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Ramanathan KR, Jin J, Giustino TF, Payne MR, Maren S. Prefrontal projections to the thalamic nucleus reuniens mediate fear extinction. Nat Commun. 2018;9:4527. doi: 10.1038/s41467-018-06970-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Zhou K, Zhu Y. The paraventricular thalamic nucleus: A key hub of neural circuits underlying drug addiction. Pharmacol Res. 2019;142:70–76. doi: 10.1016/j.phrs.2019.02.014. [DOI] [PubMed] [Google Scholar]
  • 16.Li S, Kirouac GJ. Sources of inputs to the anterior and posterior aspects of the paraventricular nucleus of the thalamus. Brain Struct Funct. 2012;217:257–273. doi: 10.1007/s00429-011-0360-7. [DOI] [PubMed] [Google Scholar]
  • 17.Dong X, Li S, Kirouac GJ. Collateralization of projections from the paraventricular nucleus of the thalamus to the nucleus accumbens, bed nucleus of the stria terminalis, and central nucleus of the amygdala. Brain Struct Funct. 2017;222:3927–3943. doi: 10.1007/s00429-017-1445-8. [DOI] [PubMed] [Google Scholar]
  • 18.Shao YF, Lin JS, Hou YP. Paraventricular Thalamus as a major thalamic structure for wake control. Neurosci Bull. 2019;35:946–948. doi: 10.1007/s12264-019-00364-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Zhang X, van den Pol AN. Rapid binge-like eating and body weight gain driven by zona incerta GABA neuron activation. Science. 2017;356:853–859. doi: 10.1126/science.aam7100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Zhu Y, Nachtrab G, Keyes PC, Allen WE, Luo L, Chen X. Dynamic salience processing in paraventricular thalamus gates associative learning. Science. 2018;362:423–429. doi: 10.1126/science.aat0481. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Ye X, Kapeller-Libermann D, Travaglia A, Inda MC, Alberini CM. Direct dorsal hippocampal–prelimbic cortex connections strengthen fear memories. Nat Neurosci. 2017;20:52–61. doi: 10.1038/nn.4443. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Otis JM, Zhu M, Namboodiri VMK, Cook CA, Kosyk O, Matan AM, et al. Paraventricular thalamus projection neurons integrate cortical and hypothalamic signals for cue-reward processing. Neuron. 2019;103(423–431):e4. doi: 10.1016/j.neuron.2019.05.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Do-Monte FH, Quiñones-Laracuente K, Quirk GJ. A temporal shift in the circuits mediating retrieval of fear memory. Nature. 2015;519:460–463. doi: 10.1038/nature14030. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.McGinins MM, Parrish BC, Chappell AM, Alexander NJ, McCool BA. Chronic ethanol differentially modulates glutamate release from dorsal and ventral prefrontal cortical inputs onto rat basolateral amygdala principal neurons. eNeuro 2020, 7: ENEURO.0132-19.2019. 10.1523/eneuro.0132-19.2019. [DOI] [PMC free article] [PubMed]
  • 25.Maeng LY, Cover KK, Taha MB, Landau AJ, Milad MR, Lebrón-Milad K. Estradiol shifts interactions between the infralimbic cortex and central amygdala to enhance fear extinction memory in female rats. J Neurosci Res. 2017;95:163–175. doi: 10.1002/jnr.23826. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Cassell MD, Freedman LJ, Shi C. The intrinsic organization of the central extended amygdala. Ann N Y Acad Sci. 1999;877:217–241. doi: 10.1111/j.1749-6632.1999.tb09270.x. [DOI] [PubMed] [Google Scholar]
  • 27.Hartley ND, Gaulden AD, Báldi R, Winters ND, Salimando GJ, Rosas-Vidal LE, et al. Dynamic remodeling of a basolateral-to-central amygdala glutamatergic circuit across fear states. Nat Neurosci. 2019;22:2000–2012. doi: 10.1038/s41593-019-0528-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Penzo MA, Robert V, Tucciarone J, De Bundel D, Wang M, Van Aelst L, et al. The paraventricular thalamus controls a central amygdala fear circuit. Nature. 2015;519:455–459. doi: 10.1038/nature13978. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Bentivoglio M, Balercia G, Kruger L. The specificity of the nonspecific thalamus: the midline nuclei. Progress in brain research. Prog Brain Res 1991, 87: 53–80. [DOI] [PubMed]
  • 30.Sierra-Mercado D, Padilla-Coreano N, Quirk GJ. Dissociable roles of prelimbic and infralimbic cortices, ventral hippocampus, and basolateral amygdala in the expression and extinction of conditioned fear. Neuropsychopharmacology. 2011;36:529–538. doi: 10.1038/npp.2010.184. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Fadok JP, Krabbe S, Markovic M, Courtin M, Xu C, Massi L, et al. A competitive inhibitory circuit for selection of active and passive fear responses. Nature. 2017;542:96–100. doi: 10.1038/nature21047. [DOI] [PubMed] [Google Scholar]
  • 32.Tang N, Ding YF, Zhang W, Hu J, Xu XH. Stay active to cope with fear: A cortico-intrathalamic pathway for conditioned flight behavior. Neurosci Bull. 2019;35:1116–1119. doi: 10.1007/s12264-019-00439-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Lai CS, Franke TF, Gan WB. Opposite effects of fear conditioning and extinction on dendritic spine remodelling. Nature. 2012;483:87–91. doi: 10.1038/nature10792. [DOI] [PubMed] [Google Scholar]
  • 34.Herry C, Ciocchi S, Senn V, Demmou L, Müller C, Lüthi A. Switching on and off fear by distinct neuronal circuits. Nature. 2008;454:600–606. doi: 10.1038/nature07166. [DOI] [PubMed] [Google Scholar]
  • 35.Ciocchi S, Herry C, Grenier F, Wolff SB, Letzkus JJ, Vlachos I, et al. Encoding of conditioned fear in central amygdala inhibitory circuits. Nature. 2010;468:277–282. doi: 10.1038/nature09559. [DOI] [PubMed] [Google Scholar]
  • 36.Haubensak W, Kunwar PS, Cai H, Ciocchi S, Wall NR, Ponnusamy R, et al. Genetic dissection of an amygdala microcircuit that gates conditioned fear. Nature. 2010;468:270–276. doi: 10.1038/nature09553. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Cai H, Haubensak W, Anthony TE, Anderson DJ. Central amygdala PKC-δ(+) neurons mediate the influence of multiple anorexigenic signals. Nat Neurosci. 2014;17:1240–1248. doi: 10.1038/nn.3767. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Chen M, Bi LL. Optogenetic long-term depression induction in the PVT-CeL circuitry mediates decreased fear memory. Mol Neurobiol. 2019;56:4855–4865. doi: 10.1007/s12035-018-1407-z. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary material 1 (PDF 162 kb) (162.7KB, pdf)

Articles from Neuroscience Bulletin are provided here courtesy of Springer

RESOURCES