Role of the PACAP system of the extended amygdala in the acoustic startle response in rats

Abstract

Anxiety-related disorders are the most prevalent mental disorders in the world and they are characterized by abnormal responses to stressors. Pituitary adenylate cyclase-activating polypeptide (PACAP) is a neuropeptide highly expressed in the extended amygdala, a brain macrostructure involved in the response to threat that includes the central nucleus of the amygdala (CeA) and the bed nucleus of the stria terminalis (BNST). The aim of this series of experiments was to systematically elucidate the role of the PACAP system of the CeA and BNST under both control, unstressed conditions and after the presentation of a stressor in rats. For this purpose, we used the acoustic startle response (ASR), an unconscious response to sudden acoustic stimuli sensitive to changes in stress which can be used as an operationalization of the hypervigilance present in anxiety- and trauma-related disorders. We found that infusion of PACAP, but not the related peptide vasoactive intestinal peptide (VIP), into either the CeA or the BNST causes a dose-dependent increase in ASR. In addition, while infusion of the antagonist PACAP(6–38) into either the CeA or the BNST does not affect ASR in non-stressed conditions, it prevents the sensitization of ASR induced by an acute footshock stress. Finally, we found that footshock stress induces a significant increase in PACAP, but not VIP, levels in both of these brain areas. Altogether, these data show that the PACAP system of the extended amygdala contributes to stress-induced hyperarousal and suggest it as a potential novel target for the treatment of stress-related disorders.

1. INTRODUCTION

Anxiety disorders, which affect approximately 40 million people and cost $42 billion a year in the United States, represent a massive public health issue [1, 2]. Symptoms of anxiety disorders include excessive fear of stimuli or perceived threats which can manifest with hypervigilance and hyperarousal. All the components of the stress systems respond to states of threatened homeostasis (i.e. presentation of a stressor) by mobilizing adaptive responses aimed to maintain the equilibrium and preparing us for immediate or potential harm [3, 4]. However, hyperactivation of these fundamental arousal and emotional systems may be key to the pathogenesis of conditions involving abnormal responses to stressors [5, 6].

Two major brain structures are known to heavily modulate the behavioral response to stress: the central nucleus of the amygdala (CeA) and bed nucleus of the stria terminalis (BNST) [7]. These brain regions are part of the extended amygdala, an anatomical construct which regulates the emotional component of the stress response [8, 9]. The CeA integrates sensory information from the environment, and sends direct projections to the BNST [9–11]. The CeA and the BNST then project information to various effector regions, such as the hypothalamus, the cortex, and the brainstem, to trigger appropriate responses, therefore coordinating the behavioral, autonomic and endocrine response to threats [12–15]. Notably, while in a non-pathological state, extended amygdala signaling is tapered appropriately to the severity of the present threat [16], hyperactivity of this brain macrostructure is hypothesized to play a critical role in the pathophysiology of anxiety and depressive disorders [17–21].

The CeA and BNST contain several important neuropeptides and neuropeptide receptors, which modulate their activity [22, 23]. One such neuropeptide is the pituitary adenylate cyclase-activating polypeptide (PACAP), a 38-amino acid peptide belonging to the secretin/glucagon/vasoactive intestinal polypeptide (VIP) superfamily. PACAP exerts its effects mainly via its cognate receptor PAC1 (PAC1R), which binds PACAP with an affinity of 1000-fold greater than VIP. On the other hand, VIP receptors (VPAC1R and VPAC2R) bind with PACAP and VIP with equal affinities [24]. In the brain, PACAP and PAC1R are highly expressed in the hypothalamus, the brainstem, and the extended amygdala [25, 26]. Dense PACAP-immunoreactive fibers are found in the dorsolateral BNST and in the capsular and lateral parts of the CeA subdivision [26, 27]. PACAP inputs to the CeA and BNST are non-local projections originating from the parabrachial nucleus [28, 29].

Human literature has shown that a single nucleotide polymorphism (SNP) in a putative estrogen response element (ERE) within the PAC1R gene was shown to predict stress disorder (PTSD) symptom severity in a human population of highly traumatized females [30]. This SNP leads to less efficient binding of estradiol activated estrogen receptor alpha at the PAC1R gene ERE, resulting in reduced expression of PAC1R and increasing risk for PTSD [31]. Recent preclinical literature has also implicated PACAP as a strong mediator of the stress response. Knockout mice studies have shown that several effects of stress, including HPA-axis activation and anxiety-like behavior, are dependent on PACAP signaling [32–34]. Central administration of PACAP to rodents evokes a stress-like response, activates the HPA axis, and induces depressive-like behavior [35–38]. Specifically in the extended amygdala, PACAP microinfusion into the CeA produces anxiety-like behavior in exploration-based tests [28, 39] as well as increased passive-coping in a shock-probe fear test [40]. PACAP microinfusion into the BNST also reduces exploratory behavior and increases acoustic startle reactivity [41, 42].

A behavioral test that has proven very useful in the investigation of the neural mechanisms of hypervigilance is the acoustic startle response (ASR). The ASR is a rapid reflex to an abrupt auditory stimulus that is mediated by very well defined neuronal pathways in the cochlear nucleus of the brainstem and spinal cord [43]. The response to these stimuli is very well conserved across species and it consists in the rapid contraction of the facial and skeletal muscles [44]. Patients with PTSD exhibit an exaggerated ASR [45–47], and in rats, ASR is a dependable measure of current emotional state and anxiety levels [48, 49]. Another advantage of using ASR is that it has a non-zero baseline, which allows bidirectional changes to be detected [50]. Importantly, drugs that are used as clinical anxiolytics, like diazepam and alprazolam, decrease startle reflex amplitude [51, 52], while drugs such as yohimbine promote anxiogenic-like effects and increase startle amplitude [53]. It has been shown that ASR can be increased as an unconditioned effect of an acute stressor (footshock), with potentiation occurring within minutes of administration [54]; this sensitization is also observed following various chronic stress paradigms [41, 55], and it is likely to reflect a heightened state of anxiety as a result of the exposure. Importantly, unlike fear-conditioning paradigms, classical sensitization of ASR by footshock is a non-associative form of learning (e.g. there are no cues involved), and it is therefore hypothesized to recruit different systems and circuits [56].

While the effects of PACAP administration in the BNST on ASR have been described, whether PACAP or PAC1R antagonism within the CeA modulates ASR is unknown. In addition, whether the PACAP/PAC1R system of either the BNST or the CeA plays a role in the effects of an acute stressor has not yet been reported. The aim of the present study was, therefore, to systematically elucidate the role of the PACAP system of the CeA and the BNST in the ASR, under both unstressed conditions and following an acute stressor.

2. MATERIAL AND METHODS

2.1. Subjects

Adult, male Wistar rats (Charles River, Wilmington, MA), weighing 301–325g upon arrival, were single-housed in wire-topped, plastic cages in a 12h:12h reverse light cycle, humidity- and temperature-controlled vivarium, with food and water available ad libitum. Experimental tests were conducted during the rats’ dark cycle. Procedures adhered to the National Institutes of Health Guide for the Care and Use of Laboratory Animals and the Principles of Laboratory Animal Care and were approved by Boston University Medical Campus Institutional Animal Care and Use Committee.

2.2. Drugs

PACAP-38 (here referred to as “PACAP”), PACAP(6–38) and VIP-28 (here referred to as “VIP”) were purchased from Bachem Americas, Inc. (Torrance, CA); doses were calculated based on the salt weights. Peptides were dissolved in sterile isotonic saline in the presence of 1% bovine serum albumin (Thermo Fisher Scientific, Waltham, MA). Drugs were administered 30 min prior to the startle session in all experiments except the sensitization to startle experiments, where PACAP(6–38) was administered 5 min after a pre-shock startle session, i.e. 30 min prior to the footshock session. PACAP-38 doses were chosen based on previous findings from our as well as other laboratories [28, 39]. PACAP(6–38) doses were chosen based on previous studies [38, 42, 57, 58]. VIP doses matched those of PACAP on a molecular weight basis; being VIP’s molecular weight lower than PACAP’s one, this ensured that VIP molar doses were not lower than those of PACAP.

2.3. Acoustic Startle Response (ASR)

Acoustic startle response (ASR) experiments were performed using the SR-Lab startle response system (San Diego Instruments, San Diego, CA), which consists of a Plexiglas cylinder (3.5” internal diameter) resting on a Plexiglas stand placed in a ventilated and sound-attenuated plywood chamber. Stimuli were delivered by a speaker located 24 cm above the cylinder; startle magnitude was detected in 200 ms bins and recorded by a piezoelectric device located beneath the cylinder. Rats were placed in the acoustic startle chamber and given 5 min of habituation with white noise only (65 dB) prior to the start of the session. An ASR session, adapted from [59], included 30 noise bursts, 50-msec in duration (10 each at 95 dB, 105 dB, and 115 dB), which were randomly presented with an average inter-trial interval of 30 sec (range of 5–39 sec). The presentation of 30 noise burst stimuli per session has been shown to cause negligible long-term habituation of startle across test days [60]. Rats were habituated to the startle chambers prior to testing and, additionally, received 2 baseline ASR sessions that were used to match groups based on their average startle in both baseline sessions. Treatment groups, therefore, did not differ in their baseline startle.

2.4. Footshock-Sensitized Acoustic Startle Response

The sensitization protocol, adapted from [54], involved a first ASR session consisting of 30 noise bursts (105dB), 50-msec in duration, presented at a fixed inter-trial interval of 30 sec. Footshocks were not administered during this “pre-shock” session. 5 min after the end of the first ASR session, rats were infused either vehicle or PACAP(6–38) and then placed back in their home cages. 30 min later, rats were returned to the startle chamber with a grid floor attached to the animal shocker controller that delivered a constant current to the grid. After 2 min habituation with white noise only (65 dB), rats received a series of 10 footshocks (500-msec in duration, 0.5 mA intensity), which were presented every 1 sec. During this footshock session, startle amplitude was measured to serve as a control measure for potential drug effects on nociceptive sensitivity. Immediately following, the “post-shock” ASR session began and was identical to the pre-shock ASR session.

2.5. Intracranial surgeries and microinfusion procedure

Intracranial surgeries

Rats were stereotaxically implanted with bilateral cannulas as described previously [38, 39, 61]. Briefly, stainless steel, guide cannulas (24 gauge, Plastics One, Inc., Roanoke, VA) were placed in the CeA or dorsal BNST using coordinates from bregma, AP −2.64 mm, ML ±4.2 mm or AP 0 mm, ML ±3.15 mm at a 14° angle, respectively. Four stainless steel screws were fastened to the rats’ skull around the cannulas with the help of dental restorative filled resin (Henry Schein Inc., Melville, NY) and acrylic cement. The incisor bar set at −3.3 mm below the interaural line. Coordinates were chosen according to Paxinos and Watson [62]. Rats were handled extensively before surgery and again at the end of the surgical recovery. Prior to microinfusion procedure, all animals underwent at least 3 sham injections to get accustomed to cannula manipulations.

Microinfusion procedure

Drugs were microinfused as previously described [38, 39, 61]. A 31-gauge internal injector connected via PE20 tubing to a Hamilton microsyringe driven by a pump (KD Scientifics, Holliston, MA), was inserted into the guide cannula for a final DV of −8.4 mm for CeA and −6.6 mm for BNST. 0.5 μl/side of drug was injected in 1 min and injectors were left inside for an additional minute to prevent backflow. All animals received one saline microinjection prior to testing to reduce stress-effects caused by the procedure which could confound the data.

Cannula placement

Cannula placement was verified at the end of all testing. Subjects were euthanized under isoflurane anesthesia and injected with the same microinjection volume of 0.5 μl/side of India ink. Brains were removed, flash frozen in a methyl butane/dry ice bath and stored at −80°C. Coronal sections of 30 μm were collected using a cryostat and placements verified under a microscope. An observer blind to drug and treatment conditions determined cannula placement. Any spillover of ink that occurred outside of the target areas was also excluded. Cannula placements that were located within the CeA or dorsal BNST for rats were included in the analyses and are depicted in Fig. 1S for CeA (A) and BNST (B).

2.6. Immunohistochemistry

Perfusion and immunohistochemistry

Rats were placed in the startle apparatus and exposed to either 10 min of white noise or to a session of footshocks; the footshock session consisted of a total of 30 footshock (500-msec in duration, 0.5 mA intensity); shocks were presented every 30 sec on average (range: 5–39 sec). These footshock parameters were chosen for this study based on their effectiveness in published studies as a mild, unconditioned stressor [63, 64]. 10 min after the session, rats were deeply anesthetized with isoflurane and then transcardially perfused as previously described [65]. The 10 min time point for sacrifice was chosen to most closely replicate the condition the cohort of rats were in during their post-footshock ASR session.

For PACAP and VIP visualization, brains were cut into 30 μm coronal sections using a cryostat and stored in a cryoprotectant solution at −20 °C. Every fourth section section (120 μm apart) of the entire BNST and every sixth section (180 μm apart) of the entire CeA were chosen and processed for immunohistochemistry. Free-floating sections were washed in tris-buffered saline (TBS) between incubations. For PACAP staining, slices were pretreated with 10 mM sodium citrate buffer (0.05% tween 20, pH 6.0) for 30 min at 80 °C. Sections were placed for 1 hr in blocking solution (3% normal goat serum, 0.4% Triton X-100) and subsequently incubated overnight at room temperature with primary antibodies raised in rabbit against either PACAP (1:500, T-4473, Peninsula Labs, San Carlos, CA) or VIP (1:1,000, 20077, Immunostar, Hudson, WI) in blocking solution. Specificity for the PACAP antibody has been well-characterized in the literature [26, 66–73]. Additionally, the manufacturer provides cross-reactivity data, obtained by radioimmunoassay, which demonstrates that the PACAP antiserum binds the C-terminal end of the PACAP-38 peptide while not showing cross-reactivity with VIP, CRF, or adrenocorticotrophic hormone. The VIP antiserum has also been extensively demonstrated to be effective in VIP antigen specificity tests in the literature [74–79]. Specificity of the VIP antibody has also been confirmed by the manufacturer through a soluble preadsorption test with VIP, which completely abolished VIP immunolabeling; on the other hand, preadsorption with CRF, calcitonin gene–related peptide, neuropeptide Y, somatostatin, or substance P has no effect on VIP immunoreactivity. Sections were incubated into a secondary antibody (1:400, Alexa Fluor 555 goat anti-rabbit, A-21429, Invitrogen, Carlsbad, CA) in blocking solution for 2 hr at room temperature. Sections were mounted onto uncoated glass slides and coverslipped with Vectashield mounting medium (Vector Laboratories, Burlingame, CA).

Densitometry

Using the Stereo Investigator software (MicroBrightField, Williston, VT), 10X objective pictures of sections containing either the CeA or BNST were taken with an Olympus BX-51 microscope equipped with a Rotiga 2000R live video camera (QImaging, Surrey, BC), a three-axis MAC6000 XYZ motorized stage (Ludl Electronics, Hawthorne, NY), and a personal computer workstation. All pictures were captured under a preset exposure and gain in order to standardize the images. Densitometry analysis was performed using ImageJ software (NIH) where images were converted to 8-bit and adjusted using the auto threshold Triangle algorithm. Once converted, mean density of immunohistochemical signal was calculated. Experimenter was blind to treatment groups.

2.7. Statistical Analysis

Data were analyzed using two-way analyses of variance (ANOVAs). Pairwise post hoc comparisons were made using Student Newman-Keuls test; Student’s t-test was used when comparing two groups. Significance was set at p≤0.05. In the dose-response experiments, noise level (3 levels) and dose (3–4 levels) were the within-subject factors. In the blockade of footshock-induced sensitization of ASR, footshock (2 levels) and treatment (2 levels) were within- and between-subject factors, respectively.

The software/graphic packages used were SigmaPlot 11.0 and Statistica 7.0. Final statistical analysis for the CeA and BNST experiments consisted of 33 and 27 rats, respectively.

3. RESULTS

3.1. PACAP, but not VIP, administered into the CeA increases ASR

As shown in Fig. 1A, PACAP microinfusion into the CeA significantly affected ASR (Noise level: F(2, 20)= 35.82, p <0.001; Dose: F(2, 20)= 3.63, p= 0.045; Noise level × Dose: F(4, 40)= 3.18, p= 0.02); post hoc analysis showed that both the 0.1 and the 0.3 μg/rat dose were effective at increasing ASR across noise intensities. Fig. 1B shows the effect of intra-CeA PACAP when the 3 noise intensities are collapsed; the highest dose of PACAP increased ASR 34.9% above vehicle.

Fig. 1.

Effect of bilateral microinfusion into the CeA of either PACAP (0, 0.1, 0.3 μg/rat) (A, B) or VIP (0, 0.1, 0.3 μg/rat) (C, D) on ASR amplitude. (B) and (D) show the effects of either PACAP or VIP, respectively, when all noise intensities are cumulated. N= 9–11/group. Bars represent Mean + SEM. *p < 0.05, **p <0.01 vs. Vehicle (Newman-Keuls).

On the other hand, as shown in Fig. 1C, VIP microinfusion into the CeA had no effect on ASR at any of the doses tested (Noise level: F(2, 16)= 31.56, p<0.001; Dose: F(2, 16)= 0.47, p = 0.63; Noise level × Dose: F(4, 32)= 0.83, p= 0.52, not significant (n.s.)). Fig. 1D shows the effect of intra-CeA VIP when the 3 noise intensities collapsed.

3.2. PACAP, but not VIP, administered into the BNST increases ASR

PACAP microinfusion into the BNST significantly increased ASR, as shown in Fig. 2A (Noise level: F(2, 18)= 23.67, p<0.001; Dose: F(3, 27)= 6.57, p =0.002; Dose × Noise level: F(6, 54)= 0.61, p= 0.72, n.s.). Post hoc analysis showed that both the 0.1 and 0.3 μg/rat doses significantly increased ASR across the various noise intensities. Fig. 2B shows the effect of intra-BNST PACAP when the 3 noise intensities are collapsed; the highest dose increased ASR 54.3% above vehicle.

Fig. 2.

Effect of bilateral microinfusion into the BNST of either PACAP (0, 0.03, 0.1, 0.3 μg/rat) (A, B) or VIP (0, 0.03, 0.1, 0.3 μg/rat) (C, D) on ASR amplitude. (B) and (D) show the effects of either PACAP or VIP, respectively, when all noise intensities are cumulated. N= 7–10/group. Bars represent Mean + SEM. *p < 0.05, **p <0.01 vs. Vehicle (Newman-Keuls).

On the other hand, as shown in Fig. 2C, VIP microinfusion into the BNST had no effect on ASR at any of the doses tested (Noise level: F(2, 12 = 27.90, p<0.001; Dose: F(2, 12)= 0.72, p= 0.51, n.s.; Noise level × Dose: F(4, 24)= 1.08, p= 0.39, n.s.). Fig. 2D shows the effect of intra-BNST VIP when the 3 noise intensities collapsed.

3.3. Footshock stress increases PACAP, but not VIP, immunoreactivity in the CeA and BNST

Densitometry analysis performed after immunohistochemistry revealed that footshock stress significantly affected PACAP levels in the CeA (t(18)= −2.25, p= 0.037) and BNST (t(17)= −4.72, p<0.001), respectively, as shown in Fig. 3A and 4A. Animals exposed to footshock showed a PACAP density increase of 39.8% in the CeA and a 40.1% increase in the BNST, compared to control, unstressed animals.

Fig. 3.

Effect of footshock stress on: (A) PACAP and (B) VIP expression in the CeA. (C) Representative images of PACAP and VIP expression in the CeA. N= 7–10/group. Bars represent Mean + SEM. *p < 0.05 vs. Control.

Fig. 4.

Effect of footshock stress on: (A) PACAP and (B) VIP expression in the BNST. (C). Representative images of PACAP and VIP expression in the BNST. N= 9–10/group. Bars represent Mean + SEM. ***p < 0.001 vs. Control.

On the other hand, footshock had no effect on VIP levels in either the CeA (t(14)= −1.74, p= 0.10, n.s.) or the BNST (t(17)= 0.03, p= 0.98, n.s.), as shown in Fig. 3B and 4B. Representative staining of PACAP and VIP density in control and footshock animals are shown in Fig. 3C for CeA and 4C for BNST.

3.4. The PAC1R/VPAC2R antagonist PACAP(6–38) administered into the CeA blocks footshock-induced sensitization of ASR

ASR was not affected by the administration of PACAP(6–38) into the CeA at any of the doses tested, as shown in Fig. 5A (Noise level: F(2, 24)= 56.54, p< 0.001; Dose: F(2, 24)= 0.57, p= 0.57, n.s.; Noise level × Dose: F(4, 48)= 1.55, p =0.20, n.s.).

Fig. 5.

Effect of bilateral microinfusion into the CeA of the PAC1/VPAC2 receptor antagonist PACAP(6–38) on: (A) basal ASR amplitude; (B) ASR amplitude post-shock (footshock-sensitized ASR) (average of 30 noise bursts (105dB), PACAP(6–38) 3 μg/rat). N= 7–13/group. Bars represent Mean + SEM. *p < 0.05 vs. Vehicle.

On the other hand, as shown in Fig. 5B, PACAP(6–38) treatment affected ASR when preceded by a short series of mild footshocks (Shock: F(1,14)= 4.76, p=0.046; Treatment: F(1,14)= 0.05, p=0.83, n.s.; Shock × Treatment: F(1,14)= 4.94, p=0.043). Indeed, while vehicle-treated animals showed a 87.1% increase in ASR following footshock compared to their pre-shock levels, animals pretreated with PACAP(6–38) (3 μg/rat) into the CeA only had a 11.1% increase. Importantly, intra-CeA vehicle- and PACAP(6–38)-treated animals had comparable startle response amplitude during the footshock session (data not shown, t(15)= −0.05, p= 0.96, n.s.); PACAP(6–38)-treated animals startled 1.04% more than vehicle-treated animals, ruling out an altered nociceptive perception.

3.5. The PAC1R/VPAC2R antagonist PACAP(6–38) administered into the BNST blocks footshock-induced sensitization of acoustic startle

ASR was not affected by the administration of PACAP(6–38) into the BNST, as shown in Fig. 6A (Noise level: F(2,18)= 28.80 p<0.001; Dose: F(2,18)= 1.51 p= 0.25 (n.s.); Noise level × Dose: F(4,36)= 0.47 p= 0.76 (n.s.)).

Fig. 6.

Effect of bilateral microinfusion into the BNST of the PAC1/VPAC2 receptor antagonist PACAP(6–38) on: (A) basal ASR amplitude; (B) ASR amplitude post-shock (footshock-sensitized ASR) (average of 30 noise bursts (105 db), PACAP(6–38) 1 μg/rat). N= 10–12/group. Bars represent Mean + SEM. *p < 0.05 vs. Vehicle.

On the other hand, as shown in Fig. 6B, intra-BNST PACAP(6–38) treatment affected ASR when preceded by a short series of mild footshocks (Shock: F(1, 18)= 8.69, p= 0.009; Treatment: F(1, 18)= 0.01, p= 0.91, n.s.; Footshock × Dose: F(1, 18)= 5.97, p= 0.025). Indeed, while vehicle-treated animals showed a 61.0% increase in ASR following footshock compared to their pre-shock levels, animals pretreated with PACAP(6–38) (1 μg/rat) into the BNST only displayed a 13.5% increase. Importantly, intra-BNST vehicle- and PACAP(6–38)-treated animals had comparable startle response amplitude during the footshock session (t(21)= −0.51, p= 0.61, n.s., data not shown); indeed, PACAP(6–38)-treated animals startled 7.57% more than vehicle-treated animals, ruling out an altered nociceptive perception.

4. DISCUSSION

The main findings of the present study were that PACAP microinfusion into either the CeA or the BNST increased ASR in rats, while the related peptide VIP in the same brain regions had no effect. In addition, the PAC1R/VPAC2R antagonist PACAP(6–38) into either the CeA or the BNST had no per se effect on baseline ASR, while it blocked footshock-induced sensitization of ASR. Finally, footshock stress increased PACAP levels in both the CeA and BNST, without altering VIP levels.

Rats microinfused with PACAP into the CeA showed a dose-dependent increase in ASR compared to vehicle-treated rats. This suggests that the PACAP drug treatment produces hyperarousal and stress-like effects in the CeA. The highest dose of 0.3 μg/rat was the most effective at increasing startle compared to vehicle-treated rats (134.9%). The finding of an effect of intra-CeA PACAP in the ASR paradigm is novel, but in line with previous studies showing that similar doses of PACAP are able to reduce exploration of the open arms of an elevated plus maze [39]. Another study had previously also reported an anxiogenic effect of intra-CeA PACAP in the same behavioral paradigm, although at significantly higher doses (2 μg/rat) [28].

Rats microinfused with PACAP into the BNST also displayed a dose-dependent increase in ASR compared to vehicle-treated rats, suggesting that PACAP is able to elicit a stress-like response also in this brain region. The 0.3 μg/rat dose was the most effective at increasing ASR. These findings confirm previous studies showing that intra-BNST PACAP increases ASR, while expanding them with a full dose-response curve and showing an effect at significantly lower doses than those previously reported [41, 80].

Notably, microinfusion of the same doses of VIP into either the CeA or the BNST was unable to elicit any changes in ASR, suggesting that PAC1R –and not VIP receptors- are likely responsible for the effects of PACAP.

Pretreatment with the PAC1R/VPAC2R antagonist PACAP(6–38) into either the CeA or the BNST did not have any effect on ASR at any of the doses tested. The doses of PACAP(6–38) chosen were based on the fact that the IC₅₀ of PACAP(6–38) is 10 times higher than the IC₅₀ of PACAP [81]. These results suggest that inhibiting PACAP signaling of the CeA and the BNST under basal, non-stressed conditions does not result in a reduction of stress states. When ASR was sensitized by a train of footshock, animals showed an ASR increase above baseline levels, in our case 87% and 58% from their unstressed ASR values in the CeA and BNST group, respectively, reflecting a rapid unconditioned response to a fear-provoking stimulus and to provide a model for a heightened state of anxiety or fear [54]. The administration of the PAC1R/VPAC2R antagonist PACAP(6–38) into either the CeA or BNST prior to the footshock stress was able to significantly block the development of the ASR sensitization, startle response during footshock was measured. PACAP(6–38) did not alter startle during footshock, ruling out that post-shock startle changes were a result of drug-induced antinociception [82].

Although PACAP(6–38) is commonly regarded as a selective PAC1R antagonist in many studies, it should be noted that it exhibits affinity also for VPAC2Rs [83, 84]. However, the efficacy of PACAP(6–38) together with the total lack of in vivo effects of VIP in the paradigms used strongly suggests that PAC1R, and not VIP receptors, mediates the anxiogenic effects of PACAP. Indeed, PAC1R selectively binds PACAP with 1000-fold greater affinity than VIP, whereas VIP receptors VPAC1R and VPAC2R bind VIP and PACAP with equal affinity [84, 85]. However, future studies utilizing highly selective PAC1R agonists and antagonists will be needed.

The results of this study suggest that endogenous PACAP is not released in these extrahypothalamic brain areas under basal, unstressed condition, and that instead this system becomes activated in response to a high-intensity, uncontrollable stress. These properties make PACAP an attractive therapeutic target, as it would ensure higher safety and tolerability in humans. Other classes of drugs, such as benzodiazepines and azapirones, have been shown to selectively reduce fear- or anxiety-enhanced ASR, while not reducing baseline ASR [86, 87]. In addition, a profile similar to PACAP(6–38) is shared also by antagonists of corticotropin-releasing factor (CRF) receptor type 1, which show efficacy in exploration-based models of anxiety under stressed, but not in non-stressed testing conditions, suggesting that the activation of stress systems is a requirement for the in vivo activity of the compounds [88–95].

Much work has been done to attempt to characterize the specific function of the CeA and BNST in terms of fear and anxiety responses. Lesions of the CeA, but not the BNST [96], block the expression of fear-potentiated startle using either visual or auditory conditioned stimuli [97, 98]. Intracerebroventricular CRF infusions are able to potentiate startle responses (CRF-enhanced startle) [99], and excitotoxic lesions of the BNST, but not CeA, completely block this enhanced startle [100]. Similarly to CRF-enhanced startle, light-enhanced startle requires the BNST but not the CeA for its expression [101]. Unfortunately, there remains some uncertainty on how the CeA and BNST differ in these types of responses. Specifically, two hypotheses for these differences are that either the CeA plays a role in mediating conditioned fear responses and the BNST unconditioned responses or, alternatively, that the CeA plays a role in mediating short-duration fear responses and the BNST longer-duration responses [102]. Interestingly, the Davis’ group has demonstrated different ways in which footshock-sensitized startle can be blocked by both lesions of the CeA [103] and BNST [104] and, therefore, our startle sensitization protocol may involve the activation of both brain regions. Despite the early notion that the CeA and the BNST have separate functions [105], recent rodent studies have shown that both these structures are activated by explicit and potential threats [106]. Future work will be needed to understand how PACAPergic inputs to these regions functionally influence the strength of signaling within these structures.

Our pharmacological findings were supported by immunohistochemical results. We found that exposure to footshock stress was able to increase PACAP levels in both the CeA and BNST of rats, unlike VIP levels. Such elevation in PACAP levels, therefore, is likely responsible for the ASR sensitization that occurs following footshock, as confirmed by the effects of the antagonist PACAP(6–38). It is important to note that similar, but not identical, protocols of footshock were used in the pharmacological vs. immunohistochemical experiment (see details in the Methods section); for instance, while rats were euthanized 25 min after the presentation of the first footshock, the post-shock session ended 15 min after the presentation of the first footshock in the pharmacological experiment. Therefore, caution should be used while interpreting the functional relevance of the PACAP elevation in the context of the effects of PACAP(6–38). In addition, since different cohorts of rats were used here for the pharmacological and for the immunohistochemical experiments, no correlations between startle amplitude and PACAP density in CeA or BNST can be made.

Our immunohistochemical study quantified fiber staining through densitometry as they are very abundant and cell bodies positive for the peptide were not visible. The PACAP released within the CeA and BNST may originate from outside sources; indeed, PACAP neurons originating from the lateral parabrachial nucleus (PBn) are known to innervate the CeA and BNST and to represent a critical source of the peptide in these areas [28, 29]. However, low to moderate levels of PACAP mRNA have also been detected in punches containing the BNST of rats [41], while the presence of PACAP mRNA in the CeA is still unconfirmed.

The present study was performed in male rats only; future follow-up studies will need to address the question of potential sex differences in the effects of PACAP on ASR, especially considering the literature showing that PACAP mRNA is regulated by estradiol and that a single nucleotide polymorphism in the PAC1R gene is associated with PTSD symptoms in females [30, 31].

Both the CeA and BNST are very heterogeneous structures with a rich diversity of cell types and complex circuitry [107–112]. Future experiments, perhaps using transgenic mice, will be needed to directly ascertain the specific cell-types and circuits activated by PACAP, to better understand how this specific peptide fits into this structural and functional complexity.

In summary, our data suggest that the PACAP system is recruited in the extended amygdala following exposure to mild stressors and that it mediates the resultant behavioral response. Although not directly tested here, these findings raise the interesting hypothesis that either the hypersecretion of PACAP or the hyperactivity of PAC1R in these brain areas may occur in certain stress- and trauma-related psychiatric disorders that are characterized by increased startle response, and may be responsible for their symptoms. More experimental evidence will be needed to confirm this hypothesis. Hence, pharmacological blockade of PACAP signaling via PAC1R may represent a potential novel avenue for the treatment of these disorders.

Highlights.

• PACAP into either the CeA or the BNST increases ASR in rats

• VIP in the same brain regions has no effect on ASR

• PAC1R/VPAC2R antagonist PACAP(6–38) in either CeA or BNST has no per se effect on ASR

• PACAP(6–38) in CeA and the BNST blocks footshock-induced sensitization of ASR

• Footshock stress increases PACAP, but not VIP, immunoreactivity in both CeA and BNST