Hormones and Behavior 63 (2013) 791–799

Contents lists available at SciVerse ScienceDirect

Hormones and Behavior

j ourna l homepage: www.e lsev ie r .com/ locate /yhbeh

Conditioned fear is modulated by CRF mechanisms in theperiaqueductal gray columns

Karina G. Borelli a,b,c,⁎, Lucas Albrechet-Souza a,b, Alessandra G. Fedoce d, Denise S. Fabri d,Leonardo B. Resstel d, Marcus L. Brandão a,b

a Instituto de Neurociências e Comportamento, Ribeirão Preto, SP, Brazilb Laboratório de Neuropsicofarmacologia, Faculdade de Filosofia, Ciências e Letras de Ribeirão Preto, Universidade de São Paulo, Ribeirão Preto, SP, Brazilc Núcleo de Cognição e Sistemas Complexos, Centro de Matemática, Computação e Cognição Universidade Federal do ABC, Santo André, SP, Brazild Departamento de Farmacologia, Faculdade de Medicina de Ribeirão Preto, Universidade de São Paulo, Ribeirão Preto, SP, Brazil

⁎ Corresponding author at: Universidade de São Paulo, AvRibeirão Preto, SP, Brazil. Fax: +55 16 36024830.

E-mail address: karinagb@usp.br (K.G. Borelli).

0018-506X/$ – see front matter © 2013 Elsevier Inc. Allhttp://dx.doi.org/10.1016/j.yhbeh.2013.04.001

a b s t r a c t

a r t i c l e i n f o

Article history:Received 4 September 2012Revised 5 April 2013Accepted 10 April 2013Available online 18 April 2013

Keywords:Corticotropin-releasing factorPeriaqueductal grayContextual fear conditioningFear-potentiated startleAutonomic responsesNonmonotonic effects

The periaqueductal gray (PAG) columns have been implicated in controlling stress responses throughcorticotropin-releasing factor (CRF), which is a neuropeptide with a prominent role in the etiology of fear-and anxiety-related psychopathologies. Several studies have investigated the involvement of dorsal PAG(dPAG) CRF mechanisms in models of unconditioned fear. However, less is known about the role of this neu-rotransmission in the expression of conditioned fear memories in the dPAG and ventrolateral PAG (vlPAG)columns. We assessed the effects of ovine CRF (oCRF 0.25 and 1.0 μg/0.2 μL) locally administered into thedPAG and vlPAG on behavioral (fear-potentiated startle and freezing) and autonomic (arterial pressureand heart rate) responses in rats subjected to contextual fear conditioning. The lower dose injected intothe columns promoted proaversive effects, enhanced contextual freezing, increased the blood pressure andheart rate and decreased tail temperature. The lower dose of oCRF into the vlPAG, but not into the dPAG, pro-duced a pronounced enhancement of the fear-potentiated startle response. The results imply that the PAG isa heterogeneous structure that is involved in the coordination of distinct behaviors and autonomic control,suggest PAG involvement in the expression of contextual fear memory as well as implicate the CRF as animportant modulator of the neural substrates of fear in the PAG.

© 2013 Elsevier Inc. All rights reserved.

Introduction

The neuropeptide corticotropin-releasing factor (CRF) has beenimplicated in the regulation of the neuroendocrine reaction (Valeet al., 1981). Over the last decades, CRF has been shown to modulatethe autonomic, immunologic, behavioral and cognitive responses toaversive stimuli (Bale and Vale, 2004). CRF receptors (CRF1 and CRF2)have been demonstrated to be broadly distributed in the structures thatcompose the limbic circuitry (De Souza et al., 1985; Hauger et al., 2006;Owens and Nemeroff, 1991; Swanson et al., 1983) and experimentalevidence demonstrated that intracerebroventricular (i.c.v.) CRF adminis-tration increases anxiety-related behavior in rodents (Momose et al.,1999; Yang et al., 2006). CRF specifically influences the reaction to aver-sive stimuli when locally injected into structures such as the amygdala(Hubbard et al., 2007; Pitts and Takahashi, 2011), hippocampus(Pentkowski et al., 2009; Todorovic et al., 2007), bed nucleus of thestria terminalis (Sink et al., 2012;Walker et al., 2009) and periaqueductal

. Bandeirantes, 3900, 14040-901,

rights reserved.

gray (PAG) (Borelli and Brandão, 2008; Carvalho-Netto et al., 2007;Martins et al., 1997, 2000; Miguel and Nunes-de-Souza, 2011).

The PAG is a limbic structure that is considered to be a final path-way of the stress reaction (Brandão et al., 2005; Carrive, 1993). Ana-tomical and histological analyses have divided this structure intolongitudinal columns parallel with the aqueduct, including the dorsal(dPAG, comprising dorsomedial and dorsolateral columns), lateral (lPAG)and ventrolateral (vlPAG) columns (Bandler and Shipley, 1994; Carrive,1993; Keay and Bandler, 2001). In rodents, PAG electrical stimulation,lesions or pharmacological manipulation alters anxiety-like behaviors inseveral paradigms, such as open field (Borelli et al., 2004; Brandão et al.,1999; Coimbra et al., 2006), elevated plus-maze (Netto and Guimarães,2004), conditioned-place aversion (Zanoveli et al., 2007), contextualconditioned fear (Kim et al., 1993; Walker and Carrive, 2003) andfear-potentiated startle (Reimer et al., 2012; Zhao et al., 2009).

Receptor autoradiography and binding studies in different centralnervous system regions in the rat demonstrated that there is a substan-tial population of both CRF receptor subtypes within the different PAGcolumns (Merchenthaler, 1984; Swanson et al., 1983). In support of aCRF neurotransmitter role within the PAG columns, the CRF receptor lo-calization correspondswellwith the distribution of CRF-immunoreactiveterminals in this structure (Steckler and Holsboer, 1999). Furthermore,

792 K.G. Borelli et al. / Hormones and Behavior 63 (2013) 791–799

perikarya with immunoreactive CRF were identified in the PAG, pre-dominantly in its ventral portion, medial to the Edinger–Westphalnucleus and CRF fibers can also be seen coursing from the amygdala,bed nucleus of the stria terminalis and hypothalamus to the dPAG(Gray and Magnuson, 1992; Swanson et al., 1983). Also, Bowers et al.(2003) using an electrophysiological recording technique showedthat CRF produced a dose dependent excitatory effect on PAG neurons,both in dorsal and in ventrolateral subdivisions.

Ovine CRF (oCRF) and cortagine, a selective CRF1 agonist, admin-istered into dPAG produced anxiogenic-like effects in the mouse de-fense test battery and rat exposure test, which are animal modelsused to investigate mice defensive patterns in the presence of naturalaversive stimuli (Carvalho-Netto et al., 2007; Litvin et al., 2007).Additionally, intra-dPAG injection of the CRF antagonist α-helical-CRFreduced anxiety-related behaviors in rodents subjected to the elevatedplus-maze (Martins et al., 2000), whereas oCRF produced oppositeresponses (Borelli and Brandão, 2008). Although a number of studieshas investigated the involvement of dPAG CRF mechanisms in modelsof unconditioned fear (Borelli and Brandão, 2008; Carvalho-Netto etal., 2007; Litvin et al., 2007; Miguel and Nunes-de-Souza, 2011), lessis known about the CRF neurotransmission in the expression of condi-tioned fear memories in the PAG columns. To further investigate thisissue, we assessed the effects of CRF locally administered into thedPAG and vlPAG on behavioral (fear-potentiated startle and freezing)and autonomic (arterial pressure and heart rate) responses in ratssubjected to contextual fear conditioning. In this paradigm, electricalfootshocks are paired with an initially neutral background context.After some pairings, the context evokes a conditioned fear reactionconsisting of behavioral and autonomic responses including freezing,urination, increased arterial blood pressure and ultrasonic vocalization(Antoniadis and McDonald, 1999; Bouton and Bolles, 1979; Fanselowand Tighe, 1988; LeDoux et al., 1988; Resstel et al., 2006; Wöhr et al.,2005).

Materials and methods

Ethics

The experiments reported in this article were performed in accor-dancewith the recommendations of the Brazilian Society ofNeuroscienceand Behavior and complied with the United States National Institutes ofHealth Guide for Care and Use of Laboratory Animals. The procedureswere approved by the Committee for Animal Care and Use, Universityof São Paulo (No. 125-2010 and 11.1.308.53.9). All efforts were made tominimize animal suffering and reduce the number of rats used.

Animals and surgical procedures

A total of 110 male Wistar rats from the animal house of theUniversity of São Paulo Ribeirão Preto campus weighing 220–270 gwere used in the study. The rats were housed in groups of four animalsper cage with food and water available ad libitum in a temperature-controlled room (23 ± 1 °C) under a 12 h/12 h light/dark cycle (lightson at 07:00 AM) for 72 h.

Before the experimental sessions, the rats were anesthetized withketamine/xylazine (100/7.5 mg/kg, intraperitoneal; Agener União,Embu-Guaçu, SP, Brazil) and fixed in a stereotaxic apparatus (DavidKopf, Tujunga, CA, USA). The upper incisor bar was set 3.3 mm belowthe interaural line so that the skull was horizontal between bregma andlambda. After scalp anesthesia with 2% lidocaine, the skull was surgicallyexposed and a stainless steel guide cannula (10 mm length, 0.6 mmouter diameter, 0.4 mm inner diameter) was unilaterally implantedinto the dPAG using lambda as the reference (angle of 16°; coordinates:anterior/posterior, 0 mm; medial/lateral, ±1.9 mm; dorsal/ventral:−4.1 mm) or vlPAG (angle of 18°; coordinates: anterior/posterior,−0.2 mm; medial/lateral, ±2.6 mm; dorsal/ventral: −3.3 mm)

according to Paxinos and Watson (2005). The cannula was fixed tothe skull with dental cement and two stainless steel screws. Aftersurgery, the guide cannula was sealed with a stainless steel wire toprevent obstruction, and the rats received an intramuscular injectionof penicillin G benzathine (Pentabiotic, 600,000 IU, 0.2 mL; FortDodge, Campinas, SP, Brazil) and a subcutaneous injection of theanti-inflammatory and analgesic Banamine (flunixin meglumine,2.5 mg/kg [10 mg/mL, 0.2 mL]; Schering-Plough, São Paulo, SP, Brazil).After surgery, the rats were returned to their home cages in groups offour and were allowed to recover over a period of five days. To assessautonomic responses, one day prior to the test day, rats were anesthe-tized with ketamine/xylazine, and a catheter (4 cm PE-10 segmentheat-bound to a 13 cm PE-50 segment; Clay Adams, Northridge, CA,USA)was inserted into the abdominal aorta through the femoral arteryfor cardiovascular recording. The catheterwas tunneled under the skin,and exteriorized on the animal's dorsum.

Drug and infusion procedure

Ovine CRF (Sigma-Aldrich) was dissolved in 0.9% physiologicalsaline shortly before use to achieve a final concentration of 0.25 μgor 1 μg/0.2 μL. Physiological saline served as the vehicle control. Thedrug was administered locally 5 min prior to testing. The doses andtimes of the injections were based on previous studies (Borelli andBrandão, 2008; Carvalho-Netto et al., 2007). Infusions were deliveredusing an infusion pump (Harvard Apparatus, Holliston, MA, USA) in aconstant volume of 0.2 μL over 30 s. A thin dental needle (0.3 mmouter diameter) attached by polyethylene tubing to a 5 μL Hamiltonsyringe was introduced through each guide cannula. The injectionneedle extended 1.0 mm (dPAG) or 3.0 mm (vlPAG) below the ven-tral tip of the implanted guide cannula. The displacement of an airbubble inside the polyethylene tubing that connected the syringe tothe injection needle was used to monitor the microinjections. The in-jection needles were left in place for 30 s after the end of the infusionto allow for diffusion.

Fear-potentiated startle test

The acoustic startle reflex is elicited by sudden, unexpected andintense auditory stimulation, and involves a series of rapid and phasiccontractions of most of the skeletal muscles throughout the body. Therapidity of the primary response suggests that a relatively simplecircuit mediates the startle reflex with few synapses interposed be-tween the auditory nerve and the spinal motoneurons (Koch, 1999;Yeomans et al., 2002). The fear-potentiated startle paradigmhas provento be useful for analyzing neural systems involved in fear and anxiety.This test measures conditioned fear by an increase in the amplitude ofthe acoustic startle reflex in the presence of an explicit or contextualcue previously paired with a footshock (Davis et al., 1993; Grillon andBaas, 2003).

Fifty-eight rats were tested to examine the involvement of thePAG CRF mechanisms in the fear-potentiated startle response in acontext previously paired with footshocks. Ovine CRF was injectedinto the dPAG or vlPAG before the test sessions. The choice of thesecolumns was based on previous data from our laboratory showingthat CRF mechanisms in the lPAG did not appear to be involved in theexpression of conditioned fear responses (Borelli et al., unpublisheddata).

A wire-grid cage (16.5 × 7.5 × 7.5 cm) fixed to a response platformby four thumb screws was used as the test cage. The floor consistedof six stainless steel bars 3.0 mm in diameter spaced 1.5 cm apart. Thecage and response platform were located inside a ventilated, sound-attenuating plywood chamber (64 × 60 × 40 cm). A loudspeaker lo-cated 10 cm behind the test cage delivered both the startle stimulus(100 dB, 50 ms burst of white noise) and continuous backgroundnoise (55 dB). The startle reaction of the rats generated pressure on

793K.G. Borelli et al. / Hormones and Behavior 63 (2013) 791–799

the platform, and the analog signals were amplified, digitized andanalyzed using the Startle Reflex software (Med Associates, St. Albans,VT, USA). The startle reaction was recorded within a time window of100 ms after the onset of the startle stimulus. Calibration procedureswere conducted before the experiments to ensure equivalent sensitiv-ities of the response platform.

MatchingFor the first 2 days, the rats were placed in the test cage for a

5 min habituation period. Afterwards, the rats received a total of 30startle stimuli with an interstimulus interval of 30 s. Each matchingsession lasted 20 min. The rats were assigned to different control anddrug groups to achieve similar average startle amplitudes among thegroups based on the last matching day. The mean startle amplitudewas defined as the baseline condition before training began.

TrainingFive days after surgery, the rats were conditioned to the back-

ground context in the previously described experimental cage. Therats were placed individually in the cage and received 10 footshocks(0.3 mA, 1 s), after a 5 min acclimation period, with a variable intertrialinterval of 60–180 s (Borelli et al., 2005; Reimer et al., 2008). Thefootshocks were delivered through the training cage floor by a constantcurrent generator built with a scrambler (Albarsh Instruments; PortoAlegre, RS, Brazil). Stimulus presentation was controlled by amicropro-cessor and an I/O board (Insight Equipment; Ribeirão Preto, SP, Brazil).Each animal was removed 5 min after the last shock and was returnedto its home cage. The duration of each training session was approxi-mately 25 min.

TestThe test sessions were conducted without footshock presentation

in the same cage used for matching and training. Twenty-four hoursafter training, the rats received intra-dPAG or vlPAG injection ofoCRF or saline and were placed in the startle test cage after 5 min.After 5 min of habituation, the rats received 30 startle stimuli (noisebursts) with a 30 s interstimulus interval. All the experimental stepswere performed in the morning between 8:00 AM and 11:00 AM.

Contextual fear conditioning test

Considerable evidence has shown that rodents acquire fear by ex-posure to a distinctive environment in which aversive unconditionedstimuli, such as a footshock, had previously been administered. In thissituation, the most prominent behavioral outcome is the freezingresponse (Blanchard and Blanchard, 1969; Kim and Fanselow, 1992;Phillips and LeDoux, 1992; Rudy, 1993).

Thirty-four rats were used to assess the involvement of CRF-relatedmechanisms in the PAG in the expression of contextual freezing andautonomic responses. Conditioning and test sessions were conducted inthe same chamber (25 × 22 × 22 cm). The side and backwalls consistedof gray acrylic and the ceiling and front door consisted of transparentPlexiglas. The grid consisted of 18 stainless steel rods spaced 1.5 cmapart to deliver computer-controlled footshocks (Insight Instruments).The chamber was cleaned with 20% alcohol after each session.

TrainingRats were individually placed in the chamber and received 3

footshocks (0.8 mA, 2 s) after a 3 min acclimation period. The intertrialinterval varied randomly between 20 and 60 s (Lisboa et al., 2010). Therats were removed 3 min after the last shock and were returned to thehome cage.

TestRats were transferred from the animal room to the test room

(distinct from the conditioning room) in their home cage and remained

undisturbed for 1 h. Following this adaptation period, the arterialcannula was connected to a pressure transducer, and baseline valuesof cardiovascular parameters were recorded over 5 min (baseline 1).After injection into the dPAG or vlPAG, autonomic measures wererecorded for 5 minwhile the rats remained in their home cage (baseline2). The rats were replaced in the footshock chamber, and freezing be-havior and autonomic responses were recorded for 10 min. The timeof the freezing response was used as the criterion to assess fear condi-tioning. Freezing was operationally defined as the total absence ofmovement with the exception of respiration (Bouton and Bolles,1979). All of the experiments weremonitored in real time by a trainedinvestigator who was unaware of the treatment condition and manu-ally scored the freezing response. The mean arterial pressure (MAP)and heart rate (HR) were recorded using an AECAD 04P amplifier(Solução Integrada Ltda.; São Carlos, SP, Brazil) and an acquisitionboard (Power lab 8/30; ADInstruments, Colorado Springs, CO, USA)connected to a personal computer. The MAP and HR values were de-rived from pulse and arterial pressure recordings and were processedonline. The cutaneous temperature of the tail was recorded each minusing a thermal camera (Multi-Purpose Thermal Imager IRI 4010;Infrared Integrated Systems Ltd., Northampton, UK) at a distance of50 cm above the tail. The testing was conducted in a roommaintainedat 26 ± 1 °C, which is the thermoneutral zone for rats (Gordon, 1990).For statistical purposes, the continuous recordingswere averaged over60-s intervals. Changes in the MAP, HR or cutaneous temperature(ΔMAP, ΔHR and ΔTemperature, respectively) were sampled at1 min intervals. Points sampled during the 5 min period before theexposure were used as control baseline values.

Histology

After the tests, the rats were deeply anesthetized with a lethaldose of chloral hydrate (500 mg/kg, i.p.) and transcardially perfusedwith 0.9% saline followed by 10% formalin. The brains were removedfrom the skulls,maintained in formalin solution for 2 h and cryoprotectedin 30% sucrose for 72 h. Serial 60 μmbrain coronal sectionswere cut usinga cryostat (−19 °C), mounted on gelatin-coated slides and stained withcresyl violet (5%, Sigma-Aldrich) to localize the microinjection sites bymicroscopic examination according to the atlas of Paxinos and Watson(2005). Eighteen rats were excluded from the analysis because of inap-propriate cannula placement.

Statistical analysis

The software used for all statistical analyses was STATISTICAversion 6.0. The data are expressed as mean ± standard error of themean (s.e.m.). Two-way analysis of variance (ANOVA) with repeatedmeasures was used to assess the effects of oCRF injections into thedPAG or vlPAG on the startle amplitude with treatments (drug andsaline) and conditions (baseline and test) as factors. Time of freezingwas analyzed using one-way ANOVA. Significant comparisons werefurther evaluated using the Newman–Keuls post-hoc test. The base-line values of the MAP, HR and cutaneous temperature recordedbefore chamber reexposure were compared using one-way ANOVA.The MAP, HR and cutaneous temperature measured during the testwere analyzed using two-way ANOVA with treatment as the main in-dependent factor and time as the repeated measurement. Post-hoccomparisonswere performedwith t-tests and a Bonferroni adjustment.In all cases, values of p b 0.05 were considered statistically significant.

Results

Representative photomicrographs of microinjections into the dPAGand vlPAG and diagrammatic representations of the saline and oCRFinjection sites in these areas are shown in Fig. 1.

Fig. 1. Photomicrographs showing representative sites of injections into the dorsal periaqueductal gray (A) and ventrolateral periaqueductal gray (C). Outline of the injection sitesin the dPAG and vlPAG (B and D, respectively) on cross-sections from the Paxinos and Watson atlas (2005). The number of points in the figure is less than the total number of ratsused because of several overlaps. Scale bar = 500 μm.

794 K.G. Borelli et al. / Hormones and Behavior 63 (2013) 791–799

Effect of intra-PAG oCRF on fear-potentiated startle test

A timeline of the procedures for testing the involvement of CRFmechanisms in the expression of the fear-potentiated startle is depictedin Fig. 2 (A). The mean startle amplitude for rats receiving intra-dPAGsaline (n = 10), oCRF 0.25 μg (n = 9), or oCRF 1.0 μg (n = 9) beforethe testing session is shown in Fig. 2 (B). Two-wayANOVA revealed sta-tistically significant effects of the conditions (F1,25 = 33.76, p b 0.01)but not of the treatments (F2,25 = 0.39, p > 0.05) or the treatmentsvs. conditions interaction (F2,25 = 0.93, p > 0.05). The Newman–Keuls post-hoc test revealed that reexposure to the environment inwhich footshocks were previously administered enhanced the startleresponse in all the groups. The mean startle amplitude for the ratsreceiving intra-vlPAG saline (n = 11), oCRF 0.25 μg (n = 10), oroCRF 1.0 μg (n = 9) before the testing session is illustrated in Fig. 2(C). Two-wayANOVA revealed statistically significant effects of the con-ditions (F1,27 = 47.78, p b 0.05), treatments (F2,27 = 3.18, p b 0.05)and the treatments vs. conditions interaction (F2,27 = 5.01, p b 0.05).

The Newman–Keuls post-hoc test revealed that reexposure to the envi-ronment in which footshocks were previously administered enhancedthe startle response in all the groups. However, the conditioned startlereflex was higher in the group treated with oCRF 0.25 μg comparedwith the control group.

Effect of intra-PAG CRF on the contextual fear conditioning test

A timeline of the procedure for testing the effects of the oCRF in-jection into the PAG columns in the expression of contextual freezingbehavior and autonomic responses is depicted in Fig. 3 (A). The meanfreezing time exhibited for rats receiving intra-dPAG saline (n = 6),oCRF 0.25 μg, or oCRF 1.0 μg (n = 5 each dose) before exposure tothe environment in which footshocks were previously administeredis illustrated in Fig. 3 (B). One-way ANOVA showed that the treat-ment produced statistically significant effect (F2,15 = 14.99, p b 0.01).Post-hoc tests revealed that oCRF 0.25 μg increased the expression ofconditioned contextual freezing compared with the saline-treated rats.

Fig. 2. Timeline of the procedures for testing the effects of oCRF injection into the PAG columns on the expression of contextual fear (A). Effects of dPAG (B) or vlPAG (C) injectionsof saline or oCRF (0.25 μg/0.2 μL and 1.0 μg/0.2 μL, respectively) on the mean startle amplitude response shown by rats submitted to contextual fear conditioning. Results areexpressed as means + s.e.m. #p ≤ 0.05 compared with the saline group and *p ≤ 0.05 compared with baseline (Newman–Keuls test).

795K.G. Borelli et al. / Hormones and Behavior 63 (2013) 791–799

The mean freezing time exhibited for rats receiving intra-vlPAG saline,oCRF 0.25 μg, or oCRF 1.0 μg (n = 6 in each group) before exposure tothe environment in which footshocks were previously administeredis shown in Fig. 3 (C). One-way ANOVA revealed that the treatment pro-duced statistically significant effect (F2,17 = 144.90, p b 0.01). Thepost-hoc test showed that oCRF 0.25 μg increased the expression of con-ditioned contextual freezing compared with the control group.

Effect of intra-PAG CRF on autonomic measures

The autonomic responses evaluated in this study are depictedin Figs. 3 (D–I). Ovine CRF injection into the dPAG or vlPAG did notaffect the baseline values of MAP, HR or the cutaneous temperature(data not shown). Reexposure to the environment inwhich footshockswere previously administered induced a sustained increase in thecardiovascular responses (dPAG-MAP: F14,195 = 12.92, p b 0.001and HR: F14,195 = 66.57, p b 0.001; vlPAG-MAP: F14,225 = 68.68,p b 0.001 and HR: F14,225 = 128.40, p b 0.001) and decreased tailtemperature during the test (dPAG—F14,195 = 73.91, p b 0.001 andvlPAG—F14,225 = 79.96, p b 0.001). There was a significant increasein the autonomic responses evoked by chamber reexposure afterthe treatment with oCRF 0.25 μg in the dPAG (MAP: F2,195 = 25.99,p b 0.001; HR: F2,195 = 87.41, p b 0.001 and tail temperature:F2,195 = 73.56, p b 0.001). A significant interactionwas also observedbetween treatment and time (MAP: F28,195 = 2.92, p b 0.001; HR:F28,195 = 3.52, p b 0.001 and tail temperature: F28,195 = 3.14,p b 0.001). Similar resultswere observedwhenoCRFwas administeredinto the vlPAG: injections of low doses of oCRF increased the autonomicresponses during reexposure to the conditioning chamber (MAP:F2,225 = 208.60, p b 0.001; HR: F2,225 = 112.40, p b 0.001 and tailtemperature: F2,225 = 78.36, p b 0.001). A significant interactionwas also observed between treatment and time (MAP: F28,225 =8.62, p b 0.001; HR: F28,225 = 4.19, p b 0.001 and tail temperature:F28,225 = 3.68, p b 0.001). Treatment of ovine CRF 1 μg in the dPAGor vlPAG did not change the autonomic responses during reexposureto the aversive context

Discussion

To assess the involvement of intra-PAG CRFergic mechanisms inthe behavioral (fear-potentiated startle and freezing) and autonomic(MAP, HR and tail temperature) expression of fear conditioning, oCRFwas injected into the dPAG and vlPAG of rats before exposure to anenvironment in which footshocks were previously administered.Although intra-dPAG oCRF did not alter the startle response exhibitedby rats exposed to contextual fear, the lower dose promoted aproaversive effect, which enhanced contextual freezing, blood pres-sure and heart rate and decreased the tail temperature. These findingsare consistent with previous studies demonstrating that intra-dPAGoCRF increased unconditioned anxiety-like responses in the EPM(Borelli and Brandão, 2008) and intra-dPAG administration of theCRF antagonist α-helical-CRF reduced anxiety-related behaviors instress-experienced rats (Martins et al., 2000).

The increased contextual freezing and autonomic responsescaused by the lower dose of the drug indicate that the lack of effecton fear-potentiated startle was not an artifact of improper dosage.Moreover, the role of CRF mechanisms in the startle response is notclear (de Jongh et al., 2003; Schulz et al., 1996; Swerdlow et al.,1989) and appears to depend on a number of factors, including thestimulated region (Bijlsma et al., 2011; Walker et al., 2009; Zhao etal., 2009), functional integrity of brain structures (Antoniadis et al.,2009; Fendt et al., 1994; Lee and Davis, 1997), participation and inter-action with other neurochemical systems (Boyson et al., 2011; Conti,2012; Gresack and Risbrough, 2011) and involvement of distinct CRFreceptors (Risbrough et al., 2003, 2004, 2009). Based on the distincteffects of oCRF injection into the dPAG on conditioned freezing andfear-potentiated startle, different neurochemicalmechanisms andneu-ral circuitries could mediate the fear responses generated in these twotests (Borelli et al., 2005; de Oliveira et al., 2012; Reimer et al., 2012).The dorsal regions of the PAG may be involved in the expression ofunconditioned and/or avoidance behaviors, whereas conditioned re-sponses may recruit the ventral parts of this mesencephalic structure.We previously demonstrated that intra-dmPAG injection of oCRF in-creased risk assessment behaviors in rats subjected to the EPM test

Fig. 3. Timeline of the procedures for testing the effects of the oCRF injection into the PAG columns in the expression of contextual freezing behavior and autonomic responses (A).Percentage of freezing time response in rats receiving intra-dPAG (B) or vlPAG (C) injection of 0.2 μL of saline, oCRF 0.25 μg, or oCRF 1.0 μg/0.2 μL subjected to contextual fear.Time-course of microinjection of saline or oCRF at doses of 0.25 μg or 1.0 μg/0.2 μL into the dPAG (left column) or vlPAG (right column) on mean arterial pressure (MAP; D and E),heart rate (HR; F and G) and cutaneous temperature (H and I) of rats exposure to the environment in which footshocks were previously administered. Results are expressed asmeans ± s.e.m. *p ≤ 0.05 compared with the control group.

796 K.G. Borelli et al. / Hormones and Behavior 63 (2013) 791–799

(Borelli and Brandão, 2008). This class of behavior has been proposedto reflect information-gathering processes triggered in potentiallythreatening situations, which optimizes the best adaptive behavioralstrategy (Albrechet-Souza et al., 2009; Blanchard et al., 1993).

In contrast to the dPAG, rats that received intra-vlPAG injections ofoCRF at the lower dose before the test session exhibited increased

fear-potentiation startle. The same dose also increased the contextualfreezing response, blood pressure and heart rate and decreased thetail temperature,which indicates proaversive effects of CRF in emotionalcircuitry areas (Griebel, 1999; Owens and Nemeroff, 1991; Yang et al.,2006). These results suggest that endogenous CRF systems in the PAGcolumns may have a role in modulating behavioral and autonomic

797K.G. Borelli et al. / Hormones and Behavior 63 (2013) 791–799

responses to stress. The results also support the specificity of PAG co-lumnar modulation of aversive responses once CRF seems to presenta distinct action in the dPAG and vlPAG. Thus, different neural circuitscould be recruited at the PAG level during the expression of the contex-tual freezing and the production of fear-potentiated startle mediatedby the CRF system.

Our findings show that, compared with dPAG, vlPAG CRF-relatedmechanisms are involved on the expression of distinct conditioned fearresponses. CRF immunoreactive terminals in the PAG originate fromthe amygdala, bed nucleus of the stria terminalis and hypothalamus(Gray and Magnuson, 1992). High-density CRF-immunoreactive neu-rons are located in the amygdaloid complex and the basolateral nucleusof the amygdala is well-established to be involved with the CS–US asso-ciation (Pitts and Takahashi, 2011). Thus, fear expression appears to orig-inate from basolateral nucleus projections crossing through the centralnucleus of the amygdala to the bed nucleus of the stria terminalis(Dong et al., 2001) and projecting into the vlPAG (Dong and Swanson,2004), which is directly involved in the expression of the conditionedfear responses (LeDoux et al., 1988; Walker and Carrive, 2003).

Considering the effects of oCRF on autonomic measurements ourresults indicated that activation of CRF receptors in the PAG columns(dPAG and vlPAG) are involved in themediation of autonomic responseto stress. In fact, the PAG contains the neural substrate required to re-cruit the sympathetic nervous system and organizes physiological andbehavioral responses required to respond to environmental challenges(Dean, 2011; Lovick, 1985). A prior study showed that CRF i.c.v. admin-istration elicits cardiovascular responses similar to the effects of stressby increasing heart rate, cardiac output and mean arterial pressure,responses mediated by an intact sympathetic nervous system (Fisheret al., 1983). Morimoto et al. (1993) clearly demonstrated that duringstress responses, central CRF modulates anxiety-related cardiovascularresponses, altered body temperature and altered locomotor activity.These responses were significantly attenuated by pretreatment with i.c.v.injection of the CRF antagonist a-helical CRF (9–41), although significantchanges were not induced in unstressed rats. Interestingly, oCRF injectedinto the dPAG and vlPAGdid not alter autonomicmeasurements in rats be-fore exposure to the aversive context, indicating the specific involvementofthis peptide in modulating the responses elicited by the PAG during thestress response. Thus, it is likely that CRF mechanisms in the PAG do notmediate cardiovascular responses and body temperature regulation dur-ing the resting state in a nonstressful environment.

In this study, however, oCRF injected into the dPAG or vlPAG athigher doses did not induce behavioral or autonomic changes, andsome possibilities can be considered to explain these results. A reason-able possibility could be that oCRF, at higher doses, may nonspecificallyact on both CRF1 and CRF2 receptors, clearly supporting the notion ofopposed information processing mechanisms of these receptors duringthe stress (Bakshi and Kalin, 2000; Koob and Heinrichs, 1999; Reul andHolsboer, 2002). Indeed, some evidence indicates that anxiety-like be-haviors mediated by CRF1 activation are reduced when CRF2 receptorsare activated (Bale et al., 2000; Kishimoto et al., 2000). Alternatively,this absence of effects may be associated to an “inverted U” dose–response mechanism, often observed for a wide variety of neuropep-tides (McGaugh, 1983; Price and Cooper, 1975). These phenomenaencompass every level of biological organization from gene expression,hormone production and cell number to changes in tissue architecture(for review see Vandenberg et al., 2012).

Altogether, our findings corroborate the previous idea of the PAGas a heterogeneous structure involved in the coordination of distinctbehaviors and autonomic control (Borelli et al., 2005; Rea et al., 2011;Vieira et al., 2011) and further indicate the participation of PAG CRFmechanisms in stress responses exhibited in threatening environments.Indeed, some anxiety disorders have been linked to abnormalities inCRF signaling. Patients with post-traumatic stress disorder exhibit ele-vated levels of CRF in the cerebrospinal fluid (de Kloet et al., 2008;Sautter et al., 2003). Mutations in CRF (Smoller et al., 2005) and the

CRF1 receptor genes (Keck et al., 2008) have been associatedwith traitspredictive for panic disorder and panic disorder diagnoses, respectively.Thus, identifying CRF receptor signaling pathways that mediate defen-sive responses is crucial to extend our understanding of the neuralbasis of physiological fear, as well as associated disorders.

Acknowledgments

This work was supported by FAPESP (Proc. # 2011/00041-3, 2011/07332-3 and 2012/09300-4) and CNPq (Proc. # 471325/11-2) grants.K.G. Borelli received a postdoctoral fellowship from FAPESP (Proc. #2006/61402-9) and CNPq (59841/2010-0), L. Albrechet-Souza receiveda postdoctoral fellowship from CNPq (Proc. # `159411/2010-6).A. Fedoce received a fellowship from FAPESP (Proc. # 2011/20762-7).

References

Albrechet-Souza, L., Borelli, K.G., Carvalho, M.C., Brandão, M.L., 2009. The anterior cin-gulate cortex is a target structure for the anxiolytic-like effects of benzodiazepinesassessed by repeated exposure to the elevated plus maze and Fos immunoreactivity.Neuroscience 164, 387–397.

Antoniadis, E.A., McDonald, R.J., 1999. Discriminative fear conditioning to contextexpressed by multiple measures of fear in the rat. Behav. Brain Res. 101, 1–13.

Antoniadis, E.A., Winslow, J.T., Davis, M., Amaral, D.G., 2009. The nonhuman primateamygdala is necessary for the acquisition but not the retention of fear-potentiated startle. Biol. Psychiatry 65, 241–248.

Bakshi, V.P., Kalin, N.H., 2000. Corticotropin-releasing hormone and animal models ofanxiety: gene–environment interactions. Biol. Psychiatry 15 (48), 1175–1198.

Bale, T.L., Contarino, A., Smith, G.W., Chan, R., Gold, L.H., Sawchenko, P.E., Koob, G.F., Vale,W.W., Lee, K.F., 2000. Mice deficient for corticotropin-releasing hormone receptor-2 dis-play anxiety-like behaviour and are hypersensitive to stress. Nat. Genet. 24, 410–414.

Bale, T.L., Vale, W.W., 2004. CRF and CRF receptors: role in stress responsivity and otherbehaviors. Annu. Rev. Pharmacol. Toxicol. 44, 525–557.

Bandler, R., Shipley, M.T., 1994. Columnar organization in the midbrain periaqueductalgray: modules for emotional expression? Trends Neurosci. 17, 379–389.

Bijlsma, E.Y., van Leeuwen, M.L., Westphal, K.G., Olivier, B., Groenink, L., 2011. Localrepeated corticotropin-releasing factor infusion exacerbates anxiety- and fear-related behavior: differential involvement of the basolateral amygdala and medialprefrontal cortex. Neuroscience 26, 82–92.

Blanchard, R.J., Blanchard, D.C., 1969. Passive and active reactions to fear-eliciting stimuli.J. Comp. Physiol. Psychol. 68, 129–135.

Blanchard, R.J., Yudko, E.B., Rodgers, R.J., Blanchard, D.C., 1993. Defense system psycho-pharmacology: an ethological approach to the pharmacology of fear and anxiety.Behav. Brain Res. 20, 155–165.

Borelli, K.G., Brandão, M.L., 2008. Effects of ovine CRF injections into the dorsomedial,dorsolateral and lateral columns of the periaqueductal gray: a functional role forthe dorsomedial column. Horm. Behav. 53, 40–50.

Borelli, K.G., Nobre, M.J., Brandão, M.L., Coimbra, N.C., 2004. Effects of acute and chronicfluoxetine and diazepam on freezing behavior induced by electrical stimulation ofdorsolateral and lateral columns of the periaqueductal gray matter. Pharmacol.Biochem. Behav. 77, 557–566.

Borelli, K.G., Gárgaro, A.C., dos Santos, J.M., Brandão, M.L., 2005. Effects of inactivationof serotonergic neurons of the median raphe nucleus on learning and performanceof contextual fear conditioning. Neurosci. Lett. 21, 105–110.

Bouton, M.E., Bolles, R.C., 1979. Role of conditioned contextual stimuli in reinstatementof extinguished fear. J. Exp. Psychol. Anim. Behav. Process. 5, 368–378.

Bowers, L.K., Swisher, C.B., Behbehani, M.M., 2003. Membrane and synaptic effects ofcorticotropin-releasing factor on periaqueductal gray neurons of the rat. BrainRes. 15 (981), 52–57.

Boyson, C.O., Miguel, T.T., Quadros, I.M., Debold, J.F., Miczek, K.A., 2011. Prevention ofsocial stress-escalated cocaine self-administration by CRF-R1 antagonist in therat VTA. Psychopharmacology 218, 257–269.

Brandão, M.L., Anseloni, V.Z., Pandóssio, J.E., De Araújo, J.E., Castilho, V.M., 1999. Neuro-chemical mechanisms of the defensive behavior in the dorsal midbrain. Neurosci.Biobehav. Rev. 23, 863–875.

Brandão, M.L., Borelli, K.G., Nobre, M.J., Santos, J.M., Albrechet-Souza, L., Oliveira, A.R.,Martinez, R.C., 2005. Gabaergic regulation of the neural organization of fear inthe midbrain tectum. Neurosci. Biobehav. Rev. 29, 1299–1311.

Carrive, P., 1993. The periaqueductal gray and defensive behavior: functional represen-tation and neuronal organization. Behav. Brain Res. 20 (58), 27–47.

Carvalho-Netto, E.F., Litvin, Y., Nunes-de-Souza, R.L., Blanchard, D.C., Blanchard, R.J.,2007. Effects of intra-PAG infusion of ovine CRF on defensive behaviors in Swiss–Webster mice. Behav. Brain Res. 25 (176), 222–229.

Coimbra, N.C., De Oliveira, R., Freitas, R.L., Ribeiro, S.J., Borelli, K.G., Pacagnella, R.C.,Moreira, J.E., da Silva, L.A., Melo, L.L., Lunardi, L.O., Brandão, M.L., 2006. Neuroana-tomical approaches of the tectum-reticular pathways and immunohistochemicalevidence for serotonin-positive perikarya on neuronal substrates of the superiorcolliculus and periaqueductal gray matter involved in the elaboration of the defen-sive behavior and fear-induced analgesia. Exp. Neurol. 197, 93–112.

798 K.G. Borelli et al. / Hormones and Behavior 63 (2013) 791–799

Conti, L.H., 2012. Interactions between corticotropin-releasing factor and the serotonin1A receptor system on acoustic startle amplitude and prepulse inhibition of thestartle response in two rat strains. Neuropharmacology 62, 256–263.

Davis, M., Falls, W.A., Campeau, S., Kim, M., 1993. Fear-potentiated startle: a neural andpharmacological analysis. Behav. Brain Res. 20 (58), 175–198.

de Jongh, R., Groenink, L., van der Gugten, J., Olivier, B., 2003. Light-enhanced and fear-potentiated startle: temporal characteristics and effects of alpha-helical corticotropin-releasing hormone. Biol. Psychiatry 54, 1041–1048.

de Kloet, C.S., Vermetten, E., Geuze, E., Lentjes, E.G., Heijnen, C.J., Stalla, G.K.,Westenberg, H.G., 2008. Elevated plasma corticotrophin-releasing hormone levelsin veterans with posttraumatic stress disorder. Prog. Brain Res. 167, 287–291.

de Oliveira, A.R., Reimer, A.E., Reis, F.M., Brandão,M.L., 2012. Conditioned fear response ismod-ulatedby a combined actionof thehypothalamic–pituitary–adrenal axis anddopamine ac-tivity in the basolateral amygdala. Eur. Neuropsychopharmacol. (Epub ahead of print).

De Souza, E.B., Insel, T.R., Perrin, M.H., Rivier, J., Vale,W.W., Kuhar, M.J., 1985. Corticotropin-releasing factor receptors arewidely distributedwithin the rat central nervous system:an autoradiographic study. J. Neurosci. 5, 3189–3203.

Dean, C., 2011. Endocannabinoidmodulation of sympathetic and cardiovascular responsesto acute stress in the periaqueductal gray of the rat. Am. J. Physiol. Regul. Integr. Comp.Physiol. 300, 771–779.

Dong, H.W., Swanson, L.W., 2004. Projections from bed nuclei of the stria terminalis,posterior division: implications for cerebral hemisphere regulation of defensiveand reproductive behaviors. J. Comp. Neurol. 12 (471), 396–433.

Dong, H.W., Petrovich, G.D., Swanson, L.W., 2001. Topography of projections fromamygdala to bed nuclei of the stria terminalis. Brain Res. Brain Res. Rev. 38, 192–246.

Fanselow, M.S., Tighe, T.J., 1988. Contextual conditioning with massed versus distributedunconditional stimuli in the absence of explicit conditional stimuli. J. Exp. Psychol.Anim. Behav. Process. 14, 187–199.

Fendt, M., Koch, M., Schnitzler, H.U., 1994. Lesions of the central gray block the sensiti-zation of the acoustic startle response in rats. Brain Res. 24 (661), 163–173.

Fisher, L.A., Jessen, G., Brown, M.R., 1983. Corticotropin-releasing factor (CRF): mechanismto elevate mean arterial pressure and heart rate. Regul. Pept. 5, 153–161.

Gordon, C.J., 1990. Thermal biology of the laboratory rat. Physiol. Behav. 47, 963–991.Gray, T.S.,Magnuson, D.J., 1992. Peptide immunoreactive neurons in the amygdala and the bed

nucleus of the stria terminalis project to themidbrain central gray in the rat. 13, 451–460.Gresack, J.E., Risbrough, V.B., 2011. Corticotropin-releasing factor and noradrenergic

signalling exert reciprocal control over startle reactivity. Int. J. Neuropsychopharmacol.14, 1179–1194.

Griebel, G., 1999. Is there a future for neuropeptide receptor ligands in the treatment ofanxiety disorders? Pharmacol. Ther. 82, 1–61.

Grillon, C., Baas, J., 2003. A review of the modulation of the startle reflex by affectivestates and its application in psychiatry. Clin. Neurophysiol. 114, 1557–1579.

Hauger, R.L., Risbrough, V.B., Brauns, O., Dautzenberg, F.M., 2006. Corticotropin releas-ing factor (CRF) receptor signaling in the central nervous system: new moleculartargets. CNS Neurol. Disord. Drug Targets 5, 453–479.

Hubbard, D.T., Nakashima, B.R., Lee, I., Takahashi, L.K., 2007. Activation of basolateralamygdala corticotropin-releasing factor 1 receptors modulates the consolidationof contextual fear. Neuroscience 150, 818–828.

Keay, K.A., Bandler, R., 2001. Parallel circuits mediating distinct emotional coping reac-tions to different types of stress. Neurosci. Biobehav. Rev. 25, 669–678.

Keck, M.E., Kern, N., Erhardt, A., Unschuld, P.G., Ising, M., Salyakina, D., Müller, M.B.,Knorr, C.C., Lieb, R., Hohoff, C., Krakowitzky, P., Maier, W., Bandelow, B., Fritze, J.,Deckert, J., Holsboer, F., Müller-Myhsok, B., Binder, E.B., 2008. Combined effectsof exonic polymorphisms in CRHR1 and AVPR1B genes in a case/control study forpanic disorder. Am. J. Med. Genet. B Neuropsychiatr. Genet. 5 (147B), 1196–1204.

Kim, J.J., Fanselow, M.S., 1992. Modality-specific retrograde amnesia of fear. Science 1(256), 675–677.

Kim, J.J., Rison, R.A., Fanselow, M.S., 1993. Effects of amygdala, hippocampus andperiaqueductal gray lesions on short- and long-term contextual fear. Behav.Neurosci. 107, 1093–1098.

Kishimoto, T., Radulovic, J., Radulovic, M., Lin, C.R., Schrick, C., Hooshmand, F.,Hermanson, O., Rosenfeld, M.G., Spiess, J., 2000. Deletion of crhr2 reveals an anxiolyticrole for corticotropin-releasing hormone receptor-2. Nat. Genet. 24, 415–419.

Koch, M., 1999. The neurobiology of startle Prog. Neurobiology 59, 107–128.Koob, G.F., Heinrichs, S.C., 1999. A role for corticotropin releasing factor and urocortin

in behavioral responses to stressors. Brain Res. 27 (848), 141–152.LeDoux, J.E., Iwata, J., Cicchetti, P., Reis, D.J., 1988. Different projections of the central

amygdaloid nucleus mediate autonomic and behavioral correlates of conditionedfear. J. Neurosci. 8, 2517–2529.

Lee, Y., Davis, M., 1997. Role of the hippocampus, the bed nucleus of the stria terminalisand the amygdala in the excitatory effect of corticotropin-releasing hormone onthe acoustic startle reflex. J. Neurosci. 17, 6434–6446.

Lisboa, S.F., Reis, D.G., da Silva, A.L., Corrêa, F.M., Guimarães, F.S., Resstel, L.B., 2010.Cannabinoid CB1 receptors in the medial prefrontal cortex modulate the expres-sion of contextual fear conditioning. Int. J. Neuropsychopharmacol. 13, 1163–1173.

Litvin, Y., Pentkowski, N.S., Blanchard, D.C., Blanchard, R.J., 2007. CRF type 1 receptorsin the dorsal periaqueductal gray modulate anxiety-induced defensive behaviors.Horm. Behav. 52, 244–251.

Lovick, T.A., 1985. Ventrolateral medullary lesions block the antinociceptive and cardio-vascular responses elicited by stimulating the dorsal periaqueductal grey matter inrats. Pain 21, 241–252.

Martins, A.P., Marras, R.A., Guimarães, F.S., 1997. Anxiogenic effect of corticotropin-releasing hormone in the dorsal periaqueductal grey. Neuroreport 10 (8), 3601–3604.

Martins, A.P., Marras, R.A., Guimarães, F.S., 2000. Anxiolytic effect of a CRH receptorantagonist in the dorsal periaqueductal gray. Depress. Anxiety 12, 99–101.

McGaugh, J.L., 1983. Hormonal influences on memory. Ann. Rev. Psychol. 27, 297–323.

Merchenthaler, I., 1984. Corticotropin releasing factor (CRF)-like immunoreactivity inthe rat central nervous system. Extrahypothalamic distribution. Peptides 5, 53–69.

Miguel, T.T., Nunes-de-Souza, R.L., 2011. Anxiogenic and antinociceptive effects in-duced by corticotropin-releasing factor (CRF) injections into the periaqueductalgray are modulated by CRF1 receptor in mice. Horm. Behav. 60, 292–300.

Momose, K., Inui, A., Asakawa, A., Ueno, N., Nakajima, M., Fujimiya, M., Kasuga, M.,1999. Intracerebroventricularly administered corticotropin-releasing factor inhibitsfood intake and produces anxiety-like behaviour at very low doses in mice. DiabetesObes. Metab. 1, 281–284.

Morimoto, A., Nakamori, T., Morimoto, K., Tan, H., Murakami, N., 1993. The central roleof corticotropin-releasing factor (CRF-41) in psychological stress in rats. J. Physiol.460, 221–229.

Netto, C.F., Guimarães, F.S., 2004. Anxiogenic effect of cholecystokinin in the dorsalperiaqueductal gray. Neuropsychopharmacology 29, 101–107.

Owens, M.J., Nemeroff, C.B., 1991. Physiology and pharmacology of corticotropin-releasing factor. Pharmacol. Rev. 43, 425–473.

Paxinos, G.,Watson, C., 2005. The Rat Brain in Stereotaxic Coordinates. Academic Press, Sydney.Pentkowski, N.S., Litvin, Y., Blanchard, D.C., Vasconcellos, A., King, L.B., Blanchard, R.J.,

2009. Effects of acidic-astressin and ovine-CRF microinfusions into the ventralhippocampus on defensive behaviors in rats. Horm. Behav. 56, 35–43.

Phillips, R.G., LeDoux, J.E., 1992. Differential contribution of amygdala and hippocampus tocued and contextual fear conditioning. Behav. Neurosci. 106, 274–285.

Pitts, M.W., Takahashi, L.K., 2011. The central amygdala nucleus via corticotropin-releasing factor is necessary for time-limited consolidation processing but notstorage of contextual fear memory. Neurobiol. Learn. Mem. 95, 86–91.

Price, M.T.C., Cooper, R.M., 1975. U-shaped functions in a shock escape task. J. Comp.Physiol. Psychol. 80, 600–606.

Rea, K., Roche, M., Finn, D.P., 2011. Modulation of conditioned fear, fear-conditionedanalgesia and brain regional c-Fos expression following administration ofmuscimol into the rat basolateral amygdala. J. Pain 12, 712–721.

Reimer, A.E., Oliveira, A.R., Brandão, M.L., 2008. Selective involvement of GABAergicmechanisms of the dorsal periaqueductal gray and inferior colliculus on the memoryof the contextual fear as assessed by the fear potentiated startle test. Brain Res. Bull. 30(76), 545–550.

Reimer, A.E., de Oliveira, A.R., Brandão, M.L., 2012. Glutamatergic mechanisms of thedorsal periaqueductal gray matter modulate the expression of conditioned freez-ing and fear-potentiated startle. Neuroscience 6, 72–81.

Resstel, L.B., Joca, S.R., Guimarães, F.G., Corrêa, F.M., 2006. Involvement of medial pre-frontal cortex neurons in behavioral and cardiovascular responses to contextualfear conditioning. Neuroscience 1 (143), 377–385.

Reul, J.M., Holsboer, F., 2002. Corticotropin-releasing factor receptors 1 and 2 in anxietyand depression. Curr. Opin. Pharmacol. 2, 23–33.

Risbrough, V.B., Hauger, R.L., Pelleymounter, M.A., Geyer, M.A., 2003. Role of corticotropinreleasing factor (CRF) receptors 1 and 2 in CRF-potentiated acoustic startle in mice.Psychopharmacology 170, 178–187.

Risbrough, V.B., Hauger, R.L., Roberts, A.L., Vale, W.W., Geyer, M.A., 2004. Corticotropinreleasing factor receptors CRF1 and CRF2 exert both additive and opposing influ-ences on defensive startle behavior. J. Neurosci. 24, 6545–6552.

Risbrough, V.B., Geyer, M.A., Hauger, R.L., Coste, S., Stenzel-Poore, M., Wurst, W.,Holsboer, F., 2009. CRF1 and CRF2 receptors are required for potentiated startleto contextual but not discrete cues. Neuropsychopharmacology 34, 1494–1503.

Rudy, J.W., 1993. Contextual conditioning and auditory cue conditioning dissociateduring development. Behav. Neurosci. 107, 887–891.

Sautter, F.J., Bissette, G., Wiley, J., Manguno-Mire, G., Schoenbachler, B., Myers, L.,Johnson, J.E., Cerbone, A., Malaspina, D., 2003. Corticotropin-releasing factor inposttraumatic stress disorder (PTSD) with secondary psychotic symptoms,nonpsychotic PTSD and healthy control subjects. Biol. Psychiatry 54, 1382–1388.

Schulz, D.W.,Mansbach, R.S., Sprouse, J., Braselton, J.P., Collins, J., Corman,M., Dunaiskis, A.,Faraci, S., Schmidt, A.W., Seeger, T., Seymour, P., Tingley III, F.D., Winston, E.N., Chen,Y.L., Heym, J., 1996. CP-154,526: a potent and selective nonpeptide antagonist of cor-ticotropin releasing factor receptors. Proc. Natl. Acad. Sci. U. S. A. 93, 10477–10482.

Sink, K.S., Walker, D.L., Freeman, S.M., Flandreau, E.I., Ressler, K.J., Davis, M., 2012. Effects ofcontinuously enhanced corticotropin releasing factor expressionwithin the bed nucleusof the stria terminalis on conditioned and unconditioned anxiety. Mol. Psychiatry 1–12.

Smoller, J.W., Yamaki, L.H., Fagerness, J.A., Biederman, J., Racette, S., Laird, N.M., Kagan, J.,Snidman, N., Faraone, S.V., Hirshfeld-Becker, D., Tsuang, M.T., Slaugenhaupt, S.A.,Rosenbaum, J.F., Sklar, P.B., 2005. The corticotropin-releasing hormone gene and behav-ioral inhibition in children at risk for panic disorder. Biol. Psychiatry 15 (57), 1485–1492.

Steckler, T., Holsboer, F., 1999. Corticotropin-releasing hormone receptor subtypes andemotion. Biol. Psychiatry 46 (11), 1480–1508.

Swanson, L.W., Sawchenko, P.E., Rivier, J., Vale, W.W., 1983. Organization of ovinecorticotropin-releasing factor immunoreactive cells and fibers in the rat brain: animmunohistochemical study. Neuroendocrinology 36, 165–186.

Swerdlow, N.R., Britton, K.T., Koob, G.F., 1989. Potentiation of acoustic startle bycorticotropin-releasing factor (CRF) and by fear are both reversed by alpha-helicalCRF (9–41). Neuropsychopharmacology 2, 285–292.

Todorovic, C., Radulovic, J., Jahn, O., Radulovic, M., Sherrin, T., Hippel, C., Spiess, J., 2007.Differential activation of CRF receptor subtypes removes stress-induced memorydeficit and anxiety. Eur. J. Neurosci. 25, 3385–3397.

Vale, W., Spiess, J., Rivier, C., Rivier, J., 1981. Characterization of a 41-residue ovinehypothalamic peptide that stimulates secretion of corticotropin and beta-endorphin.Science 213, 1394–1397.

Vandenberg, L.N., Colborn, T., Hayes, T.B., Heindel, J.J., Jacobs, D.R. Jr, Lee, D.H., Shioda,T., Soto, A.M., vom Saal, F.S., Welshons, W.V., Zoeller, R.T., Myers, J.P., 2012.Hormones and endocrine-disrupting chemicals: low-dose effects and nonmonotonicdose responses. Endocr. Rev. 33, 378–455.

799K.G. Borelli et al. / Hormones and Behavior 63 (2013) 791–799

Vieira, E.B., Menescal-de-Oliveira, L., Leite-Panissi, C.R., 2011. Functional mapping ofthe periaqueductal gray matter involved in organizing tonic immobility behaviorin guinea pigs. Behav. Brain Res. 216, 94–99.

Walker, P., Carrive, P., 2003. Role of ventrolateral periaqueductal gray neurons in thebehavioral and cardiovascular responses to contextual conditioned fear andpoststress recovery. Neuroscience 116, 897–912.

Walker, D.L., Miles, L.A., Davis, M., 2009. Selective participation of the bed nucleusof the stria terminalis and CRF in sustained anxiety-like versus phasic fear-likeresponses. Prog. Neuropsychopharmacol. Biol. Psychiatry 33, 1291–1308.

Wöhr, M., Borta, A., Schwarting, R.K., 2005. Overt behavior and ultrasonic vocalizationin a fear conditioning paradigm: a dose–response study in the rat. Neurobiol.Learn. Mem. 2005 (84), 228–240.

Yang, M., Farrokhi, C., Vasconcellos, A., Blanchard, R.J., Blanchard, D.C., 2006.Central infusion of ovine CRF (oCRF) potentiates defensive behaviors in CD-1mice in the Mouse Defense Test Battery (MDTB). Behav. Brain Res. 15 (171),1–8.

Yeomans, J.S., Li, L., Scott, B.W., Frankland, P.W., 2002. Tactile, acoustic andvestibular systems sum to elicit the startle reflex. Neurosci. Biobehav. Rev. 26, 1–11.

Zanoveli, J.M., Ferreira-Netto, C., Brandão, M.L., 2007. Conditioned place aversionorganized in the dorsal periaqueductal gray recruits the laterodorsal nucleusof the thalamus and the basolateral amygdala. Exp. Neurol. 208, 127–136.

Zhao, Z., Yang, Y., Walker, D.L., Davis, M., 2009. Effects of substance P in the amygdala,ventromedial hypothalamus and periaqueductal gray on fear-potentiated startle.Neuropsychopharmacology 34, 331–340.