Fos-like immunoreactive neurons following electrical stimulation of the dorsal periaqueductal gray at freezing and escape thresholds

Brain Research Bulletin 62 (2003) 179–189

Fos-like immunoreactive neurons following electrical stimulation of the dorsal periaqueductal gray at freezing and escape thresholds Daniel Machado Luiz Vianna, Karina Genaro Borelli, Cristina Ferreira-Netto, Carlos Eduardo Macedo, Marcus Lira Brandão∗ Laboratório de Psicobiologia, FFCLRP, Campus USP, Av. Bandeirantes 3900, 14049-901 Ribeirão Preto, SP, Brazil Received 10 May 2003; received in revised form 27 August 2003; accepted 17 September 2003

Abstract Electrical stimulation of the dorsal regions of the periaqueductal gray (PAG) leads to defensive reactions characterized as freezing and escape responses. Until recently it was thought that this freezing behavior could be due to the recruitment of neural circuits in the ventrolateral periaqueductal gray (vlPAG), while escape would be mediated by other pathways. Nowadays, this view has been changing mainly because of evidence that freezing and escape behaviors thus elicited are not altered after lesions of the vlPAG. It has been suggested that there are at least two pathways for periaqueductal gray-mediated defensive responses, one involving the hypothalamus and the cuneiform nucleus (CnF) which mediates responses to immediate danger and another one involving the amygdala and vlPAG which mediates cue-elicited responses, either learned or innate. To examine this issue further we measured Fos protein expression in brain areas activated by electrical stimulation of the dorsolateral PAG (dlPAG) at the freezing and escape thresholds. The data obtained showed that freezing-provoking stimulation caused increases in Fos expression in the dorsomedial PAG (dmPAG), while escape-provoking stimulation led to increases at both dmPAG and dlPAG. Surprisingly, neither escape- nor freezing-provoking stimulations altered Fos expression in the central nucleus of amygdala (CeA). Escape-provoking stimulation caused increased Fos expression in the ventromedial hypothalamus (VMH), dorsal premammilary nucleus (PMd) and in the cuneiform nucleus. Significant increases in Fos labeling were found in the dmPAG and PMd following freezing-provoking stimulation. Therefore, the present data support the notion of a neural segregation for defensive behaviors in the dorsal columns of PAG, with increased Fos expression in the dmPAG following freezing, while dlPAG is affected by both freezing and escape responses. dlPAG, CnF, VMH and PMd are part of a brain aversion network activated by fear unconditioned stimuli. The present data also suggests that the defensive responses generated at the dlPAG level do not recruit the neural circuits of the vlPAG and CeA usually activated by conditioned fear stimuli. © 2003 Elsevier Inc. All rights reserved. Keywords: Fear; Panic; Dorsal premammillary nucleus; Ventromedial hypothalamus; Cuneiform nucleus; Defensive behavior

1. Introduction It has been established that the longitudinal columns of the periaqueductal gray (PAG), along its rostrocaudal axis, coordinate distinct patterns of behavioral and physiological reactions critical for survival [3,15]. Electrical or chemical stimulation of the dorsal half of the periaqueductal gray of the rat leads to a vigorous defensive response characterized by alertness, freezing and escape behavior [2,7,32,43,49]. Although escape behavior has been the main focus of attention, interest in freezing behavior induced by electrical stimulation of the dlPAG has grown lately. This freezing

∗

Corresponding author. Tel.: +55-16-602-3638; fax: +55-16-6330619. E-mail address: [email protected] (M.L. Brandão).

0361-9230/$ – see front matter © 2003 Elsevier Inc. All rights reserved. doi:10.1016/j.brainresbull.2003.09.010

behavior is not mediated by ventrolateral periaqueductal gray (vlPAG), as it is not changed by vlPAG electrolytic lesions [49]. It is, thus, logical to assume that other PAG afferents and efferents should mediate this reaction [48]. Indeed, freezing behavior elicited by this procedure has been reported to resemble human panic attacks [49]. The structures with higher chance of being involved in the neural circuits of aversion are also those that when electrically stimulated, rostrally to the PAG, also give rise to defensive responses, such as the amygdala and medial hypothalamus, and the cuneiform nucleus, more caudally [22,28,40]. The medial frontal and cingulate cortices, sources of afferents to PAG, have also been implicated in the mediation of affective responses [1,27,33]. In fact, all these structures show increased Fos expression when rats are submitted to anxiogenic procedures [4,10,12,23,25,31].

180

D.M.L. Vianna et al. / Brain Research Bulletin 62 (2003) 179–189

Defensive strategies as distinct as avoidance, freezing and escape are likely to be organized by different networks [5,28,34]. Therefore, it is expected that electrical stimulation of the dlPAG either at freezing or escape thresholds should activate different sets of brain structures. To test this prediction Fos distribution was presently measured in serial sections of the brain following dlPAG electrical stimulations at freezing and escape thresholds of rats placed in an arena. Sham-stimulated animals were exposed to the same procedure in order to control for novelty of the exposure and locomotor activity. The reason for choosing electrical stimulation in this work is that this procedure makes it possible to clearly dissociate freezing and escape responses with gradual increases in the intensity of the electrical stimulation of the dlPAG. Also, we have chosen to concentrate on those nuclei known for having monosynaptic connections with PAG; that is, cingulate (Cg) and prelimbic (PrL) areas [27], central amygdaloid nucleus (CeA) [39], ventromedial (dorsomedial part, VMHdm) and premammillary (PMd) nuclei of the hypothalamus [11], cuneiform nucleus (CnF) [38], inferior colliculus (IC) [37], the locus coeruleus (LC) [26] and of course the dorsomedial (dmPAG), dorsolateral (dlPAG), lateral (lPAG) and ventrolateral (vlPAG) subdivisions of periaqueductal gray [3,15]. Microinjections of GABA receptor antagonists, such as bicuculline, into the dlPAG cause escape interspersed with freezing behavior. These effects have been related to the release of the tonic inhibition exerted by GABA mechanisms on the efferents responsible for the defense reaction [7,8]. For comparison purposes with the electrical stimulation, which stimulates fibers of passage and does not distinguish between efferents from and afferents to the region under study, chemical stimulation of the dlPAG with bicuculline, which activates only post-synaptic neurons, was also performed in this work.

Kopf, USA). A brain electrode was implanted in the midbrain, aimed at the dorsolateral PAG. The electrode was made of stainless steel wire, 160 ␮m in diameter, insulated except at the cross-section of the tip. The upper incisor bar was set at 3.3 mm below the interaural line so that the skull was horizontal between bregma and lambda. The electrode was introduced at an angle of 16◦ using the following coordinates with the lambda serving as the reference for each plane: medio-lateral, 1.9 mm; and dorso-ventral, 5.1 mm [36]. For microinjections of bicuculline and saline, a stainless steel guide-cannula (13 mm length, o.d. 0.6 mm, i.d. 0.4 mm) was similarly implanted, introduced vertically at 2.5 mm at the dorso-ventral plane. The electrode or cannula was fixed to the skull by means of acrylic resin and the three stainless steel screws. At the end of surgery each guide-cannula was sealed with a stainless steel wire to protect it from congestion. 2.3. Microinjections The animals were gently wrapped in a cloth, hand-held and a thin dental needle (o.d. 0.3 mm) was introduced through the guide-cannula until its lower end was 1 mm below its tip, reaching the same depth as the electrode tip. The injection needle was linked to a 5 ␮l Hamilton syringe by means of polyethylene tubing connected to a microinfusion apparatus (Harvard, USA). A volume of 0.2 ␮l was injected during 20 s. Following the end of the injection the microinjection needle was held inside the brain for 10 s. 2.4. Drugs Bicuculline hydrochloride (Sigma, USA) was dissolved in physiological saline (0.9%) shortly before use. The dose of bicuculline used was of 40 ng/0.2 ␮l injected into left dlPAG. Physiological saline served as vehicle control for dlPAG microinjections.

2. Materials and methods 2.5. Apparatus 2.1. Animals Naive male Wistar rats weighing 240–260 g were used. Animals were kept under controlled temperature (22 ± 2 ◦ C) and a 12-h light:12-h dark cycle (lights on at 07:00 h). They were housed in two per cage and had free access to food and water throughout the experiment. The experiments were conducted between 9.00 and 12.00 a.m. The experiments were performed in compliance with the recommendations of SBNeC (Brazilian Society of Neuroscience and Behavior), which are based on the US National Institutes of Health Guide for Care and Use of Laboratory Animals. 2.2. Surgery The animals were anaesthetized with tribromoethanol (250 mg/kg, i.p.) and fixed in a stereotaxic frame (David

Six days after surgery, the rats were placed in an open field, which was a circular enclosure 60 cm in diameter and 50 cm high. The rat was placed in the arena and had its brain electrode connected to a flexible wire cable, allowing ample movement inside the box. The cable, in turn, was connected to the stimulator by means of a mercury swivel mounted on the top of the experimental chamber. The rats were allowed a 15 min period of habituation in the enclosure. The brain was stimulated electrically by means of a sine wave stimulator [35]. The stimulation current was monitored by measuring the voltage drop across a 1 k
resistor with an oscilloscope (Philips, USA). Brain stimulation (ac, 60 Hz, 15 s) was presented at 1 min intervals with the current intensity increasing by steps of 5 ␮A for measurements of the freezing and escape thresholds. Freezing threshold was defined as the lowest intensity producing cessation of the

D.M.L. Vianna et al. / Brain Research Bulletin 62 (2003) 179–189

ongoing behavior in two consecutive series of electrical stimulation, accompanied by at least two of the following autonomic responses: piloerection, defecation, micturition and exhophtalmus. Escape threshold was defined as the lowest current intensity that produced running or jumping in two successive ascending series of electrical stimulation. Animals with an escape threshold above 120 ␮A (peak-to-peak) were discarded from the experiment. 2.6. Experimental procedure 2.6.1. Electrical stimulation On the seventh day after surgery, animals were randomly allocated to the freezing (N = 8) and escape (N = 10) groups. A control group consisted of rats submitted to the same procedure, except that no current was applied to the electrodes (N = 9). Animals of the three groups were placed in the arena and allowed 15 min for habituation. Next, the aversive thresholds were determined in one series of ascending intensity of electrical stimulation, and then the animals were returned to their home boxes. On the next day, electrical stimulation was applied at the same aversive threshold determined the day before. The animals were returned to their home cages and observed for 2 h. No apparent behavioral changes could be observed during this period. Next, their brains were processed in parallel, as described below. 2.6.2. Chemical stimulation On the seventh day after surgery, animals were randomly allocated to the bicuculline (N = 8) and saline (N = 8) groups. Animals of these groups were placed in the arena and allowed 15 min for habituation. Next, injections of saline or bicuculline (40 ng/0.2 ␮l) were done into the dlPAG and the behaviors of the animals were recorded for 30 min. Afterwards, the animals were returned to their home boxes and observed for 90 min. Next, their brains were processed in parallel, as described below. 2.7. Fos protein immunohistochemistry About 2 h after the electrical stimulation procedure, the animals were deeply anaesthetized with urethane (1.25 g/kg; Sigma, USA) and transcardially perfused with 0.9% saline followed by 4% paraformaldehyde in 0.1 M phosphate-buffered saline (PBS) (pH 7.4). Brains were removed, immersed (4 ◦ C) in the above fixative for 2 h and then kept in 30% sucrose in 0.1 M PBS until soaked. They were then quickly frozen in isopentane (−40 ◦ C) and sliced by the use of a cryostat (−15 ◦ C). Two adjacent series of 40-␮m thick brain slices were obtained, having as reference the following AP coordinates: bregma +3.7, +1.0, −1.8, −2.7, −4.2, −4.3, −5.8, −6.8, −7.6, −8.7, −9.3 mm. The sections were collected in 0.1 M PBS and subsequently processed free-floating according to the avidine–biotine procedure, using the Vectastain ABC Elite peroxidase rabbit IgG kit (Vector, USA, ref. PK 6101). All reactions were carried

181

out under agitation at room temperature. The sections were first incubated with 1% H2 O2 for 10 min, washed four times with 0.1 M PBS (5 min each) and then incubated overnight with the primary Fos polyclonal antibody (Santa Cruz, USA, SC-52) at a concentration of 1/20,000 in 0.1 M PBS enriched with 4% normal goat serum and 0.2% Triton-X. Sections were again washed three times (5 min each) with 0.1 M PBS and incubated for 1 h with biotinylated goat antirabbit antibody (one drop from the kit for 10 ml of 0.1 M PBS enriched with 4% normal goat serum and 0.2% Triton-X). After another series of three 5-min washings in 0.1 M PBS, they were incubated for 1 h with the avidine-peroxidase solution (two drops of each solution labeled A and B of the kit for 25 ml of 0.1 M PBS) and three times washed in 0.1 M PBS (5 min per wash). The sections were finally left for 5 min in a solution of 3,3 -di-aminobenzidine (DAB, 0.02%) to which hydrogen peroxide (0.04%) was added just prior to use and were washed twice with 0.1 M PBS. 2.8. Quantification of Fos-positive cells Mounted sections of the tissue were photographed with the help of an Olympus BX-50 microscope with a digital camera attached to it (Hamatsu Photonics C2400). Neuronal nuclei expressing levels of DAB reaction product above tissue background were counted by a computerized image analysis system (Image Pro Plus 4.0, Media Cybernetics, USA), as described elsewhere [34,47]. Briefly, counting of Fos-positive cells was performed at a magnification 100×. An area of the same shape and size per brain region was used for each rat. The same light and threshold conditions were employed for all sections. Fos staining could vary from one area to another. However, in order to ensure accuracy of measurement and avoid variations among same areas in different subjects, the background of each area was measured and digitally subtracted from the area to be measured. Accordingly, the threshold conditions were set for each area and maintained for all subjects [34]. All brain regions were bilaterally counted in various sections for each rat depending on the size of the structure. After that, counts for the same section indicated in Fig. 1 according to the Paxinos and Watson atlas were taken. The number of positive nuclei that fell within a 1.0 mm × 0.8 mm viewing field area (0.8 mm2 ) in each region of interest was counted. In several cases, the designated area was substantially larger than the boundaries of the viewing field. In such cases, the viewing field was placed in a fixed position within the region of interest relative to known anatomical landmarks. In other cases, the designated area to be counted was smaller than the boundaries of the viewing field. In such cases, only the region of interest, and not extraneous areas, was counted. Nuclei were counted individually and expressed as number of Fos-positive nuclei per 0.1 mm2 [34]. The nomenclature and nuclear boundaries utilized were based on the atlas of Paxinos and Watson [36] and the planes of analyzed sections were standardized as far as possible. Each region was

182

D.M.L. Vianna et al. / Brain Research Bulletin 62 (2003) 179–189

Fig. 1. Schematic diagrams, adapted from the atlas of Paxinos and Watson [36], showing the 12 areas in which Fos immunoreactivity was quantified. Open squares indicate the position of the viewing field within which the Fos counts were made. For explanation of abbreviations see Table 1.

bilaterally counted. This procedure has been used in other studies of this laboratory [34]. 2.9. Statistical analysis The data are presented as mean ± S.E.M. and were analyzed by means of a three-way analysis of variance (ANOVA), with structures, groups of stimulation (sham, freezing and escape), and side of stimulation (ipsi/contra-

lateral to the electrode placement) as the factors. For each structure this analysis was followed by a two-way ANOVA with groups as between-group factor (sham, freezing and escape) and side of stimulation (ipsi and contralateral) as within-group factor. Newman–Keuls post hoc test was done when appropriate. The same analysis was performed for the bicuculline data with the only difference being the groups (saline and bicuculline). The significance level was set at P equal to 0.05.

D.M.L. Vianna et al. / Brain Research Bulletin 62 (2003) 179–189

183

Fig. 2. Sites of stimulation electrodes in the dorsolateral periaqueductal gray with reference to the Paxinos and Watson atlas [36]. Animals submitted to electrical stimulation at freezing (䊊) or escape (䊉) thresholds. Shams are not depicted.

and 64.00 ± 2.41 (peak-to-peak), respectively. The freezing response was characterized by immobility for at least 6 s accompanied by piloerection, arching back, lowered ears and autonomic responses such as micturition, defecation or exophthalmus. The escape behavior was characterized as running, galloping or jumpings and the same autonomic responses as in the freezing behavior. Student’s t-test performed on these thresholds data revealed that they were

3. Results 3.1. Behavioral effects The tips of the electrodes were situated inside or in the immediate vicinity of the dlPAG (see Fig. 2). The intensity of the electric current applied to the dlPAG of the animals to induce freezing and escape responses was 38.13 ± 0.81 Table 1 Data represents the mean ± S.E.M. of 5–10 animals Sham

Freezing

Ipsi LC IC CnF vlPAG lPAG dlPAG dmPAG PMd VMHdm CeA Cg PrL

18.38 67.20 26.44 70.67 30.78 24.33 16.56 15.29 10.44 2.00 69.56 84.00

Contra ± ± ± ± ± ± ± ± ± ± ± ±

1.21 6.99 3.90 3.82 2.07 0.67 0.94 0.99 1.19 0.23 7.62 16.54

18.88 96.60 28.44 76.00 32.22 23.89 11.44 15.29 9.89 4.89 59.78 89.63

± ± ± ± ± ± ± ± ± ± ± ±

Escape

Ipsi 1.30 9.84 3.91 3.44 1.45 0.71 0.68 1.26 0.96 0.55 6.95 12.55

16.13 54.71 33.57 72.25 40.75 62.50 30.63 39.29 5.57 4.63 74.38 81.88

Contra ± ± ± ± ± ± ± ± ± ± ± ±

1.17 5.96 4.43 4.24 2.29 6.88∗ 2.46∗ 3.71∗ 0.60 0.60 7.47 10.25

17.88 68.29 33.38 61.63 38.88 28.63 20.75 47.86 4.43 5.88 70.38 115.00

Ipsi ± ± ± ± ± ± ± ± ± ± ± ±

1.12 7.78 3.70 4.32 1.46 1.43 1.14∗ 7.01∗ 0.57 0.92 5.88 15.15

10.86 48.33 50.00 75.10 34.90 126.60 57.60 62.60 43.20 5.60 40.50 78.80

Contra ± ± ± ± ± ± ± ± ± ± ± ±

0.66 6.77 2.54∗ 2.66 1.90 9.88∗∗ 3.83∗∗ 4.96∗ 5.06∗ 0.57 4.93 12.78

15.43 45.89 30.90 63.40 34.90 28.50 20.30 53.60 19.30 8.33 43.50 98.70

± ± ± ± ± ± ± ± ± ± ± ±

0.84 4.83 2.18 2.30 2.22 1.63∗∗ 1.60∗∗ 3.87∗ 0.98∗ 1.19 5.11 15.96

Male Wistar rats were placed in an open arena for 15 min and then electrically stimulated in the dlPAG at intensities sufficient to produce freezing or escape behavior, or not stimulated at all (sham). The animals were killed under deep anesthesia 2 h after the behavioral tests and had their brain removed and processed for the detection of Fos-like immunoreactivity. The number of Fos-positive neurons per region in the ipsilateral (ipsi) or contralateral (contra) side in relation to the electrode placement was counted using a computerized image analysis system (Image Pro Plus 4.0, Media Cybernetics). Abbreviations: CeA: central amygdaloid nucleus; IC: inferior colliculus; Cg: cingulate cortex; CnF: cuneiform nucleus; dlPAG: dorsolateral periaqueductal gray; dmPAG: dorsomedial periaqueductal gray; LC: locus coeruleus; PMd: dorsal premammillary nucleus; PrL: prelimbic cortex; vlPAG: ventrolateral periaqueductal gray; VMHdm: ventromedial hypothalamic nucleus, dorsomedial part. ∗ P < 0.05 in relation to sham group by means of ANOVA followed by Newman–Keuls test. ∗∗ P < 0.01 in relation to sham group by means of ANOVA followed by Newman–Keuls test.

184

D.M.L. Vianna et al. / Brain Research Bulletin 62 (2003) 179–189

statistically different (t = 3.25; P < 0.01). Animals injected with bicuculline into the dlPAG presented a behavioral activation immediately after the injections. This activation was characterized by escape behavior interspersed with freezing behavior for about 20 min. 3.2. Fos protein expression Immunoreactive cells exhibited a dark nucleus in neuronal nuclei expressed clearly upon the surrounding background

tissue. In the sham-stimulated group, clusters of Fos-like immunoreactive cells could already be identified in some brain regions, namely the frontal cortex, inferior colliculus and ventrolateral periaqueductal gray. The results obtained with quantitative analyses of Fos-like immunoreactivity in the brain regions studied after induction of freezing and escape responses in rats stimulated in the dlPAG are summarized in Table 1. Three-way ANOVA applied on these data revealed that there were significant effects in Fos labeling of the studied structures [F(11, 537) = 14.63;

Fig. 3. Photomicrographs of Fos-immunoreactive cells (dark dots) in coronal sections of the midbrain and diencephalons following electrical stimulation of the dlPAG. The left side of the pictures corresponds to the side where the electrode was placed. See Table 1 for abbreviations. Scale bar, 200 ␮m.

D.M.L. Vianna et al. / Brain Research Bulletin 62 (2003) 179–189

P < 0.001], groups [F(2, 520) = 3.86; P < 0.05] and no significant effect of side of stimulation [F(1, 520) = 2.20; P > 0.05]. Next, the results of the two-way ANOVA will be presented first for the structures that showed significant Fos expression following electrical stimulation of the dlPAG at the escape and then at the freezing threshold. Sample micrographs of target regions with the resulting immunoreaction can be seen in Fig. 3. 3.2.1. Escape Two-way ANOVA showed that there was a significant difference between stimulation [F = 5.74; P < 0.05], side [F = 10.87; P < 0.05] and interaction [F = 4.89; P < 0.05] after stimulation of dlPAG at the escape threshold (Fig. 3F). The same occurred with respect to dmPAG (Fig. 3F). There was a main effect of stimulation [F = 7.14; P < 0.005], side [F = 9.81; P < 0.05] and interaction between stimulation and side [F = 3.46; P < 0.05]. However, no significant effects could be detected in the lPAG (Fig. 3F) and vlPAG (Fig. 3C). Stimulation at the escape threshold also caused significant increase in Fos labeling in CnF (Fig. 3C). There was a main effect of stimulation at this nucleus [F = 5.44; P < 0.05]. Stimulation also produced a higher density of Fos-immunoreactive neurons in PMd [F = 5.81; P < 0.05] (Fig. 3I) and VMHdm [F = 6.20; P < 0.05] (Fig. 3M). Table 2 presents a complete account of all statistical analysis performed per structure in this work. 3.2.2. Freezing Stimulation of dorsal regions of PAG at the freezing threshold produced a significant increase in Fos-immunoreactivity in this nucleus itself [F = 5.74; P < 0.05] and of the dmPAG [F = 7.14; P < 0.05] (Fig. 3E). As occurred with stimulation at escape threshold, Fos counts in the LC, lPAG, inferior colliculus, CeA, Cg and PrL of rats stimulated at the freezing threshold were not different from the sham-control groups (Table 2). 3.2.3. Bicuculline Three-way ANOVA applied on these data revealed that there were significant effects in Fos labeling of the studied structures [F(11, 240) = 11.28; P < 0.001], treatments [F(1, 240) = 115.32; P < 0.05] and significant effect of side of injection [F(1, 240) = 32.12; P > 0.001]. Significant interactions occurred between structures and treatments [F(11, 240) = 6.34; P > 0.001] and treatments and side [F(1, 240) = 13.83; P > 0.001]. Two-way ANOVA showed that there was a significant difference between treatments [F(1, 17) = 10.86; P < 0.005], side [F(1, 17) = 7.72; P < 0.05] and interaction [F(1, 17) = 7.58; P < 0.05] after bicuculline microinjections into the dlPAG. The same occurred with respect to dmPAG. There was a main effect of stimulation [F(1, 17) = 12.65; P < 0.005], side [F(1, 17) = 8.35; P < 0.01] and interaction between stimulation and side [F(1, 17) = 4.81; P < 0.05]. dlPAG stimulation with

185

Table 2 Two-way ANOVA results for the 12 structures studied

LC

IC

CnF

vlPAG

lPAG

dlPAG

dmPAG

PMd

VMH

CeA

Cg

PrL

Stimulation Side Side × stimulation Stimulation Side Side × stimulation Stimulation Side Side × stimulation Stimulation Side Side × stimulation Stimulation Side Side × stimulation Stimulation Side Side × stimulation Stimulation Side Side × stimulation Stimulation Side Side × stimulation Stimulation Side Side × stimulation Stimulation Side Side × stimulation Stimulation Side Side × stimulation Stimulation Side Side × stimulation

F

d.f.

P

1.89 0.81 0.21 1.77 1.36 0.20 5.44 0.55 1.18 0.20 0.49 0.47 0.93 0.01 0.02 5.74 10.87 4.89 7.14 9.81 3.46 5.81 0.01 0.24 6.20 1.66 1.47 1.24 1.28 0.05 1.42 0.05 0.06 0.04 0.17 0.09

2, 1, 2, 2, 1, 2, 2, 1, 2, 2, 1, 2, 2, 1, 2, 2, 1, 2, 2, 1, 2, 2, 1, 2, 2, 1, 2, 2, 1, 2, 2, 1, 2, 2, 1, 2,

0.22 0.37 0.80 0.18 0.24 0.81 0.007∗ 0.46 0.32 0.81 0.48 0.62 0.40 0.97 0.96 0.005∗ 0.001∗ 0.01∗ 0.001∗ 0.002∗ 0.04∗ 0.006∗ 0.98 0.78 0.004∗ 0.20 0.23 0.30 0.26 0.94 0.25 0.82 0.94 0.96 0.68 0.91

40 40 40 40 40 40 42 42 42 48 48 48 48 48 48 48 48 48 48 48 48 42 42 42 46 46 46 44 44 44 48 48 48 42 42 42

See Table 1 for abbreviations and details. ∗ P < 0.05.

bicuculline also led to a significant increase in Fos labeling in CnF. There was a significant interaction between stimulation and side of stimulation at this nucleus [F(1, 17) = 3.74; P < 0.05]. Stimulation also produced a higher density of Fos-immunoreactive neurons in PMd [F(1, 15) = 7.84; P < 0.05] (Fig. 4F) and VMHdm [F(1, 17) = 4.67; P < 0.05].

4. Discussion The present results show that the stepwise increase in the electrical stimulation of the dorsal half of PAG elicits the characteristic pattern of behavioral activation with alertness, freezing and escape accompanied by autonomic responses, such as piloerection, exophthalmus, micturition and defecation. This stimulation led to increased Fos expression in the dmPAG and dlPAG, confirming previous reports that show Fos immunoreactivity after chemical and electrical stimulation of these regions [41,42]. The increase of Fos expression

186

D.M.L. Vianna et al. / Brain Research Bulletin 62 (2003) 179–189

Fig. 4. Photomicrographs of Fos-immunoreactive cells (dark dots) in coronal sections of the midbrain and diencephalons following microinjections of bicuculline (40 ng/0.2 ␮l) into the dlPAG. The left side of the pictures corresponds to the side where the microinjections were made. See Table 1 for abbreviations. Scale bar, 200 ␮m.

in the dmPAG and dlPAG, but not in the lPAG or vlPAG, is in harmony with the current view of parcellation of PAG into distinct subdivisions with different functional roles [3,15]. In agreement with this, the dlPAG of rats placed into semi-natural settings is preferentially activated when the animal encounters its predator [13], while the lPAG is activated when the animal is engaged in predatory behavior

[18]. On the other hand, the vlPAG is preferentially activated when the animal is under a classical fear-conditioning paradigm [17,50]. Therefore, our results are in line with the studies that indicate a role of dlPAG and dmPAG in the production of defensive behaviors during threatening situation related to immediate dangers, as when the rat faces the predator.

D.M.L. Vianna et al. / Brain Research Bulletin 62 (2003) 179–189

Electrical stimulation of the dlPAG at the escape but not at freezing threshold led to a significant increase in Fos expression in the VMHdm. This result is in line with behavioral studies that show that electrical or chemical stimulation of this hypothalamic area elicits escape, but not freezing, behavior [8,28,44]. Indeed, it has been proposed that when the animals encounter predators, dlPAG, medial hypothalamus and amygdala act in concert in the production of escape behavior or fight, when flight is not possible [6,28]. The CnF is thought to be the main relay station for dlPAG- and SC-mediated defensive responses and route to the medulla oblongata and spinal cord [20,30]. These defense-specific projections course through the midbrain tegmentum, without crossing the midline, to the ipsilateral CnF. A second pathway from the SC to the CnF, crossing the midline, also exists, but it is related to orientation and not to defensive responses [20]. Accordingly, our results showed that only the ipsilateral CnF presented significant increases in Fos immunoreactivity after escape-producing stimulation. Freezing and escape elicited by exposure to a predator leads to a rise in the expression of Fos in the PMd, and is abolished after its lesion [12]. The PMd is one of the main sources of afferent to the dlPAG [14], and is thought to be a key structure in the mediation of innate fear responses [11]. Our results showed that escape-eliciting stimulation enhanced Fos expression in the PMd, bilaterally. It has been suggested that the route from dlPAG to PMd could involve the anterior hypothalamus and VMHdm [11]. Another important source of information to the PAG is the CeA [39]. But unlike the PMd, the CeA targets primarily the vlPAG, and seems to be involved in the production of conditioned freezing [16]. In the present study there was no significant change in Fos expression in these nuclei following dlPAG electrical stimulation at freezing and escape thresholds. Therefore, the present findings support the notion that this CeA-vlPAG circuit is not involved in the expression of dlPAG-mediated defensive responses. In contrast with this, using the same technique employed in the present work, it was found that Fos expression increases in the CeA after stimulation of the IC [34]. Therefore, the fact that in the present study Fos increases in neither CeA nor IC were found can mean that different neural networks are activated by electrical stimulation of dlPAG and IC. Indeed, it has been noted that the neural organization of fear-like behaviors in mesencephalic structures is dependent on multiple factors, including the kind of aversive sensory stimuli that trigger the defensive behavioral repertoire, whether visual, acoustic or tactile [9,48]. The hypothesis that freezing and escape responses activate different sets of brain structures was confirmed by the present results. Escape-provoking stimulation caused increased Fos expression in the VMH, PMd and in the CnF, as well as in the dmPAG and dlPAG. Conversely, freezing-provoking stimulation led to a significant increase of Fos expression only in the dmPAG, while marginal in-

187

creases were observed in some other structures (PMd, dlPAG and vlPAG). It was proposed [48] that there are at least two paths for PAG-mediated defensive responses, one involving VMHdm–PMd–dlPAG–CnF and mediating responses to immediate danger, and a second one comprising CeA–vlPAG, more related to cue-elicited responses, either learned or innate. In this context, freezing and escape behaviors were produced by dlPAG stimulation in rats bearing vlPAG lesions [49]. Thus, vlPAG is not critical for the generation and elaboration of the defensive responses induced by dlPAG stimulation. This result supports the notion that dissociated neural circuits subserve different kinds of defensive behaviors. In agreement with this, increases in Fos expression in vlPAG are much more pronounced after conditioned fear [17] than after exposure to a predator [13]. On the contrary, exposure to a predator led to a clear predominance of dlPAG activation [13]. In line with these reports in the present study c-fos expressed more in the dorsal than in the ventral portions of the PAG, suggesting a predominant role of the former regions in the production of defensive responses to immediate or proximal danger. Overall, our results suggest that vlPAG could have a limited role in the mediation of dlPAG-provoked freezing, but not escape, while VMHdm and CnF have a predominant role in the mediation of escape, but not of freezing behavior. From the present results it can be suggested that PMd is implicated in the generation and elaboration of escape responses and also in the freezing induced by dlPAG stimulation. It is well established that cortical inputs to the PAG exert a control on the neural operations of this midbrain structure [3,27,45,46]. In the present study no significant change in Fos expression was detected in the cortical structures studied herein (Cg, PrL) in the mediation of freezing and escape responses elicited by dlPAG stimulation. However, a significant increase in Fos expression was reported at the Cg level after microinjection of nitric oxide donors into the dlPAG, which also causes defensive responses [21]. These apparently conflicting results could be due to differences in the techniques employed for stimulating the dlPAG, namely electrical versus chemical stimulation. Anyway, a possibility that emerges from the present results is that cortical structures have a one-way control over the neural substrates of aversion in the dorsal PAG. Indeed, to our knowledge, ascending projections to the cortical regions from the dorsal columns of the PAG have not been clearly characterized so far [3,27,48]. In agreement with this view, it has been reported that the onset of an unexpected panic attack is related to an increase of the regional cerebral blood flow in the hypothalamus, but not in the medial prefrontal cortex [29], which suggests that the activation of the hypothalamic–PAG axis is not necessarily related to prefrontal functioning. The activation of this axis, without the participation of cortical structures, could possibly be a source for panic disorder [45]. Interestingly, increased extracellular levels of dopamine in the prefrontal cortex, which has been considered as a biological marker for stressful

188

D.M.L. Vianna et al. / Brain Research Bulletin 62 (2003) 179–189

conditions, has been observed following aversive stimulation of the IC, but not following similar stimulation at dorsal PAG [19]. It cannot ruled out that the lack of significant change in Fos expression in certain regions could be due to limitations of the technique employed, as negative results of Fos immunoreactivity cannot guarantee that a given brain region is not involved in the generation of a response under study [24]. Moreover, a weak effect could have been masked by the fact that even our sham-stimulated animals exhibited appreciable number of Fos-immunoreactive cells in some regions. Besides, electrical stimulation has the disadvantage of exciting not only perikarya but also axons and fibers of passage. Rats injected with bicuculline exhibited freezing interspersed with escape behavior. Bicuculline microinjections into the dlPAG confirm the overall pattern of results obtained with electrical stimulation. Additionally, this combined behavioral reaction causes some distinctive labeling such as in the vlPAG, locus coeruleus and CeA. This pattern of effects, which is different from that obtained with dlPAG electrical stimulation, is probably related to the type of responses obtained with each procedure. Whereas electrical stimulation procedure can be switched off as soon as the response is obtained, the defense reaction elicited by microinjections of bicuculline into the dlPAG lasts as long as the drug acts so that freezing or escape alone can not be elicited by this procedure. Therefore, while dlPAG injections of bicuculline cause a full-blown reaction associated with a widespread activation of the neural substrates of aversion, the ease of eliciting distinctive responses of freezing and escape by electrical stimulation of the dlPAG allows a more specific pattern of c-fos expression associated to freezing or escape responses. This assumption is corroborated by the widespread activation of brainstem structures involved in the organization of fear following microinjections of bicuculline into the dlPAG. In view of these results we have now undertaken the task to examine the effects of excitatory amino acid compounds including compounds that stimulate only AMPA/kainate receptors and produce freezing and others that stimulate NMDA receptors and produce escape behavior. Even considering the present methodological limitations, taking into account our results we suggest that VMHdm, PMd and the CnF are part of a network activated during dlPAG-mediated defensive responses and that dlPAG and PMd mediates a particular kind of freezing that seems not to be shared by the neural substrates of the vlPAG and central nucleus of amygdala.

Acknowledgements Research supported by FAPESP (No. 02/03705-0) and CNPq (No. 140164/1999-5). D.M.L. Vianna was the recipient of a fellowship from CNPq during the execution of this work.

References [1] J.M. Allman, A. Hakeem, J.M. Erwin, E. Nimchinski, P. Hof, The anterior cingulate cortex: the evolution of an interface between emotion and cognition, Ann. N.Y. Acad. Sci. 935 (2001) 107–117. [2] R. Bandler, A. Depaulis, M. Vergnes, Identification of midbrain neurones mediating defensive behaviour in the rat by microinjections of excitatory amino acids, Behav. Brain Res. 15 (1985) 107– 119. [3] R. Bandler, M.T. Shipley, Columnar organization in the midbrain periaqueductal gray: modules for emorional expression? Trends Neurosci. 17 (1994) 379–389. [4] C.H.M. Beck, H.C. Fibiger, Conditioned fear-induced changes in behavior and in the expression of the immediate early gene c-fos: with and without diazepam pretreatment, J. Neurosci. 15 (1) (1995) 709–720. [5] D.C. Blanchard, R.J. Blanchard, Ethoexperimental approaches to the biology of emotion, Annu. Rev. Psychol. 39 (1988) 43–68. [6] M.L. Brandão, V.Z. Anseloni, J.E. Pandóssio, J.E. De Araújo, V.M. Castilho, Neurochemical mechanisms of the defensive behavior in the dorsal midbrain, Neurosci. Biobehav. Rev. 23 (1999) 863–875. [7] M.L. Brandão, J.C. de Aguiar, F.G. Graeff, GABA mediation of the anti-aversive action of minor tranquilizers, Pharmacol. Biochem. Behav. 16 (1982) 397–402. [8] M.L. Brandão, G. DiScala, M.J. Bouchet, P. Schmitt, Escape behavior produced by the blockade of glutamic acid decarboxilase (GAD) in mesencephalic central gray or medial hypothalamus, Pharmacol. Biochem. Behav. 24 (1986) 497–502. [9] M.L. Brandão, A.C. Troncoso, M.A. De Souza Silva, J.P. Huston, The relevance of neuronal substrates of defense in the midbrain tectum to anxiety and stress: empirical and conceptual considerations, Eur. J. Pharmacol. 28:463 (1–3) (2003) 225–233. [10] A.W. Bruijnzeel, R. Stam, J.C. Compaan, G. Croiset, L.M.A. Akkermans, B. Olivier, V.M. Wiegant, Long-term sensitization of Fos-responsivity in the rat central nervous system after a single stressful experience, Brain Res. 819 (1999) 15–22. [11] N.S. Canteras, The medial hypothalamic defensive system: hodological organization and functional implications, Pharmacol. Biochem. Behav. 71 (2002) 481–491. [12] N.S. Canteras, S. Chiavegatto, L.E. Ribeiro-do-Valle, L.W. Swanson, Severe reduction of rat defensive behavior to a predator by discrete hypothalamic chemical lesions, Brain Res. Bull. 44 (3) (1997) 297– 305. [13] N.S. Canteras, M. Goto, Fos-like immunoreactivity in the periaqueductal gray of rats exposed to a natural predator, Neuroreport 10 (1999) 413–418. [14] N.S. Canteras, L.W. Swanson, The dorsal premammillary nucleus: an unusual component of the mammillary body, Proc. Natl. Acad. Sci. U.S.A. 89 (1992) 10089–10093. [15] P. Carrive, The periaqueductal gray and defensive behavior: functional representation and neuronal organization, Behav. Brain Res. 58 (1993) 7–47. [16] P. Carrive, J. Lee, A. Su, Lidocaine blockade of amygdala output in fear conditioned rats reduces fos-expression in the ventrolateral periaqueductal gray, Neuroscience 95 (2000) 1071–1080. [17] P. Carrive, P. Leung, J. Harris, G. Paxinos, Conditioned fear to context is associated with increased fos expression in the caudal ventrolateral region of the midbrain periaqueductal gray, Neuroscience 78 (1) (1997) 165–177. [18] E. Comoli, E.R. Ribeiro-Barbosa, N.S. Canteras, Predatory hunting and exposure to live predator induce opposite patterns of Fos immunoreactivity in the PAG, Behav. Brain Res. 138 (2003) 17–28. [19] G. Cuadra, A. Zurita, C.E. Macedo, V.A. Molina, M.L. Brandão, Electrical stimulation of the midbrain tectum enhances dopamine release in the frontal cortex, Brain Res. Bull. 52 (2000) 413– 418.

D.M.L. Vianna et al. / Brain Research Bulletin 62 (2003) 179–189 [20] P. Dean, P. Redgrave, G.W.M. Westby, Event or emergency? Two response systems in the mammalian superior colliculus, Trends Neurosci. 12 (4) (1989) 137–147. [21] R.W. de Oliveira, E.A. Del Bel, F.S. Guimarães, Behavioral and c-fos expression changes induced by nitric oxide donors microinjected into the dorsal periaqueductal gray, Brain Res. Bull. 51 (2000) 457–464. [22] R. Depoortere, G. Sandner, G. Di Scala, Aversion induced by electrical stimulation of the mesencephalic locomotor region in the intact and freely moving rat, Physiol. Behav. 47 (3) (1990) 561–567. [23] R.A. Dielenberg, G.E. Hunt, I.S. McGregor, ‘When a rat smells a cat’: the distribution of Fos immunoreactivity in rat brain following exposure to a predatory odor, Neuroscience 104 (4) (2001) 1085– 1097. [24] M. Dragunow, R. Faull, The use of c-fos as a metabolic marker in neuronal pathway tracing, J. Neurosci. Methods 29 (1989) 261–265. [25] G.E. Duncan, D.J. Knapp, G.R. Breese, Neuroanatomical characterization of Fos induction in rat behavioral models of anxiety, Brain Res. 713 (1996) 79–91. [26] M. Ennis, M. Behbehani, M.T. Shipley, E.J. Van Bockstaele, G. Aston-Jones, Projections from the periaqueductal gray to the rostromedial pericoerulear region and nucleus locus coeruleus: anatomic and physiologic studies, J. Comp. Neurol. 306 (1991) 480–494. [27] N.S. Floyd, J.L. Price, A.T. Ferry, K.A. Keay, R. Bandler, Orbitomedial prefrontal cortical projections to distinct longitudinal columns of the periaqueductal gray in the rat, J. Comp. Neurol. 422 (2000) 556–578. [28] F.G. Graeff, Neuroanatomy and neurotransmitter regulation of defensive behaviors and related emotions in mammals, Braz. J. Med. Biol. Res. 27 (1994) 811–829. [29] M. Javanmard, J. Shlik, S.H. Kennedy, F.J. Vaccarino, S. Houle, J. Bradwejn, Neuroanatomic correlates of CCK-4-induced panic attacks in healthy humans: a comparison of two time points, Biol. Psychiatry 45 (7) (1999) 872–882. [30] K.A. Keay, R. Bandler, Parallel circuits mediating distinct emotional coping reactions to different types of stress, Neurosci. Biobehav. Rev. 25 (2001) 669–678. [31] K.J. Kovács, c-fos as a transcriptional factor: a stressful (re)view from a functional map, Neurochem. Int. 33 (1998) 287–297. [32] J.E. Krieger, F.G. Graeff, Defensive behavior and hypertension induced by glutamate in the midbrain central gray of the rat, Braz. J. Med. Biol. Res. 18 (1985) 61–67. [33] L. Lacroix, S. Spinelli, C.A. Heidbreder, J. Feldom, Differential role of the medial and lateral prefrontal cortices in fear and anxiety, Behav. Neurosci. 114 (6) (2000) 1119–1130. [34] M.R. Lamprea, F.P. Cardenas, D.M. Vianna, V.M. Castilho, S.E. Cruz-Morales, M.L. Brandão, The distribution of fos immunoreactivity in rat brain following freezing and escape responses elicited by electrical stimulation of the inferior colliculus, Brain Res. 950 (1/2) (2002) 186–194. [35] R.F. Marseillan, A solid state sine wave stimulator, Physiol. Behav. 19 (1977) 339–340.

189

[36] G. Paxinos, C. Watson, The Rat Brain in Stereotaxic Coordinates, Academic Press, New York, 1997. [37] K.W. Powell, J.B. Hatton, Projections of the inferior colliculus in cat, J. Comp. Neurol. 136 (1969) 183–192. [38] P. Redgrave, P. Dean, I.J. Mitchell, A. Odekunle, A. Clark, The projection from superior colliculus to cuneiform area in the rat I, Anat. Stud. Exp. Brain Res. 72 (1988) 611–625. [39] T.A. Rizvi, M. Ennis, M. Behbehani, M.T. Shipley, Connections between the central nucleus of the amygdala and the midbrain periaqueductal gray: topography and reciprocity, J. Comp. Neurol. 303 (1991) 121–131. [40] T.E. Robinson, Electrical stimulation of the brain stem in freely moving rats: effects on behavior, Physiol. Behav. 21 (1978) 223–231. [41] G. Sandner, G. Di Scala, B. Rocha, M.J. Angst, c-fos immunoreactivity in the brain following unilateral stimulation of the dorsal periaqueductal gray in freely moving rats, Brain Res. 573 (1992) 276–283. [42] G. Sandner, P. Oberling, M.C.L. Silveira, G. Di Scala, B. Rocha, A. Bagri, R. Deportere, What brain structures are active during emotions? Effects of brain stimulation elicited aversion on c-fos immunoreactivity and behavior, Behav. Brain Res. 58 (1993) 9–18. [43] L.C. Schenberg, M.B. Costa, P.C.L. Borges, F.S. Castro, Logistic analysis of the defense reaction induced by electrical stimulation of the rat mesencephalic tectum, Neurosci. Biobehav. Rev. 14 (1990) 473–479. [44] P. Schmitt, G. DiScala, M.L. Brandão, P. Karli, Behavioral effects of microinjections of SR95103 a new GABA—a antagonist into medial hypothalamus or the mesencephalic central gray, Eur. J. Pharmacol. 117 (1985) 149–158. [45] T.V. Sewards, M.A. Sewards, Fear and power-dominance drive motivation: neural representations and pathways mediating sensory and mnemonic inputs and outputs to premotor structures, Neurosci. Biobehav. Rev. 26 (2002) 553–579. [46] M.T. Shipley, M. Ennis, T.A. Rizvi, M.M. Behbehani, in: A. Depaulis, R. Bandler (Eds.), Topographical Specificity of Forebrain Input to the Midbrain Periaqueductal Gray: Evidence for Discrete Longitudinally Organized Input Columns, Plenum Press, New York, 1991, pp. 417–448. [47] M.C.L. Silveira, H. Zangrossi, M.B. Viana, R. Silveira, F.G. Graeff, Differential expression of Fos protein in the rat brain induced by performance of avoidance or escape in the elevated T-maze Behav, Brain Res. 126 (2001) 13–21. [48] D.M.L. Vianna, M.L. Brandão, Anatomical connections of the periaqueductal gray: specific neural substrates for different kinds of fear, Braz. J. Med. Biol. Res. 36 (5) (2003) 557–566. [49] D.M.L. Vianna, F.G. Graeff, J. Landeira-Fernandez, M.L. Brandão, Lesion of the ventral periaqueductal gray reduces conditioned fear but does not change freezing induced by stimulation of the dorsal periaqueductal gray, Learn. Mem. 8 (2001) 164–169. [50] P. Walker, F. Carrive, Role of ventrolateral periaqueductal gray neurons in the behavioral and cardiovascular responses to contextual fear and poststress recovery, Neuroscience 116 (2003) 897–912.