Regulation of forebrain GABAergic stress circuits following lesion of the ventral subiculum

BR A I N R ES E A RC H 1 1 1 6 ( 2 00 6 ) 1 3 2 –14 2

a v a i l a b l e a t w w w. s c i e n c e d i r e c t . c o m

w w w. e l s e v i e r. c o m / l o c a t e / b r a i n r e s

Research Report

Regulation of forebrain GABAergic stress circuits following lesion of the ventral subiculum Nancy K. Mueller, C. Mark Dolgas, James P. Herman ⁎ Department of Psychiatry, University of Cincinnati, 2170 E. Galbraith Rd. Cincinnati, OH 45237-0506, USA

A R T I C LE I N FO

AB S T R A C T

Article history:

Ventral subiculum (vSUB) lesions enhance corticosterone responses to psychogenic

Accepted 14 July 2006

stressors via trans-synaptic influences on paraventricular nucleus (PVN) neurons.

Available online 18 September 2006

Synaptic relays likely occur in GABA-rich regions interconnecting the vSUB and PVN. The

Keywords:

and GABA biosynthetic capacity in putative limbic–hypothalamic stress relays. Male

Novelty

Sprague–Dawley rats received bilateral ibotenate or sham lesions of the vSUB. Animals

Hypothalamic–pituitary–adrenal axis

were divided into two groups, with one group receiving exposure to novelty stress and the

c-fos

other left unstressed. Exposure to novelty stress increased c-fos mRNA expression in the

Glutamic acid decarboxylase

PVN to a greater degree in vSUB lesion relative to shams, consistent with an inhibitory role

Paraventricular nucleus

for the vSUB in the HPA stress response. However, c-fos induction was not affected in other

Bed nucleus of the stria terminalis

forebrain GABAergic stress pathways, such as the lateral septum, medial preoptic area or

current study examines whether vSUB lesions compromise stress-induced c-fos induction

dorsomedial hypothalamus. vSUB lesions increased GAD65 or GAD67 mRNA levels in several efferent targets, including anterior and posterior subnuclei of the bed nucleus of the stria terminalis and lateral septum. Lesions did not effect stress-induced increases in GAD65 expression in principal output nuclei of the amygdala. The current data suggest that loss of vSUB innervations produces a compensatory increase in GAD expression in subcortical targets; however, this up-regulation is insufficient to block lesion-induced stress hyperresponsiveness, perhaps driven by amygdalar disinhibition of the PVN. © 2006 Elsevier B.V. All rights reserved.

⁎ Corresponding author. Fax: +1 513 558 9104. E-mail address: [email protected] (J.P. Herman). 0006-8993/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.brainres.2006.07.101

BR A I N R ES E A RC H 1 1 1 6 ( 2 00 6 ) 1 3 2 –1 42

133

Abbreviations: ACTH, adrenocorticotropic hormone ANOVA, analysis of variance BST, bed nucleus of the stria terminalis CeA, central amygdaloid nucleus CORT, corticosterone CRH, corticotropin releasing hormone DG, dentate gyrus GAD, glutamic acid decarboxylase HPA, hypothalamo–pituitary– adrenal KPBS, potassium phosphatebuffered saline MeA, medial amygdaloid nucleus PVN, paraventricular nucleus vSUB, ventral subiculum

1.

Introduction

The ventral subiculum (vSUB) plays a prominent role in neuronal regulation of the hypothalamo–pituitary–adrenocortical (HPA) axis. Lesions of the vSUB enhance glucocorticoid responses to restraint stress and novelty (Herman et al., 1995, 1998; Mueller et al., 2004) and enhance corticotropin releasing hormone (CRH) expression in paraventricular hypothalamic nucleus (PVN) neurons controlling HPA axis responses to stress (Herman et al., 1995, 1998). Anatomical studies indicate that the influence of the vSUB on the PVN is trans-synaptic, involving intervening neurons in the regions such as the bed nucleus of the stria terminalis (BST), medial preoptic area, dorsomedial hypothalamus, subparaventricular zone and the peri-PVN region (Cullinan et al., 1993). Combined anterograde– retrograde tract-tracing studies indicate that the vSUB likely relays with the PVN via synaptic contacts with GABAergic neurons in these regions (Cullinan et al., 1993). Previous studies indicate that GABA plays a prominent role in inhibition of the HPA axis. Electron microscopic studies suggest that up to 50% of synapses in the PVN are GABAergic (Decavel and Van Den Pol, 1990). In addition, PVN CRH neurons express GABA-A receptor subunits (Cullinan, 2000), suggesting direct actions of GABA on hypophysiotrophic neurons. Accordingly, corticosterone responses to restraint can be attenuated by injection of the GABA-A receptor agonist muscimol into the PVN (Cullinan, 1998), whereas PVN neurons can be directly activated by microinjections of the GABA-A receptor antagonist bicuculline (Cole and Sawchenko, 2002). These data verify that GABA, acting through GABA-A receptors, is prominently involved in inhibition of the PVN. Notably, GABA-rich PVN-projecting regions are activated by stressful stimuli, supporting a putative role in HPA integration. For example, the medial preoptic area, dorsomedial hypothalamus and peri-PVN region show marked c-fos induction in response to a variety of stressors (Cole and Sawchenko, 2002; Cullinan et al., 1995; Emmert and Herman, 1999; Martinez et al., 1998; Sawchenko et al., 2000). In the case of the dorsomedial hypothalamus and medial preoptic area, c-fos immunoreactivity is localized in neurons expressing the GABA synthesizing

enzyme glutamic acid decarboxylase (GAD) (Cullinan et al., 1996), suggesting that stress activates GABAergic neurons in these regions. In addition, both GAD65 and GAD67 isoforms show marked induction by acute or chronic stress. In the case of acute restraint, GAD67 is up-regulated in the dorsomedial nucleus, medial preoptic area and anterior subnuclei of BST; the same regions showed up-regulation of GAD65 following chronic intermittent stress exposure (Bowers et al., 1998). Upregulation of GAD isoforms suggests increased synthesis of GABA, perhaps as compensation for prior stress exposure. Thus, changes in both c-fos and GAD65/67 expression support the hypothesis that PVN circuits in regions targeted by the vSUB are engaged by stressful stimuli. Affective disease states are associated with hippocampal dysfunction and HPA axis abnormalities (Bremner et al., 1995; Carroll et al., 1976; Sheline et al., 1996; Yehuda et al., 1991). As the ventral subiculum–PVN circuitry is a primary inhibitory pathway for HPA responses to psychogenic stimuli (Herman and Cullinan, 1997; Herman et al., 2003), it is highly likely that glucocorticoid hypersecretion consequent to depression is associated with altered transmission along this disynaptic pathway. Therefore, understanding the mechanism whereby the hippocampus modulates HPA tone may be of substantial importance in guiding new therapeutic and ameliorative approaches. The data summarized above suggest that the disinhibitory effects of vSUB lesion on the HPA axis may be associated with altered activation of PVN-projecting GABAergic cell populations. To test this hypothesis, the current study assesses the effects of vSUB lesion on induction of c-fos mRNA and expression of GAD65/67 mRNA isoforms following acute novelty stress exposure. The data suggest that selective regulatory changes in key GABAergic relays accompany vSUB lesion in these rats and may be involved in mediating hippocampal influences on HPA axis function.

2.

Results

Histological assessment revealed that injections of ibotenate caused extensive damage to pyramidal cells of the vSUB,

134

BR A I N R ES E A RC H 1 1 1 6 ( 2 00 6 ) 1 3 2 –14 2

not result in marked cavitation in this region, consistent with the specific actions of the neurotoxin on neuronal cell bodies. No frank lymphocyte infiltration was observed. Ventral regions of CA1 and limited portions of dentate gyrus (DG) were also compromised by the injections (Fig. 1). There was no significant damage to amygdaloid nuclei in these animals. To determine the effect of vSUB lesion on activation of hypothalamic brain regions following stress, c-fos mRNA was examined using in situ hybridization following open field exposure. Fig. 2 illustrates c-fos expression in the PVN following novelty stress in saline and vSUB-lesion animals. The red and yellow coloration in the PVN of the vSUB lesion animal reflects higher gray level values in this region, consistent with elevated c-fos mRNA expression. As can be seen in Fig. 3, c-fos activation was observed in several limbic brain regions following novelty open field stress, including the lateral septum, medial preoptic area, PVN and dorsomedial hypothalamic nucleus. vSUB lesions increased c-fos mRNA in the hypothalamic PVN following open field exposure [F1,11 = 4.850; p = 0.05], consistent with enhanced activation of this cell population. However, vSUB lesions did not affect c-fos induction in the anterior cingulate cortex, lateral septum, medial preoptic area or dorsomedial hypothalamic nucleus (Fig. 3, Table 1). Note that no significant c-fos mRNA expression was observed in unstressed animals in any group. Stress and lesion effects on GAD65 and GAD67 mRNA expression were evaluated by three-way ANOVA with

Fig. 1 – Diagram and pictures of vSUB Lesions. (A) Diagrams of largest (light circle) and smallest (dark circles) bilateral ibotenate lesions of the ventral subiculum. In all cases, the lateral component of the ventral subiculum was compromised by ibotenate injection. (B) Representative Nissl-stained section of a ventral subiculum from a saline treated rat. (C) Representative Nissl-stained section of from the ventral subiculum region of an ibotenic acid-injected rat. Arrows delineate the area of the ibotenate lesion. Reprinted with permission from Endocrinology, Stressor-Selective Role of the Ventral Subiculum in Regulation of Neuroendocrine Stress Responses, 145 (8): 3763–3768, 2004. Copyright 2004, The Endocrine Society.

marked by the absence of Nissl-stained neurons and an increased number of glial cells. Lesions were centered at the middle of the rostrocaudal extent of the vSUB and typically destroyed the vast majority of the structure. However, in all cases, there was some degree of sparing of cells located at the rostral and/or caudal extremes of the subiculum. Lesions did

Fig. 2 – Representative images illustrating c-fos mRNA expression in PVN after saline (A, Sal) or ibotenate (B, Ibo) injections into the ventral subiculum. Pseudocolor image reveals higher levels of c-fos mRNA hybridization in the PVN of ibotenate-lesion animals (arrows). In contrast, c-fos hybridization did not differ in areas outside the PVN (arrowheads).

135

BR A I N R ES E A RC H 1 1 1 6 ( 2 00 6 ) 1 3 2 –1 42

medial BST, carried by increases in the unstressed group only (Fig. 6A). Images of the lateral septum, medial amygdaloid nucleus (MeA), central amygdaloid nucleus (CeA) and hypothalamic nuclei were carefully matched for neuroanatomical level and did not show substantial variance in captured area (see Fig. 4, Tables 3 and 5). As such, these regions were assessed using integrated gray level as the endpoint; this measure accounts for both gray level intensity and the overall extent of the labeled structure and thus provides a more accurate measure of overall signal than does intensity (corrected gray level) alone. There were significant effects of both stress (F1,269 = 11.468, p < 0.05) and region (F9,269 = 514.306, p < 0.05) on GAD65 mRNA expression by three-way repeated measures ANOVA (Fig. 5B), as well as a lesion by region interaction (F9,269 = 2.710, p < 0.05). Subsequent post hoc analysis indicated significant effects of stress on GAD65 mRNA expression in the CeA, MeA, arcuate nucleus and subparaventricular nucleus (Fig. 5B). Ibotenate injection increased GAD65 mRNA expression only in the lateral septum (Fig. 5B). Significant effects of lesion (F1,269 = 13.474, p < 0.05), stress (F1,269 = 14.522, p < 0.05) and region (F9,269 = 623.660, p < 0.05) were observed for GAD67 mRNA (Fig. 6B). Stress increased GAD67 mRNA expression in the subparaventricular zone and arcuate nucleus (Fig. 6B).

Fig. 3 – c-fos mRNA levels after novelty (open field) stress in limbic brain areas. vSUB lesions increased c-fos mRNA levels in the hypothalamic paraventricular nucleus (PVN) after novelty open field stress. mPOA: medial preoptic area; LS: lateral septum; DMH: dorsomedial hypothalamic area. † p < 0.05, vSUBX is significantly different than sham.

repeated measures across regions. This strategy was designed to account for the simultaneous measurement of multiple regions in the analytical scheme and limit the opportunity for type I error due to multiple ANOVAs. Examples of nuclear parcellations for the GAD analyses are illustrated in Fig. 4. Assessment of GAD65 and GAD67 mRNA in the BST, anterior cingulate cortex and hippocampus was performed using corrected gray level measures as the areas showed substantial changes in size across the sections employed for analysis (Tables 2 and 4). In the case of GAD65 mRNA, there was a significant effect of region (F7,215 = 452.276, p < 0.05), lesion by region interaction (F7,215 = 2.382, p < 0.05) and stress by region interaction (F7,215 = 2.024, p = 0.05). Follow-up post hoc analyses revealed no significant differences within region (Fig. 5A), indicating that GAD65 mRNA was not affected by stress or lesion in individual BST subnuclei. In contrast, assessment of GAD67 mRNA revealed significant effects of lesion (F1,215 = 32.307, p < 0.05), stress (F1,215 = 5.495, p < 0.05), region (F7,215 = 168.647, p < 0.05), lesion by region interaction (F7,215 = 2.734, p < 0.05), stress by region interaction (F7,21 = 3.880, p < 0.05) and a lesion by stress by region interaction (F7,215 = 2.514, p < 0.05) (Fig. 6A). Follow-up ANOVAs indicated significant lesion-induced increases in GAD67 mRNA in the anterodorsal and anteromedial BST in the stressed group only and significant increases in GAD67 mRNA expression in the posterior lateral and posterior

3.

Discussion

The current study demonstrates that lesions of the vSUB increase both GAD65 and GAD67 mRNA expression in several brain regions implicated in stress integration. Lesions of the vSUB produce basal GAD up-regulation in its subcortical targets, including the anterior BST and lateral septum. In contrast, lesions do not affect stress-induced GAD induction in the amygdala, suggesting that the capacity for disinhibitory effects of the amygdala on the HPA axis is intact (Herman et al., 2003). Lesion of the vSUB enhances PVN c-fos responses to novelty, commensurate with the hypothesized inhibitory influence of this region on the HPA axis (Herman et al., 1995, 1998). In the subparaventricular and arcuate nuclei, noveltyinduced up-regulation of both GAD65 and GAD67 mRNA was observed. The subparaventricular region provides input to the PVN and is implicated in stress inhibition of the HPA axis (Bowers et al., 1998; Cole and Sawchenko, 2002; Herman et al.,

Table 1 – c-fos mRNA expression in forebrain stress-regulatory regions a Region Cingulate Cortex Lateral Septum MPOA PVN DMH a

Sham stressed

vSUBX stressed

97,176 ± 30,742 31,229 ± 8776 11,961 ± 1653 8828 ± 2511 b 27,888 ± 2035

106,255 ± 22,685 34,954 ± 7265 16,643 ± 2805 17,706 ± 2188 30,279 ± 2151

Data are expressed as mean integrated gray level ± standard error of the mean. b sham stressed is significantly different than vSUBX stressed.

136

BR A I N R ES E A RC H 1 1 1 6 ( 2 00 6 ) 1 3 2 –14 2

Fig. 4 – Examples of nuclear parcellations for GAD65 and 67 mRNA analyses. Examples of GAD65 mRNA X-ray images. Outlined regions are shown to indicate the areas sampled for GAD65/67 mRNA analysis. Ctx: cortex; LS: lateral septum; BSTam: anteromedial bed nucleus of the stria terminalis; BSTad: anterodorsal bed nucleus of the stria terminalis; BSTvl: ventrolateral bed nucleus of the stria terminalis; mPOA: medial preoptic area; BSTpm: posterior medial bed nucleus of the stria terminalis; BSTpl: posterior lateral bed nucleus of the stria terminalis; SCN: suprachiasmatic nucleus; RT: reticular thalamus; peri-PVN: peri-paraventricular nucleus; subPVN: subparaventricular nucleus; CeA: central amygdaloid nucleus; MeA: medial amygdaloid nucleus; DMH: dorsomedial hypothalamus; Arc: arcuate nucleus. Approximate bregma location is noted for each image (Paxinos and Watson, 1986).

2002; Roland and Sawchenko, 1993; Watts and Swanson, 1987). The arcuate nucleus is also thought to be involved in stress inhibition (Magarinos et al., 1988); however, its role remains to be thoroughly investigated. Stress induction of hypothalamic GAD67 mRNA is consistent with a previous study from our group, demonstrating elevations in GAD67 in several brain regions 1 h after initiation of restraint. However, it should be

Table 2 – GAD65 mRNA expression in cortex, hippocampus and BST a Region Cingulate cortex BST ad BST am BST vl BST pl BST pm CA1 DG a

Sham unstressed

Sham stressed

vSUBX unstressed

vSUBX stressed

36.8 ± 4.76

36.8 ± 3.10

35.1 ± 4.039

29.4 ± 3.42

91.2 ± 6.71 75.8 ± 6.83 85.3 ± 7.17 94.0 ± 3.62 118.0 ± 4.28 39.6 ± 3.76 45.4 ± 3.93

92.0 ± 4.89 76.8 ± 5.38 90.4 ± 4.65 93.4 ± 5.11 109.0 ± 5.65 45.4 ± 1.60 52.8 ± 2.08

96.6 ± 2.94 80.3 ± 3.04 91.2 ± 3.14 92.0 ± 4.78 109 ± 4.63 34.4 ± 2.48 39.1 ± 2.46

88.5 ± 5.09 79.7 ± 4.21 85.7 ± 5.85 94.8 ± 2.80 107.2 ± 4.72 37.6 ± 2.61 43.1 ± 2.50

Data are expressed as mean corrected gray level ± standard error of the mean.

noted that the regional characteristics of stress induction differed somewhat between these studies (Bowers et al., 1998). In both studies, GAD67 mRNA expression increased in the arcuate nucleus. However, in our previous study, we did not observe elevated GAD67 mRNA in the subparaventricular zone, whereas significant increases were manifest in the dorsomedial nucleus and medial preoptic area. These differences may be related to stressor intensity or the different time-course of stress exposure in the two studies (5 min in the open field vs. 1 h of restraint). Stress-induced increases in GAD65 mRNA expression were observed in the central and medial amygdaloid nuclei. Lesions of the vSUB did not affect stress-induced increases in GAD65 in either region. The amygdaloid nuclei contain GABA projection neurons (Swanson and Petrovich, 1998) and are implicated in activation of HPA stress responses (Feldman et al., 1995; Herman and Cullinan, 1997; Van de Kar and Blair, 1999). The central amygdaloid nucleus stimulates HPA axis responses to some stressors, but not others (Feldman et al., 1995; Herman and Cullinan, 1997; Van de Kar and Blair, 1999), and likely relays with the PVN through PVN-projecting neurons in the BST and possibly brainstem (Herman et al., 2003). The medial amygdala stimulates the HPA axis in response to emotional stressors (Chen and Herbert, 1995; Dayas et al., 1999) and sends a small number of projections

BR A I N R ES E A RC H 1 1 1 6 ( 2 00 6 ) 1 3 2 –1 42

137

Fig. 5 – GAD65 mRNA levels after novelty open field stress in limbic brain areas. (A) Corrected gray levels for GAD65 mRNA. (B) Integrated gray levels for GAD65 mRNA. vSUB lesion increased GAD65 mRNA levels in the lateral septum (#p < 0.05). Stress significantly increased GAD65 mRNA expression in the subparaventricular nucleus, central amygdaloid nucleus, medial amygdaloid nucleus and the arcuate nucleus (*p < 0.05).

directly to the PVN (Dayas et al., 1999). Amygdaloid neurons are thought to disinhibit the HPA axis through relays with GABA neurons in the BST and hypothalamus (Herman et al., 2003). Given that stress induction of medial and central amygdaloid GAD65 expression was not affected by vSUB lesion, it is possible that maintenance of amygdalar stress responsivity in the face of loss of hippocampal inhibition may contribute to stress hyperresponsiveness. Lesions of the vSUB increased GAD67 mRNA expression in anterior and posterior BST subdivisions and elevated GAD65 mRNA in the lateral septum, suggesting a general effect of denervation on GAD expression in these regions. The stress by lesion interaction effects seen in the anterior BST are consistent with the hypothesis that removal of the ventral subiculum primes these areas to respond to stress. Importantly, lateral septum and anterior and posterior subnuclei of the BST are prominent subcortical targets of the vSUB (Canteras and Swanson, 1992) and project either directly to

the PVN (in the case of the posterior and, to a lesser extent, anterior BST) or to other GABAergic PVN relays (lateral septum, anterior BST) (Bowers et al., 1998; Okamura et al., 1990; Risold and Swanson, 1997). Thus, denervation of subcortical targets of the ventral subiculum causes a compensatory elevation in GAD67 mRNA and/or GAD65 mRNA expression that may be of relevance to stress integration. Differential regulation of GAD isoforms has functional implications. GAD65 represents the ‘stored’ isoform of GAD, which is present in nerve terminals largely as an inactive apoenzyme (Esclapez et al., 1994; Kaufman et al., 1991). Apoenzyme is converted into holoenzyme upon phosphorylation, and as such GAD65 is thought to represent a ‘store’ of enzyme capable of being recruited upon stimulation (Esclapez et al., 1994; Kaufman et al., 1991). The GAD67 isoform is rapidly induced by stimulation of GABA neurons and is present largely as active holoenzyme (Esclapez et al., 1994; Kaufman et al., 1991). These data thus suggest that removal of vSUB input may

138

BR A I N R ES E A RC H 1 1 1 6 ( 2 00 6 ) 1 3 2 –14 2

Fig. 6 – GAD67 mRNA levels after novelty open field stress in limbic brain areas. (A) Corrected gray levels for GAD67 mRNA. GAD67 mRNA levels were increased by vSUB lesions in the anterior BST of stressed animals only, and in the posterior BST of unstressed animals only ( &p < 0.05). (B) Integrated gray levels for GAD67 mRNA. Stress increased GAD67 mRNA levels in the subPVN and arcuate nucleus (*p < 0.05).

modulate signals that promote both rapid and sustained GABA synthesis. Lesions of the vSUB enhance open field c-fos mRNA expression in the hypothalamic PVN, but not in other vSUBreceptive PVN-projecting regions (e.g., dorsomedial hypothalamus, medial preoptic area) or putative stress-regulatory regions (e.g., lateral septum, anterior cingulate cortex). The results are consistent with the observed prolongation of the stress response seen following vSUB lesion (Mueller et al., 2004). The lack of lesion-induced changes seen in extra-PVN regions suggests that lesions do not markedly affect excitability of these neural systems. However, only one time point was sampled in this study, and thus differential activation of

vSUB target regions occurring earlier or later in time may be missed. In addition, c-fos induction does not necessarily predict the net impact of an activated cell on downstream processes; this will be determined by inhibitory or excitatory neurotransmission at the level of the target (in this case, the PVN). The current study was designed to provide a broad scan of GAD regulation throughout forebrain stress circuits. There are several caveats associated with this approach that bear comment. First, the data derived from these studies are correlative in nature, designed to provide a first-level analysis of gene expression in GABA neurons consequent to removal of vSUB input. Whereas changes in GAD predict altered GABA

BR A I N R ES E A RC H 1 1 1 6 ( 2 00 6 ) 1 3 2 –1 42

139

Table 3 – GAD65 mRNA expression in diencephalon, lateral septum and amygdala a Region

Sham unstressed

Sham stressed

vSUBX unstressed

vSUBX stressed

RT SCN LS mPOA pPVN Sub PVN CeA MeA DMH Arc

27,6115 ± 23,897 54,224 ± 4751 278,930 ± 24,838 c 98,344 ± 5653 35,623 ± 4144 69,142 ± 6415 c 123,800 ± 15,421 106,057 ± 12,271 99,658 ± 8901 19,438 ± 3240

312,766 ± 26,205 53,581 ± 6054 294,315 ± 21,282 c 113,695 ± 9098 39,908 ± 2951 98,401 ± 7711 145,882 ± 9069 b 122,031 ± 3451 b 110,646 ± 17,299 28,169 ± 2332 b

254,468 ± 11,573 47,272 ± 4776 360,704 ± 32,532 95,299 ± 5878 34,285 ± 2797 82,920 ± 9651 117,306 ± 13,818 97,967 ± 14,823 106,611 ± 10,830 22,369 ± 3816

261,217 ± 16,065 52,612 ± 5686 404,649 ± 17,595 117,383 ± 6542 37,235 ± 4838 82,377 ± 6302 164,999 ± 8679 b 113,192 ± 8112 b 116,063 ± 4497 23,071 ± 772 b

a b c

Data are expressed as mean integrated gray level ± standard error of the mean. significant main effect of stress. significant main effect of lesion.

availability, these data cannot be taken to imply true changes in function of GABAergic neurons per se as other processes may compensate for alterations in message levels (e.g., protein translation, enzyme activity, packaging, etc.). Second, these animals were used for prior functional studies of HPA axis activity. One week intervened between the last stress test and the current test; in our experience, this interval is sufficient to restore normal basal HPA function (unpublished observations). In addition, all animals received the same stress exposure regimen, thus ensuring that all animals received the same treatment regimen. However, we cannot completely rule out the possibility that prior exposure to stress may differentially alter GAD regulatory responses in vSUB versus sham-lesion groups. Finally, changes in GAD mRNA expression in large regions of brain may not reflect changes in small numbers of neurons associated with discrete functions such as stress regulation. Thus, specific projections to the PVN may behave differently than large regions of relevant nuclei. In addition, changes in GAD mRNA may reflect interneurons as well as projection neurons in amygdala and hypothalamus. Nonetheless, we feel that the current dataset provides valuable insight into functional pathways innervated by the vSUB through identification of regions that do, and indeed, do not show altered capacity for GABA synthesis following lesions. The vSUB-lesion animals used in this study displayed alterations in HPA axis responses tested earlier in time (Mueller et al., 2004). In particular, corticosterone responses to elevated plus maze exposure were prolonged in the vSUBlesion rats, consistent with enhanced HPA activation and in agreement with our observed elevations in PVN c-fos mRNA expression following novelty. To summarize, the current study presents several findings of relevance to the excitatory effect of vSUB lesions on neural responses to stress. First, the ability of stress to increase GAD expression in the amygdala is not attenuated by lesions of the vSUB. This finding suggests that the downstream actions of amygdalar GABA neurons, including excitation of the HPA axis (see Herman et al., 2003), are unchanged by the hippocampal lesion. Second, vSUB lesions increase GAD expression in regions such as the lateral septum and BST, presumably as compensation for loss of input. Despite these increases, the ventral subiculum can no longer participate in driving these

neurons after stress. Finally, changes in PVN c-fos induction are not accompanied by changes in c-fos expression in other stress-regulatory regions. Taken together, these data lead us to hypothesize that, despite the denervation-induced increases in GAD levels in PVN regulatory regions, the absence of hippocampal input is insufficient to balance out an unopposed drive of the HPA axis by amygdalar circuitry. As the principal output neurons of both the central and medial amygdala are GABAergic (see Herman et al., 2003), drive is likely caused by inhibition of GABAergic projections to the PVN (from regions such as the BST), which would not be reflected in an altered c-fos signal. Overall, the data suggest that removal of the ventral subiculum causes a reorganization of GABAergic activation patterns in brain, both under resting conditions and following stress. Notably, a number of regions showing alterations in GAD expression have been associated with regulation of stress responses. Thus, the complex role of the vSUB in neuroendocrine and behavioral responsiveness to stress may be mediated at least in part by subcortical GABAergic neurons. The data are consistent with a role for these GABA circuits in mediating behavioral/endocrine dysfunction seen in affective diseases associated with hippocampal disturbances, including depression (Sheline et al., 1996) and post-traumatic stress disorder (Bremner et al., 1995).

4.

Experimental procedures

Male Sprague–Dawley rats (275–350 g) (Harlan, Indianapolis, IN) were individually housed in a constant temperature: humidity vivarium quarters on a 12:12-hour light:dark cycle with lights coming on at 6 a.m., with food and water ad libitum. One week after arrival, rats were anesthetized with ketamine (87 mg/kg)/xylazine (13 mg/kg) i.p. and received either bilateral stereotaxic injections of ibotenate (3.5 μg of a 5 mg/ml solution) or saline into the vSUB (−6.2 mm posterior to bregma, ±5.1 mm lateral and −8.2 mm ventral to the surface of the brain) over a 30-minute period from a Hamilton syringe, using the coordinate system of Paxinos and Watson (1986). It should be noted that this cohort of vSUB-lesion animals was used for acute stress studies prior to this experiment (Mueller et al., 2004). Exposure to open field testing (or control

140

BR A I N R ES E A RC H 1 1 1 6 ( 2 00 6 ) 1 3 2 –14 2

Table 4 – GAD67 mRNA expression in cortex, hippocampus and BST a Region Cingulate cortex BST ad BST am BST vl BST pl BST pm CA1 DG

Sham unstressed

Sham stressed

vSUBX unstressed

vSUBX stressed

45.0 ± 4.49

40.5 ± 4.65

42.0 ± 2.51

45.9 ± 2.87

62.6 ± 5.08 56.6 ± 5.16 63.4 ± 4.43 58.6 ± 5.17 77.4 ± 5.65 31.6 ± 5.04 36.2 ± 4.80

60.0 ± 4.74 50.2 ± 4.40 59.8 ± 5.49 70.4 ± 5.38 96.4 ± 6.97 42.0 ± 2.13 46.5 ± 1.81

b

75.3 ± 2.32 62.3 ± 2.39 70.5 ± 3.00 70.6 ± 2.65 d 95.0 ± 3.04 d 37.1 ± 2.65 40.4 ± 2.28

b

73.2 ± 2.32 61.2 ± 2.57 c 67.7 ± 2.79 70.5 ± 2.52 97.3 ± 3.10 39.6 ± 1.49 42.0 ± 2.50

a

Data are expressed as mean corrected gray level ± standard error of the mean. b vSUBX significantly different than sham. c vSUBX stressed significantly different than sham stressed. d vSUBX unstressed significantly different than sham unstressed.

conditions) in the present study occurred 7 days after the last experimental manipulation, at which point basal AM and PM corticosterone levels are within normal range (unpublished observations). All procedures were approved by the University of Cincinnati Institutional Animal Care and Use Committee.

4.1.

Stress protocol

vSUB-lesion (n = 17) and sham-lesion (n = 10) rats were divided into two groups. One group was individually placed in a novel environment for 5 min (n = 9 lesion, n = 5 sham). The novel environment was a 36 × 36 in. white Plexiglas enclosure, illuminated by standard fluorescent lighting. At the initiation of each trial, the subject was placed in the center of the field and remained in the novel environment for 5 min. Rats in the second group were rapidly decapitated without undergoing novelty exposure (n = 8 lesion, n = 5 control). Novelty-exposed rats were returned to their home cages and killed 60 min after initiation of novelty; control rats were killed unstressed. Brains were removed rapidly and frozen in isopentane at − 50 °C on dry ice. All brains were stored at − 80 °C until processing. Fourteen-micrometer-thick coronal sections were cut on a Microm cryostat. Sections were collected from the lateral septum through the ventral subiculum, thaw

mounted onto Ultrafrost slides and stored at −20 °C until processing.

4.2.

cRNA probes

Plasmids containing the cDNAs complementary to rat c-fos (subcloned in pGem4Z), GAD65 (subcloned in bluescript SK) or GAD67 (subcloned in bluescript SK) were linearized with StuI or HincII and transcribed in vitro in the presence of 35S-UTP with SP6 (c-fos) or T3 (GAD65 and GAD67) RNA polymerase. Standard labeling reactions included 1 μg linearized plasmid DNA, 1× transcription buffer (Promega (Madison, WI) or Gibco (Carlsbad, CA)), 150 μCi α-35S-UTP (Amersham (Buckinghamshire, England)), 2 mM ATP, CTP and GTP (Boehringer Mannheim (Indianapolis, IN)), 100 mM dithiothreitol, 2.5 units RNAseout (Promega) and 2 units of SP6 (Gibco) (c-fos) or T3 (Promega) (GAD65 and GAD67) RNA polymerase. The reaction was incubated for 90 min at 37 °C. Subsequently, 10 U of RNAse-free DNAse I (Promega) was added to digest the DNA template, and after 5 min at 37 °C, the reaction mix was diluted to 100 μl with diethylpyrocarbonate-treated water and ethanol precipitated with 5 M ammonium acetate.

4.3.

In situ hybridization

Prior to hybridization, sections were equilibrated to room temperature and fixed for 10 min in 4% paraformaldehyde in phosphate-buffered saline. Sections were then rinsed twice for 5 min each in 5 mM potassium phosphate-buffered saline (KPBS), pH 7.4, rinsed twice for 5 min each in KPBS containing 0.2% glycine, rinsed twice again for 5 min each in KPBS. Then, slides were incubated for 10 min in 0.1 M triethanolamine pH 8.0 with 0.25% acetic anhydride. Slides were rinsed in 2× standard saline citrate (SSC buffer; 0.25 M sodium chloride, 0.015 M sodium citrate, pH 7.2) and dehydrated in graded ethanol concentrations followed by air drying. 35S-labeled probes were diluted in hybridization buffer (50% formamide, 10% dextran sulfate, 20 mM dithiothreitol, 335 mM sodium chloride, 20 mM Tris–HCl pH 7.5, 1 mM EDTA, 1× Denhardt’s, 0.15 mg/ml yeast tRNA and 0.3 mg/ml Herring sperm DNA) to yield 1.0 × 106 cpm 35S-labeled probe/50 μl/slide. Sections were hybridized overnight at 55 °C in humidified chambers (moistened with 50% formamide). On the following day, slides

Table 5 – GAD67 mRNA expression in diencephalon, lateral septum and amygdala a Region

Sham unstressed

Sham stressed

vSUBX unstressed

RT SCN LS mPOA pPVN Sub PVN CeA MeA DMH Arc

211,075 ± 24,793 26,218 ± 2057 274,836 ± 22,532 48,538 ± 3704 21,942 ± 743 32,130 ± 1718 106,506 ± 6060 93,791 ± 4071 71,897 ± 7832 13,996 ± 2236

247,834 ± 19,678 28,663 ± 1902 261,270 ± 18,010 58,541 ± 8743 23,311 ± 1591 47,756 ± 4768 112,342 ± 2666 107,048 ± 7737 80,979 ± 5527 18,262 ± 2252

245,395 ± 10,227 28,713 ± 2399 279,081 ± 15,281 62,780 ± 5222 19,587 ± 1569 40,085 ± 2588 b 121,273 ± 6351 103,045 ± 4501 82,943 ± 5190 17,112 ± 1795 b

a b

Data are expressed as mean integrated gray level ± standard error of the mean. Significant main effect of stress.

vSUBX stressed 283,553 ± 12,838 26,777 ± 2078 339,650 ± 10,631 75,263 ± 13,174 22,101 ± 1668 46,082 ± 1281 b 126,343 ± 8001 114,698 ± 6088 98,182 ± 4563 21,544 ± 1499 b

BR A I N R ES E A RC H 1 1 1 6 ( 2 00 6 ) 1 3 2 –1 42

were rinsed twice in 2× SSC and subsequently treated with RNAse A (25 μg/ml) for 30 min at 37 °C. Sections were then washed in 2× SSC and washed three time in 0.2× SSC followed by incubation in 0.2× SSC at 65 °C for 1 h. Following that slides were briefly rinsed in 0.2× SSC at room temperature and dehydrated through graded ethanols and air-dried. Slides were then exposed to Kodak Biomax MR X-ray film for 8 days (GAD65 and GAD67) or 28 days (c-fos).

4.4.

Image analysis

Semi-quantitative analyses of in situ hybridization autoradiographs were conducted utilizing ScionImage v. 1.62 (NIH). Images were captured using TV Zoom Lens 18–180 mm lens. Sections were matched for rostrocaudal level and regions of interest delineated in accordance with the atlas of Paxinos and Watson (1986). The BST was delineated using the designations of Moga et al. (1989). Gray level measures were corrected for background by subtracting gray level values of an adjacent, hybridization-negative region and expressed as either integrated gray level (total gray level units/sampled region) or corrected gray level units/mm2. Integrated gray measures account for changes in the size of the area containing signal and are generally a more sensitive measure of total change in signal intensity but can only be used when anatomical levels and boundaries of anatomical structures can be closely matched between animals. This measure was employed under circumstances where precise and confident definition of anatomical regions could be specified (e.g., paraventricular nucleus, central amygdaloid nucleus (CeA), medial amygdaloid nucleus (MeA)). In cases where regions varied substantially in size across the 140 um interval between sections (e.g., hippocampus, BST), the more conservative corrected gray level measure was used. Mean values for all animals were determined from three to five sections (6 to 10 measures) and used in the subsequent analysis of group effects.

4.5.

Statistical analysis

The c-fos data were natural log transformed to normalize the distribution of the data across brain areas. Since multiple regions were assessed in the same in situ hybridization run, statistical analysis of c-fos data was performed by two-way repeated measures ANOVA, using lesion condition and brain region (repeated measure) as variables. Only stressed c-fos data were analyzed due to the fact that unstressed c-fos data typically did not exceed background. Analysis of GAD65 and GAD67 mRNA expression was performed by three-way repeated measures ANOVA, using lesion condition, stress condition and brain region (repeated) as independent variables. Separate three-way ANOVAs were performed on regions analyzed by integrated and corrected gray levels. GAD65 and GAD67 integrated gray levels data were natural log transformed prior to statistical analysis to normalize the distribution of the data across brain regions. The distribution of the GAD65 and GAD67 corrected gray levels passed the Kolmogorov–Smirnov test for homogeneity of variance and were therefore not transformed. Note that, for comparison, both untransformed integrated gray level and corrected gray

141

level determinations are supplied in Tables 2–5. Fisher’s PLSD was used for post hoc analysis. All analyses were accomplished using GB-STAT software. Significance was set at p < 0.05.

4.6.

Lesion analysis

Detection and analysis of the extent of the ibotenate lesions and saline injection sites were performed by histological analysis of adjacent brain sections stained for Nissl substance with cresyl violet. The absence of pyramidal neurons in the vSUB-lesion area was the main criterion used to consider a lesion complete (Figs. 1B and C).

Acknowledgments The authors would like to thank Megan E. Paskitti and the late Brian L. Bodie for their technical assistance on this project. This work was funded by NIMH grants MH049698 (JPH) and MH065770 (NKM).

REFERENCES

Bowers, G., Cullinan, W.E., Herman, J.P., 1998. Region-specific regulation of glutamic acid decarboxylase (GAD) mRNA expression in central stress circuits. J. Neurosci. 18, 5938–5947. Bremner, J.D., Randall, P., Scott, T.M., Bronen, R.A., Seibyl, J.P., Southwick, S.M., Delaney, R.C., McCarthy, G., Charney, D.S., Innis, R.B., 1995. MRI-based measurement of hippocampal volume in patients with combat-related posttraumatic stress disorder. Am. J. Psychiatry 152, 973–981. Canteras, N.S., Swanson, L.W., 1992. Projections of the ventral subiculum to the amygdala, septum, and hypothalamus: a PHAL anterograde tract-tracing study in the rat. J. Comp. Neurol. 324, 180–194. Carroll, B.J., Curtis, G.C., Mendels, J., 1976. Neuroendocrine regulation in depression, I: limbic system–adrenocortical dysfunction. Arch. Gen. Psychiatry 33, 1039–1044. Chen, X., Herbert, J., 1995. Regional changes in c-fos expression in the basal forebrain and brainstem during adaptation to repeated stress: correlations with cardiovascular, hypothermic and endocrine responses. Neuroscience 64, 675–685. Cole, R.L., Sawchenko, P.E., 2002. Neurotransmitter regulation of cellular activation and neuropeptide gene expression in the paraventricular nucleus of the hypothalamus. J. Neurosci. 22, 959–969. Cullinan, W.E., 1998. Evidence for a PVN site of action of GABA in the regulatory control of the rat stress axis. The Physiologist 41, 353. Cullinan, W.E., 2000. GABA(A) receptor subunit expression within hypophysiotropic CRH neurons: a dual hybridization histochemical study. J. Comp. Neurol. 419, 344–351. Cullinan, W.E., Herman, J.P., Watson, S.J., 1993. Ventral subicular interaction with the hypothalamic paraventricular nucleus: evidence for a relay in the bed nucleus of the stria terminalis. J. Comp. Neurol. 332, 1–20. Cullinan, W.E., Herman, J.P., Battaglia, D.F., Akil, H., Watson, S.J., 1995. Pattern and time course of immediate early gene expression in rat brain following acute stress. Neuroscience 64, 477–505. Cullinan, W.E., Helmreich, D.L., Watson, S.J., 1996. Fos expression in forebrain afferents to the hypothalamic

142

BR A I N R ES E A RC H 1 1 1 6 ( 2 00 6 ) 1 3 2 –14 2

paraventricular nucleus following swim stress. J. Comp. Neurol. 368, 88–99. Dayas, C.V., Buller, K.M., Day, T.A., 1999. Neuroendocrine responses to an emotional stressor: evidence for involvement of the medial but not the central amygdala. Eur. J. Neurosci. 11, 2312–2322. Decavel, C., Van Den Pol, A.N., 1990. GABA: a dominant neurotransmitter in the hypothalamus. J. Comp. Neurol. 302, 1019–1037. Emmert, M.H., Herman, J.P., 1999. Differential forebrain c-fos mRNA induction by ether inhalation and novelty: evidence for distinctive stress pathways. Brain Res. 845, 60–67. Esclapez, M., Tillakaratne, N.J., Kaufman, D.L., Tobin, A.J., Houser, C.R., 1994. Comparative localization of two forms of glutamic acid decarboxylase and their mRNAs in rat brain supports the concept of functional differences between the forms. J. Neurosci. 14, 1834–1855. Feldman, S., Conforti, N., Weidenfeld, J., 1995. Limbic pathways and hypothalamic neurotransmitters mediating adrenocortical responses to neural stimuli. Neurosci. Biobehav. Rev. 19, 235–240. Herman, J.P., Cullinan, W.E., 1997. Neurocircuitry of stress: central control of the hypothalamo–pituitary–adrenocortical axis. TINS 20, 78–83. Herman, J.P., Cullinan, W.E., Morano, M.I., Akil, H., Watson, S.J., 1995. Contribution of the ventral subiculum to inhibitory regulation of the hypothalamo–pituitary–adrenocortical axis. J. Neuroendocrinol. 7, 475–482. Herman, J.P., Dolgas, C.M., Carlson, S.C., 1998. Ventral subiculum co-ordinates situation-specific neuroendocrine and behavioral stress responses. Neuroscience 86, 449–459. Herman, J.P., Tasker, J.G., Ziegler, D.R., Cullinan, W.E., 2002. Local circuit regulation of paraventricular nucleus stress integration. Glutamate–GABA connections. Pharmacol. Biochem. Behav. 71, 457–468. Herman, J.P., Figueiredo, H., Mueller, N.K., Ulrich-Lai, Y., Ostrander, M.M., Choi, D.C., Cullinan, W.E., 2003. Central mechanisms of stress integration: hierarchical circuitry controlling hypothalamo–pituitary–adrenocortical responsiveness. Front. Neuroendocrinol. 24, 151–180. Kaufman, D.L., Houser, C.R., Tobin, A.J., 1991. Two forms of the gamma-aminobutyric acid synthetic enzyme glutamate decarboxylase have distinct intraneuronal distributions and cofactor interactions. J. Neurochem. 56, 720–723. Magarinos, A.M., Estivariz, F., Morado, M.I., De Nicola, A.F., 1988. Regulation of the central nervous system—pituitary–adrenal

axis in rats after neonatal treatment with monosodium glutamate. Neuroendocrinology 48, 105–111. Martinez, M., Phillips, P.J., Herbert, J., 1998. Adaptation in patterns of c-fos expression in the brain associated with exposure to either single or repeated social stress in male rats. Eur. J. Neurosci. 10, 20–33. Moga, M.M., Saper, C.B., Gray, T.S., 1989. Bed nucleus of the stria terminalis: cytoarchitecture, immunohistochemistry, and projection to the parabrachial nucleus in the rat. J. Comp. Neurol. 283, 315–332. Mueller, N.K., Dolgas, C.M., Herman, J.P., 2004. Stressor-selective role of the ventral subiculum in regulation of neuroendocrine stress responses. Endocrinology 145, 3763–3768. Okamura, H., Abitbol, M., Julien, J.F., Dumas, S., Berod, A., Geffard, M., Kitahama, K., Bobillier, P., Mallet, J., Wiklund, L., 1990. Neurons containing messenger RNA encoding glutamate decarboxylase in rat hypothalamus demonstrated by in situ hybridization, with special emphasis on cell groups in medial preoptic area, anterior hypothalamic area and dorsomedial hypothalamic nucleus. Neuroscience 39, 675–699. Paxinos, G., Watson, C., 1986. The Rat Brain in Stereotaxic Coordinates, 2nd edn. Academic Press, New York. Risold, P.Y., Swanson, L.W., 1997. Connections of the rat lateral septal complex. Brain Res. Brain Res. Rev. 24, 115–195. Roland, B.L., Sawchenko, P.E., 1993. Local origins of some GABAergic projections to the paraventricular and supraoptic nuclei of the hypothalamus of the rat. J. Comp. Neurol. 332, 123–143. Sawchenko, P.E., Li, H.Y., Ericsson, A., 2000. Circuits and mechanisms governing hypothalamic responses to stress: a tale of two paradigms. Prog. Brain Res. 122, 61–78. Sheline, Y.I., Wang, P.W., Gado, M.H., Csernansky, J.G., Vannier, M.W., 1996. Hippocampal atrophy in recurrent major depression. Proc. Natl. Acad. Sci. U. S. A. 93, 3908–3913. Swanson, L.W., Petrovich, G.D., 1998. What is the amygdala? Trends Neurosci. 21, 323–331. Van de Kar, L.D., Blair, M.L., 1999. Forebrain pathways mediating stress-induced hormone secretion. Front. Neuroendocrinol. 20, 1–48. Watts, A.G., Swanson, L.W., 1987. Efferent projections of the suprachiasmatic nucleus: II. Studies using retrograde transport of fluorescent dyes and simultaneous peptide immunohistochemistry in the rat. J. Comp. Neurol. 258, 230–252. Yehuda, R., Giller, E.L., Southwick, S.M., Lowy, M.T., Mason, J.W., 1991. Hypothalamic–pituitary–adrenal dysfunction in post-traumatic stress disorder. Biol. Psychiatry 30, 1031–1048.