Effects of social context on endocrine function and Zif268 expression in response to an acute stressor in adolescent and adult rats

Page 1

Ei

Ta

b

a

ARRA

KSSACTZ

1

ot(aBmihtr

oSf

h0

Int. J. Devl Neuroscience 35 (2014) 25–34

Contents lists available at ScienceDirect

International Journal of Developmental Neuroscience

j ourna l ho me page: www.elsev ier .com/ locate / i jdevneu

ffects of social context on endocrine function and Zif268 expressionn response to an acute stressor in adolescent and adult rats

ravis E. Hodgesa, Matthew R. Greena, Jonathan J. Simoneb, Cheryl M. McCormicka,b,∗

Department of Psychology, Brock University, CanadaDepartment of Biological Sciences, Brock University, Canada

r t i c l e i n f o

rticle history:eceived 19 December 2013eceived in revised form 4 February 2014ccepted 1 March 2014

eywords:ocial stressocial bufferingdolescenceorticosteroneestosteroneif268

a b s t r a c t

There is a paucity of studies comparing social buffering in adolescents and adults, despite their markeddifferences in social behavior. We investigated whether greater effects of social buffering on plasmacorticosterone concentrations and expression of Zif268 in neural regions after an acute stressor would befound in adolescent than adult rats. Samples were obtained before and after 1 h of isolation stress and aftereither 1 or 3 h of recovery back in the colony with either a familiar or unfamiliar cage partner. Adolescentand adult rats did not differ in plasma concentrations of corticosterone at any time point. Corticosteroneconcentrations were higher after 1 h isolation than at baseline (p < 0.001), and rats with a familiar partnerduring the recovery phase had lower corticosterone concentrations than did rats with an unfamiliarpartner (p = 0.02). Zif268 immunoreactive cell counts were higher in the arcuate nucleus in both agegroups after isolation (p = 0.007) and in the paraventricular nucleus of adolescents than adults during therecovery phase irrespective of partner familiarity. There was a significant decrease in immunoreactivecell counts after 1 h isolation compared to baseline in the basolateral amygdala, central nucleus of the

amygdala, and in the pyramidal layer of the hippocampus (all p < 0.05). An effect of partner familiarityon Zif268 immunoreactive cell counts was found in the granule layer of the dentate gyrus irrespective ofage (higher in those with a familiar partner, p = 0.03) and in the medial prefrontal cortex in adolescents(higher with an unfamiliar partner, p = 0.02). Overall, the acute stress and partner familiarity produced asimilar pattern of results in adolescents and adults, with both age groups sensitive to the social context.

© 2014 ISDN. Published by Elsevier Ltd. All rights reserved.

. Introduction

In social species, social bonds are influential in their moderationf survival, of learning, and of responses to environmental disrup-ions. Social bonds can facilitate good health and reduce mortalitye.g., Holt-Lunstad et al., 2010; Silk et al., 2010; Yee et al., 2008)nd enhance recovery from aversive experiences (e.g., Gilbert andaker, 2011; Hennessy et al., 2009; Kikusui et al., 2006). One of theeans by which recovery from aversive or stressful experiences

s gauged is by monitoring hypothalamic–pituitary–adrenal (HPA)

ormone release. When a stressor is perceived, signals are relayedo the paraventricular nucleus of the hypothalamus initiating theelease of secretagogues that act in the pituitary to release ACTH to

∗ Corresponding author at: Canada Research Chair in Neuroscience, Departmentf Psychology and Centre for Neuroscience, Brock University, 500 Glenridge Avenue,t. Catharines, Ontario, Canada L2S 3A1. Tel.: +1 905 688 5550x3700;ax: +1 905 688 6922.

E-mail address: [email protected] (C.M. McCormick).

ttp://dx.doi.org/10.1016/j.ijdevneu.2014.03.001736-5748/© 2014 ISDN. Published by Elsevier Ltd. All rights reserved.

increase the release of glucocorticoids (primarily corticosterone inrodents) from the adrenal cortex (reviewed in McEwen et al., 2012).Glucocorticoids in turn initiate various physiological mechanismssuch as redirecting energy and modulating immune, cardiovascu-lar and memory function, allowing the organism to cope with thestress exposure and encode the experience, possibly altering futurebehavior (reviewed in McEwen et al., 2012; Roozendaal et al., 2009).Efficient negative feedback mechanisms at all levels of the HPA axisand at extra-hypothalamic neural sites, notably the hippocampus,are crucial to limit over-exposure to deleterious effects of gluco-corticoids that can occur under conditions of chronic or repeatedstress (Herbert et al., 2006).

HPA responses to stressors are moderated differentially by thequality of social interactions. Negative social interactions, such associal evaluative threat in humans (e.g., Dickerson et al., 2008) orsocial defeat in rodents (Buwalda et al., 2005), are potent acti-

vators of the HPA axis. In contrast, supportive social interactionscan dampen the release of glucocorticoids (Gunnar and Donzella,2002). For example, the deleterious effects of repeated social defeat(placing a rodent into the cage of a larger aggressive male) are

Page 2

2 vl Neu

o(rgfaw

bfeb1itactTtlcsfSgtilpaM

odappa((cacetqtt

rrt(tbtuaawrRirr

6 T.E. Hodges et al. / Int. J. De

bserved in singly housed rats and are attenuated by social housinge.g., de Jong et al., 2005). The evidence of a social buffering of socialelationships in adult rodents, however, is mixed, with evidence ofreater corticosterone release in response to a stressor when with aamiliar conspecific than an unfamiliar one (Armario et al., 1983a,b)nd evidence of less corticosterone release in response to a stressorhen with a conspecific than when alone (File and Peet, 1980).

The buffering effects of social interactions on the HPA axis mighte greater in adolescence than in adulthood considering the dif-erences in social behavior at the two stages of development. Forxample, adolescents engage in more social play and affiliativeehaviors than do adults (Pellis et al., 1997; Vanderschuren et al.,997). Further, adolescent males that have been deprived of social

nteractions for a length of time will engage in more social interac-ions when a peer is next encountered than will a similarly depriveddult male (e.g., Varlinskaya and Spear, 2008). Much evidence indi-ates that social interactions are more rewarding for adolescent ratshan for adult rats (reviewed in Doremus-Fitzwater et al., 2010).he heightened sensitivity of adolescents to social input reflectshe importance of this period of development as a time of socialearning (Hall, 1998). Brain regions that are sensitive to social pro-esses, such as the medial prefrontal cortex, and that regulateocial behaviors, such as the medial amygdala, undergo change andunctional development throughout this period (Blakemore, 2012;turman and Moghaddam, 2011). Thus, adolescents may receivereater stress-reducing benefits of social interactions comparedo adults. Such an advantage in adolescence may be particularlymportant given the ongoing maturation of the HPA axis in ado-escence. For example, pre-pubertal adolescent rats have a morerolonged release of glucocorticoids to acute stressors than dodult rats (reviewed in Eiland and Romeo, 2013; McCormick andathews, 2010).Social relationships do moderate adolescent responses to stress-

rs. For example, after 24 h of single housing, adolescent (postnatalay 35) rats had reduced corticosterone release when placed in

novel environment with a familiar partner compared to whenlaced with an unfamiliar partner, and both familiar and unfamiliarairings of rats had reduced corticosterone release compared to ratslone in a novel environment (Terranova et al., 1999). Adolescentpostnatal day 30) rats had lower plasma prolactin concentrationsconsidered an indication of reduced stress) when with a conspe-ific in an open field than when alone (Wilson, 2000). Isolateddolescents, but not paired adolescents, had an increase in corti-osterone concentrations when administered nicotine (Pentkowskit al., 2011). We previously reported that after 1 h of confined isola-ion, adolescent rats (postnatal day 30) recovered to baseline moreuickly when returned to housing with a familiar cage partner thano an unfamiliar cage partner (McCormick et al., 2007). None ofhese studies, however, included an adult comparison group.

Two studies have compared the effect of social context on stressesponses at both ages. Varlinskaya et al. (2013) investigated neu-al activation in response to ethanol or to saline injections in ratsested alone or with an unfamiliar partner at postnatal day 28adolescent) or 70 (adult). Whereas the unfamiliar partner condi-ion activated regions of the extended amygdala (central nucleus,asolateral nucleus, bed nucleus of the stria terminalis), lateral sep-um, and lateral hypothalamus in adolescents and not in adults, thenfamiliar partner condition suppressed activation in the nucleusccumbens, anterior cingulate cortex, and vomeronasal organ indults and not adolescents. Although measures of HPA functionere not included, the results highlight the qualitatively different

esponses of the two age groups to social experiences. Hall and

omeo (in press) reported that adolescents showed an increase

n ACTH and corticosterone concentrations when their cage mateeturned from 30 min of restraint, but not when the cage mateeturned from a holding cage for 30 min, whereas adults responded

roscience 35 (2014) 25–34

similarly to both conditions. In a separate study in the same article,however, adolescents (adults were not tested) showed no effect ofsocial buffering on stress responses; similar recovery was found interms of ACTH and corticosterone concentrations whether with orwithout a partner after restraint stress.

In the present study, we tested the hypothesis that the effectsof social buffering on HPA responses to stress would be greater inadolescents than in adults. We directly compared the corticoste-rone response to, and recovery from, 1 h of isolation stress whenwith a familiar or unfamiliar cage partner in adolescent (postna-tal day 30) and adult (postnatal day 70) male rats. In addition,we investigated neural activation in response to the stressor andsocial contexts at both ages using the protein expression of theimmediate early gene zif268 as the marker of functional activation.Although expression of Fos is the most commonly used immediateearly gene product in mapping studies (Kovacs, 2008), we mea-sured cells immunoreactive for the protein Zif268 because of its rolein neural plasticity, which may be particularly relevant consideringthe importance of social learning and ongoing brain developmentin adolescence. Zif268 acts as a transcription factor that regulatesthe expression of several genes known to contribute to the consoli-dation and recall of long term memories (reviewed in Knapska andKaczmarek, 2004). Zif268 expression is also increased in responseto stressors and exposure to novel stimuli (reviewed in Knapska andKaczmarek, 2004). We predicted that compared to adults, adoles-cents would show more prolonged activation of nuclei associatedwith the stress response, such as the paraventricular nucleus ofthe hypothalamus, arcuate nucleus, and the basolateral and centralnucleus of the amygdala, and particularly so in brain regions thatare also important for regulating social behavior, such as the medialprefrontal cortex, hippocampus, and medial amygdala (Martinezet al., 2002).

2. Methods

2.1. Animals

Male Long–Evans rats (N = 96) were obtained from Charles River, St. Constant,Quebec, at 22 days of age. Rats were housed in pairs and maintained under a 12 hlight-dark cycle (lights on at 08:00 h) with food and water available ad libitum. Useof animals in this experiment was approved by the Brock University InstitutionalAnimal Care and Use Committee (ACUC) and was carried out in adherence to theCanadian Council of Animal Care guidelines.

2.2. Stress conditions

Rats were randomly assigned to be tested either as adolescents (PND 30) oras adults (PND 70) and to experimental conditions, which were the time points atwhich blood samples were collected and the animals were perfused for brain col-lection. Two time points were used to determine the stress response of adolescentsand adults: (i) baseline (immediately after removal from the home cage), and (ii)after isolation for 1 h (isolation). To determine the influence of partner familiarityon recovery from the isolation stress, two time points were used: (i) after isolationfor 1 h and a 1 h return to either the original cage partner (familiar partner) or to anew cage partner (unfamiliar partner) and new cage, and (ii) after isolation for 1 hand a 3 h return to either the familiar partner or to a new cage partner. Isolationconsisted of removing rats from their home cage and confining them in ventilated,round plastic containers for 1 h during the light phase of the diurnal cycle. Adoles-cent isolation containers were 14 cm in diameter and 10 cm in height, whereas adultisolation containers were 20 cm in diameter and 12 cm in height. Isolation stress issimilar to restraint stress in that both involve restricting movements, with the maindifference between the two being the shape of containers (round for our isolationprocedure, and tubular for restraint procedures). The cage partners to which ratsreturned after isolation had also undergone isolation at the same time (cage part-ners were both returning to a familiar partner or an unfamiliar partner at the sametime). Experiments involved two cohorts of rats and experiments were conductedbetween 09:00 and 13:30 h to minimize the influence of circadian variation. All ratswere housed in the same room of the animal care facility.

2.3. Plasma corticosterone and testosterone assays

A blood sample (approximately 200 �l) was obtained from a tail nick and col-lected in ice chilled blood collection tubes (Sarstedt) within 3 min. Samples were

Page 3

vl Neu

c−aftLdv0

2

dsp4shu

PaPamwaahaaa((mmwd

2

eIataobpTHPwbrBautwab

2

caf(tctbdiaf

T.E. Hodges et al. / Int. J. De

entrifuged at 1730 × g and 4 ◦C for 10 min, and plasma was collected and stored at20 ◦C until the assays. Steroids were extracted from 50 �l samples using ethyl ethernd the samples were reconstituted in 100 �l of the buffer provided in the kit, thenurther diluted (20 �l in 980 buffer). Plasma corticosterone and testosterone concen-rations were determined using enzyme-linked immunosorbent assay kits (Neogen,ansing, MI) and a Biotek Synergy plate reader. Duplicate aliquots (50 �l) of the finalilution were used for the assay. Intra-assay variance was under 10% and inter-assayariance was under 15%. Assay sensitivity was 0.05 ng/mL for corticosterone and.006 ng/mL for testosterone.

.4. Zif268 immunohistochemistry

Immediately after blood sampling, rats were deeply anaesthetized by an over-ose of sodium pentobarbital (150 mg/kg) and perfused transcardially with 0.9%aline followed by 4% paraformaldehyde in 0.1 M phosphate buffered saline (PBS;H 7.4). Brains were removed from the skulls and post-fixed in a 30% sucrose and% paraformaldehyde solution until equilibrated. Coronal sections (37 �m; sectionseparated by 222 �m) were collected throughout the medial prefrontal cortex, theypothalamus, and the hippocampus and were stored in cryoprotectant at −20 ◦Cntil the time of assay.

Free-floating coronal sections were washed thoroughly in 0.1 M PBS, then inBS-X (0.1 M PBS with 3% Triton X-100), and incubated at room temperature in

0.3% H2O2 in 0.1 M PBS-X solution for 30 min. Sections were then washed inBS-X, blocked at room temperature in 10% goat serum (Sigma) solution for 1 h,nd incubated at 4 ◦C overnight in primary anti-body (1:10,000; zif-268 rabbitAb; 15F7; Cell Signaling Technology, Inc.) in PBS-X. The next day, sections wereashed in PBS-X and then incubated for 2 h at room temperature in secondary

ntibody (biotinylated goat anti-rabbit IgG; 1:400; Vector Laboratories, Inc.). Afternother series of washes in PBS-X, sections were incubated in an avidin–biotinorseradish peroxidase complex (Vector Laboratories, Inc.) for 1.5 h at room temper-ture. Horseradish peroxidase was visualized with 3,3′-diaminobenzidine (DAB) in

3 M sodium acetate buffer containing 0.05% H2O2 (Vector Laboratories, Inc.). After final series of washes in PBS-X, sections were mounted on Superfrost Plus slidesFisher Scientific, Inc.), dried, dehydrated in increasing concentrations of ethanol70%, 95%, 100%), placed in xylenes, and coverslipped using Permount mounting

edium (Fisher Scientific, Inc.). To facilitate identification of subregions of theedial prefrontal cortex and amygdala, those sections were lightly counterstainedith neutral red before they were coverslipped. No immunoreactive cells wereetected in control sections that were not treated with the primary antibody.

.5. Microscopy and cell counting

Immunostained sections were analyzed using a Nikon Eclipse 80i microscopequipped with a digital camera (Nikon DXM1200F) and Nikon ACT-1 software.mmunoreactive (ir) cell counts were conducted blind to experimental conditionnd at 400× magnification in a 250 �m2 area in each hemisphere of the paraven-ricular nucleus (PVN), the arcuate nucleus, the medial prefrontal cortex (mPFC;nterior cingulate cortex, infralimbic cortex, prelimbic cortex), the pyramidal layerf the hippocampus (CA1, CA2, & CA3), the granule layer of the dentate gyrus, theasolateral amygdala (BLA), the central nucleus of the amygdala (CeA), the medialosterodorsal amygdala (MePD), and the medial posteroventral amygdala (MePV).hese areas were chosen for examination based on their roles in social behavior andPA function (Martinez et al., 2002). Regions were identified according to the atlas ofaxinos and Watson (Paxinos and Watson, 2005): mPFC sections used for countingere within the coordinates from bregma of 3.00 mm and 2.52 mm; for PVN, within

regma −1.44 and −1.72; for arcuate nucleus, within bregma −2.52 and −2.76; foregions of the hippocampal formation, within bregma −2.64 and −3.24; for CEA andLA, within bregma −2.52 and −2.92; for MePD and MePV, within bregma −2.76nd −3.24. The mean number of ir-cells per hemisphere per brain region per rat wassed for analysis for all regions of interest except the hippocampal formation. Forhe hippocampal formation, in each hemisphere of each of three sections, ir-cellsere counted in five separate regions, the CA1, CA2, and CA3, which were averaged

nd labeled as the pyramidal layer, and in the suprapyramidal and infrapyramidallades of the dentate gyrus, which were averaged and labeled as the granule layer.

.6. Statistical analyses

Statistical analyses were performed using SPSS version 20 software andonsisted of between group analysis of variance with the factors of Age (adolescent,dult) and Time (baseline, after 1 h isolation) for the Stress analyses, and with theactors of Age (adolescent, adult), Partner Familiarity (familiar, unfamiliar), and Timeafter 1 h isolation and 1 h return to the colony, after 1 h isolation and 3 h return tohe colony) for the Post-stress recovery analyses. Whenever multiple regions wereounted within the same sections, brain region was included as a within-subject fac-or in analyses. Post hoc test analyses consisted of F tests for simple effects and/or

etween group t-tests, where appropriate. A between group t-test also was used toetermine whether testosterone concentrations differed between baseline and 1 h

solation stress adult groups. Although the experiments initially involved n = 8 perge for each time point, final sample sizes were reduced to n = 5–8 for each measureor several reasons (e.g., unable to obtain blood sample; damaged sections; sections

roscience 35 (2014) 25–34 27

not within a region of interest). An alpha level of p ≤ 0.05 was used to determinestatistical significance, although values of p ≤ 0.10 are noted.

3. Results

Table 1 provides a summary of the results.

3.1. Corticosterone

Stress: An Age × Time ANOVA indicated that corticosteroneconcentrations were higher after isolation than at baseline(F1,22 = 27.14, p < 0.001). There was no effect of Age (p = 0.63) orinteraction of Age and Time (p = 0.93). Post-stress: An Age × Time (1or 3 h after isolation) × Partner Familiarity (familiar or unfamiliar)ANOVA indicated that corticosterone concentrations were lowerwhen returned to a familiar partner than to an unfamiliar partner(F1,44 = 5.75, p = 0.02; all other main effects and interactions werep > 0.50; see Fig. 1).

3.2. Testosterone

Testosterone concentrations were below the limit of detectionof the assay for the adolescents. Stress: In adults, testosterone con-centrations did not differ significantly between baseline and 1 hisolation groups (t8 = 0.82, p = 0.44). Post-stress: Testosterone con-centrations were higher 1 h than 3 h after isolation (F1,16 = 10.40,p = 0.004). There was no effect of Partner Familiarity (p = 0.54).The interaction of the main effects was not significant (p = 0.25),although exploratory post hoc comparisons indicated that thedecrease in testosterone concentrations from 1 h to 3 h after iso-lation stress was significant in those with a familiar partner only(p = 0.02; p = 0.24 for unfamiliar partners), and as a result testos-terone concentrations were moderately higher when with anunfamiliar cage partner (p = 0.06) than with a familiar cage partner(p = 0.78) for 3 h (see Fig. 1).

3.3. Zif268 expression

3.3.1. PVNStress: An Age × Time (baseline, 1 h isolation) ANOVA on Zif268-

ir cell counts in the PVN found no significant group differences orinteraction (all p > 0.30). Post-stress: An Age × Time (1 or 3 h afterisolation) × Partner Familiarity ANOVA on Zif268-ir cell counts inthe PVN found that adolescents had higher ir-cell counts in the PVNthan did adults (F1,38 = 4.40, p = 0.04) and ir-cell counts were higher1 h after the end of isolation than 3 h after (F1,38 = 6.52, p = 0.02).Neither the effect of Partner Familiarity nor any of the interactionswere significant (all p > 0.10; see Fig. 2).

3.3.2. Arcuate nucleusStress: Zif268-ir cell counts in the arcuate nucleus were higher

after 1 h isolation compared to baseline (F1,21 = 8.78, p = 0.007).There was no effect of Age (p = 0.79), although the Age × Time inter-action approached significance (p = 0.10). Post-stress: An Age × Time(1 or 3 h after isolation) × Partner Familiarity ANOVA on Zif268-ircell counts in the arcuate nucleus found no significant main effectsor interactions (all p > 0.15, except for trend for higher counts inadolescents than in adults p = 0.10; see Fig. 2).

3.3.3. mPFC regionsStress: An Age × mPFC region (ACC, Prelimbic, Infralim-

bic) × Time (baseline, 1 h isolation) ANOVA on Zif268-ir cell counts

found an interaction of mPFC region and Time (F2,42 = 4.66, p = 0.02).The difference between baseline and isolation was not signif-icant for any region of the mPFC, although the lower ir-cellcounts in the ACC after isolation compared to baseline approached

Page 4

28 T.E. Hodges et al. / Int. J. Devl Neuroscience 35 (2014) 25–34

Table 1Summary of the results.

Stress Post-stress

Age effect(adolescentscompared to adult)

Isolation effect(compared tobaseline)

Age effect(adolescentscompared to adult)

Partner familiarity effect(Unfam. compared to fam.partner)

Time effect(1 h compared to3 h post-stress)

EndocrineCorticosterone n.s. ↑ n.s. ↑ n.s.Testosterone ↓ n.s. ↓ n.s. ↓

Zif268-irPVN n.s. n.s. ↑ n.s. ↓Arcuate nucleus n.s. ↑ n.s. (↑) n.s. n.s.mPFC n.s. n.s. (↓) n.s ↓ (adol. only) n.s.Pyramidal layer n.s. ↓ n.s. n.s. n.s. (↑)Granule layer n.s. n.s. n.s. ↓ ↑BLA n.s. ↓ n.s. n.s. n.s.CEA n.s. ↓ n.s. n.s. n.s.MePD n.s. n.s. n.s. n.s. (↑) n.s.

n.s.

A te the

sl(fai(est

F1p(

MePV n.s. n.s.

rrows indicate the direction of an effect of p ≤ 0.05; arrows within brackets indica

ignificance (p = 0.07). Post-stress: An Age × Time (1 or 3 h after iso-ation) × Partner familiarity (familiar or unfamiliar) × PFC regionACC, Prelimbic, Infralimbic) ANOVA on Zif268-ir cell countsound an interaction of PFC region and Age (F2,86 = 6.03, p = 0.006),nd near significant interaction of Age and Partner Familiar-ty (F1,43 = 3.35, p = 0.08) and of Age × Partner Familiarity × Time

F1,43 = 2.95, p = 0.09). Thus, separate analyses were conducted forach region of the mPFC. Because interactions with age were stillignificant for the separate regions with those analyses, analyseshen were conducted for the age groups separately.

ig. 1. Mean (±S.E.M.) concentration of plasma corticosterone for adolescent and adult,

and 3 h of recovery post-stress with either a familiar or unfamiliar cage partner (n = 5 < 0.001); #higher than in rats with a familiar partner, collapsed across Age and Time Pomain effect of Time Post-Stress, p = 0.004); +post hoc comparison showed decline from 1

n.s. n.s. (↓)

direction of an effect of p ≤ 0.10. n.s. = no statistically significant effect, p > 0.05.

In adolescents, there were higher ir-cell counts in the ACC thanin the prelimbic and infralimbic regions (F2,48 = 9.69, p < 0.001), ir-cell counts were higher in those with an unfamiliar partner thana familiar partner (F1,24 = 6.52, p = 0.02), and there was an interac-tion of Partner Familiarity and Time (F1,24 = 5.38, p = 0.03), with theeffect of Partner Familiarity driven by the 1 h after isolation time

point (p = 0.02, and p = 0.83 at 3 h time point). In adults, there werehigher ir-cell counts in the ACC than in the prelimbic and infral-imbic regions (F2,38 = 21.47, p < 0.001), but no other main effect orinteraction was significant (p > 0.30).

and of plasma testosterone for adult before and after 1 h isolation stress and after–8 per group). *Higher than baseline, collapsed across Age (main effect of Time,st-Stress (main effect of Partner Familiarity, p = 0.02); @higher than 3 h post-stress

to 3 h significant only in those with a familiar cage partner, p = 0.02.

Page 5

T.E. Hodges et al. / Int. J. Devl Neuroscience 35 (2014) 25–34 29

Fig. 2. Mean (±S.E.M.) Zif268 immunoreactive cell counts in brain regions for adolescent and adult rats before and after 1 h isolation stress and after 1 and 3 h of recoverypost-stress with either a familiar or unfamiliar cage partner (n = 5–8 per group). @Higher than 3 h post-stress, collapsed across Age (main effect of Time Post-Stress, p = 0.02);& ed aco in in Si gnific

3

ufpateoeiaIcwpw

higher in adolescents (main effect of Age, p = 0.04); *higher than baseline, collapsf interest (images are used with permission from Paxinos & Watson, The Rat Branterest located under low magnification that were then photographed at 400× ma

.3.4. Hippocampal regionsStress: An Age × Hippocampal Region (Pyramidal, Gran-

le) × Time (baseline, 1 h isolation) ANOVA on Zif268-ir cell countsound an interaction of Hippocampal Region and Time (F1,25 = 4.62,

= 0.04). Zif268-ir cell counts were lower after 1 h isolation thant baseline in the pyramidal layer (p = 0.05) and did not differ byime in the granule layer (p = 0.34). Post-stress: There was no mainffect of Age, Partner Familiarity (familiar or unfamiliar), Time (1r 3 h after isolation), or significant interaction among the mainffects for cell counts in the pyramidal layer, although the higherr-cell counts 3 h after isolation compared to 1 h after isolationpproached significance (F1,53 = 3.03, p = 0.09; all other p > 0.39).n the granule layer, cell counts were higher 3 h after isolation

ompared to 1 h after isolation (F1,53 = 8.75, p = 0.005), and higherhen with a familiar cage partner than with an unfamiliar cageartner (F1,53 = 4.78, p = 0.03). No other main effect or interactionas significant (p > 0.20; see Fig. 3).

ross Age (main effect of Time, p = 0.007). Atlas images used for selection of regionsterotaxic Coordinates, 5/e, 2005, Elsevier). Boxes depict the areas with regions of

ation.

3.3.5. CeA and BLAStress: An Age × Amygdala Region (BLA, CeA) × Time (baseline,

1 h isolation) ANOVA on Zif268-ir cell counts found higher counts inthe BLA than in the CeA (F1,24 = 5.66, p = 0.03) and higher counts atbaseline than after isolation (F1,24 = 4.20, p = 0.05). Post-stress: Therewas no main effect of Age, Partner Familiarity (familiar or unfa-miliar), or Time (1 or 3 h after isolation), or significant interactionamong the main effects for cell counts in either the BLA or CeA(p > 0.14, except for the interaction of Time and Partner Familiar-ity = 0.09 in the BLA; see Fig. 4).

3.3.6. MeAStress: An Age × Medial Amygdala Region (MePD, MePV) × Time

(baseline, 1 h isolation) ANOVA on Zif268-ir cell counts foundhigher counts in the MePV than in the MePD (F1,23 = 35.41,p < 0.001). Post-stress: The effect of Partner Familiarity approachedsignificance in the MePD (F1,48 = 3.00, p = 0.09), and the effect of

Page 6

30 T.E. Hodges et al. / Int. J. Devl Neuroscience 35 (2014) 25–34

Fig. 3. Mean (±S.E.M.) Zif268 immunoreactive cell counts in brain regions for adolescent and adult rats before and after 1 h isolation stress and after 1 and 3 h of recoverypost-stress with either a familiar or unfamiliar cage partner (n = 5–8 per group). ++Higher than in rats with a familiar partner, in adolescents only (p = 0.05); *lower thanbaseline, collapsed across Age (main effect of Time, p = 0.05); #higher than in rats with an unfamiliar partner, collapsed across Age and Time Post-Stress (main effect ofPartner Familiarity, p = 0.03); @lower than 3 h post-stress, collapsed across Age (main effect of Time Post-Stress, p = 0.005). Atlas images used for selection of regions ofinterest (images are used with permission from Paxinos & Watson, The Rat Brain in Sterotaxic Coordinates, 5/e, 2005, Elsevier). Boxes depict the areas with regions of interestlocated under low magnification that were then photographed at 400× magnification.

Page 7

T.E. Hodges et al. / Int. J. Devl Neuroscience 35 (2014) 25–34 31

Fig. 4. Mean (±S.E.M.) Zif268 immunoreactive cell counts in brain regions for adolescent and adult rats before and after 1 h isolation stress and after 1 and 3 h of recoverypost-stress with either a familiar or unfamiliar cage partner (n = 5–8 per group). *Lower than baseline, collapsed across Age (main effect of Time, p = 0.05). Atlas image usedfor selection of regions of interest (images are used with permission from Paxinos & Watson, The Rat Brain in Sterotaxic Coordinates, 5/e, 2005, Elsevier) and an image of aZif268 region of interest at 400× magnification overlaying an image at 100× magnification comparable to the boxed region of the atlas image. Boxes depict the areas withregions of interest located under low magnification that were then photographed at 400× magnification.

Page 8

3 vl Neu

To

4

4

ftpciptccib

ctMcsi3eeaahcaai

tAsMfefitil

bdolcsic(cr2l(dos

2 T.E. Hodges et al. / Int. J. De

ime approached significance in the MePV (F1,48 = 2.78, p = 0.10) (allther main effects and interactions p > 0.25; see Fig. 4).

. Discussion

.1. Endocrine measures

In contrast to our hypothesis, adolescents and adults did not dif-er in corticosterone concentrations after 1 h of isolation stress or athe 1 and 3 h recovery from stress time points. In keeping with ourrevious report in 30-day-old adolescents (McCormick et al., 2007),orticosterone concentrations were higher when with an unfamil-ar cage partner during the recovery phase than with a familiar cageartner. The present results extend our previous report to show thathis difference was not moderated by age, and that the higher corti-osterone concentrations in those with an unfamiliar cage partnerontinued to be evident 3 h after isolation stress (our previous studynvolved only a 1 h recovery time point). Thus, the effect of socialuffering was similar in adolescents and adults.

Our findings are in contrast to several reports of prolonged corti-osterone release in response to stressors in adolescents comparedo adults (reviewed in Eiland and Romeo, 2013; McCormick and

athews, 2010). It is possible that a difference between adoles-ents and adults was missed because of the 1 h interval betweenamples; others reported a higher concentration of corticosteronen adolescents compared to adults during the recovery phase after0 min of restraint at the 30 min recovery time point (e.g., Binghamt al., 2011; Lui et al., 2012; Romeo et al., 2004). Others, how-ver, reported higher corticosterone concentrations both duringnd 30 min after 90 min of restraint in adolescents compared todults (Doremus-Fitzwater et al., 2009). In other studies, adults hadigher (Cao et al., 2010), or did not differ in corticosterone releaseompared to adolescents to novelty stress (Goldman et al., 1973) orfter injection of nicotine (Cruz et al., 2008), and did not differ afterdministration of ethanol (Willey et al., 2012). Thus, age differencesn HPA function are likely stressor-specific.

Testosterone concentrations were below the detection limit ofhe assay in adolescents, in keeping with their prepubertal status.lthough baseline concentrations of testosterone have suppres-ive effects on stress-induced corticosterone release in adults (e.g.,cCormick and Mahoney, 1999; Viau and Meaney, 1996), the HPA

unction of prepubertal adolescents does not appear to be influ-nced significantly by testosterone (Romeo et al., 2004). Thus, ournding of no difference between adults and adolescents in cor-icosterone concentrations in response to stress should not benterpreted as evidence of mature HPA function in prepubertal ado-escents; its regulation by sex hormones develops post-pubertally.

In adults, there was no significant change in testosterone fromaseline after 1 h of isolation stress, but a marked increase was evi-ent after 1 h return to the colony after isolation stress irrespectivef whether paired with a familiar or unfamiliar partner. Neverthe-ess, only those with a familiar partner decreased in testosteroneoncentrations from the 1 to 3 h recovery period. Although chronictress has been reported to decrease testosterone concentrationsn numerous species, there are also several reports of increasedoncentrations of testosterone after exposure to an acute stressorreviewed in Chichinadze and Chichinadze, 2008). For example,onsistent with our results, testosterone was increased in adultats 60 min after cessation of 30 min of restraint stress (Foilb et al.,011). Moreover, a rise in testosterone also is linked to the estab-

ishment of dominance relationships and territorial challenges

Gleason et al., 2009; Wingfield and Sapolsky, 2003). Thus, theecline in testosterone from 1 to 3 h after recovery was significantnly in those with a familiar partner either because of the addedtress of an unfamiliar partner slowing the return to baseline, as

roscience 35 (2014) 25–34

was found for corticosterone, or because of the initiation of dom-inance mechanisms when encountering an unfamiliar conspecificslowing decline.

4.2. Zif268 expression

Overall, the pattern of Zif268 immunoreactive (ir) cell countsbefore and after stress was similar for adolescents and adults, withthe possible exception of the PVN. An increase in ir-cell counts inthe PVN in response to isolation appeared delayed in adolescents,with higher ir-cell counts in adolescents than in adults in the recov-ery period and not in the baseline and 1 h isolation time points, andwith the significant decline in ir-cell counts from 1 to 3 h of recoverydriven primarily by the adolescent groups. There was a significantincrease in ir-cell counts in the arcuate nucleus after 1 h isolationcompared to baseline that did not differ by age, and ir-cell countsremained high in the post stress recovery period. Several studieshave reported increased expression of immediate early genes inthe PVN and arcuate nucleus in response to various stressors (e.g.,Chung et al., 2000; Cullinan et al., 1995; Kwon et al., 2006; Meliaet al., 1994). Reciprocal connections between the CRH and AVP cellsof the parvocellular PVN and the arcuate nucleus are involved inregulating stress responses (reviewed in Charmandari et al., 2005).Our results are consistent with several reports of greater expressionof the immediate early gene c-fos in the PVN in prepubertal ado-lescents (postnatal days 28–30) in response to novel environments(Novak et al., 2007), restraint (Lui et al., 2012; Romeo et al., 2006;Viau et al., 2005), and wheel running (Kim et al., 2004) comparedto adults. No age difference in Fos expression after either 15 minor 2 h of restraint, however, was reported in one study (Kellogget al., 1998). This study instead found that Fos-ir cell counts dif-fered most between the age groups in olfactory regions of the brainafter restraint (Kellogg et al., 1998).

In contrast to the increased number of Zif268-ir cells in the PVNand arcuate nucleus in response to stress, there was a significantdecrease in ir-cell counts after 1 h isolation compared to base-line in the basolateral amygdala, central nucleus of the amygdala,and in the pyramidal layer of the hippocampus. There is evidencethat expression of immediate early genes in the hippocampus maynot reflect the stressful nature of an experience as it does in thePVN, but rather the processing of stimulus features of the expe-rience (Pace et al., 2005). The decreased Zif268 ir-cell counts inthe hippocampus might be because of the reduced sensory expe-riences and limited movement afforded by the confined isolation.Decreased Fos expression in the hippocampus, medial prefrontalcortex, and amygdala accompanied by increased expression in thedorsal raphe and PVN in response to acute stress has been reportedin aged rats (Shoji and Mizoguchi, 2010). Thus, how a stressor isperceived across developmental stages may vary and determinethe pattern of neuronal activation observed.

Partner familiarity effects in the recovery period were mod-est, with ir-cell counts higher in the granule layer of the dentategyrus in those with a familiar partner compared to those with anunfamiliar partner. In contrast, ir-cell counts were higher in themPFC in those with an unfamiliar partner compared to those witha familiar partner, but the difference was significant only in ado-lescents. Zif268 expression in the mPFC has been associated withreduced anxiety in social interaction tests in adult males (Stacket al., 2010). Further, the pattern across all regions other than thehippocampal formation was for ir-cell counts to be higher in thosewith an unfamiliar partner than with a familiar partner. In adultsplaced into the context in which they had received electric shocks,

Fos-ir cell counts were reduced in the PVN when accompanied bya peer than when alone, and particularly so if the peer had notbeen shocked previously (Kiyokawa et al., 2004). Fos-ir cell countswere higher in the CeA, BLA, BNST, and lateral hypothalamus in

Page 9

vl Neu

aponi

dimfatiitlsmtaeimt(psopsaamemvs

iwecbapatzeemic2baphpip

5

h

T.E. Hodges et al. / Int. J. De

dolescents (and not adults) placed in a familiar test arena with aeer than if alone (Varlinskaya et al., 2013). Thus, greater effectsf partner familiarity may have been observed in this study hadon-stressed partners and singly-housed comparison groups been

ncluded in the post-stress recovery period.Further, in the present study the post-stress period occurred

uring the light phase (rats are nocturnal) in the colony hous-ng room, which may have limited the social interactions that

ay be required for partner familiarity effects; we previouslyound that adolescent rats (no adults were tested) isolated forn hour spent most of their time inactive and in physical con-act with their cage partner after about 30 min back in the colony,rrespective of with a new, unfamiliar partner or with their orig-nal cage partner (McCormick et al., 2007). Others have reportedhat social play increases corticosterone concentrations in ado-escent rats compared to adolescent rats that did not engage inocial play (no adults were tested) (Gordon et al., 2002). Thus, itay be that the prolonged release of corticosterone we found in

hose with an unfamiliar partner is based in greater social inter-ctions when with a novel peer. The same researchers (Gordont al., 2002), however, found no differences in c-fos expressionn the PVN, amygdala, or anterior cingulate cortex between ani-

als that played and those that did not, which they interpretedo indicate that engaging in play behaviors was not “stressful”there also was no difference between dominant and submissivelay partners in corticosterone concentrations or c-fos expres-ion); instead, those that engaged in play had higher expressionf c-fos in the tectum, dorsal striatum and ventral striatum, andarietal cortex. Recent research using c-fos expression as the mea-ure indicates that social play in adolescents engages a broadrray of neural structures associated with motivation, reward,nd motor and sensory processing (van Kerkhof et al., 2013), andany of these structures also are activated by stressors (Martinez

t al., 2002). Thus, inclusion of a broader array of neural regionsay have allowed us to better decipher the role of isolation

ersus the role of social interactions and partner familiarity in ourtudy.

Including more time points after stress also may have beennformative. For example, other studies have reported a secondary

ave of Fos (Meyza et al., 2007) and of Zif268 (Day et al., 2001)xpression in the amygdala 4 h after a stress exposure. Further, glu-ocorticoid receptor-induced hippocampal zif268 expression haseen shown to occur through the activation of both rapid-onsetnd slow-onset pathways (Revest et al., 2005). In the rapid onsetathway, GR activation induces zif268 expression within 30 min in

manner that is highly dependent on concentrations of glucocor-icoids, whereas in the slow-onset pathway, GR activation inducesif268 expression within 2 h in a MAPK-dependent manner (Revestt al., 2005). In addition, the temporal differences in immediatearly gene expression between adolescents and adults may deter-ine the extent to which age differences are observed. Earlier peaks

n c-fos expression after restraint stress were reported for adoles-ents than for adults (Lui et al., 2012; Romeo et al., 2006; Viau et al.,005). Thus, the conclusions regarding function that can be drawnased on differences or lack of differences in zif268 immunore-ctivity at specific windows of time are limited. Nevertheless, theattern of zif268 immunoreactivity across time and conditions wasighly similar in adolescents and adults, in contrast to the differentatterns of expression for adolescents and adults observed in fos

mmunoreactivity (Varlinskaya et al., 2013), and modest effects ofartner familiarity were observed.

. Conclusion

The majority of investigations of social buffering effects onypothalamic–pituitary–adrenal function have involved maternal

roscience 35 (2014) 25–34 33

buffering of neonatal stress responses (Hostinar et al., 2014). Ofthe few studies of adolescents and of adults, only one involvedboth age groups; whereas adolescent HPA function was sensitiveto the stress-status of their cage partner, adult HPA function wasnot (Hall and Romeo, in press). In a second experiment, however,adolescents (adults were not tested) showed no effect of socialbuffering on stress responses; similar recovery was found in termsof ACTH and corticosterone concentrations whether with or with-out a partner after restraint stress (Hall and Romeo, in press). Ourresults contrast those of Hall and Romeo (in press) and indicatethat adolescent HPA function is sensitive to the familiarity of thecage partner during stress recovery, replicating our previous report(McCormick et al., 2007), and extending that report to show that thesocial buffering effects on corticosterone concentrations after anacute stressor are similar in adolescents and adults. It remains pos-sible, though, that differences between adolescents and adults insocial buffering effects would emerge under conditions of chronicstress and/or chronic social instability, given the reports of long-lasting detrimental effects of such experiences when encounteredin adolescence but not when encountered in adulthood (reviewedin Green and McCormick, 2013; McCormick and Green, 2013). Wecurrently are investigating this possibility.

Acknowledgements

The research was conducted as part of the master’s degreerequirements at Brock University of TEH. We thank Julie Vasseurand Bailee Malivoire for assistance in data collection. CMM holdsa Natural Sciences and Engineering Research Council of Canada,(288348) (NSERC) Discovery Grant that supported the research.MRG holds a NSERC Postgraduate Fellowship and JJS holds anOntario Graduate Scholarship.

References

Armario, A., Luna, G., Balasch, J., 1983a. The effect of conspecifics on corticoadrenalresponse of rats to a novel environment. Behav. Neural Biol. 37, 332–337.

Armario, A., Ortiz, R., Balasch, J., 1983b. Corticoadrenal and behavioral-response toopen-field in pairs of male-rats either familiar or non-familiar to each other.Experientia 39, 1316–1317.

Bingham, B., McFadden, K., Zhang, X.Y., Bhatnagar, S., Beck, S., Valentino, R., 2011.Early adolescence as a critical window during which social stress distinctlyalters behavior and brain norepinephrine activity. Neuropsychopharmacology36, 896–909.

Blakemore, S.J., 2012. Development of the social brain in adolescence. J. R. Soc. Med.105, 111–116.

Buwalda, B., Kole, M.H., Veenema, A.H., Huininga, M., de Boer, S.F., Korte, S.M., Kool-haas, J.M., 2005. Long-term effects of social stress on brain and behavior: a focuson hippocampal functioning. Neurosci. Biobehav. Rev. 29, 83–97.

Cao, J.R., Belluzzi, J.D., Loughlin, S.E., Dao, J.M., Chen, Y.L., Leslie, F.M., 2010. Locomotorand stress responses to nicotine differ in adolescent and adult rats. Pharmacol.Biochem. Behav. 96, 82–90.

Charmandari, E., Tsigos, C., Chrousos, G., 2005. Endocrinology of the stress response.Annu. Rev. Physiol., 259–284.

Chichinadze, K., Chichinadze, N., 2008. Stress-induced increase of testosteroneContributions of social status and sympathetic reactivity. Physiol. Behav. 94,595–603.

Chung, K.K.K., Martinez, M., Herbert, J., 2000. c-fos expression behavioural,endocrine and autonomic responses to acute social stress in male rats afterchronic restraint: modulation by serotonin. Neuroscience 95, 453–463.

Cruz, F., DeLucia, R., Planeta, C., 2008. Effects of chronic stress on nicotine-inducedlocomotor activity and corticosterone release in adult and adolescent rats.Addict. Biol. 13, 63–69.

Cullinan, W.E., Herman, J.P., Battaglia, D.F., Akil, H., Watson, S.J., 1995. Pattern andtime course of immediate early gene expression in rat brain following acutestress. Neuroscience 64, 477–505.

Day, H.E.W., Badiani, A., Uslaner, J.M., Oates, M.M., Vittoz, N.M., Robinson, T.E., Wat-son, S.J., Akil, H., 2001. Environmental novelty differentially affects c-fos mRNAexpression induced by amphetamine or cocaine in subregions of the bed nucleusof the stria terminalis and amygdala. J. Neurosci. 21, 732–740.

de Jong, J.G., van der Vegt, B.J., Buwalda, B., Koolhaas, J.M., 2005. Social environment

determines the long-term effects of social defeat. Physiol. Behav. 84, 87–95.

Dickerson, S.S., Mycek, P.J., Zaldivar, F., 2008. Negative social evaluation, but not meresocial presence, elicits cortisol responses to a laboratory stressor task. HealthPsychol. 27, 116–121.

Page 10

3 vl Neu

D

D

E

F

F

G

G

G

G

G

G

H

H

H

H

H

H

K

K

K

K

K

K

K

L

M

M

M

M

M

4 T.E. Hodges et al. / Int. J. De

oremus-Fitzwater, T.L., Varlinskaya, E.I., Spear, L.P., 2009. Social and non-socialanxiety in adolescent and adult rats after repeated restraint. Physiol. Behav. 97,484–494.

oremus-Fitzwater, T.L., Varlinskaya, E.L., Spear, L.P., 2010. Motivational systems inadolescence: possible implications for age differences in substance abuse andother risk-taking behaviors. Brain Cogn. 72, 114–123.

iland, L., Romeo, R.D., 2013. Stress and the developing adolescent brain. Neuro-science 249, 162–171.

ile, S.E., Peet, L.A., 1980. The sensitivity of the rat corticosterone response to envi-ronmental manipulations and to chronic chlordiazepoxide treatment. Physiol.Behav. 25, 753–758.

oilb, A.R., Lui, P., Romeo, R.D., 2011. The transformation of hormonal stressresponses throughout puberty and adolescence. J. Endocrinol. 210, 391–398.

ilbert, M.H., Baker, K.C., 2011. Social buffering in adult male rhesus macaques(Macaca mulatta): effects of stressful events in single vs. pair housing. J. Med.Primatol. 40, 71–78.

leason, E.D., Fuxjager, M.J., Oyegbile, T.O., Marler, C.A., 2009. Testosterone releaseand social context. When it occurs and why. Front. Neuroendocrinol. 30,460–469.

oldman, L., Winget, C., Hollingshead, G.W., Levine, S., 1973. Postweaning devel-opment of negative feedback in the pituitary-adrenal system of the rat.Neuroendocrinology 12, 199–211.

ordon, N.S., Kollack-Walker, S., Akil, H., Panksepp, J., 2002. Expression of c-Fosgene activation during rough and tumble play in juvenile rats. Brain Res. Bull.57, 651–659.

reen, M.R., McCormick, C.M., 2013. Effects of stressors in adolescence on learningand memory in rodent models. Horm. Behav. 64, 364–379.

unnar, M., Donzella, B., 2002. Social regulation of the cortisol levels in early humandevelopment. Psychoneuroendocrinology 27, 199–220.

all, B.S., Romeo, R.D., 2014. The influence of post-stress social factors on hor-monal reactivity in prepubertal male rats. Dev. Psychobiol., http://dx.doi.org/10.1002/dev.21190 (in press).

all, F.S., 1998. Social deprivation of neonatal, adolescent, and adult rats has distinctneurochemical and behavioral consequences. Crit. Rev. Neurobiol. 12, 129–162.

ennessy, M.B., Kaiser, S., Sachser, N., 2009. Social buffering of the stress response:diversity, mechanisms, and functions. Front. Neuroendocrinol. 30, 470–482.

erbert, J., Goodyer, I.M., Grossman, A.B., Hastings, M.H., de Kloet, E.R., Lightman,S.L., Lupien, S.J., Roozendaal, B., Seckl, J.R., 2006. Do corticosteroids damage thebrain. J. Neuroendocrinol. 18, 393–411.

olt-Lunstad, J., Smith, T.B., Layton, J.B., 2010. Social relationships and mortality risk:a meta-analytic review. PLoS Med. 7, 1–20.

ostinar, C.E., Sullivan, R.M., Gunnar, M., 2014. Psychobiological mechanisms under-lying the social buffering of the hypothalamic–pituitary–adrenocortical axis: areview of animal models and human studies across development. Psychol. Bull.140, 256–282.

ellogg, C.K., Awatramani, G.B., Piekut, D.T., 1998. Adolescent development altersstressor-induced fos immunoreactivity in rat brain. Neuroscience 83, 681–689.

ikusui, T., Winslow, J.T., Mori, Y., 2006. Social buffering: relief from stress andanxiety. Philos. Trans. R. Soc. B 361, 2215–2228.

im, S.H., Kim, H., Kim, S.S., Shin, M.S., Chang, H.K., Lee, T.H., Jang, M.H., Shin, M.C.,Lee, H.H., Kim, Y.P., Kim, C.J., 2004. The influence of age on the treadmill exercise-induced c-Fos expression in the hippocampus of rats. Neurosci. Res. Commun.35, 41–50.

iyokawa, Y., Kikusui, T., Takeuchi, Y., Mori, Y., 2004. Partner’s stress status influ-ences social buffering effects in rats. Behav. Neurosci. 118, 798–804.

napska, E., Kaczmarek, L., 2004. A gene for neuronal plasticity in the mammalianbrain: Zif268/Egr-1/NGFI-A/Krox-24/TIS8/ZENK? Prog. Neurobiol. 74, 183–211.

ovacs, K.J., 2008. Measurement of immediate-early gene activation-c-fos andbeyond. J. Neuroendocrinol. 20, 665–672.

won, M.S., Seo, Y.J., Shim, E.J., Choi, S.S., Lee, J.Y., Suh, H.W., 2006. The effectof single or repeated restraint stress on several signal molecules in para-ventricular nucleus, arcuate nucleus and locus coeruleus. Neuroscience 142,1281–1292.

ui, P., Padow, V.A., Franco, D., Hall, B.S., Park, B., Klein, Z.A., Romeo, R.D., 2012. Diver-gent stress-induced neuroendocrine and behavioral responses prior to puberty.Physiol. Behav. 107, 104–111.

artinez, M., Calvo-Torrent, A., Herbert, J., 2002. Mapping brain response to socialstress in rodents with c-fos expression: a review. Stress 5, 3–13.

cCormick, C.M., Green, M.R., 2013. From the stressed adolescent to the anxious anddepressed adult: investigations in rodent models. Neuroscience 249, 242–257.

cCormick, C.M., Mahoney, E.M., 1999. Persistent effects of prenatal neonatal, oradult treatment with flutamide on the hypothalamic–pituitary–adrenal stressresponse of adult rats. Horm. Behav. 35, 90–101.

cCormick, C.M., Mathews, I.Z., 2010. Adolescent developmenthypothalamic–pituitary–adrenal function, and programming of adult learning

and memory. Prog. Neuro-Psychopharmacol. Biol. Psychiatry 34, 756–765.

cCormick, C.M., Merrick, A., Secen, J., Helmreich, D.L., 2007. Social instability inadolescence alters the central and peripheral hypothalamic–pituitary–adrenalresponses to a repeated homotypic stressor in male and female rats. J. Neuroen-docrinol. 19, 116–126.

roscience 35 (2014) 25–34

McEwen, B.S., Eiland, L., Hunter, R.G., Miller, M.M., 2012. Stress and anxietyStructural plasticity and epigenetic regulation as a consequence of stress. Neu-ropharmacology 62, 3–12.

Melia, K.R., Ryabinin, A.E., Schroeder, R., Bloom, F.E., Wilson, M.C., 1994. Inductionand habituation of immediate early gene expression in rat brain by acute andrepeated restraint stress. J. Neurosci. 14, 5929–5938.

Meyza, K.Z., Boguszewski, P.M., Nikolaev, E., Zagrodzka, J., 2007. The effect of age onthe dynamics and the level of c-Fos activation in response to acute restraint inLewis rats. Behav. Brain Res. 180, 183–189.

Novak, C.M., Parfitt, D.B., Sisk, C.L., Smale, L., 2007. Associations between behaviorhormones, and Fos responses to novelty differ in pre- and post-pubertal grassrats. Physiol. Behav. 90, 125–132.

Pace, T.W.W., Gaylord, R., Topczewski, F., Girotti, M., Rubin, B., Spencer, R.L., 2005.Immediate-early gene induction in hippocampus and cortex as a result of novelexperience is not directly related to the stressfulness of that experience. Eur. J.Neurosci. 22, 1679–1690.

Paxinos, G., Watson, C., 2005. The Rat Brain in Stereotaxic Coordinates. Elsevier,Sydney.

Pellis, S.M., Field, E.F., Smith, L.K., Field, V.C., 1997. Multiple differences in the playfighting of male and female rats. Implications for the causes and functions ofplay. Neurosci. Biobehav. Rev. 21, 105–120.

Pentkowski, N.S., Painter, M.R., Thiel, K.J., Peartree, N.A., Cheung, T.H.C., Deviche, P.,Adams, M., Alba, J., Neisewander, J.L., 2011. Nicotine-induced plasma cortico-sterone is attenuated by social interactions in male and female adolescent rats.Pharmacol. Biochem. Behav. 100, 1–7.

Revest, J.M., Di Blasi, F., Kitchener, P., Rouge-Pont, F., Desmedt, A., Turiault, M.,Tronche, F., Piazza, P.V., 2005. The MAPK pathway and Egr-1 mediate stress-related behavioral effects of glucocorticoids (vol. 8, p. 664, 2005). Nat. Neurosci.8, 835.

Romeo, R.D., Bellani, R., Karatsoreos, I.N., Chhua, N., Vernov, M., Conrad, C.D.,McEwen, B.S., 2006. Stress history and pubertal development interact to shapehypothalamic–pituitary–adrenal axis plasticity. Endocrinology 147, 1664–1674.

Romeo, R.D., Lee, S.J., Chhua, N., McPherson, C.R., McEwen, B.S., 2004. Testosteronecannot activate an adult-like stress response in prepubertal male rats. Neuroen-docrinology 79, 125–132.

Roozendaal, B., McEwen, B.S., Chattarji, S., 2009. Stress memory and the amygdala.Nat. Rev. Neurosci. 10, 423–433.

Shoji, H., Mizoguchi, K., 2010. Acute and repeated stress differentially regulatesbehavioral, endocrine, neural parameters relevant to emotional and stressresponse in young and aged rats. Behav. Brain Res. 211, 169–177.

Silk, J.B., Beehner, J.C., Bergman, T.J., Crockford, C., Engh, A.L., Moscovice, L.R., Wit-tig, R.M., Seyfarth, R.M., Cheney, D.L., 2010. Strong and consistent social bondsenhance the longevity of female baboons. Curr. Biol. 20, 1359–1361.

Stack, A., Carrier, N., Dietz, D., Hollis, F., Sorenson, J., Kabbaj, M., 2010. Sex differ-ences in social interaction in rats: role of the immediate-early gene zif268.Neuropsychopharmacology 35, 570–580.

Sturman, D.A., Moghaddam, B., 2011. The neurobiology of adolescence Changes inbrain architecture, functional dynamics, and behavioral tendencies. Neurosci.Biobehav. Rev. 35, 1704–1712.

Terranova, M.L., Cirulli, F., Laviola, G., 1999. Behavioral and hormonal effects ofpartner familiarity in periadolescent rat pairs upon novelty exposure. Psy-choneuroendocrinology 24, 639–656.

Vanderschuren, L.J., Niesink, R.J.M., Van Pee, J.M., 1997. The neurobiology of socialplay behavior in rats. Neurosci. Biobehav. Rev. 32, 309–326.

van Kerkhof, L.W.M., Trzza, V., Nulder, T., Gao, P., Voorn, P., Vanderschuren, L.J.M.J.,2013. Cellular activation in limbic brain systems during social play behaviour inrats. Brain Struct. Funct., 1–31, http://dx.doi.org/10.1007/s00429-013-0558-y.

Varlinskaya, E.I., Spear, L.P., 2008. Social interactions in adolescent and adultSprague-Dawley rats: impact of social deprivation and test context familiarity.Behav. Brain Res. 188, 398–405.

Varlinskaya, E.I., Vogt, B.A., Spear, L.P., 2013. Social context induces two uniquepatterns of c-Fos expression in adolescent and adult rats. Dev. Psychobiol. 55,684–697.

Viau, V., Bingham, B., Davis, J., Lee, P., Wong, M., 2005. Gender and puberty interacton the stress-induced activation of parvocellular neurosecretory neurons andcorticotropin-releasing hormone messenger ribonucleic acid expression in therat. Endocrinology 146, 137–146.

Viau, V., Meaney, M.J., 1996. The inhibitory effect of testosterone onhypothalamic–pituitary–adrenal responses to stress is mediated by themedial preoptic area. J. Neurosci. 16, 1866–1876.

Willey, A.R., Anderson, R.I., Morales, M., Ramirez, R.L., Spear, L.P., 2012. Effects ofethanol administration on corticosterone levels in adolescent and adult rats.Alcohol 46, 29–36.

Wilson, J.H., 2000. A conspecific attenuates prolactin responses to open-field expo-sure in rats. Horm. Behav. 38, 39–43.

Wingfield, J.C., Sapolsky, R.M., 2003. Reproduction and resistance to stress: whenand how. J. Neuroendocrinol. 15, 711–724.

Yee, J.R., Cavigelli, S.A., Delgado, B., McClintock, M.K., 2008. Reciprocal affiliationamong adolescent rats during a mild group stressor predicts mammary tumorsand lifespan. Psychosom. Med. 70, 1050–1059.