R

Ds

CEa

b

c

d

a

ARRAA

KMDICD

1

bosetsiaofims

MT

0d

Behavioural Brain Research 205 (2009) 406–413

Contents lists available at ScienceDirect

Behavioural Brain Research

journa l homepage: www.e lsev ier .com/ locate /bbr

esearch report

iurnal rhythm and stress regulate dendritic architecture andpine density of pyramidal neurons in the rat infralimbic cortex

laudia Perez-Cruza, Mária Simonb, Gabriele Flüggea,c,berhard Fuchsa,c,d, Boldizsár Czéha,∗

Clinical Neurobiology Laboratory, German Primate Center, Leibniz Institute for Primate Research, Kellnerweg 4, 37077 Göttingen, GermanyDepartment of Psychiatry and Psychotherapy, Medical School, University of Pécs, HungaryDFG Research Center Molecular Physiology of the Brain (CMPB), University of Göttingen, GermanyDepartment of Neurology, Medical School, University of Göttingen, Göttingen, Germany

r t i c l e i n f o

rticle history:eceived 2 June 2009eceived in revised form 14 July 2009ccepted 20 July 2009vailable online 28 July 2009

eywords:edial prefrontal cortexendrite

a b s t r a c t

The medial prefrontal cortex (mPFC) participates in several higher order cognitive functions and isinvolved in the regulation of the stress response. The infralimbic cortex (ILC), the most ventral part of themPFC, receives a strong afferent input from the master circadian pacemaker, the suprachiasmatic nucleus.This fact raises the possibility that, similarly to stress, the diurnal rhythm may affect structural plasticityof neurons in the ILC. Here we investigated, whether diurnal changes in combination with immobilizationstress have any impact on the dendritic morphology of layer III pyramidal neurons in the ILC. Prefrontalcortices were collected from control rats at two different time points of the diurnal cycle (12 h apart), andfrom rats exposed to 1-week of daily restraint stress either during their active or resting period. Dendritic

mmobilizationhronic stressiurnal changes

architecture and spine density of Golgi–Cox stained neurons were digitally reconstructed and analyzed.We found that in control rats during the active period, the basilar dendrites were always longer andmore complex, and had more spines than during the resting period. Similar although less pronounceddiurnal differences exist in the apical dendrites. Stress affected dendritic architecture in a way that thediurnal differences either disappeared or became reduced in their magnitude. Our findings indicate thatthe diurnal rhythm has a unique impact on the structural plasticity of pyramidal cells in the ILC and that

form

stress interferes with this

. Introduction

The medial prefrontal cortex (mPFC) is implicated in a num-er of higher order functions including cognitive and executiveperations such as working memory, selective attention, deci-ion making and goal-directed behavior as well as processingmotions and regulation of stress responses [13,14,45]. Many ofhese higher order functions that are regulated by the PFC areubject to diurnal rhythmicity and exhibit typical diurnal pattern-ng [24,28,39]. Disruption of the normal biological rhythms canffect both the cognitive functioning, such as short-term mem-

ry and attention [6,37], as well as regulation of mood [18]. Thesendings raise the possibility that a common underlying diurnalechanism may affect these higher level brain functions. The

uprachiasmatic nucleus (SCN) is the likely source of these diur-

∗ Corresponding author. Current address: Max-Planck-Institute of Psychiatry,olecular Neurobiology, Kraepelinstrasse 2-10, 80804 Munich, Germany.

el.: +49 551 3851 134/89 30622 643; fax: +49 551 3851 307/89 30622 610.E-mail addresses: bczeh@cnl-dpz.de, czeh@mpipsykl.mpg.de (B. Czéh).

166-4328/$ – see front matter © 2009 Elsevier B.V. All rights reserved.oi:10.1016/j.bbr.2009.07.021

of neuroplasticity.© 2009 Elsevier B.V. All rights reserved.

nal rhythms, since it is the master circadian pacemaker in thebrain [19]. Interestingly, the SCN has a pronounced multisynap-tic pathway, projecting via the paraventricular thalamic nucleus(PVT) to the mPFC [42]. Furthermore, this SCN–mPFC pathway isspecific to the infralimbic cortex (ILC), whereas the other mPFCsubareas are not involved [42]. Although the functional role of thisSCN–PVT–ILC pathway is yet unknown, it has been suggested that itmay play a specific role in modulating some higher level brain func-tions which exhibit diurnal rhythms, such as mood and attention[42].

The mPFC in rats consists of anatomically and functionallydistinct areas that are most commonly divided into three mainsubregions: the infralimbic (ILC), prelimbic (PLC), and anterior cin-gulate cortex (ACC) [11,20]. The mPFC, besides regulating higherorder cognitive functions, has a significant role in modulating auto-nomic and endocrine responses to stress, as it provides negative

feedback control to the hypothalamic–pituitary–adrenal (HPA) axisand regulates the stress responses of other structures [13,14]. It isgenerally accepted that the different mPFC subregions play func-tionally distinct roles in modulating the stress response [14]. Indeeda number of reports suggested that even neighboring areas such as

C. Perez-Cruz et al. / Behavioural Brain Research 205 (2009) 406–413 407

Fig. 1. (A) Experimental design and groups of experimental animals (for details, see Section 2). (B) Schematic illustrating reconstruction of the complete dendritic tree of aGolgi–Cox stained pyramidal neuron in layer III of the infralimbic cortex together with the Sholl circles that were used for numerical analysis. The points where dendritesc

tota

aiapuwaamtr

ross the virtual circles are called intersections.

he ILC and PLC may have somewhat opposing modulatory effectsn the stress response [34,43]. These findings highlight the impor-ance of distinguishing between the different regions when oneims to study the mPFC.

In the present study, we investigated whether the dendriticrchitecture of layer III pyramidal neurons in the rat infral-mbic cortex changes with the diurnal phases of resting andctivity. Prefrontal cortices were collected at two different timeoints of the diurnal cycle (12 h apart); the samples were stainedsing the Golgi–Cox method and pyramidal neurons in the ILCere three-dimensionally reconstructed. Spines were counted

nd a Sholl analysis was performed to determine the lengthnd complexity of the dendrites. Furthermore, with the sameethod, we examined the effect of 1-week of daily restraint stress

hat was applied either during the active or rest period of theats.

2. Materials and methods

2.1. Animals

Adult male Sprague–Dawley rats (Harlan-Winkelmann, Borchen, Germany)were housed in groups of three animals per cage with food and water ad libitum intemperature-controlled rooms (21 ± 1 ◦C). All animal experiments were conductedin accordance with the European Communities Council Directive of November 24,1986 (86/EEC) and the US National Institutes of Health Guide for the Care and Useof Laboratory Animals, and were approved by the Lower Saxony Federal State Officefor Consumer Protection and Food Safety, Germany. We used the minimum numberof animals required to obtain consistent data.

At the beginning of the experiment (habituation phase), rats weighed 150–170 g,

which according to Harlan-Winkelmann corresponds to an age of about 6–7 weeks.To investigate the impact of the light cycle on dendritic architecture, one set of ratswas kept on a normal light cycle (lights on at 0700, lights off at 1900), and anotherset was kept on an inverse light cycle (lights off at 0700, lights on at 1900). Bothsets of rats contained a control and a stressed group, resulting in four experimentalgroups: (1) control rats on a normal light cycle, (2) stressed rats on a normal light

408 C. Perez-Cruz et al. / Behavioural Brain Research 205 (2009) 406–413

F LC of ci pine dt erall,a M.

clwpcd

TN

ig. 2. Diurnal variation of the dendritic morphology of pyramidal neurons in the Intersections with the Sholl circles. (C) Total number of dendritic branches. (D–E) She light–dark phases (resting–active periods) especially in the basilar dendrites. Ovctive period (dark phase). Statistics: Student’s t-test: *P < 0.05. Data are mean ± SE

ycle, (3) control rats on an inverse light cycle, and (4) stressed rats on an inverseight cycle (Fig. 1A). Each of these four groups contained six animals. All animals were

eighed daily at 0800, and all handling procedures including stress exposure wereerformed between 0800 and 1500. In the case of the animals on the inverse lightycle, handling, weighing, and restraint stress procedures were conducted underim red light in order to avoid disruption of the light–dark cycle.

able 1umber of digitally reconstructed pyramidal neurons in the ILC layer III.

Control Stress

Resting period (light phase) 36 36Active period (dark phase) 35 35

ontrol rats. (A) Total length of basilar and apical dendrites. (B) The total number ofensity on the basilar and apical dendrites. Note the numerous differences betweenthe dendritic arbors were longer and more complex, having more spines during the

2.2. Experimental procedure

The experimental design is depicted in Fig. 1A. The first experimental phase(habituation) lasted 14 days, during which body weight was recorded daily. Duringthe second phase, which lasted for 7 days, animals were daily exposed to restraintstress as previously described [3,26,30]. Animals were immobilized daily for 6 h,from 0900 to 1500. Accordingly, for animals on the normal light cycle, the restraintstress took place during their resting period (light phase). In contrast, for the ani-mals on the inverse light cycle, the restraint stress took place during their activeperiod (dark phase). During restraint, each rat was placed in a plastic tube within

its home cage; during this time, it had no access to food or water. Control rats werehoused in separate rooms and not subjected to any kind of stressor, except dailyhandling.

On day 8, in the morning, animals were again weighed, anesthetized deeply withan intraperitoneal injection of a mixture of 50 mg/ml ketamine, 10 mg/ml xylazine,and 0.1 mg/ml atropine (0.2 ml/100 g of body weight), and perfused intracardially

l Brain Research 205 (2009) 406–413 409

w0mop

ww

2

ifiiaSwwGb

2

iefadtic

FibNdp�

C. Perez-Cruz et al. / Behavioura

ith 0.9% saline. All animals were perfused at the same time of the day, between900 and 1000, which corresponds to the beginning of the resting phase of the ani-als on the normal light cycle, and to the beginning of the active phase of the animals

n the inverse light cycle. After perfusion, brains were immediately dissected androcessed as described below.

Because increased adrenal weight is an indicator of sustained stress, adrenalsere removed from the animals before perfusion and were weighed. Adrenal weightas expressed as mg/100 g of body weight.

.3. Golgi–Cox staining

Immediately after perfusion, the frontal part of the brain was processed for mod-fied Golgi–Cox staining, as described by Gibb and Kolb [12]. Briefly, the brains wererst stored in Golgi–Cox solution in the dark for 14 days, followed by incubation

n 30% sucrose in 0.85% NaCl, for 3 days. Coronal 200 �m sections were obtainedt the level of the PFC using a vibratome (Vibracut 2, FTB, Bensheim, Germany).ections were collected on gelatin-coated glass slides, and the stain was developedith NH4OH for 30 min. Sections were immersed in Kodak film fixer for 30 min,ashed with water, dehydrated, cleared, mounted using Eukitt (Kindler, Freiburg,ermany) and cover-sliped; the slides were dried in the dark for at least 1 monthefore analyzing neuronal morphology.

.4. Analysis of neuronal morphology

Pyramidal neurons impregnated with the Golgi–Cox solution were readilydentified by their characteristic, almost triangular soma shape, apical dendritesxtending toward the pial surface, and numerous dendritic spines. The criteria

or digital reconstruction of neurons were: (1) soma location in ILC layer III; (2)

clear and completely stained dendritic tree without obviously truncated den-rites; (3) at least three main basilar and apical branches, each branching at leasto the third-degree branch order, and (4) clearly visible spines. Cells with somatan the middle third of the sections were chosen to minimize the number of trun-ated branches. Neurons were digitally reconstructed in three-dimensions using a

ig. 3. The effect of stress on the dendritic morphology of pyramidal neurons in thenfralimbic cortex. (A) Total length of basilar and apical dendrites. (B) The total num-er of intersections with the Sholl circles. (C) Total number of dendritic branches.ote that stress reduced dendritic length and complexity only in the basilar den-rites and only in animals that were stressed during their active periods (in the darkhase). Statistics: Three-way ANOVA followed by the Tukey’s post hoc test: *P < 0.05;p < 0.05 versus control in the same region and diurnal period. Data are mean ± SEM.

Fig. 4. Stress reduced spine density on the proximal apical dendrites of pyramidalneurons in layer III. (A) Basilar dendrites. (B) Apical dendrites. Statistics: Three-wayANOVA followed by the Tukey’s post hoc test: *P < 0.05; �p < 0.05 versus control inthe same region and diurnal period. Data are mean ± SEM.

computer-based neuron tracing system (Neurolucida 7, MicroBrightField, Colch-ester, VT, USA) with the experimenter blind to the condition; the morphologyof the basilar arbors was quantified using the Neurolucida Explorer (Version 7,MicroBrightField) as previously described [3,8,33]. The number of neurons thatwere reconstructed (5–6 neurons per animal) is depicted in Table 1. Accordingto an established morphometric method [3,8,33], we evaluated dendritic lengthand the degree of dendritic complexity using Sholl analysis [41]. This methodestimates the amount and distribution of dendritic material at defined distancesfrom the cell body using a virtual overlay of concentric rings centered on thesoma of the neuron (Fig. 1B). In our analysis, we set the distance between theSholl circles to 10 �m. A Sholl analysis also allows determination of the lengthof entire dendrites and of the number of intersections. Intersections are thepoints where dendrites cross the virtual Sholl circles and reflect the complexityof the arbor (Fig. 1B). The total number of dendritic branches was also calcu-lated.

Spine quantification was performed under a 100× objective (numerical aperture0.75), giving a final magnification of 7000× on the monitor. Spines were countedon dendrites longer than 10 �m in at least six proximal (at a distance of 0–30 �mfrom the soma) and six distal (at a distance of 120–200 �m from the soma) branchsegments, which represent in most cases the terminal tips of the dendrites. To countthe spines, straight branches that provided clear resolution of spines were preferred;spine density was calculated as the number of spines/�m of dendrite/neuron. Spineslocated on proximal dendrites and spines located on distal dendrites were analyzedseparately.

2.5. Statistical analysis

2.5.1. Morphometric parametersIn every animal 5–6 neurons were reconstructed. In first step, the means of

these neurons were calculated for each respective parameter; then, these meanswere used to calculate the group values. Data on total dendritic length, total numberof intersections and branch numbers were analyzed using an unpaired two-tailedStudent’s t-test to detect the effect of the diurnal changes (data in Fig. 2), or witha two-way ANOVA (stress × diurnal period) (data in Fig. 3). The effects of stressand diurnal period on spine densities were analyzed with a three-way ANOVA(stress × diurnal period × distance from soma) (see data in Fig. 4). Tukey’s post hoc

test was used to further evaluate group differences.

2.5.2. Physiological parametersBody weight was measured on day 8 and compared with day 0. Body weights

and relative adrenal gland weights (mg/100 g of body weight) were analyzed usinga two-way ANOVA (stress × diurnal period) followed by Tukey’s post hoc test.

4 l Brain

so

3

ids1tttFaa

3

pc

3

pdPtnFvbo

3

ettiP

dn

3n

eAddit

3

s((a(t

effect of stress (F1,20 = 5.68, P < 0.05) and Tukey’s post hoc analysis

10 C. Perez-Cruz et al. / Behavioura

All data are presented as group means ± SEM. The SPSS SigmaStat 3.0 statisticaloftware package (SigmaStat 3.0.1, SPSS Inc., Chicago, IL, USA) was used for analysisf the data. Differences were considered significant at P < 0.05.

. Results

We analyzed the morphology of pyramidal neurons locatedn layer III of the infralimbic cortex. Golgi–Cox stained neuronsisplayed an elaborate dendritic tree with clearly distinguishablepines. Using the Sholl analysis, with concentric circles spaced0 �m apart, we measured the total length of the dendrites, ando evaluate the complexity of the dendritic trees we determinedhe number of dendritic branches and the number of intersections,he points where dendrites cross the virtual Sholl circles (Fig. 1B).urthermore, we counted the number of spines in the proximalnd distal parts of the basilar dendrites. The results of this studyre based on the digital reconstruction of 142 neurons (Table 1).

.1. Diurnal changes of the dendritic tree in control animals

Pronounced diurnal variations were observed in several mor-hometric parameters of pyramidal neurons in layer III of the ILortex.

.1.1. Basilar dendritesTotal dendritic length was longer during the active period (dark

hase), as revealed by the Student’s t-test showing a significantifference between the active and the resting period (t10 = 4.16,< 0.01; Fig. 2A). Similarly, the number of intersections with the vir-

ual Sholl circles (t10 = 12.04, P < 0.01; Fig. 2B) as well as the branchumbers were higher during the active period (t10 = 3.14, P = 0.01;ig. 2C). Spine density on basilar dendrites also showed a diurnalariation. Spine density was always higher during the active periodoth in the proximal (t10 = 5.68, P < 0.01) as well as in the distal partf the dendrites (t10 = 3.49, P < 0.01) (Fig. 2D).

.1.2. Apical dendritesThe morphology of the apical dendritic tree showed only a mod-

st activity dependent variation. A significant difference betweenhe dark–light periods was observed in the number of intersec-ions (t10 = 10.09, P < 0.01; Fig. 2B), and there was also a differencen spine density, but only on the proximal dendrites (t10 = 10.03,< 0.01; Fig. 2E).

These data demonstrate that in control unchallenged animals,endritic length and complexity vary significantly during the diur-al cycle.

.2. The effect of stress on dendritic morphology of pyramidaleurons in the infralimbic cortex

In this experiment, we had a set of animals that were stressedither during their active or during their resting period (Fig. 1A).nalysis of dendritic morphology revealed that restraining animalsuring their active period had much more pronounced effects onendritic morphology compared to restraint stress during the rest-

ng period, and that stress changed primarily the morphology ofhe basilar dendrites (Fig. 3).

.2.1. Basilar dendritesTotal dendritic length was reduced in animals subjected to

tress during their active period (Fig. 3A). Two-way ANOVA

stress × diurnal period) revealed a significant main effect of stressF1,20 = 5.89, P < 0.05) and of diurnal period (F1,20 = 12.85, P < 0.01)s well as a significant interaction between the two factorsF1,20 = 6.35, P < 0.05). Further analysis with the Tukey’s post hocest showed a significant stress effect in the active period (q = 4.95,

Research 205 (2009) 406–413

P < 0.01) as indicated by � in Fig. 3A. A similar stress effect wasdetected when we analyzed the number of intersections with theSholl circles (Fig. 3B). Two-way ANOVA (stress × diurnal period)revealed a significant main effect of stress (F1,20 = 6.01, P < 0.05) andof diurnal period (F1,20 = 216.07, P < 0.01) as well as a significantinteraction between the two factors (F1,20 = 6.88, P < 0.05). Furtheranalysis with the Tukey’s post hoc test showed a significant stresseffect in the active period (q = 5.07, P < 0.01) as indicated by � inFig. 3B. Although stress also reduced branch numbers of the basilardendrites this effect was statistically not significant (Fig. 3C)

3.2.2. Apical dendritesStress had no significant effect on the morphology of the api-

cal dendritic tree of layer III pyramidal neurons in the infralimbiccortex.

3.3. The effect of stress on spine densities of pyramidal neurons inthe infralimbic cortex

3.3.1. Spine density on basilar dendritesStress had no effect on spine density of basilar den-

drites (Fig. 4A). Statistical analysis with three-way ANOVA(stress × diurnal period × distance from soma) revealed significantmain effects of diurnal period (F1,40 = 45.98, P < 0.01) and distancefrom soma (F1,40 = 305.26, P < 0.01), meaning that spines werealways more numerous during the active period and that spinedensities were always higher in the distal parts of the dendrites.

3.3.2. Spine density on apical dendritesStress reduced spine densities on the proximal apical den-

drites in animals that were subjected to stress during theiractive period (Fig. 4B). Statistical analysis with three-way ANOVA(stress × diurnal period × distance from soma) revealed signifi-cant main effects of stress (F1,40 = 21.35, P < 0.01), diurnal period(F1,40 = 47.34, P < 0.01) and distance from soma (F1,40 = 152.00,P < 0.01). Furthermore, a significant interaction was detectedbetween stress and distance from soma (F1,40 = 25.95, P < 0.01)meaning that stress affected spine density on the proximal den-drites. In line with that, Tukey’s post hoc test showed a significantstress effect on the proximal dendrites in the active period(q = 10.31, P < 0.05) as indicated by � in Fig. 4B. Stress had no effecton spine density of the distal apical dendrites.

3.4. Physiological effects of restraint stress

To assess the overall physiological effects of 7 days of dailyrestraint stress, body weight and adrenal weight were determinedat the end of the experiment. Stress reduced body weight in bothperiods, and two-way ANOVA (stress × diurnal period) revealeda significant main effect of stress (F1,20 = 47.04, P < 0.01). Further-more, group comparison with the Tukey’s post hoc test showeda significant difference between the control group and the stressgroup both in the rest period (q = 4.19, P < 0.05) and in the activeperiod (q = 3.95, P < 0.05) (see Fig. 5A).

As shown previously, an increased adrenal weight after stressreflects hyperactivity of the HPA axis [22]. In line with these results,we found in the present study that restraint stress during theactive phase increased relative adrenal weights (Fig. 5B). A two-way ANOVA (stress × diurnal period) revealed a significant main

showed a significant difference between the control and the stressgroup only in the active period (q = 4.37, P < 0.05). Overall, these dataimply that animals which are restrained during their active periodexperience a stronger stress than animals immobilized during theirresting period.

C. Perez-Cruz et al. / Behavioural Brain

Fig. 5. Physiological parameters of the stress response. Note that restraint stress hadaBSm

4

(tterdbpp

4p

ticsipw

stronger impact when the animals were stressed during their activity period. (A)ody weight (g). (B) Relative weight of the adrenal glands (mg/100 g body weight).tatistics: Two-way ANOVA followed by the Tukey’s post hoc test: *P < 0.05. Data areean ± SEM.

. Discussion

The present study provides the following two novel findings:1) We found profound diurnal differences in the dendritic archi-ecture of layer III pyramidal neurons in the rat infralimbic cortexhat were most pronounced in the basilar dendrites, but to a lesserxtent were also present in the apical dendritic tree. (2) Repeatedestraint stress affected dendritic architecture in a way that theiurnal changes in dendritic morphology either disappeared orecame reduced in their magnitude. The stress effects were moreronounced when the animals were stressed during their activeeriod (dark phase).

.1. Diurnal differences in the dendritic morphology of layer IIIyramidal neurons in the infralimbic cortex

In the present study, we found that during the active period ofhe day, basilar dendrites were always longer and more complex,.e., had more branches and made more intersections with the Sholl

ircles than during the resting period. Basilar dendrites also hadignificantly more spines in the active phase of the animals. Sim-lar diurnal differences exist in the apical dendrites, though lessronounced. Apical dendrites had significantly more intersectionsith the Sholl circles and had more spines on their proximal den-

Research 205 (2009) 406–413 411

drites in the active phase. To the best of our knowledge this is thefirst report on such pronounced diurnal differences in pyramidalcell morphology in the mammalian neocortex.

Such rapid and substantial fluctuation in the morphology of thedendritic tree is a truly astonishing finding, but there are previ-ous reports on similar phenomena: A rapid experience-dependentplasticity has been reported recently on pyramidal neurons of thesomatosensory cortex, where spines grow and retract within a sin-gle day [15]. Furthermore, pyramidal neurons in the hippocampi ofground squirrels and European hamsters show rapid and profoundtransformation of their apical dendrites in the course of hiberna-tion [23,32]. Dendrites are significantly shorter, less branched andhave fewer dendritic spines in the middle of the hibernation bout.Moreover, after arousal from torpor, dendrites completely restoretheir structure within 2 h. Seasonal remodeling of the hippocampaldendrites occurs repeatedly during each torpor-activity cycle.

What might be the functional significance of these diurnalchanges which are most probably driven by the afferent path-ways arriving on these neurons? The infralimbic cortex is uniquewithin the mPFC in that it is the only subarea receiving afferentsfrom the suprachiasmatic nucleus [42]. This specific input fromthe master circadian pacemaker of the brain may explain whypyramidal neurons of the ILC show a pronounced diurnal reorgani-zation in their dendritic architecture. However, one should considerthat the PFC is innervated by a variety of different neurotransmit-ter systems that all undergo diurnal changes and may modulatespine number and dendritic morphology. In this part of the cor-tex, diurnal variations have been described for transmitters suchas dopamine, noradrenalin and serotonin, with higher levels ofthese monoamines during the activity than during the rest period[10,27,38], and monoamine neurotransmitters are known to alterdendritic morphology in the mPFC [31,36,46,48]. Furthermore, therelease of neurotrophic factors like brain-derived neurotrophic fac-tor (BDNF) which is strongly associated with structural plasticity isexpressed in a diurnal manner [2,25]. In addition to neurotransmit-ters and growth factors, the glucocorticoid hormone corticosteronewhich is secreted from the adrenal cortex in a manner that followsa diurnal rhythm [4] also affects dendritic arborization of neuronsin the mPFC [47].

4.2. Stress disrupts the diurnal changes in dendritic architecture

Our study reveals that stress affects dendritic architecture insuch a way that the diurnal changes in dendritic morphology eitherdisappear or become reduced in their magnitude. It is important tonote that restraint stress, in our hands, had a more pronouncedeffect both on the dendritic arborization and on the physiologicalparameters when it was applied during the active period of theanimals. The reason, why others report on stronger stress effectson the physiological parameters when using restraint stress duringthe light period, is that, most probably, they restrain the animalsmore severely with a different apparatus.

Stress is known to have a pronounced impact on the structuraland functional plasticity of neurons in the mPFC [1,5,9,14]. There isconvincing evidence that pyramidal neurons in layers II–III of themPFC react to stress with a significant reduction in total dendriticlength of 20–35%, and with a significant decrease in branching andspine density, mainly at the distal parts of their apical dendrites[8,29,33,35]. While most of these studies report on the selectiveinvolvement of the apical dendrites in the stress response, a recent

study from our laboratory pointed out that the basilar dendritesare also affected by stress [30]. In the present study most of thetreatment effects were observed in the basilar dendrites whereasthe apical dendrites appeared to be less reactive to restraint stress.Notably, the majority of synaptic inputs to neocortical pyramidal

4 l Brain

nItpotmsuHta[mwsvensmrt[

tsta

A

t

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

12 C. Perez-Cruz et al. / Behavioura

eurons are received and processed by the basilar dendrites [21].t might be somewhat surprising that we found no stress effect onhe apical dendrites, which is in sharp contrast to the results ofrevious studies, but this might be due to the fact that we focusedn cells located exclusively in layer III of the ILC, whereas most ofhe previous studies reported on cells located in all the different

PFC subareas. In studies focusing on just the ILC, neurons fromeveral layers were selected [14,34], other animal species weresed [14,34], and different stress protocols were applied [14,34].owever, it has been suggested that the different subregions and

he different cortical layers play functionally distinct roles [14,34],nd have distinct anatomical connections with other brain regions7,11,16,42,44]. Thus, the importance of discriminating between

PFC subareas and cortical layers is increasingly acknowledgedhen analyzing, for example, the effects of stress exposure on

tructural plasticity of pyramidal cells [17,29,30]. Furthermore, aery recent study suggested that stress-induced dendritic remod-ling in the mPFC is circuit specific [40]. It was found that pyramidaleurons in the ILC projecting specifically to the entorhinal cortexhow stress-induced retraction in their apical arbor, whereas pyra-idal neurons projecting to the basolateral nucleus of the amygdala

evealed no apical dendritic remodeling with stress, suggesting thathis pathway is particularly resilient against the effects of stress40].

In summary, our findings suggest that structural plasticity inhe infralimbic cortex might be a cellular mechanism that links thetress-induced effects with disrupted biological rhythms affectinghe higher order functions of the prefrontal cortex, such as attentionnd regulation of mood.

cknowledgements

We are grateful to Simone Barsky and Cornelia Heckmann forheir excellent technical assistance.

eferences

[1] Arnsten AF. Stress signalling pathways that impair prefrontal cortex structureand function. Nat Rev Neurosci 2009;10(6):410–22.

[2] Bova R, Micheli MR, Qualadrucci P, Zucconi GG. BDNF and trkB mRNAs oscillatein rat brain during the light–dark cycle. Brain Res Mol 1998;57(2):321–4.

[3] Brown SM, Henning S, Wellman CL. Mild, short-term stress altersdendritic morphology in rat medial prefrontal cortex. Cereb Cortex2005;15(11):1714–22.

[4] Buckley TM, Schatzberg AF. On the interactions of the hypothalamic–pituitary–adrenal (HPA) axis and sleep: normal HPA axis activity andcircadian rhythm, exemplary sleep disorders. J Clin Endocrinol Metab2005;90(5):3106–14.

[5] Cerqueira JJ, Mailliet F, Almeida OF, Jay TM, Sousa N. The prefrontal cor-tex as a key target of the maladaptive response to stress. J Neurosci2007;27(11):2781–7.

[6] Cho K, Ennaceur A, Cole JC, Suh CK. Chronic jet lag produces cognitive deficits.J Neurosci 2000;20(6):RC66.

[7] Conde F, Maire-Lepoivre E, Audinat E, Crepel F. Afferent connections of themedial frontal cortex of the rat. II. Cortical and subcortical afferents. J CompNeurol 1995;352(4):567–93.

[8] Cook SC, Wellman CL. Chronic stress alters dendritic morphology in rat medialprefrontal cortex. J Neurobiol 2004;60(2):236–48.

[9] Czeh B, Perez-Cruz C, Fuchs E, Flugge G. Chronic stress-induced cellular changesin the medial prefrontal cortex and their potential clinical implications: doeshemisphere location matter? Behav Brain Res 2008;190(1):1–13.

10] Feenstra MG, Botterblom MH, Mastenbroek S. Dopamine and noradrenalineefflux in the prefrontal cortex in the light and dark period: effects of nov-elty and handling and comparison to the nucleus accumbens. Neuroscience2000;100(4):741–8.

11] Gabbott PL, Warner TA, Jays PR, Salway P, Busby SJ. Prefrontal cortex in the rat:projections to subcortical autonomic, motor, and limbic centers. J Comp Neurol2005;492(2):145–77.

12] Gibb R, Kolb B. A method for vibratome sectioning of Golgi–Cox stained wholerat brain. J Neurosci Methods 1998;79(1):1–4.

13] Herman JP, Figueiredo H, Mueller NK, Ulrich-Lai Y, Ostrander MM, Choi DC, etal. Central mechanisms of stress integration: hierarchical circuitry controllinghypothalamo-pituitary-adrenocortical responsiveness. Front Neuroendocrinol2003;24(3):151–80.

[

[

Research 205 (2009) 406–413

14] Holmes A, Wellman CL. Stress-induced prefrontal reorganization and exec-utive dysfunction in rodents. Neurosci Biobehav Rev 2009;33(6):773–83.

15] Holtmaat AJ, Trachtenberg JT, Wilbrecht L, Shepherd GM, Zhang X, Knott GW,et al. Transient and persistent dendritic spines in the neocortex in vivo. Neuron2005;45(2):279–91.

16] Hoover WB, Vertes RP. Anatomical analysis of afferent projections to the medialprefrontal cortex in the rat. Brain Struct Funct 2007;212(2):149–79.

17] Izquierdo A, Wellman CL, Holmes A. Brief uncontrollable stress causes den-dritic retraction in infralimbic cortex and resistance to fear extinction in mice.J Neurosci 2006;26(21):5733–8.

18] Kahn-Greene ET, Killgore DB, Kamimori GH, Balkin TJ, Killgore WD. The effectsof sleep deprivation on symptoms of psychopathology in healthy adults. SleepMed 2007;8(3):215–21.

19] Kalsbeek A, Palm IF, La Fleur SE, Scheer FA, Perreau-Lenz S, Ruiter M,et al. SCN outputs and the hypothalamic balance of life. J Biol Rhythms2006;21(6):458–69.

20] Krettek JE, Price JL. The cortical projections of the mediodorsal nucleusand adjacent thalamic nuclei in the rat. J Comp Neurol 1977;171(2):157–91.

21] Larkman AU. Dendritic morphology of pyramidal neurones of the visual cortexof the rat: III. Spine distributions. J Comp Neurol 1991;306(2):332–43.

22] Magarinos AM, McEwen BS. Stress-induced atrophy of apical dendritesof hippocampal CA3c neurons: comparison of stressors. Neuroscience1995;69(1):83–8.

23] Magarinos AM, McEwen BS, Saboureau M, Pevet P. Rapid and reversible changesin intrahippocampal connectivity during the course of hibernation in Europeanhamsters. Proc Natl Acad Sci USA 2006;103(49):18775–80.

24] Matchock RL, Mordkoff JT. Chronotype and time-of-day influences on thealerting, orienting, and executive components of attention. Exp Brain Res2009;192(2):189–98.

25] McAllister AK, Katz LC, Lo DC. Neurotrophin regulation of cortical dendriticgrowth requires activity. Neuron 1996;17(6):1057–64.

26] McLaughlin KJ, Gomez JL, Baran SE, Conrad CD. The effects of chronic stresson hippocampal morphology and function: an evaluation of chronic restraintparadigms. Brain Res 2007;1161:56–64.

27] Nakayama K. Diurnal rhythm in extracellular levels of 5-hydroxyindoleaceticacid in the medial prefrontal cortex of freely moving rats: an in vivo micro-dialysis study. Prog Neuropsychopharmacol Biol Psychiatry 2002;26(7-8):1383–8.

28] Owens DS, Macdonald I, Tucker P, Sytnik N, Totterdell P, Minors D, et al.Diurnal variations in the mood and performance of highly practised youngwomen living under strictly controlled conditions. Br J Psychol 2000;91(Pt 1):41–60.

29] Perez-Cruz C, Muller-Keuker JI, Heilbronner U, Fuchs E, Flugge G. Morphol-ogy of pyramidal neurons in the rat prefrontal cortex: lateralized dendriticremodeling by chronic stress. Neural Plast 2007;2007:46276.

30] Perez-Cruz C, Simon M, Czeh B, Flugge G, Fuchs E. Hemispheric differencesin basilar dendrites and spines of pyramidal neurons in the rat prelim-bic cortex: activity- and stress-induced changes. Eur J Neurosci 2009;29(4):738–47.

31] Perez-Vega MI, Feria-Velasco A, Gonzalez-Burgos I. Prefrontocortical sero-tonin depletion results in plastic changes of prefrontocortical pyramidalneurons, underlying a greater efficiency of short-term memory. Brain Res Bull2000;53(3):291–300.

32] Popov VI, Bocharova LS, Bragin AG. Repeated changes of dendritic morphologyin the hippocampus of ground squirrels in the course of hibernation. Neuro-science 1992;48(1):45–51.

33] Radley JJ, Sisti HM, Hao J, Rocher AB, McCall T, Hof PR, et al. Chronic behav-ioral stress induces apical dendritic reorganization in pyramidal neurons ofthe medial prefrontal cortex. Neuroscience 2004;125(1):1–6.

34] Radley JJ, Arias CM, Sawchenko PE. Regional differentiation of the medial pre-frontal cortex in regulating adaptive responses to acute emotional stress. JNeurosci 2006;26(50):12967–76.

35] Radley JJ, Rocher AB, Miller M, Janssen WG, Liston C, Hof PR, et al. Repeatedstress induces dendritic spine loss in the rat medial prefrontal cortex. CerebCortex 2006;16(3):313–20.

36] Robinson TE, Kolb B. Structural plasticity associated with exposure to drugs ofabuse. Neuropharmacology 2004;47(Suppl. 1):33–46.

37] Rouch I, Wild P, Ansiau D, Marquie JC. Shiftwork experience, age and cognitiveperformance. Ergonomics 2005;48(10):1282–93.

38] Rueter LE, Jacobs BL. Changes in forebrain serotonin at the light–dark transition:correlation with behaviour. Neuroreport 1996;7(5):1107–11.

39] Schmidt C, Collette F, Cajochen C, Peigneux P. A time to think: circadian rhythmsin human cognition. Cogn Neuropsychol 2007;24(7):755–89.

40] Shansky RM, Hamo C, Hof PR, McEwen BS, Morrison JH. Stress-induced den-dritic remodeling in the prefrontal cortex is circuit specific. Cereb Cortex 2009,doi:10.1093/cercor/bhp003.

41] Sholl DA. Dendritic organization in the neurons of the visual and motor corticesof the cat. J Anat 1953;87(4):387–406.

42] Sylvester CM, Krout KE, Loewy AD. Suprachiasmatic nucleus projection to themedial prefrontal cortex: a viral transneuronal tracing study. Neuroscience2002;114(4):1071–80.

43] Tavares RF, Correa FM, Resstel LB. Opposite role of infralimbic and prelimbiccortex in the tachycardiac response evoked by acute restraint stress in rats. JNeurosci Res 2009;87(11):2601–7.

l Brain

[

[

[

[prefrontal cortex after chronic corticosterone administration. J Neurobiol

C. Perez-Cruz et al. / Behavioura

44] Vertes RP. Differential projections of the infralimbic and prelimbic cortex inthe rat. Synapse 2004;51(1):32–58.

45] Vertes RP. Interactions among the medial prefrontal cortex, hippocampus andmidline thalamus in emotional and cognitive processing in the rat. Neuro-science 2006;142(1):1–20.

46] Wang HD, Deutch AY. Dopamine depletion of the prefrontal cortex inducesdendritic spine loss: reversal by atypical antipsychotic drug treatment. Neu-ropsychopharmacology 2008;33(6):1276–86.

[

Research 205 (2009) 406–413 413

47] Wellman CL. Dendritic reorganization in pyramidal neurons in medial

2001;49(3):245–53.48] Wellman CL, Izquierdo A, Garrett JE, Martin KP, Carroll J, Millstein R,

et al. Impaired stress-coping and fear extinction and abnormal corticol-imbic morphology in serotonin transporter knock-out mice. J Neurosci2007;27(3):684–91.