www.e l sev i e r . com / l oca te / eu roneu ro

European Neuropsychopharmacology (2012) 22, 441–451

Imaging trait anxiety in high anxiety F344 rats: Focuson the dorsomedial prefrontal cortexEric P. Prinssen⁎, Laurent B. Nicolas, Steffen Klein,Christophe Grundschober, Cristina Lopez-Lopez, Melanie S. Keßler,Andreas Bruns, Markus von Kienlin, Joseph G. Wettstein,Jean-Luc Moreau 1, Celine Risterucci 1

CNS Research, F. Hoffmann-La Roche Ltd., CH-4070 Basel, Switzerland

Received 9 March 2011; received in revised form 18 September 2011; accepted 5 November 2011

⁎ Corresponding author. Tel.: +41 61E-mail address: eric.prinssen@roch

1 These authors contributed equally.

0924-977X/$ - see front matter © 2011doi:10.1016/j.euroneuro.2011.11.001

KEYWORDSRats;Social behavior;fMRI;Lesion;c-Fos

Abstract

Functional magnetic resonance imaging (fMRI) has become an important method in clinical psy-chiatry research whereas there are still only few comparable preclinical investigations. Herein,we report that fMRI in rats can provide key information regarding brain areas underlying anxietybehavior. Perfusion as surrogate for neuronal activity was measured by means of arterial spinlabeling-based fMRI in various brain areas of high anxiety F344 rats and control Sprague–Dawley

rats. In one of these areas, the dorsomedial prefrontal cortex (dmPFC), c-Fos labeling was com-pared between these two strains with immunolabeling. The effects of a neurotoxic ibotenic acidlesion of the dmPFC in F344 rats were examined in a social approach–avoidance anxiety proce-dure and fMRI. Regional brain activity of high anxiety F344 rats was different in selective corticaland subcortical areas as compared to that of low anxiety Sprague–Dawley rats; the largest dif-ference (i.e. hyperactivity) was measured in the dmPFC. Independently, c-Fos labeling con-firmed that F344 rats show increased dmPFC activity. The functional role was confirmed byneurotoxic lesion of the dmPFC that reversed the high anxiety-like behavior and partially nor-malized the brain activity pattern of F344 rats. The current findings may have translationalvalue as increased activity is reported in an equivalent cortical area in patients with social anx-iety, suggesting that pharmacological or functional inhibition of activity in this brain area shouldbe explored to alleviate social anxiety in patients.© 2011 Elsevier B.V. and ECNP. All rights reserved.

68 87056.e.com (E.P. Prinssen).

Elsevier B.V. and ECNP. All ri

ght

1. Introduction

Multiple epidemiological studies have shown that geneticsplay an important role in anxiety disorders in humans(Hettema et al., 2001; Leonardo and Hen, 2006). Using

s reserved.

442 E.P. Prinssen et al.

inbred mice, this influence was confirmed by identifying sev-eral genetic factors that are associated with emotion-related behaviors in various procedures that model anxiety(Clement et al., 2002; Muigg et al., 2009; O'Mahony et al.,2010). In rats, genetic approaches for studies on anxietyhave been limited and only few studies have compared out-bred or inbred rat strains. One of the more robust findings isthat inbred Fischer 344 (F344) rats display an anxious pheno-type in several anxiety-based paradigms, such as the elevat-ed plus maze, the black–white box and the social interactiontest (Bert et al., 2002; Berton et al., 1997; Ramos et al.,1997; Rex et al., 1999), and may be a valid starting pointto identify the neuronal pathways underlying high anxietybehavior.

Brain regions of greatest interest in the pathology of anx-iety disorders, including social anxiety, are the amygdala,the anterior cingulate cortex, the ventromedial prefrontalcortex (vmPFC), the orbitofrontal cortex, the insula andthe temporal areas (Amir et al., 2005; Canteras et al.,2010; Etkin and Wager, 2007; Goldin et al., 2009b; Phan etal., 2005; Shah et al., 2009; Stein et al., 2007). Multiple pre-clinical studies using lesions or pharmacological challengeswere able to show the involvement of some of these areasin anxiety (Navarro et al., 2004; Rudebeck et al., 2007;Sajdyk et al., 1999; Sullivan and Gratton, 2002) but the find-ings are heterogeneous. A potentially valuable approach tobetter investigate the role of brain areas in high anxiety an-imals is the use of functional magnetic resonance imaging(fMRI) technology. It allows an indirect measurement ofbrain activity in many different areas at the same time in anon-invasive manner, with translational value. In contrastto the clinical setting, only few preclinical fMRI studieshave examined the neuronal pathways that may be involvedin anxiety-like behavior (Ferris et al., 2008; Kalisch et al.,2004; Nephew et al., 2009).

The main goals in the present study were: 1) to comparethe behavior of F344 rats with that of a standard rat strain(Sprague–Dawley, SD) in the social approach–avoidance(SAA) test (Nicolas and Prinssen, 2006), 2) to assess regionalbrain activity using fMRI in F344 and SD rats. Various brainareas that are associated with anxiety- and stress-relatedbehavioral or autonomic functions were examined as an un-biased approach with the assumption that the brain areashowing the largest differences between the strains [i.e.the dmPFC, thought to correspond to the anterior cingulatecortex (ACC) in humans (Seamans et al., 2008; Uylingset al., 2003)] would play a key role in anxiety, 3) to confirmthe hyperactivity of the dmPFC in F344 rats with immunola-beling of c-Fos protein, a marker for transcriptional action,and 4) to examine the effects of a neurotoxic lesion of thedmPFC in F344 rats in both the SAA test and fMRI.

2. Experimental procedures

2.1. Animals

All animal procedures were conducted in strict adherence to theSwiss federal regulations on animal protection and to the rules ofthe Association for Assessment and Accreditation of LaboratoryAnimal Care International (AAALAC), and with the explicit approvalof the local veterinary authority. Male F344 and SD rats (RCC,Switzerland; Iffa Credo, France, respectively), weighing 250 g at the

beginning of the experiment, were housed individually. According toan established protocol for the SAA test (Nicolas and Prinssen, 2006),stimulus SD rats (Iffa Credo, France) weighing 450–500 g were keptisolated for several weeks to months and were used in several differ-ent experiments (total of eight experiments). All animals were housedunder standard maintenance conditions (12:12 h light/dark cycle withlights on at 6 a.m., 21–23 °C, 55–65% relative humidity) and providedwith food and water ad libitum.

2.2. Social approach avoidance test

The experimental unit was a box divided in two sections, the non-social (20×40×30 cm) and the social (39×40×30 cm) compart-ments, connected by a sliding door. The social compartment con-tained a sub-chamber (14.5×40×30 cm) delimited by a perforatedtransparent wall confining the stimulus rat. For analysis the twocompartments were virtually divided in zones (hidden and protectedfor non-social; distal and proximal for social). The test started by in-troducing a test rat into the non-social compartment for a 3 minutehabituation period (the stimulus rat was introduced in the sub-chamber just before). At the end of the habituation period, the slid-ing door was opened allowing the test rat to freely move betweenthe two compartments for 10 min. A tracking system (Ethovisionvideo tracking, Noldus, Netherlands) measured the time spent indifferent compartment/zones and the distance traveled during ha-bituation as a measure of locomotor activity. For more details, seeNicolas and Prinssen (2006). Group sizes were between 8 and 12(see figure legends for details).

2.3. fMRI studies

On the day of MRI investigation of regional brain perfusion, animalswere anesthetized with an induction level of 4% and a maintenancelevel of 2–2.5% isoflurane (Abbott, Switzerland) in a mixture of ox-ygen (0.2 l/min) and air (1.0 l/min) administered via a face mask.Before starting the fMRI measurement, the maintenance isofluranelevel was adjusted to standardize the breathing rate to 60 breathsper minute (corresponding to 2–2.5% isoflurane). Rats were posi-tioned in a Plexiglas cradle in the magnet and their heads immobi-lized in a stereotaxic holder. Body temperature was maintained at37 °C using a feedback-regulated electric heating blanket. Breath-ing rate and concentrations of inhaled and exhaled oxygen andCO2 were continuously monitored on a PowerLab data acquisitionsystem (ADInstruments, Germany). All MRI studies were conductedon a Bruker Biospec 4.7T/40 cm instrument (Bruker Biospin,Germany), equipped with a 12 cm actively shielded gradient set. A7 cm diameter birdcage coil was used for radio-frequency excita-tion, and an actively decoupled surface coil was positioned on thehead of the animal for signal reception. For all images, the field-of-view was 4 cm and the slice thickness 1 mm.

A set of scout images (T2-weighted: repetition time (TR)=2.7 s,echo spacing (TE)=10.3 ms, RARE-factor 8, 128×64 matrix, 4 aver-ages) in axial orientation was acquired in each animal, in order to lo-cate the most rostral extension of the corpus callosum, which servedas landmark for the subsequent study. Eight coronal image planeswhich cover all areas of interest at +2.3, +1.0, −0.3, −1.6, −2.9,−5.3, −7.8 and −10.0 mm compared to bregma were selected anda set of anatomical images was obtained from these locations (T2-weighted: TR/TE=1.8 s/18.0 ms, RARE-factor 8, 256×256 matrix, 4averages). T1-maps required to quantify perfusion were also collect-ed using an inversion-recovery-snapshot-FLASH sequence with eightinversion times (TR/TE=7.5 s/1.7 ms, 128×64 matrix, 8 averages)(Haase et al., 1986). Perfusion imaging was conducted based onthe continuous arterial spin labeling (CASL) method (Alsop andDetre, 1996; Williams et al., 1992) with a single slice RARE readoutmodule (TR/TE=3 s/5.5 ms, RARE-factor=32, 128×64 matrix, 2 av-erages, 2.5 s labeling pulse). Each perfusion image plane took

443Imaging trait anxiety in high anxiety F344 rats: Focus on the dorsomedial prefrontal cortex

approximately 30 s, a complete set with the eight image planes foreach time point thus was acquired in 4 min. Three sets of imagesof basal perfusion, corresponding to a 12-min period, were recordedto assess within-subject reproducibility, and were averaged beforeentering the subsequent analyses to improve the signal-to-noiseratio. Group sizes were between 7 and 12 (see figure legends fordetails).

2.4. fMRI-data analysis

Magnetic resonance images were processed and analyzed using soft-ware developed in-house with the programming environment Inter-active Data Language (IDL version 6.4, Research Systems Inc., CO).For quantification of regional perfusion changes, the anatomicalbrain images of all animals were segmented into 32 specific areas(e.g. dorsomedial prefrontal cortex, dorsal striatum, nucleusaccumbens) using a template that was transformed by in-planetranslation and rotation to match the exact position of the brain ineach individual. T1-maps were calculated on a pixel-by-pixel basisby applying a three-parameter exponential fitting algorithm(Deichmann et al., 1999). Absolute perfusion was quantified usingthe time-series of CASL-images and the quantitative T1-map as de-scribed elsewhere (Alsop and Detre, 1996). In order to cope with sys-temic hemodynamic changes, the mean perfusion was calculated foreach image plane, and subsequently the perfusion was normalizedto this mean and expressed as a percentage. Finally for graphicalrepresentation, the perfusion images of all animals within a specificgroup were co-registered and averaged, in order to obtain represen-tative group images.

2.5. c-Fos immunohistochemistry

Animals (F344 n=4, SD n=5) were tested in the SAA test (results notshown) and placed back in their home cage for 60 min. Then, theywere deeply anesthetized and intracardial perfusion was performedwith saline followed by 4% paraformaldehyde in 0.1 M phosphatebuffer (PB, pH 7.4). Thereafter the brains were removed and fixedfor additional 24 h at 4 °C. Fifty-micrometer coronal sections con-taining the Cg1 (and part of M2 to be consistent with the fMRIstudy) and Cg2 regions from the dorsomedial prefrontal cortex(+1.9 mm to +1.5 mm from bregma), were dissected (Paxinos andWatson, 1998). Eight sections/animal were used for c-Fos stainingand quantification.

The immunocytochemical staining was performed on free-floating sections. The sections were washed 3 times in PB (pH7.4), blocked for 30 min in a solution with 10% methanol and 3% hy-drogen peroxide, washed twice in PB, and incubated overnightwith a rabbit anti-c-Fos antibody (Santa Cruz Biotechnology, Inc)1:1000 in PB 0.1 M with 0.3% Triton X-100 and 0.3% bovine serum al-bumin. The sections were then washed three times in PB with 0.3%Triton X-100 (Sigma), incubated with a secondary biotinylatedmouse anti-Rabbit IgG (Santa Cruz Biotechnology, Inc) 1:1000 inPB 0.1 M with 0.3% Triton X-100 and 0.3% bovine serum albuminfor 2 h at room temperature, washed 3 times in PB/Triton, incubat-ed with avidin–biotin complex (1:1000 in PB/Triton; Vector ABCkit) for 1 h at room temperature, and washed 3 times in PB. Theimmunostaining reaction was developed using the diaminobenzi-dine as the chromogen and the staining reaction was stopped by3 washes with PB. The reaction resulted in a dark-brown stain with-in the nuclei of c-Fos immunoreactive neurons. Tissue sectionswere mounted on gelatin-coated slides, dehydrated for observa-tion and counting was performed under bright-field microscopy.Semiquantitative analysis of c-Fos activation was performedas described previously (Lopez-Lopez et al., 2004), according tothe point counting method (Wiebel, 1979). c-Fos activation isexpressed as percent of brain surface covered with activated neu-rons, determined in the following manner: a 30-point grid covering

the entire surface of the microscopic field represents the 100% ofthe surface area. For each brain section, the number of c-Fosimmunopositive nuclei falling under a point intersection in agiven brain is counted. Four fields per section were counted andthe mean of positive c-Fos point intersections was calculated foreach rat. This number then is divided by 30 (the total points inthe grid) and multiplied by 100, giving the percent of brain surfacecovered with activated neurons. The counting of c-Fos activationwas performed in a blind manner.

2.6. Neurotoxic lesion of dmPFC

Rats were initially anesthetized with 4% isoflurane (Abbott, France)and placed in a stereotaxic frame (Kopf Instruments, CA). Anesthe-sia was maintained with 2–2.5% Isoflurane throughout the wholesurgery. Ibotenate (10 μg/μl, Sigma, Germany) in 0.2 μl artificial ce-rebrospinal fluid (ACF, Harvard Apparatus, UK) or 0.2 μl ACF onlywas bilaterally infused into the dmPFC via a 0.5 mm diameter can-nula at the following coordinates AP: +2.2 and +2.6 mm L:+/−0.6 mm V: +2.8 mm from bregma with a perfusion rate of 0.1 μl/min. After completion of injection the cannulas were left in placefor additional 3 min in order to prevent the ibotenate from spread-ing along the needle track. After suturing the skin with surgicalsilk, all animals were injected once with 1 mg/kg diazepam inorder to prevent seizures, which were regularly observed in a pilotstudy. 0.025 mg/kg Buprenorphin (Temgesic, Essex Chemie AG,Switzerland) was used as analgesic treatment for three subsequentdays. Sham animals were treated identically except that the artifi-cial cerebrospinal fluid contained no ibotenate. After approximately10 days of recovery, rats were tested in the SAA test and two dayslater in the fMRI study.

Afterwards, rats were sacrificed by CO2 anesthesia, the brainswere removed and immersed in 4% formaldehyde, additionally con-taining 30% sucrose. Fifty micrometer coronal sections were cut on aLeica cryostat and Nissl-stained with 1% cresylviolet to determinethe extent of the lesions. Only rats with selective bilateral lesionsof the dmPFC were included for further analysis in the SAA testand the fMRI study.

2.7. Statistics

Analysis of behavioral and c-Fos findings comprised group compari-sons of F344 versus SD rats and of lesioned versus sham-lesionedrats, respectively, based on unpaired, two-tailed Student's t-tests(testing residuals for normality supported parametric statisticalanalysis). fMRI-based regional blood perfusion was compared be-tween F344 and SD rats using Welch's t-test (to account for varianceinhomogeneities) separately for each brain area. Given the explor-atory character of the present study with the goal of identifyingthe area(s) being potentially most relevant to the mechanisms inquestion, we preferred accepting reasonably small numbers offalse positives over large numbers of false negatives. Hence, insteadof using the overly conservative and correspondingly low-poweredapproaches for controlling the familywise error rate (e.g.,Bonferroni, Šidàk, Holm, Hochberg) or multivariate techniques(which would have required an unacceptably large number of sub-jects per group), we applied the standard significance criterion ofpb0.05 with no correction for multiple testing, and estimated thefalse discovery rate (FDR, Kessler et al., 2011; Storey andTibshirani, 2003). Areas meeting this criterion are reported as find-ings of primary interest, being aware that they are expected to in-clude a defined small number of false positives (nFP). In thepresent study, this number is reported along with its 90% confidencebounds nFP,5% and nFP,95% (estimated via bootstrapping, Storey,2002).

Figure 1 Anxiety-related behavior of F344 versus SD rats.(A) The SAA test-box is divided in a non-social (delineated inorange) and a social (delineated in blue) compartment. Foranalysis the two compartments are virtually divided into zones(hidden and protected; distal and proximal). The social com-partment contains a sub-chamber delimited by a perforatedtransparent wall confining the stimulus rat. (B) F344 rats(n=10) spent less time in the proximal zone of the social com-partment as compared to SD rats (n=8), indicating a higherlevel of anxiety-like behavior. * Pb0.05. Data are presented inmeans+s.e.m.

444 E.P. Prinssen et al.

3. Results

3.1. Anxiety-related behavior of F344 and SD rats

It has been shown previously that time spent in the proximal andhidden zones is the key parameter in the SAA test (Fig. 1A) todetermine the anxiety state of rats (Nicolas and Prinssen,2006). Comparing the behavior of F344 versus SD rats in theSAA test, time spent in the proximal zone of the social compart-ment was lower in F344 rats (Fig. 1B), indicating increased anx-iety. The time spent in other zones was not statisticallydifferent between the two strains. The two strains showed sim-ilar baseline locomotor activity during the first 3 min in the test(mean±s.e.m.: 916±16 cm and 984±44 cm in SD and F344 rats,respectively; P>0.1).

3.2. Basal brain perfusion in F344 and SD rats

To assess within-subject reproducibility, the 3 successively ac-quired perfusion images were used to estimate, for each brainarea, the intra-class correlation (ICC) for the average of the 3 ac-quisitions, based on a random-effects model (i.e., ICC(2,3) fromShrout and Fleiss, 1979). ICC values of normalized perfusion inthe various brain areas were distributed around an overall meanof ICCNrmPrf =0.91 (ordinary standard-deviation interval: 0.86–0.94, which is asymmetric about the mean due to the skewed

distribution of correlation values). In absolute terms, the esti-mated standard deviations of within- and between-subjectvariability (expressed as percentages of global mean perfusion)were σwithin=2.6±0.1% and σbetween=7.3±0.3%, respectively.These numbers confirm that measurement noise was only a mi-nor source of variability.

To identify neurobiological correlates of different anxietystates, basal brain perfusion of F344 and SD rats was com-pared using fMRI. The brain perfusion profile of F344 and SDrats is shown in Fig. 2 presenting the 32 areas of interest.Compared to SD rats, F344 display a higher normalized perfu-sion in (areas are listed in order of magnitude; Table 1) thedmPFC, entorhinal cortex, motor cortex, vmPFC, dorsal stria-tum, and bed nucleus of stria terminalis. In contrast, the ven-tral tegmental area, piriform cortex, orbitofrontal cortex,superior colliculus, inferior colliculus, nucleus accumbens,ventral pallidum, sensory cortex, and insula show a signifi-cantly lower normalized perfusion in F344 rats compared toSD. According to the FDR analysis, 1 or 2 out of these 15 areasare expected to be false positives (cf. Table 1). No significanteffects were measured in brain areas such as the raphenuclei, hypothalamus, amygdala, locus coeruleus, septum,periaqueductal gray, hippocampus or various cortical regions(Table 1).

3.3. c-Fos activity in the dmPFC of F344 and SD rats

To confirm that the hyperperfusion in the dmPFC of F344 rats isrelated to higher neuronal activity, c-Fos activity was measuredin the cingulate cortices 1 and 2 forming together the dmPFC(Fig. 3A, B) in F344 and SD rats. As expected, F344 rats showedhigher c-Fos expression than SD rats (Fig. 3C) confirming thatF344 rats have higher dmPFC activity.

3.4. Effect of dmPFC lesions on anxiety-relatedbehavior of F344 rats

To examine the role of dmPFC hyperactivity in the highanxiety-like behavior of F344 rats, a neurotoxic lesion was in-duced in this area (Fig. 4A) and the rats were examined inthe SAA test (Fig. 4B) after a recovery period. The sham-treated animals showed a more pronounced anxiety-like be-havior as compared to the non-treated animals (i.e. less timein the social compartment and more time in the hidden zone;compare Figs. 4C and 1B), suggesting that the surgery proce-dure further amplified the baseline anxiety level. The lesionedanimals, as compared to sham-treated animals, showed a sig-nificant increase in time spent in the social compartment. Thisincrease was due to significant increases in time spent in bothdistal and proximal zones (Fig. 4C). Lesioned rats also spentsignificantly less time in the hidden zone of the non-socialcompartment (Fig. 4C). All together, the dmPFC lesion induceda remarkable shift in behavior with less avoidance and moreapproach to the stimulus rat, indicating a decrease in anxiety.These marked changes are unlikely to be explained by changesin locomotor activity. I.e., the baseline locomotor activityduring the first 3 min was not significantly different betweenthe two groups even though it tended (P=0.054) to be higherin the lesioned animals as compared to sham controls(871±26.6 cm vs. 771±41.6 cm).

Figure 2 Basal brain perfusion in 32 brain areas of F344 and SD rats. Frontal anatomical brain images (template) with correspondinganteriorities from bregma, according to the Paxinos and Watson (1998) rat brain atlas, show the delineated areas of interest(1 dmPFC, 2 entorhinal cortex, 3 ventral tegmental area, 4 piriform cortex, 5 motor cortex, 6 orbitofrontal cortex, 7 vmPFC, 8 supe-rior colliculus, 9 dorsal striatum, 10 inferior colliculus, 11 nucleus accumbens, 12 ventral pallidum, 13 bed ncl. of the stria terminalis,14 sensory cortex (S2), 15 insula, 16 median raphe, 17 paraventricular ncl. of the hypothalamus, 18 dorsal peduncular, 19 substantianigra, 20 lateral hypothalamus, 21 amygdala, 22 locus coeruleus, 23 dorsal hippocampus, 24 septum, 25 dorsal raphe, 26 periaque-ductal gray, 27 perirhinal cortex, 28 median hypothalamus, 29 sensory cortex (S1), 30 ectorhinal cortex, 31 ventral hippocampus,32 thalamus). Perfusion normalized to the mean of each image plane (expressed as percentage) in each rat was calculated andthen averaged for rats belonging to the different groups (F344 n=7 and SD n=7). Images right aside the templates show the differ-ences between the two strains (Δ (F344−SD)).

445Imaging trait anxiety in high anxiety F344 rats: Focus on the dorsomedial prefrontal cortex

3.5. Effect of dmPFC lesion on basal brain perfusionin F344 rats

Neurotoxic lesion of the dmPFC in F344 rats led to significantlydecreased % normalized perfusion in the target area. Additional-ly, the lesion led to significantly higher perfusion in the orbito-frontal cortex, sensory cortex, and insula as compared tosham-treated F344 rats (Table 2; 1 false positive expected).An interesting way to interpret these changes is to compare the-se to the control strain. When compared to SD rats (data fromthe experiment in Section 3.2), sham-treated F344 rats showedsignificant differences of % normalized perfusion in the dmPFC,entorhinal cortex, orbitofrontal cortex, vmPFC, superior and in-ferior colliculus, dorsal striatum, bed nucleus of the stria termi-nalis, sensory cortex, and insula (Table 2; 1 false positiveexpected) (qualitative differences are the same as comparedto the difference between unoperated F344 and SD rats eventhough the magnitude of difference is smaller in some cases;cf. Table 1). Furthermore, lesioning the dmPFC in F344 rats alle-viated the brain activity pattern differences observed betweenSD and sham-treated F344 rats in several areas including orbito-frontal cortex, sensory cortex, and insula (Table 2; 1 false posi-tive expected).

4. Discussion

Reinforced by our recent findings using fMRI for schizophreniaand anxiety-related research in rats (Kessler et al., 2011;Nordquist et al., 2008; Risterucci et al., 2005), and by positronemission tomography and fMRI studies in anxious patientsreporting dysregulations in emotion neurocircuits (Etkin andWager, 2007; Ressler and Mayberg, 2007), fMRI baseline perfu-sion was used to identify regional brain activity differences inrat strains with known differences in stress/anxiety levels(Dhabhar et al., 1993; Nicolas and Prinssen, 2006; Steineret al., 2011). This comparison showed that F344 rats had a dis-tinct neuronal fingerprint as compared to SD rats. Both in-creases and decreases in baseline brain activity wereobserved in several regions such as dmPFC and other corticalareas, nucleus accumbens, dorsal striatum, ventral tegmentalarea, colliculi and insula when comparing F344 rats to SD rats.It is quite likely that these areas are part of an emotional pro-cessing network with the dmPFC as a major component(Andreescu et al., 2011; Canteras et al., 2010; Ding et al.,2011; Qiu et al., 2011; Schneier et al., 2009). In addition tofMRI-measured hyperactivity, F344 rats showed higher neuro-nal activation in the dmPFC as determined by immunolabeling

Table 1 Basal brain activity of Fischer (F344) and Sprague–Dawley (SD) rats.

Areas showing statistically significant difference (ordered by magnitude)

Brain area Mean±s.e.m. normalized perfusion(%)

Δ of normalized perfusion (%) P-value

F344 (n=7) SD (n=7) Δ (F344−SD)

Dorsomedial PFC (1) # 25±4 −15±3 40 ↑ b0.0001Entorhinal cortex (2) 0±3 −25±3 25 ↑ 0.0001Ventral tegmental area (3) 30±6 51±6 21 ↓ 0.0227Piriform cortex (4) 1±3 22±5 21 ↓ 0.0039Motor cortex (5) 5±3 −14±4 19 ↑ 0.0022Orbitofrontal cortex (6) −8±2 11±2 19 ↓ b0.0001Ventromedial PFC (7) 2±3 −16±1 18 ↑ 0.0003Superior colliculus (8) 9±2 26±3 17 ↓ 0.0014Dorsal striatum (9) 11±1 −4±2 15 ↑ b0.0001Inferior colliculus (10) 14±2 29±2 14 ↓ 0.0007Nucleus accumbens (11) −3±2 10±2 12 ↓ 0.0007Ventral pallidum (12) −3±4 8±2 11 ↓ 0.0265Bed ncl of stria terminalis (13) −6±2 −15±3 9 ↑ 0.0224Sensory cortex (S2) (14) −12±3 −4±2 8 ↓ 0.0232Insula (15) −15±3 −7±1 8 ↓ 0.0197Expected number nFP (nFP,5%–nFP,95%)of false positives

1.4 (0–2.7)

Areas showing no statistically significant differences (ordered by magnitude)

Brain area Mean±s.e.m. normalized perfusion(%)

Δ of normalized perfusion (%) P-value

F344 (n=7) SD (n=7) Δ (F344−SD)

Median raphe (16) 19±4 34±6 15 ↓ 0.0766Paraventr. ncl of hypoth (17) −24±5 −11±5 13 ↓ 0.0898Dorsal peduncular (18) −3±3 −10±3 7 ↑ 0.1196Substantia nigra (19) 16±2 23±3 7 ↓ 0.1196Lateral hypothalamus (20) 9±3 16±4 7 ↓ 0.1621Amygdala (21) −14±2 −8±3 6 ↓ 0.1588Locus coeruleus (22) 19±2 25±5 6 ↓ 0.3092Dorsal hippocampus (23) −15±2 −9±3 6 ↓ 0.1511Septum (24) 1±2 −4±3 6 ↑ 0.1477Dorsal raphe (25) 3±5 8±5 5 ↓ 0.5481Periaqueductal gray (26) −20±2 −22±5 3 ↑ 0.6553Perirhinal cortex (27) −12±4 −15±3 3 ↑ 0.5570Median hypothalamus (28) 2±3 6±6 3 ↓ 0.5989Sensory cortex (S1) (29) −12±1 −11±3 1 ↓ 0.7316Ectorhinal cortex (30) −10±3 −9±3 1 ↓ 0.7463Ventral hippocampus (31) −8±2 −8±2 0 0.9706Thalamus (32) 14±3 14±2 0 0.9714

(#) Corresponds to area number in Fig. 2.; ↑↓ indicate positive and negative differences, respectively. P values obtained from Welch's t-test. Expected number nFP of false positives and 90% confidence bounds nFP,5%–nFP,95% obtained from FDR analysis.

446 E.P. Prinssen et al.

of c-Fos protein. Functional relevance of this cortical hyperac-tivity was confirmed by the finding that neurotoxic lesions ofthe dmPFC reversed high anxiety-like behavior of F344 ratsin the SAA test. It should be noted that the sham-treatedF344 rats showed a more pronounced anxiety-like behavioras compared to the non-treated F344 rats (i.e. less time inthe social compartment and more time in the hidden zone;compare Figs. 4C and 1B), suggesting that the surgery proce-dure further amplified the baseline anxiety level. There is ev-idence that surgical history can alter anxiety related behaviors(Zacharko et al., 1999). Possibly, due to this low baseline, the

effect of lesion in reversing the anxiety-like behavior was verymarked. Although a disinhibitory effect cannot be completelyruled out as we observed a tendency towards an increase inbaseline locomotion, this was a very small effect (~13%),which could not account for the marked increase in social ap-proach. fMRI analysis showed that, in addition to the expectedreduction in dmPFC activity, the lesion partially normalizedactivity in other brain areas such as orbitofrontal cortex, sen-sory cortex and insula, again pointing to those changes occur-ring within an emotion network with dmPFC being a keyplayer.

Figure 3 Comparison of neuronal activation in the dmPFC of F344 vs. SD rats. (A) Representative section showing area of c-Fosanalysis. c-Fos expression was counted in 4 fields per section, using 8 sections per animal, covering cingulate cortex area 1 (Cg1and part of M2 to be consistent with the fMRI study) and 2 (Cg2) (purple squares). (B) Representative images of c-Fos activity inCg1 and Cg2. (C) F344 rats (n=4) present a higher percentage of c-Fos activity compared to SD rats (n=5) in the dmPFC (sum ofCg1 and Cg2). * Pb0.05. Data are presented in means+s.e.m.

447Imaging trait anxiety in high anxiety F344 rats: Focus on the dorsomedial prefrontal cortex

Several findings on the role of the dmPFC in emotionalprocessing in rodents have been reported in the literature(Lacroix et al., 2000; Morgan and LeDoux, 1995). Reportedfindings are equivocal and difficult to compare to our study,due to methodological differences such as the anatomicalspecificity of lesions, behavioral read-outs or animal strain:1) many studies made lesions in both dmPFC and vmPFC, com-plicating the interpretation of the findings, 2) a large varietyof anxiety procedures were used (e.g. conditioned vs. uncon-ditioned anxiety) that may have different underlying mecha-nisms, 3) most of the studies were performed in ‘normal’strains and not in high anxiety strains such as F344 rats. Possi-bly most similar to our approach are the studies by Landgrafand colleagues, who reported that High Anxiety-related

Figure 4 Effects of ibotenic acid induced lesions of the dmPFC in Ffrom the Paxinos and Watson (1998) rat brain atlas with schematic larea including cingulate cortex area 1 (Cg1) and 2 (Cg2) of dmPFC. (Bment is further (virtually) subdivided into four zones (hidden and prosub-chamber delimited by a perforated transparent wall confining thmore time in the proximal and distal zones of the social compartmencompartment as compared to sham-treated rats (n=11), indicatingmeans+s.e.m. ** Pb0.01, *** Pb0.001.

Behavior (HAB) rats, a well characterized and validated pre-clinical model of extreme trait anxiety (Landgraf andWigger, 2003; Liebsch et al., 1998), show differences of c-Fos expression in the dmPFC following exposure to differentanxiogenic stimuli (Frank et al., 2006; Salchner et al., 2006;Salome et al., 2004) and activity differences in response to di-azepam using fMRI (Kalisch et al., 2004), as compared to con-trols. Interestingly, the here presented hyperactivity ofdmPFC in F344 rats contrasts with a lower dmPFC activityfound in HAB rats. A possible explanation is model-specific dif-ferences in the anxiety phenotype. I.e., non-selective in-breeding of F344 rats, a less extreme model of anxiety, incontrast to the behavior-specific selective breeding of theHAB rats might have led to different dysregulations in distinct

344 rats on anxiety-related behavior in the SAA test. (A) Sectionsocalization of the smallest (gray) and largest (hatched) lesioned) The SAA test-box divided in a non-social and a social compart-tected; distal and proximal). The social compartment contains ae stimulus rat. (C) Lesioned F344 rats (n=12) spent significantlyt and significantly less time in the hidden zone of the non-sociala lower level of anxiety-like behavior. Data are presented in

Table 2 dmPFC lesion-induced brain activity changes in Fischer (F344) rats.

Brain area (order as in Table 1) Mean±s.e.m. and differences (Δ) of normalized perfusion (%)

F344 lesion(n=11)

F344 sham(n=12)

Δ (lesion− sham) Δ (sham−SD) Δ (lesion−SD)

Dorsomedial PFC (1) # −7±5 11±5 18 ↓ * 26 ↑ *** 7 ↑Entorhinal cortex (2) −1±3 1±2 2 ↓ 27 ↑ *** 24 ↑ ***Ventral tegmental area (3) 42±3 38±3 3 ↑ 13 ↓ 10 ↓Piriform cortex (4) 17±3 14±4 3 ↑ 8 ↓ 5 ↓Motor cortex (5) −13±3 −8±3 5 ↓ 6 ↑ 1 ↑Orbitofrontal cortex (6) 0±2 −6±2 6 ↑ * 16 ↓ *** 11 ↓ **Ventromedial PFC (7) 1±5 3±3 2 ↓ 19 ↑ *** 17 ↑ **Superior colliculus (8) 2±2 6±3 4 ↓ 20 ↓ *** 24 ↓ ***Dorsal striatum (9) 10±2 9±1 1 ↑ 13 ↑ *** 14 ↑ ***Inferior colliculus (10) 9±2 10±2 1 ↓ 19 ↓ *** 20 ↓ ***Nucleus accumbens (11) 9±3 5±1 4 ↑ 5 ↓ 0Ventral pallidum (12) 10±4 4±2 7 ↑ 4 ↓ 2 ↑Bed ncl of stria terminalis (13) −8±2 −5±1 3 ↓ 9 ↑ * 7 ↑Sensory cortex (S2) (14) −6±1 −12±1 6 ↑ ** 8 ↓ ** 2 ↓Insula (15) −8±1 −13±1 5 ↑ ** 6 ↓ ** 1 ↓Median Raphe (16) 27±3 22±3 4 ↑ 12 ↓ 7 ↓Paraventricular ncl of hypoth (17) −19±3 −20±2 1 ↑ 9 ↓ 8 ↓Dorsal peduncular (18) −8±3 −3±2 5 ↓ 7 ↑ 2 ↑Substantia nigra (19) 24±2 25±4 2 ↓ 3 ↑ 1 ↑Lateral hypothalamus (20) 16±2 16±1 0 0 0Amygdala (21) 0±2 −5±2 5 ↑ 3 ↑ 7 ↑Locus coeruleus (22) 26±2 26±2 0 1 ↑ 1 ↑Dorsal hippocampus (23) −1±4 −5±2 5 ↑ 4 ↑ 8 ↑Septum (24) −4±2 0±1 4 ↓ 5 ↑ 1 ↑Dorsal raphe (25) 6±2 0±4 6 ↑ 7 ↓ 2 ↓Periaqueductal gray (26) −30±2 −29±3 1 ↓ 6 ↓ 8 ↓Perirhinal cortex (27) −2±2 −5±3 3 ↑ 9 ↑ * 13 ↑ **Median hypothalamus (28) 9±2 9±2 0 3 ↑ 3 ↑Sensory cortex (S1) (29) −13±2 −15±1 2 ↑ 5 ↓ 2 ↓Ectorhinal cortex (30) 5±2 −8±1 3 ↑ 1 ↑ 4 ↑Ventral hippocampus (31) 0±2 1±3 1 ↓ 9 ↑ * 8 ↑ *Thalamus (32) 11±2 12±2 1 ↓ 2 ↓ 3 ↓Expected number nFP (nFP,5%–nFP,95%)of false positives

0.9 (0–3.4) 0.8 (0–2.2) 1.1 (0–2.6)

(#) Corresponds to area number in Fig. 2.; ↑↓ indicate positive and negative differences, respectively; bold corresponds to brain areasshowing statistically significant difference in Table 1. Significances according to Welch's t-test: *Pb0.05, **Pb0.01, ***Pb0.001. Expectednumber nFP of false positives and 90% confidence bounds nFP,5%−nFP,95% obtained from FDR analysis.

448 E.P. Prinssen et al.

brain areas within the anxiety-associated neurocircuitry. Ap-parently in agreement with this hypothesis, hypo- and hyper-activity of the human ACC has been observed in differenttypes of anxiety disorders (Berkowitz et al., 2007). Therefore,both models may have validity to elucidate the neurocircuitryof anxiety disorders and emphasize the dmPFC as a key regula-tory brain area of anxiety.

In a more recent series of imaging studies using aggressivemotivation versus an intruder, the medial PFC along withareas such as the amygdala and the hypothalamus was foundto be involved in the emotional response of rats; and theanxiety-related neuronal activity could be pharmacologicallymodulated (Caffrey et al., 2010; Ferris et al., 2008; Nephewet al., 2009). Importantly, the current findings that hyperac-tivity of the dmPFC plays a key role in the social anxiety-likebehavior in F344 rats may have translational value as an in-creased neuronal response to emotional stimuli in the ACC

(generally considered as the equivalent cortical area to therat dmPFC) is one of the key findings in patients with anxiety,in particular those with social anxiety (Bishop et al., 2004;Bremner, 2004; Goldin et al., 2009a, 2009b; Liao et al.,2010; Schneier et al., 2009).

A potential drawback associated with rodent fMRI studiesis that these studies are usually conducted under anesthesia,which may constitute a confounding factor on neuronal ac-tivity per se. During the recent years, a few fMRI studies innon-anesthetized rodents have been reported (Febo et al.,2004; Lahti et al., 1998, 1999; Peeters et al., 2001; Sicardet al., 2003). However, this approach faces the issue of mo-tion artifacts, unless animals are immobilized by paralysis(Peeters et al., 2001) or tight mechanical constraint (Lahtiet al., 1998), and of high stress levels induced by immobili-zation and/or scanner noise, even after acclimation proce-dures. Due to these limitations, conducting animal fMRI

449Imaging trait anxiety in high anxiety F344 rats: Focus on the dorsomedial prefrontal cortex

studies under well-chosen anesthesia conditions (i.e. 2–2.5%isoflurane) is in our view a preferable approach to obtainreliable and meaningful data on brain neurocircuitry(Bruns et al., 2009; Nordquist et al., 2008; Risterucci etal., 2005). Several arguments support functional relevanceof our main fMRI finding indicating a key role for thedmPFC: 1) the higher neuronal activation in the dmPFC ofF344 rats was confirmed by immunolabeling of c-Fos proteinalso in the SAA-test, 2) functional relevance of this corticalhyperactivity was shown by the finding that neurotoxiclesions of the dmPFC reversed high anxiety-like behavior ofF344 rats in the SAA test, 3) excitotoxic lesion of this brainarea partly normalized the neuronal fingerprint in F344rats to that of SD rats emphasizing a major role of dmPFCwithin an emotion processing neurocircuitry.

5. Conclusion

The current study shows that F344 rats demonstrate highanxiety-like behavior that may be related to an ‘abnormal’basal neurocircuitry of which hyperactivity of the dmPFC isa key element. Our findings, together with clinical observa-tions should provide an impetus to further explore ACChyperactivity as an imaging biomarker of social anxiety aswell as to develop new non-invasive therapeutic approachesthat specifically dampen ACC activity in patients with socialanxiety (Pallanti and Bernardi, 2009; Wu et al., 2007). Ingeneral, our current findings support the emerging conceptthat preclinical fMRI research can be of great value in char-acterizing brain structures and pathways underlying psychi-atric diseases, and to help discover innovative treatments.

Role of the funding source

The study was funded by F. Hoffmann-La Roche Ltd.

Contributors

E. Prinssen, C. Risterucci and J.-L. Moreau designed the study; L.B.Nicolas, S. Klein, C. Grundschober, C. Lopez-Lopez performed theexperiments; E. Prinssen, C. Risterucci and M.S. Keßler wrote themanuscript; A. Bruns, M. von Kienlin, J. G. Wettstein and J.-L.Moreau edited the manuscript.

Conflict of interest

The authors declare that they have no competing financial interests.All authors were employed at F. Hoffmann-La Roche Ltd. at the timeof work.

Acknowledgments

We would like to thank Francoise Kahn and Sébastien Debilly for ex-pert technical assistance; Basil Künnecke, Stephanie Schöppenthauand Thomas Bielser for setting up the MRI protocols and the dataanalysis programs.

References

Alsop, D.C., Detre, J.A., 1996. Reduced transit-time sensitivity innoninvasive magnetic resonance imaging of human cerebralblood flow. J. Cereb. Blood Flow Metab. 16, 1236–1249.

Amir, N., Klumpp, H., Elias, J., Bedwell, J.S., Yanasak, N., Miller,L.S., 2005. Increased activation of the anterior cingulate cortexduring processing of disgust faces in individuals with social pho-bia. Biol. Psychiatry 57, 975–981.

Andreescu, C., Gross, J.J., Lenze, E., Edelman, K.D., Snyder, S.,Tanase, C., Aizenstein, H., 2011. Altered cerebral blood flowpatterns associated with pathologic worry in the elderly. De-press. Anxiety 28, 202–209.

Berkowitz, R.L., Coplan, J.D., Reddy, D.P., Gorman, J.M., 2007.The human dimension: how the prefrontal cortex modulatesthe subcortical fear response. Rev. Neurosci. 18, 191–207.

Bert, B., Fink, H., Huston, J.P., Voits, M., 2002. Fischer 344 and Wis-tar rats differ in anxiety and habituation but not in water mazeperformance. Neurobiol. Learn. Mem. 78, 11–22.

Berton, O., Ramos, A., Chaouloff, F., Mormde, P., 1997. Behavioralreactivity to social and nonsocial stimulations: a multivariateanalysis of six inbred rat strains. Behav. Genet. 27, 155–166.

Bishop, S., Duncan, J., Brett, M., Lawrence, A.D., 2004. Prefrontalcortical function and anxiety: controlling attention to threat-related stimuli. Nat. Neurosci. 7, 184–188.

Bremner, J.D., 2004. Brain imaging in anxiety disorders. Expert.Rev. Neurother. 4, 275–284.

Bruns, A., Künnecke, B., Risterucci, C., Moreau, J.L., von Kienlin,M., 2009. Validation of cerebral blood perfusion imaging as a mo-dality for quantitative pharmacological MRI in rats. Magn. Reson.Med. 61, 1451–1458.

Caffrey, M.K., Nephew, B.C., Febo, M., 2010. Central vasopressinV1a receptors modulate neural processing in mothers facing in-truder threat to pups. Neuropharmacology 58, 107–116.

Canteras, N.S., Resstel, L.B., Bertoglio, L.J., Carobrez Ade, P.,Guimaraes, F.S., 2010. Neuroanatomy of anxiety. Curr. Top.Behav. Neurosci. 2, 77–96.

Clement, Y., Calatayud, F., Belzung, C., 2002. Genetic basis ofanxiety-like behaviour: a critical review. Brain Res. Bull. 57, 57–71.

Deichmann, R., Hahn, D., Haase, A., 1999. Fast T1 mapping on awhole-body scanner. Magn. Reson. Med. 42, 206–209.

Dhabhar, F.S., McEwen, B.S., Spencer, R.L., 1993. Stress response,adrenal steroid receptor levels and corticosteroid-binding globu-lin levels—a comparison between Sprague–Dawley, Fischer 344and Lewis rats. Brain Res. 616, 89–98.

Ding, J., Chen, H., Qiu, C., Liao, W., Warwick, J.M., Duan, X.,Zhang, W., Gong, Q., 2011. Disrupted functional connectivity insocial anxiety disorder: a resting-state fMRI study. Magn.Reson. Imaging 29, 701–711.

Etkin, A., Wager, T.D., 2007. Functional neuroimaging of anxiety:a meta-analysis of emotional processing in PTSD, socialanxiety disorder, and specific phobia. Am. J. Psychiatry 164,1476–1488.

Febo, M., Segarra, A.C., Tenney, J.R., Brevard, M.E., Duong, T.Q.,Ferris, C.F., 2004. Imaging cocaine-induced changes in the meso-corticolimbic dopaminergic system of conscious rats. J. Neu-rosci. Methods 139, 167–176.

Ferris, C.F., Stolberg, T., Kulkarni, P., Murugavel, M., Blanchard,R., Blanchard, D.C., Febo, M., Brevard, M., Simon, N.G., 2008.Imaging the neural circuitry and chemical control of aggressivemotivation. BMC Neurosci. 9, 111.

Frank, E., Salchner, P., Aldag, J.M., Salome, N., Singewald, N.,Landgraf, R., Wigger, A., 2006. Genetic predisposition toanxiety-related behavior determines coping style, neuroendo-crine responses, and neuronal activation during social defeat.Behav. Neurosci. 120, 60–71.

Goldin, P.R., Manber-Ball, T., Werner, K., Heimberg, R., Gross, J.J.,2009a. Neural mechanisms of cognitive reappraisal of negative

450 E.P. Prinssen et al.

self-beliefs in social anxiety disorder. Biol. Psychiatry 66,1091–1099.

Goldin, P.R., Manber, T., Hakimi, S., Canli, T., Gross, J.J., 2009b.Neural bases of social anxiety disorder: emotional reactivityand cognitive regulation during social and physical threat.Arch. Gen. Psychiatry 66, 170–180.

Haase, A., Frahm, J., Matthaei, D., Hänicke, W., Merboldt, K.D.,1986. Rapid NMR imaging using low flip-angle pulses. J. Magn.Reson. 67, 258–266.

Hettema, J.M., Neale, M.C., Kendler, K.S., 2001. A review andmeta-analysis of the genetic epidemiology of anxiety disorders.Am. J. Psychiatry 158, 1568–1578.

Kalisch, R., Salome, N., Platzer, S., Wigger, A., Czisch, M., Sommer,W., Singewald, N., Heilig, M., Berthele, A., Holsboer, F., Landgraf,R., Auer, D.P., 2004. High trait anxiety and hyporeactivity to stressof the dorsomedial prefrontal cortex: a combined phMRI and Fosstudy in rats. NeuroImage 23, 382–391.

Keßler, M.S., Debilly, S., Schöppenthau, S., Bielser, T., Bruns, A.,Künnecke, B., von Kienlin, M., Wettstein, J.G., Moreau, J.L.,Risterucci, C., 2011. fMRI fingerprint of unconditioned fear-likebehavior in rats exposed to trimethylthiazoline. Eur. Neuropsy-chopharmacol doi:10.1016/j.euroneuro.2011.07.011.

Lacroix, L., Spinelli, S., Heidbreder, C.A., Feldon, J., 2000. Differ-ential role of the medial and lateral prefrontal cortices in fearand anxiety. Behav. Neurosci. 114, 1119–1130.

Lahti, K.M., Ferris, C.F., Li, F., Sotak, C.H., King, J.A., 1998. Imag-ing brain activity in conscious animals using functional MRI. J.Neurosci. Methods 82, 75–83.

Lahti, K.M., Ferris, C.F., Li, F., Sotak, C.H., King, J.A., 1999. Com-parison of evoked cortical activity in conscious and propofol-anesthetized rats using functional MRI. Magn. Reson. Med. 41,412–416.

Landgraf, R., Wigger, A., 2003. Born to be anxious: neuroendocrineand genetic correlates of trait anxiety in HAB rats. Stress 6,111–119.

Leonardo, E.D., Hen, R., 2006. Genetics of affective and anxiety dis-orders. Annu. Rev. Psychol. 57, 117–137.

Liao, W., Chen, H., Feng, Y., Mantini, D., Gentili, C., Pan, Z., Ding,J., Duan, X., Qiu, C., Lui, S., Gong, Q., Zhang, W., 2010. Selec-tive aberrant functional connectivity of resting state networksin social anxiety disorder. NeuroImage 52, 1549–1558.

Liebsch, G., Montkowski, A., Holsboer, F., Landgraf, R., 1998. Beha-vioural profiles of two Wistar rat lines selectively bred for high orlow anxiety-related behaviour. Behav. Brain Res. 94, 301–310.

Lopez-Lopez, C., LeRoith, D., Torres-Aleman, I., 2004. Insulin-likegrowth factor I is required for vessel remodeling in the adultbrain. Proc. Natl. Acad. Sci. U. S. A. 101, 9833–9838.

Morgan, M.A., LeDoux, J.E., 1995. Differential contribution of dorsaland ventral medial prefrontal cortex to the acquisition and extinc-tion of conditioned fear in rats. Behav. Neurosci. 109, 681–688.

Muigg, P., Scheiber, S., Salchner, P., Bunck, M., Landgraf, R.,Singewald, N., 2009. Differential stress-induced neuronal acti-vation patterns in mouse lines selectively bred for high, nor-mal or low anxiety. PLoS One 4, e5346.

Navarro, J.F., Rivera, A., Maldonado, E., Cavas, M., de la Calle, A.,2004. Anxiogenic-like activity of 3,4-methylenedioxy-metham-phetamine (“Ecstasy”) in the social interaction test is accompa-nied by an increase of c-fos expression in mice amygdala. Prog.Neuropsychopharmacol. Biol. Psychiatry 28, 249–254.

Nephew, B.C., Caffrey, M.K., Felix-Ortiz, A.C., Ferris, C.F., Febo,M., 2009. Blood oxygen level-dependent signal responses in cor-ticolimbic ‘emotions’ circuitry of lactating rats facing intruderthreat to pups. Eur. J. Neurosci. 30, 934–945.

Nicolas, L.B., Prinssen, E.P., 2006. Social approach–avoidance be-havior of a high-anxiety strain of rats: effects of benzodiazepinereceptor ligands. Psychopharmacology (Berl.) 184, 65–74.

Nordquist, R.E., Risterucci, C., Moreau, J.L., von Kienlin, M., Kun-necke, B., Maco, M., Freichel, C., Riemer, C., Spooren, W.,

2008. Effects of aripiprazole/OPC-14597 on motor activity, phar-macological models of psychosis, and brain activity in rats. Neu-ropharmacology 54, 405–416.

O'Mahony, C.M., Sweeney, F.F., Daly, E., Dinan, T.G., Cryan, J.F.,2010. Restraint stress-induced brain activation patterns in twostrains of mice differing in their anxiety behaviour. Behav.Brain Res. 213, 148–154.

Pallanti, S., Bernardi, S., 2009. Neurobiology of repeated transcra-nial magnetic stimulation in the treatment of anxiety: a criticalreview. Int. Clin. Psychopharmacol. 24, 163–173.

Paxinos, G., Watson, C., 1998. The rat brain in stereotaxic coordi-nates. Academic Press, San Diego, CA.

Peeters, R.R., Tindemans, I., De Schutter, E., Van der Linden, A.,2001. Comparing BOLD fMRI signal changes in the awake andanesthetized rat during electrical forepaw stimulation. Magn.Reson. Imaging 19, 821–826.

Phan, K.L., Fitzgerald, D.A., Cortese, B.M., Seraji-Bozorgzad, N.,Tancer, M.E., Moore, G.J., 2005. Anterior cingulate neurochem-istry in social anxiety disorder: 1H-MRS at 4 Tesla. Neuroreport16, 183–186.

Qiu, C., Liao, W., Ding, J., Feng, Y., Zhu, C., Nie, X., Zhang, W.,Chen, H., Gong, Q., 2011. Regional homogeneity changes insocial anxiety disorder: a resting-state fMRI study. PsychiatryRes. 194, 47–53.

Ramos, A., Berton, O., Mormede, P., Chaouloff, F., 1997. Amultiple-test study of anxiety-related behaviours in six inbredrat strains. Behav. Brain Res. 85, 57–69.

Ressler, K.J., Mayberg, H.S., 2007. Targeting abnormal neural cir-cuits in mood and anxiety disorders: from the laboratory to theclinic. Nat. Neurosci. 10, 1116–1124.

Rex, A., Voigt, J.P., Fink, H., 1999. Behavioral and neurochemicaldifferences between Fischer 344 and Harlan–Wistar rats raisedidentically. Behav. Genet. 29, 187–192.

Risterucci, C., Jeanneau, K., Schoppenthau, S., Bielser, T., Kunnecke,B., von Kienlin, M., Moreau, J.L., 2005. Functional magneticresonance imaging reveals similar brain activity changes in twodifferent animal models of schizophrenia. Psychopharmacology(Berl.) 180, 724–734.

Rudebeck, P.H., Walton, M.E., Millette, B.H., Shirley, E., Rushworth,M.F., Bannerman, D.M., 2007. Distinct contributions of frontalareas to emotion and social behaviour in the rat. Eur. J. Neurosci.26, 2315–2326.

Sajdyk, T.J., Vandergriff, M.G., Gehlert, D.R., 1999. Amygdalarneuropeptide Y Y1 receptors mediate the anxiolytic-like actionsof neuropeptide Y in the social interaction test. Eur. J. Pharma-col. 368, 143–147.

Salchner, P., Sartori, S.B., Sinner, C., Wigger, A., Frank, E., Landgraf,R., Singewald, N., 2006. Airjet and FG-7142-induced Fos expres-sion differs in rats selectively bred for high and low anxiety-related behavior. Neuropharmacology 50, 1048–1058.

Salome, N., Salchner, P., Viltart, O., Sequeira, H., Wigger, A.,Landgraf, R., Singewald, N., 2004. Neurobiological correlatesof high (HAB) versus low anxiety-related behavior (LAB): differ-ential Fos expression in HAB and LAB rats. Biol. Psychiatry 55,715–723.

Schneier, F.R., Kent, J.M., Star, A., Hirsch, J., 2009. Neural circuit-ry of submissive behavior in social anxiety disorder: a prelimi-nary study of response to direct eye gaze. Psychiatry Res. 173,248–250.

Seamans, J.K., Lapish, C.C., Durstewitz, D., 2008. Comparing theprefrontal cortex of rats and primates: insights from electro-physiology. Neurotox. Res. 14, 249–262.

Shah, S.G., Klumpp, H., Angstadt, M., Nathan, P.J., Phan, K.L.,2009. Amygdala and insula response to emotional images in pa-tients with generalized social anxiety disorder. J. PsychiatryNeurosci. 34, 296–302.

Shrout, P.E., Fleiss, J.L., 1979. Intraclass correlations: uses in asses-sing rater reliability. Psychol. Bull. 86, 420–428.

451Imaging trait anxiety in high anxiety F344 rats: Focus on the dorsomedial prefrontal cortex

Sicard, K., Shen, Q., Brevard, M.E., Sullivan, R., Ferris, C.F., King,J.A., Duong, T.Q., 2003. Regional cerebral blood flow andBOLD responses in conscious and anesthetized rats under basaland hypercapnic conditions: implications for functional MRI stud-ies. J. Cereb. Blood Flow Metab. 23, 472–481.

Stein, M.B., Simmons, A.N., Feinstein, J.S., Paulus, M.P., 2007. In-creased amygdala and insula activation during emotion proces-sing in anxiety-prone subjects. Am. J. Psychiatry 164, 318–327.

Steiner, M.A., Lecourt, H., Rakotoariniaina, A., Jenck, F., 2011.Favoured genetic background for testing anxiolytics in the fear-potentiated and light-enhanced startle paradigms in the rat.Behav. Brain Res. 221, 34–42.

Storey, J.D., 2002. A direct approach to false discovery rates. J. R.Stat. Soc. B 64, 479–498.

Storey, J.D., Tibshirani, R., 2003. Statistical significance for geno-mewide studies. Proc. Natl. Acad. Sci. U. S. A. 100, 9440–9445.

Sullivan, R.M., Gratton, A., 2002. Behavioral effects of excitotoxiclesions of ventral medial prefrontal cortex in the rat arehemisphere-dependent. Brain Res. 927, 69–79.

Uylings, H.B., Groenewegen, H.J., Kolb, B., 2003. Do rats have aprefrontal cortex? Behav. Brain Res. 146, 3–17.

Wiebel, E.R., 1979. Serological methods. Practical methods forbiological morphometry. Academic Press, London.

Williams, D.S., Detre, J.A., Leigh, J.S., Koretsky, A.P., 1992. Mag-netic resonance imaging of perfusion using spin inversion of arte-rial water. Proc. Natl. Acad. Sci. U. S. A. 89, 212–216.

Wu, L.J., Xu, H., Ren, M., Zhuo, M., 2007. Genetic and pharmaco-logical studies of GluR5 modulation of inhibitory synaptic trans-mission in the anterior cingulate cortex of adult mice. Dev.Neurobiol. 67, 146–157.

Zacharko, R.M., MacNeil, G., Mendella, P.D., Hebb, A.L., 1999. Pro-active influence of a surgical stressor on locomotor activity, ex-ploration and anxiety-related behaviour following acutefootshock in the mouse. Brain Res. Bull. 48, 283–290.