|Home | About | Journals | Submit | Contact Us | Français|
Previous human imaging studies used facial stimuli to explore the potential association between depression and fear. This study aimed at investigating brain alterations in a rodent model of depression when innate fear was induced in the form of the predator odor trimethylthiazoline (TMT). Flinders sensitive line rats (FSL), a genetic animal model of depression, and their control counterpart Flinders resistant line (FRL), were used in this functional magnetic resonance imaging (fMRI) assessment. Compared to FRL, FSL rats exhibited greater BOLD activation in the cortical amygdala and hypoactivation in the prefrontal cortex in response to TMT, suggesting cortico-amygdalar dysfunction in the depressed strain. In addition, the hyperactivation in the insular cortex in FSL rats may be the basis for enhanced neuronal responses to fear and aversion in depression. These results are evidence for the value of translational models of depression in expanding understanding of the neural circuitries sub-serving common human co-morbidities like depression and fear.
Depression is a devastating disorder with estimates of 12% to 17% of the population experiencing at least one episode in a lifetime [14, 19]. Neuroimaging studies demonstrated depressed patients exhibit amygdala hyperactivity [18, 26] and dorsal lateral prefrontal cortex (DLPFC) hypoactivity in response to negative emotional facial stimuli [2, 6, 34]. Patients recovered from depression showed an increased response to fearful facial stimuli in the DLPFC [6, 37]. This neuronal link between depression and fear processing is well supported by the high prevalence of depression in combination with other mental health disorders such as selective fears, phobias and anxiety disorders. Since human imaging studies generally involve looking at fearful faces, the exploration of innate fear processing in an animal model may facilitate better understanding of the influence of depression upon more instinctual fear response.
A very useful paradigm to study innate fear in rodents is the utilization of a predator odor, often elicited by the odor of cats [15, 45]or foxes [5, 49, 60]. For the latter effect, a synthetic compound trimethylthiazoline (TMT) isolated from fox feces, has proven successful in producing fear-related responses in rats , including freezing, avoidance, increased defensive behaviors and stress hormone release [10, 21, 50].
As a translational model to human depression studies, two genetically linked strains of rats were utilized. The Flinders sensitive line rat (FSL) is a well validated genetic animal model of depression [13, 22, 23], exhibiting a number of behavioral [12, 24, 48] and neurochemical [1, 28, 42] similarities to depression in humans. The Flinders resistant line (FRL), was developed as a control counterpart for the FSL strain . These two lines have maintained their differential sensitivities over numerous generations , supporting validity construct mechanisms including good face and predictive validity .
In this study we used functional magnetic resonance imaging (fMRI) to examine brain response to innate fear elicited by TMT in an FSL rat model of depression. This study will explore the brain network underlying the direct fear response in an animal model of depression. We hypothesize that the fearful stimulus will elicit differential activation in selected brain areas within the brain circuitry subserving emotional regulation in depressed lines compared to controls.
Male FSL (n=8) and FRL (n=8) rats (350-400g) were bred at the University of North Carolina, Chapel Hill. After shipment, all rats were housed in pairs and were given at least 2 weeks to adjust to the new environment. The housing environment was maintained at 22-24°C with a 12 hour light/dark schedule (lights on at 06:00 and off at 18:00). Food and water were provided ad libitum. All rats were acquired and cared for in accordance with the guidelines published in the NIH Guide for the Care and Use of Laboratory Animals.
In order to scan fully conscious rats in a magnetic resonance imaging (MRI) scanner, the rats must be restrained . To reduce physiologic stress and motion artifact during imaging, the animals were acclimated to the restraint and imaging procedures before imaging for eight consecutive days using a previously validated acclimation procedure .
All images were acquired using a Bruker 4.7T/40cm horizontal magnet and a 20 Gauss/cm magnetic field gradient insert (inner diameter=12cm, Bruker, Billerica, MA U.S.A). High resolution multi-slice anatomical images were obtained using rapid acquisition relaxation enhanced sequence (RARE) with relaxation time TR=2.0s, echo time TE=12ms, resolution matrix=256×256, field of view (FOV)=30mm×30mm, 12 1.2-mm slices. Two subsequent functional scans were performed using echo-planar imaging sequence (EPI) at a resolution of 64×64 with the same FOV and slice thickness. TR=2.0s, TE=55ms, total acquisition time for each EPI scan was about 6 minutes. The first EPI scan had a 2 minute baseline period when fresh air was presented, followed by 4 minutes of neutral scent (lemon). In the second EPI, lemon was replaced by TMT.
All images were examined for artifacts and motion using Stimulate  and processed using the in-house developed software MIVA (Medial Image Visualization and Analysis, (http://ccni.wpi.edu ). Motion was corrected using SPM8 . On rare occasions, EPIs with motion bigger than 1/3 of the pixel size (~0.15mm) were excluded.
Each rat within group was aligned to a standard and thereafter to a fully segmented digital rat brain atlas that delineates 12 major anatomical regions and more than 70 subregions that are involved in different neural activities, based on 2D atlas textbooks [8, 52]. The anatomy volumes were aligned to the brain atlas volume using interactive affine registration . All transformed pixel locations of the anatomy images were tagged with the segmented atlas major and sub regions creating a fully segmented map of each subject.
Structural regions of interest (ROIs) were predetermined based on previous studies with TMT [21, 30] and other regions commonly cited to be involved in depression and fear response. Results particularly presented the olfactory bulb, prefrontal cortex (PFC), insular cortex, bed nucleus of the stria terminalis (BNST), amygdala and thalamus.
EPIs were analyzed following the above preprocessing. The first few scans were discarded to eliminate T1 effects. 55 scans before and 100 scans after the presentation of TMT were specified as the control and stimulation windows, respectively. Student t-tests were performed on each subject to determine pixel signal activity, using a 99% confidence level, two-tailed distribution and heteroscedastic variance assumptions. A statistical composite map in the standard space was created for each group with the average blood-oxygenation-level dependent (BOLD) activation on a pixel-by-pixel basis. In order to estimate regional differences, one-way ANOVA was performed between groups for each brain region, with p<0.05 as level of significance (degrees of freedom df=14).
Neural responses to both lemon scent and TMT were examined in FRL and FSL animals. Figures Figures11 and and2A2A compare regional BOLD changes of both groups in their response to lemon and TMT relative to baseline (fresh air). In the olfactory area (namely the olfactory bulb), there was no significant difference between the two groups in response to either of the two odors, neither was there a difference between the two odors.
In other selected brain areas associated with fear, lemon did not induce significantly different BOLD change between the two groups (Figure 1); whereas when exposed to TMT, FSL showed stronger activations in the cortical nucleus of the amygdala (F=5.0, p=0.045) and insular cortex (F=17.2, p=0.002), compared to FRL. In the PFC, less BOLD activation was observed in FSL (F=6.024, p=0.03). No significant difference was observed in the BNST and thalamus between the two groups (Figure 2A). Representative regional time courses are shown in Figure 2B and 2C.
Figure 3 depicts a composite map of TMT odor-induced BOLD activation of positive BOLD signal response across the brain. The activated voxels (based on a t-test, thresholded at p<0.05) are overlaid onto the anatomy atlas.
This study was performed to assess the impact of depression-like phenotype on responding to an innate fear-evoking stimulus . Prior studies have demonstrated increases in fear-related behavior of the FSL rats compared to their control counterparts [17, 32], supporting our functional imaging results of differential brain activation patterns between the two lines when processing fear. BOLD activation in the olfactory system when exposed to either lemon or TMT was similar between groups, indicating the differential neural activity found in brain areas between the two phenotypic groups is not induced by olfactory deficits. Similar BOLD activation in response to the neutral scent lemon ensures any difference found in TMT exposure is related to fear induced by this particular odor. Further, any additional stress associated with acclimation and restraint was constant between baseline and stimulus periods. Since BOLD changes were calculated as a difference between baseline and stimulus periods, this potential alteration would have been taken into account.
Our results strongly support three critical brain regions, namely the cortical amygdala, PFC and insular cortex in processing fear-related stimulus in the animal model of depression versus their control counterparts. First, hyperactivity in the cortical nucleus of the amygdala in response to TMT in FSL animals is consistent with human findings that the amygdala plays an important role in fear response , especially under negative emotional states like depression. While most human studies focus on the entire amygdala (mainly the central nucleus) [2, 34, 35], others have reported the role of basolateral amygdala  as essential to fear processing. Although prior studies have linked the cortical nucleus of amygdala to smell and pheromone-processing  others reported that the posterior cortical nucleus modulates memory processing , to our knowledge the role of the cortical amygdala in depression related fear processing is not well established. The current observed exaggerated activation in this region suggests the cortical nucleus of the amygdala is a critical area in depressed subjects processing innate fear stimuli.
Decreased PFC activity in response to fear aligns well with previous work in the FSL animal model. Using the FSL strain and Sprague-Dawley (SD) rats as controls, Yadid et al. assessed the effects of stress on dopamine release from the PFC . Their findings demonstrated diminished dopamine release in response to stress in the FSL compared to controls. The observed hypoactivity of the PFC in the FSL rats after the application of fearful stimuli may be due to alterations in dopamine [16, 31, 34], norepinephrine  and/or the serotonin[3, 59].
This study also found evidence of enhanced activation in the insular cortex in response to fear induction in FSL animals. Although both gray matter reduction  and hypoactivity in the insular cortex have been documented in depression , the enhanced activation in the current study may be due to complex role of the insular cortex in environmental monitoring and negative emotional processing [39, 43, 61]. Given the emotional regulatory influence of the insular cortex [9, 25] particularly its known function (in tandem with the amygdala) in contributing to the integration of emotional salience [43, 46, 47], the observed increased activation in the insular is not unexpected.
Our results in depressed subjects demonstrate a dysfunctional relationship in the processing of information between the PFC and amygdala in an animal model . Since the PFC acts as a primary cortical control of the limbic system while the amygdala functions as a regulatory structure for the processing of emotional experiences , we speculate that a contrast of hyper- and hypoactivity seen in the amygdala and PFC respectively may be a critical circuitry sub-serving the exaggerated emotional responses often observed in depressed subjects. This proposal of a dual regulatory capacity of the cortico-limbic loop is supported by previous studies performed in a selectively bred genetic animal model for high (HAB) and low (LAB) anxiety-related behavior [27, 33]. However, the HAB rodents are genetically susceptible to high anxiety with comorbidity of depression ; whereas the FSL line is selectively bred to exhibit only depression-like similarities. The aim of this study was to investigate the neural networks of fear processing in depression, rather than the networks supporting the coexistence of the phenotypically disparate mood states of depression and anxiety. Nonetheless, the similarities in finding of amygdala-prefrontal cortex dysregulation in both lines, strengthens the argument of a dysfunctional mechanism of fear processing in subjects with a mood disorder.
In conclusion, our results suggest that depression influences the neural response to innate fear in an animal model of depression. The pre-existing negative emotional state of depression supported enhanced cortical amygdala and insular reactivity, as well as a diminished PFC activation. These results also promote the utilization of translational models of depression to expand our understanding of the neural circuitries sub-serving common human co-morbidities like depression and fear/anxiety.
This publication was made possible by the NIH Grant Number 1RO1 MH067096. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIH.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.