Repeated Ferret Odor Exposure Induces Different Temporal Patterns of Same-Stressor Habituation and Novel-Stressor Sensitization in Both Hypothalamic-Pituitary-Adrenal Axis Activity and Forebrain c-fos Expression in the Rat

Abstract

Repeated exposure to a moderately intense stressor typically produces attenuation of the hypothalamic-pituitary-adrenal (HPA) axis response (habituation) on re-presentation of the same stressor; however, if a novel stressor is presented to the same animals, the HPA axis response may be augmented (sensitization). The extent to which this adaptation is also evident within neural activity patterns is unknown. This study tested whether repeated ferret odor (FO) exposure, a moderately intense psychological stressor for rats, leads to both same-stressor habituation and novel-stressor sensitization of the HPA axis response and neuronal activity as determined by immediate early gene induction (c-fos mRNA). Rats were presented with FO in their home cages for 30 min a day for up to 2 wk and subsequently challenged with FO or restraint. Rats displayed HPA axis activity habituation and widespread habituation of c-fos mRNA expression (in situ hybridization) throughout the brain in as few as three repeated presentations of FO. However, repeated FO exposure led to a more gradual development of sensitized HPA-axis and c-fos mRNA responses to restraint that were not fully evident until after 14 d of prior FO exposure. The sensitized response was evident in many of the same brain regions that displayed habituation, including primary sensory cortices and the prefrontal cortex. The shared spatial expression but distinct temporal development of habituation and sensitization neural response patterns suggests two independent processes with opposing influences across overlapping brain systems.

Repeated exposure of rats to ferret odor leads to rapid development of stimulus-specific habituation and slower development of novel-stressor sensitization of forebrain and hypothalamic-pituitary-adrenal axis activity.

Chronic stress contributes to general health problems such as hypertension, cardiovascular disease, diabetes, stroke, accelerated carcinogenesis, and visceral fat accumulation (1,2,3). Additionally, chronic stress may adversely influence personality development and behavior (4). In some psychiatric disorders, for example post traumatic stress disorder, pathological adaptation to stress may be a primary contributor to the disorder (5). Consequently, study of the adaptive processes associated with chronic stress may be informative to understanding physical as well as mental health problems.

One of the key principles to emerge from a rich history of stress research is that repeated exposure to a particular stressor can result in habituation of the hormonal response to that same (homotypic) stressor (6, and for review see Ref. 7) but sensitization of the hormonal response to a novel (heterotypic) stressor (8,9,10,11,12,13,14,15,16,17,18). Both perceived stressfulness and stress adaptation have been characterized as being tightly coupled to relative activity of the hypothalamic-pituitary-adrenal (HPA) axis (19,20,21). However, to date, there has been very little determination of whether this principle of differential adaptive responses to homotypic vs. heterotypic stressors applies to specific regional brain activity. Immediate-early gene (IEG) expression has been used to assess the neuronal circuitry associated with HPA axis responses (22). By comparing relative stressor-induced neuronal activity of rats exposed to a stressor for the first time with that of rats having had prior experience with the same or other stressors, it has been possible to determine which brain regions show adaptive responses concurrent with adaptation of HPA axis endocrine measures. This strategy of measuring relative IEG expression throughout the brain has been applied primarily to the study of HPA axis habituation (23,24,25,26,27,28,29,30,31,32), with far fewer studies describing neuronal adaptation concurrent with HPA axis sensitization (9,11,33).

In general, repeated exposure to a homotypic challenge leads to widespread reduction in regional brain immediate-early gene expression compared with an acute exposure to the same stressor. There have been some reports of increased IEG expression in some brain regions [e.g. the orbitofrontal cortex (24)], but this finding does not appear to generalize to different situations (28). In animals given prior repeated stressor exposure, the presentation of a novel stressor producing sensitized HPA axis responses has been associated with increased IEG expression in the parabrachial nucleus, posterior paraventricular thalamic nucleus, amygdaloid subnuclei, and the paraventricular nucleus of the hypothalamus (9,11,33). However, the extent to which habituation and sensitization processes take place in the same brain regions and whether they develop simultaneously or with different time courses has not been investigated in the same study. The goal of the present study was therefore to evaluate neuroendocrine and brain indices of habituation and sensitization induced by the same repeated exposure protocol and examine these indices over time. For this study we used predator odor as the repeatedly presented stressor (ferret-exposed towels) and restraint as a novel acute stimulus. Both predator odor exposure and restraint lead to activation of the HPA axis in rats (34,35,36,37), suggesting that these stimuli are perceived as stressful. Prior studies have found that repeated ferret odor produced both habituation and sensitization of the HPA axis under some conditions (38,39).

Our first experiment examined the extent to which two weeks of repeated ferret odor (FO) exposure led to sensitization and/or habituation of the HPA axis and central c-fos gene expression response to challenge with FO or a novel stressor (restraint). We examined c-fos mRNA levels (in situ hybridization) in the paraventricular nucleus of the hypothalamus (PVN) and ventral-orbital and medial prefrontal cortices (PFC). PVN c-fos gene expression to an acute challenge has been shown to correlate strongly with relative HPA axis activity (37) and thereby provides an indication of relative activity of the neuronal component of the HPA axis. The PFC may play an important role in modulation of the HPA axis stress response (40). In addition, the PFC expresses high levels of IEGs, including c-fos, after acute exposure to either FO (41) or restraint [e.g. (36)], and it exhibits attenuation of IEG expression concurrent with HPA axis hormone habituation (28).

After observing in the first experiment both sensitization and habituation of IEG expression in PFC, as well as concurrent adaptations in the HPA axis, we undertook a follow-up time-course experiment to examine whether expression of habituation and sensitization follow the same temporal development and whether the expression of these repeated-stress adaptations were distributed similarly or differently across regions of the forebrain. In addition to the PVN and PFC, we examined c-fos expression in the barrel fields, a primary sensory cortical region related to somatosensory processing, the piriform cortex, related to olfactory processing, the paraventricular nucleus of the thalamus (PVThal), a modulator of stress-adaptive processes (42), and subnuclei of the amygdala, strongly associated with the predator odor-induced fear response (39,43,44).

Materials and Methods

Animals

Male Sprague Dawley rats (Harlan Laboratories, Indianapolis, IN) were allowed a 2-wk acclimation period after arrival to the animal facilities at the University of Colorado before experimental use (10 wk old, weight range 285–360 g at experimental onset). Rats were housed in pairs in polycarbonate cages (∼46 cm length × 24 cm width × 20 cm depth) with wood shavings until 1 d before the first day of odor presentation, after which rats were housed individually in same-sized cages. Rats were given food (Purina Rat Chow; Ralston Purina, St. Louis, MO) and tap water ad libitum. The colony room lights were maintained on a 12-h light, 12-h dark cycle, with lights on at 0700 h. Rat were weighed 1 d before the first presentation of odor stimulus and on exposure d 13 for a total weight gain interval of 13 d. Procedures for ethical treatment of animals conformed to the guidelines found in the Guide for the Care and Use of Laboratory Animals, Department of Health and Human Services publication (National Institutes of Health) 80–23, revised 1996 edition and were approved by the University of Colorado Institutional Animal Care and Use Committee.

Housing and odor exposure

Animals were housed in three adjacent rooms equipped with separate air ventilation units to avoid odor contamination. Rats were divided among rooms based on the odorants they received during the first 2 wk and on test day (i.e. rats receiving FO presentation would never be housed with animals naïve to FO or with animals not receiving FO on that day). Cages and bedding were changed weekly.

During the first 2 wk, rats were exposed to a specific regimen of strawberry and/or FO. To control for the potential acute and chronic effects of daily placement of two pieces of scented bath towels in the rat’s home cage for 30 min each day, we also placed control-scented towels in the home cage of rats that were not exposed to FO. Strawberry odor was used as a control odor stimulus (38). We previously demonstrated that strawberry odor exposure does not lead to changes in HPA axis activity (41) or defensive behaviors (38).

The strawberry odor controls were small portions of a bath towel (approximately 5 × 5 cm squares) scented with 20 μl undiluted strawberry extract (McCormick Corp., Sparks, MD). The FO towels were similarly 5 × 5 cm sections of bath towels generated by placing a bath towel in a cage with one male and one female undescented adult ferrets for approximately 1 month (gift from Dr. V. Staton, Ohio Dominican University, experiment 1), or providing bath towels to members of the Mile High Ferret Club of Colorado to be used as ferret bedding for approximately 1 month (experiment 2). Ferret towels were kept at −80 C between uses and thoroughly thawed for each use. The towels were transported to the experimental rooms immediately before testing inside sealed glass bell jars and were brushed clean each day. Strawberry towels were rescented every other day with fresh extract before use. Towels were affixed to paperclips and two were hung inside each home-cage from the wire lid at opposite ends of the long axis of the cage. Placing the towels at both ends of the cage presented the odor in a fashion that was assumed difficult to avoid. Several measures were taken to minimize systematic differences in FO potency between towels. First, the same towels were used each day in a randomized fashion so that no cage intentionally received the same towel each day. Second, the same pools of FO towels used during the first 2 wk were used again on the final test day, again in a randomized fashion. Finally, on the test day, all rats were exposed to towels of equal previous usage and freeze-thaw cycles.

Treatment groups: experiment 1

The first experiment examined the effect of 14 prior days of FO presentation on the HPA axis and c-fos mRNA response to acute challenge with FO or a novel stressor (restraint). FO or strawberry odor was presented for 30 min each morning for 14 d, and the acute challenge was presented for 30 min on the morning of d 15. Rats (N = 24) were divided into four treatment groups (n = 6) according to a 2 × 2 factorial design. The first factor was the odor exposure condition over the first 14 d: predator odor (FO) or control odor (strawberry). The second factor was the stressor challenge condition on d 15: homotypic stressor (FO) or heterotypic stressor (restraint).

Treatment groups: experiment 2

The second experiment examined the temporal development of adaptation to repeated FO presentation by comparing the HPA axis and c-fos mRNA response to acute challenge with FO or restraint after zero, two, seven, or 14 daily presentations of FO. Similar to experiment 1, odors (ferret or strawberry) were presented for 30 min per day, and then on the test day (d 15) rats were either restrained for 30 min or exposed to FO for 30 min. Rats (N = 64) were divided into eight treatment groups (n = 8) according to a 4 × 2 factorial design. The first factor was the odor exposure condition over the first 14 d: 0, 2, 7 or 14 d of predator odor (FO). Rats receiving 0 d FO exposure were exposed to strawberry odor each day for 30 min for 2 wk. Rats receiving 2 d FO exposure were exposed to strawberry odor each day for 12 d and on d 13 and 14 were exposed to FO for 30 min. Rats receiving 7 d FO exposure were exposed to strawberry odor each day for 7 d and on d 8–14 were exposed to FO for 30 min. Rats receiving 14 d FO were exposed to FO each day for 2 wk and were never exposed to strawberry odor. The second factor was the stressor challenge condition on d 15: homotypic stressor (FO) or heterotypic stressor (restraint). Due to logistical issues that included insuring that all rats received treatments within a restricted portion of the light-dark cycle, the second experiment was conducted on two separate cohorts of rats (n = 4, N = 32 per cohort), and their data were pooled.

Test day procedure (experiments 1 and 2)

On test day (d 15), rats were either exposed to 30 min of FO in their home cage or restrained for 30 min between 0900 and 1200 h. This time of day was chosen because it is near to the circadian trough of HPA hormone secretion when variability between baseline HPA axis activities is minimal. Restraint involved taking rats from the home cage and placing them in adjustable length (15.5 ± 2.5 cm long and 6.3 cm diameter) Plexiglas tubes with air holes in the front, top, and back. This stressor is considered to be primarily psychological in nature because it does not produce pain or direct physical insult (45). Restraint was administered in a separate room adjacent to the home cage room.

For experiment 2, an observer blind to treatment groups scored the number of burial mounds present at the end of 30 min in each cage of rats that were challenged on d 15 with FO. A burial mound was determined by the appearance of a wood chip pile located directly below the FO towel, at a height of approximately double (6 cm) that of normal bedding (3 cm) from the bottom of the cage.

Rats were killed via decapitation 30 min after onset of FO or restraint challenge. Previous work from our laboratory (28) has found that the magnitude of difference of HPA axis responses and c-fos mRNA levels between acutely and repeatedly restrained rats is greatest at this time. Upon decapitation, brains were collected, flash frozen in an isopentane bath maintained between −40 and −30 C and stored at −80 C. Trunk blood was taken in EDTA-coated tubes, kept on wet ice, and centrifuged at 5000 × g for 10 min within 45 min after collection. Plasma was then rapidly frozen and stored at −80 C until hormone assay. Thymi were extracted and maintained in 24-well plates over wet ice for no longer than 3 h and were then weighed. Adrenals were extracted and frozen in dry ice to be weighed in pairs at a later date.

Corticosterone (CORT) ELISA and ACTH RIA

Measurement of plasma corticosterone was conducted on 20 μl of plasma with an enzyme immunoassay kit (Assay Design, Ann Arbor, MI) according to the manufacturer’s instructions. Plasma concentrations of ACTH were determined by RIA as described previously (28). Sensitivity for the CORT assay was 130 ng per 100 ml. The intraassay coefficient of variability (CV) for the CORT assay was approximately 4%, and the interassay CV was approximately 6%. The detection limit for the ACTH assay was 15 pg/ml for a 50-μl sample and the intraassay CVs were between 2 and 7%.

In situ hybridization

Brain sections (14 μm) were cut on a cryostat (Leica Microsystems, Bannockburn, IL; model 1850) through the extent of the prefrontal cortex [∼3.20 mm anterior to bregma (46)], PVN, barrel fields, anterior PVThal, rostral amygdala, and piriform cortex (∼1.80 mm posterior to bregma), and posterior PVThal (∼3.60 mm posterior to bregma); thaw mounted onto poly l-lysine-coated slides; and stored at −80 C. In situ hybridization for c-fos mRNA was performed as described previously (28). For generation of the c-fos probe, plasmids containing a fragment of c-fos cDNA (courtesy of Dr. T. Curran, St. Jude Children’s Research Hospital, Memphis, TN) were used. [³⁵S]-labeled cRNA probes were generated using standard in vitro transcription reagents (Promega, Madison, WI). Dehydrated sections were exposed to x-ray film for 1–4 wk.

Image analysis

All analyses were performed with the aid of the atlas of Paxinos and Watson (46) for guidance. Semiquantitative analyses of autoradiographs were performed on digitized images from x-ray films (National Institutes of Health Image, Bethesda, MD) as described (47) with the following specifications: for the prefrontal cortex, a 25 × 25 pixel square was centered within the prelimbic (PrL), infralimbic (IL), or ventral orbital (VO) prefrontal cortex, and for the primary somatosensory cortex (barrel cortex), a same-sized square was centered over the two layers of cortex displaying the highest c-fos mRNA levels, layer IV or layer VI (S1BF layer IV and S1BF layer VI, respectively). The most prominent c-fos expression in amygdala sections was found in the rostral amygdala sections, from 1.8 to 2.3 mm posterior to bregma. A 30-pixel diameter circle was centered over the medial amygdala (MeA), the combined basolateral/lateral amygdala (BLA), or the central nucleus (CeA) of the amygdala. The identical method used for amygdala nuclei was also used for anterior and posterior PVThal. For the piriform cortex and PVN, the region of interest (ROI) was drawn around the brain structure using the Paxinos and Watson atlas for guidance. For all cortical analyses, an average integrated gray level was determined by averaging all measured gray levels per brain (four measurements per brain). For each subcortical analysis, an average of the three integrated gray level measurements of the greatest value (three slides, four sections per slide, bilateral measurements) was determined.

Statistical analyses

All data were analyzed using a Statistical Analysis System (SAS, Cary, NC) package for UNIX. Separate two-way ANOVAs (pretreatment condition × test day challenge) were performed when comparing gene expression within a given brain region and when comparing HPA axis hormone levels. Separate one-way ANOVAs were performed for data from experiment 2 concerning physiological measures or burial behavior. When measuring burial behavior, only those rats presented with FO on the test day were included in the ANOVA and post hoc test. Two-tailed t tests were performed on physiological data from experiment 1. When testing pairwise group differences in ACTH, CORT, thymus weight, adrenal weight, and gene expression (in situ hybridization data), post hoc tests (Fisher’s least significant difference test) were performed, and the results are indicated on the data figures. Slight differences in within-group degrees of freedom in a given experiment are due to lost or damaged tissue or plasma samples. In all statistical comparisons, an α-level of P < 0.05 was used to determine statistical significance. Data presented represent group mean ± sem.

Results

Experiment 1

Experiment 1 addressed the following: 1) whether first-time exposure to FO led to HPA axis and/or neural responses in the rat similar to restraint challenge; 2) whether repeated exposure to FO led to adrenal, thymus, and body weight changes in the rat; 3) whether the HPA axis and/or neural responses to FO habituated with repeated exposure; and 4) whether repeated FO exposure led to a sensitized HPA axis and/or neural response to subsequent restraint challenge.

FO-induced weight changes in the rat

Two weeks’ exposure to 30 min/d of FO led to a significant difference in body weight gain: t(22) < 0.0001, decreased thymus weight: t(22) < 0.001, and decreased thymus to body weight ratios: t(22) < 0.01 relative to rats presented with daily strawberry odor (Table 1). Whereas rats exposed to strawberry odor each day increased their body weight over the 2 wk, rats exposed to FO lost weight. Neither adrenal weights nor adrenal to body weights were affected by the FO exposure.

Table 1.

Physiological changes after 2 wk repeated exposure to strawberry or FO (experiment 1)

Physiological measure/odor stimulus	Change in body weight over 2 wk (g)	Thymus weight (mg)	Thymus weight (mg/g BW) × 100	Adrenal weight (mg)	Adrenal weight (mg/g BW) × 100
Strawberry	28.67 ± 2.34	312.67 ± 13.10	91.86 ± 4.62	64.5 ± 2.19	17.84 ± 0.90
Ferret	−6.08 ± 4.84a	234.5 ± 13.79a	72.98 ± 3.88a	63.25 ± 1.86	19.79 ± 0.66

Body weights were measured 1 d prior to the first presentation of odor stimulus (d 0) and on d 13. Thymus and adrenal weights were determined postmortem. Values represent the mean ± sem of each group. BW, Body weight. 

^aRats exposed to FO differed significantly from rats exposed to strawberry odor (P < 0.05). 

HPA axis hormone response and PVN c-fos response to acute FO and restraint challenge ± 2 wk prior FO exposure

First time exposure to FO led to a slightly weaker HPA axis hormone response than restraint challenge (Fig. 1). There was a trend for less ACTH secretion and significantly less CORT (P < 0.05) secretion in rats challenged with FO compared with restraint. There was no observable difference in PVN c-fos expression between first-time FO and restraint-challenged rats. Rats with 2 wk prior experience with FO responded to test day challenge of FO with habituation of the HPA axis hormone response (trend for habituation of ACTH secretion and significant habituation of CORT secretion; P < 0.01). Additionally, this group had significantly less PVN c-fos expression (P < 0.01) than rats exposed to FO for the first time. Rats repeatedly exposed to FO compared with rats repeatedly exposed to strawberry odor responded to restraint challenge with elevated HPA axis hormone responses (trend for increased ACTH secretion and significantly elevated CORT secretion; P < 0.05). This group also had significantly elevated c-fos expression in the PVN (P < 0.05), compared with restraint-challenged rats that had no prior FO experience (Fig. 1). Thus, there was a significant effect of test day challenge (FO or restraint) on ACTH [F(1, 21) = 9.37, P < 0.01]; CORT [F(1, 22) = 18.95, P < 0.001]; and PVN c-fos expression [F(1, 22) = 6.42, P < 0.05] and a significant interaction between test day challenge and repeated stimulus exposure on ACTH [F(1, 24) = 4.23, P = 0.05]; CORT [F(1, 22) = 7.26, P < 0.05]; and [PVN c-fos expression F(1, 22) = 6.99, P < 0.05].

Figure 1.

Differential effects of 2 wk of FO exposure on homotypic and heterotypic stressor-induced HPA axis hormone and PVN c-fos mRNA response (experiment 1). Rats were challenged with 30 min FO or restraint after 2 wk of daily strawberry or FO preexposure. S, Strawberry, F, ferret, R, restraint. @, FO-naïve rats exposed to restraint on the 15th day had greater plasma CORT secretion compared with FO-naïve rats exposed to FO on the 15th day (P < 0.05). *, Significantly different from a group presented with FO for the first time on d 15 (P < 0.05); #, significantly different from the group presented each day with strawberry and challenged with restraint on d 15 (P < 0.05).

PFC c-fos response to acute FO and restraint challenge ± 2 wk FO exposure

In contrast to c-fos expression patterns in the PVN, first-time FO exposure led to a greater c-fos expression in the PrL (P < 0.01), IL (P < 0.01), and VO (P < 0.01) subregions of the PFC than restraint challenge (Fig 2; see Fig. 3; for representative autoradiograms of PFC c-fos expression). Rats with 2 wk prior experience with FO responded to test-day challenge of FO with habituation of c-fos expression in all three PFC subregions (P < 0.01). Rats repeatedly exposed to FO compared with rats repeatedly exposed to strawberry odor responded to restraint challenge with significantly elevated c-fos expression in the PrL (P < 0.05) and IL (P < 0.05) PFC, with a trend for the same in the VO PFC. There was a significant interaction of test day challenge (FO or restraint) and repeated stimulus exposure (FO or strawberry odor) in the PrL, IL, and VO subregions of the PFC [PrL: F(1, 21) = 15.40, P < 0.001]; IL [F(1, 21) = 26.46, P < 0.0001]; and [VO: F(1, 21) = 14.45, P < 0.01].

Figure 2.

Differential effects of 2 wk of FO exposure on homotypic and heterotypic stressor-induced c-fos mRNA expression in the PFC (experiment 1). c-fos mRNA expression in the PrL, IL, and VO PFC was measured after rats were exposed to 30 min FO or restraint challenge after 2 wk of daily strawberry or FO preexposure. S, Strawberry, F, ferret, R, restraint. @, Rats exposed to FO for the first time on the 15th day had greater c-fos mRNA expression across the three subdivisions of the PFC than FO-naïve rats exposed to restraint on the 15th day (P < 0.05). *, Significantly different from a group presented with FO for the first time on d 15 (P < 0.05); #, significantly different from the group presented each day with strawberry and challenged with restraint on d 15 (P < 0.05).

Figure 3.

Representative autoradiograms of c-fos mRNA expression in rats exposed to FO or restraint after 2 wk daily exposure to strawberry or FO (experiment 1). Coronal sections at the levels of the PFC (left column) and at the level of the PVN (right column) illustrate that in multiple regions of the brain (including multiple regions in the same coronal section), there is habituation of c-fos expression in rats repeatedly exposed to FO and sensitization of c-fos expression in rats repeatedly exposed to FO and challenged with a novel stressor.

Behavioral responses to FO

Over the 2-wk period of placing FO towels in the cages of rats, we noticed several interesting behaviors, some of which we analyzed more carefully in experiment 2. For instance, placement of the first of two FO towels in the cage of rats led some, but not all, rats to rapidly run to the other side of the cage. This behavior was occasionally so pronounced that the rat ran into the wall of the other side of the cage. This behavior did not persist beyond the second FO exposure in these rats, and no rats developed this behavior after the first several days. Additionally, we observed that some rats began displaying burial behavior selectively directed toward FO-scented towels and that this behavior subjectively increased both in the number of rats displaying this behavior and in the extent to which a given rat displayed this behavior across days. Never over the course of 2 wk did we observe a rat bury a strawberry-scented towel.

Experiment 2

Experiment 2 examined the temporal development of the following: 1) repeated FO-induced body and organ weight changes, 2) HPA axis and neural response habituation to repeated FO, 3) FO-induced HPA axis and neural response sensitization to restraint challenge, and 4) FO-induced burial behavior.

FO-induced weight changes in the rat over time

Repeated FO exposure reduced body weight gain and this effect was already evident after 2 d of FO preexposure (Fig. 4). Less body weight gain was evident after seven preexposures to FO, after which time no further limitation of weight gain was observed. Reduction in the thymus to body weight ratio was more linear, with evidence of continued reduction of thymus to body weight ratio occurring throughout the 2-wk period. Thus, there was a significant effect of days of FO exposure on changes in body weight gain [F(3, 60) = 28.77, P < 0.0001] and thymus to body weight ratio [F(3, 60) = 3.63, P < 0.05]. As in the first experiment, there was no significant effect of days of FO exposure on changes in either adrenal weight or adrenal to body weight ratio, although a small trend for an increase in adrenal to body weight ratio across days of FO exposure is apparent (Fig. 4).

Figure 4.

Body, thymus, and adrenal weight changes across 2 wk repeated exposure to strawberry or FO (experiment 2). Rats were exposed to FO on the preceding 0, 2, 7, or 14 d. Body weights were measured 1 d before the first presentation of odor stimulus and on d 13. Thymus and adrenal weights were determined postmortem. *, Groups with significantly lower means from FO-naïve rats (P < 0.05).

Temporal development of HPA axis and PVN c-fos habituation and sensitization

Overall, the HPA axis hormone response as well as PVN c-fos expression to repeated FO exposure habituated rapidly, with evidence for significant habituation to FO by the third exposure (2 d prior exposure plus test day challenge, Fig. 5). ACTH secretion did not significantly habituate across the 2-wk period. The CORT response exhibited significant habituation by 8 d of repeated FO exposure and did not habituate further. PVN c-fos expression habituated significantly by the third FO exposure and also did not habituate further.

Figure 5.

Effect of previous FO exposure on homotypic and heterotypic stressor-induced HPA axis hormone and PVN c-fos mRNA response (experiment 2). Rats were exposed to FO on the preceding 0, 2, 7, or 14 d and challenged with 30 min restraint (circles and solid line) or FO (diamonds and dashed line) on the test day. %, Group exposed to restraint on test day with a significantly greater mean ACTH response to restraint than restrained group with two previous exposures to ferret odor; *, groups exposed to FO on test day with significantly lower mean responses from the group exposed to FO for the first time (P < 0.05).

The HPA axis hormone as well as PVN c-fos expression sensitization developed more gradually across the 2-wk period. Restraint-induced ACTH secretion was almost the same between rats with no prior FO exposure and rats that had two prior exposures to FO, but ACTH sensitization to restraint challenge gradually increased over the rest of the 2 wk, showing marked sensitization (P < 0.05) in rats with 2 wk prior exposure to FO, compared with rats with only 2 d prior FO exposure (Fig. 5). The CORT response did not demonstrate sensitization in contrast to results of the first experiment. This may be due to a heightened overall CORT responsivity of these rats to restraint compared with the responses in the first experiment. Overall, there was a significant main effect of test day challenge on ACTH secretion [F(1, 58) = 12.39, P < 0.001] and PVN c-fos expression [F(1, 62) = 51.74, P < 0.0001] and a trend for a significant interaction between days of FO and test day challenge on PVN c-fos expression [F(3, 62) = 2.35, P = 0.08)].

Temporal development of c-fos expression habituation and sensitization patterns in brain regions outside of the PVN

Habituation of FO-induced c-fos expression was evident in all brain regions investigated: PrL, IL, and VO PFC, layers IV and VI of the barrel cortex, the piriform cortex (Fig. 6), CeA, BLA, and MeA nuclei of the amygdala, and anterior and posterior sections of the paraventricular thalamus (Fig. 7). Throughout all of these brain regions, habituation developed rapidly, with the greatest extent of habituation manifest within the first few days of FO exposure. Beyond the third exposure to FO, there was generally very limited further habituation of c-fos expression. Sensitization of c-fos expression, on the other hand, developed more gradually in cortical and some subcortical regions. Whereas this was particularly evident in primary sensory cortex (barrel cortex) and the PFC, sensitization was not observed in paraventricular thalamic areas.

Figure 6.

Effect of previous FO exposure on homotypic and heterotypic stressor-induced c-fos mRNA expression in cortical subdivisions (experiment 2). Rats were exposed to FO on the preceding 0, 2, 7, or 14 d and challenged with 30 min restraint (circles and solid line) or FO (diamonds and dashed line) on the test day. c-fos mRNA expression in the PrL, IL, and VO PFC, layer IV and VI of the primary somatosensory cortex (S1BF, primary sensory barrel field), and the piriform cortex was measured. #, Groups exposed to restraint on test day with significantly greater mean responses to restraint than the FO naïve restrained group; *, groups exposed to FO on test day with significantly lower mean responses than the group exposed to FO for the first time (P < 0.05).

Figure 7.

Effect of previous FO exposure on homotypic and heterotypic stressor-induced c-fos mRNA expression in amygdala and paraventricular thalamic subdivisions (experiment 2). Rats were exposed to FO on the preceding 0, 2, 7, or 14 d and challenged with 30 min restraint (circles and solid line) or FO (diamonds and dashed line) on the test day. c-fos mRNA expression in the MeA, BLA, and CeA nuclei of the amygdala as well as in the anterior and posterior PVThal was measured. #, Groups exposed to restraint on test day with significantly greater mean responses to restraint than the FO naïve restrained group; *, groups exposed to FO on test day with significantly lower mean responses than the group exposed to FO for the first time (P < 0.05).

A main effect for test day challenge on c-fos expression was found in the PrL [F(1, 62) = 69.03, P < 0.0001], IL [F(1, 62) = 58.45, P < 0.0001 PFC], and layers IV [F(1, 62) = 12.23, P < 0.001] and VI [F(1, 62) = 22.26, P < 0.0001] of the barrel cortex as well as in the anterior and posterior PVThal [F(1, 62) = 7.53, P < 0.01; F(1, 62) = 4.18, P < 0.05, respectively]. Main effects for days of FO on c-fos expression was found in the PrL [F(3, 60) = 3.31, P < 0.05] and VO [F(3, 60) = 3.61, P < 0.05 PFC] as well as in the posterior PVThal: [F(3, 60) = 5.23, P < 0.01]. Finally, an interaction between test day challenge and days of FO exposure on c-fos expression was found in the PrL [F(3, 60) = 8.26, P < 0.001]; IL [F(3, 60) = 8.67, P < 0.0001]; VO [F(3, 60) = 7.10, P < 0.001 PFC]; layers IV [F(3, 60) = 4.75 P < 0.01] and VI [F(3, 60) = 3.89, P < 0.05] of the barrel cortex; the piriform cortex [F(3, 59) = 5.42, P < 0.01]; MeA [F(3, 59) = 7.89, P < 0.001]; BLA [F(3, 59) = 3.91, P < 0.05]; and the anterior PVThal [F(3, 60) = 3.27, P < 0.05] (Figs. 6 and 7).

Burial of FO towels

During experiment 1 it was noted that rats displayed burial behavior of FO-scented towels and that this behavior subjectively appeared to increase over time. A video recording of this observed behavior has been included in the supplemental data, published on The Endocrine Society’s Journals Online web site at http://endo.endojournals.org. We directly assessed this phenomenon in experiment 2 by counting the number of burial mounds produced by rats challenged with FO on the test day (d 15). We observed that a higher number of prior FO-exposures led to increased burial behavior of FO-scented towels on test day, although the behavior did not occur to any great extent until the rats had at least 7 prior days of exposure (Fig. 8). A one-way ANOVA based on all rats exposed to FO on test day indicates a significant difference between groups in towel burial behavior [F(3, 28) = 5.35, P < 0.01]. Post hoc tests indicate that rats with 14 d history of prior exposure to FO displayed greater burial behavior than any other group.

Figure 8.

Effect of FO exposure on towel-burying behavior (experiment 2). Rats were exposed to FO on the preceding 0, 2, 7, or 14 d and challenged with 30 min of FO on the test day. Partial burial of neither, one, or both FO towels resulted in a score of 0, 1, or 2, respectively (A). #, Group with significantly greater mean number of burial mounds than group exposed to FO for the first time (P < 0.05). Drawings illustrating examples of nonburied strawberry-scented towels and partially buried FO towels are shown in B and C, respectively.

Discussion

The main finding of this study is that repeated ferret odor exposure in rats led to both neuroendocrine and neural indices of habituation and sensitization, each developing with a distinct time course, suggesting their independence. Furthermore, the widespread regional brain c-fos mRNA reduction and facilitation induced by prior stressor exposure suggest that both adaptive processes influence brain-wide neural processing.

FO as a stressor

Previous studies have shown that rats exposed to FO respond differently than rats exposed to a nonpredator odor (39,41). These responses include increased HPA axis activation and c-fos expression in a wide range of brain regions. However, no other study has directly compared the stress responses elicited by FO with those of other stressors. We found that rats had a lower plasma CORT response (experiments 1 and 2), a trend for a lower plasma ACTH response (experiments 1 and 2), and a lower c-fos mRNA response in the PVN (experiment 2) to their first-time exposure to FO than to restraint. In contrast, first-time FO exposure compared with restraint led to similar c-fos mRNA expression or greater throughout all other brain regions examined. This indicates that whereas FO may not activate the HPA axis as strongly as restraint under these conditions, it elicits a comparable or stronger response in many brain regions. The greater c-fos mRNA response to FO was especially pronounced in ventral orbital prefrontal cortex, piriform cortex and the amygdala.

HPA axis habituation and sensitization

Rats repeatedly exposed to FO demonstrated rapid habituation within the HPA axis to homotypic stressor challenge. The CORT response decreased significantly by the eighth exposure to FO. This is similar to other paradigms using repeated FO exposures (39), although a lack of neuroendocrine habituation can also be observed under different experimental conditions (38). Interestingly, there was an even more rapid habituation of the c-fos mRNA response in the PVN than elsewhere in the HPA axis, which was evident by the third exposure to FO. In our study, although stress-induced ACTH levels declined with repeated FO exposure in experiment 1, they did not in experiment Two. This may be explained by a floor effect of ACTH levels in the second experiment, possibly due to a missed peak response that occurred before 30 min after FO presentation.

Whereas habituation of the HPA axis response to FO was evident with repeated FO exposure, there was clear evidence for a chronic effect of ferret odor on physical measures as well as sensitization of the HPA axis response to challenge with a heterotypic stressor (restraint) after the same repeated FO exposure. For instance, chronic FO exposure led to weight loss and thymic involution, physiological changes associated with chronic stress (48) and observed previously with repeated FO exposure (39). However, adrenal weights were unaffected by repeated FO exposure, in contrast to a previous study (39). This, along with the observation that CORT response to FO was modest initially and minimal by d 8 of repeated exposure, suggests that repeated FO may not produce chronic elevations in HPA axis activity. Consequently, the FO-induced development of habituation, sensitization, decreased body weight, and thymic involution may be due to processes other than hypersecretion of glucocorticoids.

Importantly, 2 wk exposure to FO led to increased restraint-induced c-fos expression in the PVN (experiment 1, trend in experiment 2); increased plasma ACTH (trend in experiment 1, significant between d 3 and 15 in experiment 2); and increased plasma CORT secretion (experiment 1). It is possible that in the second experiment, plasma CORT levels reached their physiological maximum by restraint alone, masking any additional influence chronic FO may have had on HPA axis activity. Alternatively, it is possible that, due to changes in the towel source, there was a weakening of FO potency across the two studies, leading to the partial loss of HPA axis sensitization in the second experiment. Nevertheless, taken together there appears to be sensitization evident at the level of the HPA axis hormone response, and this sensitization was generally not evident until after 2 wk of FO exposure. These results support the hypothesis of distinct processes underlying habituation and sensitization.

Habituation and sensitization of c-fos gene expression in the forebrain

Many studies have reported habituation of IEG expression throughout much of the neuroaxis (e.g. Refs. 24,25,26,27,28). A key finding in this study is sensitization of c-fos expression to heterotypic stimuli challenge across multiple forebrain regions. That sensitization of c-fos was evident at many extrahypothalamic loci in the brain including a region of primary sensory cortex attests to the widespread expression of this sensitized response. Only a few other studies have reported sensitization of IEG expression after challenge of previously stressed rats with either a homotypic (33) or heterotypic stressor (9,11). Our study uniquely contributes to the stress adaptation literature by demonstrating both habituation and sensitization of neural and HPA axis responses in rats receiving the same repeated stress regimen. Importantly, these effects were observed in the same brain regions, although it is impossible to determine whether the IEGs were in the exact same regional neuronal populations.

This is the first study to our knowledge to examine PFC IEG expression in rats expressing a sensitized HPA axis response. Sensitization of IEG expression in the PFC in rats displaying HPA axis sensitization was surprising, given the putative suppressive influence of the PFC on HPA axis activity. Several studies (e.g. Refs. 49,50,51,52) found that lesions of select subregions of the medial PFC (anterior cingulate, PrL, and IL) of rats leads to increased HPA axis responses to acute stress, suggesting that the PFC can exert a suppressive influence on stress-induced HPA axis activity. It is possible that the overall c-fos mRNA expression levels in the PFC seen in this study do not reflect the relative levels within specific pyramidal cells that may mediate a net inhibitory effect on the HPA axis.

A striking finding of this study is that there was widespread cortical and subcortical expression of neural response habituation and sensitization and that both forms of adaptation were manifest in the same neural structures that included regions of primary sensory cortex. It is unlikely, however, that the site of both adaptive processes was widely distributed throughout these various brain structures. For such to be the case, each of these structures would have to have the ability to independently discriminate between the neural input associated with the homotypic and heterotypic stressor stimuli. It seems more likely that some central brain region/network that receives sufficient multimodal sensory input necessary to discriminate between homotypic and heterotypic stressor stimuli is able to modulate the relative level of neural responses throughout the brain according to the current environmental context. The prefrontal cortex is a candidate brain region that may be able to execute a stimulus-dependent bidirectional influence on the magnitude of the system-wide neural response to specific stimuli. This system-wide modulation could be achieved by alteration of sensory gating at the thalamic level (53,54), or alternatively by alteration of the tonic activity of brain stem monoamine neuronal activity [e.g. modulation of activity of locus coeruleus noradrenergic neurons (55)]. Given this possibility, it is important to note that the prefrontal cortex, like the other forebrain regions examined, exhibited both habituation and sensitization of c-fos mRNA responses. However, the correlative nature of our immediate early gene expression pattern data does not allow for determination of whether a particular brain region is a site of adaptation and/or a mediator of the expression of adaptation within other brain regions.

The posterior PVThal, a brain region that has been shown to modulate HPA axis response habituation (42), also displayed robust habituation, although not sensitization of c-fos expression in rats repeatedly challenged with FO. The PVThal has direct neural connections with the PFC and amygdala and indirect connections with the PVN (56) and consequently is another brain region in addition to PFC that may be a good candidate for mediating widespread neural adaptation.

Taken together with lesion-based studies (e.g. Ref. 42), it appears that brain regions underlying the expression of habituation or sensitization of c-fos mRNA evident elsewhere in the brain exhibit adapted immediate early gene expression patterns themselves.

The development of habituation and sensitization follow different time courses

We observed (experiment 2) that the temporal progression for the expression of c-fos mRNA response habituation was different from that of response sensitization. Specifically, habituation of the c-fos mRNA response throughout the brain was fully evident by the third exposure to FO, with little additional habituation apparent beyond that point, whereas in many brain regions (PFC, barrel cortex, CeA, and PVN), sensitization of the c-fos mRNA response developed as a more linear function of prior FO exposure across the 2-wk period. This is strongly suggestive of separate underlying mechanisms responsible for these particular habituation and sensitization phenomena. For many years researchers have held that habituation and sensitization of motor responses to sensory stimuli depend on independent processes. Groves and Thompson (57), for example, published theory and supporting data proposing that a stimulus elicits both habituating and sensitizing neural adaptations that influences subsequent behavior. Interestingly, for the stimulus-response systems that they characterized, sensitization of responses were evident initially but dissipated with repeated exposure to a stimulus. Our work suggests something very different: a gradual development of sensitization over repeated stressor exposure.

Our study did not examine the extent to which the expression of sensitization may persist after it has been established or whether the development of sensitization requires repeated exposure to FO or simply time to develop irrespective of further stress experiences. Several studies from Armario’s group (6,58,59,60) suggested that habituation is a progressive process that does not necessarily require repeated homotypic stimulus exposure to develop. Similar to work of Armario et al., there is research suggesting that sensitization may develop over time independent of a repeated stress experience (53,54,55), and there is some evidence that this can occur with ferret odor exposure (see Ref. 38: experiment 1).

Increase in burial behavior

It was noted in experiment 1 that rats presented with FO in their home cage pushed their bedding around the hanging towel with their snouts or paws in a burial-like manner. This behavior was specific to the FO and not the strawberry odor stimulus. It appeared that the burial behavior increased across FO exposures, both at the individual and group levels. Based on these observations, we monitored burial behavior directly in experiment 2. We found that rats with greater prior experience with the FO stimulus showed on average a greater tendency to bury the towels on test day. This could be labeled behavioral sensitization. The act of burial could be considered an expression of anxiety, and the tendency toward increases in this behavior, without an increase in threat or stimulus potency may indicate a form of pathology. This pathology may relate to that seen in patients expressing generalized anxiety, a disorder that has been connected with abnormal activity patterns in the prefrontal cortex (61). On the other hand, defensive burial behavior may also be considered an active coping response (62), potentially leading to central nervous system adaptation to chronic stress (63,64). In the present study, neuronal and endocrine measures suggest that repeated exposure to FO leads to widespread expression of FO-selective habituation processes in rats. This could be due to reduced attention paid to the FO stimulus by the rat. However, rats with greater previous exposure to FO increased their burial tendencies, suggesting that, rather than reducing, rats increased their attention to the FO stimulus across days. Curiously, this increased behavioral manifestation was not reflected in a concurrent increase in any regional c-fos mRNA induction studied. It remains to be determined whether the burial behavior is directly or indirectly related to the expression of habituation or other sensitization phenomena observed.

Conclusions

Ferret odor is a negative-valence stimulus, producing similar activation or greater of cortical brain structures and the amygdala compared with restraint, and producing distinct escape and fear behaviors, whereas leading to only modest activation of the HPA axis in rats in these studies. Repeated exposure to FO leads to body weight loss, thymic involution, and increases in selective behaviors but does not induce adrenal hypertrophy. Rats express widespread habituation of c-fos expression throughout the brain in as few as three repeated homotypic stimulus presentations of FO, with little additional habituation apparent beyond this point. Repeated exposure to FO also leads to a more protracted linearly increasing sensitization of the c-fos mRNA response to restraint, manifest over multiple brain regions, including primary sensory cortices. Habituation and sensitization are expressed across many of the same brain regions, implying opposing adaptive processes manifest within the same widespread network; however, based on their independent temporal development, habituation, and sensitization are nonetheless likely to be independent processes.