Role of the posterodorsal medial amygdala in predator odour stress‐induced puberty delay in female rats

Abstract Puberty onset is influenced by various factors, including psychosocial stress. The present study investigated cat‐odour stress on puberty onset and oestrous cyclicity in rats. Female weanling rats were exposed to either soiled cat litter or fresh unused litter for 10 consecutive days. Following vaginal opening (VO), rats were smeared for 14 days to determine oestrous cyclicity. Anxiety‐like behaviour was assessed using standard anxiety tests. Brains were collected to determine corticotrophin‐releasing factor (CRF), CRF receptor 1 (CRF‐R1) and CRF receptor 2 (CRF‐R2) mRNA in the paraventricular nucleus (PVN), as well as the central nucleus of the amygdala (CEA) and the medial nucleus of the amygdala (MEA). Cat odour delayed VO and first oestrus, disrupted oestrous cycles and caused anxiogenic responses. Cat odour elicited increased CRF mRNA expression in the PVN but not in the CeA. CRF‐R1 and CRF‐R2 mRNA levels in the PVN and CeA were unaffected by cat odour; however, CRF‐R1 mRNA levels were decreased in the MeA. The role of CRF signalling in the MeA, particularly its posterodorsal subnucleus (MePD), with respect to pubertal timing was directly examined by unilateral intra‐MePD administration of CRF (0.2 nmol day‐1 for 14 days) via an osmotic mini‐pump from postnatal day 24 and was shown to delay VO and first oestrus. These data suggest that CRF signalling in the MePD may be associated with predator odour‐induced puberty delay.

amygdala. The amygdala is composed of a complex of nuclei including the central nucleus of the amygdala (CEA) and the medial nucleus of the amygdala (MEA). Numerous CRF neurones are found in the CeA and CRF-receptor (R)1 and CRF-R2 are located both in the CeA and MeA. Under chronic stress conditions, CRF expression in the amygdala is increased. 7 There is also considerable evidence to suggest that the amygdala facilitates the stress response. 8 Besides being involved in the stress response, both CRF and the amygdala are also importantly involved in fear and anxiety. 9,10 Therefore, CRF in the amygdala may play a role in disorders linked to both excessive stress, fear and anxiety, such as post-traumatic stress disorder (PTSD). 10 Chronic stress impairs reproductive function, which is a process that involves the CeA and MeA. 11 Indeed, there is a differential modulatory role for these two amygdala nuclei in terms of stress-induced suppression of pulsatile luteinising hormone (LH) secretion, with the MeA being more responsive to psychological stress and the CeA to physiological stress. 11 Recent in vivo electrophysiological studies in rats have demonstrated that the MeA had the highest frequency of response and distinct firing pattern to predator urine compared to other amygdaloid nuclei, 12 which confirms the importance of this neurocircuitry in psychological stress. Moreover, lenti-viral overexpression of CRF in the CeA in pre-pubertal female rats, mimicking the effect of chronic stress, disrupts reproductive function and leads to irregular oestrous cycles, increased anxiety and advanced puberty. 2,13 It is therefore reasonable to hypothesise that CRF signalling in the amygdala may be a key player in cross-talk between stress-induced activation of the HPA axis and suppression of the HPG axis.
The MeA, particularly its posterodorsal subnucleus (MePD) receives olfactory information from environmental and sexual cues, 14,15 which are relayed to the medial preoptic area (mPOA) where the gonadotrophin-releasing hormone (GnRH) neurones are located to bring about the appropriate reproductive response. The MePD also undergoes dramatic anatomical changes during puberty 16,17 and exerts a regulatory role on puberty timing, as indicated by studies in rats showing that MePD lesions caused puberty advancement, 18 whereas its electrical stimulation delayed puberty onset. 19 In addition, the MePD has been implicated in ovulation and oestrous cyclicity, as well as gonadotrophic hormone secretion. 18,20 In the present study, we tested the hypothesis that predator odour-induced stress would delay puberty timing and alter oestrous cyclicity. We also assessed both social and anxiety-like behaviour as proxy measures of psychological stress in predator odour-stressed prepubertal female rats. Finally, we determined the role of CRF signalling in the MeA and, more specifically, the MePD with respect to mediating changes in pubertal timing.

| Animals, predator odour stress and puberty evaluation
All procedures were conducted in accordance with the United Kingdom Home Office Animals (Scientific Procedures) Act 1986. and housed under a 12:12 hours light/dark cycle (lights on 7.00 am) at 22 ± 2°C with food and water available ad lib. Litters were reduced to eight to 12 pups on postnatal day (PND) 1 (birth, day 0). They were weaned on PND 21 and housed in groups of three or four female rats per cage and weighed every 3 days. To determine whether the timing of puberty could be affected by a predator odour stress, female pups were randomly assigned to either a treatment group (n = 12) or control group (n = 10) after weaning. For the predator odour stress (cat odour-treated), an individual rat was placed in isolator animal cage (37 × 26 × 17 cm) in which a well-soiled cat litter (in use by the cat for 3 days) was sifted to remove clumps of faeces and placed in a dish (diameter 5.5 cm) in the middle of the cage. The duration of each exposure was 15 min, after which the rat was returned to the home cage, with this procedure being repeated for 10 consecutive days  to model the effect of chronic stress. Sham exposures were conducted by placing the rat in an identical cage (used only for the control treatment) but with fresh, unused cat litter for the same amount of time. All exposures took place between 10.00 am and 12.00 pm each day.
From PND 28, the pre-pubertal female rats were monitored daily for vaginal opening. Once vaginal opening occurred, vaginal smears were taken daily for 14 consecutive days to detect first oestrus as a marker of puberty onset and the stage of the oestrous cycle. Normal oestrous cyclicity was defined as having at least two consecutive normal cycles, which lasts for 4-5 days with 1-2 days of oestrus.
Cycle length, with three stages observed in the correct order, was determined by the number of days between each occurrence of oestrus. 21

| Tissue collection and reverse transcriptasepolymerase chain reaction (RT-PCR)
To examine the effect of predator stress on levels of CRF, CRF receptor 1 (CRF-R1) and CRF receptor 2 (CRF-R2) mRNA in the PVN and amygdala (CeA and MeA), predator odour stressed and control animals were killed by decapitation 14 days post vaginal opening. The brains were then removed, snap frozen on dry ice and stored at −80°C until the brains were cut into coronal sections (thickness of 300 μm) using a cryostat (Bright Instrument Co., Luton, UK). Using the micropunch method of Palkovits, 22 bilateral punches (0.5 mm in diameter) of the CeA (bregma −2.12 to −3.14 mm), MeA (bregma −2.80 to −3.60) and PVN (bregma −0.70 to −2.10 mm) were taken according to the rat brain atlas of Paxinos and Watson. 23 For confirmation of correct punch position, sections were fixed with formalin, stained in crystal violet solution and observed under a microscope to confirm whether the specific brain regions had been correctly punched.
Total RNA was extracted from the punched tissues for each Master SYBR Green I mix, 0.8 μL of MgCl 2 and 3.7 of μL water were added together to give a total reaction volume of 8 μL. The PCR for CRF, CRF-R1 and CRF-R2 was performed under the cycling conditions: Initial denaturation and activation at 95°C for 5 minutes, followed by 40 cycles of denaturation at 95°C for 10 seconds and annealing at 56 and 72°C for 10 seconds. The reaction conditions for HPRT1 were: 5 minutes at 95°C for one cycle, then 40 cycles of 10 seconds at 95°C, 10 seconds at 56°C and 10 seconds at 72°C.
The relative expression of CRF, CRF-R1 and CRF-R2 mRNA in control and cat odour groups was determined by the threshold cycle (Ct) value in the exponential phase of the PCR reaction and normalised to the respective HPRT1 Ct value using the comparative Ct method. 4

| Evaluation of anxiety-like behaviour
To determine anxiety-like behaviour of the predator odour stressed and control rats, the elevated plus maze (EPM) and social interaction tests were performed as described by Poulin et al 24

| Elevated plus maze
The EPM tests were performed at 7-8 day intervals from PND 21-43

| Social interaction
Before starting the social interaction test, animals were habituated twice (one time per day, 2 days before the day of testing) to the behavioural room for 15 minutes and then to the social interaction arena for 5 minutes on each occasion. The social interaction test was carried out on PND 34, which was 14 days after the commencement of predator odour exposure. Female conspecifics of the same age as the test rats were used for the social interaction test. The time spent on the behavioural events comprising sniffing, chasing, following, grooming and mounting was recorded over a period of 15 minutes by two observers who were blind to the treatment. The tests were carried out between 9.00 am and 11.00 am with the pairs of unfamiliar rats, weight-matched and placed simultaneously in the arena (50 × 50 × 50 cm). The total social interaction time was calculated by the sum of the time spent in the active social behaviors noted above.
A timeline of this experimental design is shown in Figure 1.

| MePD osmotic mini-pump implantation, CRF administration and pubertal timing
All surgeries on the prepubertal rats were carried out under a combination of ketamine (Vetalar, 600 mg/kg, i.p.; Pfizer, New York, NY,
For this part of the study, a separate group of female rats was chronically implanted unilaterally with 28-gauge brain cannula to the cannula with silicone tubing and implanted s.c. in the interscapular space. In PND 24 rats, CRF (0.2 nmol) dissolved in aCSF was delivered at the rate of 6 μL day -1 or aCSF at 6 μL day -1 for a 14-day period. After the surgical procedure, the rats were allowed to recover on a heated pad until fully conscious. A further group on naïve PND 24 rats (n = 8) served as non-surgical controls. Vaginal opening was monitored and vaginal smears were taken as described above.

| Histological verification of cannula tip position for MePD osmotic mini-pump implantation
The animals implanted with osmotic mini-pumps were culled once tomical landmarks with reference guide from the rat brain atlas. 23 Animals with the cannula outside the MePD were designated as 'misplaced' and analysed as a separate group in an attempt to determine the specificity of CRF activity to the MePD.

| Statistical analysis
Comparisons between control and cat odour-treated rats with respect to puberty onset (indicated by day of vaginal opening and first oestrus), cycle length, the first EPM trial on PND 21 and social interaction behaviours were made by subjecting data to Student's t test, whereas body weight gain and data from the EPM were analysed by two-way, comparison with control, aCSF and cannula misplacement groups using a one-way ANOVA, whereas body weight gain was analysed by two-way, repeated measures ANOVA with a between-subjects factor and treatments as within-subject factor. The results are presented as the mean ± SEM. P < 0.05 was considered statistically significant for differences between groups. All statistical analyses were performed using sigmaplot (Systat Software Inc., Chicago, IL, USA).  Figure 2C).

| Effect of predator odour on puberty onset and body weight
F I G U R E 1 Timeline of the experimental design. Weanling female rats were exposed to soiled cat litter for 10 consecutive days from postnatal day (PND) 21-30 followed by daily monitoring of vaginal cytology from PND 28-42 to determine puberty onset and oestrous cyclicity. Rats were tested on the elevated plus maze (EPM) and social interaction was assessed as indicated on the timeline to determine anxiety-like behaviour. The experiment was terminated on PND 44 and brains were collected for gene expression studies  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37 38  39  40  41 42  43  44 Terminated & brain collected Daily exposure to soiled cat litter/unused litter Daily vaginal monitoring/cytology

| Effects of predator odour stress on oestrous cyclicity
After vaginal opening, oestrous cycle stage was assessed daily for  Figure 3C), which was significantly shorter than the mean oestrous cycle length of 5.7 ± 0.4 days (P < 0.05; n = 11) ( Figure 3C) observed in cat odour-treated animals.

| Effect of cat odour stress on CRF and CRF receptor expression in the PVN, CeA and MeA
Relative CRF, CRF-R1 and CRF-R2 mRNA levels were measured and compared between control and cat odour-treated groups in the PVN, CeA and MeA. Cat odour-treated animals showed a significant increase in relative mRNA expression of CRF in the PVN compared to control animals (P < 0.05; n = 10 for both) ( Figure 4A)

| Anxiety-like behaviour
To assess anxiety and fear-like behaviour in response to cat odour exposure, standard anxiety and fear tests were performed using Quantification of mRNA levels was carried out using a reverse transcriptase-polymerase chain reaction from brain micropunch samples of the PVN, CeA and MeA. CRF mRNA levels were significantly higher in the cat odour-treated group than in the control group in the PVN (P < 0.05) but not in the CeA (A). CRF-R1 mRNA levels were significantly lower in the MeA of cat odourtreated rats compared to control rats (P < 0.05), although there were no changes in the PVN or CeA (B). CRF-R2 mRNA levels in the PVN, CeA or MeA were comparable between the two experimental groups (C). Data are presented as relative levels compared to HPRT1 expression (mean ± SEM). *P < 0.05 vs control (n = 8-11 per group)

| Social interaction
A social interaction test was performed once on PND 34. The time spent in cumulative social interactive behaviour (chasing, sniffing, grooming, following and mounting) was significantly decreased in cat odour-treated animals (31.6 ± 3.38 s; n = 11) compared to control animals (75.25 ± 11.58 seconds; n = 10; P < 0.05) ( Figure 5B).

| Effect of intra-MePD administration of CRF on the timing of puberty
Cannulae placement in the MePD was verified in coronal brain sections. Of the 12 rats that received CRF, three had cannulae that were not located in the MePD (Figure 6). These animals with misplaced cannulae were separately analysed as the "misplaced" group.

| D ISCUSS I ON
The results of the present study demonstrate that chronic exposure to cat odour delays puberty onset, disrupts oestrous cyclicity and elicits fear responses in female Sprague-Dawley rats. These phenotypic and behavioural outcomes were accompanied by increased CRF expression in the PVN and decreased CRF-R1 mRNA levels in the MeA. Furthermore, intra-MePD administration of CRF mimics the same changes to pubertal timing and oestrous cyclicity as those elicited by predator odour exposure.
Predator odour stress is a model for PTSD, causing physiological and behavioural changes similar to those seen in patients with the disorder, including increased anxiety and fear, as well as decreased social interaction. 26 Predator stress also induces synaptic plasticity in brain regions involved in fear and defensive behaviour such as the amygdala. 27 In the present study, the manifestation of increased fear and decreased sociability in rats exposed to predator stress is consistent with earlier reports in rats and corroborates key features of PTSD. 28 The timing of trauma exposure relative to puberty has been shown to confer increased vulnerability to anxiety in humans, 29 which is supported by evidence of increased stress-induced c-Fos The amygdala mediates fear and defensive behaviour in rodents in response to predator odour. 27 The MeA is the first site of convergence of olfactory and vomeronasal information 32 and exhibits a robust response to cat odour. 12 However, it is not clear how predator odour signals activate the HPA axis. In response to psychological stress, the MeA is considered to exert an activating influence on HPA function, 33 which is supported by attenuation of stress-induced PVN activation and adrenocorticotrophic hormone (ACTH) secretion in rats with MeA lesions. 34 The vast majority of neural projections from the MeA to the PVN are GABAergic, which are assumed to promote CRF release via disinhibition of GABAergic inter-neurones directly presynaptic to the CRF cell populations in the PVN. 8 Specifically, chemogenetic activation of MeA neurones has been shown to increase both c-Fos expression in CRF PVN neurones and plasma levels of ACTH. 35 It is plausible that the excitatory effect of predator odour on the MeA 12 promotes HPA activation by increased CRF expression in the PVN, as observed in the present study. Indeed, up-regulation of CRF mRNA expression in the PVN is a common outcome in adult rats exposed to cat odour and may occur without any corresponding change to corticosterone levels, 36 implying a lack of feedback signal to control HPA output. The lack of effect of cat odour on CRF or CRF receptor expression in the CeA is commensurate with the minimal response of the CeA to predator odour. 12 Indeed, markers of neuronal activation were unchanged in the CeA of rats exposed to predator odour 37 and CeA lesions did not alter cat odour-induced fear responses, 38 indicating that the CeA may not be critically involved in coordinating neuroendocrine responses to predator odour. However, with restraint stress, a rapid and sustained CRF release in the CeA has been reported. 33 CRF-R1 and CRF-R2 were unchanged in the PVN and CeA, although, in the MeA, CRF-R1 was significantly down-regulated MePD in rats exposed to cat odour, which may be suggestive of an enhanced CRF signalling in the MeA. CRF is known to down-regulate CRF-R1 expression in the prefrontal cortex in response to unpredictable stress. 39 Similarly, decreased CRF-R1 expression has been observed in the frontal cortex of suicide victims suffering from depression, 40 potentially caused by chronic elevated levels of CRF in the CSF. 41 In the present study, the origin of the potential CRF inputs to the MeA are unknown. Although we observed a lack of change in CRF expression in the CeA of the cat odourexposed rats, this may reflect a situation where the endogenous CRF released in the amygdala has become habituated to chronic stress as a protective mechanism against CRF over-expression but, Further studies are required to establish the source of CRF input onto the MeA.
The phenotypic sign of delayed puberty as a result of chronic cat odour exposure may be related to changes in the neural circuits that govern puberty timing vis-à-vis the influence of the stress neuropeptide, CRF, which negatively influences the HPG axis. 45 Of note, the MeA plays an important role in puberty and its MePD subnucleus is particularly involved in regulating puberty timing. 18 Neural activity within the MeA between PND 20 and 30 is crucial with respect to determining the timing of puberty, with increased MeA activity being related to delayed puberty onset. 46 The female rats in the present study were exposed to cat odour during this critical window and, because cat odour increases neuronal into the MePD delayed puberty onset in female rats. Thus, CRF signalling specifically in the MePD is critical for timing of puberty, and further supports the important role played by this amygdaloid nucleus in controlling puberty onset. 18 The disruption in oestrous cyclicity observed in the present study agrees with previous studies using chronic mild stress 47 and may be attributed to the enhanced CRF expression in the PVN, although this contrasts with the expectation of a minimal role for PVN CRF in stress-induced reproductive dysfunction. 5 Nevertheless, CRF exerts a suppressive effect on GnRH/LH secretion 48 that may involve GABAergic inputs to GnRH neurones. 49 Over-expression of CRF in the CeA, comprising a model of chronic stress in rats, resulted in irregular oestrous cyclicity, as well as suppression of the preovulatory LH surge, 2 although the mechanisms underlying ovulatory dysfunction are unclear because there are few projections from the CeA to the hypothalamic reproductive nuclei controlling ovulation. 50 Stressful stimuli may also recruit the MeA to suppress the HPG axis. 11 The presence of CRF receptors in the MeA 51 and direct synaptic connection between MeA GABAergic neurones and GnRH neurones in the mPOA 52 might suggest that CRF signalling in the MeA could underlie the modulatory effect of the MeA on the ovulatory processes 18 by disrupting the GnRH neural network. Whether the long-term impact of exposure to cat odour on ovarian cyclicity is mediated by similar mechanisms remains to be established.
The results of the EPN and social interaction tests demonstrate the severity of cat-odour exposure on anxiety-like behaviour in female rats. CRF-R1 activation is associated with the initiation and consolidation of the long-term effects on anxiety-like behaviours caused by exposure to predator stress, 27 whereas CRF-R2 is more commonly related to modulating the recovery response. 53 Although CRF-R2 mRNA expression was unchanged in the brain nuclei examined, which may be part of a 'stress coping' mechanism, the reduced expression of CRF-R1 in the MeA may underlie a homeostatic mechanism to curtail an enhanced fear response that could be damaging to physiological processes. Remarkably, earlier studies have shown that MeA lesions attenuate anxiety-like behaviour after predator stress. 54 Therefore, CRF-R1 activity in the MeA may be implicated in predator exposure-induced fear in female rats to support a synchrony between anxiety-like behaviour and altered reproductive function under stressful conditions.
A limitation of the present study was the absence of using neuropharmacological CRF receptor antagonism to reverse the negative effect of cat odour exposure on pubertal timing. Given the technical challenge of the required surgical manipulation in small animals, our planned future experiments involving a chemogenetic approach should successfully address this limitation. In addition, the small size of the punched brain tissues hampered the possibility of a separate confirmatory assay to determine protein levels of CRF and its receptors in the rats exposed to cat odour, as well as those infused with CRF, which would have strengthened the present data set.
In conclusion, the results obtained in the present study suggest a direct involvement of CRF signalling specifically in the MePD in puberty timing that may be critical for predator odour-induced pubertal delay. We are also indebted to Dr Cathy Fernandes, King's College London, for her invaluable help and advice regarding our behavioural studies.

CO N FLI C T O F I NTE R E S T S
The authors declare that they have no conflicts of interest.