Address correspondence and reprint requests to Mariangela Serra, Department of Life and Environmental Sciences, Section of Neuroscience, University of Cagliari, Cagliari 09100, Italy. E-mail: firstname.lastname@example.org
Social isolation in male rats at weaning results in reduced basal levels of the neuroactive steroid 3α,5α-tetrahydroprogesterone (3α,5α-TH PROG) in the brain and plasma as well as increased anxiety-like behavior. We now show that socially isolated female rats also manifest a reduced basal cerebrocortical concentration of 3α,5α-TH PROG as well as an anxiety-like profile in the elevated plus-maze and Vogel conflict tests compared with group-housed controls. In contrast, despite the fact that they were raised under normal conditions, adult male offspring of male and female rats subjected to social isolation before mating exhibited an increased basal cerebrocortical level of 3α,5α-TH PROG but no difference in emotional reactivity compared with the offspring of group-housed parents. These animals also showed an increased basal activity of the hypothalamic-pituitary-adrenal axis as well as reduced abundance of corticotropin-releasing factor in the hypothalamus and of corticotropin-releasing factor receptor type 1 in the pituitary. Moreover, negative feedback regulation of hypothalamic-pituitary-adrenal axis activity by glucocorticoid was enhanced in association with up-regulation of glucocorticoid receptor expression in the hippocampus. There was also attenuation of corticosterone release induced by foot-shock stress in the offspring of socially isolated parents. The increase in the brain concentration of 3α,5α-TH PROG induced by acute stress was also blunted in these animals. Our results thus show that a stressful experience before mating can influence neuroendocrine signaling in the next generation.
Early life experiences are thought to have a profound impact on development and maturation of the CNS, including the induction of persistent plasticity of the neuroendocrine stress system (Levine 1967; Heim and Nemeroff 2002). Indeed, exposure to stress hormones during the perinatal period is thought to be a determinant of interindividual differences in emotionality, cognitive function, and stress responsiveness in adult animals and humans (Heim and Nemeroff 2001; Lupien et al. 2009; Veenema 2009). The developmental timing and context of stressful stimuli are important, however, in determining the adaptive or maladaptive consequences. Exposure to pre-natal stress has programming effects on the hypothalamic-pituitary-adrenal (HPA) axis and the brain as a result of the increase in maternal glucocorticoid secretion (Cottrell and Seckl 2009). Given the key role of glucocorticoids in brain maturation, elevated levels impair brain development and function and can render individuals more vulnerable to stress (Seckl 2008).
In spite of the many studies that have examined the effects of perinatal stress on the adult brain and emotionality, few have addressed the effects of stress experienced by parents in early life on the emotionality and stress responsiveness of offspring. Human studies have described intergeneration transmission of behavior (Harper 2005; Kim et al. 2009), and animal studies have shown transgenerational effects of environmental experiences (Anway et al. 2005; Champagne 2008; Curley et al. 2008). This type of inheritance is thought to represent a means through which adaptive changes in one generation can be transmitted across generations as an essential and dynamic component of evolution (Badyaev and Uller 2009). Recent studies have shown that stress vulnerability can influence behavior, social abilities, and serotonergic function in offspring across two generations, suggesting the possibility that exposure to environmental stress may produce epigenetic alterations (via DNA methylation) in germ cells that might then affect the phenotype of offspring (Franklin et al. 2010, 2011).
Social isolation is a model of prolonged mild stress and is associated with marked behavioral alterations—such as increased locomotor activity, anxiety, depression, and aggression in laboratory animals (Fone and Porkess 2008). In male rats, social isolation results in a decrease in the brain and plasma concentrations of neuroactive steroids that act as positive modulators at GABAA receptors (Majewska et al. 1986) and is accompanied by an enhanced response to the acute administration of ethanol as well as by enhanced neurosteroidogenesis in response to acute stressful stimuli (Serra et al. 2000, 2003). It also increases the sensitivity of the pituitary to corticotropin-releasing factor (CRF) and impairs negative feedback regulation of the HPA axis (Serra et al. 2005).
We have now investigated whether environmental influences in early life affect the stress phenotype of the next generation, in particular whether there is a transgenerational transmission of functional changes induced by chronic stress that alters the basal activity and reactivity of the HPA axis. Our previous studies on the effects of social isolation on emotional and neuroendocrine states were performed with male rats. We therefore first examined the effects of social isolation on these parameters in females. In the offspring of socially isolated parents, we then evaluated basal emotional state and measured the brain concentration of the neuroactive steroid 3α,5α-tetrahydroprogesterone (3α,5α-TH PROG). Basal activity of the HPA axis was evaluated by measurement of the plasma concentrations of adrenocorticotropic hormone (ACTH) and corticosterone as well as the expression of CRF in the hypothalamus, the type 1 receptor for CRF (CRFR1) in the pituitary, and the glucocorticoid receptor (GR) in the hippocampus. To evaluate stress reactivity, we determined the effect of acute foot-shock stress on the plasma corticosterone level as well as the efficiency of negative feedback regulation of the HPA axis with the dexamethasone suppression test. All the experiments were performed in male offspring to compare the results with our previous studies in socially isolated males.
Materials and methods
Sprague–Dawley CD rats at 25–30 days of age, immediately after weaning, were housed for 30 days either in groups of six to eight per cage (59 by 38 by 20 cm) or individually in smaller cages (42 by 26 by 15 cm). They were maintained under an artificial 12-h light, 12-h dark cycle (light on 0800–2000 hours) at a constant temperature of 23 ± 2°C and 65% humidity. Food and water were freely available until the time of experiments. After the 30-day housing period, socially isolated and group-housed females were paired for 7 days with socially isolated and group-housed males, respectively. Male pups were weaned between 25 and 30 days after birth and then housed in groups of six to eight per cage until 2 months of age (Fig. 1). Animal care and handling throughout the experimental procedures were in accordance with the European Communities Council Directive of 24 November 1986 (86/609/EEC).
Elevated plus-maze test
The plus-maze was constructed of black polyvinyl chloride and contained two open and two closed arms (12 by 60 by 3 cm) mounted 50 cm above the floor. The arms were connected by a central square (12 by 12 by 3 cm). The apparatus was located in a quiet, dimly lit room. Each rat was tested only once. The animal was placed in the central square facing a closed arm and its behavior was scored over 5 min. The number of entries into and time spent in the open and closed arms were recorded, with arm entry being defined as the presence of all four feet of the animal in the arm. The maze was cleaned after each trial.
Vogel's conflict test
Rats were deprived of water for 24 h before the conflict session. The test (3 min) was performed as previously described (Serra et al. 2000). In brief, rats were placed in a clear Plexiglas box (20 by 28 by 20 cm) with a stainless-steel grid floor, and the box was enclosed in a sound-attenuated and ventilated chamber (Lafayette Instruments, Lafayette, IN, USA). Water was provided through a stainless-steel drinking tube that extended 1 cm into the box, 3 cm above the floor. The drinking tube and the grid floor were connected to a constant-current shock generator and to a ‘drinkometer.’ The shock generator delivered one shock (0.3 mA for 0.5 s) for each cumulative period of 15 licks, termed a ‘licking period.’ Before the test, rats were habituated to the test box for 5 min; the drinking tube was then inserted and the animals were allowed to lick for three licking periods (training period) before the onset of punishment.
Spontaneous locomotor activity
The test apparatus consisted of a Plexiglas chamber (42 by 42 by 30 cm) equipped with Digiscan activity monitors (Omnitech Electronics, Columbus, OH, USA) that detect interruptions of 16 photobeams spaced 2.5 cm apart and 2.5 cm above the floor. Behavior was recorded for 10 min.
Foot shock consisted of a series of electrical impulses delivered in a box with a floor consisting of brass rods spaced 2 cm apart. Shocks (0.2 mA for 500 ms) were delivered every second over a period of 5 min. Animals were killed 25 min after the end of the foot-shock treatment.
Rats were injected intraperitoneally with dexamethasone (500 μg per kilogram of body mass) or physiological saline as vehicle and were killed 150 min later.
Extraction and assay of steroids
Rats were killed either with a guillotine (for measurement of plasma corticosterone) or by focused microwave irradiation (70 W/cm2 for 4 s) of the head (for measurement of brain 3α,5α-TH PROG). The cerebral cortices were dissected and then frozen at –80°C until steroid extraction. 3α,5α-TH PROG was extracted and purified as previously described (Serra et al. 2000). The extract residue was dissolved in 5 mL of n-hexane and applied to a SepPak silica cartridge (Waters, Milford, MA, USA), and eluted components were separated and further purified by HPLC on a 5-μm Lichrosorb-diol column (250 by 4 mm; Phenomenex, Torrance, CA, USA) with a discontinuous gradient of 2-propanol (0–30%) in n-hexane. The recovery (70–80%) of 3α,5α-TH PROG through the extraction and purification procedures was monitored by addition of a trace amount (6000–8000 cpm; 20–80 Ci/mmol) of 3H-labeled standard to the brain homogenate. 3α,5α-TH PROG was quantified by radioimmunoassay with specific antibodies generated in sheep (Purdy et al. 1990; Serra et al. 2000). Blood was also collected from the trunk of killed rats into heparinized tubes and centrifuged at 900 g for 10 min at 25°C. The resulting plasma supernatant was frozen at −80°C until steroid extraction. Steroids were extracted from plasma three times with 3 mL of ethyl acetate. The recovery (80–100%) of steroid through the extraction procedure was monitored by addition of a trace amount (6000–8000 cpm; 20–80 Ci/mmol) of [3H]corticosterone to plasma. The corticosterone content of the plasma steroid extract was quantified by radioimmunoassay with specific antibodies (MP Biomedicals, Santa Ana, CA, USA) as previously described (Serra et al. 2000).
Rats were killed with a guillotine, and blood was collected from the trunk into heparinized tubes and centrifuged at 1000 g for 10 min at 4°C. The resulting plasma supernatant was frozen at −80°C until being assayed for ACTH with an enzyme immunoassay kit (Phoenix Europe, Karlsruhe, Germany).
Animals were killed by decapitation, and the brain was rapidly removed for dissection of the diencephalic region (hypothalamus), limbic region (hippocampus), and pituitary on frosted glass kept cold on crushed ice. The dissected tissue samples were immediately frozen with dry ice and stored at −80°C until analysis. Frozen tissue was homogenized in a solution (~800 μL per 100 mg of tissue (wet weight)) containing 10 mM Tris-HCl (pH 7.4), 5 mM EDTA, 1% Triton X-100, 1% Nonidet P-40, 0.1 mM phenylmethylsulfonyl fluoride, aprotinin (1 mg/mL), 1 mM benzamidine and bacitracin (200 μg/mL). The homogenates (40 μg of protein) were then subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis on a 4–12% gradient gel (Invitrogen, Milan, Italy), and the separated proteins were transferred electrophoretically to a Hybond-P membrane (GE Healthcare Biosciences, Pittsburg, PA, USA). The membrane was incubated first for 60 min at 25°C with 5% non-fat dried milk in Tris-buffered saline containing 0.01% Tween-20 and then overnight at 4°C with rabbit polyclonal antibodies to GR (GRP-20) or to CRF (CRF-H104) or with goat polyclonal antibodies to CRFR1 (CRFRI-V14), with all antibodies being diluted 1 : 200 in 5% non-fat dried milk and obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was detected with mouse monoclonal antibodies (MAB-374, 1 : 1000 dilution; Millipore, Milan, Italy) as a loading control. The membrane was then washed with Tris-buffered saline containing 0.01% Tween-20 before incubation for 60 min at 25°C with horseradish peroxidase–linked secondary antibodies to rabbit or mouse IgG (1 : 10 000 or 1 : 5000 dilution, respectively; Millipore) or to goat IgG (1 : 5000 dilution; Santa Cruz Biotechnology) in 5% non-fat dried milk. Immune complexes were detected with enhanced chemiluminescence reagents (GE Healthcare Biosciences). Optical density was determined for the GR, CRF, CRFR1, and GAPDH bands with the use of an imaging system (Geliance 600; Perkin Elmer, Monza, Italy) and associated image acquisition (GeneSnap, Perkin Elmer) and analysis (GeneTools, Perkin Elmer) software. The values for GR, CRF, and CRFR1 were normalized by the corresponding value for GAPDH.
Quantitative data are presented as means ± SEM and were analyzed by one-way or two-way anova, with individual means being compared with Newman–Keuls post hoc test. A p < 0.05 value was considered statistically significant.
Brain 3α,5α-TH PROG level and emotional state of socially isolated female rats
To ascertain the effect of social isolation on female rats, we isolated them for 30 days immediately after weaning (at 25–30 days of age) without any additional stressor. Consistent with our previous results obtained with male rats (Serra et al. 2000), isolated females showed a significant decrease (−41%) [F(1, 31) = 11.6476; p = 0.001808] in the basal cerebral cortical concentration of 3α,5α-TH PROG compared with group-housed animals (Fig. 2a). Socially isolated females also exhibited an anxiety-like profile in the elevated plus-maze test. The time spent by socially isolated animals in the open arms of the maze was thus reduced by 61% [F(1, 51) = 10.5989; p = 0.002013] compared with group-housed rats (Fig. 2b). The number of entries into the open arms of the maze was also reduced in socially isolated females (−56%) [F(1, 51) = 10.63192; p = 0.001984], whereas the number of entries into all arms did not differ significantly between the two groups of animals (data not shown). In addition, social isolation resulted in a significant reduction (−32%) [F(1, 34) = 5.4692; p = 0.025371] in the punished consumption of water by female rats in the Vogel conflict test (Fig. 2c).
Brain 3α,5α-TH PROG concentration and emotional state in offspring of socially isolated rats
In contrast to their female (Fig. 2a) and male (Serra et al. 2000) parents, 2-month-old male offspring of socially isolated rats showed a significant increase (+35%) [F(1, 102) = 5.4400; p = 0.021645] in the basal level of 3α,5α-TH PROG in the cerebral cortex compared with offspring of group-housed rats (Fig. 3a). The time spent in the open arms of an elevated plus-maze (Fig. 3b) and the punished consumption of water in Vogel's conflict test (Fig. 3c), however, were similar for the offspring of socially isolated animals and those of group-housed rats. Moreover, locomotor activity did not differ significantly between these two experimental groups (Table 1).
Table 1. Locomotor activity in male offspring of socially isolated (ISO-O) or group-housed (GH-O) rats
Data are means ± SEM for 10 animals per group and were analyzed by one-way anova.
Total distance (cm)
530.3 ± 109.9
456.9 ± 88.1
No. of movements
35.2 ± 3.9
36.1 ± 4.7
Rest time (s)
267.8 ± 5.5
274.4 ± 4.6
Horizontal activity (cm)
547.6 ± 87.3
565.2 ± 86.5
Vertical activity (cm)
223.7 ± 26.3
285.2 ± 35.2
Margin time (s)
88.6 ± 25.4
50.8 ± 18.8
Center time (s)
214.6 ± 24.2
249.0 ± 18.9
Basal activity of the HPA axis in offspring of socially isolated rats
We next examined the basal activity of the HPA axis in male offspring of socially isolated and group-housed rats. The basal plasma concentrations of ACTH (+42%) [F(1, 16) = 17.9417; p = 0.000630] (Fig. 4a) and corticosterone (+31%) [F(1, 122) = 4.6259; p = 0.033465] (Fig. 4b) were significantly increased in offspring of socially isolated rats compared with those of group-housed rats. Intraperitoneal injection of dexamethasone reduced the plasma corticosterone concentration in both groups of animals (Fig. 4c). anova revealed a significant effect of parental isolation [F(1, 32) = 18.0700; p = 0.000172], a significant effect of treatment [F(1, 32) = 71.0741; p < 0.000001], and a significant interaction between factors [F(1, 32) = 24.8352; p = 0.00021]. However, the dexamethasone-induced decrease in the plasma corticosterone concentration was greater in the offspring of socially isolated rats (−93%) than in those of group-housed animals (−66%) [F(1, 17) = 43.0410; p = 0.000005].
We next examined the abundance of CRF and CRFR1 in the hypothalamus and pituitary, respectively, by immunoblot analysis. Immunoreactive bands of ~25 kDa and ~59 kDa were obtained with antibodies to CRF and to CRFR1, respectively, consistent with the sizes of the corresponding antigens (Thompson et al. 1987; Tsai-Morris et al. 1996). The amount of CRF in the hypothalamus (−32%) [F(1, 31) = 8.0322; p = 0.008014] (Fig. 5a) and of CRFR1 in the pituitary (−43%) [F(1, 10) = 15.1552; p = 0.002995] (Fig. 5b) were significantly lower in the male offspring of socially isolated rats compared with those of group-housed rats.
Expression of GR in the hippocampus of offspring of socially isolated rats
Changes in GR gene expression in the hippocampus contribute to negative feedback regulation of the HPA axis by glucocorticoids and thereby contribute to differences in the response of the HPA axis to stress (Oitzl et al. 2010). We therefore examined GR expression in the hippocampus of male offspring of socially isolated or group-housed rats. Immunoblot analysis revealed an ~97-kDa immunoreactive band with antibodies to GR, consistent with the size of the corresponding antigen. The hippocampal abundance of GR was significantly higher in the offspring of socially isolated rats than in those of group-housed animals [F(1, 67) = 21.3815; p = 0.000018)] (Fig. 5c).
Effects of acute stress on brain 3α,5α-TH PROG level and HPA axis activity in offspring of socially isolated rats
We next examined the effects of acute foot-shock stress on the cerebrocortical concentration of 3α,5α-TH PROG as well as on the basal level and dexamethasone-induced suppression of plasma corticosterone. At 25 min after the end of foot-shock treatment, the cerebrocortical concentration of 3α,5α-TH PROG was increased in the offspring of socially isolated or group-housed rats compared with corresponding non-shocked control animals (Fig. 6). However, whereas the foot shock–induced increase in the cerebrocortical level of 3α,5α-TH PROG was greater in socially isolated male rats than in group-housed controls (Serra et al. 2000), the percentage increase in this parameter induced by foot shock in the male offspring of socially isolated rats (+39%) was significantly smaller than that in those of group-housed rats (+120%) [F(1, 48) = 22.70316; p = 0.000018]. Similarly, the foot shock–induced increase in the circulating corticosterone level was blunted in offspring of socially isolated rats compared with those of group-housed rats (+97% vs. +149%, respectively) [F(1, 48) = 9.30369; p = 0.003717] (Fig. 7). Moreover, whereas intraperitoneal injection of dexamethasone attenuated the increase in plasma corticosterone concentration induced by foot-shock stress in the male offspring of socially isolated or group-housed rats, this increase was still significant in those of group-housed rats (+39%), but it was not in those of socially isolated animals (+19%). anova revealed a significant effect of parental isolation [F(1, 43) = 90.3253; p < 0.000001], a significant effect of dexamethasone treatment [F(1, 43) = 48.0160; p < 0.000001], and a significant interaction between factors [F(1, 43) = 29.6969; p = 0.00002] (Fig. 7).
We have shown that female rats isolated for 1 month immediately after weaning, without any additional stressor, manifested a significant decrease in the basal cerebrocortical concentration of 3α,5α-TH PROG compared with group-housed animals. Moreover, socially isolated females exhibited an anxiety-like profile in the elevated plus-maze and Vogel tests. These characteristics are thus similar to those of male rats subjected to social isolation (Serra et al. 2000), indicating that there is no apparent sex difference in the effects of this chronic mild stress.
We also found that adult male offspring of socially isolated parents showed an increased basal level of 3α,5α-TH PROG in the cerebral cortex compared with those of group-housed parents. However, the offspring of socially isolated rats showed no difference in emotional reactivity compared to those of group-housed rats, as evidenced by their similar conditioned (Vogel's test) and non-conditioned (elevated plus-maze test) anxiety responses as well as their similar locomotor activity. 3α,5α-TH PROG is one of the most potent and efficacious positive allosteric modulators of GABAA receptor function (Majewska 1992; Lambert et al. 1995), and its administration in animals, either systemically or intracerebroventricularly, induces marked anxiolytic, sedative-hypnotic, and anti-convulsant effects (Kokate et al. 1994; Bitran et al. 1995; Concas et al. 1996). These observations suggest that the increase in the brain concentration of 3α,5α-TH PROG in the offspring of socially isolated parents may be relevant for the basal emotional state. This conclusion is consistent with the observation that acute administration of 3α,5α-TH PROG attenuated the elevation of ACTH and corticosterone levels in response to emotional stress in rats (Patchev et al. 1996). Indeed, the HPA axis appears to be subject to tonic GABA-mediated inhibition at the hypothalamic level (Cullinan et al. 2008), and peripheral injection of positive modulators of the GABAA receptor (such as benzodiazepines) reduces HPA axis activity (Imaki et al. 1995; Grottoli et al. 2002). On the other hand, neuroactive steroids, including 3α,5α-TH PROG, induced CRF or arginine vasopressin (AVP) synthesis and release at the hypothalamic level, thereby increasing plasma ACTH and corticosterone concentrations (Naert et al. 2007). Consistent with these latter observations, the offspring of socially isolated parents in this study manifested increased circulating levels of ACTH and corticosterone, indicative of an activated HPA axis, as well as a reduced abundance of CRF in the hypothalamus, possibly indicative of increased CRF release (Naert et al. 2007), under the basal condition compared with the offspring of group-housed rats. This conclusion is consistent with the evidence that social isolation of male rats, in which cerebrocortical and plasma concentration of 3α,5α-TH PROG is reduced (Serra et al. 2000), showed decrease of plasma levels of both ACTH and corticosterone, as well as increase sensitivity of the pituitary to CRF (Serra et al. 2005). CRF interacts with CRFR1 in the pituitary (De Souza and Kuhar 1986; Aguilera 1988) to regulate ACTH secretion, and both CRF and glucocorticoids down-regulate CRFR1 mRNA levels in the rat anterior pituitary (Makino et al. 1995; Pozzoli et al. 1996; Sakai et al. 1996). The significant reduction in the expression of CRFR1 apparent in the pituitary of the offspring of socially isolated rats compared with those of group-housed rats is thus consistent with the notion that the basal activity of the HPA axis is increased in these animals. It remains to be determined whether adaptation of the AVP system may contribute to this altered activity of the HPA axis. Indeed, CRF and AVP act synergistically to stimulate ACTH release (Gillies et al. 1982), and such synergism was recently attributed to heterodimerization of CRFR1 and the V1b receptor for AVP (Murat et al. 2012). A reduced level of V1b expression may be consistent with some behavioral and endocrine features of the offspring of isolated rats, given that a V1b antagonist showed anti-depressive and anxiolytic properties in rats (Serradeil-Le Gal et al. 2005) and that V1b knockout mice exhibit a reduced level of ACTH secretion in response to stress (Lolait et al. 2007).
Abnormal CRF neurotransmission and CRFR1 signal transduction has been proposed to contribute to stress pathophysiology that leads to major depression. Individuals with major depressive and anxiety disorders thus manifest hyperactivity of the HPA axis with elevated plasma ACTH and cortisol concentrations under basal conditions and a blunted ACTH response in the CRF stimulation test (Condren et al. 2002; Gillespie and Nemeroff 2005; Abelson et al. 2007), the latter of which has been attributed to reduced CRFR1 expression in the anterior pituitary secondary to chronic hypersecretion of CRF from the hypothalamus.
On the other hand, offspring of socially isolated parents may be classified as resilient. It is generally accepted concept that resilience is the ability to avoid deleterious behavioral changes in response to chronic stress that can be achieved by molecular adaptations that promote normal behavioral function (see for review Russo et al. 2012). While, offspring of socially isolated parents were not challenged by chronic stress, we found that acute stress (foot shock) induced lower percentage increases in the brain concentration of 3α,5α-TH PROG and the plasma level of corticosterone in the offspring of socially isolated parents than in those of group-housed rats. Our observations that the offspring of socially isolated rats did not show either anxiety-like behavior (Fig. 2) or depressive-like behavior in the forced swim test (data not shown) seem consistent with this scenario. Thus, the constitutively high basal cerebrocortical level of 3α,5α-TH PROG found in these animals might normalize their emotional state and represent a protective mechanism against stress; 3α,5α-TH PROG might thus regulate HPA axis activity by increasing hormone levels in the basal condition and attenuating stress-induced axis activation, as suggested by Naert et al. (2007). Accordingly, up-regulation of 3α,5α-TH PROG biosynthesis appears to be one of the mechanism for the anxiolytic and anti-aggressive effects of anti-depressant drugs (Pinna et al. 2009) and molecules that exert anxiolytic effects by increasing the brain level of 3α,5α-TH PROG has been proposed as new anxiolytic drugs (Serra et al. 1999; Schüle et al. 2001). In addition, the attenuated response of the HPA axis to acute stress apparent in the offspring of socially isolated parents seems to be well correlated with the increased negative feedback sensitivity to glucocorticoid, as revealed by the increased suppressive effect of dexamethasone on the basal and stress-induced increase in plasma corticosterone level. This increased sensitivity in turn may be related to the observed up-regulation of GR expression in the hippocampus, which mediates the inhibitory effect of glucocorticoids on HPA axis activity (Jacobson and Sapolsky 1991; Malkoski and Dorin 1999), thus reflecting an increased capability of this structure to convert a hormonal signal into a feedback action (Reul and de Kloet 1985). As proposed, the endogenous levels of 3α,5α-TH PROG play a crucial role in the HPA function, thus, reduced 3α,5α-TH PROG brain levels following social isolation leads to hyperresponsive rats to stress and impair negative feedback regulation of the HPA axis (Serra et al. 2005). In offspring of socially isolated parents negative feedback regulation of HPA axis was enhanced in association with an increased basal concentration of 3α,5α-TH PROG. This conclusion is consistent with the evidence that the impairment of the negative feedback regulation of the HPA axis apparent in isolated rats (Serra et al. 2005) was prevented or normalized by administration of exogenous 3α,5α-TH PROG either during or following social isolation (Evans et al. 2012).
In addition, the negative feedback regulation of HPA axis activity, the response to stress, and the hippocampal GR expression observed in the offspring of isolated parents resemble those in adult rats that were exposed to a brief daily period of separation from their mother during the first few weeks of life (Levine et al. 1957; Levine 1967). These animals show reduced fearfulness, attenuated ACTH, and corticosterone responses to stress, and increased GR binding capacity in the hippocampus (Levine 1967; Meaney et al. 1989, 1992). All of these changes were suggested to be because of altered dam-pup interactions. Indeed, the quality of maternal care provided to offspring early in life is a direct determinant of developmental programming and calibration of the HPA axis (Liu et al. 1997; Francis et al. 1999). Our preliminary observations indicate that socially isolated dams did not differ in the total frequency of licking-grooming and arched-back nursing from group-housed dams during the first two post-partum weeks. More detailed studies are necessary, however, to rule out the possibility that maternal behavior of socially isolated dams is a key determinant of the stress responsiveness and HPA function in their offspring.
Both corticosteroids and stress have a pronounced epigenetic impact on the brain and behavior not only within individual life spans but across generations in both humans and animals (Hunter 2012). Stress vulnerability was recently proposed to be transmitted epigenetically by DNA methylation in the germ line, thus influencing behavior of offspring in the next generation (Franklin et al. 2010). In contrast, epigenetic mechanisms were found to play a limited role in the behavioral adaptations of offspring of fathers subjected to chronic social defeat (Dietz et al. 2011). We have shown that environmental perturbation can lead not only to alterations in individuals who experience such change (Serra et al. 2000, 2003, 2005, 2006; Pisu et al. 2011) but also to altered neuroendocrine responses of future generations who were not directly exposed to the challenge (present study). Further studies should provide insight into the role of epigenetic modification of the maternal or paternal germ line in this transgenerational effect.
We thank Professor Andrew J Lawrence, for critical reading of the manuscript and valuable feedback. This study was supported by the Sardinian Government (RAS grant F71J11000900002) and the Fondazione Banco di Sardegna. The authors declare no conflicts of interest.