Environmental enrichment (EE) is an experimental setting broadly used for investigating the effects of potentiating social interactions as well as increasing the exposure to sensorimotor and cognitive stimulations on brain structure and function (Rosenzweig et al., 1964; Nithianantharajah and Hannan, 2006, 2009). EE exerts beneficial effects on many behavioral (improved motor and cognitive performances, reduced reactivity to stress), morphological (increased dendritic arborization, spine number, synaptic density, and neurogenesis), and molecular (modifications in gene expression, neurotrophic factors, and neurotransmission) brain features (Nithianantharajah and Hannan, 2006, 2009; Petrosini et al., 2009; Baroncelli et al., 2010; Simpson and Kelly, 2011; Sale et al., 2014).

In spite of the vast literature on the positive effects of EE when applied immediately after weaning, in adulthood, during aging or even in the presence of brain damage, the transgenerational beneficial effects of pre-reproductive EE have been only recently examined (Arai et al., 2009; Arai and Feig, 2011; Leshem and Schulkin, 2012; Mashoodh et al., 2012; Caporali et al., 2014, 2015; Cutuli et al., 2015). And yet, this issue is noteworthy given that parental environmental experience may imprint offspring's phenotype over generations through many epigenetic processes (Weaver, 2007; Kanherkar et al., 2014).

Interactions between individual and environment take place lifelong from conception. In fact, during fetal development besides the overwhelming impact of the genetic control, even the environmental stimuli strongly influence the highly susceptible developing structures. Namely, the environment experienced by the pregnant mother exerts substantial effects on the intrauterine milieu and affects fetal organogenesis. During pregnancy, the maternal exposure either to negative (radiations, contaminants, alcohol, drugs, prenatal stress, food deprivation, or micronutrient deficiency; Rice and Barone, 2000; Meck and Williams, 2003; Van den Bergh et al., 2005; Weinstock, 2005; Mueller and Bale, 2008; Swanson et al., 2009; Thompson et al., 2009; Charil et al., 2010; Glover, 2011) or positive (complex environments, voluntary exercise, dietary supplementation; McKim and Thompson, 1975; Kiyono et al., 1985; Dell and Rose, 1987; Parnpiansil et al., 2003; Bick-Sander et al., 2006; Lee et al., 2006; Welberg et al., 2006; Sale et al., 2007; Herring et al., 2012; Leshem and Schulkin, 2012; Mychasiuk et al., 2012; Rosenfeld and Weller, 2012) environmental stimulations may exert deleterious or beneficial impact, respectively, on physical and behavioral offspring's development.

In previous studies (Caporali et al., 2014, 2015; Cutuli et al., 2015) we demonstrated the beneficial influences of pre-reproductive EE on maternal care and offspring's motor and cognitive performances as well as on neurotrophic functioning.

On such a basis the current research was aimed to analyze whether and how pre-reproductive EE of female rats may influence even the effects of a chronic stress, such as social isolation, on their adolescent offspring. In fact, it is accepted that social isolation at adolescence represents a valid tool for understanding the impact of social stress given that adolescent rodents live in groups and exhibit high levels of social behavior (Panksepp et al., 2007). To this aim glucocorticoid receptors (GRs) expression in cortico-limbic networks was assessed as a marker of stress adaptation in the offspring of pre-reproductively enriched or standard dams. It is well-known that glucocorticoids are the end product of hypothalamic-pituitary-adrenal (HPA) axis activation and have large effects on adaptive behaviors and pathophysiology of several stress-related disorders, such as depression, anxiety, and drug abuse (Holsboer, 2000; McEwen, 2000; Sapolsky, 2000; Roozendaal and McGaugh, 2011; Lee and Sawa, 2014). Furthermore, we evaluated the interactions of pre-reproductive maternal EE and chronic social isolation on the coping response of the adolescent offspring to an inescapable stressful situation (Forced Swim Test, FST), and on cortico-limbic network activation (c-Fos). Indeed, hippocampus, amygdala, and cingulate cortex, are closely involved in HPA axis regulation (Sapolsky, 2003; Sarrazin et al., 2009; Nicolaides et al., 2015). To gain further indications about the possible mechanisms through which pre-reproductive maternal rearing conditions may influence the offspring's phenotype, we also assessed the early maternal care of pre-reproductively enriched or standard reared lactating dams on the first post partum day (ppd 1).

The present study was designed to evaluate the effects of pre-reproductive maternal experience on maternal behavior and adolescent offspring's response to stress. In line with our previous studies (Caporali et al., 2014, 2015; Cutuli et al., 2015), the pre-reproductive maternal enrichment did influence maternal care and pups' behavior. In fact, in comparison to standard reared females the enriched dams exhibited a more tuned maternal care repertoire, as pointed out by their higher levels of pup-directed behaviors. Pre-reproductive exposure to EE also affected the coping response of the male adolescent offspring when exposed to an inescapable stressful situation (FST) after a chronic stress (social isolation). Namely, after 2 weeks of isolation enriched dams' offspring exhibited higher levels of Climbing during FST in comparison to not isolated offspring born to enriched dams. Furthermore, isolated and control offspring of enriched dams showed a lower amygdaloid c-Fos activation in response to FST and a higher number of GRs compared to offspring of standard reared dams.

Growing evidence indicates that EE influences maternal nurturing. For example, post-weaning enriched housing enhances Licking/Grooming (LG) behavior and oxytocin receptor binding of the low LG offspring across generations (Champagne and Meaney, 2007). Further, EE during pregnancy and lactation enhances pup-licking levels (Sale et al., 2004). Increased levels of pup nursing and licking associated with changes in BDNF and MeCP2 gene expression in the maternal hypothalamus have been also reported in standard reared females mated with an enriched male (Mashoodh et al., 2012). We recently described higher levels of licking, crouching, and nest building activities accompanied by increased BDNF levels in frontal cortex of pre-reproductively enriched dams (Cutuli et al., 2015). Furthermore, in a cross-fostering study we demonstrated that the maternal care of pre-reproductively enriched dams is markedly modulated by the bidirectional interactions between mother and pups, with the enriched dam/standard pups couple resulting in the most maladaptive encounter to shape offspring's phenotype (Caporali et al., 2015). Here, we demonstrate that the early maternal care modifications induced by pre-reproductive EE, which were previously observed using a brief mother-pups separation as eliciting condition (Cutuli et al., 2015), were still evident using undisturbed conditions. In particular, during the observations the enriched dams spent most time in Crouching, the most active and complex nursing posture that allows to nourish and contact the entire litter. Such a behavioral pattern was associated to reduced social investigative (i.e., Sniffing) and partial suckling (i.e., Nursing) behaviors. Concurrently, for the enriched dams the litter assumed a salience able to reduce the emission of non-pup-directed behaviors, such as explorative (i.e., Wall Rearing) or feeding (i.e., Drink or Eat) behaviors.

Generally, the maternal care enhanced by pre-reproductive EE are accompanied by offspring's behavioral and physical improvements. For example, better memory and learning abilities and higher visual acuity have been described in the offspring of enriched mothers compared to the offspring of standard reared dams (Sale et al., 2004; Cutuli et al., 2015). Further, adult male offspring of enriched low LG females display enhanced exploration and novelty discrimination abilities (Champagne and Meaney, 2007). It has also been found that paternal pre-reproductive EE that leads to the above-mentioned maternal care changes is linked to increased weight of their offspring at weaning (Mashoodh et al., 2012). Extending these observations, the present study describes behavioral and biochemical differences in coping response of the isolated adolescent offspring of pre-reproductively enriched females. It is well-known that in adult rats social isolation has a depressant effect that induces increased immobility in FST, while direct exposure to EE has an antidepressant effect that increases swimming and climbing behaviors (Brenes et al., 2008; Mosaferi et al., 2015). In the current research socially isolated adolescent male pups exhibited reduced body weight and paradoxically decreased passive coping responses (Immobility) regardless of maternal pre-reproductive housing. This latter finding indicates that social isolation during early developmental stages alters the response to aversive stimuli at adolescence, in accordance with previous studies (Wongwitdecha and Threenet, 2001; Wongwitdecha et al., 2006; Hong et al., 2012). More importantly, while Swimming was equally increased in both isolated groups (EF-p iso and SF-p iso), maternal enrichment induced higher levels of Climbing during FST in the isolated adolescent offspring (EF-p iso) in comparison to not isolated offspring born to EF (EF-p). Notably, the increased Climbing evident only in EF-p iso suggests that an early chronic stress is required to uncover traits transmitted from mother to pups. In line, Cymerblit-Sabba et al. (2013) found that exposure to EE during pregnancy followed by stress at adulthood improves emotional and attentional reactivity in the offspring.

It is well-known that FST active behaviors reflect the activity of multiple neurotransmitters, with norepinephrine (NE) primarily mediating climbing, and serotonin (5-HT) mediating swimming (Armario et al., 1988; Detke et al., 1995; Rénéric and Lucki, 1998; Cryan et al., 2005). Interestingly, it has been reported that EE enhances NE levels (Naka et al., 2002; Robertson, 2013). Given the increased Climbing in the isolated adolescent rats born to enriched dams, it could be interesting to verify whether the pre-reproductive EE is able to increase NE neurotransmission in the offspring.

Also biochemical analyses support the effects of social isolation and maternal rearing conditions. We found a significantly higher number of c-Fos⁺ cells in SF-p rats tested in the FST compared to no tested SF-p rats in all cortico-limbic structures investigated (Figure 6), endorsing the neuronal activation of structures, such as the hippocampus, amygdala, and cingulate cortex in response to FST (Duncan et al., 1993; Jang et al., 2009; Raineki et al., 2012). Notably, in the remaining groups (SF-p iso, EF-p iso, EF-p) no significant differences between tested and no tested animals were detected, except for SF-p iso group in cingulate cortex (Figure 6C). It is widely recognized that in standard reared animals an acute exposure to a stressful condition (as FST) produces a massive activation of stress response system (Duncan et al., 1993; Jang et al., 2009; Raineki et al., 2012) that can be altered by a persistent exposure to a stressful condition (as social isolation; Chen and Herbert, 1995; Wall et al., 2012). Also EE is able to induce physiological adaptations of stress system by decreasing neuronal activation (Solinas et al., 2009; Grimm et al., 2016). The present results extend the reduced activation of stress response even to animals not directly exposed to an enriched environment. Namely, the neuronal activation of amygdala was influenced by both social isolation and maternal EE in a summative way (Figure 7B). The findings on the amygdala exemplify how an early chronic stress uncovers the intergenerational EE effects on the response to an inescapable stress. Several studies have shown that enhancements of amygdaloid responses can be associated to develop psychopathology (LeDoux, 2000; Davis et al., 2010; Rajbhandari et al., 2016; Rincón-Cortés and Sullivan, 2016; Shackman et al., 2017). Among others, Raineki et al. (2012) demonstrated that an early-life abuse is followed by depressive-like behaviors at adolescence, which are related to amygdaloid hyperactivation and rescued by temporary pharmacological deactivation of the amygdala. Consistently, we found that EF-p iso group showed the lowest c-Fos activation index, demonstrating that a post-weaning isolation is able to reduce amygdaloid activation and that the being born to a pre-reproductive enriched female further blunts c-Fos expression.

The amygdala with its projections to the paraventricular hypothalamic nucleus, bed nucleus of the stria terminalis and periaqueductal gray (Allen and Allen, 1974; Van de Kar et al., 1991; Herman et al., 2005) is markedly involved in stress response circuitry and further clue of this modulation raises from its GRs expression. GRs are recognized as a great marker of brain plasticity, adaptation, and vulnerability (de Kloet et al., 2008). Genomic and non-genomic GRs signal transductions have considerable consequences for the individual's response and adaptation to stressors. Thus, a flourishing field of literature is rooted on corticosteroid receptor role with the goal to explain how stress can lead to multifaceted outcomes, ranging from vulnerability to stress resilience (Meaney et al., 1985; de Kloet et al., 2005; Schulte-Herbrüggen et al., 2006; Oitzl et al., 2010). We found significant differences in amygdala GRs expression associated with both offspring's social isolation and pre-reproductive rearing conditions of dams, as demonstrated by the higher amygdaloid GRs expression in SF-p iso and EF-p groups compared to SF-p group (Figure 8). Noteworthy, no significant differences were found between EF-p and EF-p iso groups, suggesting that the higher amygdaloid GRs expression of EF offspring could have been transmitted from mother to pups and no further modulated by social isolation.

Models of genetically altered animals have provided a useful tool to investigate GRs role in the stress response. Models of murine depression and stress-resistance have been generated by using the GR-heterozygous mutant mice (GR^+/−) that under-express GRs, or the YGR mice that over-express GRs (Ridder et al., 2005; Schulte-Herbrüggen et al., 2006). Namely, GR overexpression-dependent increases of BDNF in hippocampus and amygdala appear to be the dynamic correlate of enhanced stress resistance (Schulte-Herbrüggen et al., 2006). Since the transmission of positive (Caporali et al., 2014; Cutuli et al., 2015) and negative (Niknazar et al., 2016) maternal experiences affects offspring's BDNF levels, it is possible that the greater resilience of the EF-p iso group over-expressing amygdaloid GRs may be ascribed to enhanced BDNF levels.

As final note, we would like to emphasize that EF offspring's behavioral performance likely occurs through alterations in maternal care due to the pre-reproductive EE (Champagne and Meaney, 2007). In fact, the increase in Crouching levels observed in enriched dams make them similar to the high Licking and Grooming and Arched-Back Nursing (LG-ABN) mothers described by Meaney's group (Weaver et al., 2004), and it may account for the increase in Climbing exhibited by their offspring in facing the FST. Similarly, the offspring of high LG-ABN mothers exhibited enhanced exploration and reduced anxiety accompanied by attenuated HPA response to stress and an over-expression of GRs, although in the hippocampus (Liu et al., 1997, 2000; Caldji et al., 1998; Champagne and Meaney, 2007).

We are aware that the intergenerational transmission of EE effects might also occur through maternal germ-line, during fetal development, and lactation (Franklin and Mansuy, 2010; Ho and Burggren, 2010), but this topic is out of the current study scope. Anyway, the outcome of the perinatal interaction “gene × environment” depends on the degree of “matching” between intergenerationally transmitted features and environmental demands later in life. Not by chance, the main behavioral and biochemical differences observed in EF-p group become indeed evident when the pups are socially isolated.

After the beneficial effects on motor and cognitive abilities (Caporali et al., 2014; Cutuli et al., 2015), we now show the influence of pre-reproductive maternal EE on offspring's stress response.

Notably, offspring of parents exposed to negative experiences (stress or trauma) have been reported to be at a greater risk for physical, behavioral, and cognitive problems (Yehuda and Bierer, 2008; Roth et al., 2009; Debiec and Sullivan, 2014; Bowers and Yehuda, 2016). On the other hand, positive complex stimulation, such as EE, seems to intergenerationally produce a neuroplastic reserve able to counteract negative outcomes of stressful events through multiple mechanisms, such as GRs modulation. Thus, evaluating whether this reserve can be spent beyond adolescence to cope with stress during the entire lifespan is an important issue deserving investigations in the near future. In the same manner, it could be interesting to assess whether maternal care modifications induced by the pre-reproductive EE are stable in time and across generations. Further, since in the present research the amygdala has proved to be the brain area most sensitive to stress and intergenerational EE influence, the implication of amygdaloid circuits in emotional regulation should be thoroughly investigated using the paradigm of pre-reproductive EE.

All authors participated in designing the research; DC, PC, PS, DL, and FF performed behavioral evaluations; PD, EB, GP, MP, and FG performed biochemical analyses; DC, PC, PD, EB, GP, MP, and LP analyzed data; all authors discussed and approved data; DC, EB, GP, MP, DL, and LP interpreted data and wrote the paper.