Abstract
The noradrenergic system plays an important role in prefrontal cortex (PFC) function. Since early life experiences play a crucial role in programming brain function, we investigated the effects of a neonatal experience involving reward through maternal contact on the noradrenergic system of the rat PFC. Rat pups were exposed during Postnatal days (PNDs) 10–13, to a T-maze in which contact with the mother was used as a reward (RER). RER males had higher norepinephrine levels in the PFC both on PND 13 and in adulthood. The RER experience resulted in adulthood in increased levels of the active demethylase GADD45b, hypomethylation of the β1 adrenergic receptor (ADRB1) gene promoter, and consequent enhanced expression of its mRNA in the PFC. In addition, protein and binding levels of the ADRB1, as well as those of its downstream effector phosphorylated cAMP response element-binding protein were elevated in RER males. The higher activity of the PFC noradrenergic system of the RER males was reflected in their superior performance in the olfactory discrimination and the contextual fear extinction, 2 PFC noradrenergic system-dependent behavioral tasks.
Introduction
The noradrenergic system is of primary importance for the survival of the organism, since it mediates arousal and vigilance (Morilak et al. 2005), necessary for finding food, selecting a mate, and avoiding predators. Norepinephrine (NE) cell bodies are localized mainly in the locus coeruleus and project throughout the brain, including the prefrontal cortex (PFC; Berridge and Waterhouse 2003; Robertson et al. 2013). In the PFC noradrenergic neurotransmission plays a cardinal role in cognitive processes, such as arousal, attention, vigilance, behavioral flexibility, motivation, working memory, and response inhibition (Harley 1987; Aston-Jones et al. 1991, 2000; Devauges and Sara 1991; Cole and Robbins 1992; Arnsten 2000; Dalley et al. 2004).
The development of cognitive abilities is known to be influenced by early experiences, where a positive environment during the critical postnatal period has beneficial effects on adulthood, while adverse life events can lead to deficits: Increased maternal care results in improved performance in a variety of cognitive tests, such as the Morris water maze (Liu, Diorio, et al. 2000), the novel object recognition (Fenoglio et al. 2005), and contextual fear conditioning (Beane et al. 2002), while maternal deprivation has the opposite effects impairing cognitive abilities (Aisa et al. 2007; Chocyk et al. 2014). Early life experiences, within the framework of their role in regulating stress responsiveness, have been shown to affect components of the noradrenergic system (Escorihuela et al. 1995; Liu, Caldji, et al. 2000; Baamonde et al. 2002; Lucion et al. 2003). Interestingly, Coccurello et al. (2014) recently reported in mice that following maternal separation there is down-regulation of α1 adrenergic receptors in corticolimbic circuits involved in the processing of olfactory stimuli and rewards.
In our laboratory we have developed an early experience model in which rat pups on postnatal day (PND) 10 till PND13 learn to search for their mother placed at the end of 1 arm of a T-maze (Panagiotaropoulos et al. 2009); upon finding her, 1 group of pups is allowed to be retrieved by her (receiving the expected reward: RER), while the other is denied the maternal contact (denied the expected reward: DER). RER pups exhibit sustained motivation in searching for their mother, most probably due to the reward of maternal contact, which they receive repeatedly (Panagiotaropoulos et al. 2009). It should be noted that during the postnatal period the mother is the most salient, emotionally charged with positive valence, stimulus for the pups (Cirulli et al. 2003). In addition, the RER pups have increased accumbal dopamine, indicating activation of their reward system (Raftogianni et al. 2014). Relevantly, it has been shown that prefrontal NE is important in ascribing motivational salience to highly salient stimuli by regulating the release of dopamine in the accumbens, in response to rewarding stimuli (Puglisi-Allegra and Ventura 2012). Furthermore, it has been well documented that NE, acting through β adrenergic receptors, plays a crucial role in another form of learning during the neonatal period, namely associative olfactory learning (Sullivan et al. 1992, 1994, 2000; Moriceau and Sullivan 2004).
Based on the above, in the present work we investigated the effects of the early experiences of our model on the noradrenergic system in the PFC. More specifically, we determined in the PFC of 13-day-old pups the effect of the RER and DER experiences on levels of NE, as well as those of β1 adrenergic receptor (ADRB1), the most abundant receptor of the noradrenergic system in the PFC (Rainbow et al. 1984; Nicholas et al. 1993; Summers et al. 1995).
It is well known that early experiences have long-term effects manifested in adulthood (Chocyk et al. 2013; Maccari et al. 2014; Tang et al. 2014). Within this conceptual framework, we determined the levels of NE, as well as those of ADRB1 in the adult brain of male RER, DER, and control (CTR) animals. Furthermore, we assessed the levels of the phosphorylated cAMP response element-binding protein (pCREB), a downstream effector of β1adrenergic receptor (Roseboom and Klein 1995), as an index of its functionality. Given the important role of the noradrenergic system in PFC function, we proceeded to evaluate the effect of the early experiences of our model on 2 PFC-dependent behaviors (olfactory discrimination Tronel and Sara 2002, 2004 and fear extinction Do-Monte et al. 2010).
The long-term effects of early experiences have been shown to be affected through epigenetic modifications, including CpG island methylation, in the regulatory sequences of relevant genes, leading to changes in gene expression in the brain (Kundakovic and Champagne 2014). The first gene whose promoter methylation was documented to be decreased following increased maternal care (Weaver et al. 2004) and increased by maternal separation was that of the glucocorticoid receptor (GR, Nr3c1; Kember et al. 2012) in the hippocampus. Interestingly, the latter finding has also been reported for humans as a result of early life abuse (McGowan et al. 2009). Other genes whose promoter methylation pattern has been shown to be affected by the quality of maternal care include that of estrogen receptor 1 (ERα –Esr1; Champagne et al. 2006), glutamate decarboxylase (Gad1; Zhang et al. 2010), type 1 metabotropic glutamate receptor (mGluR1–Grm1; Bagot et al. 2012), corticotropin releasing factor (Crf; Wang et al. 2014), arginine vasopressin (Avp; Murgatroyd et al. 2009), neurotensin receptor (Ntsr1; Toda et al. 2014), and brain-derived neurotrophic factor (Bdnf; Roth et al. 2009) in brain areas of the limbic system, such as the hippocampus, hypothalamus, and amygdala. Most interestingly, there are no reports of early experience-induced epigenetic changes in noradrenergic system-related genes, thus posing a challenging question as to whether the development of this most important neurotransmitter system is subject to this form of regulation. In order to address this question we investigated the methylation pattern of the promoter region of the ADRB1 gene and a possible mechanism involved, as well as the levels of expression of this gene in the PFC.
Materials and Methods
Animals
Male Wistar rats born and reared in our colony were used in all experiments. Animals were kept under standard conditions (24°C, 12:12 h light/dark cycle) and received food (Kounker-Keramari Bros. & Co., Athens, Greece) and water ad libitum. Prior to the day of birth, which was designated as PND 0, each litter was assigned randomly to either of the 2 experimental groups (pups denied [DER] or receiving [RER] the expected reward), or to the CTR group (16 litters/group; animals from the same litter have been used for different biochemical, molecular, histochemical, and/or behavioral tests, see below). In order to maintain stable environmental stimulation of the pups, instead of cleaning the cage, wood chip was added every 4–5 days, without disturbing either the pups or the dam. On PND 22, animals were weaned and housed in same-sex, same group (CTR, DER, and RER) of 3–4 animals per cage. Rat pups were sacrificed on PND 13, 2 h after the last training in the T-maze, whereas adult animals when they were approximately 3 months old.
Ethics Statement
All experiments were conducted in agreement with the recommendation of the European Communities Council Directive of 22 September 2010 (2010/63/EU).
Neonatal Training in the T-maze
Animals belonging to the control group were left undisturbed with their mother throughout the lactation period. Animals of the 2 experimental groups were exposed either to the RER or DER experience, starting from PND 10 until PND 13. The training room had a temperature of 24 ± 1°C and it was moderately lit (300 lux). As previously described (Panagiotaropoulos et al. 2009), we used a custom-made T-maze whose arms led to 2 cages with dimensions 30 cm width × 22 cm length × 30 cm height. The dimensions of the start box and the arms of the T-maze were 8 cm width × 6 cm length × 6 cm height, and 7 cm width × 30 cm length × 6 cm height, respectively. At the end of the right arm of the T-maze a small sliding door (9 × 11 cm) permitted access to the mother-containing cage when pups were trained under receipt of expected reward (RER group, see below) or remained always closed, preventing entrance in the cage, when pups were trained under denial of expected reward (DER group, see below). At the end of the left arm of the T-maze another cage was placed, without access from inside the T-maze, containing a virgin female rat for control purposes.
Neonatal Training Under Receipt of Expected Reward-RER Rats
Rats trained under the continuous reinforcement schedule were exposed to 10 trials per day (which amounts to a total of 40 trials for the 4 experimental days, each trial lasted a maximum of 60 s), as previously described (Panagiotaropoulos et al. 2009). At the end of the duration of the trial, or when the pup located the entrance of the mother-containing cage the sliding-door opened and the mother could retrieve the pup. If a pup did not succeed to reach the entrance of the mother-containing cage before the end of the 60 s (maximum duration of the trial), it was gently guided to the entrance, the sliding door opened and the mother was allowed to retrieve the pup. In both cases, following the end of the trial, the pup returned to the mother-containing cage and then the next pup was exposed to the same procedure. When all pups were exposed to the first trial the same procedure was repeated for the next trial until the end of the experimental session consisting of 10 trials.
Neonatal Training Under Denial of Receipt of Expected Reward-DER Rats
The same procedure was followed for DER rats (10 trials per day for 4 training days, maximum duration of each trial 60 s). The behavioral observation was terminated at the time the pup located the entrance of the mother-containing cage. If a pup did not succeed to reach the entrance of the mother-containing cage before the end of the 60 s (maximum duration of the trial), it was gently guided to the entrance and had to remain there for 20 s. Immediately after the end of the trial the rat was placed back in the start box of the T-maze for the next trial. Following the completion of 10 trials the pup was returned to the mother-containing cage and then another pup was exposed to the same procedure. At the end of the 10 daily trials for all the pups, the mother, followed by the litter, was returned to the home cage in the animal facility.
Animals in both the RER and DER groups progressively improved their performance in the T-maze, although RER animals did so more efficiently, in accordance with previous results from our laboratory (Panagiotaropoulos et al. 2009): As training progressed animals in both groups made both more exits from the start box and more correct choices entering the arm of the maze leading to the mother-containing cage; moreover, the latency to exit from the start box as well as the latency to reach the entrance of the mother-containing cage was progressively reduced (Table 1).
Behavior of RER and DER pups during training in the T-maze
| PND10 | PND11 | PND12 | PND13 | |
|---|---|---|---|---|
| Number of exits from the start box | ||||
| RER | 0.6 ± 0.2 | 2.5 ± 0.5 | 6.0 ± 0.5 | 7.9 ± 0.4 |
| DER | 0.4 ± 0.2 | 1.5 ± 0.3 | 2.9 ± 0.4 | 5.6 ± 0.4 |
| Number of correct choices | ||||
| RER | 0.3 ± 0.1 | 1.3 ± 0.3 | 3.7 ± 0.3 | 5.2 ± 0.3 |
| DER | 0.2 ± 0.1 | 0.8 ± 0.2 | 1.6 ± 0.2 | 3.8 ± 0.3 |
| Latency to exit from the start box (s) | ||||
| RER | 57.1 ± 0.8 | 50.3 ± 1.4 | 41.4 ± 1.2 | 36.3 ± 1.5 |
| DER | 57.5 ± 0.8 | 52.9 ± 1.2 | 48.6 ± 1.2 | 40.8 ± 1.3 |
| Latency to reach the entrance of the mother-containing cage (s) | ||||
| RER | 59.3 ± 0.2 | 56.3 ± 0.9 | 50.2 ± 1.1 | 41.9 ± 1.3 |
| DER | 59.3 ± 0.2 | 57.6 ± 0.6 | 55.1 ± 0.8 | 48.3 ± 1.1 |
| PND10 | PND11 | PND12 | PND13 | |
|---|---|---|---|---|
| Number of exits from the start box | ||||
| RER | 0.6 ± 0.2 | 2.5 ± 0.5 | 6.0 ± 0.5 | 7.9 ± 0.4 |
| DER | 0.4 ± 0.2 | 1.5 ± 0.3 | 2.9 ± 0.4 | 5.6 ± 0.4 |
| Number of correct choices | ||||
| RER | 0.3 ± 0.1 | 1.3 ± 0.3 | 3.7 ± 0.3 | 5.2 ± 0.3 |
| DER | 0.2 ± 0.1 | 0.8 ± 0.2 | 1.6 ± 0.2 | 3.8 ± 0.3 |
| Latency to exit from the start box (s) | ||||
| RER | 57.1 ± 0.8 | 50.3 ± 1.4 | 41.4 ± 1.2 | 36.3 ± 1.5 |
| DER | 57.5 ± 0.8 | 52.9 ± 1.2 | 48.6 ± 1.2 | 40.8 ± 1.3 |
| Latency to reach the entrance of the mother-containing cage (s) | ||||
| RER | 59.3 ± 0.2 | 56.3 ± 0.9 | 50.2 ± 1.1 | 41.9 ± 1.3 |
| DER | 59.3 ± 0.2 | 57.6 ± 0.6 | 55.1 ± 0.8 | 48.3 ± 1.1 |
Behavior of RER and DER pups during training in the T-maze
| PND10 | PND11 | PND12 | PND13 | |
|---|---|---|---|---|
| Number of exits from the start box | ||||
| RER | 0.6 ± 0.2 | 2.5 ± 0.5 | 6.0 ± 0.5 | 7.9 ± 0.4 |
| DER | 0.4 ± 0.2 | 1.5 ± 0.3 | 2.9 ± 0.4 | 5.6 ± 0.4 |
| Number of correct choices | ||||
| RER | 0.3 ± 0.1 | 1.3 ± 0.3 | 3.7 ± 0.3 | 5.2 ± 0.3 |
| DER | 0.2 ± 0.1 | 0.8 ± 0.2 | 1.6 ± 0.2 | 3.8 ± 0.3 |
| Latency to exit from the start box (s) | ||||
| RER | 57.1 ± 0.8 | 50.3 ± 1.4 | 41.4 ± 1.2 | 36.3 ± 1.5 |
| DER | 57.5 ± 0.8 | 52.9 ± 1.2 | 48.6 ± 1.2 | 40.8 ± 1.3 |
| Latency to reach the entrance of the mother-containing cage (s) | ||||
| RER | 59.3 ± 0.2 | 56.3 ± 0.9 | 50.2 ± 1.1 | 41.9 ± 1.3 |
| DER | 59.3 ± 0.2 | 57.6 ± 0.6 | 55.1 ± 0.8 | 48.3 ± 1.1 |
| PND10 | PND11 | PND12 | PND13 | |
|---|---|---|---|---|
| Number of exits from the start box | ||||
| RER | 0.6 ± 0.2 | 2.5 ± 0.5 | 6.0 ± 0.5 | 7.9 ± 0.4 |
| DER | 0.4 ± 0.2 | 1.5 ± 0.3 | 2.9 ± 0.4 | 5.6 ± 0.4 |
| Number of correct choices | ||||
| RER | 0.3 ± 0.1 | 1.3 ± 0.3 | 3.7 ± 0.3 | 5.2 ± 0.3 |
| DER | 0.2 ± 0.1 | 0.8 ± 0.2 | 1.6 ± 0.2 | 3.8 ± 0.3 |
| Latency to exit from the start box (s) | ||||
| RER | 57.1 ± 0.8 | 50.3 ± 1.4 | 41.4 ± 1.2 | 36.3 ± 1.5 |
| DER | 57.5 ± 0.8 | 52.9 ± 1.2 | 48.6 ± 1.2 | 40.8 ± 1.3 |
| Latency to reach the entrance of the mother-containing cage (s) | ||||
| RER | 59.3 ± 0.2 | 56.3 ± 0.9 | 50.2 ± 1.1 | 41.9 ± 1.3 |
| DER | 59.3 ± 0.2 | 57.6 ± 0.6 | 55.1 ± 0.8 | 48.3 ± 1.1 |
Tissue Preparation
Animals used for in vitro binding, immunohistochemistry, and in situ hybridization were deeply anesthetized and decapitated (on PND 13 n = 6 per group, in adulthood n = 6 per group; for each age 1 animal per litter, 6 litters per group). Brains were isolated and fast frozen in −40°C isopentane. The 2 hemispheres were separated. The left hemispheres were cut into 20-μm sections on a cryostat (Leica CM1900, Nussloch, Germany) at −17°C, mounted on silane-coated slides and stored at −80°C until further use.
High Performance Liquid Chromatography—Determination of Brain Concentrations of NE
Male rat pups (n = 11 per group, 1–2 animals per litter, 8 litters per group) or adult (∼3 month; n = 10 per group, 1–2 animals per litter, 8 litters per group) male rats of the 3 experimental groups (CTR, DER, and RER) were euthanized and their brains processed simultaneously. The PFC was dissected from the right hemisphere and processed for the high performance liquid chromatography (HPLC) analysis; NE was determined following a procedure previously described (Raftogianni et al 2012, for details see Supplementary Material).
Autoradiographic In vitro Binding of β Adrenergic Receptors
Beta adrenergic receptor binding was investigated using quantitative in vitro receptor autoradiography of the (-)-CGP-12177, [5,7-3H]- radioligand, following a procedure previously described (Raftogianni et al. 2014, for details see Supplementary Material). Nonspecific binding was determined in the presence of 1 mM isoproterenol.
Immunohistochemistry
Densities of ADRB1 (β1-AR antibody [V-19]: sc-568, Santa Cruz, USA) and pCREB (pCREB-1 antibody [Ser 133]: sc-7978, Santa Cruz, USA) immunopositive cells were determined in the adult (3 month) rat brain in the area 1 of cingulate cortex (Cg1), prelimbic cortex (PrL), infralimbic cortex (IL), and medial orbital cortex (MO). In the same brain areas, the density of the demethylase (Ma, Guo, et al. 2009; Ma, Jang, et al. 2009; Wu and Sun 2009) growth arrest and DNA-damage-inducible β (GADD45b) protein (GADD 45β antibody [H-70]: sc-33172, Santa Cruz, USA) immunopositive cells was assessed, since we found hypomethylation in the ADRB1 gene promoter region (see Photomicrograph Production, Behavioral Testing in Adulthood, and Contextual Conditioned Fear Extinction). Immunohistochemistry was performed as previously described, with a few modifications depending on the antibody used (Stamatakis et al. 2015, for details see Supplementary Material).
β1 Adrenergic Receptor mRNA In situ Hybridization
Riboprobe Preparation
We used a 364 bp DNA coding fragment of the ADRB1 gene, subcloned into a pGEM-T Easy vector. The primers used for the amplification of this particular region were sense: ATC-GCC-TCT-TCG-TCT-TCT-TCA and antisense: ACC-TTG-GAC-TCG-GAG-GAG-AA. Briefly, DNA was extracted from rat brain using the NucleoSpin® Tissue kit by Macherey-Nagel. PCR conditions were as follows: 1) initial denaturation 95°C for 1 min, 2) denaturation 95°C for 30 s, 3) annealing 60°C for 30 s, 4) extension 72°C for 1 min, and 5) final extension 72°C for 3 min. Steps 2–4 were repeated for 30 cycles. The PCR product was gel extracted (NucleoSpin® Gel and PCR Clean-up, Macherey-Nagel) and subsequently ligated into the pGEM-T Easy vector (PROMEGA) with the T4 ligase (Takara) according to the manufacturer's instructions. The identity of the insert was verified by sequencing (VBC-Biotech Service GmbH, Austria). JM109 competent cells (Takara) were transformed and colonies were selected via blue/white screening. The plasmid was linearized with SalI and was transcribed in vitro by T7 for the synthesis of the digoxigenin (DIG)-UTP-labeled antisense ADRB1 riboprobe. As a control, we used the sense ADRB1 DIG-UTP-labeled probe, which was transcribed in vitro by Sp6, using as a template the same plasmid linearized with SacII. All components of the in vitro transcription were provided by Roche.
Riboprobe Hybridization–Immunohistochemical Detection of the DIG-labeled Riboprobes
For the riboprobe hybridization and the immunohistochemical detection of the probe we followed the DIG application method for nonradioactive in situ hybridization (Eisel et al. 2008) with a few modifications (details provided in Supplementary Material).
Quantification of In situ Hybridization and Immunohistochemistry
Cell counting for ADRB1)-mRNA, as well as ADRB1, pCREB, and GADD45b-protein positive cells, was performed using the image analysis program ImageJ v.1.46R, NIH, USA (Media Cybernetics, USA) by 2 independent investigators. A “threshold” was set in the image analysis system, in order to include in the counting only cells stained above the background level. Four randomly selected sections from each animal were evaluated within the following anatomical borders, according to the anatomical atlas of Paxinos and Watson (2007): For, MO from bregma 5.16 to 4.20, for area 1 of Cg1 and PrL from 4.20 to 3.24 and for IL from 3.24 to 2.76. In each brain section, cells were counted in 2 different optical fields per area. For each animal, the number of positive cells per optical field was the average value calculated from the data from all optical fields in all brain sections evaluated.
Photomicrograph Production
High-resolution microscopic images were digitally captured using a digital CCD color video camera (TK-C1381, JVC, Japan) on an optical microscope (Eclipse E400, Nikon, Japan) using the Infinite Capture v.6.0 (Lumenera Co., Ottawa, Canada) software. For in situ hybridization for β receptor binding, high-resolution images were produced from autoradiographic films scanned using a CanoScan 8000F scanner (Canon, China). Composite photomicrographs were prepared with the Adobe Photoshop CS2 (Adobe Systems, USA).
Genomic DNA Purification, Sodium Bisulfite Treatment, and Methylation Analysis
Genomic DNA was isolated from the left PFC of rats from all 3 categories (n = 7 per group, 1 animal per litter, 7 litters per group) according to the manual of NucleoSpin® Tissue kit by Macherey-Nagel. One microgram of genomic DNA was bisulfite converted using the EpiTect Bisulfite Kit by Qiagen. Afterward, a PCR using primers targeting the ADRB1 gene (Adrb1) promoter region (851 bp) was performed (sense primer: TTT-GGA-AGA-AGT-TTG-AGT-AGG-AAA-G, antisense primer: ACC-AAC-AAT-CCC-ATA-ACC-AAA-T). The PCR conditions were as follows: 1) initial denaturation 95°C for 1 min, 2) denaturation 95°C for 30 s, 3) annealing 58°C for 30 s, 4) extension 72°C for 1 min, and 5) final extension 72°C for 3 min. Steps 2–4 were repeated for 35 cycles. PCR product was confirmed on a 1% agarose gel and by sequencing (VBC-Biotech Service GmbH, Austria). Subsequently, it was gel extracted (NucleoSpin® Gel and PCR Clean-up kit, Macherey-Nagel) and ligated into the pGEM-T Easy vector (PROMEGA) with the T4 ligase (Takara) according to the manufacture's instruction. After transformation into JM109 competent cells (Takara), individual clones were selected via blue/white screening, grown in mini cultures, plasmid DNA was purified (NucleoSpin® Plasmid kit, Macherey-Nagel), and sequenced using the Sp6 primer (VBC-Biotech Service GmbH, Austria).
Behavioral Testing in Adulthood
Olfactory Discrimination Task
Training and testing were adapted from Tronel and Sara (2002) and Tronel et al. (2004). The set up includes an open-field wooden box (60 × 60 × 60 cm), painted white, in which 3 sponges with a hole cut in the center are scented with different odors (orange, mint, almond, 15 μL of essence on the 4 corners of each sponge) and placed in the 3 corners of the box. The animals had to associate a certain odor (orange) with a food reward, which was chocolate rice crispy breakfast cereal (Chocopops, Kellogg's). The flake was placed at the bottom of the hole in the sponge so the rat had to poke its snout into the hole (nose-poke), in order to obtain the reward. The sponges with the nontarget odors did not contain food. After each trial, the position of each odor within the box changed in a random order. The experimental procedure began with 2 days of habituation each lasting 20 min, in order to familiarize the rats with the experimental box. On the third day, each animal was subjected to five trials, each lasting 5 min. On the first trial only, 4 Chocopops were also placed on the corners of the sponge embedded with the target odor, as well as in the hole, while on the rest 4 trials Chocopops were placed only in the hole of the target sponge. Forty-eight hours later, memory retrieval was evaluated in a single trial, using the same procedure as during training, for example, placing Chocopos only in the hole of the target sponge. Latency to a correct response (nose-poke into the rewarded sponge) was recorded for both training and memory retrieval. The memory retrieval index was calculated using the formula [(T5− M1)/T5], in which T5 was the latency for a correct response on the fifth trial on the day of training and M1 was the latency for a correct response on the memory retrieval trial. Before the introduction of each rat into the box, the apparatus was cleaned with a 70% ethanol solution and dried thoroughly. The number of animals utilized per group was 12 for the CTR, 9 for the RER, and 11 for the DER group (1 animal/litter).
Contextual Fear Conditioning Extinction
Training and testing were adapted from Do-Monte et al. (2010). This protocol differs from the one we have previously used (Diamantopoulou, Raftogianni, Stamatakis, Tzanoulinou, et al. 2013): In the present experiments the shock is much stronger (1 mA for 2 s compared with 0.5 mA for 1 s) and results in a ceiling effect with increased freezing behavior, for all 3 animal groups (CTR, DER, RER). Briefly, on the training day, rats were placed in the conditioning box (black box, dimensions 25 × 25 × 27 cm, Pansystems) for a habituation period of 1 min. Following habituation, subjects received 1 footshock (1 mA, 2 s) and were kept in the box for an additional minute. On the following 4 days rats were placed into the conditioning box and contextual fear memory was evaluated for 10 min, without the shock administration. Rat behavior was recorded by a video camera placed above the open shock-box and the percentage of total time each animal spent freezing was considered as a memory retrieval parameter. The extinction index represents the percentage of freezing reduction from the first to the last extinction session and was calculated with the following formula: the total time of freezing on the last extinction session was subtracted from the total time of freezing during the first extinction session and the resulting difference was divided by the total time of freezing in the first extinction session. Before the introduction of each rat into the conditioning box, the apparatus was cleaned with a 70% ethanol solution and dried thoroughly. The number of animals utilized per group was: 9 for the CTR, 10 for the RER, and 10 for the DER group (1 animal/litter).
Statistical Analysis
For all immunolabeling, in situ hybridization and methylation experiments, as well as for the olfactory discrimination task, values for the DER and RER groups were expressed as the percentage of the respective values of the control group. For HPLC experiments, in those cases where 2 animals were used from 1 litter, the average value was calculated for the 2 littermates and these means served as single data points in the subsequent statistical analyses. For all other parameters, 1 animal has been used from each litter. Data were analyzed by one-way analysis of variance (ANOVA) with the group of animals (CTR, DER, RER) as the independent factor, followed by Bonferroni post hoc tests when appropriate, in order to identify specific differences between groups. The level of statistical significance was set at 0.05. All tests were performed with the SPSS software (Release 22, SPSS, USA).
Results
PND13
NE Levels
On PND 13, 2 h after the end of the last training session in the T-maze, NE levels in the PFC were determined in CTR, DER, and RER rat pups (Fig. 1A). Statistical analysis of the NE levels in the PFC, using 1-way ANOVA with the group of animals as the independent variable revealed a significant group effect (F2,21 = 11.462, P < 0.001). More specifically, the RER had higher NE levels in the PFC compared with both Control (post hoc: *RER vs. CTR P = 0.001) and DER (post hoc: #RER vs. DER P < 0.001) animals.
Levels of NE (A) and ADRB1 binding sites (B) in the PFC of control (CTR), DER and RER pups on PND 13. RER pups had higher levels of NE than the CTR (*) or DER (#) pups, *P = 0.001 #P < 0.001. Bars represent means ± standard error of the mean (SEM); Cg1, area 1 of cingulate cortex; PrL, prelimbic cortex; IL, infralimbic cortex; MO, medial orbital cortex.
In vitro Binding for Beta Adrenergic Receptors
When levels of β adrenergic receptor binding sites on PND 13 were examined, no differences were found between the 3 groups (CTR, DER, RER) in area 1 of Cg1, PrL, IL, or MO (Fig. 1B).
Adulthood
NE Levels
In adult male rats (∼90 days), there was a statistically significant group effect on NE levels in the PFC (F2,21 = 6.309, P = 0.007; Fig. 2A). Adult RER males had higher NE levels in the PFC compared with both controls (*RER vs. CTR P = 0.026) and DER (# RER vs. DER P = 0.002) animals.
Levels of NE (A), ADRB1 binding sites (B), ADRB1 immunopositive cells (C) and ADRB1 mRNA-positive cells (D) in the PFC of adult male control (CTR), DER and RER rats. RER animals had higher levels of NE as well as ADRB1 binding sites, immunopositive cells, and mRNA than the CTR or DER. *P < 0.05 for RER versus CTR, #P < 0.05 for RER versus DER. Bars represent means ± SEM; scale bars correspond to 20 μm; Cg1, area 1 of cingulate cortex; PrL, prelimbic cortex; IL, infralimbic cortex; MO, medial orbital cortex.
In vitro Binding for β Adrenergic Receptors
Furthermore (Fig. 2B), in PrL, IL, and MO levels of β adrenergic receptor binding sites were statistically different between the 3 groups of animals (main effect of group, for PrL: F2,15 = 6.655, P = 0.009; for IL: F2,15 = 7.460, P = 0.006; for MO: F2,15 = 9.009, P = 0.003). Post hoc analysis revealed that in all these 3 prefrontal areas, RER males had higher density of β adrenergic receptor binding sites than both the CTR (*RER vs. CTR P < 0.05 for all areas) and the DER (# RER vs. DER P < 0.05 for all areas) adult males.
Immunohistochemistry for β1 Adrenergic Receptor Immunopositive Cells
The observed difference in receptor binding sites was also evident in protein levels as assessed by immunohistochemistry (Fig. 2C). Statistical analysis of the density of ADRB1 immunopositive cells revealed a significant effect of group in PrL (F2,15 = 12.862, P = 0.001), IL (F2,15 = 8.390, P = 0.004), and MO (F2,15 = 68.651, P < 0.001). Further post hoc analysis showed that in all these 3 areas RER adult males had a higher density of ADRB1 immunopositive cells than either the CTR (*RER vs. CTR P < 0.05) or the DER animals (#RER vs. DER P < 0.01).
In situ Hybridization
The higher ADRB1 binding sites and protein levels of the RER adult males posed the question of whether the ADRB1 gene expression was also higher in this experimental group. Statistical analysis of the levels of ADRB1 mRNA-positive cells (Fig. 2D) revealed a significant effect of group in the PrL (F2,15 = 10.970, P = 0.001), IL (F2,15 = 15.052, P < 0.001), and MO (F2,15 = 11.923, P = 0.001). Post hoc analysis showed that in all these 3 areas, RER adult males had higher levels of ADRB1 mRNA-positive cells than either the control (*RER vs. CTR P < 0.05) or the DER animals (#RER vs. DER P = 0.001).
Immunohistochemistry for pCREB Immunopositive Cells
In addition to the increased ADRB1 binding sites, protein and mRNA levels, elevated levels of pCREB were detected in the adult RER males, indicative of elevated ADRB1 functionality (Fig. 3). Statistical analysis of the density of pCREB immunopositive cells revealed a significant effect of group in the PrL (F2,15 = 70.670, P < 0.001), IL (F2,15 = 59.134, P < 0.001), and MO (F2,15 = 81.631, P < 0.001). Further post hoc analysis showed that in all these 3 areas RER adult males had a higher density of ADRB1 immunopositive cells than either the CTR (*RER vs. CTR P < 0.001) or the DER animals (#RER vs. DER P < 0.001).
Levels of pCREB immunopositive cells in the PFC of adult male control (CTR), DER, and RER rats. RER animals had higher levels of pCREB immunopositive cells than the CTR or DER. *P < 0.05 for RER versus CTR, #P < 0.05 for RER versus DER. Bars represent means ± SEM; scale bars correspond to 20 μm; Cg1, area 1 of cingulate cortex; PrL, prelimbic cortex; IL, infralimbic cortex; MO, medial orbital cortex.
Behavioral Testing
In order to corroborate our neurochemical findings on the level of the function of the PFC, we employed 2 distinct behavioral tasks that depend on the noradrenergic neurotransmission in the PFC.
Olfactory Discrimination Task
Statistical analysis of the memory retrieval index showed a significant group effect (F2,28 = 3.474, P = 0.045; Fig. 4A). Post hoc analysis revealed that RER males remembered approximately 2 times better than both control (*RER vs. CTR P = 0.036) and DER (#RER vs. DER P = 0.022) adult animals.
Performance of control (CTR), DER, and RER adult male rats in the Olfactory Discrimination task (A) and the Contextual Conditioned Fear Extinction task (B). RER animals had a better performance in both tasks: They exhibited superior olfactory memory and extinguished fear more effectively than the CTR or the DER. *P < 0.05 for RER versus CTR, #P < 0.05 for RER versus DER. Bars represent means ± SEM.
Contextual Conditioned Fear Extinction
On the first day of memory testing, there was no statistically significant difference in the amount of freezing time between the 3 groups of animals (CTR, DER, and RER), documenting that all 3 experimental groups acquired the conditioned stimulus–unconditioned stimulus association at a similar level. However, RER males showed increased ability for fear extinction (Fig. 4B), as was evident by the fear extinction index (F2,26 = 6.589, P = 0.005; post hoc test, *RER vs. CTR P = 0.033, #RER vs. DER P = 0.006).
β1 Adrenergic Receptor Gene Promoter Methylation
Given our finding that ADRB1 gene (Adrb1) expression was increased in RER adult males compared with CTR and DER animals, we investigated the methylation profile of the promoter region of the Adrb1 gene (Fig. 5). Statistical analysis of the methylation pattern of the promoter of the Adrb1 in the PFC showed a significant group effect (F2,18 = 5.190, P = 0.017). Post hoc analysis revealed lower total methylation in the promoter region of the Adrb1 gene in the PFC of adult RER males (Fig. 5B) compared with both control (*RER vs. CTR P = 0.040) and DER (#RER vs. DER P = 0.017) adult males.
Methylation analysis of the ADRB1 gene (Adrb1) promoter region of adult male control (CTR), DER, and RER rat PFC. (A) Sequence from −389 to −261 including the core promoter (gray box; black lettered sequence). CpG dinucleotides are underlined. (B) Levels of methylation at CpG dinucleotides. The Adrb1 gene promoter region was less methylated in the RER animals compared with the CTR (*) or the DER (#), *P < 0.040, #P < 0.017. Bars represent means ± SEM.
Immunohistochemistry for GADD45b Immunopositive Cells
The decreased methylation of the ADRB1 gene (Adrb1) promoter of the RER males prompted us to examine the levels of GADD45b (Fig. 6) which has been found to actively demethylate gene promoters (Ma, Guo, et al. 2009; Ma, Jang, et al. 2009; Wu and Sun 2009). Statistical analysis of the density of GADD45b immunopositive cells revealed a significant effect of group in the PrL (F2,15 = 31.754, P < 0.001), IL (F2,15 = 26.548, P < 0.001), and MO (F2,15 = 36.205, P < 0.001). Further post hoc analysis showed that in all these 3 areas RER adult males had a higher density of ADRB1 immunopositive cells from either the control (*RER vs. CTR P < 0.001) or the DER animals (#RER vs. DER P < 0.001).
Levels of GADD45b immunopositive cells in the PFC of adult male control (CTR), DER, and RER rats. RER animals had higher levels of GADD45b immunopositive cells than the CTR or DER. *P < 0.05 for RER versus CTR, #P < 0.05 for RER versus DER. Bars represent means ± SEM; scale bars correspond to 20 μm; Cg1, area 1 of cingulate cortex; PrL, prelimbic cortex; IL, infralimbic cortex; MO, medial orbital cortex.
Discussion
In the present work we employed an experimental model, developed in our laboratory, which involves 2 early experiences the DER and the RER (for details, see Materials and Methods): Pups subjected to the RER experience are allowed entry into the mother-containing cage upon reaching it in a T-Maze and thus receive, in each trial, the expected reward of maternal contact, a potent reinforcer. In contrast, the pups exposed to the DER experience are not permitted entry into the mother-containing cage, in spite of being in close proximity to her and receiving maternally derived olfactory stimuli. Eventually, after 10 one minute trials in which the pups are denied the expected reward of maternal contact, they are reunited with their mother and the rest of the litter and receive with a delay the reward of maternal care. Notably, both the DER and RER pups, upon return to the home cage, receive increased maternal care, compared with the controls (Diamantopoulou, Raftogianni, Stamatakis, Tzanoulinou, et al. 2013).
The DER experience could be considered analogous to Amsel's “frustrative non-reward,” since there is omission of immediate reinforcement through the maternal contact (Stout et al. 2003). However, the DER experience lacks the basic characteristic of Amsel's model, the “surprisingness” of the reward omission (Papini 2003; Stout et al. 2003). If one was to compare the DER experience of our model to the postulates of Amsel's frustration theory, it is the “secondary frustration” with which it bears some resemblance (Papini 2003): through the repetition of the non-rewarded trials, the pups have come to expect the omission of the reinforcer, face an approach-avoidance conflict, manifested behaviorally as a reduced motor approach to the goal and less effective learning (Panagiotaropoulos et al. 2009). The DER experience involves incongruence between when the receipt of the reward of maternal contact is expected (upon reaching the mother-containing cage) and when it actually happens (after the end of the 10 trials). Such incongruence has been reported to be related to the dopaminergic system (Cardinal 2006). Interestingly, the DER pups have reduced dopamine levels in their PFC (Raftogianni et al. 2014), which could imply a deficit in the “wanting” component of the reinforcer (Berridge 2012) and hence the mother as a stimulus is not invested with high enough incentive salience to mobilize resources and thus increase NE (Varazzani et al. 2015) in the PFC and dopamine in the accumbens (Raftogianni et al. 2014). In contrast, in the RER experience continuous reward received by the pups increases the incentive salience of the mother as the reward stimulus. Thus reaching the mother-containing cage becomes a goal for the achievement of which the pups recruit a maximum of resources. Such energy demanding situations have been shown to increase activity in the locus coeruleus (Bouret et al. 2012; Sara and Bouret 2012; Bouret and Richmond 2015; Varazzani et al. 2015), which would lead to the observed herein increased levels of NE in the PFC. The increased NE in the PFC of the 13-day-old RER pups could in turn mediate the increase in accumbal dopamine, shown previously to be induced by the RER experience (Raftogianni et al. 2014). Our present results are in accord with the proposal of Ventura that prefrontal NE is important in ascribing motivational salience to highly salient stimuli by regulating the release of dopamine in the accumbens, in response to rewarding stimuli (Puglisi-Allegra and Ventura 2012), and furthermore document that this function of NE is already present during the neonatal period.
The increased NE levels could further enhance the salience of the mother (Cirulli et al. 2003) as the stimulus and thus act as the driving force for the increased motivation and more efficient learning in which the RER pups display in the T-maze (Panagiotaropoulos et al. 2009). Relevantly, NE, acting through β adrenergic receptors, has been well documented to play a crucial role in another form of learning during the neonatal period, namely associative olfactory learning (Sullivan et al. 1992, 1994, 2000; Moriceau and Sullivan 2004).
Interestingly, the increased levels of NE in the PFC are maintained into adulthood, albeit to a lesser degree. However, in adulthood RER males, in addition to the increased NE levels, also show increased levels of ADRB1ADRB1 as assessed by the density of both its binding sites and immunopositive cells. We furthermore demonstrate that the levels of pCREB—a downstream effector of the receptor (Roseboom and Klein 1995)—are increased in the PFC of the RER males, indicating a more effective receptor function. ADRB1 mRNA-positive cells were also more abundant, implying increased transcription of the gene, which could be related to the hypomethylation of its promoter region, also documented herein. The decreased methylation of ADRB1 gene promoter could be affected by the demethylase GADD45b, whose levels are shown by our present results to be increased by the RER experience. The neurochemical findings showing increased activity of the PFC noradrenergic system are corroborated by our behavioral results showing that adult RER males have superior performance (compared with the controls or the DER) in olfactory discrimination (Tronel and Sara 2002, 2004) and contextual fear extinction (Do-Monte et al. 2010), 2 tasks which depend on PFC noradrenergic function.
There are a few reports in the literature on the effects of other early experiences on the noradrenergic system. More specifically, maternal separation resulted in higher paraventricular hypothalamic nucleus NE responses to restrain stress, while neonatal handling had the opposite effect (Liu, Caldji, et al. 2000). In general, neonatal handling has been shown to reduce the excitability of the noradrenergic system, as assessed by the smaller volume and number of cells in the locus coeruleus (Lucion et al. 2003), the increased α2 adrenergic receptor levels on these cells (Liu, Caldji, et al. 2000), the down-regulation of α1 adrenergic receptors in the Cg1 and the hippocampus (Coccurello et al. 2014), as well as the decreased cAMP accumulation following β adrenergic receptor activation (Escorihuela et al. 1995; Baamonde et al. 2002).
The enhanced activity of the PFC noradrenergic system of the adult RER males is reflected in their behavioral phenotype. Previous results from our laboratory have shown that RER males exhibit increased climbing in the forced swimming test (Diamantopoulou, Raftogianni, Stamatakis, Oitzl, et al. 2013), a behavior reported to depend on PFC noradrenergic neurotransmission (Detke et al. 1995). Within the framework of the present work, we show that adult RER males have a superior performance in olfactory discrimination and contextual fear extinction, 2 tasks which are dependent on the PFC noradrenergic system, both the levels of NE and β adrenergic receptors (Tronel et al. 2004; Mueller et al. 2008; Do-Monte et al. 2010). More specifically, in the former task RER animals remembered better the cue-reward association and in the latter they extinguished learned fear more effectively, processes that are controlled by the PrL (Tronel et al. 2004) and IL (Barad 2005; Do-Monte et al. 2010), respectively. Notably, in both of these PFC subregions the RER experience resulted in increased ADRB1 levels and function. The enhanced performance of the RER animals in both of the above behavioral tasks—on a hyperactive noradrenergic neurobiological substrate—could reflect a state of increased arousal, leading to a greater ability to assess and ascribe salience to environmental stimuli (the cues in the tasks). This capability could be a long-term outcome of their neonatal RER experience, in which the cue (the mother) was of the outmost salience for the pups.
The long-term effects of early experiences have been shown to be mediated by epigenetic changes in the regulatory regions of target genes (Kundakovic and Champagne 2014; Maccari et al. 2014). Our results extend such findings to an additional gene, the ADRB1 gene (Adrb1), which has not been up to now identified as the target of an early experience. More specifically, we show that the RER experience results in hypomethylation of the promoter region of the Adrb1 in the PFC, an epigenetic effect which could be related to the observed increased levels of ADRB1 mRNA and protein in this same brain region. The decreased methylation of the Adrb1 promoter could be attributed to the increased GADD45b levels also found in the RER males. GADD45b has been shown to actively demethylate gene promoter regions (Ma, Guo, et al. 2009; Ma, Jang, et al. 2009; Wu and Sun 2009) and to be involved in activity-dependent neural plasticity (Nedivi et al. 1993; Lemberger et al. 2008; Ma, Jang, et al. 2009). Most interestingly, another early experience, caregiver maltreatment, has been shown to result in decreased GADD45b (Blaze and Roth 2013). Relevant to our present work are the reports that GADD45b expression is mediated through CREB (Carrier et al. 1998; Lemberger et al. 2008; Tan et al. 2012). We thus propose the following as a possible mechanism underlying our neurochemical findings: The RER experience-induced increased NE, acting through its adenylate cyclase activating receptors—among which the ADRB1 is the most abundant in the PFC—results in elevated levels of pCREB; this in turn stimulates expression of GADD45b which mediates active demethylation of the ADRB1 gene promoter leading to its enhanced transcription reflected in the increased levels of this receptor.
The RER experience-induced increase in NE in the neonatal PFC appears to act in such a way as to program the function of the noradrenergic system at an augmented level, both at the neurochemical and behavioral level, in adulthood.
Funding
This work has been supported by the John S. Latsis Public Benefit Foundation. The sole responsibility for the content lies with its authors. T.K. was supported by a “NIK. D. CHRYSOVERGIS”-“IKY” (Hellenic State Foundation for Scholarships) scholarship.
Notes
Conflict of Interest: None declared.
References
