Consequences of early life stress on genomic landscape of H3K4me3 in prefrontal cortex of adult mice
Maternal separation models in rodents are widely used to establish molecular mechanisms underlying prolonged effects of early life adversity on neurobiological and behavioral outcomes in adulthood. However, global epigenetic signatures following early life stress in these models remain unclear.
In this study, we carried out a ChIP-seq analysis of H3K4 trimethylation profile in the prefrontal cortex of adult male mice with a history of early life stress. Two types of stress were used: prolonged separation of pups from their mothers (for 3 h once a day, maternal separation, MS) and brief separation (for 15 min once a day, handling, HD). Adult offspring in the MS group demonstrated reduced locomotor activity in the open field test accompanied by reduced exploratory activity, while the HD group showed decreased anxiety-like behavior only. In a group of maternal separation, we have found a small number (45) of slightly up-regulated peaks, corresponding to promoters of 70 genes, while no changes were observed in a group of handling. Among the genes whose promoters have differential enrichment of H3K4me3, the most relevant ones participate in gene expression regulation, modulation of chromatin structure and mRNA processing. For two genes, Ddias and Pip4k2a, increased H3K4me3 levels were associated with the increased mRNA expression in MS group.
The distribution of H3K4me3 in prefrontal cortex showed relatively low variability across all individuals, and only some subtle changes were revealed in mice with a history of early life stress. It is possible that the observed long-lasting behavioral alterations induced by maternal separation are mediated by other epigenetic mechanisms, or other brain structures are responsible for these effects.
KeywordsChIP-seq Early life stress Maternal separation Handling H3K4me3 Prefrontal cortex Mice
A growing body of findings underlines the critical role of the early environment in the development of the nervous system and for the behavioral phenotype in later life [1, 2]. The adverse early life events including stress or maltreatment are known to have long-lasting effects on brain function, cognitive and emotional development and can influence the risk to develop stress-related psychopathology in adulthood [2, 3, 4]. Clinical studies have revealed a strong link between childhood maltreatment and the development the risk of psychiatric disorders and health risk behaviors including smoking, overeating, and substance abuse [4, 5, 6, 7].
However, human studies have some limits to reveal the molecular mechanisms underlying long lasting effects of early life adversity on cognitive functions and behavior. So, various animal models have been developed to investigate the molecular consequences of early life stressful events. Among them, maternal separation models in rodents are the most commonly used and established ones [8, 9, 10]. These models have contributed significantly to our knowledge of a link between early life exposures and neurobiological and behavioral outcomes in adulthood. In particular, studies on rats demonstrated that a brief maternal separation (15 min per day; handling, HD) followed by increased maternal care positively affects the development of the offspring, leading to reduced anxiety, increased exploration and communicative behavior. At the same time, prolonged maternal separation (3 h per day; MS) causes significant amounts of stress resulting in negative long-lasting changes of emotion-related behavior, stress reactivity and cognitive functions [11, 12, 13, 14]. Data on the effects of MS and HD on behavioral phenotypes, long-term memory and learning in mice are sparse and often inconsistent [15, 16, 17]. However, most works agree in that in mice, as in rats, MS leads to anxiety-related behaviors and increased stress reactivity [16, 17, 18, 19].
Epigenetic processes such as DNA methylation and post-translational histone modifications are known to play key roles in memorizing of environmental influences [20, 21]. In this way, epigenetic modifications could potentially mediate embedding early social exposure in the genome and provide the long-lasting nature of induced changes in neurobiology and behavior [4, 22, 23]. Therefore, a number of studies have assessed the epigenetic alterations following early postnatal stress, including maternal separation. However, until recently the focus has been on individual genes, mainly involved in the regulation of hypothalamic-pituitary-adrenal (HPA) axis. For example, in rats subjected to maternal separation significant reduction in histone H3 acetylation at glucocorticoid receptor (GR) promoter I7 and significant decreases in total and exon I7 GR mRNA levels were observed in the hippocampus during adulthood . In contrast, increased H3 acetylation and hypomethylation of the Crh promoter region underlying the upregulation of Crh was demonstrated in the hippocampal CA1 region of the adult rat with postnatal MS . Similarly, Chen and colleagues reported hypomethylation of the Crh promoter in the PVN of maternally deprived adult rats that correlates with enhanced Crh transcriptional response to stress in adulthood . Generally, these findings confirm the idea that early life stress induces epigenetic signatures that can lead to changes in gene expression and behaviors.
Since histone modifications can mediate physiological consequences in response to environmental pressure in early life [27, 28], we undertook a genome-wide investigation of the histone H3K4 trimethylation, a key epigenetic mark associated with transcriptional activation, in the prefrontal cortex (PFC) of adult maternally separated (MS or HD) male mice compared to control ones. Given the fact that PFC is responsible for complex cognitive functions, regulation of emotion, and adaptation to stress, we expected to reveal some alterations in promoter-associated H3K4me3 peaks in this brain region reflecting the long-lasting consequences of stress in early life.
C57BL/6 mice were housed in the Center for Genetic Resources of Laboratory Animals (RFMEFI61914X0005 and RFMEFI62114Х0010), Institute of Cytology and Genetics, the Siberian Branch of the Russian Academy of Sciences, Novosibirsk, Russia. The animals were housed under the standard conditions (12/12 h light/dark regime, switch-on at 7.30 a.m.; food and water available ad libitum). Bottom of the cage was covered with litter. Pregnant females were provided with paper towels for nest building.
Maternal separation procedure
Open field test
Open field test was conducted at PND95 – PND110. Animals were individually housed for 3 days before the test. The test was started at 1 p.m., in the middle of light phase. Open field (OF) consisted of a square arena (80 × 80 cm) with white floor and 25-cm-high walls. The arena was lit by bright light and divided into central zone (40 × 40 cm) and peripheral zone (20 cm from the walls). Each mouse was placed individually in the center of the OF arena, and the following behavioral parameters were recorded during 5-min test period: the total distance travelled, latency of the first exit from the central zone, frequency of central zone visits, time spent in central and peripheral zones; the number of rearing and self-grooming episodes; total activity (Ethovision threshold for activity: 0.2%). The OF arena was thoroughly cleaned between tests with different animals. OF behavior was monitored and scored using an automated tracking system (Noldus Ethovision 10.0, Noldus Information Technology, The Netherlands).
Four mice from each group (control, HD and MS) were chosen randomly for chromatin immunoprecipitation with antibodies to H3K4me3 (ab8580, Abcam). Animals were euthanized 24 h after the OF test. Prefrontal cortex were dissected and stored at − 70 °C until further processed. Chromatin immunoprecipitation were carried according to the method  with some modifications. To account for possible batch effects, experimental groups were randomized during the preparation of immunoprecipitation. For DNA fragmentation microccocal nuclease (MNase), that selectively spares nucleosome-bound genomic DNA, was used. Each sample (about 20 mg of murine prefrontal cortex) was homogenized in douncing buffer (10 mМ Tris-HCl pH 7.5, 4 mM MgCl2, 1 mM CaCl2) containing Halt Protease Inhibitor Cocktail (Thermo Scientific, USA) and then incubated with 1 μl of MNase (2000 gels units, NEB, USA) for 6 min at 37 °C. Reaction was stopped by adding EDTA to 10 mM. 4-fold volume excess of hypotonic lysis buffer (0.2 mM EDTA, 1.5 mМ DTT, Halt Protease Inhibitor Cocktail) was added to each sample and samples were incubated on ice for 60 min, with a brief vortexing at 10 min intervals. Debris is then removed by centrifugation at 3000×g. To reduce non-specific binding and background impurity (pre-clearing of samples) G-protein magnetic beads (NEB, USA) and 10X IP buffer (200 mM Tris HCl (pH 8.0), 20 mM EDTA, 1,5 M NaCl) were added to samples and incubated at 4 °C for 2 h with rotation. Then magnetic beads with non-specific binding fragments were collected using the magnetic stand and discarded. 10% of pre-cleared sample was saved as input DNA (Inp) for subsequent ChIP enrichment analysis. At the same time H3K4me3 antibody (3 μl) were incubate with G-protein magnetic beads with rotation at 4 °C for 2 h. Supernatant were discarded from antibody-coated beads, beads were combined with pre-cleared chromatin and incubated overnight at 4 °C with rotation. After that beads together with the captured immune complexes were washed twice by 1 ml Low Salt (0.1% SDS, 1% Тriton Х-100, 2 mМ EDTA, 20 mM Тris-HCl (рН 8.0), 150 mM NaCl), High Salt (0.1% SDS, 1% Тriton Х-100, 2 mМ EDTA, 20 mМ Тris-HCl (рН 8.0), 500 mМ NaCl), LiCl (1% IGEPAL-СА 630, 1% deoxycholic acid, 1 mМ EDTA (рН 8.0), 10 mМ Тris-HCl (рН 8.0), 0.25 М LiCl) and TE (10 mМ Тris-HCl (рН 8.0), 1 mМ EDTA) buffers. Captured IP-complexes were eluted by incubation for 1 h at 65 °C in 50 μl of freshly prepared elution buffer (100 mМ NaHCO3, 1% SDS). Supernatant was transferred in a new tube. After that, beads were washed by 50 μl of elution buffer and supernatant was collected in the same tube. Samples and input DNA were proteinase K and RNaseA treated, and then purified using phenol chloroform extraction, and DNA finally were eluted in 30 μl of 4 mM Tris-HCl (pH 8.0).
Quantitative real time PCR was carried out for each sample and inputs with positive and negative control primers for ChIP enrichment analysis. The promoter region of Rpl30, located within the active chromatin and enriched with H3K4me3 according to the UCSC Genome Browser data was selected as a positive control. The promoter region of Oosp3, which is expressed only in oocytes and is not enriched with H3K4me3 in prefrontal cortex of mice (UCSC Genome Browser data), was used as a negative control. The following primer pairs were designed: Rpl30 forward – 5-ACTTTGCACAGGGACCACAA, reverse – 5-TTACCCGTCAGCCACTTCAC; Oosp3 forward – 5-ACAGCATTGTGTCAGCATCCCTAAA, reverse – 5-GCCTGAATATGCTTGTCTAGGTGGC. Fold enrichment (FE) was estimated for each sample: FE = 2(Ct Inp – Ct IP) pos/2(Ct Inp – Ct IP) neg. Only samples with FE > 25 were used for library preparation.
Libraries were prepared according to standard New England Biolabs protocols (USA). The DNA was end-repaired using T4 DNA polymerase and Klenow DNA polymerase. After that, an A base was added to the 3′ end of the blunt phosphorylated DNA fragments, and an Illumina adaptor with a single T overhang at its 3′ end was then ligated to the end of the DNA fragment. Then USER™ enzyme was added for cutting closed adaptors. Size selection of DNA fragments was performed by means of Agencourt AMPure XP Beads (Beckman Coulter) and then PCR enrichment of the adapter-ligated library was conducted (eight cycles of PCR). The size and quantity of the library were verified on the Agilent Bioanalyser. Preliminary low-depth sequencing using the MiSeq Illumina was essential for correct quantity determination. Subsequent paired-end (2 × 100) sequencing was performed using the Illumina HiSeq2500 (Ltd Genotek, Moscow, Russia).
ChIP-seq data analysis
On average, ~ 10 million paired-end reads (8.8–10.4 million) were obtained for each sample. Paired-end sequencing data were preprocessed by trimmomatic  adapter removal tool and mapped to GRCm38/mm10 mouse reference genome using bowtie2 aligner . The quality metrics of ChIP-seq libraries (Additional file 1: Table S1) were assessed by phantompeakqualtools software (https://www.encodeproject.org/software/phantompeakqualtools/). MACS2  algorithm with nucleosome-optimized parameters (−-shift 37 --extsize 73) was applied to call both broad and narrow peaks from the pooled data. Public sequencing data on MNase-treated input samples of mouse PFC neurons (NCBI SRA accessions: SRR5032624, SRR1647895) were served as background in peak calling procedure. GENCODE comprehensive gene annotation release M13 (GRCm38.p5, https://www.gencodegenes.org) was used to link the predicted peaks to the nearest TSS of genes. An analysis of individual variability in ChIP-seq data was performed in R programming environment (https://cran.r-project.org) on a subset of broad H3K4me3 peaks located in promoter regions (±1Kb around TSS of annotated genes) and required to have a peak-calling q-value ≤0.01.
For differential enrichment (DE) analysis, DESeq2  and csaw  R packages were used. In DESeq2 workflow, reads with mapping quality (MAPQ) of at least 10 were counted at the predicted peak regions. Peaks overlapping the DAC blacklisted regions (ENCODE accession ENCFF547MET) were excluded from the subsequent analysis. Normalization, modeling, and statistical testing were done with default parameters. At most 10% FDR (assessed by Benjamini–Hochberg procedure) was allowed in the resulting list of DE peaks. Two additional filtering criteria, the location in promoter regions of known genes (±1Kbp around TSS) and the absence of substantial input bias in the corresponding locus, were applied to produce the final list of DE peaks.
Similarly, in csaw workflow, reads with MAPQ> 10 were counted in 150-bp genomic windows excluding the ones from DAC blacklisted regions. Count data were filtered by 5-fold change over global enrichment and normalized on efficiency bias. Subsequent statistical testing with controlling 10% FDR did not reveal any DE windows in the data.
Frozen tissue samples of prefrontal cortex were treated with TRIzol Reagent (Ambion, USA) following the manufacturer’s instructions and the total isolated RNA was purified by Agencourt RNAClean XP Beads (Beckman Coulter). In each sample RNA quality and yield was confirmed by NanoDrop 2000. One microgram of the total RNA was taken for cDNA synthesis using a reverse transcription kit (Syntol, Russia), with a random hexanucleotide mixture as primers. All procedures were carried out according to manufacturer’s instructions.
Real time PCR
A thermocycler CFX96 (Bio-Rad, USA) was applied for quantitative real-time PCR using TaqMan probes. Primers and TaqMan probes (Syntol, Russia) were designed (Additional file 2: Table S2) using the Primer BLAST (NCBI). The reaction mixture consisted of 0.25 mM dNTP, 2 μl 10Х PCR AS buffer-B (15 mM MgCl2, 650 mM Tris-HCl (pH 8.9), 160 mM (NH4)2SO4, 0.5% Tween 20), 10 pmol of each forward and reverse primers, 10 pmol TaqMan probes, 4 μg cDNA sample, 0.3 U SynTaq polymerase (Syntol, Russia), and ddH2O to a final volume of 20 μl. The PCR program consisted of 39 cycles of 95 °C for 30 s, and 60 °C for 30 s. All qPCR reactions were performed in duplicates. The qPCR data were analyzed using the ΔΔCt method and normalized to the housekeeping genes hypoxanthine guanine phosphoribosyl transferase (Hprt1) and ribosomal protein 16S (Rpl16S). The stability of the two housekeeping genes between and within groups was verified using the Bio-Rad CFX Manager software (Bio-Rad, USA) with the gene stability value M, set at less than 0.5 and coefficients of variation CV, at less than 0.25.
Open field behavior
In OF test, maternal separation had the greatest effect on the locomotor activity of animals (Fig. 1b–f). One way ANOVA revealed the influence of stress on distance travelled [F(2,24) = 3.6; p = 0.043] and total activity [F(2,24) = 5.2; p = 0.013]. MS group demonstrated significantly less distance travelled compared with HD group (p = 0.013) and control (a trend, p = 0.074). Decreased locomotor activity is confirmed by a lower level of total activity in MS group compared to control (p = 0.0096) and HD (p = 0.004) groups. The decreased locomotor activity accompanied by reduced exploratory activity [F(2,24) = 4.9; p = 0.017] – the duration of vertical movements (rearings) was lower in MS group (p = 0.023) than in the controls. HD mice showed normal locomotor activity but increased frequency of entries into the central zone of the open field (p = 0.017) compared with control, which indicates a decreased anxiety level in HD group [F(2,24) = 3.7; p = 0.039]. HD group also showed a reduced exploratory activity estimated by decreased duration of rearings (p = 0.014).
To assess whether the adults with a history of early life stress possess any long-term changes in H3K4me3 profile of PFC, we performed the native MNase-treated ChIP-seq analysis targeting the studied histone mark in mice subjected to brief and prolonged maternal separation and control animals. The approach produces histone profiles with a much better signal-to-noise-ratio compared to protocols utilizing formaldehyde fixation . All the libraries except one (discarded) were sequenced at sufficient depth and demonstrated high enrichment values as well as passed all the other quality estimators recommended by ENCODE Consortium  (Additional file 1: Table S1).
A correlation analysis performed on 1000 most variable broad peaks showed that most of them correlated significantly (r > =0.8, p-value<=0.01) with each other. Since positive correlation alone may arise from systemic technical bias, we further extracted a subset of negatively correlated peaks, each required also to have a positive correlation within a given subset. Seven hundred eighteen peaks met the requirements and were subsequently split into 2 opposing clusters (Fig. 3c). The latter result assumes that the fluctuations in the most variable peaks are mainly caused by a single factor. Respectively, PCA analysis on a whole set of 12,112 broad peaks showed that the corresponding PC1 factor explains 27% of observed variation (and 53% in a subset of 718 peaks). However, it does not correspond to any of experimentally controlled factors, e.g. stress condition, litter/kinship, or library enrichment.
We then assessed whether the observed covariation of H3K4me3 peaks is functionally meaningful. Functional enrichment analysis using MSigDB and DAVID resources showed high functional relatedness and tight association of corresponding genes within the clusters and a clear distinction between them (Additional file 3: Table S3). In particular, the first cluster is overrepresented by neuron-specific genes related to axon guidance, synapse assembly, and synaptic (including GABAergic and glutamatergic) transmission. The second cluster is enriched with more general nuclear transcription factors and chromatin regulators, including several histone modifiers (Mll2, Nsd1, Setd1b, Ash1l, Ezh2). Thus, the observed low individual variation appears to be functionally relevant and is partially described by a single factor, albeit not directly related to the factor of stress conditions.
Chip-seq DE analysis
To analyze the differential enrichment of H3K4me3 profiles we used both peak-based MACS-DESeq2 and window-based csaw workflows; both make use of biological replicates in statistical testing, and the latter does not require peak regions as a prerequisite. Csaw didn’t reveal any statistically significant changes in both experimental groups versus controls. The DESeq2-based approach also failed to detect any in a group of handling, but detected a small subset of modestly altered broad peaks in a MS group (Additional file 4: Table S4). 45 regions of H3K4me3 modification were found to be increased, affecting promoters of a total of 70 genes. Thereby, a substantial part of the differentially up-regulated marks (42%) was detected within the bidirectional promoters.
To compare the distribution of over-represented pathways between the stressed groups (MS and HD) we performed a Gene Set Enrichment Analysis (GSEA) of pre-ranked gene lists (using ranking by p-value and Reactome gene sets of size 5 to 500). HD group was found to be enriched with the same Reactome pathways as MS group, while the latter showed much broader set of over-represented pathways (140 vs 22, Additional file 5: Table S5). This may indicate the similarity of stress response in both studied groups, either more pronounced in MS group. Common categories are connected to RNA processing and transcription regulation, and in a less degree to metabolic processes. Additionally, MS group was enriched by some neuronal and immune pathways as well as pathways of vesicular and membrane transport.
Thus, the results of the enrichment analysis indicate non-randomness of the revealed weak changes in H3K4me3 profiles in MS group of mice. These also suggest that the consequences of early postnatal stress are manifested in epigenetic changes of genes that are one way or another regulate gene expression.
In the present study we examined a model of different type of stress in the early life period, which have been shown to have different impact on the behavior of adult mice. Indeed, in our study a prolonged maternal separation (MS) resulted in alterations of individual behavior, namely a stable decrease of locomotor and exploratory activities, while brief maternal separation (HD) induced a decreased anxiety in adulthood. Similar opposing effects of brief and prolonged early life stress were previously shown in behavioral studies on rats [41, 42, 43]. In mice, the consequences of MS and HD on behavior are much less studied, with some findings being contradictory. Nevertheless, maternal separation of pups from dams generally increases anxiety [18, 19, 44] and in some cases results in depressive-like behavior of adult male mice . Several studies also demonstrated a decreased anxiety in adult male mice experienced brief maternal separation [17, 46]. Our results confirm the presence of adverse long-term effects of prolonged maternal separation in early life on behavioral traits and beneficial behavioral effects of handling in mice.
There are provisional grounds to consider that the long-term behavioral consequences of early life stress are likely maintained by epigenetic mechanisms. Thus, a number of studies show stable changes in CpG methylation in promoter regions of genes encoding glucocorticoid receptor [45, 47, 48, 49], BDNF [45, 50, 51] and others  caused by early postnatal stress. Moreover, maternal separation alters the expression levels in conjunction with the levels of histone H3 acetylation and/or some repressive modifications (H3K9me2, H3K9me3) in promoter regions of GR, Crh, Bdnf, etc. [24, 51, 53, 54, 55, 56].
The choice of H3K4me3 modification in our study was dictated, on the one hand, by the fact that it is relatively long-lasting (compared to histone acetylation or phosphorylation) to mediate the observed durable changes of behavior . On the other hand, it is tightly associated with promoter regions, thus bringing direct link of its levels with transcriptional activity of the corresponding genes [57, 58]. The maintenance of Mll1-regulated H3K4 methylation at particular genes in PFC was shown to be critical for normal cognitive abilities and emotional state . Besides, its level was shown to be altered in rat hippocampus after the chronic stress while remain unchanged after the acute stress .
The profiles of H3K4me3 identified in our study showed surprisingly low interindividual variability. While the source of this variation appeared to be systemic and biologically relevant, it didn’t correspond to the experimental conditions of stress. Consequently, we have found only some subtle changes of H3K4me3 levels that exceeded a relaxed significance threshold (10% FDR) in a group of adult mice experienced prolonged early life stress (MS) compared to controls. Brief separation (HD) did not lead to any significant differences in the distribution of this histone mark at all.
Among the genes, whose promoter regions were found to be differentially trimethylated, there are several ones implicated in neuronal function and chromatin regulation. Thus, the level of H3K4 trimethylation was affected in several genes belonging to glutamatergic neurotransmission. Grm3 encodes the glutamate receptor mGluR3, a major regulator of PFC function and cognition , whose disruption leads to working memory defects , as well as development of bipolar disorder . Sdf2 protein was found in a complex with another glutamate receptor, mGluR1b, suggesting its involvement in folding and transport of the latter . Pclo encodes the neurospecific Piccolo protein highly expressed in cerebral cortex and participating in synaptic vesicle trafficking. Piccolo plays a pivotal role in synaptic plasticity, controlling the extracellular levels of glutamate . Phosphatidylinositol 4,5-bisphosphate (PtdIns(4,5)P2), generated by enzymatic action of Pip4k2a gene product, plays an important role in membrane trafficking, regulation of exocytotic fusion of synaptic vesicles (glutamate, dopamine) with the plasma membrane and regulates a duration of signal transduction [66, 67, 68]. Among the other affected genes closely related to neuronal functions, there are Arl6ip6, March1 and Wasl, that regulate dendritic spine formation and maintenance [69, 70, 71], as well as Zfp608, Flrt2 and Atad2b, involved in axonogenesis and neuronal cell differentiation and migration in the developing cortex [72, 73, 74].
Furthermore, the presence of several genes associated with chromatin remodeling may indicate the involvement of chromatin architecture in mediating the consequences of early life stress. A histone chaperone Supt6 is implicated in maintenance of chromatin structure [75, 76]. A product of Ubr4 appears to be a cytoskeletal component in the cytoplasm and a subunit of the chromatin scaffold in the nucleus . There are also genes related to epigenetic inactivation of transcription, encoding histone methyltransferase Ash1l, specifically methylating ‘Lys- 36’ of histone H3 (H3K36me) , Cir1 and Phf12, involved in the assembly of histone deacetylase complexes [79, 80], as well as a gene associated with activation of transcription, Wdr77, encoding a subunit of methyltransferase complex, which modifies specific arginines to dimethylarginines .
However, of 10 genes selected for differential expression analysis by qPCR, the expression of only Pip4k2a and Ddias was correspondingly increased. It should be noted that, while at the genome-wide level there is usually a strong correlation between gene expression and H3K4me3 levels, the same does not necessarily hold true for any particular gene, since the regulation of gene expression usually depends on numerous factors. Given the small amplitude of observed H3K4me3 changes, we may suppose that they are unlikely to lead to phenotypically pronounced changes in expression of corresponding genes, and thus their contribution to behavioral changes is highly questionable. On the contrary, given the high functional divergence of cortical neurons, we may assume that the relevant changes in a particular set of cells are smoothened by the rest cell mass.
The most likely explanation is that all the prominent changes in H3K4me3 landscape induced by early life stress were transient and restored to the normal levels by this age. Supporting findings were obtained on a mouse model of pronounced cognitive deficits in adult offspring with a history of maternal immune activation during prenatal development mediated by a single dose of polyriboinosinic–polyribocytidilic acid (poly IC) . In spite of the lasting working memory deficits, adult poly IC-exposed mice did not show any significant changes in H3K4me3 epigenetic landscape in the mature cerebral cortex. However the authors demonstrated that H3K4me3 profile in a number of genes was sensitive to acute IL-6 activation in primary cultures from fetal forebrain. Similarly, maternal separation in early life was shown to have no significant effect on the pattern of cortical H3K4me3 in adolescent offspring (PND30) just 2 weeks after the stress exposure , although its effects on the behavior and expression of genes being observed in adult offspring .
Summing up, the distribution of H3K4me3 in PFC, even if disturbed by maternal separation during the early life period, seems to stabilize up to a nearly normal state in adult mice, showing relatively low variability across all individuals. It is possible that this modification does not directly implicate in the observed long-lasting behavioral alterations induced by early life stress, or other brain structures are responsible for these effects. Though the possibility that the differences in H3K4me3 can only reappear under the secondary acute stress exposure is not ruled out.
We thank Pavlov K.S. (Laboratory of Experimental Models of Neurodegenerative Processes, Institute of Physiology and Basic Medicine, Novosibirsk, Russia) for his technical assistance with EthoVision XT, v.10.0.
This work was supported by the Russian Science Foundation [grant # 16–15-10131]. The publication costs for this article were funded by the Russian Science Foundation [grant # 16–15-10131].
Availability of data and materials
The sequence data were submitted to NCBI BioProject under accession number PRJNA422943.
About this supplement
This article has been published as part of BMC Genomics Volume 19 Supplement 3, 2018: Selected articles from Belyaev Conference 2017: genomics. The full contents of the supplement are available online at https://bmcgenomics.biomedcentral.com/articles/supplements/volume-19-supplement-3.
NIE and NPB contributed equally as first authors. NPB designed experiments and analyzed behavioral data, NPB and AAL conducted ChIP and performed DNA libraries, NIE performed the computational analyses, VVR made RNA extractions and primer design, JAR performed real time PCR, NIE, NPB and TIM wrote the manuscript, TIM supervised the study. All authors read and approved the final manuscript.
Ethics approval and consent to participate
All procedures were approved by the Ethics Committee of the Institute of Cytology and Genetics SB RAS (Protocol #25, December 2014) in conformity with EU Directive 2010/63/EU for animal experiments.
Consent for publication
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
- 6.Danese A, Moffitt TE, Harrington H, Milne BJ, Polanczyk G, Pariante CM, Poulton R, Caspi A. Adverse childhood experiences and adult risk factors for age-related disease: depression, inflammation, and clustering of metabolic risk markers. Arch Pediatr Adolesc Med. 2009;163(12):1135–43.CrossRefPubMedPubMedCentralGoogle Scholar
- 10.Tractenberg SG, Levandowski ML, de Azeredo LA, Orso R, Roithmann LG, Hoffmann ES, Brenhouse H, Grassi-Oliveira R. An overview of maternal separation effects on behavioural outcomes in mice: evidence from a four-stage methodological systematic review. Neurosci Biobehav Rev. 2016;68:489–503.CrossRefPubMedGoogle Scholar
- 26.Chen J, Evans AN, Liu Y, Honda M, Saavedra JM, Aguilera G. Maternal deprivation in rats is associated with corticotrophin-releasing hormone (CRH) promoter hypomethylation and enhances CRH transcriptional responses to stress in adulthood. J Neuroendocrinol. 2012;24(7):1055–64.CrossRefPubMedPubMedCentralGoogle Scholar
- 35.Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, Paulovich A, Pomeroy SL, Golub TR, Lander ES, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005;102(43):15545–50.CrossRefPubMedPubMedCentralGoogle Scholar
- 48.Daniels WM, Fairbairn LR, van Tilburg G, McEvoy CR, Zigmond MJ, Russell VA, Stein DJ. Maternal separation alters nerve growth factor and corticosterone levels but not the DNA methylation status of the exon 1(7) glucocorticoid receptor promoter region. Metab Brain Dis. 2009;24(4):615–27.CrossRefPubMedPubMedCentralGoogle Scholar
- 52.Bordner KA, George ED, Carlyle BC, Duque A, Kitchen RR, Lam TT, Colangelo CM, Stone KL, Abbott TB, Mane SM, et al. Functional genomic and proteomic analysis reveals disruption of myelin-related genes and translation in a mouse model of early life neglect. Front Psychiatry. 2011;2:18.CrossRefPubMedPubMedCentralGoogle Scholar
- 61.Walker AG, Wenthur CJ, Xiang Z, Rook JM, Emmitte KA, Niswender CM, Lindsley CW, Conn PJ. Metabotropic glutamate receptor 3 activation is required for long-term depression in medial prefrontal cortex and fear extinction. Proc Natl Acad Sci U S A. 2015;112(4):1196–201.CrossRefPubMedPubMedCentralGoogle Scholar
- 63.Kandaswamy R, McQuillin A, Sharp SI, Fiorentino A, Anjorin A, Blizard RA, Curtis D, Gurling HM. Genetic association, mutation screening, and functional analysis of a Kozak sequence variant in the metabotropic glutamate receptor 3 gene in bipolar disorder. JAMA Psychiatry. 2013;70(6):591–8.CrossRefPubMedGoogle Scholar
- 83.Carlyle BC, Duque A, Kitchen RR, Bordner KA, Coman D, Doolittle E, Papademetris X, Hyder F, Taylor JR, Simen AA. Maternal separation with early weaning: a rodent model providing novel insights into neglect associated developmental deficits. Dev Psychopathol. 2012;24(4):1401–16.CrossRefPubMedPubMedCentralGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.