Advertisement

DNA Methylation Signature of Childhood Chronic Physical Aggression in T Cells of Both Men and Women

  • Claire Guillemin,

    Affiliations: Department of Pharmacology & Therapeutics, McGill University, Montreal, Quebec, Canada, Research Unit on Children’s Psycho-Social Maladjustment and Ste-Justine Hospital Research Center, University of Montreal, Montreal, Quebec, Canada, Sackler Program for Epigenetics and Psychobiology, McGill University, Montreal, Quebec, Canada

  • Nadine Provençal,

    Affiliations: Department of Pharmacology & Therapeutics, McGill University, Montreal, Quebec, Canada, Research Unit on Children’s Psycho-Social Maladjustment and Ste-Justine Hospital Research Center, University of Montreal, Montreal, Quebec, Canada, Sackler Program for Epigenetics and Psychobiology, McGill University, Montreal, Quebec, Canada

  • Matthew Suderman,

    Affiliations: Department of Pharmacology & Therapeutics, McGill University, Montreal, Quebec, Canada, Sackler Program for Epigenetics and Psychobiology, McGill University, Montreal, Quebec, Canada, McGill Centre for Bioinformatics, McGill University, Montreal, Quebec, Canada

  • Sylvana M. Côté,

    Affiliations: Research Unit on Children’s Psycho-Social Maladjustment and Ste-Justine Hospital Research Center, University of Montreal, Montreal, Quebec, Canada, INSERM U669, Paris, France, School of Social and Preventive Medicine, University of Montreal, Montreal, Quebec, Canada

  • Frank Vitaro,

    Affiliations: Research Unit on Children’s Psycho-Social Maladjustment and Ste-Justine Hospital Research Center, University of Montreal, Montreal, Quebec, Canada, School of Psycho-Education, University of Montreal, Montreal, Quebec, Canada

  • Michael Hallett,

    Affiliation: McGill Centre for Bioinformatics, McGill University, Montreal, Quebec, Canada

  • Richard E. Tremblay ,

    richard.tremblay@ucd.ie (RT); moshe.szyf@mcgill.ca (MS)

    Affiliations: Research Unit on Children’s Psycho-Social Maladjustment and Ste-Justine Hospital Research Center, University of Montreal, Montreal, Quebec, Canada, Department of Psychology and Pediatrics, University of Montreal, Montreal, Quebec, Canada, School of Public Health and Population Sciences, University College Dublin, Dublin, Ireland, INSERM U669, Paris, France

    Equally corresponding authors.

  • Moshe Szyf

    richard.tremblay@ucd.ie (RT); moshe.szyf@mcgill.ca (MS)

    Affiliations: Department of Pharmacology & Therapeutics, McGill University, Montreal, Quebec, Canada, Sackler Program for Epigenetics and Psychobiology, McGill University, Montreal, Quebec, Canada

    Equally corresponding authors.

DNA Methylation Signature of Childhood Chronic Physical Aggression in T Cells of Both Men and Women

  • Claire Guillemin, 
  • Nadine Provençal, 
  • Matthew Suderman, 
  • Sylvana M. Côté, 
  • Frank Vitaro, 
  • Michael Hallett, 
  • Richard E. Tremblay, 
  • Moshe Szyf
PLOS
x
  • Published: January 24, 2014
  • DOI: 10.1371/journal.pone.0086822

Abstract

Background

High frequency of physical aggression is the central feature of severe conduct disorder and is associated with a wide range of social, mental and physical health problems. We have previously tested the hypothesis that differential DNA methylation signatures in peripheral T cells are associated with a chronic aggression trajectory in males. Despite the fact that sex differences appear to play a pivotal role in determining the development, magnitude and frequency of aggression, most of previous studies focused on males, so little is known about female chronic physical aggression. We therefore tested here whether or not there is a signature of physical aggression in female DNA methylation and, if there is, how it relates to the signature observed in males.

Methodology/Principal Findings

Methylation profiles were created using the method of methylated DNA immunoprecipitation (MeDIP) followed by microarray hybridization and statistical and bioinformatic analyses on T cell DNA obtained from adult women who were found to be on a chronic physical aggression trajectory (CPA) between 6 and 12 years of age compared to women who followed a normal physical aggression trajectory. We confirmed the existence of a well-defined, genome-wide signature of DNA methylation associated with chronic physical aggression in the peripheral T cells of adult females that includes many of the genes similarly associated with physical aggression in the same cell types of adult males.

Conclusions

This study in a small number of women presents preliminary evidence for a genome-wide variation in promoter DNA methylation that associates with CPA in women that warrant larger studies for further verification. A significant proportion of these associations were previously observed in men with CPA supporting the hypothesis that the epigenetic signature of early life aggression in females is composed of a component specific to females and another common to both males and females.

Introduction

The development of physical aggression has been examined within large population-based longitudinal studies from birth to adulthood. Results show that acts of physical aggression begin by the end of the first year after birth for both boys and girls, increase in frequency from 2 to 4 years of age [1][4], and then decrease in frequency from school entry to adulthood [5]. However, a minority of children (3–7%) maintain a high frequency of physical aggression from childhood to adolescence [4][6]. Although both boys and girls use physical aggression from early childhood, fewer girls manifest physically aggressive behaviors on a frequent basis and girls also tend to reduce their use of physical aggression earlier in life than boys [3], [5], [7][9]. These sex differences tend to remain stable throughout childhood and adolescence [9]. Women with atypical high levels of childhood aggression (chronic physical aggression, CPA) tend to fail in school, suffer from depression, are likely to mate with men with similar behaviour problem, become pregnant during adolescence, smoke during pregnancy, and use coercive behavior towards their children [10], [11].

Genetic epidemiological studies suggest that the frequency of childhood physical aggression is in part inherited [12][16]. Genetic association studies have also found several polymorphisms in critical genes involved in neurotransmission and hormonal regulation to associate with aggression in humans and in animals [17]. Moreover, genetics and environmental factors have been shown to interact in the expression of impulsive aggression in monkeys [12], [18] and violence in humans [19].

Very little work has been done to identify the mechanisms that might be responsible for these gene-environment associations with physical aggression. We hypothesized that DNA methylation is one such mechanism [4], [20][22]. It is now well-established that DNA sequence is complemented by epigenetic information including DNA methylation and histone modifications to program gene expression [23]. Evidence is emerging that in addition to its role in regulating gene expression during differentiation, the DNA methylation pattern is responsive to external environmental exposures including the social environment [22] in animals [24][32] and in humans [33][37].

Importantly, DNA methylation alterations associated with social exposures are not restricted to the brain but can also be detected in white blood cells (WBC) DNA [32], [33], [35], [36], [38][46]. We have recently shown that differential DNA methylation of the serotonin transporter gene promoter (SLC6A4) in T cells and monocytes is associated with in vivo measures of human brain serotonin synthesis and childhood physical aggression in men [42]. Moreover, we have shown that young adult males on a chronic physical aggression trajectory between age 6 and 15 years had differential DNA methylated regions located in the genomic loci of cytokines and related transcription factors in T cells and monocytes, compared to males with the same background who did not follow such a high physical aggression trajectory (control group) [47]. In addition, using whole genome mapping method to analyse the DNA methylation profile of men on a CPA trajectory, we have recently found that men with CPA had a T cell DNA methylation signature distinct from men from the control group (Provençal et al., under review in PLoS One).

To test the hypothesis that T cell DNA methylation profiles are associated with childhood physical aggression in females and, if so, to compare them with the association found in male DNA, we analyzed genome-wide promoter methylation profiles in blood T cells of adult female participants. We recruited women who have been on a chronic physical aggression (CPA) trajectory between 6 and 12 years of age and compared them to women with the same background who followed normal physical aggression (NPA) trajectories [5]. Methylation profiles were created using the method of methylated DNA immunoprecipitation (MeDIP) followed by microarray hybridization, and analyses of these profiles revealed regions of significant DNA methylation patterns associated with CPA in women. We then compared these results with those previously observed in men from the same cohort (Provençal et al., under review in PLoS One).

Materials and Methods

Ethics Statement

After complete description of the study to the subject, all participants provided written informed consent. The study was carried out in accordance with the Declaration of Helsinki, and was approved by the research ethics committee of the University of Montreal pediatric hospital (St-Justine Hospital).

Participants

The participants of this study are adult women of European descent, selected from a large sample of boys and girls who were first assessed when they were attending kindergarten in the province of Quebec’s French-speaking public schools in 1986–1987 [5], the Quebec Longitudinal Study of Kindergarten Children (QLSKC). The frequency of physical aggression was assessed for each participant by his school teachers between kindergarten (6 years old) and high school (12 years old). Developmental trajectories of physical aggression were previously identified for the entire sample of the longitudinal study [5]. Two groups of females were drawn from this cohort: 1) The Chronic Physical Aggression group is composed of women who were on a consistently high physical aggression trajectory from age 6 to 12 years (CPA); 2) The control group included women who did not have a history of chronic physical aggression from age 6 to 12 years (NPA). All participants were born in families with a low socioeconomic status and were living within 200 km from our laboratory at the time of recruitment in order to isolate blood cells and extract DNA from fresh blood draw. A total of 5 eligible women (out of 19) belonging to the CPA trajectory agreed to participate and were included in the study. To avoid having an imbalanced control group we included only 16 women in the control group. Due to technical reasons (insufficient DNA extracted from the sample) we had to drop two individuals from the control group. In addition to physical aggression, other behavioral problems, such as hyperactivity, were also rated from age 6 to 15 and violence at 21 years of age (See Methods S1 for details). Characteristics of the 2 groups are presented in Table 1.

thumbnail
Table 1. Characteristics of the chronic physical aggression (CPA) group and normal physical aggression (NPA) group.

doi:10.1371/journal.pone.0086822.t001

(Additional information on the characteristics of the groups can be found in Methods S1). For further validation by Illumina 450K we also included the adult males from the same longitudinal studies (8 CPA and 12 controls) that were previously used in the MeDIP-microarrays analysis (Provençal et al., under review in PLoS One).

DNA Methylation Analysis by MeDIP and Micoarrays Hybridization

Genomic DNA was extracted from T cells purified from blood draw by immunomagnetic positive selection using CD3 antibody as previously described [32], [42], [47]. Methylated DNA immunoprecipitation was performed as previously described [31], [32], [38] using anti-5methylCytosine antibody from Calbiochem. The specificity of the immunoprecipitation was assessed by regular PCR on two endogenous control genes, H19 (methylated control) and β-actin (unmethylated control) and two plasmids that were added to the DNA prior to sonication, GFP (unmethylated control) and Luciferase (in vitro methylated control) using the following primers: H19: forward (5′-GAGCCGCACCAGATCTTCAG -3′), reverse (5′-TTGGTGGAACACACTGTGATCA-3′); β-actin: forward (5′-CCAACGCCAAAACTCTCCC-3′), reverse (5′-AGCCATAAAAGGCAACTTTCG-3′); GFP: forward (CAAGGGCGAGGAGCTGTT), reverse (CGGCCATGATATAGACGTTG); Luciferase: forward (5′-AGAGATACGCCCTGGTTCC-3′), reverse (5′-CCAACACCGGCATAAAGAA-3′). The input and bound fractions were then amplified using Whole Genome Amplification kit (Sigma) and labeled for microarray hybridization with either Cy3-dUTP or Cy5-dUTP (Perkin Elmer) respectively using the CGH labeling kit (Invitrogen) following the manufacturer’s instructions. The labeled input and bound DNA samples were hybridized to a custom designed 180K promoter tiling array (Agilent technologies) that contained probes covering all transcription start sites at intervals from 1000 bp upstream to 200 bp downstream of all genes described in Ensembl (version 54) and within 250 bp of the ~400 microRNAs from miRBase, all at 100 bp-spacing. The array covered 69,818 transcription start sites corresponding to 20,318 genes. All the steps of hybridization, washing, and scanning were performed following the Agilent protocol for ChIP-on-chip analysis. After microarray scanning, probe intensities were extracted from scan images using Agilent’s Feature Extraction 9.5.3 Image Analysis Software.

MeDIP Microarray Analysis

Two replicate microarrays were hybridized and analysed for each of the 19 samples. Quality control involved visual inspection of microarray scan images, scanning software reports and generation of MvA plots (i.e. plots of log(Cy5/Cy3) vs log(Cy5 x Cy3)) to identify microarrays with dye biases or low signal. Microarrays were normalized using background subtraction followed by quantile normalization of log2-ratios of bound (Cy5) and input (Cy3) channel intensities. Differential methylation between groups of samples was determined at the probe and promoter levels to ensure both statistical significance and biological relevance as previously described [38]. At the probe level, a modified t-statistic was computed for each probe to identify probe intensity differences between CPA and NPA groups using the ‘limma’ package [48] of Bioconductor [49]. Then, promoter-level methylation differences were identified as those promoters significantly enriched with probes having positive or negative t-statistics using the Wilcoxon rank-sum test. A probe and the containing promoter were called differentially methylated if the p-value of the probe t-statistic was at most 0.05 (uncorrected for multiple testing), log2-fold difference between the groups was at least 0.25, and the false discovery rates (FDR) of the promoter-level statistic was at most 0.05. The microarray data have been deposited in Gene Expression Omnibus (GEO) at NCBI (www.ncbi.nlm.nih.gov/geo/) under the accession number GSE53193.

A Bayesian deconvolution method was used to estimate methylation levels from the microarray data [51] and, as expected, estimated promoter methylation levels were found to be significantly anti-correlated (p = 2.3e-25) with previously published gene expression levels of the same human cell type (http://biogps.org/) (Figure S1).

All functional analysis was done using Ingenuity Pathway Analysis [50] with the default parameters.

MeDIP Microarray Validation

Two approaches were used to validate the regions called differentially methylated between CPA and NPA subjects from the MeDIP-microarrays analysis; a gene-specific approach using quantitative real-time PCR on the MeDIP fractions (Q-MeDIP) and a genome-wide approach using Illumina Infinium HumanMethylation450 BeadChip® (Illumina).

Gene-specific real-time PCR validation of microarray was performed on the amplified bound fraction for the same subjects used for microarray experiments (n = 5 CPA and n = 14 NPA). Real-time PCR experiments were performed using a LightCycler and SYBR green reagent (Roche Diagnostics). Primers and detailed conditions are available upon request. The Ct (cycle thresholds) for all bound samples was converted to a relative quantity scale using a standard curve of diluted fragmented human DNA. We normalized the results by calculating the ratio of the amplified specific gene over the amplified methylated control. The methylated control used for normalization was the in vitro methylated luciferase plasmid that was added in equal amount to each sample prior to the MeDIP. The Q-MeDIP analyses of the methylated control show no difference between the groups.

Whole-genomic validation of MeDIP microarray data was performed using Illumina 450K arrays. This technique was applied to validate both the male (Provençal et al. under review in PLoS One) and female samples. Pooled samples were used since only limited amount of DNA was available for many of the subjects. Three different pools of genomic DNA were made per group: men CPA (2 or 3 different subjects per pool), men NPA (4 different subjects per pool), women CPA (2 different subjects per pool) and women NPA (4 or 5 different subjects per pool). Genomic DNA was quantified using Picogreen protocol (Quant-iT™ PicoGreen® dsDNA Products, Invitrogen) and read on SpectraMAX GeminiXS Spectrophotometer. Bisulfite conversion was performed with 500 ng of genomic DNA using the EZ-96 DNA Methylation-Gold Kit (Zymo Research) according to the manufacturer’s instructions. The Illumina Methylation 450K kit was used for the microarray experiment, as described by the manufacturer’s protocol, except that 8 µl of bisulfite converted material was used to initiate the amplification step. Arrays were scanned using the Illumina iScan Reader. Data analysis was performed using the Methylation module (version 1.9.0) of the GenomeStudio software (Illumina; version 2011.1). Significant differences in methylation levels between CPA and control groups were obtained using the Illumina custom method with a False Discovery Rate (FDR) <0.05. Beta-values with detection P-value >0.001 were removed from the analysis since they are considered to fall below the minimum intensity and threshold. All the probes that contained a SNP within 10 bp of the CpG site were excluded from the analysis to ensure equal annealing efficiency between the samples. The raw data have been deposited in Gene Expression Omnibus (GEO) at NCBI (www.ncbi.nlm.nih.gov/geo/) under the accession number GSE53193.

Validation of Illumina 450K Microarrays

Primers were designed to generate amplicons <350 bp long corresponding to the sequence containing differentially methylated CGs identified by the Illumina 450K analysis. Bisulfite conversion of individual male and female samples was performed using 500 ng of genomic DNA using the EZ-96 DNA Methylation-Gold Kit (Zymo Research) according to the manufacturer’s instructions. PCR amplifications were performed with 15 ng of bisulfite DNA using TaKaRa EpiTaq HS polymerase (Cedarlane) on individual samples using the following primers: ZNF366: forward (5′-GAAGGTATTTATTTGAGAAAAAGAG-3′), reverse (5′-CCAACTCCTAAAATCTAAATAACTACAAC-3′). To calibrate the assay and control for amplification bias, we amplified the same regions from completely unmethylated (0%) or fully methylated (100%) or mixtures of bisulfite-converted EpiTect Control DNA (Qiagen). For pyrosequencing analysis, 20 µl of the bisulfite-PCR products were processed according to the manufacturer’s standard protocol (Qiagen). Sequencing reactions were performed on a PyroMark Q96 system using the PyroMark Gold Q96 Reagent Kit (Qiagen) according to the manufacturer’s instructions. The percentage methylation at each CpG site was calculated from the raw data using PyroMark Q96-CpG Software (Qiagen).

Results

T Cell Promoter Methylation Associated with CPA in Women

Genome-wide promoter methylation profiles of T cell DNA from women on a CPA trajectory during childhood and adolescence were compared to women on normative physical aggression trajectories. We found 917 probes corresponding to 430 distinct gene promoters differentially methylated between aggression groups (P<0.05 and FDR <0.05, Spreadsheet S1). Of these promoters, 353 were less methylated and 77 were more methylated in the CPA group (Figure 1A). A heatmap of the probes in each of these gene promoters that best differentiate between the groups is shown in Figure 1B. The 25 gene promoters most significantly different between the two groups are listed in Table S1 (P<0.005 and FDR <0.01). Interestingly, the list of differentially methylated genes includes a few that have been previously associated with aggressive behavior in human and animals [17] (Table 2). For example, Tryptophan Hydroxylase 2 (TPH2), Corticotrophin Releasing Hormone binding protein (CRHBP) and the glucocorticoid receptor (NR3C1) were all found to be less methylated in the CPA group.

thumbnail
Figure 1. Gene promoters differentially methylated between women CPA (n = 5) and NPA (n = 14) in T cells.

A. Numbers of promoters with significant methylation increases and decreases in CPA versus NPA (P<0.05; FDR<0.05). B. Heatmap depicts normalized intensities of microarray probes contained in promoters that best differentiate between CPA and NPA groups. Probes were selected from gene promoters called differentially methylated with respect to aggression groups (P≤0.05; FDR ≤0.05). One probe was selected for each gene, always the probe with the most extreme t-statistic. Heatmaps are colored so the median values on each row are gray, high values are red and low values are green. Red indicates higher methylation in a row and green indicates lower methylation. Clustering was performed using Ward’s hierarchical clustering algorithm with Pearson correlation distance as the distance metric. Rows correspond to promoters and columns to subjects.

doi:10.1371/journal.pone.0086822.g001

thumbnail
Table 2. Differentially methylated gene promoters between women chronic physical aggression (CPA) and non-aggressive (NPA) groups previously shown to be associated with aggressive behavior.

doi:10.1371/journal.pone.0086822.t002

We selected 19 of these regions called differentially methylated for validation by Q-MeDIP (Figure S2A) as described in the Methods section. Of these, 10 were found to be differentially methylated by Q-MeDIP and, overall, the fold-change differences identified by Q-MeDIP were highly correlated with fold-change differences identified by microarray (R = 0.849, P<0.001 by Pearson’s correlation, Figure S2B).

Comparison between Differentially Methylated Promoters Associated with Physical Aggression in Men and Women

A previous study, conducted in our laboratory, analyzing genome-wide promoter methylation profiles in the T cells of adult males from the same longitudinal cohort as the one studied here (Provençal et al., under review in PLoS One), found 448 gene promoters to be differentially methylated in men with CPA. Of these, 31 gene promoters belong to the list of gene promoters found differentially methylated here in females (Table 3), a significant overlap (P<0.0001, Fisher’s exact test). Within the promoters of these genes, there are 32 common differentially methylated probes, indicating that a significant portion of the overlap is actually due to identical genomic sites being differentially methylated. One example is the ZNF366 gene promoter region where the MeDIP microarray analysis identified a region which methylation pattern is associated with aggression in both women and men (Figure 2).

thumbnail
Figure 2. Methylation pattern associated with aggression in women and in men of the ZNF366 gene promoter region analyzed by MeDIP microarrays and validated using Illumina 450K arrays and pyrosequencing.

Expanded views from the UCSC genome browser of ZNF366 gene promoter region located on chromosomes 5 are depicted. The first and second tracks show the average MeDIP microarray probe log2-fold differences (scale: −0.6 to 0.6 for both tracks) between chronic physical aggressive (CPA) and non-aggressive (NPA) groups for men and women T cells. In black are probes that are more methylated and in gray are those that are less methylated in the CPA group. Highlighted in blue and in pink are regions significantly differentially methylated between the groups in men and in women, respectively. The third and fourth track shows delta Beta values (average Beta CPA- average Beta NPA) obtained for each CpG site analyzed by Illumina 450K arrays in men and in women (scale: −0.2 to 0.2 for both tracks). The last track indicates the location of individual CpG sites using black lines. In this region, three CpG sites were analyzed by pyrosequencing (underlined). The bar graph shows the average methylation levels for each CpG sites in the women CPA and NPA groups as determined by pyrosequencing. Error bars show the SEM and * indicate P<0.05 and # P<0.1.

doi:10.1371/journal.pone.0086822.g002

thumbnail
Table 3. Differentially methylated gene promoters associated with childhood physical aggression in common between men and women using MeDIP-microarrays.

doi:10.1371/journal.pone.0086822.t003

Functional Relevance of the Differentially Methylated Promoters Associated with CPA

To delineate if the list of genes differentially methylated between CPA and NPA are grouped into biological functions and pathways we used Ingenuity Pathway Analysis (IPA) software. In women, IPA identified 74 functional categories significantly enriched with the affected genes including “Cell signaling and interaction”, “Inflammatory response” and “Behavior” (Table S2) as well as specific canonical signalling pathways including CCR5, TEC kinases, G-protein Gαi and IL-10 (Table S3).

Aggression might also affect the methylation status of a specific group of genes by targeting their upstream regulators. Using IPA we were able to identify a set of potential upstream regulators including transcription factors, growth factors and transmembrane receptors. These potential upstream regulators were identified to target a significant number of genes associated with aggression in women (Table S4). The list includes many molecules that play an important role in the regulation of genes involved in immune and inflammatory responses, such as transcription factors FOS, GATA1 and GATA3, hepatocyte growth factor (HGF) and many cytokines (TNF, IFNG, IL1B, IL1A, IL17R, IL10, IL13, IL18).

Almost all of the functional categories significantly enriched with genes associated with physical aggression in women were previously observed in men (Provençal et al., under review in PLoS One) (70 out of 74) with “Behavior” being one of the top 5 categories (Table 4 and Table S5; P<0.0001 Fisher’s exact test), as well as twelve out of the thirty enriched canonical pathways (Table 4 and Table S6; P<0.0001 Fisher’s exact test) and common potential upstream regulators such as TNF and IL-1β involved in inflammatory response (Table 4 and Table S7).

thumbnail
Table 4. Functions and pathways enriched with genes whose methylation is associated with aggression in both sexes from IPA analysis (women n = 430 genes and men n = 448 genes).

doi:10.1371/journal.pone.0086822.t004

Genomic Organization of Differentially Methylated Promoters Associated with CPA

In addition to functional organization, the differentially methylated promoters exhibit genomic organization. As previously observed in men (Provençal et al., under review in PLoS One), regions with higher methylation in CPA women contained higher than expected CpG densities and regions with lower methylation in CPA women contained lower than expected CpG densities (Figure 3A). Although CPA associated methylation differences are distributed throughout the genome in both men and women, the differences typically appeared in clusters such that methylation differences at any given genomic location were weakly but significantly predictive of methylation differences up to 1.5Mb away in women (Figure 3B) and 2Mb away in men (Provençal et al., under review in PLoS One).

thumbnail
Figure 3. Genomic organization of differentially methylated promoters associated with CPA.

A. CpG density of the differentially methylated promoters. Normalized CpG density is the density of CpG sites divided by the expected density calculated by multiplying the density of C sites by the density of G sites. *** indicate P<0.0001. B. The Pearson correlations of DNA methylation differences between CPA and NPA groups at various genomic distances are shown. Error bars denote 95% confidence intervals obtained from 1000 bootstraps composed of randomly selected probe pairs with replacement. The grey highlight shows the expected 95% confidence interval for correlations of probe pairs independent of their distance. Independence was simulated by with 500 random permutations of the probe coordinates. This confidence interval does not overlap with the error bars associated with distances less than 1.5Mb suggesting the existence of systematic dependencies between methylation differences at distances up to 1.5Mb.

doi:10.1371/journal.pone.0086822.g003

Validation of the MeDIP-microarray Data using Illumina Human Methylation 450K Arrays

To further confirm the results obtained using MeDIP-microarrays, we generated and analyzed new methylation profiles using the Infinium Human Methylation 450 BeadChip®. The Illumina 450K approach is an alternative to the MeDIP approach appropriate for validation because it is based on bisulfite conversion rather than antibody affinity. Although it has lower coverage than the MeDIP microarray, it has higher resolution (single-CpG resolution) and reasonable coverage with at least one CpG site in 96% of known CpG islands and 99% of RefSeq gene promoters. Due to small amounts of remaining DNA, samples were pooled to produce three sample pools per aggression group (both male and female), each pool hybridized to an Illumina 450K array and then the arrays analyzed to identify methylation differences between CPA and NPA sample pools. A total of 716 gene promoters were found to contain CpG sites differentially methylated between female CPA and NPA pools (P≤0.05 and FDR <0.05, Spreadsheet S2). Of these, 70 gene promoters were also found differentially methylated using the MeDIP-microarrays (Table S8; P<0.001, Fisher’s exact test), and 79 contain a differentially methylated CpG site in the male Illumina 450K microarrays (P<0.0001, Fisher’s exact test; Table 5). These 79 gene promoters differentially methylated in both men and women according to the Illumina platform include 7 gene promoters differentially methylated in women and 3 gene promoters differentially methylated in men according to our MeDIP-microarray data. One example is the ZNF366 gene promoter region where both the MeDIP microarray and Illumina 450K analysis identified a region which methylation pattern is associated with aggression in women (Figure 2).

thumbnail
Table 5. Differentially methylated gene promoters associated with childhood physical aggression in common between men and women using Illumina 450K arrays.

doi:10.1371/journal.pone.0086822.t005

The three CpG sites of this region were analyzed by pyrosequencing. On average, all of the CpG sites have lower methylation in the women CPA group with CG1 showing a significant difference at P<0.05 in all the techniques used: MeDIP microarray, Illumina 450K array and pyrosequencing (Figure 2).

The Illumina 450K profiles also include methylation levels for many CpG sites outside of promoters of which 5898 CpG sites are differentially methylated in women with CPA and 5795 are differentially methylated in men with CPA (P<0.05, no correction for multiple testing, Spreadsheet S3). These two sets overlap on a significant 744 CpG sites (P<0.0001; Fisher’s exact test).

Discussion

We have previously hypothesized that epigenetic mechanisms mediate aggressive behaviors by showing a signature of aggression in the T cells of adult, human males. In this study, we tested the hypothesis that a similar DNA methylation signature exists in women as well. Our study confirms this hypothesis and furthermore identifies a significant but not perfect overlap between the male and female signatures. Indeed, the methylation of 31 gene promoters was shown to associate with physical aggression in both sexes (Table 3). Interestingly, a significant portion of the overlap is due to identical genomic sites being differentially methylated in a gender-independent fashion. Adding to this list, 744 CpG sites were found differentially methylated in men and women with CPA using Illumina 450K arrays. The fact that such an overlap is observed between men and women data in spite of the fact that the methylome analyses were performed separately and independently makes it highly unlikely that the methylome signatures identified here could be derived randomly. The almost perfect overlap between functional categories represented by both the male and female signatures provides further evidence that these signatures are associated with female and male chronic physical aggression (Table 4). The existence of sex-specific and sex-independent components of the signatures is consistent with what we know about sex differences and similarities in human physical aggression, and future studies will examine the relationships of these components with gender-specific and independent aggressive behaviors.

MeDIP microarray and Illumina 450K array methods captured different population of the differentially methylated regions although both methods validated the difference between the groups. These methods measure DNA methylation differently and have different representation of CpG sites in their array. The main difference between the methods is that the MeDIP microarray method measures average differences across 250 bp and their proximal regions while Illumina 450K arrays measure CpG sites independently. One example of the different results obtained from the two techniques is illustrated in Figure 2. ZNF366 gene promoter was found differentially methylated by both techniques in both sexes but at different CpG sites. Indeed, this region contain CpG sites called differentially methylated by both techniques (Figure 2 on the left highlighted in pink) as well as regions called differentially methylated by MeDIP microarray in both sexes (Figure 2, on the right highlighted in blue and in pink) but where no CpG sites were analyzed by Illumina 450K arrays. It is therefore not surprising that these methods reveal different sensitivities to differences in DNA methylation in different promoters and therefore capture different although somewhat overlapping domains of the differentially methylated landscape associated to chronic aggression.

Three genes of the 430-gene female signature were already known to associate with aggression in animals and in humans (Table 2). Indeed, recent studies that aimed to decipher the biological mechanisms involved in the development and maintenance of chronic aggression have consistently found a key role for the serotonin (5-hydroxytryptamine, 5-HT) pathway, suggesting lower serotoninergic activity associated with aggression/impulsivity [51][54]. Tryptophan Hydroxylase is the rate-limiting enzyme in the synthesis of serotonin. In humans, Perez-Rodriguez et al. [55] tested several polymorphisms of TPH2 and found an extended haplotype associated with enhanced aggressiveness, suicidal behavior, and susceptibility to borderline personality disorder. We found that the promoter of TPH2 is less methylated in CPA women. We also found CRHBP, regulator of cortisol levels, and the glucocorticoid receptor (NR3C1, GR) to be less methylated in the CPA women. The glucocorticoid receptor, whose epigenetic changes, were associated with early-life stress (reduced maternal care in rats [31], [56], childhood abuse in human brain [37]), plays also a key role in HPA axis related stress reactivity and aggressive behavior in pigs [57]. The stress response and the HPA axis have also been associated with aggression [51], [58]. Both hyper- and hypo-active HPA axis are associated with aggression in humans [59][62]. It is perhaps not surprising that these genes are differentially methylated only in women where the HPA negative feedback control is known to be more sensitive than in males [63]. The same is true for rodents where it has been shown that female rodents have higher HPA axis activity under basal and stress-induced conditions than males [64].

As previously observed in men (Provençal et al., under review in PLoS One), several of the genes in the female aggression signature play roles in cytokine function and inflammatory response (Table 4 and Table S13). Recent studies have shown that cytokines are associated with various behavioral disorders such as anxiety, depression, suicide, childhood mood disorder and post-traumatic stress disorder (PTSD) [40], [65][75]. It was also suggested that cytokines might play a role in the neurobiology of aggression since they are expressed in brain regions already known to be involved in aggression and behavior [76][78]. For example, IL1RN have been shown to be involved in defensive aggression through its activation by IL-1 beta [79], [80]. Our results show that IL1RN belongs to both the male and female aggression signatures. Moreover, cytokines expression in blood and gene methylation in T cells was recently found to associate with men CPA in the same subjects as the one studied here [47]. Further work is needed to investigate the role of peripheral cytokines in aggression but taken together these data suggest that cytokine regulation could play an important role in human behavior and mental health.

Three of the genes in both the male and female signatures of aggression are members of the protocadherin gene family (PCDH8, PCDHB2, PCDHB8). Interestingly, we have recently shown that the protocadherin gene family is differentially methylated in hippocampi from rats exposed to low maternal care and humans subjected to child abuse [31], [81]. It is noteworthy that although these genes are clearly involved in brain function, we also observed changes in DNA methylation in T cells associated with aggressive behavior. Similar results were obtained when we recently compared rhesus monkey DNA methylation changes in prefrontal cortex and T cells in response to differences in maternal rearing [32]. This analysis revealed both tissue specific alterations as well as common differentially methylated regions in T cells and prefrontal cortex. Two previous reports suggested that brain function-specific genes were differentially methylated in peripheral blood cells in association with physical aggression [42], [47]. Further work is needed to understand why seemingly brain-specific genes would be differentially methylated in blood DNA in association with aggression and other environmental factors.

There are three important limitations to this study. First, although we recruited subjects from a large longitudinal study, we managed to assess only a small number of women with chronic physical aggression because they represent a very small percentage of the population and are difficult to recruit. Thus, replications are needed from similar longitudinal studies to confirm our results. Second, we analyzed DNA methylation profiles only in adulthood because T cells were not collected at earlier time points during the longitudinal study. New longitudinal studies are needed to help distinguish whether the observed DNA methylation profiles are present before the start of the chronic physical aggression trajectories or are an outcome of these behaviors. Third, there was no psychometric-physical evaluation at the time of blood draw. The acute psychological and/or physical status might confound our findings. Future longitudinal studies that include concurrent blood draws and psychometric-physical evaluations are required to address this question. In addition, childhood abuse is also known to increase the risk of aggression in adolescents and adults [82] and was also found to associate with DNA methylation differences [37]. Therefore, it is possible that child abuse acts as a third factor in explaining the reported association between aggression and DNA methylation. This however might be reflecting the simple fact that these behaviors are molecularly and functionally linked within the same biological pathways.

Nevertheless, this study is consistent with a DNA methylation signature of chronic aggression that is maintained into adulthood and provides justification for future longitudinal and intervention studies using T cell methylomes to investigate the causal and temporal relationships between social experiences and long-term behavioral phenotypes in humans.

Supporting Information

Figure S1.

Distributions of promoter methylation levels by published expression level. Genes expression levels were obtained from publicly available T cells expression profiles (Su, Wiltshire et al. 2004) that we used to partition genes by expression percentiles (0–5, 5–10, …, 95–100). To obtain gene promoter methylation levels, we computed the average normalized intensity of each probe across all samples and replicates, and then applied the Bayesian deconvolution algorithm mentioned above to the resulting averages (Down, Rakyan et al. 2008). Shown are the distributions of methylation levels for each expression percentile. The distributions show that genes with low or no expression (represented in green) tend to have highly methylated promoters, whereas genes with high expression (represented in red) tend to have lower promoter methylation.

doi:10.1371/journal.pone.0086822.s001

(TIF)

Figure S2.

Validation by Q-MeDIP of differentially methylated sequences between CPA (n = 5) and NPA (n = 14) groups. A. Fold differences between CPA and NPA groups obtain by Q-MeDIP (grey fill) and microarray (dark fill) analyses are shown for 19 genes predicted to be more methylated (n = 2) and less methylated (n = 17) in the CPA individuals by the microarray analysis. In bold are the genes in common between men and women MeDIP-arrays. * and # indicated the P value obtained by comparing the groups (# P≤0.1; * P<0.05; ** P<0.001; *** P<0.0001). B. Correlation of the fold differences between CPA and NPA obtain by Q-MeDIP and by MeDIP-array for 19 genes predicted to be differentially methylated between CPA and NPA groups by the microarray analysis.

doi:10.1371/journal.pone.0086822.s002

(TIF)

Table S1.

Top 25 gene promoters differentially methylated between women CPA (n = 5) and NPA (n = 14) groups from the MeDIP-microarray analysis (P<0.005 and FDR <0.01). In bold are genes also found to be differentially methylated in men MeDIP-microarray analysis. Underlined are genes validated by Illumina 450K arrays.

doi:10.1371/journal.pone.0086822.s003

(DOCX)

Table S2.

Biological functions enriched with genes whose methylation is associated with women aggression from IPA analysis (n = 430 genes).

doi:10.1371/journal.pone.0086822.s004

(DOCX)

Table S3.

Canonical pathways enriched with genes whose methylation is associated with women aggression from IPA analysis (n = 430 genes). All of the p values were calculated using a right tailed Fisher’s exact test and corrected for multiple comparison with the Benjamini-Hochberg method. Significance threshold were p = 0.05.

doi:10.1371/journal.pone.0086822.s005

(DOCX)

Table S4.

Upstream regulators showing a significant overlap with genes whose methylation is associated with aggression from IPA analysis (n = 430 genes). Upstream regulators differentially methylated between chronic and normal aggression are shown in bold. Significance threshold were P = 0.05.

doi:10.1371/journal.pone.0086822.s006

(DOCX)

Table S5.

Biological functions enriched with genes whose methylation is associated with aggression in both sexes from IPA analysis (women n = 430 genes and men n = 448 genes).

doi:10.1371/journal.pone.0086822.s007

(DOCX)

Table S6.

Canonical pathway enriched with genes whose methylation is associated with aggression in both sexes from IPA analysis (women n = 430 genes and men n = 448 genes).

doi:10.1371/journal.pone.0086822.s008

(DOCX)

Table S7.

Upstream regulators showing a significant overlap with genes whose methylation is associated with aggression in both sexes from IPA analysis (women n = 430 genes and men n = 448 genes).

doi:10.1371/journal.pone.0086822.s009

(DOCX)

Table S8.

Gene promoters with significant changes in methylation between women CPA and NPA groups from both MeDIP-microarrays and Illumina 450k arrays analysis.

doi:10.1371/journal.pone.0086822.s010

(DOCX)

Methods S1.

doi:10.1371/journal.pone.0086822.s011

(DOCX)

Spreadsheet S1.

Full table of probes differentially methylated between women CPA and women NPA by MeDIP-microarray.

doi:10.1371/journal.pone.0086822.s012

(XLSX)

Spreadsheet S2.

Full table of the 716 gene promoter CGs differentially methylated between women CPA and women NPA by Illumina 450K (CPA-NPA).

doi:10.1371/journal.pone.0086822.s013

(XLSX)

Spreadsheet S3.

Full table of all the CG sites differentially methylated between CPA and NPA for each sex by Illumina 450K (CPA-NPA). Sheet 1 corresponds to women CPA-NPA results. Sheet 2 is men CPA-NPA results.

doi:10.1371/journal.pone.0086822.s014

(XLSX)

Acknowledgments

The authors wish to thank all the participants as well as the staff of the University of Montreal Research Unit on Children’s Psychosocial Maladjustment (GRIP) for their valuable assistance and Charles-Édouard Giguère for helping with the data banks.

Author Contributions

Conceived and designed the experiments: CG NP M. Suderman SMC M. Szyf RET. Performed the experiments: CG NP. Analyzed the data: CG NP M. Suderman. Contributed reagents/materials/analysis tools: MH FV. Wrote the paper: CG NP M. Suderman SMC M. Szyf RET.

References

  1. 1. NICHD & Network ECCR (2004) Trajectories of physical aggression from toddlerhood to middle childhood: predictors, correlates, and outcomes. Monogr Soc Res Child Dev 69: vii, 1–129.
  2. 2. Alink LR, Mesman J, van Zeijl J, Stolk MN, Juffer F, et al. (2006) The early childhood aggression curve: development of physical aggression in 10- to 50-month-old children. Child Dev 77: 954–966. doi: 10.1111/j.1467-8624.2006.00912.x
  3. 3. Cote SM, Vaillancourt T, LeBlanc JC, Nagin DS, Tremblay RE (2006) The development of physical aggression from toddlerhood to pre-adolescence: a nation wide longitudinal study of Canadian children. J Abnorm Child Psychol 34: 71–85. doi: 10.1007/s10802-005-9001-z
  4. 4. Tremblay RE (2010) Developmental origins of disruptive behaviour problems: the ‘original sin’ hypothesis, epigenetics and their consequences for prevention. J Child Psychol Psychiatry 51: 341–367. doi: 10.1111/j.1469-7610.2010.02211.x
  5. 5. Broidy LM, Nagin DS, Tremblay RE, Bates JE, Brame B, et al. (2003) Developmental trajectories of childhood disruptive behaviors and adolescent delinquency: a six-site, cross-national study. Dev Psychol 39: 222–245. doi: 10.1037/0012-1649.39.2.222
  6. 6. Campbell R, Weis MA, Millet L, Powell V, Hull-Jilly D, et al. (2006) From surveillance to action: early gains from the National Violent Death Reporting System. Inj Prev 12 Suppl 2ii6–ii9. doi: 10.1136/ip.2006.012450
  7. 7. Cairns RB, Cairns BD, Neckerman HJ, Ferguson LL, et al. (1989) Growth and aggression: I. Childhood to early adolescence. Developmental Psychology 25: 320–330. doi: 10.1037/0012-1649.25.2.320
  8. 8. Baillargeon RH, Zoccolillo M, Keenan K, Cote S, Perusse D, et al. (2007) Gender differences in physical aggression: A prospective population-based survey of children before and after 2 years of age. Dev Psychol 43: 13–26. doi: 10.1037/0012-1649.43.1.13
  9. 9. Côté SM (2007) Sex Differences in Physical and Indirect Aggression: A Developmental Perspective. European Journal on Criminal Policy and Research 13: 183–200. doi: 10.1007/s10610-007-9046-3
  10. 10. Fontaine N, Carbonneau R, Barker ED, Vitaro F, Hebert M, et al. (2008) Girls’ hyperactivity and physical aggression during childhood and adjustment problems in early adulthood: a 15-year longitudinal study. Arch Gen Psychiatry 65: 320–328. doi: 10.1001/archgenpsychiatry.2007.41
  11. 11. Serbin LA, Cooperman JM, Peters PL, Lehoux PM, Stack DM, et al. (1998) Intergenerational transfer of psychosocial risk in women with childhood histories of aggression, withdrawal, or aggression and withdrawal. Dev Psychol 34: 1246–1262. doi: 10.1037/0012-1649.34.6.1246
  12. 12. Bennett AJ, Lesch KP, Heils A, Long JC, Lorenz JG, et al. (2002) Early experience and serotonin transporter gene variation interact to influence primate CNS function. Mol Psychiatry 7: 118–122. doi: 10.1038/sj/mp/4000949
  13. 13. Brendgen M, Boivin M, Vitaro F, Bukowski WM, Dionne G, et al. (2008) Linkages between children’s and their friends’ social and physical aggression: evidence for a gene-environment interaction? Child Dev 79: 13–29. doi: 10.1111/j.1467-8624.2007.01108.x
  14. 14. Dionne G, Tremblay R, Boivin M, Laplante D, Perusse D (2003) Physical aggression and expressive vocabulary in 19-month-old twins. Dev Psychol 39: 261–273. doi: 10.1037/0012-1649.39.2.261
  15. 15. Hicks BM, Krueger RF, Iacono WG, McGue M, Patrick CJ (2004) Family transmission and heritability of externalizing disorders: a twin-family study. Arch Gen Psychiatry 61: 922–928. doi: 10.1001/archpsyc.61.9.922
  16. 16. Rhee SH, Waldman ID (2002) Genetic and environmental influences on antisocial behavior: a meta-analysis of twin and adoption studies. Psychol Bull 128: 490–529. doi: 10.1037/0033-2909.128.3.490
  17. 17. Pavlov KA, Chistiakov DA, Chekhonin VP (2012) Genetic determinants of aggression and impulsivity in humans. J Appl Genet 53: 61–82. doi: 10.1007/s13353-011-0069-6
  18. 18. Suomi SJ (2006) Risk, resilience, and gene x environment interactions in rhesus monkeys. Ann N Y Acad Sci 1094: 52–62. doi: 10.1196/annals.1376.006
  19. 19. Caspi A, McClay J, Moffitt TE, Mill J, Martin J, et al. (2002) Role of genotype in the cycle of violence in maltreated children. Science 297: 851–854. doi: 10.1126/science.1072290
  20. 20. Bollati V, Baccarelli A (2010) Environmental epigenetics. Heredity (Edinb) 105: 105–112. doi: 10.1038/hdy.2010.2
  21. 21. Mill J, Petronis A (2008) Pre- and peri-natal environmental risks for attention-deficit hyperactivity disorder (ADHD): the potential role of epigenetic processes in mediating susceptibility. J Child Psychol Psychiatry 49: 1020–1030. doi: 10.1111/j.1469-7610.2008.01909.x
  22. 22. Szyf M (2011) The early life social environment and DNA methylation: DNA methylation mediating the long-term impact of social environments early in life. Epigenetics 6: 971–978. doi: 10.4161/epi.6.8.16793
  23. 23. Razin A (1998) CpG methylation, chromatin structure and gene silencing-a three-way connection. EMBO J 17: 4905–4908. doi: 10.1093/emboj/17.17.4905
  24. 24. Waterland RA, Jirtle RL (2003) Transposable elements: targets for early nutritional effects on epigenetic gene regulation. Mol Cell Biol 23: 5293–5300. doi: 10.1128/mcb.23.15.5293-5300.2003
  25. 25. Dolinoy DC, Weidman JR, Waterland RA, Jirtle RL (2006) Maternal genistein alters coat color and protects Avy mouse offspring from obesity by modifying the fetal epigenome. Environ Health Perspect 114: 567–572. doi: 10.1289/ehp.8700
  26. 26. Jirtle RL, Skinner MK (2007) Environmental epigenomics and disease susceptibility. Nat Rev Genet 8: 253–262. doi: 10.1038/nrg2045
  27. 27. MacLennan NK, James SJ, Melnyk S, Piroozi A, Jernigan S, et al. (2004) Uteroplacental insufficiency alters DNA methylation, one-carbon metabolism, and histone acetylation in IUGR rats. Physiol Genomics 18: 43–50. doi: 10.1152/physiolgenomics.00042.2004
  28. 28. Ke X, Lei Q, James SJ, Kelleher SL, Melnyk S, et al. (2006) Uteroplacental insufficiency affects epigenetic determinants of chromatin structure in brains of neonatal and juvenile IUGR rats. Physiol Genomics 25: 16–28. doi: 10.1152/physiolgenomics.00093.2005
  29. 29. Weaver IC, Diorio J, Seckl JR, Szyf M, Meaney MJ (2004) Early environmental regulation of hippocampal glucocorticoid receptor gene expression: characterization of intracellular mediators and potential genomic target sites. Ann N Y Acad Sci 1024: 182–212. doi: 10.1196/annals.1321.099
  30. 30. Sinclair KD, Allegrucci C, Singh R, Gardner DS, Sebastian S, et al. (2007) DNA methylation, insulin resistance, and blood pressure in offspring determined by maternal periconceptional B vitamin and methionine status. Proc Natl Acad Sci U S A 104: 19351–19356. doi: 10.1073/pnas.0707258104
  31. 31. McGowan PO, Suderman M, Sasaki A, Huang TC, Hallett M, et al. (2011) Broad epigenetic signature of maternal care in the brain of adult rats. PLoS One 6: e14739. doi: 10.1371/journal.pone.0014739
  32. 32. Provencal N, Suderman MJ, Guillemin C, Massart R, Ruggiero A, et al. (2012) The signature of maternal rearing in the methylome in rhesus macaque prefrontal cortex and T cells. J Neurosci 32: 15626–15642. doi: 10.1523/jneurosci.1470-12.2012
  33. 33. Terry MB, Ferris JS, Pilsner R, Flom JD, Tehranifar P, et al. (2008) Genomic DNA methylation among women in a multiethnic New York City birth cohort. Cancer Epidemiol Biomarkers Prev 17: 2306–2310. doi: 10.1158/1055-9965.epi-08-0312
  34. 34. Borghol N, Lornage J, Blachere T, Sophie Garret A, Lefevre A (2006) Epigenetic status of the H19 locus in human oocytes following in vitro maturation. Genomics 87: 417–426. doi: 10.1016/j.ygeno.2005.10.008
  35. 35. Heijmans BT, Tobi EW, Stein AD, Putter H, Blauw GJ, et al. (2008) Persistent epigenetic differences associated with prenatal exposure to famine in humans. Proc Natl Acad Sci U S A 105: 17046–17049. doi: 10.1073/pnas.0806560105
  36. 36. Oberlander TF, Weinberg J, Papsdorf M, Grunau R, Misri S, et al. (2008) Prenatal exposure to maternal depression, neonatal methylation of human glucocorticoid receptor gene (NR3C1) and infant cortisol stress responses. Epigenetics 3: 97–106. doi: 10.4161/epi.3.2.6034
  37. 37. McGowan PO, Sasaki A, D’Alessio AC, Dymov S, Labonte B, et al. (2009) Epigenetic regulation of the glucocorticoid receptor in human brain associates with childhood abuse. Nat Neurosci 12: 342–348. doi: 10.1038/nn.2270
  38. 38. Borghol N, Suderman M, McArdle W, Racine A, Hallett M, et al. (2012) Associations with early-life socio-economic position in adult DNA methylation. Int J Epidemiol 41: 62–74. doi: 10.1093/ije/dyr147
  39. 39. Kinnally EL, Feinberg C, Kim D, Ferguson K, Leibel R, et al. (2011) DNA methylation as a risk factor in the effects of early life stress. Brain Behav Immun 25: 1548–1553. doi: 10.1016/j.bbi.2011.05.001
  40. 40. Smith AK, Conneely KN, Kilaru V, Mercer KB, Weiss TE, et al. (2011) Differential immune system DNA methylation and cytokine regulation in post-traumatic stress disorder. Am J Med Genet B Neuropsychiatr Genet 156B: 700–708. doi: 10.1002/ajmg.b.31212
  41. 41. Uddin M, Koenen KC, Aiello AE, Wildman DE, de los Santos R, et al. (2011) Epigenetic and inflammatory marker profiles associated with depression in a community-based epidemiologic sample. Psychol Med 41: 997–1007. doi: 10.1017/s0033291710001674
  42. 42. Wang D, Szyf M, Benkelfat C, Provencal N, Turecki G, et al. (2012) Peripheral SLC6A4 DNA methylation is associated with in vivo measures of human brain serotonin synthesis and childhood physical aggression. PLoS One 7: e39501. doi: 10.1371/journal.pone.0039501
  43. 43. Essex MJ, Boyce WT, Hertzman C, Lam LL, Armstrong JM, et al. (2013) Epigenetic vestiges of early developmental adversity: childhood stress exposure and DNA methylation in adolescence. Child Dev 84: 58–75. doi: 10.1111/j.1467-8624.2011.01641.x
  44. 44. Naumova OY, Lee M, Koposov R, Szyf M, Dozier M, et al. (2012) Differential patterns of whole-genome DNA methylation in institutionalized children and children raised by their biological parents. Dev Psychopathol 24: 143–155. doi: 10.1017/s0954579411000605
  45. 45. Klengel T, Mehta D, Anacker C, Rex-Haffner M, Pruessner JC, et al. (2013) Allele-specific FKBP5 DNA demethylation mediates gene-childhood trauma interactions. Nat Neurosci 16: 33–41. doi: 10.1038/nn.3275
  46. 46. Ressler KJ, Mercer KB, Bradley B, Jovanovic T, Mahan A, et al. (2011) Post-traumatic stress disorder is associated with PACAP and the PAC1 receptor. Nature 470: 492–497. doi: 10.1038/nature09856
  47. 47. Provencal N, Suderman MJ, Caramaschi D, Wang D, Hallett M, et al. (2013) Differential DNA methylation regions in cytokine and transcription factor genomic Loci associate with childhood physical aggression. PLoS One 8: e71691. doi: 10.1371/journal.pone.0071691
  48. 48. Smyth GK (2005) Limma: linear models for microarray data. In: R. Gentleman VC, S Dudoit, R Irizarry, W Huber editor. Bioinformatics and Computational Biology Solutions using R and Bioconductor. New York: Springer,. 397–420.
  49. 49. Gentleman RC, Carey VJ, Bates DM, Bolstad B, Dettling M, et al. (2004) Bioconductor: open software development for computational biology and bioinformatics. Genome Biol 5: R80. doi: 10.1186/gb-2004-5-10-r80
  50. 50. Rossi RL, Rossetti G, Wenandy L, Curti S, Ripamonti A, et al. (2011) Distinct microRNA signatures in human lymphocyte subsets and enforcement of the naive state in CD4+ T cells by the microRNA miR-125b. Nat Immunol 12: 796–803. doi: 10.1038/ni.2057
  51. 51. Birger M, Swartz M, Cohen D, Alesh Y, Grishpan C, et al. (2003) Aggression: the testosterone-serotonin link. Isr Med Assoc J 5: 653–658.
  52. 52. Coccaro EF, Kavoussi RJ, Trestman RL, Gabriel SM, Cooper TB, et al. (1997) Serotonin function in human subjects: intercorrelations among central 5-HT indices and aggressiveness. Psychiatry Res 73: 1–14. doi: 10.1016/s0165-1781(97)00108-x
  53. 53. Higley JD, Mehlman PT, Poland RE, Taub DM, Vickers J, et al. (1996) CSF testosterone and 5-HIAA correlate with different types of aggressive behaviors. Biol Psychiatry 40: 1067–1082. doi: 10.1016/s0006-3223(95)00675-3
  54. 54. Tuinier S, Verhoeven WM, van Praag HM (1995) Cerebrospinal fluid 5-hydroxyindolacetic acid and aggression: a critical reappraisal of the clinical data. Int Clin Psychopharmacol 10: 147–156. doi: 10.1097/00004850-199510030-00003
  55. 55. Perez-Rodriguez MM, Weinstein S, New AS, Bevilacqua L, Yuan Q, et al. (2010) Tryptophan-hydroxylase 2 haplotype association with borderline personality disorder and aggression in a sample of patients with personality disorders and healthy controls. J Psychiatr Res 44: 1075–1081. doi: 10.1016/j.jpsychires.2010.03.014
  56. 56. Weaver IC, Cervoni N, Champagne FA, D’Alessio AC, Sharma S, et al. (2004) Epigenetic programming by maternal behavior. Nat Neurosci 7: 847–854. doi: 10.1038/nn1276
  57. 57. Murani E, Ponsuksili S, D’Eath RB, Turner SP, Kurt E, et al. (2010) Association of HPA axis-related genetic variation with stress reactivity and aggressive behaviour in pigs. BMC Genet 11: 74. doi: 10.1186/1471-2156-11-74
  58. 58. Craig IW (2007) The importance of stress and genetic variation in human aggression. Bioessays 29: 227–236. doi: 10.1002/bies.20538
  59. 59. van Bokhoven I, Van Goozen SH, van Engeland H, Schaal B, Arseneault L, et al. (2005) Salivary cortisol and aggression in a population-based longitudinal study of adolescent males. J Neural Transm 112: 1083–1096. doi: 10.1007/s00702-004-0253-5
  60. 60. Shirtcliff EA, Granger DA, Booth A, Johnson D (2005) Low salivary cortisol levels and externalizing behavior problems in youth. Dev Psychopathol 17: 167–184. doi: 10.1017/s0954579405050091
  61. 61. Loney BR, Butler MA, Lima EN, Counts CA, Eckel LA (2006) The relation between salivary cortisol, callous-unemotional traits, and conduct problems in an adolescent non-referred sample. J Child Psychol Psychiatry 47: 30–36. doi: 10.1111/j.1469-7610.2005.01444.x
  62. 62. Popma A, Vermeiren R, Geluk CA, Rinne T, van den Brink W, et al. (2007) Cortisol moderates the relationship between testosterone and aggression in delinquent male adolescents. Biol Psychiatry 61: 405–411. doi: 10.1016/j.biopsych.2006.06.006
  63. 63. Keck ME, Wigger A, Welt T, Muller MB, Gesing A, et al. (2002) Vasopressin mediates the response of the combined dexamethasone/CRH test in hyper-anxious rats: implications for pathogenesis of affective disorders. Neuropsychopharmacology 26: 94–105. doi: 10.1016/s0893-133x(01)00351-7
  64. 64. Kitay JI (1961) Sex differences in adrenal cortical secretion in the rat. Endocrinology 68: 818–824. doi: 10.1210/endo-68-5-818
  65. 65. Bauer ME, Wieck A, Lopes RP, Teixeira AL, Grassi-Oliveira R (2010) Interplay between neuroimmunoendocrine systems during post-traumatic stress disorder: a minireview. Neuroimmunomodulation 17: 192–195. doi: 10.1159/000258721
  66. 66. Dowlati Y, Herrmann N, Swardfager W, Liu H, Sham L, et al. (2010) A meta-analysis of cytokines in major depression. Biol Psychiatry 67: 446–457. doi: 10.1016/j.biopsych.2009.09.033
  67. 67. Groer MW, Morgan K (2007) Immune, health and endocrine characteristics of depressed postpartum mothers. Psychoneuroendocrinology 32: 133–139. doi: 10.1016/j.psyneuen.2006.11.007
  68. 68. Hoge EA, Brandstetter K, Moshier S, Pollack MH, Wong KK, et al. (2009) Broad spectrum of cytokine abnormalities in panic disorder and posttraumatic stress disorder. Depress Anxiety 26: 447–455. doi: 10.1002/da.20564
  69. 69. Janelidze S, Mattei D, Westrin A, Traskman-Bendz L, Brundin L (2011) Cytokine levels in the blood may distinguish suicide attempters from depressed patients. Brain Behav Immun 25: 335–339. doi: 10.1016/j.bbi.2010.10.010
  70. 70. Kawamura A, Yoshikawa T, Takahashi T, Hayashi T, Takahashi E, et al. (2001) Randomized trial of phosphodiesterase inhibitors versus catecholamines in patients with acutely decompensated heart failure. Jpn Circ J 65: 858–862. doi: 10.1253/jcj.65.858
  71. 71. Noh YH, Lee SM, Kim EJ, Kim DY, Lee H, et al. (2008) Improvement of cardiovascular risk factors in patients with type 2 diabetes after long-term continuous subcutaneous insulin infusion. Diabetes Metab Res Rev 24: 384–391. doi: 10.1002/dmrr.849
  72. 72. Misener VL, Gomez L, Wigg KG, Luca P, King N, et al. (2008) Cytokine Genes TNF, IL1A, IL1B, IL6, IL1RN and IL10, and childhood-onset mood disorders. Neuropsychobiology 58: 71–80. doi: 10.1159/000159775
  73. 73. O’Brien SM, Scott LV, Dinan TG (2004) Cytokines: abnormalities in major depression and implications for pharmacological treatment. Hum Psychopharmacol 19: 397–403. doi: 10.1002/hup.609
  74. 74. Pace TW, Hu F, Miller AH (2007) Cytokine-effects on glucocorticoid receptor function: relevance to glucocorticoid resistance and the pathophysiology and treatment of major depression. Brain Behav Immun 21: 9–19. doi: 10.1016/j.bbi.2006.08.009
  75. 75. von Kanel R, Hepp U, Kraemer B, Traber R, Keel M, et al. (2007) Evidence for low-grade systemic proinflammatory activity in patients with posttraumatic stress disorder. J Psychiatr Res 41: 744–752. doi: 10.1016/j.jpsychires.2006.06.009
  76. 76. Zalcman SS, Siegel A (2006) The neurobiology of aggression and rage: role of cytokines. Brain Behav Immun 20: 507–514. doi: 10.1016/j.bbi.2006.05.002
  77. 77. Siegel A, Bhatt S, Bhatt R, Zalcman SS (2007) The neurobiological bases for development of pharmacological treatments of aggressive disorders. Curr Neuropharmacol 5: 135–147. doi: 10.2174/157015907780866929
  78. 78. Nelson RJ, Chiavegatto S (2001) Molecular basis of aggression. Trends Neurosci 24: 713–719. doi: 10.1016/s0166-2236(00)01996-2
  79. 79. Pesce M, Speranza L, Franceschelli S, Ialenti V, Patruno A, et al. (2011) Biological role of interleukin-1beta in defensive-aggressive behaviour. J Biol Regul Homeost Agents 25: 323–329.
  80. 80. Hassanain M, Bhatt S, Zalcman S, Siegel A (2005) Potentiating role of interleukin-1beta (IL-1beta) and IL-1beta type 1 receptors in the medial hypothalamus in defensive rage behavior in the cat. Brain Res 1048: 1–11. doi: 10.1016/j.brainres.2005.04.086
  81. 81. Suderman M, McGowan PO, Sasaki A, Huang TC, Hallett MT, et al. (2012) Conserved epigenetic sensitivity to early life experience in the rat and human hippocampus. Proc Natl Acad Sci U S A 109 Suppl 217266–17272. doi: 10.1073/pnas.1121260109
  82. 82. Widom CS (1989) Child abuse, neglect, and adult behavior: research design and findings on criminality, violence, and child abuse. Am J Orthopsychiatry 59: 355–367. doi: 10.1111/j.1939-0025.1989.tb01671.x
  83. 83. Gammie SC, Seasholtz AF, Stevenson SA (2008) Deletion of corticotropin-releasing factor binding protein selectively impairs maternal, but not intermale aggression. Neuroscience 157: 502–512. doi: 10.1016/j.neuroscience.2008.09.026