**5. Stress memory and immune-like rearrangement in the human brain**

Doubts have more recently been raised whether gene transcription activated by dendritic calcium signals is sufficient to consolidate long-term functional alterations associated with memory consolidation. An alternative genomic hypothesis of memory suggests that acquired information is persistently stored within individual neurons through modifications of DNA, and that these modifications serve as the carriers of elementary memory traces. The emerging idea is therefore that lifelong behavioural memory storage may involve lasting changes in the physical, three-dimensional structure of DNA itself and chromatin alterations are emerging as a key epigenetic mechanism in the process in conjunction with usedependent synaptic plasticity (Levenson & Sweatt, 2006; Delcuve et al., 2009). The expression of immune recombination activating genes in key stress-induced memory regions in the brain suggests the adoption by the brain of this ancient pattern recognition and memory system to establish a structural basis for long-term memory through controlled chromosomal breakage at highly specific genomic regions. Fundamentally unstable processes with narrow safety margins (controlled chromosomal breakage) thus appear to underlie pattern recognition and memory consolidation in both the immune system and brain.

Unusual genetic mechanisms for diversifying recognition proteins may be a widespread characteristic of animal immunity and may have paved the way for adaptation for management of neural sensory information (Litman et al., 2005). Stress reactions form part of neuroendocrine influences that also modulate immune function. The appearance of a lymphocyte-based recombinatorial system of anticipatory immunity in vertebrates approximately 500 mya facilitated developmental and morphological plasticity in addition to the advantage conferred by the ability to recognize a larger portion of the antigenic world (Pancer & Cooper, 2006). A prototypic example of epigenetic-facilitation in memory retention pertains to memory T-cells of the mammalian immune system (reviewed in Nakayama & Yamashita (2008)). Numerous epigenetic mechanisms such as histone modifications and DNA methylation modulate gene expression and thus play a role in T-cell survival and maintenance of T-cell function in various differentiated states. These processes underlie the formation of persistent immunological memory cells in response to transient environmental stimuli (reviewed in Nakayama & Yamashita (2008). Thus, like immune Tcells, it is plausible that epigenetic mechanisms such as methylation of the cytosine base are changeable and occur in post-mitotic neurons to mediate neuronal function. However,

Stress Shaping Brains: Higher Order DNA/Chromosome

(Dietert & Dietert, 2008).

**5.2 CFS represent a network stress response** 

Mechanisms Underlying Epigenetic Programming of the Brain Transcriptome 359

particularly repeat sequences and transposable elements, and is believed to result in chromosomal instability and hypomethylation of regulatory DNA sequences activates transcription of protooncogenes, retrotransposons, as well as genes encoding proteins involved in genomic instability (Glover, 2006; Wilson et al., 2007). Retroelements represent evolutionary forces that establish and hone target gene networks of transcription factors in a species-specific manner. LTR class I endogenous retrovirus (ERV) retroelements impact considerably the transcriptional network of human tumour suppressor protein p53. A total of 1,509 of approximately 319,000 human ERV LTR regions have a near-perfect p53 DNA binding site. Human ERV p53 sites are likely part of the p53 transcriptional program and direct regulation of p53 target genes (Wang et al., 2007). Recent findings showed that key cell cycle checkpoint genes are important for genome stability at fragile sites. Altered sequences arising from chromosomal rearrangement and associated transposable element (TE) upregulation during 'cognitive stress' may result in neurospecific immune-like sequelae involving CFS as key participating regions. DNA double-strand break repair proteins were recognized 20 years ago as a major target of autoantibodies. Dysregulation of these processes can be considered to increase the risk for subsequently developing systemic inflammatory disorders through a central immunologically modified state and sensitization for increased stress responses in susceptible individuals. Because early changes may include misregulation of resident inflammatory myelomonocytic cells in the developing brain, this could be associated with prenatal–neonatal brain pathologies and neurobehavioural deficits

When data on CFS expression were analysed in a network context, it appeared that chromosomal fragile site associated genes function as part of a highly conserved stress response network (Re et al., 2006). The regulatory genome supplies an enormous computational capability with the capacity to process in parallel a vast number of regulatory inputs, comprising many thousands of processing units in the form of cis-regulatory modules. The interconnected cis-regulatory modules that control regulatory gene expression create a network that is the underlying mechanism of specification and illustrate the information processing that is done by the regulatory sequences (Ben-Tabou de-Leon & Davidson, 2007). AT islands in CFS have been shown to function as nuclear matrix attachment regions (MARs) both in vitro and in vivo (Jackson et al., 2003), which constitute the functional coordinate system for genomic regulatory regions (Liebich et al., 2002). DNA duplexes of AT islands are prone to base unpairing due to their unusual flexibility characteristics, which are necessary MAR attributes. Recent studies on the molecular mechanisms involved show that proteins of the nuclear envelope participate in regulation of transcription on several levels, from direct binding to transcription factors to induction of epigenetic histone modifications [Skalai et al., 2007]. MARs organize chromosomal loops in the interphase nucleus, are about 200 bp long, AT-rich, contain topoisomerase II consensus sequences and other AT-rich sequence motifs; often reside near cis-acting regulatory sequences, and their binding sites are abundant (greater than 10,000 per mammalian nucleus) (Blasquez et al., 1998). Cis-elements can be defined to include the repeat sequence units, the length and purity of the repeat tracts, the sequences flanking the repeat, as well as the surrounding epigenetic environment, including DNA methylation and chromatin structure (Cleary & Pearson, 2003). Contacts between cis-acting sequences through the formation of chromatin loops form the most basic level of organization that impedes or

unlike epigenetic mechanisms in the immune system, chromatin modifications in the CNS are greatly understudied. Lymphoid cells purposely introduce DNA double strand breaks into their genome to maximize the diversity and effector functions of their antigen receptor genes (Rooney & Chaudhuri, 2004). Recombinase activation gene RAG-1 directed V(D)J recombination affecting only specific recognition sequences allows the immune system to encode memories of a vast array of antigens. Research findings provide a formal demonstration that certain CFS can function as signals for RAG complex targets (Raghavan et al., 2001). Conversely, CFS were found to be enriched for genes associated with the immune response (Re et al., 2006). RAG proteins have been proposed to contribute to chromosomal translocations in general (Chatterji et al., 2004), suggesting that these may be involved in immune-like stress induced rearrangement processes following breakage at CFS. Rag1 positive-cells mainly appear in the amygdalae, hypothalamus, thalamus and hippocampus at developmental stage (Sun et al., 2007). The RAG-1 gene is also localized to neurons in the hippocampal formation and related limbic regions that are involved in spatial learning and memory as well as other parameters of neurobehavioural performance (Cushman et al., 2003). While the role of RAG-1 in learning and memory in humans still has to be determined, it remains attractive to propose that their localization in relevant anatomical areas in the brain, the importance of epigenetics changes and the postulated role of chromosomal rearrangement make this an interesting area for future studies.

#### **5.1 CFS, RAG genes and transposable elements (TE's)**

It has been motivated that the recombination system that carries out rearrangements may be a significant evolutionary force, perhaps not limited to rearrangements only at antigenreceptor loci (Roth 2000) (Chuzhanova et al., 2009). Genomic changes in V-gene structure, created by RAG recombinase acting on germline recombination signal sequences, led variously to the generation of fixed receptor specificities, pseudogene templates for gene conversion, and ultimately to Ig sequences that evolved away from Ig function (Hsu et al., 2006). RAG1 and RAG2, like the adaptive immune system itself, are found exclusively in jawed vertebrates, and are thought to have entered the vertebrate genome by horizontal transmission as components of a transposable element (Schatz, 2004). Such dynamicity allows extensive genome repatterning during transient stress phases (including oxidative stress signaling), during which some epigenetic features, such as DNA methylation, are relaxed, thus allowing transposable element (TE) amplification. Analysis of genomic rearrangement breakpoint regions has revealed specific TE repeat density patterns, suggesting that TEs may have played a significant role in chromosome evolution and genome plasticity. Hairpin DNA structures formed in palindromes (such as associated with CFS) are intermediates in V(D)J recombination and are formed by a chemical mechanism very similar to the early steps of transpositional recombination and retroviral integration. RAG proteins are able to capture exogenous target DNA molecules and carry out authentic transposition of signal ends into these targets.

Genomic instability has been indicated to involve epigenetic activation of mobile elements dispersed throughout the human genome (Stribinskis & Ramos, 2006). Barbara McClintock originally proposed that mobile elements restructure host genomes as an adaptive response to environmental challenge (McClintock, 1987; Dai et al., 2007). Retrotransposons are mobile genetic elements that can be amplified to high copy number and are considered to be an important source of genetic diversity (Grandbastien, 2004). Hypomethylation associated with genomic stress largely affects the intergenic and intronic regions of the DNA,

unlike epigenetic mechanisms in the immune system, chromatin modifications in the CNS are greatly understudied. Lymphoid cells purposely introduce DNA double strand breaks into their genome to maximize the diversity and effector functions of their antigen receptor genes (Rooney & Chaudhuri, 2004). Recombinase activation gene RAG-1 directed V(D)J recombination affecting only specific recognition sequences allows the immune system to encode memories of a vast array of antigens. Research findings provide a formal demonstration that certain CFS can function as signals for RAG complex targets (Raghavan et al., 2001). Conversely, CFS were found to be enriched for genes associated with the immune response (Re et al., 2006). RAG proteins have been proposed to contribute to chromosomal translocations in general (Chatterji et al., 2004), suggesting that these may be involved in immune-like stress induced rearrangement processes following breakage at CFS. Rag1 positive-cells mainly appear in the amygdalae, hypothalamus, thalamus and hippocampus at developmental stage (Sun et al., 2007). The RAG-1 gene is also localized to neurons in the hippocampal formation and related limbic regions that are involved in spatial learning and memory as well as other parameters of neurobehavioural performance (Cushman et al., 2003). While the role of RAG-1 in learning and memory in humans still has to be determined, it remains attractive to propose that their localization in relevant anatomical areas in the brain, the importance of epigenetics changes and the postulated role

of chromosomal rearrangement make this an interesting area for future studies.

It has been motivated that the recombination system that carries out rearrangements may be a significant evolutionary force, perhaps not limited to rearrangements only at antigenreceptor loci (Roth 2000) (Chuzhanova et al., 2009). Genomic changes in V-gene structure, created by RAG recombinase acting on germline recombination signal sequences, led variously to the generation of fixed receptor specificities, pseudogene templates for gene conversion, and ultimately to Ig sequences that evolved away from Ig function (Hsu et al., 2006). RAG1 and RAG2, like the adaptive immune system itself, are found exclusively in jawed vertebrates, and are thought to have entered the vertebrate genome by horizontal transmission as components of a transposable element (Schatz, 2004). Such dynamicity allows extensive genome repatterning during transient stress phases (including oxidative stress signaling), during which some epigenetic features, such as DNA methylation, are relaxed, thus allowing transposable element (TE) amplification. Analysis of genomic rearrangement breakpoint regions has revealed specific TE repeat density patterns, suggesting that TEs may have played a significant role in chromosome evolution and genome plasticity. Hairpin DNA structures formed in palindromes (such as associated with CFS) are intermediates in V(D)J recombination and are formed by a chemical mechanism very similar to the early steps of transpositional recombination and retroviral integration. RAG proteins are able to capture exogenous target DNA molecules and carry out authentic

Genomic instability has been indicated to involve epigenetic activation of mobile elements dispersed throughout the human genome (Stribinskis & Ramos, 2006). Barbara McClintock originally proposed that mobile elements restructure host genomes as an adaptive response to environmental challenge (McClintock, 1987; Dai et al., 2007). Retrotransposons are mobile genetic elements that can be amplified to high copy number and are considered to be an important source of genetic diversity (Grandbastien, 2004). Hypomethylation associated with genomic stress largely affects the intergenic and intronic regions of the DNA,

**5.1 CFS, RAG genes and transposable elements (TE's)** 

transposition of signal ends into these targets.

particularly repeat sequences and transposable elements, and is believed to result in chromosomal instability and hypomethylation of regulatory DNA sequences activates transcription of protooncogenes, retrotransposons, as well as genes encoding proteins involved in genomic instability (Glover, 2006; Wilson et al., 2007). Retroelements represent evolutionary forces that establish and hone target gene networks of transcription factors in a species-specific manner. LTR class I endogenous retrovirus (ERV) retroelements impact considerably the transcriptional network of human tumour suppressor protein p53. A total of 1,509 of approximately 319,000 human ERV LTR regions have a near-perfect p53 DNA binding site. Human ERV p53 sites are likely part of the p53 transcriptional program and direct regulation of p53 target genes (Wang et al., 2007). Recent findings showed that key cell cycle checkpoint genes are important for genome stability at fragile sites. Altered sequences arising from chromosomal rearrangement and associated transposable element (TE) upregulation during 'cognitive stress' may result in neurospecific immune-like sequelae involving CFS as key participating regions. DNA double-strand break repair proteins were recognized 20 years ago as a major target of autoantibodies. Dysregulation of these processes can be considered to increase the risk for subsequently developing systemic inflammatory disorders through a central immunologically modified state and sensitization for increased stress responses in susceptible individuals. Because early changes may include misregulation of resident inflammatory myelomonocytic cells in the developing brain, this could be associated with prenatal–neonatal brain pathologies and neurobehavioural deficits (Dietert & Dietert, 2008).

#### **5.2 CFS represent a network stress response**

When data on CFS expression were analysed in a network context, it appeared that chromosomal fragile site associated genes function as part of a highly conserved stress response network (Re et al., 2006). The regulatory genome supplies an enormous computational capability with the capacity to process in parallel a vast number of regulatory inputs, comprising many thousands of processing units in the form of cis-regulatory modules. The interconnected cis-regulatory modules that control regulatory gene expression create a network that is the underlying mechanism of specification and illustrate the information processing that is done by the regulatory sequences (Ben-Tabou de-Leon & Davidson, 2007). AT islands in CFS have been shown to function as nuclear matrix attachment regions (MARs) both in vitro and in vivo (Jackson et al., 2003), which constitute the functional coordinate system for genomic regulatory regions (Liebich et al., 2002). DNA duplexes of AT islands are prone to base unpairing due to their unusual flexibility characteristics, which are necessary MAR attributes. Recent studies on the molecular mechanisms involved show that proteins of the nuclear envelope participate in regulation of transcription on several levels, from direct binding to transcription factors to induction of epigenetic histone modifications [Skalai et al., 2007]. MARs organize chromosomal loops in the interphase nucleus, are about 200 bp long, AT-rich, contain topoisomerase II consensus sequences and other AT-rich sequence motifs; often reside near cis-acting regulatory sequences, and their binding sites are abundant (greater than 10,000 per mammalian nucleus) (Blasquez et al., 1998). Cis-elements can be defined to include the repeat sequence units, the length and purity of the repeat tracts, the sequences flanking the repeat, as well as the surrounding epigenetic environment, including DNA methylation and chromatin structure (Cleary & Pearson, 2003). Contacts between cis-acting sequences through the formation of chromatin loops form the most basic level of organization that impedes or

Stress Shaping Brains: Higher Order DNA/Chromosome

GN variants (Boldogkoi 2004).

gestures and oral traditions.

**6.1 Understanding the interactome** 

**network dynamics** 

Mechanisms Underlying Epigenetic Programming of the Brain Transcriptome 361

continuous restructuring of the composition of GNs rather than fixing of specific alleles or

Unlike most optimization methods working from a single point in the decision space and employing a transition method to determine the next point, in a densely interconnected system genetic algorithms work from an entire "population" of points simultaneously, trying many directions in parallel and employing a combination of several genetically-inspired methods to determine the next population of points (Cantu-Paz & Goldberg, 1999). These aspects are likely to have been linked with evolutionary recruitment of an increasing number of gene promoters as members of progressively intricate gene expression networks employing different patterns of expression of stable household genes. Such principles may reflect the human ability to *combine and recombine* highly differentiated actions, perceptions, and concepts in order to construct larger, more complex, and highly variable units in a variety of behavioural domains including language, social intelligence, tool-making, and motor sequences (Gibson, 2002). It has been suggested that speech development and visual interpretation is characterized by multipart representations formed from elementary canonical parts (e.g., phonemes in speech, geons in visual perception) (Corballis, 1992), and in such new combinations similarly later gave rise to the introduction of iconic symbols used in art, writing and reading when information management became too complex for

**6. Analytical challenges in building complex disease investigative models –** 

The reductionistic approaches which have been successful in the early history of human genetics dealt with so-called 'single gene disorders' (increasingly a challenged concept), and currently fail to uncover the information required for insight into complex gene environment interaction such as required for studies in 'brain and behaviour'. Furthermore, bioenergetic metabolism as related to mitochondrial genetics (including both mitochondrial genes and nuclear genes involved with bioenergetic crosstalk) (Wallace, 2010) as well as epigenetic modification of DNA regulatory structures are considered to be increasingly important in neuroplasticity. Most of the gene identification studies have assayed for only one type of epigenetic marker. The problem of not evaluating the entire array of epigenetic modifications at specific gene promoters, along with the fact that most available gene chips fail to cover a large portion of the genome, means current technology has not yet reached the levels needed to fully assess the gene expression changes responsible for mediating

An excellent review of the topic was recently published in Nature Genetics (Barabási et al., 2011). The potential complexity of the human interactome (all the interactions between biological entities in cells and organisms considered as a whole), is daunting. The past decade has seen an exceptional growth in human specific molecular interaction data. Network based approaches to human disease have multiple biological and clinical implications. Networks operating in biological, technological or social systems are not random, but are characterized by a core set of organizing principles. Proteins that are involved in the same disease show a high propensity to interactly directly with each other. Thus, each disease may be linked to a well defined neighbourhood of the interactome, often

many of the epigenetically-associated phenotypes in the adult brain.

permits access of factors to the genes (Dillon, 2006). It is suggested that this links to developmental remodelling of neuronal connectivity and differential network connectivity has been suggested to form the basis for species-specific network connections as key drivers of evolutionary change (Boldogkoi 2004). The behavioural phenotype manifests itself as an emergent property of such networks (Anholt, 2004).

#### **5.3 Chromosomal breakage and network assembly, gene duplication and gene copy number variation**

Analyses support a nonrandom model of chromosomal evolution associated with both recurrent small-scale duplication and large-scale evolutionary rearrangements (Hinsch & Hannenhalli, 2006). Similarly, the human brain appears to have developed anatomically by the divergent modification of pre-existing parts (Striedter, 1998) and new areas may have evolved as a result of processes likely to be linked with underlying extensive duplication of transcription factors (Babu et al., 2004) or genes. The functional characterization through analysis of the ontology of genes located at connected fragile sites clearly highlights that a great proportion of genes with significant annotated terms are involved in innate and adaptive immune responses and in particular in pathways characteristic of activated T lymphocytes (Re et al., 2006). From these findings it has been proposed that correlated breakage at fragile sites may originate in proliferating lymphocytes from a co-regulated modified expression of fragile genes; in this view the genes identified by ontological analysis may be new fragile genes; chromatin changes and DNA replication alteration at or near these genes would be produced by cellular processes connected with their coregulation performed through still unknown mechanisms. This is supported by the observation that a number of the analysed cytokine-related genes show actual functional interactions in lymphocytes or other cell types (Re et al., 2006). Duplicate genes rapidly diverge in their expression profiles in the network and contribute to maintaining network robustness as compared with singletons (Chung et al., 2006) and according to modelling analyses, duplication plays an important role in feed-forward loop evolution (Cordero et al., 2006). Gene copy number variation has been considered to underlie a significant proportion of normal human variation including differences in cognitive, behavioural, and psychological features (Lee & Lupski, 2006).

Dynamic interactions between components of living cells (e. g., proteins, genes) exist on genomic, transcriptomic, proteomic and metabolomic levels. The levels themselves are heavily interconnected, resulting in complex networks of different interacting biological entities (Bosman et al., 2007). Some novel data suggest that a large amount of genetic variation exists in the regulatory region of genes within populations. In addition, comparison of homologous DNA sequences of various species shows that evolution appears to depend more strongly on gene expression than on the genes themselves. Furthermore, it has been demonstrated in several systems that genes form functional networks, whose products exhibit interrelated expression profiles. Finally, it has been found that regulatory circuits of development behave as evolutionary units (Boldogkoi 2004).

These data demonstrate that (1) Instead of individual genes, gene networks (GNs) are responsible for the determination of traits and behaviours. (2) The primary source of microevolution is considered to be the intraspecific polymorphism in GNs and not the allelic variation in either the coding or the regulatory sequences of individual genes. (3) GN polymorphism is generated by the variation in the regulatory regions of the component genes and not by the variance in their coding sequences. (4) Evolution proceeds through

permits access of factors to the genes (Dillon, 2006). It is suggested that this links to developmental remodelling of neuronal connectivity and differential network connectivity has been suggested to form the basis for species-specific network connections as key drivers of evolutionary change (Boldogkoi 2004). The behavioural phenotype manifests itself as an

**5.3 Chromosomal breakage and network assembly, gene duplication and gene copy** 

Analyses support a nonrandom model of chromosomal evolution associated with both recurrent small-scale duplication and large-scale evolutionary rearrangements (Hinsch & Hannenhalli, 2006). Similarly, the human brain appears to have developed anatomically by the divergent modification of pre-existing parts (Striedter, 1998) and new areas may have evolved as a result of processes likely to be linked with underlying extensive duplication of transcription factors (Babu et al., 2004) or genes. The functional characterization through analysis of the ontology of genes located at connected fragile sites clearly highlights that a great proportion of genes with significant annotated terms are involved in innate and adaptive immune responses and in particular in pathways characteristic of activated T lymphocytes (Re et al., 2006). From these findings it has been proposed that correlated breakage at fragile sites may originate in proliferating lymphocytes from a co-regulated modified expression of fragile genes; in this view the genes identified by ontological analysis may be new fragile genes; chromatin changes and DNA replication alteration at or near these genes would be produced by cellular processes connected with their coregulation performed through still unknown mechanisms. This is supported by the observation that a number of the analysed cytokine-related genes show actual functional interactions in lymphocytes or other cell types (Re et al., 2006). Duplicate genes rapidly diverge in their expression profiles in the network and contribute to maintaining network robustness as compared with singletons (Chung et al., 2006) and according to modelling analyses, duplication plays an important role in feed-forward loop evolution (Cordero et al., 2006). Gene copy number variation has been considered to underlie a significant proportion of normal human variation including differences in cognitive, behavioural, and

Dynamic interactions between components of living cells (e. g., proteins, genes) exist on genomic, transcriptomic, proteomic and metabolomic levels. The levels themselves are heavily interconnected, resulting in complex networks of different interacting biological entities (Bosman et al., 2007). Some novel data suggest that a large amount of genetic variation exists in the regulatory region of genes within populations. In addition, comparison of homologous DNA sequences of various species shows that evolution appears to depend more strongly on gene expression than on the genes themselves. Furthermore, it has been demonstrated in several systems that genes form functional networks, whose products exhibit interrelated expression profiles. Finally, it has been found that regulatory

These data demonstrate that (1) Instead of individual genes, gene networks (GNs) are responsible for the determination of traits and behaviours. (2) The primary source of microevolution is considered to be the intraspecific polymorphism in GNs and not the allelic variation in either the coding or the regulatory sequences of individual genes. (3) GN polymorphism is generated by the variation in the regulatory regions of the component genes and not by the variance in their coding sequences. (4) Evolution proceeds through

circuits of development behave as evolutionary units (Boldogkoi 2004).

emergent property of such networks (Anholt, 2004).

psychological features (Lee & Lupski, 2006).

**number variation** 

continuous restructuring of the composition of GNs rather than fixing of specific alleles or GN variants (Boldogkoi 2004).

Unlike most optimization methods working from a single point in the decision space and employing a transition method to determine the next point, in a densely interconnected system genetic algorithms work from an entire "population" of points simultaneously, trying many directions in parallel and employing a combination of several genetically-inspired methods to determine the next population of points (Cantu-Paz & Goldberg, 1999). These aspects are likely to have been linked with evolutionary recruitment of an increasing number of gene promoters as members of progressively intricate gene expression networks employing different patterns of expression of stable household genes. Such principles may reflect the human ability to *combine and recombine* highly differentiated actions, perceptions, and concepts in order to construct larger, more complex, and highly variable units in a variety of behavioural domains including language, social intelligence, tool-making, and motor sequences (Gibson, 2002). It has been suggested that speech development and visual interpretation is characterized by multipart representations formed from elementary canonical parts (e.g., phonemes in speech, geons in visual perception) (Corballis, 1992), and in such new combinations similarly later gave rise to the introduction of iconic symbols used in art, writing and reading when information management became too complex for gestures and oral traditions.

### **6. Analytical challenges in building complex disease investigative models – network dynamics**

The reductionistic approaches which have been successful in the early history of human genetics dealt with so-called 'single gene disorders' (increasingly a challenged concept), and currently fail to uncover the information required for insight into complex gene environment interaction such as required for studies in 'brain and behaviour'. Furthermore, bioenergetic metabolism as related to mitochondrial genetics (including both mitochondrial genes and nuclear genes involved with bioenergetic crosstalk) (Wallace, 2010) as well as epigenetic modification of DNA regulatory structures are considered to be increasingly important in neuroplasticity. Most of the gene identification studies have assayed for only one type of epigenetic marker. The problem of not evaluating the entire array of epigenetic modifications at specific gene promoters, along with the fact that most available gene chips fail to cover a large portion of the genome, means current technology has not yet reached the levels needed to fully assess the gene expression changes responsible for mediating many of the epigenetically-associated phenotypes in the adult brain.

### **6.1 Understanding the interactome**

An excellent review of the topic was recently published in Nature Genetics (Barabási et al., 2011). The potential complexity of the human interactome (all the interactions between biological entities in cells and organisms considered as a whole), is daunting. The past decade has seen an exceptional growth in human specific molecular interaction data. Network based approaches to human disease have multiple biological and clinical implications. Networks operating in biological, technological or social systems are not random, but are characterized by a core set of organizing principles. Proteins that are involved in the same disease show a high propensity to interactly directly with each other. Thus, each disease may be linked to a well defined neighbourhood of the interactome, often

Stress Shaping Brains: Higher Order DNA/Chromosome

consistently occurring as comorbid disorders.

Mechanisms Underlying Epigenetic Programming of the Brain Transcriptome 363

chronic pain/inflammation problem arising subsequent to a hypersensitized pain and/or stress memory in genetically predisposed individuals probably as early as intrauterine life. First degree relatives have a significantly increased risk to develop FMS. Foetal programming is expected to result in severe pathophysiological hyperreactivity when exposed to subsequent stressful stimuli. This, and much other contemporary research implies that foetal and perinatal stress could have long lasting sequelae in adult life and that it also involves inappropriate immune upregulation. The 'at risk' FMS genotype may represent a risk for several non-classical FMS outcomes as central stress induced hypersensitisation appears to cause subsequent susceptibility to posttraumatic stress disorder-like phenomena, chronic pain syndromes, mood disorders and several other adult onset diseases. These effects often include, but are not limited to, anxiety, depression, ADHD, substance use disorders, and tobacco dependence as well as a dramatically increased risk for a variety of mental disorders (Bhadra & Petersel, 2010; González et al., 2010; Natelson 2010). There may exist cross reactions between various emotional stress and pain responses which involve both the immune and nervous systems. These share quite similar processes for pattern recognition and memory consolidation, and may represent a useful perspective from which to regard aetiological relationships between conditions

**7.1 Chromosomal fragility in chronic fatigue/fibromyalgia syndrome (FMS)** 

In 1995, during research on chromosomal fragile sites at the University of Pretoria, my cytogenetics collaborator, Ingrid Simonic, found an increased expression of common aphidicolin-inducible chromosomal fragile sites in FMS/"chronic fatigue" patients as opposed to unaffected intrafamilial controls (Fig 1) (Simonic & Gericke, Unpublished data).

Fig. 1. Higher frequencies of some aphidicolin-induced common chromosomal fragile sites

in FMS/chronic fatigue individuals versus first degree controls.

referred to as a 'disease module', representing a group of network components that together contribute to a cellular function and disruption which results in a particular disease phenotype.

#### **6.2 Mapping complex conditions with multiple comorbid disorders through a 'diseasome' approach**

Similarly, the systematic mapping of the network based dependencies between pathophenotypes and their disease modules has culminated in the concept of the 'diseasome' which represents disease maps whose nodes are diseases and whose links represent various molecular relationships between the disease associated cellular components (Barabási et al., 2011). Understanding such links not only helps us understand how different phenotypes, often addressed by different medical subdisciplines are linked at the molecular level but can also help us to comprehend why certain groups of diseases arise together. Diseasome - based approaches can be expected to aid drug discovery, in particular when it comes to the use of approved drugs to treat molecularly linked diseases. Single target drugs can be expected to correct some dysfunctional aspects of the disease module, but they could also alter the activity of molecules that are situated in the neighbourhood of the disease module, leading to detectable side effects. Analysis of drug target networks demonstrated that many drugs are palliative, that is they do not target the actual disease associated proteins, but proteins in their network neighbourhood. Finally, using network analytic capabilities, safer and more focused multi-target combinations can be designed e.g. for anti-inflammatory or anticancer drug combinations (Barabási et al., 2011).
