Journal of Clinical Epigenetics Open Access

Background

Epigenetics (or more specifically epigenetic landscape) was a term coined by Conrad Waddington (1942), originally used to describe his conceptual model of how genes might interact with the surrounding environment resulting in the expression of altered phenotypes. He used this term to denote permanent changes to the expression of genes that cause cell differentiation. Epigenetics now most commonly refers to the heritable changes of gene expression and activity that do not modify gene structure [1]. Epigenetic modifications have numerous functions but are most frequently associated with expression of genes that mediate embryonic development. Within developing embryos, epigenetic modifications influence erasure and re-establishment of DNA methylation; imprinting of the genome; inactivation of the X-chromosome; pluripotent stem cell development and control of somatic cell differentiation [2]. Recently, epigenetic alterations have also been implicated in disease pathology of psychiatric, neurodevelopmental and neurodegenerative disorders. In addition to this, some epigenetic alterations have been found to be heritable across several generations. As the epigenome is sensitive to environmental influences, there is a possibility that environmental factors could cause heritable epigenetic modifications that result in changes in gene expression that alter behavioural, psychiatric or psychological phenotypes in offspring.

Structure of the Genome

DNA exists within the nucleus of animal cells as a double helix structure usually tightly wrapped around proteins called histones. The four histones around which DNA is coiled are H2A, H2B, H3 and H4. These are repeated twice [3] making an organised octamer structure with DNA, which is known as a nucleosome. Linker histones from the H1 family prevent the DNA from becoming incorrectly wrapped by sitting on the entry and exit strands of DNA. Nucleosomes make up the repeating units that form eukaryotic chromatin - a highly ordered structure that ensures DNA can be packaged into the cell nucleus. One of the functions of chromatin is to control gene expression and replication [4].

Transcription and subsequently gene expression is dependent on the structure of chromatin. RNA polymerases are responsible for initiating transcription. However, RNA polymerase does not transcribe every gene within the genome. Areas of the DNA with a more open structure are more difficult to transcribe, and are known as heterochromatin. Open areas (known as euchromatin) allow the enzymes access to the genes and therefore they can be more easily transcribed and the genes expressed as RNA. Heterochromatin is basically tightly packed DNA, usually located on the periphery of the nucleus. It generally consists of genetically inactive satellite sequences of DNA [5], an excellent example of this is the inactivated X chromosome in females. There are two forms of heterochromatin: constitutive and facultative. Constitutive heterochromatin domains are highly condensed and usually consist of repetitive lengths of DNA that are mostly transcriptionally silent in every cell of an organism. For instance large stretches of the Y-chromosome, and the centromeric- and telomeric-regions of all chromosomes. In contrast, facultative heterochromatin refers to regions of the genome that are transcriptionally silent in some cell-types, but transcriptionally active in others. Once again the example of the inactive X chromosome and the active X chromosome is a good demonstration of this difference in chromatin structure.

The structural arrangement of chromatin and subsequent gene expression is subject to reversible epigenetic changes including DNA methylation, histone post-translational modifications (PTMs) and alteration by non-coding RNAs (ncRNAs). These are the epigenetic modifications of interest in this systematic review.

DNA Methylation

The 5’ position in the pyrimidine ring of cytosine bases in DNA can be modified with methyl groups to create 5-methylcytosine (m5C) by methyltransferases [6]. There are broadly two classes of DNA methyltransferase (DNMT) enzymes are responsible for this cytosine conversion; these are DNMTs responsible for de novo methylation and those that copy the methylation already existing onto daughter strands during cell replication [2]. Within mammals DNA methylation is localised to CpG dinucleotides. These are commonly found to cluster in sequences of 200 bases made from a minimum of 50% cytosine and guanine, which are known as CpG islands. In 40-50% of genes found in humans CpG islands can be found close to or within promoter regions of genes. However, such islands are often unmethylated and are transcription factor targets.

DNA methylation is generally associated with transcriptional silencing or repression and is implicated in gene regulation and development [6]. Consistent with its commonly found location on CpG islands, DNA methylation possibly blocks promoter regions and prevents the binding of activating transcription factors. Thus, DNA methylation is commonly associated with a lack of gene expression [7].

However, there is now an increasing awareness of the importance of non-CpG DNA-methylation, at so called CpH site (H = A, C, or T) and also of hydroxymethylation (5hmC), an intermediate in the recently identified active de-methylation process [8]. The existence of 5hmC in particular points towards a more dynamic role for DNA-methylation, which contrasts with prior notions of this being the most robust and stable epigenetic mechanism [9]

Histone Modifications

Post translational modifications of the histone proteins (PTMs) affect the 3 dimensional structure of chromatin and therefore alter gene expression [10] PTMs are an example of epigenetic modifications and are thought to affect gene expression through changing the structure of the condensed chromatin so different genes are exposed [11]. Histones can be altered with covalent modifications including acetylation, methylation, phosphorylation, sumoylation and ubiquitylation [12]. Histone acetylation and methylation regulate binding of effector molecules by altering chromatin structure [13]. Lysine residues on histones are subject to acetylation which neutralises the positive charge of lysine that attracts histones to the DNA backbone. Subsequently, acetylation of histones is associated with euchromatin and increased expression [14]. Histone methylation, in contrast, is associated with both enhancement of gene expression and repression depending on location of the modifications [15].

DNA methylation is often dependent on histone methylation of a particular lysine residue in histone H3 (H3-K9 methylation), which is another mechanism in which histone methylation is involved in gene repression. DNA methylation is also thought to be involved in the deacetylation of histones, further increasing its role in transcriptional silencing. Proteins can bind to methylated DNA to form complexes that result in deacetylation. Similarly, non-methylated DNA does not attract deacetylating enzymes thus chromatin structure remains more open and available for transcription.

Non-Coding RNAs

Non-Coding RNAs (ncRNAs) are RNA transcripts with no obvious protein product. ncRNAs have a role in transcriptional processes like DNA methylation and histone PTMs, but do not directly affect the structure of DNA. There are several classes based on size and function of ncRNAs including microRNAs, long RNAs (lncRNAs) and small RNAs. Small RNAs are thought to regulate gene expression as well as defensively suppress transposons. Of the numerous types of small RNAs, microRNAs (miRNAs) are thought to be most important in setting mRNA expression levels in eukaryotes through post-transcriptional modulation [16]. Generally, miRNAs are thought to bind to untranslated regions on the 3’ strand of target mRNAs, this represses the production of protein via mRNA destabilisation and silencing of translation [17]. In addition to this, miRNAs and other small RNAs target epigenetic machinery and transcription factors. Transcription factors (TF) and miRNAs can work in pairs to regulate gene expression and suppress functionally related proteins [18].

Evidence of Trans-generational Inheritance of Epigenetic Modifications

As this review has already defined epigenetics as the study of heritable changes that alter gene expression but not DNA structure, it should be elucidated that this means the epigenetic features persist through both mitosis and meiosis [19]. This systematic review will only focus on the trans-generational inheritance of epigenetic changes caused by the environment. Whilst some epigenetic markers such as DNA methylation have been established as heritable throughout these processes, others such as histone modifications have been disputed but for the purposes of this review will still be considered.

There are, generally speaking, two fundamental types of cells in the body. Somatic cells, which replicate by mitosis, and germ cells, which can replicate by both mitosis and meiosis. Germ cells produce gametes (sperm and ovum) that are able to fuse during fertilisation to produce offspring. Mitotic inheritance of epigenetic marks is common, for example the DNMTs responsible for the copying of methylation of cytosines from mother strands to daughter strands. The inheritance of histone modifications appears much more complex. Whilst epigenetic information provided by histone modifications should be inherited between mitotic cell generations in order to maintain cell fate [20], such modifications (like H3K27me3 and H3K9me3) aren’t precisely transmitted from the parent histones to the new histones. On both daughter strands, histones are randomly distributed in clusters [21]. Whilst mechanisms of histone transmission between mitotic states are unknown, since daughter cells inherit some histones from parental cells, it is worth considering histone modifications as mitotically heritable.

Meiosis is a type of cell division that allows gametogenesis or the production of gametes. During gametogenesis the germ cells are subjected to various chromatin remodelling events and eventually their chromatin is condensed, becoming transcriptionally inert [22]. This process is mediated by histones and transition proteins. Small proteins called protamines replace histones in the late haploid stage of spermatogenesis [23]. However, recent evidence suggests that in approximately 1-10% of the genome in mammals tested, histones within nucleosomes are retained throughout spermatogenesis [24], though precise structural nature of these nucleosomes is currently unknown.

After fertilisation, it is necessary that the maternal and paternal chromatin is decompressed before transcription and other somatic processes can take place. The protamines of the paternal chromatin are exchanged for histones and genome wide demethylation occurs, causing further chromatin remodelling. However, the epigenetic marks associated with imprinted genes remain intact. These processes happen rapidly after fertilisation [22]. Generally most epigenetic modifications are thought to be erased through these mechanisms between generations but DNA methylation modifications are sometimes found to be heritable. Histones maintained in sperm may persist in the zygote at least to the one-cell embryo stage [25], for example, H3K27me3 retained in sperm are thought to persist causing repression in the pre-implantation embryo. Furthermore, H3K4me3 in human sperm correlates with corresponding early gene expression in the embryo [25]. Thus such persistence of epigenetic marks could have major functional consequences in developing embryos.

Epigenetic Modifications Implicated in Behavioural, Psychiatric and Psychological Phenotypes

Firstly, various epigenetic modifications have been found to be associated with psychiatric illness and particular behavioural traits. In the study of disease discordant monozygotic twins, there is differential methylation on specific CpG sites between twins affected with bipolar, schizophrenia and those with psychosis symptoms and the unaffected groups [26]. In addition to this, there was decreased methylation in BDNF promotor exon IV and IX in the frontal cortex of schizophrenic participants compared to controls [27]. Increasingly, evidence suggests lncRNAs are involved in disease pathology including mental disorders and autism [28]. Whilst the association of epigenetic modifications and a disease phenotype may not be causal in these cases, they at least indicate that altered epigenetic regulation may play a role in disease pathology and could be key early warning indicators of disease risk or treatment targets.

The onset and development of psychiatric illness is often attributed to being induced by an interaction of environmental with an underlying genotype, as suggested by the Diathesis Stress Model [29]. Environmental and genetic factors influence the epigenome in the brain – which could potentially be the gene-environment interaction mechanism responsible for the expression of psychiatric phenotypes [30]. Various risk factors for psychiatric disease have been found to induce epigenetic changes as well as behaviours and symptoms associated with such disorders, thus implying epigenetic modifications may have a role in the onset of psychiatric conditions. Chronic social defeat stress (CSDS), a risk factor for behavioural mal-adaptations in rodent models, has been found to decrease H3K14 acetylation and histone deacetylases (HDAC2 in this case) in the nucleus accumbens, which are features also seen in depressed humans [31]. CSDS paradigms have also been found to transiently enhance then repress global H3K14 acetylation in the hippocampus, modifications which can then be reversed by chronic imipramine treatment [32], indicating histone H3 acetylation may be a target of such antidepressant medicines and possibly involved in the symptoms are depression.

Furthermore, such environmentally induced modifications of phenotype and epigenotype have been found to persist over several generations. For example, poor maternal care persists through several generations with altered DNA methylation profiles in offspring [33]. A dysmasculinized phenotype was observed in two generations of male mice after chronic prenatal stress in the first generation and this correlated with altered miRNA profiles in the second generation offspring [34]. In addition to this, altered histone PTMs have been implicated in the inheritance of a cocaine-resistant phenotype [35]. A causal link between epigenetic modifications and the production of a particular disease phenotype is plausible [36]. Injected sperm RNAs of males that had been exposed to unpredictable maternal stress and separation (MSUS), into wild-type fertilised oocytes. Behaviour observed in MSUS males were then observed in the genetically unrelated MSUS-sperm injected offspring. This has led to the development of the research question of whether transgenerational epigenetic inheritance induced by environmental factors can produce behavioural, psychiatric and psychological phenotypes in unexposed offspring.

Aims of the study

As research on trans-generational epigenetics is expanding rapidly, the need for a systematic review of the research already conducted is necessary. Currently, many systematic reviews on epigenetics focus on a single generation or on outcomes other than behavioural, psychological or psychiatric phenotypes. However, trans-generational epigenetic inheritance potentially has an important role on the transmission of psychiatric disease and will surely be the focus of many future studies. This is why a systematic review should be conducted, to summarise current methodologies, theories and understandings – the results of such a systematic review can be utilised by future researchers to create methodologically and theoretically sound experiments.

This systematic review aims to screen current literature for relevant studies that examine the role of epigenetic inheritance in the expression of behavioural, psychological or psychiatric phenotypes. This aim can be split into several components. Firstly, can epigenetic modifications induced by external stimuli be trans-generationally inherited by progeny? Secondly, if the epigenetic modification is trans-generationally inherited by the offspring, is a change in behavioural, psychological or psychiatric phenotype also present? Subsequently, this review aims to assess whether this change in phenotype can possibly be associated with the epigenetic change and through what mechanisms this is possible. Finally, the impact of the findings will be discussed in addition to a justification for further research into the field of trans-generational epigenetics. Thus, to produce a systematic review that summarises the information from these papers in a comprehensive manner, it is necessary to assess the methodological quality of the studies featured in this review. In addition to this, the conclusions of the authors featured in these papers need to be examined.

This systematic review aims to assess the hypothesis that environmentally induced epigenetic modifications can be inherited by offspring and expressed as a behavioural, psychological or psychiatric phenotype.

Methods

PICOS Structure in the Methodological Design of This Systematic Review

Whilst the research question covers a fairly novel subject area for systematic reviews, the methodology will still follow a traditional structure as recommended by the Cochrane Collaboration [37]. The research question was designed and constructed through the PICOS structure [38]. Below, the components of the PICOS structure are listed and an explanation of how each aspect pertains to the design of this systematic review is given.

Types of Participants

There was no particular participant group defined for this systematic review as the study of trans-generational epigenetics is a relatively new field, subsequently, it was expected that very few papers would be identified and narrowing participant group criteria would produce an extremely small yield. Thus it was also assumed that most papers found would feature animal trials rather than human-patient groups. The main participant criteria of this systematic review was that papers should feature at least two generations of participants for the measures of epigenetic modifications – which is essential for the study of the transgenerational aspect of the research question. In addition to this, this review did not accept papers studying unicellular populations or plant populations.

Types of Intervention

Due to predicted restriction of study yield, very few intervention requirements were stipulated. Most essentially, the experimental intervention had to be implemented with the intent of producing alterations to some aspect of the epigenome in participants. In addition to this, as this paper is interested in trans-generational inheritance, the intervention should be implemented in the parental generation i.e. there should be at least one offspring group not directly exposed to the intervention (not including controls).

Types of Comparison Group

No particular comparison group was outlined but it was assumed that most studies would have a control group of some description (i.e. a group of participants that did not receive the intervention expected to produce an epigenetic change and behavioural change). Studies without an appropriate comparison group will not be considered for further investigation. The conditions under which the comparison group was raised/kept should be the same for at least one epigenetic assay and one behavioural assay within each paper, so meaningful associations can be made. However, this is not a requirement and such paradigms will be discussed in the results section.

Types of Outcome Measures

There are two main outcome types that were required for a paper to be considered, the first being a heritable epigenetic change caused by the intervention. For example, a stressor in the first generation causes an epigenomic alteration - the study must then investigate for similar epigenetic modifications in at least one subsequent generation. Epigenetic modifications on imprinted genes are not of interest for this systematic review.

The second outcome type is the measure of phenotype change – the phenotypes of interest being either behavioural, psychological or psychiatric. A neuroanatomical, cognitive or morphological change by itself is not sufficient for inclusion, it must be linked to behaviour or to a psychological/psychiatric disorder.

Types of Studies

Multiple types of study were considered for this review. The study of trans-generational epigenetics has only been studied relatively recently due to methodological restrictions prior to this. As a result, many of the studies of interest are expected to be more in the primary stages of research using experimental design more familiar to psychological research than epidemiological research. Experimental rather than observational studies are more likely to be featured in this systematic review; limited theoretical knowledge would probably render large scale epigenetic assays hard to administer and evaluation of results of such studies would be quite difficult to interpret due to a large number of confounders that arise in such non-controlled conditions.

As most of the papers identified are likely to use experimental laboratory-research design, it is also likely that most of the studies will use one-way designs to study the direction of epigenetic and behavioural changes e.g. performance of the experimental group vs the control group on epigenetic/behavioural outcome. However, many experiments will look at other factors affecting the outcome, in addition to effect of experimental group, such as gender; these will be factorial designs. It is also probable that some paradigms will use a within-subjects design as participants will be tested on multiple occasions or on multiple assays.

Search Strategy

Though the concept of epigenetics was first mentioned in the 1940s [39] studies on epigenetics have been limited by the equipment and understanding until recent years, for example, bisulfite sequencing was implemented for chemical analysis of DNA methylation in 1994 [40]. Thus, it was deemed unnecessary to carry out a hand search of data as it is unlikely that any relevant papers would not have been written in electronic form or already converted to electronic form.

Electronic Searches

The outcome criteria of this systematic review are quite broad as initial PubMed searches did not yield many studies of relevance. As the field of trans-generational epigenetics has only recently expanded, not all studies have been fully indexed to the relevant area – i.e. not all papers referring to epigenetics will be retrieved if the term “epigenetic” is used when searching in a database (this was discovered during preliminary searches on PubMed). As a consequence of this, in this review, multiple synonyms have been used for each aspect of the research question. Every possible search term was also expanded upon within each individual database, so that as many papers as possible are identified. Several preliminary searches were conducted in PubMed, which established that different terminology or synonyms often produced additional hits with possible relevance for this review. These hits would otherwise not have been found. Whilst constructing the search term procedure, known relevant studies, such as [41] were used as a reference point for searching i.e. a search procedure should be able to retrieve this study if it was in the database, if it wasn’t able to then further relevant search terms were required.

Each synonym was entered into each individual database (using Ovid) and then expanded and mapped to subject headings to find further useful categories, this produced 3 different search term procedures. The 3 search term procedures were then reintegrated so one unified search procedure was produced. Search terms and subject headings that were valid in one or two databases invalid in the others were converted to text terms so scope was not reduced.

Search Term Procedure

The following databases were used in the identification of relevant papers:

A. Ovid Medline (R) Without Revisions (1996-Present)

B. Embase

C. Psych Info

The following search terms were used in each database for the identification of papers:

1. Epigenetic/

2. DNA acetylation.mp.

3. DNA methylation.mp.

4. DNA phosphorylation.mp.

5. Epigen*.mp

6. exp epigenetics/

7. Histone post translational modification*.mp. or chromatin. mp.

8. Histone PTM*.mp. or histone modification.mp.

9. MicroRNA*.mp.

10. miRNA*.mp. or small interfering RNA.mp.

11. ncRNA*.mp.untranslated RNA.mp. or long untranslated RNA.mp. or non coding RNA*.mp.

12. Epigenetic Inheritance.mp.

13. multigener*.mp.

14. multigenerational.mp.

15. Soft Inheritance.mp.

16. transgener*.mp.

17. trans-generational.mp. or Inheritance Patterns.mp.

18. ancest*.mp.

19. descend*.mp.

20. Hereditary.mp.

21. heritable.mp. or heritability.mp.

22. inherit* .mp.

23. progeny.mp. or Offspring.mp. or Family Members.mp. or exp Family Relations/ or Grandparents.mp. or Holocaust Survivors.mp. or Family Background.mp. or Family History. mp. or exp Mothers/ or human relation.mp. or exp family/

24. anxi*.mp. or exp anxiety disorder/

25. exp anxiety/

26. behav*.mp.

27. exp behavior/ or Mentally Ill Offenders.mp. or behaviour. mp. or exp Social Behavior/

28. brain.mp. or exp brain/

29. depress*.mp.

30. exp reactive depression/ or depression.mp. or exp endogenous depression/ or exp Depression/ or exp Treatment Resistant Depression/

31. affect*.mp. or emotion*.mp. or stress.mp. or Arousal.mp. or exp Expressed Emotion/

32. exp fear/ or exp conditioning/ or exp phobia/ or fear.mp.

33. habit*.mp.

34. exp neuroanatomy/ or neuroan*.mp.

35. neurobe*.mp.

36. exp neurobiology/ or neurobi*.mp. or exp attention deficit disorder/ or substance related disorders.mp.

37. exp neurology/ or neurolo*.mp. or exp Epilepsy/ or exp Neurology/ or exp Conversion Disorder/ or exp Neuropsychology/

38. neuron*.mp. or exp Neurons/

39. mental disease.mp. or exp posttraumatic stress disorder/ or psychiat*.mp.

40. exp child psychiatry/ or exp psychiatry/ or exp forensic psychiatry/ or psychiatry.mp. or exp Child Psychiatry/ or exp Biological Psychiatry/ or exp Psychosis/ or exp Alcohol Abuse/ or exp Drinking Behavior/ or exp Mental Health/

41. exp coping behavior/ or exp schizophrenia/ or exp cognition/ or psychol*.mp.

42. psychology.mp. or exp clinical psychology/ or exp psychology/ or exp child psychology/ or exp medical psychology/ or exp Adolescent Psychology/ or exp Trauma/ or Addiction.mp. or exp Psychopathology/ or exp Eating Disorders/ or personality.mp. or exp Motivation/ or cogniti*.mp. or exp Decision Making/ or eating*.mp.

43. 1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 or 10 or 11 12 or 13 or 14 or 15 or 16 or 17

44. 18 or 19 or 20 or 21 or 22 or 23

45. 24 or 25 or 26 or 27 or 28 or 29 or 30 or 31 or 32 or 33 or

34 or 35 or 36 or 37 or 38 or 39 or 40 or 41 or 42

46. 43 and 44 and 45 and 46

Study Yield

The abstracts and titles of the studies yielded by the search procedure were used to identify relevant papers.

A. Ovid Medline (R) Without Revisions (1996-Present):

a. 456 papers found, 109 were identified.

B. Embase:

a. 541 papers found, 124 were identified.

C. Psych Info

a. 72 papers found, 35 were identified.

At this stage review papers were included in order to ascertain whether further relevant papers could be identified from their contents that had not been found using the search procedure.

Inclusion & Exclusion Criteria

The 268 total identified papers were individually screened to ascertain whether they were relevant. The same inclusion and exclusion criteria were applied but at this stage review papers were also excluded. No further relevant results were yielded and the search procedure was thought to be sufficient. It should be noted that the phenotype related outcomes of interest needed for inclusion could be any paradigm relating to behaviour, psychology or psychiatry; these included tests of locomotive activity, sucrose consumption (as a measure of depression), etc. Memory assays and brain chemical assays were not included in this review. As the outcome criteria was so wide and may not have been explicitly mentioned in the abstracts, many studies were not excluded at the initial identification stage when abstracts were scanned. However, imprinting disorders were excluded. Epigenetic modifications of interest also excluded changes at imprinted genes.

Papers would not be considered unless at least two generations were studied. In addition to this, at least one generation must not have been exposed to the experimental manipulation. Furthermore, the study population could not be plants or unicellular. Books or reviews were excluded if they did not contain meaningful data. The inclusion and exclusion criteria are summarised in Table 1.

Inclusion Criteria	Exclusion Criteria
• Presence of trans-generational inheritance of traits i.e. across more than one generation.	• Any of the five Inclusion Criteria were not met.
• Epigenetic modification studied in 2 generations.	• If trans-generational inheritance refers to in utero effects between mothers and offspring and no subsequent generation is studied.
• Behavioural, psychiatric of psychological phenotype studied.	• If behaviours/epigenetic modifications refers to imprinting disorders.
• Appropriate control group for comparison of results of epigenetic assay and behavioural/psychological/ psychiatric assay.	• If behavioural assay refers to memory.
• At least one generation must not have direct or in utero exposure to experimental manipulation.	• If paper was from book or other medium that does not contain meaningful data.
	• If study population is not animals/humans e.g. unicellular organisms and plants.

Table 1 Inclusion and Exclusion Criteria.

	Embase	MedLine	PsychInfo
Identified	109	124	35
Inappropriate Participant Group	84	80	0
No Epigenetic Outcome of Interest	14	13	3
No Trans-generational Aspect	10	9	7
No PhenotypeOutcome of Interest	145	210	2
No Epigenetic&Trans-generational	4	4	6
No Epigenetic&Phenotype of Interest	27	36	0
No Phenotype of Interest&Trans-generational	42	50	0
No Epigenetic, Phenotype or Trans-generational	21	15	5
Book/No Data	0	0	14

Table 2: Numbers of Papers Identified and Excluded by Scanning of Abstracts.

	Number of Papers
Identified	6
Inappropriate Epigenetic Measure	19
Inappropriate Behavioural Outcome	8
No Trans-generationalStudy of EpigeneticInheritance	3
No Trans-generational Aspect / Generationis not Removed from Experimental Exposure	2
Review/No Data	68

Table 3: Numbers of Papers Identified and Excluded by Reading of Full Paper.

Results

Study Selection Process

The initial search procedure yielded 1068 papers in total. In the scanning of abstracts, 109 were found to be relevant in Embase; 124 in MedLine and 35 in PsychInfo. The articles were then retrieved and read fully.

Of the 268 papers (162 were duplicates) identified using the search procedure outlined in Section 2.7-2.9, only 106 were retrieved. Some papers could not be found using the references provided (N=17) and some were Non-English Language papers that could not be translated (N=24). From these 100 retrieved papers, six papers met all five inclusion criteria and did not meet the parameters of the exclusion criteria. Please refer to Appendix 1 for the list of excluded papers from this stage.

Description of studies

All of the 6 papers identified were heterogeneous laboratory conducted experiments. Within these the following behavioural paradigms were studied: Accelerating Rotarod (n=1); Auditory Fear Conditioning (n=1); Cocaine Self-Administration (n=1); Ethanol Consumption/Preference (n=1); Elevated Plus Maze (n=2); Forced Swim Test (n=2); Free Exploratory Paradigm (n=1); Light-Dark Box (n=1); Maternal Behaviour Assay (n=3); Olfactory Fear Potentiated Startle (OPS; n=1); Olfactory Sensitivity Assay (n=1); Open Field Paradigm (n=2); Sensorimotor Assay (n=1); Sucrose Consumption (n=1); Sucrose-Self Administration (n=1); Tail Suspension Test (n=1) & Two-Bottle Choice Assay (n=1). Please note that from now on as no psychiatric or psychological outcomes found, the altered phenotype will be referred to as a behavioural phenotype. In addition to this the following epigenetic features were examined: DNA methylation (n=3); Histone H3 Acetylation (n=2); miRNA/ncRNA levels (n=2).

The results of the data extraction of each paper featured in this systematic review can be seen in five parts. The behavioural and epigenetic assay results included manually assessed interpretations of visual data. Such interpretations were included as many authors did not include statistical results that were non-significant, it is generally good practise to publish nonsignificant results thus as much information pertaining to the direction of non-significant results was recorded for this review. Whilst some authors such as [42] often recorded results that were approaching significance as “a trend in” other authors did not do so. As this systematic review aims to provide information for other researchers, it was decided that any available information should be extracted should other experimenters wish to use similar experimental paradigms in the future.

Design of Identified Studies

As predicted, the studies yielded were not epidemiological in nature. Instead experiments conducted in laboratories were found. Whilst systematic reviews traditionally examine randomised control trial (RCT) designs and other epidemiological studies, it was still decided that a systematic review of the laboratory studies was possible as they can be viewed as the nonhuman equivalent of RCT studies (Kramer, 1998). At this stage of research into the field of epigenetics, the use of laboratory experiments is more appropriate for assaying the effects of epigenetic inheritance on behavioural phenotype as extraneous variables can be controlled, measured and their effect on the results assessed. Table 4 provides a general summary of relevant methodological design features of the identified papers.

	Dias & Ressler [41]	Finegersh & Homanics [42]	Franklin et al [43]	Gapp et al [36]	Vassoler et al [44]	Yao et al [9]
Subject Species	Mice	Mice	Mice	Mice	Rats	Rats
Breed(s) of Subject Species	M71-LacZ transgenic. C57Bl/6J.	C57Bl/6J (Sires) C57Bl/6J x Strain 129Sv/ImJ (Offspring)	C57Bl6/J	C57Bl/6J	Sprague-Dawley	Long-Evans Hooded.
Environmental Stressor Paradigm Used (Experimental Manipulation of Interest)	Olfactory Fear Conditioning (acetophenone)	Vapour EtOH exposure.	MSUS	MSUS	Self-administered cocaine infusion	Stress by Restraint & Forced Swim.
Group(s) Subjected to Environmental Stressor	F0 Males. F0 Females.	F0 Males.	F0 dams & F1 litters.	F1 Litters.	F0	F0. F1. F2.
Manipulation for Control Group	Left in Home Cage or Olfactory Fear Conditioning with Propanol.	Room air exposure.	Undisturbed dams and litters	Left in Home Cage.	Self-administered Saline infusion	Not Stressed.
Epigenetic Modification of Interest	DNA Methylation. Histone H3 Acetylation.	DNA Methylation.	DNA Methylation.	miRNA expression levels.	Histone H3 Acetylation.	miRNAs.
Area Epigenetic Modification(s) Assayed.	Sperm. MOE. Epididymis.	Sperm. mPFC. VTA.	Brain. Sperm.	Serum. Sperm. Hippocampus. Hypothalamus. Cortex.	mPFC Sperm	Brain. Placenta. Uterus.
Gene(s)/Promoters Epigenetic Modification(s) Studied On.	Olfr151 (M71). Olfr6.	Bdnf. Dlk1.	5HT1A CB1 CRFR2 MAOA MeCP2	N/A.	Bdnf exon I, IV & VI.	N/A.
Generation  Epigenetic Modifications Studies In.	F0 F1 F2	F0 F1	F1 F2	F1 F2 RNAinj	F0 F1	F0 F1 F2
Initial Inheritance through parental Germline	Both.	Paternal.	Paternal.	Paternal.	Paternal.	Maternal.
Generations Not Directly Exposed to Experimental Manipulation.	F1. F2.	F1.	F2.	F2.	F1.	F1-SN F2-SNN F2-SSN F3-SNNN F3-SSNN F3-SSSN
Gender Used in Study for Epigenetic Assays	Male.	Both.	Both.	Male.	Male.	Female.
Behavioural Assay(s)	OPS. Odour Sensitivity Assay. Elevated Plus Maze. Auditory Fear Conditioning Assay.	Accelerating Rotarod. Elevated Plus Maze. Open Field. Two-Bottle Choice.	Maternal Behaviour. Forced Swim. Sucrose Consumption. Tail Suspension Test. Free Exploratory Paradigm. Open Field.	Elevated Plus Maze. Light-Dark Box. Forced Swim.	Cocaine Self Administration. Maternal Behaviour. Sucrose Self Administration.	Maternal Post-Partum. Sensorimotor Behaviour.
Gender of Offspring Used for Behavioural Assays	Male.	Both.	Both.	Female.	Both.	Both.
Generation(s) Behavioural Assay Tested in.	F1-Ace F1-Prop F2-Ace F1-Ace-nfost F1-Ace-fost F1-Prop-nfost F1-Prop-fost	F1.	F0 F1 F2 F3	F1 F2 RNA-inj RNA-inj Offspring	F1	F0 F1 F1 F2 F3

Table 4: Summary of Identified Papers

Systematic Review of Participant Groups

Unfortunately, total participant number (n) for all experiments cannot be reported as all data pertaining to n is not available in identified papers (see Table S1, Appendix 2). Table 5 summarises all information available about participant groups, including species and breeds, reason for selecting these breeds and species and how subsequent generations were produced. Four identified papers examined mice [41-,43,36] and two examined rats [44,9].

Study	Experimental Procedure	Control Group Procedure	Groups Exposed to Experimental Procedure	Age Exposed to Experimental Procedure	Unexposed Progeny of Experimental Group
Dias & Ressler [41]	Trained to associate acetophenone (10s) presentation with co-terminating mild foot-shock (0.25-s 0.4mA). Average interval: 120s. Trained on 3 consecutive days: 5 odour presentation trials/day.	Left in home cage or conditioned with propanol.	F0 C57Bl/6J Dams F0 C57Bl/6J Sires F0 M71 Dams F0 M71 Sires	2 months old 2 months old	F1-C57-Ace(nfost) F1-C57-Ace(fost) F1-IVF F1-C57-Ace F2-C57-Ace F1-M71-Ace F1-M71IVF
Finegersh & Homanics [42]	~250μl/min vapor EtOH inhalation in custom 0.5” plexiglass vapour chambers (16”x16”x24”) from 8:00-16:00 5d/5wks at 78?F.	Placed in equivalent chamber but exposed to room air.	F0 C57Bl/6J sires	2 months old	F1-E-Sired
Franklin et al [43]	MSUS 3hr/day	Undisturbed (1 cage change/week).	F0 dams and F1 litters.	Dams not specified. Litters PND1-14	F2-MSUS F3-MSUS
Gapp et al [36]	MSUS: 3hr/day proximal separation	Undisturbed (1 cage change/week).	F0 C57Bl/6J dams & F1 litters	Dams: 2-3months Litters: PND1-14	F2-MSUS (C57Bl/6J) F1-RNAinj (B6ND2F1)
Vassoler et al [44]	Cocaine self-administration in operant chamber through indwelling silastic catheter in right jugular vein (0.25mg dose over 5s, 20s intervals). Limited to 75 infusions/2hr/d for 60d. Mean infusions at start= 3-4mg/day. Mean infusions D60=7-8mg/day.	Yoked intravenous saline injections.	F0-Sires	Not specified.	F1-Coc-Sired
Yao et al [9]	Pregnant dams GD-12-18. Restraint in plexiglass 6cm inner diameter container for 20mins (prevented turning and maintains standing) in time period 8:00-9:00. Between 16:00-17:00 forced swim in tank (45x77x50cm, water 21?C) for 5mins.	Non-Stressed, otherwise not specified.	F0-S F1-SS F2-SSS (and therefore their in utero pups)	Not specified but on GD12-18	F2-SNN F3-SNNN F3-SSNN

Table 5: Summary of Experimental Manipulation.

All four papers using mice used a C57Bl/6J strain. [42] outbred these to Strain 1295v/lmJ mice to produce hybrids. [36] also used an addition strain for IVF component of their study (where fertilised eggs of BD62F1 mice were injected with C57Bl/6J sperm RNAs). Dias and Ressler used an additional strain of mice, M71-IRES-tauLac Transgenic Mice (M71-LacZ), which were involved in a variety of behavioural and neuroanatomical experiments but not epigenetic assays. [44]) bred Sprague-Dawley male rats who were subjected to the cocaine self-administration paradigm to naïve females 3 days after the last administration of cocaine (3/18 males bred continued cocaine-self administration but offspring of these produced results no different from the other sires thus groups were combined). Finally, [9] used Long-Evans hooded rats in their experiment.

Systematic Review of Experimental Manipulation

Firstly, please note that experimental manipulation is equivalent to the environmental factor aspect of the research questions. Two studies used MSUS as the experimental manipulation [43] [36] Dias & Ressler used an olfactory fear conditioning paradigm; Finegersh & Homanics used vapour ethanol exposure; [44] exposed rats to cocaine-self administration & [9] used a stress paradigm during gestation. Only the F0 generation was exposed to the experimental manipulation in studies conducted by [41,42,44] In the MSUS studies, both F0 dams and F1 litters were exposed to the experimental manipulation. Finally, [9] had several generations exposed to the experimental manipulation: for each generation the participants were split into stressed and non-stressed groups.

Only [41] and [42] used at least one control group that had followed exactly the same experimental procedure as the experimental manipulation group. [44] used a similar stressful procedure in controls (intravenous injection of saline). Franklin et al (2010) and Gapp et al (2014) merely left the control groups in the home cage with 1 cage change per week (though given the nature of MSUS alternative control group procedures are difficult). Yao et al (2014) did not specify further than “not stressed”.

[41] studied multiple different groups of unexposed progeny of F0- conditioned sires. [42] [44] studied only one unexposed group of progeny of F0-exposed sires each. [43] [36] studied 2 unexposed group of progeny of F1-exposed males. Yao et al (2014) examined 3 unexposed progeny of F1 females who were stressed in utero.

Heritable Behavioural Phenotypes

Although there are limitations associated with the tests used (see Appendix 2, Table S2), all the experimental manipulations induced a behavioural phenotype in at least one gender of offspring (Table 6). Inheritance of behavioural phenotypes was observed through the maternal, and the paternal line depending on the procedure used. Inheritance of behaviour was often sex linked, though many studies only examined behavioural transmission in one gender.

Overall Outcome of Interest	Experimental Manipulation	Paradigm Used and Results Indicate	Result	Generations Outcome Observed In Relative to Respective Control Group
Behavioural Manifestations of Early Life Stress	MSUS of F0 Dams and F1 Litters	EPM: Decreased anxiety/reduced behavioural control.	Decreased Latency to Enter Open Arm	F1-MSUS male# F2-MSUS male (Gapp Only) F2-MSUS females (Franklin Only) F3-MSUS females
			Increased Latency to Enter	F3-MSUS males (Franklin Only)
			No Differences in Locomotive Activity	F1-MSUS male# F2-MSUS male# F1-MSUS-RNAinj males (Gapp Only)
		Forced Swim: Increased depressive like symptoms.	Increased Time Floating	F1-MSUS male# F2-MSUS female (Franklin Only) F2-MSUS male# F2-MSUS-RNAinj (Gapp Only)
			Decreased time floating after desipramine treatment	F2-MSUS-Des vs. MSUS-Sal (Franklin Only)
			Increased time floating after desipramine treatment	F2-MSUS-Sal vs. F2-C-Sal (Franklin Only)
		Index of Maternal Care & Index of Absence of Care	Decreased time actively nursing on PND1-7	F0-MSUS females
			Increased time spent off nest on PND1-7	F0-MSUS females
		Light-Dark Box: Altered response to behavioural phenotype seen in two generations of males.	Increased Time in Illuminated Compartment	F1-MSUS males F2-MSUS males F1-MSUS-RNAinj males
		Open Field: Inheritance of MSUS induced F1 male phenotype by F2 and F3 females. Test showed reduced behavioural control. (Franklin Only)	Decreased Latency to Enter Unfamiliar Arm	F1-MSUS males F2-MSUS females F3-MSUS females
			No Difference in Distance Travelled	F1-MSUS males F2-MSUS females on Trial 3
			Increased Difference Travelled	F2-MSUS females on Trials 1 and 2.
		Sucrose Consumption: Showed MSUS produces anhedonia in first generation only. (Franklin Only)	Decreased Sucrose Consumption	F1-MSUS male
		TST (Franklin Only): Increased depressive like symptoms.	Decreased time immobile after acute and chronic saline treatment	F2-MSUS-Sal vs. F2-C-Sal
			Increased time immobile after acute and chronic desipramine treatment	F2-MSUS-Des vs. F2-MSUS-Sal
Behavioural Manifestations of Preterm Birth	Chronic Stress	Inclined Plane Test: Delayed Sensorimotor Development	Increased time spent in downward position (increases with every successive generation).	F1-SN F2-SNN F2-SSN F3-SNNN F3-SSNN F3-SSSN
		Maternal Post-Partum Behaviour	Decreased Tail Chasing and Rotational Behaviours	F1-SN F1-SS F2-SNN F2-SSN F2-SSS
Drug Related Behaviours	Cocaine-Self Administration in F0 males	Cocaine Self-Administration (FR1): Cocaine Resistant Phenotype	Decreased Cocaine Infusions at 0.5mg	F1-C females F1-C males F1-C-DMSO
			Decreased Cocaine Infusions at 1.0mg	F1-C males
		Cocaine Self-Administration (PR1): Cocaine Resistant Phenotype	Reduction of breakpoint at 1.0mg/g	F1-C males
		Sucrose Self-Administration (FR1): Cocaine-resistant phenotype in cocaine self-administration paradigms not confounded by learning deficits.	No Difference Between Sucrose Self-Administration	F1-Coc-Sired males vs. F1-Sal-Sired males
		Sucrose Self-Administration (PR): Cocaine-resistant phenotype in cocaine self-administration paradigms not confounded by learning deficits.	No Differences in Breakpoint of Sucrose Administration	F1-Coc-Sired males vs. F1-Sal-Sired males
		Maternal Behaviour	No Differences in Licking/Grooming	F0-Dams of Cocaine Sired Offspring
			No Differences in Time off Nest	F0-Dams of Cocaine Sired Offspring
	Ethanol Exposure in F0 males.	Accelerating Rotarod: Paternal alcohol exposure subtly increases motor enhancement after ethanol exposure.	Increased Time Spent on Rotarod on 5th Trial	F1-E-Sired-EtOHinj Males vs. F1-E-Sired-salineinj
		EPM: Paternal alcohol exposure increases anxiolytic and enhanced locomotor effects of alcohol.	Increased Time in Open Arm	F1-E males F1-E-ETOH-inj males
			Increased Proportion of Open Arm Entries	F1-E males F1-E-ETOH-inj males
			Increased Number of Total Arm Entries	F1-E-ETOH-inj males
		Open Field: Basal Activity not increased in mice found to have ethanol sensitive phenotype on other assays.	No Differences in Distance Travelled	F1-E-ETOH-inj males vs F1-E-Sired-salineinj
		Two-Bottle Choice Assay (Ethanol):	Decreased Preference at 3% Concentrations	F1-E males
			Decreased Preference at 6% Concentrations	F1-E males
			Decreased Preference at 9% Concentrations	F1-E males
			Decreased Consumption at 12% Concentrations	F1-E males
		Two-Bottle Choice Assay (Quinine and Saccharin):	No Differences in Quinine or Saccharin Preference at 0.03 and 0.06mM Concentrations	F1-E males vs F1-C males
Olfactory Sensitivity	Olfactory Conditioning to Acetophenone in F0 males.	Auditory Fear Conditioning	No Significant Difference in Acquisition, Consolidation and Extinction Retention of Aversive Memory Cue.	F1-Ace-C57 vs. F1-Home-C57
		EPM	Equal Time in Open & Closed Arms and Equal Entries into Open Arms	All Groups
		OPS	Increased OPS to acetophenone	F1-Ace-C57 F2-Ace-C57 F1-Ace-M71 F1-Ace-C57-nfost F1-Ace-C57-fost
			Increased OPS to propanol	F2-Prop-C57 F1-Prop-M71
		Olfactory Sensitivity Assay	Increased Aversion Index to 0.003% and 0.006% acetophenone concentration.	F1-Ace-C57
			Increased Aversion Index to 0.003% and 0.006% propanol concentration.	F1-Prop-C57

Table 6: Summary of Behavioural Results.

Heritable Epigenetic Modifications

Again, although there are limitations associated with the techniques used (see Appendix 2, Tables S3) overall, most studies have demonstrated that some epigenetic modifications can be transmitted to subsequent generations. However, transmission appears complex in most cases and often a direct transmission to a subsequent generation. The alteration and inheritance of miRNA/ncRNA profiles can differ in direction between generations and between neuroanatomical area (Table 7).

Experimental Manipulation	Area assayed	Generation Change Observed In	miRNA
MSUS	Frontal Cortices	F0-MSUS male	miR-24-1-5p miR-138-3p miR-145-3p miR-375
	Hippocampus	F1-MSUS male	miR-375-3p miR-466c-5p
		F2-MSUS male	miR-375-3p miR-375-5p miR-466c-5p
		F1-RNA-inj	miR-375-5p
	Hypothalamus	F1-MSUS male	miR-375-5p
	Sperm	F1-MSUS male	miR-200b-3p miR-375-5p miR-466c-5p miR-672-5p
		F2-MSUS male	miR-375-3p
	Serum	F1-MSUS male	miR-375-3p miR-375-5p miR-466c-5p miR-672-5p
		F2-MSUS male	miR-466c-5p
Stress GD12-18	Frontal Cortices	F1-SS	miR-23b miR-200c
		F2-SSS	miR-96 miR-141 miR-182 miR-183 miR-200a miR-200b miR-429 miR-451
	Placenta	F2-SNN	miR-181a
		F2-SSS	miR-181a
	Uterus	F1-SS	miR-200b miR-429
		F2-SSS	miR-200b miR-429

Please note that green refers to increased levels and blue refers to decreased levels of miRNAs.

Table 7: Summary of miRNA and ncRNA Results

Experimental manipulations modified DNA methylation levels at various loci in both directions (Table 8). Altered DNA methylation levels in specific areas tend to be inherited in the same direction (i.e. either increased or decreased) as they are found in the parental generation.

	Experimental Manipulation	Direction of Change Compared to Controls	Location of Analysis	Generation and Gender
Bdnf Exon IXa	EtOHExposure	Decrease	Sperm	F0 male
			VTA	F1 male F1 female
CB1	MSUS	Increased	Sperm	F1 male
			Brain	F2 female
CRFR2	MSUS	Decreased	Sperm	F1 male F2 male
			Brain	F2 female
IG DMR of Dlk1	MSUS	Decreased	Sperm	F0 male
MeCP2	MSUS	Increased	Sperm	F1 males F2 males
			Brain	F2 females
Olfr151	Conditioning to Acetophenone	Decreased	Sperm	F0 male F1 male
		No Difference	MOE	F1 male F2 male

Table 8: Summary of DNA Methylation Results

[44] found significantly increased AcH3 associated with Bdnf promoter IV in the medial prefrontal cortex (mPFC) of F1 males. Furthermore increased AcH3 levels were found in the testes of F0 males and specifically (though global levels were not increased) at Bdnf promoters I, IV and VI in F0 sperm. This indicates a possible mode of transmission of Histone H3 acetylation (AcH3) through the paternal germline should histones persist through spermatogenesis (Table 9).

Experimental Manipulation	Direction of Changed of Histone PTM	Area Assayed	Gene PromoterAssayed	Histone PTM Type	Generation Studied In
Olfactory Fear Conditioning	Decrease	Epidydymis	N/A	Histone H3 Acetylation	F0 male
				Histone H3 Methylation	F0 male
Cocaine Self-Administration	Increase	mPFC	Bdnf Promoter IV	Histone H3 Acetylation	F1 male
		Sperm	Bdnf Promoters I, IV & VI	Histone H3 Acetylation	F0 male
		Testes	N/A.	Histone H3 Acetylation	F0 males

Table 9: Summary of Histone PTM Results

Discussion

This systematic review summarised all data from six papers examining trans-generational inheritance of epigenetic marks and their association with behavioural outcomes (Table 10). Specifically we focused on the role of environmentally induced modifications (i.e. those changes caused by experimental manipulation) in the production of altered behavioural phenotype. Despite much discussion in the literature, original studies were limited and most studies included are in their early stages of methodological design, thus, interpretations of the results extracted are mainly associative. Furthermore, it is difficult at this point in time to extend any findings of experimental research from laboratories to humans. For example, [44] acknowledge that their findings of a cocaine resistant phenotype varies considerably from evidence of cocaine-related behaviours found in humans, where addiction is found to be heritable (Kendler et al, 2003). Currently, there are too many confounding variables in human life to be able to draw parallels from laboratory experiments. However, some promising evidence has been provided by the identified studies. Whilst the evidence is not conclusive proof of the trans-generational inheritance of environmentally induced epigenetic modifications producing altered behavioural phenotypes, it at least provides an evidence-based foundation for future studies to expand upon.

Authors	Experimental Manipulation	Behavioural Phenotype	Associated Epigenetic Mechanisms	Evidence Supporting Epigenetic Contribution to Expression of Behaviour	Evidence Confounding/Negating Epigenetic Contribution to Expression of Behaviour
Dias & Ressler [41]	F0 odour fear conditioning to acetophenone.	Olfactory sensitivity to F0 conditioned odour.	Decreased DNA methylation of Olfr151	OIfr151 (M71) responds to acetophenone. No methylation of Olfr6 which is unresponsive to acetophenone. Hypomethylation of sperm inherited from F0 to F1. Transmission of behavioural phenotype through the paternal line and in F2 and cross-fostered generation suggest biological transmission.	Mechanism for inducing methylation change unknown. Lack of hypomethylation in MOE. Lack of alterations to histone PTMs in sperm. Mechanism for inheriting epigenetic change is unknown.
Finegersh & Homanics [42]	F0 sire exposure to ethanol.	Decreased ethanol preference/consumption and increased sensitivity to ethanol induced anxiolysis and locomotor enhancement.	Decreased DNA methylation of Bdnf exon IXa promoter.	Bdnf regulates drinking behaviour. Bdnf mRNA levels in VTA and on sperm increased: corresponds to hypomethylation. Reduced alcohol consumption seen.	Females have hypomethylation but no changes to phenotype.
Franklin et al [43]	MSUS with F0 dams and F1 litters	Non-specific altered behavioural phenotype on various tests.	DNA methylation	Altered methylation at CB1, CRFR2 and MeCP2 in F1 sperm and female brain. CRFR2 and MeCP2 in F2 sperm.	Mechanism for inheriting epigenetic change is unknown. Normal methylation profiles in sperm of F2 males and all areas tested in F3 males but still expressed anxiety related phenotype.
Gapp et al [36]	MSUS with F0 dams and F1 litters	Non-specific altered behavioural phenotype on various tests.	sncRNAs	Injection of MSUS-sperm RNAs into oocytes produces behavioural phenotype.	Normal miRNA profiles in sperm of F2 males and all areas tested in F3 males but still expression of behavioural phenotype.
Vassoler et al [44]	F0 sire cocaine-self administration	Cocaine-resistant phenotype	AcH3	Increased association of AcH3 with Bdnf exon IV in both mPFC of offspring and sperm of sires. Bdnf exon I and VI also more acetylated in sperm of sires.
Yao et al [9]	Successive generations exposed to restraint and forced swim stress on GD 12-18	Decreased postpartum maternal behaviours and increased sensorimotor deficits. Both of which were increasingly different from controls with each successive generation.	miRNAs	Altered behavioural phenotype and altered miRNA profiles.	Confounding due to maternal behaviour and in utero effects. Only one group removed from in utero effects had miRNA profiles measured and this was only in one area.

Table 10: Summary of All Results Providing Evidence or Negating that Environmentally Induced Epigenetic Modifications can be inherited and Contribute to the Expression of a Behavioural Phenotype in Offspring

Of all the papers identified, [36] provides the most convincing support for the hypothesis that environmentally induced epigenetic modifications can be passed to subsequent generations and expressed as behavioural phenotypes. Their experimental procedure provides the only direct example of an epigenetic induced change to behaviour. Firstly, the authors found an altered non-specific behavioural phenotype was present in successive generations (this includes measures of altered responses to aversive, stress inducing and anxiogenic stimuli), the second generation of which were not directly exposed to the environmental stressor (in this case MSUS). In addition to this, altered miRNAs levels were found in neural areas associated with stress response in two successive generations of mice. One of the altered miRNAs (miR-375) and its predicted target catenin β1 (Ctnnb) may be particularly implicated in the altered phenotype, as both are associated in the corticotropin-releasing factor signalling pathways. These pathways are involved in stress response. They also observed modified levels of both miR-375 and Ctnnb1 in the hippocampus – an area directly involved in stress response. A possible causal link between MSUS altered miRNAs and the expression of this modified phenotype can be supported by their findings that show an injection of RNAs from F0-MSUS sperm into naïve fertilised oocytes produces the same behavioural phenotype in the adult progeny as that observed in F0 males. The MSUS RNA-injected test group, also demonstrated similar metabolic phenotypes to the MSUS mice and upregulation of miR-375 in the hippocampus. It is possible that other mechanisms also mediate the transmission of this behavioural phenotype. There was no difference, compared to controls, of sncRNA levels found in F2-MSUS sperm and serum. Furthermore, normal sncRNA profiles were found in F3 progeny. As the miRNA profile did not differ from controls, the behaviour of the F3-MSUS offspring should not be altered. However, F3 progeny still exhibit the same behaviour found in the F1 and F2-MSUS generation. The authors suggest miRNAs alter other epigenetic aspects i.e. DNA methylation, and these are responsible for further transmission. Whilst posited, these alternative modes of inheritance have not been further explored.

[44] provide some interesting evidence of an environmentally induced (cocaine exposure) epigenetic modification (increased AcH3) that is expressed in the sperm and testes of F0 males and subsequently found in functionally relevant areas of the brain in offspring (mPFC). The AcH3 increase was found on gene promoters that correspond to the observed cocaine-resistant phenotype found both in the F0 and F1 male test subjects. Increased AcH3 was associated specifically to Bdnf promoters (I, VI and VI), which would increase BDNF expression (AcH3 is generally thought of as facilitative to transcription). Furthermore, the authors found that decreased BDNF activity at the TrkB receptor reversed the observed cocaine-resistant phenotype in cocaine-exposed offspring, implicating a causal role of BDNF. Whilst this evidence seems fairly convincing, the data still lacks a clear mechanism for which the histone modifications could be inherited as histones are mostly replaced by protoamines during spermatogenesis. Thus such results may not be as a common occurrence as it is thought. Only 1-10% of histones persist throughout spermatogenesis [24] It is unclear whether epigenetic modifications persist in both genders as only male profiles were tested – the inheritance of epigenotype and expression of phenotype may be much more complex than the data presented in this paper would first imply. However, the results are very much worth considering for further research as potentially this paradigm could easily be used whilst controlling for confounders with multifactorial analysis.

Similarly, [42] also found a male-only behavioural phenotype. Ethanol exposure induced hypomethylation of Bdnf exon IXa in F0 sperm, which provides a possible mechanism of inheritance of the epigenetic modification and therefore possible transmission of phenotype in offspring. In addition to this, Bdnf exon IXa transcripts were found to be upregulated in functionally relevant areas of the brain (the VTA). As Bdnf is thought to be a regulator of drinking behaviour, this supports the implications that ethanol induced hypomethylation may contribute to the behavioural phenotype observed [33]. Interestingly, hypomethylation of the Bdnf IXa promoter is found in both female and male offspring of ethanol exposed mice. This result can be interpreted in two ways. Firstly, hormonal or other biological differences between males and females may interact with epigenetic modifications to produce either no change (in the case of females) or the change observed in male offspring. Secondly, the hypomethylation observed may have little or no influence on the behavioural phenotype observed. Unfortunately, the evidence linking hypomethylation to the behaviour is still only speculative and thus interpreting this complex sex-dependent transmission of behaviour is particularly difficult.

[41] also begin to provide convincing evidence of a possibly causative role of environmentally induced epigenetic modifications in the expression of an altered behavioural phenotype. An odour-sensitive behavioural phenotype specific to the F0 conditioned odour was observed in two subsequent generations, through both the maternal and paternal line and also in offspring that were cross-fostered by non-exposed mothers. These results taken together suggest biological transmission of behaviour rather than behavioural transmission. In addition to this, in these generations and an IVF-derived generation (sperm from acetophenone conditioned males used to fertilise oocyte of naïve females) neuroanatomical changes corresponded accordingly to this odour sensitive phenotype – which would be the biological basis possibly responsible for the observed behavioural phenotype. Bisulfite sequencing also indicated hypomethylation in sperm at Olfr151. As hypomethylation was located on sperm, this indicates a mechanism that changes could be inherited from F0 sires to F1 offspring and then to F2 offspring. Hypomethylation of Olfr151 theoretically corresponds to increased Olfr151 expression that was observed in the brain. However, there was no difference in methylation status of Olfr151 in the MOE of F1 and F2 animals. Like [36] [41] attribute the lack of expected epigenetic modification in the main olfactory epithelia (MOE) to alternative modes of epigenetic inheritance. This was a reasonable assumption as other types of epigenetic inheritance have been found to alter olfactory receptor loci in the MOE [34]. Based on research suggesting a relationship between DNA methylation and histone modifications [6] the authors investigated histone H3 acetylation and methylation but ChIP revealed no differences on either measure in sperm chromatin – which the authors suggest could be due to incorrect immunoprecipitation. As a result, a mechanism is needed that explains how hypomethylation produces the neuroanatomical and subsequent behavioural changes before conclusions can be drawn on this data.

Despite a lack of hypomethylation of the Olfr151 locus in the MOE and no evidence of histone modification mediated inheritance, [41]still imply the observed neuroanatomical and behavioural changes may be associated with epigenetic mechanisms. They suggest that ncRNA inheritance may be mediating the transmission of odour-sensitivity, citing [45] who found ncRNA inheritance was observed in the Kit locus. Whilst the evidence found in their paper is quite convincing, there are several components missing in order to conclusively say that an environmentally induced epigenetic modification has produced a behavioural phenotype in offspring. Firstly, a mechanism for odorants to produce an epigenetic change is required. [41] suggest blood-borne odorants may activate receptors in sperm due to the fear conditioning protocol [46] but there is no conclusive evidence yet.

[43] perhaps provide the most difficult results to interpret. Their behavioural assays can only be taken as suggestive of the inheritance of a “non-specific behavioural phenotype”. Not all results are provided and this makes it hard to comment on the specific inheritance of a phenotype. For example, F1 female behavioural data is missing but sometimes referred to and statistical analysis of performance on a tail suspension test (TST) by F1 males is also absent. From the data available, it appears alterations of behaviour (in all tests) are generally observed in F1 MSUS males and their F2 female but not male progeny. The F3 MSUS generation males but not females showed inheritance of a depressive-like phenotype seen in their F1-MSUS grandsires which their F2-MSUS sires do not exhibit. In contrast F3-MSUS females but not males tend to show similar performance on locomotor assays to their F1-MSUS grandsires. Similarly, epigenetic assays show various mixed results. MeCP2 hypermethylation and CRFR2 hypomethylation is seen in F1 and F2-MSUS sperm and in F2- MSUS female brains. CB1 hypermethylation is observed in F1- MSUS sperm and F2-MSUS female brains but not in F2-MSUS sperm. These candidate genes are quite promising examples of areas that might have a role in inducing the observed behavioural phenotype as they are commonly associated with stress response, depressive and affective disorders [47,48,49,50,51]. However, such epigenetic contributions to behaviour are very complex and there are many variations in possible phenotype that they could potentially produce. In order to better understand the role of altered DNA methylation of these candidate genes, an assay of the brain of F1 and F2-MSUS males and F3 offspring would have been prudent to see if this could clarify the complex sexdependent results. The authors suggest alternative epigenetic modifications such as ncRNAs could be responsible but no additional assays were conducted. Currently, the most conclusive thing that can be said of such data is there is a general association between the MSUS induced behavioural phenotype and altered DNA methylation profiles.

The results from [9] do not contribute much to the argument. Unfortunately, it is hard to distinguish any evidence of causality as much of their data is subject to confounding. Only the F2-SNN (stress, non-stress, non-stress), F3-SNNN (stress, non-stress, non-stress, non-stress), and F3-SSNN (stress, stress, non-stress, non-stress) generations would have not been directly affected by the experimental manipulation (to avoid in utero confounding F1-SN, F1-SS, F2-SSN, F2-SSS and F3-SSSN were excluded from consideration in this systematic review). Of these experimental groups, miRNA data was only taken for F2-SNN and this was only in the placenta. Thus, it is still unclear whether miRNAs could contribute to the increasingly distinctive altered phenotype observed in successive generations of stressed mice. Certainly at least, this study contributes results that are useful for future research to build upon but the design of the study does not allow for conclusions pertaining to this systematic review’s research question to be drawn.

Conclusion

This study aimed to examine the role of trans-generational epigenetic inheritance in the expression of psychiatric, psychological and behavioural phenotypes in progeny. A key take home message is that currently there is a lot of hype and overlyzealous attribution of the role of epigenetics in contributing to various phenomena [52-84]. This is clearly evidenced by the fact that of the 100 retrieved papers in this systematic review, only six original articles were identified, with 68 being review articles and/or containing no original data.

However, taking all the results from the identified papers together, this systematic review has found strong associations between the inheritance of environmentally induced epigenetic modifications and the presence of altered behavioural phenotypes in offspring of experimentally manipulated subjects. Many of the neuroanatomical, physiological and behavioural outcomes also provide support for an epigenetic causality or contribution. Unfortunately, confounding is a major problem in many papers and thus definitive conclusions should not be drawn. Moreover, experimental research currently lacks clear mechanisms illuminating how environmental stressors alter the epigenotype and how this altered epigenotype can be inherited. Nevertheless, the results of this systematic review begin to provide a basis on which future research can build.

Acknowledgements

The authors should like to thank Paul Buckland and Holly Thomas for help with analyses, and Jan Chatting for support throughout this study.

Funding

This work was funded by the Medical Research Council (MRC) Centre for Neuropsychiatric Genetics and Genomics (G0801418)

Competing and Conflicting Interests

The authors declare they have no competing or conflicting interests.

References

1. Wu, C. & Morris, J.R. (2001). Genes, genetics, and epigenetics: a correspondence. Science 239: 1103-1105.
2. Phillips, T. (2008). The Role of Methylation in Gene Expression. Nature Education1-116.
3. Luger, KMader, A.W, Richmond, RK, Sargent, DF. & Richmond, TJ (1997)Crystal structure of nucleosome core particle 2.8 A resolution. Nature 389: 251-260.
4. Li, Q. & Zhang, Z. (2012). Linking DNA replication to heterochromatin silencing and epigenetic inheritance. Acta Biochimica et Biophysica Sinica, 44: 3-13.
5. Tost, J. DNA methylation: an introduction to the biology and disease-associated changes of a promising biomarker. Methods in Molecular Biology 507: 3-20.
6. Bergman, Y. & Cedar, H. (2013). DNA methylation dynamics in health and disease. Nature Structural and Molecular Biology 20: 274-281.
7. Suzuki, M.M.  & Bird, A. (2008). DNA methylation lanscapes: provovative insights from epigenomics. Nature Reviews Genetics 9: 465-476.
8. Kohli, R.M. & Zhang, Y. (2013) TET enzymes, TDG and the dynamics of DNA demethylation. Nature 502: 472-479.
9. Yao, Y., Robinson, A., Zucchi, F., Robbins, J., Babenko, O., & Kovalchuk, O. et al. (2014). Ancestral exposure to stress epigenetically programs preterm birth risk and adverse maternal and newborn outcomes. BMC Medicine1: 121
10. Jenuwein, T. & Allis, CD (2001). Translating the histone code. Science 293: 1074-1080.
11. Greer, E.L. & Shi, Y. (2012). Histone methylation: a dynamic mark in health, disease and inheritance. Nature Reviews Genetics 13: 342-357.
12. Kouzarides, T. & Berger, S.L. (2007). In Allis, C.D., Jenuwein, T. & Reinberg, D. Epigenetics. 81-100. New York: Cold Spring Harbor Laboratory Press.
13. Bannister, A.J. & Kouzarides, T. (2011). Regulation of chromatin by histone modifications. Cell Research, 21: 381-395.
14. Wang ,Z., Zang, C., Cui, K., Schones, D.E., Barski, A., Peng, W. & Zhao, K. (2009). Genome-wide mapping of HATs and HDACs reveals distinct functions in active and inactive genes. Cell 138: 1019–1103.
15. Cedar, H. & Bergman, Y. (2009). Linking DNA methylation and histone modification: patterns and paradigm. Nature Reviews Genetics 10: 295-304.
16. Fabian, M.R., Sonenberg, N. & Filipowicz, W. (2010). Regulation of mRNA Translation and Stability by microRNAs. Annual Review of Biochemistry 79:351-379.
17. Cannell, I.G., Kong, Y.W. & Bushell, M. (2008). How do microRNAs regulate gene expression? Biochemical Society Transactions 36: 1224-1231.
18. Zhou, X., Wang, G., Sutoh, K., Zhu, J.K. & Zhang, W. (2008) Identification of cold-inducible microRNAs in plants by transcriptome analysis. Biochimica et Biophysica Acta, 1779: 780–788.
19. Fineberg, A.P. & Tycko, B. (2004). The history of cancer epigenetics. Nature Reviews Cancer 4: 143-153.
20. Margueron, R. & Reinberg, D. (2010). Chromatin structure and the inheritance of epigenetic information. Nature Reviews Genetics 11: 285-296.
21. Annunziato, A.T. (2012). Assembling Chromatin: the Long and Winding Road. Biochimica et Biophysica Acta 1819:196-210.
22. Ooi, S.L. & Henikoff, S. (2007). Germline histone dynamics and epigenetics. Current Opinion in Cell Biology 19: 257-265.
23. Churikov, D., Zalenskaya, I.A. & Zalensky, A.O. (2004). Male germline-specific histones in mouse and man. Cytogenetic and Genome Research 105: 203-214.
24. Erkek, S., Hisano, M., Liang, C.Y., Gill, M., Murr, R., Dieker, J., Schubeler, D., Vlag, J., Stadler, M.B. & Peters, A.H. (2013). Molecular determinants of nucleosomes retention at CpG-rich sequences in mouse spermatozoa. Nature Structural and Molecular Biology 20:868-875.
25. van de Werken, C., van der Heijden, G.W., Eleveld, C., Teeuwssen, M., Albert,M., Baarends, W.M., Laven, J.S.E., Peters, A.H.F.M. & Baart, E.B. (2014) Paternal heterochromatin formation in human embryos is H3K9/HP1 directed and primed by sperm-derived histone modificaitons. Nature Communications 5: 58-68.
26. Dempster, E.L., Pidsley, R., Schalkwyk. L.C., Owens, S., Georgiades, A., Kane, F., Kalidindi, S., Picchioni, M., Kravariti, E., Toulopoulou, T., Murray, R.M. & Mill, J. (2011). Disease-associated epigenetic changes in monozygotic twins discordant for schizophrenia and bipolar disorder. Human Molecular Genetics 15: 4786-4796.
27. Ikegame, T. Bundo, M. Sunaga, F. Asai, T. Nishimura, F. Yoshikawa, A. Kawamura, Y. Hibino, H. Tochigi, M. Kakiuchi, C. Sasaki, T. Kato, T. Kasai, K. Iwamoto K. (2013). DNA methylation analysis of BDNF gene promoters in peripheral blood cells of schizophrenia patients. Neuroscience Research 77: 208-214.
28. Ziats, M.N. & Rennert, O.M. (2013). Aberrant expression of long noncoding RNAs in autistic brain. Journal of Molecular Neuroscience 49: 589-593.
29. Oatley, K., Keltner, D. & Jenkins, J.M. (2006) Understanding Emotions (2nd Ed). 321-351. Oxford, UK: Blackwell Publishing.
30. Petronis, A. (2010). Epigenetics as a unifying principle in the aetiology of complex traits and diseases. Nature 465: 721-727.
31. Uchida, S., Hara, K., Kobayashi, A., Otsuki, K., Yamagata, H., Hobara, T., Suzuki, Miyata, N. & Watanabe, Y. (2011). Neuron 69: 359-372.
32. Cocington, H.E., Maze, I., Sun, H., Bomze, H.M., DeMaio, K., Wu, E.Y., Dietz, D.M., Lobo, M.K., Ghose, S., Mouzon, E., Neve, R.L., Tamminga, E.J. & Nestler, E.J. (2011). A role for repressive histone methylation in cocaine-induced vulnerability to stress. Neuron 71: 656-670.
33. Logrip, M.L., Janak, P.H. & Ron, D. (2009). Escalating ethanol intake is associated with altered corticostriatal BDNF expression. Journal Neurochemistry 109: 1459-1468.
34. Magklara, A., Yen, A., Colguitt, B.M., Clowney, E.J., Allen, W., Markenscoff-Papadimitriou, E., Evans, Z.A., Kheradpour, P., Mountufaris, G., Carey, C., Barnea, G., Kellis, M. & Lomvardas, S. (2011). An epigenetic signature for monoallelic olfactory receptor expression. Cell 145: 555-570.
35. Novikova, S.I, He, F., Bai, J., Cutrufello, N.J., Lidow, M.S. & Undieh, A.S. (2008). Maternal cocaine administration in mice altered DNA methylation and gene expression in hippocampal neurons of neonatal and prepubertal offspring, PloS One 3: e1919.
36. Gapp, K., Jawaid, A., Sarkies, P., Bohacek, J., Pelczar, P., & Prados, J. et al. (2014). Implication of sperm RNAs in trans-generational inheritance of the effects of early trauma in mice. Nature Neuroscience17: 667-669.
37. Higgins, J.P.T. & Green, S. (editors). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0.
38. Wright, R.W., Brand, R.A., Dunn, W. & Spindler, K.P (2007). How to write a systematic review. Clinical Orthopaedics And Related Research 455: 23-29.
39. Waddington, C.H. (1942). The epigenoyupe. Endeavour 1: 18-20.
40. Frommer, M., Paul, C.L., Harrison, J. & Clark, S.J. (1994). High sensitivity mapping of methylated cytosines. Nucleic Acids Research 22: 2990-2997.
41. Dias, B., & Ressler, K. (2013). Parental olfactory experience influences behavior and neural structure in subsequent generations. Nature Neuroscience17: 89-96.
42. Finegersh, A., & Homanics, G. (2014). Paternal Alcohol Exposure Reduces Alcohol Drinking and Increases Behavioral Sensitivity to Alcohol Selectively in Male Offspring. Plos ONE, 9.
43. Franklin, T., Russig, H., Weiss, I., Gräff, J., Linder, N., & Michalon, A. et al. (2010). Epigenetic Transmission of the Impact of Early Stress Across Generations. Biological Psychiatry, 68: 408-415.
44. Vassoler, F., White, S., Schmidt, H., Sadri-Vakili, G., & Pierce, R. (2012). Epigenetic inheritance of a cocaine-resistance phenotype. Nature Neuroscience16: 42-47.
45. Rassoluzadgan, M., Grandjean, V., Gounon, P., Vincent, S., Gillot, I. & Cuzin, F. (2006). RNA-mediated non-mendelian inheritance of an epigenetic change in the mouse. Nature 441: 469-474.
46. Vanderhaeghen, P., Schurmans, S., Vassart, G. & Parmentier, M. (1997). Specific repertoire of olfactory receptor genes in the male germ cells of several mammalian species. Genomics 39: 239-46.
47. Urdinguio, R.G., Sanchez-Mut, J.V. & Esteller, M. (2009). Epigenetic mechanisms in neurological diseases: Genes, syndromes, and therapies. Lancet Neurol 8: 1056-1072.
48. McGill, B.E., Bundle, S.F, Yaylaoglu, M.B., Carson, J.P, Thaller, C. & Zoghbi, H.Y. (2006). Enhanced anxiety and stress-induced corticosterone release are associated with increased Crh expression in a mouse model of Rett syndrome. Proceedings of the National Academy of Science USA 103: 18267-18272.
49. Bale, T.L & Vale, W.W. (2003). Increased depression-like behaviours in corticotropin-releasing factor receptor-2-deficient mice: sexually dichotomous responses. Journal of Neuroscience 23: 5295-5301.
50. Pelleymounter, M.A., Joppa, M., Ling, N. & Foster, A.C. (2002). Pharmacological evidence supporting a role for central corticotropin-releasing factor(2) receptors in behavioural, but not endocrine, response to environmental stress. Journal Pharmacology and Experimental Therapeutics 302: 145-152.
51. Pelleymounter, M.A., Joppa, M., Ling, N. & Foster, A.C. (2004). Behavioral and neuroendocrine effects of the selective CRF2 receptor agonists urocortin II and urocortin III, Peptides 25: 659-666.
52. Isles, A. R. (2015). Neural and Behavioral Epigenetics; What it is, and What is Hype. Genes, Brain and Behavior 14: 64-72.
53. Charney, D.S., Nestler, E.J., Sklar, P. & Buxbaum, JD. (2013). Neurobiology of Mental Illness. Oxford: Oxford University Press.
54. Allis, C.D., Jenuwein, T. & Reinberg, D. (2007). Epigenetics. 23-62. New York: Cold Spring Harbor Laboratory Press.
55. Anagnostaras. S.G, Josselyn, S.A., Frankland, P.W. & Silva, A.J. (2000). Computer-Assisted Behavioural Assessment of Pavlovian Fear Conditioning in Mice. Learning & Memory 7: 58-72.
56. Bailey, K.R. & Crawley, J.N. (2009). Anxiety-Related Behaviors in Mice. In: Buccafusco J.J., editor. Methods of Behavior Analysis in Neuroscience. 2nd ed. Boca Raton (FL): CRC Press.
57. Bourin, M. & Hascoet, M. (2003). The mouse light/dark box test. European Journal of Pharmacology 463: 55-65.
58. Braff, D; Stone, C; Callaway, E; Geyer, M; Glick, I; Bali, L (1978). "Prestimulus effects on human startle reflex in normals and schizophrenics". Psychophysiology 15: 339–343.
59. Bustin, S.A. & Nolan, T. (2004) Pitfalls of Real-Time Reverse-Transcription Polymerase Chain Reaction. Journal of Biomolecular Technology 15: 155-166.
60. Caldji, C., Diorio, J. & Meaney, M.J. (2000). Variations in maternal care in infancy regulate the development of stress reactivity. Biological Psychiatry 48: 1164-1174.
61. Can, A., Dao, D.T., Arad, M., Terrillion, C.E., Piantadosi, S.C. & Gould, T.D. (2012a). The Mouse Forced Swim Test. Journal of Visualized Experiments 59: 3638-3645.
62. Can, A., Dao, D. T., Terrillion, C. E., Piantadosi, S. C., Bhat, S., & Gould, T. D. (2012b). The Tail Suspension Test. Journal of Visualized Experiments 59: 3769-3774.
63. Crabbe, J., Phillips, T.J. & Belknap, J.K. (2010). The Complexity of Alcohol Drinking: Studies in Rodent Genetic Models. Behavioural Genetics 40: 737-750.
64. Deacon, R.M.J. (2013). Measuring Motor Coordination in Mice. Journal of Visualized Experiments 75: 2069-2077.
65. Green, A.S. & Grahame, N.J. (2007). Reprogramming of genetic networks during initiation of Fetal Alcohol Syndrome. Developmental Dynamics 236: 613-631.
66. Hackett, J.A., Sengupta, R., Zylicz, J.J., Murakami, K., Lee, C., Down, T. & Surani, M.A. (2012). Germline DNA demethylation dynamics and imprint erasure through 5-hydroxymethylcytosine. Science 339: 448-452.
67. Huang, Y., Pastor, W.A., Shen, Y., Tahiliani, M., Liu, D.R. & Rao, A. (2010) The behaviour of 5-hydroxymethylcytsoine in bisulfite sequencings. PLoS 5: e8888
68. Jaluria, P., Konstantopoulos, K., Betenbaugh, M. & Shiloach, J. (2007). A perspective on microarrays: current applications, pitfalls, and potential uses. Microbial Cell Factories 6: 4.
69. Kamangar, F. (2012). Confounding variables in epidemiologic studies: basics and beyond. Archives of Iranian Medicine 8: 508-516.
70. Kendler, K.S., Jacobson, K.B., Prescott, C.A. & Neale, M.C. Specificity of genetic and environmental risk factors for use and abuse/dependence of cannabis, cocaine, hallucinogens, sedatives, stimulants, and opiates in male twins. American Journal of Psychiatry 160: 687-695.
71. Koob, G.F., Kenneth Lloyd, G. & Mason, B.J. (2009). Development of pharmacotherapies for drug addiction: a Rosetta Stone approach. Nature Reviews Drug Discovery 8: 500-515.
72. Liu, D., Diorio, J., Day, J.C., Francis, D.D., Meaney, M.J. (2000). Maternal care, hippocampal synaptogenesis and cognitive development in rats. Natural Neuroscience 3: 799-806.
73. Miranda, T.B. & Jones, P.A. (2007). Journal of Cell Physiology 213: 384-390.
74. Misra. S. (2012) Randomized double blind placebo control studies, the “Gold Standard” in intervention based studies. Indian Journal of Sexually Transmitted Diseases 33: 131-134.
75. Olek, A., Oswald, J. & Walter, J. (1996). A modified and improved method for bisulfite based cytosine methylation analysis. Nucleic Acids Research 24: 5064-5066.
76. Paylor, R., Spencer, C.M., Yuva-Paylor, L.A. & Pieke-Dahl, S. (2006). The use of behavioural test batteries, II: Effect of test interval. Physiology & Behaviour 87: 95-102.
77. Pellow, S. & File, S.E. (1986). Anxiolytic and anxiogenic drug effects on exploratory activity in an elevated plus-maze: A novel test of anxiety in the rat. Pharmacology, Biochemistry, and Behaviour 24: 525-529.
78. Rustay, R., Wahlsten, D. & Crabbe, J.C. (2003). Assessment of genetic susceptibility to ethanol intoxication in mice. Proceedings of the National Academy of Science USA 100: 2917-2922.
79. Smith, C.J. & Osborn, A.M. (2009) Advantages and limitations of quantitative PCR (Q-PCR)-based approaches in microbial ecology. FEMS Microbiology Ecology 67: 6-20.
80. Teif, V., Beshnova, D.A., Vainshtein, Y., Marth, C., Mallm, J., Höfer, T. & Rippe, K. (2014). Nucleosome repositioning links DNA (de)methylation and differential CTCF binding during stem cell development. Genome Research 24: 1285–1295.
81. Turner B (2001) ChIP with Native Chromatin: Advantages and Problems Relative to Methods Using Cross-Linked Material. In: Mapping Protein/DNA Interactions by Cross-Linking [Internet]. Paris: Institut national de la santé et de la recherche médicale.
82. Ward ID, Zucchi, F.C.R., Robbins, J., Falkenberg, E.A., Olson, D.M., Benzies, K. & Metz, G.A. (2013). Trans-generational programming of maternal behaviour by prenatal stress. BMC Pregnancy Childbirth 12: S9.
83. Weaver, I.C.G., Cervoni, N., Champagne, F.A., D’Alessio, A.C., Sharma, S., Seckl, J.R., Dymov, S., Szyf, M. & Meaney, M. (2004). Epigenetic programming by maternal behaviour. Nature Neuroscience 7: 847-854.
84. Yu, M., Hon, G.C., Szulwach, K.E., Song, C., Jin, P Ren, B., He, C. (2012). Tet-assisted bisulfite sequencing of 5-hydroxymethylcytosine. Nature Protocols 7: 21-59.