wenz iD - Proefschrift Lotte Houtepen

L.C. Houtepen

The interplay between environmental factors and DNA methylation in psychotic disorders Environmental orchestration of the epigenome

The interplay between environmental factors and DNA methylation in psychotic disorders Environmental orchestration of the epigenome Copyright Š L.C. Houtepen 2016

All rights reserved. No part of this thesis may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording or otherwise, without prior permission of the author.

ISBN:

9789462955158

Cover illustration: Marion Westerman, The Paper Moon Factory (www.papermoonfactory.nl) Layout: wenz iD || www.wenzid.nl Printed by: Proefschriftmaken.nl || Uitgeverij BOXPress

The interplay between environmental factors and DNA methylation in psychotic disorders Environmental orchestration of the epigenome

Het samenspel tussen omgevingsfactoren en DNA methylatie in psychotische stoornissen (met een samenvatting in het Nederlands)

Proefschrift ter verkrijging van de graad van doctor aan de Universiteit Utrecht op gezag van de rector magnificus, prof. dr. G.J. van der Zwaan, ingevolge het besluit van het college voor promoties in het openbaar te verdedigen op donderdag 20 oktober 2016 des ochtends te 10.30 uur

door

Lotte Cathelijne Houtepen geboren op 10 maart 1987 te Maastricht

Promotoren:

Prof.dr. R.S. Kahn Prof.dr. M. JoĂŤls

Copromotoren:

Dr. M.P.M. Boks Dr. C.H. Vinkers

C

hioda Cristina

M

ozart’s compositions are original, but the number of interpretations made on them is countless. Epigenetics is like an orchestral director giving a different flavour to the same notes by modulating different instruments. In this respect, the DNA - the genetic information - is like sheets of music, and epigenetics is the ensemble of conductors and musicians.

Prelude

A

piece or movement that serves as an introduction to another section or composition and establishes the key.

CHAPTER 1 General introduction

CHAPTER 1

Environment in psychotic disorders Both bipolar disorder and schizophrenia are psychotic disorders caused by complex interactions between biological and environmental factors. Heritability estimates from twin studies indicate high albeit not complete heritability of these disorders, with concordance rates in identical twins of 40-70% for bipolar disorder (Craddock & Sklar, 2013) and around 50% for schizophrenia (Cardno et al., 1999). Thus, two individuals with identical DNA and genetic risk do not automatically both develop a psychotic disorder.

Clearly not all disease risk can be explained via genetic variation (O’Donovan, Craddock, & Owen, 2009), and the environment (in interaction with genetic variation, so called gene-environment interactions) plays an important role (Caspi & Moffitt, 2006). A wellknown example of environmental impact on disease risk for psychotic disorders is the two-fold increased incidence of schizophrenia in offspring of mothers exposed to famine during pregnancy (Brown & Susser, 2008).

In general, traumatic experiences during early-life increase the risk for almost all psychiatric disorders, including schizophrenia and bipolar disorder (Kessler et al., 2010).

The mechanisms underlying this increased risk by traumatic experiences have only been partly identified. Childhood trauma affects stress responses at adult age in healthy controls as well as patients with a psychiatric disorder (Carpenter et al., 2007; Heim et al., 2000; Lovallo, Farag, Sorocco, Cohoon, & Vincent, 2012). Stressful experiences later in life often precipitate episodes in schizophrenia (Nugent, Chiappelli, Rowland, & Hong, 2015) and bipolar patients (Proudfoot, Doran, Manicavasagar, & Parker, 2011), suggesting that the response to stress is impaired. Based on these observations, early-life stress is hypothesized to alter the stress response system and consequently increase the risk for a psychotic disorder, especially when individuals are exposed to stressful conditions later in life. If so, stress reactivity could be a marker for disease vulnerability and examining the acute response to stress could further our understanding of disease risk (Allen, Kennedy, Cryan, Dinan, & Clarke, 2014).

Acute stress activates the hypothalamus- pituitary- adrenal (HPA) axis, which results in the release of the hormone cortisol (McEwen, 2004) (see Figure 1.1). Cortisol stress reactivity is altered in several psychiatric disorders (Brenner et al., 2009; Lange et al., 2013; Petrowski, Wintermann, Schaarschmidt, Bornstein, & Kirschbaum, 2013), underlining the importance of the stress response system in psychiatry. When assessing cortisol stress reactivity, multiple other factors, such as genetic background (DeRijk & de Kloet, 2008; van Leeuwen et al., 2011), number of episodes (Morris, Rao, & Garber, 2012), medication use (Lange et al., 2013) and previous stressful experiences (Heim et al., 2000), should be considered. Assessing childhood trauma is especially relevant due to its association with altered cortisol stress reactivity (Carpenter et al., 2007; Heim et al., 2000; Lovallo et al., 2012) as well as increased risk for psychiatric disorders (Kessler et al., 2010).

It remains unclear which biological mechanism connects childhood trauma exposure to changes in cortisol stress reactivity and/or the development of a psychiatric disorder later in life.

10

Introduction

1

Figure 1.1 Schematic overview of the hypothalamus- pituitary- adrenal (HPA) axis. Stress can induce the release of corticotropin-releasing factor (CRF) from the hypothalamus. CRF activates the release of adrenocorticotropic hormone (ACTH) from the anterior pituitary gland. ACTH stimulates cells of the adrenal glands to release the stress hormone cortisol. Cortisol exerts a negative feedback on CRF and ACTH release. + = activates, - = inhibits

Epigenetics A potential mechanism linking the environment to health outcomes is a variety of molecular processes coined ‘epigenetics’ by Waddington in the early 20th century (Waddington, 1942). Epigenetics studies “a stably heritable phenotype resulting from changes in a chromosome without alterations in the DNA sequence”. Chromosomes densely pack DNA to fit the entire DNA sequence in a single cell. Altering the DNA packaging influences gene expression and ultimately the translation of DNA to proteins and molecules. In fact, the tightly regulated gene expression changes that occur during neurodevelopment are intricately linked to epigenetic variation (Bale, 2015; Hannon et al., 2016; Slieker et al., 2015). Epigenetics is not a single switch to turn on or off, but rather encompasses many different processes that influence the accessibility of the chromosome environment without changing the DNA sequence. Specifically, chromatin accessibility depends on the chromatin structure which can change from the standard condensed transcriptionally silent state called heterochromatin to the less condensed more active euchromatin state (B. E. Bernstein, Meissner, & Lander, 2007). Many epigenetic processes can affect gene expression, such as histone modifications and DNA methylation, and each specific epigenetic mark should be interpreted in relation to the others (see Figure 1.2) (for review see (Kofink, Boks, Timmers, & Kas, 2013)). For instance there are several different types of histone marks. While histone 3-lysine 4 mono-, diand tri-methylation (H3K4me, H3K4me2, H3K4me3), H3K9me and H3K27me marks tend to indicate transcriptionally active gene sites and di-, tri-methylated H3K9 and H3K27 marks are more frequent around inactive genes, combined information of several histone marks is the most useful indicator of chromatin state (Kundaje et al., 2015). In general, DNA methylation is substantially correlated with other epigenetic marks such as chromatin modifications (Hoffmann, Zimmermann, & Spengler, 2015; Kundaje et al., 2015) and tends to add stability to epigenetic states (Jones, 2012). DNA methylation is 11

CHAPTER 1

extensively studied because it is well preserved in stored DNA samples and array-based technologies enable large scale genome-wide studies (Bibikova et al., 2011). Therefore, the current thesis focuses on DNA methylation.

Figure 1.2 Two well-studied epigenetic mechanisms: histone modifications and DNA methylation.

DNA methylation DNA methylation involves the addition of a methyl group to a DNA base (for review see (Schubeler, 2015)). Of the four DNA bases (cytosine, guanine, adenine, thymine), the only DNA bases known to be highly methylated in mammals are cytosines preceding guanines (CpG sites). The human genome comprises ~600 million cytosine bases per strand of which ~5 % are CpG sites. The majority of CpG sites is methylated (70-80%) (Bird, 2002), but the unmethylated CpG sites tend to cluster in special “islands�. Typically, 85% of these CpG islands are hypomethylated (reviewed in states (Jones, 2012)). CpG islands are defined as 200 base pair (bp) windows with a >50% GC content and an observed (within a given sequence) to expected (within the genome) CpG ratio >60% (Fatemi, Alizad, & Greenleaf, 2005). CpG islands can vary in size between 300 and 3000 bp and are found upstream of ~40% of transcriptional start-sites (TSS) (Frigola et al., 2006; Jones, 2012).

The classical view that DNA methylation represses gene activity stems from cancer research that links lower gene activity to higher DNA methylation at the edges of the promoter associated CpG islands (Irizarry et al., 2009). It quickly became apparent that DNA methylation differences outside these CpG islands are also relevant for gene expression, as DNA methylation is higher in the gene body of actively transcribed genes (Maunakea et al., 2010). The precise role for this gene body DNA methylation is unclear, but appears to involve other genomic functions, such as adaptations in gene transcript 12

Introduction

length (alternative splicing) and promoter usage. However, the relationship between DNA methylation and gene expression is more complex than originally thought. A recent report showed that the correlation between DNA methylation and gene activity can be positive as well as negative, irrespective of the location at which DNA methylation changes occur (van Eijk et al., 2012; Wagner et al., 2014). Thus, the exact location on the genome may indicate the likelihood that DNA methylation affects gene activity, but does not predict whether more methylation will stimulate or repress gene expression.

DNA methylation changes do not occur independently for each individual differentially methylated position (DMP), but are often correlated within 1000 bases (Eckhardt et al., 2006). This can be used to identify differentially methylated regions (DMR). Different types of DMRs are classified according to the feature that was used to distinguish the DMR. For example, some DMRs were identified when comparing cancer to normal tissue and are labelled disease associated DMRs (Irizarry et al., 2009), while other DMRs differentiate between cell types (cell-type DMR) or tissue types (tissue-specific DMR) (Lister et al., 2009). A useful extension of cell-type specific DMRs is the calculation of different blood cell-type populations from whole-blood DNA methylation patterns (Houseman et al., 2012). In a comprehensive study of healthy human tissue, higher DMR methylation was related to lower gene expression (Schultz et al., 2015). This negative correlation was stronger for DMRs closer to the transcription start site, although the strongest negative correlation was not in gene promoters but downstream of the promoter up to 8 kilobases (kb) away in regions inside a gene (intragenic). The exact role of these intragenic DMRs is unknown, but based on their overlap with other genomic elements they can be subdivided into intragenic DMRs near CpG islands (19%), promoters (22%) or enhancers (23%), and regions with unknown genomic function (35%). Clearly location is important when interpreting DNA methylation differences in individual loci as well as regions, but it can be equally useful to identify methylation patterns with methods that do not rely on a priori knowledge of for instance genomic location. DNA methylation loci can also be clustered according to the similarity in DNA methylation changes (Langfelder & Horvath, 2008). This weighted correlation network analysis (WGCNA) method successfully identified age related DNA methylation clusters regardless of the tissue type (Horvath et al., 2012). Whether looking at DNA methylation differences on one location, a region or clusters, linking DNA methylation to the phenotype of interest is essential for interpretation.

DNA methylation and the environment Since DNA methylation is influenced by genetic variants (Bell et al., 2011; Boks et al., 2009), random stochastic factors as well as environmental exposures (Heijmans et al., 2008), it is important to identify the contribution of different sources. In identical twins some DNA methylation differences are already present at birth indicating some individual differences already develop in utero (Fraga et al., 2005). More importantly, most of the DNA methylation pattern changes over time are under influence of the environment (for review see (Yet, Tsai, Castillo-Fernandez, Carnero-Montoro, & Bell, 2016)). In a recent comprehensive genome-wide twin study of adipose tissue (n=4798) the environment 13

1

CHAPTER 1

not shared between twins accounted for >60% of DNA methylation variation and shared environment played a minor role. The clustering of the environmentally related DNA methylation changes also suggests it is a true environmental signal and not some form of biological noise (Busche et al., 2015). A textbook example of the environmental influence on DNA methylation is the link between DNA methylation and early-life trauma in animal models. Trauma exposed pups have altered DNA methylation of the glucocorticoid receptor gene (NR3C1) and increased cortisol stress reactivity later in life (J. Chen et al., 2012; Szyf, Weaver, Champagne, Diorio, & Meaney, 2005; Weaver et al., 2004). In humans childhood trauma affects DNA methylation in epigenome wide screens (Labonte et al., 2012) as well as HPA-axis related candidate genes (Vinkers et al., 2015).

In fact, prenatal stress exposure altered DNA methylation around the NR3C1 gene and these NR3C1 DNA methylation changes were subsequently linked to increased cortisol stress reactivity in the offspring (Oberlander et al., 2008). This supports the hypothesis of DNA methylation as an underlying mechanism for these early-life exposures-related changes in the stress response system and perhaps ultimately in the risk for a psychiatric disorder (Teicher & Samson, 2013).

An important aspect of this hypothesis is the temporal stability of DNA methylation. The question is: will later life DNA methylation patterns still reflect these early-life events?

Some DNA methylation differences are quite stable over time, as evident from the ability to separate liver and brain tissue on DNA methylation patterns throughout life (Rakyan et al., 2008). However, DNA methylation is in a dynamic state, constantly influenced by both deterministic as well as stochastic processes, making it difficult to tease the underlying factors apart at any given time. In utero environmental exposures are hypothesized to have a long lasting impact (Waterland & Michels, 2007), because major DNA methylation changes occur in all cells during the first trimester of pregnancy (Reik, Dean, & Walter, 2001). In line with this hypothesis, in utero famine in early, but not in mid or late gestation is associated with extensive DNA methylation differences in the offspring of famine-exposed mothers (Heijmans et al., 2008; Tobi et al., 2015). Preliminary evidence even connects famine related DNA methylation differences to later life outcomes, as famine exposure was associated with carnitine palmitoyltransferase-1A (CPT1A) DNA methylation which was in turn correlated with LDL cholesterol levels (Tobi et al., 2014).

This indicates that studying DNA methylation changes can facilitate insights into disease etiology.

DNA methylation in psychotic disorders It is clear that neither genes nor the environment act in isolation to increase susceptibility for psychotic disorders and this inherent complexity of gene-environment interactions is reflected in the epigenome. By understanding DNA methylation marks that are influenced by the environment and contribute to the development of psychotic disorders, we can increase our understanding of the mechanisms underlying the environmental impact in psychotic disorders. In support of DNA methylation reflecting disease associated changes, twins discordant for psychosis also differ in global DNA methylation (Dempster et al., 2011). Several studies reported DNA methylation differences in brain tissue of 14

Introduction

psychotic patients compared to brain tissue of healthy controls (Jaffe et al., 2016; Melka et al., 2014; Mill et al., 2008; Pidsley et al., 2014). Such studies are consistent with the neurodevelopmental origin of psychotic disorders, as the differentially methylated DNA loci identified in brain tissue of schizophrenia patients were enriched for loci that undergo dynamic DNA methylation changes during human fetal brain development (Pidsley et al., 2014). Indeed, the majority of schizophrenia related DNA methylation differences in a recent genome-wide analysis of prefrontal cortex (PFC) tissue were linked to both genes associated with schizophrenia as well as early developmental processes. In fact, overall the strongest enrichment was for early developmental processes rather than schizophrenia, suggesting that environmental factors play a more important role in these schizophrenia associated DNA methylation patterns than genetic differences (Jaffe et al., 2016). This corresponds with another genome-wide study published around the same time, where genetic variation that regulates DNA methylation variation during development was enriched for schizophrenia related genetic mutations (Hannon et al., 2016). If DNA methylation differences that contribute to disease risk are already present before disease development, identifying these DNA methylation differences could provide early detection. DNA methylation can also be useful to detect new treatment targets and potentially help optimize treatment by identifying who will benefit.

Epigenetic interventions Tentative support that altering DNA methylation is important in psychotic disorders comes from a mouse model where the mothers were exposed to stress during pregnancy.

The mice that were exposed to in-utero stress exhibited behavioural and epigenetic alterations later in life akin to some of the symptoms and epigenetic changes observed in schizophrenia and bipolar disorder patients (for review see (Guidotti & Grayson, 2014) and (Dong et al., 2015)). Specifically, the behavioural phenotype in the in-utero stressed mice is characterized by hyperactivity, stereotyped and compulsive behaviours, hypersensitivity to N-methyl d-aspartate receptor blockers and deficits in social interaction, pre-pulse inhibition and fear conditioning. The DNA methylation profile common to the in-utero stressed mice and psychotic patients includes hypermethylation at the glutamate decarboxylase 1 (GAD1), REELIN (RELN) and brain-derived neurotrophic factor (BDNF) gene promoters. Indeed, recently GAD1, RELN and BDNF promoter methylation in these in-utero stressed mice was correlated with locomotor activity and social interaction, suggesting that DNA methylation underlies the psychotic-like behaviour in these mice (Dong, Tueting, Matrisciano, Grayson, & Guidotti, 2016). Moreover, the behavioural and DNA methylation changes in the in-utero stressed mice were normalized by treatment with the atypical antipsychotic clozapine that has chromatin remodelling properties, but were unaffected by the typical antipsychotic haloperidol that does not have the same chromatin remodelling properties. Interestingly, valproate acid, a psychiatric medication often prescribed to bipolar disorder patients, also abolished the DNA hypermethylation at the GAD1, RELN and BDNF promoters as well as the altered behaviours in the in-utero stress mice model (Dong et al., 2016). Taken together, these data show that in-utero stress in mice induces abnormalities in DNA methylation and 15

1

CHAPTER 1

schizophrenia-like behaviours that are sensitive to medication treatment. In the majority of human DNA methylation studies, patients are taking medication and it is not possible to distinguish treatment effects from disease-related differences in DNA methylation.

However, studies with medication free (Kinoshita et al., 2013) or medication naĂŻve (Abdolmaleky et al., 2014) schizophrenia patients also reported DNA methylation differences between patients and controls. Indeed, schizophrenia patients treated with psychiatric medication had higher global DNA methylation levels (60.5%) that were more similar to healthy controls (65.7%) than schizophrenia patients who did not receive antipsychotic treatment (37.6%) (BĂśnsch et al., 2012), indicating that treatment can normalize some of the DNA methylation differences seen in psychotic patients. At the moment it is still too early to disentangle these treatment effects from etiological changes in DNA methylation, although some common psychiatric medications such as valproic acid are known to affect epigenetic regulators (for review see (Boks et al., 2012) and (Szyf, 2015)). Since DNA methylation marks are introduced and removed by enzymes and therefore potentially reversible, drugs targeting epigenetic mechanisms are already under investigation for the treatment of leukemia (Minucci & Pelicci, 2006). In psychiatry, we are further away from practical applications, but contextual fear conditioning studies in animals have shown promising results for several DNA methylation inhibitors that suggest inhibiting DNA methylation could be translated into a treatment for posttraumatic stress disorder (PTSD) in humans (Szyf, 2015).

Aim of this thesis Environmental exposures early-in-life increase the risk for a psychotic disorder, including bipolar disorder and schizophrenia. DNA methylation is a potential mechanism mediating this environmental influence. Examining DNA methylation differences induced by environmental exposures relevant for psychotic disorders can therefore increase our understanding of disease etiology and ultimately help improve treatment. In this thesis we examined the interplay between environmental exposures relevant for psychotic disorders and DNA methylation.

In chapter 2 we investigated whether acute stress responses are related to bipolar disorder by examining the determinants of stress reactivity differences between bipolar patients, siblings of bipolar patients, and controls. Since medication can have a profound influence on DNA methylation, we examined global DNA methylation signatures of bipolar patients in chapter 3. In chapter 4 we employed a genome-wide approach to identify DNA methylation differences related to cortisol stress reactivity in healthy controls with a specific focus on the role of childhood trauma. Finally, in chapter 5 we detected DNA methylation changes in famine exposed schizophrenia patients and discuss how these reflect the environment (in utero) as well as disease risk.

16

Introduction

1

Figure 1.3 Overview of the main subjects discussed in this thesis and the connection between them. Dotted lines are used to denote the relationships that are examined in this thesis. Briefly, chapter 2 discusses the influence of antipsychotic medication on cortisol stress reactivity in bipolar patients, chapter 3 indicates medication use can also affect DNA methylation levels of bipolar patients, chapter 4 investigates the role of DNA methylation in childhood trauma related differences in cortisol stress reactivity and chapter 5 explores DNA methylation differences that are related to schizophrenia risk after in utero famine exposure.

17

Exposition

T

he initial presentation of the thematic material of a musical composition, movement, or section. The use of the term generally implies that the material will be developed or varied.

CHAPTER 2 Antipsychotic use is associated with a blunted cortisol stress response: A study in euthymic bipolar disorder patients and their unaffected siblings European Neuropsychopharmacology 2015 Jan;25(1):77-84 DOI: 10.1016/j.euroneuro.2014.10.005

L.C. Houtepen, M.P.M. Boks, R.S. Kahn, M. JoĂŤls, C.H. Vinkers

CHAPTER 2

Abstract There is ample evidence that the acute stress response is altered in schizophrenia and bipolar disorder. However, it is not clear whether such changes are related to the illness, a genetic vulnerability, or is the result of medication that is used in the majority of these patients. Therefore, we investigated determinants of the acute endocrine and autonomic stress response in healthy controls (n=48), euthymic BD1 patients (n=49) and unaffected siblings of BD1 patients (n=27). All participants completed a validated psychosocial stress task, the Trier Social Stress Test for Groups (TSST-G). Saliva levels of alpha-amylase and cortisol were measured before, during, and after exposure to stress. Compared to controls, we found a significantly blunted cortisol stress response in BD1 patients. Conversely, BD1 patients displayed exaggerated alpha-amylase levels in response to stress. Antipsychotic use was a significant contributing factor to the blunted cortisol stress response in BD1 patients. Unaffected BD1 siblings displayed similar stress-induced cortisol and alpha-amylase levels as controls, suggesting that familial risk for BD1 did not have a large effect on the functionality of the stress system.

In conclusion, this study shows that euthymic BD1 patients have a substantially blunted endocrine stress response but an exaggerated autonomic stress response and that the endocrine stress response differences can be largely contributed to antipsychotic use rather than constitute a specific BD1 phenotype or vulnerability.

Introduction In psychiatric disorders such as schizophrenia, bipolar disorder and major depressive disorder, blunted and increased stress responsivity have been reported (Brenner et al., 2009; Petrowski, et al., 2013). However, our understanding of the neurobiological background altered stress response in these disorders is limited. There are indications in bipolar disorder 1 (BD1) patients that stress exposure results in a blunted endocrine stress response (Wieck et al., 2013) and reduced heart rate variability (Cohen et al., 2003). BD1 patients also display an increased basal activity of the endocrine and autonomic systems as reflected by an increased cortisol awakening response (Cervantes, Gelber, Kin, Nair, & Schwartz, 2001) and increased basal sympathetic nervous system (SNS) activity (Lake et al., 1982). Even though there are indications that stress system functionality is altered in BD1, it is currently unknown whether stress reactivity is consistently altered in BD1 patients and which factors may influence stress reactivity.

In other psychiatric populations, several factors have been identified that influence stress reactivity, including childhood maltreatment (Heim et al., 2000), and the number of depressive episodes (Morris et al., 2012). Investigating the influence of psychiatric medication such as antidepressants (Lange et al., 2013) and antipsychotics (Cohrs et al., 2006) is also relevant as these drugs can alter HPA axis activity. Moreover, multiple studies in healthy controls have clearly linked genetic markers to altered stress reactivity in the TSST task (see for example (van Leeuwen et al., 2011; DeRijk & de Kloet, 2008)). In light 20

Cortisol stress response in bipolar disorder

of the high heritability of BD1 and alterations in the stress system, impaired stress system functionality in BD1 may also be the result of a genetic predisposition for BD1 (Ising & Holsboer, 2006). If a genetic BD1 predisposition would result in altered stress reactivity, it is expected that unaffected adult siblings of BD1 patients, who on average share 50% of risk genes with probands, display a similar altered stress response.

This study aims to establish i) indices for both HPA-axis and ANS reactivity to stress in euthymic BD1 patients and ii) examine to what extent disease characteristics, medication, and genetic vulnerability contribute to aberrant stress reactivity of BD1 patients. To this end, we exposed euthymic BD1 patients to a group-wise Trier Social Stress Test (TSST-G), and compared this group to unrelated unaffected healthy siblings of BD1 patients and healthy controls.

Experimental procedures Participants Participants were eligible for participation if they had 3 or more Dutch grandparents and were fluent in Dutch, in view of the verbal test. Participants were instructed to refrain from heavy meals, drinks other than water and heavy exercise at least 2Â h before the stress task, as well as to refrain from caffeine use at least 4Â h before the experimental session. All participants reported that they had adhered to these instructions. Participants had not previously been enrolled in stress-related research and were unfamiliar with each other. Between December 2011 and June 2013, participants were recruited through flyers at general hospitals, psychiatric institutions, and patient associations. Siblings (n=27) were not related to BD1 participants (n=49). Healthy controls (n=48) were matched based on age and gender. The study was approved by the Utrecht Medical Center ethical review board and performed according to the ICH guidelines for Good Clinical Practice and the Declaration of Helsinki and its latest amendments. All participants gave their written informed consent prior to their inclusion in the study and were financially compensated.

Clinical characteristics and drug use On the day of testing lifetime DSM-IV diagnoses in unaffected siblings and healthy controls group were assessed with the Mini International Neuropsychiatric Interview (MINI) plus (Sheehan et al., 1998). For all BD1 participants clinical characteristics including BD1 diagnosis, comorbid psychiatric diagnosis, number of manic and depressed episodes, and age of disease onset were established with the Structured Clinical Interview for DSM-IV (SCID) (First, Spitzer, Gibbon, & Williams, 2002). The interviews were conducted by at least one well-trained independent rater.

Just before the experiments euthymic mood status in BD1 patients was confirmed using the 30-item Inventory of Depressive Symptomatology (IDS-C30, range 0-24 (Rush, Gullion, Basco, Jarrett, & Trivedi, 1996)) and Young Mania Rating Scales (YMRS, range 0-7 (Tohen et al., 2000; R.C. Young, Biggs, Ziegler, & Meyer, 1978)) interviews. Participation 21

2

CHAPTER 2

was postponed if participants scored above predefined cut off values on the YMRS (above 12) or IDS-C30 (above 24). Based on these criteria, 2 participants were rescheduled for a different test day.

Current use of psychoactive substances (amphetamines, MDMA, barbiturates, cannabinoids, benzodiazepines, cocaine, and opiates) was determined with a urine multidrug screening device (InstantView) and self-report questionnaire. All participants reporting psychiatric medication use (antidepressants, benzodiazepines, anticonvulsants and antipsychotics), were on a stable (at least one month) dosing schedule.

Benzodiazepine use was confirmed by the urine screen. The dose of antipsychotics was transformed to chlorpromazine (CPZ) equivalents as previously described (Woods, 2003).

If participants smoked daily, they were defined as a smoker. Furthermore, female participants were explicitly asked for information about their menstrual cycle and contraceptive medication use as this can influence HPA axis activity (Kirschbaum, Kudielka, Gaab, Schommer, & Hellhammer, 1999).

Four participants (3 controls and 1 BD1 patient) scored positive for cannabis use. Exclusion of the cannabis users did not change any of the results. Four participants (3 BD1 patients and 1 sibling of BD1 patient) taking beta blockers for high blood pressure were excluded from alpha-amylase analysis as its use impairs the alpha-amylase response (van Stegeren, Rohleder, Everaerd, & Wolf, 2006). Since beta blocker use does not affect the cortisol stress response (Kudielka et al., 2007), participants (N=4) taking beta blockers were included for cortisol analysis. Nevertheless, exclusion of these participants did not change any of the cortisol results.

Procedures General Participants were tested between 1300h and 1700h to control for diurnal variations of cortisol secretion. Participants completed an individual assessment of current mood and symptomology based on psychiatric diagnostic interviews (see paragraph Clinical characteristics and drug use). Participants were instructed not to communicate with each other before being escorted to a waiting area. Ten minutes prior to the TSST-G, participants received instructions and subsequently completed the (TSST-G) task as previously described (Vinkers et al., 2013). In short, participants were instructed to carry out a public speaking test and subsequently perform an arrhythmic task. All participants were called upon in random order.

Saliva samples Saliva was collected using salivettes (Sarstedt, Nümbrecht, Germany) for analysis of cortisol and alpha-amylase levels. In total, eight saliva samples were collected in each subject over a 90 minutes’ time period (Figure 2.1). Samples were directly stored at −80 °C and analyzed as previously described (Vinkers et al., 2013). In short, cortisol was measured without extraction using an in house competitive radio-immunoassay, and alpha-amylase was measured using a Beckman-Coulter AU5811 chemistry analyzer.

If one (out of 8) saliva samples was missing, the missing value was obtained by averaging 22

Cortisol stress response in bipolar disorder

the surrounding data points. This procedure was used for ten participants (4 controls, 2 siblings and 4 BD1 patients). Exclusion of these 10 participants altogether from the analyses did not alter any of the results. Three participants (2 controls and 1 BD1 patient) were excluded because more than one of the eight salivettes did not contain enough saliva for analysis.

Questionnaires Perceived levels of stress and anxiety were repeatedly assessed at baseline (t = −10 min) and during the stress test (t = +8 min) using visual analog scales (VAS, 118 mm scale).

Moreover, previous exposure to traumatic stress influences stress reactivity (Carpenter et al., 2007; Heim et al., 2000; Lovallo et al., 2012), therefore exposure to childhood maltreatment (Childhood Trauma Questionnaire (CTQ)(Bernstein et al., 2003) and major life events (Lifetime Stressor Checklist-Revised (LSC-R) (Wolfe, Kimerling, & Brown, 1996; Vinkers et al., 2014)) were investigated with questionnaires in all participants.

Statistical analyses All statistics were carried out using SPSS version 20 (SPSS Inc., Chicago, IL, USA). One way ANOVAs were used to analyze differences in baseline parameters between groups.

When applicable, Bonferroni post-hoc analysis was performed. If the assumption of homogeneity of variance was violated, Welch’s F test was performed followed up with a Games-Howell post-hoc analysis (indicated by non-integer F values). In the case of significant baseline differences in childhood maltreatment, major life events, menopausal status, contraceptive medication use, cannabis use or smoking status, main analyses were performed both with and without these indicators as covariate. Changes in subjective stress rating (VAS during the stress task – VAS baseline) were analyzed using a one way ANOVA with group (control, sibling or BD1) as between-subject factor. Cortisol and alpha-amylase levels were analyzed using repeated measures ANOVA with group (control/sibling/BD1) and sex as between-subject factor and age as a covariate. Sex and age were included in the model as both parameters have been reported to influence cortisol and alpha-amylase levels (Nater, Hoppmann, & Scott, 2013). Planned follow up analysis consisted of comparing BD1 patients to controls and BD1 patients to siblings.

Results were corrected by the more stringent Greenhouse-Geisser procedure where appropriate (indicated by an ε value). If there was a significant effect in the repeated measures analysis, data were analyzed per time point using a one-way ANOVA with group (control, sibling or BD1) as between-subject factor.

Furthermore in a stratified analysis of the BD1 group, medication use and clinical characteristics (age of onset, number of manic and depressed episodes and comorbid psychiatric diagnosis) were added as covariates to the repeated measure model. For medication use, lithium (dose in mgs), antidepressant use (yes/no), benzodiazepine use (yes/no), anticonvulsant use (yes/no), and antipsychotic use (yes/no) were used as separate indicators in a single model. Pearson’s correlations were used to investigate the relation between medication use and clinical characteristics. If there was a significant correlation between a clinical characteristic and medication use, these measures were 23

2

CHAPTER 2

both included in the repeated measures model. The significance level was set at P < 0.05.

All results are presented as average and standard error of the mean.

Results Lifetime psychiatric disorders and current mood status Four healthy controls met criteria for a lifetime but not current psychiatric disorder and had recovered completely without professional help: alcohol/drug abuse (n=3) and agoraphobia (n=1). Exclusion of these four participants with a history of a psychiatric disorder did not alter any of the results (data not shown). A total of nine siblings met criteria for a previous depressive episode, of which 5 experienced recurrent depressive episodes (and 4 of these 5 participants used antidepressants at the time of participation).

Three of the siblings who met the criteria for a previous depressive disorder, also had a comorbid anxiety (n=2) or alcohol abuse (n=1) diagnosis.

The BD1 diagnosis was confirmed in all recruited BD1 patients. On the experimental day, current mood in the BD1 patients was euthymic as determined by the YMRS (average 1.2, range 0-7) and IDS-C30 (average 5.3, range 0-24). Other clinical characteristics collected in the BD1 sample were: number of episodes (average 6.4, range 1-25), age of onset (average 26.6, range 7-60) and comorbid disorders. Nine patients (18%) had either a comorbid anxiety (n=7) or obsessive compulsive (n=2) disorder.

Baseline characteristics Age, sex, contraceptive use and cannabis use did not significantly differ between controls and BD1 patients. The BD1 siblings group contained significantly more females, had a higher average age, and a lower number of smokers compared to the other groups (Table 2‑1). Inclusion of smoking status (all p-values >0.12), menopausal status (all p-values >0.16) or contraceptive use (all p-values >0.20) as covariate had no effect on either cortisol or alpha-amylase levels and were not included as a covariate in further analyses.

Childhood trauma and major life events Siblings and BD1 patients reported more stressful life events compared to controls (LSC-R effect F2, 123 =7.4 p=0.001; post hoc: BD1-control p=0.004 and sibling-control p=0.005).

In contrast, childhood trauma levels were comparable across all groups (CTQ effect F2, =1.3 p=0.27). Inclusion of LSC-R as covariate had no effect on any significance level 123 in the current study and was not included in subsequent analyses.

Perceived stress levels Basal subjective stress ratings were comparable between healthy controls (mean 1.74±2.3), siblings of BD1 patients (mean 1.76±1.9), and BD1 patients (mean 2.36±2.4) (F2,123=1.014 p=0.37). Moreover, in response to stress, there was a similar increase in subjective stress levels across all groups (F2, 121=1.2 p=0.292).

24

Cortisol stress response in bipolar disorder

Table 2.1 Sample characteristics.

Control(n=48)

Siblings(n=27)

BD1(n=49)

52%

70%

51%

Age (years, range)

43.5 (21-69)

54.5 (32-66)

43.4 (19-67)

Childhood maltreatment (CTQ total score, range)

33.4 (24-63)

34.4 (24-82)

36.9 (24-74)

Major Life Events (LSC-R total score, range)

3.90 (0-10)

5.85 (1-11)**

5.61 (0-11)**

Cannabis use (yes/no, %)

3 (6.3%)

0

1 (2.0%)

Smoker(yes/no, %)

11 (23%)

1 (3.7%)*

18 (36%)

Contraceptive use (yes/no, %)

5 (20%)

0*

5 (20%)

Lithium (yes/no, %)

-

-

34 (69%)

Antipsychotics (yes/no, %)

-

-

29 (61%)

Antidepressants (yes/no, %)

1 (2.1%)

4 (15%)

10 (20%)

Benzodiazepines (yes/no, %)

-

2 (7.4%)

8 (14%)

Anticonvulsants (yes/no, %)

-

-

14 (28%)

Age of onset (years, range)

-

-

16.1 (7-60)

Episodes (number, range)

-

-

6.3 (1-25)

YMRS score (total score, range)

-

-

1.2 (0-7)

IDS-C30 score (total score, range)

-

-

5.3 (0-24)

Sex (% female )

2

*: p<0.05; **: p <0.01; compared to controls

Cortisol Groups differed in the endocrine stress response (group effect F2, 116=6.2 p=0.003; time x group interaction F14, 742=3.2 p=0.010 ε=0.317; see Figure 2.1). Planned follow up analysis revealed that this difference resulted from a significantly blunted cortisol responses in BD1 patients (BD1-control time x group F7, 588=4.6 p=0.008 ε =0.324). Post hoc analysis confirmed significant different cortisol levels between BD1 patients and controls from 20 minutes after TSST-G until the end of measurement (t=90min) (p<0.05). Siblings showed a similar cortisol stress response as healthy controls (Sibling- control: time xgroup F7, 476=2.2 p=0.111; ε =0.298). As expected, women exhibited an overall blunted cortisol stress response compared to men (sex effect F1, 116=9.6 p=0.002; time x sex interaction F7, 472=3.7 p=0.021 ε =0.317). A similar sex effect was found after exclusion of women who used contraceptives (sex effect F1, 109=9.4 p=0.003; time x sex interaction F7, 763=3.7 p=0.026 ε =0.302).

Alpha-amylase Stress-induced alpha-amylase levels differed between groups over time (group effect F2, 114=5.7 p=0.004; time x group interaction F14, 798 = 3.6 p=0.004 ε =0.353; see Figure 2.2). Planned follow up analysis showed that this difference originated from a significantly increased alpha-amylase response in BD1 patients (BD1-control time x group interaction F7, 609 =7.1 p<0.001 ε = 0.322). Significant differences were present between BD1 patients and healthy controls during all time points (p<0.05) except at t=80 min (p=0.076).

25

CHAPTER 2

In contrast, the alpha-amylase stress response of siblings was not significantly different from the alpha-amylase stress response in healthy controls (Sibling vs.-control: time x group interaction F7, 469=1.6 p=0.170 ε = 0.452). No significant sex effects or interactions were found (sex effect F1, 114=0.4 p=0.842, time x sex F7, 472=0.2p=0.846 ε =0.353). The alpha-amylase stress response was enhanced in older participants independent of group status (age effect F1, 114=8.5 p=0.004; time x age F7, 472=2.9 p=0.044 ε =0.353).

Figure 2.1 The cortisol response to the TSST-G in BD1 patients, unaffected siblings, and healthy controls. *: p<0.05; #: p<0.10.

Figure 2.2 The effect of the TSST-G on a-amylase levels in saliva of BD1 patients, unaffected siblings, and healthy controls. *: p<0.05; #: p<0.10

Medication All BD1 patients used medication, most commonly lithium (69%) and antipsychotics (61%, average CPZ equivalent 112.8 mg), followed by anticonvulsants (28%), antidepressants (20%), and benzodiazepines (14%). The cortisol stress response in BD1 patients was not influenced by use of lithium (all p-values >0.31), benzodiazepines (all p values>0.24), antidepressants (all p values>0.38), or anticonvulsants (all p values>0.295). Similarly, stress-induced alpha-amylase measures were not influenced by use of lithium (all p values >0.16), benzodiazepines (all p values>0.08), antidepressants (all p values>0.41), or anticonvulsants (all p values>0.45).

26

Cortisol stress response in bipolar disorder

However, antipsychotic use was associated with a blunted stress-induced cortisol response in BD1 patients (time*antipsychotic use F7, 315=4.2 p=0.007 Îľ =0.369; see Figure 2.3). This antipsychotic effect was specific for stress-induced cortisol fluctuations over time since there was no significant main effect of antipsychotic use on overall or basal cortisol levels in the BD1 group (F1, 45=2.8 p=0.100). Moreover, CPZ equivalents showed similar blunted stress-induced cortisol responses in BD1 patients (time*CPZÂ F7, 315=5.8 p=0.002 Îľ =0.350). Antipsychotic use was not significantly associated with either overall or time-dependent levels of stress-induced alpha-amylase (all p values >0.17). No significant correlations were found between antipsychotic use (CPZ equivalents) and clinical characteristics (number of episodes r=-.008 p=0.960; age of onset r=-0.038 p=0.813), therefore clinical characteristics were not used as a covariate in the analysis of antipsychotics.

Figure 2.3 The effect of antipsychotic use on the cortisol response to the TSST-G in BD1 patients.

The cortisol response of healthy controls is added for graphical comparison only and was not used in the statistical model. *: p<0.05; #: p<0.10

Figure 2.4 Number of mood episode (more or less than 5) on the cortisol stress response in BD1 patients. The cortisol response of healthy controls is added for graphical comparison only and was not used in the statistical model. *: p<0.05

27

2

CHAPTER 2

Clinical BD1 disease parameters The number of previous mood episodes was significantly associated with overall decreased cortisol levels in BD1 patients independent of stress exposure (episode effect F1, 37=6.6, p=0.014; see Figure 2.4). In contrast, no significant effect of number of previous mood episodes was seen on the stress-induced cortisol response (time x episodes interaction F7, 259=1.1 p=0.313 ε =0.381). For graphical representation, number of episodes were divided with a median split in high (>5) and low (≤5) number of episodes. Age of onset (all p values>0.26) and comorbid psychiatric diagnosis (all p values>0.19) had no significant effect on either cortisol or alpha amylase levels.

Discussion We found that in response to stress, euthymic BD1 patients perceived similar stress levels as healthy controls but displayed a blunted cortisol response and increased alphaamylase levels. This blunted cortisol response is in line with a recent study that investigated 13 euthymic female BD1 patients (Wieck et al., 2013). A striking finding of the current study is that antipsychotic use was the most important factor associated with the reduced cortisol stress response in BD1 patients, accounting for most of the differences between stress-induced cortisol levels in BD1 patients and healthy controls.

After inclusion of antipsychotic use as a covariate, the blunted cortisol stress response in BD1 patients no longer significantly differed from the one in healthy controls. To our knowledge, this is the first study to show that antipsychotic use plays an important role in endocrine stress reactivity. Studies investigating stress reactivity in schizophrenia have consistently found a blunted cortisol stress response (Jansen et al., 1998; Jansen, Gispende Wied, & Kahn, 2000; Ritsner et al., 2005), and antipsychotics have been implicated in this attenuated cortisol response (Brenner et al., 2009). In support, the majority of studies found reduced baseline cortisol levels in schizophrenia patients who used antipsychotics (for review see (Walker, Mittal, & Tessner, 2008)), and increased cortisol levels are a predictor of antipsychotic treatment effects in schizophrenia (Ritsner et al., 2005). An advantage of studying the effects of antipsychotic use in BD1 patients is that – in contrast to schizophrenia patients - not all patients use antipsychotics. This enabled the direct comparison of the stress response between euthymic BD1 patients that use antipsychotics and those who do not. In our BD1 sample use of antipsychotics was not correlated to clinical characteristics, such as age of onset or number of manic and depressed episodes, indicating the blunted stress induced cortisol response in the antipsychotic group is not related to these clinical characteristics. Our data raise the question whether blunted cortisol response in BD1 disorder and schizophrenia may be only the result of antipsychotic drug use. This seems unlikely, since a blunted cortisol response was also present in medication naïve first episode psychotic patients (van Venrooij et al., 2012).

In contrast to the use of antipsychotics, use of lithium, anticonvulsants, benzodiazepines, and antidepressants did not significantly affect the endocrine and autonomic stress response in our sample. Even though acute administration of alprazolam or lamotrigine 28

Cortisol stress response in bipolar disorder

results in blunted stress-induced cortisol levels in healthy controls (Fries, Hellhammer, & Hellhammer, 2006; Kudielka et al., 2007; Maheu, Joober, & Lupien, 2005; Makatsori et al., 2004; Walker et al., 2008), such effects were not always detected (Lange et al., 2013; Watson, Gallagher, Ritchie, Ferrier, & Young, 2004; Roelofs et al., 2009). The chronic use of psychiatric medication in BD1 patients or the limited statistical power due to a rather small sample size may also have contributed to the fact that we found no significant effects for most medication groups in our study.

No significant effect was found of either age of onset or comorbid psychiatric disorders in BD1 patients. Nevertheless, an increased number of previous mood episodes was associated with overall reduced cortisol levels independent of stress. This adjustment of basal HPA axis activity after an increasing number of mood episodes is in line with the kindling hypothesis (Post, 1992), which suggests that during initial affective episodes, alterations in stress related systems occur which lead to an increased risk to develop subsequent mood episodes independent of stressors (for a recent review of the kindling hypothesis in BD1 see (Bender & Alloy, 2011)).

Traumatic stress plays an important role in the etiology and course of BD1 (Proudfoot et al., 2011) and also affects stress reactivity (Heim et al., 2000; Nater, Rohleder, Schlotz, Ehlert, & Kirschbaum, 2007; van Winkel, van Nierop, Myin-Germeys, & van Os, 2013).

Nevertheless, our data do not support a role for traumatic stress exposure or major life events on the acute stress response in BD1 patients. First, childhood maltreatment was similar across the experimental groups and the number of reported major life events was comparable between siblings and BD1 patients. Thus, even though levels of traumatic stress or major life events were comparable, differences were present in the endocrine and autonomic stress response in BD1 patients compared to healthy controls and siblings respectively. Moreover, adding traumatic stress scores or major life events as a covariate did not change any of the results.

Previous studies have reported increased basal cortisol levels in healthy offspring of BD1 patients (Ostiguy, Ellenbogen, Walker, Walker, & Hodgins, 2011), indicating that a genetic predisposition for BD1 may result in alterations in HPA axis functionality. However, we found no indication of changes in stress-induced cortisol and alpha-amylase levels in unaffected siblings of BD1 patients. Therefore, this study supports and extends an earlier study in which young offspring (age 13-21) displayed a similar cortisol stress response compared to healthy controls (Ellenbogen, Hodgins, Walker, Couture, & Adam, 2006).

In light of these data, an intermediate stress reactivity phenotype in unaffected siblings of BD1 patients is not likely.

Our findings should be interpreted in light of the existing data on HPA-axis activity in bipolar disorder. Several studies in which the HPA-axis was pharmacologically challenged showed that there is a reduced negative feedback in bipolar disorder patients, associated with hypercortisolism (Watson et al., 2004; Rush et al., 1996; Rybakowski & Twardowska, 29

2

CHAPTER 2

1999; Schmider et al., 1995). This disrupted feedback has further been related to decreased glucocorticoid receptor (GR) sensitivity in bipolar disorder (for review see (Spijker et al., 2011)). Reduced negative feedback would be expected to result in elevated basal cortisol levels and enhanced cortisol response after stress. Since we did not use any HPA-axis related pharmacological intervention, our data can only indirectly assess feedback and GR sensitivity. Nevertheless, from our data there is no indication for a reduced negative feedback in patients with bipolar disorder, as basal cortisol levels were unaffected and there was an overall blunted cortisol stress response. A possible explanation for the blunted cortisol stress response is chronic exposure to increased cortisol levels that gradually decrease the dynamic capacity of the HPA-axis. However, confirmation of this hypothesis would require prospective follow-up of a population with a high-risk for developing bipolar disorder. Importantly, antipsychotic use explained the variation in cortisol stress reactivity in our study to a large extent. This suggests that antipsychotic use rather than bipolar disorder per se may be involved in a gradual blunting of the cortisol response. Finally, in bipolar offspring, basal cortisol levels but not stressinduced cortisol levels were altered (Ellenbogen et al., 2006), indicating that these cortisol measures cannot be compared directly.

The results of the present study should be interpreted in the context of several potential limitations. Firstly, as a result of the cross-sectional nature of our study, we cannot determine the order of adaptation of the endocrine or autonomic system. A second limitation is that this study was not specifically set up to compare the effects of antipsychotic use in a BD1 population. It is possible that more severely ill and treatmentresistant BD1 patients were included who â&#x20AC;&#x201C; as a result â&#x20AC;&#x201C; used increased doses of antipsychotic drugs. However, antipsychotic use did not significantly correlate with clinical characteristics such as age of onset, comorbid psychiatric diagnosis, and number of mood episodes. Thirdly, alpha-amylase was included as a measure for autonomic stress reactivity whereas heart rate is a more common readout parameter. Nevertheless, alpha-amylase is a validated autonomic stress marker which consistently correlates to autonomic stress responses and noradrenaline release (Nater & Rohleder, 2009; Rohleder, Nater, Wolf, Ehlert, & Kirschbaum, 2004). Unfortunately, unaffected and unrelated siblings were more difficult to recruit, particularly as a result of daytime jobs, resulting in a smaller sample size compared to the other groups, a significant higher number of female participants and a significant higher age. Although we adjusted for this we cannot rule out residual confounding (G.A. Miller & Chapman, 2001). Strengths of the current study are the relatively large overall sample size, the inclusion of unaffected siblings to investigate a possible genotype effects, inclusion of both male and female participants, and assessment of current and lifetime diagnosis.

In conclusion, the current study shows a blunted cortisol stress response and increased autonomic stress response in euthymic BD1 patients. Our data indicate that antipsychotic use but not genetic loading is the most important determinant of the blunted cortisol stress response in BD1 patients. Even though it remains to be determined whether 30

Cortisol stress response in bipolar disorder

antipsychotics exert a protective or detrimental role in stress-related processes in BD1 patients, our findings underscore the possible importance of antipsychotic use on endocrine stress reactivity.

Funding body agreements and policies This study was made possible by a seed money grant of the Neuroscience and Cognition Initiative of the University Utrecht (http://www.neuroscience-cognition.org).

Contributors Authors CHV, MPM and LCH designed the study and wrote the protocol. Authors LCH, CHV and MPM carried out the study. Authors LCH, CHV, RSK, MPM and MJ discussed and managed the literature searches and analyses. Authors LCH, CHV and MPM undertook the statistical analysis, and author LCH wrote the first draft of the manuscript. All authors contributed to and have approved the final manuscript.

Conflicts of interest The authors declare no conflict of interest.

Acknowledgements The authors would like to thank Mrs. M. Litjens, Mrs. I. Maitimu and Mr. E. Strengman for their valuable technical support on this project.

31

2

Variation

F

ormal technique where material is altered during repetition.

CHAPTER 4 Genome-wide DNA methylation levels and altered cortisol stress reactivity following childhood trauma in humans Nature Communications 2016 Mar;21(7):10967 doi:10.1038/ncomms10967

L.C. Houtepen, C.H. Vinkers, T. Carrillo-Roa, M. Hiemstra, P.A. van Lier, W. Meeus, S. Branje, C.M. Heim, C.B. Nemeroff, J. Mill, L.C. Schalkwyk, M.P. Creyghton, R.S. Kahn, M. JoĂŤls, E.B. Binder, M.P.M. Boks

CHAPTER 4

Abstract DNA methylation likely plays a role in the regulation of human stress reactivity. Here we show that in a genome-wide analysis of blood DNA methylation in 85 healthy individuals, a locus in the Kit ligand gene (KITLG; cg27512205) showed the strongest association with cortisol stress reactivity (P=5.8 × 10−6). Replication was obtained in two independent samples using either blood (N=45, P=0.001) or buccal cells (N=255, P=0.004). KITLG methylation strongly mediates the relationship between childhood trauma and cortisol stress reactivity in the discovery sample (32% mediation). Its genomic location, a CpG island shore within an H3K27ac enhancer mark, and the correlation between methylation in the blood and prefrontal cortex provide further evidence that KITLG methylation is functionally relevant for the programming of stress reactivity in the human brain. Our results extend preclinical evidence for epigenetic regulation of stress reactivity to humans and provide leads to enhance our understanding of the neurobiological pathways underlying stress vulnerability.

Exposure to childhood trauma is a major risk factor for the development of almost all psychiatric disorders (Kessler et al., 2010), including depression (Burke, Davis, Otte, & Mohr, 2005), post-traumatic stress disorder (PTSD) (Petrowski et al., 2013) and schizophrenia (Jansen et al., 2000). Childhood trauma is also associated with blunted or increased activity of the hypothalamic–pituitary–adrenal (HPA) axis (Carpenter et al., 2007; Heim et al., 2000) (Supplementary Table S4.1 for a literature overview). These neuroendocrine changes may underlie the increased risk for psychiatric disorders across the life span. However, our understanding of how early life trauma can have such persistent detrimental effects is currently limited.

Epigenetic alterations may at least partially be involved in the lasting impact of childhood trauma. Preclinical studies have shown a consistent link between the early life environment, DNA methylation changes and adult stress reactivity and behaviour (J. Chen et al., 2012; Weaver et al., 2004). In humans, the long-term impact of traumatic stress on DNA methylation patterns is supported by several studies, which mainly focused on single genes (Vinkers et al., 2015), particularly on the glucocorticoid receptor gene that is pivotal for adequate HPA-axis functionality (Edelman et al., 2012; Klengel et al., 2013; McGowan et al., 2009; Oberlander et al., 2008; Perroud et al., 2011; Tyrka, Price, Marsit, Walters, & Carpenter, 2012). Even though hypothesis-driven studies have convincingly demonstrated a relation between traumatic stress and DNA methylation, the persistent detrimental influence of childhood trauma is unlikely to result from epigenetic modifications in a single gene (Mehta et al., 2013). Recently, two clinical studies investigated genome-wide methylation changes associated with childhood trauma (Labonte et al., 2012) and trauma exposure in PTSD (Mehta et al., 2013), but no study has investigated functional changes in endocrine stress reactivity using an unbiased genome-wide approach.

66

DNA methylation and cortisol stress reactivity

The main aim of this study is to provide an unbiased investigation of the role of DNA methylation in cortisol stress reactivity and its relationship with childhood trauma. To this end, we perform a genome-wide DNA methylation analysis for cortisol stress reactivity in healthy individuals. We identify a locus on the KITLG gene (cg27512205) that is not only associated to cortisol stress reactivity, but also partly mediates the relationship between childhood trauma and cortisol stress reactivity. Furthermore, we replicate the association between cortisol stress reactivity and methylation at the KITLG locus in two independent samples measuring methylation in either whole blood or buccal (crosstissue) DNA.

4

Figure 4.1 First we performed a genome-wide analysis of the association between cortisol stress reactivity and DNA methylation in the discovery sample (N=85). On the basis of the P value distribution, we sought replication of the top three loci in two independent samples (N=45/N=255) and replicated the negative association between the top KITLG locus and cortisol stress reactivity. Then we investigated the influence of childhood trauma on KITLG methylation and cortisol stress reactivity in the discovery and blood replication sample. On finding an association for childhood trauma with KITLG methylation and cortisol stress reactivity in the discovery sample and Caucasian of the blood replication sample, we examined whether the KITLG locus is a mediator for the blunted cortisol stress response after childhood trauma exposure.

Results DNA methylation and cortisol stress reactivity Our workflow is listed in Figure 4.1. After quality control, 385,882 DNA methylation loci were investigated for their association with cortisol stress reactivity (Supplementary Data S4.1 shows the results for the 22,425 loci with P values <0.05 in a linear regression 67

CHAPTER 4

model). Since none of the CpG sites survived adjustment for multiple testing, we selected the three loci that stood out in the P-value distribution of the genome-wide cortisol stress reactivity analysis (for QQ plot see Supplementary Figure S4.1;

Figure 4.2 The association between cortisol stress reactivity and DNA methylation at the KITLG/cg27512205 locus in the discovery (top panel), blood replication (middle panel) and cross-tissue replication (bottom panel) samples.

A linear regression line is plotted through the individual methylation values.

68

DNA methylation and cortisol stress reactivity

Table 4.1 Characteristics of the replicated kit ligand (KITLG) locus from the cortisol stress reactivity epigenome-wide association study (EWAS).

Cg number

27512205

Gene

KITLG

Location

Chr 12: 88579621 north-shore CpG island

Discovery mean methylation* (range)

0.15 (0.12–0.19)

Replication mean methylation* (range)

0.14 (0.11–0.18)

Cross-tissue mean methylation* (range)

0.09 (0.07–0.12)

Discovery association cortisol AUCi

B=−1,161, P=5.8 × 10−6†

Blood replication association cortisol AUCi

B=−1,040, P=0.006†

Cross-tissue association cortisol AUCi

B=−104, P=0.003†

Discovery association CTQ

B=0.005, P=0.04†

Blood replication association CTQ

B=0.001, P=0.146

4

AUCi, area under the curve (AUC) with respect to the increase; CTQ, Childhood Trauma Questionnaire. *Methylation in percentage (beta). †Denotes a nominal association in a linear regression model (P<0.05).

cg27512205 B=−1,162, P=5.8 × 10−6; cg05608730 B=−935, P=6.0 × 10−6; cg26179948 B=−1,009, P=8.0 × 10−6 in linear regression models) and were associated with childhood trauma (P<0.05 in linear regression models; Supplementary Table S4.2). The Kit ligand (KITLG) locus showed the strongest association with cortisol stress reactivity (cg27512205 chr12: 88579621; B=−1,162, P=5.8 × 10−6, empirical P value=2 × 10−6, model fit: F(3,81)=15.28, P=5.8 × 10−8, R2=0.34 in a linear regression model). This locus was also negatively associated with cortisol stress reactivity in two independent replication samples: a blood replication sample (N=45; B=−1,039, P=0.005, model fit: F(5,39)=5.6, P=0.0005, R2=0.35 in a linear regression model) and a cross-tissue replication sample that used mouth swaps to obtain buccal DNA (N=255; B=−104, P=0.004, model fit: F(3,251)=3.5, P=0.02, R2=0.03 in a linear regression model; Table 4.1; Figure 4.2). Visual inspection of cortisol stress reactivity measures pointed to five potential outliers in the discovery sample. However, Cook’s distance was lower than 1 in all analyses, suggesting that these observations did not affect the results (Supplementary Figure S4.2). Moreover, removal of these five potential outliers did not affect the association of cortisol stress reactivity with either childhood trauma (before removal: B=−14.6, P=0.007; after removal: B=−9.0, P=0.01 in a linear regression model) or KITLG methylation (before removal: B=−1,161, P=5.8 × 10−6; after removal: B=−617, P=7.0 × 10−4 in a linear regression model). To quantify the chance of finding these P values in the three independent samples, we used Fisher’s method to calculate the combined P value of the three samples, yielding an overall significance level of P=5.9 × 10−8 for the association between cortisol stress reactivity and methylation at the KITLG locus.

69

CHAPTER 4

Ethnicity and cortisol stress reactivity in the replication samples In light of the influence of current major depressive disorder (MDD) (Heim & Nemeroff, 2001) and ethnicity on cortisol stress reactivity (Hostinar, McQuillan, Mirous, Grant, & Adam, 2014; Melhem et al., 2015), we examined the contribution of these factors to our results in the blood replication sample, which included non-Caucasian individuals (Table 4.2). In the blood replication sample, cortisol stress reactivity was significantly lower in the African-American than the Caucasian individuals (B=−300, P=0.02, model fit: F(4,40)=4.1, P=0.007, R2=0.22 in a linear regression model; Supplementary Figure S4.3), though not related to current MDD (B=−19, P=0.89 in a linear regression model).

Therefore, we performed stratified analyses for ethnicity (Figure 4.3; Supplementary Note 4.1 and Supplementary Figure 4.4). In Caucasian individuals (N=17), there was a significant negative association between cortisol stress reactivity and methylation of the KITLG locus (B=−1,961, P=0.02, model fit: F(3,13)=2.8, R2=0.26 in a linear regression model; Supplementary Figure 4.4). Moreover, childhood trauma was significantly associated with blunted cortisol stress reactivity only in the Caucasian individuals (B=−7.9, P=0.003, model fit: F(3,13)=4.8, P=0.02, R2=0.42 in a linear regression model; Figure 4.3) and increased methylation at the KITLG locus (B=0.002, P=0.04, model fit: F(8,8)=2.3, R2=0.39 in a linear regression model). Inclusion of all ethnicities (N=45) rendered the relation between childhood trauma and cortisol stress reactivity nonsignificant (B=−3.7, P=0.09 in a linear regression model), as well as the association of childhood trauma with KITLG methylation (B=0.001, P=0.17 in a linear regression model).

Table 4.2 Sample description.

Characteristic

Discovery sample (N=85)

Blood replication sample (N=45)

Cross-validation sample (N=255)

Sex (% of female)

50.5

80

45

Age (mean in years, range)

33 (18 to 69)

28 (19 to 45)

17 (15 to 18)

Caucasian ethnicity (%)

100

38

100

Current MDD (%)

0

24

NA

56.8 (25 to 110)

NA

Childhood trauma (mean total score, range) 31.9 (24 to 63)

Cortisol stress reactivity (mean AUCi, range) 242.3 (−1,030 to 1,876) 1,185 (378 to 2,045) −37 (−426 to 313) On the basis of these findings in the blood replication sample, we examined the influence of ethnicity in the cross-tissue replication sample with a sensitivity analysis. The association between cortisol stress reactivity and methylation at the cg27512205 locus did not change after inclusion of non-Caucasian individuals and addition of ethnicity as a covariate (N=267; B=−101, P=0.004, F(4,262)=2.5, P=0.04, R2=0.02 in a linear regression model).

Mediation by KITLG methylation We carried out an in-depth analysis of the potentially mediating role of KITLG methylation in association between childhood trauma and cortisol stress reactivity (Figure 4.4). In the discovery sample childhood trauma was associated with lower cortisol stress reactivity (childhood trauma B=−14.7, P=0.007, model fit: F(3,81)=8.9, P=3.8 × 10−5, R2=0.22 in a linear regression model; Figure 4.3) and increased DNA methylation at the KITLG locus 70

DNA methylation and cortisol stress reactivity

4 Figure 4.3 Correlation between childhood trauma (total CTQ score) and cortisol stress reactivity (AUCi) in the discovery and replication sample. Colour indicates methylation levels at the cg27512205 (KITLG) locus. In the replication sample (right panel) differences in ethnicity are visualized. AUCi, area under the curve (AUC) with respect to the increase; CTQ, Childhood Trauma Questionnaire.

(childhood trauma B=0.0045, P=0.04, model fit: F(3,81)=1.8, P=0.15, R2=0.03 in a linear regression model). Moreover, the KITLG cg27512205 locus mediated 32% of the influence of childhood trauma on cortisol stress reactivity in the discovery sample (indirect effect=−4.8, P=0.04; total effect=−14.6, P=0.01; proportion mediated=0.32, P=0.05 in the mediation model; Figure 4.4).

Although in the blood replication sample childhood trauma was significantly associated with KITLG methylation and cortisol stress reactivity in the Caucasian individuals (N=17), the KITLG locus did not mediate the relationship between childhood trauma and cortisol stress reactivity (indirect effect=−1.8, P=0.31; total effect=−7.9, P<0.001; proportion mediated=0.22, P=0.31 in the mediation model). Moreover, mediation by the KITLG locus could not be established in the complete replication sample (N=45; indirect effect=−1.1, P=0.20; total effect=−3.7, P=0.09; proportion mediated=0.26, P=0.24 in the mediation model).

Figure 4.4 Model used to investigate mediation by the KITLG locus in the discovery sample. For graphical representation only, we did not add the sex and age covariates that were included in all statistical analyses.

71

CHAPTER 4

Influence of the age of trauma exposure In the blood replication sample, we found no evidence that age of onset of trauma affected the association between KITLG methylation, childhood trauma and cortisol stress reactivity. The age of first trauma exposure (both general and specific trauma) did not alter the association between childhood trauma and cortisol stress reactivity (age of general trauma × childhood trauma interaction B=1.9, P=0.16 in a linear regression model; age of specific trauma × childhood trauma interaction B=1.3, P=0.18 in a linear regression model). Similarly, the association between KITLG DNA methylation and childhood trauma was not influenced by the age of childhood trauma (age of general trauma × childhood trauma interaction B=−0.0006, P=0.26 in a linear regression model; age of specific trauma × childhood trauma interaction B=−0.0006, P=0.14 in a linear regression model). In support, adding age of trauma as a covariate interacting with childhood trauma did not improve linear regression model fits (measured as a reduction of residual sum of squares) for either cortisol stress reactivity as outcome (age of general trauma P=0.18; age of specific trauma P=0.12) or KITLG DNA methylation as outcome (age of general trauma P=0.38; age of specific trauma P=0.53). In the discovery sample, adult trauma (Life Stressor Checklist-Revised (LSC-R) score mean=3.4, s.d.=2.1) was not significantly associated with cortisol stress reactivity (B=−23, P=0.28 in a linear regression model) or DNA methylation of the KITLG locus (B=0.005, P=0.60 in a linear regression model), indicating that adult trauma has less impact on cortisol stress reactivity and KITLG methylation levels than childhood trauma.

Blood–brain correlation of the KITLG locus Blood cg27512205 methylation levels were positively correlated to cg27512205 methylation levels in the prefrontal cortex (PFC; r=0.293, P=0.01 in a linear regression model) and negatively to cg27512205 methylation levels in the superior temporal gyrus (STG; r=−0.267, P=0.02 in a linear regression model; Supplementary Figure S4.5). No significant correlations were found for the entorhinal cortex and the cerebellum (P<0.05 in a linear regression model; Supplementary Figure S4.5).

Histone mark analysis of the KITLG locus To further analyse the potential functional relevance of the DNA containing the cg27512205 probe, we compared its location to the location of epigenomic signatures that typically cover functional non-coding DNA (Kundaje et al., 2015). We identified enrichment for histone 3 lysine 27 acetylation (H3K27ac), overlapping the cg27512205 probe location in lateral hypothalamus tissue (Figure 4.5). This histone mark was previously found to be selectively present at active gene regulatory DNA suggesting that our probe is located in a functional sequence (Creyghton et al., 2010).

Regional analysis of KITLG gene methylation Cg27512205 was the only KITLG probe (out of eighteen present on the methylation array) associated with cortisol stress reactivity in all three independent samples (Figure 4.5).

72

DNA methylation and cortisol stress reactivity

4

Figure 4.5 Overview of the 1,500 base pair area downstream and upstream of the cg27512205 KITLG locus. The top panel contains the −log P values for the association between DNA methylation and cortisol stress reactivity in the discovery (blue, N=85), blood replication (black, N=45) and cross-tissue replication (magenta, N=255) samples per locus (total of 14 loci in the depicted area). The other panels indicate the presence of coding exons (blue blocks) and non-coding introns (grey line) of the KITLG gene (second panel), location of a CpG island (third panel) and the percentage of G (guanine) and C (cytosine) bases (fourth panel) in the area around the cg27512205 locus extracted from the UCSC website (Kent et al., 2002) with the Gviz R package (Hahne et al., 2015). The bottom panel indicates the location of a H3K27ac histone modification in lateral hypothalamus tissue (data from (Creyghton et al., 2010)). Our locus of interest is shaded across all panels by a red rectangle. In the top panel all points above the horizontal dashed grey line are nominally associated (P<0.05) loci in a linear regression model. chr, chromosome.

KITLG-related methylation network analyses By using weighted gene co-expression network analysis (WGCNA), we derived 40 consensus modules, based on the correlation patterns among probes in the discovery and cross-tissue samples. These 40 consensus modules were significantly related to cortisol stress reactivity in the discovery sample (multiple analysis of covariance (MANCOVA) Pillai’s trace=0.72, F(40,37)=2.4, P=0.004) and borderline significant in the cross-tissue replication sample (MANCOVA Pillai’s trace=0.21, F(40,212)=1.4, P=0.051).

Subsequently, we analysed the module containing the KITLG probe (the ‘red’ module; Figure 4.6 and Supplementary Data 4.2). This red module was significantly associated with cortisol stress reactivity in both the discovery (F(1,76)=4.4, P=0.04 in the follow-up analysis of variance (ANOVA)) and the cross-tissue replication sample (F(1,251)=4.3, P=0.04 in the follow-up ANOVA).

73

CHAPTER 4

To further understand the biology of the red module, we examined enrichment for gene ontology (GO) terms and the potential regulation by microRNAs (miRNAs) of the KITLG network within the red module. The red module contained 21,211 probes linked to 9,494 genes, which were significantly enriched for GO terms related to metabolism and regulation of transcription (Supplementary Data 4.2). Two thousand seven hundred fortyeight of these probes were nominally associated with cortisol stress reactivity in the discovery sample. Selection of the 5% strongest connections yielded a 21-gene network around the KITLG probe (Figure 4.6). With the webGestalt tool, we found that the 21-gene network around KITLG is a preferred target for three miRNAs: miR449 (genes COL12A1, SHKBP1 and KITLG FDRhypergeometric (FDR, false discovery rate)=0.0012), miR23A/ miR23B (genes EYA1, HMGN2 and KITLG FDRhypergeometric=0.0018) and miR9 (genes COL12A1, CCDC43 and KITLG FDRhypergeometric=0.0019; Supplementary Table S4.3).

The entire red module (containing 9,494 of 19,815 genes) was enriched for genes related to these three miRNAs (miR449 Fisher’s exact test, odds ratio (OR)=1.4, P=0.0015, FDR=0.009, miR23A/miR23B Fisher’s exact test, OR=1.4, P=7.3 × 10−6, FDR=9.8 × 10−5 and miR9 Fisher’s exact test, OR=1.3, P=9.5 × 10−5, FDR=0.0006). Several other methylation modules were also enriched for these miRNAs (10 modules enriched for

CCDC43

FAM198B

DPF3 MRPL22 IL17RD

TGDS

INTS6 EXO1 MCOLN1

SHKBP1

KITLG TES

KCNC3

ACAA1

REV3L

COL12A1

HMGN2

EYA1

TARSL2

LOC344595 TPX2

NUP93

Figure 4.6 Graphical depiction of the connection between the KITLG-related probe and its direct neighbours within the red module (Supplementary Data 4.2). The node size indicates the association with cortisol stress reactivity, while the width of the lines indicates the connection strength between the nodes.

74

DNA methylation and cortisol stress reactivity

miR449 Fisher’s exact test FDR<0.05; 19 modules enriched for miR23A/miR23B Fisher’s exact test FDR<0.05 and 20 modules enriched for miR9 Fisher’s exact test FDR<0.05; Supplementary Table S4.4–S4.6).

Discussion By using a unique and unbiased approach, we analysed the relationship between DNA methylation levels and cortisol stress response in three independent samples (total N=385) using an experimental stress paradigm. Genome-wide analysis of the association of whole blood DNA methylation with cortisol stress reactivity in the discovery cohort identified a locus (cg27512205) in the KITLG gene. This locus was also associated with cortisol stress reactivity in two independent samples: one replication sample in blood and another replication sample using buccal cell DNA. Even though the observed DNA methylation differences in KITLG were small, the impact was considerable since the model accounted for 35% of the variation in cortisol stress reactivity in the discovery and the blood replication sample. Moreover, KITLG methylation was a mediator in the association between childhood trauma and cortisol stress reactivity. The identified methylation locus (cg27512205, chr12: 88579621) is located on the north shore of a CpG island near the KITLG gene (Figure 4.5). This gene codes for a ligand of the C-kit receptor and is involved in cellular developmental processes such as hematopoiesis by activating the C-kit receptor (Lennartsson & Ronnstrand, 2012). The involvement of KITLG protein in stress-induced HPA-axis activity is biologically plausible, because KITLG levels correlate with glucocorticoid receptor expression in response to in vitro stress-induced erythropoiesis (Varricchio et al., 2012). In mice, early life stress increased both anxiety and KITLG expression in the hippocampus (Suri, Bhattacharya, & Vaidya, 2014). Also, C-kit-positive hematopoietic progenitors proliferate in response to chronic stress, resulting in higher levels of inflammatory leukocytes in mice (Heidt et al., 2014).

Recent studies highlight the complexity of epigenetic regulation and indicate that the interplay between DNA methylation and enhancers may trigger cascades of transcriptional events that are highly relevant for neurodevelopment (Kundaje et al., 2015; Thakurela, Sahu, Garding, & Tiwari, 2015; Vermunt et al., 2014). The identified KITLG locus is located in a region enriched for the histone mark H3K27ac in the human hypothalamus, which is pivotal for cortisol stress reactivity (Creyghton et al., 2010); this supports a biologically relevant and functional signal. H3K27ac is typically found at active regulatory DNA, such as enhancer and promoter regions, and is associated with a more open chromatin structure indicative of gene transcription (Creyghton et al., 2010; Kundaje et al., 2015; Thakurela et al., 2015; Vermunt et al., 2014). Interestingly, the cg27512205 CpG is the only KITLG probe located both in the H3K27ac-enriched region and on the shore of a CpG island. As CpG island shores are frequently linked to DNA methylation differences that affect gene transcription and expression (Irizarry et al., 2009), the co-occurrence of these two epigenetic signatures suggests that methylation differences at this CpG 75

4

CHAPTER 4

location could alter gene regulatory DNA near KITLG. Interestingly, exposure of young animals to early life stress altered histone modifications at the KITLG promoter, specifically an increase in H3K9ac and a decrease of the repressive H3K9me; this change is associated with increased hippocampal KITLG expression (Suri et al., 2014). Another potential insight into the biological mechanisms related to KITLG methylation comes from our coexpression network analyses showing that KITLG is part of a gene network enriched for genes regulated by miRNAs 449 (miR449), 23A/23B (miR23A/miR23B) and 9 (miR9).

Notably, in rodents two of these miRNAs were previously linked to stress system functionality (Nemoto, Kakinuma, & Shibasaki, 2015; Rinaldi et al., 2010). Specifically, miR449 is involved in the regulation of corticotropin-releasing factor type 1 receptor in the anterior pituitary and HPA-axis activation (Nemoto et al., 2015). Also, miR9 is upregulated in the frontal cortex of mice in response to acute stress (Rinaldi et al., 2010).

The fact that the KITLG locus was significantly related to cortisol stress reactivity in three independent samples is noteworthy considering the substantial differences in study characteristics between these samples. The blood replication sample was smaller, ethnically diverse and included individuals selected for either low or high levels of childhood trauma. In addition, cortisol was measured in blood and, even though differences in cortisol stress reactivity can be detected in both blood and saliva (Petrowski et al., 2013), this may have contributed to the difference in cortisol values between the discovery and blood replication sample. The cross-tissue replication sample measured methylation in DNA extracted from buccal cells in relatively young participants (15â&#x20AC;&#x201C;18 years) and used a public speaking task without an arithmetic stressor. Despite these differences, the KITLG locus was in all cases related to cortisol stress reactivity, which supports the robustness of the observation. This is further supported by the fact that a significant association between cortisol stress reactivity and KITLG methylation was observed in buccal and blood DNA. Some recommend buccal samples for population epigenetic studies, as they contain more hypomethylated DNA regions, which tend to cluster around disease associated single-nucleotide polymorphisms (SNPs) (Lowe et al., 2013); others, however, argue that demographic factors may be better reflected in blood DNA methylation patterns (Jiang et al., 2015). Blood and buccal cells are peripheral tissues and do not necessarily reflect changes in DNA methylation in the central nervous system.

However, there are several reasons why KITLG methylation in peripheral tissues can be informative for the neurobiological mechanisms underlying cortisol stress reactivity. First, cortisol is released into the periphery by the pituitary and is known to affect multiple tissue types. Second, DNA methylation co-expression network analyses with the module containing the KITLG probe (red module) demonstrated that this module was associated with cortisol stress reactivity in both tissue types, suggesting that a broader methylation network around KITLG is biologically relevant for stress reactivity. Previous reports on a variety of traits such as age (Horvath et al., 2012) also indicate that methylation coexpression networks are stable indicators for epigenetic regulation across tissue types.

Third, HPA-axis genes are abundantly expressed in peripheral blood mononuclear cells (Gladkevich et al., 2004). Peripheral changes in methylation may therefore at least partially 76

DNA methylation and cortisol stress reactivity

be a proxy of epigenetic processes in the brain. Indeed, previous studies have shown that childhood trauma-related changes in methylation obtained from peripheral blood mononuclear cells were significantly enriched for central nervous system pathways (Mehta et al., 2013). From the four brain areas that we examined, significant correlations with blood methylation were found in the PFC and the STG, which are biologically relevant brain regions for stress. Thus, a wealth of literature points to the PFC as a pivotal regulator of the stress response (for review see (McKlveen, Myers, & Herman, 2015)); in agreement, altered cortisol stress responses have been found after lesions in the PFC (Buchanan et al., 2010). Regarding the STG: a recent meta-analysis supported a link of this area to stress susceptibility (Kogler et al., 2015). In light of the opposing bloodâ&#x20AC;&#x201C;brain correlations, it may be hypothesized that the STG and PFC have opposing roles in the regulation of cortisol stress reactivity, but this cannot be inferred from the present study and warrants further research.

It is particularly interesting that epigenetic regulation of the stress response was found to be related to childhood trauma. In the discovery sample, increased levels of childhood trauma were significantly related to blunted cortisol stress reactivity and higher methylation at the KITLG locus (Figure 4.4). In the blood replication sample, a similar result was only obtained in Caucasian individuals. In the complete blood replication sample, this association was (just) not significant, suggesting that ethnic diversity influences analyses on the relationship between childhood trauma and cortisol stress reactivity. This may be the result of overall lower cortisol stress reactivity in Afro-American individuals (Hostinar et al., 2014; Melhem et al., 2015). Unlike the discovery sample, there was no evidence that KITLG methylation is a mediator of the association between childhood trauma and cortisol stress reactivity in either the entire (N=45) or Caucasianonly (N=17) replication sample. Overall non-replication of the mediation analysis may be due to a more heterogeneous ethnicity, smaller sample sizeâ&#x20AC;&#x201D;below the recommended N=50 (Fritz & MacKinnon, 2007)â&#x20AC;&#x201D;and unfavourable distribution of childhood trauma due to the inclusion of individuals based on either low or high levels of childhood trauma (Supplementary Figure S4.6).

The relationship between childhood trauma and a blunted cortisol response in the present study is in concordance with some but not all of the published literature (16 studies; Supplementary Table S4.1). For example, in the discovery sample and in Caucasian individuals from the replication sample, the explained variance of cortisol stress reactivity by childhood trauma was 34 and 26%, respectively. This is comparable to the 30% explained variance in the study of (Carpenter et al., 2007) who also included healthy individuals without a psychiatric diagnosis. Previous studies have pointed to specific periods during which individuals are particularly sensitive to trauma exposure (Lupien, McEwen, Gunnar, & Heim, 2009). However, in the blood replication sample, we did not find any indications that our KITLG results were related to age of onset of childhood trauma. Moreover, in the discovery sample, cortisol stress reactivity and KITLG methylation were not significantly related to traumatic experiences in adult life, suggesting that early 77

4

CHAPTER 4

life is a more sensitive period for the persistent impact of trauma on HPA-axis activity.

In conclusion, this study shows that altered stress reactivity following childhood trauma in humans is related to altered DNA methylation levels at the KITLG locus. Identification of such epigenetic marks may help to identify inter-individual differences in susceptibility to traumatic stress in early life and elucidate the neurobiological pathways underlying its long-lasting detrimental effects.

Methods Study population For discovery, 85 healthy individuals were recruited from the general population at the University Medical Center, Utrecht, The Netherlands (see Table 4.1 for sample characteristics). Participants had three or more Dutch grandparents, were not taking any prescription medication and had not been enrolled in stress-related research before participation. The absence of any mental or physical disorder was confirmed by an independent rater in an interview according to the Mini-International Neuropsychiatric Interview (MINI) plus criteria (Sheehan et al., 1998). Participants abstained from heavy meals, drinks other than water or heavy exercise for at least 2â&#x20AC;&#x2030;h before the study protocol.

Current use of psychoactive substances (amphetamines, 3,4-methylenedioxymethamphetamine (MDMA), barbiturates, cannabinoids, benzodiazepines, cocaine and opiates) was assessed by self-report and verified with a urine multi-drug screening device (InstantView).

The blood replication sample consisted of 45 individuals who were part of a larger study, the Conte Center Study for the Psychobiology of Early-Life Trauma (MH58922) and included some individuals with exposure to childhood trauma before the age of 13 years and with/without a diagnosis of MDD (Heim et al., 2009) (Table 4.1). Depressed mood was assessed with the 21-item self-report Beck Depression Inventory (BDI) (Beck, Ward, Mendelson, Mock, & Erbaugh, 1961). Eleven subjects with a score above 9 on the BDI were classified as current MDD. Exclusion criteria include: current medical illness, lifetime diagnosis of psychosis or bipolar disorder, alcohol or substance abuse within 6 months or eating disorders within the last year. None of the participants was receiving psychiatric treatment or currently taking psychiatric medication. Heavy smokers (>20 cigarettes per day) were excluded. The blood replication sample was ethnically more diverse and included predominantly Afro-American (N=23, 52%) and Caucasian (N=17, 38%) individuals.

The cross-tissue validation sample consisted of 255 healthy adolescents participating in the longitudinal RADAR-Y (Research on Adolescent Development and Relationships Young cohort) study (Table 4.1). DNA samples were collected for 414 subjects of whom 314 completed the stress task. Exclusion criteria were adolescents who currently received prescription medication (N=42) or were of non-Caucasian ethnicity (N=17).

78

DNA methylation and cortisol stress reactivity

All studies were approved by an ethical review board and performed according to the ICH guidelines for Good Clinical Practice and the Declaration of Helsinki. More specifically, the discovery and RADAR-Y studies were approved by the medical ethical committee of the University Medical Center Utrecht, while the Conte Center Study for the Psychobiology of Early-Life Trauma was approved by the Institutional Review Board of Emory University School of Medicine. For all studies, participants gave written informed consent before inclusion and were financially compensated. The data used to replicate the findings of the discovery sample are available in Supplementary Data 4.3.

Stress procedure Both the discovery and blood replication sample used a version of the Trier Social Stress Test (TSST) as a stress induction task, consisting of a public speaking test (PST) and subsequent arithmetic task. In the discovery study, the TSST was adapted to a group format and carried out as previously described (Vinkers et al., 2013). Cortisol levels were measured with an in-house radioimmunoassay in eight saliva samples (Salivette) collected over a time period of 90 min (Supplementary Figure S4.7) (Vinkers et al., 2013). In the blood replication study, an individual TSST was conducted (Pace et al., 2006). Cortisol levels were examined in blood samples obtained from an indwelling venous catheter during eight 15-min intervals (15 min before the TSST up to 90 min afterwards). Blood was collected into chilled EDTA-coated Monovette and centrifuged immediately before cortisol was measured with a radioimmunoassay. In the cross-tissue sample, a PST based on the Leiden-PST was used (Westenberg et al., 2009). Cortisol levels over a time period of 45 min were measured with a radioimmunoassay in seven saliva samples obtained by passive drooling into a plastic tube (0.5 ml SaliCap) (Westenberg et al., 2009).

In all studies, participants were tested in the afternoon to mitigate the influence of diurnal variations in cortisol secretion and the area under the curve (AUC) with respect to the increase (AUCi) of cortisol was calculated based on the consecutive data points as previously described (Pruessner, Kirschbaum, Meinlschmid, & Hellhammer, 2003).

Trauma exposure In the cross-tissue sample, childhood trauma exposure was not measured. In the discovery and blood replication sample, childhood trauma exposure was assessed using the short version of the Childhood Trauma Questionnaire (CTQ) (D. P. Bernstein et al., 2003). The validity of the 25 clinical CTQ items, including a Dutch translation, has been demonstrated in clinical and population samples (D. P. Bernstein et al., 2003; Thombs, Bernstein, Lobbestael, & Arntz, 2009). In the discovery sample, one translated item (‘I believe I was molested’) was excluded as this translation was found to be an invalid indicator of childhood sexual abuse in a previous validation study (Thombs et al., 2009).

In the blood replication sample, age of onset of general trauma and age of onset of any specific childhood trauma exposure were assessed with the Early Trauma Inventory (Bremner, Vermetten, & Mazure, 2000). In the discovery sample, data on adult trauma (>16 years) were available for 69 of 85 individuals who completed the LSC-R self-report questionnaire (Wolfe et al., 1996).

79

4

CHAPTER 4

DNA methylation measurement In all studies genome-wide DNA methylation levels were assessed using Illumina Infinium HumanMethylation450K BeadChip (Illumina) arrays. X chromosome, Y chromosome and nonspecific binding probes were removed (Y. A. Chen et al., 2013). Failed probes were excluded based on a detection P value >0.001 and bead count <5. In addition, probes with SNPs of minor allele frequency >5% within 10 base pairs of the primer were excluded after constructing ancestry estimates based on their principal components as proposed by (Barfield et al., 2014) (list of CpG sites is available at http://genetics.emory.edu/ research/conneely/annotation-r-code.html). In the discovery sample 385,882 DNA methylation probes survived quality control and were used for further genome-wide analysis. Finally, in all studies DNA methylation data were normalized and batch effects were removed based on inspection of the association of the principal components of the methylation levels with plate, sentrix array and position using multivariate linear regression and visual inspection of heat maps (see Supplementary Figures S4.8-S4.10 and Supplementary Notes S4.1-S4.3 for the model summaries per sample). Quality control and analysis were performed with the wateRmelon (Schalkwyk et al., 2013), the Minfi (Aryee et al., 2014), the Limma (Smyth, 2005) and the sva packages (Leek, Johnson, Parker, Jaffe, & Storey, 2014) from the Bioconductor platform in R. More specifically, in the discovery sample whole blood was obtained before the stress test and DNA was extracted with the Gentra Puregene kit (Qiagen, Valencia, CA, USA).

DNA concentration and integrity was assessed using Ribogreen and Bioanalyzer.

Bisulphite conversion was conducted with Zimo kits (Zymo Research, Orange, CA, USA) using standard procedures. Samples were distributed over the twelve 450K arrays according to gender and age to reduce batch effects to the minimum. Intensity read outs, quality control parameters and methylation measures were obtained from the GenomeStudio software. In total, 20,845 probes with failed detection in more than 1% of the participants or <5 beads in 5% of samples were excluded. All samples were included as none of the samples had more than 1% of probes failed (Schalkwyk et al., 2013). The data were normalized to remove systematic differences in overall signal distribution due to probe design bias using the Beta MIxture Quantile dilation (BMIQ) normalization (Teschendorff et al., 2013) as implemented in the wateRmelon package (Schalkwyk et al., 2013). After removing batch effects of Sentrix array and position with the Combat procedure from the sva package no batches were apparent (Johnson et al., 2007) (Supplementary Figure S4.8). Finally, cell-type composition estimates (another wellknown potential confounder in whole blood samples) were calculated using a Minfibased implementation of the Houseman algorithm (Houseman et al., 2012) with FACSsorted DNA methylation data as a reference set and related to DNA methylation levels (see Supplementary Figure S4.8 and Supplementary Note S4.2 for model summary).

In the replication study, blood was obtained before the stress test. DNA was extracted and genome-wide DNA methylation levels were assessed using Illumina 450K DNA methylation arrays as previously published (Klengel et al., 2013; Mehta et al., 2013).

Intensity read outs, normalization, cell-type composition estimation, beta and M values were obtained using the Minfi package (version 1.10.2) in Bioconductor (Aryee et al., 80

DNA methylation and cortisol stress reactivity

2014). In total, 233 failed probes were excluded based on a detection P value >0.01 in at least 75% of the samples. We removed probes located within 10 bp from a SNP with a minor allele frequency of ≥0.05 in any of the populations represented in the sample. The data were then normalized with functional normalization; an extension of the quantile normalization procedure implemented in the Minfi R package (Fortin et al., 2014). Sentrix array and position-related batch effects were identified by linear regression analysis with the first principal component of the methylation levels and visual inspection of principal component analysis (PCA) plots. Batch effect removal was performed with the Combat procedure as implemented in the sva package (Johnson et al., 2007).

In the cross-tissue study, buccal swaps were obtained before the PST and DNA was extracted using the chemagic saliva isolation kit on a Chemagen Module I workstation (Chemagen Biopolymer Technologie AG, Baesweiler, Germany). Samples were equally distributed over thirty-five 450K arrays according to gender and age to reduce batch effects to the minimum. Intensity read outs, quality control parameters and methylation measures were obtained using the methylumi package (version 2.14.0) in Bioconductor (Davis & Bilke, 2010). In total, 3,574 probes with failed detection in more than 1% of the participants or <5 beads in 5% of samples were excluded as were three samples where more than 1% of probes failed (Schalkwyk et al., 2013). The three excluded samples were already identified before data analysis, as there were technical difficulties during bisulphite conversion. Therefore, technical duplicates were present on the array and we obtained high-quality methylation data for all participants in the cross-tissue study. The data were normalized to remove systematic differences in overall signal distribution due to probe design bias using BMIQ (Teschendorff et al., 2013) as implemented in the wateRmelon package (Schalkwyk et al., 2013). After removing batch effects related to Sentrix array and position with the Combat procedure from the sva package no remaining batches were apparent (Johnson et al., 2007) (see Supplementary Note 4.3 for model summaries; Supplementary Figures S4.9 and S4.10).

Blood-brain sample For the identified KITLG methylation locus, we compared blood–brain correlation in a database containing 78 individuals (40–105 years old) described in more detail in a previous study (Hannon, Lunnon, Schalkwyk, & Mill, 2015). In short, whole blood samples were collected before death, as well as PFC (N=74), entorhinal cortex (N=69), STG (N=75) and/or cerebellum (N=69) tissue post mortem. Approximately, 500 ng DNA from each sample was extracted and assessed using 450K Illumina DNA methylation arrays. Raw signals were extracted with Illumina GenomeStudio software and further pre-processed with the methylumi and wateRmelon (Schalkwyk et al., 2013) packages in R. Initial quality control checks were performed using functions in the methylumi package to assess concordance between reported and genotyped gender. Non-CpG SNP probes on the array were also used to confirm that all four brain regions and matched bloods were sourced from the same individual. Array data for each of the tissues was normalized separately using the dasen function from the wateRmelon (Schalkwyk et al., 2013) package and initial analyses were performed separately by tissue. The effects of age and 81

4

CHAPTER 4

sex were regressed out before blood and brain methylation levels were compared using linear regression modelling as previously described (Hannon et al., 2015).

Histone mark in hypothalamus H3K27ac data determined with ChIP-sequencing analysis on post-mortem hypothalamus tissue was downloaded from a previous study (Creyghton et al., 2010) and overlaid with our probe data. Enrichment was found at the identified cg27512205 probe in the KITLG locus and visualized using the Gviz R package (Hahne et al., 2015).

Statistical analyses All statistical analyses were carried out in R version 3.2.2 (R-Core-Team, 2014). All regression modelling was performed with the Limma64 package and outliers were defined using Cookâ&#x20AC;&#x2122;s distance with a cutoff value of 1. We report the regression coefficient (B) and P value for all analyses. If relevant individual parameters have a significant association (P<0.05), we also report the percentage of variance explained by the complete model (R2) with the corresponding F statistic and P value. Beta values of methylation were used for graphical display, but analyses were conducted with M values (log2 ratio of methylation probe intensity) for better statistical validity (Du et al., 2010). To account for potential confounding by blood cell type in the discovery and whole blood replication sample (Supplementary Note 4.1 and Supplementary Note 4.2), we calculated standardized residuals for the M values using cell count estimates from the Houseman algorithm as the independent variables. Because cortisol stress reactivity and methylation levels may vary with age and sex (Boks et al., 2009; Hostinar et al., 2014), both factors were included as covariates in all analyses. In previous studies, both current MDD (Heim & Nemeroff, 2001) and ethnicity (Hostinar et al., 2014; Melhem et al., 2015) influenced cortisol stress reactivity. In light of the ethnic diversity and current MDD individuals in the blood replication sample, we investigated whether cortisol stress reactivity was associated with current MDD or ethnicity. If there was a significant association with either current MDD or ethnicity, stratified analyses were performed and the determinant (current MDD or ethnicity) was included as covariate in all analyses of the complete blood replication sample. In the Caucasian discovery sample population, stratification did not play a role, therefore the methylation-based population principal components as proposed in the Barfield study (Barfield et al., 2014) were not included as covariate (Supplementary Figure S4.8). In the cross-tissue replication sample, non-Caucasian individuals (N=12) were excluded a priori to make the cross-tissue replication sample more comparable to the discovery sample regarding ethnicity. Moreover, we also performed a sensitivity analysis by including the 12 non-Caucasian individuals (N=267) and adding ethnicity as a covariate.

To investigate DNA methylation and cortisol stress reactivity, we first performed a genome-wide association analysis in the discovery sample with cortisol stress reactivity (AUCi) as the outcome and DNA methylation, age and sex as the determinants in a linear regression model. We considered a FDR at the 0.05 level as genome-wide significant.

Visual inspection of the QQ plot in the discovery sample did not indicate a deviant 82

DNA methylation and cortisol stress reactivity

distribution of P values (Îť=1.10; Supplementary Figure S4.1). On the basis of the P value distribution, we sought replication for loci of which the strength of the association stood out. In the two independent replication samples, we implemented the same linear regression model with cortisol stress reactivity as dependent and methylation, age and sex as indicators. On the basis of the association with cortisol stress reactivity (Supplementary Note 4.1), ethnicity was added as covariate to the model in the whole blood replication sample. Finally, we interrogated potential type-I error inflation for the replicated methylation loci in the discovery sample by calculating an empirical P value based on 1,000,000 label-swapping permutations.

To examine the influence of age at trauma exposure in the discovery sample, adult trauma was also associated with either cortisol stress reactivity or DNA methylation of the identified locus using linear regression models with sex and age as covariates. In the replication sample, we also investigated whether age of onset of childhood trauma modified its relationship with either cortisol stress reactivity or DNA methylation of the identified locus. To this end, the interaction between age of trauma with childhood trauma levels was examined in linear regression models with DNA methylation or cortisol stress reactivity as outcomes and sex, age and ethnicity as covariates.

In all samples, we investigated the association between cortisol stress reactivity and DNA methylation of the other probes located on the KITLG gene using a linear regression model with age and sex as covariates. On the basis of the association with cortisol stress reactivity (Supplementary Note 4.1), ethnicity was added as covariate to the model in the whole blood replication sample.

We hypothesized that DNA methylation of the KITLG locus mediates the association between childhood trauma and cortisol stress reactivity. Therefore, we investigated the association between childhood trauma and cortisol stress reactivity using a linear regression model with sex and age as covariates. Next, the association between childhood trauma and methylation was investigated. On the basis of the association with cortisol stress reactivity (Supplementary Note S4.1), ethnicity was added as covariate to the model in the whole blood replication sample.

To quantify the average causal mediation effect of DNA methylation, we performed a model-based mediation analysis in the discovery and replication sample, using the mediation package in R (Tingley et al., 2014). This method uses the information of two linear regression models: (1) DNA methylation as outcome and childhood trauma, age and sex as determinants and (2) cortisol stress responsivity as the outcome variable and DNA methylation, childhood trauma, age and sex as determinants. The algorithm estimates the presence of mediation (average causal mediation effect/indirect effect) as well as the proportion of the link between childhood trauma and cortisol stress reactivity mediated by DNA methylation by using a quasi-Bayesian Monte Carlo method with 10,000 simulations.

Finally, in the discovery and cross-tissue samples, weighted gene co-expression network analysis was performed with the WGCNA package in R to identify consensus methylation clusters (Langfelder & Horvath, 2008). We did not use the blood replication sample for the identification of consensus methylation clusters, since the sample size was relatively 83

4

CHAPTER 4

small with a heterogeneous background with regard to ethnicity and current depression.

The consensus clusters containing loci of interest were further characterized based on their relationship with cortisol stress reactivity and biological processes. To link the methylation cluster to biological processes a GO-term-enrichment analysis was conducted with the missMethyl (Phipson, Maksimovic, & Oshlack, 2015) package in R. First, all loci surviving quality control were mapped to genes. Next, the relationship between the number of probes per gene and the probability of selection was calculated with an adapted GOseq method by (M. D. Young, Wakefield, Smyth, & Oshlack, 2010).

Finally, the probes in the module of interest were selected and the other loci used as a reference set to perform a modified version of a hypergeometric test to incorporate the over-representation of the selected genes in each GO category.

To examine the association between the WGCNA methylation modules and cortisol stress reactivity in the discovery and cross-tissue replication samples, we performed MANCOVA with the participant score on the methylation consensus modules as outcomes and cortisol stress reactivity, sex and age as determinants. For the discovery sample, cell count estimates were also added as covariates. Separate follow-up ANOVA analyses were carried out for the individual modules containing loci of interest, but only if the methylation modules were significantly associated to cortisol stress reactivity.

To establish the connections of the replicated loci within these methylation clusters, we selected all probes with a nominal association to cortisol stress reactivity in the module containing a replicated locus. Then connection strength was established based on the correlation between individual loci and only the top 5% strongest connections were used for plotting and enrichment analysis. On the basis of these criteria, the neighbours of the replicated locus were selected and evaluated for miRNA regulation using the WebGestalt tool based on the miRTarBase website (http://mirtarbase.mbc.nctu.edu.tw/).

Enrichment for genes related to a particular miRNA was investigated with a Fisher’s exact test for the presence of the selected genes present in the module of interest compared to the presence of the selected genes present in all other modules.

Acknowlegements We acknowledge Ruben Van ‘t Slot for his practical assistance with the methylation analysis, Dr Eilis Hannon for her invaluable assistance with the blood–brain correlations and Josine Vaes for her help with the identification of the childhood trauma studies in the literature. The RADAR cohort has been financially supported by main grants from the Netherlands Organisation for Scientific Research to RADAR PI’s and the CID consortium. The blood replication sample study was supported by the Silvio O. Conte Center for the Psychobiology of Major Psychiatric Disorders (National Institutes of Health Grant MH58922). C.H.V. and M.P.M.B. acknowledge a seeding grant of the Brain and Cognition programme of the University Utrecht. Statistical analyses were carried out on the Genetic Cluster Computer (http://www.geneticcluster.org), which is hosted on the Dutch National e-Infrastructure with the support of SURF Cooperative and financial support by the Netherlands Scientific Organization (NWO 480-05-003 PI: Posthuma).

84

DNA methylation and cortisol stress reactivity

Author contributions All authors have written and approved the manuscript. M.P.M.B., C.H.V. and L.C.H. designed and collected the data for the discovery study and carried out all genome-wide analyses. T.C.-R. performed all analyses in the blood replication sample under supervision of E.B.B. M.H. supplied the methylation and cortisol data collected by the RADAR consortium principle investigators P.v.L., W.M. and S.B. C.M.H. and C.B.N. were responsible for the experimental design and data collection of the blood replication sample and participated in preparing the manuscript. J.M. supplied the bloodâ&#x20AC;&#x201C;brain data. L.C.S. helped with the quality control of the genome-wide methylation analyses. M.P.C. supplied the histone 3 lysine 27 acetylation (H3K27ac) data. R.S.K. and M.J. supervised and commented on the manuscript at all stages.

Competing financial interests C.B.N.â&#x20AC;&#x2122;s work has been funded by the NIH. He is a member of the scientific advisory board of the American Foundation for Suicide Prevention (AFSP), Brain & Behavior Research Foundation (BBRF), Xhale, Anxiety Disorders Association of America (ADAA), Skyland Trail, Clintara LLC and RiverMend Health LLC, and has received income sources or equity of $10,000 or more from American Psychiatric Publishing, Xhale and Clintara.

He is also a member of the board of directors of the AFSP, Gratitude America and ADAA and has consulted for Xhale, Takeda, SK Pharma, Shire, Roche, Lilly, Allergan, Mitsubishi Tanabe Pharma Development America, Taisho Pharmaceutical Inc., Lundbeck, Prismic Pharmaceuticals, Clintara LLC and Total Pain Solutions (TPS). C.H.V. is an advisor for DynaCorts. M.P.M.B., T.C.-R., M.H., P.v.L., W.M., S.B., C.M.H., J.M., L.C.S., M.P.C., R.S.K., M.J., E.B.B. and L.C.H. declare no potential conflict of interest.

85

4

CHAPTER 4

Supplemental figures and tables

Figure S4.1 QQplot showing the p-value distribution for the genome-wide cortisol stress reactivity analysis in the discovery sample (N=85). The three methylation loci for which replication was sought are denoted by red dots.

Figure S4.2 Cook plot of the leverage and standardized residuals for the association between childhood trauma and cortisol stress reactivity (left panel) and for the association between methylation at the KITLG locus and cortisol stress reactivity (right panel).

86

DNA methylation and cortisol stress reactivity

4

Figure S4.3 Cortisol stress reactivity (AUCi) is significantly lower in Afro-American individuals (n=23) compared to Caucasian individuals (n=17) (p=0.002 in a linear regression model). In this boxplot the whiskers extend to the most extreme data point which is no more than 1.5 times the length of the box away from the box. Abbreviations: AUCi= Area under the curve(AUC) with respect to the increase.

87

CHAPTER 4

Figure S4.4 Overview of the stratified analyses for ethnicity and current major depressive disorder (MDD) status in the replication sample. Significant correlations are present for the Caucasians (solid green lines in panels A, C, E) and for the non-MDD subjects (cortisol~methylation only; solid black line in panel F). Dashed lines indicate non-significant correlations. Abbreviations: CTQ= childhood trauma questionnaire, AUCi=Area under the curve (AUC) with respect to the increase.

88

DNA methylation and cortisol stress reactivity

4

Figure S4.5 Correlation of KITLG methylation levels at the cg27512205 locus in whole blood and four brain regions in a linear regression model. The top panel is a boxplot with the DNA methylation levels per tissue type. The prefrontal cortex (PFC; upper left panel), superior temporal gyrus (STG; lower left panel), entorhinal cortex (EC; upper right panel) and cerebellum (CER; lower right panel).

89

CHAPTER 4

Figure S4.6 Distribution of the childhood trauma questionnaire (CTQ) scores in the discovery (N=85) and replication (N=45) samples.

Figure S4.7 The cortisol response over time in the discovery sample of 85 healthy controls. The grey shading indicates the standard error of measurement (SEM) per time point. Abbreviations: min= minutes.

90

DNA methylation and cortisol stress reactivity

4

Figure S4.8 The correlation between batch effects, potential confounders and general methylation levels (principal components) before and after batch effect removal in the discovery sample. Positive correlations are blue and negative correlations are red, the color intensity is related to the size of the correlation coefficient. Significant values are denoted by * p<0.05, ** p<0.001, ***p<0.0001. Abbreviations: AUCi= cortisol stress response area under the curve(AUC) with respect to the increase,CTQ= Childhood Trauma Questionnaire, CD8T= CD8 T cell proportion, CD4T=CD4 T cell proportion, NK=Natural Killer cell proportion, Mono=Monocytes cell proportion,Gran=Granulocytes cell proportion, Bar= ancestry estimates calculated according to (Barfield et al.

2014), PC= principal component. The mention of ‘pre’ after the principal component number (e.g. PC1pre), stands for ‘before batch effect removal’.

Barfield, R.T., Almli, L.M., Kilaru, V., Smith, A.K., Mercer, K.B., Duncan, R., Klengel, T., Mehta, D., Binder, E.B., Epstein, M.P., Ressler, K.J., Conneely, K.N., 2014. Accounting for population stratification in DNA methylation studies. Genet.

Epidemiol. 38, 231-241.

91

CHAPTER 4

Figure S4.9 Correlation between our variables of interest and potential confounders before batch correction in the cross-tissue sample. Significant values are denoted by * p<0.05, ** p<0.001, ***p<0.0001. Abbreviations: Cort_AUCi= cortisol stress response area under the curve(AUC) with respect to the increase, Col= sentrix column,pc= principal component

Figure S4.10 Correlation between our variables of interest and potential confounders after batch correction for sentrix and position in the cross-tissue sample. Significant values are denoted by * p<0.05, ** p<0.001, ***p<0.0001.

Abbreviations: AUCi= cortisol stress response area under the curve(AUC) with respect to the increase, Col= sentrix column, pc= principal component

92

DNA methylation and cortisol stress reactivity

Table S4.1 Overview of studies investigating childhood trauma and cortisol stress reactivity.

Reference

Population (N)

Ethnicity

Stress task

Childhood trauma assessment

Relation childhood trauma and cortisol

(Carpenter et al.

2007)

HC (N=50)

Cauc (76%); Black (6%); Asian (8%); Other (10%)

TSST

CTQ

↓

(Carpenter et al.

2011)

HC (N=78)

Unknown

TSST

CTQ

↓

(Elzinga et al. 2008)

HC (N=80)

Unknown

TSST

TEC

↓

(Heim et al. 2000)

HC (N=26) vs MDD (N=23)

Afr-Am(27%); Cauc (73%)

TSST

ETI

Ns in HC, ↑ in MDD

(Moran-Santa Maria et al. 2010)

HC (N=43) and Cocaine dependence (N=42)

Unknown

TSST

ETI-SR

Ns in HC, Ns in cocaine dependent

(Pesonen et al. 2010)

HC vs WWII separation(N=282)

Unknown

TSST

WWII separation

↑ separated both parents

(Witt et al. 2011)

HC (N=228)

Cauc (99%)

TSST

Enriched family adversity index

N/A focus NPY genotype: ↓ TT allele

(Buchmann et al.

2014)

HC (N=195)

Unknown

TSST

CTQ

N/A focus FKBP5 genotype: ↓ C allele, ns T allele

(Kraft and Luecken, 2009)

HC (N=94)

Cauc(78%); Afr-Am(2%); His(14%); Oth(6%)

Speech

Parental divorce

↓

(Luecken, 1998)

HC (N=61)

Cauc(66%); Black(25%); Asian(7%); Other(3%)

Speech/ video

Parental loss

↑

(Otte et al. 2005)

HC (N=76)

Unknown

Video

LSCR interview; age <14 yrs

Ns

(Goldman-Mellor et al. 2012)

HC (N=543)

Unknown

Two 5-min tasks: Stroop test/ Mirror tracing

ELS questions in longitudinal study

↑ELS no distress ↓ELS distress

(Elzinga et al. 2010)

Social Anxiety Disorder(N=50)

Unknown

TSST

TEC

↑SAD patient

(Videlock et al.

2009)

Early life stress mixed sample (HC& IBS N=83)

Unknown

Sigmoidoscopy

Trauma history questionnaire

Ns

(Bremner et al.

2003)

HC vs PTSD (N=41)

Unknown

20 min Cognitive challenge

ETI

Ns

(Santa Ana et al.

2006)

HC vs PTSD (adult/ childhood index trauma N=89)

Cauc(73%); Afr-Am(25%); Other(2%)

CPT(= Cold pressure test)

CAPS determine index trauma childhood vs adult

Ns

4

Abbreviations: Cauc=Caucasian, Afr-Am=Afro-American, CTQ= childhood trauma questionnaire, ELS=early life stress, DHS= Daily hassles scale; LES= Life experiences survey; LSCR= Life stressor checklist revised, SAD= social anxiety disorder, MDD= major depressive disorder, HC= healthy control, PTSD= post traumatic stress disorder, IBS= Irritable Bowel Syndrome, TSST= Trier social stress test, ETI= Early trauma inventory; CPT= Cold pressure test; WWII= second world war; Ns= non significant; N/A= not applicable.

93

CHAPTER 4

List of studies included in Table S4.1 Bremner, J.D., Vythilingam, M., Vermetten, E., Adil, J., Khan, S., Nazeer, A., Afzal, N., McGlashan, T., Elzinga, B., Anderson, G.M., Heninger, G., Southwick, S.M., Charney, D.S., 2003. Cortisol response to a cognitive stress challenge in posttraumatic stress disorder (PTSD) related to childhood abuse. Psychoneuroendocrinology 28, 733-750.

Buchmann, A.F., Holz, N., Boecker, R., Blomeyer, D., Rietschel, M., Witt, S.H., Schmidt, M.H., Esser, G., Banaschewski, T., Brandeis, D., Zimmermann, U.S., Laucht, M., 2014. Moderating role of FKBP5 genotype in the impact of childhood adversity on cortisol stress response during adulthood. Eur.Neuropsychopharmacol. 24, 837-845.

Carpenter, L.L., Carvalho, J.P., Tyrka, A.R., Wier, L.M., Mello, A.F., Mello, M.F., Anderson, G.M., Wilkinson, C.W., Price, L.H., 2007. Decreased adrenocorticotropic hormone and cortisol responses to stress in healthy adults reporting significant childhood maltreatment. Biol.Psychiatry 62, 1080-1087.

Carpenter, L.L., Shattuck, T.T., Tyrka, A.R., Geracioti, T.D., Price, L.H., 2011. Effect of childhood physical abuse on cortisol stress response. Psychopharmacology (Berl) 214, 367-375.

Elzinga, B.M., Roelofs, K., Tollenaar, M.S., Bakvis, P., van, P.J., Spinhoven, P., 2008. Diminished cortisol responses to psychosocial stress associated with lifetime adverse events a study among healthy young subjects.

Psychoneuroendocrinology 33, 227-237.

Elzinga, B.M., Spinhoven, P., Berretty, E., de, J.P., Roelofs, K., 2010. The role of childhood abuse in HPA-axis reactivity in Social Anxiety Disorder: a pilot study. Biol.Psychol. 83, 1-6.

Goldman-Mellor, S., Hamer, M., Steptoe, A., 2012. Early-life stress and recurrent psychological distress over the lifecourse predict divergent cortisol reactivity patterns in adulthood. Psychoneuroendocrinology 37, 1755-1768.

Heim, C., Newport, D.J., Heit, S., Graham, Y.P., Wilcox, M., Bonsall, R., Miller, A.H., Nemeroff, C.B., 2000. Pituitaryadrenal and autonomic responses to stress in women after sexual and physical abuse in childhood. JAMA 284, 592-597.

Kraft, A.J., Luecken, L.J., 2009. Childhood parental divorce and cortisol in young adulthood: evidence for mediation by family income. Psychoneuroendocrinology 34, 1363-1369.

Luecken, L.J., 1998. Childhood attachment and loss experiences affect adult cardiovascular and cortisol function.

Psychosom.Med. 60, 765-772.

Moran-Santa Maria, M.M., McRae-Clark, A.L., Back, S.E., DeSantis, S.M., Baker, N.L., Spratt, E.G., Simpson, A.N., Brady, K.T., 2010. Influence of cocaine dependence and early life stress on pituitary-adrenal axis responses to CRH and the Trier social stressor. Psychoneuroendocrinology 35, 1492-1500.

Otte, C., Neylan, T.C., Pole, N., Metzler, T., Best, S., Henn-Haase, C., Yehuda, R., Marmar, C.R., 2005. Association between childhood trauma and catecholamine response to psychological stress in police academy recruits. Biol.

Psychiatry 57, 27-32.

Pesonen, A.K., Raikkonen, K., Feldt, K., Heinonen, K., Osmond, C., Phillips, D.I., Barker, D.J., Eriksson, J.G., Kajantie, E., 2010. Childhood separation experience predicts HPA axis hormonal responses in late adulthood: a natural experiment of World War II. Psychoneuroendocrinology 35, 758-767.

Santa Ana, E.J., Saladin, M.E., Back, S.E., Waldrop, A.E., Spratt, E.G., McRae, A.L., LaRowe, S.D., Timmerman, M.A., Upadhyaya, H., Brady, K.T., 2006. PTSD and the HPA axis: differences in response to the cold pressor task among individuals with child vs. adult trauma. Psychoneuroendocrinology 31, 501-509.

Videlock, E.J., Adeyemo, M., Licudine, A., Hirano, M., Ohning, G., Mayer, M., Mayer, E.A., Chang, L., 2009. Childhood trauma is associated with hypothalamic-pituitary-adrenal axis responsiveness in irritable bowel syndrome.

Gastroenterology 137, 1954-1962.

Witt, S.H., Buchmann, A.F., Blomeyer, D., Nieratschker, V., Treutlein, J., Esser, G., Schmidt, M.H., Bidlingmaier, M., Wiedemann, K., Rietschel, M., Laucht, M., Wust, S., Zimmermann, U.S., 2011. An interaction between a neuropeptide Y gene polymorphism and early adversity modulates endocrine stress responses. Psychoneuroendocrinology 36, 1010-1020.

94

DNA methylation and cortisol stress reactivity

Table S4.2 Characteristics of the top three DNA methylation loci from the genome-wide cortisol stress reactivity methylation analysis.

Cg number

27512205

05608730

261779948

Gene

KITLG

C1QTNF2

JAZF1

Location

Chr 12: 88579621

Chr 5: 159797775

Chr 7: 28181035

North-shore CpG island

South-shore CpG island

South-shore CpG island

Discovery (N=85) mean methylation# [range]

0.15 [0.12 - 0.19]

0.38 [0.29 - 0.47]

0.12 [0.09 - 0.16]

Replication (N=45) mean methylation# [range]

0.14 [0.11-0.18]

0.37 [0.31-0.46]

0.10 [0.08-0.14]

Cross-tissue (N=255) mean methylation# [range]

0.09 [0.07-0.12]

0.08 [0.05-0.11]

0.23 [0.16-0.31]

Discovery association cortisol AUCi

B= -1161, p=5.8x10-6*

B= -935, p=6.0x10-6*

B= -1009, p=8.0x10-6*

Blood replication association Cortisol AUCi

B=-1040, p=0.006*

B=433, p= 0.15

B=223, p=0.48

Cross-tissue association cortisol AUCi

B=-104, p=0.003*

B=38, p=0.13

B=-2, p=0.95

Discovery association CTQ

B=0.005, p=0.04*

B=0.006, p=0.04*

B=0.006, p=0.02*

Blood replication association CTQ

B=0.001, p=0.146

B=-0.0001, p=0.903

B=-1.4 x10-5, p=0.990

4

* denotes a nominal association (p<0.05) in a linear regression model # Methylation in percentage (betas) Abbreviations: AUCi= Area under the curve(AUC) with respect to the increase; CTQ=Childhood trauma questionnaire; KITLG=kit ligand; C1QTNF2= C1q and tumor necrosis factor related protein 2; JAZF1= Juxtaposed with another zinc finger protein 1

95

CHAPTER 4

Table S4.3 The webgestalt output for the enrichment analysis of microRNA-related genes in the 19 gene network around KITLG (EntrezGene id 4254). We used a hypergeometric test with all Entrez genes as background and a significance level of 0.05 (FDR correction). The results for each enriched gene sets are listed in this table. For each gene set, the first row lists database name, gene set name, and corresponding gene set ID. The second row lists the following statistics: C: the number of reference genes in the category; O: the number of genes in the gene set and also in the category; E: the expected number in the category; R: ratio of enrichment; rawP: p value from hypergeometric test; adjP: p-value adjusted by the multiple test adjustment.

Database:microRNA Target Name:hsa_CACTGCC,MIR-34A,MIR-34C,MIR-449 ID:DB_ID:673 C=277; O=3; E=0.14; R=21.23; rawP=0.0004; adjP=0.0012 Index

UserID

Gene Symbol

Gene Name

EntrezGene

Ensembl

1

KITLG

KITLG

KIT ligand

4254

ENSG00000049130

2

SHKBP1

SHKBP1

SH3KBP1 binding protein 1

92799

ENSG00000160410

3

COL12A1

COL12A1

collagen, type XII, alpha 1

1303

ENSG00000111799

Database:microRNA Target Name:hsa_AATGTGA,MIR-23A,MIR-23B ID:DB_ID:683 C=417; O=3; E=0.21; R=14.10; rawP=0.0012; adjP=0.0018 Index

UserID

Gene Symbol

Gene Name

EntrezGene

Ensembl

1

KITLG

KITLG

KIT ligand

4254

ENSG00000049130

2

EYA1

EYA1

eyes absent homolog 1 (Drosophila)

2138

ENSG00000104313

3

HMGN2

HMGN2

high mobility group nucleosomal binding domain 2

3151

ENSG00000198830

Database:microRNA Target Name:hsa_ACCAAAG,MIR-9 ID:DB_ID:809 C=493; O=3; E=0.25; R=11.93; rawP=0.0019; adjP=0.0019 Index

UserID

Gene Symbol

Gene Name

EntrezGene

Ensembl

1

KITLG

KITLG

KIT ligand

4254

ENSG00000049130

2

CCDC43

CCDC43

coiled-coil domain containing 43

124808

ENSG00000180329

3

COL12A1

COL12A1

collagen, type XII, alpha 1

1303

ENSG00000111799

96

DNA methylation and cortisol stress reactivity

Methylation cluster

Number of miR449 related genes

Number of not miR449 related genes

Percentage of miR449 genes present

Odds Ratio

Fisher p value

FDR corrected p value

Table S4.4 Enrichment for micro RNA 449 (miR449) related genes for all the methylation clusters (denoted by color).

All

268

19547

100

black

109

6370

1.682358

1.245691

0.061912

0.130342

blue

191

11991

1.567887

1.153744

0.134

0.243636

brown

186

11605

1.577474

1.153496

0.143219

0.249077

cyan

82

3891

2.063932

1.536893

0.001115*

0.008338*

darkgreen

20

703

2.766252

2.074587

0.005235*

0.021351*

darkgrey

15

820

1.796407

1.333978

0.284564

0.392502

darkmagenta

6

360

1.639344

1.215595

0.643431

0.756978

darkolivegreen

4

319

1.23839

0.914569

1

1

darkorange

4

462

0.858369

0.631463

0.537006

0.683596

darkred

21

995

2.066929

1.539304

0.071587

0.143174

darkturquoise

19

749

2.473958

1.850108

0.016935*

0.060097

green

197

10399

1.859192

1.38066

0.000835*

0.008338*

greenyellow

95

4636

2.008032

1.493732

0.001251*

0.008338*

grey

198

11512

1.690863

1.251039

0.018029*

0.060097

grey60

73

4013

1.786588

1.326762

0.035661*

0.089153

lightcyan

62

2984

2.035456

1.515091

0.005338*

0.021351*

lightgreen

22

958

2.244898

1.674892

0.025257*

0.07355

lightyellow

14

1006

1.372549

1.014867

0.889579

0.961707

magenta

72

4019

1.759961

1.306167

0.050033

0.117724

midnightblue

81

4393

1.81046

1.344817

0.025743*

0.07355

orange

6

436

1.357466

1.003714

1

1

orangered4

1

57

1.724138

1.279568

0.546877

0.683596

paleturquoise

4

356

1.111111

0.819519

1

1

pink

87

5337

1.603982

1.187151

0.171955

0.286591

plum1

3

105

2.777778

2.083792

0.182146

0.291433

purple

139

6852

1.988271

1.479495

0.000265*

0.006305*

red

175

9319

1.843269

1.369061

0.001504*

0.008593*

royalblue

26

977

2.592223

1.94093

0.003444*

0.017219*

saddlebrown

7

620

1.116427

0.823483

0.859114

0.954572

salmon

90

4195

2.10035

1.563072

0.000473*

0.006305*

sienna3

4

250

1.574803

1.166974

0.780178

0.891632

skyblue

10

638

1.54321

1.143197

0.605118

0.733477

skyblue3

3

112

2.608696

1.953568

0.206203

0.305485

steelblue

9

332

2.639296

1.977104

0.055395

0.123101

tan

103

5792

1.747243

1.296961

0.0293*

0.078133

turquoise

180

12050

1.471791

1.078191

0.434365

0.579153

violet

1

294

0.338983

0.24807

0.194646

0.299456

white

15

697

2.106742

1.569591

0.099306

0.189154

yellow

219

11446

1.877411

1.394889

0.000321*

0.006305*

yellowgreen

5

234

2.09205

1.558406

0.264274

0.377535

4

*p<0.05 in a Fisherâ&#x20AC;&#x2122;s exact test

97

CHAPTER 4

Methylation cluster

Number of miR23A/ miR23B related genes

Number of not miR23A/miR23B related genes

Percentage of miR23A/miR23B genes present

Odds Ratio

Fisher p value

FDR corrected p value

Table S4.5 Enrichment for micro RNA 23A/23B (miR23A/miR23B) related genes for all the methylation clusters (denoted by color).

All

388

18888

100

Black

125

6354

1.92931

0.980542

0.876933

0.97437

Blue

275

11907

2.257429

1.145333

0.090175

0.150291

Brown

309

11482

2.620643

1.336477

0.000194*

0.000864*

Cyan

110

3863

2.768689

1.421835

0.001925*

0.005922*

Darkgreen

34

689

4.702628

2.463708

1.11E-05*

0.000111*

Darkgrey

27

808

3.233533

1.668413

0.016055*

0.033799*

darkmagenta

6

360

1.639344

0.832311

0.848498

0.97437

darkolivegreen

4

319

1.23839

0.626196

0.538432

0.694751

darkorange

13

453

2.7897

1.432982

0.234532

0.347454

Darkred

30

986

2.952756

1.519366

0.037978*

0.066048

darkturquoise

36

732

4.6875

2.45566

5.94E-06*

9.78E-05*

Green

268

10328

2.529256

1.294863

0.001427*

0.004756*

greenyellow

124

4607

2.62101

1.343338

0.005541*

0.014776*

Grey

273

11437

2.331341

1.188753

0.031269*

0.05956

grey60

104

3982

2.545277

1.304249

0.021412*

0.042825*

lightcyan

117

2929

3.841103

1.994367

1.21E-09*

4.85E-08*

lightgreen

31

949

3.163265

1.631224

0.013915*

0.030923*

lightyellow

19

1001

1.862745

0.947735

0.908

0.981622

magenta

121

3970

2.957712

1.521468

0.000135*

0.000755*

midnightblue

130

4344

2.905677

1.494441

0.000151*

0.000755*

orange

9

433

2.036199

1.037991

0.86174

0.97437

orangered4

3

55

5.172414

2.723566

0.106554

0.170487

paleturquoise

8

352

2.222222

1.134953

0.699199

0.847514

pink

122

5302

2.249263

1.147337

0.191923

0.295265

plum1

2

106

1.851852

0.942235

1

1

purple

189

6802

2.703476

1.387493

0.000384*

0.001534*

red

267

9227

2.812303

1.4444

7.33E-06*

9.78E-05*

royalblue

36

967

3.589232

1.859137

0.001183*

0.004301*

saddlebrown

15

612

2.392344

1.223963

0.463515

0.61802

salmon

106

4179

2.473746

1.265298

0.037522*

0.066048

sienna3

5

249

1.968504

1.002777

1

1

skyblue

10

638

1.54321

0.782737

0.562594

0.703242

skyblue3

4

111

3.478261

1.799455

0.292754

0.41822

steelblue

14

327

4.105572

2.137922

0.010054*

0.025134*

tan

170

5725

2.8838

1.482782

3.63E-05*

0.000242*

turquoise

241

11989

1.970564

0.99597

1

1

violet

8

287

2.711864

1.391961

0.392259

0.541046

white

24

688

3.370787

1.74197

0.013647*

0.030923*

yellow

315

11350

2.700386

1.385325

2.77E-05*

0.000222*

yellowgreen

12

227

5.020921

2.639601

0.003268*

0.009338*

*p<0.05 in a Fisherâ&#x20AC;&#x2122;s exact test

98

DNA methylation and cortisol stress reactivity

Methylation cluster

Number of miR9 related genes

Number of not miR9 related genes

Percentage of miR9 genes present

Odds Ratio

Fisher p value

FDR corrected p value

Table S4.6 Enrichment for micro RNA 9 (miR9) related genes for all the methylation clusters (denoted by color).

All

471

18805

100

Black

168

6311

2.592993

1.091209

0.352824

0.431453

Blue

346

11836

2.840256

1.1922

0.014562*

0.03077*

Brown

356

11435

3.019252

1.266976

0.001035*

0.003764*

Cyan

129

3844

3.246917

1.378097

0.001885*

0.0058*

darkgreen

23

700

3.181189

1.349202

0.172396

0.246279

darkgrey

32

803

3.832335

1.636352

0.011275*

0.028189*

darkmagenta

16

350

4.371585

1.877447

0.023084*

0.046167*

darkolivegreen

12

311

3.71517

1.584633

0.137311

0.211248

darkorange

15

451

3.218884

1.365857

0.220659

0.304357

Darkred

35

981

3.444882

1.465245

0.036233*

0.065878

darkturquoise

28

740

3.645833

1.553862

0.030762*

0.058594

Green

341

10255

3.218196

1.364611

2.03E-05*

0.000247*

greenyellow

148

4583

3.128303

1.325515

0.003827*

0.010935*

Grey

348

11362

2.971819

1.254436

0.001833*

0.0058*

grey60

135

3951

3.303965

1.403281

0.000844*

0.003375*

lightcyan

120

2926

3.939593

1.683951

2.02E-06*

8.1E-05*

lightgreen

38

942

3.877551

1.656696

0.00554*

0.014773*

lightyellow

21

999

2.058824

0.863202

0.596849

0.663166

magenta

125

3966

3.055488

1.293958

0.013142*

0.03077*

midnightblue

135

4339

3.017434

1.277805

0.014616*

0.03077*

orange

14

428

3.167421

1.343389

0.269089

0.358785

orangered4

1

57

1.724138

0.720537

1

1

paleturquoise

11

349

3.055556

1.294451

0.381217

0.448491

pink

153

5271

2.820796

1.192889

0.067614*

0.11269

plum1

2

106

1.851852

0.774915

1

1

purple

223

6768

3.189815

1.353138

0.000322*

0.001609*

red

301

9193

3.170423

1.344136

9.54E-05*

0.000636*

royalblue

34

969

3.389831

1.44101

0.045612*

0.079326*

saddlebrown

14

613

2.232855

0.937981

1

1

salmon

152

4133

3.547258

1.508769

2.62E-05*

0.000247*

sienna3

3

251

1.181102

0.490888

0.295266

0.380989

skyblue

21

627

3.240741

1.375513

0.151841

0.224949

skyblue3

4

111

3.478261

1.479937

0.355948

0.431453

steelblue

9

332

2.639296

1.113338

0.718295

0.776535

tan

194

5701

3.290925

1.397485

0.00015*

0.000859*

turquoise

313

11917

2.55928

1.067382

0.39254

0.448617

violet

12

283

4.067797

1.741421

0.079458

0.127133

white

37

675

5.196629

2.25118

2.19E-05*

0.000247*

yellow

370

11295

3.171882

1.344763

3.08E-05*

0.000247*

yellowgreen

15

224

6.276151

2.749983

0.000769*

0.003375*

*p<0.05 in a Fisherâ&#x20AC;&#x2122;s exact test

99

4

CHAPTER 4

Supplementary Note S4.1, S4.2 and S4.3 Due to the length of Supplemental Note 4.1, 4.2 and 4.3, they are not displayed here.

The complete notes are available online as a supplement to the article and can be obtained from the author on request.

S4.1 Possible confounders in the blood replication First the potential confounding by cellcount on average DNA methylation was examined (analysis 1).

Next we investigated the potential influence of current MDD and ethnicity per outcome variable (cg27512205 methylation levels, cortisol stress reactivity and childhood trauma) (analyses 2-7).

Moreover stratified analyses for current MDD and ethnicity were performed to investigate potential confounding (analyses 8-19).

Finally, to establish the influence of current MDD and ethnicity, they were added as covariates to the model examining the association between KITLG and cortisol stress reactivity (analysis 20).

S4.2 Influence of cell count in the discovery sample The association of cellcount with cortisol stress reactivity (analysis 1), childhood trauma (analysis 2), average methylation (analysis 3) and KITLG methylation (analysis 4) is examined. Finally, to establish the influence of cellcount on the association between KITLG and cortisol stress reactivity, we show the model summary for the entire model (analysis 5).

S4.3 Summary of the KITLG model in the cross-tissue replication sample

100

DNA methylation and cortisol stress reactivity

Supplementary data S4.1 Due to the length of supplementary data S4.1, only the first page is displayed here. The complete list is available online as a supplement to the article and can be obtained from the author on request.

Table containing a list of all the probes nominally associated with cortisol stress reactivity (AUCi) in the discovery sample with the regression coefficient and p-value.

ProbeID

B

P.Value

ProbeID

B

P.Value

cg27512205

-1161.82

5.78E-06

cg19497501

-1061.38

3.99E-05

cg05608730

-935.119

6.04E-06

cg21977075

800.4065

4.5E-05

cg26179948

-1009.39

7.95E-06

ch.10.1965163R

-740.964

4.93E-05

cg07780534

-1070.35

1.85E-05

cg00229532

686.4244

4.94E-05

cg09573795

-692.594

1.93E-05

cg07539200

-817.253

5.05E-05

cg05739167

-729.363

2.08E-05

cg17739552

-668.438

5.19E-05

cg26168148

645.7768

2.56E-05

cg00344209

-1249.91

5.27E-05

cg06422529

640.3179

2.93E-05

cg25120210

-583.623

5.56E-05

cg02621907

-1253.38

3.1E-05

cg10575441

-660.894

5.57E-05

cg06957003

949.9938

3.18E-05

cg01270241

712.6224

5.97E-05

cg10822172

-1098.92

3.26E-05

cg03740162

-695.063

6.04E-05

cg03009397

-992.554

3.29E-05

cg19695521

-971.197

6.45E-05

cg14862454

-1260.97

3.4E-05

cg22538778

712.5977

6.82E-05

cg24460489

863.0936

3.72E-05

cg23232612

-1158.58

6.86E-05

cg17294330

-1058.8

3.89E-05

cg23809597

-821.949

6.88E-05

4

101

CHAPTER 4

Supplementary data S4.2 Due to the length of supplementary data S4.2, only the first page is displayed here.

The complete list is available online as a supplement to the article and can be obtained from the author on request.

List of all the Gene Ontology (GO)-terms significantly enriched (FDR<0.05) in the KITLG-probe containing red Weighted Gene Coexpression Network Analysis (WGCNA) methylation cluster.

Term

Ontology

N

DE

P.DE

FDR

GO:0044260

cellular macromolecule metabolic process

BP

7161

4352

2.18494E-66

2.64514E-63

GO:0006139

nucleobase-containing BP compound metabolic process

4992

3140

4.3946E-60

4.72908E-57

GO:0090304

nucleic acid metabolic process

BP

4449

2837

2.64524E-59

2.69675E-56

GO:0034641

cellular nitrogen compound metabolic process

BP

5476

3395

2.72806E-58

2.64212E-55

GO:0046483

heterocycle metabolic process

BP

5162

3210

3.10948E-55

2.86813E-52

GO:0010467

gene expression

BP

4416

2802

4.79079E-55

4.21807E-52

GO:0006725

cellular aromatic compound metabolic process

BP

5176

3210

5.66788E-54

4.77334E-51

GO:0044237

cellular metabolic process

BP

9052

5255

1.10129E-52

8.88835E-50

GO:0006807

nitrogen compound metabolic process

BP

5780

3530

4.34446E-52

3.36609E-49

GO:0043170

macromolecule metabolic process

BP

7825

4617

1.21309E-51

9.03755E-49

GO:0016070

RNA metabolic process

BP

3987

2538

1.26601E-50

9.08244E-48

GO:1901360

organic cyclic compound metabolic process

BP

5380

3299

1.63459E-50

1.13078E-47

GO:0044238

primary metabolic process

BP

8926

5137

1.13025E-45

7.29765E-43

GO:0071704

organic substance metabolic process

BP

9257

5295

1.60335E-43

9.7053E-41

GO:0008152

metabolic process

BP

10676

6017

3.9894E-43

2.34166E-40

GO:0009059

macromolecule biosynthetic process

BP

4427

2744

1.88576E-41

1.01464E-38

102

DNA methylation and cortisol stress reactivity

Supplementary data S4.3 Excel file with three sheets, one per dataset (discovery, blood-replication and buccalreplication). Each sheet contains phenotype data and, for the top 3 loci (cg27512205, cg05608730 and cg26179948), the DNA methylation values used for analyses (after normalization and batch effect correction) as well as signal intensities and detection p-values. Detection p-values represent the confidence that a given transcript is expressed above the background defined by negative control probes. Abbreviations: Cort_AUCi= cortisol stress response area under the curve (AUC) with respect to the increase, CTQ= Childhood Trauma Questionnaire, LSCR= Life Stressor Checklist- Revised, CD8T= CD8 T cell proportion, CD4T=CD4 T cell proportion, NK=Natural Killer cell proportion, Mono=Monocytes cell proportion, Gran=Granulocytes cell proportion, Bcell= B cell proportion. The mention of ‘beta’ after the cg number (e.g. cg27512205_beta), stands for ‘DNA methylation levels expressed in beta’. Similarly, ‘meth’, ‘unmeth’ and ‘det_pval’ indicate respectively the methylated, unmethylated and detection p-value for the specified probe (e.g. cg27512205_unmeth).

Due to the length of supplementary data S4.3, only the first page is displayed here. The complete excel file is available online as a supplement to the article and can be obtained from the author on request.

103

4

104

0.166

0.180

0.161

0.173

0.182

0.125

0.143

0.163

0.134

0.161

0.167

0.183

0.164

0.160

0.167

DISCOVERY-02

DISCOVERY-03

DISCOVERY-04

DISCOVERY-05

DISCOVERY-06

DISCOVERY-07

DISCOVERY-08

DISCOVERY-09

DISCOVERY-10

DISCOVERY-11

DISCOVERY-12

DISCOVERY-13

DISCOVERY-14

DISCOVERY-15

0.428

0.347

0.408

0.426

0.405

0.402

0.358

0.360

0.408

0.353

0.469

0.406

0.375

0.357

0.354

cg27512205 cg05608730

DISCOVERY-01

Sheet 1 discovery sample

Sample name

0.113

0.124

0.127

0.142

0.155

0.111

0.113

0.134

0.107

0.115

0.108

0.137

0.110

0.126

0.135

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

cg26179948 Race

Male

Male

Male

Male

Male

Male

Male

Male

Male

Male

Male

Male

Male

Male

Male

Gender

21

56

29

24

22

25

26

22

48

25

22

59

58

22

19

age

35.4

1165.2

765.15

264.4

-131.35

120.25

742.75

230.2

503.7

614.5

14.2

386.15

674.85

539.8

372.4

Cort AUCi

24

28

25

24

31

24

24

28

24

29

27

26

29

30

24

CTQ total

#N/A

6

#N/A

1

#N/A

0

1

0

2

1

#N/A

2

6

#N/A

1

0.11

0.06

0.09

0.11

0.10

0.08

0.07

0.12

0.08

0.09

0.11

0.11

0.11

0.09

0.13

0.07

0.06

0.13

0.14

0.11

0.20

0.15

0.08

0.13

0.13

0.04

0.13

0.12

0.12

0.08

0.17

0.15

0.18

0.17

0.22

0.17

0.21

0.21

0.26

0.20

0.14

0.15

0.16

0.24

0.15

LSCR total CD8T Mono NK

0.08

0.06

0.08

0.06

0.08

0.06

0.10

0.07

0.08

0.10

0.05

0.05

0.07

0.07

0.08

0.59

0.71

0.54

0.56

0.52

0.51

0.50

0.56

0.50

0.52

0.69

0.60

0.58

0.49

0.61

CD4T Bcell

0.11

0.06

0.09

0.11

0.10

0.08

0.07

0.12

0.08

0.09

0.11

0.11

0.11

0.09

0.13

Gran

8622007124

8622007124

8622007124

8622007124

8622007046

8622007046

8622007046

8622007046

8622007046

8622007046

8622007046

8622007046

8622007046

8622007046

8622007046

sentrix

R04C02

R03C02

R02C02

R01C02

R06C02

R06C01

R05C02

R05C01

R04C02

R04C01

R03C02

R02C02

R02C01

R01C02

R01C01

position

CHAPTER 4

cg27512205

0.140

0.127

0.131

0.129

0.148

0.134

0.159

0.139

0.140

0.111

BLOOD REPLICATION-01

BLOOD REPLICATION-02

BLOOD REPLICATION-03

BLOOD REPLICATION-04

BLOOD REPLICATION-05

BLOOD REPLICATION-06

BLOOD REPLICATION-07

BLOOD REPLICATION-08

BLOOD REPLICATION-09

BLOOD REPLICATION-10

0.329

0.368

0.396

0.380

0.363

0.367

0.355

0.321

0.424

0.377

cg05608730

Sheet 2 blood replication sample

Sample name

0.100

0.111

0.132

0.095

0.108

0.106

0.099

0.113

0.094

0.107

cg26179948

FALSE

FALSE

FALSE

FALSE

FALSE

TRUE

FALSE

FALSE

FALSE

FALSE

MDD

AfroAmerican

Caucasian

AfroAmerican

Caucasian

AfroAmerican

AfroAmerican

AfroAmerican

Caucasian

Other

AfroAmerican

Race

Male

Male

Male

Male

Female

Female

Male

Male

Female

Female

Gender

25

32

40

19

28

26

33

24

28

24

1082.63

1534.58

1582.58

1380.82

1211.03

870.6

809.63

1822.5

1775.93

543

Age Cort AUCi

41

67

40

66

37

61

27

26

25

96

CTQ total

0.05

0.02

0.10

0.09

0.10

0.04

0.02

0.06

0.00

0.06

CD8T

0.12

0.09

0.12

0.08

0.08

0.09

0.09

0.11

0.11

0.06

Mono

0.08

0.06

0.15

0.09

0.15

0.07

0.07

0.05

0.03

0.08

NK

0.24

0.15

0.10

0.14

0.09

0.13

0.17

0.14

0.04

0.21

CD4T

Gran 0.51 0.80 0.61 0.63 0.64 0.53 0.55 0.49 0.64 0.44

Bcell 0.09 0.02 0.03 0.04 0.04 0.06 0.05 0.06 0.04 0.06

position

6164647077 R05C02

6164647077 R05C01

6164647077 R04C02

6164647077 R04C01

6164647077 R03C02

6164647077 R03C01

6164647077 R02C02

6164647062 R06C02

6164647062 R04C02

6164647062 R02C02

sentrix

DNA methylation and cortisol stress reactivity

4

105

106

cg27512205 beta

0.100820951

0.102881909

0.098117964

0.086683942

0.094715606

0.089254493

0.09813365

0.075504012

0.086780358

0.074399951

0.092333928

0.089877388

0.085401304

0.099000701

0.09106659

0.111042878

0.08485008

0.070332974

0.101773372

0.102657901

0.094933838

0.089773671

0.076008324

0.082724538

0.092344236

0.097417138

Sample name

BUCCAL_REPLICATION-01

BUCCAL_REPLICATION-02

BUCCAL_REPLICATION-03

BUCCAL_REPLICATION-04

BUCCAL_REPLICATION-05

BUCCAL_REPLICATION-06

BUCCAL_REPLICATION-07

BUCCAL_REPLICATION-08

BUCCAL_REPLICATION-09

BUCCAL_REPLICATION-10

BUCCAL_REPLICATION-11

BUCCAL_REPLICATION-12

BUCCAL_REPLICATION-13

BUCCAL_REPLICATION-14

BUCCAL_REPLICATION-15

BUCCAL_REPLICATION-16

BUCCAL_REPLICATION-17

BUCCAL_REPLICATION-18

BUCCAL_REPLICATION-19

BUCCAL_REPLICATION-20

BUCCAL_REPLICATION-21

BUCCAL_REPLICATION-22

BUCCAL_REPLICATION-23

BUCCAL_REPLICATION-24

BUCCAL_REPLICATION-25

BUCCAL_REPLICATION-26

Sheet 3 buccal replication sample

0.223947338

0.217438692

0.254623546

0.208258403

0.208518046

0.234465444

0.174706143

0.257061125

0.184685016

0.204302968

0.196236705

0.256998975

0.271804933

0.210119455

0.223340794

0.290223737

0.259726958

0.215444073

0.227798032

0.211914129

0.260101296

0.270746863

0.19595179

0.258249409

0.240062191

0.24923682

cg05608730 beta

0.066148544

0.064050219

0.076601121

0.083771845

0.077782368

0.081241766

0.079132106

0.081084201

0.05525367

0.072116842

0.067718649

0.085640332

0.08627797

0.070037461

0.087997203

0.082520291

0.083158904

0.080666559

0.066936479

0.070604492

0.087248152

0.070051304

0.076894315

0.089538131

0.073610032

0.08313066

cg26179948 beta

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Caucasian

Race

Male

Male

Male

Female

Male

Female

Male

Female

Male

Female

Male

Female

Male

Female

Male

Female

Female

Male

Male

Male

Female

Female

Male

Male

Female

Female

Gender

12.89

13.13

13.27

12.86

13.04

13.07

13.37

13.22

12.54

12.73

13.03

12.77

12.9

13.23

13.81

13.22

12.77

13.26

12.64

12.42

12.84

12.62

13.23

13.24

12.91

13.08

age

9.33

1.52

-46.05

46.7

-336.4

-98.47

-23

-42.62

-1.88

-87.55

-4.9

38.05

-6.27

-39.62

-225.4

15.1

-65.28

-0.15

-107.75

-16.1

-83.02

-158.07

62.35

-16.52

-75

-56.1

Cort AUCi

9971851117

9971851088

9971851117

9971851088

9971851088

9971851117

9971851117

9971851117

9971851117

9971851088

9971851088

9971851088

9971851117

9971851117

9971851088

9971851088

9971851133

9971851133

9971851133

9971851133

9971851133

9971851133

9971851133

9971851133

9971851133

9971851133

sentrix

R01C02

R06C02

R06C01

R05C02

R04C02

R04C01

R03C01

R02C01

R01C01

R02C02

R01C02

R05C01

R06C02

R05C02

R03C01

R01C01

R06C02

R05C02

R03C02

R02C02

R01C02

R05C01

R04C01

R03C01

R02C01

R01C01

position

CHAPTER 4

4

107

Notes

S

heet music is the basic form in which Western classical music is notated so that it can be learned and performed by solo singers, instrumentalists or musical ensembles.

CHAPTER 9 References

CHAPTER 9

Abdolmaleky, H. M., Nohesara, S., Ghadirivasfi, M., Lambert, A. W., Ahmadkhaniha, H., Ozturk, S., … Thiagalingam, S. (2014). DNA hypermethylation of serotonin transporter gene promoter in drug na??ve patients with schizophrenia.

Schizophrenia Research, 152(2-3), 373–380. http://doi.org/10.1016/j.schres.2013.12.007 Aberg, K. A., McClay, J. L., Nerella, S., Clark, S., Kumar, G., Chen, W., … van den Oord, E. J. C. G. (2014). MethylomeWide Association Study of Schizophrenia. JAMA Psychiatry, 71(3), 255. http://doi.org/10.1001/ jamapsychiatry.2013.3730 Aberg, K. A., Xie, L. Y., McClay, J. L., Nerella, S., Vunck, S., Snider, S., … van den Oord, E. J. C. G. (2013). Testing two models describing how methylome-wide studies in blood are informative for psychiatric conditions. Epigenomics, 5(4), 367–77. http://doi.org/10.2217/epi.13.36 Acosta, G. B., & Rubio, M. C. (1994). GABAA receptors mediate the changes produced by stress on GABA function and locomotor activity. Neuroscience Letters, 176(1), 29–31. http://doi.org/10.1016/0304-3940(94)90863-X Alger, J. R. (2010). Quantitative proton magnetic resonance spectroscopy and spectroscopic imaging of the brain: a didactic review. Topics in Magnetic Resonance Imaging : TMRI, 21(2), 115–28. http://doi.org/10.1097/ RMR.0b013e31821e568f Allen, A. P., Kennedy, P. J., Cryan, J. F., Dinan, T. G., & Clarke, G. (2014). Biological and psychological markers of stress in humans: Focus on the Trier Social Stress Test. Neuroscience and Biobehavioral Reviews, 38(1873-7528 (Electronic)), 94–124. http://doi.org/10.1016/j.neubiorev.2013.11.005 Anderson, O. S., Sant, K. E., & Dolinoy, D. C. (2012). Nutrition and epigenetics: An interplay of dietary methyl donors, one-carbon metabolism and DNA methylation. Journal of Nutritional Biochemistry. http://doi.org/10.1016/j. jnutbio.2012.03.003 Andreasen, N. C., Flaum, M., & Arndt, S. (1992). The Comprehensive Assessment of Symptoms and History (CASH).

An instrument for assessing diagnosis and psychopathology. Archives of General Psychiatry, 49(8), 615–23. http:// doi.org/10.1001/archpsyc.1992.01820080023004 Andreychenko, A., Boer, V. O., Arteaga De Castro, C. S., Luijten, P. R., & Klomp, D. W. J. (2012). Efficient spectral editing at 7 T: GABA detection with MEGA-sLASER. Magnetic Resonance in Medicine, 68(4), 1018–1025. http:// doi.org/10.1002/mrm.24131 Andreychenko, A., Klomp, D. W. J., De Graaf, R. A., Luijten, P. R., & Boer, V. O. (2013). In vivo GABA T2 determination with J-refocused echo time extension at 7 T. NMR in Biomedicine, 26(11), 1596–1601. http://doi.org/10.1002/ nbm.2997 Aryee, M. J., Jaffe, A. E., Corrada-Bravo, H., Ladd-Acosta, C., Feinberg, A. P., Hansen, K. D., & Irizarry, R. A. (2014).

Minfi: A flexible and comprehensive Bioconductor package for the analysis of Infinium DNA methylation microarrays. Bioinformatics, 30(10), 1363–1369. http://doi.org/10.1093/bioinformatics/btu049 Ashburner, J., & Friston, K. J. (2005). Unified segmentation. NeuroImage, 26(3), 839–851. http://doi.org/10.1016/j. neuroimage.2005.02.018 Bale, T. L. (2015). Epigenetic and transgenerational reprogramming of brain development. Nature Reviews.

Neuroscience, 16(6), 332–344. http://doi.org/10.1038/nrn3818 Barfield, R. T., Almli, L. M., Kilaru, V., Smith, A. K., Mercer, K. B., Duncan, R., … Conneely, K. N. (2014). Accounting for population stratification in DNA methylation studies. Genetic Epidemiology, 38(3), 231–241. http://doi. org/10.1002/gepi.21789

174

References

Beck, A. T., Ward, C. H., Mendelson, M., Mock, J., & Erbaugh, J. (1961). An inventory for measuring depression.

Archives of General Psychiatry, 4(6), 561–571. http://doi.org/10.1001/archpsyc.1961.01710120031004 Becker, J. (1996). Hungry ghosts : China’s secret famine. New York: Free Press.

Bedse, G., Romano, A., Tempesta, B., Lavecchia, M. A., Pace, L., Bellomo, A., … Gaetani, S. (2015). Inhibition of anandamide hydrolysis enhances noradrenergic and GABAergic transmission in the prefrontal cortex and basolateral amygdala of rats subjected to acute swim stress. Journal of Neuroscience Research, 93(5), 777–787. http://doi.org/10.1002/jnr.23539 Bell, J. T., Pai, A. A., Pickrell, J. K., Gaffney, D. J., Pique-Regi, R., Degner, J. F., … Pritchard, J. K. (2011). DNA methylation patterns associate with genetic and gene expression variation in HapMap cell lines. Genome Biology, 12(1), R10. http://doi.org/10.1186/gb-2011-12-1-r10 Bender, R. E., & Alloy, L. B. (2011). Life stress and kindling in bipolar disorder: Review of the evidence and integration with emerging biopsychosocial theories. Clinical Psychology Review, 31(3), 383–398. http://doi.org/10.1016/j. cpr.2011.01.004 Bernstein, B. E., Meissner, A., & Lander, E. S. (2007). The Mammalian Epigenome. Cell, 128(4), 669–681. http:// doi.org/10.1016/j.cell.2007.01.033 Bernstein, D. P., Stein, J. A., Newcomb, M. D., Walker, E., Pogge, D., Ahluvalia, T., … Zule, W. (2003). Development and validation of a brief screening version of the Childhood Trauma Questionnaire. Child Abuse and Neglect, 27(2), 169–190. http://doi.org/10.1016/S0145-2134(02)00541-0 Bian, Q., Kato, T., Monji, A., Hashioka, S., Mizoguchi, Y., Horikawa, H., & Kanba, S. (2008). The effect of atypical antipsychotics, perospirone, ziprasidone and quetiapine on microglial activation induced by interferon-?? Progress in Neuro-Psychopharmacology and Biological Psychiatry, 32(1), 42–48. http://doi.org/10.1016/j.pnpbp.2007.06.031 Bibikova, M., Barnes, B., Tsan, C., Ho, V., Klotzle, B., Le, J. M., … Shen, R. (2011). High density DNA methylation array with single CpG site resolution. Genomics, 98(4), 288–295. http://doi.org/10.1016/j.ygeno.2011.07.007 Biggio, G., Concas, A., Follesa, P., Sanna, E., & Serra, M. (2007). Stress, ethanol, and neuroactive steroids.

Pharmacology and Therapeutics. http://doi.org/10.1016/j.pharmthera.2007.04.005 Bird, A. P. (1986). CpG-rich islands and the function of DNA methylation. Nature, 321(6067), 209–213. http://doi. org/10.1038/321209a0 Bird, A. P. (2002). DNA methylation patterns and epigenetic memory. Genes and Development, 16(1), 6–21. http:// doi.org/10.1101/gad.947102 Bjornsson, H. T., Sigurdsson, M. I., Fallin, M. D., Irizarry, R. A., Aspelund, T., Cui, H., … Feinberg, A. P. (2008). Intraindividual change over time in DNA methylation with familial clustering. Jama, 299(24), 2877–83. http://doi. org/10.1001/jama.299.24.2877 Boer, V. O., van Lier, A. L. H. M. W., Hoogduin, J. M., Wijnen, J. P., Luijten, P. R., & Klomp, D. W. J. (2011). 7-T 1H MRS with adiabatic refocusing at short TE using radiofrequency focusing with a dual-channel volume transmit coil. NMR in Biomedicine, 24(9), 1038–1046. http://doi.org/10.1002/nbm.1641 Boks, M. P. M., de Jong, N. M., Kas, M. J. H., Vinkers, C. H., Fernandes, C., Kahn, R. S., … Ophoff, R. A. (2012). Current status and future prospects for epigenetic psychopharmacology. Epigenetics : Official Journal of the DNA Methylation Society, 7(1), 20–8. http://doi.org/10.4161/epi.7.1.18688

175

9

CHAPTER 9

Boks, M. P. M., Derks, E. M., Weisenberger, D. J., Strengman, E., Janson, E., Sommer, I. E., … Ophoff, R. A. (2009).

The relationship of DNA methylation with age, gender and genotype in twins and healthy controls. PLoS ONE, 4(8), e6767. http://doi.org/10.1371/journal.pone.0006767 Boksa, P. (2013). A way forward for research on biomarkers for psychiatric disorders. Journal of Psychiatry & Neuroscience : JPN, 38(2), 75–77. http://doi.org/10.1503/jpn.130018 Bonder, M. J., Luijk, R., Zhernakova, D., Moed, M., Deelen, P., Vermaat, M., … Heijmans, B. T. (2015). Disease variants alter transcription factor levels and methylation of their binding sites. bioRxiv, 033084. http://doi. org/10.1101/033084 Bönsch, D., Wunschel, M., Lenz, B., Janssen, G., Weisbrod, M., & Sauer, H. (2012). Methylation matters? Decreased methylation status of genomic DNA in the blood of schizophrenic twins. Psychiatry Research, 198(3), 533–537. http://doi.org/10.1016/j.psychres.2011.09.004 Booth, M. J., Ost, T. W. B., Beraldi, D., Bell, N. M., Branco, M. R., Reik, W., & Balasubramanian, S. (2013). Oxidative bisulfite sequencing of 5-methylcytosine and 5-hydroxymethylcytosine. Nature Protocols, 8(10), 1841–51. http:// doi.org/10.1038/nprot.2013.115 Borsini, A., Zunszain, P. A., Thuret, S., & Pariante, C. M. (2015). The role of inflammatory cytokines as key modulators of neurogenesis. Trends in Neurosciences, 38(3), 145–157. http://doi.org/10.1016/j.tins.2014.12.006 Borsini, F., Mancinelli, A., D’Aranno, V., Evangelista, S., & Meli, A. (1988). On the role of endogenous GABA in the forced swimming test in rats. Pharmacology, Biochemistry, and Behavior, 29(2), 275–9. Retrieved from http:// www.ncbi.nlm.nih.gov/pubmed/3362921 Breen, M. S., Maihofer, A. X., Glatt, S. J., Tylee, D. S., Chandler, S. D., Tsuang, M. T., … Connor, D. T. O. (2015). Gene networks specific for innate immunity define post-traumatic stress disorder. Molecular Psychiatry, 20(10 March 2015), 1–8. http://doi.org/10.1038/mp.2015.9 Bremner, J. D., Vermetten, E., & Mazure, C. M. (2000). Development and preliminary psychometric properties of an instrument for the measurement of childhood trauma: The early trauma inventory. Depression and Anxiety, 12(1), 1–12. http://doi.org/10.1002/1520-6394(2000)12:1<1::AID-DA1>3.0.CO;2-W Brenner, K., Liu, A., Laplante, D. P., Lupien, S. J., Pruessner, J. C., Ciampi, A., … King, S. (2009). Cortisol response to a psychosocial stressor in schizophrenia: Blunted, delayed, or normal? Psychoneuroendocrinology, 34(6), 859–868. http://doi.org/10.1016/j.psyneuen.2009.01.002 Brown, A. S., & Susser, E. S. (2008). Prenatal nutritional deficiency and risk of adult schizophrenia. Schizophrenia Bulletin, 34(6), 1054–1063. http://doi.org/10.1093/schbul/sbn096 Buchanan, T. W., Driscoll, D., Mowrer, S. M., Sollers, J. J., Thayer, J. F., Kirschbaum, C., & Tranel, D. (2010). Medial prefrontal cortex damage affects physiological and psychological stress responses differently in men and women.

Psychoneuroendocrinology, 35(1), 56–66. http://doi.org/10.1016/j.psyneuen.2009.09.006 Burke, H. M., Davis, M. C., Otte, C., & Mohr, D. C. (2005). Depression and cortisol responses to psychological stress: A meta-analysis. Psychoneuroendocrinology, 30(9), 846–856. http://doi.org/10.1016/j.psyneuen.2005.02.010 Busche, S., Shao, X., Caron, M., Kwan, T., Allum, F., Cheung, W. a, … Grundberg, E. (2015). Population whole-genome bisulfite sequencing across two tissues highlights the environment as the principal source of human methylome variation. Genome Biology, 16(1), 290. http://doi.org/10.1186/s13059-015-0856-1 Cai, K., Pr Nanga, R., Lamprou, L., Schinstine, C., Elliott, M., Hariharan, H., … Epperson, C. N. (2012). The Impact of

176

References

Gabapentin Administration on Brain GABA and Glutamate Concentrations: A 7T 1H-MRS Study.

Neuropsychopharmacology, 37(10), 2764–2771. http://doi.org/10.1038/npp.2012.142 Cardno, A. G., Marshall, E. J., Coid, B., Macdonald, A. M., Ribchester, T. R., Davies, N. J., … Murray, R. M. (1999).

Heritability estimates for psychotic disorders: the Maudsley twin psychosis series. Archives of General Psychiatry, 56(2), 162–168. http://doi.org/10.1001/archpsyc.56.2.162 Carpenter, L. L., Carvalho, J. P., Tyrka, A. R., Wier, L. M., Mello, A. F., Mello, M. F., … Price, L. H. (2007). Decreased Adrenocorticotropic Hormone and Cortisol Responses to Stress in Healthy Adults Reporting Significant Childhood Maltreatment. Biological Psychiatry, 62(10), 1080–1087. http://doi.org/10.1016/j.biopsych.2007.05.002 Carroll, J. F. X. (2008). Development of the Mental Health Screening Form III. International Journal of Mental Health and Addiction, 6(1), 72–76. http://doi.org/10.1007/s11469-007-9102-8 Caspi, A., & Moffitt, T. E. (2006). Gene-environment interactions in psychiatry: joining forces with neuroscience.

Nature Reviews. Neuroscience, 7(7), 583–90. http://doi.org/10.1038/nrn1925 Cedar, H., & Bergman, Y. (2009). Linking DNA methylation and histone modification: patterns and paradigms.

Nature Reviews. Genetics, 10(5), 295–304. http://doi.org/10.1038/nrg2540 Cervantes, P., Gelber, S., Ng Ying Kin, F. N. K., Nair, V. N. P., & Schwartz, G. (2001). Circadian secretion of cortisol in bipolar disorder. Journal of Psychiatry and Neuroscience, 26(5), 411–416.

Chen, A. J., Zhou, G., Juan, T., Colicos, S. M., Cannon, J. P., Cabriera-Hansen, M., … Belmont, J. W. (2002). The dual specificity JKAP specifically activates the c-Jun N-terminal kinase pathway. Journal of Biological Chemistry, 277(39), 36592–36601. http://doi.org/10.1074/jbc.M200453200 Chen, J., Evans, A. N., Liu, Y., Honda, M., Saavedra, J. M., & Aguilera, G. (2012). Maternal Deprivation in Rats is Associated with Corticotrophin-Releasing Hormone (CRH) Promoter Hypomethylation and Enhances CRH Transcriptional Responses to Stress in Adulthood. Journal of Neuroendocrinology, 24(7), 1055–1064. http://doi. org/10.1111/j.1365-2826.2012.02306.x Chen, Q., Ouyang, D., Geng, M., Xu, L., Zhang, Y., Wang, F., & He, X. (2011). Valproic acid exhibits biphasic effects on apoptotic cell death of activated lymphocytes through differential modulation of multiple signaling pathways.

Journal of Immunotoxicology, 8(3), 210–218. http://doi.org/10.3109/1547691X.2011.568979 Chen, Y. A., Lemire, M., Choufani, S., Butcher, D. T., Grafodatskaya, D., Zanke, B. W., … Weksberg, R. (2013). Discovery of cross-reactive probes and polymorphic CpGs in the Illumina Infinium HumanMethylation450 microarray.

Epigenetics, 8(2), 203–209. http://doi.org/10.4161/epi.23470 Choi, C., Dimitrov, I. E., Douglas, D., Patel, A., Kaiser, L. G., Amezcua, C. A., & Maher, E. A. (2010). Improvement of resolution for brain coupled metabolites by optimized 1H MRS at 7T. NMR in Biomedicine, 23(9), 1044–1052. http://doi.org/10.1002/nbm.1529 Cohen, H., Kaplan, Z., Kotler, M., Mittelman, I., Osher, Y., & Bersudsky, Y. (2003). Impaired heart rate variability in euthymic bipolar patients. Bipolar Disorders, 5(2), 138–143. http://doi.org/10.1034/j.1399-5618.2003.00027.x Cohrs, S., Röher, C., Jordan, W., Meier, A., Huether, G., Wuttke, W., … Rodenbeck, A. (2006). The atypical antipsychotics olanzapine and quetiapine, but not haloperidol, reduce ACTH and cortisol secretion in healthy subjects. Psychopharmacology, 185(1), 11–18. http://doi.org/10.1007/s00213-005-0279-x Craddock, N. J., & Sklar, P. (2013). Bipolar Disorder 1 - Genetics of bipolar disorder. The Lancet, 381(9878), 1654–1662. http://doi.org/10.1016/S0140-6736(13)60855-7

177

9

CHAPTER 9

Creyghton, M. P., Cheng, A. W., Welstead, G. G., Kooistra, T., Carey, B. W., Steine, E. J., … Jaenisch, R. (2010). Histone H3K27ac separates active from poised enhancers and predicts developmental state. Proceedings of the National Academy of Sciences of the United States of America, 107(50), 21931–21936. http://doi.org/10.1073/ pnas.1016071107 Davies, M. N., Volta, M., Pidsley, R., Lunnon, K., Dixit, A., Lovestone, S., … Mill, J. (2012). Functional annotation of the human brain methylome identifies tissue-specific epigenetic variation across brain and blood. Genome Biology, 13(6), R43. http://doi.org/10.1186/gb-2012-13-6-r43 Davis, S., & Bilke, S. (2010). Methylumi: Handle Illumina Methylation Data. Bioconductor R Package Version 132.

Day, J. J., Kennedy, A. J., & Sweatt, J. D. (2014). DNA Methylation and Its Implications and Accessibility for Neuropsychiatric Therapeutics. Annual Review of Pharmacology and Toxicology, (October 2014), 1–21. http:// doi.org/10.1146/annurev-pharmtox-010814-124527 de Groote, L., & Linthorst, A. C. E. (2007). Exposure to novelty and forced swimming evoke stressor-dependent changes in extracellular GABA in the rat hippocampus. Neuroscience, 148(3), 794–805. http://doi.org/10.1016/j. neuroscience.2007.06.030 Dempster, E. L., Pidsley, R., Schalkwyk, L. C., Owens, S., Georgiades, A., Kane, F., … Mill, J. (2011). Disease-associated epigenetic changes in monozygotic twins discordant for schizophrenia and bipolar disorder. Human Molecular Genetics, 20(24), 4786–4796. http://doi.org/10.1093/hmg/ddr416 DeRijk, R. H., & de Kloet, E. R. (2008). Corticosteroid receptor polymorphisms: Determinants of vulnerability and resilience. European Journal of Pharmacology, 583(2-3), 303–311. http://doi.org/10.1016/j.ejphar.2007.11.072 Dong, E., Dzitoyeva, S. G., Matrisciano, F., Tueting, P., Grayson, D. R., & Guidotti, A. (2015). Brain-derived neurotrophic factor epigenetic modifications associated with schizophrenia-like phenotype induced by prenatal stress in mice. Biological Psychiatry, 77(6), 589–596. http://doi.org/10.1016/j.biopsych.2014.08.012 Dong, E., Grayson, D. R., Guidotti, A., & Costa, E. (2009). Antipsychotic subtypes can be characterized by differences in their ability to modify GABAergic promoter methylation. Epigenomics, 1(1), 201–11. http://doi.org/10.2217/epi.09.2 Dong, E., Guidotti, A., Grayson, D. R., & Costa, E. (2007). Histone hyperacetylation induces demethylation of reelin and 67-kDa glutamic acid decarboxylase promoters. Proceedings of the National Academy of Sciences of the United States of America, 104(11), 4676–81. http://doi.org/10.1073/pnas.0700529104 Dong, E., Nelson, M., Grayson, D. R., Costa, E., & Guidotti, A. (2008). Clozapine and sulpiride but not haloperidol or olanzapine activate brain DNA demethylation. Proceedings of the National Academy of Sciences of the United States of America, 105(36), 13614–9. http://doi.org/10.1073/pnas.0805493105 Dong, E., Tueting, P., Matrisciano, F., Grayson, D. R., & Guidotti, A. (2016). Behavioral and molecular neuroepigenetic alterations in prenatally stressed mice: relevance for the study of chromatin remodeling properties of antipsychotic drugs. Translational Psychiatry, 6(1), e711. http://doi.org/10.1038/tp.2015.191 Doyle, E. L., Stoddard, B. L., Voytas, D. F., & Bogdanove, A. J. (2013). TAL effectors: Highly adaptable phytobacterial virulence factors and readily engineered DNA-targeting proteins. Trends in Cell Biology. http://doi.org/10.1016/j. tcb.2013.04.003 Drouet, J. B., Fauvelle, F., Maunoir-Regimbal, S., Fidier, N., Maury, R., Peinnequin, A., … Canini, F. (2015). Differences in prefrontal cortex GABA/glutamate ratio after acute restraint stress in rats are associated with specific behavioral and neurobiological patterns. Neuroscience, 285(1873-7544 (Electronic)), 155–165. http://doi.org/10.1016/j. neuroscience.2014.10.058

178

References

Du, P., Zhang, X., Huang, C.-C., Jafari, N., Kibbe, W. A., Hou, L., & Lin, S. M. (2010). Comparison of Beta-value and M-value methods for quantifying methylation levels by microarray analysis. BMC Bioinformatics, 11(1), 587. http:// doi.org/10.1186/1471-2105-11-587 Eckhardt, F., Lewin, J., Cortese, R., Rakyan, V. K., Attwood, J., Burger, M., … Beck, S. (2006). DNA methylation profiling of human chromosomes 6, 20 and 22. Nature Genetics, 38(12), 1378–85. http://doi.org/10.1038/ng1909 Edelman, S., Shalev, I., Uzefovsky, F., Israel, S., Knafo, A., Kremer, I., … Ebstein, R. P. (2012). Epigenetic and Genetic Factors Predict Women’s Salivary Cortisol following a Threat to the Social Self. PLoS ONE, 7(11), e48597. http:// doi.org/10.1371/journal.pone.0048597 Ellenbogen, M. A., Hodgins, S., Walker, C. D., Couture, S., & Adam, S. (2006). Daytime cortisol and stress reactivity in the offspring of parents with bipolar disorder. Psychoneuroendocrinology, 31(10), 1164–1180. http://doi. org/10.1016/j.psyneuen.2006.08.004 Ernst, J., Kheradpour, P., Mikkelsen, T. S., Shoresh, N., Ward, L. D., Epstein, C. B., … Bernstein, B. E. (2011).

Mapping and analysis of chromatin state dynamics in nine human cell types. Nature, 473(1476-4687 (Electronic)), 43–49.

Falcon, S., & Gentleman, R. (2007). Using GOstats to test gene lists for GO term association. Bioinformatics., 23(1367-4811 (Electronic)), 257–258.

Fatemi, M., Alizad, A., & Greenleaf, J. F. (2005). Characteristics of the audio sound generated by ultrasound imaging systems. J.Acoust.Soc.Am., 117(0001-4966 (Print)), 1448–1455.

Ferguson-Smith, A. C. (2011). Genomic imprinting: the emergence of an epigenetic paradigm. Nature Reviews.

Genetics, 12(8), 565–575. http://doi.org/10.1038/nrg3032 First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. W. (2002). Structured Clinical Interview for DSM-IV-TR Axis I Disorders, Research Version, Patient edition with Psychotic Screen (SCID-I/P W/PSY SCREEN). New York: Biometrics Research, New York State Psychiatric Insitute.

Flanagan, R. J., & Dunk, L. (2008). Haematological toxicity of drugs used in psychiatry. Hum.Psychopharmacol., 23 Suppl 1(1099-1077 (Electronic)), 27–41.

Flavell, S. W., & Greenberg, M. E. (2008). Signaling mechanisms linking neuronal activity to gene expression and plasticity of the nervous system. Annu.Rev.Neurosci., 31(0147-006X (Print)), 563–590.

Foley, P., & Kirschbaum, C. (2010). Human hypothalamus-pituitary-adrenal axis responses to acute psychosocial stress in laboratory settings. Neurosci.Biobehav.Rev., 35(1873-7528 (Electronic)), 91–96.

Fortin, J.-P., Labbe, A., Lemire, M., Zanke, B. W., Hudson, T. J., Fertig, E. J., … Hansen, K. D. (2014). Functional normalization of 450k methylation array data improves replication in large cancer studies. Genome Biol., 15(1474760X (Electronic)), 503.

Fraga, M. F., Ballestar, E., Paz, M. F., Ropero, S., Setien, F., Ballestar, M. L., … Esteller, M. (2005). Epigenetic differences arise during the lifetime of monozygotic twins. Proc.Natl.Acad.Sci.U.S.A, 102(0027-8424 (Print)), 10604–10609.

Fries, E., Hellhammer, D. H., & Hellhammer, J. (2006). Attenuation of the hypothalamic-pituitary-adrenal axis responsivity to the Trier Social Stress Test by the benzodiazepine alprazolam. Psychoneuroendocrinology, 31(03064530 (Print)), 1278–1288.

179

9

CHAPTER 9

Frigola, J., Song, J., Stirzaker, C., Hinshelwood, R. A., Peinado, M. A., & Clark, S. J. (2006). Epigenetic remodeling in colorectal cancer results in coordinate gene suppression across an entire chromosome band. Nat.Genet., 38(1061-4036 (Print)), 540–549.

Fritz, M. S., & MacKinnon, D. P. (2007). Required sample size to detect the mediated effect. Psychol.Sci., 18(09567976 (Print)), 233–239.

Fukuchi, M., Nii, T., Ishimaru, N., Minamino, A., Hara, D., Takasaki, I., … Tsuda, M. (2009). Valproic acid induces up- or down-regulation of gene expression responsible for the neuronal excitation and inhibition in rat cortical neurons through its epigenetic actions. Neurosci.Res., 65(1872-8111 (Electronic)), 35–43.

Gaunt, T. R., Shihab, H. A., Hemani, G., Min, J. L., Woodward, G., Lyttleton, O., … Relton, C. L. (2016). Systematic identification of genetic influences on methylation across the human life course. Genome Biology, 17(1), 1–14. http://doi.org/10.1186/s13059-016-0926-z Gibbs, J. R., Brug, M. P., Hernandez, D. G., Traynor, B. J., Nalls, M. A., & Lai, S. L. (2010). Abundant quantitative trait loci exist for DNA methylation and gene expression in human brain. PLoS Genet., 6. http://doi.org/10.1371/journal. pgen.1000952 Giovanoli, S., Engler, H., Engler, A., Richetto, J., Voget, M., Willi, R., … Meyer, U. (2013). Stress in puberty unmasks latent neuropathological consequences of prenatal immune activation in mice [Supplementary Materials]. Science, 339(6123), 1095–9. http://doi.org/10.1126/science.1228261 Gladkevich, A., Kauffman, H. F., & Korf, J. (2004). Lymphocytes as a neural probe: potential for studying psychiatric disorders. Prog.Neuropsychopharmacol.Biol.Psychiatry, 28(0278-5846 (Print)), 559–576.

Goodwin, G. M. (2009). Evidence-based guidelines for treating bipolar disorder: revised second edition-recommendations from the British Association for Psychopharmacology. J.Psychopharmacol., 23(0269-8811 (Print)), 346–388.

Gottlicher, M. (2004). Valproic acid: an old drug newly discovered as inhibitor of histone deacetylases. Ann. Hematol., 83 Suppl 1(0939-5555 (Print)), S91–S92.

Govindaraju, V., Young, K., & Maudsley, A. A. (2000). Proton NMR chemical shifts and coupling constants for brain metabolites. NMR Biomed., 13(0952-3480 (Print)), 129–153.

Grayson, D. R., & Guidotti, A. (2013). The dynamics of DNA methylation in schizophrenia and related psychiatric disorders. Neuropsychopharmacology, 38(1740-634X (Electronic)), 138–166.

Guidotti, A., Dong, E., Kundakovic, M., Satta, R., Grayson, D. R., & Costa, E. (2009). Characterization of the action of antipsychotic subtypes on valproate-induced chromatin remodeling. Trends Pharmacol.Sci., 30(0165-6147 (Print)), 55–60.

Guidotti, A., & Grayson, D. R. (2014). DNA methylation and demethylation as targets for antipsychotic therapy.

Dialogues in Clinical Neuroscience, 16(3), 419–429.

Guintivano, J., Aryee, M. J., & Kaminsky, Z. A. (2013). A cell epigenotype specific model for the correction of brain cellular heterogeneity bias and its application to age, brain region and major depression. Epigenetics., 8(15592308 (Electronic)), 290–302.

Guintivano, J., & Kaminsky, Z. A. (2016). Role of epigenetic factors in the development of mental illness throughout life. Neurosci.Res., 102(1872-8111 (Electronic)), 56–66.

180

References

Gunn, B. G., Brown, A. R., Lambert, J. J., & Belelli, D. (2011). Neurosteroids and GABA(A) Receptor Interactions: A Focus on Stress. Front Neurosci, 5, 131. http://doi.org/10.3389/fnins.2011.00131 Hahne, F., Durinck, S., Ivanek, R., Mueller, A., Lianoglou, S., Tan, G., & Parsons, L. (2015). Gviz:Plotting data and annotation information along genomic coordinates.

Hannon, E., Lunnon, K., Schalkwyk, L. C., & Mill, J. (2015). Interindividual methylomic variation across blood, cortex, and cerebellum: implications for epigenetic studies of neurological and neuropsychiatric phenotypes. Epigenetics., (1559-2308 (Electronic)), 0.

Hannon, E., Spiers, H., Viana, J., Pidsley, R., Burrage, J., Murphy, T. M., … Mill, J. (2016). Methylation QTLs in the developing brain and their enrichment in schizophrenia risk loci. Nat.Neurosci., 19(1546-1726 (Electronic)), 48–54.

Hao, Y., Creson, T., Zhang, L., Li, P., Du, F., Yuan, P., … Chen, G. (2004). Mood stabilizer valproate promotes ERK pathway-dependent cortical neuronal growth and neurogenesis. J.Neurosci., 24(1529-2401 (Electronic)), 6590– 6599.

Harris, H. R., Willett, W. C., & Michels, K. B. (2013). Parental smoking during pregnancy and risk of overweight and obesity in the daughter. International Journal of Obesity (2005), 37(10), 1356–63. http://doi.org/10.1038/ ijo.2013.101 Hasler, G., van der Veen, J. W., Grillon, C., Drevets, W. C., & Shen, J. (2010). Effect of acute psychological stress on prefrontal GABA concentration determined by proton magnetic resonance spectroscopy. Am.J.Psychiatry, 167(1535-7228 (Electronic)), 1226–1231.

Heidt, T., Sager, H. B., Courties, G., Dutta, P., Iwamoto, Y., Zaltsman, A., … Nahrendorf, M. (2014). Chronic variable stress activates hematopoietic stem cells. Nat.Med., 20(1546-170X (Electronic)), 754–758.

Heijmans, B. T., Tobi, E. W., Stein, A. D., Putter, H., Blauw, G. J., Susser, E. S., … Lumey, L. H. (2008). Persistent epigenetic differences associated with prenatal exposure to famine in humans. Proc.Natl.Acad.Sci.U.S.A, 105(10916490 (Electronic)), 17046–17049.

Heim, C. M., Bradley, B., Mletzko, T. C., Deveau, T. C., Musselman, D. L., Nemeroff, C. B., … Binder, E. B. (2009).

Effect of Childhood Trauma on Adult Depression and Neuroendocrine Function: Sex-Specific Moderation by CRH Receptor 1 Gene. Front Behav.Neurosci., 3(1662-5153 (Electronic)), 41.

Heim, C. M., & Nemeroff, C. B. (2001). The role of childhood trauma in the neurobiology of mood and anxiety disorders: preclinical and clinical studies. Biol.Psychiatry, 49(0006-3223 (Print)), 1023–1039.

Heim, C. M., Newport, D. J., Heit, S., Graham, Y. P., Wilcox, M., Bonsall, R., … Nemeroff, C. B. (2000). Pituitaryadrenal and autonomic responses to stress in women after sexual and physical abuse in childhood. JAMA, 284(0098-7484 (Print)), 592–597.

Hernando-Herraez, I., Heyn, H., Fernandez-Callejo, M., Vidal, E., Fernandez-Bellon, H., Prado-Martinez, J., …

Marques-Bonet, T. (2015). The interplay between DNA methylation and sequence divergence in recent human evolution. Nucleic Acids Research, 43(17), 8204–8214. http://doi.org/10.1093/nar/gkv693 Het, S., Rohleder, N., Schoofs, D., Kirschbaum, C., & Wolf, O. T. (2009). Neuroendocrine and psychometric evaluation of a placebo version of the “Trier Social Stress Test.” Psychoneuroendocrinology, 34(1873-3360 (Electronic)), 1075–1086.

Hoffmann, A., Zimmermann, C. A., & Spengler, D. (2015). Molecular epigenetic switches in neurodevelopment in health and disease. Front Behav.Neurosci., 9(1662-5153 (Electronic)), 120.

181

9

CHAPTER 9

Horiuchi, Y., Nakayama, J., Ishiguro, H., Ohtsuki, T., Detera-Wadleigh, S. D., Toyota, T., … Arinami, T. (2004). Possible association between a haplotype of the GABA-A receptor alpha 1 subunit gene (GABRA1) and mood disorders.

Biol.Psychiatry, 55(0006-3223 (Print)), 40–45.

Horvath, S., Zhang, Y., Langfelder, P., Kahn, R. S., Boks, M. P. M., Van Eijk, K. R., … Ophoff, R. A. (2012). Aging effects on DNA methylation modules in human brain and blood tissue. Genome Biol., 13(1465-6914 (Electronic)), R97.

Hostinar, C. E., McQuillan, M. T., Mirous, H. J., Grant, K. E., & Adam, E. K. (2014). Cortisol responses to a group public speaking task for adolescents: variations by age, gender, and race. Psychoneuroendocrinology, 50(18733360 (Electronic)), 155–166.

Houseman, E. A., Accomando, W. P., Koestler, D. C., Christensen, B. C., Marsit, C. J., Nelson, H. H., … Kelsey, K. T. (2012). DNA methylation arrays as surrogate measures of cell mixture distribution. BMC.Bioinformatics., 13(14712105 (Electronic)), 86.

Houseman, E. A., Kim, S., Kelsey, K. T., & Wiencke, J. K. (2015). DNA Methylation in Whole Blood: Uses and Challenges. Curr.Environ.Health Rep., 2(2196-5412 (Electronic)), 145–154.

Hu, Y., Huang, K., An, Q., Du, G., Hu, G., Xue, J., … Fan, G. (2016). Simultaneous profiling of transcriptome and DNA methylome from a single cell. Genome Biology, 17, 88. http://doi.org/10.1186/s13059-016-0950-z Huynh, M., Chee, K., & Lau, D. H. (2005). Thrombotic thrombocytopenic purpura associated with quetiapine. Ann. Pharmacother., 39(1060-0280 (Print)), 1346–1348.

Illingworth, R. S., Gruenewald-Schneider, U., De Sousa, D., Webb, S., Merusi, C., Kerr, A. R., … Bird, A. P. (2015).

Inter-individual variability contrasts with regional homogeneity in the human brain DNA methylome. Nucleic Acids Res., 43(1362-4962 (Electronic)), 732–744.

Irizarry, R. A., Ladd-Acosta, C., Wen, B., Wu, Z., Montano, C., Onyango, P., … Feinberg, A. P. (2009). The human colon cancer methylome shows similar hypo- and hypermethylation at conserved tissue-specific CpG island shores. Nat.Genet., 41(1546-1718 (Electronic)), 178–186.

Ising, M., & Holsboer, F. (2006). Genetics of stress response and stress-related disorders. Dialogues.Clin.Neurosci., 8(1294-8322 (Print)), 433–444.

Jaehne, E. J., Corrigan, F., Toben, C., Jawahar, M. C., & Baune, B. T. (2015). The effect of the antipsychotic drug quetiapine and its metabolite norquetiapine on acute inflammation, memory and anhedonia. Pharmacol.Biochem.

Behav., 135(1873-5177 (Electronic)), 136–144.

Jaffe, A. E., Gao, Y., Deep-Soboslay, A., Tao, R., Hyde, T. M., Weinberger, D. R., & Kleinman, J. E. (2016). Mapping DNA methylation across development, genotype and schizophrenia in the human frontal cortex. Nat.Neurosci., 19(1546-1726 (Electronic)), 40–47.

Jaffe, A. E., Murakami, P., Lee, H., Leek, J. T., Fallin, M. D., Feinberg, A. P., & Irizarry, R. A. (2012). Bump hunting to identify differentially methylated regions in epigenetic epidemiology studies. Int.J.Epidemiol., 41(1464-3685 (Electronic)), 200–209.

Jansen, L. M., Gispen-de Wied, C. C., Gademan, P. J., De Jonge, R. C., van der Linden, J. A., & Kahn, R. S. (1998).

Blunted cortisol response to a psychosocial stressor in schizophrenia. Schizophr.Res., 33(0920-9964 (Print)), 87–94.

Jansen, L. M., Gispen-de Wied, C. C., & Kahn, R. S. (2000). Selective impairments in the stress response in schizophrenic patients. Psychopharmacology (Berl), 149(0033-3158 (Print)), 319–325.

182

References

Jiang, R., Jones, M. J., Chen, E., Neumann, S. M., Fraser, H. B., Miller, G. E., & Kobor, M. S. (2015). Discordance of DNA Methylation Variance Between two Accessible Human Tissues. Scientific Reports, 5, 8257. http://doi. org/10.1038/srep08257 Johnson, W. E., Li, C., & Rabinovic, A. (2007). Adjusting batch effects in microarray expression data using empirical Bayes methods. Biostatistics., 8(1465-4644 (Print)), 118–127.

Jones, P. A. (2012). Functions of DNA methylation: islands, start sites, gene bodies and beyond. Nat.Rev.Genet., 13(1471-0064 (Electronic)), 484–492.

Joubert, B. R., Felix, J. F., Yousefi, P., Bakulski, K. M., Just, A. C., Breton, C., … London, S. J. (2016). DNA Methylation in Newborns and Maternal Smoking in Pregnancy: Genome-wide Consortium Meta-analysis. The American Journal of Human Genetics, 680–696. http://doi.org/10.1016/j.ajhg.2016.02.019 Karst, H., Berger, S., Turiault, M., Tronche, F., Schutz, G., & Joels, M. (2005). Mineralocorticoid receptors are indispensable for nongenomic modulation of hippocampal glutamate transmission by corticosterone. Proceedings of the National Academy of Sciences, 102(52), 19204–19207. http://doi.org/10.1073/pnas.0507572102 Karst, H., & Joëls, M. (2005). Corticosterone slowly enhances miniature excitatory postsynaptic current amplitude in mice CA1 hippocampal cells. Journal of Neurophysiology, 94(5), 3479–3486. http://doi.org/10.1152/ jn.00143.2005 Kent, W. J., Sugnet, C. W., Furey, T. S., Roskin, K. M., Pringle, T. H., Zahler, A. M., & Haussler, D. (2002). The human genome browser at UCSC. Genome Res., 12(1088-9051 (Print)), 996–1006.

Kessler, R. C., McLaughlin, K. A., Green, J. G., Gruber, M. J., Sampson, N. A., Zaslavsky, A. M., … Williams, D. R. (2010). Childhood adversities and adult psychopathology in the WHO World Mental Health Surveys. Br.J.Psychiatry, 197(1472-1465 (Electronic)), 378–385.

Kim, B., Rincon Castro, L. M., Jawed, S., & Niles, L. P. (2008). Clinically relevant concentrations of valproic acid modulate melatonin MT(1) receptor, HDAC and MeCP2 mRNA expression in C6 glioma cells. Eur.J.Pharmacol., 589(0014-2999 (Print)), 45–48.

Kim, J. K., Samaranayake, M., & Pradhan, S. (2009). Epigenetic mechanisms in mammals. Cell Mol.Life Sci., 66(14209071 (Electronic)), 596–612.

Kinoshita, M., Numata, S., Tajima, A., Shimodera, S., Ono, S., Imamura, A., … Ohmori, T. (2013). DNA methylation signatures of peripheral leukocytes in schizophrenia. Neuromolecular.Med., 15(1559-1174 (Electronic)), 95–101.

Kirschbaum, C., Kudielka, B. M., Gaab, J., Schommer, N. C., & Hellhammer, D. H. (1999). Impact of gender, menstrual cycle phase, and oral contraceptives on the activity of the hypothalamus-pituitary-adrenal axis. Psychosom.Med., 61(0033-3174 (Print)), 154–162.

Kirschbaum, C., Pirke, K. M., & Hellhammer, D. H. (1993). The “Trier Social Stress Test”--a tool for investigating psychobiological stress responses in a laboratory setting. Neuropsychobiology, 28(0302-282X (Print)), 76–81.

Klengel, T., Mehta, D., Anacker, C., Rex-Haffner, M., Pruessner, J. C., Pariante, C. M., … Binder, E. B. (2013). Allelespecific FKBP5 DNA demethylation mediates gene-childhood trauma interactions. Nat.Neurosci., 16(1546-1726 (Electronic)), 33–41.

Klose, R. J., & Bird, A. P. (2006). Genomic DNA methylation: the mark and its mediators. Trends Biochem.Sci., 31(0968-0004 (Print)), 89–97.

183

9

CHAPTER 9

Kofink, D., Boks, M. P. M., Timmers, H. T., & Kas, M. J. H. (2013). Epigenetic dynamics in psychiatric disorders: environmental programming of neurodevelopmental processes. Neurosci.Biobehav.Rev., 37(1873-7528 (Electronic)), 831–845.

Kogler, L., Muller, V. I., Chang, A., Eickhoff, S. B., Fox, P. T., Gur, R. C., & Derntl, B. (2015). Psychosocial versus physiological stress - Meta-analyses on deactivations and activations of the neural correlates of stress reactions.

Neuroimage., 119(1095-9572 (Electronic)), 235–251.

Kowatch, R. A., Fristad, M., Birmaher, B., Wagner, K. D., Findling, R. L., & Hellander, M. (2005). Treatment guidelines for children and adolescents with bipolar disorder. J.Am.Acad.Child Adolesc.Psychiatry, 44(0890-8567 (Print)), 213–235.

Kreis, R., Ernst, T., & Ross, B. D. (1993). Development of the human brain: in vivo quantification of metabolite and water content with proton magnetic resonance spectroscopy. Magn Reson.Med., 30(0740-3194 (Print)), 424–437.

Kudielka, B. M., Fischer, J. E., Metzenthin, P., Helfricht, S., Preckel, D., & von, K. R. (2007). No effect of 5-day treatment with acetylsalicylic acid (aspirin) or the beta-blocker propranolol (Inderal) on free cortisol responses to acute psychosocial stress: a randomized double-blind, placebo-controlled study. Neuropsychobiology, 56(1423-0224 (Electronic)), 159–166.

Kundaje, A., Meuleman, W., Ernst, J., Bilenky, M., Yen, A., Heravi-Moussavi, A., … Kellis, M. (2015). Integrative analysis of 111 reference human epigenomes. Nature, 518(1476-4687 (Electronic)), 317–330.

Kundakovic, M., Gudsnuk, K., Herbstman, J. B., Tang, D., Perera, F. P., & Champagne, F. A. (2015). DNA methylation of BDNF as a biomarker of early-life adversity. Proceedings of the National Academy of Sciences of the United States of America, 112(22), 6807–13. http://doi.org/10.1073/pnas.1408355111 Labonte, B., Suderman, M., Maussion, G., Navaro, L., Yerko, V., Mahar, I., … Turecki, G. (2012). Genome-wide epigenetic regulation by early-life trauma. Arch.Gen.Psychiatry, 69(1538-3636 (Electronic)), 722–731.

Laird, P. W. (2010). Principles and challenges of genomewide DNA methylation analysis. Nat Rev Genet, 11(3), 191–203. http://doi.org/10.1038/nrg2732 Lake, C. R., Pickar, D., Ziegler, M. G., Lipper, S., Slater, S., & Murphy, D. L. (1982). High plasma norepinephrine levels in patients with major affective disorder. Am.J.Psychiatry, 139(0002-953X (Print)), 1315–1318.

Lam, L. L., Emberly, E., Fraser, H. B., Neumann, S. M., Chen, E., Miller, G. E., & Kobor, M. S. (2012). Factors underlying variable DNA methylation in a human community cohort. Proc.Natl.Acad.Sci.U.S.A, 109 Suppl(1091-6490 (Electronic)), 17253–17260.

Lange, C., Zschucke, E., Ising, M., Uhr, M., Bermpohl, F., & Adli, M. (2013). Evidence for a normal HPA axis response to psychosocial stress in patients remitted from depression. Psychoneuroendocrinology, 38(1873-3360 (Electronic)), 2729–2736.

Langfelder, P., & Horvath, S. (2008). WGCNA: an R package for weighted correlation network analysis. BMC. Bioinformatics., 9(1471-2105 (Electronic)), 559.

Langfelder, P., & Horvath, S. (2012). Fast R Functions for Robust Correlations and Hierarchical Clustering. J.Stat.

Softw., 46(1548-7660 (Print)).

Leek, J. T., Johnson, W. E., Parker, H. S., Jaffe, A. E., & Storey, J. D. (2014). sva: Surrogate Variable Analysis R package version 3.10.0. Http://www.bioconductor.org/packages/release/bioc/html/sva.html.

184

References

Lemire, M., Zaidi, S. H., Ban, M., Ge, B., Aissi, D., Germain, M., … Hudson, T. J. (2015). Long-range epigenetic regulation is conferred by genetic variation located at thousands of independent loci. Nat.Commun., 6(2041-1723 (Electronic)), 6326.

Leng, Y., Liang, M. H., Ren, M., Marinova, Z., Leeds, P., & Chuang, D. M. (2008). Synergistic neuroprotective effects of lithium and valproic acid or other histone deacetylase inhibitors in neurons: roles of glycogen synthase kinase-3 inhibition. J.Neurosci., 28(1529-2401 (Electronic)), 2576–2588.

Lennartsson, J., & Ronnstrand, L. (2012). Stem cell factor receptor/c-Kit: from basic science to clinical implications.

Physiol Rev., 92(1522-1210 (Electronic)), 1619–1649.

Leung, D., Jung, I., Rajagopal, N., Schmitt, A. D., Selvaraj, S., Lee, A. Y., … Ren, B. (2015). Integrative analysis of haplotype-resolved epigenomes across human tissues. Nature, 518(7539), 350–354. http://doi.org/10.1038/ nature14217 Li, J., Guo, Y., Schroeder, F. A., Youngs, R. M., Schmidt, T. W., Ferris, C., … Akbarian, S. (2004). Dopamine D2-like antagonists induce chromatin remodeling in striatal neurons through cyclic AMP-protein kinase A and NMDA receptor signaling. J.Neurochem., 90(0022-3042 (Print)), 1117–1131.

Linthorst, A. C. E., & Reul, J. M. (2008). Stress and the brain: Solving the puzzle using microdialysis. Pharmacology Biochemistry and Behavior. http://doi.org/10.1016/j.pbb.2007.09.019 Lipska, B. K., Deep-Soboslay, A., Weickert, C. S., Hyde, T. M., Martin, C. E., Herman, M. M., & Kleinman, J. E. (2006).

Critical Factors in Gene Expression in Postmortem Human Brain: Focus on Studies in Schizophrenia. Biological Psychiatry, 60(6), 650–658. http://doi.org/10.1016/j.biopsych.2006.06.019 Lister, R., Pelizzola, M., Dowen, R. H., Hawkins, R. D., Hon, G., Tonti-Filippini, J., … Ecker, J. R. (2009). Human DNA methylomes at base resolution show widespread epigenomic differences. Nature, 462(1476-4687 (Electronic)), 315–322.

Liu, K., Liu, Y., Lau, J. L., & Min, J. (2015). Epigenetic targets and drug discovery Part 2: Histone demethylation and DNA methylation. Pharmacology and Therapeutics, 151, 121–140. http://doi.org/10.1016/j. pharmthera.2015.04.001 Lountos, G. T., Cherry, S., Tropea, J. E., & Waugh, D. S. (2015). Structural analysis of human dual-specificity phosphatase 22 complexed with a phosphotyrosine-like substrate. Acta Crystallogr.F.Struct.Biol.Commun., 71(2053-230X (Electronic)), 199–205.

Lovallo, W. R., Farag, N. H., Sorocco, K. H., Cohoon, A. J., & Vincent, A. S. (2012). Lifetime adversity leads to blunted stress axis reactivity: studies from the Oklahoma Family Health Patterns Project. Biol.Psychiatry, 71(1873-2402 (Electronic)), 344–349.

Lowe, R., Gemma, C., Beyan, H., Hawa, M. I., Bazeos, A., Leslie, R. D., … Ramagopalan, S. V. (2013). Buccals are likely to be a more informative surrogate tissue than blood for epigenome-wide association studies. Epigenetics : Official Journal of the DNA Methylation Society, 8(4), 445–454. http://doi.org/10.4161/epi.24362 Lunnon, K., Hannon, E., Smith, R. G., Dempster, E. L., Wong, C. C. Y., Burrage, J., … Mill, J. (2016). Variation in 5-hydroxymethylcytosine across human cortex and cerebellum. Genome Biology, 17, 27. http://doi.org/10.1186/ s13059-016-0871-x Luo, C., Xu, H., & Li, X. M. (2005). Quetiapine reverses the suppression of hippocampal neurogenesis caused by repeated restraint stress. Brain Res., 1063(0006-8993 (Print)), 32–39.

185

9

CHAPTER 9

Lupien, S. J., McEwen, B. S., Gunnar, M. R., & Heim, C. M. (2009). Effects of stress throughout the lifespan on the brain, behaviour and cognition. Nat.Rev.Neurosci., 10(1471-0048 (Electronic)), 434–445.

Lutz, P. E., & Turecki, G. (2014). DNA methylation and childhood maltreatment: From animal models to human studies. Neuroscience. http://doi.org/10.1016/j.neuroscience.2013.07.069 Maeyama, H., Hirasawa, T., Tahara, Y., Obata, C., Kasai, H., Moriishi, K., … Kubota, T. (2015). Maternal restraint stress during pregnancy in mice induces 11β-HSD1-associated metabolic changes in the livers of the offspring. Journal of Developmental Origins of Health and Disease, 6(02), 105–114. http://doi.org/10.1017/S2040174415000100 Maheu, F. S., Joober, R., & Lupien, S. J. (2005). Declarative memory after stress in humans: differential involvement of the beta-adrenergic and corticosteroid systems. J.Clin.Endocrinol.Metab, 90(0021-972X (Print)), 1697–1704.

Makatsori, A., Duncko, R., Moncek, F., Loder, I., Katina, S., & Jezova, D. (2004). Modulation of neuroendocrine response and non-verbal behavior during psychosocial stress in healthy volunteers by the glutamate releaseinhibiting drug lamotrigine. Neuroendocrinology, 79(0028-3835 (Print)), 34–42.

Marjanska, M., Auerbach, E. J., Valabregue, R., Van de Moortele, P. F., Adriany, G., & Garwood, M. (2012). Localized 1H NMR spectroscopy in different regions of human brain in vivo at 7 T: T2 relaxation times and concentrations of cerebral metabolites. NMR Biomed., 25(1099-1492 (Electronic)), 332–339.

Marsman, A., Mandl, R. C., Van Den Heuvel, M. P., Boer, V. O., Wijnen, J. P., Klomp, D. W. J., … Hulshoff-Pol, H. E. (2013). Glutamate changes in healthy young adulthood. Eur.Neuropsychopharmacol., 23(1873-7862 (Electronic)), 1484–1490.

Martin, D. I. K., Singer, M., Dhahbi, J., Mao, G., Zhang, L., Schroth, G. P., … Boffelli, D. (2011). Phyloepigenomic comparison of great apes reveals a correlation between somatic and germline methylation states. Genome Research, 21(12), 2049–2057. http://doi.org/10.1101/gr.122721.111 Mathers, J. C., Strathdee, G., & Relton, C. L. (2010). Induction of Epigenetic Alterations by Dietary and Other Environmental Factors. Advances in Genetics (Vol. 71). http://doi.org/10.1016/B978-0-12-380864-6.00001-8 Matrisciano, F., Dong, E., Gavin, D. P., Nicoletti, F., & Guidotti, A. (2011). Activation of group II metabotropic glutamate receptors promotes DNA demethylation in the mouse brain. Mol.Pharmacol., 80(1521-0111 (Electronic)), 174–182.

Matrisciano, F., Tueting, P., Dalal, I., Kadriu, B., Grayson, D. R., Davis, J. M., … Guidotti, A. (2013). Epigenetic modifications of GABAergic interneurons are associated with the schizophrenia-like phenotype induced by prenatal stress in mice. Neuropharmacology, 68(1873-7064 (Electronic)), 184–194.

Maunakea, A. K., Nagarajan, R. P., Bilenky, M., Ballinger, T. J., D’Souza, C., Fouse, S. D., … Costello, J. F. (2010).

Conserved role of intragenic DNA methylation in regulating alternative promoters. Nature, 466(1476-4687 (Electronic)), 253–257.

McEwen, B. S. (2004). Protection and damage from acute and chronic stress: allostasis and allostatic overload and relevance to the pathophysiology of psychiatric disorders. Ann.N.Y.Acad.Sci., 1032(0077-8923 (Print)), 1–7.

McGowan, P. O., Sasaki, A., D’Alessio, A. C., Dymov, S., Labonte, B., Szyf, M., … Meaney, M. J. (2009). Epigenetic regulation of the glucocorticoid receptor in human brain associates with childhood abuse. Nat.Neurosci., 12(15461726 (Electronic)), 342–348.

McGregor, K., Bernatsky, S., Colmegna, I., Hudson, M., Pastinen, T., Labbe, A., & Greenwood, C. M. T. (2016). An evaluation of methods correcting for cell-type heterogeneity in DNA methylation studies. Genome Biology, 17, 84. http://doi.org/10.1186/s13059-016-0935-y

186

References

McKlveen, J. M., Myers, B., & Herman, J. P. (2015). The medial prefrontal cortex: coordinator of autonomic, neuroendocrine and behavioural responses to stress. J.Neuroendocrinol., 27(1365-2826 (Electronic)), 446–456.

McRae, A. F., Powell, J. E., Henders, A. K., Bowdler, L., Hemani, G., & Shah, S. (2014). Contribution of genetic variation to transgenerational inheritance of DNA methylation. Genome Biol., 15. http://doi.org/10.1186/gb-201415-5-r73 Mehta, D., Klengel, T., Conneely, K. N., Smith, A. K., Altmann, A., Pace, T. W., … Binder, E. B. (2013). Childhood maltreatment is associated with distinct genomic and epigenetic profiles in posttraumatic stress disorder. Proc.

Natl.Acad.Sci.U.S.A, 110(1091-6490 (Electronic)), 8302–8307.

Melhem, N. M., Keilp, J. G., Porta, G., Oquendo, M. A., Burke, A., Stanley, B., … Brent, D. A. (2015). Blunted HPA Axis Activity in Suicide Attempters Compared to those at High Risk for Suicidal Behavior. Neuropsychopharmacology, (1740-634X (Electronic)).

Melka, M. G., Laufer, B. I., McDonald, P., Castellani, C. A., Rajakumar, N., O’Reilly, R., & Singh, S. M. (2014). The effects of olanzapine on genome-wide DNA methylation in the hippocampus and cerebellum. Clin.Epigenetics., 6(1868-7075 (Print)), 1.

Mill, J., Tang, T., Kaminsky, Z. A., Khare, T., Yazdanpanah, S., Bouchard, L., … Petronis, A. (2008). Epigenomic profiling reveals DNA-methylation changes associated with major psychosis. Am.J.Hum.Genet., 82(1537-6605 (Electronic)), 696–711.

Miller, B. R., & Hen, R. (2015). The current state of the neurogenic theory of depression and anxiety. Curr.Opin.

Neurobiol., 30(1873-6882 (Electronic)), 51–58.

Miller, G. A., & Chapman, J. P. (2001). Misunderstanding analysis of covariance. J.Abnorm.Psychol., 110(0021-843X (Print)), 40–48.

Minucci, S., & Pelicci, P. G. (2006). Histone deacetylase inhibitors and the promise of epigenetic (and more) treatments for cancer. Nat.Rev.Cancer, 6(1474-175X (Print)), 38–51.

Mitchell, C. P., Chen, Y., Kundakovic, M., Costa, E., & Grayson, D. R. (2005). Histone deacetylase inhibitors decrease reelin promoter methylation in vitro. J.Neurochem., 93(0022-3042 (Print)), 483–492.

Mody, I., & Maguire, J. (2012). The reciprocal regulation of stress hormones and GABAA receptors. Frontiers in Cellular Neuroscience, 6(January), 1–6. http://doi.org/10.3389/fncel.2012.00004 Moleres, A., Campion, J., Milagro, F. I., Marcos, A., Campoy, C., Garagorri, J. M., … Marti, A. (2013). Differential DNA methylation patterns between high and low responders to a weight loss intervention in overweight or obese adolescents: the EVASYON study. FASEB J., 27(1530-6860 (Electronic)), 2504–2512.

Montano, C., Taub, M. A., Jaffe, A. E., Briem, E., Feinberg, J. I., Trygvadottir, R., … Feinberg, A. P. (2016). Association of dna methylation differences with schizophrenia in an epigenome-wide association study. JAMA Psychiatry, 73(5), 506–514. Retrieved from http://dx.doi.org/10.1001/jamapsychiatry.2016.0144 Moore, L. D., Le, T., & Fan, G. (2013). DNA methylation and its basic function. Neuropsychopharmacology : Official Publication of the American College of Neuropsychopharmacology, 38(1), 23–38. http://doi.org/10.1038/ npp.2012.112 Morris, M. C., Rao, U., & Garber, J. (2012). Cortisol responses to psychosocial stress predict depression trajectories: social-evaluative threat and prior depressive episodes as moderators. J.Affect.Disord., 143(1573-2517 (Electronic)), 223–230.

187

9

CHAPTER 9

Mullins, P. G., McGonigle, D. J., O’Gorman, R. L., Puts, N. A., Vidyasagar, R., Evans, C. J., & Edden, R. A. (2014).

Current practice in the use of MEGA-PRESS spectroscopy for the detection of GABA. Neuroimage., 86(1095-9572 (Electronic)), 43–52.

Nater, U. M., Hoppmann, C. A., & Scott, S. B. (2013). Diurnal profiles of salivary cortisol and alpha-amylase change across the adult lifespan: Evidence from repeated daily life assessments. Psychoneuroendocrinology, (1873-3360 (Electronic)).

Nater, U. M., & Rohleder, N. (2009). Salivary alpha-amylase as a non-invasive biomarker for the sympathetic nervous system: Current state of research. Psychoneuroendocrinology. http://doi.org/10.1016/j.psyneuen.2009.01.014 Nater, U. M., Rohleder, N., Schlotz, W., Ehlert, U., & Kirschbaum, C. (2007). Determinants of the diurnal course of salivary alpha-amylase. Psychoneuroendocrinology, 32(0306-4530 (Print)), 392–401.

Nemoto, T., Kakinuma, Y., & Shibasaki, T. (2015). Impaired miR449a-induced downregulation of Crhr1 expression in low-birth-weight rats. J.Endocrinol., 224(1479-6805 (Electronic)), 195–203.

Niculescu, M. D., & Zeisel, S. H. (2002). Diet, Methyl Donors and DNA Methylation : Interactions between Dietary Folate, Methinine and Choline. The Journal of Nutrition, 132, 2333–2335.

Nievergelt, C. M., Maihofer, A. X., Mustapic, M., Yurgil, K. A., Schork, N. J., Miller, M. W., … Baker, D. G. (2015).

Genomic predictors of combat stress vulnerability and resilience in U.S. Marines: A genome-wide association study across multiple ancestries implicates PRTFDC1 as a potential PTSD gene. Psychoneuroendocrinology, 51(1873-3360 (Electronic)), 459–471.

Nugent, K. L., Chiappelli, J., Rowland, L. M., & Hong, L. E. (2015). Cumulative stress pathophysiology in schizophrenia as indexed by allostatic load. Psychoneuroendocrinology, 60(1873-3360 (Electronic)), 120–129.

O’Donovan, M. C., Craddock, N. J., & Owen, M. J. (2009). Genetics of psychosis; insights from views across the genome. Hum.Genet., 126(1432-1203 (Electronic)), 3–12.

Oberlander, T. F., Weinberg, J., Papsdorf, M., Grunau, R., Misri, S., & Devlin, A. M. (2008). Prenatal exposure to maternal depression, neonatal methylation of human glucocorticoid receptor gene (NR3C1) and infant cortisol stress responses. Epigenetics., 3(1559-2308 (Electronic)), 97–106.

Ostiguy, C. S., Ellenbogen, M. A., Walker, C. D., Walker, E. F., & Hodgins, S. (2011). Sensitivity to stress among the offspring of parents with bipolar disorder: a study of daytime cortisol levels. Psychol.Med., 41(1469-8978 (Electronic)), 2447–2457.

Otero Losada, M. E. (1988). Changes in central GABAergic function following acute and repeated stress.

Br.J.Pharmacol., 93(0007-1188 (Print)), 483–490.

Ouyang, B., Baxter, C. S., Lam, H. M., Yeramaneni, S., Levin, L., Haynes, E., & Ho, S. M. (2012). Hypomethylation of dual specificity phosphatase 22 promoter correlates with duration of service in firefighters and is inducible by low-dose benzo[a]pyrene. J.Occup.Environ.Med., 54(1536-5948 (Electronic)), 774–780.

Oyesanmi, O., Kunkel, E. J., Monti, D. A., & Field, H. L. (1999). Hematologic side effects of psychotropics.

Psychosomatics, 40(0033-3182 (Print)), 414–421.

Pace, T. W., Mletzko, T. C., Alagbe, O., Musselman, D. L., Nemeroff, C. B., Miller, A. H., & Heim, C. M. (2006).

Increased stress-induced inflammatory responses in male patients with major depression and increased early life stress. Am.J.Psychiatry, 163(0002-953X (Print)), 1630–1633.

188

References

Peedicayil, J. (2008). Epigenetic biomarkers in psychiatric disorders. British Journal of Pharmacology, 155(6), 795–6. http://doi.org/10.1038/bjp.2008.254 Perisic, T., Zimmermann, N., Kirmeier, T., Asmus, M., Tuorto, F., Uhr, M., … Zschocke, J. (2010). Valproate and amitriptyline exert common and divergent influences on global and gene promoter-specific chromatin modifications in rat primary astrocytes. Neuropsychopharmacology, 35(1740-634X (Electronic)), 792–805.

Perroud, N., Paoloni-Giacobino, A., Prada, P., Olie, E., Salzmann, A., Nicastro, R., … Malafosse, A. (2011). Increased methylation of glucocorticoid receptor gene (NR3C1) in adults with a history of childhood maltreatment: a link with the severity and type of trauma. Transl.Psychiatry, 1(2158-3188 (Electronic)), e59.

Petrowski, K., Wintermann, G. B., Schaarschmidt, M., Bornstein, S. R., & Kirschbaum, C. (2013). Blunted salivary and plasma cortisol response in patients with panic disorder under psychosocial stress. Int.J.Psychophysiol., 88(1872-7697 (Electronic)), 35–39.

Petty, F., & Sherman, A. D. (1981). GABAergic modulation of learned helplessness. Pharmacol.Biochem.Behav., 15(0091-3057 (Print)), 567–570.

Phipson, B., Maksimovic, J., & Oshlack, A. (2015). missMethyl: an R package for analyzing data from Illumina’s HumanMethylation450 platform. Bioinformatics., (1367-4811 (Electronic)).

Pidsley, R., Viana, J., Hannon, E., Spiers, H., Troakes, C., Al-Sarraj, S., … Mill, J. (2014). Methylomic profiling of human brain tissue supports a neurodevelopmental origin for schizophrenia. Genome Biol., 15(1474-760X (Electronic)), 483.

Popkie, A. P., Zeidner, L. C., Albrecht, A. M., D’Ippolito, A., Eckardt, S., Newsom, D. E., … Phiel, C. J. (2010).

Phosphatidylinositol 3-kinase (PI3K) signaling via glycogen synthase kinase-3 (Gsk-3) regulates DNA methylation of imprinted loci. J.Biol.Chem., 285(1083-351X (Electronic)), 41337–41347.

Popoli, M., Yan, Z., McEwen, B. S., & Sanacora, G. (2012). The stressed synapse: the impact of stress and glucocorticoids on glutamate transmission. Nat.Rev.Neurosci., 13(1471-0048 (Electronic)), 22–37.

Post, R. M. (1992). Transduction of psychosocial stress into the neurobiology of recurrent affective disorder.

Am.J.Psychiatry, 149(0002-953X (Print)), 999–1010.

Proudfoot, J., Doran, J., Manicavasagar, V., & Parker, G. (2011). The precipitants of manic/hypomanic episodes in the context of bipolar disorder: a review. J.Affect.Disord., 133(1573-2517 (Electronic)), 381–387.

Provencher, S. W. (1993). Estimation of metabolite concentrations from localized in vivo proton NMR spectra.

Magn Reson.Med., 30(0740-3194 (Print)), 672–679.

Provencher, S. W. (2015). LCModel & LCMgui User’s Manual. S-Provencher.com/pages/lcm-Manual.shtml. Retrieved from http://s-provencher.com/pub/LCModel/manual/manual.pdf Pruessner, J. C., Kirschbaum, C., Meinlschmid, G., & Hellhammer, D. H. (2003). Two formulas for computation of the area under the curve represent measures of total hormone concentration versus time-dependent change.

Psychoneuroendocrinology, 28(0306-4530 (Print)), 916–931.

Punzi, G., Ursini, G., Shin, J. H., Kleinman, J. E., Hyde, T. M., & Weinberger, D. R. (2014). Increased expression of MARCKS in post-mortem brain of violent suicide completers is related to transcription of a long, noncoding, antisense RNA. Mol.Psychiatry, (1476-5578 (Electronic)).

189

9

CHAPTER 9

Rakyan, V. K., Down, T. A., Thorne, N. P., Flicek, P., Kulesha, E., Gräf, S., … Beck, S. (2008). An integrated resource for genome-wide identification and analysis of human tissue-specific differentially methylated regions (tDMRs).

Genome Res., 18(1088-9051 (Print)), 1518–1529.

Ran, F. A., Hsu, P. D., Wright, J., Agarwala, V., Scott, D. A., & Zhang, F. (2013). Genome engineering using the CRISPR-Cas9 system. Nature Protocols, 8(11), 2281–2308. http://doi.org/10.1038/nprot.2013.143 Rao, S. S., Huntley, M. H., Durand, N. C., Stamenova, E. K., Bochkov, I. D., Robinson, J. T., … Aiden, E. L. (2014). A 3D map of the human genome at kilobase resolution reveals principles of chromatin looping. Cell, 159(1097-4172 (Electronic)), 1665–1680.

R-Core-Team. (2014). R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing. Retrieved from http://www.r-project.org/ Reik, W. (2007). Stability and flexibility of epigenetic gene regulation in mammalian development. Nature, 447(14764687 (Electronic)), 425–432.

Reik, W., Dean, W., & Walter, J. (2001). Epigenetic reprogramming in mammalian development. Science, 293(00368075 (Print)), 1089–1093.

Reik, W., & Lewis, A. (2005). Co-evolution of X-chromosome inactivation and imprinting in mammals. Nature Reviews. Genetics, 6(5), 403–10. http://doi.org/10.1038/nrg1602 Reinius, L. E., Acevedo, N., Joerink, M., Pershagen, G., Dahlen, S. E., Greco, D., … Kere, J. (2012). Differential DNA methylation in purified human blood cells: implications for cell lineage and studies on disease susceptibility. PLoS. One., 7(1932-6203 (Electronic)), e41361.

Reus, G. Z., Fries, G. R., Stertz, L., Badawy, M., Passos, I. C., Barichello, T., … Quevedo, J. (2015). The role of inflammation and microglial activation in the pathophysiology of psychiatric disorders. Neuroscience, 300(18737544 (Electronic)), 141–154.

Rinaldi, A., Vincenti, S., De Vito, F., Bozzoni, I., Oliverio, A., Presutti, C., … Mele, A. (2010). Stress induces region specific alterations in microRNAs expression in mice. Behav.Brain Res., 208(1872-7549 (Electronic)), 265–269.

Ritsner, M., Gibel, A., Maayan, R., Ratner, Y., Ram, E., Biadsy, H., … Weizman, A. (2005). Cortisol/ dehydroepiandrosterone ratio and responses to antipsychotic treatment in schizophrenia.

Neuropsychopharmacology, 30(0893-133X (Print)), 1913–1922.

Roelofs, K., van Peer, J., Berretty, E., de Jong, P., Spinhoven, P., & Elzinga, B. M. (2009). Hypothalamus-pituitaryadrenal axis hyperresponsiveness is associated with increased social avoidance behavior in social phobia. Biol.

Psychiatry, 65(1873-2402 (Electronic)), 336–343.

Rohleder, N., Nater, U. M., Wolf, J. M., Ehlert, U., & Kirschbaum, C. (2004). Psychosocial stress-induced activation of salivary alpha-amylase: An indicator of sympathetic activity? In Annals of the New York Academy of Sciences (Vol. 1032, pp. 258–263). http://doi.org/10.1196/annals.1314.033 Roseboom, T. J., De Rooij, S. R., & Painter, R. C. (2006). The Dutch famine and its long-term consequences for adult health. Early Hum.Dev., 82(0378-3782 (Print)), 485–491.

Roseboom, T. J., Painter, R. C., Van Abeelen, A. F. M., Veenendaal, M. V. E., & De Rooij, S. R. (2011). Hungry in the womb: What are the consequences? Lessons from the Dutch famine. Maturitas. http://doi.org/10.1016/j. maturitas.2011.06.017

190

References

Roth, T. L., & Sweatt, J. D. (2009). Regulation of chromatin structure in memory formation. Curr.Opin.Neurobiol., 19(1873-6882 (Electronic)), 336–342.

Ruiz, M. A., Peters, R. H., Sanchez, G. M., & Bates, J. P. (2009). Psychometric Properties of the Mental Health Screening Form Iii Within a Metropolitan Jail. Criminal Justice and Behavior, 36(6), 607–619. http://doi. org/10.1177/0093854809334013 Rush, A. J., Giles, D. E., Schlesser, M. A., Orsulak, P. J., Parker Jr., C. R., Weissenburger, J. E., … Vasavada, N. (1996).

The dexamethasone suppression test in patients with mood disorders. J.Clin.Psychiatry, 57(0160-6689 (Print)), 470–484.

Rush, A. J., Gullion, C. M., Basco, M. R., Jarrett, R. B., & Trivedi, M. H. (1996). The Inventory of Depressive Symptomatology (IDS): psychometric properties. Psychol.Med., 26(0033-2917 (Print)), 477–486.

Rutten, B. P., & Mill, J. (2009). Epigenetic mediation of environmental influences in major psychotic disorders.

Schizophr.Bull., 35(1745-1701 (Electronic)), 1045–1056.

Rybakowski, J. K., & Twardowska, K. (1999). The dexamethasone/corticotropin-releasing hormone test in depression in bipolar and unipolar affective illness. J.Psychiatr.Res., 33(0022-3956 (Print)), 363–370.

Sanchez-Mut, J. V., Aso, E., Heyn, H., Matsuda, T., Bock, C., Ferrer, I., & Esteller, M. (2014). Promoter hypermethylation of the phosphatase DUSP22 mediates PKA-dependent TAU phosphorylation and CREB activation in Alzheimer’s disease. Hippocampus, 24(1098-1063 (Electronic)), 363–368.

Schalkwyk, L. C., Pidsley, R., Wong, C. C. Y., By, with functions contributed, Touleimat, N., Defrance, M., …

Maksimovic, J. (2013). wateRmelon: Illumina 450 methylation array normalization and metrics. R package version 1.4.0. http://www.bioconductor.org/packages/release/bioc/html/wateRmelon.html.

Schizophrenia Working Group of the Psychiatric Genomics Consortium. (2014). Biological insights from 108 schizophrenia-associated genetic loci. Nature, 511(1476-4687 (Electronic)), 421–427.

Schmider, J., Lammers, C. H., Gotthardt, U., Dettling, M., Holsboer, F., & Heuser, I. J. (1995). Combined dexamethasone/corticotropin-releasing hormone test in acute and remitted manic patients, in acute depression, and in normal controls: I. Biol.Psychiatry, 38(0006-3223 (Print)), 797–802.

Schubeler, D. (2015). Function and information content of DNA methylation. Nature, 517(1476-4687 (Electronic)), 321–326.

Schultz, M. D., He, Y., Whitaker, J. W., Hariharan, M., Mukamel, E. A., Leung, D., … Ecker, J. R. (2015). Human body epigenome maps reveal noncanonical DNA methylation variation. Nature, 523(1476-4687 (Electronic)), 212–216.

Schür, R. R., Draisma, L. W. R., Wijnen, J. P., Boks, M. P. M., Koevoets, M. G. J. C., Joëls, M., … Vinkers, C. H. (2016).

Brain GABA levels across psychiatric disorders: A systematic literature review and meta-analysis of 1H-MRS studies.

Human Brain Mapping, n/a–n/a. http://doi.org/10.1002/hbm.23244 Schwartzman, O., & Tanay, A. (2015). Single-cell epigenomics: techniques and emerging applications. Nat Rev Genet, 16(12), 716–726. Retrieved from http://dx.doi.org/10.1038/nrg3980 Serretti, A., & Mandelli, L. (2008). The genetics of bipolar disorder: genome “hot regions,” genes, new potential candidates and future directions. Mol.Psychiatry, 13(1476-5578 (Electronic)), 742–771.

Shankar, B. R. (2007). Quetiapine-induced leucopenia and thrombocytopenia. Psychosomatics, 48(0033-3182 (Print)), 530–531.

191

9

CHAPTER 9

Sheehan, D. V, Lecrubier, Y., Sheehan, K. H., Amorim, P., Janavs, J., Weiller, E., … Dunbar, G. C. (1998). The MiniInternational Neuropsychiatric Interview (M.I.N.I.): the development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. J.Clin.Psychiatry, 59 Suppl 2(0160-6689 (Print)), 22–33.

Shen, Y., Matsuno, Y., Fouse, S. D., Rao, N., Root, S., Xu, R., … Fan, G. (2008). X-inactivation in female human embryonic stem cells is in a nonrandom pattern and prone to epigenetic alterations. Proceedings of the National Academy of Sciences of the United States of America, 105(12), 4709–14. http://doi.org/10.1073/pnas.0712018105 Skilbeck, K. J., Johnston, G. A., & Hinton, T. (2010). Stress and GABA receptors. J.Neurochem., 112(1471-4159 (Electronic)), 1115–1130.

Slieker, R. C., Roost, M. S., Van Iperen, L., Suchiman, H. E. D., Tobi, E. W., Carlotti, F., … Chuva de Sousa Lopes, S. (2015). DNA Methylation Landscapes of Human Fetal Development. PLoS.Genet., 11(1553-7404 (Electronic)), e1005583.

Smil, V. (1999). China’s great famine: 40 years later. BMJ, 319(0959-8138 (Print)), 1619–1621.

Smith, A. K., Kilaru, V., Kocak, M., Almli, L. M., Mercer, K. B., Ressler, K. J., … Conneely, K. N. (2014). Methylation quantitative trait loci (meQTLs) are consistently detected across ancestry, developmental stage, and tissue type.

BMC Genomics, 15, 145. http://doi.org/10.1186/1471-2164-15-145 Smyth, G. K. (2004). Linear models and empirical bayes methods for assessing differential expression in microarray experiments. Stat.Appl.Genet.Mol.Biol., 3(1544-6115 (Electronic)), Article3.

Smyth, G. K. (2005). Limma: linear models for microarray data. In R. Gentleman, V. Carey, S. Dudoit, R. Irizarry, & W. Huber (Eds.), Bioinformatics and computational biology solutions using R and Bioconductor (pp. 397–420).

New York: Springer.

Spijker, A. T., Giltay, E. J., van Rossum, E. F., Manenschijn, L., DeRijk, R. H., Haffmans, J., … Hoencamp, E. (2011).

Glucocorticoid and mineralocorticoid receptor polymorphisms and clinical characteristics in bipolar disorder patients. Psychoneuroendocrinology, 36(1873-3360 (Electronic)), 1460–1469.

St. Clair, D., & He, L. (2005). Rates of Adult Schizophrenia Following of 1959-1961. Jama, 294(5), 557–562. http:// doi.org/10.1001/jama.294.5.557 St-Cyr, S., & McGowan, P. O. (2015). Programming of stress-related behavior and epigenetic neural gene regulation in mice offspring through maternal exposure to predator odor. Frontiers in Behavioral Neuroscience, 9(June), 145. http://doi.org/10.3389/fnbeh.2015.00145 Strahl, B. D., & Allis, C. D. (2000). The language of covalent histone modifications. Nature, 403(6765), 41–45. http://doi.org/10.1038/47412 Sun, Y. V, Turner, S. T., Smith, J. A., Hammond, P. I., Lazarus, A., Van De Rostyne, J. L., … Kardia, S. L. (2010).

Comparison of the DNA methylation profiles of human peripheral blood cells and transformed B-lymphocytes.

Hum.Genet., 127(1432-1203 (Electronic)), 651–658.

Suri, D., Bhattacharya, A., & Vaidya, V. A. (2014). Early stress evokes temporally distinct consequences on the hippocampal transcriptome, anxiety and cognitive behaviour. Int.J.Neuropsychopharmacol., 17(1469-5111 (Electronic)), 289–301.

Suzuki, M. M., & Bird, A. P. (2008). DNA methylation landscapes: provocative insights from epigenomics. Nat.Rev. Genet., 9(1471-0064 (Electronic)), 465–476.

192

References

Szyf, M. (2015). Prospects for the development of epigenetic drugs for CNS conditions. Nat.Rev.Drug Discov., 14(1474-1784 (Electronic)), 461–474.

Szyf, M., Weaver, I. C., Champagne, F. A., Diorio, J., & Meaney, M. J. (2005). Maternal programming of steroid receptor expression and phenotype through DNA methylation in the rat. Front Neuroendocrinol., 26(0091-3022 (Print)), 139–162.

Teh, A. L., Pan, H., Chen, L., Ong, M. L., Dogra, S., Wong, J., … Holbrook, J. D. (2014). The effect of genotype and in utero environment on interindividual variation in neonate DNA methylomes. Genome Research, 24(7), 1064– 1074. http://doi.org/10.1101/gr.171439.113 Teicher, M. H., & Samson, J. A. (2013). Childhood maltreatment and psychopathology: A case for ecophenotypic variants as clinically and neurobiologically distinct subtypes. Am.J.Psychiatry, 170(1535-7228 (Electronic)), 1114– 1133.

Teroganova, N., Girshkin, L., Suter, C. M., & Green, M. J. (2016). DNA methylation in peripheral tissue of schizophrenia and bipolar disorder: a systematic review. BMC Genetics, 17(1), 27. http://doi.org/10.1186/s12863-016-0332-2 Teschendorff, A. E., Marabita, F., Lechner, M., Bartlett, T., Tegner, J., Gomez-Cabrero, D., & Beck, S. (2013). A betamixture quantile normalization method for correcting probe design bias in Illumina Infinium 450 k DNA methylation data. Bioinformatics., 29(1367-4811 (Electronic)), 189–196.

Thakurela, S., Sahu, S. K., Garding, A., & Tiwari, V. K. (2015). Dynamics and function of distal regulatory elements during neurogenesis and neuroplasticity. Genome Res., 25(1549-5469 (Electronic)), 1309–1324.

Thombs, B. D., Bernstein, D. P., Lobbestael, J., & Arntz, A. (2009). A validation study of the Dutch Childhood Trauma Questionnaire-Short Form: factor structure, reliability, and known-groups validity. Child Abuse Negl., 33(1873-7757 (Electronic)), 518–523.

Tingley, D., Yamamoto, T., Hirose, K., Keele, L., & Imai, K. (2014). mediation: R Package for Causal Mediation Analysis.

Journal of Statistical Software, 59(5), 1–38. Retrieved from http://www.jstatsoft.org/v59/i05/ Tobi, E. W., Goeman, J. J., Monajemi, R., Gu, H., Putter, H., Zhang, Y., … Heijmans, B. T. (2014). DNA methylation signatures link prenatal famine exposure to growth and metabolism. Nat.Commun., 5(2041-1723 (Electronic)), 5592.

Tobi, E. W., Lumey, L. H., Talens, R. P., Kremer, D., Putter, H., Stein, A. D., … Heijmans, B. T. (2009). DNA methylation differences after exposure to prenatal famine are common and timing- and sex-specific. Human Molecular Genetics, 18(21), 4046–4053. http://doi.org/10.1093/hmg/ddp353 Tobi, E. W., Slieker, R. C., Stein, A. D., Suchiman, H. E. D., Slagboom, P. E., van Zwet, E. W., … Lumey, L. H. (2015).

Early gestation as the critical time-window for changes in the prenatal environment to affect the adult human blood methylome. Int.J.Epidemiol., 44(1464-3685 (Electronic)), 1211–1223.

Tohen, M., Jacobs, T. G., Grundy, S. L., McElroy, S. L., Banov, M. C., Janicak, P. G., … Breier, A. (2000). Efficacy of olanzapine in acute bipolar mania: a double-blind, placebo-controlled study. The Olanzipine HGGW Study Group.

Arch.Gen.Psychiatry, 57(0003-990X (Print)), 841–849.

Tremolizzo, L., Doueiri, M. S., Dong, E., Grayson, D. R., Davis, J. M., Pinna, G., … Guidotti, A. (2005). Valproate corrects the schizophrenia-like epigenetic behavioral modifications induced by methionine in mice. Biol.Psychiatry, 57(0006-3223 (Print)), 500–509.

193

9

CHAPTER 9

Tsankova, N., Renthal, W., Kumar, A., & Nestler, E. J. (2007). Epigenetic regulation in psychiatric disorders. Nat.Rev. Neurosci., 8(1471-003X (Print)), 355–367.

Tyrka, A. R., Price, L. H., Marsit, C. J., Walters, O. C., & Carpenter, L. L. (2012). Childhood adversity and epigenetic modulation of the leukocyte glucocorticoid receptor: preliminary findings in healthy adults. PLoS.One., 7(19326203 (Electronic)), e30148.

Ulrich-Lai, Y. M., & Herman, J. P. (2009). Neural regulation of endocrine and autonomic stress responses. Nat.Rev. Neurosci., 10(1471-0048 (Electronic)), 397–409.

Urnov, F. D., Rebar, E. J., Holmes, M. C., Zhang, H. S., & Gregory, P. D. (2010). Genome editing with engineered zinc finger nucleases. Nature Reviews. Genetics, 11(9), 636–646. http://doi.org/10.1038/nrg2842 van de Bank, B. L., Emir, U. E., Boer, V. O., van Asten, J. J., Maas, M. C., Wijnen, J. P., … Scheenen, T. W. (2015).

Multi-center reproducibility of neurochemical profiles in the human brain at 7 T. NMR Biomed., 28(1099-1492 (Electronic)), 306–316. van Dongen, J., Ehli, E. A., Slieker, R. C., Bartels, M., Weber, Z. M., Davies, G. E., … Boomsma, D. I. (2014). Epigenetic variation in monozygotic twins: A genome-wide analysis of DNA methylation in buccal cells. Genes, 5(2), 347–365. http://doi.org/10.3390/genes5020347 van Dongen, J., Nivard, M. G., Willemsen, G., Hottenga, J.-J., Helmer, Q., Dolan, C. V, … Boomsma, D. I. (2016).

Genetic and environmental influences interact with age and sex in shaping the human methylome. Nat Commun, 7. Retrieved from http://dx.doi.org/10.1038/ncomms11115 van Eijk, K. R., de Jong, S., Boks, M. P. M., Langeveld, T., Colas, F., Veldink, J. H., … Ophoff, R. A. (2012). Genetic analysis of DNA methylation and gene expression levels in whole blood of healthy human subjects. BMC.Genomics, 13(1471-2164 (Electronic)), 636. http://doi.org/10.1186/1471-2164-13-636 van Leeuwen, N., Bellingrath, S., de Kloet, E. R., Zitman, F. G., DeRijk, R. H., Kudielka, B. M., & Wust, S. (2011). Human mineralocorticoid receptor (MR) gene haplotypes modulate MR expression and transactivation: implication for the stress response. Psychoneuroendocrinology, 36(1873-3360 (Electronic)), 699–709.

Van Os, J., Rutten, B. P., Myin-Germeys, I., Delespaul, P., Viechtbauer, W., Van Zelst, C., … Mirjanic, T. (2014).

Identifying gene-environment interactions in schizophrenia: Contemporary challenges for integrated, large-scale investigations. Schizophrenia Bulletin, 40(4), 729–736. http://doi.org/10.1093/schbul/sbu069 van Stegeren, A., Rohleder, N., Everaerd, W., & Wolf, O. T. (2006). Salivary alpha amylase as marker for adrenergic activity during stress: effect of betablockade. Psychoneuroendocrinology, 31(0306-4530 (Print)), 137–141. van Venrooij, J. A., Fluitman, S. B., Lijmer, J. G., Kavelaars, A., Heijnen, C. J., Westenberg, H. G., … Gispen-de Wied, C. C. (2012). Impaired neuroendocrine and immune response to acute stress in medication-naive patients with a first episode of psychosis. Schizophr.Bull., 38(1745-1701 (Electronic)), 272–279. van Winkel, R., van Nierop, M., Myin-Germeys, I., & van Os, J. (2013). Childhood trauma as a cause of psychosis: linking genes, psychology, and biology. Can.J.Psychiatry, 58(0706-7437 (Print)), 44–51.

Varricchio, L., Tirelli, V., Masselli, E., Ghinassi, B., Saha, N., Besmer, P., & Migliaccio, A. R. (2012). The expression of the glucocorticoid receptor in human erythroblasts is uniquely regulated by KIT ligand: implications for stress erythropoiesis. Stem Cells Dev., 21(1557-8534 (Electronic)), 2852–2865.

Vasudev, K., Keown, P., Gibb, I., & McAllister-Williams, R. H. (2010). Hematological effects of valproate in psychiatric patients: what are the risk factors? J.Clin.Psychopharmacol., 30(1533-712X (Electronic)), 282–285.

194

References

Veldic, M., Kadriu, B., Maloku, E., Agis-Balboa, R. C., Guidotti, A., Davis, J. M., & Costa, E. (2007). Epigenetic mechanisms expressed in basal ganglia GABAergic neurons differentiate schizophrenia from bipolar disorder.

Schizophr.Res., 91(0920-9964 (Print)), 51–61.

Venero, C., & Borrell, J. (1999). Rapid glucocorticoid effects on excitatory amino acid levels in the hippocampus: A microdialysis study in freely moving rats. European Journal of Neuroscience, 11(7), 2465–2473. http://doi. org/10.1046/j.1460-9568.1999.00668.x Vermunt, M. W., Reinink, P., Korving, J., De Bruijn, E., Creyghton, P. M., Basak, O., … Creyghton, M. P. (2014). Largescale identification of coregulated enhancer networks in the adult human brain. Cell Rep., 9(2211-1247 (Electronic)), 767–779.

Vinkers, C. H., Joëls, M., Milaneschi, Y., Kahn, R. S., Penninx, B. W., & Boks, M. P. M. (2014). Stress exposure across the life span cumulatively increases depression risk and is moderated by neuroticism. Depress.Anxiety., 31(15206394 (Electronic)), 737–745.

Vinkers, C. H., Kalafateli, A. L., Rutten, B. P., Kas, M. J. H., Kaminsky, Z. A., Turner, J. D., & Boks, M. P. M. (2015).

Traumatic stress and human DNA methylation: a critical review. Epigenomics., 7(1750-192X (Electronic)), 593–608.

Vinkers, C. H., Zorn, J. V, Cornelisse, S., Koot, S., Houtepen, L. C., Olivier, B., … Joëls, M. (2013). Time-dependent changes in altruistic punishment following stress. Psychoneuroendocrinology, 38(1873-3360 (Electronic)), 1467– 1475.

Vojta, A., Dobrinić, P., Tadić, V., Bočkor, L., Korać, P., Julg, B., … Zoldoš, V. (2016). Repurposing the CRISPR-Cas9 system for targeted DNA methylation. Nucleic Acids Research . http://doi.org/10.1093/nar/gkw159 Waddington, C. H. (1942). The Epigenotype. Endeavour, 1, 18–20. http://doi.org/10.1093/ije/dyr184 Wadhwa, P. D., Buss, C., Entringer, S., & Swanson, J. M. (2009). Developmental origins of health and disease: Brief history of the approach and current focus on epigenetic mechanisms. Seminars in Reproductive Medicine. http:// doi.org/10.1055/s-0029-1237424 Wagner, J. R., Busche, S., Ge, B., Kwan, T., Pastinen, T., & Blanchette, M. (2014). The relationship between DNA methylation, genetic and expression inter-individual variation in untransformed human fibroblasts. Genome Biol., 15(1474-760X (Electronic)), R37.

Walker, E., Mittal, V., & Tessner, K. (2008). Stress and the hypothalamic pituitary adrenal axis in the developmental course of schizophrenia. Annu.Rev.Clin.Psychol., 4(1548-5943 (Print)), 189–216.

Walton, E., Hass, J., Liu, J., Roffman, J. L., Bernardoni, F., Roessner, V., … Ehrlich, S. (2015). Correspondence of DNA Methylation Between Blood and Brain Tissue and its Application to Schizophrenia Research. Schizophr.Bull., (1745-1701 (Electronic)).

Waterland, R. A., & Michels, K. B. (2007). Epigenetic epidemiology of the developmental origins hypothesis. Annual Review of Nutrition, 27, 363–88. http://doi.org/10.1146/annurev.nutr.27.061406.093705 Watkins, C. C., Sawa, A., & Pomper, M. G. (2014). Glia and immune cell signaling in bipolar disorder: insights from neuropharmacology and molecular imaging to clinical application. Transl.Psychiatry, 4(2158-3188 (Electronic)), e350.

Watson, S., Gallagher, P., Ritchie, J. C., Ferrier, I. N., & Young, A. H. (2004). Hypothalamic-pituitary-adrenal axis function in patients with bipolar disorder. Br.J.Psychiatry, 184(0007-1250 (Print)), 496–502.

195

9

CHAPTER 9

Weaver, I. C., Cervoni, N., Champagne, F. A., D’Alessio, A. C., Sharma, S., Seckl, J. R., … Meaney, M. J. (2004).

Epigenetic programming by maternal behavior. Nat.Neurosci., 7(1097-6256 (Print)), 847–854.

Wen, L., Li, X., Yan, L., Tan, Y., Li, R., Zhao, Y., … Qiao, J. (2014). Whole-genome analysis of 5-hydroxymethylcytosine and 5-methylcytosine at base resolution in the human brain. Genome Biology, 15(3), R49. http://doi.org/10.1186/ gb-2014-15-3-r49 Westenberg, P. M., Bokhorst, C. L., Miers, A. C., Sumter, S. R., Kallen, V. L., van Pelt, J., & Blote, A. W. (2009). A prepared speech in front of a pre-recorded audience: subjective, physiological, and neuroendocrine responses to the Leiden Public Speaking Task. Biol.Psychol., 82(1873-6246 (Electronic)), 116–124.

Wieck, A., Grassi-Oliveira, R., Do Prado, C. H., Rizzo, L. B., de Oliveira, A. S., Kommers-Molina, J., … Bauer, M. E. (2013). Differential neuroendocrine and immune responses to acute psychosocial stress in women with type 1 bipolar disorder. Brain Behav.Immun., 34(1090-2139 (Electronic)), 47–55.

Wijtenburg, S. A., Yang, S., Fischer, B. A., & Rowland, L. M. (2015). In vivo assessment of neurotransmitters and modulators with magnetic resonance spectroscopy: application to schizophrenia. Neurosci.Biobehav.Rev., 51(1873-7528 (Electronic)), 276–295.

Wolfe, J., Kimerling, R., & Brown, P. J. (1996). Psychometric review of the life stressor checklist-revised.

Instrumentation in Stress, Trauma, and Adaptation. Sidran Press.

Wong, C. C. Y., Meaburn, E. L., Ronald, A., Price, T. S., Jeffries, A. R., Schalkwyk, L. C., … Mill, J. (2014). Methylomic analysis of monozygotic twins discordant for autism spectrum disorder and related behavioural traits. Mol. Psychiatry, 19(1476-5578 (Electronic)), 495–503.

Woods, S. W. (2003). Chlorpromazine equivalent doses for the newer atypical antipsychotics. J.Clin.Psychiatry, 64(0160-6689 (Print)), 663–667.

Wu, H., & Sun, Y. E. (2006). Epigenetic regulation of stem cell differentiation. Pediatric Research, 59(4 Pt 2), 21R–5R. http://doi.org/10.1203/01.pdr.0000203565.76028.2a Yan, H., Tian, S., Slager, S. L., Sun, Z., & Ordog, T. (2015). Genome-Wide Epigenetic Studies in Human Disease: A Primer on -Omic Technologies. American Journal of Epidemiology. http://doi.org/10.1093/aje/kwv187 Yanagisawa, M., Planel, E., Ishiguro, K., & Fujita, S. C. (1999). Starvation induces tau hyperphosphorylation in mouse brain: implications for Alzheimer’s disease. FEBS Lett., 461(0014-5793 (Print)), 329–333.

Yasuda, S., Liang, M. H., Marinova, Z., Yahyavi, A., & Chuang, D. M. (2009). The mood stabilizers lithium and valproate selectively activate the promoter IV of brain-derived neurotrophic factor in neurons. Mol.Psychiatry, 14(1476-5578 (Electronic)), 51–59.

Yet, I., Tsai, P. C., Castillo-Fernandez, J. E., Carnero-Montoro, E., & Bell, J. T. (2016). Genetic and environmental impacts on DNA methylation levels in twins. Epigenomics., 8(1750-192X (Electronic)), 105–117.

Young, M. D., Wakefield, M. J., Smyth, G. K., & Oshlack, A. (2010). Gene ontology analysis for RNA-seq: accounting for selection bias. Genome Biol., 11(1474-760X (Electronic)), R14.

Young, R. C., Biggs, J. T., Ziegler, V. E., & Meyer, D. A. (1978). A rating scale for mania: reliability, validity and sensitivity. Br.J.Psychiatry, 133(0007-1250 (Print)), 429–435.

196

References

Yuan, P. X., Huang, L. D., Jiang, Y. M., Gutkind, J. S., Manji, H. K., & Chen, G. (2001). The mood stabilizer valproic acid activates mitogen-activated protein kinases and promotes neurite growth. J.Biol.Chem., 276(0021-9258 (Print)), 31674–31683.

Ziller, M. J., Gu, H., Müller, F., Donaghey, J., Tsai, L. T.-Y., Kohlbacher, O., … Meissner, A. (2013). Charting a dynamic DNA methylation landscape of the human genome. Nature, 500(7463), 477–81. http://doi.org/10.1038/nature12433 Zwanzger, P., Zavorotnyy, M., Gencheva, E., Diemer, J., Kugel, H., Heindel, W., … Pfleiderer, B. (2013). Acute shift in glutamate concentrations following experimentally induced panic with cholecystokinin tetrapeptide--a 3T-MRS study in healthy subjects. Neuropsychopharmacology, 38(1740-634X (Electronic)), 1648–1654.

9

197

Soloist background

O

verview of the soloists career.

CHAPTER 10 Curriculum vitae

CHAPTER 10

Lotte Houtepen was born in Maastricht, The Netherlands, on 10th March 1987. In 2005 she finished her secondary education at the Porta Mosana (Maastricht, The Netherlands) and started studying Pharmacy at the University of Utrecht (Utrecht, The Netherlands).

She obtained her bachelor degree in 2008 and started with the research master Drug Innovation at the University of Utrecht in 2009. During her master Lotte did her scientific internships at the department of Psychopharmacology (Utrecht University, The Netherlands) and the department of Physiology (University of Gothenburg, Sweden). In 2011 she started working as a research assistant at the University Medical Center Utrecht (UMC Utrecht, The Netherlands), under supervision of dr. M.P.M. Boks and dr. C.H. Vinkers on the epistress study. The epistress study examined the association of (epi)genetic factors on the acute stress response after psychosocial stress in schizophrenia and bipolar disorder. After one year, she continued working on this project as a PhD student under supervision of dr. C.H. Vinkers, dr M.P.M. Boks, prof. dr. M. JoĂŤls and prof. dr. R.S. Kahn.

On 1st June 2016 Lotte started working as a research associate at the MRC Integrative Epidemiology Unit (University of Bristol, England).

1987

2005

2009

2011

2016

Maastricht

Bachelor Pharmacy, University Utrecht

Master Drug innovation, University Utrecht

PhD University Medical Center Utrecht

Research associate University of Bristol

200

10

203

Orchestra

A

large instrumental ensemble.

CHAPTER 11 Dankwoord

CHAPTER 11

Vier jaar werken aan onderzoek doe je niet alleen, zowel op werk als daarbuiten heb ik veel hulp gehad. Hoewel het onmogelijk is iedereen te bedanken die (in)direct heeft bijgedragen, ga ik toch een aantal mensen extra in het zonnetje zetten.

Allereerst hartelijk dank aan alle proefpersonen die naar het UMCU zijn gekomen voor het (zoveelste) onderzoek. Vragenlijsten, bloed doneren en een stresstest zijn geen favoriete vrijetijdsbestedingen, maar de investering maakte dit proefschrift mogelijk.

Ook wil ik de leescommissie bestaande uit prof. Binder, prof. Burbach, prof. Meeus, prof.

Pasterkamp en dr. Creyghton bedanken voor de tijd en moeite die ze gestoken hebben in het beoordelen van dit proefschrift. Als ook dr. Rutten hartelijk dank voor het plaatsnemen als opponent.

Daarnaast is dit de laatste kans om op te scheppen over mijn fijne promotieteam. In alfabetische volgorde: Christiaan, Marco, Marian en René. In wisselende samenstelling heb ik iets gehad aan al jullie raden en adviezen.

Christiaan, in 2009 toen je zelf nog moest promoveren startte ik als masterstudent met een studie over stress in ratten. Op dat moment combineerde je al alles, maar nog intimiderender was het schijnbare gemak waarmee je zowel de grote lijn als inhoudelijke kennis op een rijtje had. Toendertijd verwachtte ik niet dat je me uiteindelijk als copromotor zou begeleiden bij humaan onderzoek. Dat geeft maar weer aan wat een ongelofelijk kameleonvermogen je hebt om voor jezelf en anderen kansen te creëren in nieuwe organisaties en op andere gebieden. Uiteraard ben ik alleen maar blij dat je dat vermogen hebt ingezet om samen met Marco mij aan te nemen als onderzoeksassistent in 2011, het startschot tot dit alles. Dank voor al het vertrouwen.

Marco, in het begin was het even wennen dat het duo een trio werd, maar al snel kon ik me niet meer anders voorstellen. Het gemak waarmee Christiaan en jij elkaar scherp hielden (of het nu over kapsels, muzieksmaak of onderzoeksmethodiek ging) heeft alleen maar bijgedragen aan onze wekelijkse overleggen. Die heb je vanaf het begin ingevoerd en zijn steeds waardevoller geworden. Uiteindelijk werd het nerd alert Christiaan iets te gortig en hadden we een extra halfuur om samen te Rrrrrrren. Mede dankzij je geduld, véle mailtjes en voorbeeldscripts, vind ik het ondertussen fantastisch om een nieuw raadsel op te lossen. Gelukkig trok je me weer op tijd uit de analyses, zodat er ook nog wat op papier is gekomen. Ook jij hebt je altijd ten volle ingezet en dat heb ik zeer gewaardeerd. Zelfs toen het tijd werd om te zoeken naar een nieuwe baan heb je me over de grens heen geholpen. Hartstikke bedankt! J Marian, in 2012 kwam ik officieel onder je hoede. Vanaf het begin was je kristalhelder over het belang van een goede planning en de nadruk op de grotere lijn, hoewel je juist ook de details tijdens onze maandelijkse meetings altijd scherp had. Ik kwam terecht in een groepsoverstijgende stressmeet, waar brainstormen centraal stond. Dank voor het 206

Acknowledgments

creëren van zo’n creatieve omgeving, maar nog meer voor de structuur die nodig was om jaren onderzoek om te zetten tot een logisch verhaal.

René, de overleggen waren altijd kort maar doeltreffend. Spijkers met koppen moesten er geslagen worden en die daadkrachtige aanwezigheid gaf soms net de doorslag die nodig was. Ook de kritische blik op de stukken heeft zowel het aantal spelfouten teruggebracht als het wetenschappelijk niveau verhoogd.

Alle studenten die meegeholpen hebben hartelijk dank! Lotte, Lieske en André jullie waren de allereersten. Daarna hebben Merle en Maaike met veel doorzettingsvermogen het stokje opgepakt en in rap tempo geïncludeerd, waarna Mathilde en Mirjam de laatste inclusies op het Epistress project verzorgden. Jasja en Caitlyn, jullie hebben veel bijgedragen aan de COLuMBuS studie en nooit teruggedeinsd voor de scantijden in weekenden/ avonden.

Tijdens het opzetten van de studies en bespreken van resultaten heb ik veel gehad aan diverse (oud) stressgroep leden. In het bijzonder Sandra, je zeer praktische uitleg is zo goed blijven hangen dat zure citroenen een vast onderdeel zijn van mijn eigen uitleg over speekselafname aan anderen. Jelle, we startten ongeveer tegelijkertijd met includeren, dus dan krijg je te maken met dezelfde chaos. Supergezellig dat je later nog terugkwam in UMCU, hoewel je nu helemaal op je plek bent in Frankrijk. Susanne, als staartje aan je promotietraject nam je het stokje van Jelle over voor de vrouweninclusie in CHOICE. Het was altijd heerlijk kletsen en chouffejes drinken met zijn 3en, oh ja we werkten ook nog tussendoor. Jolien, Lotte, Marloes, Angela, Femke, Rixt, Jiska, Henk en Manilla, hoewel we nooit samen aan een studie hebben gewerkt waren er vele sociale maar ook werktripjes waar ik altijd hartelijk welkom was. Dank daarvoor.

Voor sommige expertise waren we afhankelijk van andere afdelingen. Manja en Lot, heel fijn dat jullie de weg in het stratenum doolhof gebaand hebben van vriezeropslag tot hoe meet ik IL6 in speeksel. Yuije, you quickly impressed me with your meticulous excel skills but even more your kind personality. Thank you for all your help with the fibroblast cell work and the nice tea breaks. Ook super dat Ruben, Eric en Flip de genetica logistiek in de gaten hielden en Inge M. de speekselbepalingen in goede banen leidde. Dank aan de RADAR PIs prof. Branje, prof. Pol en prof. Meeus voor de soepele samenwerking. Vooral ook aan Marieke voor de praktische hulp. Voor de spectroscopie studie was er de bezielende leiding van Dennis in de spectromeet, Anouk voor inwerken op de 7T, Vincent die geduldig een matlab vraag of scanner error oploste en zeker aan het einde Jannie die met raad en daad beschikbaar was (ik zie ons nog GABA averages checken). Ook de hulp van René M. en Bas vanuit Psychiatrie bij het scannen op 7T en logistiek daaromtrent heb ik altijd zeer gewaardeerd. And of course I could always count on Katy and Mel, for (mental) support in case of a scan failure on one of those Mondays. Irene, het praktisch uitvoeren van een studie met alle bijbehorende 207

11

CHAPTER 11

regelgeving is niet altijd even makkelijk. Dank voor je geduld, begrip en flexibiliteit tijdens onze overleggen, als ook je nauwkeurige controle van al onze protocollen.

Mijn UMCU tijd was naast heel leerzaam, ook erg gezellig met (oud) psychiatrie collega’s. Lucija, dankzij jou voelde ik me meteen thuis op het kleinste kantoor ooit.

Je doorzettingsvermogen en niet zeuren maar doen houding is (gelukkig) nooit verdwenen. Met Annabel erbij was nog gezelliger, maar het is volgens mij onmogelijk Annabel niet acuut aardig te vinden. Ons kantoor was toch echt te klein, dus ben ik verhuisd naar B01.101. Inge, ik heb bewondering voor je organisatietalent als vrouw met kleine kinderen die nog een leuk mens bleek ook: Are you kidding me! Anne Lotte, onze Lotte logica en gemeenschappelijke gevoel voor humor waren meteen raak. Ook toen je het UMCU nest verliet om carrierétijger te worden, bleven we elkaar vaak zien.

Je nuchtere kijk op promoveren (‘niets voor mij’) zette me altijd met beide beentjes op de grond. Er zijn altijd belangrijkere dingen in het leven, zoals de bovenkant van de klimmuur halen. Charissa deed gelijk mee met onze gekkigheid, waardoor we nog altijd mini reünies organiseren. Eigenlijk was de hele “onderzoeksgang” te vinden voor voetbalpartijtjes, mega mannelijke movies, stappen, festivals, recepten uitwisselen, theeleuten, nespresso aanschaf... In het bijzonder Maya, Sanne K., Marieke, Wendy en Katy jullie maakten het altijd een feestje binnen en buiten werk. Toen er een echte “stresskamer” gecreëerd werd verhuisde ik door. Remmelt, al snel waren we samen door het MRS moeras aan het waden, dat ging niet altijd zonder stootje maar we zijn er doorheen gekomen. Ondertussen ben je guru GABA, chief MRS, grote leider NR3C1 en uiteraard een wizard met R, dat kan niet anders met je toewijding en inzet voor onderzoek. Gelukkig ben je vooral ook een heel gezellige toevoeging op elke vegetarische barbecue, ook al noemde je Lunetten ooit de banlieu van Utrecht. Judith, ontzettend veel respect voor je nuchtere aanpak waarmee je zoveel georganiseerd krijgt in zo weinig tijd. Ik weet hoe moeizaam een studie opstarten kan zijn, bij jou lijkt het haast makkelijk. In mijn laatste jaar ben ik weer op een echte meidenkamer terecht gekomen met Annabel, Lucija, Annet en Sanne V. Samen vormden jullie al een hecht team, maar ik heb me altijd thuis gevoeld. Ik was blij met de morele steun, heel veel thee en ook nog gezellig kletsen. Wonderbaarlijk dat er nog iets gedaan is in dat laatste jaar. Annet, die medicatiepaper is ons gedeelde project, wat tot stand kwam gedurende je zwangerschap. Toch heb ik je nooit anders gezien als nuchter en doeltreffend, dat was heerlijk samenwerken. Sanne V., heel lang hebben we met zijn tweëen wel en wee op kantoor gedeeld. Ik heb veel bewondering voor hoe je op een respectvolle manier voor jezelf blijft opkomen, ook al is dat tijdens een promotietraject niet altijd even makkelijk. Ik zal onze nespresso momenten missen.

Naast werk, hebben ook privé veel mensen bijgedragen aan dit traject. Jits, als oudhuisgenoot kennen we elkaar door en door. Ik weet dat je echt altijd voor me klaarstaat en dat is kostbaar. Als ik weer eens aan het stressen was riep je als eerste ‘Dat doe je altijd en het komt altijd goed’. Gelukkig sta je ook weer als paranimf naast me tijdens de verdediging en kan ik dus rekenen op een stevige por als ik mezelf gek aan 208

Acknowledgments

het maken ben. Martje, vanaf het eerste jaar farmacie hebben we alles gedeeld tijdens bachelor, master en PhD. Ongeacht de tijd ertussenin en locatie (Utrecht, Zeist, Maastricht, Den Bosch of Göteborg), het praat alsof we elkaar gisteren nog gezien hebben. Hoewel we nooit genoeg tijd hebben en chronisch te laat komen op afspraken die we daarna plannen. Met alle andere dudes hebben we zoveel barbeques, etentjes en verjaardagen gedaan, echt supermooi dat we zo’n mooie club over hebben gehouden aan onze stages bij Psychofarmacologie. Vooral met Daniëlle heb ik lief en leed gedeeld. Het schept toch een band als je blootgesteld wordt aan dr. Vinkers lingo. Uiteraard heb ook ik er ‘alle vertrouwen in’ dat jouw doctoraat helemaal goedkomt. Noortje, onze zwemsessies in de afgelopen twee jaar hebben me niet alleen fysiek fitter gehouden, maar ook de broodnodige cafeïneshots gegeven. Kan niet wachten op jouw 20 hoofdstuk tellende proefschrift.

Mijn familie, de mensen die zo dichtbij staan dat een simpel bedankje aan het einde van dit proefschrift nooit genoeg kan zijn. Maar ja, dat is wat ik te bieden heb. Mam, waar ik ook ben, ik weet dat ik altijd thuis terecht kan. Je nieuwsgierigheid, inzet en aanpassingsvermogen zijn voor mij altijd inspirerend geweest. Pap, gelukkig ben je nooit gestopt met het beantwoorden van mijn wa,wa,wa,waarom vragen. En de planmatigheid die ik duidelijk van jou heb geërfd was nog best nuttig voor de totstandkoming van dit proefschrift. Sanne, gelukkig sta je als grote zus naast me tijdens de verdediging. Hoewel we altijd roepen hoe verschillend we zijn, doen we vooral heel veel leuke dingen samen.

Met de komst van Abel is veel veranderd, maar de band tussen ons is even sterk. Ik vind het dan ook een eer om je zusje te zijn. David, via Sanne heb je me ook tijdens alle ups en downs meegekregen. Misschien paste je meteen zo goed in de familie door die Maastrichtse roots. Hoe dan ook ik ben blij dat ik altijd bij jullie terecht kon.

Last but most definitely not least. Douwe, alle muurtegeltjes kan ik opschrijven: je kent me door en door, je bent mijn beste vriend, je houdt me in balans, mijn leven is nog mooier met jou erbij en echt thuis ben ik pas als jij er ook bent. Voor iemand die altijd kletst heb ik verrassend weinig originele uitdrukkingen om je te bedanken voor de steun in de afgelopen jaren, zelfs toen ik besloot naar Bristol te vertrekken voor mijn nieuwe baan. Gelukkig begrijp je me ook perfect zonder woorden.

209

11