UNDERSTANDING SUICIDE BIOMARKER SKA2: DISCOVERY OF A SUICIDE

BIOSIGNATURE TO AUGMENT SKA2 SUICIDE PREDICTION AND

INVESTIGATING SKA2 REGULATORY MECHANISMS

by Makena L. Clive

A dissertation submitted to Johns Hopkins University in conformity with the requirements for the degree of Doctor of Philosophy.

Baltimore, Maryland May, 2017

© 2017 Makena L. Clive All Rights Reserved

Abstract…..

Suicide is the 2nd leading cause of death for ages 10-34 in the United States. Suicide rates have risen dramatically since 1999, with an estimated 25 attempts per single completed suicide, calling for increased efforts to improve prevention strategies. Suicide is a state that is associated with dysregulation of the hypothalamic-pituitary-adrenal (HPA) axis.

The spindle and kinetochore associated complex 2 (SKA2) gene was recently discovered as a biomarker of suicide and HPA-axis dysregulation. Genetic and epigenetic variation at rs7208505, a single nucleotide polymorphism located in the 3’ untranslated region of

SKA2, interacted with stress/anxiety metrics to predict suicidal behavior. Additionally,

SKA2 was downregulated in the brains of suicide decedents. Little is known concerning the regulation and function of SKA2 in regards to HPA-axis dysregulation or suicide. The goal of this dissertation was to improve suicide prevention by enhancing our existing

SKA2 suicide prediction model to better identify at-risk individuals, and to increase understanding of underlying suicide biology by investigating the regulation of SKA2.

We used a custom bioinformatics approach to derive a DNA methylation biosignature that both interacted with rs7208505 methylation in post mortem prefrontal cortex and predicted suicide attempt in peripheral blood samples. Replacing stress/anxiety metrics in the SKA2 suicide prediction model with this biosignature improved prediction of suicidal behaviors. Additionally, the biosignature showed associations with immune modulation and HPA-axis metrics, suggesting that it may represent a biological state that mediates suicidal behavior. To better understand SKA2 regulation in the context of suicide, we screened various pathway agonists for alterations of SKA2 expression. We observed

ii SKA2 downregulation in response to both anoxia and low iron conditions. Knockdown experiments revealed that this regulation was not due to the canonical hypoxia-inducible factor (HIF) pathway.

Together, these studies offer an enhanced suicide prediction model and new understanding of underlying biology and molecular mechanisms. This improved suicide prediction model will allow for and enhances ability to identify at-risk individuals and implement interventions to prevent future suicide attempts. With further development, knowledge of SKA2 and other molecular signaling pathways that are involved in HPA- axis dysregulation will allow for development of treatments targeting these pathways.

Advisors: Joseph P. Bressler, PhD & Zachary A. Kaminsky, PhD

Thesis Readers:

John D. Groopman, PhD

Peter P. Zandi, PhD

Richard S. Lee, PhD

Alternates:

Winnie Wan-Yee Tang, PhD

Sabra L. Klein, PhD

iii Acknowledgements

The journey through this PhD program has been anything but easy. From late nights studying and endless hours preparing for exams to failed experiments, feelings of inadequacy, and the seemingly constant weight of stress on my shoulders, there were moments that succeeding appeared impossible. Nevertheless, I persisted through these difficulties, borne up by the support of wonderful advisors, colleagues, friends, and family, to which I would be remiss if I did not acknowledge their contributions to my success.

First, I would like to offer an enormous thank you to Joseph Bressler. Time and time again, he has gone far above and beyond the call of advisor. I thank him for making himself constantly available to address my endless questions and concerns, introducing me to valuable mentors when experiments exceeded my abilities, and sacrificing his time and sanity to culture cells and conduct experiments in my absence. I’m truly indebted to him for the success of my project and for helping me navigate difficult moments during my time at Hopkins.

I would also like to thank Zach Kaminsky for being a wonderful thesis advisor and mentor. I owe many of my skills to his tutelage, including excellent lab technique, performing bioinformatics analysis in R, impeccable efficiency and time management, and the ability to persevere when your experiments seem to be conspiring against your success. And I thank him for providing an excellent example of a dedicated scientist who manages to maintain a work-life balance, which has inspired me to pursue a future career in the sciences.

In addition to my wonderful advisors, I’m grateful for the support of lab members past and present. Richard Lee has been incredibly supportive, acting both as a thesis committee member to guide my progress, and as a dear friend, providing stimulating conversation, life advice, delicious lunch outings, and acting as a therapist when frustrations hit their boiling point. Olivia Cox has

iv been a good friend, providing words of encouragement and endless hours of entertaining anecdotes. I would also like to thank Ilenna Jones, Tori Brown, and Jerry Guintivano for their assistance with my projects, and to the Mood Disorders Group for their support.

I thank John Groopman for participating in my thesis committee and offering words of wisdom on life beyond graduate school. And thank you to my additional thesis readers Peter Zandi,

Winnie Tang, and Sabra Klein for donating their time and support.

I’m grateful to the EHS Department, students past and present, staff, and faculty for their support, and for providing me with funding throughout my program on the NIEHS T32 Training Grant.

Thank you to individuals at KKI who have provided me with help and mentorship: Hernando

Lopez-Bertoni, Shuli Xia, and Larry Frelin.

I’m extremely grateful for my classmates and dear friends, Meena Aladdin and Suzanne Martos.

We laughed together, we cried together (well, mostly Meena), and somehow we all survived the past five years of torture and are only moderately jaded as a result. Their friendship and support mean everything to me, and I wish them both marvelous success in their future careers. I’m also very grateful for the friendship and support of Ben Davis, Cissy Li, Katie Kuhns, Jonathan

Coulter, and David and Annaliesa Peterson.

Last, but certainly not least, I thank my family for their unwavering support. Thank you to my wonderful parents, Jeff and Sherrie Hinrichs, for their constant backing of my ambitions, and for my siblings Jessica, Bridger, and Matthew for always being there. Thank you to my superb husband, Geoff, for taking me on many adventures, encouraging me to pursue my dreams, and tolerating my PhD-induced insanity for the past five years. And thank you to my catchild,

Copernicus, for finding me when I needed him the most.

v Table of Contents

Abstract…...... ii

Acknowledgements ...... iv

Table of Contents ...... vi

List of Tables ...... ix

List of Figures ...... x

List of Abbreviations ...... xiii

1. INTRODUCTION & BACKGROUND ...... 1

1.1 Suicide Overview ...... 1

1.2 Suicide Biomarkers: HPA-axis Dysregulation ...... 4

1.3 SKA2 as a Psychiatric Biomarker ...... 7

1.4 SKA2 Gene Function...... 12

2. DISCOVERY AND REPLICATION OF A PERIPHERAL TISSUE DNA METHYLATION

BIOSIGNATURE TO AUGMENT A SUICIDE PREDICTION MODEL ...... 16

2.1 Abstract ...... 17

2.2 Background ...... 19

2.3 Results ...... 22

2.4 Discussion ...... 28

2.5 Conclusions ...... 35

2.6 Materials & Methods ...... 36

vi 2.7 Tables and Figures ...... 40

2.8 Supplementary Material ...... 50

3. SUICIDE BIOMARKER GENE SKA2 IS REGULATED BY HIF-INDEPENDENT

HYPOXIA PATHWAY ...... 58

3.1 Abstract ...... 58

3.2 Background ...... 60

3.3 Results ...... 62

3.4 Discussion ...... 66

3.5 Conclusions ...... 71

3.6 Materials & Methods ...... 72

3.7 Tables and Figures ...... 76

3.8 Supplementary Figures ...... 88

4. CONCLUSIONS, PERSPECTIVES, AND FUTURE STUDIES ...... 93

4.1 Summary & Conclusions ...... 93

4.2 Significance & Innovation ...... 94

4.3 Future Studies ...... 95

4.4 Perspectives on Suicide Biomarker Research & Implications of Findings .... 99

5. APPENDIX A: CHROMATIN CONFORMATION CAPTURE IN SKA2 ...... 104

5.1 Introduction ...... 104

5.2 Results ...... 105

vii 5.3 Discussion & Conclusions ...... 107

5.4 Materials & Methods ...... 108

5.5 Tables and Figures ...... 112

References ...... 122

Curriculum Vitae ...... 147

viii List of Tables

2. DISCOVERY AND REPLICATION OF A PERIPHERAL TISSUE DNA METHYLATION

BIOSIGNATURE TO AUGMENT A SUICIDE PREDICTION MODEL………………………..16

Table 1. Gene ontology results ………………………………………………….40

Table 2. Overrepresentation analysis ……………………………………………41

Table 3. Probes from interaction biosignature contributing to the PCA model….42

Table 4. Prediction model results………………………………………………...45

Table S1. Sample overview……………………………………………………...51

Table S2. Pyrosequencing assay primer sequences……………………………...52

Table S3. Probes with significant correlations to the interaction biosignature

across all data sets………………………………………………………………..53

3. SUICIDE BIOMARKER GENE SKA2 IS REGULATED BY HIF-INDEPENDENT

HYPOXIA PATHWAY……………………………………………………………………..58

Table 1. shRNA Sequences ……………………………………………………...76

Table 2. qRT-PCR Primers ……………………………………………………...77

5. APPENDIX A: CHROMATIN CONFORMATION CAPTURE IN SKA2…………….104

Table 1. PCR Primer Sequences ……………………………………………….112

ix List of Figures

2. DISCOVERY AND REPLICATION OF A PERIPHERAL TISSUE DNA METHYLATION

BIOSIGNATURE TO AUGMENT A SUICIDE PREDICTION MODEL………………………..16

Table 1. Gene ontology results ………………………………………………….40

Figure 1. Discovery of interaction biosignature probes and prediction of suicidal

ideation using interaction biosignature in multiple cohorts……………………...46

Figure 2. Myeloid-derived cell proportions correlated with the interaction

biosignature in all cohorts, and were predictive of suicidal behavior……………48

Figure 3. Interaction biosignature and methylation at SHP1 (cg24437859)

correlated with inflammatory markers and stress measures in the PPD cohort

(trimester > 1)……………………………………………………………………49

Figure S1. Summary of algorithm identifying Interaction Proxy………………..54

Figure S2. HPA axis model depiction……………………………………………55

Figure S3. ROC curves of prediction using myeloid-derived cell proportion in

place of interaction biosignature in prediction model……………………………56

Figure S4. DDR1 (cg08469255) array methylation validation by sodium bisulfite

pyrosequencing…………………………………………………………………..57

3. SUICIDE BIOMARKER GENE SKA2 IS REGULATED BY HIF-INDEPENDENT

HYPOXIA PATHWAY……………………………………………………………………..58

Figure 1. Forskolin stimulation of CREB does not alter SKA2 expression in SH-

SY5Y or HEK293………………………………………………………………..78

Figure 2. PPAR-α agonist fenofibrate did not alter SKA2 expression in SH-SY5Y

or HEK293……………………………………………………………………….80

x Figure 3. Anoxia treatment in SH-SY5Y, but not HEK293, downregulated SKA2

expression………………………………………………………………………..81

Figure 4. SKA2 expression was downregulated by DFO in a dose-responsive

manner……………………………………………………………………………82

Figure 5. SKA2 expression was downregulated by DFO in a time-dependent

manner……………………………………………………………………………83

Figure 6. SKA2 expression and hypoxia pathways are controlled by glucose

concentration……………………………………………………………………..84

Figure 7. SKA2 downregulation by DFO is not mediated by HIF-1α…………...85

Figure 8. SKA2 downregulation by DFO is not mediated by HIF-1β…………..86

Figure 9. SKA2 does not mediate hypoxia signaling pathways…………………87

Figure S1. HEK293 anoxia time course and screening of THP-1 and D425 for

SKA2 response to anoxia treatment……………………………………………..88

Figure S2. HIF1A knockdown confirmation…………………………………….90

Figure S3. HIF1B knockdown confirmation…………………………………….91

Figure S4. SKA2 knockdown confirmation…………………………………...... 92

5. APPENDIX A: CHROMATIN CONFORMATION CAPTURE IN SKA2…………….104

Figure 1. Overview of Hypothesis……………………………………………...113

Figure 2. Overview of 3C Technique…………………………………………..114

Figure 3. Detection of SKA2 transcription start site fragment in 4C product

enriched for rs7208505 fragment……………………………………………….115

Figure 4. Detection of promoter-3’UTR (rs7208505) junction in 3C product....117

Figure 5. Detection of SKA2 promoter-3’ interaction in HEK293 and THP-1...118

xi Figure 6. HEK293 and THP-1, but not SH-SY5Y responded to dexamethasone treatment………………………………………………………………………..119

Figure 7. Dexamethasone treatment did not alter SKA2 promoter or rs7208505 methylation……………………………………………………………………..120

xii List of Abbreviations

3C, chromatin conformation capture

AMPK, 5'-adenosine monophosphate-activated protein kinase

APOA2, apolipoprotein A-II

ARHGEF10, rho guanine nucleotide exchange factor 10

AUC, area under the curve

BDNF, brain-derived neurotrophic factor

CNS, central nervous system

CpG, cytosine-guanine dinucleotide

CPT1A, carnitine palmitoyltransferase IA

CREB1 (CREB), cAMP responsive element binding protein 1

CSF, cerebrospinal fluid

CTQ, Childhood Trauma Questionnaire

DDR1, discoidin domain receptor tyrosine kinase 1

DFO, desferrioxamine mesylate

DST, dexamethasone suppression test

ERK, extracelluar signal-regulated kinase

FKBP5, FK506 binding protein 5

GenRED, Genetics of Recurrent Early-Onset Depression

GEO, NCBI Gene Expression Omnibus

GLUT1, glucose transporter 1

GR, glucocorticoid receptor (NR3C1)

GTP, Grady Trauma Project

xiii HCA, hierarchical clustering analysis

HCC, hair cortisol concentrations

HIF1A (HIF-1α), hypoxia inducible factor 1 alpha

HM450, Infinium HumanMethylation450 BeadChip array hnRNP A1, heterogeneous nuclear ribonucleoprotein A1

HPA-axis, hypothalamic-pituitary-adrenal axis

IL-6, interleukin 6

KT, kinetochore

LDA, linear discriminant analysis

MAPK, mitogen-activated protein kinase

MDD, major depressive disorder

MEOX2, homeobox protein MOX-2 miR, microRNA

PCA, principal component analysis

PHD, prolyl hydroxylase

PPARA (PPAR-α), peroxisome proliferator-activated receptor alpha

PTSD, post-traumatic stress disorder

RT, room temperature

SA, suicidal attempt

SCARED, Self-Report for Childhood Anxiety Related Disorders

SGK1, serum/glucocorticoid regulated kinase 1

SHP1, protein tyrosine phosphatase, non-receptor type 6

SI, suicidal ideation

xiv SKA1, spindle and kinetochore associated complex subunit 1

SKA2, spindle and kinetochore associated complex subunit 2

SKA3, spindle and kinetochore associated complex subunit 3

SNP, single nucleotide polymorphism

TSS, transcription start site

UTR, untranslated region

VEGFA, vascular endothelial growth factor

VHL, von Hippel-Lindau tumor suppressor

xv 1. INTRODUCTION & BACKGROUND

1.1 Suicide Overview

1.1.1 Suicide as a Public Health Crisis

In the United States, suicide is reported as the 10th leading cause of death for all ages, the 2nd leading cause of death for ages 10-34, and the 4th leading cause of death for ages 35-54, accounting for 42,773 total deaths in 2014,1 posing a significant financial burden for the country each year.2 Between 1999 and 2015, the suicide rate has increased for all ages 10-74, with the highest increase in ages 50-59 (13.1 to 20.9 per 100,000), and an average increase of 3.9 per 100,000 across ages 15-69 between these years.1 Between

2008 and 2009 and estimated 8.3 million adults in the United States reported having suicidal thoughts, while 1 million reported attempting suicide.3 Additionally, males are

3.5 times more likely to commit suicide than females, resulting in a large gender disparity between suicide victims. There are an estimated 25 attempts per single completed suicide,2 indicating a larger population that would benefit from improved early interventions. This dramatic rise in the suicide rate in recent years and the even higher rate of unsuccessful suicide attempts call for increased research efforts to improve understanding of the underlying biology and to develop prevention strategies.4

1.1.2 Suicide Neurobiology Overview

Suicide is a complex physiologic and psychologic state that has a variety of risk factors, including: mental illness,5, 6 especially mood disorders,7 psychosocial factors,8 genetic factors,9, 10 aggressive-impulsive behavior,11 stress,12 and history of childhood abuse.13 Suicide is characterized by structural and functional changes in the brain, which

1 include alterations to the serotonergic system, neuroendocrine system, neuroimmune system, and second messenger systems such as brain-derived neurotrophic factor (BDNF) and tropomycin receptor kinase B (TrkB).14 Importantly, suicide is associated with dysregulation of the hypothalamic-pituitary-adrenal (HPA) axis, a system that orchestrates the body’s response to stress.

1.1.3 The HPA-Axis

The HPA-axis plays an important role in the body’s stress response, and functions as follows: the hypothalamus secretes corticotropin-releasing hormone (CRH), which stimulates the secretion of adrenocorticotrophic hormone (ACTH) from the pituitary gland, which causes the adrenal cortex to produce cortisol, the body’s stress hormone.

Cortisol is responsible for important processes in the body, which include recovering homeostasis after stress exposure, regulating blood glucose levels and muscular glycogen metabolism, maintaining pH, and suppressing the immune system. Cortisol represses

CRH and ACTH production in the hypothalamus and pituitary gland to negatively regulate the stress response. On a molecular level, cortisol binds to the mineralocorticoid

(MR; NR3C2 gene) and glucocorticoid receptors (GR; NR3C1 gene) to alter cellular function and induce the expression of stress-responsive genes.14

Although not yet completely understood, it is widely accepted that prenatal maternal depression/anxiety can result in an altered HPA-axis response in offspring both after birth and into adulthood, which has been demonstrated in both humans and rodent models. In a rat model established by the Meaney group, increased maternal care during a period of ten days after birth was associated with a decreased stress response to stimuli.

2 This time period corresponds to the human third trimester. Furthermore, adult offspring were characterized by increased GR expression in the hippocampus and an altered epigenetic regulation of NR3C1 in the hippocampus. These effects were reversed by cross fostering experiments, suggesting a direct effect of maternal care, not genetics.15-17 An association has been identified in humans between antenatal maternal depression/anxiety and cognitive, behavioral, and emotional alterations in children,18 altered neonatal measures of urinary cortisol, norepinephrine, dopamine, and serotonin,19 increased salivary cortisol measures in response to stress at 3 months of age,20 and increased risk of developing an anxiety disorder and comorbid depression at the age of 18.21 Additionally, altered methylation at NR3C1 exon 1F in infant cord blood was associated with metrics of maternal depression during the second trimester.20 These studies provide evidence that prenatal maternal mood disorders and stress could modify an individual’s HPA-axis before birth, resulting in dysregulation later in life.

Many studies have reported dysregulation of the HPA-axis in suicide, describing an association between suicide risk and the lack of cortisol suppression in the dexamethasone suppression test (DST),22-25 and an increased cortisol response following stimulation with 5HTP, a serotonin precursor.26 HPA-axis dysregulation has also been described in post-traumatic stress disorder (PTSD) and major depressive disorder

(MDD),27, 28 both of which have high incidence of suicide.29, 30 Despite the current knowledge of the wide variety of risk factors and neurobiological abnormalities associated with suicide, the rates of suicide and suicide attempts are still on the rise, suggesting the need for improved understanding of the underlying biology of suicide to

3 develop effective biomarkers that will help identify at-risk individuals, allowing for timely intervention and prevention.

1.2 Suicide Biomarkers: HPA-axis Dysregulation

1.2.1 Introduction to Biomarkers

A biomarker can be defined as “any substance, structure, or process that can be measured in the body or its products and influence or predict the incidence of outcome or disease.”31 In the context of suicide, ideal biomarkers are predictive of suicidal behavior, defined as suicidal ideation (SI) and suicidal attempt (SA), or are markers of suicide completion. In the case of SI and SA prediction, biomarkers are derived from the peripheral tissues, since the organ of interest, the brain, is not accessible in living individuals. Peripheral tissues used for suicide biomarkers include blood (plasma, white blood cells, and platelets), cerebrospinal fluid (CSF), saliva, and urine, which can be obtained by non-invasive or minimally-invasive means. Current suicide biomarkers include measurable alterations of the serotonergic, neuroendocrine, and neuroimmune systems, which have been extensively reviewed by Oquendo et al.32 Here, we highlight several biomarkers of suicide as related to dysregulation of the HPA-axis, specifically the glucocorticoid receptor, NR3C1, and its chaperone, FK506 binding protein 5 (FKBP5).

1.2.2 Suicide Biomarker: NR3C1

As stated previously, the GR is encoded by the NR3C1 gene and plays an important role in the HPA-axis stress response by binding to cortisol and altering the expression of downstream genes. NR3C1 is expressed in most tissues in the body,33 and

4 has 15 distinct mRNA isoforms, including major isoforms GR-α, GR-β, and GR-γ,34 which have similar function, but are differentially expressed, as described below. NR3C1 mRNA and associated GR proteins have been established as important biomarkers of

HPA-axis dysregulation, exhibiting altered expression in association with completed suicide or suicidal behaviors. Park et al. discovered a polymorphism in the NR3C1 Bcl1 region that was associated with suicide following cancer diagnosis,35 unveiling a potential genetic component of suicide for further investigation. McGowan et al. reported decreased hippocampal expression of total NR3C1 (all isoforms) in suicide victims with a history of abuse, and observed a similar reduction in the NR3C1 exon 1F splice variant.36

Corroborating these results, Labonte et al. observed decreased expression of GR exon 1B,

1C, and 1H variants in the hippocampus of suicide completers with a history of abuse, which was associated with altered methylation in each of these exon 1 variants.37 In addition to the hippocampus, several groups reported altered GR expression in other brain regions. Pérez-Ortiz et al. found decreased NR3C1 mRNA and protein expression in the amygdala of male suicide completers.38 Similarly, Pandey et al. reported decreased

GR-α expression in the amygdala and prefrontal cortex of teenage suicide victims.39

Together, these findings establish the downregulation of NR3C1 mRNA and proteins as a biomarker of suicide in various brain regions, which is consistent with HPA-axis dysregulation. In terms of predictive efficacy, Melhem et al. discovered decreased

NR3C1-α mRNA in blood samples from patients with SA,40 laying the foundation for future development of an NR3C1 biomarker for predicting suicidal behavior in live individuals.

5 1.2.3 Suicide Biomarker: FKBP5

Also important in glucocorticoid signaling is FKBP5, a protein that interacts with heat-shock protein 90, a molecular chaperone of GR, which reduces GR binding affinity to glucocorticoids. FKBP5 is regulated by the GR, creating a negative feedback loop for glucocorticoid signaling in the cell. Polymorphisms across the gene alter its expression and downstream GR regulation, resulting in a modified glucocorticoid response.41 Studies have found associations between specific FKBP5 polymorphisms and suicide or suicidal behaviors. Supriyanto et al. discovered an FKBP5 haplotype (set of polymorphisms) that was associated with completed suicide in a Japanese population.42 Similarly, Yin et al. identified a FKBP5 haplotype that was associated with suicidal behavior in live and deceased cohorts.9 The FKBP5 polymorphisms best characterized in suicide are rs3800373 and rs1360780, which exist in linkage disequilibrium, meaning specific alleles are often inherited together. Fudalej et al. reported that the rs3800373 risk allele (C) was associated with completed suicide,43 while Brent et al. reported that the same allele was associated with suicidal events in depressed adolescents,44 providing replication and further evidence of the importance of this allele in suicide risk. With respect to rs1360780,

Perroud et al. reported an association between this polymorphism and increased suicidality with use of antidepressants.45 Correspondingly, Yeo et al. derived lymphoblastoid cell lines from suicidal patients or controls and found that rs1360780 risk allele was associated with increased expression of FKBP5,46 which would result in decreased GR activation in the presence of glucocorticoids, consistent with previously reported findings.41 Additional studies have reported associations between other FKBP5 polymorphisms and suicidal behaviors.47, 48 One study observed decreased FKBP5

6 mRNA and protein expression in the amygdala of male suicide completers.38 Collectively, these studies present promising targets for further investigation of FKBP5 genetic and expression variation as candidates for development of a predictive suicide biomarker.

1.3 SKA2 as a Psychiatric Biomarker

1.3.1 Discovery of SKA2 Biomarker

Furthering our discussion of suicide biomarkers, the gene SKA2 was originally discovered as a biomarker for suicide by Guintivano et al. in their 2014 landmark study.49

In this work, authors obtained Infinium HumanMethylation450 data from post mortem prefrontal cortex from suicide victims and matched controls, and with this data performed a genome-wide screen, ultimately discovering cg13989295, a cytosine-guanine dinucleotide (CpG) located in the 3’ untranslated region (UTR) of the SKA2 gene, which had significant associations with suicide in prefrontal cortex neurons. This CpG is part of a single nucleotide polymorphism (SNP), rs7208505, which has two alleles: C, which creates a CpG with the neighboring G residue, and T, which abrogates the CpG. The C can be methylated by DNA methyltransferases, which generally acts as an additional layer of gene regulation. Both rs7208505 genetic and epigenetic variation were associated with suicide in a linear model, which was replicated in two independent cohorts, identifying SKA2 rs7208505 as a biomarker of suicide. The authors also observed lower expression of SKA2 mRNA in the brains of suicide decedents, along with a negative association of rs7208505 methylation with SKA2 expression, suggesting that not only does rs7208505 serve as a biomarker for suicide, but that the SKA2 gene itself may somehow be involved in suicide biology. In addition to the brain, the authors examined

7 rs7208505 in blood, and described an association between variation in rs7208505 methylation and both waking cortisol measures and with perceived stress and anxiety scores. This observation allowed the authors to create a prediction model in which rs7208505 genetic and epigenetic variation in the blood interacted with metrics of stress

(waking cortisol, perceived stress/anxiety) to predict suicidal ideation or attempt, which predicted suicidal behaviors with an area under the curve (AUC) of >70% in all cohorts examined. This model suggests that rs7208505 genetic and epigenetic variation may represent an individual “trait”, which interacts with a current “state”, such as stress and anxiety, to alter one’s behavior. Overall, this study established rs7208505 as a biomarker that could be used in combination with stress and anxiety metrics to predict suicidal behavior, and identified SKA2 as a gene potentially involved in suicide biology.

1.3.2 Replication Studies in Suicide

Because several studies have replicated the findings described above, SKA2 is now an accepted biomarker of suicide and related HPA-axis disorders. More studies are needed because of the complexity of the risk factors that drive suicide. Kaminsky et al. observed poor prediction of suicidal behavior in a large cohort with high rates of trauma exposure by using the Guintivano et al. SKA2 prediction model, which the authors attributed to the high rates of trauma in this cohort.50 By altering the prediction model to include a childhood trauma variable, which was associated with SA and anxiety, suicidal behavior prediction was improved to an AUC of > 70%.50 The authors also reported that rs7208505 methylation and a trauma metric interacted to predict cortisol suppression following the DST,50 furthering the SKA2 trait-state argument set forth by Guintivano et

8 al.49 Sadeh et al. reported an independent replication of the SKA2 prediction model, describing an association of rs7208505 methylation with current, but not past, suicidal thoughts and behaviors, and with internalizing, but not externalizing, disorders. They also reported enhanced prediction of suicide phenotypes with rs7208505 methylation compared to other risk factors.51 Yin et al. discovered three additional SNPs in the SKA2 gene that showed genotype-dependent but non-significant associations with suicidal attempt. They did not examine rs7208505 for association with suicidal behaviors in this sample.9

Beyond the replication of SKA2 rs7208505 as a suicide biomarker, several additional studies replicated the observation of altered SKA2 expression in suicide.

Niculescu et al. tested the expression of blood-based suicide biomarkers at low and high

SI states in males with various psychiatric disorders (bipolar disorder, MDD, schizoaffective disorder and schizophrenia). They reported SKA2 among the top differentially-expressed genes in all disorders.52 SKA2 expression was found to be specific to SI, and showed good diagnostic prediction for SI across all psychiatric groups.52 Pandey et al. sought to further characterize the changes in SKA2 expression in post mortem brains of suicide victims, and to determine the specificity of SKA2 to suicide compared to other psychiatric disorders.53 The authors reported that decreased SKA2 mRNA and protein expression was specific to suicide independent of psychiatric diagnosis,53 bolstering the observations of Guintivano et al. and further suggesting a role for SKA2 in suicide biology. Contrary to these studies, Melhelm et al. recently reported increased expression of SKA2 mRNA in the blood of patients admitted following a suicidal attempt.40 These contrasting SKA2 expression results signify a need for further

9 validation of SKA2 expression changes in both blood and brain in a variety of cohorts to characterize its role in suicide biology and HPA-axis dysregulation, and to determine how SKA2 expression is altered in relation to SI, SA, and traumatic events.

1.3.3 Replication Studies in PTSD

Several studies have also investigated the usefulness of SKA2 as a biomarker for

PTSD, which similar to suicide, involves dysregulation of the HPA-axis. Kaminsky et al. reported predictability of PTSD using the SKA2 prediction model modified to incorporate childhood trauma scores. They observed an association between PTSD with SKA2 promoter methylation but not rs7208505 methylation.50 Sadeh et al. observed an association between rs7208505 genotype-adjusted methylation and reduced prefrontal cortex cortical thickness, which is thought to result from neuron damage caused by trauma-exposure induced hypercortisolism.54 This study also reported a negative correlation between PTSD symptom severity and both prefrontal cortical thickness and rs7208505 genotype-adjusted methylation, hypothesizing that SKA2 represents a potential biomarker of stress exposure and susceptibility, and is not specific to suicide.54 Boks et al. tested SKA2 prediction of PTSD in pre- and post-deployment blood samples of a military cohort and reported an average 5% decrease in rs7208505 methylation with development of PTSD post-deployment, with a SKA2 prediction model AUC of 0.66 for PTSD.55 The authors also observed opposite associations of SKA2 methylation changes post- deployment with traumatic exposure during deployment, with increased methylation in non-PTSD individuals and decreased methylation in PTSD individuals.55 The authors attributed this observation to a differential adaptation of the HPA-axis response to

10 traumatic experience. Collectively, these studies suggest that SKA2 may be a biomarker of HPA-axis dysregulation, which is observed in both suicide and PTSD, but that its relationship with PTSD needs to be further investigated to elucidate the nature of the relationship.

1.3.4 Importance of Studying SKA2 Biomarker

In conclusion, SKA2 has been established as a promising biomarker of suicide, with potential in prediction of PTSD. Importantly, both SKA2 expression and rs7208505 methylation show associations in suicide and PTSD, suggesting that SKA2 is a prediction biomarker, but may also be involved in the biology of these disorders. Further studies are needed to further characterize the rs7208505 biomarker. A longitudinal study would allow us to understand whether there is a temporal relationship between HPA-axis dysregulation and altered methylation at rs7208505, and whether this methylation is altered with treatment. The field would benefit from additional, large replication studies in both suicide and PTSD cohorts to further distinguish the role of SKA2 as a biomarker in these disorders and to improve prediction capabilities. Doing this would improve methods of identifying at-risk individuals, allowing for implementation of early intervention. Likewise, investigating the molecular mechanism underlying the involvement of SKA2 in suicide biology and HPA-axis dysregulation would be helpful in developing therapeutic targets and prevention approaches.

11 1.4 SKA2 Gene Function

1.4.1 SKA2 in Cell Cycle

SKA2 was discovered in 2006 by Hanisch et al. in a proteomics screen to identify novel proteins involved in the mitotic spindle complex.56 This screen identified the spindle and kinetochore associated complex subunit 1 (SKA1), which was found to localize to the kinetochores during mitosis.56 The SKA2 protein was subsequently identified as a SKA1 binding partner in a yeast two-hybrid screen, and was renamed from its predicted gene name, FAM33A.56 SKA2 was discovered to bind SKA1 both in vitro and in vivo, and was shown to localize to the kinetochore (KT) complex with SKA1 during mitosis.56 Gaitanos et al. later discovered a third member of the SKA complex, spindle and kinetochore associated complex subunit 3 (SKA3), which was identified as a binding partner of SKA2, and immunoprecipitated with SKA1 and SKA2.57 The SKA proteins intertwine to form the SKA complex,58 which behaves as a 700-kDa protein that localizes to the kinetochores during mitosis. It functions to establish the metaphase plate, upon which chromosomes are aligned in preparation for cell division, and to silence the spindle checkpoint, allowing the cells to progress through metaphase to anaphase.56, 57

Knockdown of any one member of the SKA complex resulted in delayed progression through mitosis, with cells stalling in metaphase due to the lack of silencing of the spindle checkpoint,56, 57, 59 indicating that these proteins are required for timely cell cycle progression, but are not required for cell survival, as these knockdowns did not alter the localization of any other KT protein.56, 57 Knockdown of any two complex members resulted in cell death,57 suggesting that, although a single member can be lost, the SKA complex must exist in some form for cell survival. Additionally, knockdown of SKA1

12 resulted in decreased expression of SKA2.56 Knockdown of SKA2 did not result in decreased expression of SKA1, however, the loss of SKA2 resulted in the absence of

SKA1 at the spindle and kinetochores.56 Knockdown of SKA3 affected both SKA1 and

SKA2 expression. Overall, SKA complex function depends on the expression of each family member.

SKA2 is expressed by many different cell types, including HEK293, A549, various breast cancer cell lines, and HeLa cells, but surprisingly it was not detected in normal lung tissue, HepG2, mouse liver, Rat-1, or skin fibroblast cells.59 SKA2 is primarily located in the cytoplasm in normal tissues, but is found predominantly in the nucleus in cancerous tissues.59 Two splice variants of SKA2 have been reported (HeLa

S3 cells) in mitotic cells, but only one in cells not undergoing mitosis.56 Splicing is due to a binding site on SKA2 mRNA for heterogeneous nuclear ribonucleoprotein A1 (hnRNP

A1). Binding modifies the pattern of splicing at exon 3 under certain conditions, leading to a higher ratio of isoform 1 to isoform 2.60 SKA2 has also been reported to be expressed throughout meiosis in mice.61

1.4.2 SKA2 Regulation

Despite the important role of SKA2 in the cell cycle, only a few studies have sought to understand the regulation of SKA2 expression. Micro RNAs miR-141 and miR-

301a have been shown to regulate SKA2. miR-141 (chr12:7073260-7073354, hg19), a tumor-suppressive miR, was shown to target the SKA2 3’UTR by Bian et al. and regulated SKA2 mRNA and protein expression in gliomas.62 Shi et al. reported increased expression of miR-301a and SKA2 in 84% of tested breast cancer tumors.63 Cao et al.

13 proposed a mechanism in which miR-301 regulates SKA2 expression through a signaling pathway involving CREB1. miR-301a (chr17:57228497-57228582, hg19), encoded by

SKA2 intron 1, inhibits transcription of MEOX2, which relieves MEOX2 inhibition of the

ERK1/2 (MAPK3/1) kinase cascade. The subsequent MAPK3/1 activation results in the binding of CREB1 to the SKA2 promoter and increased SKA2 transcription.64 Zhuang et al. also reported CREB1-dependent SKA2 expression. They reported that CREB1 binds to the SKA2 promoter and that knocking down CREB1 results in decreased expression of

SKA2 .65 Presumably CREB1 could be activated by other pathways, which would increase SKA2 expression. CREB1 is the substrate of a large number protein kinases.33

Studies have shown the regulation of SKA2 through peroxisome proliferator-activated receptor alpha (PPAR-α) and hypoxia-inducible factor 1 alpha (HIF-1α) pathways.66

Gonsalves et al. reported increased levels of SKA2 mRNA with placental growth factor

(PlGF) treatment, which was eliminated by knocking down PPARA or HIF1A.66

Moreover, SKA2 was upregulated by treatment with PPAR-α agonist fenofibrate,66 suggesting a regulatory role for PPAR-α in SKA2 expression.

1.4.3 SKA2 as a GR Chaperone

In addition to its involvement in the cell cycle, SKA2 is also a binding partner of the glucocorticoid receptor (GR; NR3C1 gene),59 which is also responsive to glucocorticoids. SKA2 was bound by the GR in a yeast two-hybrid assay, which was confirmed by further experiments showing co-localization of SKA2 with GR in the nucleus of HEK293 cells transfected by yellow fluorescent protein-tagged GR stimulated by 100 nM dexamethasone, a GR agonist.59 The authors also showed increased

14 dexamethasone-induced glucocorticoid transactivation with SKA2 overexpression, and decreased transactivation with SKA2 knockdown,59 suggesting an important role for

SKA2 in regulating the GR and related stress responsive pathways. Interestingly, a dose- dependent decrease was observed in SKA2 expression in A549 cells treated with dexamethasone. Although many more studies are needed, SKA2 appears to offer a negative-feedback mechanism regulating GR signaling. A decrease in SKA2 expression in cells treated with glucocorticoid would attenuate further glucocorticoid signaling by preventing GR transactivation. A better understanding of the involvement of SKA2 in glucocorticoid signaling could improve our understanding of suicide biology.

1.4.4 Importance of Studying SKA2 Regulation

In summary, the involvement of SKA2 in mitosis is well-characterized, but its involvement in other functions is largely unknown. Similarly, SKA2 regulation is poorly understood. SKA2 has been shown to be regulated by microRNAs miR-141 and miR-

301a,62, 64 the latter of which is associated with the MAPK pathway and CREB1.65 SKA2 is also regulated by HIF-1α, PPAR-α, and GR.59, 66 More studies are needed. Delving deeper into the regulation of SKA2 and its function outside of cell division will expand the understanding of SKA2 as a suicide biomarker and perhaps increase knowledge of suicide biology, leading to better methods of identifying at-risk individuals and the development of improved prevention approaches.

15 2. DISCOVERY AND REPLICATION OF A PERIPHERAL TISSUE DNA

METHYLATION BIOSIGNATURE TO AUGMENT A SUICIDE PREDICTION MODEL

Reproduced from:

Clive ML, Boks MP, Vinkers CH, Osborne LM, Payne JL, Ressler KJ et al. Discovery and replication of a peripheral tissue DNA methylation biosignature to augment a suicide prediction model. Clinical Epigenetics 2016; 8(1): 113.

The article presented has been reproduced here in agreement with the BioMed Central license agreement.

16 2.1 Abstract

Background. Suicide is the 2nd leading cause of death among adolescents in the United

States and rates are rising. Methods to identify individuals at risk are essential for implementing prevention strategies and the development of a biomarker can potentially improve prediction of suicidal behaviors. Prediction of our previously reported SKA2 biomarker for suicide and PTSD is substantially improved by questionnaires assessing perceived stress or anxiety, and is therefore reliant on psychological assessment.

However, such stress related states may also leave a biosignature that could equally improve suicide prediction. In genome-wide DNA methylation data we observed significant overlap between waking cortisol-associated and suicide-associated DNA methylation in blood and brain, respectively.

Results. Using a custom bioinformatic brain to blood discovery algorithm, we derived a

DNA methylation biosignature that interacts with SKA2 methylation to improve the prediction of suicidal ideation in our existing suicide prediction model across both blood and saliva data sets. This biosignature was independently validated in the Grady Trauma

Project cohort, and interacted with HPA-axis metrics in the same cohort. The biosignature showed a relationship with immune status by its correlation with myeloid- derived cell proportions in all data sets, and with IL-6 measures in a prospective post- partum depression cohort. Three probes showed significant correlations with the biosignature: cg08469255 (DDR1), cg22029879 (ARHGEF10), and cg24437859 (SHP1), of which SHP1 methylation correlated with immune measures.

17 Conclusions. We conclude that this biosignature interacts with SKA2 methylation to improve suicide prediction and may represent a biological state of immune and HPA-axis modulation that mediates suicidal behavior.

18 2.2 Background

Suicide accounts for 1.4% of worldwide deaths annually, posing a serious public health issue.67 Based on 2014 data, it is the 2nd leading cause of death among adolescents, and the 10th leading cause of death for all ages in the United States.1 Given the rising rates of suicide in the United States, methods to identify individuals at risk for implementing prevention strategies are urgently needed.4

Recently, our laboratory identified a DNA methylation mark that is associated with suicide in a post mortem brain tissue cohort at a CpG (cg13989295) located within a single nucleotide polymorphism (SNP), rs7208505, in the spindle and kinetochore associated protein 2 (SKA2) where the reference allele of rs7208505 eliminates the CpG.

The observed epigenetic association with suicide was replicated in two additional brain tissue cohorts and with suicidal behaviors including suicidal ideation (SI) and attempt

(SA) in peripheral blood in three living cohorts.49 In our original work, gene expression of SKA2 was correlated with DNA methylation at this position and was significantly decreased in suicide decedents. Several recent independent studies have observed decreased expression of SKA2 in both the blood of violent suicide completers 68 and in the prefrontal cortex of suicide victims,49, 69 the latter of which was also associated with decreased protein levels.

The SKA2 protein is thought to interact with the hypothalamic-pituitary-adrenal

(HPA) axis by chaperoning the glucocorticoid receptor (GR) from the cytoplasm to the nucleus upon cortisol binding.70 Once in the nucleus, the GR can interact with genomic

DNA and influence gene expression involved in negative feedback regulation of the HPA axis response. In two independent cohorts with high levels of childhood trauma, elevated

19 SKA2 DNA methylation in peripheral blood before administration of the TRIER social stress test was significantly associated with a blunted post-test cortisol level, while SKA2

DNA methylation before the dexamethasone suppression test (DST) was significantly associated with elevated post-test cortisol levels.50, 71 These data support the interpretation that SKA2 DNA methylation state may be an important contributor to individual stress response.

In an attempt to identify at-risk individuals, we previously generated a suicide prediction model, which describes suicidal behavior as a function of both genotype and methylation at the single nucleotide polymorphism (SNP) rs7208505 in SKA2 which interacts with a state level metric of stress or trait level metric of anxiety to confer risk.72

Notably, some studies demonstrate that state level stress can be influenced by trait level anxiety.73 Model predictive accuracies vary between ~70-85% in various cohorts and are consistent with SKA2 gene expression based prediction accuracies reported by other groups.52, 71 The statistical interaction with stress is likely related to the physiological role

SKA2 plays in mediating HPA axis activity. In this context, it is hypothesized that epigenetic variation of SKA2 may represent an underlying trait vulnerability of the HPA axis that must interact with a state of stress to elicit risk. In our previous work, we have identified significant interactions of SKA2 with various self-reported psychological scales to influence suicide risk. The scales vary by study cohort and include the Child Trauma

Questionnaire (CTQ), the Perceived Stress Scale, waking salivary cortisol levels, and various metrics of anxiety including self-reported binary estimates and those quantified by the Self-Report for Childhood Anxiety Related Disorders (SCARED).49, 50, 71

Furthermore, our work and others have noted an increased model efficacy in subgroups

20 of individuals having experienced childhood trauma.50, 51, 54, 71 It is possible that high values in the stress metrics represent a biological state that may be related to HPA axis function. Despite the possibility to assess these states using questionnaires, the use of self-reported scales has many drawbacks including a lack of standardization across studies, variability in psychometric properties, and variability in the subjective rating of stress levels. In the clinical context, the administration of questionnaires requires time and patient compliance.

Recent attempts have been made to circumvent the use of psychological assessments and develop biomarker proxies.52, 74 A challenge for the identification of peripheral tissue based epigenetic biomarkers in the context of psychiatry is the generalizability of the identified peripheral epigenetic variation in the brain. We have hypothesized previously that circulating steroid hormones such as cortisol may mark peripheral tissue DNA on the epigenetic level while affecting behavior through central nervous system (CNS) specific actions.75-77 In support of this hypothesis, the initial objective of this study was to evaluate if cortisol associated DNA methylation levels in peripheral tissues, blood and saliva, are enriched among suicide associated DNA methylation levels in the brain.

While systemic factors like cortisol may influence epigenetic patterns across tissues and may represent relevant biomarkers interacting with SKA2, we did not wish to limit our analysis to cortisol associated probes alone. Thus, the second major objective of this study was to generate an epigenetic biosignature of SKA2 interacting state markers in a bioinformatically driven and unbiased manner and to understand the underlying biological context driving any identified biosignatures.

21 2.3 Results

Overrepresentation of peripheral cortisol associated loci among brain associated suicide genes

To address our first objective, we attempted to address the degree to which peripheral blood or saliva based DNA methylation profiles are indicative of epigenetic profiles in the brain related to suicidal behaviors. One potential substrate for peripheral tissue-brain overlap is a cross tissue reprogramming by the systemic influence of hormones. For

Genetics of Recurrent Early-Onset Depression (GenRED) Offspring, we identified

20,146 and 22,865 probes that were nominally associated with the area under the curve

(AUC) of waking weekday cortisol in blood and saliva samples, respectively (Additional

Files 1-2). To increase statistical power, we performed a combined analysis of blood and saliva samples using a linear model with age, sex, tissue type (blood or saliva), and cell type proportion as covariates and identified 22,425 loci associated with the AUC of waking weekday cortisol (Additional File 3). A pathway enrichment analysis of the genes significantly associated (P < 0.001) with the AUC of weekday waking cortisol in

GenRED Offspring blood and saliva using the tool g:Profiler revealed an enrichment of neural development pathways at varying levels of significance (Table 1).78, 79 Given the importance of dysregulated cortisol biology to suicidal behaviors, cortisol-associated methylation probes in peripheral blood (N = 18) and saliva (N = 20) from the GenRED

Offspring cohorts were assessed for an overrepresentation with those probes significantly associated with completed suicide separately in post mortem prefrontal cortical neurons and non-neurons (N = 45). Cortisol-associated probes within genes or gene regulatory sites were significantly overrepresented among prefrontal neuron suicide-associated

22 genes and genes previously identified as associated with cortisol stress reactivity (Table

2).80 These findings indicate that there may be common pathways between cortisol biology and suicidal behavior and that the epigenetic marks of suicide associated hormonal changes may be detectable in peripheral tissues.

Algorithmic identification of SKA2-interacting biosignature for DNA methylation- based suicidal behavior prediction

In light of the above findings, our strategy to approach our second objective of generating a biosignature of SKA2 interacting state markers was to identify epigenetic variation interacting with SKA2 that was consistent across brain and peripheral tissues.

The full algorithm is explained in Figure S1. Briefly, DNA methylation in prefrontal cortex neurons at each probe was assessed for statistical interaction with rs7208505 CpG methylation (chr17:59110368, hg38) in a linear model controlling for age, sex, and rs7208505 genotype, and identified 669 probes below a p-value cutoff of 0.005 (Figure

1A, Result S1). Of these 669 probes, 72 exhibited an AUC prediction for SA in the top

25th percentile (AUC > 0.825) in the GenRED Offspring blood cohort (Figure 1B;

Table 3). The methylation at these 72 probes was used to train a principal component analysis (PCA) model on the GenRED Offspring blood data, which was then used to predict PCA models in the other data sets. The first eigenvector was used to assess suicidal behavior prediction in the original prediction model, replacing the stress measure with the PCA first eigenvector. All steps were evaluated by the Monte Carlo method and found to be statistically significant (P < 0.001, Result S1). This approach predicted SI in

PPD cohort blood with an AUC of 0.88 (95% CI: 0.75-1; P = 0.041), and in GenRED

23 Offspring saliva with an AUC of 0.81 (95% CI: 0.59-1; P = 0.011) (Figure 1C; Table 4).

These high prediction AUCs provide evidence that the PCA first eigenvector may represent a methylation SKA2 “interaction biosignature” that is predictive of suicidal behavior in the existing suicide prediction model and replaces the need for a stress questionnaire.

Independent validation of SKA2 model interaction biosignature performance

The interaction biosignature model was validated using methylation array data from the

Grady Trauma Project (GTP), a sample of urban minorities with low socioeconomic status and high rates of traumatic experience and PTSD. On the entire sample, the prediction model generated an AUC of 0.50 (95% CI: 0.42-0.58; P = 0.724) for prediction of SI in all 376 individuals. Based on recent publications that have provided evidence that both PTSD and substance abuse may confound SKA2 methylation,50, 81 we selected a subset of the GTP sample with no history of PTSD or drug use (N = 115; 6 cases, 109 controls), where a combination of SKA2 and the interaction biosignature predicted SI with an AUC of 0.73 (95% CI: 0.59-0.87; P = 0.050) (Figure 1D; Table

4).54 Although our interaction biosignature model was unsuccessful in suicidal behavior prediction across the entire GTP cohort, prediction was successful in a subset without

PTSD. This altered suicidal behavior prediction with PTSD is consistent with previously published findings.50

24 Association of interaction biosignature metrics with HPA axis function

To improve our understanding of the biological relevance of the interaction biosignature, we assessed biosignature loci for a relationship with various metrics of HPA axis function in two cohorts with high levels of childhood trauma as assessed by the CTQ.

The interaction biosignature eigenvector interacted with CTQ scores to associate with post-test AUC cortisol levels following the administration of the TRIER social stress test

(Biosignature = 3446.9 + 1631.2, P= 0.038, CTQ = -40.6 + 12.9, P= 0.002, Interaction

= -92.8 + 45.0, P= 0.043, F= 4.5, df= 4/81, Model P= 0.038) (Figure S2A). In the GTP sample subset, the interaction biosignature eigenvector interacted with CTQ scores to associate with the natural log of the day 2 cortisol following administration of the DST

(Biosignature = -6.4 + 2.8, P= 0.027, CTQ = 0.096 + 0.037, P= 0.012, Interaction =

0.22 + 0.095, P= 0.027, F= 2.4, df= 4/49, Model P= 0.027) (Figure S2B). Taken together, the data suggest that SKA2 interaction biosignature values associate with early life trauma status to influence HPA axis sensitivity.

Assessment of biological relevance of SKA2 model interaction biosignature

We reasoned that the biological underpinnings of our SKA2 interaction biosignature may be related to variation in peripheral immune cells, as inflammation may be influenced by state factors like stress. The predicted proportion of granulocyte and monocyte content

(myeloid-derived cells) showed a negative association with the interaction biosignature across all data sets (Figure 2 A-D), with significant correlations in GenRED Offspring blood (rho = -0.76, P = 2.7×10-4), PPD cohort blood (rho = -0.29, P = 0.034) and GTP blood (rho = -0.57, P = 2.4×10-7), and a non-significant association in GenRED Offspring

25 saliva (rho = -0.39, P = 0.092). Substituting the proportion of myeloid-derived cells for the interaction biosignature in the prediction model yielded comparable predictions of SI

(Figure S3; Table 4) in GenRED Offspring saliva (AUC = 0.79; 95% CI: 0.56-1; P =

0.28), PPD cohort blood, (AUC = 0.83; 95% CI: 0.61-1; P = 0.003), and GTP subset

(AUC = 0.73; 95% CI: 0.55-0.9; P = 0.99). The PPD cohort (trimester > 1) showed a non-significant correlation between peripheral blood interleukin-6 (IL-6) levels and the predicted myeloid-derived cell proportion (Figure 3A; rho = -0.32, P = 0.054), however, there was no correlation between IL-6 and the interaction biosignature (rho = 0.05, P =

0.80). Increased granulocyte and monocyte counts along with altered IL-6 levels may indicate increased inflammation and implies that our interaction biosignature could be an indicator of an immune state involved in suicide etiology.82 The PPD cohort also showed a significant correlation between perceived stress and the interaction biosignature (Figure

3B; rho = 0.33, P = 0.019), suggesting that the interaction biosignature may represent a biological state of both stress and inflammation.

Identification of probes driving the SKA2 model interaction biosignature

Each of the 72 probes comprising the interaction biosignature were tested for correlation to the first eigenvector of the PCA model across each dataset to identify subsets of probes that may be driving a majority of the variation. Three probes exhibited significant correlations (P < 0.05) consistent across all cohorts (Table S3): cg08469255 (DDR1), cg22029879 (ARHGEF10), and cg24437859 (SHP1). Microarray-derived DNA methylation values were validated by pyrosequencing in select loci (Figure S4).

Methylation at cg24437859 used in place of the interaction biosignature predicted SI in

26 GenRED Offspring saliva with an AUC of 0.77 (95% CI: 0.54-1; P = 0.20), and in PPD cohort blood with an AUC of 0.84 (95% CI: 0.63-1; P = 0.001). Probe cg24437859 is located within the promoter of SHP1, which has known immune system functions, providing a plausible biological explanation for the statistical interaction with SKA2 identified in our data. This relationship was investigated further in the PPD cohort, where plasma cytokine levels were available. CpG methylation at cg24437859 correlated with IL-6 levels (rho = -0.37, P = 0.035; Figure 3C) and the predicted myeloid-derived cell proportions (rho = 0.74, P = 1.4×10-8; Figure 3D) in PPD cohort blood collected during the 2nd or 3rd trimester.

27 2.4 Discussion

In this study, we used brain, saliva, and whole blood DNA methylation data of several cohorts to derive a biosignature of a stress state that may aid in the prediction of suicide.

Using a statistically oriented approach that analyzed cross-tissue epigenetic reprogramming by cortisol and interaction with the previous reported SKA2 suicide biomarker, we generated an epigenetic biosignature. To assess the effects of cortisol on

DNA methylation patterns, we performed a pathway enrichment analysis of genes with methylation significantly associated with AUC weekday cortisol in the GenRED

Offspring samples, which revealed an enrichment of neural developmental pathways.

This data is consistent with the notion that there are a number of genes that regulate both cortisol and neural development. Early life stress, for example, is known to affect both brain development and HPA-axis function later in life. Several recent studies in mice reported impaired neurogenesis and cognition with early life stress,83 as well as altered

CpG methylation in NR3C1, the gene encoding the glucocorticoid receptor.84 Similar adverse effects have been observed in humans with exposure to early life stress.85, 86 This link was further bolstered by an overrepresentation analysis that showed an enrichment of

AUC weekday cortisol associated genes in GenRED Offspring blood and saliva with suicide associated genes in prefrontal neurons as well as previously identified genes associated with cortisol stress reactivity in blood,80 indicating that there are consistent cross-tissue DNA methylation changes with cortisol dysregulation and a behavioral outcome such as suicide. Our results are consistent with a model whereby suicide associated HPA-axis dysregulation causes an overproduction of circulating cortisol, which causes DNA methylation changes in various tissues, resulting in behavioral

28 changes through the actions of DNA methylation in the brain, while leaving measurable marks in the periphery that enable the biomarker based prediction of suicidal ideation and behaviors.

We present a biosignature that is representative of probes with both a significant interaction with SKA2 genotype and methylation in prefrontal neurons and is predictive of suicidal ideation in three cohorts. Using this biosignature in interaction with SKA2 can replace the stress questionnaire metrics used as interactive covariates in our original suicide prediction model. We used Monte Carlo based testing for significance by generating a similar PCA eigenvector from randomly selected sets of 72 probes.

Randomly selected probes yielded predictions inferior to that of the biosignature in almost all bootstraps, suggesting that improved model prediction accuracy is not due to the underlying data structure. The biosignature performance did not reach significance by this method in saliva, possibly due to the confounding influence of buccal derived cell types influencing the variation generated at biosignature loci. This interaction biosignature showed correlation with percent granulocyte and monocyte (both myeloid lineage derived cells) content in all peripheral tissue data sets, suggesting a possible immune modulation associated with the methylation at these 72 probes. Both the interaction biosignature and myeloid-derived cell levels also correlated with serum interleukin-6 (IL-6) in the PPD cohort, further suggesting that a pro-inflammatory immune modulation is interacting with SKA2 methylation to mediate suicidal behavior.

Consistent with our findings, increased pro-inflammatory cytokines, including IL-6, have been observed in the CSF of suicide attempters,87 and in the prefrontal cortex of teenage suicide victims.88, 89 Additionally, PTSD is known to show increased levels of C-reactive

29 protein and IL-6, which are both signs of increased inflammation.90-95 Biological changes, such as inflammation and immune system modulation, are known to be associated with suicide and related mood disorders. For example, lymphocytes are known to play a role in HPA-axis dysfunction and have been used to assess many different psychiatric disorders, including depression and suicide.14, 96 Suicidal behavior is known to be associated with an inflammatory state which, if measured, may improve the prediction of such behavior. Substituting the percentage of myeloid derived cells for the interaction biosignature in the prediction model was successful in predicting suicidal ideation in all of the cohorts, suggesting that this interaction biosignature may be indicative of a biological state that interacts with the trait of SKA2 genotype and methylation to influence behavior.

In further efforts to reduce these 72 probes to a smaller number that would help us better understand the biology and facilitate practical implementation we assessed the genes displaying epigenetic variation most closely mimicking the PCA first eigenvector.

We discovered that there were three probes within the 72-probe interaction biosignature with significant correlations to the interaction biosignature in all cohorts: cg22029879, cg08469255, and cg24437859. ARHGEF10 (rho guanine nucleotide exchange factor 10; cg22029879) was identified as one of 21 genes located on chromosome 8p, a region that is thought to contribute to neuropsychiatric disorders, including depression.97 Although little evidence exists tying ARHGEF10 to suicide etiology, this locus may be worthy of further investigation due to its consistent association with the interaction biosignature across all data sets. DDR1 (discoidin domain receptor tyrosine kinase 1; cg08469255) is primarily involved in cell adhesion and extracellular matrix remodeling, but also has

30 known roles in immune and inflammatory pathways. DDR1 was shown in a cell culture model to induce the expression of cyclooxygenase (COX2), which is involved in the synthesis of prostaglandins and has a known role in inflammation.98 COX2 also activates the NF-κB pathway, which is involved in inflammatory pathways and cytokine production,98 and has been shown as a downstream target of DDR1 to cause infiltrating macrophages to produce chemokines. Additionally, DDR1 was also shown in a mouse model of kidney obstruction to mediate the development of inflammation and fibrosis following kidney injury.99 Given this evidence, DDR1 methylation could account for the correlation of myeloid-derived cell content with the interaction biosignature, and could also represent a target for further investigation.

SHP1 (protein tyrosine phosphatase, non-receptor type 6; cg24437859) has been implicated in modulating neutrophil recruitment to inflamed tissues through modulation of the phosphoinositol pathway,100, 101 and has been shown to play an inhibitory role in cytokine induced activation of the HPA axis through the JAK-STAT pathway.102

Furthermore, SHP1 methylation correlated with IL-6 in the PPD cohort as well as the myeloid-derived cell proportion in all cohorts, altogether demonstrating biological evidence for the statistical interaction with HPA axis relevant SKA2 identified in our data.

Critically, IL-6 contributes to hematopoietic stem cell fate decisions and helps to differentiate myeloid from non-myeloid cells.101, 103 As such, the possibility remains that epigenetic variation in genes like SHP1 may be important not only for differentiation of hemetopeitic stem cells into meyloid cells but for regulation of pro-inflammatory cytokines and may moderate the influence of pro-inflammatory cytokines on HPA axis activity. This supposition is supported by data demonstrating that SKA2 interaction

31 biosignature data interacted with CTQ scores to predict HPA axis responsivity in two stress challenges from multiple datasets. The relationship between the interaction biosignature signals and HPA axis sensitivity is very similar to previously reported findings related to the influence of SKA2 DNA methylation on HPA axis activity from the same cohorts.50, 71

In our independent validation in the GTP cohort (N=376), we observed that our interaction biosignature model was only able to predict suicidal ideation in a smaller subset of drug-naïve individuals without a PTSD diagnosis (N = 115), but was unsuccessful predicting in the full GTP sample. Several recent papers on the usefulness of SKA2 in predicting PTSD have observed a confounding effect of marijuana use on

SKA2 methylation, which potentially inhibits the accurate prediction of suicidal behaviors.50, 71 Kaminsky et al. observed a significant interaction between substance abuse and SKA2 methylation in predicting suicidal behavior in several cohorts, and Boks et al. observed altered SKA2 methylation with smoking, alcohol consumption, and use of medications.50, 71 Based on these findings, it is reasonable to assume that without taking into account substance abuse, using SKA2 methylation to predict suicidal behaviors could produce inaccurate results. Along with substance abuse, trauma exposure has recently been shown to influence SKA2 methylation.50, 54, 71 Boks et al. showed that development of PTSD symptoms was associated with longitudinal decreases in SKA2 methylation after military deployment, which is the opposite of the positive association between SKA2 methylation and suicide risk.49, 71 Furthermore, Sadeh et al. observed a positive association between genotype adjusted SKA2 methylation and PTSD symptom severity in one military cohort, while this PTSD association was not replicated in a second military

32 cohort,51 adding to the complexity that is the interaction between SKA2 methylation,

PTSD, and suicidal behavior.54 One biological explanation for this interaction is altered

HPA-axis function in PTSD, as shown by increased clearance of dexamethasone in the

DST.104 Kaminsky et al. recently observed decreased day two cortisol in the DST in subjects diagnosed with PTSD but not suicidal behavior, indicating increased HPA-axis sensitivity.50 In another study, van Zuiden et al. observed a longitudinal increase in sensitivity to dexamethasone in T-cells collected from a Dutch military cohort pre- and post-deployment.105 Another biological explanation for the SKA2 methylation-PTSD relationship is that PTSD is associated with increased inflammation, which has been observed in many studies showing increased levels of C-reactive protein and IL-6 in the blood of both military and non-military cohorts.90-95 The relationship between SKA2 methylation and PTSD should be studied further to better understand the impact on suicidal behaviors.

This study has many limitations. Sodium bisulfite modification cannot distinguish between 5-methyl cytosine (5-MC) and 5-hydroxy methyl cytosine (5-HMC) levels. Like

5-MC, 5-HMC can vary in the brain in response to stress 106, 107 and has been identified in various psychiatric disorders.108, 109 Brain tissue analyses have the potential to be confounded by post mortem factors such as the method and timing of tissue preservation.

Psychiatric diseases can often be co-morbid with other illnesses such as cancer and heart disease, among others.110, 111 Despite the implication that inflammatory factors may be interacting with SKA2, we did not assess for the health status of the study subjects and any potential impact this might have on our results. This study is limited by using small samples that are not representative of the general population and are biased towards

33 controls due to a low ratio of cases to controls, and only validated findings in a single independent sample in which suicidal behavior is only predicted in small subsets. Ideally, these findings would be further validated in a large sample that is more representative of the general population to prove its usefulness in prediction of suicidal behavior.

34 2.5 Conclusions

We present a biosignature that predicts suicide consistently across multiple, highly variable data sets, specifically, youth at high risk for depression, pregnant women at high risk for PPD, and middle-aged urban population with high incidence of trauma and PTSD.

This biosignature is cross-tissue in that it predicts suicidal behavior in both blood and saliva samples, and is based on probes that are associated with suicide in prefrontal neurons. To our knowledge, this is the first prediction model to date that works in both blood and saliva, and the first suicide prediction model to use only DNA methylation to predict suicidal behavior. Additionally, correlations of the interaction biosignature with myeloid proportion and stress metrics may indicate a fuller integration of suicide etiology into the existing SKA2 suicide prediction model. Finally, this biosignature allows us to predict suicidal behavior without using a stress questionnaire or assessment. Although the biosignature produces lower prediction AUCs than the stress questionnaires, it represents a single measure that allows us to predict suicidal behavior across all data sets, providing consistency and better allowing for comparison across populations. Ultimately this work will add to the development of early diagnostics tests that may aid in the early identification and prevention of suicide.

35 2.6 Materials & Methods

Human Samples. Peripheral blood and saliva samples were obtained from separate individuals in the GenRED Offspring cohort from Johns Hopkins 49, 112-115. Post mortem prefrontal cortex neurons (cases, N=22; controls, N=23) were obtained as previously described 49; data available at NCBI Gene Expression Omnibus (GEO) accession

GSE41826. Peripheral blood samples (cases, N=8; controls, N=43) were obtained from consenting individuals in a Johns Hopkins prospective cohort of pregnant women (PPD cohort), as previously described 116; data available from GEO accession GSE44132. Data from individuals in the Grady Trauma Project (cases, N=63; controls, N=313) were downloaded from the NCBI GEO accession GSE72680 112-115. Data on TRIER social stress test cohort (N = 85) was downloaded from GEO accession GSE77445 80. All cohorts used in model generation and validation are described in detail in Table S1.

DNA Methylation Analysis. Study data was derived from genome-wide DNA methylation data using the Infinium HumanMethylation450 BeadChip Array (Illumina

Inc., San Diego, CA). DNA methylation profiles for GenRED Offspring cohort blood

(cases, N=4; controls, N=14) and saliva (cases, N=5; controls, N=15), respectively, were generated as described below.

Infinium Chip Assay. Bisulfite-converted DNA was analyzed using Illumina’s Infinium

HM450 BeadChip Kit (WG-314-1001) by following manufacturer’s protocol. The

BeadChip contains 485,577 CpG loci in human genome. Briefly, 4 µl of bisulfite- converted DNA was added to a 0.8 ml 96-well storage plate (Thermo Scientific), denatured in 0.014 N sodium hydroxide, neutralized and amplified with kit-provided reagents and buffer at 37 °C for 20-24 hours. Samples were fragmented using kit-

36 provided reagents and buffer at 37 °C for one hour and precipitated by adding 2-propanol.

Re-suspended samples were denatured in a 96-well plate heat block at 95 °C for 20 minutes. 12 µl of each sample was loaded onto a 12-sample chip and the chips were assembled into hybridization chamber as instructed in the manual. After incubation at

48 °C for 16-20 hours, chips were briefly washed and then assembled and placed in a fluid flow-through station for primer-extension and staining procedures. Polymer-coated chips were image-processed in Illumina’s iScan scanner.

Data Acquisition. Data were extracted using Methylation Module of GenomeStudio v1.0

Software and processed using the “minfi” and “wateRmelon” packages in R 117-119. Raw data was trimmed of probes failing quality assessment, followed by scale-based data correction for Illumina type I relative to type II probes. Methylated and un-methylated intensity values were then quantile normalized separately prior to the calculation of the β

(beta) value based on following definition:

 value = (signal intensity of methylation-detection probe)/(signal intensity of

methylation- detection probe + signal intensity of non-methylation-detection

probe + 100).

Sodium Bisulfite Pyrosequencing. Microarray data was validated at select probes in the

GenRED Offspring saliva cohort to corroborate array data (Figure S4). Bisulfite conversion was carried out using EZ DNA Methylation-Gold Kit (Zymo Research, Irvine,

CA) according to the manufacturer’s instructions on N=51 subjects from the Johns

Hopkins Prospective cohort. Nested PCR amplifications were performed with a standard

PCR protocol in 25 µl volume reactions containing 3-4 l of sodium-bisulfite-treated

DNA, 0.2 µM primers, and master mix containing Taq DNA polymerase (Sigma Aldrich,

37 St. Louis, MO). Primer sequences can be found in Table S2. PCR amplicons were processed for pyrosequencing analysis according to the manufacturer’s standard protocol using a PyroMark Q96 MD system (QIAGEN, Germantown, MD) with Pyro Q-CpG

1.0.9 software (QIAGEN) for CpG methylation quantification. Only data passing internal quality checks for sodium bisulfite conversion efficiency, signal to noise ratio, and the observed versus expected match of the predicted pyrogram peak profile using reference peaks were incorporated in subsequent analyses. Data generated derive from one technical replicate.

Blood Analysis in PPD Cohort. Participant blood was collected at each visit in four

10ml EDTA tubes. Blood samples were non-fasting, and collection times were arranged at the convenience of the participant. All occurred during the working day (9:00 am to

5:00 pm). Cytokine levels have a recognized circadian rhythm but are lowest during the daytime; we were unable to control further for time of day in our analyses 120. Samples were immediately centrifuged at 4 °C for 30 minutes. The plasma was then aliquoted in 2 mL microcentrifuge tubes, snap frozen on dry ice, and immediately stored in a -80 °C freezer. Cytokines were analyzed using BD Cytometric Bead Array. Plasma samples from patients were diluted 1:10 and incubated with capture beads coated with antibodies specific for IL-6. Beads were then incubated with a phycoerythrin-conjugated detection reagent containing antibodies specific to each capture bead. The capture bead + analyte + detection reagent complexes produced an individual fluorescent signal for each cytokine and were acquired on a FACSCalibur instrument. The data were analyzed using FCAP

Array software. The limit of detection for IL-6 was 1.6 pg/mL. Proportions of our samples that fell below the limit of detection were as follows: IL-6, 18.9%. Samples that

38 fell below the limit of detection were coded “0.” All samples were run in duplicate and the coefficient of variation between samples was <10%. Analyses were repeated using the lowest detectable dose for those below the limit of detection, and results did not change.

Statistical Analysis. All statistical tests were performed in R (http://www.r-project.org/).

Cross-reactive Illumina probes were removed from data prior to further analysis 121.

Using an Anderson-Darling test from the “nortest” package 122, all distributions of data that rejected the null hypothesis of normality were subsequently evaluated with non- parametric tests. Variance between case and control groups in each sample was assumed to be equal. All statistical tests performed were two tailed and a P < 0.05 was considered significant. All statistics presented are a result of either a linear regression model or a

Monte Carlo method permutation test (1,000 permutations). Unless otherwise specified + denotes the standard deviation of the mean. Cell sub-fraction percentages were quantified for CD8 T cells, CD4 T cells, B cells, NK cells, monocytes, and granulocytes using an algorithm designed by Houseman et al. for the quantification of the cell-types using DNA methylation proxies 123. A buccal cell epigenetic profile was derived by taking the mean of N=109 buccal derived HM450 microarray profiles from a dataset in GEO accession

GSE25892. Incorporation of the buccal derived profile at N=500 HM450 loci into the

Houseman algorithm generated training set and incorporation of a buccal covariate was used to retrain the Houseman algorithm to quantify buccal profiles.

39 2.7 Tables and Figures

Table 1. Gene ontology results.

Probability Expected Rank Estimate Probability P-value Term ID Term Name 1 0.211 0.053 1.74×10-5 GO:0007399 Nervous system development 2 0.15 0.059 3.63×10-5 GO:0030182 Neuron differentiation 3 0.156 0.057 6.40×10-5 GO:0048699 Generation of neurons 4 0.161 0.056 8.19×10-5 GO:0022008 Neurogenesis 5 0.128 0.061 1.42×10-4 GO:0048666 Neuron development 6 0.809 0.036 2.94×10-4 GO:0044424 Intracellular part 7 0.111 0.063 3.77×10-4 GO:0000904 Cell morphogenesis involved in differentiation Cell morphogenesis involved in neuron 8 0.093 0.066 9.76×10-4 GO:0048667 differentiation 9 0.224 0.048 1.08×10-3 GO:0009653 Anatomical structure morphogenesis 10 0.085 0.067 2.16×10-3 GO:0007409 Axonogenesis 11 0.184 0.05 3.32×10-3 GO:0048468 Cell development 12 0.093 0.064 3.37×10-3 GO:0048812 Neuron projection morphogenesis 13 0.271 0.045 3.62×10-3 GO:0065008 Regulation of biological quality 14 0.817 0.035 5.78×10-3 GO:0005622 Intracellular 15 0.271 0.045 6.13×10-3 GO:0009893 Positive regulation of metabolic process 16 0.085 0.065 6.92×10-3 GO:0061564 Axon development 17 0.362 0.041 0.0105 GO:0048856 Anatomical structure development 18 0.709 0.036 0.0107 GO:0043229 Intracellular organelle 19 0.108 0.058 0.0112 GO:0031175 Neuron projection development 20 0.279 0.044 0.0161 GO:0030154 Cell differentiation

Gene ontology analysis of genes significantly associated with weekday AUC cortisol measured in GenRED Offspring blood and saliva samples using the tool g:Profiler.

40 Table 2. Overrepresentation analysis.

Significance Expected Probability Binomial Test Cohort Comparison Level Probability Estimate P-value

0.05 0.507 0.614 3.47×10-102 GenRED Neuron, Offspring 0.01 0.158 0.242 8.70×10-28 Suicide Saliva, Cortisol 0.001 0.019 0.026 0.332

0.05 0.470 0.569 2.97×10-86 GenRED Neuron, Offspring 0.01 0.132 0.209 1.30×10-26 Suicide Blood, Cortisol 0.001 0.015 0.039 9.71×10-3

Combined 0.05 0.522 0.618 1.68×10-81 GenRED Neuron, Offspring Blood 0.01 0.187 0.251 1.34×10-15 Suicide & Saliva, Cortisol 0.001 0.032 0.035 0.851

0.05 0.470 0.578 2.01×10-105 Houtepen et al., Neuron, 0.01 0.119 0.193 6.02×10-32 2016, Cortisol Suicide 0.001 0.011 0.029 3.15×10-3

Combined 0.05 0.522 0.603 6.70×10-60 GenRED Houtepen et Offspring Blood al., 2016, 0.01 0.187 0.264 1.08×10-25 & Saliva, Cortisol Cortisol 0.001 0.032 0.061 4.96×10-3

Overrepresentation analysis of genes significantly associated with AUC weekday cortisol measures in GenRED Offspring blood and saliva samples, genes significantly associated with suicide in post-mortem prefrontal cortex neurons, and genes associated with AUC cortisol in blood from Houtepen et al. 2016.

41 Table 3. Probes from interaction biosignature contributing to the PCA model.

Probe ID Location (hg19) Gene Symbol Gene Name cg01060471 chr10:103911733 NOLC1 nucleolar and coiled-body phosphoprotein 1 cg01252219 chr12:110302105 GLTP glycolipid transfer protein cg02068690 chr2:25600451 DTNB dystrobrevin, beta cg02097235 chr16:1116799 SSTR5-AS1 SSTR5 antisense RNA 1 cg02246725 chr11:2014127 HOTS H19 opposite tumor suppressor cg02340818 chr8:145808436 ARHGAP39 Rho GTPase activating protein 39 cg02464608 chr3:122631723 SEMA5B semaphorin 5B cg02516957 chr8:128011063 cg02953125 chr2:1079100 SNTG2 syntrophin, gamma 2 cg03198117 chrX:152939967 PNCK pregnancy up-regulated nonubiquitous CaM kinase cg03351894 chrX:48686200 ERAS ES cell expressed Ras cg03782771 chr4:152801238 cg03887528 chr2:231090531 SP140 SP140 nuclear body protein cg06363485 chr6:41207376 TREML4 triggering receptor expressed on myeloid cells like 4 cg06774087 chr5:79647614 CRSP8P mediator complex subunit 27 pseudogene cg07483709 chr10:29439573 cg07589300 chr4:1404128 cg07787977 chr1:962651 AGRN agrin cg08119631 chrX:118822815 SEPT6 septin 6 cg08469255 chr6:30851069 DDR1 discoidin domain receptor tyrosine kinase 1 cg08674415 chr16:34430905 cg08720806 chr11:125142671 PKNOX2 PBX/knotted 1 homeobox 2

ADAM metallopeptidase with thrombospondin type cg08795752 chr9:136293271 ADAMTS13 1 motif, 13 cg09105334 chr17:15683060

42 cg11450546 chr19:43965012 LYPD3 LY6/PLAUR domain containing 3 cg11496113 chr5:34627766 cg11780096 chr2:178976254 RBM45 RNA binding motif protein 45 cg11880367 chr10:80141181 LINC00856 cg12378817 chr10:133961235 JAKMIP3 Janus kinase and microtubule interacting protein 3 cg12622680 chr7:158819620 LINC00689 cg12833118 chr4:2572546 cg12967319 chr14:101291997 MEG3 maternally expressed 3 (non-protein coding) cg13466253 chr20:3731334 HSPA12B heat shock 70kD protein 12B cg13798679 chr1:36617570 TRAPPC3 trafficking protein particle complex 3

guanine nucleotide binding protein (G protein), q cg14439102 chr9:80360719 GNAQ polypeptide cg14509196 chr15:25494565 SNORD115-44 small nucleolar RNA, C/D box 115-44

calcium/calmodulin-dependent protein kinase kinase cg15071166 chr17:3771325 CAMKK1 1, alpha cg15302379 chr10:102821848 KAZALD1 Kazal-type serine peptidase inhibitor domain 1 cg15508401 chr2:239997380 HDAC4 histone deacetylase 4 cg15677087 chr20:61584850 SLC17A9 solute carrier family 17, member 9 cg15838142 chr17:77184010 RBFOX3 RNA binding protein, fox-1 homolog 3 cg16900102 chr7:98283412 cg16943635 chr11:62067691 SCGB1D4 secretoglobin, family 1D, member 4 cg17267737 chr7:158819531 LINC00689 cg17813840 chr15:79852956

eukaryotic elongation factor, selenocysteine-tRNA- cg18144654 chr3:127995346 EEFSEC specific cg18546110 chr4:10763497 cg19688321 chrX:70476177 BCYRN1 brain cytoplasmic RNA 1

43 cg20704602 chr6:29635371 MOG myelin oligodendrocyte glycoprotein cg20929545 chr11:118958046 HMBS hydroxymethylbilane synthase cg21066636 chr17:4675292 TM4SF5 transmembrane 4 L six family member 5 cg21352158 chr8:832917 ERICH1-AS1 ERICH1 antisense RNA 1 cg21407196 chr1:46751975 LRRC41 leucine rich repeat containing 41 cg21729235 chrX:100739272 ARMCX4 armadillo repeat containing, X-linked 4 cg22029879 chr8:1790861 ARHGEF10 Rho guanine nucleotide exchange factor 10 cg22133973 chr6:170789640 PSMB1 proteasome subunit beta 1 cg22184136 chr6:30038720 RNF39 ring finger protein 39 cg22939113 chr19:40719949 MAP3K10 mitogen-activated protein kinase kinase kinase 10 cg22954621 chr10:13514626 BEND7 BEN domain containing 7 cg23043139 chr3:13678918 FBLN2 fibulin 2 cg23374863 chr14:94984140 SERPINA12 serpin family A member 12 cg24112628 chr6:150174215 LRP11 low density lipoprotein receptor-related protein 11 cg24416190 chr4:186007379 cg24437859 chr12:7066614 PTPN6 protein tyrosine phosphatase, non-receptor type 6 cg25132241 chr14:92396859 FBLN5 fibulin 5 cg25133951 chr1:178575267 cg25215117 chr17:11461665 SHISA6 shisa family member 6 cg25447359 chr22:30790057 cg25477843 chr8:145061318 PARP10 poly (ADP-ribose) polymerase family, member 10 cg26224354 chr7:1096374 GPR146 G protein-coupled receptor 146 cg26305504 chr19:947612 ARID3A AT rich interactive domain 3A

mitogen-activated protein kinase 8 interacting cg26493346 chr16:1812273 MAPK8IP3 protein 3

44 Table 4. Prediction model results.

Cohort N Outcome Cases Controls Interaction AUC 95% CI P-value

GenRED Offspring blood 18 Attempt 4 14 Biosignature 0.768 0.47-1 0.052

GenRED Offspring saliva 20 Ideation 5 15 Biosignature 0.807 0.59-1 0.011

PPD cohort blood 51 Ideation 8 43 Biosignature 0.884 0.75-1 0.041

GTP blood 376 Ideation 63 313 Biosignature 0.500 0.42-0.58 0.724

GTP subset blood 115 Ideation 6 109 Biosignature 0.727 0.59-0.87 0.050

GenRED Offspring blood 18 Attempt 4 14 % Myeloid 0.732 0.40-1 0.28

GenRED Offspring saliva 20 Ideation 5 15 % Myeloid 0.788 0.56-1 0.28

PPD cohort blood 51 Ideation 8 43 % Myeloid 0.834 0.61-1 0.003

GTP subset blood 115 Ideation 6 109 % Myeloid 0.728 0.55-0.9 0.99

Prediction model results for all cohorts using either the interaction biosignature or the myeloid-derived proportion as the interaction term. P-values are derived from Monte-

Carlo permutation.

45

46 Figure 1. Discovery of interaction biosignature probes and prediction of suicidal ideation using interaction biosignature in multiple cohorts. (A) Volcano plot in prefrontal cortex neurons (cases, N=22; controls, N=23) of the interaction of individual probe methylation with rs7208505 methylation and genotype. (B) Probes with an interaction p-value <

0.005 (N = 669) were optimized for prediction of SA in GenRED Offspring blood.

Probes with an AUC prediction above 0.825 (N=72) were used to train a PCA model. (C)

ROC curves of prediction of SI in GenRED Offspring saliva and PPD cohort blood using the first eigenvectors of predicted PCAs. (D) ROC curves of prediction of SI in the whole

GTP cohort and a subset of drug-naïve, non-PTSD individuals (cases=6; controls=109).

47

Figure 2. Myeloid-derived cell proportions correlated with the interaction biosignature in all cohorts, and were predictive of suicidal behavior. Correlations were observed between the interaction biosignature and the proportion of myeloid-derived cells in (A) GenRED

Offspring blood (P = 2.7×10-4), (B) saliva (P = 0.092), (C) PPD cohort blood (P = 0.034), and (D) GTP blood (P = 2.4×10-7).

48

Figure 3. Interaction biosignature and methylation at SHP1 (cg24437859) correlated with inflammatory markers and stress measures in the PPD cohort (trimester > 1). (A)

Levels of IL-6, an inflammatory marker, correlated with myeloid-derived cell proportion

(P = 0.054). (B) The interaction biosignature correlated with the Perceived Stress metric

(P = 0.019). SHP1 methylation correlated with both (C) IL-6 levels (P = 0.035), and (D) myeloid-derived cell proportion (P = 1.4×10-8).

49 2.8 Supplementary Material

Result S1. Monte Carlo permutation was performed at each selection step of the algorithm for discovering the interaction proxy.

The first cutoff of P < 0.005 in prefrontal cortex neuron probes resulted in 669 probes.

Permutation of this cutoff value was performed ×1000 by selecting 669 random probes from the list of prefrontal cortex neuron probes, then selecting from these probes those with an AUC prediction of suicide attempt in the GenRED Offspring blood cohort above the top 25% quantile. The resulting list of probes was used to build a principal component analysis (PCA) model in the GenRED Offspring blood cohort, which was then used to predict PCA models in the other data sets. The first eigenvector was used to assess suicidal ideation prediction in the original prediction model, replacing the stress measure with the PCA first eigenvector. This permutation found the algorithm to be significant or nearly significant in all cohorts assessed: GenRED Offspring blood (P = 0.186), GenRED

Offspring Saliva (P = 0.147), PPD blood (P = 0.018), and GTP subset (P = 0.162).

The second cutoff of selecting AUC prediction greater than the top 25% quantile, which resulted in 72 probes, was permuted by selecting 72 random probes, building a PCA model in GenRED Offspring blood using these random probes, and then following the remainder of the algorithm. This permutation found this algorithm step to be significant or nearly significant in all cohorts: GenRED Offspring blood (P = 0.052), GenRED

Offspring saliva (P = 0.146), PPD blood (P = 0.041), and GTP subset (P = 0.050).

50 GenRED GenRED Grady Trauma Prospective PPD Offspring Blood Offspring Saliva Project

Depression: 50% Depression: 21% Depression: 41% Depression: 73% Diagnosis Bipolar: 6% Bipolar: 0% Bipolar: 4% Bipolar: 27% Control: 44% Control: 79% Control: 55%

Sample size 18 20 52 391

White blood cell, Cell type White blood cell White blood cell White blood cell buccal cell

Cauc: 70% Cauc: 95 % AA: 24% Ethnicity Cauc: 100% AA: 100% AA: 5% Hisp: 2% Asian: 4%

rs7208505 T/T: 16.7% T/T: 36.8% T/T: 42.3% T/T: 56.3% genotype C/T: 50.0% C/T: 42.1% C/T: 50.0% C/T: 40.7% distribution C/C: 33.3% C/C: 21.1% C/C: 7.7% C/C: 3.1%

Suicidal Ideation/Attempt Ideation/Attempt Ideation Ideation/Attempt behavior

Traumatized 22% 16% 13% 42%

Prospective No No Yes No

Table S1. Sample overview.

51 Locus Primer Sequence cg24437859 cg24437859_Fo GAGGTATAGAGTAGGGTATTG cg24437859_Ro2 CAAATCTTATTTCCTCCAAAACC cg24437859_Fib2 /5Biosg/TAGGGTATTGTGGTTTATTTTAGGTT cg24437859_Ri ATCACCTATCTCTATAAACTCCTC cg24437859_seqR1 AAAACAATACCTACCCTAACAA cg22029879 cg22029879_Fo TTTAGATTTAGTTGGGATTAGGTG cg22029879_Ro CTATAATCCCATATACTTTTCCTAAT cg22029879_Fib /5Biosg/TATTAGTTTGAGTGTGAGTTAGATGTTA cg22029879_Ri AAACCTAAAACTAATACCTAATCCCC cg22029879_seqR1 TTATTTCACCCTAAATAACTCTAA cg08469255 cg08469255_Fo GTTGTTTGTTAGTAGGTTTTTTAGT cg08469255_Ro TACTTTTTCCCCACTCAACAC cg08469255_Fib /5Biosg/GGGTTTTTGAATTAGTGTTTATGTGTG cg08469255_Ri TCTCCACACCCCCACACCTA cg08469255_seqR1 CTCACCACCTAAACAATATCA

Table S2. Pyrosequencing assay primer sequences.

52 GenRED Offspring, GenRED Offspring, Prospective PPD, GTP subset, blood blood saliva blood Probe Rho P-value Rho P-value Rho P-value Rho P-value cg08469255 -0.54 0.021 -0.46 0.047 -0.28 0.046 -0.42 3.4×10-4 cg22029879 -0.5 0.036 -0.96 1.1×10-10 -0.55 3.2×10-5 -0.75 < 2.2×10-16 cg24437859 -0.69 0.002 0.67 0.002 -0.49 2.0×10-4 -0.46 6.7×10-5

Table S3. Probes with significant correlations to the interaction biosignature across all data sets.

53

Figure S1. Summary of algorithm identifying Interaction Proxy.

54

Figure S2. HPA axis model depiction

Three dimensional simulations of significant models demonstrating the interaction of

SKA2 interaction proxy Eigen vector 1 with CTQ scores to influences cortisol suppression in the GTP cohort (A) and the AUC cortisol following the TRIER social stress test (B). Data for the GTP cohort derive from drug free individuals without a diagnosis of PTSD.

55

Figure S3. ROC curves of prediction using myeloid-derived cell proportion in place of interaction biosignature in prediction model for GenRED Offspring saliva, and PPD cohort blood, GTP blood and GTP subset blood.

56

Figure S4. DDR1 (cg08469255) array methylation validation by sodium bisulfite pyrosequencing in GenRED Offspring saliva cohort (P = 0.0041; N = 17).

57 3. SUICIDE BIOMARKER GENE SKA2 IS REGULATED BY HIF-

INDEPENDENT HYPOXIA PATHWAY

3.1 Abstract

Background: Suicide is the 10th leading cause of death, and rates are rising, signifying a need for an improved understanding of the underlying biology to improve approaches to suicide prevention. Our laboratory recently identified the spindle and kinetochore associated complex subunit 2 gene (SKA2) as an important suicide biomarker.

Decreased SKA2 expression was associated with suicide in the prefrontal cortex of suicide decedents. SKA2 expression has been shown to be regulated by CREB, PPAR-α, or HIF-1α in various non-neuronal models; however, has not been investigated in neuroblastoma cell lines.

Hypothesis: We hypothesized that SKA2 is regulated by CREB, PPAR-α, or HIF-1α in a manner consistent with the downregulation observed in suicide victims.

Methods: HEK293 and SH-SY5Y cells were screened for responsiveness to forskolin, fenofibrate, and anoxia to identify regulating pathways. Hypoxia signaling was further investigated by treatment with desferrioxamine (DFO), culturing under low glucose, and shRNA knockdown of HIF1A and HIF1B.

Results: Cells were not responsive to treatment with forskolin or fenofibrate, suggesting

SKA2 is not regulated by CREB or PPAR-α. SH-SY5Y, but not HEK293, showed downregulation of SKA2 with anoxia exposure. Both HEK293 and SH-SY5Y showed downregulation of SKA2 in both a time- and dose-dependent manner in response to DFO treatment. Culturing HEK293 under low glucose for seven days also downregulated

58 SKA2. Knockdown of HIF1A and HIF1B did not alter the response of SKA2 to DFO in

HEK293.

Conclusions: We conclude that SKA2 is regulated by a HIF-independent hypoxia responsive pathway.

59 3.2 Background

In the United States, suicide is the 10th leading cause of death for all ages, and the

2nd leading cause of death for ages 10-24, posing a threat to adolescent health.1 Rising suicide rates call for increased research efforts to improve understanding of the underlying biology and to develop prevention strategies.4

Our laboratory recently identified a DNA methylation mark at rs7208505 in the spindle and kinetochore associated complex subunit 2 gene (SKA2) that was associated with suicide in the prefrontal cortex of suicide decedents.49 SKA2 is a cell cycle gene with roles in silencing the spindle checkpoint during mitosis.56, 59 In our original work, SKA2 mRNA expression was negatively correlated with rs7208505 methylation and was significantly decreased in suicide decedents compared to health controls. These findings were consistent with those of subsequent independent studies reporting decreased SKA2 mRNA in the blood of violent suicide completers and both decreased mRNA and protein expression in the prefrontal cortex of suicide victims,49, 68, 69, 124 implying potential involvement SKA2 in suicide etiology. Rice et al. previously identified SKA2 as a cytosolic binding partner and nuclear chaperone for the glucocorticoid receptor (GR), which is involved in regulating the stress response as part of the hypothalamic-pituitary- adrenal (HPA) axis.59 Dysregulation of the HPA-axis and the stress response is a consistently observed endophenotype in suicidal individuals.125-127

Despite compelling evidence that SKA2 may be involved in the biology of suicide, few studies have investigated its regulation in the context of baseline or stress related signaling pathways. Cao et al. reported decreased expression and promoter activity of

SKA2 with CREB (cAMP response element binding protein) siRNA,64 which was

60 supported by the study by Zhuang et al. that reported decreased SKA2 mRNA and protein with CREB knockdown and showed binding of CREB to the SKA2 promoter by chromatin immunoprecipitation.65 A recent study reported that an increase in cAMP levels in striatal neurons in the mouse forebrain was associated with an altered response to stressors, providing evidence of a link between CREB and the stress response.128

Gonsalves et al. identified putative response elements for PPAR-α (peroxisome proliferator-activated receptor alpha) and HIF-1α (hypoxia inducible factor 1 alpha) in the SKA2 promoter region, and demonstrated SKA2 dependence on PPAR-α and HIF-1α for placental growth factor-induced expression of SKA2 with shRNA knockdown experiments.66 PPAR-α has been shown to interfere with NF-κB transcriptional activity,129 which could alter many downstream inflammatory roles of NF-κB,130 consistent with literature on abnormal neuroimmune function in suicide and recent findings of an interaction between an immune modulation biosignature and SKA2 methylation.14, 131 Hypoxia has important implications in fetal brain development,132 and

HIF is involved in inflammation pathways,133 suggesting roles for hypoxia signaling in neuroimmune function and SKA2 regulation.14

Given the experimental evidence of CREB, PPAR-α, and HIF-1α regulation of

SKA2 and the potential involvement of each of these genes in the stress response and immune pathways, we hypothesized that CREB, PPAR-α, or HIF-1α regulates SKA2 expression in a manner consistent with the downregulation of SKA2 mRNA and protein observed in suicide decedents.49, 68, 69, 124 The goal of this study was to increase knowledge of SKA2 regulation and identify pathways by which SKA2 may be involved in suicide biology.

61 3.3 Results

3.3.1 Screening of candidate pathways to determine involvement in SKA2 regulation

To test the involvement of CREB in the regulation of SKA2, we treated HEK293 and SH-SY5Y cells with for four hours with 25 µM forskolin to increase intracellular levels of cyclic AMP (cAMP). Increased expression was observed in two CREB- regulated genes, brain-derived neurotrophic factor (BDNF) and proto-oncogene c-FOS, in both SH-SY5Y and HEK293. No change was observed in expression of SKA2 (Figure

1 A-F). Longer treatments also did not modify expression of SKA2 in either cell line

(Figure 1 G-L).

To investigate SKA2 regulation by PPAR-α, cells were stimulated with the

PPAR-α agonist fenofibrate at 50 µM. A 24-hr treatment of SH-SY5Y cells resulted in increased expression of PPAR-α regulated genes apolipoprotein A-II (APOA2) and carnitine palmitoyltransferase IA (CPT1A), but did not alter expression of SKA2 (Figure

2 A-C). Treatment of HEK293 for 8-48 hours resulted in increased expression of APOA2 and CPT1A between 16 and 48 hours, which was associated with modest increases in

SKA2 expression at 16 and 24 hours, but not at the 48 hour time point (Figure 2 D-F).

Due to this inconsistency of the SKA2, APOA2 and CPT1A responses to fenofibrate treatment in HEK293 and the lack of any response in SH-SY5Y, this line of investigation was not further pursued.

To assess the involvement of HIF-1α in SKA2 regulation, SH-SY5Y and HEK293 were incubated in an atmosphere without oxygen (5% CO2, 95% N2) for 24 hours. Both cell lines showed robust increases in the expression of genes regulated by HIF-1α including vascular endothelial growth factor (VEGFA) and glucose transporter 1

62 (GLUT1). In contrast, a decrease was observed in SKA2 expression SH-SY5Y cells. In

HEK293, no change was observed in SKA2 expression even after a 24 hour incubation without oxygen (Figure 3). Increased exposure time (24-72 hours) of HEK293 to anoxia resulted in a time-dependent increase in VEGFA expression, but no change in SKA2 expression (Figure S1 A-B). In THP-1 (monocytic leukemia) and D425

(medulloblastoma) cell lines, the anoxic conditions resulted in increased expression of

VEGFA and GLUT1, but no effects were in SKA2 response (Figure S1 C-H).

3.3.2 Characterization of SKA2 response to hypoxia

In an attempt to further characterize the response of SKA2 to hypoxia stimulus, we treated cells with an intracellular iron chelator, desferrioxamine mesylate (DFO). Iron chelators stabilize HIF-1α by inhibiting the iron-sensitive enzyme prolyl hydroxylase

(PHD), thereby preventing ubiquitination and subsequent degradation of HIF-1α by ubiquitin ligase von Hippel-Lindau tumor suppressor (VHL). Treatment of SH-SY5Y and

HEK293 with 20-100 µM DFO for 24 hours resulted in dose-dependent activation of

HIF-1α downstream genes VEGFA and GLUT1, which corresponded with decreased expression of SKA2 in both cell lines (Figure 4). Treatment of SH-SY5Y and HEK293 with 50 µM DFO for 8-24 hours, followed by removal of DFO after 24 hours, revealed a time-dependent decrease in SKA2 expression until removal of DFO at 24 hours, followed by recovery of expression in the 24 hours following DFO removal (Figure 5 A, C). This was consistent with the VEGFA response at these time points, which showed increased expression of VEGFA up to 24 hours, then decreased expression following DFO removal

63 (Figure 5 B, D), suggesting that these responses are orchestrated by the activation of the same iron-dependent pathway.

Given the known ties of HIF activation to metabolic pathways, and the co- occurrence of hypoxia and glucose deprivation in the body,134 we investigated the role of glucose concentration in regulating the expression of SKA2 under normoxic and anoxic conditions. Culturing HEK293 in low glucose (1.0 g/L) DMEM for one week resulted in decreased expression of SKA2 and increased expression of VEGFA compared to normal high glucose (4.5 g/L) DMEM (Figure 6). This downregulation of SKA2 in low glucose conditions was also observed in cells cultured in anoxia, which was associated with an enhanced VEGFA response (Figure 6). These observations provide evidence of SKA2 regulation by hypoxia-related metabolic pathways.

3.3.3 Determination of genes involved in SKA2 downregulation by hypoxia

To test the hypothesis that SKA2 downregulation by anoxia and DFO was due to

HIF-1α signaling, we generated stable HIF1A knockdown (KD) cell lines in HEK293 using shRNA against HIF1A (Table 1), which gave >60% KD efficiency in the four shRNAs screened (Figure S2 A), and 75-85% efficiency in the two stable KD lines created (Figure S2 B). Treatment of HIF1A KD lines with 50 µM DFO for 8 and 24 hours showed a blunted response in VEGFA and GLUT1 expression compared to empty vector (EV), indicating a functional knockdown of HIF-1α (Figure 7). However, knockdown of HIF1A did not alter the SKA2 response at 24 hours (Figure 7 D), indicating that HIF-1α is not involved in the downregulation of SKA2 in HEK293 induced by DFO. Although HIF-1α is responsible for the majority of hypoxia-induced

64 alterations of gene expression, other members of the HIF family, HIF-2α or HIF-3α, dimerize with HIF-1β (ARNT) to bind to DNA.135 To assess the involvement of these other HIF family members, we generated HIF1B KD lines in HEK293 using shRNA against HIF1B (Table 1). A 70-90% KD efficiency was achieved in the two stable KD lines selected (Figure S3). Consistent with the results from the HIF1A KD, treatment of

HIF1B KD lines with 50 µM DFO for 8 and 24 hours showed a blunted response in

VEGFA and GLUT1 expression compared to EV, indicating a functional knockdown of

HIF-1β (Figure 8). Also consistent with the HIF1A KD results, HIF1B KD did not alter the SKA2 response at 24 hours (Figure 8 D), leading us to conclude that the downregulation of SKA2 in response to DFO and anoxia does not occur through HIF pathways.

3.3.4 Investigation of SKA2 involvement in hypoxia signaling pathways

Given the small number of studies investigating the roles of SKA2 outside of the cell cycle, we tested the hypothesis that SKA2 is involved in HIF signaling pathway. We generated stable SKA2 knockdown (KD) cell lines in HEK293 using shRNA against

SKA2 (Table 1), which gave 50-95% KD efficiency in the four shRNAs screened

(Figure S4 A), and 70-90% efficiency in the two stable KD lines created (Figure S4 B).

Treatment of SKA2 KD lines with anoxic gas for 24 hours revealed no alterations in the

VEGFA response compared to EV (Figure 9 A). Treatment of SKA2 KD lines with 25-

100 µM DFO for 24 hours also showed no alteration of the VEGFA response compared to

EV (Figure 9 B). We concluded that SKA2 is not involved in hypoxia signaling pathways.

65 3.4 Discussion

In this study we sought to improve understanding of the regulation of SKA2 by testing the hypothesis that SKA2 is regulated by CREB, PPAR-α, of HIF-1α, in a manner consistent with multiple reports of downregulation of SKA2 observed in suicide decedents.49, 68, 69, 124 We tested this hypothesis by screening stimulants of each pathway-- alterations were not observed in SKA2 expression after treatment with the CREB activator forskolin, or the PPAR-α agonist fenofibrate. This screen did, however, reveal downregulation of SKA2 by anoxia in SH-SY5Y, but not HEK293. Treatment of cells with DFO, an intracellular iron chelator used to induce hypoxia pathways, also showed downregulation of SKA2, providing further evidence of SKA2 regulation by hypoxia pathways. To confirm that the downregulation of SKA2 by anoxia and DFO was occurring through HIF-1α, we used shRNA to knockdown HIF1A, then treated cells with

DFO and observed no difference in SKA2 downregulation, suggesting that SKA2 is not, in fact, regulated by HIF-1α. We then used shRNA to knockdown HIF-1β, another member of the HIF complex, and observed no alteration in the SKA2 response, giving additional evidence that SKA2 is regulated by a hypoxia pathway independent of HIF. Further experiments revealed that SKA2 is also downregulated in low-glucose conditions, potentially providing evidence of other regulating pathways, and that knockdown of

SKA2 did not alter the hypoxia signaling, implying that SKA2 is only downstream of hypoxia pathways, but is not involved in the upstream hypoxia response.

In our model, SKA2 did not respond to treatment with CREB activator, forskolin, despite a strong response in upregulated expression of CREB positive control genes,

BDNF and FOS. This was contrary to a previous report by Cao et al., which reported

66 decreased SKA2 expression with knockdown of CREB.64 We did not detect any alteration of SKA2 expression with the activation of CREB, potentially indicating that

CREB is not involved in the regulation of SKA2 under stress conditions. In addition to a different approach to evaluating CREB regulation of SKA2, Cao et al. utilized A549 cells, a lung carcinoma cell line, which likely has different characteristics than the cell lines we tested, SH-SY5Y and HEK293, potentially accounting for differences in response of

SKA2 to CREB.

Similar to CREB, PPAR-α experiments did not present consistent or compelling evidence for a major regulatory role of PPAR-α in the cell lines examined. Previously,

Gonsalves et al. identified a putative response element in the SKA2 promoter for PPAR-α, and demonstrated that PPAR-α mediated the SKA2 response to placental growth factor stimulation using shRNA knockdown.66 PPAR-α may regulate SKA2 expression in the context of placental growth factor stimulation, but may not be responsible for SKA2 expression outside of this pathway, consistent with our observations. Additionally, due to low baseline expression of PPARA in the human brain compared to the heart, kidney, and liver,136 it is probable that PPAR-α may not regulate SKA2 expression in the brain.

Culturing multiple cell lines under anoxic gas revealed a significant downregulation of

SKA2 in SH-SY5Y, consistent with the decreased expression of SKA2 observed in suicide decedents. 49, 68, 69, 124 This decrease in SKA2 expression coincided with increased expression of hypoxia positive control genes VEGFA and GLUT1. The fold change of the

VEGFA and GLUT1 responses in SH-SY5Y (>30-fold and >10-fold, respectively) compared to the other lines tested (<10-fold and <8-fold, respectively) suggests higher sensitivity of SH-SY5Y to anoxia than other cell lines, providing rationale for the

67 decreased SKA2 expression observed only in this cell line. Increased sensitivity of SH-

SY5Y to anoxia is consistent with previously reported results.137

Given literary evidence that hypoxia-related pathways are potentially related to suicide, the regulation of SKA2 by hypoxia may have important implications for improving understanding of suicide biology. Decreased levels of VEGF, a gene directly regulated by hypoxia, are associated with suicide attempt and completion,138 providing a potential link between hypoxia signaling, SKA2 response, and suicide. Maternal stress has been linked to alterations of the intrauterine environment, including gestational hypoxia, which can reprogram the offspring HPA-axis response and anxiety-like behaviors later in life.139, 140 Given the link between anxiety in adults and risk of suicidal ideation,141 maternal stress and resulting fetal hypoxia could represent a topic for future investigation to understand early-life risk factors for suicide.

In addition to the SKA2 response to anoxia, we also observed SKA2 downregulation with treatment of iron chelator, DFO. Although we did not observe a change in SKA2 expression in HEK293 with anoxia treatment, we did observe decreased

SKA2 expression with DFO treatment, which was also observed in SH-SY5Y. This could indicate that HEK293 are more sensitive to iron chelation than anoxia, which activate the same hypoxia sensitive pathways.142 Low intracellular iron levels have been previously shown to slow cell growth,143 which could account for the observed decrease in SKA2 expression, given the role of SKA2 in the cell cycle,56, 59 however, the relevance of decreased cell proliferation may not be relevant in the brain, the primary organ of concern in suicide. Interestingly, cellular iron homeostasis, which is partially regulated by

68 HIF-1α, has been linked to oxidative stress, inflammation, and major depressive disorder, further implying a potential role of hypoxia signaling in suicide.144

Interestingly, knocking down HIF subunits, HIF1A and HIF1B, did not change the SKA2 response to DFO, even though the VEGF and GLUT1 responses were attenuated. This was surprising, given the pervasive role of HIF in orchestrating the cellular response to hypoxia conditions.142 Previously, Gonsalves et al. observed increased expression of SKA2 in endothelial cells stimulated by placental growth factor

(PlGF). They identified two hypoxia response elements (HREs) in the SKA2 promoter region that attenuated the SKA2 response to PlGF when removed from the promoter.66

This study, however, did not attempt to stimulate the HIF pathway to alter expression of

SKA2, but only deleted HREs from the SKA2 promoter, suggesting that HIF may be important for control of SKA2 expression in the context of PlGF stimulation, but not under hypoxic conditions. There are a number of other major cellular pathways that exhibit HIF-independent hypoxia responses and may be responsible for the regulation of

SKA2 under hypoxic conditions: P53,145 NF-κB,146 NRF2,147 and mTOR.148 NF-kB subunits NFKB1 and RelA both have annotated binding sites in the SKA2 promoter region, and mTOR-associated transcription repressor protein YY1 has two binding sites in the SKA2 promoter region,149 suggesting that these hypoxia-responsive pathways may regulate SKA2 expression. We recommend an investigation of these pathways in the context of hypoxia to determine their potential role in regulating SKA2.

We discovered further evidence of the relationship of SKA2 and hypoxia signaling by varying the amount of glucose in the cell media. Cells were cultured in various concentrations of glucose under normoxic or anoxic conditions for one week, and

69 we observed a decrease in SKA2 expression in low glucose treatment, which was consistent with increased VEGFA expression in both normoxic and anoxic conditions.

This provided evidence that another stimulus, low glucose, induced the same response as anoxia and DFO, suggesting that all of these stimuli activate a common regulatory pathway to induce changes in SKA2 and VEGFA expression. A previous study reported that VEGF is independently regulated by both hypoxia and glucose starvation,134 providing support for our observations. Yun et al. reported that VEGF mRNA stabilization under glucose starvation conditions occurs through 5'-adenosine monophosphate-activated protein kinase (AMPK),150 which is an upstream regulator of the mTOR pathway,151 a candidate pathway for SKA2 regulation in hypoxia.

In addition to investigating the role of hypoxia in SKA2 regulation, we also tested the hypothesis that SKA2 may be involved in HIF signaling pathways, given the previous report of the role of SKA2 as a nuclear chaperone for the GR.59 Cells with knockdown of

SKA2 did not exhibit any differences in the VEGFA response to anoxia or DFO treatment, suggesting that SKA2 is not involved in HIF signaling, but is only responsive to hypoxia stimuli. Previous studies have reported growth retardation in A549 and HeLa S3 cells with knockdown of SKA2,56, 57, 59 which should be tested in our HEK293 knockdown model to confirm consistency with previous models. Given the limited knowledge of the

SKA2 function, further experiments are recommended to characterize the roles of SKA2 in the cell, especially roles outside of cell cycle.

This study is limited by the choice of model cell lines—HEK293 and SH-SY5Y.

Although useful for studying signaling pathways, neither of these lines is ideal for studying suicide-related pathways in our organ of interest, the brain. SKA2 was observed

70 to be downregulated in the prefrontal cortex neurons of suicide decedents,49 which are post-mitotic cells and would not be affected by growth alterations associated with SKA2 dysregulation. The downregulation of SKA2 observed in this study could reflect growth retardation induced by low oxygen, low iron, or low glucose conditions, which would not be related to SKA2 downregulation in post-mitotic neurons. The development of a post- mitotic neuron model by differentiating SH-SY5Y with retinoic acid and BDNF would allow us to better understand the role of SKA2 downregulation in suicide.

3.5 Conclusions

We conclude that SKA2 is downregulated by hypoxia stimuli, including anoxia, low iron, and low glucose, in a manner consistent with SKA2 downregulation observed in suicide decedents. This regulation, however, occurs through a HIF-independent hypoxia responsive pathway, and will require further investigation to determine the specific pathway involved. We further conclude that SKA2 in not involved in hypoxia signaling pathways.

71 3.6 Materials & Methods

Cell Culture. HEK293 were maintained in high glucose DMEM (ThermoFisher Scientific,

Waltham, MA, Cat. No. 11995) supplemented with 10% FBS (Sigma-Aldrich, St. Louis,

MO, Cat. No. F4135) and 1× Penn-Strep (ThermoFisher Scientific, Waltham, MA, Cat.

No. 15140122). SH-SY5Y were maintained in low glucose DMEM media (ThermoFisher

Scientific, Cat. No. 11885) supplemented with 10% FBS and 1× Penn-Strep. THP-1 were maintained in RPMI-1640 (ThermoFisher Scientific, Cat. No. 11875) supplemented with

10% FBS and 1× Penn-Strep. D425 were maintained in MEM media (ThermoFisher

Scientific, Cat. No. 11095) supplemented with 10% FBS, 1× Penn-Strep, 1 mM sodium pyruvate (ThermoFisher Scientific, Cat. No. 11360), and 1× MEM non-essential amino acids (ThermoFisher Scientific, Cat. No. 11140). All cells were incubated at 37 °C with

5% CO2.

Cell Treatments. Cell treatments included forskolin (Cell Signaling Technology, Danvers,

MA, Cat. No. 3828), fenofibrate (Sigma-Aldrich, St. Louis, MO, Cat. No. F6020), and desferrioxamine mesylate (Sigma-Aldrich, St. Louis, MO, Cat. No. D9533). For anoxia treatment, cells were placed in a hypoxia incubation chamber, and then filled with an anoxic gas mixture (95% N2 and 5% CO2) and incubated at 37 °C.

shRNA Knockdown by Lipofectamine. shRNA plasmids (pLKO.1 vector) from the The

RNAi Consortium (TRC) library (Broad Institute, Cambridge, MA) were obtained as bacterial glycerol stocks from the Johns Hopkins ChemCORE (Baltimore, MD). Bacteria were grown in 50 mL overnight cultures in LB Broth (Quality Biological, Gaithersburg,

72 MD, Cat. No. 340-004-101) supplemented with 100 µg/mL ampicillin (Corning Cellgo,

Corning, NY, Cat. No. 61-238-RH). Plasmids were isolated using the QIAfilter Plasmid

Midi Kit (QIAGEN, Germantown, MD, Cat. No. 12243) following manufacturer’s instructions. To knockdown gene expression, HEK293 cells were plated at 1×105 per 3.5 cm well and grown overnight. A lipofectamine mixture, added dropwise to wells, was prepared by mixing 3 µL Lipofectamine 3000, 2 µL P3000 reagent (ThermoFisher

Scientific, Waltham, MA, Cat. No. L3000008), 1 µg shRNA plasmid, and 250 µL Opti-

MEM media (ThermoFisher Scientific, Cat. No. 31985070). Lipofectamine mixture was left on cells for 24 hours. Cells were collected for RNA expression analysis after 48 hours.

Generation of Stable Cell Lines. Lentiviral particles were generated by transfecting 2×106

HEK293 cells with 3 µg of the shRNA plasmid of interest, 2 µg of lentiviral packaging plasmid psPAX2 (Addgene, Cambridge, MA, Cat. No. 12260), and 1 µg lentiviral envelope package pMD2.G (Addgene, Cambridge, MA, Cat. No. 12259) using 12 µL

Lipofectamine 3000 and 18 µL P3000 reagent prepared in 1 mL Opti-MEM media.

Media containing lentiviral particles was collected after two days and spun down to remove cell debris. HEK293 and SH-SY5Y cells in 6-well tissue culture dishes at 60-

70% confluence were treated with 0.5 mL virus media, 4.5 mL high glucose DMEM, and

8 µg/mL polybrene (Santa Cruz Biotechnology, Dallas, TX, Cat. No. sc-134220).

Transduced cells were incubated for 24 hours, then media was replaced to remove virus and polybrene. After reaching confluence, transduced cells were moved to T-25 tissue culture flasks and maintained in normal media. Cells were selected for successful viral

73 transduction using 0.5-1.0 µg/mL puromycin (Sigma-Aldrich, St. Louis, MO, Cat. No.

P8833).

Nucleic Acid Isolation. RNA was isolated using TRIzol reagent (QIAGEN, Germantown,

MD, Cat. No. 79306) following the manufacturer’s protocol.

Quantitative Real Time PCR (qRT-PCR). cDNA was made from 1 µg of RNA using 5 mM MgCl2, GeneAmp 10X PCR Buffer II (ThermoFisher Scientific, Waltham, MA, Cat.

No. N8080130), 175 µM dNTP mix (ThermoFisher Scientific, Waltham, MA, Cat. No.

R0141, R0151, R0161, R0171), 20 U RNase Inhibitor (ThermoFisher Scientific,

Waltham, MA, Cat. No. N8080119), 2.5 µM Oligo d(T )16 primer (ThermoFisher

Scientific, Waltham, MA, Cat. No. N8080128), and 50 U MuLV Reverse Transcriptase

(ThermoFisher Scientific, Waltham, MA, Cat. No. N8080018). Samples were incubated at 37 °C for 60 minutes, and then inactivated at 95 °C for 5 minutes. cDNA was diluted two-fold prior to further use. qRT-PCR was performed on 2 µL cDNA using Power

SYBR Green PCR Master Mix (ThermoFisher Scientific, Waltham, MA, Cat. No.

4367659) and 160 nM of each primer. All samples were run in triplicate. Fold-change expression was calculated using the −ΔΔCt method (Applied Biosystems), where triplicate Ct values for each sample were averaged and subtracted from the geometric mean of reference gene Ct values according to previously published methods.152 Unless otherwise indicated, RPLP0 was used at the reference gene in all experiments. Primer sequences for all genes tested are listed in Table 2.

74 Statistical Analysis. Unless otherwise noted, all statistics presented result from a two- group Student’s t-test. A P-value below 0.05 was considered to be significant.

75 3.7 Tables and Figures

Table 1. shRNA Sequences

TRC Library No. Target Gene Target Sequence pLKO_TRC001 Empty Vector N/A TRCN0000003810 HIF1A GTGATGAAAGAATTACCGAAT TRCN0000010819 HIF1A TGCTCTTTGTGGTTGGATCTA TRCN0000003819 HIF1B (ARNT) GAGAAGTCAGATGGTTTATTT TRCN0000003820 HIF1B (ARNT) AGCCTCATCATCGTTCAAGTT TRCN0000072570 SKA2 GCGGCAGAGCAATTCAAATTT TRCN0000072571 SKA2 TGTCAGTGATAAAGTCTCGAT

76 Table 2. qRT-PCR Primers

Target Forward Primer (5’-3’) Reverse Primer (5’-3’) Gene SKA2 GCCGCATTTGTGCTACTGTG CTCTGCCGCAGTTTTCTCTT APOA2 GGAGCCATGTGTGGAGAGC CAGTTCCGTTCCAGCCTTCT FOS GGGGCAAGGTGGAACAGTTAT CCGCTTGGAGTGTATCAGTCA VEGFA AGGGCAGAATCATCACGAAGT AGGGTCTCGATTGGATGGCA GLUT1 ATTGGCTCCGGTATCGTCAAC GCTCAGATAGGACATCCAGGGTA ACTB ACGACATGGAGAAAATCTGGCA AGGCGTACAGGGATAGCACA GAPDH ACAACTTTGGTATCGTGGAAGG GCCATCACGCCACAGTTTC RPLP0 GCAGCATCTACAACCCTGAAG CACTGGCAACATTGCGGAC HIF1A ATCCATGTGACCATGAGGAAATG TCGGCTAGTTAGGGTACACTTC HIF1B CAGAGGCCACAACTAGGTC GGAATGATTGTAGCTGGCCAGT TET1 GAACAGCCATCAGATCTGTAAG ACTGTAGTCCATGGATTCTGAC

77

78 Figure 1. Forskolin stimulation of CREB does not alter SKA2 expression in SH-SY5Y or

HEK293. Cells were treated with 25 µM forskolin or DMSO vehicle for 4 or 24 hours.

SH-SY5Y treated for 4 hours were evaluated for (A) SKA2 expression, and CREB activation positive control genes (B) BDNF and (C) FOS. HEK293 were treated for 4 hours and evaluated for expression of (D) SKA2, (E) BDNF, and (F) FOS. SH-SY5Y were treated for 24 hours and evaluated for expression of (G) SKA2, (H) BDNF, and (I)

FOS. HEK293 were treated for 24 hours and evaluated for expression of (J) SKA2, (K)

BDNF, and (L) FOS. N = 3 per treatment group; fold-change expression normalized to

ACTB. Significance determined by two group Student’s t-test.

79

Figure 2. PPAR-α agonist fenofibrate did not alter SKA2 expression in SH-SY5Y or

HEK293. SH-SY5Y were treated with 50 µM fenofibrate or DMSO vehicle for 24 hours.

Gene expression was evaluated for (A) SKA2 and PPAR-α activation positive control genes (B) CPT1A and (C) APOA2, with fold-change expression normalized to ACTB.

HEK293 were treated with 50 µM fenofibrate for 8-24 hours, then mRNA expression was determined by qRT-PCR, and fold-change expression was normalized to the geometric mean of ACTB and RPLP0. Gene expression was evaluated for (D) SKA2, (E)

CPT1A, and (F) APOA2, with fold-change expression normalized to the geometric mean of ACTB and RPLP0. *P<0.05; N=3 per treatment group.

80

Figure 3. Anoxia treatment in SH-SY5Y, but not HEK293, downregulated SKA2 expression. Cells were treated with anoxic gas (5% CO2, 95% N2) for 24 hours. Gene expression in SH-SY5Y was determined for (A) SKA2 and HIF-1α activation positive control genes (B) VEGFA and (C) GLUT1. Gene expression in HEK293 was determined for (D) SKA2, (E) VEGFA and (F) GLUT1. N=3 per treatment group.

81

Figure 4. SKA2 expression was downregulated by DFO in a dose-responsive manner.

Cells were treated with 20-50 µM DFO for 24 hours to activate hypoxia pathways. SH-

SY5Y were evaluated for expression of (A) SKA2, (B) VEGFA, and (C) GLUT1 expression. Gene expression in HEK293 was determined for (D) SKA2, (E) VEGFA and

(F) GLUT1. N=3 per group, outliers removed. *P<0.05, **P<0.01, **P<0.001, significance compared to untreated control.

82

Figure 5. SKA2 expression was downregulated by DFO in a time-dependent manner.

Cells were treated with 50 µM DFO for 8-24 hours, followed by removal of DFO and measurement of gene expression up to 48 hours. SH-SY5Y (A, B) and HEK293 (C, D) were evaluated for changes in SKA2 and VEGFA expression. N=3 per group; error bars represent standard error. *P<0.05, **P<0.01, **P<0.001, significance compared to vehicle time point control.

83

Figure 6. SKA2 expression and hypoxia pathways are controlled by glucose concentration. HEK293 cells were cultured in DMEM with high glucose (4.5 g/L) or low glucose (1.0 g/L) under normoxic or anoxic conditions for 7 days. Expression was measured for SKA2 and VEGFA.

84

Figure 7. SKA2 downregulation by DFO is not mediated by HIF-1α. HEK293 cells with stable knockdown of HIF1A were treated with 50 µM DFO for 8 hours (A-C) and 24 hours (D-F), and evaluated for expression changes in SKA2, VEGFA, and GLUT1.

*P<0.05, **P<0.01, **P<0.001, significance compared to vehicle control.

85

Figure 8. SKA2 downregulation by DFO is not mediated by HIF-1β. HEK293 cells with stable knockdown of HIF1B were treated with 50 µM DFO for 8 hours (A-C) and 24 hours (D-F), and evaluated for expression changes in SKA2, VEGFA, and GLUT1. N=3, outliers removed. *P<0.05, **P<0.01, **P<0.001, significance compared to vehicle control.

86

Figure 9. SKA2 does not mediate hypoxia signaling pathways. (A) HEK293 cells with stable knockdown of SKA2 were treated with anoxia gas for 24 hours and evaluated for

VEGFA expression. (B) HEK293 cells with stable knockdown of SKA2 were treated with

50 µM DFO for 24 hours and evaluated for changes in VEGFA expression. N=3 per group; significance compared to group control, *P<0.05, **P<0.01, **P<0.001.

87 3.8 Supplementary Figures

88 Figure S1. HEK293 anoxia time course and screening of THP-1 and D425 for SKA2 response to anoxia treatment. HEK293 cells were cultured under anoxia or normoxia conditions for 24-72 hours and evaluated for (A) SKA2 and (B) VEGFA expression; N=3 per group. THP-1 (monocytic leukemia) cells were cultured under anoxia or normoxia for

24 hours and evaluated for (C) SKA2, (D) VEGFA, and (E) GLUT1 expression; N=3 per group. D425 cells (medulloblastoma) were cultured under anoxia or normoxia for 24 hours and evaluated for (F) SKA2, (G) VEGFA, and (H) GLUT1 expression; N=6 per group. Error bars represent group standard error. *P<0.05, **P<0.01, **P<0.001, significance compared to time point control.

89

Figure S2. HIF1A knockdown confirmation. (A) HEK293 cells were transfected with shRNA plasmids targeting HIF1A or pLKO empty vector (EV) control; N=2 per group.

(B) Confirmation of HIF1A knockdown in HEK293 cells transduced with lentivirus containing shRNA plasmids targeting HIF1A or EV; N=3 per group. *P<0.05, **P<0.01,

**P<0.001, significance compared to EV.

90

Figure S3. HIF1B knockdown confirmation. Confirmation of HIF1B knockdown in

HEK293 cells transduced with lentivirus containing shRNA plasmids targeting HIF1B or

EV. N=3 per group, outliers removed. *P<0.05, **P<0.01, **P<0.001, significance compared to EV.

91

Figure S4. SKA2 knockdown confirmation. (A) HEK293 cells were transfected with shRNA plasmids targeting SKA2 or EV. (B) Confirmation of SKA2 knockdown in

HEK293 cells transduced with lentivirus containing shRNA plasmids targeting SKA2 or

EV. N=2 per group.

92 4. CONCLUSIONS, PERSPECTIVES, AND FUTURE STUDIES

4.1 Summary & Conclusions

The work presented in Chapters 2 and 3 is largely built on the previous study in our lab by Guintivano et al. that established SKA2 as a suicide biomarker, presenting a prediction model that incorporates rs7208505 genotype and blood methylation to predict suicidal behavior, and describing reported decreased expression of SKA2 in the brains of suicide decedents.49 To further these findings, we first expanded our previous genome- wide DNA methylation analysis to investigate underlying suicide biology, as presented in

Chapter 2, and then shifted our focus to an in vitro system to increase our knowledge of

SKA2 regulation, as presented in Chapter 3. The ultimate goal of both of these studies was to add to the body of knowledge of suicide biology, allowing for improved identification and treatment of at-risk individuals.

In Chapter 2, we sought to improve our existing SKA2 suicide prediction model by replacing the perceived stress/anxiety questionnaires used in the original prediction model. Using a custom bioinformatic brain to blood discovery algorithm, we derived a

DNA methylation biosignature that interacted with SKA2 methylation to improve the prediction of suicidal ideation in our existing suicide prediction model across both blood and saliva data sets, interacted with HPA-axis and immune modulation metrics, that was validated in an independent cohort. We conclude that this biosignature interacts with

SKA2 methylation to improve suicide prediction and may represent a biological state of immune and HPA-axis modulation that mediates suicidal behavior.

In Chapter 3 we examined the involvement of different signaling pathways in the regulation of SKA2 based on consensus sequences identified on the UCSC genome

93 browser. We did not find convincing evidence for the involvement of PPAR-α or CREB in the regulation of SKA2. Data was presented showing the sensitivity of SKA2 expression to hypoxia. Interestingly, SKA2 downregulation was observed in response to anoxia, low glucose, and to the iron chelator desferrioxiamine. However, neither HIF-1α nor its associated transcription factor, HIF-1β, were found to be involved in the response.

We concluded that SKA2 is regulated by a HIF-independent hypoxia-responsive pathway in a manner consistent with the downregulation of SKA2 in the brains of suicide decedents.

4.2 Significance & Innovation

The goal of the work presented in this dissertation was to improve knowledge of suicide biology and improve a suicide prediction model. In the studies presented, we discovered a novel suicide biosignature that both improved the prediction of suicidal behavior by SKA2 and correlated with immune modulation and HPA-axis metrics. This enhanced suicide prediction model allows, for the first time, prediction of suicidal behaviors using only unbiased biological measures, innovatively replacing the use of questionnaires in suicide prediction. Our observations of biosignature associations with

HPA-axis metrics and immune modulation improve the general knowledge of underlying suicide biology and provide new insight into the molecular mechanisms that may be involved. We also present SKA2 regulation by hypoxia, which provides new information to this limited field concerning the role and regulation of this important suicide biomarker.

This study not only concludes that SKA2 is modified by hypoxia in a unique HIF- independent mechanism, but is also the first to study SKA2 regulation in the context of a

94 disease. Together, these studies offer an enhanced suicide prediction model and new understanding of underlying biology and molecular mechanisms. Ideally, identifying molecular signaling pathways that are involved in HPA-axis dysregulation will allow for development of treatments targeting these pathways, allowing for intervention to normalize HPA-axis regulation to prevent future suicide attempts.

4.3 Future Studies

The results presented in Chapters 2 and 3 make a significant contribution to the field of suicide research, but more research will be needed to fully comprehend the results and translate the findings into products that can be directly used to improve the lives of at-risk individuals. The biosignature model presented in Chapter 2 will require additional validation in independent cohorts before it can be established as a valid measure to improve suicide prediction. Ideally, independent cohorts would represent various at-risk populations, such as groups with childhood trauma, PTSD, MDD, or bipolar disorder. Such cohorts represent groups with higher rates of suicide than the general population, allowing for validation of the biosignature in populations where identification of at-risk individuals is needed the most. Additionally, these disorders each exhibit unique underlying biology which may or may not be consistent with suicide, and therefore could affect the prediction accuracy of the biosignature. For example, suicidal ideation is associated with lack of cortisol suppression in response to the DST,22-25 which is consistent with DST in individuals with a history of childhood trauma153 or MDD,154 but inconsistent with increased cortisol suppression observed in PTSD.155 Given associations of the biosignature with HPA-axis metrics, differences in HPA-axis

95 regulation between cohorts with these disorders could present challenges to accurately predicting suicide using the biosignature model. Conversely, PTSD,156, 157 childhood adversity,158-162 MDD,163-166 and bipolar disorder167-169 exhibit increased levels of pro- inflammatory cytokines (e.g. IL-6, C-reactive protein, TNF-α) compared to healthy controls. These similarities between disorders may be reflected in the biosignature, allowing for increased prediction accuracy in these at-risk populations. Further replication of the biosignature in large cohorts will be essential in determining its validity for future use in suicide prediction. The biosignature should also be further evaluated in replication cohorts for associations with immune modulation and HPA-axis measures to improve understanding of its biological relevance.

Since the biosignature was generated and validated in several small cohorts, it is possible that it will not replicate in further studies due to the inherent limitations of using small cohorts representing a specific population. In this event, a similar approach to that presented in Chapter 3 could be taken in a larger discovery cohort to generate a more robust DNA methylation biosignature. Using a larger cohort would allow for improved identification of differences of DNA methylation at various loci in respect to suicidal behavior and/or interaction with SKA2, leading to a biosignature that represents consistent DNA methylation alterations related to underlying suicide biology. Alternate approaches could also be used in generating a DNA methylation biosignature in a larger cohort, such as linear discriminate analysis (LDA), hierarchical clustering analysis

(HCA) or an alternative use of principal component analysis (PCA). LDA is used to recognize patterns in data to create a linear combination that explains variation in the data.

This approach could be used to combine loci that are differentially methylated in respect

96 to suicidal behavior, HPA-axis dysregulation, or immune modulation metrics to create a biosignature representing these changes. HCA is used in large data sets to cluster similar measures by a given variable, thereby allowing one to visualize how data groups are related to one another in respect to that variable. In the context of biosignature generation,

HCA could be used to cluster loci that show similar DNA methylation changes with respect to suicidal behavior, and then each of these clusters could be tested for associations with metrics of HPA-axis dysregulation or immune modulation to determine groups of loci with consistent DNA methylation changes with respect to these metrics.

Clusters would then be tested for interaction with SKA2 to identify candidate clusters for the creation of a biosignature. PCA, an approach used to generate the presented biosignature, is used to identify linear combinations of variables to explain large amounts of variation in data. A PCA approach could be applied to any combination of loci, perhaps generated by an HCA approach, to create a single measure that explain variation across these loci to generate a novel biosignature.

While we discovered in Chapter 3 that SKA2 is responsive to hypoxia, we failed to identify the pathway through which this regulation occurs. Additional experiments are needed to elucidate which hypoxia-responsive HIF-independent pathway is responsible for this regulation. Candidate pathways include P53,145 NF-κB,146 NRF2,147 and mTOR,148 which are discussed in Chapter 3. Identification of the responsible pathway would provide further insight into SKA2 regulation and related suicide biology, potentially identifying targets for the development of risk factors and possibly treatments.

Additionally, SKA2 regulation should be studied in post-mitotic cells to evaluate the relevance of our findings in the brain. Beyond this study, future experiments should seek

97 to understand the purpose of SKA2 downregulation in the brains of suicidal individuals.

Although there are not currently models of suicide in rodents, models of chronic variable stress may have relevance to aspects of suicide biology in humans and prove useful in discovering the role of SKA2 in suicide. Chronic variable (or unpredictable) stress has been shown to increase HPA-axis activity and corticosterone (rodent cortisol equivalent) in rodent models,170, 171 and promotes proliferation of inflammatory leukocytes,172 recapitulating important endophenotypes of human suicide biology. SKA2 could be investigated in these models to further elucidate its role in suicide biology and possible reasons for its downregulation in the brains of suicide decedents.

Although SKA2 has been established as an important biomarker in the prediction of suicidal behavior, it remains unclear whether SKA2 is etiologically important in suicide. Guintivano et al. and others have reported decreased expression of SKA2 mRNA in the prefrontal cortex of suicide decedents,49 suggesting a role for SKA2. Additionally,

Rice et al. reported SKA2 involvement in glucocorticoid signaling,59 implying that SKA2 may be involved in HPA-axis pathways. Beyond these studies, however, there is little evidence that SKA2 is anything more than a predictive biomarker of suicidal behavior

Additionally, it is currently unknown whether SKA2 rs7208505 methylation or expression changes with alteration of suicidal ideation or attempt status. These questions could be answered by animal models, as mentioned above, and longitudinal studies. Ideally, longitudinal studies would regularly assess suicidal behavior and collect specimen for tracking of SKA2 expression and rs7208505 methylation status to determine changes with alteration of suicidal behavior.

98 4.4 Perspectives on Suicide Biomarker Research & Implications of Findings

Currently, the field of suicide biomarkers is rapidly expanding with regards to the number of biomarkers, but suffers from a deficiency in replication studies, and a lack of understanding of the biology of these markers and how they fit into the bigger picture of suicidal behavior. Numerous studies have identified promising biomarkers of suicide, which include gene expression (mRNA or protein), epigenetic, and genetic markers.

Gene expression biomarkers include NR3C1,36-39 SKA2,49 SLC4A,52 SAT1,173 BCL2,74

S100A10,174 PTEN,175 MARCKS,175 MAP3K3,175 and VEGF.138 Epigenetic biomarkers include SKA2 and NR3C1,37, 49 and genetic biomarkers include NR3C1 and FKBP5.9, 35,

42-48 Although these many biomarkers have associations with suicidal behavior, few of these markers have been replicated or shown to be useful in predicting suicidal behavior in a prospective manner. For example, Niculescu et al. identified SLC4A4 expression as a biomarker of suicidal ideation (SI) in male psychiatric patients--changes in SLC4A4 expression predicted SI with an AUC of 93% in bipolar patients.52 Despite this high prediction rate, this biomarker has not been replicated in further studies. The field would benefit from a concerted effort to validate these biomarkers by replication in independent cohorts. It is also important to differentiate between suicidal attempt (SA) and suicidal ideation (SI) when characterizing biomarkers. The incidence of SI is much higher than

SA, with 8.3 million adults reporting SI between 2008 and 2009, and an estimated 1 million reporting attempting suicide.3 Though closely related, there are biological disparities between SI and SA that should be taken into account when discovering or validating biomarkers. For example, Melhem et al. reported decreased hair cortisol concentrations, GR-α mRNA expression, and increased levels of CRP, TNF-α, and IL-6

99 in the blood of patients with a history of SA compared to healthy controls, with none of these differences occurring when comparing to patients with a history of SI.40 The data from this study indicates that biological differences between SA and SI that affect suicide biomarkers, suggesting a need to differentiate between these groups in replication studies.

Additionally, it would be beneficial to improve understanding of the biology underlying these biomarkers to discover how they could work together in an effort to enhance suicide prediction. For example, individual biomarkers identified in these studies may have associations with suicidal behavior (SA or SI), but may actually represent a specific aspect or endophenotype of suicide, such as immune modulation or inflammation, HPA- axis dysregulation, stress or anxiety, genetic risk, or childhood abuse. Increased knowledge of what aspect of suicide these biomarkers represent will enhance the ability to integrate them into suicide prediction models to identify at-risk individuals, and potentially pinpoint individual vulnerabilities contributing to suicidal behavior, allowing for more individualized treatment. Such associations could be identified in large longitudinal studies with serial time points in which biomarkers are compared to suicidal behavior changes and suicide endophenotype measures, such as pro-inflammatory cytokines, HPA-axis metrics, stress/anxiety metrics, or trauma exposure. Comparing changes in biomarker genes with these metrics could improve understanding of how each biomarker interacts with the underlying biology and identify biomarkers of endophenotypes that are predictive of future suicide. To date, studies have not been reported using the suggested approach to improve our understanding of existing suicide biomarkers. However, Melhem et al. recently measured hair cortisol concentrations

(HCC) to quantify stress exposure over 3 months prior to suicide attempt, and compared

100 this to a number of the above mentioned suicide biomarkers, including NR3C1, SKA2, and FKBP5, and pro-inflammatory cytokines.40 The authors observed decreased HCC with suicide attempt, which was associated with decreased NR3C1 mRNA in blood,40 providing evidence for a link between biomarker gene NR3C1 and a longitudinal measure of stress. Longitudinal studies or studies using longitudinal measures of suicide endophenotypes, such as that presented by Melhem et al., should be performed to increase understanding of suicide biomarkers.

Although promising, our biosignature prediction model and other existing suicide prediction models52, 74, 175 do not yet have strong enough prediction capabilities for use in a clinical setting. Such models give an average prediction area-under-the-curve (AUC) of

70-90%, with variable true positive (sensitivity) and false positive (specificity) rates.

Ideally, a prediction model would have 100% sensitivity and 100% specificity, translating into detecting 100% of all true positives with 0% false positives. Current prediction models, however, have less than ideal sensitivity and specificity, leading to mischaracterization of cases and controls. If implemented in a clinical setting, a model with 80% specificity would falsely categorize 20,000 out of 100,000 people tested, which is drastically higher than the low suicide base rate of 10-20 per 100,000, even with an estimated attempt rate of 25 attempts per completed suicide. Therefore, prediction models need to be dramatically improved before they are implemented in a clinical setting.

Deficiencies in specificity and sensitivity may arise from interindividual variation that is not accounted for in the prediction model. In the case of a biosignature prediction model, the biosignature could be trained on a much larger data set to account for variability.

Additionally, variation may derive from current psychosocial status which may not be

101 determined by biological measures. In addition to predictive capabilities, current biomarker models are limited by a lack of longitudinal data concerning how the biomarker is altered by changes in SI/SA status. This deficit could be rectified by large, longitudinal studies that seek to understand biological changes that occur with alterations of SI or SA status. Though small in scale, Niculescu et al. recently demonstrated the usefulness of this type of approach.52 The authors evaluated psychiatric patients at time points with changes in SI status (high, low, or none), and identified SLC4A4 as the top biomarker predictive of SI and future hospitalizations due to suicide attempt across multiple psychiatric disorders. This study demonstrates the usefulness of longitudinal studies in identifying predictive biomarkers that are altered by SI or SA status.

In addition to improving suicide prediction, the studies presented have implications for the development of novel suicide treatments. We observed associations between our biosignature and both HPA-axis and immune modulation metrics, which implies that these pathways could be targeted for treatment of suicidal behavior.

Currently, treatments for suicide are largely based on psychotherapy and a number of biologic treatments that seek to treat underlying mental illness: antipsychotics including clozapine, antidepressants, lithium, electroconvulsive therapy, transcranial magnetic stimulation, and ketamine.176 Interestingly, ketamine is an anesthetic that has shown fast- acting antidepressant properties when used at sub-anesthetic doses, and has shown promise as a suicide treatment.176 Although the evidence is limited, studies suggest that ketamine may act to regulate the inflammatory response,177 thereby targeting a key aspect of suicide biology. Additionally, a recent meta-analysis reported that glucocorticoid therapy treatment of adolescents for acute lymphoblastic leukemia caused adrenal

102 insufficiency following the cessation of treatment,178 providing evidence for pharmacological alteration of HPA-axis function. As suggested by our biosignature, developing new treatments that target inflammatory pathways or the HPA-axis could be effective in treating suicidal individuals. As discussed in Chapter 3, hypoxia and low iron conditions regulate SKA2 expression, and have postulated links to suicide,138-140, 144 and may represent additional pathways to be further investigated for development of treatments.

103 5. APPENDIX A: CHROMATIN CONFORMATION CAPTURE IN SKA2

5.1 Introduction

Our laboratory recently identified a DNA methylation mark at cg13989295

(rs7208505) in the spindle and kinetochore associated complex subunit 2 gene (SKA2) that was associated with suicide in the prefrontal cortex of suicide decedents.49 The single nucleotide polymorphism (SNP) rs7208505 is located in the 3’ untranslated region (UTR) of the SKA2 gene and has two genotypes: C or T. The C genotype creates a cytosine- guanine dinucleotide (CpG), which can be methylated. In this study, increased methylation was associated with decreased expression of SKA2 in prefrontal cortex neurons, implying a potential regulatory role for SKA2 3’UTR methylation.49

Additionally, Guintivano et al. reported immunoprecipitation enrichment for the glucocorticoid receptor (GR) in both the 3’UTR and promoter regions of the SKA2 gene, and observed an association between rs7208505 methylation and methylation near the transcription start site (TSS) and intron 1 regions.49 Together, these results suggest an interaction between the SKA2 3’UTR and promoter regions, mediated by the GR, by which methylation at rs7208505 alters promoter methylation and influences gene expression (Figure 1).

To test this hypothesis, we employed chromatin conformation capture (3C). This technique involves cross-linking of DNA to proteins, thereby preserving any three- dimensional chromatin interactions, which is followed by digestion with a restriction enzyme to fragment DNA and then ligation. The resulting 3C product represents regions

104 of DNA that are proximal in three dimensional space due to chromatin conformation, but distant in terms of sequence (Figure 2).

5.2 Results

To determine whether or not there exists a chromatin interaction between the

SKA2 promoter and 3’UTR, we performed 4C, a 3C method that includes an additional restriction enzyme digestion and ligation step to shorten end product, on SK-N-BE and

SH-SY5Y neuroblastoma cells treated with 100 nM dexamethasone or ethanol vehicle for

1 hour. SK-N-BE cells have a T/T genotype at rs7208505, and SH-SY5Y have C/T genotype, allowing us to account for differences in chromatin interaction due to genotype.

To test for the interaction of various promoter sequences with the 3’UTR, we used a nested PCR design in which the 4C product was amplified outwards from rs7208505, presumably amplifying all interacting regions in the ligated sequence, then performed a second round of PCR with primers targeting sequential promoter and intronic regions beginning at the transcription start site. PCR results revealed the presence of promoter region 1 (chr17:57232524-57232686) in the rs7208505 4C product, but showed little amplification of other promoter and intronic regions (Figure 3). There were no differences observed between the two cell lines tested or between treatment conditions.

To confirm these findings, 3C products were generated for SK-N-BE and SH-SY5Y under the same treatment conditions. If two regions of DNA are found to interact by 3C, the two sequences will be joined into one, creating a junction of sequences that doesn’t exist in genomic DNA. To detect this junction, we performed PCR with primers designed against the promoter and 3’ regions (Table 1), and detected the expected sequences based

105 on the hypothesized interaction between the promoter (chr17: 57232473-57233460) and rs7208505 (chr17: 57187711-57187931) sequences (Figure 4), with no differences between cell lines or treatment groups. These results were confirmed by Sanger sequencing (results not shown).

Given that we observed the promoter-3’UTR interaction in both SH-SY5Y and

SK-N-BE cells regardless of treatment group, we sought to test the hypothesis that the 3C interaction was unique to these neuroblatoma cell lines. We performed 3C on HEK293

(human embryonic kidney; rs7208505 T/T) and THP-1 (monocytic leukemia; rs7208505

C/T) cells treated with 1 µM dexamethasone or ethanol vehicle for 24 hours. PCR analysis showed that the promoter-3’UTR interaction existed in both of these cells types regardless of treatment (Figure 5). These results were confirmed by Sanger sequencing

(results not shown). Together, these results suggest that this promoter-3’UTR interaction exists in multiple cell types and therefore may be integral to the regulation of SKA2.

In all cell lines examined, the promoter-3’ interaction was found to exist regardless of treatment group, cell type, or genotype. To determine the effectiveness of the dexamethasone treatment used, we measured changes in expression of SKA2 and glucocorticoid response positive control gene SGK1 in SH-SY5Y, HEK293, and THP-1

(Figure 6). We observed expected upregulation of SGK1 in HEK293 and THP-1, which was associated with downregulation of SKA2 expression in these lines (Figure 6 C-F), but did not observe any response in SH-SY5Y (Figure 6 A-B). We concluded that the promoter-3’ interaction observed was not altered by dexamethasone treatment.

Finally, we evaluated SKA2 rs7208505 and promoter region methylation to evaluate whether dexamethasone treatment-induced SKA2 expression changes were

106 associated with methylation changes in the promoter and 3’ interaction regions. We observed no changes in HEK293 promoter region methylation or in THP-1 promoter or rs7208505 methylation with 1 µM dexamethasone treatment for 24-48h (Figure 7 C-G).

Additionally, we did not observe any changes in SH-SY5Y promoter or rs7208505 methylation with 100 nM dexamethasone treatment for 1-24h (Figure 7 A-B). Taken together, these results lead us to conclude that although there is a promoter-3’ interaction observed in all samples, this interaction may not have a regulatory function in the context of dexamethasone treatment. Further experiments are needed to investigate this hypothesis.

5.3 Discussion & Conclusions

In this study we tested the hypothesis that a three-dimensional interaction between the SKA2 promoter and 3’UTR regions is responsible for mediating an association between 3’UTR rs7208505 methylation and SKA2 expression. We confirmed, through both PCR analysis and Sanger sequencing in multiple cell types that an interaction exists between a 3’UTR fragment containing rs7208505 and a fragment in the SKA2 promoter region, allowing us to conclude that there is, indeed, a chromatin conformation basis for a potential effect of rs7208505 methylation on SKA2 gene expression. This observation is novel and will provide the basis for future research on the role of the rs7208505 methylation biomarker and regulation of SKA2 in the context of suicide. Interestingly,

Klengel et al. recently reported a regulatory mechanism in the FKBP5 gene in which rs1360780 genotype controlled long-range interactions of the FKBP5 TSS with introns 2 and 7, which was associated with altered transcriptional activation by the GR in the

107 context of childhood trauma.179 This report gives us confidence that our findings are real and should be further investigated to determine the full regulatory potential.

This study is limited by the lack of robust treatment effect on the methylation at rs7208505 and expression of SKA2 in SH-SY5Y. This cell line was used in the bulk of our experiments due to its unique heterozygous (C/T) genotype at rs7208505, which allowed us to investigate alterations in methylation and expression with dexamethasone treatment. Unfortunately, we did not observe altered methylation at rs7208505 or changes in expression with the dexamethasone treatment used, which prevented us from further investigating our hypothesis of a regulatory role of rs7208505 methylation in SKA2 expression. To successfully investigate this hypothesis in the future, the authors recommend establishing a model in which cells are highly responsive to the treatment used (e.g. dexamethasone), then investigating the time responsiveness of changes in methylation and gene expression. Additionally, site-directed mutagenesis could be used to alter the rs7208505 genotype to allow for different ranges of methylation (0% for T/T,

~50% for C/T, >80% for C/C) to be tested for influence on gene expression.

5.4 Materials & Methods

Cell Culture. SH-SY5Y and SK-N-BE cells were maintained in low glucose DMEM media (ThermoFisher Scientific, Cat. No. 11885) supplemented with 10% FBS (Sigma-

Aldrich, St. Louis, MO, Cat. No. F4135) and 1× Penn-Strep (ThermoFisher Scientific,

Cat. No. 15140122). HEK293 were maintained in high glucose DMEM (ThermoFisher

Scientific, Cat. No. 11995) supplemented with 10% FBS and 1× Penn-Strep. THP-1 were

108 maintained in RPMI-1640 (ThermoFisher Scientific, Cat. No. 11875) supplemented with

10% FBS and 1× Penn-Strep. All cells were incubated at 37 °C with 5% CO2.

Cell treatments. Cells were treated with varying concentrations of dexamethasone

(Sigma-Aldrich, Cat. No. D1756) or ethanol vehicle for time indicated in experiments.

Chromatin Conformation Capture (3C). Method was adapted from 4C technique by

Splinter et al., completed through the first digestion and ligation step.180 Briefly, cells were fixed in 2% formaldehyde, lysed, and nuclei isolated. Restriction enzyme DpnII

(New England Biolabs, Ipswich, MA, Cat. No. R0543L) was used to digest DNA. T4

DNA Ligase (New England Biolabs, Cat. No. M0202S) was used to ligate DNA.

Following ligation, crosslinks were removed by incubation with Proteinase K (Epicentre,

Madison, WI, Cat. No. MPRK092) and RNase A (Epicentre, Cat. No. MRNA092). 3C product was isolated by ethanol precipitation, then purified using QIAquick PCR

Purification Kit (QIAGEN, Cat. No. 28104). Samples for which 4C experiments were completed used NlaIII (New England Biolabs, Cat. No. R0125S) as the second restriction enzyme.

Nucleic Acid Isolation. RNA was isolated using TRIzol reagent (QIAGEN, Germantown,

MD, Cat. No. 79306) following the manufacturer’s protocol. DNA was isolated using

Epicentre MasterPure DNA Purification Kit (Illumina, Cat. No. MCD85201) following the manufacturer’s protocol.

109 Polymerase Chain Reaction (PCR). PCR was performed in 25 µL reactions using 200

µM dNTP mix (ThermoFisher Scientific, Cat. No. R0141, R0151, R0161, R0171), 400 nM primers, and 1.25 U Taq Polymerase with 1× ThermoPol Reaction Buffer (New

England Biolabs, Cat. No. M0267). The thermocycler profile was as follows: 95 °C for

30 sec, then 95 °C for 20 sec, primer annealing temperature (58-63 °C) for 30 sec, 68 °C for 30 sec repeated for 30-40 cycles, followed by 68 °C for 5 min. Primers are listed in

Table 1; primers to evaluate SKA2 rs7208505 methylation were previously published.49

Products were run on a 2% agarose gel containing ethidium bromide to determine enrichment of target sequences.

Sanger Sequencing. Samples were submitted to the Johns Hopkins School of Medicine

Genetic Resources Core Facility for Sanger sequencing analysis.

Quantitative Real Time PCR (qRT-PCR). cDNA was made from 1 µg of RNA using the

QuantiTect Reverse Transcription Kit (QIAGEN, Cat. No. 205311). cDNA was diluted two-fold prior to further use. qRT-PCR was performed on 2 µL cDNA using TaqMan

Universal Master Mix (ThermoFisher Scientific, Cat. No. 4440038) and 1 µL of each

TaqMan Probe. TaqMan Probes used are as follows: SKA2 - Hs00735057_m1, SGK1 -

Hs00985033_g1, ACTB - Hs99999903_m1. All samples were run in triplicate. Fold- change expression was calculated using the −ΔΔCt method (Applied Biosystems), where triplicate Ct values for each sample were averaged and subtracted from the geometric mean of reference gene Ct values according to previously published methods.152 Unless otherwise indicated, ACTB was used at the reference gene in all experiments.

110

DNA Methylation Analysis. DNA bisulfite treatment and pyrosequencing were performed as previously described.49

Statistical Analysis. Unless otherwise noted, all statistics presented result from a two- group Student’s t-test. A P-value below 0.05 was considered to be significant.

111 5.5 Tables and Figures

Table 1. PCR Primer Sequences

Target Forward Primer (5’-3’) Reverse Primer (5’-3’) Region rs7208505 4C AGATGGCTCTGGGATGTGAT CAATCCCAACAATCATCATCAC product SKA2 Promoter CAGAGGGGCGTTACCCAG TAGTTGACATTCCGCAGACCG Region 1 SKA2 Promoter AATGATTCCTGTGCTATCTTCTGA ACCACCCCATCATTAAGTGAAG Region 2 SKA2 Promoter GCGTTTAATCAGTAAATTGCCTTG TGCATAGATGTAGAAAGGAATTTTGT Region 3 SKA2 Promoter TGTCCAAAATGGCAACCCTACT CAGACGTGTTTCATAATGCAGTAA Region 4 SKA2 Promoter GAGACAGGGTCTCACTGTGT CAAGACTTCGGCTTTGCTTAAC Region 5 SKA2 Promoter ATTGACAGTGACATCCTATCTCAA ATTATAAGCAGACCTACACACTGA Region 6 Promoter-3’ CCCAAAGTGCTGGGATTACA TTATTTCTTTGTTATCTGAGCTCG Outside Promoter-3’ GCTCTTGTCATGAGATGGCTCT TGCTCGGCCTGGCTTCATCT Inside

112

Figure 1. Overview of Hypothesis. We hypothesize that there exists an interaction between the SKA2 3’UTR and promoter regions, mediated by the GR, by which methylation at rs7208505 influences gene expression.

113

Figure 2. Overview of 3C Technique.

114

115 Figure 3. Detection of SKA2 transcription start site fragment in 4C product enriched for rs7208505 fragment. 4C was performed in SK-N-BE and SH-SY5Y treated with 100 nM dexamethasone or ethanol vehicle for 1 hour. PCR was performed on 4C product to amplify fragments interacting with the rs7208505 fragment. Promoter region fragments were amplified by PCR in the rs7208505 enriched product to detect promoter fragments interacting with rs7208505 in the 4C product. Samples: SK-N-BE (C1, C2, D1, D2), SH-

SY5Y (C3, C4, D3, D4), gDNA positive control (+), negative control (-). (A) promoter region 1 (chr17:57232524-57232686), (B) promoter region 2 (chr17:57228993-

57230161), (C) promoter region 3 (chr17:57228794-57228992), (D) promoter region 4

(chr17:57227965-57228793), (E) promoter region 5 (chr17: 57227483-57227964), (F) promoter region 6 (chr17: 57225492-57225846). Products were run on a 2% agarose gel.

N=2 per cell line, per treatment group; all samples not shown on gel.

116

Figure 4. Detection of promoter-3’UTR (rs7208505) junction in 3C product. 3C was performed in SK-N-BE and SH-SY5Y treated with 100 nM dexamethasone or ethanol vehicle for 1 hour. PCR was performed with one primer located in the 3’UTR fragment containing rs7208505 (chr17: 57187711-57187931) and the other located in the promoter fragment (chr17:57232473-57233460). Products were run on a 2% agarose gel. N=2 per cell line, per treatment group; all samples not shown on gel.

117

Figure 5. Detection of SKA2 promoter-3’ interaction in HEK293 and THP-1. 3C was performed in HEK293 and THP-1 treated with 1 µM dexamethasone or ethanol vehicle for 24 hours. PCR was performed with one primer located in the 3’UTR fragment containing rs7208505 (chr17: 57187711-57187931) and the other located in the promoter fragment (chr17:57232473-57233460). Products were run on a 2% agarose gel. N=2 per cell line, per treatment group.

118

Figure 6. HEK293 and THP-1, but not SH-SY5Y responded to dexamethasone treatment.

Cells were treated with 1 µM dexamethasone or ethanol vehicle for 24 hours. Expression of SGK1, glucocorticoid activation positive control gene, and SKA2, were determined by qRT-PCR in (A-B) SH-SY5Y, (C-D) THP-1, and (E-F) HEK293. N=3 per group.

119

120 Figure 7. Dexamethasone treatment did not alter SKA2 promoter or rs7208505 methylation. SH-SY5Y were treated with 100 nM dexamethasone evaluated for changes in (A) promoter methylation at 6 hours, and (B) rs7208505 methylation at 1-24 hours.

THP-1 cells were treated with 1 µM dexamethasone and evaluated for changes in promoter methylation at (C) 24 hours and (D) 48 hours, and (E) rs7208505 methylation at

24 and 48 hours. HEK293 cells were treated with 1 µM dexamethasone and evaluated for changes in promoter methylation at (F) 24 hours and (G) 48 hours. N=3 per group, outliers removed. *P<0.05, **P<0.01, ***P<0.001.

121 References

1. Web-based Injury Statistics Query and Reporting System (WISQARS) [online].

Centers for Disease Control and Prevention: National Center for Injury Prevention

and Control. 2016, [June 30, 2016]. Available from URL:

http://www.cdc.gov/injury/wisqars.

2. American Foundation for Suicide Prevention: Suicide Statistics [Online]. (2017).

Available from: URL: https://afsp.org/about-suicide/suicide-statistics/ . [2017

Mar 27].

3. Crosby A, Han B, Ortega L, Park S, Gfroerer J. Suicidal Thoughts and Behaviors

Among Adults Aged ≥18 Years — United States, 2008–2009 MMWR 2011;

60(13).

4. Pringle B, Colpe LJ, Heinssen RK, Schoenbaum M, Sherrill JT, Claassen CA et al.

A Strategic Approach for Prioritizing Research and Action to Prevent Suicide.

Psychiatric Services 2013; 64(1): 71-75.

5. Conwell Y, Duberstein P, Cox C, Herrmann J, Forbes N, Caine E. Relationships

of age and axis I diagnoses in victims of completed suicide: a psychological

autopsy study. American Journal of Psychiatry 1996; 153(8): 1001-1008.

6. Harris E, Barraclough B. Suicide as an outcome for mental disorders. A meta-

analysis. Br J Psychiatry 1997; 170: 205-228.

7. Jamison K. Suicide and bipolar disorder. J Clin Psychiatry 2000; 61: 47-51.

8. Chen EYH, Chan WSC, Wong PWC, Chan SSM, Chan CLW, Law YW et al.

Suicide in Hong Kong: a case-control psychological autopsy study. Psychological

Medicine 2006; 36(6): 815-825.

122 9. Yin H, Galfalvy H, Pantazatos SP, Huang Y-y, Rosoklija GB, Dwork AJ et al.

Glucocorticoid receptor-related genes: genotype and brain expression

relationships to suicide and major depressive disorder. Depression and Anxiety

2016; 33(6): 531-540.

10. Bondy B, Buettner A, Zill P. Genetics of suicide. Mol Psychiatry 2006; 11(4):

336-351.

11. Joiner TJ, Brown J, Wingate L. The psychology and neurobiology of suicidal

behavior. Annu Rev Psychol 2005; 56(287-314 ).

12. Mann JJ. Neurobiology of suicidal behaviour. Nat Rev Neurosci 2003; 4(10):

819-828.

13. O’Brien B, Sher L. Child sexual abuse and the pathophysiology of suicide in

adolescents and adults. International Journal of Adolescent Medicine and Health

2013 25(3): 201-205.

14. Pandey GN. Biological basis of suicide and suicidal behavior. Bipolar Disorders

2013; 15: 524-541.

15. Liu D, Diorio J, Tannenbaum B, Caldji C, Francis D, Freedman A et al. Maternal

Care, Hippocampal Glucocorticoid Receptors, and Hypothalamic-Pituitary-

Adrenal Responses to Stress. Science 1997; 277(5332): 1659.

16. Francis D, Diorio J, Liu D, Meaney MJ. Nongenomic Transmission Across

Generations of Maternal Behavior and Stress Responses in the Rat. Science 1999;

286(5442): 1155.

123 17. Weaver ICG, Cervoni N, Champagne FA, D'Alessio AC, Sharma S, Seckl JR et

al. Epigenetic programming by maternal behavior. Nat Neurosci 2004; 7(8): 847-

854.

18. Van den Bergh BRH, Mulder EJH, Mennes M, Glover V. Antenatal maternal

anxiety and stress and the neurobehavioural development of the fetus and child:

links and possible mechanisms. A review. Neuroscience & Biobehavioral

Reviews 2005; 29(2): 237-258.

19. Field T, Diego M, Hernandez-Reif M, Vera Y, Gil K, Schanberg S et al. Prenatal

maternal biochemistry predicts neonatal biochemistry. International Journal of

Neuroscience 2004; 114(8): 933-945.

20. Oberlander TF, Weinberg J, Papsdorf M, Grunau R, Misri S, Devlin AM. Prenatal

exposure to maternal depression, neonatal methylation of human glucocorticoid

receptor gene (NR3C1) and infant cortisol stress responses. Epigenetics 2008;

3(2): 97-106.

21. Capron LE, Glover V, Pearson RM, Evans J, O’Connor TG, Stein A et al.

Associations of maternal and paternal antenatal mood with offspring anxiety

disorder at age 18 years. Journal of Affective Disorders 2015; 187: 20-26.

22. Yerevanian BI, Feusner JD, Koek RJ, Mintz J. The dexamethasone suppression

test as a predictor of suicidal behavior in unipolar depression. Journal of Affective

Disorders; 83(2): 103-108.

23. Coryell W, Schlesser M. The dexamethasone suppression test and suicide

prediction. Am J Psychiatry 2001; 158(5): 748-753.

124 24. Lester D. The dexamethasone suppression test as an indicator of suicide: a meta-

analysis. Pharmacopsychiatry 1992 25(6): 265-270.

25. Norman W, Brown W, Miller I, Keitner G, Overholser J. The dexamethasone

suppression test and completed suicide. Acta Psychiatrica Scandinavica 1990; 81:

120-125

26. Meltzer HY, Perline R, Tricou B, Lowy M, Robertson A. Effect of 5-

hydroxytryptophan on serum cortisol levels in major affective disorders: Ii.

relation to suicide, psychosis, and depressive symptoms. Archives of General

Psychiatry 1984; 41(4): 379-387.

27. Pariante CM, Miller AH. Glucocorticoid receptors in major depression: relevance

to pathophysiology and treatment. Biological Psychiatry; 49(5): 391-404.

28. Mehta D, Binder EB. Gene × environment vulnerability factors for PTSD: The

HPA-axis. Neuropharmacology 2012; 62(2): 654-662.

29. Ramsawh HJ, Fullerton CS, Mash HBH, Ng THH, Kessler RC, Stein MB et al.

Risk for suicidal behaviors associated with PTSD, depression, and their

comorbidity in the U.S. Army. Journal of Affective Disorders; 161: 116-122.

30. Subramaniam M, Abdin E, Seow E, Picco L, Vaingankar J, Chong S. Suicidal

ideation, suicidal plan and suicidal attempts among those with major depressive

disorder. Ann Acad Med Singapore 2014; 43(8): 412-421.

31. WHO International Programme on Chemical Safety Biomarkers in Risk

Assessment: Validity and Validation. 2001. Retrieved

from http://www.inchem.org/documents/ehc/ehc/ehc222.htm.

125 32. Oquendo MA, Sullivan GM, Sudol K, Baca-Garcia E, Stanley BH, Sublette ME

et al. Toward a Biosignature for Suicide. American Journal of Psychiatry 2014;

171(12): 1259-1277.

33. Coordinators NR. Database resources of the National Center for Biotechnology

Information. Nucleic Acids Research 2016; 44(Database issue): D7-D19.

34. Pruitt KD, Brown GR, Hiatt SM, Thibaud-Nissen F, Astashyn A, Ermolaeva O et

al. RefSeq: an update on mammalian reference sequences. Nucleic Acids

Research 2014; 42(Database issue): D756-D763.

35. Park S, Hong JP, Lee J-K, Park Y-M, Park Y, Jeon J et al. Associations between

the neuron-specific glucocorticoid receptor (NR3C1) Bcl-1 polymorphisms and

suicide in cancer patients within the first year of diagnosis. Behavioral and Brain

Functions : BBF 2016; 12: 22.

36. McGowan PO, Sasaki A, D’Alessio AC, Dymov S, Labonté B, Szyf M et al.

Epigenetic regulation of the glucocorticoid receptor in human brain associates

with childhood abuse. Nature neuroscience 2009; 12(3): 342-348.

37. Labonte B, Yerko V, Gross J, Mechawar N, Meaney MJ, Szyf M et al.

Differential Glucocorticoid Receptor Exon 1B, 1C, and 1H Expression and

Methylation in Suicide Completers with a History of Childhood Abuse.

Biological Psychiatry 2012; 72(1): 41-48.

38. Pérez-Ortiz JM, García-Gutiérrez MS, Navarrete F, Giner S, Manzanares J. Gene

and protein alterations of FKBP5 and glucocorticoid receptor in the amygdala of

suicide victims. Psychoneuroendocrinology; 38(8): 1251-1258.

126 39. Pandey GN, Rizavi HS, Ren X, Dwivedi Y, Palkovits M. Region-specific

alterations in glucocorticoid receptor expression in the postmortem brain of

teenage suicide victims. Psychoneuroendocrinology 2013; 38(11): 2628-2639.

40. Melhem NM, Munroe S, Marsland A, Gray K, Brent D, Porta G et al. Blunted

HPA axis activity prior to suicide attempt and increased inflammation in

attempters. Psychoneuroendocrinology 2017; 77: 284-294.

41. Binder EB. The role of FKBP5, a co-chaperone of the glucocorticoid receptor in

the pathogenesis and therapy of affective and anxiety disorders.

Psychoneuroendocrinology 2009; 34: S186-S195.

42. Supriyanto I, Sasada T, Fukutake M, Asano M, Ueno Y, Nagasaki Y et al.

Association of FKBP5 gene haplotypes with completed suicide in the Japanese

population. Progress in Neuro-Psychopharmacology and Biological Psychiatry

2011; 35(1): 252-256.

43. Fudalej S, Kopera M, Wołyńczyk-Gmaj D, Fudalej M, Krajewski P, Wasilewska

K et al. Association between FKBP5 Functional Polymorphisms and Completed

Suicide. Neuropsychobiology 2015; 72(2): 126-131.

44. Brent D, Melhem N, Ferrell R, Emslie G, Wagner KD, Ryan N et al. Association

of FKBP5 Polymorphisms With Suicidal Events in the Treatment of Resistant

Depression in Adolescents (TORDIA) Study. The American journal of psychiatry

2010; 167(2): 190-197.

45. Perroud N, Bondolfi G, Uher R, Gex-Fabry M, Aubry J-M, Bertschy G et al.

Clinical and genetic correlates of suicidal ideation during antidepressant treatment

in a depressed outpatient sample. Pharmacogenomics 2011; 12(3): 365-377.

127 46. Yeo S, Enoch MA, Gorodetsky E, Akhtar L, Schuebel K, Roy A et al. The

influence of FKBP5 genotype on expression of FKBP5 and other glucocorticoid-

regulated genes, dependent on trauma exposure. Genes Brain Behav 2016; 16(2):

223-232.

47. Breen ME, Gaynor SC, Monson ET, de Klerk K, Parsons MG, Braun TA et al.

Targeted Sequencing of FKBP5 in Suicide Attempters with Bipolar Disorder.

PLoS ONE 2016; 11(12): e0169158.

48. Willour VL, Chen H, Toolan J, Belmonte P, Cutler DJ, Goes FS et al. Family-

based association of FKBP5 in bipolar disorder. Molecular psychiatry 2009;

14(3): 261-268.

49. Guintivano J, Brown T, Newcomer A, Jones M, Cox O, Maher BS et al.

Identification and Replication of a Combined Epigenetic and Genetic Biomarker

Predicting Suicide and Suicidal Behaviors. American Journal of Psychiatry 2014;

171(12): 1287-1296.

50. Kaminsky Z, Wilcox HC, Eaton WW, Van Eck K, Kilaru V, Jovanovic T et al.

Epigenetic and genetic variation at SKA2 predict suicidal behavior and post-

traumatic stress disorder. Transl Psychiatry 2015; 5: e627.

51. Sadeh N, Wolf EJ, Logue MW, Hayes JP, Stone A, Griffin LM et al. Epigenetic

Variation at Ska2 Predicts Suicide Phenotypes and Internalizing Psychopathology.

Depress Anxiety 2016; 33(4): 308-315.

52. Niculescu AB, Levey DF, Phalen PL, Le-Niculescu H, Dainton HD, Jain N et al.

Understanding and predicting suicidality using a combined genomic and clinical

risk assessment approach. Mol Psychiatry 2015; 20(11): 1266-1285.

128 53. Pandey GN, Rizavi HS, Zhang H, Bhaumik R, Ren X. The Expression of the

Suicide-Associated Gene SKA2 Is Decreased in the Prefrontal Cortex of Suicide

Victims but Not of Nonsuicidal Patients. International Journal of

Neuropsychopharmacology 2016.

54. Sadeh N, Spielberg JM, Logue MW, Wolf EJ, Smith AK, Lusk J et al. SKA2

methylation is associated with decreased prefrontal cortical thickness and greater

PTSD severity among trauma-exposed veterans. Mol Psychiatry 2016; 21(3):

357-363.

55. Boks MP, Rutten BPF, Geuze E, Houtepen LC, Vermetten E, Kaminsky Z et al.

SKA2 Methylation is Involved in Cortisol Stress Reactivity and Predicts the

Development of Post-Traumatic Stress Disorder (PTSD) After Military

Deployment. Neuropsychopharmacology 2016; 41(5): 1350-1356.

56. Hanisch A, Silljé HHW, Nigg EA. Timely anaphase onset requires a novel

spindle and kinetochore complex comprising Ska1 and Ska2. The EMBO Journal

2006; 25(23): 5504-5515.

57. Gaitanos TN, Santamaria A, Jeyaprakash AA, Wang B, Conti E, Nigg EA. Stable

kinetochore–microtubule interactions depend on the Ska complex and its new

component Ska3/C13Orf3. The EMBO Journal 2009; 28(10): 1442-1452.

58. Jeyaprakash AA, Santamaria A, Jayachandran U, Chan Ying W, Benda C, Nigg

Erich A et al. Structural and Functional Organization of the Ska Complex, a Key

Component of the Kinetochore-Microtubule Interface. Molecular Cell 2012;

46(3): 274-286.

129 59. Rice L, Waters CE, Eccles J, Garside H, Sommer P, Kay P et al. Identification

and functional analysis of SKA2 interaction with the glucocorticoid receptor.

Journal of Endocrinology 2008; 198(3): 499-509.

60. Bruun GH, Doktor TK, Borch-Jensen J, Masuda A, Krainer AR, Ohno K et al.

Global identification of hnRNP A1 binding sites for SSO-based splicing

modulation. BMC Biology 2016; 14(1): 1-19.

61. Zhang Q-H, Qi S-T, Wang Z-B, Yang C-R, Wei Y-C, Chen L et al. Localization

and function of the Ska complex during mouse oocyte meiotic maturation. Cell

Cycle 2012; 11(5): 909-916.

62. Bian E-B, Ma C-C, He X-J, Wang C, Zong G, Wang H-L et al. Epigenetic

modification of miR-141 regulates SKA2 by an endogenous ‘sponge’ HOTAIR in

glioma. Oncotarget; Vol 7, No 21 2016.

63. Shi W, Gerster K, Alajez NM, Tsang J, Waldron L, Pintilie M et al. MicroRNA-

301 Mediates Proliferation and Invasion in Human Breast Cancer. Cancer

Research 2011; 71(8): 2926-2937.

64. Cao G, Huang B, Liu Z, Zhang J, Xu H, Xia W et al. Intronic miR-301 feedback

regulates its host gene, ska2, in A549 cells by targeting MEOX2 to affect

ERK/CREB pathways. Biochemical and Biophysical Research Communications

2010; 396(4): 978-982.

65. Zhuang H, Meng X, Li Y, Wang X, Huang S, Liu K et al. Cyclic AMP responsive

element-binding protein promotes renal cell carcinoma proliferation probably via

the expression of spindle and kinetochore-associated protein 2. Oncotarget; Vol 7,

No 13 2016.

130 66. Gonsalves Caryn S, Li C, Malik P, Tahara Stanley M, Kalra Vijay K. Peroxisome

proliferator-activated receptor-α-mediated transcription of miR-301a and miR-

454 and their host gene SKA2 regulates endothelin-1 and PAI-1 expression in

sickle cell disease. Bioscience Reports 2015; 35(6).

67. Suicide Data [online]. World Health Organization. 2016, [June 30,

2016]. Available from

URL: http://www.who.int/mental_health/prevention/suicide/suicideprevent/en/.

68. Niculescu AB, Levey D, Le-Niculescu H, Niculescu E, Kurian SM, Salomon D.

Psychiatric blood biomarkers: avoiding jumping to premature negative or positive

conclusions. Mol Psychiatry 2015; 20: 286-288.

69. Pandey GN, Rizavi HS, Zhang H, Bhaumik R, Ren X. The Expression of the

Suicide-Associated Gene SKA2 Is Decreased in the Prefrontal Cortex of Suicide

Victims but Not of Nonsuicidal Patients. International Journal of

Neuropsychopharmacology 2016: 1-10.

70. Rice L, Charlotte E., Eccles J, Garside H, Sommer P, Kay P, Blackhall FH et al.

Identification and functional analysis of SKA2 interaction with the glucocorticoid

receptor. Journal of Endocrinology 2008; 198: 499-509.

71. Boks MP, Rutten BP, Geuze E, Houtepen LC, Vermetten E, Kaminsky Z et al.

SKA2 Methylation is Involved in Cortisol Stress Reactivity and Predicts the

Development of Posttraumatic Stress Disorder (PTSD) After Military

Deployment. Neuropsychopharmacology 2015; 41(5): 1350-1356.

72. Guintivano J, Brown T, Newcomer A, Jones M, Cox O, Maher BS et al.

Identification and Replication of a Combined Epigenetic and Genetic Biomarker

131 Predicting Suicide and Suicidal Behaviors. The American journal of psychiatry

2014.

73. Chung KC, Springer I, Kogler L, Turetsky B, Freiherr J, Derntl B. The influence

of androstadienone during psychosocial stress is modulated by gender, trait

anxiety and subjective stress: An fMRI study. Psychoneuroendocrinology 2016;

68: 126-139.

74. Levey DF, Niculescu EM, Le-Niculescu H, Dainton HL, Phalen PL, Ladd TB et

al. Towards understanding and predicting suicidality in women: biomarkers and

clinical risk assessment. Mol Psychiatry 2016; 21(6): 768-785.

75. Guintivano J, Kaminsky ZA. Role of epigenetic factors in the development of

mental illness throughout life. Neurosci Res 2014.

76. Kimmel M, Clive M, Gispen F, Guintivano J, Brown T, Cox O et al. Oxytocin

Receptor DNA Methylation in Postpartum Depression.

Psychoneuroendocrinology 2016; 69: 150-160.

77. Osborne L, Clive M, Kimmel M, Gispen F, Guintivano J, Brown T et al.

Replication of Epigenetic Postpartum Depression Biomarkers and Variation with

Hormone Levels. Neuropsychopharmacology 2016; 41(6): 1648-1658.

78. Reimand J, Kull M, Peterson H, Hansen J, Vilo J. g:Profiler--a web-based toolset

for functional profiling of gene lists from large-scale experiments. Nucleic Acids

Research 2007; 35(Web Server issue): W193-W200.

79. Reimand J, Arak T, Vilo J. g:Profiler--a web server for functional interpretation

of gene lists (2011 update). Nucleic Acids Research 2011.

132 80. Houtepen LC, Vinkers CH, Carrillo-Roa T, Hiemstra M, van Lier PA, Meeus W

et al. Genome-wide DNA methylation levels and altered cortisol stress reactivity

following childhood trauma in humans. Nature communications 2016; 7: 10967.

81. Boks MP, Rutten BP, Geuze E, Houtepen LC, Vermetten E, Kaminsky Z et al.

SKA2 Methylation is Involved in Cortisol Stress Reactivity and Predicts the

Development of Posttraumatic Stress Disorder (PTSD) After Military

Deployment. Neuropsychopharmacology 2015.

82. Rosenblat JD, Cha DS, Mansur RB, McIntyre RS. Inflamed moods: A review of

the interactions between inflammation and mood disorders. Progress in Neuro-

Psychopharmacology and Biological Psychiatry 2014; 53: 23-34.

83. Naninck EFG, Hoeijmakers L, Kakava-Georgiadou N, Meesters A, Lazic SE,

Lucassen PJ et al. Chronic early life stress alters developmental and adult

neurogenesis and impairs cognitive function in mice. Hippocampus 2015; 25(3):

309-328.

84. Bockmühl Y, Patchev AV, Madejska A, Hoffmann A, Sousa JC, Sousa N et al.

Methylation at the CpG island shore region upregulates Nr3c1 promoter activity

after early-life stress. Epigenetics 2015; 10(3): 247-257.

85. Hsiao Y-M, Tsai T-C, Lin Y-T, Chen C-C, Huang C-C, Hsu K-S. Early life stress

dampens stress responsiveness in adolescence: Evaluation of neuroendocrine

reactivity and coping behavior. Psychoneuroendocrinology 2016; 67: 86-99.

86. Baes CW, Martins CMS, Tofoli SMC, Juruena MF. Early life stress in depressive

patients: HPA axis response to GR and MR agonist. Frontiers in Psychiatry 2014;

5.

133 87. Lindqvist D, Janelidze S, Hagell P, Erhardt S, Samuelsson M, Lennart M et al.

Interleukin-6 is Elevated in the Cerebrospinal Fluid of Suicide Attempters and

Related to Symptom Severity. Biological Psychiatry 2009; 2009(3): 287-292.

88. Tonelli LH, Stiller J, Rujescu D, Giegling I, Schneider B, Maurer K et al.

Elevated cytokine expression in the orbitoprefrontal cortex of victims of suicide.

Acta Psychiatrica Scandinavica 2008; 117(3): 198-206.

89. Pandey GN, Rizavi HS, Ren X, Fareen J, Hoppensteadt DA, Roberts RC et al.

Proinflammatory cytokines in the prefrontal cortex of teenage suicide victims.

Journal of Psychiatric Research 2012; 46(1): 57-63.

90. Gill J, Lee H, Barr T, Baxter T, Heinzelmann M, Rak H et al. Lower health

related quality of life in U.S. military personnel is associated with service-related

disorders and inflammation. Psychiatry Research 2014; 216(1): 116-122.

91. O’Donovan A, Chao LL, Paulson J, Samuelson KW, Shigenaga JK, Grunfeld C et

al. Altered inflammatory activity associated with reduced hippocampal volume

and more severe posttraumatic stress symptoms in Gulf War veterans.

Psychoneuroendocrinology 2015; 51: 557-566.

92. Heath NM, Chesney SA, Gerhart JI, Goldsmith RE, Luborsky JL, Stevens NR et

al. Interpersonal violence, PTSD, and inflammation: Potential psychogenic

pathways to higher C-reactive protein levels. Cytokine 2013; 63(2): 172-178.

93. Michopoulos V, Rothbaum AO, Jovanovic T, Almli LM, Bradley B, Rothbaum

BO et al. Association of CRP Genetic Variation and CRP Level With Elevated

PTSD Symptoms and Physiological Responses in a Civilian Population With

High Levels of Trauma. American Journal of Psychiatry 2014; 172(4): 353-362.

134 94. Eraly SA, Nievergelt CM, Maihofer AX, et al. ASsessment of plasma c-reactive

protein as a biomarker of posttraumatic stress disorder risk. JAMA Psychiatry

2014; 71(4): 423-431.

95. Plantinga L, Bremner JD, Miller AH, Jones DP, Veledar E, Goldberg J et al.

Association between posttraumatic stress disorder and inflammation: A twin

study. Brain, Behavior, and Immunity 2013; 30: 125-132.

96. Gladkevich A, Kauffman HF, Korf J. Lymphocytes as a neural probe: potential

for studying psychiatric disorders. Progress in Neuro-Psychopharmacology

& Biological Psychiatry 2004; 28: 559-576.

97. Tabares-Seisdedos R, Rubenstein J. Chromosome 8p as a potential hub for

developmental neuropsychiatric disorders: implications for schizophrenia, autism

and cancer. Molecular Psychiatry 2009; 14: 563-589

98. Das S, Ongusaha PP, Yang YS, Park J-M, Aaronson SA, Lee SW. Discoidin

Domain Receptor 1 Receptor Tyrosine Kinase Induces Cyclooxygenase-2 and

Promotes Chemoresistance through Nuclear Factor-κB Pathway Activation.

Cancer Research 2006; 66(16): 8123-8130.

99. Guerrot D, Kerroch M, Placier S, Vandermeersch S, Trivin C, Mael-Ainin M et al.

Discoidin Domain Receptor 1 Is a Major Mediator of Inflammation and Fibrosis

in Obstructive Nephropathy. The American Journal of Pathology 2011; 179(1):

83-91.

100. Stadtmann A, Block H, Volmering S, Abram C, Sohlbach C, Boras M et al.

Cross-Talk between Shp1 and PIPKIgamma Controls Leukocyte Recruitment.

Journal of immunology 2015; 195(3): 1152-1161.

135 101. Watson NB, Schneider KM, Massa PT. SHP-1-dependent macrophage

differentiation exacerbates virus-induced myositis. Journal of immunology 2015;

194(6): 2796-2809.

102. Bousquet C, Susini C, Melmed S. Inhibitory roles for SHP-1 and SOCS-3

following pituitary proopiomelanocortin induction by leukemia inhibitory factor.

The Journal of clinical investigation 1999; 104(9): 1277-1285.

103. Lodish H. Molecular cell biology. 5 edn. W. H. Freeman and Co: New York 2003.

104. de Kloet CS, Vermetten E, Geuze E, Kavelaars A, Heijnen CJ, Westenberg HGM.

Assessment of HPA-axis function in posttraumatic stress disorder:

Pharmacological and non-pharmacological challenge tests, a review. Journal of

Psychiatric Research 2006; 40(6): 550-567.

105. van Zuiden M, Kavelaars A, Vermetten E, Olff M, Geuze E, Heijnen C. Pre-

deployment differences in glucocorticoid sensitivity of leukocytes in soldiers

developing symptoms of PTSD, depression or fatigue persist after return from

military deployment. Psychoneuroendocrinology 2015; 51: 513-524.

106. Papale LA, Li S, Madrid A, Zhang Q, Chen L, Chopra P et al. Sex-specific

hippocampal 5-hydroxymethylcytosine is disrupted in response to acute stress.

Neurobiology of disease 2016; 96: 54-66.

107. Li S, Papale LA, Zhang Q, Madrid A, Chen L, Chopra P et al. Genome-wide

alterations in hippocampal 5-hydroxymethylcytosine links plasticity genes to

acute stress. Neurobiology of disease 2016; 86: 99-108.

136 108. Dong E, Gavin DP, Chen Y, Davis J. Upregulation of TET1 and downregulation

of APOBEC3A and APOBEC3C in the parietal cortex of psychotic patients.

Translational psychiatry 2012; 2: e159.

109. Tseng PT, Lin PY, Lee Y, Hung CF, Lung FW, Chen CS et al. Age-associated

decrease in global DNA methylation in patients with major depression.

Neuropsychiatric disease and treatment 2014; 10: 2105-2114.

110. Kang HJ, Kim SY, Bae KY, Kim SW, Shin IS, Yoon JS et al. Comorbidity of

depression with physical disorders: research and clinical implications. Chonnam

medical journal 2015; 51(1): 8-18.

111. Kapfhammer HP. [Comorbid depressive and anxiety disorders in patients with

cancer]. Der Nervenarzt 2015; 86(3): 291-292, 294-298, 300-291.

112. Edgar R, Domrachev M, Lash AE. Gene Expression Omnibus: NCBI gene

expression and hybridization array data repository. Nucleic Acids Research 2002;

30(1): 207-210.

113. Gillespie CF, Bradley B, Mercer K, Smith AK, Conneely K, Gapen M et al.

Trauma exposure and stress-related disorders in inner city primary care patients.

Gen Hosp Psychiatry 2009; 31(6): 505-514.

114. Ressler KJ, Mercer KB, Bradley B, Jovanovic T, Mahan A, Kerley K et al. Post-

traumatic stress disorder is associated with PACAP and the PAC1 receptor.

Nature 2011; 470(7335): 492-497.

115. Binder EB, Bradley RG, Liu W, Epstein MP, Deveau TC, Mercer KB et al.

Association of FKBP5 polymorphisms and childhood abuse with risk of

137 posttraumatic stress disorder symptoms in adults. JAMA 2008; 299(11): 1291-

1305.

116. Guintivano J, Arad M, Gould TD, Payne JL, Kaminsky ZA. Antenatal prediction

of postpartum depression with blood DNA methylation biomarkers. Molecular

Psychiatry 2014; 19: 560-567.

117. Aryee MJ, Jaffe AE, Corrada-Bravo H, Ladd-Acosta C, Feinberg AP, Hansen KD

et al. Minfi: a flexible and comprehensive Bioconductor package for the analysis

of Infinium DNA methylation microarrays. Bioinformatics 2014; 30(10): 1363-

1369.

118. Schalkwyk LC, Pidsley R, Wong CCY, Touleimat N, Defrance M, Teschendorff

A et al. wateRmelon: Illumina 450 methylation array normalization and metrics.

119. Pidsley R, CC YW, Volta M, Lunnon K, Mill J, Schalkwyk LC. A data-driven

approach to preprocessing Illumina 450K methylation array data. BMC Genomics

2013; 14: 293.

120. Dinarello CA. Proinflammatory Cytokines. Chest 2000; 118(2): 503-508.

121. Chen Y-a, Lemire M, Choufani S, Butcher DT, Grafodatskaya D, Zanke BW et al.

Discovery of cross-reactive probes and polymorphic CpGs in the Illumina

Infinium HumanMethylation450 microarray. Epigenetics 2013; 8(2): 203-209.

122. Gross J, Ligges U. nortest: Tests for Normality. R package version 1.0-4 edn2015.

123. Houseman EA, Accomando WP, Koestler DC, Christensen BC, Marsit CJ,

Nelson HH et al. DNA methylation arrays as surrogate measures of cell mixture

distribution. BMC Bioinformatics 2012; 13(86).

138 124. Pandey GN, Rizavi HS, Zhang H, Bhaumik R, Ren X. The Expression of the

Suicide-Associated Gene SKA2 is Decreased in the Prefrontal Cortex of Suicide

Victims, but Not of Non-Suicidal Patients. Int J Neuropsychopharmacol 2016.

125. Mann JJ, Currier D, Stanley B, Oquendo MA, Amsel LV, Ellis SP. Can biological

tests assist prediction of suicide in mood disorders? International Journal of

Neuropsychopharmacology 2006; 9(4): 465-474.

126. Mann JJ, Arango VA, Avenevoli S, Brent DA, Champagne FA, Clayton P et al.

Candidate endophenotypes for genetic studies of suicidal behavior. Biol

Psychiatry 2009; 65(7): 556-563.

127. Lindqvist D, Isaksson A, Lil Träskman B, Brundin L. Salivary cortisol and

suicidal behavior--A follow-up study. Psychoneuroendocrinology 2008; 33(8):

1061-1068.

128. Plattner F, Hayashi K, Hernandez A, Benavides DR, Tassin TC, Tan C et al. The

role of ventral striatal cAMP signaling in stress-induced behaviors. Nat Neurosci

2015; 18(8): 1094-1100.

129. Delerive P, De Bosscher K, Besnard S, Vanden Berghe W, Peters JM, Gonzalez

FJ et al. Peroxisome Proliferator-activated Receptor α Negatively Regulates the

Vascular Inflammatory Gene Response by Negative Cross-talk with Transcription

Factors NF-κB and AP-1. Journal of Biological Chemistry 1999; 274(45): 32048-

32054.

130. Lawrence T. The Nuclear Factor NF-κB Pathway in Inflammation. Cold Spring

Harbor Perspectives in Biology 2009; 1(6): a001651.

139 131. Clive ML, Boks MP, Vinkers CH, Osborne LM, Payne JL, Ressler KJ et al.

Discovery and replication of a peripheral tissue DNA methylation biosignature to

augment a suicide prediction model. Clinical Epigenetics 2016; 8(1): 113.

132. Nyakas C, Buwald B, Luiten PGM. Hypoxia and brain development. Progress in

Neurobiology 1996; 49(1): 1-51.

133. Palazon A, Goldrath Ananda W, Nizet V, Johnson Randall S. HIF Transcription

Factors, Inflammation, and Immunity. Immunity; 41(4): 518-528.

134. Shweiki D, Neeman M, Itin A, Keshet E. Induction of vascular endothelial growth

factor expression by hypoxia and by glucose deficiency in multicell spheroids:

implications for tumor angiogenesis. Proceedings of the National Academy of

Sciences of the United States of America 1995; 92(3): 768-772.

135. Loboda A, Jozkowicz A, Dulak J. HIF-1 and HIF-2 Transcription Factors -

Similar but Not Identical. Mol Cells 2010; 29(5): 435-442.

136. Fagerberg L, Hallström BM, Oksvold P, Kampf C, Djureinovic D, Odeberg J et al.

Analysis of the Human Tissue-specific Expression by Genome-wide Integration

of Transcriptomics and Antibody-based Proteomics. Molecular & Cellular

Proteomics 2014; 13(2): 397-406.

137. Stolze I, Berchner-Pfannschmidt U, Freitag P, Wotzlaw C, Rössler J, Frede S et al.

Hypoxia-inducible erythropoietin gene expression in human neuroblastoma cells.

Blood 2002; 100(7): 2623.

138. Isung J, Aeinehband S, Mobarrez F, Martensson B, Nordstrom P, Asberg M et al.

Low vascular endothelial growth factor and interleukin-8 in cerebrospinal fluid of

suicide attempters. Transl Psychiatry 2012; 2: e196.

140 139. Fan JM, Chen XQ, Jin H, Du JZ. Gestational hypoxia alone or combined with

restraint sensitizes the hypothalamic–pituitary–adrenal axis and induces anxiety-

like behavior in adult male rat offspring. Neuroscience 2009; 159(4): 1363-1373.

140. Edwards LJ, McMillen IC. Impact of Maternal Undernutrition During the

Periconceptional Period, Fetal Number, and Fetal Sex on the Development of the

Hypothalamo-Pituitary Adrenal Axis in Sheep During Late Gestation1. Biology of

Reproduction 2002; 66(5): 1562-1569.

141. Sareen J, Cox BJ, Afifi TO, et al. Anxiety disorders and risk for suicidal ideation

and suicide attempts: A population-based longitudinal study of adults. Archives of

General Psychiatry 2005; 62(11): 1249-1257.

142. Weidemann A, Johnson RS. Biology of HIF-1[alpha]. Cell Death Differ 2008;

15(4): 621-627.

143. Pourcelot E, Lénon M, Mobilia N, Cahn J-Y, Arnaud J, Fanchon E et al. Iron for

proliferation of cell lines and hematopoietic progenitors: Nailing down the

intracellular functional iron concentration. Biochimica et Biophysica Acta (BBA) -

Molecular Cell Research 2015; 1853(7): 1596-1605.

144. Rybka J, Kędziora-Kornatowska K, Banaś-Leżańska P, Majsterek I, Carvalho LA,

Cattaneo A et al. Interplay between the pro-oxidant and antioxidant systems and

proinflammatory cytokine levels, in relation to iron metabolism and the erythron

in depression. Free Radical Biology and Medicine 2013; 63: 187-194.

145. Sermeus A, Michiels C. Reciprocal influence of the p53 and the hypoxic

pathways. Cell Death and Dis 2011; 2: e164.

141 146. Culver C, Sundqvist A, Mudie S, Melvin A, Xirodimas D, Rocha S. Mechanism

of Hypoxia-Induced NF-κB. Molecular and Cellular Biology 2010; 30(20): 4901-

4921.

147. Kim T-H, Hur E-g, Kang S-J, Kim J-A, Thapa D, Lee YM et al. NRF2 Blockade

Suppresses Colon Tumor Angiogenesis by Inhibiting Hypoxia-Induced Activation

of HIF-1α. Cancer Research 2011; 71(6): 2260.

148. Wouters BG, Koritzinsky M. Hypoxia signalling through mTOR and the unfolded

protein response in cancer. Nat Rev Cancer 2008; 8(11): 851-864.

149. Rosenbloom KR, Sloan CA, Malladi VS, Dreszer TR, Learned K, Kirkup VM et

al. ENCODE Data in the UCSC Genome Browser: year 5 update. Nucleic Acids

Research 2013; 41(D1): D56-D63.

150. Yun H, Lee M, Kim S-S, Ha J. Glucose Deprivation Increases mRNA Stability of

Vascular Endothelial Growth Factor through Activation of AMP-activated Protein

Kinase in DU145 Prostate Carcinoma. Journal of Biological Chemistry 2005;

280(11): 9963-9972.

151. Xu J, Ji J, Yan X-H. Cross-Talk between AMPK and mTOR in Regulating

Energy Balance. Critical Reviews in Food Science and Nutrition 2012; 52(5):

373-381.

152. Vandesompele J, De Preter K, Pattyn F, Poppe B, Van Roy N, De Paepe A et al.

Accurate normalization of real-time quantitative RT-PCR data by geometric

averaging of multiple internal control genes. Genome Biology 2002; 3(7):

research0034.0031.

142 153. Lu S, Gao W, Huang M, Li L, Xu Y. In search of the HPA axis activity in

unipolar depression patients with childhood trauma: Combined cortisol

awakening response and dexamethasone suppression test. Journal of Psychiatric

Research 2016; 78: 24-30.

154. Jarcho MR, Slavich GM, Tylova-Stein H, Wolkowitz OM, Burke HM.

Dysregulated diurnal cortisol pattern is associated with glucocorticoid resistance

in women with major depressive disorder. Biological psychology 2013; 93(1):

150-158.

155. Yehuda R, Southwick S, Krystal J, Bremner D, Charney D, Mason J. Enhanced

suppression of cortisol following dexamethasone administration in posttraumatic

stress disorder. American Journal of Psychiatry 1993; 150(1): 83-86.

156. Hoge EA, Brandstetter K, Moshier S, Pollack MH, Wong KK, Simon NM. Broad

spectrum of cytokine abnormalities in panic disorder and posttraumatic stress

disorder. Depression and Anxiety 2009; 26(5): 447-455.

157. Gill JM, Saligan L, Woods S, Page G. PTSD is associated with an excess of

inflammatory immune activities. Perspectives in Psychiatric Care 2009; 45(4):

262-277.

158. Miller GE, Chen E. Harsh family climate in early life presages the emergence of

pro-inflammatory phenotype in adolescence. Psychological science 2010; 21(6):

848-856.

159. Slopen N, Lewis TT, Gruenewald TL, Mujahid MS, Ryff CD, Albert MA et al.

Early life Adversity and Inflammation in African Americans and Whites in the

143 Midlife in the United States Survey. Psychosomatic medicine 2010; 72(7): 694-

701.

160. Kiecolt-Glaser JK, Gouin J-P, Weng N-p, Malarkey WB, Beversdorf DQ, Glaser

R. Childhood Adversity Heightens the Impact of Later-Life Caregiving Stress on

Telomere Length and Inflammation. Psychosomatic medicine 2011; 73(1): 16-22.

161. Danese A, Pariante CM, Caspi A, Taylor A, Poulton R. Childhood maltreatment

predicts adult inflammation in a life-course study. Proceedings of the National

Academy of Sciences of the United States of America 2007; 104(4): 1319-1324.

162. Slopen N, Kubzansky LD, McLaughlin KA, Koenen KC. Childhood adversity

and inflammatory processes in youth: A prospective study.

Psychoneuroendocrinology 2013; 38(2): 188-200.

163. Dowlati Y, Herrmann N, Swardfager W, Liu H, Sham L, Reim EK et al. A Meta-

Analysis of Cytokines in Major Depression. Biological Psychiatry 2010; 67(5):

446-457.

164. Hiles SA, Baker AL, de Malmanche T, Attia J. A meta-analysis of differences in

IL-6 and IL-10 between people with and without depression: Exploring the causes

of heterogeneity. Brain, Behavior, and Immunity 2012; 26(7): 1180-1188.

165. Howren M, Lamkin D, Suls J. Associations of depression with C-reactive protein,

IL-1, and IL-6: a meta-analysis. Psychosom Med 2009; 71(2): 171-186.

166. Slavich GM, Irwin MR. From Stress to Inflammation and Major Depressive

Disorder: A Social Signal Transduction Theory of Depression. Psychological

bulletin 2014; 140(3): 774-815.

144 167. Bai Y-M, Su T-P, Tsai S-J, Wen-Fei C, Li C-T, Pei-Chi T et al. Comparison of

inflammatory cytokine levels among type I/type II and

manic/hypomanic/euthymic/depressive states of bipolar disorder. Journal of

Affective Disorders 2014; 166: 187-192.

168. Ortiz-Domínguez A, Hernández ME, Berlanga C, Gutiérrez-Mora D, Moreno J,

Heinze G et al. Immune variations in bipolar disorder: phasic differences. Bipolar

Disorders 2007; 9(6): 596-602.

169. Kim Y-K, Jung H-G, Myint A-M, Kim H, Park S-H. Imbalance between pro-

inflammatory and anti-inflammatory cytokines in bipolar disorder. Journal of

Affective Disorders 2007; 104(1): 91-95.

170. Monteiro S, Roque S, de Sá-Calçada D, Sousa N, Correia-Neves M, Cerqueira JJ.

An Efficient Chronic Unpredictable Stress Protocol to Induce Stress-Related

Responses in C57BL/6 Mice. Frontiers in Psychiatry 2015; 6: 6.

171. Marin MT, Cruz FC, Planeta CS. Chronic restraint or variable stresses differently

affect the behavior, corticosterone secretion and body weight in rats. Physiology

& Behavior 2007; 90(1): 29-35.

172. Heidt T, Sager HB, Courties G, Dutta P, Iwamoto Y, Zaltsman A et al. Chronic

variable stress activates hematopoietic stem cells. Nat Med 2014; 20(7): 754-758.

173. Klempan TA, Rujescu D, Mérette C, Himmelman C, Sequeira A, Canetti L et al.

Profiling brain expression of the spermidine/spermine N1-acetyltransferase 1

(SAT1) gene in suicide. American Journal of Medical Genetics Part B:

Neuropsychiatric Genetics 2009; 150B(7): 934-943.

145 174. Zhang L, Su T-P, Choi K, Maree W, Li C-T, Chung M-Y et al. P11 (S100A10) as

a potential biomarker of psychiatric patients at risk of suicide. Journal of

Psychiatric Research 2011; 45(4): 435-441.

175. Le-Niculescu H, Levey DF, Ayalew M, Palmer L, Gavrin LM, Jain N et al.

Discovery and validation of blood biomarkers for suicidality. Mol Psychiatry

2013.

176. Griffiths JJ, Zarate CA, Rasimas JJ. Existing and Novel Biological Therapeutics

in Suicide Prevention. American journal of preventive medicine 2014; 47(3 0 2):

S195-S203.

177. De Kock M, Loix S, Lavand'homme P. Ketamine and Peripheral Inflammation.

CNS Neuroscience & Therapeutics 2013; 19(6): 403-410.

178. Gordijn MS, Rensen N, Gemke RJBJ, van Dalen EC, Rotteveel J, Kaspers GJL.

Hypothalamic-pituitary-adrenal (HPA) axis suppression after treatment with

glucocorticoid therapy for childhood acute lymphoblastic leukaemia. Cochrane

Database of Systematic Reviews 2015; (8).

179. Klengel T, Mehta D, Anacker C, Rex-Haffner M, Pruessner JC, Pariante CM et al.

Allele-specific FKBP5 DNA demethylation mediates gene-childhood trauma

interactions. Nat Neurosci 2013; 16(1): 33-41.

180. Splinter E, de Wit E, van de Werken HJG, Klous P, de Laat W. Determining long-

range chromatin interactions for selected genomic sites using 4C-seq technology:

From fixation to computation. Methods 2012; 58(3): 221-230.

146 Curriculum Vitae MAKENA L. CLIVE

4322 Flint Hill Dr. Apt. 104 · Owings Mills, MD 21117 [email protected] · 435.901.1895 www.linkedin.com/in/makena-clive

EDUCATION

Ph.D. in Environmental Health Sciences 2012 - 2017 Program in Molecular & Translational Toxicology Johns Hopkins Bloomberg School of Public Health, Baltimore, MD Advisors: Joseph P. Bressler, PhD & Zachary A. Kaminsky, PhD

B.S. in Biochemistry 2008 - 2012 Brigham Young University, Provo, UT  GPA of 3.6 out of 4.0  Completed courses in analytical, organic, and physical chemistry, and biochemistry, including laboratory classes, and instrumental analysis

PUBLICATIONS & PRESENTATIONS

Clive, M.L., Boks, M.P., Vinkers, C.H., Osborne, L.M., Payne, J.L., Ressler, K.J., Smith, A.K., et al. (2016). Discovery and Replication of a Peripheral Tissue DNA Methylation Biosignature to Augment a Suicide Prediction Model. Clinical Epigenetics, 2016; 8(1): 113.

Kimmel, M., Clive, M., Gispen, F., Guintivano, J., Brown, T., Cox, O., Beckmann, M.W., et al. (2016). Oxytocin Receptor DNA Methylation in Postpartum Depression. Psychoneuroendocrinology, 69, 150-160. doi: 10.1016/j.psyneuen.2016.04.008

Osborne, L., Clive, M., Kimmel, M., Gispen, F., Guintivano, J., Brown, T., Cox, O., Judy, J., Meilman, S., Braier, A., Beckmann, M.W., Kornhuber, J., Fasching, P.A., Goes, F., Payne, J.L., Binder, E.B., Kaminsky, Z. (2015). Replication of Epigenetic Postpartum Depression Biomarkers and Variation with Hormone Levels. Neuropsychopharmacology, 1-11. doi: 10.1038/npp.2015.333

Clive, M., Kaminsky, Z.A., Bressler, J.P. (2017). Hypoxia Regulates the Microtubule Associated Protein SKA2 in Neuroblastoma Cells. Poster presented at the Society of Toxicology Annual Meeting, Baltimore, MD.

Clive, M., Eaton, W. W., Payne, J. L., Wilcox, H. C., & Kaminsky, Z. A. (2015). Discovery of a DNA Methylation Stress Proxy for Use in Suicide Prediction Model. Presented at the International Summit on Suicide Research, New York, NY.

147

Clive, M. L., Sysa Shah, P., Johnston, N., Hargest, T., Bedja, D., Gabrielson, K. (2014). Epigenetic Mechanisms of Cadmium-Induced Placental Insufficiency. Poster presented at Society of Toxicology Annual Meeting, Phoenix, AZ.

Hinrichs (Clive), M. L., Cline, T., Hansen, J. C., Hansen, L. D., & Mayo, M. (2012). Thermal and photolytic degradation of promethazine hydrochloride solutions. Poster presented at the American Chemical Society National Meeting, San Diego, CA.

TECHNICAL SKILLS

Laboratory: Tissue culture, sterile technique, bacterial culture, plasmid manipulation, lentiviral preparation, creation of stable knockdown lines, DNA processing, RNA processing, bisulfite conversion, PCR optimization, PCR primer design, pyrosequencing (methylation and genotype analysis), pyrosequencing assay design, gene expression assay design, qRT-PCR, Sanger sequencing, gel electrophoresis, chromatin conformation capture (3C), chromatin immunoprecipitation (ChIP), Western blot, immunoprecipitation, mouse handling, mouse necropsy, ELISA. Computer: proficient in R, analysis of gene expression & methylation array data, basic statistics, Microsoft Office

RESEARCH & LABORATORY EXPERIENCE

Graduate Research Assistant April 2014 - May 2017 Labs of Drs. Zachary A. Kaminsky & Joseph P. Bressler Johns Hopkins University, Baltimore, MD  Analyze genome-wide methylation data to develop a DNA methylation biosignature that improved accuracy of a suicide prediction model from 80 to 88%  Design and validate PCR and pyrosequencing assays used to quantify DNA methylation and improve a suicide prediction model  Research the regulation and function of the SKA2 gene through tissue culture, creating lentiviral knockdown models, designing qRT-PCR assays, and analyzing gene expression

Graduate Research Assistant March 2013 - April 2014 Labs of Dr. Kathleen Gabrielson Johns Hopkins University, Baltimore, MD  Designed experiments to evaluate effect of cadmium exposure on placental and fetal epigenetics in a mouse model  Western blot, immunoprecipitation, and mouse handling techniques  Collaborated with ICP-MS technician to develop method to measure cadmium in mouse tissues

148 Graduate Research Assistant December 2012 - March 2013 Lab of Dr. DeLisa Fairweather Johns Hopkins University, Baltimore, MD  Evaluated effects of high fat diet in a mouse model of myocarditis by qRT-PCR, ELISA, and pathology in various tissues  Performed mouse necropsy and prepared H&E slides for pathology analysis

Graduate Research Assistant August 2012 - December 2012 Lab of Dr. James D. Yager Johns Hopkins University, Baltimore, MD  Cultured human breast cancer cells  Isolated nucleic acids, determined changes in gene expression using qRT-PCR  Analyzed metabolomics data to determine pathways activated by treatments

Undergraduate Research Assistant April 2010 - April 2012 Lab of Dr. Jaron C. Hansen Brigham Young University, Provo, UT  Developed a method for biogas quantization and analyzed gases produced by anaerobic digestion using developed method  Optimized calorimetric substrate analysis method  Prepared and analyzed data collected in experiments  Trained assistants on analytical methods specific to biogas quantization

Intern for Dustin Heslop, Laboratory Supervisor June 2009 - August 2009 Nutraceutical Corporation, Odgen, UT  Conducted research and performed data analysis in a laboratory setting  Learned basic operations of HPLC, FTIR, ICP-MS, and ICP-OES systems  Assisted in sample preparation for quality testing in microbiology laboratory

MENTORSHIP EXPERIENCE

Mentor for Student in the Center for Talented Youth Program Summers 2014-2016 Johns Hopkins University, Baltimore, MD  Trained high school students in basic laboratory techniques  Assisted students in planning and executing research projects

Teaching Assistant to Dr. Joseph P. Bressler Spring 2014 Principles of Environmental Health Course Johns Hopkins University, Baltimore, MD  Assisted course director in preparing course materials  Collaborated on writing and grading course exams

149 LEADERSHIP EXPERIENCE

President Spring 2015 - Spring 2016 Environmental Health Sciences Student Organization Johns Hopkins University, Baltimore, MD  Organized activities for ~50 students on a regular basis  Recruited 8 board members  Planned and led monthly board meetings  Identified and communicated with community partners  Created and distributed media related to activities

Vice President Spring 2014 - Spring 2015 Environmental Health Sciences Student Organization Johns Hopkins University, Baltimore, MD  Assisted President in planning meetings, events, and communicating with community partners

Vice President Spring 2011 - Spring 2012 Y Chem, Student Chapter of American Chemical Society Brigham Young University, Provo, UT  Planned educational enrichment activities and publicized club events  Organized and performed chemistry demonstrations for local elementary schools

Program Director, Adopt-a-Grandparent Program Spring 2011 - Spring 2012 Center for Service and Learning Brigham Young University, Provo, UT  Managed 200 volunteers  Established program at 10 assisted-living centers in the Utah Valley area  Recruited volunteers, organized activities, coordinated with community partners  Visited and assisted “adopted” grandma weekly

GRADUATE COURSEWORK

Molecular Biology  Molecular Biology and Genomics  Macromolecular Structure and Analysis  Pathways and Regulation  Cell Structure and Dynamics  Statistics for Laboratory Scientists  Statistics for Genomics

Toxicology  Public Health Toxicology  Fundamentals of Human Physiology  Xenobiotic Metabolism and Biomarker Development  Environmental Toxicological Pathology 150

Public Health  Principles of Environmental Health  Environmental Health: The Molecular Basis  Environmental Health In Neurological and Mental Disorders  Public Health Perspectives on Research

AWARDS

NIEHS T32 Training Grant 2012 - 2017 Molecular & Translations Toxicology Program Johns Hopkins Bloomberg School of Public Health, Baltimore, MD

Undergraduate Research Award 2011 Department of Chemistry & Biochemistry Brigham Young University, Provo, UT

Undergraduate Research Award 2010 Department of Chemistry & Biochemistry Brigham Young University, Provo, UT

Heritage Scholarship 2008 - 2010 Brigham Young University, Provo, UT

PROFESSIONAL SOCIETY MEMBERSHIPS

Society of Toxicology 2013 - Present

American Chemical Society 2011 - 2012

151