Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Genome-wide DNA methylation and imprinted gene analysis in babies conceived by ART Gooding, Luke David 2019

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
24-ubc_2019_november_gooding_luke.pdf [ 1.22MB ]
Metadata
JSON: 24-1.0384565.json
JSON-LD: 24-1.0384565-ld.json
RDF/XML (Pretty): 24-1.0384565-rdf.xml
RDF/JSON: 24-1.0384565-rdf.json
Turtle: 24-1.0384565-turtle.txt
N-Triples: 24-1.0384565-rdf-ntriples.txt
Original Record: 24-1.0384565-source.json
Full Text
24-1.0384565-fulltext.txt
Citation
24-1.0384565.ris

Full Text

GENOME-WIDE DNA METHYLATION AND IMPRINTED GENE ANALYSIS IN BABIES CONCEIVED BY ART by  LUKE DAVID GOODING  B.Sc., The University of British Columbia, 2014  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Reproductive and Developmental Sciences)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  October 2019  © Luke David Gooding, 2019  ii  The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, a thesis entitled:  Genome-wide DNA methylation and imprinted gene analysis in babies conceived by ART  submitted by Luke David Gooding in partial fulfillment of the requirements for the degree of Master of Science in Reproductive and Developmental Sciences  Examining Committee: Dr. Paul Yong Supervisor  Dr. Patrice Eydoux Supervisory Committee Member  Dr. Wan Lam Supervisory Committee Member Dr. Angela Devlin Additional Examiner   Additional Supervisory Committee Members: Dr. Dan Rurak Supervisory Committee Member iii  Abstract Assisted reproductive technologies (ARTs) are associated with a number of adverse pregnancy, neonatal, and long-term outcomes. One plausible explanation is that ART procedures may be causing alterations in epigenetic mechanisms, including the regulation of imprinted genes, that persist during development. Therefore, I investigated gene expression of the imprinted genes PLAGL1, CDKN1C, KCNQ1OT1, and H19 in cord blood from 24 IVF, 18 ICSI, 9 IUI, and 26 naturally conceived babies as well as genome-wide DNA methylation using Illumina’s EPIC methylation array in cord blood from 10 IVF, 9 ICSI, and 10 naturally conceived babies. All the samples were procured from healthy newborn singletons. No differences in the gene expression of the imprinted genes were observed across conception modes. The genome-wide DNA methylation analysis revealed an overall stability of DNA methylation following ART; however, a small number of CpG sites exhibited hypervariability in the ICSI (47 CpG sites) and the IVF (4 CpG sites) groups. Furthermore, the mCSEA method for detecting DMRs revealed 101 promoter associated DMRs in the ICSI group and 101 promoter associated DMRs in the IVF group. 35 DMRs overlapped between the ICSI and IVF groups, suggesting some regions may be susceptible to DNA methylation alterations following ART. Four imprinted gene DMRs were also found to overlap DMRs between conception modes. Overall this analysis revealed that a small number of genomic regions may be impacted by ART. These regions may be significant as genes associated with neurodevelopmental disorders, intrauterine programming of adult onset obesity, and male infertility were observed to be altered in both the IVF and ICSI groups. Although validation of these regions is required, this analysis provides support for ART impacting DNA methylation that persists to birth in genes related to adverse outcomes and mediating transmission of DNA methylation alterations from infertile parents to babies.  iv  Lay Summary Assisted reproductive technologies (ART) are associated with a small increase in the risk for adverse outcomes that may have some impacts on the long-term health of individuals born via these technologies. One plausible explanation for this increased risk of adverse outcomes following ART is that ART procedures may be affecting environmentally sensitive mechanisms that regulate genes, called epigenetics. I show that epigenetic mechanisms are mostly unaltered in babies born via ART. However, I found a small number of epigenetic alterations that may be linked to the increased risk of adverse outcomes seen in ART conceived individuals. It is possible that these alterations are caused, at least in part, by the infertility of the parents undergoing the ART procedures. These alterations, however, require validation with other measurement techniques before conclusions can be drawn. There is also the possibility that these alterations have other causes that were not considered in this analysis.  v  Preface Research in this thesis was approved by the University of British Columbia Children’s and Women’s Research Ethics Board (certificate number: H06- 03668) and was conducted in Dr. Sai Ma’s laboratory with input from Dr. Sai Ma, Dr. Paul Yong, Dr. Wan Lam, Dr. Patrice Eydoux, and Dr. Dan Rurak.  Sample collection and clinical information data base management were performed by Kate Watt, Samuel Schafer, Kenny Louie, Rebecca Vincent, and I. DNA extractions were performed by Kate Watt, Samuel Schafer, Rebecca Vincent, and I.  In chapter 2, the design of the analysis was proposed by Rebecca Vincent, Samuel Schafer, and I with input from Dr. Sai Ma. I performed all RNA extractions, cDNA library constructions, and qPCR experiments. I performed data processing, statistical analysis, and production of figures with assistance from Kenny Louie. In chapter 3, the design of the analysis was proposed by Kenny Louie and myself with input from Dr. Sai Ma. I performed the sample preparation and collaborated with the Genome Quebec Innovation Centre for EPIC array data procurement. I researched and employed the analytical pipeline with input from Kenny Louie. vi  Table of Contents Abstract ......................................................................................................................................... iii Lay Summary ............................................................................................................................... iv Preface .............................................................................................................................................v Table of Contents ......................................................................................................................... vi List of Tables ................................................................................................................................ xi List of Figures .............................................................................................................................. xii List of Abbreviations ................................................................................................................. xiii Acknowledgements .................................................................................................................... xvi Dedication .................................................................................................................................. xvii Chapter 1: INTRODUCTION ......................................................................................................1 1.1 Infertility ......................................................................................................................... 1 1.1.1 Lifestyle factors impacting infertility ......................................................................... 1 1.1.2 Male factor infertility .................................................................................................. 3 1.1.3 Female factor infertility .............................................................................................. 6 1.2 Assisted reproductive technologies................................................................................. 8 1.2.1 Intrauterine insemination ............................................................................................ 9 1.2.2 Ovulation induction .................................................................................................. 10 1.2.3 In vitro fertilization ................................................................................................... 10 1.2.4 Intracytoplasmic sperm injection .............................................................................. 11 1.3 Outcomes: assisted reproductive technologies ............................................................. 12 1.3.1 ART singleton outcomes........................................................................................... 13 1.3.2 Long-term outcomes ................................................................................................. 14 vii  1.4 Causes for poor outcomes in ART conceived babies ................................................... 17 1.4.1 Infertility ................................................................................................................... 17 1.4.2 ART procedures ........................................................................................................ 18 1.4.2.1 Mechanisms by which ART procedures may impact outcomes ....................... 20 1.5 Epigenetics .................................................................................................................... 23 1.5.1 DNA methylation ...................................................................................................... 24 1.5.1.1 Genomic regions regulated by DNA methylation ............................................ 25 1.5.1.2 DNA methyltransferases ................................................................................... 26 1.5.1.3 DNA demethylation .......................................................................................... 27 1.5.1.4 Environmental influences on DNA methylation ............................................... 27 1.5.1.5 DNA methylation and age ................................................................................ 28 1.5.2 Genomic imprinting .................................................................................................. 29 1.5.2.1 Imprinting disorders .......................................................................................... 31 1.5.3 Epigenetic reprogramming........................................................................................ 33 1.6 Epigenetics and ART .................................................................................................... 36 1.6.1 ART and imprinting disorders .................................................................................. 36 1.6.2 Cause of epigenetic alterations in ART .................................................................... 41 1.6.2.1 Ovarian stimulation ........................................................................................... 41 1.6.2.2 In vitro culture................................................................................................... 43 1.6.2.3 Male factor infertility ........................................................................................ 45 1.6.3 Imprinted gene alterations in non-imprinting disorder ART conceived babies ........ 47 1.6.3.1 Imprinted gene network .................................................................................... 50 1.6.4 Genome-wide DNA methylation in ART conceived babies..................................... 52 viii  1.7 Rational and hypotheses ............................................................................................... 55 1.7.1 Objectives ................................................................................................................. 57 Chapter 2: GENE EXPRESSION ANALYSIS OF PLAGL1 AND AN IMPRINTED GENE NETWORK IN CORD BLOOD FROM IVF, ICSI, AND IUI CONCEIVED BABIES COMPARED TO NC CONTROLS ...........................................................................................58 2.1 Introduction ................................................................................................................... 58 2.2 Methods......................................................................................................................... 61 2.2.1 Study participants...................................................................................................... 61 2.2.2 Sample preparation ................................................................................................... 62 2.2.3 Reverse transcription and cDNA library preparation ............................................... 62 2.2.4 Quantitative real-time polymerase chain reaction (qPCR) ....................................... 63 2.2.5 Statistical analysis ..................................................................................................... 64 2.3 Results ........................................................................................................................... 65 2.3.1 Clinical information .................................................................................................. 65 2.3.2 Gene expression analysis of imprinted genes in cord blood from ART babies ........ 66 2.3.3 Analysis of an imprinted gene network in cord blood from ART babies ................. 70 2.4 Discussion ..................................................................................................................... 72 Chapter 3: GENOME-WIDE DNA METHYLATION ANALYSIS IN CORD BLOOD FROM IVF AND ICSI CONCEIVED BABIES COMPARED TO NC CONTROLS ..........76 3.1 Introduction ................................................................................................................... 76 3.2 Methods......................................................................................................................... 78 3.2.1 Study participants...................................................................................................... 78 3.2.2 DNA extraction from umbilical cord blood .............................................................. 80 ix  3.2.3 Illumina Infinium MethylationEPIC BeadChip array analysis ................................. 80 3.2.4 Data preprocessing .................................................................................................... 80 3.2.5 Cell composition estimation ..................................................................................... 81 3.2.6 Confounding variable correction .............................................................................. 81 3.2.7 Differentially methylated position (DMP) analysis .................................................. 82 3.2.8 Differentially variable position (DVP) analysis ....................................................... 82 3.2.9 Differentially methylated region (DMR) analysis .................................................... 82 3.2.10 Metastable epiallele and imprinted gene analysis ................................................. 83 3.2.11 Gene enrichment analysis ..................................................................................... 84 3.2.12 Statistical analysis ................................................................................................. 84 3.3 Results ........................................................................................................................... 86 3.3.1 Clinical information .................................................................................................. 86 3.3.2 Differentially methylated position analysis .............................................................. 87 3.3.3 Differentially variable position analysis ................................................................... 88 3.3.4 Metastable epiallele and imprinted gene analysis of DVPs ...................................... 93 3.3.5 Differentially methylated region analysis ................................................................. 94 3.3.6 Metastable epiallele and imprinted gene analysis of DMRs ..................................... 97 3.3.7 Gene enrichment analysis ......................................................................................... 98 3.3.8 DNA methylation and expression of H19 ............................................................... 100 3.4 Discussion ................................................................................................................... 102 Chapter 4: CLOSING REMARKS AND FUTURE DIRECTIONS .....................................112 Bibliography ...............................................................................................................................118 Appendix .....................................................................................................................................140 x  A.1 Promoter associated DMRs in the ICSI and NC group comparison at an FDR less than 0.05...................................................................................................................................... 140 A.2 Promoter associated DMRs in the IVF and NC group comparison at an FDR less than 0.05...................................................................................................................................... 143 A.3 Overlapping DMRs between the DMRs in the ICSI and IVF groups at an FDR less than 0.05...................................................................................................................................... 146  xi  List of Tables Table 1.  2010 WHO 5th centile sperm parameters.. ....................................................................... 4 Table 2. Male infertility classifications.. ........................................................................................ 4 Table 3. Relative risk for adverse outcomes in ART singletons................................................... 14 Table 4. Summary of known imprinting disorders.. ..................................................................... 32 Table 5. Studies investigation imprinting disorders following ART. ........................................... 40 Table 6. Summarized maternal and pregnancy characteristics ..................................................... 66 Table 7. Gene expression mean log10 RQ values, standard deviations, and p-values ................. 67 Table 8. Summarized maternal, pregnancy characteristics and cell type proportions .................. 86 Table 9. Differential variable positions ........................................................................................ 90 Table 10. Top promoter DMRs ..................................................................................................... 95 Table 11. DMRs promoter, gene body, and CGI regions ............................................................. 97 Table 12. Promoter DMRs overlapping with imprinted gene DMR ............................................ 98 Table 13. Top DMRs in the nervous system development GO term (GO:0007399) .................... 99  xii  List of Figures Figure 1. Gene expression boxplot. .............................................................................................. 68 Figure 2. Correlation between KCNQ1OT1 and PLAGL1 expression in cord blood.. ................. 70 Figure 3. Correlation between H19 and PLAGL1 expression in cord blood.. .............................. 71 Figure 4. Correlation between CDKN1C and PLAGL1 expression in cord blood........................ 72 Figure 5. Analytical pipeline for EPIC array analysis. ................................................................. 85 Figure 6. Volcano plot of DNA methylation variance between the ICSI and NC groups. ........... 92 Figure 7. Volcano plot of DNA methylation variance between the IVF and NC groups. ............ 93 Figure 8. Correlation between H19 DNA methylation and expression ...................................... 102 xiii  List of Abbreviations 450k     Illumina Infinium Methylation450 BeadChip  5hmc     5-hydroxymethylcytosine 5-mc     5-methylcytosine ADAD1   adenosine deaminase domain containing 1 ADAMTS    ADAM metallopeptidase with thrombospondin  ADAMTS16    ADAM metallopeptidase with thrombospondin type 1 motif 16 ANOVA    analysis of variance ART     assisted reproductive technology AS     Angelman syndrome  ASD                                        autism spectrum disorders AZF     azoospermia factor  BH     Benjamini-Hochberg  BMI     body mass index  BPA     bisphenol A BWS     Beckwith-Wiedemann syndrome  C17orf97    chromosome 17 open reading frame 97 C9orf3                                     chromosome 9 open reading frame 3 CDKN1C    cyclin-dependent kinase inhibitor 1C  cDNA     complementary DNA  CFTR     cystic fibrosis transmembrane regulator  CGI     CpG island CI     confidence interval  CpG     cytosine-phosphate-guanine  Cq    quantification cycle ddCt     delta-delta Ct  DIRAS3    DIRAS family, GTP-binding RAS-like protein 3  DLK1     delta-like homolog 1  DMP     differentially methylated position DMR     differentially methylated regions  DNA     deoxyribonucleic acid  DNMT    DNA methyltransferases  DOHaD    Developmental Origins of Health and Disease DVP     differentially variable position E2     enhancer 2 eBayes    Empirical Bayesian  EDTA    ethylenediaminetetraacetic acid  Elp3     elongator acetyltransferase complex subunit 3 EPIC     Illumina Infinium MethylationEPIC BeadChip ES     enrichment score FDR     false discovery rate  FSH     follicle-stimulating hormone  GNAS     guanine nucleotide-binding protein alpha stimulating  GO     Gene Ontology  xiv  GRB10    growth factor receptor-bound protein 10  GSTM5    glutathione S-transferase mu 5 hCG     human chorionic gonadotropin HIV     human immunodeficiency virus HYMAI    hydatidiform mole associated and imprinted ICR     Imprinting control region  ICSI     intracytoplasmic sperm injection  IGF2     insulin-like growth factor 2  IGF2R    IGF2 receptor  IGN     imprinted gene network IUGR     intrauterine growth restriction  IUI     intrauterine insemination  IVF     in vitro fertilization  KCNQ1OT1    KCNQ1-overlapping transcript 1  LH     luteinizing hormone  LINEs    long interspersed transposable elements Lit1     mouse KCNQ1OT1 lncRNA    long non-coding RNA  LOS     large offspring syndrome  LTR     long terminal repeat MBD3    methyl-CpGs binding domain protein 3 MCCC1    methylcrotonoyl-CoA carboxylase 1 mCSEA    methylated CpG set enrichment analysis MCTS2   MCTS family member 2, pseudogene Me-CP2    methyl CpG-binding protein 2 MeDIP    methylated DNA immunoprecipitation MEG3    maternally expressed 3 MEST     mesoderm specific transcript  MGMT    O-6-methylguanine-DNA methyltransferase mRNA    messenger RNA  MRPL39    mitochondrial ribosomal protein L39 NC     natural conception  NES     normalized enrichment score NNAT     neuronatin NPY     neuropeptide Y NRG1     neuregulin 1 NTD     neural tube defect OR     odds ratio P57Kip2    mouse CDKN1C PCDH    protocadherin PCDHA    protocadherin alpha PCDHB    protocadherin beta PCDHB11    protocadherin beta 11 PCDHG    protocadherin gamma PCDHGA1   protocadherin gamma subfamily A, 1 xv  PCDHGA2    protocadherin gamma subfamily A, 2 PCDHGA3    protocadherin gamma subfamily A, 3 PCDHGA4    protocadherin gamma subfamily A, 4 PCDHGA5    protocadherin gamma subfamily A, 4 PCDHGA6    protocadherin gamma subfamily A, 6 PCDHGB1    protocadherin gamma subfamily B, 1 PCDHGB2    protocadherin gamma subfamily B, 2  PCDHGB3    protocadherin gamma subfamily B, 3 PCR     polymerase chain reaction  PEG10    paternally expressed gene 10  PEG3     paternally expressed 3 PGC     Primordial germ cells  PGS/PGD    pre-implantations genetic screening/diagnosis  PLAGL1    pleiomorphic adenoma gene-like 1  PRCP2    prolylcarboxypeptidase PWS     Prader-Willi syndrome  RNA     ribonucleic acid  RQ     relative quantitation  RR     relative risk RT     reverse transcriptase  RTL1     retrotransposon gag like 1 Rt-qPCR    reverse transcriptase real-time quantitative PCR RUNX     RUNX family transcription factor 3 SD     standard deviation  SDK1     sidekick cell adhesion molecule 1 SET    single embryo transfer SNORD52    small nucleolar RNA, C/D box 52 SNP     single nucleotide polymorphisms  SNRPN    small nuclear ribonucleoprotein-associated protein N  SRS     Silver-Russell syndrome TET     ten-eleven translocation  TNDM    transient neonatal diabetes mellitus  TNP1     transition protein 1 TRIM28    tripartite motif containing 28 TTP     time to pregnancy UBC     ubiquitin C UBE3A    ubiquitin protein ligase E3A  UPD     uniparental disomy WHO     World Health Organization  Zac1     mouse PLAGL1 ZFP57    zinc finger protein 57 homolog ZNF597    zinc finger protein 597  xvi  Acknowledgements I am truly grateful to many individuals for their support, guidance, patience, and love in the completion of this thesis. My sincerest gratitude to the faculty members, staff, and students at UBC and the BC Children’s Hospital Research Institute who make work in this field possible, foster a community of excellence, and contribute immensely to increasing human well-being. Thank-you Dr. Sai Ma for opening many doors for me and under whom I learned a great deal about research and life. I am indebted to my supervisor, Dr. Paul Yong, and the members of my supervisory committee, Dr. Dan Rurak, Dr. Patrice Eydoux, and Dr. Wan Lam, for their patience with my progress, insights into my research, and interest in my future. I offer my enduring appreciation for enabling me the opportunity to close a chapter of my life. Further thanks to my colleagues, fellow students, and former students especially Kenny Louie, Samuel Schafer, Richard Ng, Rebecca Vincent, Annie Ren, Kate Watt, Rowena Ho, and Paloma Stanar for their being invaluable sources of help, friendship, and good times.  I would finally like to extend a heartfelt thank-you to my family and loved ones – especially my mother, father, and Karla – for their abiding love, motivation, support, and understanding.   xvii  Dedication  To curiosity and the search for answers, both without and within.   We roar out in triumph Blind as we are To our collective impotence The conquering catalysts In these bags of flesh Have won naught but servitude To the God of Chaos Entropy, entropy, entropy Your insatiable thirst consumes us all Chained to your march ever onward We corrode in the wake This toil towards order Meaningless We are doomed to fail You win, you always do1  Chapter 1: INTRODUCTION 1.1 Infertility Infertility – defined as the inability to have children after a prolonged period (generally 12 months) – is a significant global health problem affecting an estimated 48.5 million couples worldwide and one in six couples in developed countries (Mascarenhas et al., 2012; Thoma et al., 2013). In Canada, 15.7% of couples not using contraceptives, where the female partner is between the age of 18 and 49, reported not falling pregnant in a 12 month period from 2009 to 2010 – an increase of 10% over the 1984 Canadian infertility rate of 5.4% based on a similar definition (Bushnik et al., 2012). Many western countries are facing similar upwards trends of infertility with rates that are expected to continue to increase in the coming years (Butler et al., 2003). The causes of infertility are many and varied, occurring in both men and women; however, the increasing infertility rates in western countries are likely largely attributed to the decision to delay having children as female age is a well-known risk factor for infertility (Gougeon et al., 2003). However, increasing sexually transmitted infection rates, rising obesity levels, and other lifestyle factors may also contribute to rising infertility levels (Anderson, Nisenblat et al., 2010). Generally, the cause of infertility is attributable to the female partner in 40% of the cases, the male partner in 20% of cases, both partners in 30% of the cases, and of unexplained etiology in 10% of cases (Seli, 2011; Patel et al., 2011).  1.1.1 Lifestyle factors impacting infertility The obesity epidemic – dramatically increasing rates for obesity, clinically defined as body mass index (BMI) equal or greater than 30, in developed countries all around the world, reaching levels as high as 60% of the population in some countries – may be significantly 2  impacting infertility rates (Flegal et al., 2012). Women who are overweight (BMI 25-29.99), obese, and very obese (BMI>35) have been found to have longer time to pregnancy measurements than normal weight women (BMI 18.5 – 24.9) (Seli, 2011). Furthermore, direct correlations have been shown between a higher BMI and poorer fertility prognosis (Rittenberg et al., 2011).  Smoking may be attributable to 13% of infertility cases (Seli, 2011). In a large population study of women (n=8,515), increasing number of cigarettes smoked per day is significantly correlated to increasing delay to conception (Hull et al., 2000). Smoking is also associated with advance time of menopause, increased risk of pregnancy complications, and decrease in sperm concentration and motility (Waylen et al., 2008; Seli, 2011). Infertility has also been linked to caffeine and alcohol consumption; however, data is conflicting with respect to the amount of alcohol and caffeine consumption needed to effect infertility (Anderson, Norman et al., 2010).  High caffeine intake, 500mg or 5 cups of coffee a day, in women has been linked to an increased time to pregnancy (Seli, 2011). A prospective study of 221 couples across six fertility centers found that increasing alcohol consumption by one additional alcoholic drink per day in women was associated with a 13% decrease in eggs aspirated, an increase in the risk of not achieving a pregnancy by 2.86 times, and a 2.21 times increase in the risk of miscarriage (Klonoff-Cohen et al., 2003). Similarly, one additional drink per day in men increased the risk of not achieving a live birth by 2.28 to 8.32 times (Klonoff-Cohen et al., 2003). However, the majority of literature does not support a relationship between moderate caffeine and alcohol intake with markers of male and female infertility (Curtis et al., 1997; Olsen et al., 1997; Jensen & Henriksen et al., 1998; Jensen & Hjollund et al., 1998; Klonoff-Cohen et al., 2002; Jensen & Gottschau et al., 2014; Jensen & Swan et al., 2014). 3  Interestingly a recent ongoing prospective cohort study of 171 men reported increased pregnancy rates in men with moderate alcohol consumption while decreased pregnancy rates in men with moderate caffeine consumption compared to men with low to no consumption following reproductive technologies (Karmon et al., 2017).   1.1.2 Male factor infertility Male infertility is the sole cause of infertility in 20% of infertile couples; however, around 30% of infertile couples exhibit both male and female infertility, thus, male infertility is involved in half of all infertile couples (Patel et al., 2011). Despite the high prevalence of male infertility – 10% to 15% of men in prime reproductive years are affected by infertility – approximately 50% of the cases are of unexplained causes (Esteves et al., 2011; Brandes et al., 2009). Investigation into the conditions affecting male fertility usually includes patient history, physical examination, and a semen analysis (Patel et al., 2011). Semen analyses measure semen volume, total sperm number in the semen, sperm concentration, vitality (% of sperm alive), progressive motility, total motility, and morphology (Cooper et al., 2010). In an attempt to define normal semen parameters to be used as reference for determining abnormal semen parameters, the World Health Organization (WHO) used semen parameter data in nearly 1900 fertile men to generate percentile distributions for the various parameters among a fertile male population (Table 1) (Cooper et al., 2010). The fifth centile cutoff was established as the lower cutoff limit for normal sperm parameters, making measurements outside of these cutoff values abnormal and indicative of aberrant sperm production (Table 1) (Cooper et al., 2010).    4  Table 1.  2010 WHO 5th centile sperm parameters. Table adapted from Cooper et al. (2010).   Classification terms for reduced sperm count, reduced sperm motility, increased abnormal sperm forms, all sperm variables abnormal, and no sperm in the semen are listed in Table 2.   Table 2. Male infertility classifications. Table adapted from Hirsh et al. (2003). Infertility Classification Semen parameters Oligozoospermia Reduced sperm counta Asthenozoospermia Reduced sperm motilitya Teratozoospermia Increased abnormal formsa Oligoasthenoteratozoospermia All sperm variables abnormala Azoospermia No sperm in semen aValues outside of 5th centile  However, a semen analysis is only a surrogate marker for infertility and does not provide robust information on male fertility potential or causes for infertility as evidenced by infertile men with normal semen parameters (15%) and the high (50%) rates of idiopathic male infertility (Patel et al., 2011; Esteves et al., 2011).  Semen parameters 5th centileSemen volume 1.5mLSperm count 15 Million/mLTotal sperm count 39 MillionTotal motility 40% motileProgressive motility 32%Vitality 58% aliveMorphology 4% normal forms5  The list of factors that can affect male infertility is long and varied with non-genetic and genetic components.  Non-genetic factors include developmental complications, diabetes, hypertension, obesity, genital tract obstruction, endocrinopathies, and varicoceles (Sharlip et al., 2002). Varicoceles – enlargement of the network of small veins around the spermatic cord – are found in 40% of infertile men compared to 15% in the general male population and are implicated in male infertility related to the increase in testicular temperature due to pooling of venous blood in the vein network around the spermatic cord (Patel et al., 2011). Genetic factors of male infertility include karyotype abnormalities, Y-chromosome microdeletions, and cystic fibrosis transmembrane regulator (CFTR) mutations. Approximately 6% of all infertile men have karyotype abnormalities – aberrant chromosomal numbers, deletions, inversions, and translocations – the most common of which is an extra X chromosome in XXY Klinefelter syndrome (Bourrouillou et al., 1992). Micro deletions in the azoospermic factor (AZF) loci of the Y-chromosome –AZFa, AZFb, and AZFc regions – which contain genes necessary for spermatogenesis, germ cycle regulation, and meiosis is responsible for 7% of infertility in men (Simoni et al., 2004). Deletions in AZFa and AZFb result in sterility, whereas sperm can be found in men with AZFc mutations, allowing for treatment via assisted reproduction; however, male offspring of men with Y-microdeletions will also carry these mutations and the associated subfertility phenotype (Bourrouillou et al., 1992). Mutations in the 190 kilo-base gene CFTR (encoding a chloride channel transmembrane protein) are common, with more than 1 in 25 people a carrier for a mutation (Cuppens et al., 2004). Cystic fibrosis disease occurs when both CFTR alleles that are inherited contain a mutation and is a multi-system disorder, affecting multiple organs (Ahmad et al., 2013). 98% of men with cystic fibrosis 6  are infertile due to congenital bilateral absence of the vas deferens which disconnects the sperm producing testes from the penis and results in obstructive azoospermia (Boyd et al., 2004).   1.1.3 Female factor infertility Female factor infertility can be broken down into the following etiologies: ovulation disorders (40%), fallopian tubal factors (30%), endometriosis (15%) – the presence of endometrial-like tissue outside of the uterus – and other factors including uterine/cervical abnormalities (15%) (Jose-Miller et al., 2007). Of the women of reproductive years with ovulation disorders, 85% are attributed to dysfunction of the hypothalamic-pituitary-ovarian axis (the endocrine axis responsible for the secretion of the hormones follicle stimulating hormone (FSH), luteinizing hormone (LH), estrogen, and progesterone which regulate ovulation and the menstrual cycle) caused most commonly by the endocrine disorder poly-cystic ovary syndrome (WHO group II) (Lindsay et al., 2015). 10% are attributed to hypogonadotropic hypogonadism, diminished functioning of the ovaries in response to impairment in the secretion of gonadotropin hormones FSH and LH by the pituitary gland, most commonly related to stress, exercise, and weight loss (WHO group I) (Lindsay et al., 2015). The remaining 5% of women with ovulation disorders can be attributed to premature ovarian failure, the loss of ovarian reserve of eggs before the age of 40, inducing menopause like symptoms (WHO group III) (Lindsay et al., 2015). Infertility attributed to fallopian tube factors are caused by scarring —previous surgeries or sexually transmitted infections— inflammatory conditions, endometriosis, and congenital anomalies that impede the descent of fertilized/un-fertilized oocytes into the uterus (Briceag et al., 2015).  7  Population studies have consistently noted that birth rates decline rapidly when women reach the age of 35 (Menken et al., 1986). Although the usefulness of these studies as a representation of a woman’s maximum fertility potential is questioned by non-reproductive factors such as desire to prevent pregnancy, coital frequency, and ageing partners not taken into account, it is still clear that infertility rates increase with age – rising from 11% at 34 to 33% at 40 and 87% at age 45 (Tietze et al., 1957).  As age does not seem to significantly affect the response of the uterine endometrium to hormonal stimulation and the ability of the uterus to maintain a pregnancy, age-associated infertility appears to be primarily caused by the progressive decrease in the number and quality of oocytes from fetal life until menopause (Liu et al., 2011). The number of oocytes in the ovary decreases over time (via apoptosis and programmed cell death)  from 6-7 million at 20 weeks gestation, 1-2 million at birth, 500,000 at puberty, to ~1000 during the onset of menopause at mean age 51 (Liu et al., 2011; Johnson et al., 2012). As the number of oocytes in the ovary declines, cycle shortening, menstrual irregularity, and infertility follows (Liu et al., 2011). Oocyte quality also decreases with age, with an increase in the number oocytes exhibiting aneuploidy (aberrant number of chromosomes) from 10% at ages less than 35 to 40% at age 43 (Pellestor et al., 2003). Due to the high frequency of aneuploidy in older women, pregnant women of advanced age (>35 years) have a higher risk of stillbirth with a relative risk (RR) of 1.83 for stillbirth in women aged 40+ compared to women aged 20-29 (Kenny et al., 2013).  Furthermore, older women are more likely to develop pathologies and lifestyle issues that impact infertility including endometriosis, tubal disease, and obesity (Johnson et al., 2012).  The above-mentioned aspects of maternal age on fertility are especially concerning in the face of the growing trend in Canada and other developed nations for childbearing to occur later 8  in women’s lives. According to Canadian census data, the average age of mothers at childbirth has increased from 27 in 1981 to 30.2 in 2011, the highest age to date (Milan, 2013). The average age at first birth in 2011 – 28.5 years – was also the oldest recoded age (Milan, 2013). As of 2010, more Canadian women aged 35-39 are giving birth than women aged 20 -24, with 19.2% of total Canadian births occurring to women aged 35 and older compared to 4.9% in 1981 (Milan, 2013).  Overall, infertility is a major health concern as it impacts quality of life in both men and women, significantly associated with impairments in marital relationships, sexual dis-satisfaction, and impaired psychological well-being (Nelson et al., 2008; Khademi et al., 2008; Tan et al., 2008; Chachamovich et al., 2010). If the trend towards delayed childbearing continues, developed nations can anticipate increased demand for reproductive assistance to overcome the challenges posed by infertility.   1.2 Assisted reproductive technologies Assisted reproductive technologies (ART) is the term used to encompass technologies that facilitate the fertilization of gametes with the goal of achieving a pregnancy, generally through the in vitro handling of gametes and/or embryos, and include in vitro fertilization (IVF), intra-cytoplasmic sperm injection (ICSI), and intrauterine insemination (IUI) procedures. Since Louise Brown’s birth in 1978, the first human conceived via ART, these reproductive technologies have been instrumental in the treatment of infertility, allowing female, male, and idiopathic infertile patients (even in severe cases) the ability to conceive biological children - revolutionizing the reproductive health field. With the before mentioned increasing infertility rates, the maturation of the ART science allowing for better pregnancy rates, and greater 9  acceptance among populations, ART births have dramatically increased world-wide to over 7 million total babies born (ESHRE, 2018). In Canada, the number of IVF and ICSI cycles has increased from around 9,000 in 2002 to over 27,000 in 2012 which resulted in around 8,000 infants born in 2012 – approximately 2% of all Canadian births for that year (Gunby, 2012).    1.2.1 Intrauterine insemination Intrauterine insemination is the simplest, least invasive, least expensive, and most commonly performed assisted conception method (Seli, 2011).  IUI involves the direct insemination of washed sperm into the uterus via a thin catheter entering the uterus through the cervix, a process that is timed with natural ovulation or ovarian stimulation (Seli, 2011). The procedure is most often used – and most effective – in treating idiopathic infertility, mild male infertility, and mechanical issues related to fertilization through coitus (Seli, 2011). However, it may be used to treat several infertilities given sperm of sufficient quality and a non-obstructed fallopian tube (Seli, 2011). In the case of male infertility, IUI is a first line treatment for oligozoospermia, asthenozoospermia, and teratozoospermia where sperm washing and direct injection into the uterus enhances sperm concentration, selects normal sperm, as well as decreases sperm migration work and sperm attrition (Seli, 2011). For female factor infertilities, IUI may be used to bypass immunologic, mucosal, and mechanical cervical barriers as well as ovulation timing issues. IUI may also be utilized in conjunction with ovulation induction and/or controlled ovarian hyper stimulation (COH) when ovulatory disorders are present (Seli, 2011).  10  1.2.2 Ovulation induction Ovulation induction is the stimulation of follicular growth and subsequent ovulation of oocytes from the ovaries and can be used to treat ovulatory disorders. Stimulation of follicular growth is achieved through administration of exogenous gonadotropin hormones, FSH and LH/ hCG (human chorionic gonadotropin, an analog of LH), or through reducing the negative feedback on the production of gonadotropin hormones by estradiol – thereby increasing the production of gonadotropins (Seli, 2011). The latter is achieved via administration of clomiphene citrate, an estrogen antagonist that binds to estrogen receptors and decreases the ability of circulating estradiol to negatively feedback on gonadotropin production, or via administration of aromatase inhibitors which prevent the production of estradiol (Casper, 2007). Administration of exogenous LH/hCG artificially recreates the LH surge that is required to facilitate ovulation (Seli, 2011). Controlled ovarian hyperstimulation or superovulation refers to techniques used to stimulate the growth and ovulation of multiple follicles/oocytes. The drugs and methods used are the same as mentioned above for ovulation induction; however, doses, and attenuation of the doses, of drugs administered are altered to stimulate the production of more follicles (Seli, 2011).   1.2.3 In vitro fertilization In vitro fertilization refers to the procedures that allow for the in vitro control of ovulating multiple oocytes, fertilization, and early embryogenesis before subsequent transfer of embryos into the uterus. As such IVF can be used to treat most male factor, female factor, and idiopathic infertilities (Seli, 2011). Controlled ovarian hyperstimulation is followed by oocyte-cumulus retrieval via aspiration through a trans-vaginal, ultrasound guided, long needle. Oocytes are then cultured with washed sperm, minimum 10,000 to 100,000 sperm, until fertilization. 11  Resulting embryos are cultured to either the day 3 cleavage stage or the day 5/6 blastocyst stage before being transferred into a primed uterus (Seli, 2011). Multiple embryos or a single embryo may be transferred at a time; however, those embryos not transferred fresh need to be frozen and cryopreserved before being thawed and transferred at a later date, if necessary, to achieve a pregnancy (Seli, 2011). Although still an area of intense research, the in vitro nature of IVF allows for the analysis of embryo quality through observational and molecular means, enabling the transferal of the embryos most likely to implant in the endometrium. Pre-implantation genetic screening/diagnosis (PGS/PGD) is the analysis of DNA obtained from early embryonic cells at the cleavage or blastocyst stage that allows for the selection of embryos with the greatest chance of implantation based on genetic information and screen embryos for a specific genetic disease (Huntington’s disease) (Seli, 2011).   1.2.4 Intracytoplasmic sperm injection ICSI follows the same protocol as described in IVF, the difference being, in ICSI, a single sperm is micromanipulated into each oocyte to facilitate fertilization. As such, men with low sperm count, under the necessary 10,000 for IVF, may still achieve fertilization and is used effectively to treat severe male infertilities. Cumulus cells are removed from oocytes prior to ICSI, with the denuded oocytes held steady by a holding pipette while a micropipette injects a single spermatozoon into the oocyte (Seli, 2011). In conjunction with microdissection testicular sperm extraction, where sperm is extracted from testicular tissue, ICSI enables the treatment of all but the most severe cases of male infertility (Esteves et al., 2011). Due to the relatively guaranteed success of fertilization with ICSI, currently, ICSI represents two-thirds of all treatments world-wide with IVF representing one-third (ESHRE, 2018).  12  1.3 Outcomes: assisted reproductive technologies Although ART have been instrumental in treating infertility and has exploded world-wide, these technologies were introduced with little evaluation of their effects on maternal and fetal health (Willem et al., 2005). The safety of these procedures is still unclear and necessitates further scrutiny as the number of ART children born continues to increase. The majority of babies born through ART are perceivably healthy; however, ART pregnancies are at a higher risk than naturally conceived (NC) pregnancies for a number of complications and adverse outcomes.  This increased risk is mainly due to the higher incidence of multiple pregnancies following ART (Qin et al., 2015). Due to the large number of embryos produced through IVF and ICSI, multiple embryos have historically been transferred per cycle in order to maximize the rate of implantation and pregnancy. However, this practice significantly increases the chance of multiple gestation pregnancies with most countries exhibiting a multiple birth rate of 20-30% following ART up until 2010 (Bhattacharya et al., 2014). Multiple pregnancies are at a significantly increased risk for a number of adverse maternal and neonatal outcomes over singleton pregnancies, with higher order multiple pregnancies (triplets, quadruplets, etc.) associated with greater risk (Bhattacharya et al., 2014). Twin, triplet and higher order multiple pregnancies have much higher rates of preeclampsia, gestational diabetes, neonatal intensive care admission, perinatal mortality, intracranial bleeding, respiratory distress syndrome, maternal mortality, pre-term delivery, and congenital malformations (Bhattacharya et al., 2014). As such, multiple pregnancies are a significant burden on the health of the mother and fetus as well as on the health care system.  For the above reasons, in recent years, fertility clinics in most countries have made efforts to lower the number of multiple pregnancies following ART by decreasing the number of 13  embryos transferred and adopting the use of single embryo transfer (SET) and frozen embryo cycles (Bhattacharya et al., 2014). Although still much higher than the 1% multiple pregnancy rate in the general population, Canadian ART multiple pregnancy rates have dropped from 34.8% in 2002 to 15.1% in 2012 (Gunby, 2012).   1.3.1 ART singleton outcomes With the advent of increased SET, researchers have become interested in determining the risk of adverse outcomes among singleton pregnancies conceived from ART compared to those conceived naturally. The results from recent cohort studies examining the association between ART singleton pregnancies and obstetric risk has been mixed, with some reporting similar outcomes to natural/spontaneous conceptions while most report worse outcomes (Qin et al., 2016). However, all six meta-analyses analyzing data from 1978-2012, collectively, have reported increased risk for a number of neonatal and maternal outcomes including preterm birth, very preterm birth, low birth weight, very low birth weight, and small for gestational age (Helmerhorst et al., 2004; Jackson et al.,  2004; McGovern et al., 2004; McDonald et al., 2005; Pandey et al., 2012). Results from a recent meta-analysis – analyzing 50 cohort studies involving 161,370 ART singletons and 2,280,241 NC singletons birthed between 1982-2012 across European, North American, and east Asian countries – revealed significantly higher, confidence interval (CI) of 95%, relative risk for 17 adverse pregnancy outcomes (Table 3) including preterm birth (1.71 RR; 1.67-1.83), very preterm birth (2.75; 2.62-2.88), low birth weight (1.69; 1.64-1.73), very low birth weight (RR 2.18; 2.06-2.30), small for gestational age (1.49; 1.44-1.54), and congenital malformations (1.32: 1.27-1.36) (Qin et al., 2016).   14  Table 3. Relative risk for adverse outcomes in ART singletons. Table adapted from Qin et al. (2016). Outcome Number of ART SC RR (95% CI) studies (n) singletons (n) singletons (n) Preterm birth 36 133,338 1,289,549 1.70(1.67-1.74) Very preterm birth 25 128,547 1,253,013 2.75(2.62-2.88) Low birth weight 36 130,147 1,062,445 1.69(1.64-1.73) Very low birth weight 30 127,088 980,322 2.18(2.06-2.30) Small for gestational age 14 81,090 753,771 1.49(1.33-1.54) Perinatal mortality 22 106,267 1,262,997 1.57(1.46-1.70) Congenital malformations 28 77,697 724,300 1.32(1.27-1.36)  Another recent metanalysis characterizing congenital malformations in 315,402 ART cases against 5,154,779 NC controls revealed significantly higher odds ratios for the ART group than the NC groups for cardiac abnormalities 1.43 (1.27-1.62), central nervous system abnormalities 1.36 (1.10-1.70), urogenital system abnormalities 1.58 (1.28-1.94), and musculoskeletal disorders 1.35 (1.12-1.64) (Hooersan et al., 2017).  ART singletons remain at a significantly higher relative risk for the above adverse outcomes in matched subgroup analyses for maternal age and parity, controlling for these confounders (Pandey et al., 2012). Interestingly, a meta-analysis of 3 cohort studies and 8 case-control studies found that the use of ART is associated with a higher percentage of autism spectrum disorders (ASD), (RR 1.35; 1.09-1.68, p-value = 0.007); however, the authors note that further large prospective studies are needed to confirm this risk (Gao et al., 2017).  1.3.2 Long-term outcomes The developmental origins of health and disease hypothesis (DOHaD), formerly known as the Barker hypothesis, was put forward by David Barker in 1990 to provide an explanation for the correlation between adverse neonatal outcomes – particularly low birth weight and premature 15  births – and coronary heart disease (Barker & Osmond, 1986). The hypothesis proposes a causal relationship between an infant not reaching its growth potential, as a result of adverse peri-conception and uterine environments, and later life onset of diseases such as coronary heart disease, hypertension, and type II diabetes (Hales et al., 1992). This proposal relies on the assumption that exposure to adverse environments at critical stages of development in uterine life will lead to adaptive change (Hart & Norman, 2013). If the post-natal conditions are different than the conditions in utero these changes may then become mal-adaptive and increase the risk of developing disease (Bateson et al., 2004). Analyses of the population born during and after the Dutch Winter Famine of 1944-1945, where nutrition fell well below 1000 calories for 6 months, revealed a doubled rate of coronary heart disease, increased type II diabetes, raised glucose levels, and other diseases that were dependent on the gestational period of caloric restriction while in utero (Rosenboom et al., 2011). These Dutch Winter Famine studies provide support for the DOHaD hypothesis and allow for insight into windows of sensitivity for specific developmental processes (Rosenboom et al., 2011). Due to increased risk of low birth weight, preterm births, and intra uterine growth restriction among singleton ART pregnancies – even when controlling for maternal age and parity – there is concern that ART conceived embryos/fetus may be exposed to adverse environments during stages of development that may lead to later life negative outcomes as proposed by the DOHaD hypothesis (Hart & Norman, 2013).  However, meaningful longer-term studies of ART populations are difficult due to the youthfulness of the population as well as difficulties in controlling for differences in the motivation of parents to enroll children, threshold of parents to seek medical attention for their children, parental age, and socioeconomic status between ART and NC groups (Hart & Norman, 16  2013).  The literature on the longer-term health outcomes of children conceived through ART is conflicting with some studies concluding there are minimal differences between the health of ART conceived children and those conceived naturally, while others have found differences in measurements associated with cardio-metabolism as well as in cancer rates (Williams & Sutcliffe, 2009; Wilson et al., 2011).  Cardio-metabolic investigations consist of measurements in cohorts involving less than 250 participants with mean ages ranging from 5.9 to 21.2 years of age. These investigations reveal evidence for raised blood pressure (Ceelen et al., 2008; Sakka et al., 2010), elevated fasting glucose levels (Ceelen et al., 2008), increase in total body fat composition (Ceelen et al., 2007), and potentially increased growth velocity in early life (Ceelen et al., 2009; Hart & Norman, 2013).  Although investigations into the incidence of cancer among ART children include much larger samples, population studies of up to 26,000 ART children, results are still mixed with some investigations reporting lower or insignificant relative risk of cancer while others have reported an increased overall cancer risk and risk for specific cancers (Moll et al., 2003; Källén et al., 2005a; Källén et al., 2010; Reigstad et al., 2016). A recent and large investigation analyzed cancer diagnoses in all children born between 1984 and 2011 in Norway (1,602,876 non-ART and 25,782 ART children) (Reigstad et al., 2016). Maternal age, birth order, gestational age, and birth weight were controlled in their statistical model. No significant difference in overall cancer risk was found between ART and non-ART children; however, increased risk for leukemia (1012 non ART cases, 17 ART cases; 1.67 hazard ratio) –acute myeloid leukemia (768, 9; 2.63) and other leukemias (71, 3; 5.12) – as well as Hodgkin’s lymphoma (258, 3; 3.63) was found in ART children (Reigstad et al., 2016).  17  Overall, children conceived by ART appear to be healthy. However, there is some evidence to suggest that ART children may be at an increased risk for cardiometabolic alterations and cancer. As the ART population ages, more studies will be necessary to better elucidate the risk ART poses to long term health.  1.4 Causes for poor outcomes in ART conceived babies The causal mechanisms involved in the association between poor outcomes and ART in singleton pregnancies and children is unclear. Possible explanations include the underlying infertility of patients undergoing ART, demographic factors of the ART population, and factors associated with the actual ART procedures themselves (Pinborg et al., 2013). It is imperative that the underlying mechanisms for the increased adverse outcomes in ART are elucidated as this will provide a better understanding of the safety of ART, better ability to council patients, and better understanding of what directions to take to increase their safety.  1.4.1 Infertility Most studies analyzing pregnancy outcomes in ART compare outcomes from infertile couples that underwent ART procedures to natural conception (NC) outcomes in pregnancies from fertile couples (Pinborg et al., 2013).  This has made it difficult to disentangle the effects of infertility per se and ART procedures per se on the adverse outcomes associated with ART.  Therefore, a number of studies have compared neonatal outcomes from pregnancies conceived spontaneously with a time to pregnancy of less than or equal to 1 year (TPP<1 year) – fertile couple – to neonatal outcomes from pregnancies conceived spontaneously with a time to pregnancy of greater than 1 year (TPP>1 year) – subfertile couple (Pinborg, et al. 2013). A meta-18  analysis on a number of these studies indicates that subfertile couples who conceive naturally after a prolonged time to pregnancy (TPP>1 year) have greater risk for preterm births (pooled odds ratio (OR): 1.3; 95% CI: 1.2-1.4), low birth weight (1.3; 1.2-1.5) and small for gestational age (1.2; 1.0-1.3) (Messarlian et al., 2012). Although the above meta-analysis provides strong evidence that infertility, and subfertility, affects neonatal outcomes, infertility has many varied underlying causes with potentially varied effects on neonatal outcomes (Hansen & Bower, 2014). Indeed, children born after ART for treating tubal factor infertility were 30% more likely to be born premature than children born after ART for treating male factor infertility (Kawwass et al., 2013). Furthermore, infertile couples that undergo ART may differ from subfertile couples that are able to conceive naturally in ways that could significantly impact the ability to extrapolate the risk subfertility poses for adverse neonatal outcomes onto the risk infertile couples would pose. Subfertile couples have also been found to have increased prevalence of obesity, smoking, and alcohol consumption when compared to ART conceiving couples – all of which are known to affect birth outcomes (Raatikainen et al., 2012). The literature supports an association between infertility and adverse outcomes observed after ART; however, the underlying cause of how infertility may impact outcomes and the degree to which it does on a case by case basis is still unclear.   1.4.2 ART procedures The complexity and variation in ART treatments across clinics and countries hinders the ability to elucidate the aspects of ART procedures that may be causing adverse neonatal outcomes (Hansen et al., 2014). Furthermore, the rapid development and change of ART procedures, combined with the lag in understanding of how ART procedures may be affecting 19  neonatal outcomes, hinders the ability of this research to effect change in the technology. However, a meta-analyses of six studies where ART singleton births were compared with non-ART singleton births to subfertile parents (TTP>1 year) showed increased odds ratios for pre-term birth (pooled OR: 1.6; 95%% CI: 1.3-1.8) – indicating that ART procedures may increase the risk for adverse neonatal outcomes above the risk attributed to subfertility (Pinborg et al., 2013). Furthermore, in an attempt to control for all parental factors that may be impacting neonatal outcomes, a meta-analysis of studies comparing neonatal outcomes between consecutive siblings conceived naturally and through ART to the same mother revealed an increased odds ratio (1.3; 1.1-1.5) for preterm birth among the siblings conceived through ART (Pinborg et al., 2013). It is unclear how much each component of the ART procedure is contributing to adverse outcomes; however, ovulation induction is consistently shown to affect outcomes. Ovulation induction is associated with increased risk of poor neonatal outcomes when compared to natural conceptions (OR: 1.4; 95% CI: 1.2-1.7) (Pinborg et al., 2013). Furthermore, risks for a number of adverse neonatal outcomes – pre-term birth, small for gestational age, and low birth weight – are shown to be lower in pregnancies conceived via frozen embryo transfers compared to fresh embryo transfer (Maheshwari et al., 2012; Pinborg et al., 2013). As fresh embryos are transferred after superovulation while frozen embryos are transferred in cycles without ovarian stimulation, lower risk in frozen embryo cycles may be due to the absence of the effects of ovarian stimulation. However, it is not clear what impact freezing of embryos may have on outcomes in the comparison above as frozen embryo transfer has been shown to increase the risk for large for gestational age in resulting pregnancies (Wennerholm et al., 2013). Studies of the effects of embryo culture on birth outcomes are limited by differences in media between fertility clinics, although one study found significant differences in mean birth weight between 20  pregnancies conceived from embryos grown on two different embryo culture media (Dumoulin et al., 2010). Recently, analysis of perinatal outcomes in a large Nordic population of ART singletons (n=62, 379) and twins (29, 758) born from 1988-2007 revealed a decline in the risk of being born preterm in ART singletons over this period, a decrease in the proportion of ART singletons born low birthweight, and a decline in the number of ART stillborns (Henningsen et al., 2015). This decrease in adverse outcomes over time is encouraging; however, the decline was most steep in the early years of ART use and has plateaued in recent years. A number of possible explanations are possible for this decline in adverse outcomes. Patient characteristics may be changing as ART use increases over time and includes more of the population, adoption of single embryo transfer has led to an increase in the number of frozen cycles likely accounting for some of the decrease in adverse outcomes over time, changes to laboratory conditions and milder ovarian stimulation protocols may also be contributing to this decrease (Pinborg et al., 2013).   1.4.2.1 Mechanisms by which ART procedures may impact outcomes Many factors are involved in the development of the embryo and in the establishment of a dialogue between the conceptus and maternal environment that support the growth of a healthy pregnancy. Assisted reproduction technologies introduce a number of variables –supraphysiological levels of hormones after super ovulation, in vitro fertilization, micromanipulation of sperm into a denuded oocyte in ICSI, and embryo culture – that may impact the developmental processes and trajectories of the resulting embryos and concepti. However, most of the literature on this topic focuses on the impact of embryo culture and ovarian stimulation in the mouse or bovine, limiting the application to humans (Van Montfoort et al., 2012). The fallopian tube and uterine environments, where cleavage and blastocyst 21  development of the embryo take place in vivo, are different from each other with respect to a number of different variables – nutrient availability, pH, growth factors, and oxygen tension – which change dynamically as the embryo moves down the fallopian tube (Feuer & Rinaudo, 2012). The difficulty of modeling these dynamic changes in embryo culture, most embryos are cultured to the blastocyst stage in a single culture media, may impact developmental processes in the embryo (Bloise et al., 2014). Indeed, changes to culture media that better recapitulate the in vivo environment have been documented to improve blastocyst success rates and pregnancy success rates (Hentemann et al., 2011). Embryos are able to progress from the fallopian tube to the uterus and develop to the blastocyst stage on embryo culture media that does not recapitulate the in vivo environment, due to a high metabolic plasticity (Leese, 2012).  This plasticity is thought to be achieved through flexibility in glycolytic activity, oxidative phosphorylation, and membrane transport – allowing embryos to survive stressful environments (Leese, 2012). Studies investigating the effects of IVF, ICSI, and embryo culture on gene expression in mouse blastocysts indicate that the majority of changes to gene expression are involved in metabolic processes (Rinaudo et al., 2006; Giritharan et al., 2012; Schwarzer et al., 2012). A study that investigated the impact of ovarian stimulation in bovine also revealed impairment of embryo development as well as global changes in the expression of genes involved in growth and proliferation (Gad et al., 2011). Murine and bovine studies demonstrate that ovulation stimulation may delay the timing of endometrial maturation and impact the duration of the implantation window – period when the uterus is receptive to implantation of the blastocyst (Simon et al., 2003: Papanikolaou et al., 2005). This may provide a possible explanation for the increased rates of pre-eclampsia (disease associated with abnormal placentation) and preterm births.  Furthermore, culture media composition has been shown to impact growth, cell fate, 22  post-implantation development, birth weights, and post-natal growth curves in mice (Banrezes et al., 2011). This continuation of alterations at each developmental level in response to changes in the nutrients of cell media suggests that a “memory” of the ART procedure is preserved (Bloise et al., 2014). Indeed, murine studies have observed culture-specific transcriptional signatures in placental gene expression, suggesting the impact of embryo culture on embryonic cells were “remembered” by differentiated placental villi cells (Fauque et al., 2010; Schwarzer et al., 2012). Cellular mechanisms that allow for environmental adaptations to be preserved and inherited mitotically are widely believed to be epigenetic in nature (Van Montfoort et al., 2012). Indeed, recent studies in humans have revealed that environments in infancy and early childhood (socioeconomic and psychosocial) are associated with molecular changes in epigenetic mechanisms that may impact genomic regulation and development (Essex et al., 2013; Boyce & Kobor, 2015; Needham et al., 2015; Esposito et al., 2016; McDade et al., 2017). Investigations into the impact of ART procedures on epigenetic mechanisms have revealed alterations in epigenetic regulation at a number of important genes (Van Montfoort et al., 2012).  Although knowledge of the exact cause for the increased risk of adverse outcomes in singleton pregnancies following assisted reproductive technologies is not yet elucidated, given the complexity of the underlying infertilities and ART procedures, it is clear that multiple avenues of research will be required to illuminate this consequential issue. However, the findings of alterations in epigenetic mechanisms in ART embryos and the ability for epigenetic mechanisms to serve as a cellular “memory” – allowing for changes in cellular processes to stably be passed on to future cellular generations – possibly linking ART procedures to neonatal/longer-term outcomes suggests that epigenetic mechanisms are a prescient avenue of research in this space (Waterland & Michels., 2007). 23  1.5 Epigenetics Epigenetics refers to the study of changes in gene function that are mitotically and/or meiotically heritable and cannot be explained by changes in the DNA sequence, leading to the continuation of heritable changes in phenotype (Riggs & Porter, 1996). As most cells in an organism share the same genetic code, epigenetic mechanisms allow for cellular differentiation whereby each cell has a particular lineage-specific epigenome that enables specific expression profiles for different cell types (Jenuwein, 2006). Regulation of gene function through epigenetic mechanism can either be static, as in X-inactivation where the X-chromosome is consistently inactivated or dynamic (Klar, 1998). Dynamic epigenetic regulation allows for an epigenetic memory – heritable change in gene expression or behavior that is induced by a previous stimulus (D’Urso & Brickner, 2014). Due to the stability of genetic information, dynamic epigenetic regulation is important for adaptation to environmental and developmental stimuli allowing for rapid and mitotically heritable responses. Indeed, transient – lasting at least 30 hours but less than 2 weeks – epigenetic alterations (primarily DNA methylation) have been found in response to short-term diesel exhaust exposure (Jiang et al., 2014; Clifford et al., 2016). DNA methylation alterations associated with prenatal maternal stressors such as maternal smoking during pregnancy and prenatal intimate partner violence have been observed to persist to at least late childhood demonstrating that DNA methylation alterations associated with environmental changes can also be long lasting (Radtke et al., 2011; Joubert et al., 2016). These epigenetic alterations associated with different environmental stimulus may play a role in mediating pathologies associated with these exposures. Epigenetic memory can occur through multiple means but usually requires chromatin-based changes, such as DNA methylation and histone modifications (D’Urso & Brickner, 2014).  24   Chromatin refers to the structure in which DNA is packaged inside the nucleus of the cell. The nucleosome is the fundamental unit of chromatin, made up of 147 base pairs of DNA wrapped around an octamer of histones that consists of two of each histone H3, H4, H2A, and H2B (Kouzarides, 2007). Nucleosomes are organized, with the help of a linker histone protein H1, into further higher order chromatin structures. The degree to which the nucleosomes are condensed dictates whether transcription can occur, with dense chromatin (heterochromatin) transcriptionally repressed and loose chromatin (euchromatin) transcriptionally active. The movement between euchromatin and heterochromatin is dynamically regulated by a network of epigenetic mechanisms, including DNA methylation and histone tail modifications (Van Montfoort et al., 2012). Histones tail modifications are many and include methylation, acetylation, phosphorylation, and ubiquitination which are added and removed by many enzymes (Kouzarides, 2007). Histone tail modification can be both repressive and activating by either interacting with the chromatin directly or via recruitment of transcription factors and chromatin remodeling proteins (Clapier & Cairns, 2009).  1.5.1 DNA methylation DNA methylation is the addition of a methyl group to the fifth carbon of cytosine, producing 5-methylcytosine (5mc) at cytosine-guanine dinucleotides (CpG sites) throughout the genome. There are 28 million CpG sites, the majority of which are methylated, occurring at a low frequency (1 per 100 base pairs) throughout the genome (Laird et al., 2004). These CpG sites are most commonly found in dense clusters that run for 500-2000 base pairs termed CpG islands which are often associated with promoters (Deaton & Bird, 2011). DNA methylation follows a bi-modal distribution where 60-90% of single, isolated, CpG sites are hypermethylated 25  (high methylation levels) while CpG islands are predominantly hypomethylated (low methylation levels) (Deaton & Bird, 2011). CpG islands are often associated with gene promoters for developmental or reference genes where their hypomethylated state is associated with transcription factor binding and gene expression (Weber et al., 2007). Although DNA methylation is generally thought of as a transcriptionally repressive marker, the density of CpG clusters and the length of these clusters are important in the relationship between DNA methylation and transcription (Meissner et al., 2008). The addition of methylation to high CpG density promoters, which are generally rarely methylated, is associated with efficient gene silencing – supporting DNA methylation as a repressive mechanism. Intermediate CpG density promoters are also inactive when methylated; however, at pluripotency gene promoters, DNA methylation can be present without changing transcription (potentially performing the role of a safety mechanism for silencing pluripotency genes in differentiated cells) (Weber et al., 2007; Meissner et al., 2008). Low CpG density promoters are generally hypermethylated and remain transcriptionally active independent of DNA methylation (Weber et al., 2007; Meissner et al., 2008). The mechanisms by which DNA methylation represses transcription are not entirely clear; however, DNA methylation is thought to act as a barrier to transcriptional machinery and can also recruit histone modification complexes such as the methyl CpG-binding protein 2 (Me-CP2) (Fuks et al., 2003).  1.5.1.1 Genomic regions regulated by DNA methylation DNA methylation has been shown to be important in controlling the transcription of specialized regions. Around 40% of the mammalian genome is composed of transposable elements (Lander et al. 2001). DNA methylation is evolutionarily conserved to suppress these 26  transposable elements, particularly long interspersed transposable elements (LINEs) and long terminal repeat (LTR) elements as they carry strong promoters (Messerschmidt et al., 2014). DNA methylation is also necessary in regulating the inactivation of the X-chromosome, occurring in mammalian female pre-implantation cells where two X-chromosomes are present (one X-chromosome is inactivated). X-chromosome inactivation is mediated by DNA methylation after a cascade of events triggered by the coating the X-chromosome with the non-coding RNA Xist (Messerschmidt et al., 2014). The final outcome of this cascade is methylation at the CpG sites of promoters; however, it appears that X-chromosome methylation may function as long-term repression insurance as loss of methylation on the X-chromosome has only moderate effects on gene expression from genes on the X-chromosome (Messerschmidt et al., 2014). DNA methylation is also vital for controlling the expression of a third group of specialized regions throughout the genome – imprinted genes (discussed below).  1.5.1.2 DNA methyltransferases  Methylation of DNA occurs through the activity of DNA methyltransferase enzymes – DNA methyltransferase 1 (DNMT1), DNMT2, DNMT3A, DNMT3B, and DNMT3L (Chen & Li, 2004). During DNA replication, DNA methylation is transmitted, semi-conservatively, from the methylated parent strand to the un-methylated daughter strand (in the hemi-methylated double stranded DNA helix) via DNAMT1 (Hermanm et al., 2004). Therefore, DNAMT1 serves as a DNA maintenance machine. DNMT3A and DNMT3B establish de-novo DNA methylation patterns, specifically during stages of gametogenesis and early development in a sequence specific fashion (Okano et al., 1999). DNMT3A has been shown to be necessary for establishing DNA methylation at maternally imprinted and paternally imprinted gene in gametes (Kaneda et 27  al., 2004). Deletion of DMT3B causes embryonic lethality in mice, indicating the importance of methylation establishment to proper embryogenesis (Li et al., 1992). DNMT3L is molecularly similar to DNMT3A and DNMT3B, however, DNMT3L has no enzymatic activity and is thought to act as a co-factor to DNMT3A (Jia et al., 2007).  1.5.1.3 DNA demethylation  Loss of DNA methylation is not as straightforwardly enzymatically controlled as methylation with most demethylation occurring through passive, replication-dependent dilution, or indirect active demethylation (Wu et al., 2014). Passive demethylation is achieved by down regulation of DNA methylation maintenance machinery (DNMT1) in actively dividing cells and can only happen in a global fashion (Messerschmidt et al., 2014). The indirect active demethylation occurs with the use of ten-eleven translocation (TET) enzymes that can oxidize 5-methylcytosine to 5-hydroxymethylcytsine (5hmc) which further enzymes can then remove (Ito et al., 2010).   1.5.1.4 Environmental influences on DNA methylation  DNA methylation is dependent on the network of biochemical reactions involved in one-carbon metabolism that allow for the availability of methyl groups. As such, nutrients central to one-carbon metabolism – folate (vitamin B9), vitamins B2, B6, B12 – are important contributors to maintaining DNA methylation as well as enabling de novo DNA methylation (Michels, 2012). Additional environmental exposures shown to be associated with DNA methylation alterations are alcohol/ethanol exposure, tobacco use, endocrine disruptors such as bisphenol A (BPA), benzene, radiation, arsenic, heavy metals, air pollution, and asbestos (Michels, 2012). However, 28  the mechanisms underlying the changes in DNA methylation are as yet not entirely clear and may be context specific with respect to age, genotype, and exposure type (Michels, 2012). One mechanism hypothesized is increased reactive oxygen species as a result of an inflammatory response which can oxidize 5-mc to 5-hmc, not maintained by DNMT1, and result in altered DNA methylation profiles (Valinluck & Sowers, 2007). Due to the environmental influence on epigenetic signatures and the potential of these signatures to impact physiology and pathology, the field epigenetic epidemiology has emerged to study the associations between epigenetic variation and the risk of disease in humans with the potential to identify factors that create disease-specific epigenetic patterns in the hopes of providing new avenues of treatment and prevention (Michels, 2012). As such, there has been inroads in identifying epigenetic patterns associated with numerous pathologies including cancer, infectious diseases (helicobacter pylori, hepatitis viruses, epstein-barr virus, papilloma virus), rheumatoid arthritis, asthma, type I diabetes, metabolic disorders (type II diabetes, obesity), and psychiatric disorders (depression, suicide, schizophrenia, bipolar disorder, Alzheimer’s disease) (Michels, 2012).   1.5.1.5 DNA methylation and age  Interestingly, though perhaps unsurprisingly, DNA methylation has been shown to change with age. Although DNA methylation dynamics across the genome with respect to age are not entirely mapped, it is clear that DNA methylation levels increase (for the most part) from birth to adulthood with the first year of life showing increased changes than subsequent years (Martino et al., 2011). Both the aging process itself and the accumulation of environmental influences are thought to impact DNA methylation profiles during aging. The divergence of methylation profiles between monozygotic twins after the first year of life and more similar 29  DNA methylation profiles between younger twins than older twins illustrated the ability of the environment to shape the epigenome (Fraga et al., 2005; Martino et al., 2013). Recently, using DNA methylation array datasets, Horvath created an algorithm using methylation levels at 353 CpG sites across the genome to predict age (with a Pearson correlation of 0.96 between DNA methylation age and chronological age) (Horvath, 2013). This age predictor can be used across multiple tissues and cell types and is proposed to measure the cumulative effect of an epigenetic maintenance system where perturbations in the epigenetic maintenance system would accelerate DNA methylation age (although the details are not known). The Horvath Clock, as it has become to be known as, has been proposed to measure biological age as evidenced by offspring of semi-supercentenarians (individuals that reach ages between 105-109 years) having a younger DNA methylation age than age matched controls (Horvath et al., 2015). Furthermore, centenarians are younger based on the Horvath Clock (DNA methylation age) than would be expected based on their chronological age (Horvath et al., 2015).  Indeed, further work by Horvath and others has shown that DNA methylation age is accelerated with respect to chronological age in disease phenotypes such as cancer (Horvath, 2013), obesity and metabolic syndrome (Horvath et al., 2014), Down syndrome (Horvath et al., 2015), Alzheimer’s disease (Levine et al., 2015), Huntington’s disease (Horvath et al., 2016), and HIV infection (Horvath & Levine, 2015).   1.5.2 Genomic imprinting Genomic imprinting is an epigenetic process that enables genes to be expressed in a parent-of-origin specific manner (Reik & Walter, 2001). Genes expressed in a parent-of-origin fashion are termed imprinted genes, consisting of both maternally expressed imprinted genes and paternally expressed imprinted genes. Genomic imprinting was first identified in pronuclear 30  transfer experiments where it was discovered that both maternal and paternal genomes are required for normal fetal development, later it was understood that the factors necessary for development was the parent-specific expression of imprinted genes. In these experiments, pronuclei were transplanted between one-cell-stage mouse embryos, creating diploid embryos with either two female pronuclei (biparental gynogenones) or two male pronuclei (biparental androgenones) (McGrath & Solter, 1984). Both the biparental gynogenones and biparental androgenones were non-viable; however, the phenotypes of the resulting abortuses were different (McGrath & Solter, 1984). The biparental gynogenones exhibited underdeveloped and retarded extraembryonic tissues while the biparental androgenones displayed underdeveloped and retarded embryonic tissues (McGrath & Solter, 1984). This study highlighted the crucial role parental-specific genes play in the development of both the placenta and the fetus. Knock out studies of imprinted genes have revealed that, in general, paternally expressed imprinted genes enhance fetal growth while maternal expressed imprinted genes restrict fetal growth (Reik et al., 2003). Keeping in line with these observations, the kinship theory, a widely cited theory, was proposed as a means to explain the origin of genomic imprinting (Moore & Haig, 1991). The kinship theory proposes that there is a conflict between the interests of the paternal and maternal genes in a fetus when it is reliant upon resources from the mother (Moore & Haig, 1991). As allocating more maternal resources to the fetus may allow for better fitness of the offspring, it is optimal for the paternal genes to maximize the acquisition of maternal resources; however, it is optimal for maternal genes to be sparing in demands for maternal resources to better the chances of producing further offspring (Moore & Haig, 1991).  Theories for the evolution of imprinting, however, remain contested and it is possible that no one theory can explain all the observations associated with imprinted genes (Peters, 2014). Another widely 31  cited theory, the coadaptive theory, proposes that imprinted genes act coadaptively in optimizing fetal development and maternal care – explaining the role of imprinted genes in maternal care behavior in relation to their role in intra-uterine growth (Keverne & Curley, 2008).  Imprinted genes are usually organized in clusters with 80% of the known murine imprinted genes clustering together (Reik & Walter, 2001). Parent-specific expression of multiple imprinted genes in a cluster is determined by a cis acting imprinted control region (ICR) (Barlow, 2011). These ICRs exhibit parental-specific DNA methylation levels, where one parental allele is hypermethylated and one parental allele hypomethylated (Barlow, 2011). As such these regions may be termed imprinted differentially methylated regions (DMR). Most ICRs acquire methylation in the female germ line during oogenesis and male ICRs acquire methylation in the male germline (Barlow, 2011). ICRs that are unmethylated are active whereas ICRs that are methylated are inactive, controlling transcription of imprinted genes under its command in opposite fashions often through long non-coding RNA silencing of a region in cis (Peters, 2014).  As imprinted genes have been shown to be important in diverse areas, including fetal and placental growth, post-natal feeding, regulation of metabolism, and behaviours that optimize maternal care, altered expression of imprinted genes contributes to a wide range of diseases – imprinting disorders, intrauterine growth restriction, obesity, diabetes, psychiatric disorders and cancer (Peters, 2014).   1.5.2.1 Imprinting disorders As imprinted genes are controlled by DNA methylation and other epigenetic mechanisms at ICRs, both genetic mechanisms and epigenetic mechanisms can result in altered expression of 32  imprinted genes.  There are three main causes for alterations in imprinted gene expression (Peters, 2014). 1) Deletion of an imprinted gene or of the regulatory region controlling the imprinted gene. 2) The presence of both sets of chromosomes, or part of a chromosome, from only one parent and none from the other (uniparental disomy (UPD)) results in loss of both sets of parental-specific control of imprinted genes on that chromosome and altered expression. 3) Alteration of DNA methylation and other epigenetic mechanisms at imprinted gene clusters (ICRs) can alter the expression of one or more genes in this region (Peters, 2014). Alteration in expression of imprinted genes has been shown to cause a number of disorders – imprinting syndromes – relating to development, growth, and cancer. A list of human imprinted syndromes, their major features, and causes is in Table 4.  Table 4. Summary of known imprinting disorders. Table adapted from Peters (2014). Imprinting disorder Clinical characteristics Causes Location Angelman Syndrome (AS) Delayed development, microcephaly, intellectual disability, and happy behavioural profile (Buiting, 2010) Loss of maternally expressed UBE3A. UBE3A mutation paternal UPD15. Epigenetic defects not common 15q11-3 Beckwith-Wiedemann Syndrome (BWS) Overgrowth, tongue enlargement, defects of abdominal wall, embryonal tumors (Peters, 2014) Loss of maternal expression CDKN1C  Gain of maternal expression IGF2 Loss of maternal expression H19. Epigenetic defects common 11p15.5 Maternal UPD14 Growth restriction, obesity, precocious puberty (de Rocha et al., 2008) Loss of paternal expression of RTL1 and DLK1 14q32 33  Paternal UPD14 Dysmorphism, defects of abdominal wall, increased amniotic fluid and placenta size (Kogami et al., 2012) Increased expression of RTL1 14q32 Prader-Willi Syndrome (PWS) Distinctive features, increased appetite, delay in development, obesity, poor sexual development, cognitive impairment (Buiting, 2010) Loss of paternal expression of many genes Paternal deletion of chromosome 15q11-13 genes Maternal UPD15. Epigenetic defects rare 15q11-13 Pseudohypoparathyroidism type 1a Dysmorphism, cognitive issues, obesity and resistance to parathyroid hormone (Kelsey, 2010) Maternal inactivating mutations of GNAS, Loss of imprinting of GNAS. 20q13 Pseudohypoparathyroidism type 1b Resistance to parathyroid hormone (Kelsey, 2010) Loss of maternal GNAS methylation and loss of expression of imprinted GNAS 20q13 Silver-Russell Syndrome (SRS) Dysmorphism, IUGR, learning disabilities and delated, stunted growth (Peters, 2014) Hypomethylation at DMR of H19. Maternal UPD7 11p15.5 Transient Neonatal Diabetes Mellitus (TNDM) IUGR and high neonatal blood sugar levels (Mackay & Temple, 2010) Increased expression of PLAGL1 and HYMAI 6q24   1.5.3 Epigenetic reprogramming In contrast to the transmission of DNA methylation to daughter cells through mitosis, the inheritance of DNA methylation and other epigenetic information between generations is 34  generally prevented (Van Montfoort et al., 2012). This is likely due to the need for germ cells and embryos to acquire totipotency, the ability to differentiate into any cell, which necessitates the removal of differentiating epigenetic marks (Van Montfoort et al., 2012). As epigenetic mechanisms are dynamic and may change with response to stressors, the prevention of inheritance of epigenetic marks through meiosis may also serve to protect the next generation from a potentially harmful parental epigenetic burden (Lange & Schneider, 2010). The genome-wide removal of these marks occurs through two phases. The first phase of genome-wide DNA methylation (and other epigenetic marks) removal begins at the primordial germ cell (PGC) stage and ends after PGCs enter into the gonads (Van Montfoort et al., 2012). The second phase of CpG demethylation occurs after the fusion of sperm and oocyte in the resulting zygote (the first cleavage divisions) (Van Montfoort et al., 2012). As there is limited research done on human primordial germ cells and embryos with respect to epigenetic reprogramming, due to ethical considerations, most of the knowledge on DNA demethylation and subsequent reprogramming comes from studies on mice. As such, the following summarization of DNA demethylation and subsequent reprogramming will be from studies in the mouse. For reference, human PGCs form during week 2 of development, reach the gonadal ridge at the end of week 5, female germ cells begin meiosis at week 10 (finishing meiosis just before ovulation postnatally), and male germ cells enter meiosis postnatally (Allegrucci et al., 2005).  In the mouse, the first phase of genome-wide demethylation, beginning with germ cell development from epiblast cells and continuing until 3 days after the PGCs have reached the gonads, is thought to be an active process that results in less than 10% of CpG sites retaining a methyl mark (Seki et al., 2005; Popp et al. 2010). Demethylation occurs across most sequence elements, where, importantly, sex-specific imprinted marks are erased and reset (Hajkova et al., 35  2008). As PCGs continue into gametogenesis, re-methylation of the genome begins to take place, including the sex-specific re-methylation of imprinted regions according to the sex of the fetus (Shovlin et al., 2006). Complete re-establishment of sex-specific DNA methylation at imprinted genes in the male germline occurs by birth; however, re-establishment of sex-specific DNA methylation of imprinted genes in the female germline is not fully completed until just prior to the ovulation of each oocyte (Allegrucci et al., 2005).    The second phase of removing epigenetic marks and demethylation of the genome takes place after sperm and egg fuse in the resulting zygote (Van Montfoort et al., 2012). Paternal DNA is actively de-methylated, facilitated through active conversion of 5-methylcytosine CpG sites to 5-hydroxymethylcytosine; however, maternal DNA is passively de-methylated until the morula stage due to decrease in DNMT1 activity and takes longer to reach lower DNA methylation levels (Wossidlo et al., 2011; Cirio et al., 2008). At the point of implantation (blastocyst), DNAMT3b catalyzes de novo methylation that mediates the differentiation of cells into respective lineages (Borgel et al., 2010).  During the second phase of genome-wide demethylation, methylation at imprinted CpG sites is maintained in order to conserve the sex-specific expression patterns. The mechanisms by which imprinted regions are able to conserve methylation status is not clear (Van Montfoort et al., 2012). DNMT1 and a form of DNMT1 inherited from the oocyte (DNMT1o) appear to play a role as loss of DNMT1o leads to loss of imprinting at many imprinted loci (Howell et al., 2001). The oocyte is indicated to play a further role in the maintenance of imprinting in the supplying of proteins that may allow for imprinting regions to be targeted for DNA methylation preservation. Key maternal effect proteins that protect imprinted DNA methylation sites during preimplantation development have been identified: STELLA (Prayer et al., 2003), zinc finger 36  protein 57 (ZFP57) (Li et al., 2008), tripartite motif-containing 28 protein (TRIM28)(Messerschmidt et al., 2012), CpG binding protein MBD3 (Reese et al., 2007), and RNA elongation factor Elp3 (Okada et al., 2010).   1.6 Epigenetics and ART ART procedures such as ovarian stimulation, in vitro fertilization, and embryo culture overlap with the second phase of genome-wide DNA demethylation, subsequent re-methylation as well as with the final establishment of imprints (prior to ovulation) and maintenance of imprinted DMR methylation. Therefore, there is concern that the stresses oocytes and embryos are exposed to during ART procedures may lead to altered methylation at imprinted genes as well as other genes throughout the genome.   1.6.1 ART and imprinting disorders The first evidence that ART may be impacting epigenetic mechanisms, and specifically imprinting, came from animal studies in the late 1990’s and early 2000’s. Sheep derived from in vitro cultured embryos were shown to be at a greater risk for large offspring syndrome (LOS) – a ruminant specific disorder characterized by increased birth weight, congenital abnormalities, and placental dysfunction (Young et al., 1998). LOS was later identified by the same group to be associated with loss of expression of the imprinted gene insulin-like growth factor 2 receptor (IGF2R) – imprinted in sheep but not in humans – as well as loss of DNA methylation at the IGF2R imprinting center (Young et al., 2001). As LOS has similar phenotypic characteristics with the human imprinting disorder Beckwith-Wiedemann syndrome (BWS), research into the prevalence of imprinting disorders in humans with respect to ART and embryo manipulation was 37  initiated. The two imprinting disorders found to be most associated with ART are Beckwith-Wiedemann syndrome (BWS) and Angelman syndrome (AS). As listed in Table 4, BWS is a congenital overgrowth syndrome that occurs in approximately 1 in 13,700 births in the general population (Weksberg et al., 2010). Characteristics of BWS in children include macroglossia, macrosomia, midline abdominal wall defects, neonatal hypoglycemia as well as an increased risk of developing embryonal tumors in childhood (Weksberg et al., 2010). AS, also listed in Table 4, is a neurogenetic disorder that occurs in 1 in 16,000 births in the general population and is characterized by severe mental retardation, frequent laughter, seizure disorder, and gait disturbances (Clayton-Smith & Pembrey, 1992).  The first study to show an association between an imprinting disorder and ART in humans came in a case report in 2002 where loss of methylation at the maternally methylated SNPRN locus was observed in two ICSI conceived children – for male factor infertility – who developed Angelman syndrome (Cox et al., 2002). Another case report in 2003 of AS after ICSI with absence of maternal pattern methylation at SNRPN provided further evidence that ART may be associated with an increased risk of imprinting disorders (Ørstavik et al., 2003). In 2005, Ludwig et al. reported that subfertility may play a large role in AS cases as they found 20% of AS patients were born to subfertile couples in an analysis of the German Angelman Syndrome Support group where 16 out of 79 members reported subfertility (Ludwig et al., 2005). Imprinted defects at SNRPN were found in four of the 16 responders that were subfertile with the following conception breakdown: one ICSI conceived child, one child born with the aid of hormonal therapy, and two children born without reproductive technologies (Ludwig et al., 2005). As the relative risk for an imprinting defect in AS children born to subfertile couples was calculated to be 6.25 and the relative risk of an imprinting defect in AS children born to couples who received 38  fertility treatments (including hormonal therapy) was 12.5, the authors speculated that infertility and ART may be associated with imprinting disorders (Ludwig et al., 2005). A later investigation into 384 British families with AS children, found three cases to be born through ART with one case found to have altered methylation at the maternal SNRPN gene (Sutcliffe et al., 2006). Although a subsequent study in 63 children with AS study revealed none were conceived with ARTs, they found that 12 out of 63 AS children (19%) had parents that suffered from fertility issues – larger than the 5.6% of children that have parents with fertility issues in the general population at the time – indicating that infertility may have be associated with imprinting disorders (Doornbos et al., 2007). Due to the rarity of AS, the absolute risk of AS after ART is still low with only seven reported cases in all of AS after ART; however, there is compelling evidence that the risk is increased in the ART population compared to the general population. Furthermore, of these seven cases, five showed altered imprinting (much higher than the percentage of AS due to imprinting defects in the general population, 6%) suggesting an association between ART and altered imprinting at the SNRPN locus. However, more studies in AS populations are needed for conclusive results with respect to the risk of imprinting defects following ART and the extent to which infertility and/or the ART procedures themselves contribute to this risk (Michels. 2012). A link between ART and BWS was first brought to light when seven individuals with BWS were found to be conceived after ART – two IVF, five ICSI, indicating that the prevalence of ART in the BWS cohort may be 4.6% which was six-fold higher than the 0.76% prevalence in the general population at the time (DeBaun et al., 2003; Michels, 2012). Six of these patients were then tested for the molecular cause of the syndrome and five of the six showed hypomethylation at the imprinted KCNQ1OT1/KvDMR1 locus (DeBaun et al., 2003). In the 39  general population, approximately 50-60% of BWS cases are due to loss of maternal methylation at KvDMR1, which controls expression of KCNQ1OT1 – a lower percentage than observed in this study (Weksberg et al., 2005). These results added to the evidence that KCNQ1OT1 may be vulnerable to imprinting errors during the preimplantation developmental period as BWS caused by imprinting defects at KCNQ1OT1 is found to be more prevalent in monozygotic twins (Weksberg et al., 2002). Two subsequent similar studies in the UK and France of 149 BWS cases per study revealed similar results calculating an odds ratio of around 3 for risk of BWS after ART and finding a high incidence of methylation alterations at KvDMR1 (Maher et al., 2003; Gicquel et al., 2003). Further case control studies added to the evidence that ART prevalence was higher in the BWS populations than in the general population and that ART BWS patients had a higher prevalence of KvDMR1 methylation alterations (Halliday et al., 2004; Rossignol et al., 2006; Bowdin et al., 2007; Lim et al., 2009). However, these initial findings of an association between ART and imprinting disorders was not supported by further large cohort studies. A large Danish registry study of 6,052 IVF children and 442,349 naturally conceived children born between 1995-2001 did not find any BWS cases (Lidegaard et al., 2005). A large Swedish study of 16,280 ART conceived children also found no cases of BWS (Källén et al., 2005b).  Although almost all of the ART conceived BWS cases were found to have hypomethylation of the maternal allele at KvDMR1, there is conflicting data as to whether widespread methylation alterations occurs in ART conceived children and what the prevalence of BWS is in the ART conceived population (Michels, 2012). A summary of the studies that have analyzed imprinting disorders in human populations following ART is presented in Table 5 below.    40  Table 5. Studies investigation imprinting disorders following ART. Table adapted from Michels (2012) Author Study Type Number of cases Prevalence  of ART in cases (%) Prevalence of ART  in general population (%) Frequency of  imprinting defect  Ludwig et al., (2005) Survey 79 AS; 16/79 subfertile NA  NA 4/16 Debaun et al., (2003) Case series 65 BWS 4.6 0.76 5/6 Maher et al., (2003) Case series 149 BWS 4 1.2 2/2 Gicquel at al., (2003) Case series 149BWS 4 1.3 6/6 Halliday et al., (2004) Case series 37 BWS 10.81 0.67 3/3 Lidegaard et al., (2005) Cohort 6052 singleton  ART,  0 BWS found 0 1.3 0 Chang et al., (2005) Case series 341 BWS 5.6 NR NR Sutcliffe et al., (2006) Survey 79 BWS, 74 AS 2.9 0.8 8/8 BWS; 1/3 AS Rossignol et al., (2006) Cohort 11 BWS 27.5 NA 11/11 Bowdin et al., (2007) Survey 1/1524 BWS in ART; 0/1524 AS in ART  NA  NA 1/47 Doornbos et al., (2007) Survey 71 BWS; 63 AS 5.6 (BWS) 0.92 BWS NA Lim et al., (2008) Case control 25 ART BWS NA  NA 24/25 Tenorio et al., (2016) Case series 187 BWS 9.1 NA 15/17  41  1.6.2 Cause of epigenetic alterations in ART As invasive ART procedures occur simultaneously with the epigenetic reprogramming events – erasure and addition of DNA methylation – including the establishment/maintenance of DNA methylation at imprinted DMRs, it is plausible that these procedures may alter epigenetic and gene expression signatures in resulting embryos with potential developmental consequences. Due to constrains on studying human oocytes and embryos, the majority of research on the impact individual ART procedures have on epigenetic mechanisms and gene expression comes from murine and bovine studies (Anckaert & Fair, 2015). However, investigations in sheep have also linked alterations in growth and placentation following ART to epigenetics and imprinted genes (Ptak et al., 2012; Fidanza et al., 2014). Reduction in growth and survival of sheep fetuses following ART were shown to be concurrent with a reduction in DNMT1 expression which is required for proper imprinted gene maintenance (Ptak et al., 2012).  1.6.2.1 Ovarian stimulation Although there are imprinted genes that are imprinted in the male germline, most imprinted domains contain a gametic DMR where DNA methylation is established in the female germline during oocyte growth (Hiura et al., 2006). The establishment of DNA methylation at maternal gametic imprinted genes occurs at different times for each gene during the development of antral follicles from primordial follicles (Hiura et al., 2006). Bovine studies have revealed that DNA methylation at maternal gametic DMRs is linked to the diameter of the oocyte, where acquisition of DNA methylation at maternal imprints occurs during the final phase of oocyte growth in the tertiary follicle (O’Doherty et al., 2012). Furthermore, SNRPN and MEST appear to be fully methylated before IGF2, PEG10, and PLAGL1 (O’Doherty et al., 2012). In humans, 42  SNRPN, KCNQ1OT1, PLAGL1, and MEST are observed to be methylated in fully grown germinal vesicle oocytes, with PLAGL1 also found to be fully methylated at an earlier pre-antral follicle stage (Geuns et al., 2007; Sato et al., 2007; Arima et al., 2006).  Due to imprinted gene DNA methylation establishment in the oocyte occurring around the same time as superovulation and ovarian stimulation ART procedures, there is concern that these procedures may impact imprinted gene regulation and expression. Although, ovarian stimulation has been shown to be associated with global DNA methylation changes in two-cell mouse embryos (Shi & Haaf, 2002), murine studies on the impact ovarian stimulation has on oocyte imprinting establishment have been conflicting. One study investigating methylation at H19, Mest, Lit1, and Plagl1 after ovarian stimulation showed aberrant methylation at only H19 in the oocytes (Sato et al., 2007). Recent studies in meiosis II oocytes and two-cell stage embryos after ovarian stimulation did not find aberrant methylation at the H19, Snrpn, Peg3, (Denomme et al., 2011) and H19, Snrpn, Igf2r loci, respectively (El Hajj et al., 2011). Differences between these studies may be explained by differences in the protocols administering the luteinizing hormone analog to induce ovulation, differences in the susceptibility of mouse strains to imprinting defects, and whether studies corrected for cumulus cell DNA contamination (Anckaert & Fair, 2015). However, there are several studies that suggest ovarian stimulation may impact the ability of imprinted genes to maintain imprinting after fertilization. In one study, biallelic expression of Snrpn and H19 was observed in murine placentas during the gestational period following ovulation stimulation (Fortier et al., 2008). In another study, loss of methylation at maternally methylated Snrpn, Peg3, and Kcnq1ot1 and gain of methylation at paternally methylated H19 was observed in mouse blastocysts, the degree of which correlated with the doses of drugs administered during superovulation (Market-Velker, Zhang et al., 2010). Altered DNA methylation and/or gene 43  expression of H19, Peg3, and Snrpn was also observed in different tissues (liver, brain, brain, respectively) in mice born from oocytes obtained after superovulation (de Waal et al., 2012). Interestingly, following ovarian stimulation in mice, aberrant DNA methylation has been observed at Snrpn, H19, and Mest in the spermatozoa of first-generation and second-generation male offspring, suggesting that there may be transgenerational effects from ovarian stimulation on imprinted gene maintenance (Stouder et al., 2009; Anckaert & Fair, 2015). Although there are mouse studies that show that ovarian stimulation has no effect on imprinting establishment, there are some studies that suggest ovarian stimulation may interfere with imprinting maintenance after fertilization. The mechanisms by which superovulation may be impacting imprinting maintenance is not clear; however, it is possible that ovarian stimulation may impact the ability of the oocyte to synthesize or store maternal effect products that are necessary for imprinting maintenance after fertilization.  1.6.2.2 In vitro culture The impact of embryo in vitro culture on genomic imprinting has mostly, due to ethical constraints, been studied in the mouse (Anckaert & Fair, 2015). Similar to the studies in ovarian stimulation, reports of imprinted alterations following in vitro culture are conflicting (Anckaert & Fair, 2015). Aberrant imprinting of the maternally methylated Snrpn, Igf2r, Mest, Peg3, and the paternally imprinted H19 have been previously demonstrated following in vitro culture (Doherty et al., 2000; Young et al., 2001; Khosla et al., 2001; Karjean et al., 2003; Mann et al., 2004; Rivera et al., 2008; Suziki et al., 2009). Altered methylation at H19, Snrpn, and at the KvDMR1 locus as well as altered expression of Igf2, Kcnq1ot1, H19, Snrpn, and Peg3 has been observed in placental and fetal tissues derived from embryos that were exposed to in vitro culture 44  conditions, providing evidence that aberrant DNA methylation from embryo culture may be maintained in pregnancy (Mann et al., 2004; Rivera et al., 2008). However, comparable DNA methylation at Snrpn, Mest, Peg3, Igf2r, and H19 between oocytes from in vitro follicle culture, in vivo grown superovulated oocytes, and in vivo grown mouse oocytes has also been reported (Lucifero et al, 2002; Lucifero et al., 2004; Hiura et al., 2006; Anckaert et al., 2009b). Recent studies support stability of imprinted genes in fully grown germinal vesicle oocytes (Trapphoff et al., 2010) and 2-cell embryos subjected to in vitro culture conditions (El Hajj et al., 2011). The type of culture media used during embryo and oocyte culture conditions appears to be important with respect to DNA methylation alterations as mouse embryos cultured in different culture media were observed to be impacted differently – embryos cultured in Whitten’s media were observed to have loss of methylation at H19, while embryos cultured in KSOM+AA had normal methylation at H19 (Doherty et al., 2000).  A more recent study provided additional evidence for the importance of the embryo culture media where differences in the severity of methylation aberrancies at H19, Snrpn, and Peg3 were found in a comparison of six different embryo culture media where embryos were cultured from the two cell stage to the blastocyst stage (Market-Velker, Fernandes et al., 2010). However, imprinted gene methylation has also been found to be stable when oocytes are subjected to in vitro culture under various treatments and suboptimal conditions (Anckaert et al., 2009a; Anckaert et al., 2009b; Anckaert et al., 2010). The differences between the findings of altered methylation at imprinted genes in embryos following embryo culture and the stability of methylation at imprinted genes in oocytes following oocyte culture may indicate a different susceptibility of the oocyte and the embryo to in vitro culture effects on imprinting (Anckaert & Fair, 2015). This may be due to alterations in the ability of maternal effect proteins to maintain DNA methylation at imprinted genes, after superovulation or embryo 45  culture, during epigenetic reprogramming events in the embryo (Anckaert & Fair, 2015). The amount of time and/or the stage of in vitro culture during embryogenesis may also be important with respect to DNA methylation alterations. Indeed, embryos grown in vivo to the blastocyst stage and then transferred and subjected to in vitro culture for less than 1.5 hours displayed lower DNA methylation aberrancies than embryos that were subject to in vitro culture from the two-cell stage to the blastocyst stage (Rivera et al., 2008). Although results are conflicting, there is evidence in murine studies for the interference of in vitro culture on DNA methylation, especially at imprinted genes, that seems to be more prevalent in embryos and more developed tissues than in oocytes, suggesting that maternal effect proteins may be impacted by ART procedures. Obtaining a clear overall picture of how human oocytes and embryos may be impacted by embryo culture is difficult due to ethical constraints and the differences in proprietary culture media and protocols used in different IVF laboratories. Furthermore, improvements in culture media and superovulation protocols with respect to pregnancy rates and in simulating the in vivo environment in recent years, which is still an ongoing process, may be reducing the adverse impacts ART procedures may have on DNA methylation establishment/maintenance over time.  1.6.2.3 Male factor infertility Findings of altered genome-wide DNA methylation levels, DNA methylation at imprinted genes, and DNA damage in the sperm of infertile men has led to concern that the use of sperm from infertile men in ART, particularly ICSI, may be passing genetic and epigenetic aberrancies onto resulting offspring. A number of studies have shown alteration in DNA methylation at both paternally and maternally expressed genes – H19, MEG3, MEST, 46  KCNQ1OT1, PLAGL1, PEG3, SNRPN, and IGF2 – in infertile men (Marques et al., 2004; Kobayashi et al., 2007; Marques et al., 2008; Poplinski et al., 2010; Hammoud et al., 2010; Minor et al., 2011). These studies have also highlighted that the risk of imprinted gene alterations appears to increase with the severity of the male infertility, particularly with respect to sperm count. A large study by Kobayashi et al. (2007) showed that sperm from infertile men, particularly men with low sperm count (oligozoospermia), had higher rates of altered DNA methylation at imprinted genes when compared to fertile men. Out of 97 infertile men, this study found abnormal DNA methylation at paternal imprints – H19 and MEG3 – in 14 patients and abnormal DNA methylation at maternal imprints – MEST, KCNQ1OT1, PLAGL1, PEG3, and SNRPN – in 20 patients (Kobayashi et al., 2007).  In an effort to shed more light on DNA methylation alterations in infertile men where semen parameters are normal, a recent meta-analysis investigated controlled clinical trials evaluating sperm DNA methylation at imprinted genes in men with idiopathic infertility (normal semen analysis) compared to fertile controls. In an analysis of 24 studies, H19 methylation levels were significantly lower, MEST methylation levels were significantly higher, and SNRPN methylation levels were significantly higher in idiopathic infertile men than in controls (Santi et al., 2017).  Although early studies on sperm DNA methylation focused on imprinted genes due to the link between ART and imprinting disorders, altered sperm DNA methylation patterns have also been found in non-imprinted genes, including genes associated with spermatogenic impairment (Houshdaran et al., 2007; Navarro-Costa et al., 2010; Wu et al., 2010; Nanassy & Carrell, 2011; Pacheco et al., 2011; Heyn et al., 2012). As such, genome-wide methylation patterns in sperm of infertile men has become of interest as a means to characterize male infertility, understand the 47  mechanisms behind infertility, and elucidate the safety of using sperm from infertile men in ART procedures. Of particular interest is the potential for using DNA methylation alterations found in sperm as a means to better understand idiopathic infertility where men that are infertile have normal semen parameters. A recent genome-wide methylation study investigating genome-wide DNA methylation in 29 normospermic infertile men compared to 17 normospermic fertile men using the Illumina Infinium HD Human Methylation 450k array found 2752 CpG sites that were differently methylated between patient groups (Urdinguio et al., 2015). These differentially methylated CpG sites were also significantly associated with sites that are methylated specifically in sperm and not somatic cells (Urdinguio et al., 2015). The authors concluded that DNA methylation may contribute to infertility impairment in couples with unexplained infertility.   Evidence of increased DNA methylation abnormalities in the sperm of infertile men when compared to fertile controls is concerning as these abnormalities may be inherited by subsequent offspring. Indeed, a study comparing paternal sperm methylation to methylation abnormalities found in ART concepti revealed seven out of seventeen cases where abnormal DNA methylation in an ART sample was identical to alterations were found in the paternal sperm (Kobayashi et al., 2009). Therefore, DNA methylation alterations found in human ART pregnancies (imprinting disorders), may in part be due to the use of sperm from infertile men with intrinsic imprinting mutations.   1.6.3 Imprinted gene alterations in non-imprinting disorder ART conceived babies The findings of an increased risk for imprinting disorders, caused by altered methylation at imprinted genes, in ART conceptions catalyzed research on imprinted gene methylation in 48  ART conceptions that did not have imprinting disorders. As imprinting disorders were rare and populations difficult to study, interrogating DNA methylation and gene expression of imprinted genes in non-imprinting disorder ART populations may provide clearer insight into the impact ART may have on imprinted genes as larger, better controlled populations could be used. Furthermore, as imprinted genes are important for growth and development, investigating DNA methylation and gene expression at imprinted genes following ART is a potential avenue to explain the adverse pregnancy and neonatal outcomes observed following ART. The majority of studies were conducted on cord blood and/or placenta due to these tissues being unwanted and relatively easy to collect as well as being derived from the trophectoderm and inner cell mass of the blastocyst respectively.  Initially, efforts were focused on analyzing DNA methylation at imprinted genes associated with BWS and AS (the imprinting disorders linked to ART). One of the first analysis of DNA methylation at imprinted genes in non-imprinted syndrome ART pregnancies analyzed methylation at KvDMR1 (associated with BWS) in peripheral blood and cord blood from phenotypically normal children conceived via IVF and ICSI (Gomes et al., 2009). This study revealed hypomethylation at KvDMR1 in 3/18 (2 IVF, 1 ICSI) children conceived through ART where mean methylation levels were 41.5% in NC controls and 14% in the IVF/ICSI group (Gomes et al., 2009). However, following studies have yielded conflicting results. Most studies report an overall high stability of methylation at imprinted genes in the cord blood and placenta of ART conceived neonates (Tierling et al., 2010; Wong et al., 2011; Rancourt et al., 2012; Oliver et al., 2012; Puumela et al., 2012; Camprubí et al., 2013; Nelissen et al., 2013; Sakian et al., 2015; Melamed et al., 2015; Vincent et al., 2016). Although these studies revealed robustness at DMRs of imprinted genes, slight differences in methylation between ART conceived and NC controls was still noted at MEST, KCNQ1OT1, SNRPN, H19, 49  PEG3, MCTS2, and PLAGL1 (Tierling et al.,  2010; Rancourt et al., 2012; Camprubí et al., 2013; Nelissen et al., 2013; Vincent et al., 2016). Marked differences in DNA methylation at imprinted genes have also been observed in several studies (Gomes et al., 2009; Katari et al., 2009; Turan et al., 2010; Song et al., 2015). In an investigation of 66 ART and 49 NC conceived neonates, placental methylation at 37 CpG sites across 16 genes, including CpG sites at five imprinted genes, Song et al. (2015) found 49% of the CpGs to be significantly different between infertile ART and fertile NC groups. In another analysis where DNA methylation at imprinted genes in cord blood from 45 ART conceived and 56 matched NC controls was compared, aberrant methylation patterns were observed at the H19/IGF2 locus (Turan et al., 2010). A recent meta-analysis review of the literature on DNA methylation at imprinted genes following ART conception (18 publications included) revealed no difference in the weighted mean difference in methylation between ART and NC groups at H19, PEG1-MEST, GRB10, IGF2, SNRPN, KvDMR/KCNQ1OT1, and PEG3 (Lazaraviciute et al., 2014). However, the authors concluded that their ability to determine the full effect of ART on DNA methylation and imprinting was hampered by heterogeneity in the types of infertility treatments, tissues used, and measurement methods (Lazaraviciute et al., 2014). They called for more controlled studies, using standardized methodologies, in larger, better clinically defined populations (Lazaraviciute et al., 2014).   Although most of the literature points to DNA methylation at imprinted genes following ART in humans being relatively stable, several studies have found altered expression of imprinted genes without any alterations in methylation at the regulatory DMR (Katari et al., 2009; Turan et al., 2010;  Nelissen et al., 2014; Sakian et al., 2015; Vincent et al., 2016). Indeed, other epigenetic mechanisms such as histone modifications and noncoding RNAs have been implicated in the regulation of imprinted genes, highlighting the importance of studying gene 50  expression of imprinted genes in relevant tissues (Lewis et al., 2004; Lui et al., 2008; Court et al., 2014). Altered expression of H19, IGF2, PLHDA2, KCNQ1OT1, and PLAGL1 have been observed in placental and/or cord blood tissues when comparing ART conceptions to NC controls (Rancourt et al., 2013; Nelissen et al., 2013; Nelissen et al., 2014; Sakian et al., 2015; Vincent et al., 2016).   1.6.3.1 Imprinted gene network The notion of an imprinted gene network between imprinted genes located on chromosomal band 6q24 (PLAGL1) and chromosome band 11p15.5 (KCNQ1OT1, CDKN1C, H19, and IGF2) started when patients with mutations in PLAGL1 and CDKN1C resulting in transient neonatal diabetes mellitus (TNDM) and BWS, respectively, were noted to share many of the same characteristics (Arima et al., 2005). This group went on to show that the mouse Zac1 (PLAGL1 in humans) and p57Kip2 (CDKN1C in humans) shared expression patterns. PLAGL1, a paternally expressed imprinted gene – methylated in the female germ line – encodes a zinc finger transcription factor with potential binding sites at CDKN1C and KCNQ1OT1; therefore, Arima et al. (2005) tested to see if PLAGL1 was interacting with either CDKN1C or KCNQ1OT1 regions. Although, they could not show human PLAGL1 protein binding to CDKN1C (via band shift assays), they showed PLAGL1 binds to the KCNQ1OT1 CpG island region (preferentially to an unmethylated CpG island) and that the binding of PLAGL1 to the unmethylated KCNQ1OT1 CpG island region promoted expression of KCNQ1OT1 (Arima et al., 2005). KCNQ1OT1 is an anti-sense RNA that negatively regulates imprinted genes on chromosome 11p15.5, including the CDKN1C gene (Arima et al., 2005). Therefore, they concluded that PLAGL1, CDKN1C, and KCNQ1OT1 could be a part of a signaling pathway regulating cell 51  growth (Arima et al. 2005). A later study characterized the phenotype of Zac1 knock out mice where Zac1 was knocked out on the paternal allele of mice (Varrault et al., 2006). Resulting pups showed extreme growth restriction, altered gross morphology, and high rates of neonatal lethality (Varrault et al., 2006). In order to gain insight into the mechanism of Zac1, they then performed a meta-analysis of 116 mouse microarray gene expression data sets in order to find genes frequently co-expressed with Zac1, finding strong co-expression of Zac1 with several imprinted genes. Varrault et al. (2006) also demonstrated that expression of Zac1 upregulated expression of Igf2, Cdkn1c, H19, and Dlk1 and that Zac1 under expression resulted in downregulation of these same genes – CDKN1C, IGF2, H19 and DLK1 (Varrault et al., 2006). Furthermore, they showed that Zac1 binds to a shared enhancer (E2) that induces expression of H19 and Igf2 (Varrault et al., 2006). The authors concluded that they identified a network of imprinted genes that are highly involved in embryonic growth where Zac1 (PLAGL1) may be a master regulator (Varrault et al., 2006). In humans, Iglesias-Platas et al. (2014) recently demonstrated significant positive correlations in gene expression between PLAGL1 and H19, IGF2, and CDKN1C and showed binding of PLAGL1 to the enhancer region for the IGF2/H19 locus, thus supporting PLAGL1 as a regulator of this imprinted gene cluster/IGN in humans. Interestingly, a small subgroup of their study populations that were conceived via IVF or ICSI (n=13) had significantly lower PLAGL1 expression in term placental villi compared with controls; however, the number of ART cases was too small to be conclusive (Iglesias-Platas et al., 2014). Although no differences in PLAGL1 expression and methylation were found in the placenta, a more recent investigation in PLAGL1 methylation and expression in ART (101 IVF, 81 ICSI, 82 NC) conceived neonates revealed significantly higher methylation at PLAGL1 in cord blood from IVF conceived groups compared 52  to NC control as well as lower expression of PLAGL1 in IVF and ICSI groups compared to controls (Vincent et al., 2016).  KCNQ1OT1 was also shown to have lower expression in the IVF group when compared to controls and PLAGL1 expression was found to be correlated to KCNQ1OT1 expression but not to CDKN1C or IGF2 expression (Vincent et al., 2016). These studies suggest that PLAGL1 and the imprinted gene network may be altered following ART, potentially contributing to adverse outcomes after ART.   1.6.4 Genome-wide DNA methylation in ART conceived babies Epigenetic mechanisms are also thought to be the mediator through which intra-uterine and neonatal life can shape adult and long-term health in the developmental origins of health and disease hypothesis (DOHaD). Furthermore, DNA methylation dynamics during the epigenetic reprogramming periods of conception and early embryogenesis occurs in a genome-wide fashion. Therefore, ART related interventions that occur simultaneously to this epigenetic reprogramming event could have far reaching genome-wide impacts on DNA methylation  With the reduction in cost of interrogating genome-wide methylation, more studies have focused on determining global methylation changes associated with ART conception. An initial study utilized the Illumina GoldenGate assay to examine DNA methylation at 1536 CpG sites in 10 IVF and 13 NC samples, finding hypermethylation in placental and cord blood tissues, at both imprinted and non-imprinted genes, from IVF children (Katari et al., 2009). This study also demonstrated gene expression differences in a subset of the differentially methylated genes (Katari et al., 2009). Another study used the GoldenGate assay to interrogate 25 imprinted DMRs in a larger population consisting of 73 ART and 121 NC cord blood and placental samples (Camprubí et al., 2013). However, no significant differences in DNA methylation at the 53  25 imprinted DMRs between the ART and NC groups were observed (Camprubí et al., 2013). More recently, emergence of Illumina arrays capable of interrogating 27,000 and 450,000 CpGs (27K and 450K arrays) have allowed analysis of more global DNA methylation sites. In an analysis of 10 ART and 8 NC cord blood samples using the Illumina Infinium HumanMethylation 27K array, 24 genes with at least two differentially methylated CpG sites were found to be significantly different between IVF and NC controls (Melamed et al., 2015). Furthermore, this study observed increased variance in methylation in the ART group suggesting stochastic genome-wide changes in DNA methylation (Melamed et al., 2015). A later study analyzed blood spots from 18 IUI, 38 ICSI with frozen embryo transfer, 38 ICSI with fresh embryo transfer, and 43 naturally conceived neonates using the more comprehensive Illumina Infinium HumanMethylation450 bead array, investigating DNA methylation at over 450,000 CpG sites (Estill et al., 2016). This study revealed different methylation profiles between ART conceived and NC controls, particularly in promoter regions and exons of protein coding loci (Estill et al., 2016). The DNA methylation profiles of ICSI-frozen and IUI bloodspots were similar, suggesting that cryopreservation may mitigate epigenetic alterations associated with ART procedures (Estill et al., 2016). Interestingly, this study also found differences in DNA methylation at metastable epialleles – loci where DNA methylation is stochastically established within blastocysts, influenced by peri-conceptional nutritional states – suggesting that periconception states may have a lasting impact on the epigenome of ART conceived individuals (Estill et al., 2016). Another study utilized the Illumina 450k array to interrogate genome-wide DNA methylation in cord blood from 48 ICSI and 46 NC singleton pregnancies, finding 0.11% of analyzed CpG sites to be differentially methylated between conception modes with enrichment in CpG islands with low methylation values (0-20%) and in imprinting control 54  regions (El Hajj, Haertle et al., 2017). However, another recent investigation into genome-wide DNA methylation after ART using genome-wide methylated DNA immunoprecipitation coupled with deep sequencing (MeDIP) in whole cord blood cells and cord mononuclear cells in 47 IVF and 60 NC newborn twins (54 total twin pairs) found only one region to be significantly different in methylation at a false discovery rate of 5% (Castillo-Fernandez et al., 2017). The authors suggested that infertility associated genes may have altered methylation in ART newborns due to this region being ~3kb upstream of a gene previously linked to male infertility (TNP1); furthermore, of the 46 genes associated with ART at a false discovery rate (FDR) of 25%, one gene (C9orf3), is also associated with infertility (Castillo-Fernandez et al., 2017). Three studies have investigated genome-wide methylation in placental tissue using the Illumina 450k array; however, these studies have failed to find significant relationships between CpG methylation and ART (Litzky et al., 2017; Xu et al., 2017; Choufani et al., 2018).   A recent review of genome-wide DNA methylation studies in ART conceived populations highlighted the remarkable inconsistencies in the results (Mani et al., 2019). The authors compiled a list of the 237 unique genes found to be differentially methylated in ART conceived groups in genome-wide DNA methylation studies to date. Of these, only four genes – GNAS, PEG10, PRCP, and RUNX3 – were overlapped between studies and only one, GNAS, was found to be significantly altered in ART conceived groups in three studies (Mani et al., 2019). Inconsistent reports on DNA methylation alterations in ART samples may be due to the stochastic nature of the DNA methylation disturbances, the assays used, heterogeneity of clinical variables, heterogeneity of ART procedures, and insufficient sample sizes (Mani et al., 2019).  55  1.7 Rational and hypotheses Assisted reproductive procedures, ovarian stimulation and in vitro culture, overlap with establishment of regulatory DNA methylation at imprinted gene DMRs, imprinted gene DMR DNA methylation maintenance, and genome-wide epigenetic reprogramming (Van Montfoort et al., 2012). Animal studies have shown that ovarian stimulation and in vitro culture may impact DNA methylation and gene expression of imprinted genes (Van Montfoort et al., 2012). There have been reports of an increased prevalence of imprinting disorders caused by DNA methylation alterations in ART conceived individuals (Michels, 2012). Infertility has also been linked to alterations in epigenetic mechanisms, including genomic imprinting (Kobayashi et al., 2007). As imprinted genes are crucial for growth and development of the fetus in utero, alterations in the regulation and expression of imprinted genes, potentially caused by ART procedures or inherited from infertile parents, may be related to the adverse pregnancy outcomes observed following ART. PLAGL1 is an imprinted gene encoding a transcription factor that has been proposed to regulate a growth implicated imprinted gene network of genes on the chromosome 11p15.5 band – CDKN1C, KCNQ1OT1, and H19 (Varrault et al., 2006). Alterations in DNA methylation and expression of PLAGL1 has previously been observed in IUGR pregnancies, suggesting a link between altered PLAGL1 regulation and adverse pregnancy outcomes related to growth (Iglesias-Platas et al., 2014). Alterations in DNA methylation and expression of PLAGL1 have also been found in cord blood from ART pregnancies (Vincent et al., 2016). However, this finding was confounded by the inclusion of twins and pregnancy complications in the ART groups only, therefore, it is difficult to determine whether the ART procedures/infertilities of the parents, or the pregnancy complications are associated with the alterations in PLAGL1. However, we suspect that the ART procedures may be causing 56  alterations in the regulation of PLAGL1 and associated genes in the IGN in cord blood and that ART procedures with an in vitro step (IVF and ICSI) may impact IGN imprinted gene regulation more than ART procedures without this step (IUI). Due to link between infertility severity and imprinted gene alterations, we further suspect that ICSI and IVF conceived babies may be at a greater risk for imprinted gene alterations due to these procedures treating more severe infertilities than IUI. As such, I hypothesized that 1) levels of mRNA in genes involved in the imprinted gene network may be altered, particularly PLAGL1, in cord blood from IVF, ICSI, and/or IUI conceived pregnancies compared to natural conceptions. Cord blood from IVF and ICSI babies may have more alterations in mRNA levels of imprinted genes than in cord blood from IUI babies.  Alterations in DNA methylation have also been observed at non-imprinted genes (Melamed et al., 2015; Estill et al., 2016; El Hajj et al., 2017). Increasing evidence suggests that DNA methylation and other epigenetic mechanisms may be altered by peri conception and intra uterine environments that may lead to long-term adverse outcomes in health (Hart & Norman, 2013; Dominguez-Salas et al., 2014; Silver et al., 2015; Estill et al., 2016). As ART procedures introduce gametes and embryos to non-physiological in vitro environments during epigenetic reprogramming, there is concern that ART may induce genome-wide DNA methylation alterations. Several studies have shown small but significant changes in genome-wide DNA methylation in ART conceived babies (Melamed et al., 2015; Estill et al., 2016; El Hajj et al., 2017); however, more studies covering larger proportions of the CpG sites in the genome and controlling for clinical confounders are warranted. We suspect that a genome-wide approach, using the novel Illumina Infinium MethylationEPIC BeadChip array covering over 850,000 CpG sits across the genome, to interrogating DNA methylation in ART conceived babies may provide 57  insight into the susceptibility of the epigenome to periconception ART procedures as well as alterations in genes that may be related to adverse outcomes observed following ART. As such, I hypothesized that 2) DNA methylation patterns in cord blood from ART conceived pregnancies may be altered compared to NC controls and may correlate with ontologies related to negative outcomes associated with ART pregnancies.   1.7.1 Objectives Objective 1a. To investigate the mRNA levels of genes in the imprinted gene network – PLAGL1, H19, KCNQ1OT1, and CDKN1C – in cord blood from children conceived via IVF, ICSI, IUI and natural conceptions. Objective 1b. To assess the severity of alterations in gene expression of PLAGL1, H19, KCNQ1OT1, and CDKN1C across ICSI, IVF, and IUI procedures to determine the risk each procedure poses for alterations in the expression of imprinted gene network associated imprinted genes. Objective 2a. To assess the genome-wide DNA methylation patterns in cord blood from children conceived via IVF and ICSI compared to those conceived naturally.  Objective 2b. To assess DNA methylation at imprinted gene DMRs in cord blood from children conceived via IVF and ICSI compared to those conceived naturally.  Objective 2c. To determine if alterations in DNA methylation identified in cord blood between ART and naturally conceived babies may be linked to adverse outcomes following ART.   58  Chapter 2: GENE EXPRESSION ANALYSIS OF PLAGL1 AND AN IMPRINTED GENE NETWORK IN CORD BLOOD FROM IVF, ICSI, AND IUI CONCEIVED BABIES COMPARED TO NC CONTROLS 2.1  Introduction ART singleton pregnancies are at a higher risk than NC pregnancies for a number of pregnancy and neonatal adverse outcomes including perinatal mortality, preterm birth, small for gestational age, and congenital abnormalities (Qin et al., 2016). Although infertility of the parents and the actual ART procedures themselves are thought to play a role in the adverse outcomes following ART, the degree to which either of these factors impact adverse outcomes and the mechanisms involved are not known. Observations of a higher rate of imprinting disorders caused by altered DNA methylation at imprinted genes has led to the hypothesis that ART children may be at an increased risk of epigenetic abnormalities, which may be contributing to the adverse outcomes in ART pregnancies (Michels, 2012). Indeed, adverse outcomes including pre-eclampsia, intra-uterine growth restriction, and small for gestational age have been associated with DNA methylation and imprinted gene alterations (Peters, 2014). This hypothesis has been strengthened by animal studies that have identified DNA methylation and imprinted gene alterations after ART procedures (embryo culture, superovulation) (Van Montfoort et al., 2012).  Genomic imprinting (imprinted genes) is an epigenetic mechanism that allows for genes to be expressed in a parent-of-origin specific manner and is crucial for proper placental and fetal development/growth (Reik & Walter, 2001). Expression of imprinted genes from either the paternally inherited or maternally inherited allele is regulated by the presence or absence of DNA 59  methylation at CpG islands termed differentially methylated regions (DMRs) – each parental allele has opposite methylation to the other (either hypo or hyper methylated) (Barlow, 2011).  Sex-specific methylation at imprinted DMRs are established during gametogenesis, although completed prior to birth in the male germline, methylation of imprinted genes in the female germline is not completely established until prior to ovulation in each menstrual cycle (Anckaert & Fair, 2015). Interestingly, DNA methylation at imprinted genes is maintained during fertilization and early embryogenesis – a period of genome-wide DNA methylation reprogramming (Van Montfoort et al., 2012). The procedures used during ART (ovarian stimulation, in vitro maturation, embryo culture) overlap with the establishment and maintenance of imprinted genes; therefore, it is plausible that these procedures may be altering the regulation and expression of imprinted genes and may be contributing to the observations of higher rates of imprinting disorders found in ART conceived individuals and altered imprinted gene regulation/expression in ART animal studies. Although investigations into DNA methylation and gene expression alterations of imprinted genes in ART populations have yielded conflicting results, a recent investigation by our group reported significantly higher methylation at the imprinted gene PLAGL1 – paternally expressed, methylated in the maternal germ line, gene encoding a zinc finger transcription factor – in cord blood from ART neonates when compared to NC controls (Vincent et al., 2016). Lower expression of PLAGL1 was also found in cord blood from ART conceived neonates when compared to NC controls (Vincent et al., 2016). A similar result was found in a subgroup analysis of ART pregnancies within another study where PLAGL1 expression was found to be lower in ART placental villi when compared to NC controls, however, the sample size of this subgroup analysis was small (n=13) (Iglesias-Platas et al., 2014). PLAGL1 has been shown to 60  upregulate a number of growth-related imprinted genes located on the human chromosome 11p15.5 band (H19, IGF2, CDKN1C, and KvDMR1) (Iglesias-Platas et al., 2014). As such, PLAGL1 has been proposed to potentially serve as a master regulator of a network of imprinted genes termed the imprinted gene network (IGN) that could function to fine-tune growth and development of the fetus in utero (Iglesias-Platas et al., 2014). Alterations in the expression of PLAGL1 and genes in the IGN may occur higher in ART conceived pregnancies than in NC, potentially contributing to the adverse outcomes observed after ART with respect to growth and development. However, the study conducted by Vincent et al. (2016) was complicated by the inclusion of twins in only the IVF and ICSI groups (not in the NC group), inclusion of complicated pregnancies (pre-term birth, small for gestational age, pre-eclampsia), and significant differences in maternal age between IVF, ICSI and NC conception modes. Furthermore, mRNA of PLAGL1 was measured across all transcripts in cord blood, including mRNA transcribed from the bi-allelically expressed P2 promoter, a promoter that is not imprinted and not regulated by the DNA methylation at the DMR. The study design also made it difficult to differentiate between the impact of the underlying infertility of the parents and the effect of the ART procedures themselves on imprinted gene regulation/expression. The research was therefore hindered with respect to the ability to make conclusions on the expression of the imprinted gene PLAGL1 in ART pregnancies, necessitating further research. I therefore examined gene expression of mono-allelically paternally expressed PLAGL1 as well as the expression of genes in the IGN – KCNQ1OT1, H19, and CDKN1C – in the cord blood collected from phenotypically normal singleton IVF, ICSI, IUI, and NC neonates. The inclusion of an IUI group, an ART that does not include in vitro culture, allows for comparisons to be made between ART procedures in order to 61  determine the effects superovulation and in vitro culture may have on the expression of these imprinted genes.  2.2 Methods 2.2.1 Study participants Study participants consisted of 77 cases (24 IVF, 18 ICSI, 9 IUI, and 26 NC) where IVF, ICSI, and IUI pregnancies were recruited from several IVF centers across Canada and NC pregnancies were recruited from hospitals in the greater Vancouver area, Canada. Written informed consent was received from pregnant women, those that wished to participate, for the collection of cord blood for research purposes. Gene expression and statistical analysis was performed at the BC Children’s Hospital Research Institute in Vancouver, British Columbia. The University of British Columbia Research Ethics Board approved this study.  IVF, ICSI, IUI, and naturally conceived pregnancies resulting in live singleton births were included in this study only. In order to exclude gene expression alterations caused by pregnancy complications, pregnancies with perinatal complications including pre-eclampsia, preterm birth, fetal malformation, intra-uterine growth restriction, maternal smoking during pregnancy, and known chromosomal abnormalities were excluded from the analysis. NC pregnancies were also excluded if any sign of infertility was present – fertility drugs were used for conception and/or if the pregnancy took over one year to conceive. Maternal/paternal characteristics and pregnancy history was collected via a self-administered questionnaire. Pregnancy outcome information was collected by delivery room staff or taken from the patients’ medical chart. Information regarding some of the particulars of the ART treatment – day of 62  embryo transfer, hormonal dosage for ovarian stimulation, embryo quality, and pre-implantation genetic diagnosis results – were not known.   2.2.2 Sample preparation Cord blood was collected by delivery room staff immediately following delivery into EDTA vacuum tubes (BD Vacutainer®, NJ). Within 24 hours of delivery, 2.5ml of cord blood was transferred into PAXgene™ blood RNA tubes (PreAnalytix, Switzerland) in order to preserve RNA in the blood cells. The PAXgene™ blood RNA tubes were incubated overnight at room temperature prior to being stored at -20˚C or -80 ˚C. If PAXgene™ blood RNA tubes were immediately frozen at -20˚C or -80 ˚C, they were thawed overnight before RNA extraction.   2.2.3 Reverse transcription and cDNA library preparation Cord blood RNA was extracted using the PAXgene™ Blood RNA kit (PreAnalytix, Switzerland) according to the manufacturer protocol, where 2 mL of Paxgene™ tube cord blood was used per extraction, and stored at -20°C. Prior to cDNA conversion, RNA samples were quality control tested on 1.5% agarose gel and then randomized in order to minimize the risk of batch effect across conception modes (IVF, ICSI, IUI, NC). RNA concentrations were determined using a NanoSpec spectrometer (Nanovue by General Electric Inc., CT, USA). 1µg of RNA from each sample was converted into cDNA, in duplicate, using the High Capacity cDNA Reverse Transcription kit (Applied Biosystems, Foster City, CA), with the following thermal profile: 10 minutes at 25°C, 2 hours at 37°C, and hold at 4°C. The cDNA product concentrations were determined using a NanoSpec spectrometer (Nanovue) and then diluted one in ten times prior to being used for quantitative real-time PCR. 63  2.2.4 Quantitative real-time polymerase chain reaction (qPCR) Gene expression analysis was carried out using reverse transcription and quantitative real-time polymerase chain reaction (RT-qPCR). All cDNA samples were run in duplicate on a clear MicroAmp® Optical 96-Well Reaction Plate (Applied Biosystems, Foster City, CA) with TaqmanTM Universal PCR master mix (Applied Biosystems) on a 7500 Fast Real-Time PCR system with 7500 software v2.0.6 (Applied Biosystems). The thermal cycling conditions were the following: 1 cycle at 50°C for 2 min and 95°C for 10 min, followed by 40 cycles of 95°C for 15s and 60°C for 1 min. As cDNA conversion was run in duplicate, a total of four gene expression measurements were performed for each sample (per gene). PLAGL1, CDKN1C, KCNQ1OT1, and H19 gene expression assays were performed with PrimeTime® Std qPCR Assays, FAM-labelled TaqmanTM Probes (Applied Biosystems). The following will list, per gene, the PrimeTime® Std qPCR Assay, the RefSeq acquisition number of the mRNA variants analyzed, and the exons that the assay spans. The PLAGL1 assay, Hs.PT.58.26731331, detects the variants NM_006718, NM_002656, NM_001080953, NM_001080952, and NM_0010809516 which cover exons 2-5c and are all expressed from the imprinted P1 promoter. The KCNQ1OT1 assay, Hs.PT.58.4572396.g, detects the lncRNA NR_002728 covering exon 1. The CDKN1C assay, Hs.PT.58.1677181, detects all variants (NM_001122631, NM_001122630, NM_000076) covering exons 1-2. The H19 assay, Hs.PT.58.2694336.g, detects the lncRNA NR_002196 covering exons 4-5. The endogenous control, reference gene UBC Hs.PT.39a.22214853 PrimeTime® Std qPCR Assay (Applied Biosystems), detects the only mRNA variant NM_021009 covering exons 1-2. The amount of reagent per well of the 96 well plate for the target genes were as follows: 10µl TaqManTM Gene Expression Master Mix (Applied Biosystems), 1µl target gene primer, and 9 µl of the cDNA sample (diluted one in ten 64  times). The amount of reagent per well of the 96 well plate for the endogenous control gene was 10 µl TaqManTM Gene Expression Master Mix (Applied Biosystems), 1µl endogenous control gene primer, 5µl water, and 4 µl of the cDNA sample. No-reverse-transcription controls were also run in duplicate for each sample run (per gene) where reverse transcriptase was not added before cDNA library construction in order to control for genomic DNA contamination. Calibration wells were included on the plate for each gene in order to perform inter-plate calibration and control for variation between qPCR runs. In these wells, cDNA was added from a calibrator sample – a pool of cDNA from all the natural conception control samples. Two no-template-controls were included on each plate (water added in place of cDNA). Quantification cycle (Cq) values were determined by the Applied Biosystems 7500 software v2.0.6 for each gene per plate. Inter-plate calibration, qPCR repeat averaging, normalization with the reference gene UBC, reverse transcriptase repeat averaging, and relative quantification were performed using GenEx (TATAA Biocentre, Göteborg, Sweden) (MultiD Analyses AB, Göteborg, Sweden) qPCR data analysis software (version 7.0). The relative quantification (RQ) values were determined by the GenEx (version 7.0) software using the ddCT method where the sample with the highest delta Cq value was used as the calibrator (Livak & Schmittgen, 2001).  2.2.5 Statistical analysis Statistical analysis and graphics rendering were performed in R statistical program (version 3.6) and in conjunction with packages ggplot2 (version 3.1.1; Wickham, 2016) and car (version 3.0-3; Fox & Weisberg, 2018). Clinical information and RNA expression relative quantification (RQ) value differences between conception modes was calculated with one-way analysis of variance (ANOVA) or Kruskal-Wallis, depending on data distribution. If significant 65  differences were found, pairwise differences were calculated with the Tukey’s honest significant difference post hoc when ANOVA was used or Mann-Whitney test following Kruskal-Wallis. Fisher’s exact test was used to calculated differences between categorical variables. The Bonferroni correction was used to correct for multiple group comparisons. Correlation analyses were determined with Pearson’s r or Spearman’s rho, depending on data distribution. A minimum significance level of 5% (95% confidence interval) was used across all analyses.   2.3 Results 2.3.1 Clinical information The relative gene expression levels of four genes – KCNQ1OT1, H19, CDKN1C, and PLAGL1 – involved in an imprinted gene network, were measured in cord blood from 77 neonates (24 IVF, 18 ICSI, 9 IUI, and 26 NC) via RT-qPCR after standardizing against the reference gene UBC. Maternal and pregnancy characteristics were not significantly different across conception modes except for delivery mode (vaginal vs caesarean section) and gestational age (p-value<0.01, p-value=0.04 respectively). Maternal and pregnancy characteristics are summarized Table 6. Differences in delivery mode between conception mode was observed to be driven by significant differences between the ICSI and NC groups (p-value<0.01), while delivery mode was not observed to be significantly different in comparisons between other conception modes. Likewise, significant differences in gestational ages between conception modes was only found in the ICSI vs NC comparison (p-value=0.02).     66  Table 6. Summarized maternal and pregnancy characteristics   IVF ICSI IUI NC Adjusted p-value n = 77 24 18 9 26  Mean Maternal Age  (years ± SD) 36.5 ± 4.7 36.3 ± 3.0 33.8 ± 4.3 34.9 ± 4.2 NS Mean Gestational Age (days ± SD) 274 ± 10 278 ± 8 277 ± 8 271 ± 6 ICSI vs. NC  0.017* Sex Ratio  (boys: girls) 7:17 3:15 6:3 10:16 NS Delivery Mode a  (caesarean: vaginal) 13:8 5:12 4:4 20:5 ICSI vs. NC  <0.01** Mean Birth Weight b (grams ± SD) 3358 ± 579 3688 ± 505 3387 ± 566 3367 ± 527 NS  a Number of participants with delivery mode information: 21 IVF, 17 ICSI, 8 IUI, and 25 NC bNumber of participants with birth weight information: 22 IVF, 16 ICSI, 7 IUI, and 24 NC *Kruskal-Wallis followed by Mann-Whitney and Bonferroni correction was used to calculate pairwise differences **Fisher’s exact test followed by Bonferroni correction was used to calculate pairwise differences  2.3.2 Gene expression analysis of imprinted genes in cord blood from ART babies No significant differences were observed in RQ values for KCNQ1OT1, H19, CDKN1C, and PLAGL1 between conception modes. The p-value across conception modes when comparing RQ values were p-value=0.95 for PLAGL1, p-value=0.09 for KCNQ1OT1, p-value=0.67 for H19, and p-value=0.09 for CDKN1C. Mean log10 RQ values for PLAGL1 in the NC control group was 0.57 with a standard deviation (SD) of ± 0.37 compared to mean log10 RQ values and SD of 0.63±0.58, 0.62±0.56, and 0.72±0.63 for IVF, ICSI, and IUI respectively. Mean log10 RQ values, standard deviations, and p-values for PLAGL1, KCNQ1OT1, H19, and CDKN1C are 67  listed in Table 7. A boxplot of log10RQ for each conception mode, per gene, is depicted in Figure 1.  Table 7. PLAGL1, KCNQ1OT1, H19, and CDKN1C mean log10 RQ values, standard deviations, and p-values   IVF ICSI IUI NC p-value n = 77 24 18 9 26  PLAGL1 (mean log10 RQ ± SD) 0.63 ± 0.58 0.62 ± 0.56 0.72 ± 0.63 0.57 ± 0.37 0.95 KCNQ1OT1 (mean log10 RQ ± SD) 1.06 ± 0.22 0.86 ± 0.38 0.71 ± 0.43 1.09 ± 0.47 0.09 H19 (mean log10 RQ ± SD) 1.86 ± 0.70 2.05 ± 0.82 2.17 ± 0.87 2.11 ± 0.65 0.67 CDKN1C (mean log10 RQ ± SD) 0.73 ± 0.26 0.55 ± 0.31 0.66 ± 0.25 0.81 ± 0.47 0.09  68   Figure 1. Gene expression boxplot. log10RQ values of CDKN1C (red), H19 (green), KCNQ1OT1 (blue), and PLAGL1 (purple) in cord blood from ICSI, IUI, IVF, and NC babies. The horizontal black line in each box represents the median. The upper and lower edges of the box represent the upper and lower quartiles, respectively. The whiskers represent the range to the highest and lowest data points within 1.5 interquartile range from the upper and lower quartiles, respectively.   As delivery mode and gestational age have been linked to gene expression changes, I analyzed if delivery mode and gestational age differences between conception modes may be confounding our results. I performed linear model analyses for each gene to ascertain differences in RQ values between conception modes where delivery mode was used as a covariate, gestational age was used as a covariate, and both delivery mode and gestational age were used as covariates. No model revealed significant differences in RQ values between conception modes. I also analyzed if there was a significant correlation between gestational age and RQ expression 69  values with Spearman’s ranked correlation test, per gene, across all samples. The correlation analysis between gestational age and RQ expression values was only significant for the CDKN1C gene, p-value of 0.03. Although, the Spearman’s rho value of -0.26 in this analysis suggests that the correlation between gestational age and RQ expression in the CDKN1C gene is weak, the analysis shows that higher gestational age is significantly correlated with lower expression of CDKN1C. For the CDKN1C gene, an almost significant p-value of 0.09 was found for the comparison between conception modes. This difference in expression of CDKN1C is driven by differences in ICSI and IVF as well as ICSI and NC comparison where lower CDKN1C expression is observed in the ICSI group. However, gestational age is significantly higher in the ICSI cohort (278 ± 8) than the NC cohort (271 ± 6), p-value of 0.02, suggesting that the differences in gene expression between conception modes for the CDKN1C gene could be driven/confounded by differences in gestational age between conception modes, specifically ICSI and NC cohorts. Further analyses were performed to determine if there were significant differences in gene expression, per gene, across delivery modes. I performed a Mann-Whitney-Wilcoxon test on RQ expression values against delivery mode for each gene. No significant differences in gene expression were found between delivery modes (vaginal versus c-section). Overall, these analyses show that the differences in summary statistics are not impacting the result of no significant differences in gene expression between conception modes. Instead, it appears that gestational age differences may be driving the small, not significant difference, in CDKN1C expression between conception modes.   70  2.3.3 Analysis of an imprinted gene network in cord blood from ART babies As PLAGL1, an imprinted transcription factor, has been shown, in the murine model, to bind to imprinted domains and upregulate gene expression of a network of important growth associated imprinted genes (KCNQ1OT1, H19, CDKN1C, and IGF2), I analyzed how the RQ values of PLAGL1 correlated with RQ values from KCNQ1OT1, H19, and CDKN1C analyses in neonatal cord blood. PLAGL1 RQ values correlated with KCNQ1OT1 RQ values (r=0.26, p-value=0.02) (Figure 2). PLAGL1 RQ values further correlated with H19 RQ values (r= 0.46, p-value<0.01) (Figure 3). Finally, PLAGL1 RQ values also correlated with CDKN1C RQ values (r=0.24, p-value=0.04) (Figure 4).    Figure 2. Correlation between KCNQ1OT1 and PLAGL1 expression in cord blood. Red dots represent ICSI samples, green dots represent IUI samples, blue dots represent IVF samples, and purple dots represent NC samples. 71   Figure 3. Correlation between H19 and PLAGL1 expression in cord blood. Red dots represent ICSI samples, green dots represent IUI samples, blue dots represent IVF samples, and purple dots represent NC samples.  72   Figure 4. Correlation between CDKN1C and PLAGL1 expression in cord blood. Red dots represent ICSI samples, green dots represent IUI samples, blue dots represent IVF samples, and purple dots represent NC samples.  2.4 Discussion PLAGL1 has been found to have altered DNA methylation and expression in cord blood from an ART cohort that included multiple gestation pregnancies and pregnancies with adverse outcomes. As such, I hypothesized that PLAGL1 may have altered gene expression in cord blood when comparing IVF, ICSI, and/or IUI conceived singleton pregnancies with NC singleton controls where no adverse pregnancy outcomes are reported in any group. I further hypothesized that genes implicated in the IGN – KCNQ1OT1, CDKN1C, and H19 – may also be altered in cord blood from ART conceived babies and may correlate with the expression of PLAGL1. The comparison between IVF, ICSI, and IUI also made it possible to potentially identify whether 73  procedures with in vitro culture (IVF and ICSI) may be impacting imprinted genes more than ART procedures without an in vitro culture step (IUI). However, no significant differences in gene expression were found between any conception mode in any of the imprinted genes analyzed.  Although, this investigation did not analyze DNA methylation or the total complement of imprinted genes, the finding of no difference in expression of PLAGL1, KCNQ1OT1, CDKN1C, and H19 between conception modes suggests a preservation of imprinting in cord blood from ART conceived babies. This analysis, therefore, provides support for imprinted gene stability in ART conceived babies, particularly in term pregnancies where no adverse outcomes are apparent, as reported previously (Tierling et al., 2010; Wong et al., 2011; Rancourt et al., 2012; Oliver et al., 2012; Puumela et al., 2012; Camprubí et al., 2013; Nelissen et al., 2013; Sakian et al., 2015; Melamed et al., 2015; Vincent et al., 2016; Castillo-Fernandez et al., 2017). However, slight differences (Tierling et al.,  2010; Rancourt et al., 2012; Camprubí et al., 2013; Nelissen et al., 2013; Vincent et al., 2016) and robust differences (Gomes et al., 2009; Katari et al., 2009; Turan et al., 2010; Song et al., 2015) in imprinted gene DNA methylation and expression have also previously been reported. Previous studies in PLAGL1 imprinting have reported mixed results. Iglesias-Platas et al. (2014) found lower expression of PLAGL1 in term placental villa from a small subgroup of ART conceived babies(n=13). More recently, our group found higher DNA methylation at the PLAGL1 DMR and lower expression of PLAGL1 in cord blood from 66 IVF babies when compared to 56 NC controls (Vincent et al., 2016). However, Camprubí et al. (2013) did not find any differences in DNA methylation or allelic expression of PLAGL1 in either placenta villi or cord blood between 73 ART and 121 NC babies. These discrepancies may be due to the differences in size, ethnic background, and infertilities of the study populations as 74  DNA methylation at imprinted genes have been observed to be associated with the sperm of infertile men. Furthermore, the studies that identified differences in PLAGL1 methylation and or expression were confounded by the inclusion of IUGR, preterm, and multiple gestation pregnancies making it difficult to associate PLAGL1 alterations with ART procedures as PLAGL1 alterations have also been linked to adverse pregnancy outcomes and gestational age (Iglesias-Platas et al., 2014; Moore et al., 2015). It is possible, however, that excluding pregnancies with adverse outcomes may select for individuals that are less susceptible to epigenetic alterations. Indeed Ghosh et al. (2015) identified outlier individuals that exhibited greater alterations in DNA methylation and that these individuals were significantly more common in babies with lower birth weights and children conceived via ART. It would be interesting for future studies to interrogate DNA methylation and gene expression changes of IGN imprinted genes in a controlled comparison between ART and NC low birth weight/IUGR groups.  Although alterations in expression of IGN imprinted genes were not observed, PLAGL1 expression was found to be significantly correlated with KCNQ1OT1, CDKN1C, and H19 expression, supporting previous observations of these genes being regulated together, potentially by PLAGL1 (Arima et al., 2005; Varrault et al., 2006; Iglesias-Platas et al., 2014).   The conclusions of this study are limited by the lack of allele-specific expression and targeted DNA methylation analyses of the genes in the IGN which could provide more insight into the impact ART may have on imprinting at these genes. Furthermore, cell type proportions in the cord blood was not known, which may impact results as different cell type proportions have been shown to influence gene expression in leukocytes (Kong et al., 2019). Although the biological relevance of this study is limited by this analysis being done on cord blood, cord blood 75  represents an accessible tissue where an IGN expression analysis may provide insight into regulation of imprinted genes by more stable markers (DNA methylation) across tissues derived from the inner cell mass and may also serve as an accessible tissue for the use of IGN imprinted genes as biological markers for ART induced growth alterations. A further limitation to this study is the possibility of cord blood contamination with maternal blood which could impact results.   76  Chapter 3: GENOME-WIDE DNA METHYLATION ANALYSIS IN CORD BLOOD FROM IVF AND ICSI CONCEIVED BABIES COMPARED TO NC CONTROLS 3.1 Introduction As the frequency of ART increases world-wide, both the short and long-term health outcomes associated with ART conception are under increased scrutiny and much research has been dedicated to ascertaining the safety of these technologies (Qin et al., 2016). Although contradictory, there appears to be an increase in risk for pregnancy, neonatal, and long-term adverse health outcomes following ART (Qin et al., 2016). Epigenetic mechanisms have become a focus of concern with respect to ART safety, especially DNA methylation, initially due to the link between ART conception and imprinting disorders (Michels, 2012). As such, most of the research has focused on imprinted gene DNA methylation and expression. Although animal studies provide evidence for altered DNA methylation at imprinted genes after ART procedures, investigations in humans have been inconclusive with respect to altered DNA methylation at imprinted genes following ART (Van Montfoort et al., 2012; Lazaraviciute et al., 2014). However, given that both the short term and long-term adverse health outcomes are associated with epigenetic mechanisms, an in-depth assessment of the influence ART has on DNA methylation profiles may elucidate potential underlying mechanisms related to health outcomes are ART (Michels, 2012). Epigenetic mechanisms are thought to be the mediator through which intra-uterine and neonatal life can shape adult and long-term health in the developmental origins of health and disease hypothesis (DOHaD) (Barker & Osmond, 1986; Haugen et al., 2015). Furthermore, DNA methylation dynamics during the epigenetic reprogramming during 77  conception and early embryogenesis occurs in a genome-wide fashion (Van Montfoort et al., 2012). Therefore, ART related interventions that occur simultaneously to this epigenetic reprogramming event could have far reaching genome-wide impacts on DNA methylation; however, few studies have investigated the impact of ART on genome-wide DNA methylation. In a study analyzing 27,000 CpG sites in cord blood from 10 IVF cases and 8 NC controls using the Illumina Infinium HumanMethylation27 bead array, 24 genes with at least two differentially methylated CpG sites were found to be significantly different between IVF and NC controls (Melamed et al., 2015). A later study analyzed blood spots from 18 IUI, 38 ICSI with frozen embryo transfer, 38 ICSI with fresh embryo transfer, and 43 naturally conceived neonates using the more comprehensive Illumina Infinium HumanMethylation450 bead array, investigating DNA methylation at over 450,000 CpG sites (Estill et al., 2016). This study revealed differential methylation profiles between ART conceived and NC controls and found difference in DNA methylation at metastable epialleles – loci where DNA methylation is stochastically established within blastocysts, influenced by peri-conceptional nutritional states. A more recent investigation into genome-wide DNA methylation after ART involved analyzing DNA methylation using genome-wide methylated DNA immunoprecipitation coupled with deep sequencing (MeDIP) in whole cord blood cells and cord mononuclear cells in 47 IVF and 60 NC newborn twins (54 total twin pairs) (Castillo-Fernandez et al., 2017). This investigation found only one region to be significantly different in methylation at a false discovery rate of 5%. However, the Castillo-Fernandez et al. (2017) suggested that infertility associated genes may have altered methylation in ART newborns due to this region being ~3kb upstream of a gene previously linked to male infertility (TNP1); furthermore, of the 46 genes associated with ART at an FDR of 25%, one gene (C9orf3), is also associated with infertility. All prior studies, to the best of my knowledge, 78  interrogating genome-wide DNA methylation alterations in ART conceived populations have focused on identifying differentially methylated positions (DMPs) and differentially methylated regions (DMRs), which correspond to differential methylation between densely clustered regions of CpGs. However, difference in the variation or range of methylation at CpG sites between groups and not necessarily in the mean DNA methylation describes a further epigenetic marker termed differentially variable positions (DVPs) that has not previously been interrogated in ART newborn populations. The importance of DVPs as an epigenetic marker has been revealed in the association of DVPs to type I diabetes, cancers, and rheumatoid arthritis (Paul et al., 2016, Hansen et al., 2011, Webster et al., 2018). In order to build a more complete understanding of how DNA methylation may be altered across the genome, more studies with higher CpG coverage and tightly controlled for confounders is needed. Here, I analyze genome-wide DNA methylation for DMPs, DMRs, and DVPs in IVF, ICSI, and NC cord blood collected at birth using the Illumina Infinium MethylationEPIC bead array, covering more than 850,000 CpG sites across the genome – the most single resolution CpGs analyzed in ART newborns at the time of writing.   3.2 Methods 3.2.1 Study participants Study participants consisted of 29 cases (10 IVF, 9 ICSI, and 10 NC) where IVF and ICSI pregnancies were recruited from several IVF centers across Canada and NC pregnancies were recruited from hospitals in the greater Vancouver area, Canada. Written informed consent was received from pregnant women, those that wished to participate, for the collection of cord blood for research purposes. Cord blood was collected at birth by delivery room staff. Cord 79  blood DNA extraction and statistical analysis was performed at the BC Children’s Hospital Research Institute in Vancouver, British Columbia. Bisulfite conversion and hybridization of the samples to the Illumina Infinium MethylationEPIC array were carried out at McGill University and the Genome Quebec Innovation Centre. The University of British Columbia Research Ethics Board approved this study. IVF, ICSI and naturally conceived pregnancies resulting in live singleton births were included in this study only. Pregnancies complicated by major fetal or pregnancy related anomalies were excluded from the study – pre-eclampsia, very preterm birth, fetal malformation, intra-uterine growth restriction, maternal smoking during pregnancy, and known chromosomal abnormalities were excluded from the analysis. NC pregnancies were also excluded if any sign of infertility was present – fertility drugs were used for conception and/or if the pregnancy took over one year to conceive. Maternal/paternal characteristics and pregnancy history was collected via a self-administered questionnaire. Pregnancy outcome information was collected by delivery room staff or taken from the patients’ medical chart. Collected information included sex of the newborn, maternal age, paternal age, delivery mode (vaginal vs. caesarean section), birth weight, gestational age. Information regarding some of the other particulars of the ART treatment – day of embryo transfer, fresh or frozen embryo transfer, cause of infertility, maternal BMI, hormonal dosage for ovarian stimulation, embryo quality, and pre-implantation genetic diagnosis results – were partially known or not known.   80  3.2.2 DNA extraction from umbilical cord blood Cord blood DNA extraction was carried out using the Qiagen Puregene Blood Core Kit C (Qiagen, Mississauga, ON), according to manufacturer protocol and stored at -20°C. Concentrations were then recorded using a NanoSpec spectrometer (Nanovue).  3.2.3 Illumina Infinium MethylationEPIC BeadChip array analysis Genomic DNA (2.5 µg) was sent to McGill University and Genome Quebec Innovation Centre (Montreal, QB, Canada) for sodium bisulfite modification and hybridization to the Illumina Infinium MethylationEPIC BeadChip array following manufacture protocols.    3.2.4 Data preprocessing Raw IDAT files were imported to and processed in R (version 3.6) using the minfi package (version 1.30.0; Aryee et al., 2014). Probes with low signal quality (detection p-value > 0.05) in more than 10% of samples were filtered out of the analysis (1391 probes). Probes overlapping known SNPs were also removed (15,717 probes). Lastly, 40,914 non-specific probes known to bind to multiple genomic regions were removed from the analysis, leaving a total number of 793,197 probes for subsequent analyses (Pidsley et al., 2016).  Normalization of the type I and type II probes used to interrogate methylation in the EPIC array was done using the stratified quantile normalization method in the minfi package which was developed for analyses where small changes in methylation are expected (Aryee et al., 2014). The stratified quantile normalization method also allows for probes on the X and Y chromosomes to be normalized separately between males and females, allowing for their inclusion in subsequent analyses (Aryee et al., 2014). Annotation for the EPIC array was 81  retrieved from the IlluminaHumanMethylationEPICanno.ilm10b4.hg19 R package using Bioconductor (version 3.9; Hansen, 2016). The annotation uses the Genome Reference Consortium Human Build 37 (GRCh37) or the equivalent hg19 reference genome.   3.2.5 Cell composition estimation Given that DNA methylation varies between cell types, it is imperative to control for cell type compositions in samples undergoing genome-wide DNA methylation analyses (Houseman et al., 2012). The cord blood cell compositions for each newborn were estimated using the estimateCellCounts function in the minfi package implementing the algorithm developed by Houseman et al. (2012). Cell compositions between conception modes were compared by the Wilcoxon-Mann-Whitney test for each cell type and multiple testing was corrected for by the Bonferroni method.  3.2.6 Confounding variable correction In order to control for differences in clinical information that could confound results, a model that included sex, gestational age, delivery mode, and nucleated red blood cell proportions as covariates was implemented in subsequent analyses. To correct for latent variables relating to batch effects and/or differences in cell type proportions, the number of surrogate variables to be included in downstream analysis was estimated using the SVA R package (version 3.32.1; Leek & Storey, 2007). The single surrogate variable identified in this analysis was included as a covariate in following analyses.  82  3.2.7 Differentially methylated position (DMP) analysis To determine the impact of conception mode on DNA methylation at individual CpG sites, differentially methylated position (DMP) analysis, a statistical regression model, linear model with empirical Bayesian statistics (eBayes), was implemented using the limma R package (version 3.40.2; Ritchie et al., 2015). False discovery rates (FDRs) were calculated using the method of Benjamini and Hochberg where an FDR less than 0.05 was considered significant (Benjamini & Hochberg, 1995).  3.2.8 Differentially variable position (DVP) analysis  The impact of conception mode on the range, or variation, at individual CpG sites, differentially variable position (DVP) analysis, was estimated using the DiffVar method from the MissMethyl R package using absolute deviations (version 1.18.0; Phipson et al., 2014). False discovery rates (FDRs) were calculated using the method of Benjamini and Hochberg where an FDR less than 0.05 was considered significant (Benjamini & Hochberg, 1995).  3.2.9 Differentially methylated region (DMR) analysis  The impact of conception mode on clustered groups of CpG sites, differentially methylated region (DMR) analysis, was estimated using the bumphunter and mCSEA methods from the Bumphunter (version 1.26.0; Jaffe et al., 2012) and mCSEA (version 1.4.0; Martorell-Marugan et al., 2018) R packages. The bumphunter algorithm defines clusters of probes as groups of probes where no two probes are separated by more than a set distance. Then the algorithm determines a t-statistic at each probe between groups and identifies candidate differentially methylated regions as clusters of probes where all the t-statistics exceed a set 83  threshold. Significance of the region is determined with permutations or bootstrapping to create a null distribution. The bumphunter algorithm was applied in this study using a 15% difference in methylation threshold and a bootstrap value of 1000. The mCSEA (methylated CpGs Set Enrichment Analysis) method was designed to detect moderate, but consistent methylation changes. This method takes a ranked list of t-statistics assigned to each CpG probe, determined from a linear model, and defines regions of interest (DMRs) as predefined regions associated with promoters, CpG islands (CGI), or gene bodies whose CpG sites are over-represented in the top or bottom of the list of t-statistics. The more CpG sites in the top or bottom of the ranked list of t-statistics, the higher the enrichment score (ES) of the region. Significance of the ES is calculated by permuting the set of CpGs and recomputing the ES in order to get a null distribution for the ES (Martorell-Marugan et al., 2018). Sex chromosome probes were removed prior to the mCSEA analysis due to differences in sex between the conception mode groups; therefore, the mCSEA analysis was performed on 775,301 probes. False discovery rates (FDRs) were calculated using the method of Benjamini and Hochberg where an FDR less than 0.05 was considered significant (Benjamini & Hochberg, 1995).  3.2.10 Metastable epiallele and imprinted gene analysis Positions of 109 high-confidence candidate metastable epialleles characterized by (Silver et al., 2015) were matched to the 775,301 EPIC array probes in this analysis. Of these 109 metastable epiallele regions, 25 were informative and were interrogated by a total of 48 probes. The names of the probes that map to the 50 known imprinted DMRs on the EPIC array, recently characterized by (Hernandez Mora et al., 2018), were matched to the probes in this analysis. 752 probes mapping to these 50 imprinted DMRs were informative in my analysis. 84  3.2.11 Gene enrichment analysis RefGene names associated with the differential methylation between conception modes was analyzed for enrichment in the 2018 Gene Ontology (GO) Biological Processes data base using the enrichR R package (version 1.0; Chen et al., 2013).  3.2.12 Statistical analysis Comparisons between conception modes for clinical information was performed using two-tailed Fisher’s exact test, ANOVA, or Kruskal-Wallis – depending on data type and distribution. The Bonferroni method was used to adjust p-values for multiple comparisons. Correlation analysis were performed with Pearson’ r. A minimum significance level of 5% (95% confidence interval) was used across all analyses. An overview of the analytical pipeline used in this analysis is outlined in Figure 5.         85   Figure 5. Analytical pipeline for EPIC array analysis. Raw data is first imported into R where poor detection probes, probes binding to SNP regions, and non-specific probes are removed. Cell type proportions are then estimated and corrected for alongside other confounding variables. Investigations into differentially variable positions, differentially methylated positions, and differentially methylated regions are performed in parallel. Subsequent hits were subjected to gene enrichment analysis and literature review86  3.3 Results 3.3.1 Clinical information Interrogation of methylation at over 850 thousand CpG sites across the genome with the Illumina EPIC array was conducted on 29 samples – 10 IVF, 9 ICSI, and 10 NC. Maternal and pregnancy characteristics are summarized in Table 8. Maternal age, sex of the newborn, and birth weight were not significantly different between conception modes. Gestational age of the newborns was significantly lower in the NC control group (269 ± 3 days) than in both the IVF (278 ± 5 days, p-value<0.01) and the ICSI (281 ± 7 days, p-value<0.01) groups. Delivery mode was significantly different between the ICSI and NC groups (p-value<0.01) as 100 percent of the naturally conceived newborns were delivered via caesarean sections (n=10) while only 22 percent of the ICSI conceived newborns were delivered via caesarean section (n=2).  The mean proportions of CD8T, CD4T, natural killer, B cell, monocyte, granulocyte, and nucleated red blood cells (nRBCs), estimated from the DNA methylation data obtained from cord blood, in each conception mode is listed in Table 8. Of the cell types analyzed, only nRBCs were observed to be present in significantly different proportions between conception modes. This difference was driven by the difference in estimated proportions of nRBCs between the ICSI (0.17) and NC (0.08) groups (p-value<0.01).   Table 8. Summarized maternal, pregnancy characteristics and cell type proportions   IVF ICSI NC Adjusted p-value n = 29 10 9 10  Mean Maternal Age 34.0 ± 3.7 35.5 ± 2.4 34.1 ± 3.3 NS (years ± SD) Mean Gestational Age 278 ± 5 281 ± 7 269 ± 3 ICSI vs. NC <0.01* 87  (days ± SD)  IVF vs NC <0.01* Sex Ratio 4:6 2:7 6:4 NS (boys: girls) Delivery Mode  6:4 2:7 10:0 ICSI vs. NC (caesarean: vaginal)  <0.01** Mean Birth Weight a 3287 ± 595 3605 ± 337 3381 ± 617 NS (grams ± SD) Cell type and proportions    CD8T 0.12 0.15 0.12 NS CD4T 0.23 0.15 0.22 NS Natural killer <0.01 <0.01 <0.01 NS B cell 0.09 0.11 0.11 NS Monocyte 0.08 0.10 0.10 NS Granulocyte 0.41 0.36 0.43 NS Nucleated red blood cell 0.10 0.17 0.08 ICSI vs. NC <0.01* a Number of participants with delivery mode information: 10 IVF, 8 ICSI, 10 NC *Kruskal-Wallis followed by Mann-Whitney and Bonferroni correction was used to calculate pairwise differences **Fisher’s exact test followed by Bonferroni correction was used to calculate pairwise differences  3.3.2 Differentially methylated position analysis Following the removal of poor detecting probes, SNP containing probes, and cross-reactive probes, 793,197 probes were analyzed for differences in methylation across conception modes. In order to control for differences in clinical information that could confound results, a linear model that included sex, gestational age, delivery mode, and nRBC proportions as covariates was used to identify deferentially methylated positions between conception modes. A surrogate variable is a covariate constructed from the data that controls for unknown or latent sources of noise. A surrogate variable analysis identified one surrogate variable – possibly 88  related to batch effects – which was included as an additional covariate in the linear model. After controlling for multiple comparisons, none of the 793,197 positions interrogated were identified to be differentially methylated between the ICSI group and the NC group using a false discovery rate (FDR) of 0.05. Similarly, no DMPs were identified between the IVF group and the NC group using an FDR of 0.05. In the comparison between ICSI and NC groups, relaxation of the FDR in order to investigate probes at the cost of a higher chance of false positives was not conducted due to the lowest adjusted p-value being 0.86. In the comparison between IVF and NC groups, no probes were found to have an adjusted p-value less than the 0.05 cutoff.  3.3.3 Differentially variable position analysis CpG sites exhibiting differences in the range of DNA methylation values between comparison groups have been identified in type I diabetes, rheumatoid arthritis, and cancers, illustrating that differentially variable positions are a valuable epigenetic marker for disease (Paul et al., 2016, Hansen et al., 2011, Webster et al., 2018). I used the DiffVar method (Phipson & Oshlack, 2014), which allows for a three-group comparison while controlling for confounding variables, to interrogate probes for differences in methylation variation. As in the DMP analysis, sex, gestational age, delivery mode, nRBC proportions, and the surrogate variable were included as confounding variables. In the comparison between the ICSI and NC groups, 49 DVPs were identified at an FDR of less than 0.05. Of these DVPs, 47 were hypervariable, exhibiting greater variance, in the ICSI group. In the comparison between the IVF and NC groups, four DVPs were identified using an FDR of less than 0.05, three of which were hypervariable in the IVF group when compared to the NC group. A full list of the DVPs identified between conception modes with annotations are contained in Table 9. Previous analysis of genome-wide methylation in 89  ART newborns revealed methylation differences in genes that are associated with both male and female infertility, suggesting that methylation changes in children conceived via ART could reflect the infertility of the parents undergoing these procedures (Castillo-Fernandez et al., 2017). An investigation into the DVPs identified between conception modes and associations with infertility revealed that four of the DVPs between the ICSI and NC group comparison were associated with genes that are linked to infertility. The ICSI group was found to be hypervariable in a CpG site located in a CpG island 200-1500 base pairs upstream from the transcriptional start site of ADAMTS16, a gene highly expressed in the ovary that encodes a protein with disintegrin, metalloproteinase, and thrombospondin motifs. Serum ADAMTS16 levels were previously found to be significantly higher in patients with habitual abortions than in control patients, indicating that pregnancy loss rate may be affected by ADAMTS16 expression (Pekcan et al., 2017). Additionally, SNPs in ADAMTS16 have been previously found to be associated with premature ovarian failure (Pyun et al., 2014). Interestingly, abnormal DNA methylation in ADAMTS gene family has previously been found in IVF patients who deliver pre-term (Mani et al., 2018). A CpG site located in the body of the ADAD1 gene, encoding a testis specific RNA binding protein which has been found to be differentially expressed in the spermatozoa of infertile men when compared to fertile men, was found to be hypervariable in the ICSI group when compared to the NC group (Bansal et al., 2015).  An additional gene, MRPL39 – also differentially expressed in the spermatozoa of infertile men when compared to fertile men – was found to contain a hypervariable gene body CpG site in the ICSI group when compared to the NC group (Bansal et al., 2015). The ICSI group was further found to be hypervariable in a CpG site located in the body of NRG1, a gene encoding a membrane glycoprotein that mediates cell to cell signaling, is important in the development of multiple organ systems, and has been linked to 90  both oocyte maturation and spermatogenesis (Umehara et al., 2016). None of the hypervariable CpG sites in the IVF and NC group comparison were found to be linked to infertility. Due to the micromanipulation of sperm in the ICSI procedure as well as the use of ICSI for treatment of both male and female factor infertilities, ICSI has been assumed to have a more adverse impact on the epigenome of resulting newborns than IVF. In keeping with this theory, we observed more hypervariable CpG positions when comparing the ICSI group of newborns to the NC group than when comparing the IVF group of newborns to the NC group. This difference in CpG site variability is illustrated in the following volcano plots depicting the variability of all the probes compared between the ICSI and NC groups (Figure 6) and the IVF and NC groups (Figure 7).  Table 9. Differential variable positions between ICSI and NC, IVF and NC comparisons at an FDR < 0.05 Probe Log2 variance ratio Adjusted p-value Hypervariable group Chromosome Relationship to gene Gene  cg09208162 4.902 0.0009 IVF chr2 3'UTR NDUFA10 cg18315527 3.180 0.0036 ICSI chr17 5'UTR ARMC7 cg27636047 5.295 0.0036 ICSI chr4   cg22479546 3.442 0.0047 IVF chr12 TSS1500 LASS5 cg16210302 3.514 0.0063 IVF chr4 Body WHSC1 cg09526912 2.643 0.0190 ICSI chr2 TSS1500 SIX2 cg17944161 2.785 0.0190 ICSI chr19 TSS200 NOTCH3 cg17139861 2.683 0.0190 ICSI chr4 Body ADAD1 cg13850625 3.761 0.0190 ICSI chr4 5'UTR TMEM144 cg02633079 2.518 0.0210 ICSI chr8 Body NRG1 cg02858606 1.760 0.0210 ICSI chr16 5'UTR CDH8 cg12055680 2.598 0.0210 ICSI chr14   cg26454563 2.241 0.0210 ICSI chr6 TSS1500 SLC17A3 cg25033364 1.855 0.0210 ICSI chr11 Body KCNC1 cg04525508 2.860 0.0210 ICSI chr3 TSS200 HEG1 cg06912990 2.607 0.0213 ICSI chr4 5'UTR TMEM144 cg03969957 3.489 0.0213 ICSI chr10 TSS1500 C10orf4 cg16744168 2.646 0.0218 ICSI chr17 TSS1500 PPP1R1B cg22039458 2.176 0.0252 ICSI chr2 TSS1500 HTRA2 91  cg09520904 -4.587 0.0258 NC(IVF-NC) chr11 Body CCND1 cg20817150 2.394 0.0327 ICSI chr21 Body IFNGR2 cg21199852 2.021 0.0359 ICSI chr3   cg20517123 2.773 0.0373 ICSI chr16   cg04700648 3.320 0.0373 ICSI chr19 3'UTR ZNF880 cg10188349 2.921 0.0373 ICSI chr19   cg13708497 2.052 0.0386 ICSI chr14 TSS1500 KIF26A cg25654242 1.753 0.0386 ICSI chr21 Body MRPL39 cg14040899 3.231 0.0393 ICSI chr10 Body ADARB2 cg24705286 2.101 0.0393 ICSI chr16 5'UTR JMJD5 cg07834743 2.778 0.0406 ICSI chr5 TSS1500 ADAMTS16 cg14215625 2.915 0.0415 ICSI chr15 TSS1500 DYX1C1 cg13512607 3.395 0.0415 ICSI chr11 Body OPCML cg26337430 2.525 0.0415 ICSI chr1 TSS1500 LEPRE1 cg18449964 2.011 0.0427 ICSI chr18 Body ZADH2 cg11287055 3.761 0.0427 ICSI chr21 Body DSCR3 cg00143623 1.278 0.0427 ICSI chr1 5'UTR SLC6A9 cg04337611 1.985 0.0427 ICSI chr4 TSS200 DDX60L cg20091620 2.764 0.0427 ICSI chr10 TSS1500 C10orf4 cg22294740 2.618 0.0427 ICSI chr19 5'UTR LINGO3 cg20959920 1.395 0.0461 ICSI chr17   cg26115987 3.016 0.0461 ICSI chr3 Body MUC13 cg01923218 2.615 0.0461 ICSI chr11 5'UTR CCDC67 cg11366788 1.572 0.0461 ICSI chr12 TSS1500 NDUFA12 cg20475917 1.913 0.0461 ICSI chr10   cg02702481 2.598 0.0461 ICSI chr16 Body SPIRE2 cg21808053 2.358 0.0461 ICSI chr1 5'UTR DIRAS3 cg10998570 2.515 0.0461 ICSI chr5   cg19789919 -4.482 0.0461 NC(ICSI-NC) chr12 TSS200 ATF7IP cg08345402 1.419 0.0463 ICSI chr6 5'UTR NRSN1 cg02558623 1.430 0.0497 ICSI chr15   cg09520904 -3.803 0.0497 NC(ICSI-NC) chr11 Body CCND1 cg13050884 1.738 0.0497 ICSI chr3 Body FNDC3B cg05218283 2.788 0.0497 ICSI chr7 Body MKLN1  92   Figure 6. Volcano plot of DNA methylation variance between ICSI and NC in cord blood. -log10 p-value and log2 variance ratio for 793,197 CpG sites. Blue dots represent CpG sites that 1) exhibit greater than 3.5 and less than -3.5 log2 variance ratios between ICSI and NC groups and 2) adjusted p-values less than 0.05. The dotted blue horizontal line is set at a -log10 p-value of 5.49 which corresponds to an adjusted p-value of 0.05. Red dots represent CpG sites that exhibit less than 3.5 and greater than -3.5 log2 variance ratios between ICSI and NC groups. 93   Figure 7. Volcano plot of DNA methylation variance between IVF and NC in cord blood. -log10 p-value and log2 variance ratio for 793,197 CpG sites. Blue dots represent CpG sites that 1) exhibit greater than 3.5 and less than -3.5 log2 variance ratios between IVF and NC groups and 2) adjusted p-values less than 0.05. The dotted blue horizontal line is set at a -log10 p-value of 6.80 which corresponds to an adjusted p-value of 0.05. Red dots represent CpG sites that exhibit less than 3.5 and greater than -3.5 log2 variance ratios between IVF and NC groups.  3.3.4 Metastable epiallele and imprinted gene analysis of DVPs DNA methylation at metastable epialleles and imprinted genes are established around conception, have previously been shown to be altered by periconception environments, and are stable across mitosis and tissues. These characteristics allow for investigating the impact periconception stressors may have on the epigenome using measurements in peripheral tissues.  Methylation at metastable epialleles and imprinted genes have previously been found to be altered in ART groups (Estill et al., 2016). Therefore, I analyzed the list of DVPs found between 94  the ART and NC groups with the aim of determining if the DVPs were located in any of the 109 high-confidence metastable epialleles or located in imprinted gene DMRs. Out of the 109 high-confidence metastable epialleles characterized by Silver et al. (2015), 25 were informative in my analysis. None of the DVPs in either of the conception mode comparisons mapped to the 48 probes that interrogated CpG methylation at these 25 metastable epialleles. Of the 789 EPIC array probes recently identified to coincide with 50 imprinted DMRs by Hernandez Mora et al. (2018), 752 probes mapping to these 50 imprinted DMRs were informative in my analysis. One of the hypervariable DVPs in the ICSI group was mapped to the imprinted DMR of the gene DIRAS3.  DIRAS3 is a paternally expressed imprinted gene that inhibits growth and inhibits lactation when overexpressed in the murine model (Xu et al., 2000).   3.3.5 Differentially methylated region analysis In order to determine whether DNA methylation was altered at densely clustered regions of CpG sites (DMRs), the well-established bumphunter method from the minfi R package was used. At an FDR of less than 0.05, no significant DMRs were found between either of the conception mode comparisons, suggesting a stability across genome-wide CpG dense regions in ART conceived newborns or subtle, small changes in methylation between groups that bumphunter may not be sensitive enough to detect with the sample sizes in this analysis. A novel method for identifying DMRs, called mCSEA, that implements a Gene Set Enrichment Analysis Approach has been developed to detecting subtle, but consistent, methylation differences between groups.  In a methylation dataset of child sibling pairs, where one sibling had been exposed to maternal diabetes during gestation while the other sibling was not, expected subtle differences in DMR methylation were not detected by DMR detecting methods, including 95  bumphunter (Martorell-Marugan et al., 2018). The developers of mCSEA showed that their mCSEA method was able to detect 1055 significant DMRs in this dataset which were enriched in pathways related to onset of diabetes in the young (Martorell-Marugan et al., 2018). Using the mCSEA method while controlling for sex, gestational age, delivery mode, nRBC proportions, and the surrogate variable, I detected 101 DMRs associated with gene promoter regions between the ICSI and NC groups at an FDR of less than 0.05 (Appendix A.1). Similarly, 101 DMRs were found between the IVF and NC groups at an FDR of less than 0.05 (Appendix A.2). The top 50 promoter DMRs in both the IVF and ICSI comparisons, sorted by percent difference in the mean methylation of the leading edge CpGs (the set of CpGs that contribute to the enrichment score of the region) for each comparison, are listed in Table 10. Interestingly, 35 promoter DMR associated genes overlapped between the ICSI and IVF groups, suggesting potential regions that may be sensitive to ART procedures (Appendix A.3). In both the ICSI and IVF comparisons with the NC group, no DMRs were detected in gene body or in CGI regions at an FDR of less than 0.05. However, relaxing the FDR in the IVF and NC comparison revealed 52 gene body DMRs (FDR < 0.068) and 78 DMRs in CGI regions (FDR < 0.065). Relaxing the FDR in the ICSI and NC group comparison resulted in 60 gene body DMRs (FDR < 0.057) and 175 DMRs in CGI regions (FDR <0.059) (Table 11).  Table 10. Top 50 promoter DMRs between conception modes ranked by percent methylation difference Comparison NES DMR #CpG Chromosome Adjusted p-value Beta difference Beta % difference Gene IVF-NC -2.259 16 chr3 0.039 -0.059 -47.394 MCCC1 IVF-NC -2.464 14 chr17 0.039 -0.111 -43.885 C17orf97 IVF-NC -2.242 16 chr3 0.039 -0.055 -36.042 PLSCR1 IVF-NC -2.053 5 chr1 0.039 -0.111 -25.662 GSTM5 IVF-NC -2.228 12 chr1 0.039 -0.104 -22.633 PM20D1 96  IVF-NC -2.049 7 chr6 0.039 -0.075 -22.526 SNORD52 IVF-NC -2.183 11 chr3 0.039 -0.031 -19.369 STAC IVF-NC -2.121 14 chr4 0.039 -0.070 -18.623 TRAM1L1 IVF-NC 2.066 13 chr6 0.039 0.050 16.672 HCG4P6 IVF-NC -2.214 7 chr1 0.039 -0.051 -14.053 AKR7L IVF-NC -2.456 15 chr19 0.039 -0.038 -14.002 ZNF577 IVF-NC -2.160 9 chr1 0.039 -0.055 -13.622 C1orf173 IVF-NC -2.223 15 chr7 0.039 -0.024 -12.499 NPY IVF-NC 2.194 30 chr6 0.039 0.017 12.367 RNF5 IVF-NC -2.248 17 chr19 0.039 -0.016 -12.343 ZNF350 IVF-NC 1.889 5 chr19 0.039 0.072 12.225 C3 IVF-NC 2.176 22 chr8 0.039 0.033 11.897 TP53INP1 IVF-NC -2.004 8 chr12 0.039 -0.034 -10.910 TMBIM4 IVF-NC -2.173 9 chr5 0.039 -0.036 -10.822 LOC134466 IVF-NC 2.042 12 chr8 0.039 0.040 10.747 HTRA4 IVF-NC 2.073 9 chr15 0.039 0.025 10.519 C15orf26 IVF-NC -2.332 9 chr17 0.039 -0.041 -10.267 SKAP1 IVF-NC 2.278 16 chr5 0.039 0.023 9.355 CCNG1 IVF-NC 2.136 31 chr6 0.039 0.015 9.293 RNF5P1 IVF-NC -2.089 13 chr2 0.039 -0.024 -9.015 CASP10 IVF-NC -2.042 15 chr2 0.039 -0.017 -8.871 DUSP19 IVF-NC -2.074 6 chr21 0.039 -0.057 -8.310 ABCC13 IVF-NC -1.985 8 chr16 0.039 -0.030 -7.990 IL17C ICSI-NC -2.046 5 chr1 0.040 -0.157 -38.298 GSTM5 ICSI-NC 2.946 19 chr10 0.040 0.049 36.019 C10orf4 ICSI-NC -2.017 7 chr6 0.040 -0.071 -21.407 SNORD52 ICSI-NC -2.324 14 chr17 0.040 -0.083 -21.355 C17orf97 ICSI-NC -2.233 7 chr1 0.040 -0.062 -17.494 AKR7L ICSI-NC -2.293 15 chr7 0.040 -0.030 -15.098 NPY ICSI-NC -2.079 14 chr4 0.040 -0.050 -13.812 TRAM1L1 ICSI-NC -2.109 39 chr8 0.040 -0.051 -13.357 FDFT1 ICSI-NC -2.172 7 chr1 0.040 -0.020 -11.711 AKR7A3 ICSI-NC -2.438 13 chr11 0.040 -0.031 -10.989 NAALAD2 ICSI-NC 2.004 11 chr16 0.040 0.010 10.915 CDR2 ICSI-NC -2.267 9 chr18 0.040 -0.051 -9.988 LOC100130522 ICSI-NC 1.958 6 chr1 0.040 -0.044 -9.210 GPR88 ICSI-NC -2.432 15 chr19 0.040 -0.025 -9.159 ZNF577 ICSI-NC 1.926 8 chr22 0.040 0.012 9.105 DGCR6L ICSI-NC 2.030 13 chr5 0.040 0.026 8.721 PCDHGB3 ICSI-NC -2.329 15 chr5 0.040 -0.024 -8.688 LRRC14B ICSI-NC -2.289 11 chr5 0.040 -0.057 -8.551 SLC6A19 ICSI-NC -2.004 9 chr4 0.040 -0.028 -8.475 C4orf49 ICSI-NC 2.113 34 chr13 0.040 0.016 8.208 CCNA1 97  ICSI-NC -2.103 9 chr17 0.040 -0.033 -8.067 SKAP1 ICSI-NC 2.115 10 chr4 0.040 0.033 7.837 TRIM61 Comparison: ICSI and NC comparison (ICSI-NC). IVF and NC comparison (IVF-NC) NES: normalized enrichment score of each DMR DMR # CpG: The number of CpG sites in the DMR Beta difference: The difference in mean DNA methylation (beta values) between the ICSI or IVF groups and the NC groups. Positive value corresponds to hypermethylation while a negative value corresponds to hypomethylation in the ICSI or IVF group. Beta % difference: The percent difference in mean DNA methylation (beta values)  Table 11. Number of hits in the promoter, gene body, and CGI analyses at specified FDR values Conception mode comparison Promoter Gene body CpG Island IVF-NC 101a 0a ; 52b 0 a ; 78c  ICSI-NC 101a 0a ; 60d  0 a ; 175e a Benjamini-Hochberg (BH) adjusted p-value < 0.05 bBH adjusted p-value < 0.068  cBH adjusted p-value < 0.065   dBH adjusted p-value < 0.057  eBH adjusted p-value < 0.059    3.3.6 Metastable epiallele and imprinted gene analysis of DMRs None of the significant DMRs in promoter regions of either the ICSI or IVF group comparisons overlapped with the 25 metastable epialleles that were informative in this analysis. However, in the CGI analysis of both the IVF and ICSI groups, a DMR were detected that included CpGs that mapped to the metastable epiallele on chromosome 7 (chr7:4305001-4305200) associated with the SDK1 gene. The SDK1 associated CpG island DMR consisted of 25 CpGs covering 1983 base pairs (chr7: 4303079-4305062) and included both CpGs informative to the metastable epiallele in this region. This DMR was found to be hypomethylated in both the ICSI (2.00 % methylation difference, FDR of 0.06) and the IVF (2.82 % methylation difference, FDR of 0.06) groups.  98   Four of the significant DMRs associated with promoters in the ICSI group were found to overlap with the imprinted DMRs of NNAT, BLCAP, ZNF597, and H19. In the IVF group, promoter associated DMRs were also found to overlap with the imprinted DMRs of NNAT, and H19. The annotation of these DMRs overlapping with imprinted DMRs are in Table 12.  Table 12. Promoter associated DMRs overlapping with imprinted gene DMR Comparison NES DMR #CpG DMR location Adjusted p-value Beta difference Beta % difference Gene ICSI-NC 2.326 36 chr11: 2019116-2020560 0.040 0.025 4.072 H19 ICSI-NC 2.165 14 chr16: 3493614-3494094 0.040 0.014 3.048 ZNF597 ICSI-NC -1.996 105 chr20: 36148154-36157266 0.040 -0.006 -1.184 BLCAP ICSI-NC -2.692 41 chr20: 36148154-36149750 0.040 -0.010 -1.494 NNAT IVF-NC 2.492 36 chr11: 2019129-2020560 0.039 0.020 3.370 H19 IVF-NC -2.006 41 chr20: 36148133-36149750 0.039 -0.002 -0.308 NNAT Comparison: ICSI and NC comparison (ICSI-NC). IVF and NC comparison (IVF-NC) NES: normalized enrichment score of each DMR DMR # CpG: The number of CpG sites in the DMR Beta difference: The difference in mean DNA methylation (beta values) between the ICSI or IVF groups and the NC groups. Positive value corresponds to hypermethylation while a negative value corresponds to hypomethylation in the ICSI or IVF group. Beta % difference: The percent difference in mean DNA methylation (beta values)  3.3.7 Gene enrichment analysis For the gene sets identified to be differentially methylated between conception modes in this study, a gene enrichment analysis was performed using Enrichr and the 2018 Gene Ontology (GO) Biological Process data base. Of the 49 probes determined to be significantly variable between ICSI and NC groups in the DifVar analysis, 40 mapped to known RefGene names. These 40 genes were not found to be significantly enriched in known GO biological processes. 99  Similarly, no significant enrichment of genes in biological processes was observed in the IVF or ICSI promoter DMR gene sets of 101 genes, respectively. I then combined the gene sets of DMRs at promoters, CGIs, and gene bodies in both the ICSI and IVF mCSEA analyses into a total gene set of 516 genes. An enrichment analysis on this set of 516 genes revealed an almost significant (adjusted p-value of 0.07) association with the GO biological process, nervous system development (GO:0007399), with 21 out of the 456 genes associated with this GO term in my total combined gene set. Of the 21 genes linked to the nervous system development GO term, six were uniquely associated with DMRs found in the IVF group analyses, 14 were uniquely associated with DMRs found in the ICSI group analyses, and one (SDK1) was associated with DMRs found in both the ICSI and IVF analyses. The top eight DMRs, with respect to percent difference in methylation, are listed in Table 13.     Table 13. Top DMRs in the nervous system development GO term (GO:0007399) Comparison NES DMR #CpG Chromosome Adjusted p-value Beta difference Beta % difference Gene ICSI-NC  promoter 2.030 13 chr5 0.040 0.026 8.721 PCDHGB3 ICSI-NC promoter 2.287 16 chr5 0.040 0.025 7.379 PCDHGB1 ICS-NC  CGI 2.278 16 chr5 0.057 0.025 7.379 PCDHGB1 ICSI-NC  CGI 2.073 12 chr5 0.057 0.021 6.733 PCDHGB2 IVF-NC  gene body -2.264 32 chr1 0.068 -0.027 -5.555 NBL1 IVF-NC  gene body -2.079 13 chr9 0.039 -0.013 -5.465 BARHL1 100  IVF-NC  gene body -2.287 14 chr5 0.039 -0.024 -5.096 PCDHB11 IVF-NC  gene body -2.127 36 chr1 0.039 -0.026 -5.019 NBL1 Comparison: ICSI and NC comparison (ICSI-NC). IVF and NC comparison (IVF-NC). Promoter, CGI, or gene body analysis. NES: normalized enrichment score of each DMR DMR # CpG: The number of CpG sites in the DMR Beta difference: The difference in mean DNA methylation (beta values) between the ICSI or IVF groups and the NC groups. Positive value corresponds to hypermethylation while a negative value corresponds to hypomethylation in the ICSI or IVF group. Beta % difference: The percent difference in mean DNA methylation (beta values)  3.3.8 DNA methylation and expression of H19 Of the 29 IVF, ICSI, and NC newborns included in the genome-wide DNA methylation analysis, 28 (9 IVF, 9 ICSI, and 10 NC) were also included in the imprinted gene expression analysis in chapter 2. As both ICSI and IVF groups were found to be hypermethylated in DMRs associated with H19, I investigated whether H19 gene expression may be correlated with mean methylation of the 18 CpG sites in the leading edge of the DMR found between the ICSI and NC groups, and the 18 CpG sites (one CpG site different from the ICSI DMR) in the DMR found between the IVF and NC groups. H19 is an imprinted gene, maternally expressed, that is growth restrictive (Tycko & Morison, 2002). DNA methylation at the imprinted DMR regulates the expression of H19, where methylation at the DMR, as seen on the paternal allele, silences the expression of H19 (Sasaki et al., 2000).  Both DMRs associated with H19 found between conception modes overlapped with the imprinted DMR at the H19/IGF2 locus. Pearson’s correlation analysis revealed that mean methylation of the H19 associated DMR found to be hypermethylated (4.1 % higher mean methylation than NC) in the ICSI group and expression of H19 in the subgroup of 28 newborns was significantly correlated (r = -0.41, p-value = 0.03), as seen in Figure 8, where higher levels of methylation correlated with lower levels of expression. 101  Mean methylation of the H19 DMR found to be hypermethylated in the IVF group (3,4% higher mean methylation) was also correlated with H19 gene expression in this subgroup (r = -0.44, p-value = 0.02). A further analysis (Kruskal-Wallis followed by Mann-Whitney test) into the expression levels of  H19 between IVF, ICSI, and NC newborns in this subgroup revealed a slight, but not significant, increase in mean expression log10RQ values between the IVF and NC group (log10RQ = 1.88 and 1.80 respectively; p-value=.28). An almost significant increase in log10RQ values was observed in the ICSI group when compared to the NC group (log10RQ = 1.97 and 1.80, respectively; p-value= 0.05). In contrast to the model of epigenetic regulation of H19 where higher DNA methylation at the imprinted DMR leads to silencing, hypermethylation in DMRs associated with H19 in both the ICSI and IVF groups was not associated with lower expression of H19 in the IVF and ICSI groups when compared to the NC group. Instead, slightly higher expression of H19 was observed in the IVF and ICSI group, suggesting the difference in mean methylation at the H19 associated DMRs observed in the IVF and ICSI groups does not impact the maintenance of imprinting in this region. Indeed, previous reports indicate that mean methylation levels between 30% and 60% are considered in line with imprinting maintenance (Rancourt et al., 2012).   102   Figure 8. Correlation between H19 methylation as measured by mean beta values and H19 expression as measured by log10RQ values in cord blood. Red dots represent ICSI samples, green dots represent IVF samples, and blue dots represent NC samples.    3.4  Discussion Since ART procedures are carried out during an important period of epigenetic reprogramming in early development, I hypothesized that ART procedures may induce DNA methylation alterations that persist to birth, potentially correlating with ontologies related to negative outcomes associated with ART. Therefore, in this analysis, I interrogated DNA methylation at 793,197 CpG sites throughout the human genome in cord blood from 10 IVF, 9 ICSI, and 10 NC newborns with the Illumina Infinium MethylationEPIC BeadChip array.  No significant differences in mean DNA methylation was observed at individual CpG sites between ICSI and NC as well as between IVF and NC newborns, suggesting an overall 103  stability of DNA methylation following ART and supporting previous reports of stable DNA methylation in ART conceived newborns (Tierling et al., 2010; Wong et al., 2011; Rancourt et al., 2012; Oliver et al., 2012; Puumela et al., 2012; Camprubí et al., 2013; Nelissen et al., 2013; Sakian et al., 2015; Melamed et al., 2015; Vincent et al., 2016; Castillo-Fernandez et al., 2017). However, significant differences in the variance, or range, of DNA methylation was found at a small number of CpG sites between the ART groups and the NC controls. Although the 49 CpG sites that were found to have differences in variance between the ICSI and NC groups represent a small fraction of the total CpG sites interrogated (0.006%), 47 of these CpGs were hypervariable (higher variance) in the ICSI group. Similarly, three of the four CpG sites found to have significant differences in variance between the IVF and NC groups were hypervariable in the IVF group. This increased variance in methylation in the ART groups could represent a stochastic, or random, model of DNA methylation alterations following ART where CpG sites could be more labile to DNA methylation alterations in either direction (hypomethylation or hypermethylation), leading to potentially no differences in mean DNA methylation between groups. An increase in variance in DNA methylation in ART conceived newborns has previously been reported in a genome-wide analysis of cord blood using the Illumina 450k array as well as an analysis of several imprinted loci in humans, providing support for stochastic alterations in ART newborns (Turan et al., 2010; Melamed et al., 2015). The higher number of hypervariable CpG sites found in the ICSI comparison than in the IVF comparison implies that ICSI may have a greater impact on genome-wide methylation than IVF as suggested by El Hajj et al. (2017). However, it is unclear whether the more invasive ICSI procedure itself or the potential increased severity of infertility in patients undergoing ICSI may be related to this finding. Indeed, 4 of the 49 hypervariable sites identified in the ICSI group were found in infertility associated genes, 104  suggesting that the increase in range of DNA methylation found in the ICSI group may, at least in part, be linked to infertility.  Gene expression is usually regulated by DNA methylation of densely clustered CpG regions termed CpG Islands, therefore interrogation into regional DNA methylation alterations in ART groups was performed. The bumphunter approach, which defines de novo DMRs as clusters of probes where no two probes are separated by more than a set distance and exhibit differential methylation between groups, did not reveal any significant differences in regional DNA methylation between conception modes, perhaps due to the small sample sizes in this study. However, the bumphunter method may fail to detect DMRs where there are small, but consistent changes in DNA methylation across groups as seen in an analysis of discordant twins for gestational diabetes (Martorell-Marugan et al., 2018). Using the novel mCSEA method for DMR detection designed to identify moderate to small, but consistent changes in DNA methylation using a Gene Enrichment Analysis approach and predefined regions of clustered CpGs, I identified 101 promoter associated DMRs in the ICSI and NC comparison as well as 101 promoter associated DMRs in the IVF and NC comparison. Hypomethylation was observed in 55% of the ICSI DMRs and 59% of the IVF DMRs. Furthermore, of the 15 DMRs in promoter regions identified to have a 15% difference in DNA methylation between the ART groups and the NC groups, 13 exhibited hypomethylation.  Although both hypomethylated and hypermethylated DMRs were observed in promoter regions, a trend towards hypomethylation may suggest a more targeted effect of ART on DNA methylation, perhaps tangential to the stochastic impact of ART on DNA methylation, and has previously been observed in another genome-wide investigation in cord blood from ART newborns (Melamed et al., 2015). In further support of a targeted effect of ART on DNA methylation, 35 of the significant promoter DMRs 105  overlapped between the ICSI and IVF comparisons. However, the significance of these potentially targeted regions is unclear as gene enrichment analysis on the genes associated with these overlapping DMRs did not reveal any significantly enriched biological processes, cellular components, or molecular functions.  Instead, a gene enrichment analysis on the total number of genes, 516 genes, found to be associated with DMRs in promoter (FDR < 0.05), CGI (FDR < 0.07), and gene bodies (FDR < 0.07) of both IVF and ICSI groups revealed an almost significant (p-value = 0.07) enrichment for genes involved in nervous system development (GO:0007399). Interestingly, the association with DMRs between conception modes and the nervous system development GO biological process seems to be driven mostly by DMRs identified in the ICSI group at promoter, CpG island, and gene body regions of protocadherin (PCDH) genes – a novel finding in ART studies. PCDHs are a group of cell to cell adhesion proteins that are mostly expressed in the nervous system, belong to the cadherin family – transmembrane proteins important in binding cells together – and are important in neural circuit formation during development (Weiner & Jontes, 2013). PCDHs are divided into clustered and non-clustered PCDHs, where clustered PCDHs are organized in the chromosome 5q31 region containing PCDHA, PCDHB, and PCDHG genes that are important in generating unique single cell identifiers for self-recognition in the nervous system (El Hajj, Dittrich et al., 2017; Lomvardas & Maniatis, 2016). In the DMR analysis with the mCSEA method, I identified 3 promoter associated, 9 gene body associated, and 3 CGI associated DMRs in the ICSI group, all of which were hypermethylated, that overlapped with clustered PCDH genes PCDHGA1, PCDHGA2, PCDHGA3, PCDHGA4, PCDHGA5, PCDHGA6, PCDHGB1, PCDHGB2, PCDHGB3. A hypomethylated promoter associated DMR in the IVF group overlapped with the non-clustered PCDH gene PCDHB11 located on chromosome 5. Epigenetic 106  alterations at PCDH genes has been associated with child maltreatment, prenatal alcohol exposure, the neurodevelopmental disorder Rett Syndrome, and is thought to potentially play a role in autism spectrum disorders (El Hajj, Dittrich et al., 2017). It is interesting to speculate if epigenetic alterations at these genes may provide insight into the higher observed odds ratios for nervous system congenital malformations as well as an increased risk for autism spectrum disorders in ART populations, however, further validation of DNA methylation alteration in these regions are required (Hooersan et al., 2017; Gao et al., 2017).  Of the 15 DMRs in promoter regions identified to have a greater than 15% DNA methylation difference between conception modes, 12 were associated with unique genes as 3 overlapped in the ICSI and IVF analysis (SNORD52, C17orf97, GSTM5). Although a gene enrichment analysis did not reveal any significant biological processes associated with these 12 genes, one the genes – GSTM5 – has been shown to have altered DNA methylation in peripheral blood from infertile men (Sarkar et al., 2019), suggesting that DNA methylation alterations in ART newborns may be linked, at least in part, to infertility of the parents. Indeed, of the 49 hypervariable CpGs in the ICSI group, four were found to be associated with infertility. This supports a previous finding of altered DNA methylation in genes associated with infertility in ART newborns in another genome-wide DNA methylation analysis where the authors suggest that an epigenetic signature influencing DNA methylation could be transmitted from parent to offspring (Castillo-Fernandez et al., 2017). Castillo-Fernandez et al. (2017) went on to estimate that the heritability of a DMR associated with an infertility gene – C9orf3 – was around 25%, although the mechanism of transmission is unclear.  However, two of the genes associated with altered promoter DNA methylation in ART groups – MCCC1 and NPY – may be linked to ART procedures impacting epigenetic 107  reprogramming. MCCC1, the gene with the largest percent difference in methylation between conception modes (percent differences of 47.4%, hypomethylated in the IVF group), has recently been shown to be a placental specific imprinted gene with a novel mechanism of imprint that occurs simultaneously with ART procedures (Tayama et al., 2014).  MCCC1 encodes the alpha subunit of the 3-MCC carboxylase enzyme important in protein catabolism and has been shown to be imprinted in the placenta, where it is maternally expressed and the paternal allele is hypermethylated (Yuen et al., 2011; Tayama et al., 2014). Unlike well-characterized imprinted genes where DNA methylation levels at each parental allele is conferred by epigenetic marks in the gametes and maintained during epigenetic reprogramming, placental-specific DMRs do not inherit methylation from gametes and are instead devoid of methylation in human embryonic stem cells (Tayama et al., 2014). As such, the authors proposed a novel imprinting mechanism that may involve histone modifications or transcription factor recruitment during trophoblast differentiation and speculate that these mechanisms may therefore be particularly sensitive to ART procedures as the first differentiation step that results in the trophectoderm occurs when blastocysts are in in vitro culture (Tayama et al., 2014). However, only one of the 17 identified placental-specific imprinted genes was found to be associated with DMRs in ART groups in this analysis. Furthermore, it is not clear how alterations in cord blood methylation, tissue derived from the inner cell mass, may relate to alterations in the epigenetic machinery necessary to instill imprinted DNA methylation during trophoblast differentiation. It would be interesting to investigate the placental DNA methylation of these genes following ART. Another of the genes observed to be associated with promoter DMR hypomethylation in the ICSI group, NPY – encoding a neuropeptide that is widely expressed in the central nervous system and influences many physiological processes including the stress response and appetite regulation – is linked to 108  the intrauterine programming of adult onset obesity and metabolic disorders (Gali Ramamoorthy et al., 2015; Crujeiras et al., 2013). In rodents, maternal undernourishment is associated with increased hypothalamic NPY expression (Shin et al., 2012). Similarly, the neonatal feeding of rats with high-carbohydrate formula milk resulted in increased hypothalamic expression of NPY which also correlated with hypomethylation of the promoter region of NPY in hypothalamic tissues (Mahmood et al., 2013). As such, NPY regulation through methylation, along with other hypothalamic appetite regulators, has been thought to potentially underlie the developmental programming for obesity in rat neonates subjected to a high-carbohydrate diets (Mahmood et al., 2013).  In humans, hypomethylation at the promoter region of the obesity associated gene NPY has been observed in men who regained weight after dieting when compared to obese men who did not regain weight after dieting (Crujeiras et al., 2013). Although further validation of DNA methylation alteration at this region in ART newborns is necessary, my finding of hypomethylation (15 % lower methylation) at promoter DMR associated with NPY in the ICSI group may be linked to epigenetic reprogramming following altered nutritional conditions of the early embryo during ART procedures which may have long-term implications for the resulting children and may help explain the increase in total body fat compositions found in ART groups  increase in total body fat composition (Ceelen et al., 2007).  As imprinted genes are associated with fetal and placental growth and development, are regulated by gametic DNA methylation that is maintained during early embryonic epigenetic reprogramming, and are associated with imprinting disorders (potentially linked to ART conceived children), DNA methylation at imprinted gene DMRs have been investigated in ART groups by several studies. However, there has been mixed evidence on the effect of ART on imprinted genes with some studies reporting DNA methylation alterations at imprinted genes 109  (Gomes et al., 2009; Katari et al., 2009; Turan et al., 2010; Song et al., 2015) while others have reported little or no alterations in DNA methylation at imprinted genes following ART (Tierling et al., 2010; Wong et al., 2011; Rancourt et al., 2012; Oliver et al., 2012; Puumela et al., 2012; Camprubí et al., 2013; Nelissen et al., 2013; Sakian et al., 2015; Melamed et al., 2015; Vincent et al., 2016). This analysis did not reveal an overall destabilization of DNA methylation at imprinted DMRs, but instead, that specific imprinted DMRs – DIRAS3, NNAT, BLCAP, ZNF597, H19 – are hypervariable or overlapped with regions found to be DMRs between conception modes, suggesting small but significant alterations may be associated with ART conception.  The significance of these changes is not clear as the percent difference in methylation at the DMRs associated with these genes were small (1-4%) and may not alter the regulation of the imprinted genes. Indeed, the promoter associated DMRs overlapping with the H19 imprinted DMR were found to correlate with H19 expression; however, no significant alterations in H19 expression were found between conception modes.  Recently, metastable epialleles have become of interest in DNA methylation studies with respect to epigenetic reprogramming as these regions have epigenetic states that are mitotically inherited after establishment in early development and are therefore shared across tissues (Rakyan et al., 2002). Furthermore, nutritional conditions during conception have previously been reported to be important in the establishment of DNA methylation at metastable epialleles (Dominguez-Salas et al., 2014). A recent 450k analysis on dried blood spots from 137 ART and NC newborns found that 18 out of 22 metastable epialleles informative in their analysis exhibited altered DNA methylation in ART conceived newborns (Estill et al., 2017). Of the 25 metastable epialleles informative in this analysis, one – associated with the SDK1 gene – was shown to overlap with DMRs identified at an FDR of 0.06. However, as only 2 of the 25 CpGs in this DMR mapped to the metastable epiallele associated with SDK1, 110  it is not clear that DNA methylation at the metastable epiallele is altered in our analysis. This discrepancy could be due to the smaller sample size of this analysis not being powered to detect DNA methylation changes at these regions or to the potential confounding effect of including multiple gestation and preterm pregnancies in the study conducted by Estill et al. (2017). Another genome-wide analysis of DNA methylation in whole cord blood cells and cord blood mononuclear cells of ART newborns also observed no differences in metastable epiallele DNA methylation between ART and NC groups, suggesting that further interrogation into DNA methylation alterations in metastable epialleles following ART is required (Castillo-Fernandez et al., 2017). Discrepancies between genome-wide DNA methylation analyses in ART conceived individuals has been widespread. Mani et al. (2019) highlighted that out of the 237 genes identified to be significantly associated with ART in nine separate genome-wide DNA methylation studies, only four overlapped between studies with only one – GNAS – was shown to be altered in three separate studies. This investigation adds to these inconsistencies as no significant differences in DNA methylation between conception modes was observed in the previous overlapping genes between studies. However, our finding of significant differences in DNA methylation at the NNAT and MGMT regions overlapped with previous reports by Katari et al. (2009) and Melamed et al. (2015), respectively. These inconsistencies may be explained by the heterogeneity of the ART procedures, heterogeneity of the populations included in the studies, stochastic nature of DNA methylation, lack of validation, and insufficient sample sizes (Mani et al., 2019). Another potential explanation is that individuals may differ in their susceptibility to environmental influences in epigenetic marks in early development (Ghosh et al., 2015). Outlier DNA methylation has been previously observed in cord blood in a subset of 111  the population (Ghosh et al., 2015). These outlier individuals, potentially driving DNA methylation changes between ART and NC babies, could be linked to higher rates of adverse outcomes following ART; therefore, the strict inclusion criterion used in this analysis where pregnancies with adverse outcomes including pre-term births, growth restriction, pre-eclampsia, and other poor outcomes were excluded may have limited the detection of marked DNA methylation changes following ART in this analysis.  A limitation to this study was the small sample size of IVF (n=10) and ICSI (n=9) newborns which limited the power of this study to identify effects of ART on DNA methylation. An additional limitation was the lack of clinical information pertaining to ethnicity, day of embryo transfer, fresh or frozen embryo transfer, and the specifics of the infertilities of the parents. Although, an attempt was made to control for confounding variables, differences in DNA methylation found between groups may be confounded by differences in sex, gestational age, and cell type compositions. Cord blood contamination with maternal blood during collection is also a possibility. Sex estimation during preprocessing allows for identification of contamination in male samples, if the sample clusters incorrectly by sex; however, in female samples, maternal blood contamination is not able to be detected and may impact results. The identified DMRs found using the novel technique mCSEA should also be interpreted with caution as this method lacks widespread validation. Furthermore, targeted DNA methylation and gene expression analyses, in tissues of function, for the identified DMRs in the ART groups is required to validate these results.    112  Chapter 4: CLOSING REMARKS AND FUTURE DIRECTIONS Infertility impacts 1 in 6 couples in developed countries (Thoma et al., 2013). Assisted reproductive technologies is an effective treatment for both male and female infertilities and now accounts for over 7 million babies world-wide since the first ART birth in 1978 (ESHRE, 2018). However, observations of adverse outcomes following ART are concerning for the long-term health of ART conceived individuals, as per the DOHaD hypothesis. These adverse outcomes may be caused by alterations in epigenetic mechanisms induced by subjecting gametes/embryos to ovarian stimulation and in vitro culture procedures during important epigenetic reprogramming events. Epigenetic defects have also been found to be associated with infertility, causing concern that ART may mediate the transmission of these defects onto babies (Kobayashi et al., 2007). Imprinted genes have become a model to study growth related epigenetic alterations in ART babies; however, results remain conflicting and are confounded by inclusion of twins and pregnancies with adverse outcomes in many studies. Furthermore, the genome-wide nature of the epigenetic reprogramming events overlapping ART procedures suggest that alterations in genome-wide DNA methylation, or at genes difficult to detect without a genome-wide approach, may be associated with ART procedures. In this thesis both targeted and genome-wide approaches were employed to interrogate imprinted gene defects and genome-wide DNA methylation alterations in singletons conceived via ART. With the use of the Illumina Infinium MethylationEPIC BeadChip array, this thesis describes the widest coverage of single resolution CpG sites interrogated in ART conceived babies at the time of writing.   The major findings of this thesis suggest that imprinted genes and genome-wide DNA methylation are stable following ART. However, findings of small but significant DNA 113  methylation alterations in imprinted genes, genes linked to development, and genes associated with infertility suggest that ART procedures may 1) impact DNA methylation that persists to birth in genes related to adverse outcomes and could 2) mediate transmission of DNA methylation alterations from infertile parents to babies.   In Chapter 2, I investigated the expression of imprinted genes involved in a proposed imprinted gene network important for in utero growth – PLAGL1, KCNQ1OT1, CDKN1C, and H19 – in cord blood from 24 IVF, 18 ICSI, 9 IUI, and 26 naturally conceived babies. No differences in gene expression of these imprinted genes were observed in any of the ART conception modes indicating a stability of the imprinted gene network in cord blood from ART babies, regardless of ART procedure. However, this analysis was done in ART conceived babies from normal singleton pregnancies, adverse outcomes were excluded, and cannot rule out an association between adverse outcomes observed following ART and altered regulation of these imprinted genes. The strength of this analysis is the strict exclusion criterion which enables comparisons to be made between ICSI, IVF, IUI and NC groups that are not confounded by multiple gestation pregnancies or pregnancies where adverse outcomes are observed. However, limitations to this analysis include lack of targeted DNA methylation and allele-specific expression data which could provide a more in-depth investigation of imprinting maintenance. Furthermore, due to ethical constraints on the tissues that can be analyzed for gene expression alterations in the IGN, cord blood was used in this analysis. However, the biological relevance of stable cord blood gene expression between conception modes is limited as IGN genes are expressed in and regulate development of fetal tissues including bone and skeletal muscle (Arima et al., 2005). 114  Overall this study adds to the literature in that it shows that imprinted genes involved in the growth implicated IGN are not altered following ART in cord blood, an inner cell mass derived tissue. Furthermore, invasive procedures such as ICSI and IVF which subject embryos to in vitro culture were not observed to impact gene expression of IGN imprinted genes more than the less invasive IUI procedure. Although these results suggest a safety of ART with respect to PLAGL1, CDKN1C, KCNQ1OT1, and H19 regulation, further work is needed in more biologically relevant tissues. As PLAGL1 has been found to be altered in a population of ART babies with adverse pregnancy outcomes (Iglesias-Platas et al., 2014; Vincent et al., 2016), interrogating alterations in DNA methylation and expression of IGN imprinted genes in a controlled comparison between ART conceived low birthweight or IUGR babies and NC low birthweight/IUGR babies may be warranted to investigate the role the PLAGL1 IGN may have in these adverse outcomes associated with ART.  In chapter 3, I analyzed DNA methylation at 793,197 CpG sites in cord blood from IVF, ICSI, and NC babies. No significant differences in individual CpG sites and no significant DMRs (using the bumphunter method) were found between conception modes, suggesting a high stability of DNA methylation following ART. However, a small number of CpG sites exhibited higher range, or variance, in DNA methylation in the ICSI (47 CpG sites) and IVF (4 CpG sites) groups. These hypervariable sites in the ART groups indicate that ART procedures could cause some CpG sites to become more labile to DNA methylation changes in either direction (increased methylation or decreased methylation). It is difficult, however, to decouple the impact of ART and the impact of infertility on DNA methylation in ART conceived babies. Indeed, four of the genes found to be associated with hypervariable sites in the ICSI group were linked to 115  infertility. It is possible that the increased range of DNA methylation at these CpG sites could be due to variance in DNA methylation of the parents or an increased predisposition to DNA methylation alterations in gametes, embryos, and resulting babies from infertile parents. A further analysis using the novel mCSEA method that implements a Gene Set Enrichment method for detecting DMRs, where methylation changes are expected to be small but consistent, identified 101 significant promoter associated DMRs in the ICSI group and 101 significant promoter associated DMRs in the IVF group. Interestingly, 25 DMRs overlapped between the IVF and ICSI groups. This finding suggests a targeted impact of ART on DNA methylation. A gene enrichment analysis on these overlapping genes did not, however, identify any significant associations with biological process, cellular component, or molecular function GO terms.   Instead, an almost significant association (p-value = 0.07) with the GO term nervous system development (GO:0007399) was found in a gene enrichment analysis of the 516 genes associated with DMRs in promoter, gene body, and CGI regions. This link seemed to be driven by DMRs found in a PCDH gene cluster. As epigenetic alterations at PCDH genes have been linked to neurodevelopmental disorders and autism spectrum disorders, DNA methylation alterations at these genes following ART could provide a potential mechanism for the observed increase of ASD in ART conceived individuals. Another gene, NPY, associated with a hypomethylated promoter DMR in both IVF and ICSI groups is linked to intrauterine programming of adult onset obesity. Hypomethylation (greater than 10% DNA methylation difference between ART and NC groups) in the promoter of this gene in both ICSI and IVF groups may indicate a targeted DNA methylation alteration following ART, potentially caused by altered nutritional conditions of the early embryo in ART, that could impact long-term 116  outcomes of ART conceived individuals. This analysis also provides support for the transmission of epigenetic alterations linked to male infertility as observed by Castillo-Fernandez et al (2017) as another gene found to be hypomethylated (greater than 25% difference between ART and NC groups) in both the IVF and ICSI groups – GSTM5 – has been shown to have altered methylation in peripheral blood from infertile males. GSTM5 is an antioxidant gene and plays an important role in spermatogenesis and normal sperm function (Oakley, 2011).  This analysis did not reveal an overall destabilization of imprinted genes. However, of the 50 imprinted genes interrogated, DMRs between ART and NC conception modes were found to overlap with the imprinted DMRs of H19, NNAT, BLCAP, and ZNF597. This suggests small but significant alterations at imprinted genes may be associated with ART. The strengths of this analysis include the large number of clinical characteristics that were controlled for via linear regression modelling and the large number of CpG sites interrogated. The weaknesses of this analysis include the small number of samples which limits the power of this study to detect significant differences between conception modes. Furthermore, the results from the novel mCSEA method should be interpreted with caution as this method does not yet have widespread validation. False positives are an issue with large array investigations; therefore, the results of this analysis should again be interpreted with caution due to the lack of validation of significant DMRs with more targeted DNA methylation investigations. Overall the analysis in Chapter 3 provides support to previous findings of stable genome-wide DNA methylation and imprinted gene DNA methylation. However, this analysis revealed that a small number of genomic regions may be impacted by ART. These regions may be significant as genes associated with neurodevelopmental disorders, intrauterine programming of adult onset obesity, and male infertility were observed to be altered in both the IVF and ICSI 117  groups. As such, this analysis provides some support for the notion that ART may be impacting DNA methylation that persists to birth in genes related to adverse outcomes and mediating transmission of DNA methylation alterations from infertile parents to babies. Furthermore, this analysis provides novel genomic regions to be analyzed in the context of adverse outcomes following ART. Future studies validating the findings of this analysis are warranted. Furthermore, if validated, targeted DNA methylation analysis of the NPY and PCDH genes in obese and neurodevelopmentally altered ART conceived individuals, respectively, may provide insight into the mechanism by which ART may be increasing the risk for these poor outcomes. My finding of a small number of regions with altered DNA methylation between conception modes may be due to this analysis being done in normal singleton pregnancies. The exclusion of adverse outcomes may select for babies with DNA methylation profiles similar to those conceived naturally. Therefore, future genome-wide methylation studies in pregnancies with adverse outcomes may provide a clearer picture of the extent to which ART may alter DNA methylation in ART conceived babies. 118  Bibliography 1. Ahmad, A., Ahmed, A., & Patrizio, P. (2013). Cystic fibrosis and fertility. Current Opinion in Obstetrics and Gynecology, 25(3), 167-172. 2. Allegrucci, C., Thurston, A., Lucas, E., & Young, L. (2005). Epigenetics and the germline. Reproduction, 129(2), 137-149. 3. Anckaert, E., & Fair, T. (2015). DNA methylation reprogramming during oogenesis and interference by reproductive technologies: studies in mouse and bovine models. Reproduction, Fertility and Development, 27(5), 739-754. 4. Anckaert, E., Adriaenssens, T., Romero, S., & Smitz, J. (2009a). Ammonium accumulation and use of mineral oil overlay do not alter imprinting establishment at three key imprinted genes in mouse oocytes grown and matured in a long-term follicle culture. Biology of reproduction, 81(4), 666-673. 5. Anckaert, E., Adriaenssens, T., Romero, S., Dremier, S., & Smitz, J. (2009b). Unaltered imprinting establishment of key imprinted genes in mouse oocytes after in vitro follicle culture under variable follicle-stimulating hormone exposure. International Journal of Developmental Biology, 53(4), 541-548. 6. Anckaert, E., Romero, S., Adriaenssens, T., & Smitz, J. (2010). Effects of low methyl donor levels in culture medium during mouse follicle culture on oocyte imprinting establishment. Biology of reproduction, 83(3), 377-386. 7. Anderson, K., Nisenblat, V., & Norman, R. (2010). Lifestyle factors in people seeking infertility treatment–a review. Australian and New Zealand journal of obstetrics and gynaecology, 50(1), 8-20. 8. Anderson, K., Norman, R. J., & Middleton, P. (2010). Preconception lifestyle advice for people with subfertility. Cochrane database of systematic reviews, (4). 9. Arima, T., & Wake, N. (2006). Establishment of the primary imprint of the HYMAI/PLAGL1 imprint control region during oogenesis. Cytogenetic and genome research, 113(1-4), 247-252. 10. Arima, T., Kamikihara, T., Hayashida, T., Kato, K., Inoue, T., Shirayoshi, Y., ... & Wake, N. (2005). ZAC, LIT1 (KCNQ1OT1) and p57 KIP2 (CDKN1C) are in an imprinted gene network that may play a role in Beckwith–Wiedemann syndrome. Nucleic acids research, 33(8), 2650-2660. 11. Aryee, M. J., Jaffe, A. E., Corrada-Bravo, H., Ladd-Acosta, C., Feinberg, A. P., Hansen, K. D., & Irizarry, R. A. (2014). Minfi: a flexible and comprehensive Bioconductor package for the analysis of Infinium DNA methylation microarrays. Bioinformatics, 30(10), 1363-1369. 12. Banrezes, B., Sainte-Beuve, T., Canon, E., Schultz, R. M., Cancela, J., & Ozil, J. P. (2011). Adult body weight is programmed by a redox-regulated and energy-dependent process during the pronuclear stage in mouse. PloS one, 6(12), e29388. 13. Bansal, S. K., Gupta, N., Sankhwar, S. N., & Rajender, S. (2015). Differential genes expression between fertile and infertile spermatozoa revealed by transcriptome analysis. PloS one, 10(5), e0127007. 14. Barker, D. J., & Osmond, C. (1986). Infant mortality, childhood nutrition, and ischaemic heart disease in England and Wales. The Lancet, 327(8489), 1077-1081. 119  15. Barlow, D. P. (2011). Genomic imprinting: a mammalian epigenetic discovery model. Annual review of genetics, 45, 379-403. 16. Bateson, P., Barker, D., Clutton-Brock, T., Deb, D., D'Udine, B., Foley, R. A., ... & McNamara, J. (2004). Developmental plasticity and human health. Nature, 430(6998), 419. 17. Benjamini, Y., & Hochberg, Y. (1995). Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the Royal statistical society: series B (Methodological), 57(1), 289-300. 18. Bhattacharya, S., & Kamath, M. S. (2014). Reducing multiple births in assisted reproduction technology. Best Practice & Research Clinical Obstetrics & Gynaecology, 28(2), 191-199. 19. Bloise, E., Feuer, S. K., & Rinaudo, P. F. (2014). Comparative intrauterine development and placental function of ART concepti: implications for human reproductive medicine and animal breeding. Human reproduction update, 20(6), 822-839. 20. Borgel, J., Guibert, S., Li, Y., Chiba, H., Schübeler, D., Sasaki, H., ... & Weber, M. (2010). Targets and dynamics of promoter DNA methylation during early mouse development. Nature genetics, 42(12), 1093. 21. Bourrouillou, G., Bujan, L., Calvas, P., Colombies, P., Mansat, A., & Pontonnier, F. (1992). Role and contribution of karyotyping in male infertility. Progres en urologie: journal de l'Association francaise d'urologie et de la Societe francaise d'urologie, 2(2), 189-195. 22. Bowdin, S., Allen, C., Kirby, G., Brueton, L., Afnan, M., Barratt, C., ... & Reardon, W. (2007). A survey of assisted reproductive technology births and imprinting disorders. Human Reproduction, 22(12), 3237-3240. 23. Boyce, W. T., & Kobor, M. S. (2015). Development and the epigenome: the ‘synapse’of gene–environment interplay. Developmental science, 18(1), 1-23. 24. Boyd, J. M., Mehta, A., & Murphy, D. J. (2004). Fertility and pregnancy outcomes in men and women with cystic fibrosis in the United Kingdom. Human Reproduction, 19(10), 2238-2243. 25. Brandes, M., Hamilton, C. J. C. M., De Bruin, J. P., Nelen, W. L. D. M., & Kremer, J. A. M. (2009). The relative contribution of IVF to the total ongoing pregnancy rate in a subfertile cohort. Human Reproduction, 25(1), 118-126. 26. Briceag, I., Costache, A., Purcarea, V. L., Cergan, R., Dumitru, M., Briceag, I., ... & Ispas, A. T. (2015). Current management of tubal infertility: from hysterosalpingography to ultrasonography and surgery. Journal of medicine and life, 8(2), 157. 27. Buiting, K. (2010, August). Prader–Willi syndrome and Angelman syndrome. In American Journal of Medical Genetics Part C: Seminars in Medical Genetics (Vol. 154, No. 3, pp. 365-376). Hoboken: Wiley Subscription Services, Inc., A Wiley Company. 28. Bushnik, T., Cook, J. L., Yuzpe, A. A., Tough, S., & Collins, J. (2012). Estimating the prevalence of infertility in Canada. Human reproduction, 27(3), 738-746. 29. Butler, P., & Khanna, J. (2003). Assisted reproduction in developing countries-facing up to the issues. Progress in reproductive health research, 63, 1-8. 30. Camprubí, C., Iglesias-Platas, I., Martin-Trujillo, A., Salvador-Alarcon, C., Rodriguez, M. A., Barredo, D. R., ... & Monk, D. (2013). Stability of genomic imprinting and 120  gestational-age dynamic methylation in complicated pregnancies conceived following assisted reproductive technologies. Biology of reproduction, 89(3), 50-1. 31. Casper, R. F. (2007). Aromatase inhibitors in ovarian stimulation. The Journal of steroid biochemistry and molecular biology, 106(1-5), 71-75. 32. Castillo-Fernandez, J. E., Loke, Y. J., Bass-Stringer, S., Gao, F., Xia, Y., Wu, H., ... & Saffery, R. (2017). DNA methylation changes at infertility genes in newborn twins conceived by in vitro fertilisation. Genome medicine, 9(1), 28. 33. Ceelen, M., Van Weissenbruch, M. M., Prein, J., Smit, J. J., Vermeiden, J. P., Spreeuwenberg, M., ... & Delemarre-van De Waal, H. A. (2009). Growth during infancy and early childhood in relation to blood pressure and body fat measures at age 8–18 years of IVF children and spontaneously conceived controls born to subfertile parents. Human reproduction, 24(11), 2788-2795. 34. Ceelen, M., van Weissenbruch, M. M., Roos, J. C., Vermeiden, J. P., van Leeuwen, F. E., & Delemarre-van de Waal, H. A. (2007). Body composition in children and adolescents born after in vitro fertilization or spontaneous conception. The Journal of Clinical Endocrinology & Metabolism, 92(9), 3417-3423. 35. Ceelen, M., van Weissenbruch, M. M., Vermeiden, J. P., van Leeuwen, F. E., & Delemarre-van de Waal, H. A. (2008). Cardiometabolic differences in children born after in vitro fertilization: follow-up study. The Journal of Clinical Endocrinology & Metabolism, 93(5), 1682-1688. 36. Chachamovich, J. R., Chachamovich, E., Ezer, H., Fleck, M. P., Knauth, D., & Passos, E. P. (2010). Investigating quality of life and health-related quality of life in infertility: a systematic review. Journal of Psychosomatic Obstetrics & Gynecology, 31(2), 101-110. 37. Chang, A. S., Moley, K. H., Wangler, M., Feinberg, A. P., & DeBaun, M. R. (2005). Association between Beckwith-Wiedemann syndrome and assisted reproductive technology: a case series of 19 patients. Fertility and sterility, 83(2), 349-354. 38. Chen, E. Y., Tan, C. M., Kou, Y., Duan, Q., Wang, Z., Meirelles, G. V., ... & Ma’ayan, A. (2013). Enrichr: interactive and collaborative HTML5 gene list enrichment analysis tool. BMC bioinformatics, 14(1), 128. 39. Chen, T., & Li, E. (2004). Structure and function of eukaryotic DNA methyltransferases. In Current topics in developmental biology (Vol. 60, pp. 55-89). Academic Press. 40. Choufani, S., Turinsky, A. L., Melamed, N., Greenblatt, E., Brudno, M., Bérard, A., ... & Monnier, P. (2018). Impact of assisted reproduction, infertility, sex and paternal factors on the placental DNA methylome. Human molecular genetics, 28(3), 372-385. 41. Cirio, M. C., Ratnam, S., Ding, F., Reinhart, B., Navara, C., & Chaillet, J. R. (2008). Preimplantation expression of the somatic form of Dnmt1 suggests a role in the inheritance of genomic imprints. BMC developmental biology, 8(1), 9. 42. Clapier, C. R., & Cairns, B. R. (2009). The biology of chromatin remodeling complexes. Annual review of biochemistry, 78, 273-304. 43. Clayton-Smith, J., & Pembrey, M. E. (1992). Angelman syndrome. Journal of medical genetics, 29(6), 412. 44. Clifford, R. L., Jones, M. J., MacIsaac, J. L., McEwen, L. M., Goodman, S. J., Mostafavi, S., ... & Carlsten, C. (2017). Inhalation of diesel exhaust and allergen alters human bronchial epithelium DNA methylation. Journal of Allergy and Clinical Immunology, 139(1), 112-121. 121  45. Cooper, T. G., Noonan, E., Von Eckardstein, S., Auger, J., Baker, H. W., Behre, H. M., ... & Vogelsong, K. M. (2010). World Health Organization reference values for human semen characteristics. Human reproduction update, 16(3), 231-245. 46. Court, C., Tayama, C., Romanelli, V., Martin-Trujillo, A., Iglesias-Platas, I., Okamura, K., Sugahara, N., ... & Sanchez-Mut, J. V. (2014). Genome-wide parent-of-origin DNA methylation analysis reveals the intricacies of human imprinting and suggests a germline methylation-independent mechanism of establishment. Genome research, 24(4), 554-569. 47. Cox, G. F., Bürger, J., Lip, V., Mau, U. A., Sperling, K., Wu, B. L., & Horsthemke, B. (2002). Intracytoplasmic sperm injection may increase the risk of imprinting defects. The American Journal of Human Genetics, 71(1), 162-164. 48. Crujeiras, A. B., Campion, J., Díaz-Lagares, A., Milagro, F. I., Goyenechea, E., Abete, I., ... & Martínez, J. A. (2013). Association of weight regain with specific methylation levels in the NPY and POMC promoters in leukocytes of obese men: a translational study. Regulatory peptides, 186, 1-6. 49. Cuppens, H., & Cassiman, J. J. (2004). CFTR mutations and polymorphisms in male infertility. International journal of andrology, 27(5), 251-256. 50. Curtis, K. M., Savitz, D. A., & Arbuckle, T. E. (1997). Effects of cigarette smoking, caffeine consumption, and alcohol intake on fecundability. American journal of epidemiology, 146(1), 32-41. 51. D’Urso, A., & Brickner, J. H. (2014). Mechanisms of epigenetic memory. Trends in genetics, 30(6), 230-236. 52. da Rocha, S. T., Edwards, C. A., Ito, M., Ogata, T., & Ferguson-Smith, A. C. (2008). Genomic imprinting at the mammalian Dlk1-Dio3 domain. Trends in Genetics, 24(6), 306-316. 53. de Waal, E., Yamazaki, Y., Ingale, P., Bartolomei, M. S., Yanagimachi, R., & McCarrey, J. R. (2012). Gonadotropin stimulation contributes to an increased incidence of epimutations in ICSI-derived mice. Human molecular genetics, 21(20), 4460-4472. 54. Deaton, A. M., & Bird, A. (2011). CpG islands and the regulation of transcription. Genes & development, 25(10), 1010-1022. 55. DeBaun, M. R., Niemitz, E. L., & Feinberg, A. P. (2003). Association of in vitro fertilization with Beckwith-Wiedemann syndrome and epigenetic alterations of LIT1 and H19. The American Journal of Human Genetics, 72(1), 156-160. 56. Denomme, M. M., Zhang, L., & Mann, M. R. (2011). Embryonic imprinting perturbations do not originate from superovulation-induced defects in DNA methylation acquisition. Fertility and sterility, 96(3), 734-738. 57. Doherty, A. S., Mann, M. R., Tremblay, K. D., Bartolomei, M. S., & Schultz, R. M. (2000). Differential effects of culture on imprinted H19 expression in the preimplantation mouse embryo. Biology of reproduction, 62(6), 1526-1535. 58. Dominguez-Salas, P., Moore, S. E., Baker, M. S., Bergen, A. W., Cox, S. E., Dyer, R. A., ... & Swan, G. E. (2014). Maternal nutrition at conception modulates DNA methylation of human metastable epialleles. Nature communications, 5, 3746. 59. Doornbos, M. E., Maas, S. M., McDonnell, J., Vermeiden, J. P., & Hennekam, R. C. (2007). Infertility, assisted reproduction technologies and imprinting disturbances: a Dutch study. Human Reproduction, 22(9), 2476-2480. 122  60. Dumoulin, J. C., Land, J. A., Van Montfoort, A. P., Nelissen, E. C., Coonen, E., Derhaag, J. G., ... & Evers, J. L. (2010). Effect of in vitro culture of human embryos on birthweight of newborns. Human reproduction, 25(3), 605-612. 61. El Hajj, N., Dittrich, M., & Haaf, T. (2017, September). Epigenetic dysregulation of protocadherins in human disease. In Seminars in cell & developmental biology (Vol. 69, pp. 172-182). Academic Press. 62. El Hajj, N., Haertle, L., Dittrich, M., Denk, S., Lehnen, H., Hahn, T., ... & Haaf, T. (2017). DNA methylation signatures in cord blood of ICSI children. Human Reproduction, 32(8), 1761-1769. 63. El Hajj, N., Trapphoff, T., Linke, M., May, A., Hansmann, T., Kuhtz, J., ... & Eichenlaub-Ritter, U. (2011). Limiting dilution bisulfite (pyro) sequencing reveals parent-specific methylation patterns in single early mouse embryos and bovine oocytes. Epigenetics, 6(10), 1176-1188. 64. ESHRE, A. Fact Sheet 2018. 65. Esposito, E. A., Jones, M. J., Doom, J. R., MacIsaac, J. L., Gunnar, M. R., & Kobor, M. S. (2016). Differential DNA methylation in peripheral blood mononuclear cells in adolescents exposed to significant early but not later childhood adversity. Development and Psychopathology, 28(4pt2), 1385-1399. 66. Essex, M. J., Thomas Boyce, W., Hertzman, C., Lam, L. L., Armstrong, J. M., Neumann, S. M., & Kobor, M. S. (2013). Epigenetic vestiges of early developmental adversity: childhood stress exposure and DNA methylation in adolescence. Child development, 84(1), 58-75. 67. Esteves, S. C., Miyaoka, R., & Agarwal, A. (2011). An update on the clinical assessment of the infertile male. Clinics, 66(4), 691-700. 68. Estill, M. S., Bolnick, J. M., Waterland, R. A., Bolnick, A. D., Diamond, M. P., & Krawetz, S. A. (2016). Assisted reproductive technology alters deoxyribonucleic acid methylation profiles in bloodspots of newborn infants. Fertility and sterility, 106(3), 629-639. 69. Fauque, P., Mondon, F., Letourneur, F., Ripoche, M. A., Journot, L., Barbaux, S., ... & Jammes, H. (2010). In vitro fertilization and embryo culture strongly impact the placental transcriptome in the mouse model. PLoS One, 5(2), e9218. 70. Feuer, S., & Rinaudo, P. (2012). Preimplantation stress and development. Birth Defects Research Part C: Embryo Today: Reviews, 96(4), 299-314. 71. Fidanza, A., Toschi, P., Zacchini, F., Czernik, M., Palmieri, C., Scapolo, P., ... & Ptak, G. E. (2014). Impaired placental vasculogenesis compromises the growth of sheep embryos developed in vitro. Biology of reproduction, 91(1), 21-1. 72. Flegal, K. M., Carroll, M. D., Kit, B. K., & Ogden, C. L. (2012). Prevalence of obesity and trends in the distribution of body mass index among US adults, 1999-2010. Jama, 307(5), 491-497. 73. Fortier, A. L., Lopes, F. L., Darricarrere, N., Martel, J., & Trasler, J. M. (2008). Superovulation alters the expression of imprinted genes in the midgestation mouse placenta. Human molecular genetics, 17(11), 1653-1665. 74. Fox, J., & Weisberg, S. (2018). An R companion to applied regression. Sage Publications. 123  75. Fraga, M. F., Ballestar, E., Paz, M. F., Ropero, S., Setien, F., Ballestar, M. L., ... & Boix-Chornet, M. (2005). Epigenetic differences arise during the lifetime of monozygotic twins. Proceedings of the National Academy of Sciences, 102(30), 10604-10609. 76. Fuks, F., Hurd, P. J., Deplus, R., & Kouzarides, T. (2003). The DNA methyltransferases associate with HP1 and the SUV39H1 histone methyltransferase. Nucleic acids research, 31(9), 2305-2312. 77. Gad, A., Besenfelder, U., Rings, F., Ghanem, N., Salilew-Wondim, D., Hossain, M. M., ... & Schellander, K. (2011). Effect of reproductive tract environment following controlled ovarian hyperstimulation treatment on embryo development and global transcriptome profile of blastocysts: implications for animal breeding and human assisted reproduction. Human Reproduction, 26(7), 1693-1707. 78. Gali Ramamoorthy, T., Begum, G., Harno, E., & White, A. (2015). Developmental programming of hypothalamic neuronal circuits: impact on energy balance control. Frontiers in neuroscience, 9, 126. 79. Gao, J., He, X., Cai, Y., Wang, L., & Fan, X. (2017). Association between assisted reproductive technology and the risk of autism spectrum disorders in the offspring: a meta-analysis. Scientific reports, 7, 46207. 80. Geuns, E., Hilven, P., Van Steirteghem, A., Liebaers, I., & De Rycke, M. (2007). Methylation analysis of KvDMR1 in human oocytes. Journal of medical genetics, 44(2), 144-147. 81. Ghosh, J., Mainigi, M., Coutifaris, C., & Sapienza, C. (2015). Outlier DNA methylation levels as an indicator of environmental exposure and risk of undesirable birth outcome. Human molecular genetics, 25(1), 123-129. 82. Gicquel, C., Gaston, V., Mandelbaum, J., Siffroi, J. P., Flahault, A., & Le Bouc, Y. (2003). In vitro fertilization may increase the risk of Beckwith-Wiedemann syndrome related to the abnormal imprinting of the KCNQ1OT gene. The American Journal of Human Genetics, 72(5), 1338-1341. 83. Giritharan, G., Piane, L. D., Donjacour, A., Esteban, F. J., Horcajadas, J. A., Maltepe, E., & Rinaudo, P. (2012). In vitro culture of mouse embryos reduces differential gene expression between inner cell mass and trophectoderm. Reproductive sciences, 19(3), 243-252. 84. Gomes, M. V., Huber, J., Ferriani, R. A., Amaral Neto, A. M., & Ramos, E. S. (2009). Abnormal methylation at the KvDMR1 imprinting control region in clinically normal children conceived by assisted reproductive technologies. Molecular human reproduction, 15(8), 471-477. 85. Gougeon, A. (2005). The biological aspects of risks of infertility due to age: the female side. Revue d'epidemiologie et de sante publique, 53, 37-45. 86. Gunby, J. (2012). Assisted reproductive technologies (ART) in Canada: 2012 results from the Canadian ART Register. CARTR Annual reports. CARTER Annual report – 2012. Retrieved May 2, 2019 from Canadian Fertility & Andrology Society: https://cfas.ca/_Library/_documents/CARTR_2012.pdf 87. Hajkova, P., Ancelin, K., Waldmann, T., Lacoste, N., Lange, U. C., Cesari, F., ... & Surani, M. A. (2008). Chromatin dynamics during epigenetic reprogramming in the mouse germ line. Nature, 452(7189), 877. 124  88. Hales, C. N., & Barker, D. J. (1992). Type 2 (non-insulin-dependent) diabetes mellitus: the thrifty phenotype hypothesis. Diabetologia, 35(7), 595-601. 89. Halliday, J., Oke, K., Breheny, S., Algar, E., & Amor, D. J. (2004). Beckwith-Wiedemann syndrome and IVF: a case-control study. American journal of human genetics, 75(3), 526. 90. Hammoud, S. S., Purwar, J., Pflueger, C., Cairns, B. R., & Carrell, D. T. (2010). Alterations in sperm DNA methylation patterns at imprinted loci in two classes of infertility. Fertility and sterility, 94(5), 1728-1733. 91. Hansen, K. D. (2016). IlluminaHumanMethylationEPICanno.ilm10b2.hg19: Annotation for Illumina's EPIC methylation arrays. R package version 0.6.0, https://bitbucket.com/kasperdanielhansen/Illumina_EPIC. 92. Hansen, K. D., Timp, W., Bravo, H. C., Sabunciyan, S., Langmead, B., McDonald, O. G., ... & Briem, E. (2011). Increased methylation variation in epigenetic domains across cancer types. Nature genetics, 43(8), 768. 93. Hansen, M., & Bower, C. (2014, August). The impact of assisted reproductive technologies on intra-uterine growth and birth defects in singletons. In Seminars in Fetal and Neonatal Medicine (Vol. 19, No. 4, pp. 228-233). WB Saunders. 94. Hart, R., & Norman, R. J. (2013). The longer-term health outcomes for children born as a result of IVF treatment: Part I–General health outcomes. Human reproduction update, 19(3), 232-243. 95. Haugen, A. C., Schug, T. T., Collman, G., & Heindel, J. J. (2015). Evolution of DOHaD: the impact of environmental health sciences. Journal of developmental origins of health and disease, 6(2), 55-64. 96. Helmerhorst, F. M., Perquin, D. A., Donker, D., & Keirse, M. J. (2004). Perinatal outcome of singletons and twins after assisted conception: a systematic review of controlled studies. Bmj, 328(7434), 261. 97. Henningsen, A. A., Gissler, M., Skjaerven, R., Bergh, C., Tiitinen, A., Romundstad, L. B., ... & Pinborg, A. (2015). Trends in perinatal health after assisted reproduction: a Nordic study from the CoNARTaS group. Human Reproduction, 30(3), 710-716. 98. Hentemann, M., Mousavi, K., & Bertheussen, K. (2011). Differential pH in embryo culture. Fertility and sterility, 95(4), 1291-1294. 99. Hermann, A., Goyal, R., & Jeltsch, A. (2004). The Dnmt1 DNA-(cytosine-C5)-methyltransferase methylates DNA processively with high preference for hemimethylated target sites. Journal of Biological Chemistry, 279(46), 48350-48359. 100. Hernandez Mora, J. R., Tayama, C., Sánchez-Delgado, M., Monteagudo-Sánchez, A., Hata, K., Ogata, T., ... & Esteller, M. (2018). Characterization of parent-of-origin methylation using the Illumina Infinium MethylationEPIC array platform. Epigenomics, 10(7), 941-954. 101. Heyn, H., Ferreira, H. J., Bassas, L., Bonache, S., Sayols, S., Sandoval, J., ... & Larriba, S. (2012). Epigenetic disruption of the PIWI pathway in human spermatogenic disorders. PloS one, 7(10), e47892. 102. Hirsh, A. (2003). Male subfertility. Bmj, 327(7416), 669-672. 103. Hiura, H., Obata, Y., Komiyama, J., Shirai, M., & Kono, T. (2006). Oocyte growth‐dependent progression of maternal imprinting in mice. Genes to Cells, 11(4), 353-361. 125  104. Hoorsan, H., Mirmiran, P., Chaichian, S., Moradi, Y., Hoorsan, R., & Jesmi, F. (2017). Congenital malformations in infants of mothers undergoing assisted reproductive technologies: A systematic review and meta-analysis study. Journal of Preventive Medicine and Public Health, 50(6), 347. 105. Horvath, S. (2013). DNA methylation age of human tissues and cell types. Genome biology, 14(10), 3156. 106. Horvath, S., & Levine, A. J. (2015). HIV-1 infection accelerates age according to the epigenetic clock. The Journal of infectious diseases, 212(10), 1563-1573. 107. Horvath, S., Erhart, W., Brosch, M., Ammerpohl, O., von Schönfels, W., Ahrens, M., ... & Deloukas, P. (2014). Obesity accelerates epigenetic aging of human liver. Proceedings of the National Academy of Sciences, 111(43), 15538-15543. 108. Horvath, S., Garagnani, P., Bacalini, M. G., Pirazzini, C., Salvioli, S., Gentilini, D., ... & Franceschi, C. (2015). Accelerated epigenetic aging in Down syndrome. Aging cell, 14(3), 491-495. 109. Horvath, S., Langfelder, P., Kwak, S., Aaronson, J., Rosinski, J., Vogt, T. F., ... & Choi, O. W. (2016). Huntington's disease accelerates epigenetic aging of human brain and disrupts DNA methylation levels. Aging (Albany NY), 8(7), 1485. 110. Horvath, S., Pirazzini, C., Bacalini, M. G., Gentilini, D., Di Blasio, A. M., Delledonne, M., ... & De Rango, F. (2015). Decreased epigenetic age of PBMCs from Italian semi-supercentenarians and their offspring. Aging (Albany NY), 7(12), 1159. 111. Houseman, E. A., Accomando, W. P., Koestler, D. C., Christensen, B. C., Marsit, C. J., Nelson, H. H., ... & Kelsey, K. T. (2012). DNA methylation arrays as surrogate measures of cell mixture distribution. BMC bioinformatics, 13(1), 86. 112. Houshdaran, S., Cortessis, V. K., Siegmund, K., Yang, A., Laird, P. W., & Sokol, R. Z. (2007). Widespread epigenetic abnormalities suggest a broad DNA methylation erasure defect in abnormal human sperm. PloS one, 2(12), e1289. 113. Howell, C. Y., Bestor, T. H., Ding, F., Latham, K. E., Mertineit, C., Trasler, J. M., & Chaillet, J. R. (2001). Genomic imprinting disrupted by a maternal effect mutation in the Dnmt1 gene. Cell, 104(6), 829-838. 114. Hull, M. G., North, K., Taylor, H., Farrow, A., & Ford, W. C. L. (2000). Delayed conception and active and passive smoking. Fertility and sterility, 74(4), 725-733. 115. Iglesias-Platas, I., Martin-Trujillo, A., Petazzi, P., Guillaumet-Adkins, A., Esteller, M., & Monk, D. (2014). Altered expression of the imprinted transcription factor PLAGL1 deregulates a network of genes in the human IUGR placenta. Human molecular genetics, 23(23), 6275-6285. 116. Ito, S., D’Alessio, A. C., Taranova, O. V., Hong, K., Sowers, L. C., & Zhang, Y. (2010). Role of Tet proteins in 5mC to 5hmC conversion, ES-cell self-renewal and inner cell mass specification. Nature, 466(7310), 1129. 117. Jackson, R. A., Gibson, K. A., Wu, Y. W., & Croughan, M. S. (2004). Perinatal outcomes in singletons following in vitro fertilization: a meta-analysis. Obstetrics & Gynecology, 103(3), 551-563. 118. Jaffe, A. E., Murakami, P., Lee, H., Leek, J. T., Fallin, M. D., Feinberg, A. P., & Irizarry, R. A. (2012). Bump hunting to identify differentially methylated regions in epigenetic epidemiology studies. International journal of epidemiology, 41(1), 200-209. 126  119. Jensen, T. K., Gottschau, M., Madsen, J. O. B., Andersson, A. M., Lassen, T. H., Skakkebæk, N. E., ... & Jørgensen, N. (2014). Habitual alcohol consumption associated with reduced semen quality and changes in reproductive hormones; a cross-sectional study among 1221 young Danish men. BMJ open, 4(9), e005462. 120. Jensen, T. K., Henriksen, T. B., Hjollund, N. H. I., Scheike, T., Kolstad, H., Giwercman, A., ... & Olsen, J. (1998). Caffeine intake and fecundability: a follow-up study among 430 Danish couples planning their first pregnancy. Reproductive toxicology, 12(3), 289-295. 121. Jensen, T. K., Hjollund, N. H. I., Henriksen, T. B., Scheike, T., Kolstad, H., Giwercman, A., ... & Skakkebæk, N. E. (1998). Does moderate alcohol consumption affect fertility? Follow up study among couples planning first pregnancy. Bmj, 317(7157), 505-510. 122. Jensen, T. K., Swan, S., Jørgensen, N., Toppari, J., Redmon, B., Punab, M., ... & Irvine, D. S. (2014). Alcohol and male reproductive health: a cross-sectional study of 8344 healthy men from Europe and the USA. Human reproduction, 29(8), 1801-1809. 123. Jenuwein, T. (2006). The epigenetic magic of histone lysine methylation: delivered on 6 July 2005 at the 30th FEBS Congress in Budapest, Hungary. The FEBS journal, 273(14), 3121-3135. 124. Jia, D., Jurkowska, R. Z., Zhang, X., Jeltsch, A., & Cheng, X. (2007). Structure of Dnmt3a bound to Dnmt3L suggests a model for de novo DNA methylation. Nature, 449(7159), 248. 125. Jiang, R., Jones, M. J., Sava, F., Kobor, M. S., & Carlsten, C. (2014). Short-term diesel exhaust inhalation in a controlled human crossover study is associated with changes in DNA methylation of circulating mononuclear cells in asthmatics. Particle and fibre toxicology, 11(1), 71. 126. Johnson, J. A., Tough, S., Wilson, R. D., Audibert, F., Blight, C., BrockS, J. A., ... & Murphy-Kaulbeck, L. (2012). Delayed child-bearing. Journal of obstetrics and gynaecology Canada, 34(1), 80-93. 127. Jose-Miller, A. B., Boyden J. W., Frey, K. A. (2007). Infertility. American family physician, 75(6), 849-856. 128. Joubert, B. R., Felix, J. F., Yousefi, P., Bakulski, K. M., Just, A. C., Breton, C., ... & Küpers, L. K. (2016). DNA methylation in newborns and maternal smoking in pregnancy: genome-wide consortium meta-analysis. The American Journal of Human Genetics, 98(4), 680-696. 129. Kagami, M., Matsuoka, K., Nagai, T., Yamanaka, M., Kurosawa, K., Suzumori, N., ... & Kato, F. (2012). Paternal uniparental disomy 14 and related disorders: placental gene expression analyses and histological examinations. Epigenetics, 7(10), 1142-1150. 130. Källén, B., Finnström, O., Lindam, A., Nilsson, E., Nygren, K. G., & Olausson, P. O. (2010). Cancer risk in children and young adults conceived by in vitro fertilization. Pediatrics, 126(2), 270-276. 131. Källén, B., Finnström, O., Nygren, K. G., & Olausson, P. O. (2005a). In vitro fertilization in Sweden: child morbidity including cancer risk. Fertility and sterility, 84(3), 605-610. 132. Källén, B., Finnström, O., Nygren, K. G., & Olausson, P. O. (2005b). In vitro fertilization (IVF) in Sweden: risk for congenital malformations after different IVF 127  methods. Birth Defects Research Part A: Clinical and Molecular Teratology, 73(3), 162-169. 133. Kaneda, M., Okano, M., Hata, K., Sado, T., Tsujimoto, N., Li, E., & Sasaki, H. (2004). Essential role for de novo DNA methyltransferase Dnmt3a in paternal and maternal imprinting. Nature, 429(6994), 900. 134. Karmon, A. E., Toth, T. L., Chiu, Y. H., Gaskins, A. J., Tanrikut, C., Wright, D. L., ... & Earth Study Team. (2017). Male caffeine and alcohol intake in relation to semen parameters and in vitro fertilization outcomes among fertility patients. Andrology, 5(2), 354-361. 135. Katari, S., Turan, N., Bibikova, M., Erinle, O., Chalian, R., Foster, M., ... & Sapienza, C. (2009). DNA methylation and gene expression differences in children conceived in vitro or in vivo. Human molecular genetics, 18(20), 3769-3778. 136. Kawwass, J. F., Crawford, S., Kissin, D. M., Session, D. R., Boulet, S., & Jamieson, D. J. (2013). Tubal factor infertility and perinatal risk after assisted reproductive technology. Obstetrics and gynecology, 121(6), 1263. 137. Kelsey, G. (2010, August). Imprinting on chromosome 20: Tissue‐specific imprinting and imprinting mutations in the GNAS locus. In American Journal of Medical Genetics Part C: Seminars in Medical Genetics (Vol. 154, No. 3, pp. 377-386). Hoboken: Wiley Subscription Services, Inc., A Wiley Company. 138. Kenny, L. C., Lavender, T., McNamee, R., O’Neill, S. M., Mills, T., & Khashan, A. S. (2013). Advanced maternal age and adverse pregnancy outcome: evidence from a large contemporary cohort. PloS one, 8(2), e56583. 139. Kerjean, A., Couvert, P., Heams, T., Chalas, C., Poirier, K., Chelly, J., ... & Poirot, C. (2003). In vitro follicular growth affects oocyte imprinting establishment in mice. European journal of human genetics, 11(7), 493. 140. Keverne, E. B., & Curley, J. P. (2008). Epigenetics, brain evolution and behaviour. Frontiers in neuroendocrinology, 29(3), 398-412. 141. Khademi, A., Alleyassin, A., Amini, M., & Ghaemi, M. (2008). Evaluation of sexual dysfunction prevalence in infertile couples. The journal of sexual medicine, 5(6), 1402-1410. 142. Khosla, S., Dean, W., Brown, D., Reik, W., & Feil, R. (2001). Culture of preimplantation mouse embryos affects fetal development and the expression of imprinted genes. Biology of reproduction, 64(3), 918-926. 143. Klar, A. J. (1998). Propagating epigenetic states through meiosis: where Mendel's gene is more than a DNA moiety. Trends in genetics, 14(8), 299-301. 144. Klonoff-Cohen, H., Bleha, J., & Lam-Kruglick, P. (2002). A prospective study of the effects of female and male caffeine consumption on the reproductive endpoints of IVF and gamete intra-Fallopian transfer. Human reproduction, 17(7), 1746-1754. 145. Klonoff-Cohen, H., Lam-Kruglick, P., & Gonzalez, C. (2003). Effects of maternal and paternal alcohol consumption on the success rates of in vitro fertilization and gamete intrafallopian transfer. Fertility and sterility, 79(2), 330-339. 146. Kobayashi, H., Hiura, H., John, R. M., Sato, A., Otsu, E., Kobayashi, N., ... & Yaegashi, N. (2009). DNA methylation errors at imprinted loci after assisted conception originate in the parental sperm. European journal of human genetics, 17(12), 1582. 128  147. Kobayashi, H., Sato, A., Otsu, E., Hiura, H., Tomatsu, C., Utsunomiya, T., ... & Arima, T. (2007). Aberrant DNA methylation of imprinted loci in sperm from oligospermic patients. Human molecular genetics, 16(21), 2542-2551. 148. Kong, Y., Rastogi, D., Seoighe, C., Greally, J. M., & Suzuki, M. (2019). Insights from deconvolution of cell subtype proportions enhance the interpretation of functional genomic data. PloS one, 14(4), e0215987. 149. Kouzarides, T. (2007). Chromatin modifications and their function. Cell, 128(4), 693-705. 150. Laird, C. D., Pleasant, N. D., Clark, A. D., Sneeden, J. L., Hassan, K. A., Manley, N. C., ... & Stöger, R. (2004). Hairpin-bisulfite PCR: assessing epigenetic methylation patterns on complementary strands of individual DNA molecules. Proceedings of the National Academy of Sciences, 101(1), 204-209. 151. Lander, E. S., Linton, L. M., & Birren, B. (2001). Initial sequencing and analysis of the human genome [published correction appears in Nature. 2001; 411 (6838): 720]. Nature, 409(6822), 860-921. 152. Lange, U. C., & Schneider, R. (2010). What an epigenome remembers. Bioessays, 32(8), 659-668. 153. Lazaraviciute, G., Kauser, M., Bhattacharya, S., Haggarty, P., & Bhattacharya, S. (2014). A systematic review and meta-analysis of DNA methylation levels and imprinting disorders in children conceived by IVF/ICSI compared with children conceived spontaneously. Human reproduction update, 20(6), 840-852. 154. Leek, J. T., & Storey, J. D. (2007). Capturing heterogeneity in gene expression studies by surrogate variable analysis. PLoS genetics, 3(9), e161. 155. Leese 156. , H. J. (2012). Metabolism of the preimplantation embryo: 40 years on. Reproduction, 143(4), 417-427. 157. Levine, M. E., Lu, A. T., Bennett, D. A., & Horvath, S. (2015). Epigenetic age of the pre-frontal cortex is associated with neuritic plaques, amyloid load, and Alzheimer’s disease related cognitive functioning. Aging (Albany NY), 7(12), 1198. 158. Lewis, A., Mitsuya, K., Umlauf, D., Smith, P., Dean, W., Walter, J., ... & Reik, W. (2004). Imprinting on distal chromosome 7 in the placenta involves repressive histone methylation independent of DNA methylation. Nature genetics, 36(12), 1291. 159. Li, E., Bestor, T. H., & Jaenisch, R. (1992). Targeted mutation of the DNA methyltransferase gene results in embryonic lethality. Cell, 69(6), 915-926. 160. Li, X., Ito, M., Zhou, F., Youngson, N., Zuo, X., Leder, P., & Ferguson-Smith, A. C. (2008). A maternal-zygotic effect gene, Zfp57, maintains both maternal and paternal imprints. Developmental cell, 15(4), 547-557. 161. Lidegaard, Ø., Pinborg, A., & Andersen, A. N. (2005). Imprinting diseases and IVF: Danish National IVF cohort study. Human Reproduction, 20(4), 950-954. 162. Lim, D., Bowdin, S. C., Tee, L., Kirby, G. A., Blair, E., Fryer, A., ... & Reik, W. (2008). Clinical and molecular genetic features of Beckwith–Wiedemann syndrome associated with assisted reproductive technologies. Human reproduction, 24(3), 741-747. 163. Lindsay, T. J., & Vitrikas, K. R. (2015). Evaluation and treatment of infertility. Am Fam Physician, 91(5), 308-314. 129  164. Litzky, J. F., Deyssenroth, M. A., Everson, T. M., Armstrong, D. A., Lambertini, L., Chen, J., & Marsit, C. J. (2017). Placental imprinting variation associated with assisted reproductive technologies and subfertility. Epigenetics, 12(8), 653-661. 165. Liu, K., Case, A., Cheung, A. P., Sierra, S., AlAsiri, S., Carranza-Mamane, B., ... & Lee, F. (2011). Advanced reproductive age and fertility. Journal of Obstetrics and Gynaecology Canada, 33(11), 1165-1175. 166. Livak, K. J., & Schmittgen, T. D. (2001). Analysis of relative gene expression data using real-time quantitative PCR and the 2− ΔΔCT method. methods, 25(4), 402-408. 167. Lomvardas, S., & Maniatis, T. (2016). Histone and DNA modifications as regulators of neuronal development and function. Cold Spring Harbor perspectives in biology, 8(7), a024208. 168. Lucifero, D., Mann, M. R., Bartolomei, M. S., & Trasler, J. M. (2004). Gene-specific timing and epigenetic memory in oocyte imprinting. Human molecular genetics, 13(8), 839-849. 169. Lucifero, D., Mertineit, C., Clarke, H. J., Bestor, T. H., & Trasler, J. M. (2002). Methylation dynamics of imprinted genes in mouse germ cells. Genomics, 79(4), 530-538. 170. Ludwig, M., Katalinic, A., Gross, S., Sutcliffe, A., Varon, R., & Horsthemke, B. (2005). Increased prevalence of imprinting defects in patients with Angelman syndrome born to subfertile couples. Journal of medical genetics, 42(4), 289-291. 171. Lui, J. C., Finkielstain, G. P., Barnes, K. M., & Baron, J. (2008). An imprinted gene network that controls mammalian somatic growth is down-regulated during postnatal growth deceleration in multiple organs. American Journal of Physiology-Regulatory, Integrative and Comparative Physiology, 295(1), R189-R196. 172. Mackay, D. J., & Temple, I. K. (2010, August). Transient neonatal diabetes mellitus type 1. In American Journal of Medical Genetics Part C: Seminars in Medical Genetics (Vol. 154, No. 3, pp. 335-342). Hoboken: Wiley Subscription Services, Inc., A Wiley Company. 173. Maher, E. R., Brueton, L. A., Bowdin, S. C., Luharia, A., Cooper, W., Cole, T. R., ... & Hawkins, M. M. (2003). Beckwith-Wiedemann syndrome and assisted reproduction technology (ART). Journal of medical genetics, 40(1), 62-64. 174. Maheshwari, A., Pandey, S., Shetty, A., Hamilton, M., & Bhattacharya, S. (2012). Obstetric and perinatal outcomes in singleton pregnancies resulting from the transfer of frozen thawed versus fresh embryos generated through in vitro fertilization treatment: a systematic review and meta-analysis. Fertility and sterility, 98(2), 368-377. 175. Mahmood, S., Smiraglia, D. J., Srinivasan, M., & Patel, M. S. (2013). Epigenetic changes in hypothalamic appetite regulatory genes may underlie the developmental programming for obesity in rat neonates subjected to a high-carbohydrate dietary modification. Journal of developmental origins of health and disease, 4(6), 479-490. 176. Mani, S., Ghosh, J., Coutifaris, C., Sapienza, C., & Mainigi, M. (2019). Epigenetic Changes and Assisted Reproductive Technologies. Epigenetics, (just-accepted). 130  177. Mani, S., Ghosh, J., Lan, Y., Senapati, S., Ord, T., Sapienza, C., ... & Mainigi, M. (2018). Epigenetic changes in preterm birth placenta suggest a role for ADAMTS genes in spontaneous preterm birth. Human molecular genetics, 28(1), 84-95. 178. Mann, M. R., Lee, S. S., Doherty, A. S., Verona, R. I., Nolen, L. D., Schultz, R. M., & Bartolomei, M. S. (2004). Selective loss of imprinting in the placenta following preimplantation development in culture. Development, 131(15), 3727-3735. 179. Market-Velker, B. A., Fernandes, A. D., & Mann, M. R. W. (2010). Side-by-side comparison of five commercial media systems in a mouse model: suboptimal in vitro culture interferes with imprint maintenance. Biology of reproduction, 83(6), 938-950. 180. Market-Velker, B. A., Zhang, L., Magri, L. S., Bonvissuto, A. C., & Mann, M. R. (2010). Dual effects of superovulation: loss of maternal and paternal imprinted methylation in a dose-dependent manner. Human molecular genetics, 19(1), 36-51. 181. Marques, C. J., Carvalho, F., Sousa, M., & Barros, A. (2004). Genomic imprinting in disruptive spermatogenesis. The lancet, 363(9422), 1700-1702. 182. Marques, C. J., Costa, P., Vaz, B., Carvalho, F., Fernandes, S., Barros, A., & Sousa, M. (2008). Abnormal methylation of imprinted genes in human sperm is associated with oligozoospermia. MHR-Basic Science of Reproductive Medicine, 14(2), 67-74. 183. Martino, D. J., Tulic, M. K., Gordon, L., Hodder, M., Richman, T. R., Metcalfe, J., ... & Saffery, R. (2011). Evidence for age-related and individual-specific changes in DNA methylation profile of mononuclear cells during early immune development in humans. Epigenetics, 6(9), 1085-1094. 184. Martino, D., Loke, Y. J., Gordon, L., Ollikainen, M., Cruickshank, M. N., Saffery, R., & Craig, J. M. (2013). Longitudinal, genome-scale analysis of DNA methylation in twins from birth to 18 months of age reveals rapid epigenetic change in early life and pair-specific effects of discordance. Genome biology, 14(5), R42. 185. Martorell-Marugan, J., Gonzalez-Rumayor, V., & Carmona-Saez, P. (2018). mCSEA: Detecting subtle differentially methylated regions. bioRxiv, 293381. 186. Mascarenhas, M. N., Flaxman, S. R., Boerma, T., Vanderpoel, S., & Stevens, G. A. (2012). National, regional, and global trends in infertility prevalence since 1990: a systematic analysis of 277 health surveys. PLoS medicine, 9(12), e1001356. 187. McDade, T. W., Ryan, C., Jones, M. J., MacIsaac, J. L., Morin, A. M., Meyer, J. M., ... & Kuzawa, C. W. (2017). Social and physical environments early in development predict DNA methylation of inflammatory genes in young adulthood. Proceedings of the National Academy of Sciences, 114(29), 7611-7616. 188. McDonald, S. D., Murphy, K., Beyene, J., & Ohlsson, A. (2005). Perinatal outcomes of singleton pregnancies achieved by in vitro fertilization: a systematic review and meta-analysis. Journal of obstetrics and gynaecology Canada, 27(5), 449-459. 189. McGovern, P. G., Llorens, A. J., Skurnick, J. H., Weiss, G., & Goldsmith, L. T. (2004). Increased risk of preterm birth in singleton pregnancies resulting from in vitro fertilization–embryo transfer or gamete intrafallopian transfer: a meta-analysis. Fertility and sterility, 82(6), 1514-1520. 190. McGrath, J., & Solter, D. (1984). Completion of mouse embryogenesis requires both the maternal and paternal genomes. Cell, 37(1), 179-183. 131  191. Meissner, A., Mikkelsen, T. S., Gu, H., Wernig, M., Hanna, J., Sivachenko, A., ... & Gnirke, A. (2008). Genome-scale DNA methylation maps of pluripotent and differentiated cells. Nature, 454(7205), 766. 192. Melamed, N., Choufani, S., Wilkins-Haug, L. E., Koren, G., & Weksberg, R. (2015). Comparison of genome-wide and gene-specific DNA methylation between ART and naturally conceived pregnancies. Epigenetics, 10(6), 474-483. 193. Menken, J., Trussell, J., & Larsen, U. (1986). Age and infertility. Science, 233(4771), 1389-1394. 194. Messerlian, C., Maclagan, L., & Basso, O. (2012). Infertility and the risk of adverse pregnancy outcomes: a systematic review and meta-analysis. Human reproduction, 28(1), 125-137. 195. Messerschmidt, D. M., de Vries, W., Ito, M., Solter, D., Ferguson-Smith, A., & Knowles, B. B. (2012). Trim28 is required for epigenetic stability during mouse oocyte to embryo transition. Science, 335(6075), 1499-1502. 196. Messerschmidt, D. M., Knowles, B. B., & Solter, D. (2014). DNA methylation dynamics during epigenetic reprogramming in the germline and preimplantation embryos. Genes & development, 28(8), 812-828. 197. Michels, K. B. (Ed.). (2012). Epigenetic epidemiology. Springer Science & Business Media. 198. Milan, A. (2013). Fertility: Overview, 2009 to 2011. Statistics Canada (Catalogue number 91-209-X). Retrieved May 2, 2019 from Statistics Canada: https://www150.statcan.gc.ca/n1/en/pub/91-209-x/2013001/article/11784-eng.pdf?st=BfxpXM1Q 199. Minor, A., Chow, V., & Ma, S. (2011). Aberrant DNA methylation at imprinted genes in testicular sperm retrieved from men with obstructive azoospermia and undergoing vasectomy reversal. Reproduction, 141(6), 749-757. 200. Moll, A. C., Imhof, S. M., Cruysberg, J. R., Schouten-van Meeteren, A. Y., Boers, M., & van Leeuwen, F. E. (2003). Incidence of retinoblastoma in children born after in-vitro fertilisation. The Lancet, 361(9354), 309-310. 201. Moore, G. E., Ishida, M., Demetriou, C., Al-Olabi, L., Leon, L. J., Thomas, A. C., ... & Duncan, A. J. (2015). The role and interaction of imprinted genes in human fetal growth. Philosophical Transactions of the Royal Society B: Biological Sciences, 370(1663), 20140074. 202. Moore, T., & Haig, D. (1991). Genomic imprinting in mammalian development: a parental tug-of-war. Trends in Genetics, 7(2), 45-49. 203. Nanassy, L., & Carrell, D. T. (2011). Analysis of the methylation pattern of six gene promoters in sperm of men with abnormal protamination. Asian journal of andrology, 13(2), 342. 204. Navarro-Costa, P., Nogueira, P., Carvalho, M., Leal, F., Cordeiro, I., Calhaz-Jorge, C., ... & Plancha, C. E. (2010). Incorrect DNA methylation of the DAZL promoter CpG island associates with defective human sperm. Human reproduction, 25(10), 2647-2654. 205. Needham, B. L., Smith, J. A., Zhao, W., Wang, X., Mukherjee, B., Kardia, S. L., ... & Diez Roux, A. V. (2015). Life course socioeconomic status and DNA methylation in 132  genes related to stress reactivity and inflammation: The multi-ethnic study of atherosclerosis. Epigenetics, 10(10), 958-969. 206. Nelissen, E. C., Dumoulin, J. C., Busato, F., Ponger, L., Eijssen, L. M., Evers, J. L., ... & van Montfoort, A. P. (2014). Altered gene expression in human placentas after IVF/ICSI. Human Reproduction, 29(12), 2821-2831. 207. Nelissen, E. C., Dumoulin, J. C., Daunay, A., Evers, J. L., Tost, J., & van Montfoort, A. P. (2013). Placentas from pregnancies conceived by IVF/ICSI have a reduced DNA methylation level at the H19 and MEST differentially methylated regions. Human Reproduction, 28(4), 1117-1126. 208. Nelson, C. J., Shindel, A. W., Naughton, C. K., Ohebshalom, M., & Mulhall, J. P. (2008). Prevalence and predictors of sexual problems, relationship stress, and depression in female partners of infertile couples. The journal of sexual medicine, 5(8), 1907-1914. 209. Oakley, A. (2011). Glutathione transferases: a structural perspective. Drug metabolism reviews, 43(2), 138-151. 210. O'Doherty, A. M., O'Shea, L. C., & Fair, T. (2012). Bovine DNA methylation imprints are established in an oocyte size-specific manner, which are coordinated with the expression of the DNMT3 family proteins. Biology of reproduction, 86(3), 67-1. 211. Okada, Y., Yamagata, K., Hong, K., Wakayama, T., & Zhang, Y. I. (2010). A role for the elongator complex in zygotic paternal genome demethylation. Nature, 463(7280), 554. 212. Okano, M., Bell, D. W., Haber, D. A., & Li, E. (1999). DNA methyltransferases Dnmt3a and Dnmt3b are essential for de novo methylation and mammalian development. Cell, 99(3), 247-257. 213. Oliver, V. F., Miles, H. L., Cutfield, W. S., Hofman, P. L., Ludgate, J. L., & Morison, I. M. (2012). Defects in imprinting and genome-wide DNA methylation are not common in the in vitro fertilization population. Fertility and sterility, 97(1), 147-153. 214. Olsen, J., Bolumar, F., Boldsen, J., Bisanti, L., & European Study Group on Infertility and Subfecundity. (1997). Does moderate alcohol intake reduce fecundability? A European multicenter study on infertility and subfecundity. Alcoholism: Clinical and Experimental Research, 21(2), 206-212. 215. Ørstavik, K. H., Eiklid, K., Van Der Hagen, C. B., Spetalen, S., Kierulf, K., Skjeldal, O., & Buiting, K. (2003). Another case of imprinting defect in a girl with Angelman syndrome who was conceived by intracytoplasmic sperm injection. The American Journal of Human Genetics, 72(1), 218-219. 216. Pacheco, S. E., Houseman, E. A., Christensen, B. C., Marsit, C. J., Kelsey, K. T., Sigman, M., & Boekelheide, K. (2011). Integrative DNA methylation and gene expression analyses identify DNA packaging and epigenetic regulatory genes associated with low motility sperm. PloS one, 6(6), e20280. 217. Pandey, S., Shetty, A., Hamilton, M., Bhattacharya, S., & Maheshwari, A. (2012). Obstetric and perinatal outcomes in singleton pregnancies resulting from IVF/ICSI: a systematic review and meta-analysis. Human reproduction update, 18(5), 485-503. 218. Papanikolaou, E. G., Bourgain, C., Kolibianakis, E., Tournaye, H., & Devroey, P. (2005). Steroid receptor expression in late follicular phase endometrium in GnRH antagonist IVF cycles is already altered, indicating initiation of early luteal phase 133  transformation in the absence of secretory changes. Human Reproduction, 20(6), 1541-1547. 219. Patel, Z. P., & Niederberger, C. S. (2011). Male factor assessment in infertility. The Medical clinics of North America, 95(1), 223-234. 220. Paul, D. S., Teschendorff, A. E., Dang, M. A., Lowe, R., Hawa, M. I., Ecker, S., ... & Burden, F. (2016). Increased DNA methylation variability in type 1 diabetes across three immune effector cell types. Nature communications, 7, 13555. 221. Payer, B., Saitou, M., Barton, S. C., Thresher, R., Dixon, J. P., Zahn, D., ... & Surani, M. A. (2003). Stella is a maternal effect gene required for normal early development in mice. Current Biology, 13(23), 2110-2117. 222. Pekcan, M. K., Sarıkaya, E., Tokmak, A., Alışık, M., Alkan, A., Özakşit, G., & Erel, Ö. (2017). ADAMTS-3,-13,-16, and-19 levels in patients with habitual abortion. The Kaohsiung journal of medical sciences, 33(1), 30-35. 223. Pellestor, F., Andréo, B., Arnal, F., Humeau, C., & Demaille, J. (2003). Maternal aging and chromosomal abnormalities: new data drawn from in vitro unfertilized human oocytes. Human genetics, 112(2), 195-203. 224. Peters, J. (2014). The role of genomic imprinting in biology and disease: an expanding view. Nature Reviews Genetics, 15(8), 517. 225. Phipson, B., & Oshlack, A. (2014). DiffVar: a new method for detecting differential variability with application to methylation in cancer and aging. Genome biology, 15(9), 465. 226. Pidsley, R., Zotenko, E., Peters, T. J., Lawrence, M. G., Risbridger, G. P., Molloy, P., ... & Clark, S. J. (2016). Critical evaluation of the Illumina MethylationEPIC BeadChip microarray for whole-genome DNA methylation profiling. Genome biology, 17(1), 208. 227. Pinborg, A., Wennerholm, U. B., Romundstad, L. B., Loft, A., Aittomaki, K., Söderström-Anttila, V., ... & Bergh, C. (2012). Why do singletons conceived after assisted reproduction technology have adverse perinatal outcome? Systematic review and meta-analysis. Human reproduction update, 19(2), 87-104. 228. Poplinski, A., Tüttelmann, F., Kanber, D., Horsthemke, B., & Gromoll, J. (2010). Idiopathic male infertility is strongly associated with aberrant methylation of MEST and IGF2/H19 ICR1. International journal of andrology, 33(4), 642-649. 229. Popp, C., Dean, W., Feng, S., Cokus, S. J., Andrews, S., Pellegrini, M., ... & Reik, W. (2010). Genome-wide erasure of DNA methylation in mouse primordial germ cells is affected by AID deficiency. Nature, 463(7284), 1101. 230. Ptak, G. E., D'Agostino, A., Toschi, P., Fidanza, A., Zacchini, F., Czernik, M., ... & Loi, P. (2012). Post-implantation mortality of in vitro produced embryos is associated with DNA methyltransferase 1 dysfunction in sheep placenta. Human Reproduction, 28(2), 298-305. 231. Puumala, S. E., Nelson, H. H., Ross, J. A., Nguyen, R. H., Damario, M. A., & Spector, L. G. (2012). Similar DNA methylation levels in specific imprinting control regions in children conceived with and without assisted reproductive technology: a cross-sectional study. BMC pediatrics, 12(1), 33. 134  232. Pyun, J. A., Kim, S., Cha, D. H., & Kwack, K. (2014). Epistasis between polymorphisms in TSHB and ADAMTS16 is associated with premature ovarian failure. Menopause, 21(8), 890-895. 233. Qin, J., Liu, X., Sheng, X., Wang, H., & Gao, S. (2016). Assisted reproductive technology and the risk of pregnancy-related complications and adverse pregnancy outcomes in singleton pregnancies: a meta-analysis of cohort studies. Fertility and sterility, 105(1), 73-85. 234. Qin, J., Wang, H., Sheng, X., Liang, D., Tan, H., & Xia, J. (2015). Pregnancy-related complications and adverse pregnancy outcomes in multiple pregnancies resulting from assisted reproductive technology: a meta-analysis of cohort studies. Fertility and sterility, 103(6), 1492-1508. 235. Raatikainen, K., Kuivasaari-Pirinen, P., Hippeläinen, M., & Heinonen, S. (2012). Comparison of the pregnancy outcomes of subfertile women after infertility treatment and in naturally conceived pregnancies. Human reproduction, 27(4), 1162-1169. 236. Radtke, K. M., Ruf, M., Gunter, H. M., Dohrmann, K., Schauer, M., Meyer, A., & Elbert, T. (2011). Transgenerational impact of intimate partner violence on methylation in the promoter of the glucocorticoid receptor. Translational psychiatry, 1(7), e21. 237. Rakyan, V. K., Blewitt, M. E., Druker, R., Preis, J. I., & Whitelaw, E. (2002). Metastable epialleles in mammals. Trends in Genetics, 18(7), 348-351. 238. Rancourt, R. C., Harris, H. R., & Michels, K. B. (2012). Methylation levels at imprinting control regions are not altered with ovulation induction or in vitro fertilization in a birth cohort. Human reproduction, 27(7), 2208-2216. 239. Reese, K. J., Lin, S., Verona, R. I., Schultz, R. M., & Bartolomei, M. S. (2007). Maintenance of paternal methylation and repression of the imprinted H19 gene requires MBD3. PLoS genetics, 3(8), e137. 240. Reigstad, M. M., Larsen, I. K., Myklebust, T. Å., Robsahm, T. E., Oldereid, N. B., Brinton, L. A., & Storeng, R. (2016). Risk of cancer in children conceived by assisted reproductive technology. Pediatrics, 137(3), e20152061. 241. Reik, W., & Walter, J. (2001). Genomic imprinting: parental influence on the genome. Nature Reviews Genetics, 2(1), 21. 242. Reik, W., Constância, M., Fowden, A., Anderson, N., Dean, W., Ferguson‐Smith, A., ... & Sibley, C. (2003). Regulation of supply and demand for maternal nutrients in mammals by imprinted genes. The Journal of physiology, 547(1), 35-44. 243. Riggs, A. D., & Porter, T. N. (1996). Overview of epigenetic mechanisms. Cold Spring Harbor Monograph Archive, 32, 29-45. 244. Rinaudo, P. F., Giritharan, G., Talbi, S., Dobson, A. T., & Schultz, R. M. (2006). Effects of oxygen tension on gene expression in preimplantation mouse embryos. Fertility and sterility, 86(4), 1265-e1. 245. Ritchie, M. E., Phipson, B., Wu, D., Hu, Y., Law, C. W., Shi, W., & Smyth, G. K. (2015). limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic acids research, 43(7), e47-e47. 246. Rittenberg, V., Seshadri, S., Sunkara, S. K., Sobaleva, S., Oteng-Ntim, E., & El-Toukhy, T. (2011). Effect of body mass index on IVF treatment outcome: an updated systematic review and meta-analysis. Reproductive biomedicine online, 23(4), 421-439. 135  247. Rivera, R. M., Stein, P., Weaver, J. R., Mager, J., Schultz, R. M., & Bartolomei, M. S. (2008). Manipulations of mouse embryos prior to implantation result in aberrant expression of imprinted genes on day 9.5 of development. Human molecular genetics, 17(1), 1-14. 248. Roseboom, T. J., Painter, R. C., van Abeelen, A. F., Veenendaal, M. V., & de Rooij, S. R. (2011). Hungry in the womb: what are the consequences? Lessons from the Dutch famine. Maturitas, 70(2), 141-145. 249. Rossignol, S., Steunou, V., Chalas, C., Kerjean, A., Rigolet, M., Viegas-Pequignot, E., ... & Gicquel, C. (2006). The epigenetic imprinting defect of patients with Beckwith—Wiedemann syndrome born after assisted reproductive technology is not restricted to the 11p15 region. Journal of medical genetics, 43(12), 902-907. 250. Sakian, S., Louie, K., Wong, E. C., Havelock, J., Kashyap, S., Rowe, T., ... & Ma, S. (2015). Altered gene expression of H19 and IGF2 in placentas from ART pregnancies. Placenta, 36(10), 1100-1105. 251. Sakka, S. D., Loutradis, D., Kanaka-Gantenbein, C., Margeli, A., Papastamataki, M., Papassotiriou, I., & Chrousos, G. P. (2010). Absence of insulin resistance and low-grade inflammation despite early metabolic syndrome manifestations in children born after in vitro fertilization. Fertility and sterility, 94(5), 1693-1699. 252. Santi, D., De Vincentis, S., Magnani, E., & Spaggiari, G. (2017). Impairment of sperm DNA methylation in male infertility: a meta‐analytic study. Andrology, 5(4), 695-703. 253. Sarkar, S., Sujit, K. M., Singh, V., Pandey, R., Trivedi, S., Singh, K., ... & Rajender, S. (2019). Array-based DNA methylation profiling reveals peripheral blood differential methylation in male infertility. Fertility and sterility. 254. Sasaki, H., Ishihara, K., & Kato, R. (2000). Mechanisms of Igf2/H19 imprinting: DNA methylation, chromatin and long-distance gene regulation. The Journal of Biochemistry, 127(5), 711-715. 255. Sato, A., Otsu, E., Negishi, H., Utsunomiya, T., & Arima, T. (2007). Aberrant DNA methylation of imprinted loci in superovulated oocytes. Human reproduction, 22(1), 26-35. 256. Schwarzer, C., Esteves, T. C., Arauzo-Bravo, M. J., Le Gac, S., Nordhoff, V., Schlatt, S., & Boiani, M. (2012). ART culture conditions change the probability of mouse embryo gestation through defined cellular and molecular responses. Human reproduction, 27(9), 2627-2640. 257. Seki, Y., Hayashi, K., Itoh, K., Mizugaki, M., Saitou, M., & Matsui, Y. (2005). Extensive and orderly reprogramming of genome-wide chromatin modifications associated with specification and early development of germ cells in mice. Developmental biology, 278(2), 440-458. 258. Seli, E., & Wiley Online Library. (2011). infertility. Hoboken, NJ;Chichester, West Sussex;: Wiley-Blackwell. 259. Sharlip, I. D., Jarow, J. P., Belker, A. M., Lipshultz, L. I., Sigman, M., Thomas, A. J., ... & Overstreet, J. W. (2002). Best practice policies for male infertility. Fertility and sterility, 77(5), 873-882. 136  260. Shi, W., & Haaf, T. (2002). Aberrant methylation patterns at the two‐cell stage as an indicator of early developmental failure. Molecular Reproduction and Development: Incorporating Gamete Research, 63(3), 329-334. 261. Shin, B. C., Dai, Y., Thamotharan, M., Gibson, L. C., & Devaskar, S. U. (2012). Pre‐and postnatal calorie restriction perturbs early hypothalamic neuropeptide and energy balance. Journal of neuroscience research, 90(6), 1169-1182. 262. Shovlin, T. C., Bourc’his, D., La Salle, S., O’Doherty, A., Trasler, J. M., Bestor, T. H., & Walsh, C. P. (2006). Sex-specific promoters regulate Dnmt3L expression in mouse germ cells. Human reproduction, 22(2), 457-467. 263. Silver, M. J., Kessler, N. J., Hennig, B. J., Dominguez-Salas, P., Laritsky, E., Baker, M. S., ... & Gong, Y. Y. (2015). Independent genomewide screens identify the tumor suppressor VTRNA2-1 as a human epiallele responsive to periconceptional environment. Genome biology, 16(1), 118. 264. Simon, C., Domínguez, F., Valbuena, D., & Pellicer, A. (2003). The role of estrogen in uterine receptivity and blastocyst implantation. Trends in Endocrinology & Metabolism, 14(5), 197-199. 265. Simoni, M., Bakker, E., & Krausz, C. (2004). EAA/EMQN best practice guidelines for molecular diagnosis of y‐chromosomal microdeletions. State of the art 2004. International journal of andrology, 27(4), 240-249. 266. Song, S., Ghosh, J., Mainigi, M., Turan, N., Weinerman, R., Truongcao, M., ... & Sapienza, C. (2015). DNA methylation differences between in vitro-and in vivo-conceived children are associated with ART procedures rather than infertility. Clinical epigenetics, 7(1), 41. 267. Stouder, C., Deutsch, S., & Paoloni-Giacobino, A. (2009). Superovulation in mice alters the methylation pattern of imprinted genes in the sperm of the offspring. Reproductive Toxicology, 28(4), 536-541. 268. Sutcliffe, A. G., Peters, C. J., Bowdin, S., Temple, K., Reardon, W., Wilson, L., ... & Maher, E. R. (2006). Assisted reproductive therapies and imprinting disorders—a preliminary British survey. Human Reproduction, 21(4), 1009-1011. 269. Suzuki, J., Therrien, J., Filion, F., Lefebvre, R., Goff, A. K., & Smith, L. C. (2009). In vitro culture and somatic cell nuclear transfer affect imprinting of SNRPN gene in pre-and post-implantation stages of development in cattle. BMC developmental biology, 9(1), 9. 270. Tan, S., Hahn, S., Benson, S., Janssen, O. E., Dietz, T., Kimmig, R., ... & Elsenbruch, S. (2008). Psychological implications of infertility in women with polycystic ovary syndrome. Human Reproduction, 23(9), 2064-2071. 271. Tayama, C., Romanelli, V., Martin-Trujillo, A., Iglesias-Platas, I., Okamura, K., Sugahara, N., ... & Sanchez-Mut, J. V. (2014). Genome-wide parent-of-origin DNA methylation analysis reveals the intricacies of human imprinting and suggests a germline methylation-independent mechanism of establishment. Genome research, 24(4), 554-569. 272. Tenorio, J., Romanelli, V., Martin‐Trujillo, A., Fernández, G. M., Segovia, M., Perandones, C., ... & Gordo, G. (2016). Clinical and molecular analyses of Beckwith–Wiedemann syndrome: Comparison between spontaneous conception and assisted reproduction techniques. American journal of medical genetics Part A, 170(10), 2740-2749. 137  273. Thoma, M. E., McLain, A. C., Louis, J. F., King, R. B., Trumble, A. C., Sundaram, R., & Louis, G. M. B. (2013). Prevalence of infertility in the United States as estimated by the current duration approach and a traditional constructed approach. Fertility and sterility, 99(5), 1324-1331. 274. Tierling, S., Souren, N. Y., Gries, J., LoPorto, C., Groth, M., Lutsik, P., ... & Griesinger, G. (2010). Assisted reproductive technologies do not enhance the variability of DNA methylation imprints in human. Journal of medical genetics, 47(6), 371-376. 275. Tietze, C. (1957). Reproductive span and rate of reproduction among Hutterite women. Obstetrical & Gynecological Survey, 12(5), 727-728. 276. Trapphoff, T., El Hajj, N., Zechner, U., Haaf, T., & Eichenlaub-Ritter, U. (2010). DNA integrity, growth pattern, spindle formation, chromosomal constitution and imprinting patterns of mouse oocytes from vitrified pre-antral follicles. Human reproduction, 25(12), 3025-3042. 277. Turan, N., Katari, S., Gerson, L. F., Chalian, R., Foster, M. W., Gaughan, J. P., ... & Sapienza, C. (2010). Inter-and intra-individual variation in allele-specific DNA methylation and gene expression in children conceived using assisted reproductive technology. PLoS genetics, 6(7), e1001033. 278. Tycko, B., & Morison, I. M. (2002). Physiological functions of imprinted genes. Journal of cellular physiology, 192(3), 245-258. 279. Umehara, T., Kawashima, I., Kawai, T., Hoshino, Y., Morohashi, K. I., Shima, Y., ... & Shimada, M. (2016). Neuregulin 1 regulates proliferation of Leydig cells to support spermatogenesis and sexual behavior in adult mice. Endocrinology, 157(12), 4899-4913. 280. Urdinguio, R. G., Bayon, G. F., Dmitrijeva, M., Toraño, E. G., Bravo, C., Fraga, M. F., ... & Fernández, A. F. (2015). Aberrant DNA methylation patterns of spermatozoa in men with unexplained infertility. Human reproduction, 30(5), 1014-1028. 281. Valinluck, V., & Sowers, L. C. (2007). Endogenous cytosine damage products alter the site selectivity of human DNA maintenance methyltransferase DNMT1. Cancer research, 67(3), 946-950. 282. Van Montfoort, A. P. A., Hanssen, L. L. P., De Sutter, P., Viville, S., Geraedts, J. P. M., & De Boer, P. (2012). Assisted reproduction treatment and epigenetic inheritance. Human reproduction update, 18(2), 171-197. 283. Varrault, A., Gueydan, C., Delalbre, A., Bellmann, A., Houssami, S., Aknin, C., ... & Pavlidis, P. (2006). Zac1 regulates an imprinted gene network critically involved in the control of embryonic growth. Developmental cell, 11(5), 711-722. 284. Vincent, R. N., Gooding, L. D., Louie, K., Wong, E. C., & Ma, S. (2016). Altered DNA methylation and expression of PLAGL1 in cord blood from assisted reproductive technology pregnancies compared with natural conceptions. Fertility and sterility, 106(3), 739-748. 285. Waterland, R. A., & Michels, K. B. (2007). Epigenetic epidemiology of the developmental origins hypothesis. Annu. Rev. Nutr., 27, 363-388. 286. Waylen, A. L., Metwally, M., Jones, G. L., Wilkinson, A. J., & Ledger, W. L. (2008). Effects of cigarette smoking upon clinical outcomes of assisted reproduction: a meta-analysis. Human reproduction update, 15(1), 31-44. 138  287. Weber, M., Hellmann, I., Stadler, M. B., Ramos, L., Pääbo, S., Rebhan, M., & Schübeler, D. (2007). Distribution, silencing potential and evolutionary impact of promoter DNA methylation in the human genome. Nature genetics, 39(4), 457. 288. Webster, A. P., Plant, D., Ecker, S., Zufferey, F., Bell, J. T., Feber, A., ... & Worthington, J. (2018). Increased DNA methylation variability in rheumatoid arthritis-discordant monozygotic twins. Genome medicine, 10(1), 64. 289. Weiner, J. A., & Jontes, J. (2013). Protocadherins, not prototypical: a complex tale of their interactions, expression, and functions. Frontiers in molecular neuroscience, 6, 4. 290. Weksberg, R., Shuman, C., & Beckwith, J. B. (2010). Beckwith–Wiedemann syndrome. European journal of human genetics, 18(1), 8. 291. Weksberg, R., Shuman, C., & Smith, A. C. (2005, August). Beckwith–wiedemann syndrome. In American Journal of Medical Genetics Part C: Seminars in Medical Genetics (Vol. 137, No. 1, pp. 12-23). Hoboken: Wiley Subscription Services, Inc., A Wiley Company. 292. Weksberg, R., Shuman, C., Caluseriu, O., Smith, A. C., Fei, Y. L., Nishikawa, J., ... & Ives, E. (2002). Discordant KCNQ1OT1 imprinting in sets of monozygotic twins discordant for Beckwith–Wiedemann syndrome. Human Molecular Genetics, 11(11), 1317-1325. 293. Wennerholm, U. B., Henningsen, A. K. A., Romundstad, L. B., Bergh, C., Pinborg, A., Skjaerven, R., ... & Tiitinen, A. (2013). Perinatal outcomes of children born after frozen-thawed embryo transfer: a Nordic cohort study from the CoNARTaS group. Human reproduction, 28(9), 2545-2553. 294. Wickham, H. (2016). ggplot2: elegant graphics for data analysis. Springer. 295. Willem, O., Karen, P., De Sutter, P., Jan, G., Eugene, B., Guy, M., ... & Wilfried, G. (2005). Perinatal outcome of ICSI pregnancies compared with a matched group of natural conception pregnancies in Flanders (Belgium): a cohort study. Reproductive biomedicine online, 11(2), 244-253. 296. Williams, C., & Sutcliffe, A. (2009). Infant outcomes of assisted reproduction. Early human development, 85(11), 673-677. 297. Wilson, C. L., Fisher, J. R., Hammarberg, K., Amor, D. J., & Halliday, J. L. (2011). Looking downstream: a review of the literature on physical and psychosocial health outcomes in adolescents and young adults who were conceived by ART. Human Reproduction, 26(5), 1209-1219. 298. Wong, E. C., Hatakeyama, C., Robinson, W. P., & Ma, S. (2011). DNA methylation at H19/IGF2 ICR1 in the placenta of pregnancies conceived by in vitro fertilization and intracytoplasmic sperm injection. Fertility and sterility, 95(8), 2524-2526. 299. Wossidlo, M., Nakamura, T., Lepikhov, K., Marques, C. J., Zakhartchenko, V., Boiani, M., ... & Walter, J. (2011). 5-Hydroxymethylcytosine in the mammalian zygote is linked with epigenetic reprogramming. Nature communications, 2, 241. 300. Wu, H., & Zhang, Y. (2014). Reversing DNA methylation: mechanisms, genomics, and biological functions. Cell, 156(1-2), 45-68. 139  301. Wu, W., Shen, O., Qin, Y., Niu, X., Lu, C., Xia, Y., ... & Wang, X. (2010). Idiopathic male infertility is strongly associated with aberrant promoter methylation of methylenetetrahydrofolate reductase (MTHFR). PloS one, 5(11), e13884. 302. Xu, F., Xia, W., Luo, R. Z., Peng, H., Zhao, S., Dai, J., ... & Parlow, A. F. (2000). The human ARHI tumor suppressor gene inhibits lactation and growth in transgenic mice. Cancer research, 60(17), 4913-4920. 303. Xu, N., Barlow, G. M., Cui, J., Wang, E. T., Lee, B., Akhlaghpour, M., ... & Goodarzi, M. O. (2017). Comparison of genome-wide and gene-specific DNA methylation profiling in first-trimester chorionic villi from pregnancies conceived with infertility treatments. Reproductive Sciences, 24(7), 996-1004. 304. Young, L. E., Fernandes, K., McEvoy, T. G., Butterwith, S. C., Gutierrez, C. G., Carolan, C., ... & Sinclair, K. D. (2001). Epigenetic change in IGF2R is associated with fetal overgrowth after sheep embryo culture. Nature genetics, 27(2), 153. 305. Young, L. E., Sinclair, K. D., & Wilmut, I. (1998). Large offspring syndrome in cattle and sheep. Reviews of reproduction, 3(3), 155-163. 306. Yuen, R. K., Jiang, R., Peñaherrera, M. S., McFadden, D. E., & Robinson, W. P. (2011). Genome-wide mapping of imprinted differentially methylated regions by DNA methylation profiling of human placentas from triploidies. Epigenetics & chromatin, 4(1), 10.  140  Appendix A.1 Promoter associated DMRs in the ICSI and NC group comparison at an FDR less than 0.05 NES DMR size Chr Adjusted p-value ICSI mean beta NC mean beta Beta difference % beta difference Gene -2.69 41 chr20 0.040 0.697 0.708 -0.010 -1.494 NNAT -2.48 12 chr6 0.040 0.869 0.875 -0.006 -0.675 HUS1B -2.44 13 chr11 0.040 0.264 0.294 -0.031 -10.989 NAALAD2 -2.43 15 chr19 0.040 0.266 0.291 -0.025 -9.159 ZNF577 -2.33 15 chr5 0.040 0.269 0.293 -0.024 -8.688 LRRC14B -2.32 14 chr17 0.040 0.345 0.428 -0.083 -21.355 C17orf97 -2.29 15 chr7 0.040 0.186 0.217 -0.030 -15.098 NPY -2.29 13 chr5 0.040 0.543 0.557 -0.013 -2.427 ESM1 -2.29 11 chr5 0.040 0.638 0.695 -0.057 -8.551 SLC6A19 -2.27 9 chr5 0.040 0.345 0.369 -0.024 -6.819 LOC134466 -2.27 9 chr18 0.040 0.486 0.537 -0.051 -9.988 LOC100130522 -2.26 26 chr5 0.040 0.295 0.319 -0.023 -7.595 RUFY1 -2.23 7 chr1 0.040 0.326 0.388 -0.062 -17.494 AKR7L -2.19 20 chr14 0.040 0.317 0.318 -0.001 -0.203 MRPL52 -2.17 7 chr1 0.040 0.161 0.181 -0.020 -11.711 AKR7A3 -2.15 18 chr8 0.040 0.224 0.228 -0.003 -1.495 PDGFRL -2.14 17 chr3 0.040 0.329 0.350 -0.021 -6.282 SSR3 -2.13 11 chr5 0.040 0.484 0.492 -0.008 -1.603 BNIP1 -2.12 18 chr22 0.040 0.213 0.222 -0.009 -4.224 TBX1 -2.11 39 chr8 0.040 0.356 0.407 -0.051 -13.357 FDFT1 -2.11 15 chr2 0.040 0.774 0.798 -0.024 -3.106 KCNE4 -2.11 19 chr5 0.040 0.418 0.430 -0.012 -2.772 ADAMTS16 -2.10 9 chr17 0.040 0.388 0.420 -0.033 -8.067 SKAP1 -2.08 33 chr6 0.040 0.766 0.770 -0.004 -0.484 EXOC2 -2.08 14 chr4 0.040 0.335 0.384 -0.050 -13.812 TRAM1L1 -2.08 15 chr11 0.040 0.673 0.694 -0.021 -3.073 TRIM29 -2.07 16 chr19 0.040 0.309 0.333 -0.024 -7.528 KLK7 -2.07 16 chr2 0.040 0.096 0.102 -0.006 -6.237 POLR1A -2.05 5 chr1 0.040 0.331 0.488 -0.157 -38.298 GSTM5 -2.05 13 chr16 0.040 0.743 0.748 -0.005 -0.703 CCDC154 -2.03 50 chr16 0.040 0.423 0.449 -0.027 -6.125 ABAT -2.03 69 chr6 0.040 0.509 0.516 -0.007 -1.382 HIVEP2 -2.02 7 chr6 0.040 0.298 0.369 -0.071 -21.407 SNORD52 141  -2.01 15 chr20 0.040 0.293 0.296 -0.003 -1.052 DPM1 -2.00 9 chr4 0.040 0.314 0.342 -0.028 -8.475 C4orf49 -2.00 8 chr19 0.040 0.493 0.500 -0.006 -1.262 CLEC11A -2.00 105 chr20 0.040 0.530 0.536 -0.006 -1.184 BLCAP -1.98 8 chr5 0.040 0.410 0.431 -0.020 -4.869 ITK -1.98 8 chr15 0.040 0.769 0.800 -0.031 -3.925 FGF7 -1.98 8 chr16 0.040 0.825 0.847 -0.022 -2.596 MSLNL -1.93 53 chr1 0.040 0.469 0.486 -0.017 -3.638 RAP1GAP -1.91 6 chr17 0.040 0.773 0.809 -0.036 -4.566 THRA1/BTR -1.89 5 chr21 0.040 0.720 0.739 -0.019 -2.557 KRTAP12-1 -1.87 6 chr10 0.040 0.727 0.732 -0.005 -0.697 DHX32 -1.86 6 chr6 0.040 0.762 0.784 -0.022 -2.885 FHL5 -1.86 5 chr2 0.040 0.100 0.103 -0.003 -3.347 LOC400940 -1.86 6 chr17 0.040 0.501 0.505 -0.004 -0.794 LOC284009 -1.84 5 chr19 0.040 0.841 0.843 -0.002 -0.204 KIR2DL1 1.89 59 chr6 0.040 0.560 0.548 0.012 2.221 DDR1 1.93 8 chr22 0.040 0.133 0.122 0.012 9.105 DGCR6L 1.94 86 chr2 0.040 0.694 0.688 0.006 0.894 SH3BP4 1.94 53 chr6 0.040 0.557 0.548 0.009 1.683 PPT2 1.95 5 chr12 0.040 0.576 0.562 0.014 2.391 NTS 1.96 6 chr1 0.040 0.451 0.494 -0.044 -9.210 GPR88 1.99 9 chr3 0.040 0.712 0.709 0.003 0.444 PRR23B 2.00 11 chr16 0.040 0.095 0.085 0.010 10.915 CDR2 2.02 9 chr12 0.040 0.613 0.597 0.015 2.534 RASSF9 2.02 13 chr11 0.040 0.265 0.250 0.016 6.075 ALKBH3 2.02 8 chr13 0.040 0.314 0.311 0.003 0.963 MLNR 2.03 13 chr5 0.040 0.312 0.286 0.026 8.721 PCDHGB3 2.03 14 chr4 0.040 0.263 0.256 0.007 2.792 TMEM144 2.06 13 chr4 0.040 0.284 0.275 0.008 3.001 SLC10A4 2.06 23 chr10 0.040 0.382 0.359 0.023 6.251 MGMT 2.06 12 chr12 0.040 0.101 0.098 0.003 3.340 MIRLET7I 2.07 8 chr5 0.040 0.689 0.668 0.021 3.112 PRDM9 2.08 9 chr15 0.040 0.656 0.635 0.021 3.305 SPESP1 2.08 8 chr5 0.040 0.583 0.565 0.018 3.058 DMGDH 2.08 12 chr13 0.040 0.707 0.706 0.001 0.117 F10 2.09 12 chr16 0.040 0.547 0.544 0.003 0.483 ZNF205 2.09 10 chr19 0.040 0.688 0.678 0.010 1.453 BEST2 2.10 12 chr2 0.040 0.701 0.696 0.005 0.749 IL1RL2 2.10 10 chr8 0.040 0.259 0.256 0.003 1.241 CHMP4C 2.10 13 chr5 0.040 0.439 0.439 0.000 -0.001 MARVELD2 2.11 22 chr11 0.040 0.168 0.168 -0.001 -0.375 ZNF214 142  2.11 34 chr13 0.040 0.204 0.188 0.016 8.208 CCNA1 2.12 10 chr4 0.040 0.442 0.408 0.033 7.837 TRIM61 2.12 31 chr6 0.040 0.626 0.616 0.009 1.521 PRRT1 2.12 34 chr17 0.040 0.469 0.463 0.006 1.198 HOXB6 2.13 13 chr13 0.040 0.614 0.603 0.011 1.865 DLEU7 2.14 9 chr13 0.040 0.654 0.656 -0.002 -0.307 F7 2.15 11 chr4 0.040 0.472 0.469 0.003 0.620 PF4 2.16 14 chr16 0.040 0.454 0.441 0.014 3.048 ZNF597 2.18 73 chr12 0.040 0.618 0.615 0.003 0.449 TSPAN9 2.18 14 chr3 0.040 0.669 0.678 -0.009 -1.340 PRR23C 2.18 14 chr6 0.040 0.449 0.431 0.018 4.119 ARMC2 2.20 13 chr3 0.040 0.764 0.766 -0.002 -0.326 PRR23A 2.22 15 chr5 0.040 0.602 0.587 0.015 2.540 PCDHGA4 2.22 36 chr7 0.040 0.677 0.674 0.003 0.461 HOXA5 2.25 15 chr12 0.040 0.365 0.356 0.009 2.517 PLBD1 2.29 16 chr5 0.040 0.358 0.333 0.025 7.379 PCDHGB1 2.33 22 chr8 0.040 0.302 0.291 0.011 3.691 TP53INP1 2.33 36 chr11 0.040 0.624 0.599 0.025 4.072 H19 2.37 30 chr2 0.040 0.710 0.703 0.007 0.943 BOLL 2.37 26 chr17 0.040 0.665 0.646 0.020 2.992 WFIKKN2 2.38 9 chr12 0.040 0.493 0.480 0.013 2.582 LOC144571 2.41 25 chr19 0.040 0.809 0.809 0.001 0.073 LYPD4 2.47 13 chr17 0.040 0.380 0.385 -0.005 -1.259 MYCBPAP 2.50 13 chr19 0.040 0.779 0.735 0.045 5.878 AURKC 2.51 22 chr19 0.040 0.809 0.809 0.001 0.073 DMRTC2 2.61 19 chr4 0.040 0.213 0.203 0.010 4.767 DDX60L 2.95 19 chr10 0.040 0.161 0.112 0.049 36.019 C10orf4 NES: normalized enrichment score of each DMR DMR size: The number of CpG sites in the DMR Chr: Chromosome Beta difference: The difference in mean DNA methylation (beta values) between the ICSI and NC groups. Positive value corresponds to hypermethylation while a negative value corresponds to hypomethylation in the ICSI group. Beta % difference: The percent difference in mean DNA methylation (beta values)   143  A.2 Promoter associated DMRs in the IVF and NC group comparison at an FDR less than 0.05 NES DMR size Chr Adjusted p-value IVF mean beta NC mean beta Beta difference % difference beta Gene -2.51 37 chr22 0.039 0.527 0.544 -0.017 -3.19 GRAP2 -2.46 14 chr17 0.039 0.197 0.308 -0.111 -43.88 C17orf97 -2.46 15 chr19 0.039 0.253 0.291 -0.038 -14.00 ZNF577 -2.36 13 chr16 0.039 0.652 0.682 -0.030 -4.48 IL32 -2.36 12 chr6 0.039 0.875 0.879 -0.003 -0.38 HUS1B -2.33 9 chr17 0.039 0.379 0.420 -0.041 -10.27 SKAP1 -2.32 17 chr5 0.039 0.481 0.504 -0.023 -4.64 PCDHB7 -2.29 14 chr5 0.039 0.462 0.486 -0.024 -5.10 PCDHB11 -2.26 16 chr3 0.039 0.095 0.154 -0.059 -47.39 MCCC1 -2.25 17 chr19 0.039 0.123 0.139 -0.016 -12.34 ZNF350 -2.24 16 chr3 0.039 0.126 0.181 -0.055 -36.04 PLSCR1 -2.24 15 chr3 0.039 0.362 0.391 -0.029 -7.64 LRRC15 -2.24 23 chr1 0.039 0.598 0.621 -0.023 -3.72 PRKACB -2.23 12 chr1 0.039 0.407 0.511 -0.104 -22.63 PM20D1 -2.22 15 chr7 0.039 0.182 0.207 -0.024 -12.50 NPY -2.21 7 chr1 0.039 0.337 0.388 -0.051 -14.05 AKR7L -2.20 8 chr16 0.039 0.771 0.788 -0.017 -2.12 MSLNL -2.19 17 chr11 0.039 0.452 0.465 -0.013 -2.74 FXYD2 -2.18 15 chr7 0.039 0.437 0.460 -0.023 -5.13 DNAJC2 -2.18 11 chr3 0.039 0.146 0.178 -0.031 -19.37 STAC -2.18 9 chr1 0.039 0.812 0.824 -0.012 -1.47 SEC16B -2.17 9 chr5 0.039 0.318 0.354 -0.036 -10.82 LOC134466 -2.16 13 chr5 0.039 0.543 0.562 -0.019 -3.40 ESM1 -2.16 9 chr1 0.039 0.374 0.429 -0.055 -13.62 C1orf173 -2.16 46 chr4 0.039 0.463 0.475 -0.012 -2.53 RHOH -2.14 33 chr6 0.039 0.810 0.813 -0.002 -0.26 EXOC2 -2.14 23 chr13 0.039 0.593 0.604 -0.011 -1.81 HTR2A -2.14 8 chr19 0.039 0.604 0.620 -0.016 -2.69 CLEC4G -2.13 36 chr1 0.039 0.507 0.533 -0.026 -5.02 NBL1 -2.12 13 chr11 0.039 0.272 0.284 -0.013 -4.51 NAALAD2 -2.12 14 chr4 0.039 0.343 0.414 -0.070 -18.62 TRAM1L1 -2.12 16 chr4 0.039 0.365 0.373 -0.008 -2.07 ARAP2 -2.11 13 chr19 0.039 0.384 0.403 -0.019 -4.74 PINLYP -2.11 16 chr19 0.039 0.308 0.327 -0.019 -5.94 KLK7 -2.09 13 chr2 0.039 0.250 0.273 -0.024 -9.02 CASP10 144  -2.08 13 chr9 0.039 0.240 0.253 -0.013 -5.47 BARHL1 -2.08 15 chr11 0.039 0.706 0.712 -0.006 -0.84 TRIM29 -2.07 6 chr21 0.039 0.653 0.709 -0.057 -8.31 ABCC13 -2.07 6 chr10 0.039 0.709 0.732 -0.023 -3.21 DHX32 -2.06 15 chr5 0.039 0.400 0.425 -0.025 -6.06 ZNF300 -2.06 17 chr6 0.039 0.464 0.471 -0.007 -1.54 RNGTT -2.05 15 chr2 0.039 0.739 0.760 -0.021 -2.80 KCNE4 -2.05 5 chr1 0.039 0.377 0.488 -0.111 -25.66 GSTM5 -2.05 8 chr8 0.039 0.640 0.649 -0.008 -1.32 DEFA4 -2.05 7 chr6 0.039 0.295 0.369 -0.075 -22.53 SNORD52 -2.04 15 chr2 0.039 0.186 0.203 -0.017 -8.87 DUSP19 -2.04 13 chr3 0.039 0.256 0.276 -0.020 -7.47 LARS2 -2.01 41 chr20 0.039 0.690 0.692 -0.002 -0.31 NNAT -2.00 8 chr12 0.039 0.294 0.327 -0.034 -10.91 TMBIM4 -1.98 8 chr16 0.039 0.357 0.386 -0.030 -7.99 IL17C -1.97 7 chr12 0.039 0.456 0.481 -0.026 -5.47 C12orf45 -1.96 39 chr8 0.039 0.363 0.390 -0.027 -7.20 FDFT1 -1.95 8 chr7 0.039 0.475 0.501 -0.026 -5.34 GIMAP1 -1.92 6 chr16 0.039 0.278 0.281 -0.003 -1.07 BRD7 -1.87 6 chr1 0.039 0.708 0.722 -0.014 -1.93 LCE2D -1.87 6 chr6 0.039 0.436 0.460 -0.024 -5.43 TXLNB -1.86 6 chr17 0.039 0.766 0.781 -0.015 -1.90 C17orf112 -1.85 5 chr3 0.039 0.618 0.622 -0.004 -0.58 LOC101929337 -1.85 5 chr4 0.039 0.617 0.627 -0.010 -1.57 ADH1B 1.84 86 chr2 0.039 0.688 0.681 0.008 1.12 SH3BP4 1.89 5 chr1 0.039 0.486 0.475 0.012 2.41 CHTOP 1.89 5 chr19 0.039 0.621 0.549 0.072 12.23 C3 1.93 8 chr19 0.039 0.817 0.809 0.008 0.95 APOC4 1.96 47 chr11 0.039 0.669 0.658 0.010 1.53 P2RY6 1.99 9 chr22 0.039 0.688 0.665 0.022 3.32 C22orf26 2.00 8 chr20 0.039 0.478 0.447 0.031 6.72 LOC149837 2.01 12 chr1 0.039 0.708 0.695 0.013 1.81 CDCP2 2.04 8 chr2 0.039 0.713 0.694 0.019 2.71 LOC100189589 2.04 12 chr8 0.039 0.395 0.355 0.040 10.75 HTRA4 2.04 10 chr11 0.039 0.583 0.554 0.029 5.02 SPDYC 2.05 9 chr21 0.039 0.775 0.743 0.032 4.18 PCBP3 2.05 13 chr18 0.039 0.865 0.823 0.042 5.01 TCEB3B 2.06 10 chr4 0.039 0.427 0.408 0.019 4.51 TRIM61 2.07 13 chr6 0.039 0.327 0.277 0.050 16.67 HCG4P6 2.07 9 chr15 0.039 0.253 0.227 0.025 10.52 C15orf26 2.07 25 chr14 0.039 0.529 0.517 0.011 2.16 JDP2 145  2.08 57 chr6 0.039 0.277 0.263 0.014 5.10 AGPAT1 2.08 13 chr19 0.039 0.751 0.742 0.008 1.13 AURKC 2.10 8 chr20 0.039 0.688 0.677 0.011 1.68 SNORA51 2.12 11 chr1 0.039 0.755 0.743 0.012 1.56 NR0B2 2.14 31 chr6 0.039 0.173 0.158 0.015 9.29 RNF5P1 2.15 40 chr4 0.039 0.490 0.485 0.005 1.07 PRDM8 2.17 30 chr17 0.039 0.425 0.403 0.023 5.52 TEX2 2.18 22 chr8 0.039 0.290 0.257 0.033 11.90 TP53INP1 2.18 22 chr11 0.039 0.163 0.157 0.007 4.10 ZNF214 2.19 19 chr1 0.039 0.279 0.268 0.011 4.05 SGIP1 2.19 12 chr13 0.039 0.727 0.713 0.014 2.01 F10 2.19 30 chr6 0.039 0.144 0.127 0.017 12.37 RNF5 2.21 12 chr16 0.039 0.570 0.547 0.022 4.00 ZNF205 2.22 13 chr13 0.039 0.576 0.555 0.021 3.69 DLEU7 2.22 26 chr17 0.039 0.696 0.682 0.014 2.02 WFIKKN2 2.23 34 chr17 0.039 0.491 0.482 0.009 1.86 HOXB6 2.23 13 chr17 0.039 0.374 0.385 -0.011 -2.81 MYCBPAP 2.23 13 chr12 0.039 0.132 0.123 0.010 7.45 C12orf61 2.25 9 chr12 0.039 0.535 0.520 0.015 2.83 LOC144571 2.28 16 chr5 0.039 0.253 0.231 0.023 9.35 CCNG1 2.30 25 chr7 0.039 0.293 0.287 0.006 1.98 HOXA2 2.32 14 chr1 0.039 0.797 0.785 0.012 1.53 GPR37L1 2.34 73 chr12 0.039 0.678 0.668 0.009 1.39 TSPAN9 2.49 36 chr11 0.039 0.610 0.590 0.020 3.37 H19 2.57 19 chr10 0.039 0.139 0.134 0.005 3.87 C10orf4 NES: normalized enrichment score of each DMR DMR size: The number of CpG sites in the DMR Chr: Chromosome Beta difference: The difference in mean DNA methylation (beta values) between the IVF and NC groups. Positive value corresponds to hypermethylation while a negative value corresponds to hypomethylation in the IVF group. Beta % difference: The percent difference in mean DNA methylation (beta values)    146  A.3 Overlapping DMRs between the DMRs in the ICSI and IVF groups at an FDR less than 0.05 ICSI adj. p-val  IVF adj. p-val  ICSI NES IVF NES ICSI % beta difference IVF % beta difference DMR size Chr Gene 0.040 0.039 -2.23 -2.21 -17.49 -14.05 7 chr1 AKR7L 0.040 0.039 2.50 2.08 5.88 1.13 13 chr19 AURKC 0.040 0.039 2.95 2.57 36.02 3.87 19 chr10 C10orf4 0.040 0.039 -2.32 -2.46 -21.36 -43.88 14 chr17 C17orf97 0.040 0.039 -1.87 -2.07 -0.70 -3.21 6 chr10 DHX32 0.040 0.039 2.13 2.22 1.86 3.69 13 chr13 DLEU7 0.040 0.039 -2.29 -2.16 -2.43 -3.40 13 chr5 ESM1 0.040 0.039 -2.08 -2.14 -0.48 -0.26 33 chr6 EXOC2 0.040 0.039 2.08 2.19 0.12 2.01 12 chr13 F10 0.040 0.039 -2.11 -1.96 -13.36 -7.20 39 chr8 FDFT1 0.040 0.039 -2.05 -2.05 -38.30 -25.66 5 chr1 GSTM5 0.040 0.039 2.33 2.49 4.07 3.37 36 chr11 H19 0.040 0.039 2.12 2.23 1.20 1.86 34 chr17 HOXB6 0.040 0.039 -2.48 -2.36 -0.67 -0.38 12 chr6 HUS1B 0.040 0.039 -2.11 -2.05 -3.11 -2.80 15 chr2 KCNE4 0.040 0.039 -2.07 -2.11 -7.53 -5.94 16 chr19 KLK7 0.040 0.039 -2.27 -2.17 -6.82 -10.82 9 chr5 LOC134466 0.040 0.039 2.38 2.25 2.58 2.83 9 chr12 LOC144571 0.040 0.039 -1.98 -2.20 -2.60 -2.12 8 chr16 MSLNL 0.040 0.039 2.47 2.23 -1.26 -2.81 13 chr17 MYCBPAP 0.040 0.039 -2.44 -2.12 -10.99 -4.51 13 chr11 NAALAD2 0.040 0.039 -2.69 -2.01 -1.49 -0.31 41 chr20 NNAT 0.040 0.039 -2.29 -2.22 -15.10 -12.50 15 chr7 NPY 0.040 0.039 1.94 1.84 0.89 1.12 86 chr2 SH3BP4 0.040 0.039 -2.10 -2.33 -8.07 -10.27 9 chr17 SKAP1 0.040 0.039 -2.02 -2.05 -21.41 -22.53 7 chr6 SNORD52 0.040 0.039 2.33 2.18 3.69 11.90 22 chr8 TP53INP1 0.040 0.039 -2.08 -2.12 -13.81 -18.62 14 chr4 TRAM1L1 0.040 0.039 -2.08 -2.08 -3.07 -0.84 15 chr11 TRIM29 0.040 0.039 2.12 2.06 7.84 4.51 10 chr4 TRIM61 0.040 0.039 2.18 2.34 0.45 1.39 73 chr12 TSPAN9 0.040 0.039 2.37 2.22 2.99 2.02 26 chr17 WFIKKN2 0.040 0.039 2.09 2.21 0.48 4.00 12 chr16 ZNF205 0.040 0.039 2.11 2.18 -0.37 4.10 22 chr11 ZNF214 0.040 0.039 -2.43 -2.46 -9.16 -14.00 15 chr19 ZNF577 147  ICSI adj. p-val: Adjusted p-value between ICSI and NC group, FDR using Benjamini-Hochberg IVF adj. p-val: Adjusted p-value between IVF and NC group, FDR using Benjamini-Hochberg ICSI NES: Normalized enrichment score for the ICSI and NC group comparison IVF NES: Normalized enrichment score for the IVF and NC group comparison ICSI % beta difference: The percent difference in mean DNA methylation (beta values) between the ICSI and the NC groups. Positive value corresponds to hypermethylation while a negative value corresponds to hypomethylation in the ICSI group. IVF % beta difference: The percent difference in mean DNA methylation (beta values) between the IVF and the NC groups. Positive value corresponds to hypermethylation while a negative value corresponds to hypomethylation in the IVF group. DMR size: The number of CpG sites in the DMR Chr: chromosome   

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0384565/manifest

Comment

Related Items