Terminology

CpG

Region of DNA where a cytosine nucleotide is followed by aguanine in the linear sequences of bases

DMCpGs

Differentially methylated CpGs

RFM

Respiratory function monitor

VT

Tidal volume

Overview

Objectives

To determine whether the amount of oxygen provided during postnatal stabilization changes the DNA methylome in preterm infants.

Study design

This prospective, observational study included 32 preterm infants ≤32 weeks of gestation who received oxygen in the delivery room.

Patients were monitored using a respiratory function monitor to determine the amount of oxygen received upon stabilization.

Blood samples were processed for comparison of DNA methylation before and after resuscitation using a DNA methylation high-resolution micro array InfiniumHuman DNA methylation EPIC850KBeadChip.

Results

The median amount oxygen provided to preterm infants during stabilization was 644 mLO₂/kg.

Male sex and vaginal delivery were associated with increased oxygen needs.

There were 2626 differentially methylated CpGs representing 1567 genes that showed an association with oxygen load selected and, of these, 85% were hypomethylated.

We found that oxygen loads of >500 mLO₂/kg changed the methylation pattern of the selected CpGs.

Genes associated with these CpGs were “enriched” in KEGG pathways involved in cell cycle progression, DNA repair, and oxidative stress.

Conclusions

The oxygen load provided upon resuscitation modified the DNA methylome.

Differential methylation may lead to altered expression of genes related to cell cycle progression, oxidative stress, and DNA repair.

The reversibility of these early epigenetic changes is unknown but merits further study.

Introduction

Significance

Prematurity is one of the leading causes of morbidity and mortality in the first year of life.1

Methylation

Epigenetic changes refer to modifications in the DNA that do not involve changes in the sequence but have an impact on gene transcription and function and have been linked to the development of specific diseases.2-4

The most widely studied epigenetic change is DNA methylation, which is characterized by the addition of a methyl group to the 5′ position of a cytosine in a cytosine guanine dinucleotide (region of DNA where a cytosine nucleotide is followed by a guanine in the linear sequences of bases[CpG]) site.

Increased methylation levels in promoters silence gene transcription and demethylation creates a transcriptionally permissive state by relaxing the chromatin.5

Oxygen Therapy

The stabilization of preterm infants after birth often requires oxygen.

However, preterm infants frequently spend a substantial time either in hypoxia or hyperoxia.6

Hyperoxia

Hyperoxia is associated with the generation of free radicals that cause structural and functional alterations to biomolecules.7

Preterm infants have immature antioxidant defenses and are very susceptible to the harmful effects of oxidative stress, including DNA guanine base oxidation.8,9

There is a relationship between oxygen, oxidative stress, and epigenetics during the perinatal period.

Histone demethylases require oxygen as a cofactor,directly linking epigenetic processes to oxygen gradients during development.10

In addition, oxygen free radicals cause changes in the cellular redox status and could also be considered environmental factors capable of modifying DNA methylation pattern, thus affecting the expression of specific genes after neonatal supplementation of oxygen.10

Hypothesis

We hypothesized that oxygen load during post natal stabilization might cause a general pattern of hypomethylation.

In a proof-of-concept study, we calculated the oxygen load during stabilization in the delivery room using a respiratory function monitor (RFM) and assessed its impact on DNA methylation at the genomic level using a high-resolution microarray in the human genome.

Methods

Location

This prospective, observational study was performed in the neonatal intensive care unit of the UPH La Fe (Valencia, Spain) between June 1, 2016 and May 31, 2017.

Subjects

We studied 32 eligible infants <32 weeks of gestation who required positive pressure ventilation and oxygen in the delivery room.

Spontaneously breathing babies or those who initiated effective spontaneous breathing during stabilization not requiring positive pressure ventilation and oxygen were excluded.

In addition, infants with severe neonatal depression (Apgar score at 1 minute of <3), having severe malformations or chromosomopathies, or not following the study protocol were also excluded.

Immediately after birth, a pulse oximeter sensor was attached to the right wrist to monitor SpO2 and heart rate. Positive pressure ventilation was provided using silicon facemasks coupled to a T-piece resuscitator (Giraffe T-piece resuscitator, GE Health Care, Helsinki, Finland) with an initial pressure of 6 cm H2O. Pressure limits were set at 20-25 cm H2O and the positive end-expiratory pressure was set at 5-6 cm H2O. The initial FiO2 of 0.3 was titrated to achieve SpO2 targets of 70%-75%,75%-80%,80%-85%,and 85%-95% at 3,4,5, and 10 minutes, respectively.11 Simultaneously, an RFM (Newlifebox, Weener, Germany) was connected between the inspiratory line and the face mask interface (Figure 1; available at www.jpeds.com). The RFM also received input from the air/oxygen blender, the oxygen analyzer (Teledyne Analytical Instruments, City of Industry, California) located in the inspiratory limb of the T-piece resuscitator, pulse oximeter (Masimo Corporation, Irvine, California) and retrieved respiratory graphs, respiratory rate, expiratory tidal volume (VT), peak inspiratory pressure, end expiratory pressure, FiO2, heart rate, and SpO2. All the information was exported to a correlational database.

The oxygen load provided during the initial resuscitation was retrieved using an RFM. DNA methylome analyses before and after oxygen supplementation were completed. Oxygen load was defined as the milliliters of O2 per kilogram provided during delivery room stabilization (heart rate >100 bpm; SpO2 ≥90% for >1 minute). We retrieved theVT per breath and total breaths. The oxygen analyzer provided information of the oxygen content in the inspired gas. If babies were on room air, VT was multiplied by 0.21. Oxygen load per breath (B) was calculated as: OLB = [VT × FiO2]/kg, and the total oxygen load during stabilization was calculated as: S (B1 + B2 + ...Bn), where n is the number of breaths during stabilization. Investigators were trained in the use of the RFM in mannequins and with live patients. Reproducibility in 3 successive measurements of RFM variables were assessed at the end of the training period. Moreover, readings of the monitor were stored and the total oxygen load calculated as breath × breath to confirm algorithm’s accuracy.

Blood (0.5 mL) was sampled using a heparinized syringe at (t1) before oxygenation in umbilical artery and (t2) after admission in the neonatal intensive care unit. Blood was centrifuged (1500g × 10 minutes) at 4°C to separate plasma from the cell pellet. Cell fractions were stored at-80°C until processed. Epigenomic studies were performed with the Infinium Human DNA Methylation EPIC 850K BeadChip arrays (Illumina Inc, San Diego, California) used for the interrogation of >850 000 CpG sites. The protocol followed is detailed in the Supplementary Methods (Appendix; available at www.jpeds.com).

Statistical Analyses

Clinical data were summarized using mean, median, and first and third quartiles for continuous variables and relative and absolute frequencies for categorical variables.

Changes in the methylation status of specific CpGs were assessed using mixed beta regression models.12 Beta regression naturally accommodates responses between 0 and 1 such as percent methylation values (beta values) obtained by the methylation 850K array.12 The methylation score for each CpG was represented as a beta value according to the fluorescent intensity ratio. Beta values may take any value between 0 (no methylated) and 1 (completely methylated). The raw data (IDAT files) were normalized using functional normalization as implemented in the R-package minfi (version 1.22.1, The R Foundation for Statistical Computing, Vienna, Austria). Because observations were paired, the individuals were added as a random factor to the models to account for the non independence of observations in our pre–post design. All P-values were adjusted using the Benjamini-Hochberg procedure to control the False Discovery Rate. Adjusted P values of <.05 were considered statistically significant.

For the search of the KEGG Pathways of Gene Ontology, we used the annotations found by GeneCodis3.[^13] P values of <.03 have been obtained through hypergeometric analysis corrected by the false discovery rate method.

Results

Summary

A total of 32 eligible preterm infants <32 weeks of gestation were included in the study (Figure 2; available at www.jpeds.com).

The clinical characteristics of the population and perinatal interventions and the individual gestational age, sex, birth weight, and type of delivery and oxygen load are shown in Tables I and II,respectively.

The mean and standard deviation for delivery room stabilization was 11.5 ± 2.5 minutes.

Oxygen load ranged from 275 to 976 O2 mL/kg, with a median of 644 mL O2/kg.

No significant differences were found in relation to the oxygen load during stabilization relative to birth weight (Figure 3, A and B).

However, male sex (Figure 3, C) and vaginal delivery (Figure 3, D) were associated with significantly greater oxygen load during stabilization (P < .01).

We assessed the effect of oxygen load on epigenetic status, considering oxygen load as a continuous variable to improve statistical power and accuracy and allow for better interpret ability of the results. A total of 2626 differentially methylated CpGs(DMCpGs)were identified(falsediscoveryrate-adjusted P<.05)and were associated with 1567RefSeqgenes(Supple ment;availableatwww.jpeds.com).

We assessed the methylation profileof pretermi nfants rela tive to oxygen load and found a gradual effect of oxygen load on methylation status,with a diffusely defined threshold around 500mLO2/kg of weight (Figure4,A).At oxygen loads of≥500mLO2/kg of weight, we identified a clear loss of meth ylation of 2196 CpGs(representing83.6%) and again of meth ylation of 430CpGs (representing16.4%;Figure4,B). Interestingly, the distribution of DMCpGs was not random (P<.001).

To address the functional implications of the signature found inDNAmethylation,we performed an enrichment analysis of functions using GeneCodis13withtheDMCpGs with potential role in gene expression that is found in islands and shores.Enrichmentanalysisofgenesassociatedwiththe lossofmethylationDMCpGsforannotationsinvolvingKEGG pathways revealed a significant enrichment for genes involved in pathways in cancer, interaction between cells or between cells and the extracellular matrix,andactincytoskel eton (Figure5, A). Pathways in cancer included relevant genes involved in the maintenance of genomic integrity and theDNAdamage checkpoint, such as BRCA1, CDKN2B, PERP, ATMIN, FOXM1, andTP53BP1, andDNA repair pathways including XAB2, XRCC6, GTF2H1, POLR2H, MTHFD1,ERCC5,andRAD51AP1(Supplement). Ingenes associated with gain of methylationDMCpGs, we found enrichment for functions involving cell adhesionandcell cycle progression (Figure5, B) such as CHEK1, POLD2, TP53,TOP2A,andRASSF1(Supplement).Finally,wefurther identified several genes encoding proteins involved in mito chondrial electron transport such asNDUFS2, NDUFB4, ATP5F1, ATP5A1, andCOX15; and response to oxidative stress such as DUOX1,DUOXA1,PRDX3,andATOX1.

DMCpGs of the MethylationEPIC array were classified based on their gene name and gene region of UCSC annotations and CpG context previously defined17 (Supplement). With respect to CpG context,CpGsites canbe located in islands, shores (regions flanking CpG islands ≤2kb distant with lower CpG density),shelves(regions≤2kb long flanking shores with a lower density of CpG sites), andopensea (very low CpG density regions throughout the genome).18 Almost one-half of the selected DMCpGs are found in the open sea (47%), one-third in the shores of CpG islands (32%), 15%withinCpG islands, and the rest in shelves (Figure6,A). We focused on DMCpGs within CpG islands and shores because shelves andopenseaplay a relatively minor role in the control of gene transcription.5 The overall distribution of theseDMCpGs revealed similar enrichment TableI. Clinical characteristics and perinatal interven tions in preterm infants≤32weeksofgestationen rolled in the study Variables Preterminfants (n=32) Gestationalage(weeks) 28(25-32) Birthweight(g) 1273.3±468.4 Apgarat1min 7(4-9) Apgarat5min 8(6-10) UmbilicalarterypH 7.31±0.05 Malesex 15(46.9) Typeofdelivery:vaginal/cesearean 10/22(31.1/68.9) Twinpregnancy 20(62.5) Antenatalsteroids* 29(90.6) Hypertensivestateduringpregnancy† 6(18.8) Chorioamnionitis‡ 10(31.2) Intubationinthedeliveryroom 12(37.5) Time after birth for obtaining secondbloodsample(min)† 125±18 Valuesaremedian(IQR),mean(SD),orn(%). *Full courseofantenatal steroids:12-mgdosesofbetamethasonegivenintramuscularly24 hoursapartorfour6-mgdosesofdexamethasoneadministeredintramuscularlyevery12hours.14 †Followingthedefinitionof theCanadianObstetricGuidelines.15 ‡Chorioamnionitisorintra-amnioticinfectionwasdefinedasanacuteinflammationofthemem branesandchorionof theplacentawithclinical,histologic,ormicrobiological findings.16 TableII. Preterminfants≤32weeksofgestationpost natalcharacteristicsandoxygenloadretrievedwithan RFMinthedeliveryroom Case numbers Gestational age (weeks) Birth weight (g) Sex (M/F) Oxygenload (mLO2perkg) Typeof delivery 1 25 750 Female 359 Cesearean 2 25 940 Female 393 Cesearean 3 27 795 Male 393 Vaginal 4 28 1200 Female 781 Vaginal 5 30 1000 Female 297 Cesearean 6 27 940 Female 475 Cesearean 7 32 1930 Female 408 Cesearean 8 32 2610 Male 546 Vaginal 9 32 1660 Female 470 Cesearean 10 30 1400 Male 275 Cesearean 11 25 850 Male 665 Vaginal 12 25 875 Female 857 Cesearean 13 27 925 Female 925 Cesearean 14 28 1025 Male 831 Vaginal 15 28 1150 Female 678 Cesearean 16 27 1090 Male 673 Vaginal 17 25 965 Male 819 Cesearean 18 32 1500 Male 408 Cesearean 19 32 1840 Female 398 Cesearean 20 28 1315 Female 955 Cesearean 21 29 1600 Male 885 Vaginal 22 29 1320 Male 623 Vaginal 23 25 870 Male 966 Vaginal 24 27 970 Male 848 Vaginal 25 26 680 Male 611 Cesearean 27 31 1685 Female 708 Vaginal 28 32 1860 Female 340 Cesearean 29 26 990 Female 389 Cesearean 30 32 1955 Female 343 Cesearean 31 32 1980 Female 674 Cesearean 32 27 730 Male 944 Cesearean

Figure 3. Correlation between oxygen received upon stabilization in preterm infants with clinical characteristics of the popu lation and perinatal interventions. A, Gestational age. B, Birth weight. C, Sex. D, Type of delivery. **P < .01 by 2-tailed t test. of loss and gain of methylation CpGs in promoters, body gene, and intergenic regions (Figure 6, B).

Discussion

The amount of oxygen provided during delivery room stabi lization was titrated to avoid hyperoxia or hypoxemia1 and quantified as an oxygen load using an RFM and an ad hoc al gorithm. This approach, in conjunction with the use of beta regression models, which improve the detection of differential DNAmethylation and take advantage of the continuous nature of both the response and the predictor variables,19 allowed us to establish an association between oxygen load and induced epigenetic changes. The median oxygen load (644 mL O2/kg; range, 275-976 mL O2/kg) was significantly higher than oxygen 4 load reported for term healthy newborn infants (~300 mL O2/kg).20

Wefound that male infants or those born vaginally needed significantly more oxygen to achieve stabilization (Figure 3, C). This finding may reflect the greater pulmonary and vas cular maturity of female neonates.21 Cesarean delivery has been advocated as protective for very preterm infants, although results have not been conclusive.22 Our results suggest one effect is a decreased oxygen load in the delivery room.

The effects of hyperoxia have been widely studied in animal models. For instance, hyperoxia-exposed newborn rats have arrested lung alveolarization secondary to epigenetic disrup tion of signaling pathways relevant to lung and neural devel opment and neuroprotection.23-25 We now show that human infants display short-term epigenetic disruption after neona tal resuscitation with oxygen.

Figure 4. Selection of DMCpGs in preterm newborns depending on the oxygen load received during stabilization. A, Heatmap including the CpGs displaying statistically significant differences in beta regression models. Rows (CpGs) and columns (indi viduals) are ordered according to the results of hierarchical clustering. Color scale ranges from red for hypermethylation to green for the hypomethylation. Different methylation profiles clustered higher and lower doses of oxygen load. Changes in methyla tion appear progressively as oxygen load increases with a diffuse common threshold for most CpGs around 500 mL O2/kg. B, The number of loss and gain of methylation DMCpGs with higher oxygen load.

Our epigenome-wide approach identified a signature of DMCpGsites with an association with the oxygen load during resuscitation. We found that exceeding a boundary of approxi mately 500 mL O2/kg was associated with altered methyla tion status; most of DMCpG sites lost methylation and others gained it. It has been reported that DNA oxidative damage pro vokes global hypomethylation and aberrant hypermethylation of some gene promoter regions,26 which is in agreement with our results. Base oxidation of the CpGs modifies the interac tions between the CpGs sites and transcription factors leading to changes in gene expression.27 In this scenario, the use of a higher FiO2 of 0.9 vs 0.3 in extremely preterm infants caused significant oxidative damage to DNA and led to increased re spiratory morbidity.9 Interestingly, preterm infants have limited nucleotide excision repair capacity.28 Persistent oxidized DNA triggers the genome-wide hypomethylation that has been as sociated with the development of tumors, obesity, diabetes, or autoimmune diseases.29-31 Epidemiologic studies in the US and Sweden established an association between the use of pure oxygen in the first minutes after birth and childhood cancer.32-34

Widespread methylation differences between term and ex tremely preterm infants have been reported, and these differences may interact with environmental and social risk factors to establish risk for future complications. However, follow-up studies of preterm and term infants through the end of adolescence showed that methylation differences are largely resolved by 18 years of age, although specific probes associ ated with preterm individuals remain and reflect a long-term epigenetic legacy of preterm birth.35

We found that a brief exposure to oxygen caused specific enrichment of functions involved in defense against oxida tive stress, interaction between cells or between cells and the extracellular matrix and actin cytoskeleton, cell adhesion, cell cycle progression, and cancer. We identified several genes that have been previously associated with oxidative stress and oxygen regulation such as DUOX1 responsible for the synthesis of dual oxidases and NADPH oxidase and associ ated with the production of reactive oxygen species36; and PRDX3 that encodes a mitochondrial protein involved in redox regulation and has an antioxidant function protecting cells against oxidative stress,37 and ATOX1 which encodes a copper chaperone that plays a role in the incorporation of copper to Cu-dependent antioxidant enzymes and also func tions as an antioxidant against superoxide and hydrogen peroxide.38 The Oxygen Load Supplied during Delivery Room Stabilization of Preterm Infants Modifies the DNA Methylation Profile

Figure 5. Gene ontology analysis of the genes associated to the selected DMCpGs present in CpG islands and shores. KEGGpathways significantly enriched in genes associated to A, loss of methylation and B, gain of methylation DMCpGs. All the analyses were performed using GeneCodis3 web tool with an adjusted P value from false discovery rate of <.03. ECM, extracellular matrix.

Our study has some limitations. The sample size is rela tively small. However, the lack of waiver of consent and the need for mask ventilation coupled with the RFM favored the exclusion of a substantial number of newborn infants. In addition, we only evaluated some of the specific antenatal stressors influencing DNA methylation at a single point in time.

Our results show that oxygen in the immediate postnatal period potentially induces changes in the DNA methylation pattern, especially when oxygen supplementation is >500 mL O2/kg. The durability of these methylome changes remains an important question and warrants further study. ■ 6 A B % DMCpGs 15% 47% 32% 6% CpG Island Shore Shelf Open sea Hypomethylated 35.00 30.00 25.00 20.00 15.00 10.00 5.00

Promotor Body Intergenic Figure 6. Genomic distribution of selected DMCpGs accord ing to the effect of oxygen load. A, Number of the selected DMCpGs by CpG context: CpG islands, shores, shelves and open sea. B, Relative abundance of loss and gain of DMCpGs according to gene promoters, gene bodies or intergenic regions. Submitted for publication Mar 13, 2018; last revision received Jun 13, 2018; accepted Jul 3, 2018

References

1. Vento M.Oxygen supplementation in the neonatal period: changing the paradigm. Neonatology 2014;105:323-31.
2. Murgatroyd C,Patchev AV,Wu Y,Micale V,Bockmühl Y,Fischer D,et al. Dynamic DNAmethylation programs persistent adverse effects of early life stress. Nat Neurosci 2009;12:1559-66.
3. Portela A, Esteller M. Epigenetic modifications and human disease. Nat Biotechnol 2010;28:1057-68.
4. Rakyan VK,Down TA,Balding DJ,Beck S.Epigenome-wide association studies for common human diseases. Nat Rev Genet 2011;12:529-41.
5. Paluch BE, Naqash AR, Brumberger Z, Nemeth MJ, Griffiths EA. Epigenetics: a primer for clinicians. Blood Rev 2016;30:285-95.
6. White LN, Thio M, Owen LS, Kamlin CO, Sloss S, Hooper SB, et al. Achievement of saturation targets in preterm infants <32 weeks’ gesta tional age in the delivery room. Arch Dis Child Fetal Neonatal Ed 2017;102:F423-7.
7. Saugstad OD, Sejersted Y, Solberg R, Wollen EJ, Bjørås M. Oxygenation of the newborn: a molecular approach. Neonatology 2012;101:315
9. Vento M,Aguar M,Escobar J,Arduini A,Escrig R,Brugada M,et al. An tenatal steroids and antioxidant enzyme activity in preterm infants: influence of gender and timing. Antioxid Redox Signal 2009;11:2945
11. Vento M,MoroM,Escrig R,Arruza L,Villar G,Izquierdo I,et al. Preterm resuscitation with low oxygen causes less oxidative stress, inflammation, and chronic lung disease. Pediatrics 2009;124:e439-49.
12. Hitchler MJ, Domann FE. An epigenetic perspective on the free radical theory of development. Free Radic Biol Med 2007;43:1023-36.
13. Wyckoff MH, Aziz K, Escobedo MB, Kapadia VS, Kattwinkel J, Perlman JM, et al. Part 13: neonatal resuscitation: 2015 American Heart Associa tion guidelines update for cardiopulmonary resuscitation and emer gency cardiovascular care. Circulation 2015;132(18 Suppl 2):S543-60.
14. Moran S, Arribas C, Esteller M. Validation of a DNA methylation microarray for 850,000 CpG sites of the human genome enriched in en hancer sequences. Epigenomics 2016;8:389-99.
15. Carmona-Saez P, Chagoyen M, Tirado F, Carazo JM, Pascual-Montano A. GENECODIS:aweb-based tool for finding significant concurrent an notations in gene lists. Genome Biol 2007;8:R3.
16. American College of Obstetricians and Gynecologists’ Committee on Prac tice Bulletins—Obstetrics. Practice Bulletin No. 171: management of preterm labor. Obstet Gynecol 2016;128:e155-64.
17. Magee LA,Pels A, Helewa M,Rey E,von Dadelszen P,SOGC Hyperten sion Guideline Committee. Diagnosis, evaluation, and management of the hypertensive disorders of pregnancy: executive summary. J Obstet Gynaecol Can 2014;36:575-6.
18. Tita ATN, Andrews WW. Diagnosis and management of clinical chorioamnionitis. Clin Perinatol 2010;37:339-54.
19. Sandoval J, Heyn H, Moran S, Serra-Musach J, Pujana MA, Bibikova M, et al. Validation of a DNA methylation microarray for 450,000 CpG sites in the human genome. Epigenetics 2011;6:692-702.
20. Sandoval J, Esteller M. Cancer epigenomics: beyond genomics. Curr Opin Genet Dev 2012;22:50-5.
21. Triche TJ, Laird PW, Siegmund KD.Beta regression improves the detec tion of differential DNA methylation for epigenetic epidemiology. bioRxiv [Internet]. http://biorxiv.org/content/early/2016/05/23/054643.abstract. Accessed April 27, 2018. 2016.
22. Vento M, Asensi M, Sastre J, Lloret A, García-Sala F, Viña J. Oxidative stress in asphyxiated term infants resuscitated with 100% oxygen. J Pediatr 2003;142:240-6.
23. Vento M, Cubells E, Escobar JJ, Escrig R, Aguar M, Brugada M, et al. Oxygen saturation after birth in preterm infants treated with continu ous positive airway pressure and air: assessment of gender differences and comparison with a published nomogram. Arch Dis Child Fetal Neona tal Ed 2013;98:F228-32.
24. Lee HC,Gould JB.Survival rates and mode of delivery for vertex preterm neonates according to small- or appropriate-for-gestational-age status. Pe diatrics 2006;118:e1836-44.
25. Chen C-M,Liu Y-C,Chen Y-J,Chou H-C.Genome-wide analysis of DNA methylation in hyperoxia-exposed newborn rat lung. Lung 2017;195:661
27. Sengoku T,Murray KM,Wilson ME.Neonatal hyperoxia induces altera tions in neurotrophin gene expression. Int J Dev Neurosci 2016;48:31-7.
28. Stockwell J, Jakova E, Cayabyab FS. Adenosine A1 and A2A receptors in the brain: current research and their role in neurodegeneration. Mol ecules 2017;22:pii: e676.
29. Menezo YJR, Silvestris E, Dale B, Elder K. Oxidative stress and altera tions in DNA methylation: two sides of the same coin in reproduction. Reprod Biomed Online 2016;33:668-83.
30. Dattilo M, Giuseppe D, Ettore C, Ménézo Y. Improvement of gamete quality by stimulating and feeding the endogenous antioxidant system: mechanisms, clinical results, insights on gene-environment interactions and the role of diet. J Assist Reprod Genet 2016;33:1633-48.
31. Vande Loock K, Ciardelli R, Decordier I, Plas G, Haumont D, Kirsch Volders M. Preterm newborns show slower repair of oxidative damage and paternal smoking associated DNA damage.Mutagenesis 2012;27:573
33. Eden A,Gaudet F,Waghmare A,Jaenisch R.Chromosomal instability and tumors promoted by DNA hypomethylation. Science 2003;300:455.
34. Vickers MH. Early life nutrition, epigenetics and programming of later life disease. Nutrients 2014;6:2165-78.
35. Knight AK,Smith AK.Epigenetic biomarkers of preterm birth and its risk factors. Genes (Basel) 2016;7:pii:e15. doi:10.3390/genes7040015.
36. Naumburg E, Bellocco R, Cnattingius S, Jonzon A, Ekbom A. Supple mentary oxygen and risk of childhood lymphatic leukaemia. Acta Paediatr 2002;91:1328-33.
37. Spector LG, Klebanoff MA, Feusner JH, Georgieff MK, Ross JA. Child hood cancer following neonatal oxygen supplementation. J Pediatr 2005;147:27-31.
38. Kang M-G, Kim Y-N, Lee JH, Szardenings M, Baek H-J, Kook H, et al. Clinicopathological implications of mitochondrial genome alterations in pediatric acute myeloid leukemia. Ann Lab Med 2016;36:101-10.
39. Cruickshank MN, Oshlack A, Theda C, Davis PG, Martino D, Sheehan P, et al. Analysis of epigenetic changes in survivors of preterm birth reveals the effect of gestational age and evidence for a long term legacy. Genome Med 2013;5:96.
40. DeDeken X,Corvilain B,Dumont JE,Miot F.Roles of DUOX-mediated hydrogen peroxide in metabolism, host defense, and signaling. Antioxid Redox Signal 2014;20:2776-93.
41. Rhee SG, Kil IS. Multiple functions and regulation of mammalian peroxiredoxins. Annu Rev Biochem 2017;86:749-75.
42. Muller PAJ, Klomp LWJ. ATOX1: a novel copper-responsive transcrip tion factor in mammals? Int J Biochem Cell Biol 2009;41:1233-6. The Oxygen Load Supplied during Delivery Room Stabilization of Preterm Infants Modifies the DNA Methylation Profile

Appendix

Oxygen Load during Stabilization of Preterm Infants Modifies Epigenetic Profile Supplementary Methods. DNAExtraction. Total DNA was isolated from the cell pellet with MagNAPure Compact NucleicAcid Isolation Kit 1-Larg Volume (03730972001, Roche Molecular Systems Inc., Basel, Switzerland) and MagNA Pure Compact Instrument (03731146001, Roche Molecular Systems Inc.), following the manufacturer’s instructions. Purified DNA was quantified with NanoDrop and RNA was removed with ribonuclease A from a bovine pancreas for molecular biology (R6513, Sigma Aldrich Co., Saint Louis, Missouri) at 10 µg/mL during 30 minutes at 60°C, was quantified by the fluorometric method (Quant-iT PicoGreen dsDNA Assay, Life Technologies, Carls bad, California), and assessed for purity with NanoDrop (Thermo Scientific,Waltham, Massachusetts) 260/280 and 260/ 230 ratio measurements. The DNA integrity of fresh frozen samples was checked by electrophoresis in 1.3% agarose gel. Epigenomic Studies. Infinium DNA MethylationEPIC beadchip array shares the Infinium HD chemistry Assay (Illumina Inc, San Diego, California) were used to interro gate the cytosine markers with HumanMethylation450 beadchip. Thus, the applicable protocol for MethylationEPIC is the same as for HumanMethylation450,which is the Infinium HDMethylation Assay Protocol that has been previously es tablished as a reliable technology to detect epigenetic altera tion in our and other laboratories.1 TheEPICarrayused interrogates >850 000 CpG sites (dinucleotides that are the main target for methylation) that covers 99% of genes described and 95% of CpG islands. The 850K array includes the informa tion of projects such as ENCODE and FANTOM5,which have identified regions as critical sites for differential methylation. 7.e1 This technology is highly reliable detecting epigenetic alterations.2 There were 600 ng of purified DNA distributed randomly on a96-well plate and processed using the EZ-96 DNA Meth ylation kit (Zymo Research Corp., Irvine, California) follow ing the manufacturer’s recommendations for Infinium assays. We processed 4 µL of bisulfite-DNA following the Illumina Infinium HD Methylation Assay Protocol, as previously described.2 This full protocol consisted of a whole genome am plification step followed by enzymatic endpoint fragmenta tion, precipitation, and resuspension. The resuspended samples were hybridized on Human Methylation 850K EPIC BeadChips at 48°C for 16 hours. Then, unhybridized and nonspecifically hybridized DNA were washed away,followed by a single nucleo tide extension using the hybridized bisulfite-treated DNA as a template. The nucleotides incorporated were labeled with biotin (ddCTP and ddGTP) and 2,4-dinitrophenol (DNP) (ddATP and ddTTP). After the single base extension, re peated rounds of staining were performed with a combina tion of antibodies that differentiated DNP and biotin by fixing them different fluorophores. Finally, the BeadChip was washed and protected to scan it. Scanning Beadchips. The Illumina HiScan SQ scanner is a 2-color laser (532 nm/660 nm) fluorescent scanner with a 0.375 µmspatial resolution capable of exciting the fluorophores gen erated during the staining step of the protocol. References

1. Sandoval J, Heyn H, Moran S, Serra-Musach J, Pujana MA, Bibikova M, et al. Validation of a DNA methylation microarray for 450,000 CpG sites in the human genome. Epigenetics 2011;6:692-702.
2. MoranS,Arribas C,Esteller M.Validation of a DNA methylation microarray for 850,000 CpG sites of the human genome enriched in enhancer se quences. Epigenomics 2016;8:389-99.

Figure 1. Ventilation and monitoring set up in the delivery room. Ventilation was provided from an air/oxygen blender, passed to a T-piece resuscitator and to the RFM transductor with a silicon mask interface. RFM stored videotaping of the resuscitation and clinical measures. Figure 2. Selection process for the patients studied in the present work. In a 12-month period, 184 preterm newborns of ≤32 weeks of gestation were attended in the delivery room. Of these, 92 were eligible for the study, 45 were recruited, and 32 com pleted the epigenomic study. The Oxygen Load Supplied during Delivery Room Stabilization of Preterm Infants Modifies the DNA Methylation Profile 7.e2