Erin C. Macaulay

Hypomethylation of functional retrotransposon-derived genes in the human placenta

DNA hypomethylation is assumed to be a feature of the mammalian placenta; however, its role in regulating placental gene expression is not well defined. In this study, MeDIP and Sequenom MassARRAY were used to identify hypomethylated gene promoters in the human placenta. Among the genes identified, the hypomethylation of an alternative promoter for KCNH5 was found to be restricted to the placenta and chorion. Complete methylation of this promoter correlates with a silenced KCNH5 transcript in embryonic tissues, including the amnion. Unusually, this hypomethylated promoter and the alternative first exon are derived from a SINE (AluY) retrotransposon. Examination of additional retrotransposon-derived gene promoters in the placenta confirmed that retrotransposon hypomethylation permits the placenta-specific expression of these genes. Furthermore, the lineage-specific methylation displayed by KCNH5, INSL4, and ERVWE1 revealed that dichotomous methylation establishes differential retrotransposon silencing between the extra-embryonic and embryonic lineages. The hypomethylation of the retrotransposons that regulate these genes, each of which arose during recent primate evolution, is consistent with these genes having functional roles that are unique to the invasive haemochorial placentas of humans and recent primates.

Placental Hypomethylation Is More Pronounced in Genomic Loci Devoid of Retroelements

The human placenta is hypomethylated compared to somatic tissues. However, the degree and specificity of placental hypomethylation across the genome is unclear. We assessed genome-wide methylation of the human placenta and compared it to that of the neutrophil, a representative homogeneous somatic cell. We observed global hypomethylation in placenta (relative reduction of 22%) compared to neutrophils. Placental hypomethylation was pronounced in intergenic regions and gene bodies, while the unmethylated state of the promoter remained conserved in both tissues. For every class of repeat elements, the placenta showed lower methylation but the degree of hypomethylation differed substantially between these classes. However, some retroelements, especially the evolutionarily younger Alu elements, retained high levels of placental methylation. Surprisingly, nonretrotransposon-containing sequences showed a greater degree of placental hypomethylation than retrotransposons in every genomic element (intergenic, introns, and exons) except promoters. The differentially methylated fragments (DMFs) in placenta and neutrophils were enriched in gene-poor and CpG-poor regions. The placentally hypomethylated DMFs were enriched in genomic regions that are usually inactive, whereas hypermethylated DMFs were enriched in active regions. Hypomethylation of the human placenta is not specific to retroelements, indicating that the evolutionary advantages of placental hypomethylation go beyond those provided by expression of retrotransposons and retrogenes.

The importance of early life in childhood obesity and related diseases: a report from the 2014 Gravida Strategic Summit

Obesity and its related non-communicable diseases (NCDs), such as type 2 diabetes, heart disease and cancer, impose huge burdens on society, particularly the healthcare system. Until recently, public health and policy were primarily focused on secondary prevention and treatment of NCDs. However, epidemiological and experimental evidence indicates that early-life exposures influence the risk of childhood obesity and related diseases later in life, and has now focused attention on the health of both mother and child. During pregnancy and the early neonatal period, individuals respond to their environment by establishing anatomical, physiological and biochemical trajectories that shape their future health. This period of developmental plasticity provides an early window of opportunity to mitigate the environmental insults that may increase an individual’s sensitivity to, or risk of, developing obesity or related diseases later in life. Although much investigation has already occurred in the area of Developmental Origins of Health and Disease research, the science itself is still in its infancy. It remains for researchers to tackle the important outstanding questions and translate their knowledge into workable solutions for the public good. The challenge, however, is to decide which areas to focus on. With these opportunities and challenges in mind, the 2014 Gravida Summit convened to examine how its early-life research program can determine which areas of research into mechanisms, biomarkers and interventions could contribute to the international research strategy to fight childhood obesity and its related diseases.

Retrotransposon Hypomethylation in Melanoma and Expression of a Placenta-Specific Gene

In the human placenta, DNA hypomethylation permits the expression of retrotransposon-derived genes that are normally silenced by methylation in somatic tissues. We previously identified hypomethylation of a retrotransposon-derived transcript of the voltage-gated potassium channel gene KCNH5 that is expressed only in human placenta. However, an RNA sequence from this placental-specific transcript has been reported in melanoma. This study examined the promoter methylation and expression of the retrotransposon-derived KCNH5 transcript in 25 melanoma cell lines to determine whether the acquisition of ‘placental’ epigenetic marks is a feature of melanoma. Methylation and gene expression analysis revealed hypomethylation of this retrotransposon in melanoma cell lines, particularly in those samples that express the placental KCNH5 transcript. Therefore we propose that hypomethylation of the placental-specific KCNH5 promoter is frequently associated with KCNH5 expression in melanoma cells. Our findings show that melanoma can develop hypomethylation of a retrotransposon-derived gene; a characteristic notably shared with the normal placenta.

Comparative assessment of DNA methylation patterns between reduced representation bisulfite sequencing and Sequenom EpiTyper methylation analysis

AIM Validation of sequencing-based DNA methylation data is an important step for meaningful translation of findings. However, there has been limited assessment of different platforms to validate methylation data from next generation sequencing. METHODS We performed a comparative methylation analysis between the genome-wide platform of reduced representation bisulfite sequencing with a targeted, Sequenom EpiTyper platform (four genes were analyzed in 15 cell lines covering 52 CpG sites). RESULTS We show that the accuracy of validation substantially improves if results from multiple and adjacent CpG sites are combined rather than at single CpG sites. We demonstrate increased read number improves accuracy of reduced representation bisulfite sequencing results. Further, by using series of replicates, we document variation in samples analyzed by Sequenom EpiTyper, indicating the importance of including replicates to increase precision. CONCLUSION The results reveal potential sources of bias and provide a guideline for refining study design for DNA methylation analysis.

The Genes of Life and Death: A Potential Role for Placental-Specific Genes in Cancer

The placenta invades the adjacent uterus and controls the maternal immune system, like a cancer invades surrounding organs and suppresses the local immune response. Intriguingly, placental and cancer cells are globally hypomethylated and share an epigenetic phenomenon that is not well understood – they fail to silence repetitive DNA sequences (retrotransposons) that are silenced (methylated) in healthy somatic cells. In the placenta, hypomethylation of retrotransposons has facilitated the evolution of new genes essential for placental function. In cancer, hypomethylation is thought to contribute to activation of oncogenes, genomic instability, and retrotransposon unsilencing; the latter, we postulate, is possibly the most important consequence. Activation of placental retrotransposon‐derived genes in cancer underpins our hypothesis that hypomethylation of these genes drives cancer cell invasion. This alludes to an interesting paradox, that while placental retrotransposon‐derived genes are essential for promoting early hominid life, the same genes promote disease‐susceptibility and death through cancer.

Human trophoblasts are primarily distinguished from somatic cells by differences in the pattern rather than the degree of global CpG methylation

ABSTRACT The placenta is a fetal exchange organ connecting mother and baby that facilitates fetal growth in utero. DNA methylation is thought to impact placental development and function. Global DNA methylation studies using human placental lysates suggest that the placenta is uniquely hypomethylated compared to somatic tissue lysates, and this hypomethylation is thought to be important in conserving the unique placental gene expression patterns required for successful function. In the placental field, methylation has frequently been examined in tissue lysates, which contain mixed cell types that can confound results. To better understand how DNA methylation influences placentation, DNA from isolated first trimester trophoblast populations underwent reduced representation bisulfite sequencing and was compared to publicly available data of blastocyst-derived and somatic cell populations. First, this revealed that, unlike murine blastocysts, human trophectoderm and inner cell mass samples did not have significantly different levels of global methylation. Second, our work suggests that differences in global CpG methylation between trophoblasts and somatic cells are much smaller than previously reported. Rather, our findings suggest that different patterns of CpG methylation may be more important in epigenetically distinguishing the placenta from somatic cell populations, and these patterns of methylation may contribute to successful placental/trophoblast function. Summary: The placenta may not be as uniquely hypomethylated as previously reported, rather differences in the pattern of CpG methylation are what make it epigenetically distinct.

Unravelling the Link Between the Placental Epigenome and Pregnancy Outcomes

Placental dysfunction underlies common and serious pregnancy complications such as fetal growth restriction (FGR), pre-eclampsia, preterm birth, and pregnancy loss. Although these complications most frequently manifest in midto late gestation, the origins of placental dysfunction often arise in early pregnancy and include maternal factors such as vascular insufficiency, smoking, and metabolic challenge [1]. Unfortunately, the mechanisms behind these conditions remain poorly understood. There is, however, increasing evidence that placental dysfunction from a variety of causes is associated with an altered epigenetic profile in the late-gestation placenta [2–5]. Recent work is beginning to uncover how the placental epigenome responds to the pregnancy environment. The epigenome contains three classes of tractable, quantifiable, and accessible information about the exposures of pregnancy. First, placental DNA contains programmed epigenetic foundations that are distinct from those of the fetus and essential for normal placental function, such as the hypomethylation of retrotransposon-derived genes [6–12]. Second, DNA methylation in the early embryo can be altered in response to environmental factors, leading to lifelong phenotypic change, as shown in mouse paradigms [13]. Third, the fetal epigenome can respond to different environments of pregnancy (obesity, insulin sensitivity, hypoxia, etc.) such as through changes in placental transporter genes [1]. The responsive nature of the placental epigenome may benefit fetal growth in suboptimal pregnancy conditions; however, it also may contribute to the development of pregnancy-related pathologies, such as FGR. The importance of epigenetics in the placenta has been highlighted by findings that show that epigenetic aberrations interfere with normal placental and fetal growth [14, 15]. However, it is unknown which specific epigenetic events compromise placental function, induce pregnancy-related pathologies, and have long-term effects on the fetus. Care, therefore, must be taken when attributing causality for these outcomes to epigenetic changes in late gestation [16]. The manuscript by Leeuwerke et al. [17] in a recent issue of Biology of Reproduction adds important information to this question by analyzing placental methylation patterns of genes associated with FGR from first trimester chorionic villus biopsy samples in pregnancies that continued to term. In contrast to the DNA methylation changes that have been reported in several genes in term placentae of infants born with either intrauterine growth restriction or small for gestational age (SGA) (reviewed in [18]), Leeuwerke et al. did not find any methylation differences in the placental samples obtained at a median gestational age of 11.3 wk between those pregnancies that went on to deliver an SGA baby at term and those that delivered an appropriately grown baby. These data indicate that the methylation changes associated with FGR must arise in the second or third trimesters. Also of note is that one-third of the mothers of SGA pregnancies were smokers whereas none of the control pregnancies were. Smoking in pregnancy is a major cause of FGR and has been associated with changes in the methylation profile of placental DNA at term [5]. If smokingrelated epigenetic changes arose in early pregnancy, one might have expected to see differences in methylation profiles between SGA and control cases. Overall, the data reported by Leeuwerke et al. support the hypothesis that the methylation changes previously reported in term placentae are a consequence of the pathological pregnancy and may not be causally related to the pathology or, indeed, even the consequences, such as growth restriction. This last point requires further research. Although most FGR becomes apparent in the third trimester, with more severe cases presenting in the second trimester, there is circumstantial evidence suggesting that size at birth may be determined, at least in part, in the first trimester [19, 20]. It clearly is not possible to obtain placental samples beyond the first trimester, except at delivery, but newer techniques such as profiling of fetal and placental DNA in maternal plasma [21] may facilitate further insights into the timing of the changes in the epigenetic profile in pregnancies complicated by placental dysfunction and a greater understanding of cause and consequence. Searching for a methylation change in the epigenome can be like looking for a needle in a haystack. Although Leeuwerke et al. analyzed candidate genes for FGR that were supported by the literature, their study may have benefited more from performing a genomewide analysis to detect methylation changes associated with FGR in first trimester placenta. The rapidly advancing field of genomics has greatly improved the tools available for genomewide epigenetic analysis. The cost of genomewide studies has drastically reduced in recent years, making it a relatively small investment for an incredible amount of data. Genomewide methylation data, generated from either reduced representation bisulfite sequencing [22] or whole genome bisulfite sequencing [23, 24], would be a tremendous resource to investigate placental dysfunction in FGR pregnancies. Mining of such data could reveal how the placental epigenome responds to environmental stimuli, which Correspondence: E-mail: f.bloomfield@auckland.ac.nz