Objective
The purpose of this study was to investigate imprinting patterns in first-trimester human placentas.
Study Design
Using samples of 17 first-trimester and 14 term placentas from uncomplicated pregnancies, we assessed loss of imprinting (LOI) at the RNA level in a panel of 14 genes that are known to be imprinted in the placenta with the use of a quantitative allele-specific reverse transcriptase polymerase chain reaction analysis of those genes that contained readout single nucleotide polymorphisms in their transcripts.
Results
There is significant LOI (ie, biallelic expression) in all 14 genes in first-trimester placentas. LOI was more variable and generally at lower levels at term. Although there is little difference in gene expression, the level of LOI is higher in the first-trimester placentas, compared with term placentas.
Conclusion
Genomic imprinting appears to be a dynamic maturational process across gestation in human placenta. In contrast with prevailing theories, epigenetic imprints may continue to evolve past 12 weeks of gestation.
Genomic imprinting refers to the silencing of 1 parental allele in the fertilized embryo, depending on the parent of origin. Although most human genes are expressed in biallelic fashion (ie, both maternal and paternal alleles are concomitantly expressed), a small subset of genes that are known as imprinted genes are monoallelically expressed. Although there is a relatively small number of imprinted genes in the mammalian genome (approximately 1%), they play a critical role in reproduction and fetal development.
Imprinting is predominantly driven by epigenetic mechanisms, such as DNA methylation, histone modification, and RNA silencing. DNA methylation has been described as the primary epigenetic signal that controls genomic imprinting. For most genes, both the paternal and maternal genomes undergo a wave of demethylation in the preimplantation embryo that erases all but the imprinted methylation marks that are inherited from the parents. Shortly after implantation, a wave of de novo methylation establishes the new global methylation pattern; the imprinted parentally derived methylation setting is preserved in its differential allele-specific status. Later in fetal development, a small subset of cells is selected to become primordial germ cells. These cells undergo a second de-/remethylation wave that exclusively targets the imprinted domains. In this step, parental allele-specific imprinting marks are removed and biallelicaly replaced according to the sex of the developing embryo to be later allocated to haploid mature germ cells, as shown in animal models. At the same time during the preimplantation phase, histone reprogramming takes place. It replaces paternal protamines with histones and resets histone signaling across the entire genome. This takes place in a highly coordinated process that is intertwined heavily with the resetting of DNA methylation. The same enzymatic machinery is used in both processes. These events also happen jointly with the activation of the transcription from specific loci that produce both coding and noncoding RNAs with associated gene silencing/activation functions.
This imprinting process results in a stable parent (or gamete)-of-origin–specific mark on the genome that ultimately produces a functional difference between the genetic information that is contributed by each parent. The monoallelically expressed imprinted genes are functionally haploid. Thus, genomic imprinting is counter to classic Mendelian genetic theory (equal inheritance of both parental traits) because there is monoallelic parent-of-origin–specific expression from an imprinted autosomal locus. Because these epigenetic methylation marks on imprinted genes are believed to be maintained after fertilization, any acquired changes in the intrauterine environment may lead to stable transgenerational effects. Imprinted regions of the human genome appear to be affected by highly complex and long-range mechanisms (such as antisense RNA silencing and methylation-sensitive boundary elements) that are understood poorly at this time. This regulatory complexity of imprinted genes may render them particularly sensitive to environmental changes, such as diet and nutrition. Interestingly, emerging evidence implicates aberrant imprinting in many common human diseases, which include complications of pregnancy such as intrauterine growth restriction (IUGR), preeclampsia, and postnatal disorders such as obesity, cardiovascular disease, and type 2 diabetes mellitus.
We previously measured loss of imprinting (LOI) in imprinted genes that are expressed in human placenta and found that LOI was a common phenomenon in full-term placentas from uncomplicated human pregnancies. We also demonstrated that a different set of genes showed LOI in placentas from pregnancies that were complicated by preeclampsia and/or IUGR, compared with placentas from appropriately grown uncomplicated term pregnancies. It is fascinating to note that the shared pattern of LOI is consistent with the fact that both preeclampsia and IUGR have their pathophysiologic roots in the periimplantation period. Both disorders seem to be associated with shallow trophoblast invasion and aberrant placental implantation. This is in contrast to other complications of pregnancy (such as preterm labor or chorioamnionitis) in which pathophysiologic pathways appear to be of later onset and of different pathophysiologic origin. Therefore, for this study, we hypothesized that the normal pattern of LOI that is seen in term uncomplicated pregnancies would also be seen in the first trimester, as suggested by prevailing theories in the epigenetics literature that imprinting marks are stable across gestation. Once normal patterns of genomic imprinting in the first trimester of pregnancy are established, a “signature” pattern of aberrant imprinting in the first trimester can be determined for IUGR and preeclampsia in future studies.
The purpose of this study was to measure LOI in placental tissue in a cohort of uncomplicated first-trimester placentas from elective terminations of pregnancy to compare with LOI in uncomplicated third-trimester placentas with the use of a 14 imprinted gene panel that we previously developed. We also quantified the expression levels of these 14 imprinted genes in first-trimester human placental tissue as compared with term placenta.
Materials and Methods
Tissue collection and processing
After institutional review board approval, placental villous tissue was collected anonymously from women who underwent elective termination of pregnancy between 10 and 12 weeks of gestation. Women with a history of diabetes mellitus, hypertension, collagen-vascular disease, or other medical problems were excluded. Missed abortions, incomplete abortions, or fetuses with abnormal karyotype or anomalies were also excluded. After curettage, the products of conception were washed in sterile saline solution to clear tissue of blood and debris. Chorionic villi were identified, collected in sterile saline solution, and processed within 3 hours of the surgical procedure. The villous tissue was blotted to remove excess liquid and aliquoted for snap freezing for molecular studies. The tissue was kept on ice before being frozen in liquid nitrogen and stored at –80°C. We previously tested RNA stability in placental tissue and found it to be stable after collection for 3-6 hours when the samples were kept on ice before being frozen. Frozen villous tissue samples were later ground in a mortar (Scienceware/Bel-Art, Pequannock, NJ) and cooled by liquid nitrogen; the tissue powder was transferred to 2-mL screw-cap tubes and stored at –80°C until assayed.
Institutional review board approval was also obtained to use discarded samples from a previous study that used third-trimester placental collection. For that protocol, healthy women with uncomplicated pregnancies were identified and consented for placental donation. After delivery, the placenta was weighed and sampled with the use of an established protocol under the guidance of our Department of Pathology. Approximately 10 g of the placenta were collected for research purposes. Any areas on either maternal or fetal surface that appeared grossly abnormal were not sampled for research. Four biopsy samples that were free of maternal decidua were collected from each placental quadrant midway between the cord insertion and the placenta rim. The excised sections were approximately 1.5 cm in diameter, with an average thickness of 2.0-2.5 cm. The collected tissue was rinsed with cold saline solution and aliquoted for snap freezing for molecular studies. From this point forward the third-trimester placental tissue was handled in the same fashion as described for the first-trimester samples earlier.
RNA isolation
Total RNA was extracted in 3 steps: (1) the tissue powder was thawed in lysis buffer. (2) RNA was extracted with the RNeasy Mini Kit (Qiagen, Valencia, CA), according to the manufacturer’s instruction. (3) To ensure high RNA purity, 2 on-column DNA digestions were performed with DNase I. Isolated RNA was kept in RNase-free water at –80°C. We have shown that this procedure eliminates detectable DNA contamination.
DNA isolation
Genomic DNA was extracted from placental tissue with the QIAmp DNA Mini Kit (Qiagen), in accordance with the manufacturer’s instructions. DNA concentrations for the extraction were then measured with the DNA-specific dye PicoGreen in a fluorimeter setting to quantify the exact amount of DNA that had been isolated for each sample.
Genotyping
Informative single nucleotide polymorphisms (SNPs) that were used for LOI quantification were genotyped by direct sequencing. Briefly, primers that would bracket the SNP regions were designed. Genomic DNA was amplified and submitted to the Mount Sinai DNA Sequencing Core Facility (New York, NY). The amplicons were sequenced in both directions alternatively with the same forward and reverse genotyping primers.
Measuring LOI
To measure LOI quantitatively, complementary DNA (cDNA) was generated from total RNA by random primers with the AffinityScript Multiple Temperature cDNA Synthesis Kit (Stratagene, La Jolla, CA). The cDNA was cleaned with the PCR Purification Kit (Qiagen). CDNA was stored in diethylpyrocarbonate water at approximately 20ng/L concentration at –20°C.
In the previous work from our research group, we demonstrated that quantitative allele-specific polymerase chain reaction (qASPCR) was robust for the determination of allele frequencies in pooled DNA samples. 28 On the basis of similar technology, we developed a sensitive and functional assay for measuring allelic imbalance of imprinted genes (ie, LOI) on messenger RNA. Briefly, qASPCR is used on the reverse transcriptase-polymerase chain reaction (RT-PCR) product that contained a reporter marker, such as a readout SNP . SNP is defined as a minor variation in the DNA sequence in which a single base in the DNA differs from the usual base in that position that can be used to differentiate individuals. This mRNA-based assay that is used to measure LOI is independent of the mechanism of imprinting, which makes it more biologically relevant.
The LOI assay is a 2-step process. First, a common readout polymorphism that resides in the transcript of the studied gene is selected. Normal RT-PCR is used to amplify the sequence that contains the readout polymorphism. After that, qASPCR is applied to measure the relative abundance of the 2 heterozygous alleles, which allows quantification of LOI ( Figure 1 ). In individuals who are heterozygous for this readout polymorphism, LOI is a measurement of the presence of the supposedly silenced allele. LOI is calculated as: <SPAN role=presentation tabIndex=0 id=MathJax-Element-1-Frame class=MathJax style="POSITION: relative" data-mathml='LOI=2−|△Ct|’>LOI=2−|△Ct|LOI=2−|△Ct|
LOI = 2 − | △ Ct |
, where <SPAN role=presentation tabIndex=0 id=MathJax-Element-2-Frame class=MathJax style="POSITION: relative" data-mathml='|△Ct|’>|△Ct||△Ct|
| △ Ct |
is the difference between the allele-specific Ct values on cDNA, corrected for the specificity of the allele-specific PCR. This RNA-based assay can achieve the sensitivity of 1% LOI.
