Differential DNA Methylation as a Tool for Noninvasive Prenatal Diagnosis (NIPD) of X Chromosome Aneuploidies

The demographic tendency in industrial countries to delay childbearing, coupled with the maternal age effect in common chromosomal aneuploidies and the risk to the fetus of invasive prenatal diagnosis, are potent drivers for the development of strategies for noninvasive prenatal diagnosis. One breakthrough has been the discovery of differentially methylated cell-free fetal DNA in the maternal circulation. We describe novel bisulfite conversion- and methylation-sensitive enzyme digestion DNA methylation-related approaches that we used to diagnose Turner syndrome from first trimester samples. We used an X-linked marker, EF3, and an autosomal marker, RASSF1A, to discriminate between placental and maternal blood cell DNA using real-time methylation-specific PCR after bisulfite conversion and real-time PCR after methylation-sensitive restriction digestion. By normalizing EF3 amplifications versus RASSF1A outputs, we were able to calculate sex chromosome/autosome ratios in chorionic villus samples, thus permitting us to correctly diagnose Turner syndrome. The identification of this new marker coupled with the strategy outlined here may be instrumental in the development of an efficient, noninvasive method of diagnosis of sex chromosome aneuploidies in plasma samples.Aneuploidies, ie, an abnormal number of chromosomes, are responsible for a range of genetic disorders. The most frequent aneuploidies compatible with life are represented by trisomy 21 causing Down syndrome, trisomy 13 causing Patau syndrome, trisomy 18 causing Edwards syndrome, and sex chromosome aneuploidies.Sex chromosome aneuploidies include 45,X causing Turner syndrome (1/2500 living females), 47,XXX associated with triple X syndrome (1/1000 live births), 47,XXY associated with Klinefelter syndrome (prevalence 1/500 live males) (for prevalence and incidence data refer to http://www.orpha.net/consor/cgi-bin/index.php, last accessed on March 9, 2010) and the 47,XYY karyotype (the incidence generally reported is 1/1000 live births). Compared to trisomy 21, 13, and 18, sex chromosome aneuploidies show less severe clinical phenotypes, but taken together, the incidence of sex chromosome aneuploidies is high and, in the case of 47,XXX, 47,XXY and 47,XYY, this is largely underestimated. Moreover, although the mortality in utero of fetuses with Turner syndrome is high, most fetuses with other sex chromosome aneuploidies survive to term.¹ Major malformations may occur in Turner syndrome, but not in the XXY Klinefelter, XXX and XYY syndromes.²To date, conventional prenatal diagnosis of genetic disorders has been based on the analysis of fetal cells obtained using invasive procedures such as amniocentesis and chorionic villus sampling (CVS). These techniques are very reliable, but the downside is that both are associated with a small but significant risk of fetal loss, ie, in the order of 0.5 to 1.0% of cases. For this reason, invasive prenatal diagnosis is offered only if the perceived risk of abnormal pregnancy, estimated by maternal age, ultrasonography and other noninvasive methods, exceeds the miscarriage risk.³Several groups have investigated noninvasive methods of prenatal diagnosis.³ Attempts have been made to isolate fetal nucleated cells from maternal blood^4,5,6 but their rarity and the possibility of cells persistent from previous pregnancies have so far made this strategy unreliable.Recent strategies for noninvasive prenatal diagnosis (NIPD) have been based on the observation in 1948 of the presence of cell free circulating nucleic acid in blood plasma⁷ and the increase in this plasma DNA in cancer.^8,9,10 More recently, Lo et al¹¹ demonstrated the presence of male fetal DNA in maternal plasma, by amplifying Y specific sequences. Moreover, it was found that cell free fetal DNA (cffDNA) in maternal plasma is fragmented¹² and the half-life is in the order of 16 minutes after delivery.¹³ The amount of cffDNA in maternal plasma DNA ranges between approximately 3 to 6% with a mean of 25.4 genome copies/ml of maternal plasma during early pregnancy.¹⁴ cffDNA from maternal plasma has been successfully used to determine fetal rhesus D (RhD) blood type,¹⁵ for determination of fetal sex,^14,16,17 thus limiting the need for invasive diagnosis in cases of sex specific pathologies as well as for the identification of some fetal disorders due to paternal genetic mutations or recessive conditions where parents are compound heterozygotes.^18,19 However, the presence of a great excess of free maternal DNA complicates the use of such methods.¹⁴The most important source of cffDNA released in maternal plasma during pregnancy appears to be the placenta,^20,21 whereas it has been suggested that the cell free maternal DNA (cfmDNA) originates from hematopoietic cells.²² On the basis of the placental origin of free fetal nucleic acids (cffDNA and cffRNA²³) and the finding of a chromosome 21 placenta-specific mRNA marker in maternal plasma,²⁴ an approach based on RNA-SNP allelic ratio has been reported, to detect aneuploidies of this chromosome.²⁴ Recently, Lo and co-workers²⁵ reported the use of digital PCR to determine the over-representation of chromosome 21 in trisomy 21 samples in mixtures of placental and maternal blood cell DNA, using samples containing at least 25% of fetal DNA, a concentration many fold higher than that present in a first trimester maternal plasma sample. A similar strategy, based on microfluidic digital PCR platform has been applied by Fan and co-workers²⁶ in the set-up of diagnosis of chromosome number abnormalities, on amniotic fluid and CVS samples.Recently, high throughput technologies, such as those based on parallel DNA-sequencing²⁷ are being applied to NIPD strategies. Read depth analysis was used to successfully identify chromosome 13, 18, and 21 aneuploidies of the fetus.²⁸ Chiu and co-workers²⁹ also applied high throughput DNA-sequencing to quantify the amount of unique chromosome 21 sequences from plasma (maternal and fetal), revealing a potential trisomy. However, high throughput DNA-sequencing is still costly and difficult to manage in a routine laboratory, due to the large bioinformatic and computer resources required for analysis. For the foreseeable future, it may be difficult to translate this approach for widespread noninvasive diagnosis of aneuploidies.Strategies exploiting differential DNA methylation may also be used to discriminate tissues/cells of different origin. DNA methylation, the major post-biosynthetic modification found throughout mammalian genomes, is involved in many important biological phenomena, such as X chromosome inactivation and genomic imprinting, and in controlling tissue-specific expression in adult somatic tissues.^30,31,32Differential DNA methylation between maternal and fetal DNA has been investigated for use in NIPD. Differential methylation between fetal CVS and maternal blood cell DNA was first reported in 2002³³ and three years later the first universal epigenetic marker of fetal DNA in maternal plasma, SERPINB5 was described. The promoter region of SERPINB5 is hypomethylated in placenta and hypermethylated in maternal blood cells. Fetal SERPINB5 was distinguished from maternal SERPINB5 in maternal plasma DNA³⁴ after bisulfite modification³⁵ and methylation specific PCR.³⁶ Subsequently, the application of bisulfite-independent methods were also reported.^37,38Despite the appeal of using epigenetic differences between maternal and fetal DNA to develop NIPD for aneuploidies, the search for chromosome specific markers is challenging: this is particularly true for X linked sequences. An alternative high-throughput approach for identifying chromosome specific methylated markers is based on immunoprecipitation of methylated DNA (MeDiP³⁹) coupled with high resolution tiling oligonucleotide array analysis. More than 2000 differentially methylated regions between placenta and maternal blood cells, on respectively chromosome 13, 18, 21, X and Y, both in non-genic regions and in the CpG islands have been recently reported using this approach.⁴⁰Taking advantage of these latter data, we selected three putative X-linked markers that are differentially methylated between placenta and maternal blood cell DNA, and after further validation we developed a NIPD strategy to determine copy number of fetal sex chromosomes.