Aberrant DNA methylation of imprinted loci in human spontaneous abortions after assisted reproduction techniques and natural conception.

STUDY QUESTION: Do assisted reproduction techniques (ARTs) affect DNA methylation of imprinted genes and does aberrant methylation of imprinted genes account for the incidence of human spontaneous abortion (SA)? SUMMARY ANSWER: Our results show that imprinting errors of imprinted genes may contribute to human SA, and the occurrence of aberrant methylation of imprinted genes in ART pregnancies was comparable with that in natural pregnancies. WHAT IS KNOWN ALREADY: Animal data and human studies demonstrated that in vitro culture of embryos can cause methylation defects in individual genes, which might affect subsequent embryonic development and contribute to SA. However, our previous studies showed an abnormal methylation pattern of PEG1 in human aborted chrionic villus samples (CVS) but an increased occurrence of aberrant methylation in CVS from ART-derived pregnancies was not observed. STUDY DESIGN, SIZE AND DURATION: CVS were collected from women who underwent abortion procedures in the Department of Gynecology and Obstetrics in Nanfang Hospital from May 2008 to July 2011. Muscle samples (MS) were obtained from aborted fetuses and stillbirths. The samples were divided into four experimental groups: (A) SA/stillbirth after ART (n = 75), (B) multi-fetal reduction after ART (n = 73), (C) SA/stillbirth of natural pregnancies (n = 90) and (D) induced abortion (IA) of natural pregnancies (n = 82). PARTICIPANTS/MATERIALS, SETTING AND METHODS: The mean ± SD age of patients was 31.0 ± 4.1 (range: 18-45 years). The DNA methylation patterns of one paternally methylated (H19) and two maternally methylated (LIT1 and SNRPN) genes were analyzed in CVS and MS using pyrosequencing and bisulfite sequencing PCR. MAIN RESULTS AND THE ROLE OF CHANCE: Clear hypo-methylation (<10%) or hyper-methylation (>90%) were not detected in LIT1 and SNRPN but two regions of hyper-methylation (91.7 and 91.4%) were observed in H19. The mean percentage of methylation in the SA samples (groups A and C) was higher than that in the IA samples (groups B and D; P<0.05). Box plot analyses showed that in the 165 SA samples, methylation values for 40/495 (8.1%) differentially methylated regions of the three genes represented outliers. The incidence of outlier was highest for LIT1 (13.3%, 22/165). In contrast, no outliers were found in the 155 IA samples. The receiver operating characteristic curve analyses showed a positive correlation between percentage methylation of all three genes and incidence of SA (P<0.05). In addition, the conception modes (natural versus ART) and the fertilization methods used in ART (IVF and ICSI) did not affect the methylation patterns of the imprinted genes. No increase in the rate of abnormal methylation was found in the ART samples. LIMITATIONS AND REASONS FOR CAUTION: The studied loci represent only a small fraction of developmentally important genes. Further studies are needed to evaluate changes in the expression and the methylation status of other genes that may lead to SA. WIDER IMPLICATIONS OF THE
FINDINGS: The findings provide new insights into the etiology of human SA. The possibility that the abnormal methylation seen is a consequence of the defect that led to the SA cannot be excluded. STUDY FUNDING/COMPETING INTEREST(S): None of the authors has any competing interest. This study was supported by National Natural Science Foundation of China (81170574), The National Key Basic Research Development Plan of China (973 Program) (2007CB948104), Comprehensive strategic sciences cooperation projects of Guangdong Province and Chinese Academy (04020416) and Guangzhou Science and Technology Program key projects (11C22120737).