Maternal effect mutations of NLRP7 or KHDC3L, two genes that are not present in rodents, cause recurrent biparentally inherited molar pregnancies (BiHM) with characteristic absent DNA methylation at imprinted germline differentially-methylated regions (gDMRs) that normally gain methylation in oocytes. This indicates that NLRP7 and KHDC3L are required for the determination of which gDMRs need to have their imprinting marks reprogrammed in the developing oocyte, and/or for the reprogramming process itself. These fundamental aspects of genomic imprinting are still poorly understood, especially in humans and other primates. Because rodents do not have NLRP7 and KHDC3L genes, precluding generation of mouse models, we established a human embryonic stem cell (hESC) model, using the H9 (WA09) line (NIH reg. 0062) to study their function. We discovered that stable NLRP7 knockdown in hESCs changed DNA methylation levels at many CpG sites and augmented BMP4-induced differentiation of hESCs into trophoblast. We also found that NLRP7 binds to KHDC3L, YY1, CTCF and the CpG-binding protein CFP1. Preliminary gene-expression and DNA methylation profiling data in hESC indicate that NLRP7 levels influence expression and methylation of a subset of genes that include epigenetic regulators with known or putative roles in imprinting. Furthermore, NLRP7 and KHDC3L are upregulated in dividing cells where they co-dependently localize to the mitotic spindle, an intriguing discovery considering that reprogramming of imprinting occurs during meiosis, a form of cell division, and needs to be maintained through mitotic cell divisions. These new findings led us to formulate our hypothesis, that NLRP7 and KHDC3L directly and cooperatively act in establishment or maintenance of imprinting marks at maternal gDMRs and that they are critical for recognition of a defined set of gDMRs that need to acquire DNA methylation or maintain it post-fertilization. We propose three specific aims to address this hypothesis.
Specific aim 1 is to investigate the cooperative function of KHDC3L with NLRP7 in reprogramming of gDMRs by generating a KHDC3L knockdown hESC model and compare its cellular phenotype to that of the existing NLRP7 knockdown. We will also characterize in detail how NLRP7 and KHDC3L interact, and determine if they can rescue each other's loss.
Specific Aim 2 is to identify the protein binding partners of KHDC3L and NLRP7, and how their functions, specifically DNA-binding, are affected by KHDC3L and NLRP7. This will be accomplished by testing candidate interactors, by characterizing the complexes in which these proteins carry out their role in germline DMR repogramming, and by ChIPseq of selected interactors.
Specific aim 3 is to perform high-throughput sequencing-based profiling to study how NLRP7 or KHDC3L affect the transcriptome and methylome. We anticipate that the integrated analysis of results from all these experiments will reveal how NLRP7 and KHDC3L affect gDMR reprogramming. This may result in a significant breakthrough in the understanding basic mechanisms of human genomic imprinting.
Mutations in two genes, NLRP7 and KHDC3L cause recurrent biparental hydatidiform molar pregnancies (BiHM), in which the placenta is very abnormal and the fetus does not develop. We propose to examine how loss of function of these genes causes these abnormal pregnancies through studies in embryonic stem cell culture models. We will search for the proteins with which KHDC3L and NLRP7 interact and study how they cooperate to regulate the expression of many other genes to help us understand fundamental principles of the genetic regulation of reproduction and the causes of BiHM and other similar diseases.
|Veeraragavan, Surabi; Wan, Ying-Wooi; Connolly, Daniel R et al. (2016) Loss of MeCP2 in the rat models regression, impaired sociability and transcriptional deficits of Rett syndrome. Hum Mol Genet 25:3284-3302|