Hydatidiform Mole, Recurrent, 1

A number sign (#) is used with this entry because of evidence that recurrent hydatidiform mole-1 (HYDM1) is caused by homozygous or compound heterozygous mutation in the NLRP7 gene (609661) on chromosome 19q13.

Description

A hydatidiform mole is an abnormal pregnancy characterized by hydropic placental villi, trophoblastic hyperplasia, and poor fetal development. Familial recurrent hydatidiform mole is an autosomal recessive condition in which women experience recurrent pregnancy losses, predominantly complete hydatidiform mole (CHM). However, unlike sporadic CHMs, which are androgenetic with 2 paternal chromosome complements, CHMs associated with familial recurrence are genetically biparental in origin with both a maternal and a paternal contribution to the genome. Other pregnancy losses in this condition include partial hydatidiform mole, stillbirths, ectopic pregnancies, early neonatal deaths, and miscarriages, some of which may be undiagnosed molar pregnancies. Normal pregnancies are extremely rare in families with this condition (summary by Fallahian et al., 2013).

Genetic Heterogeneity of Recurrent Hydatidiform Mole

Another form of recurrent complete hydatidiform mole (HYDM2; 614293) is caused by mutation in the KHDC3L gene (611687) on chromosome 6q13. HYDM3 (618431) is caused by mutation in the MEI1 gene (608797) on chromosome 22q13. HYDM4 (618432) is caused by mutation in the C11ORF80 gene (616109) on chromosome 11q13.

Clinical Features

In India, Ambani et al. (1980) observed gestational trophoblastic disease in multiple pregnancies of sisters in 3 unrelated kindreds. In 1 family a first cousin of 2 'affected' sisters also had a mole pregnancy and the 3 husbands of the 'affected' females had a common ancestral couple, i.e., were related as second cousins, although not related to their wives. Kajii and Ohama (1977) presented evidence for solely paternal genome in hydatidiform moles. On the basis of genetic origin, hydatidiform mole can be divided into 3 types. Approximately 25% of hydatidiform moles ascertained clinically are partial moles. They are triploid, the additional set of chromosomes generally being paternally derived. They do not seem to be associated with the development of choriocarcinoma. A second type of mole that can be classified pathologically is the complete mole. These moles are genetically diploid but are unusual in that all chromosomes are paternally derived, although the cytoplasm has been shown to be maternally derived as in normal conceptions. Complete hydatidiform moles may have 1 of 2 different origins. Most, about 90%, are homozygous, arising from duplication of a haploid sperm. More rarely, complete mole arises by dispermy, the fertilization of an anucleate egg by 2 sperm, and are therefore heterozygous. In 1 choriocarcinoma following pregnancy with hydatidiform mole, Fisher et al. (1988) demonstrated homozygosity.

Moglabey et al. (1999) found reports of 7 familial cases.

Helwani et al. (1999) provided a partial pedigree of a Lebanese family with recurrent hydatidiform moles involving 3 sibships, the offspring of consanguineous parents. They pointed out that the same family had been reported by Vejerslev et al. (1991), Sunde et al. (1993), and Seoud et al. (1995). Using microsatellite markers amplified by PCR, they performed a genetic study on 8 independent molar tissues occurring in 2 sisters. Karyotype and genotype data demonstrated a diploid and biparental constitution in 7 of the analyzed moles, suggesting a common mechanism underlying the etiology of the various molar pregnancies in this family. The data suggested that complete and partial hydatidiform moles are not always separate entities and that women with familial recurrent hydatidiform moles are homozygous for an autosomal recessive mutation. In this pedigree, not only were the parents of the molar pregnancies consanguineous, but the women were in each case the product of a consanguineous mating. One of their patients had had at least 8 molar pregnancies and several abortions but no viable children. Women with recurrent hydatidiform moles usually fail to have normal pregnancies. Helwani et al. (1999) suggested that the defective gene may be required in the fertilized/unfertilized ovum or in the maternal reproductive tract. They noted that the initial development of the mammalian zygote is under the control of maternally inherited proteins and mRNA produced and stored in the oocyte during oogenesis. Moreover, the progression of the fertilized ovum through cleavage, blastocyst formation, and implantation is dependent on the successful interaction between the preimplantation embryo and the maternal reproductive tract. Therefore, a defective maternal gene at any of these levels might deregulate the imprinting process in diploid zygotes and lead to abnormal embryonic development and to a phenotype similar to that observed in androgenetic diploid and diandric triploid conceptuses.

In a family in southern Italy, Sensi et al. (2000) confirmed that recurrent familial hydatidiform moles are diploid and biparental and arise from independent conceptions. The 2 sisters were related in each case to their husbands and all 4 were related to each other. One sister experienced 8 reproductive failures, including 6 complete moles. One pregnancy was attempted by ovum donation, but STS analysis and HLA molecular typing of the molar conceptus established that it was originated by the fertilization of a maternal ovum. This mole was persistent and treated with methotrexate. The proband's sister reported the recurrence of 3 molar pregnancies.

Slim and Mehio (2007) reviewed the history and genetics of hydatidiform mole.

Population Genetics

The incidence of hydatidiform mole varies among ethnic groups and reaches 1 in every 250 pregnancies in eastern Asia. The frequency in the US is approximately 1 in every 1,500 pregnancies (summary by Moglabey et al., 1999).

Mapping

To map the hydatidiform mole locus, Moglabey et al. (1999) performed a genomewide scan on the Lebanese family (MoLb1) reported by Helwani et al. (1999) and on a previously reported German family (MoGe2). They demonstrated that a defective maternal gene is responsible for recurrent hydatidiform moles. This gene mapped to 19q13.3-q13.4 in a 15.2-cM interval flanked by D19S924 and D19S890. They claimed that this was the first genetic mapping of a maternal locus involved in early embryogenesis in mammals.

In a family in southern Italy, Sensi et al. (2000) confirmed that recurrent familial hydatidiform moles are diploid and biparental and arise from independent conceptions. A narrowing of the gene interval on chromosome 19q13.3-q13.4 was suggested by haplotype analysis in 2 sisters.

Molecular Genetics

By fine mapping, Murdoch et al. (2006) narrowed the hydatidiform mole candidate region to a 0.65-Mb region of chromosome 19q13.4. By screening genes in this region, they identified in NLRP7 (609661) 2 different splice site mutations in 2 families (609661.0001 and 609661.0002, respectively). Screening of 2 additional families and a single family member with recurrent moles demonstrated 3 different missense mutations (609661.0003-609661.0005, respectively). NLRP7 is a member of the CATERPILLAR protein family involved in inflammation and apoptosis. Murdoch et al. (2006) pointed out that NLRP7 is the first maternal effect gene identified in humans and is also responsible for recurrent spontaneous abortions, stillbirths, and intrauterine growth retardation.

Djuric et al. (2006) analyzed molar tissues from 2 Lebanese sisters, in whom Murdoch et al. (2006) had previously identified a splice site mutation in the NLRP7 gene (609661.0001), and demonstrated normal postzygotic DNA methylation patterns at major repetitive and long interspersed nuclear elements, genes on the inactive X chromosome, 3 cancer-related genes, and CpG-rich areas surrounding the PEG3 (601483) differentially methylated region (DMR). Djuric et al. (2006) concluded that postzygotic DNA methylation and de novo methylation are normal in familial hydatidiform moles with defects in NLRP7, and that abnormal DNA methylation in these tissues is restricted to imprinted DMRs.

Deveault et al. (2009) reported 10 novel nonsynonymous variants/mutations and 1 truncation mutation (609661.0006) of the NLRP7 gene in sporadic and familial patients with hydatidiform mole. Diploid biparental, diploid androgenetic, triploid, and tetraploid conceptions were seen in patients. In vitro and in vivo early embryo cleavage abnormalities were documented in 3 patients. The authors proposed a 2-hit mechanism at the origin of androgenetic moles. This mechanism consists of variable degrees of early embryo cleavage abnormalities leading to chaotic mosaic aneuploidies, with haploid, diploid, triploid, and tetraploid blastomeres. Surviving embryonic cells that reach implantation may then be subject to the maternal immune response. Because of the patients' impaired inflammatory response, androgenetic cells, which are complete allograft, may grow and proliferate.

Wang et al. (2009) analyzed the NLRP7 gene in affected individuals from 20 families with a confirmed diagnosis of familial recurrent hydatidiform mole and identified 16 different mutations in 17 of the families (see, e.g., 609661.0003-609661.0012), including in 2 Asian sisters previously studied by Fisher et al. (2002) (609661.0009) and in 2 Italian sisters previously reported by Sensi et al. (2000) (609661.0010). Affected members from 14 of the 17 mutation-positive families were homozygous for the identified mutation, even though only 1 family reported consanguinity. Most pregnancies in the affected women were complete hydatidiform mole, although other reproductive losses were reported, including miscarriages, partial hydatidiform mole, and 1 stillbirth. None of the women had pregnancies resulting in normal live births.

Fallahian et al. (2013) performed genetic analysis of tissue from the complete hydatidiform mole pregnancies of a woman who was previously studied by Wang et al. (2009) and found to be homozygous for a 14-bp duplication in the NLRP7 gene (609661.0011). Her first and third were diploid biparental CHMs, whereas the second was a digynic triploid conceptus, with 1 paternal and 2 maternal alleles. Fallahian et al. (2013) stated that these findings were consistent with a role for NLRP7 in setting and/or maintaining the maternal imprint.

Andreasen et al. (2012) analyzed the NLRP7, NLRP2 (609364), and KHDC3L genes in 11 Danish women with hydatidiform mole, including 8 with mosaic diploid androgenetic/diploid biparental (PP/PM) moles and 3 with diploid biparental (PM) moles. Homozygosity for a splice site mutation in the NLRP7 gene (609661.0001) was identified in 1 woman with a PM mole who had a positive family history for HYDM and who had experienced 7 HYDMs. The 10 other women had no family history of HYDM and had only experienced 1 HYDM. Andreasen et al. (2012) concluded that although NLRP7 and KHDC3L mutations are associated with recurrent diploid biparental HYDMs, these genes and the NLRP2 gene are not associated with diploid HYDMs with biparental contributions to the molar genome in general.

Pathogenesis

In a series of patients with biparental complete HYDM, Fisher et al. (2002) observed dramatic underexpression of p57(KIP2) (CDKN1C; 600856) identical to the pattern seen in complete HYDM of androgenetic origin. The series included 2 sisters, both of whom had biparental complete HYDM. Genotyping of this family identified a 15-cM region of homozygosity for 19q13.3-q13.4 similar to that found in 3 other families with recurrent biparental complete HYDM. Fisher et al. (2002) concluded that biparental complete HYDM, like HYDM of androgenetic origin, may result from abnormal expression of imprinted genes (such as CDKN1C), and that a locus on 19q13.3-q13.4 may regulate expression of imprinted genes on other chromosomes.

Nguyen et al. (2014) found variable expression of CDKN1C in 35 conceptuses from 17 patients with biallelic mutations in the NLRP7 gene. Of the informative samples, 19 (59%) did not express CDKN1C and 13 (41%) displayed variable levels (20-100%) of CDKN1C. All tissue contained a diploid biparental genome. Some NLRP7 missense mutations did not completely repress CDKN1C expression, and these samples were associated with the presence of embryonic tissue of inner cell mass origin, mild trophoblastic proliferation, and low expression of CDKN1C. In contrast, truncating NLRP7 mutations were associated with lack of CDKN1C expression, absence of embryonic tissue of inner cell mass origin, and the presence of excessive trophoblastic proliferation. The findings suggested that NLRP7 regulates the imprinted expression of CDKN1C and consequently the balance between tissue differentiation and proliferation during early human development.

El-Maarri et al. (2003) reported the methylation status of 4 imprinted genes in 2 biparental complete HYDMs from 2 sisters, a 16-year-old normal offspring, and 2 sporadic biparental complete HYDMs from unrelated patients. Using 2 bisulfite-based methods, the authors demonstrated a general trend of abnormal hypomethylation at the paternally expressed genes PEG3 (601483) and SNRPN (182279), and hypermethylation at the maternally expressed genes NESP55 (see 139320) and H19 (103280), in 2 to 4 biparental complete HYDMs. Using single-nucleotide polymorphisms, the authors provided evidence that SNRPN, NESP55, and H19 were abnormally methylated on the maternal alleles in biparental complete HYDMs. They showed, in biparental complete HYDMs from the 2 sisters, that the abnormally methylated H19 allele was inherited from a maternal grandparent. These data suggested that the abnormal methylation in biparental complete HYDM may not be due to an error in erasing the parental imprinting marks, but rather in the reestablishment of the new maternal marks during oogenesis or their postzygotic maintenance. The defective 19q13.4 locus may have led to the development of variable degrees of faulty paternal marks on the maternal chromosomes.