Beckwith-Wiedemann Syndrome
A number sign (#) is used with this entry because Beckwith-Wiedemann syndrome (BWS) can be caused by mutation or deletion of imprinted genes within the chromosome 11p15.5 region. Specific genes involved include p57(KIP2) (CDKN1C; 600856), H19 (103280), and LIT1 (KCNQ1OT1; 604115). Hypermethylation and variation in the H19/IGF2-imprinting control region (ICR1; 616186) on chromosome 11p15.5, which regulates imprinted expression of H19 and IGF2 (147470), is also associated with BWS.
See also Silver-Russell syndrome (SRS; 180860), which is caused by hypomethylation defects at 11p15.
DescriptionBeckwith-Wiedemann syndrome is a pediatric overgrowth disorder involving a predisposition to tumor development. The clinical presentation is highly variable; some cases lack the hallmark features of exomphalos, macroglossia, and gigantism as originally described by Beckwith (1969) and Wiedemann (1969) (summary by Weksberg et al., 2010).
Mussa et al. (2016) provided a review of Beckwith-Wiedemann syndrome, including the wide spectrum of phenotypic manifestations, delineation of the frequencies of manifestations according to genotype, and discussion of the molecular and epigenetic defects that underlie the disorder.
Clinical FeaturesIndividuals with BWS may grow at an increased rate during the latter half of pregnancy and in the first few years of life, but adult heights are generally in the normal range. Abnormal growth may also manifest as hemihypertrophy and/or macroglossia. Hypoglycemia is reported in 30 to 50% of babies with BWS. There is an increased frequency of malformations and medical complications, including abdominal wall defects (omphalocele, umbilical hernia, and diastasis recti) and visceromegaly involving liver, spleen, pancreas, kidneys, or adrenals. Fetal adrenocortical cytomegaly is a pathognomonic finding. Renal anomalies may include primary malformations, renal medullary dysplasia, nephrocalcinosis, and nephrolithiasis. There is a predisposition to embryonal malignancies, with Wilms tumor and hepatoblastoma the most common (review by Weksberg et al., 2010).
Irving (1967, 1970) initially described the 'typical linear indentations of the lobe' that have become one of the diagnostic criteria, also well documented by Best and Hoekstra (1981). Peculiar posterior helical ear pits were first described in the BWS by Kosseff et al. (1972) and later by many others (see Best, 1991).
Two reported patients had hearing loss due to fixation of the stapes (Paulsen, 1973 and Daugbjerg and Everberg, 1984). In 3 patients, BWS and type III polycystic kidney disease occurred simultaneously (Mulvihill et al., 1989). An adult woman developed a progressive virilization due to her androgen-secreting adrenal carcinoma (Clouston et al., 1989).
A review of 31 patients with BWS and malignant tumors showed that 18 had Wilms tumor (Sotelo-Avila et al., 1980). Wiedemann (1983) reported that of 388 children, 29 developed 32 neoplasms. Of these tumors, 26 were intraabdominal, 14 being Wilms tumors and 5 adrenocortical carcinoma. Hemihypertrophy, partial or complete, was noted in 12.5% of the cases but in more than 49% of the children with neoplasms.
Wiedemann (1989) commented on overgrowth of the external genitalia in both males and females with BWS. Sippell et al. (1989) reported longitudinal data on height, bone maturation, weight, and pubertal development in 7 children with BWS. The children reached an average height of 2.5 SD above the mean at or after puberty. Growth velocity was above the ninetieth percentile until 4 to 6 years of age, and normal thereafter. Bone age was significantly advanced in all patients studied. One of the patients had latent hypothyroidism. The association of BWS and thyroid disorders may be more than coincidental (Leung, 1985 and Leung and McArthur, 1989). Emery et al. (1983) reported 2 affected sibs, one with thoracic neuroblastoma and the other who died at age 2 months of cardiomyopathy and respiratory failure.
A 'new' aspect of the natural history of BWS was reported by Chitayat et al. (1990) who observed 2 infants who were apparently normal at birth but later developed characteristics of the disorder. Both had hypoglycemia neonatally and gradually developed coarse facial changes, umbilical hernia, and macroglossia. Renal sonography done after the macroglossia developed showed large kidneys in both. The placenta was carefully examined in each case but findings described as typical of BWS were found only in one. Chitayat et al. (1990) postulated that the cellular hyperplasia and hypertrophy characteristic of BWS is caused by persistent rests of embryonal cells that secrete paracrine and/or endocrine growth factors and that the effects may not become evident until postnatal life. Neuroblastoma is another form of embryonal neoplasm that occurs in BWS (Chitayat et al., 1990). Falik-Borenstein et al. (1991) described an affected infant with congenital gastric teratoma.
In a study of 53 affected children, Carlin et al. (1990) suggested that this disorder may be milder in many cases than one would guess from published descriptions. In 11 families (21%), more than one child had BWS, including 2 sets of twins, one monozygotic and one dizygotic. Additionally, 24 families had one or sometimes both parents, and/or other relatives, affected with one or more signs of BWS. They suggested that hemihypertrophy is an underappreciated diagnostic clue for BWS in the relatives of probands. Knight et al. (1980) and Watanabe and Yamanaka (1990) described prune belly syndrome (100100) in association with BWS.
Mental retardation was documented in 6 of 39 cases observed by Martinez-y-Martinez et al. (1992), one being related to neonatal hypoglycemia.
Elliott et al. (1994) observed 76 patients with Beckwith-Wiedemann syndrome. The criteria for diagnosis were the presence of 3 major findings (macroglossia, pre- or postnatal growth greater than the 90th centile, and abdominal wall defects) or 2 major findings plus minor manifestations. In this preselected group, macroglossia was found in 97% of the patients, overgrowth in 88%, and abdominal wall defects in 80%. Hypoglycemia occurred in 48 patients and neoplasias in 3. Intellectual development was normal in all. Congenital heart defects were reported in 5 patients. Three patients had postaxial polydactyly of the foot. In one family, the mother of the index case had an ear pit and macroglossia as a child. In one family, 2 first cousins were affected. In 2 other families, 2 sibs were affected. Of 68 apparently sporadic cases, 15 had a relative with minor features of the syndrome. Elliott et al. (1994) suggested that incomplete penetrance may lead to underdiagnosis of familial cases.
Weng et al. (1995) reported the results of a follow-up study on 15 patients with WBS. They found that the pregnancies in these cases tended to have polyhydramnios with large placentas that were almost twice the normal placental weight. The large fetal size and polyhydramnios often resulted in early delivery with occasional perinatal mortality (observed in 3 cases). Excessive umbilical cord length was a manifestation of the increased placental size and was a useful sign in suspecting WBS before delivery. Abdominal wall defects and/or macroglossia helped confirm the diagnosis at birth. The newborn patients were almost 2 standard deviations above the expected mean length and weight for gestational age. The trend to increased size continued through early childhood and became less dramatic with increasing age. No cytogenetic abnormality was detected in 9 patients studied and the only tumor detected was a gastric teratoma evident in one infant at birth. Four of 15 patients had surgery for macroglossia. The findings were compared with those of Pettenati et al. (1986), who studied 22 patients.
Drut and Drut (1996) studied affected members of a family in which 4 members had WBS as a result of trisomy 11p15. Clinical examination showed nonimmune hydrops and placentomegaly in 2 sibs and multiple phenotypic abnormalities consistent with WBS in the 2 other relatives.
Moore et al. (2000) performed craniofacial anthropometric analyses on 19 patients with BWS and their relatives. The authors concluded that a unique, though variable, pattern of facial morphology can be defined in this syndrome, and that this phenotype does not diminish with age.
Everman et al. (2000) conducted a retrospective study that compared the serial alpha-fetoprotein (AFP; 104150) concentrations from 22 children with BWS with levels established for healthy children. The AFP concentration was greater with BWS and declined during the postnatal period at a significantly slower rate than what had been reported in healthy children. AFP levels obtained in the course of routine tumor screening in children with BWS should be interpreted with a normal curve established specifically for BWS rather than with previously published data for healthy infants and children.
Reddy et al. (1972) described a cardiac hamartoma in a 2-year-old child with BWS. Williams et al. (1990) found hamartoma of the urinary bladder in an infant with BWS. Jonas and Kimonis (2001) described a girl with a left chest wall hamartoma, macroglossia, nevus flammeus of the middle forehead, and a small umbilical hernia who developed left lower limb hemihypertrophy by 1 year of age and was presumed to have BWS.
Poole et al. (2012) reported 2 brothers with classic BWS. Both had very high birth weight (greater than 99th percentile), macroglossia requiring surgical correction, undescended testes, diastasis recti, and neonatal hypoglycemia. The older brother had large kidneys with unilateral cysts, a minor right-sided ear anomaly, and attention deficit-hyperactivity disorder. At age 24 years, he had a large head, prominent supraorbital ridges, a large mouth, and large hands and feet. The younger brother had mild right-sided hemihypertrophy and generalized joint hypermobility. He developed a Wilms tumor and needed special education. Puberty was delayed.
Other FeaturesGardiner et al. (2012) retrospectively identified 7 unrelated patients with BWS and posterior fossa brain abnormalities from a large cohort of 450 patients. Two cases were found to have Blake pouch cyst; 2 had Dandy-Walker variant, or hypoplasia of the inferior part of the vermis; 2 had Dandy-Walker malformation (DWM); and 1 had a complex of DWM, dysgenesis of the corpus callosum, and brainstem abnormality. One pregnancy was terminated, 2 patients died within 2 weeks of age, and a third patient died at age 2.5 years. Two living children had autism and developmental delay, respectively. Only 1 patient had normal development. Molecular studies showed that 3 cases had loss of methylation at IC2, 2 had CDKN1C mutations, and 1 had loss of methylation at IC2 and a microdeletion. In 1 case, no mutation or methylation abnormality was detected. These findings suggested that genes in imprinted domain 2 on 11p15.5 are involved in normal midline development of several organs, including the brain.
InheritanceThe mode of inheritance of BWS is complex. Possible patterns include autosomal dominant inheritance with variable expressivity, contiguous gene duplication at 11p15, and genomic imprinting resulting from a defective or absent copy of the maternally derived gene.
Wiedemann (1964) reported 3 affected sibs, and Irving (1967) observed a family with 2 affected sibs and an affected second cousin.
Autosomal dominant inheritance was suggested by Kosseff et al. (1972), Forrester (1973), and Lubinsky et al. (1974). Mausuura et al. (1975) presented a pedigree in which each of 3 sibships related as second cousins had 1 case. Kosseff et al. (1976) reviewed the pedigrees of this syndrome and invoked premutation (a special form of dominant inheritance) to explain the findings in some pedigrees.
Chemke (1976) described 8 cases in 2 sibships of an inbred kindred. Seven of those affected died in the neonatal period.
Piussan et al. (1980) described a family in which all 6 infants (including twins) appeared to have had BWS. Five who died in the neonatal period had congenital omphalocele. The surviving twin presented only minimal umbilical hernia, which reduced spontaneously. Autopsy in the last-born confirmed the diagnosis of BWS. The parents were normal and not related.
Sommer et al. (1977) reported a kindred in which 3 normal sisters gave birth to 8 infants with BWS. Autosomal dominant inheritance with the purported phenomenon of 'delayed mutation' was proposed.
Best and Hoekstra (1981) described BWS in a mother, her brother, and 2 of her children by different fathers.
Nivelon-Chevallier et al. (1983) described a family in which 4 offspring of 3 normal sisters were affected. In 2 cases, antenatal diagnosis was established by ultrasonography which showed exomphalos. In one of these cases, histologic examination of the abortus confirmed the diagnosis.
Niikawa et al. (1986) analyzed 5 unrelated kindreds with 18 affected persons and 19 families from the literature, each with more than 1 affected person. The clinical findings were highly variable, tending to become less distinctive with age. The syndrome was transmitted directly and vertically through 3 generations in 4 families and through 2 generations in 7 families. Male-to-male transmission was noted once. The sex ratio was not significantly different from 1. The segregation ratio among sibs of the proband was 0.571 +/- 0.066.
Pettenati et al. (1986) examined 22 cases of BWS clinically and cytogenetically and compared the findings with those in 226 previously reported cases. All 22 patients were chromosomally normal. The authors thought that transmission was most consistent with autosomal dominant inheritance with incomplete penetrance. There were 2 affected sisters in one family and 3 affected sisters in another; in many families, there were isolated manifestations such as macroglossia, ear crease, or omphalocele.
Olney et al. (1988) reported 3 pairs of monozygotic (MZ) twins and reviewed 3 previously reported MZ twin pairs, in which one twin showed typical BWS with minimal or no expression of the condition in the cotwin. Indeed, phenotypic concordance in MZ twin pairs had not been reported. Olney et al. (1988) concluded that the most likely mode of inheritance is autosomal dominant. Clayton-Smith et al. (1992) brought to 11 the total number of sets of monozygotic twins with BWS. Ten of these were female. Since the likelihood of this occurring due to chance alone was less than 1 in 200, the authors thought that this supported an association between MZ twinning in females and the BWS locus. Lubinsky and Hall (1991) suggested that the process of X inactivation may be responsible for imprinting at the autosomal locus for BWS as a result of a 'spillover' effect; thus would discordance for BWS in monozygotic twins be explicable, particularly in light of the fact that the so-called unaffected twin may have minor anomalies such as earlobe creases or mild macroglossia. Orstavik et al. (1995) described 13-year-old MZ female twins who were discordant for WBS. They used PCR analysis at the androgen receptor locus to determine the pattern of X inactivation. Only the paternal allele was active in all cells of the affected child. In her sister, the pattern of X inactivation was moderately skewed in the same direction, whereas in their mother, the pattern was random. Orstavik et al. (1995) suggested that nonrandom X inactivation may be related to expression of the autosomal locus for WBS.
Aleck and Hadro (1989) reported a 4-generation family in which affected children, both male and female, occurred among the offspring of each of 4 sisters, 2 of whom were monozygotic twins. Affected individuals also occurred among the children of 2 affected females. No microsigns of BWS were found in presumptive carrier women. Aleck and Hadro (1989) concluded that their findings supported the hypothesis of Lubinsky et al. (1974) that BWS is an irregular dominant resulting from the segregation of a stable premutation passed through males and females but with expression (telomutation) only in offspring of female carriers of the premutation. They suggested that an ovum-mediated sex-associated factor may be involved in the process of telomutation. One child with BWS in full-blown form had juvenile fibromatosis (220600) of the face. In addition to earlobe creases, some affected members of the family had 'dents in their ears,' punched-out lesions on the posterior aspect of the pinnae; see also Best and Hoekstra (1981). Meckel diverticulum and bicornuate uterus were also noted in one fully affected woman.
Koufos et al. (1989) discussed the peculiarities of genetic transmission of familial BWS. Whether a pleiotropic mutation at 11p15.5 or a variety of allelic mutations at 11p15.5 underlies the pathogenesis of BWS as well as of the related tumors (rhabdomyosarcoma, hepatoblastoma, adrenal tumors) or, alternatively, whether these diseases may be due to defects at closely linked but separate loci, was unclear.
Henry et al. (1991) used 11p15.5 markers to determine the parental origin of chromosome 11 in 8 sporadic cases of BWS. Probands in 3 informative families had uniparental paternal disomy for region 11p15.5. Furthermore, an overall greatly increased frequency of homozygosity for several 11p15.5 markers in 21 sporadic BWS patients suggested that isodisomy probably accounts for an even higher proportion of BWS sporadic cases. The findings were compared with the lack of paternal alleles in the Prader-Willi syndrome (176270) and the lack of maternal alleles in Angelman syndrome (105830). Paternal duplication in trisomic BWS patients, retention of paternal alleles in tumors, and higher penetrance in individuals born to female carriers in familial BWS (Lubinsky et al., 1974; Brown et al., 1990) corroborate the involvement of maternal genomic imprinting. An unbalanced dosage of maternal and paternal alleles may be the common factor in the different etiologic forms of BWS and associated tumors.
On the basis of an analysis of 19 published pedigrees suggesting autosomal dominant inheritance, Moutou et al. (1992) confirmed an excess of transmitting females and showed that there were 2 reasons for this excess: reduced fecundity in affected males compared to affected females in a ratio of 1 to 4.6; and a smaller risk of being affected, in a ratio of 1 to 3, for subjects who inherited the gene from their father. The latter finding suggested genomic imprinting. These results, together with the occurrence of uniparental disomy in some sporadic cases and the fact that all cases trisomic for the region 11p15.5 have had the duplicated region derived from the father, support the proposal that overgrowth in BWS patients and malignant proliferation in associated tumors reflect an imbalance between paternal and maternal alleles.
Viljoen and Ramesar (1992) also examined the case for imprinting in BWS. They presented a new pedigree that appeared to support paternal imprinting. An unaffected male had affected grandchildren through unaffected daughters born of different wives. Under the postulate of paternal imprinting, the phenotypically normal progenitor was a carrier for the imprinted gene and had a 50% chance of passing it to progeny, all of whom would lack stigmata of BWS. According to the hypothesis, change of the imprint status would occur in the germline of female carriers, and their children, both male and female, would have a 50% risk of manifesting the disorder, that is, 100% of those carrying the gene would be affected. The pedigree was consistent with this interpretation. An affected granddaughter of the progenitor had an affected daughter. Viljoen and Ramesar (1992) also reviewed 27 previously published pedigrees with 2 or more affected persons and in all but 4 concluded that paternal imprinting would explain the nonmendelian inheritance of BWS.
Elliott and Maher (1994) reviewed the subject of BWS. They pointed out that it is useful to consider the complications of the syndrome by the age of the subject. Multiple births are more common in BWS, with an excess of both monozygotic and dizygotic twins. Twin pairs are invariably discordant for BWS, although the second twin may occasionally show minor features. An excess of female monozygotic twin pairs (13 female, 1 male) has been observed among twin pairs with normal chromosomes.
Weksberg et al. (2002) showed that the incidence of female monozygotic twins among patients with BWS is dramatically increased over that of the general population. In skin fibroblasts from 5 monozygotic twin pairs discordant for BWS, each affected twin had an imprinting defect at the KCNQ1OT1 gene (604115) on 11p15, whereas the unaffected twin did not. Five additional monozygotic twin pairs, for whom only blood was available, also displayed an imprinting defect at KCNQ1OT1. The authors hypothesized that discordance for BWS in monozygotic twins may be due to unequal splitting of the inner cell mass during twinning, thereby causing differential maintenance of imprinting at KCNQ1OT1. Alternatively, KCNQ1OT1 may be especially vulnerable to a loss of imprinting event, caused by a lack of maintenance DNA methylation at a critical stage of preimplantation development, and that this loss of imprinting may predispose to twinning as well as to discordance for BWS. Weksberg et al. (2002) recommended continued surveillance of children born following assisted reproductive technologies that may impact the preimplantation embryo.
Superovulation (ovarian stimulation) is an assisted reproductive technology (ART) for human subfertility/infertility treatment, which has been correlated with increased frequencies of imprinting disorders such as Angelman syndrome and BWS. Market-Velker et al. (2010) examined the effects of superovulation on genomic imprinting in individual mouse blastocyst stage embryos. Superovulation perturbed genomic imprinting of both maternally and paternally expressed genes. Loss of Snrpn (182279), Peg3 (601483), and Kcnq1ot1 and gain of H19 (103280) imprinted methylation were observed. This perturbation was dose-dependent, with aberrant imprinted methylation more frequent at higher hormone dosage. Maternal as well as paternal H19 methylation was perturbed by superovulation. Market-Velker et al. (2010) postulated that superovulation may have dual effects during oogenesis, disrupting acquisition of imprints in growing oocytes, as well as maternal-effect gene products subsequently required for imprint maintenance during preimplantation development.
CytogeneticsIn 2 unrelated children with features of the Beckwith-Wiedemann syndrome, Waziri et al. (1983) found partial duplication of 11p. They reviewed 6 other reported cases of partial duplication of chromosome 11p and found description of features suggesting BWS. Their first patient had deletion of 11q23.33-qter and duplication of 11p13-p15. In the second case, duplication of 11p15 was suspected. Since the duplicated region presumably contains the insulin locus (INS; 176730) and perhaps also the locus for insulin-like growth factor-2 (IGF2; 147470), the finding suggested that the EMG syndrome may be 'caused' by excess of one or both of these.
Pueschel and Padre-Mendoza (1984) described a child with this syndrome and a balanced 11/22 translocation: 46,XX,t(11p;22q). The phenotypically normal mother had the same balanced translocation. Turleau et al. (1984) found trisomy 11p15 in 2 cases of Beckwith-Wiedemann syndrome. One was an instance of de novo duplication of 11p15; the other was the result of t(4;11)(q33;p14)pat.
Turleau and de Grouchy (1985) found no clear evidence of phenotypic differences between patients with and those without chromosome abnormality. Okano et al. (1986) described an infant with partial trisomy of the terminal part of 11p. The father had a balanced translocation between chromosomes 4 and 11 with a breakpoint at 11p13. The proband had duplication of the 11pter-p13 segment. Okano et al. (1986) analyzed the clinical features of this and 14 other reported cases; findings in 13 patients with duplication of 11p15 resembled those of BWS.
Henry et al. (1989) reported the first 2 cases of BWS with dup11p15 and adrenocortical carcinoma (ADCC). Together with evidence for somatic chromosomal events leading to loss of 11p15.5 alleles in familial ADCC cases, they hypothesized that a gene involved in predisposition to ADCC maps to region 11p15.5.
Norman et al. (1992) observed a phenotypically normal mother and 2 offspring with BWS, all 3 of whom carried the same paracentric inversion, inv(11)(p11.2;p15.5). The phenotypically normal mother was a carrier of the translocation; both of her parents had normal chromosomes. The breakpoint on 11p was mapped by in situ hybridization to a site proximal to the insulin and insulin-growth factor 2 loci, and distal to D11S12. That the translocation t(9;11) had originated in the grandfather was proven because the derivative translocation chromosome 9 had a large block of methyl green/DAPI-positive heterochromatin on the long arm which could be traced back to him. Testing for loci on chromosome 11 likewise indicated that the derivative chromosome 11 originated from the grandfather. Thus, BWS associated with balanced chromosome translocations is transmitted in the same sex-dependent pattern as are noncytogenetic forms of familial BWS. Tommerup et al. (1993) reported the case of a patient with BWS and a reciprocal translocation t(9;11)(p11;p15.5).
Fryns et al. (1993) observed a patient with BWS and duplication 4q/deficiency 18p as the result of an unbalanced paternal translocation. The authors suggest that other contiguous gene duplication/deletions exist in BWS.
Drut and Drut (1996) used fluorescence in situ hybridization for interphase analysis to demonstrate trisomy 11p15 in interphase nuclei of formalin-fixed paraffin-embedded placenta from a stillborn fetus and in peripheral blood lymphocytes from 2 liveborn female relatives with WBS. Karyotype of the father of 2 stillborn sibs demonstrated a balanced reciprocal translocation: 46,XY,t(10;11)(q26;p15).
Slavotinek et al. (1997) described a 3-generation family in which a father and son had a balanced reciprocal translocation t(5;11)(p15.3;p15.3). Two family members with unbalanced products of this translocation resulting in trisomy for chromosome 11pter-p15.3 had features of BWS, with prenatal overgrowth, macroglossia, coarse facial features, and broad hands. Slavotinek et al. (1997) reviewed the clinical features of BWS patients with paternally derived duplications of 11p15.5 compared to those with normal chromosomes and found that patients with duplications had a decreased incidence of hypoglycemia, facial nevus flammeus, and hemihypertrophy, but an increased incidence of moderate to severe learning difficulties. They also noted persistent physical differences, including prominent occiput, prominent forehead, round face with full cheeks, deep-set eyes with epicanthic folds, hypertelorism, broad, flat nasal bridge, and micrognathia.
Chiesa et al. (2012) described 2 maternal 11p15.5 microduplications with contrasting phenotypes. In the first case, a 1.2-Mb inverted duplication of chromosome 11p15 derived from the maternal allele resulted in Silver-Russell syndrome (SRS; 180860). The duplication encompassed the entire 11p15.5 imprinted gene cluster, and hypermethylation of CpGs throughout the ICR2 region was observed. These findings were consistent with the maintenance of genomic imprinting, with a double dosage of maternal imprinting and resulting in a lack of KCNQ1OT1 transcription. In the second case, a maternally inherited 160-kb inverted duplication that included only ICR2 and the most 5-prime 20 kb of KCNQ1OT1 resulted in a BWS phenotype in 5 individuals in 2 generations. This duplication was associated with hypomethylation of ICR2 resulting from partial loss of the imprinted methylation of the maternal allele, expression of a truncated KCNQ1OT1 transcript, and silencing of CDKN1C. Chromatin RNA immunopurification studies suggested that the KCNQ1OT1 RNA interacts with chromatin through its most 5-prime 20-kb sequence, providing a mechanism for the silencing activity of this noncoding RNA. The finding of similar duplications of ICR2 resulting in different methylation imprints suggested that the ICR2 sequence is not sufficient for establishing DNA methylation on the maternal chromosome, and that some other property, possibly orientation-dependent, is needed.
MappingIn a family with multiple cases of BWS in multiple sibships and at least 2 generations (Aleck and Hadro, 1989), Koufos et al. (1989) demonstrated tight linkage of the mutation causing the syndrome to markers located at 11p15.5. Stimulated by this observation, Koufos et al. (1989) studied Wilms tumors of sporadic nature, not associated with BWS, and identified a subset of these tumors which had attained somatic homozygosity through mitotic recombination, with the smallest shared region of overlap being distal to the beta-globin complex at 11p15.5. The data suggested that there is a second locus, distinct from that involved in the WAGR syndrome, that plays a role in the association of Wilms tumor with BWS. See also Henry et al. (1989).
In 2 families, Ping et al. (1989) found linkage of BWS to the insulin gene, which is located at 11p15.5 (maximum lod = 3.6; theta = 0.00). Weksberg et al. (1990) found molecular evidence consistent with a duplication of a portion of 11p in 2 patients. In one, the duplicated segment was inherited; in the second, it had originated de novo and contained a portion of 11p15. Weksberg et al. (1990) showed that the duplicated segment included the genes for insulin-like growth factor-2, beta-globin (141900), calcitonin, and parathyroid hormone (168450), but did not include the gene for myogenic differentiation factor (MYOD1; 159970). These data place the BWS gene and the other genes mentioned distal to MYOD1 on 11p15.
Molecular GeneticsJeanpierre et al. (1985) found no obligate or consistent duplication of any 11p markers in BWS and concluded that duplication of INS (176730), HRAS1 (190020), and IGF2 (147470) cannot be directly responsible for the hyperinsulinism, predisposition to neoplasm, or gigantism in this disorder. Spritz et al. (1986) studied 7 patients and found no evidence of extra dosage of the insulin or IGF2 genes. Mannens et al. (1987) found that a tumor from a BWS patient had loss of heterozygosity for all 11p markers tested and for one 11q marker, PGA (169700); APOA1 (107680) remained heterozygous.
Hayward et al. (1988) found loss of heterozygosity at the HRAS1 locus (190020) on 11p in an adrenal adenoma from a 45-year-old woman with BWS. Little et al. (1988) found homozygosity for a region of 11p defined by the calcitonin (114130) and insulin (176730) genes in a hepatoblastoma from a child with BWS.
Litz et al. (1988) described monozygotic twins of whom only one had BWS. No cytogenetic or molecular abnormality of 11p could be detected in either normal or affected tissues obtained from the BWS twin. Using cloned DNA fragments homologous to 4 genes located on 11p, namely, catalase, parathyroid hormone, insulin-like growth factor II, and HRAS, Schofield et al. (1989) could find no evidence of large-scale deletions or amplifications in this chromosome region in 4 patients with BWS.
Since the IGF2 gene is parentally imprinted in the mouse, it has been suggested that, in the human, duplication of the nonimprinted locus might lead to diploid expression of the gene and consequent general hyperplasia. However, using RFLPs for 4 linked markers on 11p and genomic clones internal to the IGF2 locus, Nystrom et al. (1992) found no evidence of alteration or amplification in 11 patients. In one patient who developed Wilms tumor, they found no evidence for loss of material on 11p. The possibility of mutation in unknown transacting factors affecting the expression of IGF2 remained. In a related study, Nystrom et al. (1992) found that in 1 of 14 sporadic cases of BWS, both copies of chromosome 11 were derived from the father, indicating paternal isodisomy. Schinzel (1993) commented that BWS was, to that date, the only example of partial parental disomy, which he referred to as 'mosaic-segmental uniparental disomy.' The paternal isodisomy suggests that the BWS gene is maternally imprinted. The IGF2 gene is also maternally imprinted, functions as a fetal growth factor, and is overexpressed in Wilms tumor.
Weksberg et al. (1993) studied allele-specific expression of the IGF2 gene, using an ApaI polymorphism in the 3-prime untranslated region of IGF2. Control skin fibroblasts were shown to maintain monoallelic expression of the paternal IGF2 allele, whereas skin fibroblasts from 3 out of 5 patients with BWS demonstrated biallelic IGF2 expression. In a sixth BWS patient, fresh tongue tissue as well as the fibroblasts derived from this tissue demonstrated biallelic expression, whereas the tongue tissue obtained from a control showed monoallelic expression. Weksberg et al. (1993) excluded paternal heterodisomy, using RFLPs in the IGF2, insulin, and tyrosine hydroxylase (191290) genes. They concluded that biallelic expression reflects disruption of IGF2 imprinting and that the BWS phenotype can result from the loss of normal suppression of the maternally inherited IGF2 gene. In contrast to previous reports in which imprinting of IGF2 has been invoked as the mechanism to explain sporadic cases of BWS (especially in situations where uniparental disomy and trisomy of the 11p15.5 region has occurred), Ramesar et al. (1993) suggested that paternal imprinting of a growth suppressor gene, e.g., H19 (103280), may be one of the causes of familial BWS.
Brown et al. (1996) described evidence of alteration of imprinting in a BWS family with an inversion, inv(11)(p11.2;p15.5). The carrier mother had no BWS manifestations, and had apparently inherited the inversion from her father. The 2 children who inherited the inversion from the mother were affected. The maternally inherited inversion was located approximately 300 kb centromeric to the IGF2 gene. Allele-specific expression analysis revealed that in the affected children the IGF2 was expressed from both parental alleles. Brown et al. (1996) demonstrated that the inversion led to biallelic expression of IGF2 and altered DNA replication patterns in the IGF2 region. The H19 imprinting in affected individuals was normal, suggesting an H19-independent pathway to biallelic IGF2 transcription. DNA methylation in IGF2 remained monoallelic, suggesting to the investigators that the mutation caused by the translocation had uncoupled allele-specific methylation from gene expression.
Working from the proposal that the paternally derived gene(s) at 11p15.5 are selectively expressed in BWS while the maternally transmitted gene(s) are inactive, Kubota et al. (1994) examined 18 patients for the parental origin of their 11p15 regions. Two patients had duplications of 11p15 from their respective fathers and 1 from the mother, indicating the transmission of an excessive dosage of the paternal gene in each. In the series of 12 sporadic cases, uniparental paternal disomy, either total constitutional or segmental, was not observed.
Algar et al. (1999) reported 2 patients with mosaic paternal isodisomy of the 11p15 region. These patients had reduced levels of CDKN1C expression in the liver and kidney, respectively. Some expression from the paternally derived CDKN1C allele was evident, consistent with incomplete paternal imprinting. One patient showed maternal allele silencing, in addition to allele imbalance. Algar et al. (1999) concluded that CDKN1C expression is reduced in patients with BWS with allele imbalance, and suggested that CDKN1C haploinsufficiency contributes to the BWS phenotype in patients with mosaic paternal isodisomy of chromosome 11.
Three regions on 11p15 (BWSCR1, BWSCR2, and BWSCR3) may play a role in the development of BWS. BWSCR2 and BWSCR3 map, respectively, 5 Mb and 7 Mb proximal to BWSCR1, which is located 200 to 300 kb proximal to the IGF2 gene on 11p15.5 (Redeker et al., 1994). BWSCR2 is defined by 2 BWS breakpoints. By sequence analysis of 73 kb containing BWSCR2, followed by screening a cDNA library, Alders et al. (2000) isolated cDNAs encoding 2 zinc finger genes, ZNF214 (605015) and ZNF215 (605016). Alders et al. (2000) demonstrated that 2 of the 5 alternatively spliced ZNF215 transcripts are disrupted by both BWSCR2 breakpoints. Parts of the 3-prime end of these splice forms are transcribed from the antisense strands of ZNF214. Alders et al. (2000) showed that the ZNF215 gene is imprinted in a tissue-specific manner, whereas ZNF214 is not imprinted. These data supported a role for ZNF215, and possibly for ZNF214, in the etiology of BWS.
Catchpoole et al. (2000) performed sequence analysis of DNA from 15 individuals with BWS and found no pathogenic mutations in the H19 gene. A total of 21 BWS patients were also analyzed for mutations in the NAP1L4 gene (601651); again, no mutations were found. Finally, Catchpoole et al. (2000) found no mutations in the conserved differentially methylated region (DMR) of IGF2 in 13 patients with BWS. They concluded that IGF2 loss of imprinting seen in BWS patients was not due to mutations in any of these sequences.
Bliek et al. (2001) studied the methylation status of H19 and KCNQ1-overlapping transcript 1 (KCNQ1OT1) in a large series of Beckwith-Wiedemann syndrome patients. Different patient groups were identified: group I patients (20%) with uniparental disomy and aberrant methylation of H19 and KCNQ1OT1; group II patients (7%) with a BWS imprinting center 1 (BWSIC1) defect causing aberrant methylation of H19 only; group III patients (55%) with a BWS imprinting center 2 (BWSIC2) defect causing aberrant methylation of KCNQ1OT1 only; and group IV patients (18%) with normal methylation patterns for both H19 and KCNQ1OT1. Of 31 patients with KCNQ1OT1 demethylation only (group III), none developed a tumor. However, tumors were found in 33% of patients with H19 hypermethylation (group I and II) and in 20% of patients with no detectable genetic defect (group IV). All 4 familial cases of BWS showed reduced methylation of KCNQ1OT1, suggesting to the authors that in these cases the imprinting switch mechanism may be disturbed.
Murrell et al. (2004) screened conserved sequences between human and mouse DMRs of the IGF2 gene for variants in order to find other genetic predispositions to BWS. Four SNPs were found in DMR0 (T123C, G358A, T382G, and A402G), which occurred in 3 out of 16 possible haplotypes: TGTA, CATG, and CAGA. DNA samples from a cohort of sporadic BWS patients and healthy controls were genotyped for the DMR0 SNPs. There was a significant increase in the frequency of the CAGA haplotype and a significant decrease in the frequency of the CATG haplotype in the patient cohort compared to controls. These associations were still significant in a BWS subgroup with KvDMR1 loss of methylation, suggesting that the G allele at T382G SNP (CAGA haplotype) may be associated with loss of methylation at KvDMR1. Murrell et al. (2004) proposed either a genetic predisposition to loss of methylation or interactions between genotype and epigenotype that impinge on the disease phenotype.
BWS, like Sotos syndrome (117550), is an overgrowth syndrome. Deregulation of imprinted growth regulatory genes within the 11p15 region is the major cause of BWS. Similarly, defects of the NSD1 gene (606681) account for more than 60% of cases of Sotos syndrome. In 2 patients with a clinical diagnosis of BWS, Baujat et al. (2004) identified mutations in the NSD1 gene (see 606681.0011-606681.0012).
Sparago et al. (2004) found a 1.8-kb deletion in the H19 differentially methylated region (DMR) in 2 individuals with BWS (103280.0001). These deletions abolished 2 CTCF target sites, and maternal transmission resulted in hypermethylation of the H19 DMR, biallelic IGF2 expression, H19 silencing, and BWS, indicative of loss of function of the IGF2-H19 imprinting control element. The 1.8-kb deletion was not detected in any of 14 individuals with BWS with defects other than H19 methylation or in any of 50 healthy individuals.
In 3 sibs with BWS and Wilms tumor and 2 unaffected sibs from a 3-generation family, Prawitt et al. (2005) identified a 2.2-kbp microdeletion in the H19/IGF2 imprinting center-1 which abolished 3 CTCF target sites. Maternal inheritance of the deletion was associated with IGF2 (147470) loss of imprinting and upregulation of IGF2 mRNA. However, in at least 1 affected family member a second lesion was identified, duplication of maternal 11p15, which was accompanied by a further increase in IGF2 mRNA levels (35-fold higher than control values). Prawitt et al. (2005) suggested that the combined effects of the BWSIC1 microdeletion and 11p15 duplication were necessary for the manifestation of BWS in this family.
Cerrato et al. (2005) emphasized that hypermethylation and silencing of H19, as shown by Sparago et al. (2004), likely also contributed to the BWS phenotype, and noted that the 2.2-kb deletion reported by Prawitt et al. (2005) did not affect DNA methylation of H19. Cerrato et al. (2005) identified a 1.4-kb deletion in a patient with BWS that eliminated only a subfragment of the interval missing in the 2.2-kb deletion reported by Prawitt et al. (2005), but was still associated with hypermethylation of the H19 promoter. Cerrato et al. (2005) concluded that BWS can result from maternally inherited deletions causing loss of imprinting of IGF2 only if associated with either 11p15 duplication or with hypermethylation and silencing of H19.
Niemitz et al. (2004) reported a microdeletion involving the entire LIT1 gene (604115.0001), thus providing genetic confirmation of the importance of this gene region in BWS. When inherited maternally, the deletion caused BWS with silencing of p57(KIP2) (600856), indicating deletion of an element important for the regulation of p57(KIP2). When inherited paternally, there was no phenotype, suggesting that LIT1 RNA itself is not necessary for normal development in humans.
Cerrato et al. (2005) showed that maternal germline methylation at IC2 and imprinted expression of 5 genes of the IC2 domain were correctly reproduced on an 800-kb YAC transgene when transferred outside of their normal chromosomal context. The authors determined that key imprinting control elements were located within a 400-kb region centromeric of IC2 and that each of the 2 domains of the cluster contained the cis-acting elements required for the imprinting control of its own genes. Maternal, but not paternal, transmission of the transgene resulted in fetal growth restriction, suggesting that during evolution the acquisition of imprinting may have been facilitated by the opposite effects of the 2 domains on embryo growth.
Azzi et al. (2009) studied the methylation status of 5 maternally and 2 paternally methylated loci in a series of 167 patients with 11p15-related fetal growth disorders. Seven of 74 (9.5%) Russell-Silver (RSS; 180860) patients and 16 of 68 (24%) Beckwith-Wiedemann (BWS) patients showed multilocus loss of methylation (LOM) at regions other than ICR1 and ICR2 11p15, respectively. Moreover, over two-thirds of multilocus LOM RSS patients also had LOM at a second paternally methylated locus, DLK1/GTL2 IG-DMR. No additional clinical features due to LOM of other loci were found, suggesting an (epi)dominant effect of the 11p15 LOM on the clinical phenotype for this series of patients. Surprisingly, 4 patients displayed LOM at both ICR1 and ICR2 11p15; 3 of them had a RSS and 1 patient had a BWS phenotype. The authors concluded that multilocus LOM can also concern RSS patients, and that LOM can involve both paternally and maternally methylated loci in the same patient.
Cerrato et al. (2008) reported 12 BWS cases with BWSIC1 hypermethylation in which there was no deletion or other nearby mutation; similarly, no BWSIC1 mutation was detected in 40 sporadic nonsyndromic Wilms tumors. Detailed methylation analysis of the BWS patients showed that the hypermethylation extended over the entire or only the 3-prime half of the IC1 region, did not affect other imprinted loci, generally occurred in the mosaic form, and was never present in the unaffected relatives. All of the BWS cases were sporadic, and in at least 2 families, affected and unaffected individuals shared the same maternal BWSIC1 allele but not the abnormal maternal chromosome epigenotype. In addition, the chromosome with the imprinting defect derived from either the maternal grandfather or maternal grandmother. Cerrato et al. (2008) concluded that, in the absence of deletions, BWSIC1 hypermethylation generally occurs as sporadic epimutation and is associated with low recurrence risk.
Demars et al. (2010) investigated the CTCF (604167) gene and the ICR1 domain in 21 BWS patients with ICR1 gain of methylation and 16 SRS patients with ICR1 loss of methylation. There were 4 constitutional ICR1 genetic defects in BWS patients, including a familial case. Three of those defects were imprinting defects consisting of small deletions and a single mutation, which did not involve one of the CTCF binding sites. Moreover, 2 of those defects affected OCT (PLXNA2; 601054)-binding sequences, which may normally maintain the unmethylated state of the maternal allele. A single-nucleotide variation was identified in a SRS patient.
In a 15-year-old girl with BWS, Zollino et al. (2010) identified a 900-kb de novo deletion at chromosome 11p15.5 on the maternal allele, spanning A_14_P130713 to A_14_P123179 and encompassing ICR2 and 16 genes, including CDKN1C (600856). DNA methylation analysis showed complete absence of methylation at ICR2 in the patient, with normal methylation at ICR1. The patient had mild psychomotor delay and a peculiar facial appearance, with horizontal eyebrows with synophrys, downslanting palpebral fissures with epicanthic folds, narrow nasal bridge, hypoplastic philtrum and prominent jaw, low posterior hairline, and hypertrichosis. Her tongue was slightly asymmetric, with one half larger than the other. Zollino et al. (2010) stated that only 1 other BWS patient had been reported with deletion of ICR2 (Niemitz et al., 2004).
In 2 brothers with BWS, Poole et al. (2012) identified a heterozygous A-to-C transversion in the A2 repeat of ICR1 that was demonstrated to alter the binding of nuclear factors, most likely OCT4 (POU5F1; 164177). The mutation was inherited from the unaffected mother, who carried it on the paternal allele. The patients had hypermethylation of the ICR1 region. DNA sequencing of 9 additional patients with BWS and H19 hypermethylation did not identify mutations in the H19 ICR or promoter region.
Mutations in the CDKN1C Gene
Studying DNA samples from 9 unrelated Japanese patients with BWS, Hatada et al. (1996) analyzed the entire coding region of p57(KIP2) (CDKN1C; 600856), including intron/exon boundaries, by direct PCR using 5 PCR primer sets. They detected mutations in 2 patients (e.g., 600856.0001). In one other patient, Hatada et al. (1996) demonstrated reduced expression of the p57(KIP2) gene in adrenal gland. They concluded that the studies provided evidence for a new mechanism for producing a phenotype with dominant transmission and little or no gene product: one allele with an inactive product is expressed and the other allele is repressed by genomic imprinting. Hatada et al. (1996) commented that other loci may possibly be involved in BWS, since there are 3 other known balanced translocations leading to BWS which map several megabases from the p57(KIP2) region.
Lam et al. (1999) sequenced the CDKN1C gene in 70 patients with BWS. Fifty-four were sporadic with no evidence of uniparental disomy and 16 were familial from 7 kindreds. Novel germline CDKN1C mutations were identified in 5 probands, 3 of 7 familial cases and 2 of 54 sporadic cases. There was no association between germline CDKN1C mutations and IGF2 or H19 abnormalities. There was a significantly higher frequency of exomphalos in the CDKN1C mutation cases as compared to cases with other types of molecular pathology. There was no association between germline CDKN1C mutations and risk of embryonal tumors. No CDKN1C mutations were identified in 6 non-BWS patients with overgrowth and Wilms tumor.
Romanelli et al. (2010) identified 7 novel mutations in the CDKN1C gene in 8 of 50 patients with BWS who did not have epigenetic alterations at chromosome 11q15. Six patients inherited the mutation from apparently asymptomatic mothers, 1 was de novo, and 1 could not be determined. Three of the mutations involved nucleotide 845 (see, e.g., 600856.0004 and 600856.0005), suggesting a possible mutation hotspot. In additional to classic features of the disorder, 2 patients had polydactyly, 2 had an extra nipple, and 3 had cleft palate. No mutations were found in 22 patients with isolated hemihypertrophy, omphalocele, or macroglossia.
PathogenesisGardner (1973) pointed out some similarities between the Beckwith-Wiedemann syndrome and the disorder in the offspring of diabetic mothers. The similarities suggest that fetal hyperinsulinism may be involved in the latter condition, inasmuch as the insulin gene (INS; 176730) is located in the 11p15 region. However, see the work of Jeanpierre et al. (1985) and Spritz et al. (1986) outlined earlier.
The finding of loss of IGF2 imprinting in sporadic Wilms tumor (Rainier et al., 1993) further strengthens the view that IGF2 overexpression plays an important role in somatic overgrowth and the development of embryonal tumors.
In connection with studies of the methylation of the H19 (103280) and IGF2 genes in BWS, Reik et al. (1994) outlined the following distinctive features of the disorder. In approximately 15% of patients, BWS is inherited as an autosomal dominant with variable expressivity and incomplete penetrance. The penetrance is affected by parental transmission with an increase on maternal and a decrease on paternal transmission. In familial cases, the BWS gene maps to 11p15. In some sporadic cases, there are 11p15 chromosomal abnormalities, including partial trisomy, where the duplicated segment is always of paternal origin, balanced translocations and inversions that are maternally inherited, and paternal isodisomy for 11p, which occurs in approximately 20% of sporadic BWS patients. These observations suggest that genes for BWS are located on 11p15 and that they are either maternally imprinted genes with growth enhancing activity or paternally imprinted genes with growth suppressing activity. Reik et al. (1994) analyzed the allele-specific methylation patterns in the IGF2 gene and in the neighboring and reciprocally imprinted H19 gene in 42 BWS patients, 10 of whom represented mosaic uniparental disomy (UPD) cases. They found that allelic methylation of both genes was normal in all non-UPD cases, with the paternal allele being methylated, and was increased in UPD cases in proportion with the disomic lineage. These findings suggested to Reik et al. (1994) that sporadic BWS is not associated with a general alteration of methylation imprinting of the IGF2 and H19 genes. The methylation assay used in this study thus offers a simple and reliable diagnostic test of UPD for 11p15.5.
Reik et al. (1995) identified BWS patients who had inherited a normal biparental chromosome complement of the 11p15.5 region, where IGF2 and H19 reside, but had an altered pattern of allelic methylation of both genes, with the maternal chromosome carrying a paternal imprinting pattern. In fibroblasts, IGF2 was expressed from both parental alleles and H19 was not expressed, precisely as predicted from the altered pattern of allelic methylation. DNA replication patterns in the 11p15.5 region remained asynchronous as in controls, however. Reik et al. (1995) stated that the results provided the first example of a dissociation of regional control of DNA replication from regional control of allelic methylation and expression in imprinting. The authors suggested that the altered pattern of allelic methylation and expression arises in the germline or in the early embryo from defects in resetting or setting of imprinting in the maternal germline. Potential candidate regions for mutations include the previously identified translocation breakpoint clusters and the H19 gene itself. The finding of possible 'imprinting mutations' in BWS raises the prospect of identifying genetic factors that control imprinting in this region.
Hoovers et al. (1995) isolated YACs, and from the YACs cosmid libraries, representing the region of loss of heterozygosity in embryonal tumors associated with BWS. They isolated 5 germline balanced chromosomal rearrangement breakpoint sites from BWS patients, as well as a balanced chromosomal translocation breakpoint from a rhabdoid tumor, within a 295- to 320-kb cluster defined by a complete cosmid contig crossing these breakpoints. This breakpoint cluster terminated approximately 100 kb centromeric to the imprinted gene IGF2 and 100 kb telomeric to p57(KIP2) (600856), an inhibitor of cyclin-dependent kinases, and was located within subchromosomal transferable fragments that suppressed the growth of embryonal tumor cells in genetic complementation experiments. They identified 11 transcribed sequences in this BWS/tumor suppressor coincident region of 11p15, one of which corresponded to p57(KIP2). However, 3 additional BWS breakpoints were more than 4 megabases centromeric to the other 5 breakpoints and were excluded from the tumor suppressor region defined by subchromosomal transferable fragments. Thus, multiple genetic loci defined BWS and tumor suppression on 11p15. The authors speculated that, most likely, a group of cancer-related genes falls within a several megabase region, similar to 1p, 3p, and 9p.
Several lines of evidence suggest that BWS may be caused by relative overexpression of the maternally imprinted IGF2 gene. Although cytogenetic and molecular findings are normal in most cases, a few show paternally inherited duplication of 11p15, segmental or mosaic paternal uniparental disomy for 11p15, or a maternal translocation with one breakpoint at 11p15. In familial cases, furthermore, transmission is always through the mother. These observations suggest that 2 paternal copies of IGF2 result from either paternal duplication or paternal UPD, that deletion or disruption associated with maternal translocation results in activation of the maternal gene, and that the mutations which are maternally transmitted activate the maternal IGF2 gene. Another overgrowth syndrome, Simpson-Golabi-Behmel syndrome (SGBS; 312870), has been shown to be due to mutations in an extracellular proteoglycan, glypican-3 (300037), that is inferred to play an important role in growth control in embryonic mesodermal tissues in which it is selectively expressed. It forms a complex with IGF2 and probably modulates IGF2 action. Thus there may be a commonality in the pathogenesis of BWS and SGBS.
Further evidence of the role of IGF2 in BWS was presented by Sun et al. (1997). They introduced Igf2 transgenes into the mouse genome by using embryonic stem (ES) cells and thereby caused transactivation of the endogenous Igf2 gene. The consequent overexpression of Igf2 resulted in most of the symptoms of Beckwith-Wiedemann syndrome, including prenatal overgrowth, polyhydramnios, fetal and neonatal lethality, disproportionate organ overgrowth including tongue enlargement, and skeletal abnormalities. This was presented as evidence that IGF2 overexpression is a key determinant of BWS. The Igf2 dosage-dependent phenotypes seen in the transgenic mouse models BWS closely. These phenotypes overlap with other fetal overgrowth syndromes attributed to increased IGF2 levels, such as the IGF2 overgrowth syndrome (Morison et al., 1996), in which symptoms are less severe, and the Simpson-Golabi-Behmel syndrome, in which symptoms are more severe, than in BWS. Sun et al. (1997) commented that the transgenic Igf2 mouse is a better model than the H19 mouse, which shows overgrowth only, or the Igf2r mouse (147280), which shows disproportionate overgrowth of the heart but not other organs, and an overall increase in body weight to 130%. Exomphalos or omphalocele, enlargement of the adrenal cortex, and renal medullary dysplasia were observed in the p57(KIP2) knockout mouse together with some other phenotypes that might constitute minor features of BWS. Overexpression of IGF2 and underexpression of p57(KIP2) account for almost all features of BWS, with the possible exception of tumors and hypoglycemia.
Catchpoole et al. (1997) studied the molecular pathology of 106 sporadic BWS cases. Fourteen of 83 informative cases had segmental paternal isodisomy for a region of 11p15.5 bounded by D11S861 and D11S2071. Isodisomy for 11q was detected in 3 cases, but only as far telomerically as 11q13-11q21. The allele-specific methylation status of the H19 gene was determined in 80 sporadic BWS cases. Thirteen had UPD and showed H19 hypermethylation. Of 63 cases with biparental inheritance, 5 showed H19 hypermethylation consistent with an imprinting center mutation or imprinting error lesion. The phenotype of these 5 patients overlapped with that of sporadic BWS cases without UPD and with normal H19 methylation; however, exomphalos was more common (p less than 0.05) in the latter group. Catchpoole et al. (1997) concluded that the expression of imprinted genes in this region may be dependent on the precise molecular pathology and that H19 methylation is useful in diagnosis of UPD or altered imprinting in BWS.
Maher and Reik (2000) reviewed 'imprinting in clusters' in relation to BWS. They pointed to the fact that biallelic expression of IGF had been reported in this syndrome. The IGF2 (147470) and CDKN1C (600856) genes lie in the same imprinted domain. They speculated that the link between the functions of the IGF2 and CDKN1C gene products that are revealed by their involvement in BWS may turn out to be the first of many examples of functional interactions between genes with an imprinted gene cluster. Thus, clustering of imprinted genes enables coordinate regulation of imprinting across large domains and more local mechanisms within specific regions of the domain. Evidence reviewed by Maher and Reik (2000) suggested that many cases of BWS are entirely epigenetic in origin; reports of monozygotic twins who are discordant for this syndrome are consistent with this view.
The speculative model for imprinting of genes in the BWS cluster on chromosome 11p15.5 presented in Figure 3 of the review by Maher and Reik (2000) was originated by Lee et al. (1999) in a paper in which they described a novel antisense gene within CDKN1C, which they termed LIT1 (604115), and associated CpG island, that undergoes loss of imprinting in most patients with BWS.
Itoh et al. (2000) performed a study of multiple organs and tissues in a case of BWS with a high degree of mosaic paternal 11p15 UPD. The proportion of cells with paternal 11p15 UPD correlated with the degree of organ enlargement. The authors concluded that mosaicism may explain the variability of phenotypes including hemihyperplasia and predisposition to childhood cancers in BWS.
Horike et al. (2000) generated modified human chromosomes carrying a targeted deletion of the LIT1 CpG island using recombination-proficient chicken DT40 cells. The mutation abolished LIT1 expression on the paternal chromosome, accompanied by activation of the normally silent paternal alleles of multiple imprinted loci at the centromeric domain including KVLQT1 (KCNQ1; 607542) and p57(KIP2). The deletion had no effect on imprinting of H19 located at the telomeric end of the cluster. The authors hypothesized that the LIT1 CpG island can act as a negative regulator in cis for coordinate imprinting at the centromeric domain, thereby suggesting a role for the LIT1 locus in a BWS pathway leading to functional inactivation of p57(KIP2).
Engel et al. (2000) identified loss of methylation at the KvDMR1 region of the KVLQT1 gene (termed 'BWSIC2 defects' by the authors) in 35 of 69 sporadic cases of BWS without UPD. This was associated often with loss of imprinting of IGF2 and always with a normal H19 methylation pattern. The incidence of exomphalos in those with putative BWSIC2 defects was not significantly different from that in patients with CDKN1C mutations, but was significantly greater than in those with BWSIC1 defects. Engel et al. (2000) concluded that BWSIC2 defects result in epigenetic silencing of CDKN1C and variable loss of imprinting of IGF2. They also found that no BWS patients with embryonal tumors had BWSIC2 defects.
The most common constitutional abnormalities in BWS are epigenetic, involving abnormal methylation of either H19 (103280) or LIT1 (604115), both of which encode untranslated RNAs on 11p15. DeBaun et al. (2002) hypothesized that different epigenetic alterations would be associated with specific phenotypes in BWS. To test this hypothesis, they performed a case-cohort study, using the BWS Registry. The cohort consisted of 92 patients with BWS who had had molecular analysis of both H19 and LIT1; these patients showed the same frequency of clinical phenotypes as those patients in the Registry from whom biologic samples were not available. The frequency of altered DNA methylation in H19 in patients with cancer was significantly higher than the frequency in patients without cancer and cancer was not associated with LIT1 alterations. The frequency of altered DNA methylation of LIT1 in patients with midline abdominal wall defects and macrosomia was significantly higher than in patients without such defects. DeBaun et al. (2002) also found that paternal uniparental disomy of 11p15 was associated with hemihypertrophy, cancer, and hypoglycemia. These results defined an epigenotype-phenotype relationship in BWS related to cancer risk and specific birth defects.
Cox et al. (2002) noted that assisted reproductive technology (ART) may affect the epigenetics of early embryogenesis and may cause birth defects--specifically, Angelman syndrome. After nuclear transfer or even exposure to in vitro environments by tissue culture, unusually large offspring have been born; this has been referred to as 'large offspring syndrome,' or LOS (Young et al., 1998). Young et al. (2001) linked overgrowth in ART offspring in sheep to loss of imprinting in the Igf2r gene, although this gene is not imprinted in humans.
DeBaun et al. (2003) provided the first evidence, to their knowledge, that ART is associated with a human overgrowth syndrome: namely, BWS. In a prospective study, 3 of 65 (4.6%) BWS patients were conceived by ART, versus the background rate of 0.8% in the United States. A total of 7 children with BWS were born after ART; 5 of these were conceived after intracytoplasmic sperm injection. Molecular studies of 6 of the children indicated that 5 had specific epigenetic alterations associated with BWS: 4 were at the LIT1 gene and 1 was at both LIT1 and H19.
Weksberg et al. (2003) discussed the mechanisms of growth control, oncogenesis, and genomic imprinting as revealed through studies of BWS, with emphasis on methylation and chromatin modification and possible epigenetic mechanisms associated with the early stages of embryogenesis.
Gicquel et al. (2003) studied a series of 149 patients referred for overgrowth syndromes and diagnosed as BWS. Six of the 149 patients were born following ART. The representation of ART (4%) in this series was 3 times higher than that in the general population (1.3%), as reported by the French Ministry of Health. DeBaun et al. (2003) and Maher et al. (2003) likewise analyzed BW registries and found the proportion of individuals with BWS conceived using in vitro fertilization (IVF) to be 3/65 and 6/149, respectively. The data suggested that approximately 4% of individuals with BWS are conceived using IVF, a figure greater than the generally accepted usage of IVF in these centers. Halliday et al. (2004) conducted a study in the State of Victoria, Australia, where a single clinical genetics service and laboratory provide molecular tests for BWS. This allowed complete ascertainment of children born in Victoria between 1983 and 2003 and diagnosed with BWS by a clinical geneticist. Their results indicated that if a child has BWS, the odds that the child was conceived using IVF was approximately 18 times greater than that for a child without BWS, although the magnitude of this odds ratio should be cautiously interpreted, given the wide confidence intervals (CI). During the period of study, 14,894 babies were born as a result of an IVF procedure (excluding gamete intrafallopian transfer). Using the population-based data, they could then estimate the absolute risk of having a liveborn baby with BWS when IVF is used as the means of conception to be 4/14,894. Halliday et al. (2004) concluded that the fact that the overall risk of BWS in children conceived using IVF remains low and that BWS is, in most cases, associated with a good long-term outcome makes it unlikely that this finding will deter couples from using IVF.
Syndromes involving epigenetic changes reported in animals conceived by assisted reproductive technology (ART) include large offspring syndrome in ruminants (Young et al., 2001). Rossignol et al. (2006) investigated whether the epigenetic imprinting error that occurs with ART is random or is restricted to a specific imprinted domain. They analyzed the methylation status of various imprinted genes in 40 patients with BWS showing loss of methylation at KCNQ1OT1 (11 patients with BWS born after the use of ART and 29 patients with BWS conceived naturally). Three of the 11 (27%) patients conceived using ART and 7 of the 29 (24%) patients conceived normally display an abnormal methylation pattern at a locus other than KCNQ1OT1. The mosaic distribution of epimutations suggested that imprinting is lost after fertilization owing to a failure to maintain methylation marks during preimplantation development.
In a metaanalysis of 5 studies examining genotype/phenotype correlations for the risk of tumor formation in BWS, Rump et al. (2005) found that 55 (13.7%) of 402 BWS patients developed tumors. The majority were Wilms tumors (67%), followed by hepatoblastomas (11%), rhabdomyosarcomas (5%), and neuroblastomas (4%). Compared to patients with normal methylation patterns, who were assigned an odds ratio of 1.00 for tumor development, those with loss of H19 imprinting alone or with loss of H19 and LIT1 imprinting had increased risk for tumor development (odds ratio of 4.01 and 2.63, respectively), whereas patients with loss of LIT1 imprinting alone had a decreased risk of tumor development (odds ratio of 0.33), and no Wilms tumors were seen in them. Patients with normal methylation patterns and CDKN1C mutations also had a lower risk of tumor development. The findings suggested that Wilms tumor development in BWS is primarily associated with dysregulation at the telomeric domain of 11p15 (i.e., H19), rather than at the centromeric domain.
DiagnosisDiagnosis is based on clinical findings. A 'mild' presentation may include prominent tongue and umbilical hernia (Weksberg et al., 2010). A careful cytogenetic analysis of the 11p15 region is recommended. Prenatal diagnosis by ultrasonography is possible (Nivelon-Chevallier et al., 1983; Winter et al., 1986; Cobelis et al., 1988). When the pregnancy is not terminated, the prenatal diagnosis helps to prevent neonatal complications (Viljoen et al., 1991).
Clinical ManagementSince neonatal hypoglycemia is frequent (1 in 3 cases) and potentially deleterious for the CNS, Martinez-y-Martinez et al. (1992) proposed monitoring the glycemia in BWS newborns every 6 hours during the first 3 days in order to correct blood glucose levels below 2.6 mmol/l (46.8 mg/dl).
Adrenal carcinoma, nephroblastoma, hepatoblastoma, and rhabdomyosarcoma occur with increased frequency and justify biannual abdominal ultrasound examinations (Azouz et al., 1990). Wiedemann (1983) recommended that children with this syndrome be examined with renal sonography: first, at 3-month intervals, and after the third year of life, at 6-month intervals. Although less frequent, thoracic neuroblastoma occurs. A periodic chest radiograph is necessary (Sirinelli et al., 1989).
DeBaun and Tucker (1998) studied 183 children with Beckwith-Wiedemann syndrome followed for 482 person-years. Thirteen children (7.1%) were identified with cancers before the fourth year of life, and 6 of the tumors were Wilms tumors. The relative risk of Wilms tumor in Beckwith-Wiedemann syndrome patients over the general population was 816. The relative risk for neuroblastoma was 197 and the relative risk for hepatoblastoma was 2,280. Asymmetry of the limbs, or hemihypertrophy, was the only clinical feature associated with an increased relative risk of cancer, the relative risk being 4.6 with 95% confidence interval, 1.5 to 14.2. In a smaller series by Schneid et al. (1997), 8 of 38 (21%) children with Beckwith-Wiedemann syndrome had tumors, 5 of which (13%) were Wilms tumors. DeBaun et al. (1998) suggested that nephromegaly detected on early renal sonograms may distinguish the subset of patients with BWS at risk for developing Wilms tumor. Twelve of 16 patients with nephromegaly detected on early sonograms subsequently had Wilms tumor. None of the 27 with BWS whose early sonograms revealed normal kidney size subsequently had Wilms tumor. Kidney size was compared with norms for age rather than height. In an accompanying editorial, Beckwith (1998) suggested that until the association of nephromegaly in the neonatal period with Wilms tumor is confirmed in a larger sample, screening BWS patients by renal ultrasound every 3 months for the first 7 years of life should be continued.
Choyke et al. (1998) performed a retrospective review of 152 patients with Beckwith-Wiedemann syndrome ranging in age from 1 day to 30 years old to determine the spectrum of nonmalignant renal disease in patients with this disorder. Thirty-eight (25%) of 152 patients had 45 nonmalignant renal abnormalities, including 19 with medullary renal cysts (13%), 2 with caliceal diverticula (1%), 18 with hydronephrosis (12%), and 6 with nephrolithiasis (4%). Of the 38 patients with nonmalignant renal disease, 33 (87%) were asymptomatic. Of the remaining 5 patients, 4 had urinary tract infections and 1 had flank pain due to obstructive stone disease. Nonmalignant renal disease was mistaken for Wilms tumor in 2 patients, resulting in unnecessary nephrectomies. Seven of the children (18%) had Wilms tumor and nonmalignant renal disease. Choyke et al. (1998) concluded that nonmalignant renal abnormalities occur in approximately 25% of patients with BWS but are generally asymptomatic.
In a series of 18 consecutive patients with BWS, Goldman et al. (2003) found hypercalciuria in 22%, compared with a predicted rate of 7 to 10% in the general population. Three of the 4 patients with hypercalciuria had abnormal renal imaging: 2 with nephrocalcinosis and 1 with hyperechoic kidneys.
Choyke et al. (1999) used a case series analysis to compare the proportion of late-stage Wilms tumor in 15 patients with BWS/idiopathic hemihypertrophy who were screened with sonography to the proportion of late-stage Wilms tumor in 59 unscreened patients with BWS/idiopathic hemihypertrophy. The patients were identified from the BWS Registry and from previously published studies. Screened patients had sonograms at intervals of 4 months or less. None of the 12 screened children with Wilms tumor had late stage disease, whereas 25 of 59 (42%) of unscreened children had late-stage Wilms tumor, a difference that was statistically significant (p less than 0.003). Three children had false-positive screening studies. They were operated on for suspected Wilms tumor but the lesions proved to be complicated renal cysts or nephroblastomatosis. Choyke et al. (1999) concluded that children with BWS/idiopathic hemihypertrophy may benefit from screening sonograms at intervals of 4 months or less; however, false-positive screening results may result in unnecessary surgery.
Population GeneticsThorburn et al. (1970) described 6 cases in Jamaican blacks and estimated a population incidence of 1 in 13,700 births. Weksberg et al. (2010) noted that this figure is likely an underestimate as milder phenotypes may not be ascertained. The incidence is equal in males and females with the notable exception of monozygotic twins that show a dramatic excess of females.
Animal ModelZhang et al. (1997) produced targeted disruption of the p57(KIP2) gene (CDKN1C; 600856) in mice and demonstrated that they have altered cell proliferation and differentiation, leading to abdominal muscle defects; cleft palate; endochondral bone ossification defects with incomplete differentiation of hypertrophic chondrocytes; renal medullary dysplasia; adrenal cortical hyperplasia and cytomegaly; and lens cell hyperproliferation and apoptosis. Since many of these phenotypes are observed in patients with BWS, Zhang et al. (1997) suggested that the observations support a loss of p57(KIP2) expression as having a role in that disorder. Omphalocele was a feature of the mutant mice; mutant embryos showed umbilical abnormalities as early as E16.5. Neonatal lethality was due to defects in the closure of the secondary palate, with aspiration of milk and swallowing of air causing inflation and stretching of the stomach and intestines. Renal medullary dysplasia caused enlargement of the kidneys. Zhang et al. (1997) noted that type X collagen (120110) is expressed in hypertrophic chondrocytes and has been implicated in proper bone development. In mutant mice, expression of type X collagen was significantly reduced in the mutant hypertrophic zone. Thus, the investigators concluded that p57(KIP2) is required for expression of collagen X, and perhaps for other genes that facilitate the ossification of chondrocytes. Expression of p57(KIP2) is restricted to the fetal adrenal cortex and presumably plays a role in controlling cell proliferation; its absence leads to adrenal cortex hyperplasia and cytomegaly. The adrenal gland is among the most consistently enlarged organs in BWS patients. Some other manifestations of BWS are not so easily explained by the loss of control by this cyclin-dependent kinase inhibitory protein, e.g., the defects in kidney development and formation of the secondary palate.
In human and mouse, most imprinted genes are arranged in chromosomal clusters. Their linked organization suggests coordinated mechanisms controlling imprinting and gene expression. Identification of local and regional elements responsible for the epigenetic control of imprinted gene expression is important for understanding the molecular basis of disorders associated with imprinting such as BWS. Paulsen et al. (1998) established a complete contig of clones along the murine imprinting cluster on distal chromosome 7 syntenic with the human imprinting region at chromosome 11p15.5 (BWCR) associated with BWS. The cluster comprises approximately 1 Mb of DNA, contains at least 8 imprinted genes, and is demarcated by the 2 maternally expressed genes Ipl (602131) and H19 (103280), which are directly flanked by the nonimprinted genes Nap1l4 (601651) and L23mrp (600789), respectively. Paulsen et al. (1998) also localized Kvlqt1 (KCNQ1; 607542) and Tapa1 (186845) between Cdkn1c (600856) and Mash2 (601886). The mouse Kvlqt1 gene was maternally expressed in most fetal tissues but biallelically transcribed in most neonatal tissues, suggesting relaxation of imprinting during development.
In connection with reports that in vitro fertilization may increase the risk of BWS, it is noteworthy that in sheep and cattle, epigenetic abnormalities have been shown to be involved in large offspring syndrome (LOS) (Young et al., 1998). Affected animals exhibit various phenotypes, including large size at birth. In both species, the syndrome is caused by the in vitro exposure of embryos, between fertilization and the blastocyst stage, to various unusual environments. LOS is related to the loss of imprinting of the IGF2 receptor gene (147280), which ensures internalization and degradation of IGF2 (147270) and displays an antiproliferative function (Young et al., 2001).
The proximal imprinting center IC1 is located about 2-kb upstream of the H19 gene, and the distal imprinting center IC2 is located within intron 10 of the Kcnq1 gene. Lefebvre et al. (2009) engineered an interstitial deletion of the approximately 280-kb intervening region between the 2 imprinting centers IC1 and IC2 on mouse chromosome 7. The deletion was flanked by the Ins2 and Ascl2 (601886) genes. The deletion allele, Del(7AI), was silent with respect to epigenetic marking at the 2 flanking imprinting centers. Reciprocal inheritance of Del(7AI) demonstrated that the deleted region, which represents more than a quarter of the previously defined imprinted domain, is associated with intrauterine growth restriction in maternal heterozygotes. In homozygotes, the deficiency behaved as a tyrosine hydroxylase (TH; 191290)-null allele and could be rescued pharmacologically by bypassing the metabolic requirement for tyrosine hydroxylase in utero. Lefebvre et al. (2009) concluded that the deleted interval is not required for normal imprinting on distal mouse chromosome 7.
HistoryFor biographical and autobiographical accounts of Hans-Rudolf Wiedemann, see Opitz and Mullen (1992) and Wiedemann (1992), respectively.