Beckwith-Wiedemann syndrome (BWS) is characterized by cancer predisposition, overgrowth and highly variable association of macroglossia, abdominal wall defects, nephrourological anomalies, nevus flammeus, ear malformations, hypoglycemia, hemihyperplasia, and organomegaly. BWS molecular defects, causing alteration of expression or activity of the genes regulated by two imprinting centres (IC) in the 11p15 chromosomal region, are also heterogeneous. In this paper we define (epi)genotype-phenotype correlations in molecularly confirmed BWS patients. The characteristics of 318 BWS patients with proven molecular defect were compared among the main four molecular subclasses: IC2 loss of methylation (IC2-LoM, n = 190), IC1 gain of methylation (IC1-GoM, n = 31), chromosome 11p15 paternal uniparental disomy (UPD, n = 87), and cyclin-dependent kinase inhibitor 1C gene (CDKN1C) variants (n = 10). A characteristic growth pattern was found in each group; neonatal macrosomia was almost constant in IC1-GoM, postnatal overgrowth in IC2-LoM, and hemihyperplasia more common in UPD (Po0.001). Exomphalos was more common in IC2/CDKN1C patients (Po0.001). Renal defects were typical of UPD/IC1 patients, uretheral malformations of IC1-GoM cases (Po0.001). Ear anomalies and nevus flammeus were associated with IC2/CDKN1C genotype (Po0.001). Macroglossia was less common among UPD patients (Po0.001). Wilms' tumor was associated with IC1-GoM or UPD and never observed in IC2-LoM patients (Po0.001). Hepatoblastoma occurred only in UPD cases. Cancer risk was lower in IC2/CDKN1C, intermediate in UPD, and very high in IC1 cases (P = 0.009).In conclusion, (epi)genotype-phenotype correlations define four different phenotypic BWS profiles with some degree of clinical overlap. These observations impact clinical care allowing to move toward (epi) genotype-based follow-up and cancer screening.

In this study we applied a new analytical strategy to investigate the relations between stochastic epigenetic mutations (SEMs) and aging. We analysed methylation levels through the Infinium HumanMethylation27 and HumanMethylation450 BeadChips in a population of 178 subjects ranging from 3 to 106 years. For each CpG probe, epimutated subjects were identified as the extreme outliers with methylation level exceeding three times interquartile ranges the first quartile (Q1-(3 × IQR)) or the third quartile (Q3+(3 × IQR)). We demonstrated that the number of SEMs was low in childhood and increased exponentially during aging. Using the HUMARA method, skewing of X chromosome inactivation (XCI) was evaluated in heterozygotes women. Multivariate analysis indicated a significant correlation between log(SEMs) and degree of XCI skewing after adjustment for age (β = 0.41; confidence interval: 0.14, 0.68; p-value = 0.0053). The PATH analysis tested the complete model containing the variables: skewing of XCI, age, log(SEMs) and overall CpG methylation. After adjusting for the number of epimutations we failed to confirm the well reported correlation between skewing of XCI and aging. This evidence might suggest that the known correlation between XCI skewing and aging could not be a direct association but mediated by the number of SEMs.

BackgroundMultiple (epi)genetic defects affecting the expression of the imprinted genes within the 11p15.5 chromosomal region underlie Silver–Russell (SRS) and Beckwith–Wiedemann (BWS) syndromes. The molecular diagnosis of these opposite growth disorders requires a multi-approach flowchart to disclose known primary and secondary (epi)genetic alterations; however, up to 20 and 30 % of clinically diagnosed BWS and SRS cases remain without molecular diagnosis. The complex structure of the 11p15 region with variable CpG methylation and low-rate mosaicism may account for missed diagnoses. Here, we demonstrate the relevance of complementary techniques for the assessment of different CpGs and the importance of testing multiple tissues to increase the SRS and BWS detection rate.ResultsMolecular testing of 147 and 450 clinically diagnosed SRS and BWS cases provided diagnosis in 34 SRS and 185 BWS patients, with 9 SRS and 21 BWS cases remaining undiagnosed and herein referred to as “borderline.” A flowchart including complementary techniques and, when applicable, the analysis of buccal swabs, allowed confirmation of the molecular diagnosis in all borderline cases. Comparison of methylation levels by methylation-specific multiplex ligation-dependent probe amplification (MS-MLPA) in borderline and control cases defined an interval of H19/IGF2:IG-DMR loss of methylation that was distinct between “easy to diagnose” and “borderline” cases, which were characterized by values ≤mean −3 standard deviations (SDs) compared to controls. Values ≥mean +1 SD at H19/IGF2: IG-DMR were assigned to borderline hypermethylated BWS cases and those ≤mean −2 SD at KCNQ1OT1: TSS-DMR to hypomethylated BWS cases; these were supported by quantitative pyrosequencing or Southern blot analysis. Six BWS cases suspected to carry mosaic paternal uniparental disomy of chromosome 11 were confirmed by SNP array, which detected mosaicism till 10 %. Regarding the clinical presentation, borderline SRS were representative of the syndromic phenotype, with exception of one patient, whereas BWS cases showed low frequency of the most common features except hemihyperplasia.ConclusionsA conclusive molecular diagnosis was reached in borderline methylation cases, increasing the detection rate by 6 % for SRS and 5 % for BWS cases. The introduction of complementary techniques and additional tissue analyses into routine diagnostic work-up should facilitate the identification of cases undiagnosed because of mosaicism, a distinctive feature of epigenetic disorders.Electronic supplementary materialThe online version of this article (doi:10.1186/s13148-016-0183-8) contains supplementary material, which is available to authorized users.