Marinesco-Sjögren syndrome is a rare autosomal recessive multisystem disorder featuring cerebellar ataxia, early-onset cataracts, chronic myopathy, variable intellectual disability and delayed motor development. More recently, mutations in the SIL1 gene, which encodes an endoplasmic reticulum resident co-chaperone, were identified as the main cause of Marinesco-Sjögren syndrome. Here we describe the results of SIL1 mutation analysis in 62 patients presenting with early-onset ataxia, cataracts and myopathy or combinations of at least two of these. We obtained a mutation detection rate of 60% (15/25) among patients with the characteristic Marinesco-Sjögren syndrome triad (ataxia, cataracts, myopathy) whereas the detection rate in the group of patients with more variable phenotypic presentation was below 3% (1/37). We report 16 unrelated families with a total of 19 different SIL1 mutations. Among these mutations are 15 previously unreported changes, including single- and multi-exon deletions. Based on data from our screening cohort and data compiled from the literature we found that SIL1 mutations are invariably associated with the combination of a cerebellar syndrome and chronic myopathy. Cataracts were observed in all patients beyond the age of 7 years, but might be missing in infants. Six patients with SIL1 mutations had no intellectual disability, extending the known wide range of cognitive capabilities in Marinesco-Sjögren syndrome to include normal intelligence. Modestly constant features were somatic growth retardation, skeletal abnormalities and pyramidal tract signs. Examination of mutant SIL1 expression in cultured patient lymphoblasts suggested that SIL1 mutations result in severely reduced SIL1 protein levels irrespective of the type and position of mutations. Our data broaden the SIL1 mutation spectrum and confirm that SIL1 is the major Marinesco-Sjögren syndrome gene. SIL1 patients usually present with the characteristic triad but cataracts might be missing in young children. As cognitive impairment is not obligatory, patients without intellectual disability but a Marinesco-Sjögren syndrome-compatible phenotype should receive SIL1 mutation analysis. Despite allelic heterogeneity and many families with private mutations, the phenotype related to SIL1 mutations is relatively homogenous. Based on SIL1 expression studies we speculate that this may arise from a uniform effect of different mutations on protein expression.

Among the clusters of imprinted genes in humans, one of the most relevant regions involved in human growth is localised in 11p15. Opposite epigenetic and genomic disturbances in this chromosomal region contribute to two distinct imprinting disorders associated with disturbed growth, Silver–Russell and Beckwith–Wiedemann syndromes. Due to the complexity of the 11p15 imprinting regions and their interactions, the interpretation of the copy number variations in that region is complicated. The clinical outcome in case of microduplications or microdeletions is therefore influenced by the size, the breakpoint positions and the parental inheritance of the imbalance as well as by the imprinting status of the affected genes. Based on their own new cases and those from the literature, the authors give an overview on the genotype–phenotype correlation in chromosomal rearrangements in 11p15 as the basis for a directed genetic counselling. The detailed characterisation of patients and families helps to further delineate risk figures for syndromes associated with 11p15 disturbances. Furthermore, these cases provide us with profound insights in the complex regulation of the (imprinted) factors localised in 11p15.

In all known congenital imprinting disorders an association with aberrant methylation or mutations at specific loci was well established. However, several patients with transient neonatal diabetes mellitus (TNDM), Silver-Russell syndrome (SRS) and Beckwith-Wiedemann syndrome (BWS) exhibiting multilocus hypomethylation (MLH) have meanwhile been described. Whereas TNDM patients with MLH show clinical symptoms different from carriers with isolated 6q24 aberrations, MLH carriers diagnosed as BWS or SRS present only the syndrome-specific features. Interestingly, SRS and BWS patients with nearly identical MLH patterns in leukocytes have been identified. We now report on the molecular findings in DNA in three SRS patients with hypomethylation of both 11p15 imprinted control regions (ICRs) in leukocytes. One patient was a monozygotic (MZ) twin, another was a triplet. While the hypomethylation affected both oppositely imprinted 11p15 ICRs in leukocytes, in buccal swab DNA only the ICR1 hypomethylation was visible in two of our patients. In the non-affected MZ twin of one of these patients, aberrant methylation was also present in leukocytes but neither in buccal swab DNA nor in skin fibroblasts. Despite mutation screening of several factors involved in establishment and maintenance of methylation marks including ZFP57, MBD3, DNMT1 and DNMT3L the molecular clue for the ICR1/ICR2 hypomethylation in our patients remained unclear. Furthermore, the reason for the development of the specific SRS phenotype is not obvious. In conclusion, our data reflect the broad range of epimutations in SRS and illustrate that an extensive molecular and clinical characterization of patients is necessary.

In addition to patients with the classical Silver-Russell syndrome phenotype fulfilling the Silver-Russell syndrome-specific scores, genetic testing for the 11p15 epimutation and/or maternal uniparental disomy of chromosome 7 should also be considered in case of "Silver-Russell syndrome-like" phenotypes, for example, mild intrauterine growth restriction and postnatal growth retardation associated with a prominent forehead and triangular face or asymmetry as the only clinical signs. In particular, the lack of intrauterine growth restriction in patients with a Silver-Russell syndrome-like phenotype should not automatically result in exclusion from molecular testing.

No abstract