A framework to identify contributing genes in patients with Phelan-McDermid syndrome

Abstract

Phelan-McDermid syndrome (PMS) is characterized by a variety of clinical symptoms with heterogeneous degrees of severity, including intellectual disability (ID), absent or delayed speech, and autism spectrum disorders (ASD). It results from a deletion of the distal part of chromosome 22q13 that in most cases includes the SHANK3 gene. SHANK3 is considered a major gene for PMS, but the factors that modulate the severity of the syndrome remain largely unknown. In this study, we investigated 85 patients with different 22q13 rearrangements (78 deletions and 7 duplications). We first explored the clinical features associated with PMS, and provide evidence for frequent corpus callosum abnormalities in 28% of 35 patients with brain imaging data. We then mapped several candidate genomic regions at the 22q13 region associated with high risk of clinical features, and suggest a second locus at 22q13 associated with absence of speech. Finally, in some cases, we identified additional clinically relevant copy-number variants (CNVs) at loci associated with ASD, such as 16p11.2 and 15q11q13, which could modulate the severity of the syndrome. We also report an inherited SHANK3 deletion transmitted to five affected daughters by a mother without ID nor ASD, suggesting that some individuals could compensate for such mutations. In summary, we shed light on the genotype-phenotype relationship of patients with PMS, a step towards the identification of compensatory mechanisms for a better prognosis and possibly treatments of patients with neurodevelopmental disorders.

Fig. 1

Clinical and genetic description of the cohort. a Spider plot of the clinical profile of the patients with PMS as a percentage of cases having each feature over those tested. b Mechanisms and inheritance of identified 22q13 CNVs in the cohort. ABCR apparently balanced chromosomal rearrangement, der(22) derived from unbalanced reciprocal translocation, ring(22) caused by a ring chromosome 22

Fig. 2

Multivariate analysis of the 22q13 deletion size, sex and clinical features of patients. a Hierarchical clustering based on multivariate analysis shows four clusters. b Tree view showing the four clusters of patients and the main genomic and clinical features of each cluster. * indicate patients carrying mosaic CNVs

Fig. 3

Mapping of genomic regions at 22q13 associated with high risk of presenting clinical features. a Prevalence is measured each 50 kb, within overlapping windows of 1.5 Mb (Materials and Methods). Dashed lines represent the global prevalence of each feature, measured as the fraction of patients with 22q13 deletions presenting the feature over all patients with 22q13 deletions. Dotted lines and colored areas represent standard errors of the proportion. Black solid lines show the numbers of informative patients for each window, and black dotted lines correspond to the minimum number of individuals required by window (n = 3) to reduce interpretation biases. b The 22q13 region is represented with the genes (block: exon, line: intron, arrow: strand) and the regions corresponding to a higher than global prevalence for each feature (top). Regions significantly associated with absence of speech and ophthalmic features are also shown in more details (bottom). HI haploinsufficiency, pLI probability of loss-of-function intolerance. Color code similar to a

Fig. 4

Additional CNVs identified in the cohort and including at least one NP-gene. Deletions (red) and duplications (blue) are represented for each patient and each NP-gene. Green squares indicate individuals with seizures. * indicate patients carrying mosaic CNVs

Fig. 5

Multiplex family with inherited SHANK3 and NRXN1 deletions and KAL1 duplication. a The nine studied individuals are displayed in the pedigree of the family (red border). b CNVs detected by OmniExpress Illumina arrays and including the NRXN1, SHANK3 and KAL1 genes (log2 scale)