NSD1 mutations generate a genome-wide DNA methylation signature

Abstract

Sotos syndrome (SS) represents an important human model system for the study of epigenetic regulation; it is an overgrowth/intellectual disability syndrome caused by mutations in a histone methyltransferase, NSD1. As layered epigenetic modifications are often interdependent, we propose that pathogenic NSD1 mutations have a genome-wide impact on the most stable epigenetic mark, DNA methylation (DNAm). By interrogating DNAm in SS patients, we identify a genome-wide, highly significant NSD1(+/-)-specific signature that differentiates pathogenic NSD1 mutations from controls, benign NSD1 variants and the clinically overlapping Weaver syndrome. Validation studies of independent cohorts of SS and controls assigned 100% of these samples correctly. This highly specific and sensitive NSD1(+/-) signature encompasses genes that function in cellular morphogenesis and neuronal differentiation, reflecting cardinal features of the SS phenotype. The identification of SS-specific genome-wide DNAm alterations will facilitate both the elucidation of the molecular pathophysiology of SS and the development of improved diagnostic testing.

Figure 1. DNA methylation signature associated with NSD1^+/− mutations.

Unsupervised hierarchical clustering of 72 samples using the differentially methylated CpG sites comprising the NSD1^+/−-specific signature is shown using Pearson correlation. Note that two patients (A1208 and DL179067) harbour the same mutation in NSD1 at the end of the gene (exon 22) and display slightly different DNA methylation changes compared with the other SS patients. Orange indicates high DNA methylation and blue indicates low DNA methylation. Pink bar represents SS patients with NSD1 whole-gene deletion and truncating mutations. Blue bar represents controls.

Figure 2. Testing the sensitivity and the specificity of the NSD1^+/− DNAm signature.

Using 19 SS and 53 healthy subjects from the discovery cohort (labelled as ‘primary'), we generated the median-methylation profiles of SS and control, respectively, on the CpG sites comprising the NSD1^+/−-specific signature. We then estimated the specificity of these profiles on a validation set of GEO blood samples (n=1,056, blue dots), all of which were more similar to the control profile (specificity 100%). We also estimated the sensitivity on a separate SS validation set (n=19, magenta dots), all of which were more similar to the SS profile (100% sensitivity). Similarity was computed as the Pearson correlation to either the SS or the control DNAm profile. Out of 16 missense samples (green squares), 9 classify with SS and 7 with controls. Also shown are Weaver patients with EZH2 mutations (orange triangles) as well as the classification of the primary 53 controls (dark blue squares) and 19 SS (light magenta squares) used in the derivation of the NSD1^+/−-specific signature.

Figure 3. NSD1^+/−-specific signature can be identified in SS fibroblasts.

Unsupervised hierarchical clustering of three SS-derived fibroblasts (Supplementary Data 1) and four control-derived fibroblasts clearly distinguish SS from controls. Orange indicates high DNA methylation and blue indicates low DNA methylation. Pink bar represents Sotos samples with NSD1 whole-gene deletion and truncating mutations. Blue bar represents controls.

Figure 4. Photographs of two patients carrying NSD1 missense variants.

(a) Patient (DL136303) with a missense mutation of uncertain significance, clinical presentation characteristic of Sotos syndrome and a positive SS score based on our analysis. (b) Patient (DL181344) with a missense mutation of uncertain significance, clinical presentation that is not characteristic of Sotos syndrome and a negative SS score based on our analysis. Specific consent to publish facial photographs was obtained for the two patients.

Figure 5. Genomic distribution of the NSD1^+/−-specific CpG sites.

Bar chart representing the percentage distribution of the CpG sites according to genomic annotations extracted from the Illumina 450K array annotation file. We compared the distribution of the CpGs between the data set (424,586 CpGs) and NSD1^+/− CpGs (7,085 CpGs) for the regulatory feature group, relation to CpG island and other functional categories such as overlapping enhancer region, DNase hypersensitive sites (DHS) and type of differentially methylated regions (Reprogrammed DMR (RDMR), cancer DMR (cDMR), other DMR) in addition to relation to RefSeq group annotation. The numbers at the top of each bar represent the percentage distribution of CpGs within each category. N refers to north and S refers to south.

Figure 6. Enrichment analysis of the genes overlapping NSD1^+/− DNAm signature.

We identified 2,167 unique genes that overlap the CpGs comprising the NSD1^+/−-specific signature and used DAVID (http://david.abcc.ncifcrf.gov) to identify the biological processes most enriched within our data set. Over-represented functional categories were visualized in Cytoscape (http://www.cytoscape.org) using the Enrichment Map plugin (www.baderlab.org/Software/EnrichmentMap). Network nodes represent statistically significant Gene Ontology—Biological Process terms (Benjamini-corrected P value <0.05), with node size proportional to the number of SS-associated genes annotated for each term. Edges represent overlaps between these gene sets, with edge thickness proportional to their Jaccard index. The results demonstrate enrichment in functional terms that relate to cellular morphogenesis, neuronal development and cellular differentiation, involving highly overlapping subsets of genes. In addition, we identified an enrichment of genes with roles in organ development, ion transport as well as embryonic developmental pathways.