Endocrine Abstracts

Background/objectives: Kabuki Syndrome 1 (KS1) is a neurodevelopmental disorder caused by loss-of-function of histone 3 lysine 4 mono-methyltransferase KMT2D. In addition, to neurodevelopmental features, some Kabuki Syndrome patients also exhibit endocrine-related phenotypes, such as hypoglycaemia. KMT2D is involved in global gene regulation, therefore, it is important to have a systems-based approach to understand pathomechanisms of KS1.

Methods: DNA methylation samples from blood were used to extract differentially methylated points (DMPs), which were used for systems analysis.

Results: In KS1 cohort (n=22) we identified 2002 DMPs (753 hypermethylated and 1249 hypomethylated; adjusted P-value < 1×10−4), compared to Control (n=138). Focussing on genomic regions with >7 contiguous (in cis) hypermethylated or hypomethylated DMPs, we identified 17 significant differentially methylated regions (FWER<0.05, 11 hypermethylated and 6 hypomethylated) associated with organ morphogenesis and skeletal system development pathways, with 13 DMRs being novel for KS1. This approach, however, failed to extract functional relevance of >90% DMPs form our data, therefore, we performed hypernetwork analysis to identify indirect co-ordinations between DMPs. This revealed 986/2002 DMPs to be highly co-ordinated (strongly correlated but majority in trans and not necessarily methylated in the same direction) in KS1. These DMPs were enriched for genes associated with extracellular matrix organization, cartilage development and neuronal migration. Finally, using an iterative analysis of 1000 network simulations we detected significantly lower Shannon entropy in KS1 compared to controls (P<1×10−4). This shows more ordered and less diverse co-ordination of DMPs in KS1 compared to Control. The analysis of entropy within Gene Ontology processes has shown that DNA methylation associated with Interleukin-17-mediated signalling pathway and Microtubule bundle formation is more ordered in KS1 samples compared to controls.

Conclusions: Hypernetwork approach is useful in quantifying network-level differences, and in extracting deeper mechanistic insights into the fundamental pathophysiology of genetic disorders, which might be overlooked with traditional DNA methylation analyses. This is especially important with rapid increase in patient-derived epigenome data, as this approach may have significant translational potential.