ESPE Abstracts (2019) 92 FC12.6

ESPE2019 Free Communications Growth and Syndromes (to include Turner Syndrome) (6 abstracts)

An Integrated Systems Biology Analysis of the Genome, Epigenome and Transcriptome Identifies a Distinct Pattern of Hypermethylation Associated with Low Childhood Growth

Terence Garner , Robert Sellers , Hui Guo , Andrew Whatmore , Peter Clayton , Adam Stevens & Philip Murray


University of Manchester, Manchester, United Kingdom


Background: Current data from genome wide association studies (GWAS) explains 24.6% of the variation in adult height from 3290 single nucleotide polymorphisms (SNPs)1. Data on the genetic control of growth velocity during childhood is more limited and no previous studies have linked childhood growth to changes in the transcriptome (gene expression) or epigenome (DNA methylation). Here we present a systems biology approach to understand mid-childhood growth velocity using genome wide SNP, transcriptome and epigenome analysis.

Study Design: Height velocity (HV) between ages-7 and 9-years was modelled using a mixed-effects model including gender, age, height at start of analysis and pubertal stage in 6487 normal children from the Avon Longitudinal Study of Parents and Children. Blood transcriptome was available in 947 children age-9 and blood epigenome from 980 children age-7. Differentially methylated regions (DMRs), genomic regions that contain multiple methylation sites (CpGs), were defined using the R package ChAMP. A hypernetwork approach was used to integrate differentially expressed genes and methylated CpGs to assess functional association between omic data layers. A GWAS was conducted (n=4781) to identify SNPs correlated with HV as a continuous variable. Expression quantitative trait loci (eQTLs) were identified linking GWAS findings to levels of gene expression at age-9.

Results: Height velocity between 7 and 9 years was associated with multi-omic patterns distinguishing children with low (LHV) and high HV (HHV). A greater number of transcriptomic differences (p-value range, 6.9x10-5 to <1.0x10-3) were identified in LHV children than HHV (34 LHV vs 23 HHV). This was also apparent in the epigenome, where there was a greater impact with a 7-fold larger number of identified CpGs with differences in methylation (1618 LHV vs 140 HHV, p-range 3.0x10-7 to <1.0x10-3). A highly significant DMR was identified in the LHV: in HOXA5 (corrected-P<5x10-3), a homeobox regulator involved in morphogenesis. GWAS identified 1264 significant SNPs (p-range 8.4x10-8 to<1x10-4); 485 eQTLs were detected across 253 unique genes. GWAS results included 116 significant SNPs associated with the growth hormone receptor (corrected-P=1.1x10-3), and 52 SNPs associated with MSRA (corrected-P=5.8x10-5), a gene implicated in biological aging and insulin resistance – both of which also had eQTLs.

Conclusions: We have demonstrated that in normal children (1) a distinct hypermethylation pattern is associated with low, but not high, height velocity and (2) HV-related polymorphisms in growth and metabolic genes are linked to levels of gene expression across the genome.

1Yengo, L., et al. Hum.Mol.Gen 2018;27(20):3641-3649.

Volume 92

58th Annual ESPE

Vienna, Austria
19 Sep 2019 - 21 Sep 2019

European Society for Paediatric Endocrinology 

Browse other volumes

Article tools

My recent searches

No recent searches.