Open Access Highly Accessed Research article

Multilocus loss of DNA methylation in individuals with mutations in the histone H3 Lysine 4 Demethylase KDM5C

Daria Grafodatskaya1, Barian HY Chung23, Darci T Butcher1, Andrei L Turinsky4, Sarah J Goodman3, Sana Choufani3, Yi-An Chen1, Youliang Lou1, Chunhua Zhao1, Rageen Rajendram1, Fatima E Abidi5, Cindy Skinner5, James Stavropoulos6, Carolyn A Bondy7, Jill Hamilton89, Shoshana Wodak104, Stephen W Scherer1011, Charles E Schwartz5 and Rosanna Weksberg29*

Author Affiliations

1 Genetics and Genome Biology Program, Hospital for Sick Children, Toronto, ON, Canada

2 Division of Clinical and Metabolic Genetics, Hospital for Sick Children, Toronto, ON, Canada

3 Centre of Reproduction, Growth & Development, Department of Pediatrics & Adolescent Medicine, The University of Hong Kong, Hong Kong, Hong Kong

4 Program in Molecular Structure and Function, Hospital for Sick Children, Toronto, ON, Canada

5 J.C. Self Research Institute, Greenwood Genetic Center, Greenwood, SC, USA

6 Department of Pediatric Laboratory Medicine, Hospital for Sick Children, Toronto, ON, Canada

7 Developmental Endocrinology Branch, National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, MD, USA

8 Division of Endocrinology, Department of Pediatrics, Hospital for Sick Children, Toronto, ON, Canada

9 Department of Pediatrics, University of Toronto, Toronto, ON, Canada

10 Department of Molecular and Medical Genetics, University of Toronto, Toronto, ON, Canada

11 The Centre for Applied Genomics, Hospital for Sick Children, Toronto, ON, Canada

For all author emails, please log on.

BMC Medical Genomics 2013, 6:1  doi:10.1186/1755-8794-6-1

Published: 28 January 2013

Additional files

Additional file 1:

Table S1. Clinical features and demographic information of the 10 male patients with X-linked intellectual disability due to KDM5C mutations, 2 male patients with a variant of unknown significance. Table S2: Primer Sequences. Table S3: Number of significant CpG sites with loss and gain of DNA methylation detected by multivariate permutation analysis for different levels of confidence (1-α) and false discovery proportion limit (γ). Table S4: The top 53 most significant CpG sites that are differentially methylated between the KDM5C mutations and normal controls with the lowest FDP = 0 and the highest confidence level of 99.5%. Table S5: The top 53 most significant CpG sites with additional CpG sites within the same genes. Table S6: GEO studies used to assess DNA methylation at at FBXL5, SCMH1 and CACYBP in blood samples of population controls. Table S7: Analysis of sex specific DNA methylation differences in blood samples from diabetes study (GSE20067). Table S8: Analysis of sex specific DNA methylation differences in brain in FBXL5, SCMH1 and CACYBP promoters.

Format: XLS Size: 147KB Download file

This file can be viewed with: Microsoft Excel Viewer

Open Data

Additional file 2:

Figure S1. Scatter plots of DNA methylation determined by Illumina (X-axis) and pyrosequencing (Y-axis) at overlapping CpG sites. Grey diamonds are cases with KDM5C mutations, black squares are controls and grey triangles are individuals with p.R1546Q variant. Figure S2: Visual mapping of all 31 samples to the coordinate space defined by the first two principal components reveals that there is no clear separation between samples with and without KDM5C mutations. The principal component analysis (PCA) was performed using all 23,837 CpG sites. The red dots represent the 10 mutation cases, the 19 light green dots represent controls and the two dark green dots represent the benign mutation variants (p.R1546Q). Figure S3: Unsupervised hierarchical clustering of methylation data at 23, 837 CpG sites reveals that there is no clear separation between samples with and without KDM5C mutations. C1-16, are unrelated controls, UN-R1-2 are unaffected relative. R1546Q 1–2 are cases with benign variant, and the rest of the samples are KDM5C mutation cases. Figure S4: Visual mapping of all 31 samples to the coordinate space defined by the first two principal components. The principal component analysis (PCA) was performed using only the methylation levels at the 53 most significant CpG sites (the same CpG sites as shown in Figure 1). The red dots represent the 10 KDM5C mutation cases, the 19 light green dots represent controls, and the two dark green dots represent the benign mutation variants (p.R1546Q). Although the PCA procedure did not use any information on the mutation status, the data distribution shows a clear separation between aberrant mutations and benign mutations and controls. Figure S5: DNA methylation levels at 3 CpG sites in promoter of Long Interspersed Element-1 (LINE-1) as determined by pyrosequencing. The Y-axis is DNA methylation%. Two groups of samples, controls (C; N = 19) and KDM5C mutation cases (K;N = 10) are shown on the X-axis. Figure S6: Regional DNA methylation in SCMH1 promoter. A) Screenshot from the UCSC genome browser showing location of SCMH1 promoter exon1, intron1, CpG island, Illumina 27 K microarray probes and pyrosequencing assays and tracks for H3K4me1 and H3K4me3 in lymphoblastoid cell line (GM12878) and ES cell line (H1 h-ES) (Broad Institute Histone data). B&C) Boxplots of Illumina DNA methylation data for two microrray probes within SCMH1 gene. The Y axis shows DNA methylation levels presented as C/C + T and ranging from 0 to 1. The bottom and the top of the box are 25th and 75th percentiles respectively, the whiskers are within the 1.5 interquartile range (IQR) of the data, and the circles, are outlier data points above or below 1.5 IQR. C are controls (N = 19), K are cases with KDM5C mutations (N = 10), V are cases with p.R1546Q sequence variant. D) DNA methylation upstream of SCMH1 transcription start site as determined by pyrosequencing assay (pyro). The order of CpG sites are shown from the further upstream towards the transcription start site. Each column is an average for each of three groups of 1)10 cases with KDM5C mutations (mutation), 2)19 controls (control), and 3) two individuals with the p.R1546Q variant of unknown clinical significance (VUS). The arrows show the CpG sites from the microarray. P-values were determined by Kruskal-Wallis test between mutation cases and controls. Figure S7: Regional DNA methylation in CACYBP promoter. A) Screenshot from the UCSC genome browser showing location of CACYBP promoter, exons1&2, introns 1&2, CpG island, Illumina 27 K microarray probes and pyrosequencing assays and tracks for H3K4me1 and H3K4me3 in lymphoblastoid cell line (GM12878) and ES cell line (H1 h-ES) (Broad Institute Histone data). B&C) Boxplots of Illumina DNA methylation data for two microrray probes within CACYBP gene. The Y axis shows DNA methylation levels presented as C/C + T and ranging from 0 to 1. The bottom and the top of the box are 25th and 75th percentiles respectively, the whiskers are within the 1.5 interquartile range (IQR) of the data, and the circles, are outlier data points above or below 1.5 IQR. C are controls (N = 19), K are cases with KDM5C mutations (N = 10), V are cases with p.R1546Q sequence variant. D) DNA methylation upstream of CACYBP transcription start site as determined by pyrosequencing assay (pyro). The order of CpG sites are shown from the further upstream towards the transcription start site. Each column is an average for each of three groups of 1)10 cases with KDM5C mutations (mutation), 2)19 controls (control), and 3) two individuals with the p.R1546Q variant of unknown significance (VUS). P-values were determined by Kruskal-Wallis test between mutation cases and controls. **** is p <0.0001, * is p <0.05. Figure S8: Regional DNA methylation in ZMYMD12/PPCS promoter. A) Screenshot from the UCSC genome browser showing location of ZMYND12 and PPCS promoter, CpG island, Illumina 27 K microarray probes and pyrosequencing assays and tracks for H3K4me1 and H3K4me3 in lymphoblastoid cell line (GM12878) and ES cell line (H1 h-ES) (Broad Institute Histone data). B&C) Boxplots of Illumina DNA methylation data for two microrray probes within ZMYND12 gene. The Y axis shows DNA methylation levels presented as C/C + T and ranging from 0 to 1. The bottom and the top of the box are 25th and 75th percentiles respectively, the whiskers are within the 1.5 interquartile range (IQR) of the data, and the circles, are outlier data points above or below 1.5 IQR. C are controls (N = 19), K are cases with KDM5C mutations (N = 10), V are cases with p.R1546Q sequence variant. D) DNA methylation upstream of PPCS transcription start site and within exon1/intrion1 of ZMYND12 as determined by pyrosequencing assay (pyro). The order of CpG sites are shown from the further upstream towards the PPCS transcription start site. Each column is an average for each of three groups of 1)10 cases with KDM5C mutations (mutation), 2)19 controls (control), and 3) two individuals with the p.R1546Q variant of unknown significance (VUS). P-values were determined by Kruskal-Wallis test between mutation cases and controls. **** is p <0.0001. Figure S9: Regional DNA methylation in DYDC1/DYDC2 promoter. A) Screenshot from the UCSC genome browser showing location of DYDC1/DYDC2 promoter, CpG island, Illumina 27 K microarray probes and pyrosequencing assays and tracks for H3K4me1 and H3K4me3 in lymphoblastoid cell line (GM12878) and ES cell line (H1 h-ES) (Broad Institute Histone data). B&C) Boxplots of Illumina DNA methylation data for two microrray probes within DYDC1 gene. The Y axis shows DNA methylation levels presented as C/C + T and ranging from 0 to 1. The bottom and the top of the box are 25th and 75th percentiles respectively, the whiskers are within the 1.5 interquartile range (IQR) of the data, and the circles, are outlier data points above or below 1.5 IQR. C are controls (N = 19), K are cases with KDM5C mutations (N = 10), V are cases with p.R1546Q sequence variant. D) DNA methylation upstream of DYDC1 transcription start site as determined by pyrosequencing assay (pyro) overlapping Illumina CpG site cg17703212. Column is an average for each of three groups of 1)10 cases with KDM5C mutations (mutation), 2)19 controls (control), and 3) two individuals with the p.R1546Q variant of uknown significance (VUS). P-values were determined by Kruskal-Wallis test between mutation cases and controls. **** is p <0.0001. Figure S10: Expression of FBXL5, SCMH1 and CACYBP in a panel of human tissues: BT -brain total, FB- fetal brain, CC-cerebral cortex, CB-cerebellum, TL- temporal lobe, FL –frontal lobe, SC –spinal cord, K-kidney, H-heart, L-liver, SM-smooth muscle, SK-skeletal muscle, LCL-lymphoblastoid cell line. Error bars are standard deviation of duplicated q-PCR experiments. Figure S11: DNA methylation levels of three CpG sites cg02630888 (FBXL5), cg03387723 (SCMH1), cg16743289 (CACYBP) in different blood cell types (GSE35069). The cell types are whole blood, peripheral blood mononuclear cells (PBMC), granulocytes and seven isolated cell populations (CD4+ T cells, CD8+ T cells, CD56+ NK cells, CD19+ B cells, CD14+ monocytes, neutrophils (Neu), and eosinophils (Eos)) which are each shown by different colour. Y-axis is beta methylation value determined by Illumina methylation450 array (GSE35069). Each column represents mean methylation from 6 samples from healthy males, error bars are standard deviation from the mean.

Format: PPT Size: 1.1MB Download file

This file can be viewed with: Microsoft PowerPoint Viewer

Open Data