Introduction
Co
nclu
sio
n
Brain tumours are the most common paediatric solid tumours, and are the leading cause of cancer-related death
in children under 14 years of age (1,2). Although molecular-targeted treatments have been developed for many
kinds of cancer, the impact for brain tumours in children has been limited. We are conducting a detailed genetic
and epigenetic investigation of a large set of paediatric low-grade astrocytomas to improve our understanding of
the biology of these tumours, and identify targets that can be used to develop new forms of treatment.
Astrocytomas are classified into four grades on the basis of biological and histological features (3). Low-grade
astrocytomas (grades I & II) predominate in children, while high-grade tumours (grades III & IV) are mainly found
in adults. Pilocytic astrocytomas (grade I) are the most common type of astrocytoma in children aged 0-19, and
are well-circumscribed, slow growing, cystic lesions which are usually successfully treated by surgical excision (4).
They can arise throughout the neuraxis, but are usually situated in the cerebellum. In contrast, childhood diffuse
astrocytomas (grade II) infiltrate the brain and complete excision may not be possible, leading to a lower 5-year
survival rate than for pilocytic astrocytomas (4). The key molecular changes identified so far in paediatric low
grade astrocytomas are RAF gene fusions that activate the MAPK signaling pathway in pilocytic astrocytomas
(5,6), and abnormalities of the MYB oncogene in certain grade II astrocytomas (7).
To obtain a comprehensive understanding of gene deregulation in paediatric low grade astrocytomas, we have
performed a genome-wide study of DNA methylation in 20 pilocytic astrocytomas and 10 diffuse astrocytomas
using the Illumina 450K BeadChip (8). This system measures methylation at 450,000 CpG sites within gene
promoter regions, external regions such as shores, all known differentially methylated (DMR) and cancer-specific
methylated (CMR) regions, as well as intergenic CpGs specific for stem cells. We present here our findings which
show distinct methylation patterns for genes involved in brain development and drug resistance in pilocytic and
diffuse astrocytomas.
We are grateful to Gabriel Doctor and
Mohamad Ayub Bin Abd Gani
for assistance with data analysis
Method 20 pilocytic astrocytomas and 10 diffuse astrocytomas were obtained as surgical specimens. Age of patients at
diagnosis ranged from 3 to 20 years. Access to tumours and linked clinical data was given in accordance with
Institutional Review Board and MREC regulations: St Jude Children’s Research Hospital (USA) XPD07-107/IRB,
and Tissue Resource Request No 07-007; Newcastle (UK) REC ref No 2002/112; Blizard Institute (UK)
ICMS/PR/09/77. Control samples were foetal cerebellum, brain and frontal lobe; adult brain and neural stem cells
(BioChain). Normal human astrocyte data was obtained from the UCSC browser.The DNA methylation data were
analysed using Illumina Genome Studio software, as well as MethLab
(http://genetics.emory.edu/conneely/MethLAB). Validation was performed using pyrosequencing. Gene
expression was analysed using Affymetrix U133 plus 2.0 arrays, GeneSpring software and RT-PCR.
Tumour cohort
20 Pilocytic astrocytomas
10 Diffuse astrocytomas
&
Controls
Bisulphite Converted
DNA
Array Hybridisation
Quality control
for bisulphite conversion and signal
detection.
Beta values used to produce
methylation profiles for 17 Pilocytic
and 10 Diffuse astrocytomas
Differentially methylated region
(DMR) identification
Delta beta values obtained by
comparison to foetal controls.
DMR
>0.3 delta beta value
>3 consecutive CPG sites
DNA methylation profiling of low grade astrocytomas using the illumina 450K BeadChip reveals
changes in genes involved in brain development and drug resistance Jennie N Jeyapalan1, Isabel CF Morley1, Alfred A Hill1, Ruth G Tatevossian2, Ibrahim Qaddoumi2, David W Ellison2*, Denise Sheer1*
1Barts & The London School of Medicine & Dentistry, Queen Mary University of London, London E1 2AT. 2St Jude Children’s Research Hospital, Memphis, TN38105-3678. *Corresponding Authors
Clustering according to the differentially methylated regions (DMRs)
showed the tumours grouped according to location. The paediatric tumours did not show the IDH1-methylator phenotype found in adult grade II
astrocytomas (9).
Figure 2. Hierarchical clustering of methylation profiles.
CpG sites were selected from genes that had DMRs including
>3 consecutive CpG sites, with a delta beta value >0.3 in
comparison to foetal control. The heatmap indicates sites of
low (green) to high (red) methylation.
The samples are shown as: Control, foetal brain controls; AB,
adult brain control; NSC, human neural stem cells; PA,
pilocytic astrocytoma; DA, diffuse astrocytomas.
Tumour features are shown as: Tumour type, Pilocytic
astrocytoma - grey, Diffuse astrocytoma – black, Controls -
white; Tumour location, infratentorial – grey, supratentorial -
black; Age,, <5 years – light grey, 5-13 years – dark grey, >13
years - black; Sex, female - grey, male - black; BRAF status,
BRAF/RAF fusion - grey, BRAF V600E - black; MYB status,
darker shade - higher MYB expression (7).
The overall methylation profiles of pilocytic and diffuse astrocytomas are similar,
although there are significant differences at specific genes.
Results
B
Figure 1. Differentially methylated regions in low grade astrocytomas
A) Density plots using beta values to show correlation between a) pilocytic
astrocytomas vs. normal cerebellum, b) diffuse astrocytomas vs. normal frontal lobe,
c) infratentorial pilocytic vs. diffuse astrocytomas, d) supratentorial pilocytic vs.
diffuse astrocytomas.
B) Manhattan plot for association between methylation and tumour type. CpG sites
are plotted across the chromosome, with sites showing significant difference above
the dotted line (Benjamini-Hochberg corrected p-values <0.05).
A a
b
c
d
Genes involved in development show differential DNA methylation,
e.g. EMX2OS, EMX2, PRKCZ, HES3, PAX6 and HOX gene family.
Figure 4. Differential methylation in the EMX2 gene and the associated lncRNA EMX2OS is
associated with down-regulation of both genes in pilocytic astrocytomas. EMX2, a
transcription factor expressed in the embryonic intermediate mesoderm and CNS, is essential for
patterning in the neocortex (13,14). EMX2 is regulated by the antisense transcript EMXOS.
A) Diagrammatic representation of EMX2OS and EMX2 transcripts.
B) Methylation plot showing DMR. The rectangular blocks represent CpG sites showing low or high
methylation.
C) Scatter graphs showing expression levels for EMX2 and EMX2OS measured by Affymetrix U133
plus 2.0 arrays
B
A EMX2OS EMX2
Foetal control
Adult control
Pilocytic astrocytomas
Diffuse astrocytomas
Gene body 5’UTR Gene body 3’UTR
C
EMX2
EMX2OS
Adult control
Foetal control
Pilocytic astrocytomas
Diffuse astrocytomas
Log t
ransfo
rmed n
orm
aliz
ed v
alu
es
Log t
ransfo
rmed n
orm
aliz
ed v
alu
es
References
1. Horner MJ et al (2009) SEER Cancer Statistics Review 1975-2006, National Cancer Institute. Bethesda, MD, http://seer.cancer.gov/csr/1975_2006/
2. CBTRUS. Statistical Report: Primary Brain Tumors in the United States, 2000–2004. Published by the Central Brain Tumor Registry of the United States. 2008.
3. Louis DN et al (2007) The 2007 WHO classification of tumours of the central nervous system. Acta Neuropathol. 114(2):97-109.
4. Qaddoumi I et al (2009) Outcome and prognostic features in pediatric gliomas: a review of 6212 cases from the surveillance, epidemiology, and end results database. Cancer.115: 5761–5770.
5. Forshew T et al (2009) Activation of the ERK/MAPK pathway: a signature genetic defect in posterior fossa pilocytic astrocytomas. J Pathol. 218(2):172-81.
6. Jones DT et al (2012) MAPK pathway activation in pilocytic astrocytoma. Cell Mol Life Sci. 69(11):1799-811.
7. Tatevossian RG et al (2010) MYB upregulation and genetic aberrations in a subset of pediatric low-grade gliomas. Acta Neuropathol 120(6):731-43.
8. Sandoval J et al (2011) Validation of a DNA methylation microarray for 450,000 CpG sites in the human genome. Epigenetics 6(6):692-702.
9. Turcan S et al (2012) IDH1 is sufficient to establish the gliomas hypermethylator phenotype. Nature. 483(7390):479-83.
10. Sharma MK et al (2007) Distinct genetic signatures among pilocytic astrocytomas relate to the brain region origin. Cancer Res. 67(3):890-900.
11. Liu HK, et al (2008) The nuclear receptor tailless is required for neurogenesis in the adult subventricular zone. Genes Dev. 22:2473–2478.
12. Miyawaki T, et al. (2004) Tlx, an orphan nuclear receptor, regulates cell numbers and astrocyte development in the developing retina. J. Neurosci.24:8124–8134.
13. Spigoni G et al (2010) Regulation of Emx2 Expression by Antisense Transcripts in Murine Cortico-Cerebral Precursors. PLoS One 5(1):e8658.
14. Hamasaki T et al (2004) EMX2 regulates sizes and positioning of the primary sensory and motor areas in neocortex by direct specification of cortical progenitors. Neuron 43:359-372.
15. Scortegagna M et al (2009) Hypoxia-inducible factor 1 alpha suppresses squamous carcinogenic progression and epithelial-mesenchymal transition. Cancer Res. 69(6):2638-46.
16. Silvers AL et al (2010) Decreased selenium-binding protein 1 in esophageal adenocarcinoma results from posttranscriptional and epigenetic regulation and affects chemosensitivity. Clin Cancer Res.
16(7):2009-21.
17. Huang KC (2006) Selenium binding protein-1 in ovarian cancer. Int J Cancer. 118(10):2433-40.
18. Fischer H et al (2008) Fibroblast growth factor receptor-mediated signals contribute to the malignant phenotype of non-small lung cell cancer cells: therapeutic implications and synergism with
epidermal growth factor receptor inhibition. Mol Cancer Ther. 7: 3408-3419.
• Overall methylation profiles of pilocytic and diffuse astrocytomas suggest that these tumours arise from a similar cell type
• Significant differences between pilocytic and diffuse astrocytomas are found in genes involved in CNS development. Developmental processes may thus contribute to the phenotypic disparity between these tumour types
• Alterations of genes involved in drug resistance should be investigated further as a possible means to improve therapeutic efficacy
• Taken together with the recently discovered genetic abnormalities, these epigenetic changes provide a framework for understanding mechanisms of gene deregulation in low grade astrocytomas
Summary DNA methylation is a crucial regulator of gene expression The genome-wide DNA methylation profiles of a large cohort of paediatric low grade
astrocytomas have been analysed While the majority of genomic sites show similar levels of methylation in pilocytic and
diffuse astrocytomas, significant differences are present at specific genes Hierarchical clustering according to the differentially methylated regions showed the
tumours grouped according to location in the brain Genes involved in development and drug resistance show differential methylation,
providing new avenues for understanding the biology of low grade astrocytomas
Genes involved in drug resistance show differential DNA methylation,
e.g. SELENBP1, RUNX1 and TNFRSF1A
SELENBP1
B
C
Gene Body 5’UTR TSS200 TSS1500
A
Foetal control
Adult control
Pilocytic astrocytomas
Diffuse astrocytomas
Figure 5. Differential methylation in the transcription start site in
SELENBP1 is associated with lower expression in pilocytic astrocytomas.
Down-regulation of SELENBP1 occurs in many cancers and which has been
shown to lead to chemoresistance in lung cancer.
A) Methylation plot showing methylated TSS in pilocytic astrocytomas against
foetal and adult controls. The circles represent CpG sites showing low or
high methylation.
B) Scatter graphs showing the methylation status for tumours and controls.
C) Scatter graphs showing SELENBP1 expression.
Adult control
Foetal control
Pilocytic astrocytomas
Diffuse astrocytomas
Be
ta v
alu
es
Be
ta v
alu
es
Chromosome location
Chromosome location
Pilocytic astrocytomas
Adult control
Foetal control
Normal human astrocytes
Diffuse astrocytomas
Adult control
Foetal control
Normal human astrocytes
Log t
ransfo
rmed n
orm
aliz
ed v
alu
es
Acknowledgements
American Lebanese Syrian Associated Charities
Figure 3. Hierarchical clustering of methylation profiles of location specific genes identifies methylation of NR2E1 gene in supratentorial tumours.
A) The heatmap indicates sites of low (green) to high (red) methylation in the above six genes. Tumour features are shown as: Tumour type, Pilocytic astrocytoma - grey,
Diffuse astrocytoma – black, Controls - white; Tumour location, infratentorial – grey, supratentorial – black.
B) NR2E1 gene body methylation. The methylation plots indicate the gene, with the arrow showing transcription start site and level of activity is represented by the size of the
arrowhead. The circles represent CpG sites showing low or high methylation. The red box indicates the region of differential methylation.
NR2E1
5’UTR Gene Body
Infratentorial pilocytic
astrocytomas
Supratentorial pilocytic
astrocytomas
Supratentorial Diffuse
astrocytomas
B
NR2E1, a gene involved in forebrain development, shows differential DNA
methylation within the gene body in supratentorial tumours, which correlates with
enhanced expression of the gene. Expression profiles of six genes, NR2E1, LHX2, SIX3, IRX3, IRX5 and PAX3, have been shown to be
sufficient to distinguish supra- and infratentorial pilocytic astrocytomas (10). We find that of this group, only
NR2E1 is differentially methylated. NR2E1 encodes a transcription factor that is crucial for development of
the brain and retina (11,12).
A
Tumour type
Location
Top Related