Epigenomics and cancer heterogeneity in Ewing sarcoma

Nathan Sheffield, PhD
www.databio.org/slides

Today's outline

Epigenomics and cancer heterogeneity in Ewing sarcoma

Ewing sarcoma

Rare, aggressive pediatric bone cancer
Small cell tumor
neuroblastoma, Wilms' tumor, lymphoma
desmoplastic small-round-cell tumor
Driven by a single, well-characterized oncogenic fusion transcription factor (EWS-FLI1)

Rosa and Shaw 2013. Biology

Epigenomics and cancer heterogeneity in Ewing sarcoma

Ewing sarcoma clinical heterogeneity

  • Ewing (Ewing 1921) vs. Peripheral primitive neuro-ectodermal tumor (pPNETs; Stout 1918)
    • "ES and pPNET are considered as two tumor phenotypes along a gradient of limited neuroglial maturation that arise from the same stem cell” (Kovar 2005)
  • osseous vs. extra-osseous (Tefft et al. 1969)
  • site of origin: Askin tumor of the chest wall (Askin et al. 1979)
Lawrence et al. 2013

RRBS dataset

  • 140 primary Ewing sarcoma tumors
  • 32 primary mesenchymal stem cell samples (3 sources: bone marrow, umbilical cord, placenta)
  • 16 Ewing sarcoma cell lines (6 derived from these tumors)

RRBS reference panel

  • Cancer comparison: 105 samples from 7 other cancer types
  • Cell-type diversity: 266 samples from 50+ tissues

Why RRBS?

Nice cost/benefit ratio
Single-nucleotide resolution

Acknowledgments

St. Anna's Child Cancer Research Institute, Vienna
Heinrich Kovar
Eleni Tomazou
Peter Ambros
Inge Ambros
Diana Walder
Paracelsus Medical University, Salzburg
Dirk Strunk
Katharina Schallmoser
Medical University of Graz
Beate Rinner
Bernadette Liegl-Atzwanger
Berthold Huppertz
Andreas Leithner
CeMM Research Center for Molecular Medicine, Vienna
Christoph Bock
Johanna Klughammer
Andreas Schönegger
Michael Schuster
Paul Datlinger
Johanna Hadler
Münster University Hospital
Uta Dirksen
Institute of Biomedicine of Seville, Spain
Ana T. Amaral
Enrique de Álava
Institut Curie, Paris
Olivier Delattre
Franck Tirode
Sandrine Grossetete
Funding
Human Frontier Science Program
Kapsch NGS Grant
FWF Lise Meitner Fellowship

Locus Overlap Analysis

http://code.databio.org/LOLA/
Sheffield and Bock (2016). Bioinformatics.
Nagraj, Magee, and Sheffield (2018). Nucleic Acids Research.

Proportion Intermediate Methylation

Conclusions
  • DNA methylation quantifies cancer heterogeneity
  • DNA methylation can be used to infer regulatory activity
  • EWS-FLI suppresses mesenchymal regulatory activity
  • Mesenchymal regulatory elements are more suppressed in STAG2-mutants
  • More heterogeneity (higher PIM) signals cancer progression

Questions?

nsheffield@virginia.edu
www.databio.org