Determining the Mechanism of 5-Azacytidine Resistance in Myeloid Malignancies Using Single-cell DNA Methylation Sequencing Paired With Flow Cytometry
K O'Neill1,4, T Hui2, D Deng4, J Parker4, M Moksa3, M Hirst3,4, A Karsan1,4
1Pathology Department, 2Genome Science and Technology Program, 3Department of Microbiology & Immunology, University of British Columbia, 4Michael Smith Genome Sciences Centre, BC Cancer Agency, Vancouver, Canada
Myeloid malignancies, especially myelodysplastic syndromes (MDS) are often treated with the DNA methyltransferase inhibitor, 5-azacytidine (5-Aza). But only half of those patients respond, and resistance typically develops within two years, followed by rapid disease progression and death  . Recent evidence suggests 5-Aza resistance originates in rare MDS propagating cells (MDS-PC) within hematopoeitic progenitors . However, the exact phenotype of these cells is unclear  , and there has been limited research into their role in resistance.
We have index-sorted and sequenced 192 CD34+CD38-Lin- single cells each from bone marrow samples before and after 5-Aza treatment from a patient with relapse after initial response, and performed whole-genome bisulfite sequencing (SC-WGBS) of each cell. We observed that while pre-treated cells have similar global methylation (mean=76.5%, SD=1.2%), 30% of the post-treated cells were globally hypomethylated. This provides the first SC-WGBS study of 5-Aza response in a clinical context, and contrasts with the much more consistent 5-Aza hypomethylation response seen in vitro in SC-WGBS studies of K562 .
We were also able to resolve single cell copy number alterations to 5Mbp resolution, subclonally detecting a trisomy 9 only present in the post-treatment sample, and previously reported from clinical metaphase spreads.
Neither the trisomy 9 nor the clustering of cells into hypo- and normally methylated could have been detected in bulk data. This work validates the use of single-cell DNA methylation analysis for investigating response to 5-Aza therapy. Future work includes sequencing more patients pre and post 5-Aza treatment, and more in-depth analysis of differential methylation, intercellular heterogeneity, and the role (if any) of subclonal copy number alterations.
1. Bejar, Steensma (2014) Blood 124:2793–2803.
2. Zeidan et al. (2014) British Journal of Haematology 166:352–359.