Cancers develop from the accumulation of somatic mutations, yet it remains unclear how oncogenic lesions cooperate to drive cancer progression. Using a mouse model harboring NRasG12D and EZH2 mutations that recapitulates leukemic progression, we employ single-cell transcriptomic profiling to map cellular composition and gene expression alterations in healthy or diseased bone marrows during leukemogenesis. At cellular level, NRasG12D induces myeloid lineage-biased differentiation and EZH2-deficiency impairs myeloid cell maturation, whereas they cooperate to promote myeloid neoplasms with dysregulated transcriptional programs. At gene level, NRasG12D and EZH2-deficiency independently and synergistically deregulate gene expression. We integrate results from histopathology, leukemia repopulation, and leukemia-initiating cell assays to validate transcriptome-based cellular profiles. We use this resource to relate developmental hierarchies to leukemia phenotypes, evaluate oncogenic cooperation at single-cell and single-gene levels, and identify GEM as a regulator of leukemia-initiating cells. Our studies establish an integrative approach to deconvolute cancer evolution at single-cell resolution in vivo.
Keyphrases
- single cell
- acute myeloid leukemia
- gene expression
- rna seq
- bone marrow
- high throughput
- transcription factor
- copy number
- papillary thyroid
- genome wide
- dna methylation
- genome wide identification
- mouse model
- long non coding rna
- long noncoding rna
- squamous cell
- dendritic cells
- stem cells
- public health
- cell cycle arrest
- cell proliferation
- soft tissue
- smoking cessation