Transcriptional Heterogeneity Governs Cell Fate Diversification during Pre-Leukemia to Leukemia Progression

Acute myeloid leukemia (AML) is a heterogenous clonal disorder of hematopoietic progenitor cells with a dismal survival. It has a strong reliance on epigenetic and transcriptional factors for disease progression. Accordingly, we have previously identified KAT2A, a histone acetyl-transferase, as a requirement for AML maintenance; where chemical inhibition of KAT2A promotes differentiation of AML cell lines (Tzelepis et al., 2016, Cell Reports 17, 1193–1205). More recently, using a conditional knockout mouse model for Kat2a we showed that it sustains KMT2A/MLLT3 AML stem cells. Kat2a is a classical regulator of transcriptional variability, it’s loss leads to cell-to-cell heterogeneity in transcription levels specifically from genes involved in ribosomal biogenesis and translation (Domingues et al., 2020, eLife 9:e51754). No recurrent mutations in the KAT2A gene have been described in AML, and it is unclear if and how it participates in pre-leukemia-to-AML progression.

Herein, we use our conditional Kat2a knockout mouse model to analyze the effects of Kat2a loss in biology of RUNX1-RUNX1T1(9a) and Idh1R132H-initiated AML. These models represent forms of human disease with a prolonged pre-leukemia phase that typically require additional mutations for leukemia progression. We observed that loss of Kat2a accelerates leukemia initiation and progression in vivo. This acceleration was a consequence of fixation of transformed Kat2a KO cells in vivo which reflects as enhanced self-renewal capacity in vitro as measured by serial re-plating colony forming assay.

Given the central role of Kat2a in limiting cell-to-cell transcription heterogeneity, we interrogated a potential link between loss of Kat2a, its consequent increase in transcriptional heterogeneity and pre-leukemia progression. For this, we performed single-cell RNA sequencing (scRNA-seq) of early-stage Kat2a WT and Kat2a KO RUNX1-RUNX1T1(9a) pre-leukaemia. Compatible with our previous observation, we observed that Kat2a KO cells were more heterogenous transcriptionally. Interestingly, this was accompanied by diversification of cell fates towards B-lymphocytes and monocytes. Furthermore, pseudo-temporal ordering of single Kat2a KO cells revealed highly branched trajectory heavily populated with intermediate stages of transformation; including accumulation of leukemia progenitors with RUNX1-RUNX1T1 signature. In contrast, Kat2a WT cells have linear normal hematopoiesis trajectory with minimal branching and an abrupt transition towards candidate leukemia progenitor state.

Pathway analysis of Kat2a KO leukemia progenitor cells indicated perturbation of ribosomal biogenesis and translation associated genes. In order to test how these changes contributed to transformation, we performed S6K1 inhibition on Kat2a WT cells which transiently promoted transformation in vitro in both RUNX1-RUNX1T1(9a) and Idh1R132H cells, thus, phenocopying the effects of Kat2a loss. This suggested a mechanistic contribution of observed transcriptional changes in protein synthesis machinery towards leukemia progression.

Taken together, our work suggests that loss of Kat2a results in diversification of cell fates, including with increased accessibility to cell states prone to transformation. Furthermore, these cells, prone to transformation, may benefit from a low biosynthetic activity that promotes their progression to leukemia state. We hypothesize that Kat2a loss may function similarly in the context of other malignancies. In the future, this knowledge may aid in development of early diagnostic tools and suggest bespoke therapeutic interventions.