Researchers at Kumamoto University have demonstrated that lysine-specific demethylase 1 (LSD1) shapes the variation in metabolic phenotypes seen in distinct acute myeloid leukemia cell types. Their approach integrates epigenomic and transcriptomic data to reveal novel epigenetic mechanisms underlying metabolic heterogeneity in leukemia, highlighting a potential therapeutic strategy.
Metabolic heterogeneity in acute myeloid leukemia
Acute myeloid leukemia (AML) is a form of blood cancer caused by haematopoietic malignancies in the bone marrow. Different AML subtypes arise from genetic and epigenetic mutations at various stages of haematopoietic stem cell differentiation, or lineage identities. Variations in metabolic features are known to exist among AML subtype. This may explain the drastic variability in the success of therapy between subtypes. Therefore, targeting subtype-specific metabolic phenotypes may represent a viable strategy for the development of new and highly efficacious AML therapies.
Previous research has suggested that AML metabolic phenotypes depend on lineage identity. However, the regulatory mechanisms underlying this metabolic heterogeneity have been elusive. Depletion of LSD1 was recently shown to impair haematopoietic lineage differentiation, implicating a significant role for LSD1 in human leukemogenesis. LSD1 demethylates histone H3 lysine 4 (H3K4) and is involved in the epigenetic regulation of several biological processes. Previously, the researchers behind this study have also shown that LSD1 regulates both metabolic genes and transcription factors involved in haematopoiesis.
Reduced LSD1 and glycolytic gene expression in acute myeloid leukemia
In this study, a team of researchers at Kumamoto University in Japan revealed the molecular role of LSD1 in determining the unique metabolic phenotypes of AML subtypes. They specifically investigated erythroid leukemia (EL), a subtype that develops from the cells that differentiate into red blood cells.
Analysing gene expression profiles of AML patients and cell culture lines revealed that LSD1 and glycolytic gene expression was high in EL. The researchers then generated doxycycline-inducible LSD1-knockdown EL cells and treated EL cell lines with selective LSD1 inhibitors. LSD1 inhibition reproducibly resulted in downregulation of glycolytic genes and reduced glucose uptake and glycolytic capacity in the cells. Therefore, this suggested that elevated glycolytic activity in EL is directly linked to higher LSD1. To elucidate the mechanisms by which this occurs, the researchers employed a multi-omics approach.
LSD1 regulates EL lineage transcription factors
RNA-seq of LSD1-inhibited EL cells revealed that LSD1 inhibition significantly downregulates genes involved in heme biosynthesis, an essential metabolic pathway in erythrocyte function. Markers of non-erythroid cells were also upregulated in LSD1-inhibited cells. Collectively, this suggested that LSD1 confers erythroid-specific features to EL.
Next, ChIP-seq analysis revealed that LSD1-bound sites in EL cells were enriched in consensus motifs for GATA family transcription factors. One member, GATA1, is erythroid-specific and is strongly expressed in EL cells. Functional analysis revealed that LSD1 prevents the proteasomal degradation of GATA1, which then promotes glycolysis and heme synthesis in EL cells.
The researchers also investigated the significant upregulation seen in CEBPA in LSD1-inhibited EL. CEBPA encodes for C/EBPα, a key transcription factor of the granulocyte-monocyte leukocytic lineage. Its knockdown in EL cell lines suggested that C/EBPα suppresses GATA1, which in turn inhibits glycolysis and heme synthesis and ultimately reverses the EL metabolic phenotype. ChIP-seq revealed that genomic regions enriched in LSD1 binding sites were located near mono- and di-methylated H3K4, markers associated with enhancer functions. LSD1 inhibition increased the levels of these enhancer markers near the CEBPA transcription start site. This suggests that LSD1 demethylates this region and suppresses C/EBP.
Combined, these results show that LSD1 facilitates GATA1 expression and suppresses C/EBPα in EL cells to promote erythroid-associated metabolic phenotypes. By balancing the expression of these lineage transcription factors, LSD1 therefore shapes the metabolic phenotype of EL cells.
LSD1 controls acute myeloid leukemia metabolic programming
Using publicly available transcriptome datasets of different AML subtypes, the researchers then gained a broader perspective on the role of LSD1 in AML metabolic phenotypes. They observed that expression of LSD1 is positively correlated to GATA1 and metabolic genes, and negatively correlated to CEBPA. Single-cell RNA-seq data from AML patients additionally revealed that LSD1 correlates with lineage and metabolic gene expression in both normal and malignant cell populations. Altogether, the results suggested that by regulating lineage transcription factors, LSD1 determines the unique metabolic features of different AML subtypes.
Altogether, this study elucidated the molecular mechanism by which LSD1 determines the metabolic phenotype of EL. Integrating epigenomic and transcriptomic data revealed that LSD1 balances lineage-specific transcription factors to generate the metabolic heterogeneity seen in AML. With this rationale, coupling LSD1 inhibition with lineage-specific metabolic may be a highly effective treatment for different forms of AML. Importantly, this study has highlighted the value of multi-omics in revealing novel molecular insights and therapeutic targets in cancer.
Image credit: FreePik