Metabolic profiling of myeloproliferative neoplasms reveals subtype specific metabolic traits

Introduction: Myeloproliferative neoplasms (MPN) are haematopoietic disorders leading to aberrant expansion of myeloid, erythroid, and megakaryocytic lineages, which are classified into three phenotypes namely Polycythaemia Vera (PV), Essential Thrombocythemia (ET) and Primary Myelofibrosis (PMF). The evolution of MPN mainly originates from three somatic mutations in haematopoietic stem cells (HSCs) including Janus Kinases2 (JAK2), Calreticulin (CALR), Myeloproliferative Leukemia Virus Oncogene (MPL). The JAK2 inhibitor Ruxolitinib, despite being the state-of-the- art therapy is unable to completely eradicate leukemic stem cells (LSCs). . Metabolic alterations are crucial in the development of cancer by creating different anaplerotic pathways supporting cancer cell proliferation. Metabolic vulnerabilities can be utilized to target MPN mutational hematopoietic stem cells.

Methods: We preformed Seahorse analysis on three independent biological replicates for Jak2^V617F wild type (WT), heterozygous (HET, represents ET) and homozygous (HOM, represent PV) c-kit positive cells and for even more primitive lineage negative, c-kit positive, Sca1 positive (LSK). We additionally used Liquid Chromatography with tandem mass spectrometry (LC-MS-MS) and RNA sequencing to investigate metabolic differences in LSCs.

Results: We see increased respiration in Jak2^V617F HET c-kit cells compared to WT and Jak2^V617F HOM cells. However, mitochondrial numbers as determined my mitotracker green flow cytometry are similar between all genotypes. Moreover, mitochondrial membrane potential, as measured my TRMT staining and flow cytometry, does not differ between genotypes in c-kit positive and in LSK cells. These data indicate that differences in respiration between genotypes is not due to changes in the mitochondrial compartment, but might instead be caused by changes in e.g. gene expression levels of respiratory enzymes or other players involved in cellular respiration. We therefore performed bulk RNA-sequencing on c-Kit positive and LSK cells. We found the glutamate receptor signalling pathway elevated in Jak2^V617F Hom, with single cell RNASeq data from highly purified Hom Jak2^V617F LSCs showing significant upregulation in glutathione metabolism, feeding into the same metabolic processes.

To elucidate changes in metabolism in more detail, we generated data from two independent biological replicates of c-kit positive, murine cells from all three genotypes for LC-MS. We found that Jak2^V617F Het cells show increased activity of the TCA cycle compared to Hom. This is in accordance with our Seahorse data showing increased respiration in Het compared to Hom. We therefore added an OxPhos inhibitor (IACS) to Het and Hom c-kit positive cells to characterise a possible vulnerability in the Het cells. Jak2^V617F Het c-kit positive cells seem indeed more sensitive to IACS compared to Hom cells, where IACS yields little effect. In contrast, higher glutamate levels were discovered medium of Jak2^V617F Hom cells after 24 hours of culture, indicating glutamate evacuation out of the cells.

Conclusions Our data demonstrate, for the first time, differences in metabolic needs in subtypes of MPNs, with ET-like murine cells depending on the TCA cycle compared to PV-like murine cells, which deregulate the glutamate/ glutathione metabolism pathway.

No relevant conflicts of interest to declare.