Nutrient-Sensitising Drug Repurposing Screen Identifies Lomerizine As a Mitochondrial Metabolism Inhibitor in Chronic Myeloid Leukaemia

In chronic myeloid leukaemia (CML), the persistence of leukaemic stem cells (LSCs) following first line treatment with tyrosine kinase inhibitors (TKIs) such as imatinib, can lead to disease relapse. We have previously demonstrated that therapy-resistant CML LSCs rely on oxidative phosphorylation (OXPHOS) for survival and that targeting mitochondrial metabolism sensitises CML LSCs to imatinib treatment. However, current OXPHOS inhibitors have demonstrated limited efficacy, or are associated with adverse effects in clinical trials, highlighting that identification of clinically safe oxidative pathway inhibitors is warranted.

To identify clinically applicable compounds that target cells dependent on mitochondrial metabolism, we used a nutrient-sensitised repurposing screening strategy. Upon replacement of glucose with galactose as the only sugar source in cell culture media, there is a shift in cellular metabolism towards OXPHOS (as cells are unable to effectively use galactose for glycolysis). Therefore, CML cells grown in galactose-containing media are more sensitive to inhibition of mitochondrial respiration than glucose-grown cells. Using this approach, we screened total of >1,100 unique and structurally diverse FDA-approved drugs in CML cells, cultured in the presence of 11mM glucose or 11mM galactose. This uncovered lomerizine dihydrochloride, a L-type Ca2+ channel blocker used as an anti-migraine drug, as a candidate compound effective against OXPHOS-dependent CML cells.

Subsequently, patient derived CML CD34+ cells were shown to have increased ER mass (the main intracellular Ca2+ store) and an increase in mitochondrial Ca2+ levels compared with normal cells. Moreover, transcriptional analysis revealed that Ca2+ channel genes such as voltage-gated CACNA1D and receptor-activated TRPC6 are significantly upregulated in CML LSCs (CD34+CD38-) when compared with normal haematopoietic stem cells (HSCs), which was not reverted following imatinib treatment. Therefore, we hypothesised that the observed increase in mitochondrial metabolism in CML LSCs was dependent on CACNA1D and TRPC6 activity, and the increase in mitochondrial Ca2+ levels, which could be inhibited by lomerizine treatment. Notably, CRISPR-Cas9-mediated CACNA1D or TRPC6 knockout significantly reduced cytoplasmic and mitochondrial Ca2+ levels, and decreased cellular oxygen consumption rate, indicative of reduction in OXPHOS. Similarly, while imatinib treatment did not affect intracellular Ca2+ levels in primary CML CD34+ cells, lomerizine treatment significantly reduced cytoplasmic and mitochondrial Ca2+ levels, and mitochondrial respiration/OXPHOS.

Applying liquid chromatography-mass spectrometry and isotope assisted metabolomics, we traced labelled nutrients (glucose and palmitate) in CD34+ cells. This revealed a significant reduction in activity of mitochondria-located Ca2+ dependent dehydrogenases and overall reduction in tricarboxylic acid (TCA) cycle activity in primary CML cells following lomerizine treatment.

Of clinical relevance, using combination of cell surface markers, cell division trackers and functional assays, including long-term culture initiating cell (LTC-IC) assay, we demonstrate that lomerizine treatment, alone or in combination with imatinib, targeted CML LSCs when cultured in physiological medium in vitro, without affecting normal HSCs. Notably, combination treatment with imatinib and lomerizine reduced CML tumour burden, targeted CML LSCs and extended survival in xenotransplantation model of human CML.

In summary, we reveal for the first time that Ca2+ regulation is disrupted in CML stem cells and influx via CACNA1D and TRPC6 is crucial for maintaining high mitochondrial Ca2+ levels and enhanced mitochondrial OXPHOS. This renders CML stem cells sensitive to lomerizine, an FDA-approved Ca2+ inhibitor identified following metabolism-specific drug-repurposing screening.