A New computational approach for drug repurposing identified novel compounds for differentiation therapy in acute myeloid leukaemia

Repurposing of drugs for new therapeutic use is a scheme that has received considerable attention as it can reduce time and cost of drug development. In my thesis I developed a new strategy to identify compounds able to induce differentiation in leukaemic cells. Since Acute Myeloid Leukaemia (AML) results from a block in myeloid differentiation, finding new drugs capable of inducing a terminal maturation of blast cells is considered a worthwhile strategy. This conclusion has been reinforced as clinical trials of All-Trans Retinoic Acid (ATRA) treatments have shown remarkable efficacy in ameliorating the condition of Acute Promyelocytic Leukaemia patients. We applied a new computational strategy to identify compounds approved by the Food and Drug Administration (FDA) that promote myeloid differentiation. To this end I took advantage of the Connectivity Map, a large reference catalogue of gene expression data from cultured human cancer cells perturbed with many FDA approved compounds (Lamb et al., 2006). Starting from the gene expression profile of HL-60 (a human promyelocytic cell line) perturbed by approximately 1000 drugs, we characterized the modulation of the transcription profiles after incubation with each drug. By listing the genes that are significantly up- or down-regulated we could infer the transcription factors that are activated or inactivated by the drugs. Compounds that promoted the activation of transcription factors that play a positive role in myeloid differentiation were considered candidate pro-differentiation drugs. This approach yielded a list of chemicals ranked according to the potential of inducing differentiation in leukaemic progenitor cells. Well known prodifferentiation drugs such as for instance all trans retinoic acid are in the top positions of this ranked list. To validate our strategy, I tested the in vitro differentiation potential of 22 candidate compounds using the HL-60 human cell line as a myeloid differentiation model. These cells are classified as promyelocytes and can be induced to differentiate into granulocyte-like cells in vitro. The differentiation analysis identified 10 compounds, out of the 22 in the top position of the ranked list, that are able to induce a significant differentiation of HL-60. Some of these compounds cause DNA damage, a mechanism that was already implicated in myeloid differentiation. These results underscore the potential of our approach to accelerate the drug discovery process and could identify novel uses of existing drugs reducing experimental time and costs.