The main challenges in treating acute promyelocytic leukemia (APL) are currently early mortality, relapse, refractory disease after induction therapy, and drug resistance to ATRA and ATO. In this study, a computational chemogenomics approach was used to identify new molecular targets and drugs for APL treatment. The transcriptional profiles induced by APL were compared with those induced by genetic or chemical perturbations. The genes that can reverse the transcriptional profiles induced by APL when perturbed were considered to be potential therapeutic targets for APL. Drugs targeting these genes or proteins are predicted to be able to treat APL if they can reverse the APL-induced transcriptional profiles. To improve the target identification accuracy of the above correlation method, we plotted the functional protein association networks of the predicted targets by STRING. The results determined PML, RARA, SPI1, HDAC3, CEBPA, NPM1, ABL1, BCR, PTEN, FOS, PDGFRB, FGFR1, NUP98, AFF1, and MEIS1 to be top candidates. Interestingly, the functions of PML, RARA, HDAC3, CEBPA, NPM1, ABL, and BCR in APL have been previously reported in the literature. This is the first chemogenomics analysis predicting potential APL drug targets, and the findings could be used to guide the design of new drugs targeting refractory and recurrent APL.
Keywords: Acute promyelocytic leukemia; Chemogenomics; Drug target; Transcriptional profile.