Prediction of drug sensitivity in AML subtypes in response to Mitoxantrone from global gene expression, an impact on personalized cancer chemotherapy

Introduction: Acute myelogenous leukemia (AML) is the most common acute leukemia affecting adults, and its incidence increases with age. A combination treatment with an anthracycline, combined with cytarabine (ara-C), has been the cornerstone of AML treatment since the 1960s. The goal of the study: In line with personalized cancer chemotherapy, in the current study, using computational methods, possible drug sensitivity of AML subtypes to Mitoxantrone was predicted through gene expression and phylogenic analysis. Moreover, biological pathways related to drug resistance of AML cells to Mitoxantrone were investigated Methods: A global gene expression dataset from cells sensitive to Mitoxantrone and cells with resistance to this drug with the accession number of GSE48876 was obtained from the gene expression omnibus (GEO) database. Expression data of AML subtypes were also obtained from GEO with the accession number of GSE59808. The CEL files were normalized and summarized with RMA method. Differential gene expression analyses for genes in sensitive and resistance cells were performed using linear regression models in the Limma R package. For the phylogenetic tree construction, we only considered genes that show significant differences (determined with a one-way ANOVA) in their expression profiles. To perform the prediction of drug sensitivity in AML subtypes, we included expression data of sensitive cell as an out group for the phylogenetic tree construction. Pairwise distances of the cancer subtypes were computed with the Pearson Correlation Distance. The phylogenetic tree was constructed with Fitch-Margoliash methodand visualized with Dendroscope. Results and Discussion: The results of phylogenetic studies showed that M0 and M7 subtypes of AML were placed closer to sensitive cell that others. With the Pearson Correlation Distance of zero from sensitive cells, we cloud consider M0 and M7 subtypes as sensitive types to Mitoxantrone. On the other hand, the phylogeny ranks the AML subtypes according to their dissimilarity from sensitive cell as M6, M1, M2, M3, M4 and M5. Moreover, the study showed that Insulin and Neurotrophin signaling pathways were the most important pathways were significantly affected by differential expressed genes in AML resistance cells to Mitoxantrone. Conclusion: The results of this study could help prediction of the associated drug response, which could improve the efficiency of chemotherapy, and may improve cancer treatment.

Keywords:

Personalized cancer chemotherapy , Gene expression , AML subtypes , Drug sensitivity , Mitoxantrone