Expression Signature of Therapy Resistant AML: Use of In Vitro Selected Ara-C Resistant BXH-2 Strain AML Cell Lines.

Resistance to Ara-C, a deoxynucleoside analogue used to treat acute myeloid leukemia (AML) plays an important role in AML treatment failure and disease recurrence. Two different murine AML cell lines (B117P and B140P) derived from BXH2 mice were used to create highly Ara-C resistant cell lines (B117H and B140H) with consecutive in vitro selection. We performed oligonucleotide microarray analysis (430A GeneChip, Affymetrix) on triplicate samples of each of the four lines. Significance analysis of microarray, principal components analysis, hierarchical clustering, t-test and fold change analysis were performed (SAM and GeneData Expressionist Analyst) to identify an evaluable list of 555 genes that differed by 2-fold or greater with a p-value of 0.001 or less. Quantitative RT-PCR to confirm expression levels for the most relevant genes is in progress. Genes from several potentially interesting and relevant pathways (Ara-C metabolism, asparagine biosynthesis, cholesterol metabolism, DNA damage repair, and apoptosis) were identified. Nt5e and Nt5c2, involved in nucleotide catabolism and transport, were up-regulated in Ara-C resistant cells. Gstm1 and Gstt2, suspected to be involved in chemotherapy resistance, were also up-regulated in resistant cells. Interestingly, the gene asparagine synthetase (Asns) and several others involved in asparagine and glutamine metabolism (Nars, Asph, Got1, Glrp1) were altered in Ara-C resistant cells. L-Asparaginase, a drug commonly used to treat ALL, has previously been reported to be effective in treatment of relapsed or refractory AML. A number of genes involved in cholesterol and fatty acid metabolism were altered in resistant cells (Fdps, Acadl, Elovla6, Hdlbp, Hsd17b1). Recently farnesyl transferase and cholesterol biosynthesis inhibitors have shown activity in myeloid malignancies. The known inhibitors of apoptosis, Bcl2 and Bclxl, were up-regulated in Ara-C resistant cells. Additionally, a number of genes involved in the regulation of stress reponse were altered (Dnajc3, Dnajb10). Many genes involved in repair of DNA damage were down-regulated in resistant cells (Ddb1, Msh2, Msh6, Xpc, Xrcc5). Other genes known to be invovled in the pathogenesis of human AML and other malignancies (Kit, Kit/Scf, Cmet) were over-expressed, suggesting an additional role in chemoresistance. One of the most dramatically and significantly up-regulated genes was the amyloid beta (A4) precursor protein gene (App), which has previously been described in relation to its role in Alzheimer’s disease, although there has been another recent study describing its over-expression in AML with complex karyotype. In conclusion, we were able to clearly differentiate Ara-C sensitive from resistant murine AML cells with microarray analysis. Our data corroborate a number of genes well described in AML and chemotherapy resistance. We have also identified a number of novel genetic targets for future study in the therapeutic intervention of chemotherapy-resistant AML.