Development of an in vitro Relapse Model for Identification of Novel Therapeutics in Acute Myeloid Leukemia

Abstract

AML is a cancer of the blood and bone marrow characterized by the presence of highly proliferative and abnormally differentiated myeloblasts. Previous work from the Bhatia lab utilized the orthotopic xenograft model in order to isolate a population of leukemic regenerating cells (LRC) that exists prior to relapse. Affymatrix analysis of LRCs revealed up-regulation of 248 genes that can act as unique targets to prevent relapse. In order to screen compounds against all 248 targets, it is important to develop an in vitro model that is able to appropriately recapture the functional and molecular markers of LRCs. Primary AML samples were treated with 5-doses of 0.15 μM, 1 μM AraC, or DMSO control and several outcomes were measured. In vitro AraC treatment was not able to recapitulate the progenitor frequency curve and CD34 expression curve observed in vivo. Additionally, we were not able to see a consistent increase in select LRC targets DRD2, GLUT2, FUT3, and FASL via flow cytometry. Despite an increase in the mRNA levels of LRC genes 24h after treatment with 0.15 μM AraC, long term analysis could not be completed due to poor RNA quality and low expression of LRC-targets. Primary AML cells were co-culture with mouse MS-5 stromal cell line order to study the effects of mesenchymal stromal cells on AML response to AraC. Co-culture with MS-5 cells had different effects on select primary AML cells. AML 14939 showed an increase in CD34 and LRC targets DRD2 and FUT3 following AraC treatment when co-cultured with MS-5 cells; while A374 showed no differences between DMSO and AraC treated groups. Overall, these findings suggest the LRC signature is not induced by treatment with AraC alone. Complex interactions between AML cells and their bone marrow niche during AraC treatment plays an important role in the development of LRCs prior to AML relapse.