A Translational 3D Assay Mimicking the Bone Marrow Microenvironment for Drug Screening and Mechanistic Studies of Hypoxia Induced Resistance in Acute Myeloid Leukemia

Abstract

Acute Myeloid Leukemia (AML) is a complex disease that despite incredibly toxic treatments continues to have an overall 5-year survival rate of 25% in adults and 65% in children. It is important to study the mechanisms of resistance to continue developing effective anticancer agents and provide an improved prognosis for AML. One of the initial steps towards the development of novel anticancer agents is lab bench cancer research that better recapitulates the microenvironment of the suspension malignancy, AML. Typically, a recognized novel approach to study anticancer agents is in vitro high-throughput drug screening (HTS). However, HTS has been designed and performed utilizing traditional two-dimensional (2D) cell culture in which AML cells are grown in suspension on flat plastic surfaces with atmospheric oxygen (21%). These laboratory conditions although allowing for HTS, lack many components of the bone marrow microenvironment (BMM). The BMM is a 3D structure that contains a complex set of cellular, chemical, structural, and physical cues that maintain the viability and function of the hematopoietic system. The BMM has been shown to play a critical role in the development and survival of AML cells and also contributes to drug resistance and relapse. Current advances have led to the development of 3D cell culture systems using Matrigel® in 24- and 48-well culture systems (Parikh et. al, 2014), as well as an in vitro HTS, to measure of novel anticancer agents. Preliminary studies in the Institute of Molecular Medicine Laboratory at Phoenix Children’s Hospital Department of Child Health and the University of Arizona College of Medicine-Phoenix showed that AML cell lines grown in 2D under hypoxia developed resistance to the survivin inhibitor YM155 in 6-, 12-, 24 and 96-well format. This effect was increased in a 3D Matrigel culture. Furthermore, the 3D hypoxia condition altered the response of AML cell lines to both Dovitinib and cytarabine. In the present thesis, we have confirmed these results using a scale-up to 384-well plate format and we used this format to study its suitability for HTS of a small library of 49 drugs in 3D normoxia. This 3D assay was used to mimic BMM and screen for drug responses in AML. Interestingly, we observed: irrespective of the two oxygen conditions optimal cell growth and viability occurred with an initial cell density of 24,000 per well over a time course culture of six days. Additionally, the adaptation of 3D CTG as an alternative to the CRS extraction technique was effective in measuring ATP as an indicator of cell viability. Additionally, in vitro studies were carried out in MV-4-11 AML cell lines in a 384-well format 3D functional screen in normoxia with a 49-compound library and found the drugs YM155 (survivin inhibitor), Cytarabine (Ara-C, antimetabolic standard of care chemotherapy agent) to have statistically significant (p < 0.05) lower RLU values than DMSO, thus confirming previous studies in the lab using 96-well format. Next, we selected YM155, Ara-C and Dovitinib (multi-kinase inhibitor) to further interrogate their therapeutic effect in 96-well single-dose treatment and 384-well drug dose-response with conditions such as dimensionality (2D vs. 3D) and oxygen levels (normoxia vs. hypoxia). Intriguingly, both the 96- and 384-well drug studies reproduced the observation of hypoxia and 3D induced drug resistance to MV-4-11 cells treated with pro-apoptotic drug YM155, suggesting that our 3D scale-up assay assessed altering drug response profiles from 2D and 21% O2 cell cultures and a potential mechanistic pathway worth further interrogating. We utilized flow cytometry to distinguish the stages of cell death in YM155-mediated apoptosis in both oxygen conditions and found no significant cell arrest in hypoxia despite a 5-fold increased cytotoxic dose. Next, qPCR/western blot studies around a set of previously identified deregulated genes and proteins in AML, found hypoxia-induced resistance to YM155 treatment may be influenced by the upregulation of growth factor HIF family to increase the expression of anti-apoptotic protein MCL-1, suggesting pro-survival, thus confirming preliminary studies in the lab that demonstrated upregulation of MCL-1 upon treatment of MV-4-11 cells in 3D hypoxia conditions but not in 2D. There may exist a ‘rescue’ pathway as a mechanism by which constitutive MCL-1 signaling in MV-4-11 cells cultured in hypoxia may help protect specific YM155 targets resulting in a treatment-resistant phenotype. The stabilization of MCL-1 may be the “rescue pathway” involved in hypoxia-induced resistance to apoptosis-mediated drug treatment. Collectively our results indicate that MCL-1 signaling may be an important factor mediating YM155 response in AML. Mono- or combinational treatment inhibitors for transcription factors (such as HIF) could be a future direction for targeted therapy.