Abstract
Introduction
While the majority of patients diagnosed with acute myeloid leukemia (AML) will respond to standard induction chemotherapy, recurrent and refractory disease rates are high. Much needed new therapies for AML are currently in development; however, many drugs indiscriminately target rapidly proliferating cells and are ineffective against indolent, more quiescent leukemic stem cells. To identify novel therapies specific for AML blasts and leukemic stem cells (LSC), we developed a high-throughput, multi-parametric flowcytometry(FC) based method that simultaneously distinguishes specific cell populations and their differential drug responses. Using this method, we could distinguish healthy cells from leukemic stem cells and assess the effects of several drugs on these cell populations as well as the immune profile of individual patients. In addition, we identified drugs that could target both blasts and LSC and could potentially be repositioned for the treatment of AML.
Methods
Mononuclear cells were enriched from bone marrow (BM) aspirates collected from 10 patients by density gradient centrifugation and seeded to pre-drugged 96-well plates at 100,000 cells/well. The drugs were tested in a 10,000-fold concentration range and included standard chemotherapy agents (cytarabineandidarubicin), corticoid steroid dexamethasone,rapalogtemosirolimus, PI3K/mTORinhibitoromipalisib, JAK1/2 inhibitorruxolitinib,Srckinase inhibitordasatiniband Bcl-2 inhibitorvenetoclax. The cells were incubated with the drugs for 3 days in RPMI medium supplemented with conditioned medium from the human BM stromal cell line HS-5. The cells were stained using six different antibodies to identify blasts (CD45+, CD34+/CD33+, CD123+), LSC (CD45+, CD34+, CD38-, CD96+), healthy stem cells (HSC) (CD45+, CD34+ CD38-, CD96-), normal BM cells (CD45+, CD34-) and lymphocytes (CD45+ high, CD34-/CD33-, SSH low). We usedAnnexinV and 7-AAD to detect and remove dead and apoptotic cells. The stained cells were measured within the 96-well plates using theiQueScreener PLUS instrument (Intellicyt) in 20 minutes and at least 50,000 events acquired per well. Live cell numbers for different cell populations were calculated and dose response curves generated for each population and drug. For comparison, the viability of the samples was also assessed using theCellTiterGlo(CTG) assay.
Results
Assay set up and plate reading were accomplished within a relatively short time (staining 1 h, plate reading 20 min) with more than 5 million cell events acquired from each plate. Dose response curves generated from the live cell counts of the FC assay and viability readouts of the CTG assay were highly similar for cytotoxic drugs.(Fig. 1A).However, for targeted therapies, CTG overestimated the effect of JAK andmTORinhibitors by approximately 20% compared to the FC readouts (Fig. 1B). High variation in drug responses was observed between patient samples indicating patient and sample heterogeneity. In addition, cell populations within the same patient samples showed very different responses to the tested drugs. For example lymphocytes were sensitive tovenetoclax, but unresponsive to the other tested drugs. In contrast, blast cells were sensitive to most drugs and especially tovenetoclaxandruxolitinibwhen compared to other cell populations, suggesting that these drugs are particularly effective against this rapidly proliferating cell population.(Fig. 1C).Interestingly,ruxolitinibwas also effective at targeting LSCs, although HSCs were equally sensitive to this drug.
Conclusion
By applying a high-throughput FC based method, we could easily assess the effects of several drugs on AML patient samples and distinguish the functional impact on different cell populations. The method differentiated leukemic and healthy stem cells and identified drugs that could target LSC, which are believed to be a main cause of relapses in AML. Interestingly, JAK1/2 inhibitorruxolitinib, approved for the treatment ofmyelofibrosisand polycythemiavera, was effective at targeting AML blasts and LSC. In addition, this method accounts for inter- and intra-patient variability, and may be useful for a precision medicine based approach to identify optimal, patient-specific treatment strategies.
Heckman:Celgene: Research Funding; Pfizer: Research Funding.
