Most patients diagnosed with acute myeloid leukemia (AML) suffer from disease relapses within two years after the initial remission. Treatment options for relapsed AML are limited and the clinical outcome associated with this stage of disease is dismal. Partly due to the lack of experimental models, the biology underlying AML relapses is not well understood. We hypothesized that the cancer stem-like (CSL) cells, which represent a minority of the neoplastic cell population, are the major contributors of AML relapses. SORE6, a reporter designed to detect the expression and activity of inducible pluripotent stem cell factors, was used to identify/purify CSL cells in two FLT3-mutated AML cell lines, MOLM-13 and MV4-11. Both cell lines contained ~10% of SORE6+ cells at the steady state, as revealed by their expression of the reporter green fluorescence protein. Compared to SORE6- cells, SORE6+ cells had significantly more spheroid formation in the hanging drop assay (75 vs 35%, p=.006), and were significantly more resistant to Ara-C (inhibitory concentration at 50%, 60 vs 25 nM, p=.0005) and Venetoclax (2.6 vs 8.2 nM, p=.002) in the presence of low-dose Ara-C. By western blots, SORE6+ cells were found to express substantially higher levels of Myc/FLT3/pSTAT5 than SORE6- cells. Myc appeared to be a key driver of SORE6 activity, and the reporter activity was effectively suppressed by Myc pharmacologic inhibitors or shRNA. By subjecting a mixture of bar-coded SORE6- and SORE6+ MOLM-13 cells (in 9:1 ratio to mimic that at the steady state) to the lowest concentration of Ara-C that resulted in no detectable viability in the cell culture, we consistently found evidence of cell regeneration after 8-12 days of undetectable cell viability. Regeneration initially occurred slowly, and the cell growth went into a log phase after 1-2 weeks. Using PCR to detect the relative proportion of bar-coded SORE6- and SORE6+ cells, we found that the regenerated cell population contained a substantially higher proportion of SORE6+ cells, with 64% SORE6+ cells in the initial slow-growth phase. In comparison, the proportion of cells originating from SORE6+ cells increased to an average of 86% during the log-growth phase. Importantly, cells in the log-growth phase were found to have a significantly higher resistance to Ara-C and had significantly more spheroid formation in the hanging drop assay compared to untreated cells. To further validate the concept that CSL cells are the key contributors to the AML relapse, we collected pairs of initially diagnostic and relapsed bone marrow specimens derived from the same patients. By immunocytochemistry, we found that relapsed AML blasts were more frequently positive for Myc, a surrogate marker of SORE6 activity, compared to those in the pre-treatment samples. To conclude, using SORE6, we have developed a technically simple experimental model that capitulates some of the key features of AML relapses. Our model has provided direct evidence that CSL cells are key contributors to the disease relapse in AML. We believe that this in-vitro model is valuable in testing the efficacy of new therapeutic agents against AML, especially those designed to target the CSL cell population.

Brandwein:Abbvie: Honoraria; BMS/Celgene: Honoraria; Taiho: Honoraria; Pfizer: Honoraria; Astellas: Honoraria; Amgen: Honoraria; Jazz: Honoraria; Merck: Honoraria. Chu:Bristol Myers Squibb: Honoraria; Gilead: Honoraria; Sanofi: Honoraria; Janssen: Honoraria. Sandhu:Pfizer: Honoraria; Amgen: Honoraria; Takeda: Honoraria; Forus Therapeutics: Honoraria; Sanofi: Honoraria; Janssen: Honoraria; BMS/Celgene: Honoraria. Wang:Abbvie: Honoraria.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution