Acute myeloid leukemia drug-tolerant persister cells survive chemotherapy by transiently increasing plasma membrane rigidity, that also increases their sensitivity to immune cell killing

Abstract

Resistance to chemotherapy remains a major hurdle to the cure of patients with acute myeloid leukemia (AML). Recent studies indicate that a minority of malignant cells, termed drug-tolerant persisters (DTP), stochastically upregulate stress pathways to evade cell death upon acute exposure to chemotherapy without acquiring new genetic mutations. This chemoresistant state is transient and the cells return to the baseline state after removal of chemotherapy. Nevertheless, the mechanisms employed by DTP to resist chemotherapy are not well understood and it is largely unknown whether these mechanisms are also seen in patients receiving chemotherapy. Here, we used leukemia cell lines, primary AML patients' samples and samples from patients with AML receiving systemic chemotherapy to study the DTP state. We demonstrated that a subset of AML cells transiently increases membrane rigidity to resist killing due to acute exposure to daunorubicin and Ara-C. Upon removal of the chemotherapy, membrane rigidity returned to baseline and the cells regained chemosensitivity. Although resistant to chemotherapy, the increased membrane rigidity rendered AML cells more susceptible to T-cell-mediated killing. Thus, we identified a novel mechanism by which DTP leukemic cells evade chemotherapy and a strategy to eradicate these persistent cells.

Figure 1.

An in-vitro model for acute myeloid leukemia drug-tolerant persister cells. (A) MOLM13 drug-tolerant persistent (DTP) cells were treated with a combination of daunorubicin (40 nM) and Ara-C (200 nM) to achieve a 90% inhibitory concentration (IC₉₀). Residual viable cells were collected on day 6 after the daunorubicin and Ara-C treatment and plated in fresh medium. Growth and viability of the cells were measured over time by trypan blue exclusion staining. Data represent the mean ± standard deviation (SD) of three independent experiments. (B) Growth and viability of persisting MOLM13 over time starting 20 days after daunorubicin and Ara-C treatment as measured by trypan blue exclusion staining. Data represent the mean ± SD from three independent experiments. (C) Persisting MOLM13 cells were collected on day 6 after daunorubicin (40 nM) and Ara-C (200 nM) treatment and plated in fresh medium to recover until day 20. On day 20 cells were retreated with increasing concentrations of daunorubicin for 72 hours and growth and viability were measured using Alamar Blue staining. Data represent the mean ± SD growth and viability from a representative experiment (N=2). (D) Cell cycle analysis as measured by propidium iodide staining and flow cytometry in persisting MOLM13 cells collected on day 6 after daunorubicin and Ara-C treatment. Data represent mean ± SD from a representative experiment (N=2). (E) Wild-type and persisting MOLM13 cells were labeled with 5-ethynyl-2′-deoxyuridine (EdU). EdU uptake was measured by flow cytometry 2 and 72 hours after treatment. Data represent the mean ± SD from a representative experiment (N=2). (F) MOLM13 control and DTP cells were labeled with CellTrace Violet stain. Staining intensity and dilution over time were measured at days 1 and 5 after staining using flow cytometry.

Figure 2.

Lipid composition of leukemic drug-tolerant persister cells. (A) MOLM13 cells were treated with a combination of daunorubicin (40 nM) and Ara-C (200 nM) to achieve a 90% inhibitory concentration (IC₉₀). Residual viable cells were collected on day 6 after daunorubicin and Ara-C treatment for lipidomic analysis. Difference in proportion of phosphatidylcholine species between drug-tolerant persister (DTP) and control samples. Species above the x-axis are enriched in DTP samples, and species below are enriched in untreated samples. (B) Difference in proportion of triglyceride species between DTP and control samples. Species above the x-axis are enriched in DTP samples, and species below are enriched in untreated samples. (C) Difference in proportion of phosphatidylethanolamine species between DTP and control samples. Species above the x-axis are enriched in DTP samples, and species below are enriched in untreated samples. (D) Heatmap showing abundance of sphingomyelin species in MOLM13, MV4-11, THP-1 and NB4 cells in the DTP phase (day 6 after chemotherapy) and in the recovery phase (day 21 after chemotherapy). Relative fold-change levels are indicated on the color scale, with numbers indicating the log₂ fold change normalized to untreated cells. (E) MOLM13 DTP cells were collected on day 6 after daunorubicin and Ara-C treatment and incubated with BODIPY-FL-C₁₆. The graph shows BODIPY-FL-C₁₆ mean fluorescence intensities measured by flow cytometry. Data represent the mean ± standard deviation from three independent experiments. PC: phosphatidylcholines; TG: triglycerides; PE: phosphatidylethanolamines; SM: sphingomyelins; MFI: mean fluorescent intensity.

Figure 3.

Leukemic drug-tolerant persister cells exhibit decreased cell membrane fluidity. (A) MOLM13, NB4, OCI-AML2, MV4-11, THP-1 and OCI-AML3 cells were treated with daunorubicin + AraC to achieve a 90% inhibitory concentration (IC₉₀). Residual viable cells were collected on day 6 after treatment and stained using the fluorescent PDA probe to assess membrane fluidity. Drug-tolerant persister (DTP) recovery represents DTP cells collected on day 6 after treatment, plated in fresh medium to recover until day 20. The emission fluorescence was quantified using a plate reader in triplicate. Data represent the mean ± standard deviation (SD) of three independent experiments. **P<0.01; ****P<0.0001 by the Student t test. (B) Schematic illustration showing a patient-derived xenograft model. Primary acute myeloid leukemia (AML) cells were injected into the right femur of sub-lethally irradiated NSG mice. Eight weeks after cell injection, mice were treated with Ara-C 60 mg/kg/day or vehicle for 5 days. DTP cells were collected on day 8 after Ara-C treatment. (C) Primary AML cells were engrafted into NSG mice and treated with Ara-C as in Figure 3B. The percent of AML cells (CD45⁺CD33⁺) engrafted into the mouse femur was measured by flow cytometry. ****P<0.0001 by the Student t test. (D) Pooled human CD45⁺ patient-derived DTP cells were collected from mice on day 8 after Ara-C treatment. Cells were stained using the fluorescent PDA probe to assess membrane fluidity. The emission fluorescence was quantified using a plate reader in triplicate. Data represent the mean ± SD. ****P<0.0001 by the Student t test. (E) Membrane fluidity of blast cells in paired patients’ bone marrow samples collected on day 0 and day +5 after chemotherapy. Cells were stained using the fluorescent PDA probe and membrane fluidity was assessed using flow cytometry and analysis of the mean fluorescence intensity of CD45^dimCD34⁺CD117⁺ cells. MFI: mean fluorescence intensity.

Figure 4.

Non-pharmacological models that transiently increase membrane rigidity confer chemoresistance in acute myeloid leukemia. (A) MOLM13 cells were cultured at 31°C for 24 hours prior to analysis, at 37°C for 24 hours, or at 31°C for 24 hours and then transferred back to 37°C (31°C→37°C). Growth and viability of the cells were measured by trypan blue exclusion staining. Data represent the mean ± standard deviation (SD) of three independent experiments. (B) MOLM13 cells from Figure 4A were retreated with increasing concentrations of daunorubicin or Ara-C for 72 hours and growth and viability were measured using a sulforhodamine B (SRB) assay. Data represent the mean ± SD growth from a representative experiment (N=2). (C) MOLM13 cells cultured at 31°C, 37°C or 31°C→37°C were stained using the fluorescent PDA probe to assess membrane fluidity. Emission fluorescence was quantified using a plate reader in triplicate. Data represent the mean ± SD of three independent experiments. ****P<0.0001 by the Student t test. (D) MOLM13 cells treated with cholesterol (60 µg/mL) were stained using the fluorescent PDA probe to assess membrane fluidity. Data represent the mean ± SD of three independent experiments. ****P<0.0001 by the Student t test. (E) Control and cholesterol-treated MOLM13 cells incubated with daunorubicin at increasing concentrations. Daunorubicin uptake was assessed by flow cytometry. Data represent the mean ± SD of three independent experiments. ****P<0.0001 by two-way analysis of variance test. MFI: mean fluorescence intensity.

Figure 5.

Increasing cell membrane rigidity sensitizes acute myeloid leukemia cells to T-cell-mediated cytotoxicity. (A) MOLM13 and OCI-AML2 cells cultured at 37°C or 31°C (24 hours) were incubated with double-negative T (DNT) cells for 2 hours. Percent specific killing of acute myeloid leukemia (AML) cells by DNT cells was determined by annexin V staining and flow cytometry. Data represent the mean ± standard deviation (SD) from a representative experiment (N=3). **P<0.01 by the Student t test. (B) MOLM13 cells engrafted and collected from mice after injection with Ara-C ± DNT cells. Percent engraftment was determined by flow cytometry gating on live human CD3^–CD45⁺ cells. Data represent the mean ± SD (N=7-10). **P<0.001 by the Student t test. (C) A patient’s sample (AML 166315) engrafted and collected from mice at day +8 after Ara-C treatment. Pooled human CD45⁺ cells were incubated with DNT cells for 2 hours. Percent specific killing of AML cells by DNT cells was determined by annexin V staining in flow cytometry. Data represent the mean ± SD (N=3). ***P<0.01 by the Student t test. (D) CD34⁺ blast cells were separated from paired bone marrow samples (patient AML#3) collected on day 0 and day +5 after chemotherapy. Cells were incubated with DNT cells for 2 hours. Percent specific killing of CD34⁺CD117⁺ blast cells by DNT cells was determined by annexin V staining in flow cytometry. Data represent the mean ± SD (N=3). *P<0.05 by the Student t test. PBS: phosphate-buffered saline.