Advertisement

Inhibition of Pol I transcription treats murine and human AML by targeting the leukemia-initiating cell population

Nadine Hein, Donald P. Cameron, Katherine M. Hannan, Nhu-Y N. Nguyen, Chun Yew Fong, Jirawas Sornkom, Meaghan Wall, Megan Pavy, Carleen Cullinane, Jeannine Diesch, Jennifer R. Devlin, Amee J. George, Elaine Sanij, Jaclyn Quin, Gretchen Poortinga, Inge Verbrugge, Adele Baker, Denis Drygin, Simon J. Harrison, James D. Rozario, Jason A. Powell, Stuart M. Pitson, Johannes Zuber, Ricky W. Johnstone, Mark A. Dawson, Mark A. Guthridge, Andrew Wei, Grant A. McArthur, Richard B. Pearson and Ross D. Hannan

Data supplements

Article Figures & Data

Figures

  • Figure 1.

    MLL-driven leukemias are sensitive to inhibition of hyperactivated Pol I transcription in vitro. (A) Pol I transcription is upregulated in malignant myeloid (M/E, GFP+-sorted) cells compared with normal myeloid (Mac-1+) cells from the bone marrow (BM) of C57Bl/6 mice. rRNA FISH with DAPI counterstain in malignant myeloid (GFP+) cells compared with normal myeloid (Mac-1+) cells. Images were taken with a ×63 objective. (B) Culture-adapted MLL-driven AML cells (M/E and M/A9) (M/E either p53WT or null) were treated with vehicle or CX-5461. The cells were treated with increasing concentration of CX-5461 for 1 hour, and 47S precursor rRNA synthesis was analyzed by 32P-orthophosphate labeling (average IC50 of 100 to 200 nM in p53WT and 400 to 500 nM in p53null AML). Graph represents mean ± SEM; n = 3. (C) Cell death was determined by PI incorporation after 24 hours of CX-5461 treatment (M/E IC50 = 280 nM; M/A9 IC50 = 37 nM; M/E p53null IC50 = 5328 nM; and M/E p53null IC50 = 3800 nM). Graph represents mean ± SEM; n = 3. (D) Apoptotic cell death was analyzed by Annexin V/PI staining using flow cytometry, after cells were treated with 100 nM and 1000 nM CX-5461 for 24 hours. Representative dot plots from n = 3. (E) Western blot analysis of total p53 and p21 protein after M/E and M/A9 p53WT cells were treated with 100 nM CX-5461 for the times indicated. Representative western blot from n = 3. a.u., arbitrary units; IC50, 50% inhibitory concentration; SEM, standard error of the mean.

  • Figure 2.

    CX-5461 therapy delays disease progression and significantly extends survival in leukemic mice. (A) Leukemia progression and engraftment of M/E p53WT and M/A9 p53WT AML in recipient C57Bl/6 mice was followed by bioluminescence imaging at day 7 (7 d, baseline), 14 (14 d), and 21 (21 d) posttransplant (pt). Representative images of n = 10 per group. (B) Kaplan-Meier survival curves for CX-5461 (40 mg/kg every 3 days, start of therapy day 7 pt, last dose day 28 pt in M/E p53WT, and last dose day 20 pt in M/A9 p53WT); cytarabine/doxorubicin (5 days of 50 mg/kg cytarabine in combination with the first 3 days of 1.5 mg/kg doxorubicin, start of therapy day 7 pt, and last dose day 11 pt) and vehicle-treated leukemic mice (M/E p53WT median survival, 17 days for vehicle vs 36 days for CX-5461, P < .0001; M/A9 p53WT median survival, 13 days for vehicle vs 24 days for CX-5461, ****P < .0001; n = 8 to 10 per group). (C) WBC prior to therapy initiation and after 3 doses of CX-5461 treatment (****P < .0001; n = 8 to 10 per group). (D) Spleen weights of mice when euthanized (****P < .0001; n = 8 to 10 per group; WT nonmalignant). (E) Overall survival of CX-5461 (40 mg/kg every 3 days, 6 doses total, start of therapy day 20 pt, and last dose day 35 pt) and vehicle-treated AML1/ETO9a Nras p53WT leukemia-bearing mice (****P < .0001; n = 10 per group). (F) GFP+ circulating tumor cells prior to therapy and after 3 doses of treatment in the peripheral blood of leukemic mice (AML1/ETO9a Nras p53WT (****P < .0001; n = 7 to 10 per group). Gray indicates time of CX-5461 and vehicle treatment (B,E). Log-rank test (B,E) and unpaired 2-tailed Student t test were performed (C,D,F). Graphs represent mean ± SEM. Cyt, cytarabine; d, day; Dox, doxorubicin.

  • Figure 3.

    CX-5461 exhibits therapeutic potential in mouse and human AML independent of their p53 status. (A) Leukemia progression and engraftment of M/E p53null AML in recipient C57Bl/6 mice was followed by bioluminescence imaging at day 7 (7 d) prior to initiation of treatment, 14 (14 d), and 21 (21 d) posttransplant (pt). Representative images from n = 20 per group. (B) Kaplan-Meier survival curves for CX-5461 (35 mg/kg every 3 days, start of therapy day 7 pt, and last dose day 31 pt) and vehicle-treated leukemic mice (median survival, 11 days for vehicle vs 24 days for CX-5461; ****P < .0001; n = 20 per group). Gray indicates time of CX-5461 and vehicle treatment. (C) Spleen weight of mice when euthanized (*P = .0132; n = 20 per group). Log-rank test (B) and unpaired 2-tailed Student t test (C) were performed. Graphs represent mean ± SEM. (D) Human leukemia cell lines differ in their sensitivity to Pol I inhibition 48 hours after drug treatment as assessed by PI exclusion (cell viability). IC50 value, the concentration of CX-5461 that decreases cell viability by 50% compared to the control, was calculated using a three-parameter log vs the inhibition nonlinear regression method in GraphPad Prism software. AML cells are graphed in groups: IC50 <100 nM (blue), IC50 100 nM to 1000 nM (green), and IC50 >1000 nM (red). IC50 values are expressed as the best-fit values for at least n = 3. Graph represents mean ± SEM. (E) The average IC50 for CX-5461 does not correlate with p53 status according to an unpaired 2-tailed Student t test. TP53 mutation status was provided by the Cancer Cell Line Encyclopedia and the Catalogue of Somatic Mutations in Cancer. Only cell lines with known p53 status via these databases are shown in the figure. (F) Cell viability in response to BMH-21 (100 nM, 500 nM, and 1000 nM) and ActD (5 nM) in KASUMI-1, ML-2, MOLM-13, THP-1, MV4-11, SHI-1, and KG-1 was determined by PI exclusion after 48 hours of treatment. DMSO, dimethyl sulfoxide.

  • Figure 4.

    A single CX-5461 administration reduces the tumor burden in M/E leukemic mice. M/E p53WT AML was transplanted into recipient C57Bl/6 mice (A-E). (A) Mac-1+/GFP+ double-positive tumor cells within the BM (1.45 × 107 ± 0.05 cells for vehicle vs 0.18 × 107 ± 0.04 cells for CX-5461; ****P < .0001; n = 5) and (B) spleen weights were determined after 24 hours of CX-5461 treatment (40 mg/kg) (****P < .0001; n = 5). (C) Analysis of apoptotic cell death via TUNEL staining of femoral BM sections from mice treated for 12 hours and stained with hematoxylin and eosin (H&E). Sections shown are representative of n = 3. Scale bar represents 50 μm (D) Cell death was determined by examining the SubG1 DNA content in the BM 10 hours post-drug administration (12.2% ± 2.2 for CX-5461 vs 1.9% ± 0.3 for vehicle; *P = .0102; n = 5). (E) Total p53 protein induction in response to CX-5461 single-dose treatment in the BM of M/E leukemic mice (n = 5 per group). (F) Quantitation of cell-cycle distribution in the BM cells by BrdU incorporation 24 hours post–CX-5461 treatment (***P = .0005; n = 5). Graph represents mean ± SEM. M/E p53null AML was injected into recipient C57Bl/6 mice (G-J). (G) The total number of Mac1+/GFP+ double-positive tumor cells within the BM (3.2 × 107 ± 0.2 cells for vehicle vs 3.1 × 107 ± 0.3 cells for CX-5461 at 24 hours, and 3.9 × 107 ± 0.5 cells for vehicle and 0.5 × 107 ± 0.2 cells for CX-5461 at 72 hours; **P = .0053), and (H) spleen weights (*P = .0134; n = 4 to 6) were determined after CX-5461 administration (35 mg/kg). Graphs represent mean ± SEM. Cell death (I) was analyzed by determining SubG1 DNA content in the BM 10 hours post–CX-5461 treatment (n = 6). Quantitation of cell-cycle distribution (J) in the BM cell was analyzed by BrdU incorporation 24 hours post–CX-5461 administration (***P = .0007). Graphs represent mean ± SEM (n = 5). In all cases, an unpaired 2-tailed Student t test was used. n.s., not significant.

  • Figure 5.

    Pol I inhibition activates ATM/ATR-dependent signaling and alters cell-cycle progression in human AML cell lines. (A) Cell-cycle analysis by BrdU incorporation after 24 hours of Pol I inhibition (n = 4). (B) Western blot analysis of CHK1 S345, CHK2 T68, and pp53 S15 phosphorylation, and total p53 abundance in 6 human AML cell lines with varied p53 status after treatment with 500 nM CX-5461 for the times indicated (n = 2). (C) Quantitation of cell-cycle distribution by BrdU incorporation in SHI-1, KG-1, and THP-1 after 30 minutes of pre-treatment with either ATMi (5 μM KU-5593), ATRi (5 μM VE-821), or both ATMi/ATRi (5 μM) in combination with CX-5461 (100 nM) treatment of 24 hours (n = 3). (D) Analysis of CHK1 S345, and CHK2 T68 phosphorylation in THP-1 and SHI-1 after 30 minutes pre-treatment with either ATMi (5 μM KU-5593), ATRi (5 μM VE-821), or both ATMi/ATRi (5 μM) in combination with 100 nM or 1000 nM CX-5461 for 1 hour by western blotting (n = 2). Graphs represent mean ± SEM.

  • Figure 6.

    Transcriptome analysis of M/E p53WT AML treated with CX-5461 by RNA sequencing revealed induction of myeloid maturation. (A) Schematic diagram of the experimental design (n = 3 mice per group). M/E p53WT AML were treated with a single dose of CX-5461 (40 mg/kg). (B) Scatter-plots for p53WT M/E illustrating logFC over CPM at 10 hours, and (C) GeneGo pathway analysis. (D) M/E p53WT or p53null AML were engrafted into recipient C57Bl/6 mice and treated with CX-5461 for 48 hours. Differential cell classification was performed on May-Grünwald Giemsa-stained cytospins prepared from GFP+-sorted BM (****P < .0001, ***P = .0008, **P = .0066; n.s. P = .0036; n = 5). Graphs represent mean ± SEM. Insert is a representative image from 1 mouse per group. Scale bar represents 10 μm. (E) Expression of Mac-1, as determined by flow cytometry, in the BM-derived tumor cells (M/E p53WT and p53null, MFI, ****P < .0001; n = 5). (F) Representative flow cytometry dot plots from 1 mouse per group and quantitation of c-Kit expression in GFP+ tumor cells from the BM of M/E p53WT leukemic mice (****P < .0001; n = 5). (G) Mac-1 cell surface expression in KG-1, THP-1, ML-2, and MOLM-13 cells in response to CX-5461 or CX-5447 treatment after 24 hours (n = 3). Graph represents mean ± SEM. Unpaired 2-tailed Student t test (A,D) was used. CPM, counts per million; CRTH2, chemoattractant receptor-homologous molecule expressed on T helper type 2 cells; FACS, fluorescence-activated cell sorting; FDR, false discovery rate; G-CSF, granulocyte colony-stimulating factor; logFC, log fold-change; MFI, mean fluorescence intensity; n.s., not significant; PTAFR, platelet-activating factor receptor; RNA-seq, RNA sequencing; Th2, T helper 2.

  • Figure 7.

    CX-5461 reduces the L-GMP and LIC population in MLL-driven AML. M/E p53WT or p53null AML cells were engrafted into recipient C57Bl/6 mice and treated with CX-5461 for 48 or 72 hours (p53WT 40 mg/kg, p53null 35 mg/kg). (A) The L-GMP population (% of the GFP+ tumor cells that are GMP+) was determined by flow cytometry at 48 and 72 hours (p53WT: *P = .0224, **P = .0011, n = 5; p53null: **P = .0017; n = 4). (B) Serial dilution transplant experiment. M/E tumor cells from the BM of either a CX-5461 or vehicle-treated mouse were re-injected into recipient C57Bl/6 mice (500 000 cells: n = 5 per group; 10 to 100 000 cells: n = 9 per group) (500 000 cells vehicle vs CX-5461, P = .0027; 100 000 cells vehicle vs CX-5461, P < .0001; 10 000 cells vehicle vs CX-5461, P = .0078; and 1000 cells vehicle vs CX-5461, P < .0001). (C) Single-dose CX-5461 treatment reduces LIC frequency. (D) Drug treatment reduces the clonogenic capacity in methylcellulose (****P < .0001; n = 5). (E) CX-5461 (500 nM) reduces colony forming potential in primary human MLL and non-MLL rearranged AML. Representative image from 1 patient sample. (F) Schematic overview of the patient-derived AML (relapsed/refractory, NOS, FAB M1, and non-MLL) xenograft experiments. (G) CX-5461 (40 mg/kg) therapy (total of 6 doses) reduced the percentage of hCD45+ tumor cells in the BM of mice with established human disease (**P = .0075; n = 6). (H) Colony formation of sorted hCD45+ tumor cells is significantly reduced post–CX-5461 therapy ex vivo (**P = .0048; n = 4). (I) Delayed disease latency in secondary recipient transplanted with sorted CX-5461–treated hCD45+ tumor cells assessed by analysis of disease burden in the BM of secondary recipient (****P < .0001; n = 6). Unpaired 2-tailed Student t test. Graphs represent mean ± SEM. CI, confidence interval; FAB, French-American-British; FACS, fluorescence-activated cell sorting; NOS, not otherwise specified; n.s., not significant; pb, peripheral blood.