Preclinical efficacy of MEK inhibition in Nras-mutant AML

Abstract

Oncogenic NRAS mutations are highly prevalent in acute myeloid leukemia (AML). Genetic analysis supports the hypothesis that NRAS mutations cooperate with antecedent molecular lesions in leukemogenesis, but have limited independent prognostic significance. Using short hairpin RNA-mediated knockdown in human cell lines and primary mouse leukemias, we show that AML cells with NRAS/Nras mutations are dependent on continued oncogene expression in vitro and in vivo. Using the Mx1-Cre transgene to inactivate a conditional mutant Nras allele, we analyzed hematopoiesis and hematopoietic stem and progenitor cells (HSPCs) under normal and stressed conditions and found that HSPCs lacking Nras expression are functionally equivalent to normal HSPCs in the adult mouse. Treating recipient mice transplanted with primary Nras(G12D) AMLs with 2 potent allosteric mitogen-activated protein kinase kinase (MEK) inhibitors (PD0325901 or trametinib/GlaxoSmithKline 1120212) significantly prolonged survival and reduced proliferation but did not induce apoptosis, promote differentiation, or drive clonal evolution. The phosphatidylinositol 3-kinase inhibitor GDC-0941 was ineffective as a single agent and did not augment the activity of PD0325901. All mice ultimately succumbed to progressive leukemia. Together, these data validate oncogenic N-Ras signaling as a therapeutic target in AML and support testing combination regimens that include MEK inhibitors.

Figure 1

NRAS/Nras expression is required for maintenance of AML harboring oncogenic mutations. (A) Human AML cell lines with mutations in NRAS (OCI-AML3, HL60, THP-1), KRAS (NB4), or FLT3 (MV4-11 and MOLM-14) were infected with control shRenilla targeting Renilla luciferase and 2 independent NRAS shRNAs shown in panel B. Note depletion of NRAS-mutant AML cells expressing either NRAS hairpin. Max, maximum. (B) Western blot demonstrating N-Ras protein knockdown in each of the human AML cell lines using shNRAS.30, 34, and shRenilla control. (C) Western blot demonstrating N-Ras protein knockdown in primary murine Nras^G12D AML #6695 expressing shNras.26, shNras.30, or the control shRenilla. We infected AML #6695 with shNras.26, shNras.30, or shRenilla at equivalent multiplicities of infection (input mCherry-positive cells 17% to 29% of total AML), bulk transplanted into recipient mice (n = 4 mice per shRNA), and measured the percentage of mCherry-positive blasts (D) in BM and spleen (E) 3 weeks later. Note the selective dropout of shNras-expressing cells with 2 independent shRNAs (P < .0001). (F) Survival of mice transplanted with Nras^G12D AML #6695 (red lines, n = 10) or with a control AML generated in a WT mouse (black lines, n = 6). AMLs were infected with shNras.30 (solid lines) or control shRenilla (broken lines) and sorted to near purity before transplant. Nras knockdown prolongs survival in Nras-mutant AML (P = .0052). (G) Nras message knockdown in primary mouse AML cells before transplant (shNras.30) assessed by quantitative polymerase chain reaction. Early relapse (day 62) shNras.30 mCherry-positive leukemia (relapse) partially restored Nras expression relative to Renilla controls.

Figure 2

Nras is dispensable for normal function of HSCs. (A) Total number of CD150⁺CD48⁻LSK HSCs, CD150⁻CD48⁻LSK MPPs, and LSK cells in BM and spleens (SP) of WT (Nras^+/+) and Mx1-cre; Nras^fl/fl (Nras^Δ/Δ) mice 2 weeks after pIpC treatment (n = 10). (B) BrdU incorporation in WT and Nras^Δ/Δ HSCs after 24-hour BrdU incorporation (n = 6). SLAM, signaling lymphocyte activation molecule. (C) BM and spleen cellularity in WT and Nras^Δ/Δ mice (n = 10). (D) 5 × 10⁵ donor BM cells from Mx1-cre; Nras^fl/fl (Nras^Δ/Δ) or littermate control mice at 2 weeks after pIpC treatment were transplanted into irradiated recipient mice with 5 × 10⁵ recipient BM cells. Donor cell reconstitution in the myeloid (Gr-1⁺ or Mac-1⁺ cells), B- (B220⁺), and T- (CD3⁺) cell lineages was monitored for 4 to 20 weeks after transplantation (n = 5 recipients/genotype). Only the week 4 time point is significant for B- (P < .001) and T-cell (P = .024) repopulation. (E) Competitive repopulation of Mx1-cre; Nras^G12D/fl; (Nras^G12D/Δ), Mx1-cre; Nras^G12D/+ (Nras^G12D/+), or littermate control BM cells (n = 5 recipients/genotype). Two-tailed Student t tests were used to assess statistical significance and P < .001 between Nras^G12D/+ or Nras^G12D/Δ and control at 8, 12, 16, and 20 weeks with no significant difference between Nras^G12D/+ and Nras^G12D/Δ.

Figure 3

MEK inhibition prolongs survival in mice transplanted with Nras^G12D AML. (A) Survival of secondary recipient mice engrafted with 2 transplantable MPNs that were treated with PD901 (n = 8), GDC-0941 (n = 8), a combination of PD901 and GDC-0941 (n = 8), or control vehicle (n = 10). PD901 significantly prolonged the survival of transplant recipients (P = .0009). (B) Survival of recipient mice transplanted with 3 independent aggressive Nras^G12D AML lines treated with PD901 (n = 11), GDC-0941 (n = 11), a combination of PD901 and GDC-0941 (n = 8), or the control vehicle (n = 9). PD901 significantly prolonged the survival of transplant recipients (P = .0003). (C) Southern blot analysis of AML #6768 reveals an identical pattern of retroviral integrations in recipients treated with PD901 compared with controls. (D) WT mice treated with vehicle, PD901 (5 mg/kg), or trametinib (0.5 mg/kg) were euthanized at time 0 and 2, 8, and 24 hours, BM was collected, and c-Kit⁺ cells were assayed for ERK phosphorylation (pERK) under basal conditions and after stimulation with stem cell factor (SCF). (E) Survival of recipient mice transplanted with the 3 aggressive Nras^G12D AMLs shown in panel C treated with 0.5 mg/kg per day trametinib (n = 12) or the control vehicle (n = 10) demonstrating a significant improvement in overall survival (P = .0063).

Figure 4

MEK inhibition reduces proliferation in Nras-mutant AML without inducing differentiation. (A) Serial white blood cell counts in secondary transplant recipients of myeloid neoplasms #6730 and #8064 treated with PD901 (red line, n = 8) or a control vehicle (black line, n = 10), with error bars representing the 95% confidence interval (*statistical significance by unpaired t test). Next, mice transplanted with AML #6695 were treated with 4 daily doses of vehicle (n = 4) or PD901 (n = 5) starting 10 days posttransplant. (B-C) PD901 treatment effectively reduced white blood cell counts (P = .045) and spleen size (P = .0057). (D) Normalized viable cell count after 4 days of growth in liquid culture with increasing doses of PD901 in 5 independent N-Ras^G12D myeloid neoplasms (n ≥ 3 mice per leukemia). (E) Western blot analysis of granulocyte macrophage colony-stimulating factor (GM-CSF) stimulated pERK and phosphorylated Akt (pAkt) in BM cells isolated from transplant recipients of Nras^G12D myeloid neoplasms and WT BM control. BM from moribund mice was assayed for level of pERK, total ERK, pAkt, total Akt, and β-actin at B (basal conditions), following S (starvation), and subsequent GM-CSF stimulation in the absence or presence of increasing concentrations of PD901 (0.01, 0.1, and 1 μM). (F) Mice engrafted with AML #6695 were treated with PD901 (n = 5) or vehicle (n = 4) for 4 days (as in panels B and C) and total BM was stained for c-Kit, Mac-1 (CD11b), and Gr-1 surface expression. WT untreated BM is shown as a control. Representative plots are shown, and numbers represent the mean percentage for each gate across replicates.

Figure 5

MEK inhibition does not induce apoptosis in Nras-mutant AML. (A) Cell-cycle analysis using in vivo BrdU incorporation comparing vehicle- (n = 2) and PD901- (n = 3) treated Nras^G12D AML #6730 after 24 hours. (B-C) The percentage of BM cells collected from recipients of AML #6695 after PD901 (n = 5) or control vehicle (n = 4) treatment expressing cleaved Caspase-3 (P = .43) and Annexin V/7-AAD (P = .18). (D) Induction of apoptosis in short-term ex vivo culture measured by percentage activated Caspase-3 in 3 independent Nras^G12D AML lines (n = 2 per AML) treated with control vehicle, staurosporine (STS), or PD901.