NRAS proteins are central regulators of proliferation, survival, and self-renewal in leukemia. Previous work demonstrated that the effects of oncogenic NRAS in mediating proliferation and self-renewal are mutually exclusive within leukemia subpopulations and that levels of oncogenic NRAS vary between highly proliferative and self-renewing leukemia subpopulations. These findings suggest that NRAS activity levels may be important determinants of leukemic behavior. To define how oncogenic NRAS levels affect these functions, we genetically engineered an acute myeloid leukemia (AML) cell line, THP-1, to express variable levels of NRASG12V. We replaced the endogenous NRASG12D gene with a tetracycline-inducible and dose-responsive NRASG12V transgene. Cells lacking NRASG12V oncoprotein were cell-cycle arrested. Intermediate levels of NRASG12V induced maximal proliferation; higher levels led to attenuated proliferation, increased G1 arrest, senescence markers, and maximal self-renewal capacity. Higher levels of the oncoprotein also induced self-renewal and mitochondrial genes. We used mass cytometry (CyTOF) to define the downstream signaling events that mediate these differential effects. Not surprisingly, we found that the levels of such canonical RAS-effectors as pERK and p4EBP1 correlated with NRASG12V levels. β-Catenin, a mediator of self-renewal, also correlated with NRASG12V levels. These signaling intermediates may mediate the differential effects of NRASG12V in leukemia biology. Together, these data reveal that oncogenic NRAS levels are important determinants of leukemic behavior explaining heterogeneity in phenotypes within a clone. This system provides a new model to study RAS oncogene addiction and RAS-induced self-renewal in AML.

Implications:

Different levels of activated NRAS may exert distinct effects on proliferation and self-renewal.

This article is featured in Highlights of This Issue, p. 1589

RAS proteins are small GTPases, which orchestrate signal transduction pathways to proliferation and survival. Mutations of RAS family genes lead to constitutive activation of RAS proteins and are observed in 6% to 96% of cancers, depending on cancer type (1–4). NRAS is the most commonly mutated RAS gene in hematologic malignancies and is mutated in 12% to 24% of de novo acute myeloid leukemia (AML; ref. 5). In addition, NRAS mutations commonly arise as a mechanism of treatment resistance in FLT3-mutant AML treated with FLT3-inhibitors (6–8). Mutations in NRAS almost exclusively affect codons 12, 13, or 61 in AML (5, 9). In addition, mutant NRAS can drive self-renewal in both normal (10) and leukemia (11) stem cells. However, the self-renewal effects of oncogenic NRAS are only observed in poorly proliferating cells whereas NRAS-induced proliferation is accompanied by the absence of self-renewal capacity (10, 12).

Oncogene levels may be important determinants of oncogene function through gene duplication and amplification. Gene amplifications in RAS genes have been described in solid tumors (13, 14) and in AML, where NRAS mutations have been associated with uniparental disomy (15, 16) and increased NRAS transcript levels (17). In murine models, homozygous NRAS mutations increase leukemia aggressiveness and cytokine-independent growth, relative to heterozygous mutations (17, 18). However, these studies focused on genetic modifications of NRAS. The effects of oncoprotein levels in mediating oncogenesis have not been well-studied. In a single-cell RNA-sequencing (RNA-seq) study of a murine Mll-AF9/NRASG12V model of AML, NRASG12V transcript levels were associated with distinct functions, suggesting that oncoprotein levels within a leukemic clone dictate phenotypic outcomes (12). Within the leukemia stem cell (LSC) compartment, LSCs expressing high levels of NRASG12V transcript possessed self-renewal capacity and were able to transplant leukemia, but this self-renewing LSC subset was poorly proliferative. In contrast, LSCs expressing low levels of NRASG12V transcript were incapable of self-renewal, but were highly proliferative. Although this study revealed a correlation between oncogene dose and oncogene function, the effect of oncoprotein levels on these critical leukemic functions has not been tested.

THP-1 is a human monocytic leukemia cell line bearing an MLL-AF9 fusion gene and NRASG12D, which encodes a constitutively active NRAS protein (19). In this study, we genetically modified THP-1 cells to express a tunable, activated NRAS oncogene. We used CRISPR/Cas9 to delete the endogenous NRASG12D and re-introduce NRASG12V under the control of a tetracycline-inducible promoter. The exogenous G12V allele allows us to discriminate this gene from the endogenous G12D allele in our analyses. We established two clones, B11 and G6, that express minimal endogenous NRAS and express NRASG12V in response to doxycycline (Dox) in a dose-dependent manner. In the absence of Dox, these cells express minimal NRAS protein and do not proliferate, as expected. We used this system to interrogate the impact of NRAS oncoprotein dose on intracellular signaling, proliferation, and self-renewal in AML. Moderate levels of NRASG12V led to optimal proliferation and cell-cycle progression while maximal NRASG12V levels led to enhanced self-renewal. In CyTOF analysis, NRASG12V levels were correlated with levels of canonical RAS effectors (pERK, p4EBP1, and pAKT) and a mediator of self-renewal (β-catenin). These findings suggest that the levels of these mediators may be important determinants of cellular response to NRASG12V and that the functional impact of oncogenic NRAS is determined by oncoprotein level.

Cell culture, NRAS expression vector, and dual CRISPR lentiviral vectors

THP-1 cells were purchased by ATCC and maintained in RPMI1640 medium with 10% FBS and 1% penicillin/streptomycin. Lentivirus induction was performed as described previously (20). Two clones were derived from THP-1: B11 and G6, which express minimal endogenous NRAS and express NRASG12V in response to doxycycline (Dox) in a dose-dependent manner. The endogenous NRASG12D allele was knocked out using CRISPR/Cas9 and replaced with a mutant NRAS under the control of a tetracycline-inducible promoter using pENTR-NRASG12V and TripZ-TRE-DEST-IRES-GFP- Ubq -rtTA vectors via the Invitrogen Gateway LR Clonase reaction (Thermo Fisher Scientific) following manufacturer instructions. The nucleotides of the gRNA resistant NRAS were amplified from pENTR221-NRAS by the inverse PCR with 5′-CgaAtaTaaGTTAgtAgtAgttggagcagttggtgttgg-3′ (Capital letters were replaced) and 5′-gtcatggtctcgaggaatcg-3′.

Transcriptome deep sequencing and analysis (RNA-seq)

RNA was isolated using RNAeasy Mini Kit (Qiagen). Each sample was harvested in three replicates per cell line, per condition. Sequencing was performed on an Illumina NextSeq High Output instrument (75 bp single reads; 14.7–20.7 million reads/sample). Mapping and expression calculations were generated using the Gopher rnaseq pipelines (https://bitbucket.org/jgarbe/gopher-pipelines), which executed HISAT2 (21) and Cuffnorm (22) using the Ensembl version GRCh38.84 of the human reference genome (23). Mutation abundance was calculated from the mapped reads displayed by Integrative Genomic Viewer (IGV; ref. 24). For the variability analysis, only protein coding genes with at least one sample in the group having an FPKM of greater than 1.0 were selected. Fastq files and the Cuffnorm output were deposited at Gene Expression Omnibus (GEO, RRID:SCR_005012, https://scicrunch.org/resolver/SCR_005012, GSE115911). For differential gene expression and gene set enrichment analysis (GSEA), raw read counts were normalized to counts per million (CPM) using custom R codes. Differentially expressed genes were defined using edgeR (v3.32.1; ref. 25). GSEA (v4.1.0; https://www.gsea-msigdb.org/gsea/index.jsp) was performed using MsigDB (v7.2, https://www.gsea-msigdb.org/gsea/msigdb/; refs. 11, 26, 27).

CyTOF analysis

CyTOF experiments were performed as described previously (11, 12, 28). Cells treated with Dox were harvested after 96 hours of treatment. Cells were labeled with cisplatin, a dead-cell exclusion reagent for mass cytometry (Fluidigm), blocked with Human Fc Block (anti-CD16/CD32 antibody, Catalog No. 564219; BD Pharmingen), and incubated with metal-tagged primary antibodies to cell surface markers for 30 minutes at room temperature. Immunolabeled cells were fixed in 1.6% paraformaldehyde (Electron Microscopy Sciences) for 10 minutes at room temperature, permeabilized in 95% ice-cold methanol, and stored at −80°C. Permeabilized cells were labeled with antibodies to intracellular proteins and intercalated with Cell-ID Intercalator-Ir-125 μmol/L (a DNA intercalator used for recognition of single cells in mass cytometry data). Samples were analyzed with a CyTOF2 mass cytometer (Fluidigm). CyTOF-based cell-cycle assessments were performed by quantitatively measuring levels of incorporated 5-Iodo-2′-deoxyuridine (IdU), phosphorylated retinoblastoma (pRb), phosphorylated Histone 3 (pH3), and cyclin B1 as described previously (29, 30). All antibodies were purchased from Fluidigm unless otherwise noted (Supplementary Table S1). CyTOF data analysis were performed using Cytobank (http://www.cytobank.org/index.html, www.cytobank.org, Cytobank RRID: SCR_014043, Inc., https://scicrunch.org/resolver/SCR_014043).

Tissue culture

THP-1, and the derived cell line clones B11 and G6, cells were grown in complete medium consisting of RPMI supplemented with 10% FBS and 1% penicillin/streptomycin. Clones B11 and G6 were maintained in medium containing 0.5 μg/mL of Dox to maintain minimal expression of the NRASG12V transgene. For all experiments, THP-1, B11, and G6 cells were treated with 0, 1, and 10 μg/mL of Dox for 96 hours prior to harvesting for analysis. Parental THP-1 and established cell lines were frozen with aliquots and passage time was designed for less than 1 month. Cell lines were not tested for mycoplasma.

Colony forming assays

THP-1, B11, and G6 cells were plated in Methocult H4435 (Stemcell Technologies) supplemented with 100 U/mL of penicillin/streptomycin as well as Dox at 0, 1, or 10 μg/mL. Each condition was plated in three replicates in nontissue culture-treated 24-well plates. Colonies are defined as a cluster of 10 or more cells and were scored under the microscope. For serial colony forming assays, colonies were picked, dispersed, and replated in Methocult H4435. At the first plating, cells were plated at 333 cells/well, at the secondary plating, cells were plated at 167 cells/well, and at the tertiary plating, cells were plated at 33 cells/well. To normalize colony counts across replicates, the number of colonies obtained at each condition divided by the total number of colonies obtained by all conditions at that replicate.

Western blot analysis and RAS activity assay

Whole cell lysates were run on 10% PAGE gels and transferred to a PVDF using the NuPAGE and iBlot systems (Life Technologies). Blots were blocked and stained according to manufacturer's recommendations. Blots were developed using the SuperSignal West Pico ECL Kit (Thermo Fisher Scientific) and visualized, signals were quantified using the LI-COR imaging system (LI-COR Biosciences). The following antibodies were used: anti-actin (I-19) goat polyclonal (Santa Cruz Biotechnology, Catalog No. sc-1616), NRAS (F-155) mouse monoclonal (Santa Cruz Biotechnology, Catalog No. sc-31), p44/42 MAPK (Erk1/2) rabbit (137F5, Cell Signaling Technologies, Catalog no. 4695), phospho-p44/42 MAPK(Erk1/2)(Thr202/Tyr204) rabbit (D13.14.4E, Cell Signaling Technologies, Catalog No. 4370), Akt(pan)(11E7) rabbit, phospho-Akt (Ser473) rabbit (D9E, Cell Signaling Technologies, Catalog No. 4060), and GAPDH (14C10, Cell Signaling Technologies, Catalog No. 2118). HRP-conjugated secondary antibodies were procured from Cell Signaling Technologies. Estimation of RAS activity was performed by Active Ras Pull-Down and Detection Kit (Thermo Fisher Scientific).

qPCR analysis

RNA was isolated using RNeasy Mini Kit (Qiagen) according to the manufacturer's instructions. Complementary DNA (cDNA) was generated from RNA with ReverTra AceR qPCR RT Master Mix (Toyobo). The following primers were used for NRAS: 5′-ggccgatattaatccggtgt-3′ (forward), 5′-cactgggcctcacctctatg-3′ (reverse), TNFα: 5′-ccccagggacctctctctaa-3′ (forward), 5′-tgggctacaggcttgtcact-3′ (reverse), β-ACTIN: 5′-cacagagcctcgcctttgcc-3′ (forward) and 5′-cacagagcctcgcctttgcc-3′ (reverse) as a loading control. qPCR using SYBR mix was performed on an ABI Prism 7900HT (Applied Biosystems). NRAS copy number was assessed using Genomic DNA TaqMan Copy Number Assays (Hs05807163_cn, Hs00006651_cn; Thermo Fisher Scientific). TaqMan Copy Number Reference Assay, human, TERT (Thermo Fisher Scientific) was used for normalization. Control cells were de-identified peripheral blood mononuclear cells obtained from G-CSF mobilized peripheral blood (MPB) donors from the University of Minnesota Leukemia Tissue Bank according to protocols approved by the University of Minnesota Institutional Review Board. They were provided as deidentified specimens.

Flow cytometry and senescence analysis

For staining of intracellular antigens, cells were fixed with 1.6% paraformaldehyde (Electron Microscopy Sciences) and permeabilized with 90% methanol (Sigma). Cells were analyzed using LSR II or Fortessa digital flow cytometer (BD Biosciences RRID:SCR_013311, https://scicrunch.org/resolver/SCR_013311), and cell sorting was performed using BD FACS Aria II (BD Biosciences RRID:SCR_013311, https://scicrunch.org/resolver/SCR_013311) at the University of Minnesota Flow Cytometry Resource. To detect senescent cells, the fluorometric senescent cell detection reagent SPIDER-βGal (Dojindo Laboratory) was used. Cells were fixed with 4% paraformaldehyde and treated with SPIDER-βGal, following the manufacturer's instructions.

Establishing NRAS-inducible clones

THP-1 cells harbor a mutation in NRAS that leads to a glycine to aspartic acid substitution at position 12 (NRASG12D; ref. 19). To confirm this NRAS mutation in THP-1 cells, classical Sanger sequencing was performed and revealed a mixed signal of glycine and alanine in codon 12, indicating the presence of both wild-type and mutant alleles (Supplementary Fig. S1A). Genomic wild-type and mutant NRAS alleles were quantitated in parental THP-1s by qRT-PCR and showed approximately five copies of NRAS per cell (Supplementary Fig. S1B). In AML, NRAS is most frequently activated by missense mutations but copy-number gains can be seen as well. Among 54 AML cases with genomic-level NRAS activation in the cBioPortal database, 11% were copy-number gains (Supplementary Fig. S1C; refs. 31, 32).

THP-1 cells were genetically modified to generate cell lines with tunable levels of mutant NRAS. The endogenous NRASG12D allele was knocked out using CRISPR/Cas9 and replaced with a lentivirally encoded NRASG12V transgene under the control of a tetracycline-inducible promoter. The NRASG12V transgene was inserted randomly in the genome and was not directed to the endogenous NRAS locus. First, cells were infected with a lentivirus encoding Tet-inducible NRASG12V and GFP transgenes. The NRASG12V allele was used to distinguish the inducible allele from the native NRASG12D allele. The G12V allele is also found in a subset of NRAS mutant AMLs (5, 33). The Dox-inducible NRASG12V allele harbors a synonymous mutation in gRNA binding site (Fig. 1A), rendering this allele resistant to CRISPR/Cas9 editing. Cells were grown in Dox and selected for GFP expression. We used a dual gRNA system. One gRNA (#2) was designed to target the NRAS coding region in exon 2, and the other (#1) to target upstream of the start codon. In combination, these gRNAs induced deletion of endogenous NRASG12D (Supplementary Fig. S1D). Clones harboring the knocked-out alleles were selected using puromycin for 3 to 4 weeks. Dox was included in the culture, at 1 μg/mL, to support expression of the inducible NRASG12V (which supports growth in these cells). Puromycin-resistant clones were harvested and plated in two replicates: one with Dox and the other without Dox. Dox-dependent clones were selected by choosing those that grew in the presence of Dox but did not grow in the absence of Dox (Fig. 1B). We established two clones, B11 and G6, that express minimal mutant NRAS transcript in the absence of Dox and express NRASG12V in response to Dox. Another clone, E5, developed only a partial knockout of the endogenous NRASG12D alleles; this clone retains Dox-independent, endogenous NRASG12D.

Figure 1.

Establishment of THP-1 subclones with inducible NRASG12V. A, Schema of Dox-inducible vector of NRASG12V. The CRISPR-gRNA-resistant variant is indicated (top). The Lenti-CRISPR vector used to delete endogenous NRASG12D alleles (bottom). Two gRNA were integrated in the Lenti-CRISPR vector. B, Pipeline of establishment Dox-inducible sub-clones. C, Schematic representating the state of the NRAS loci in the derivative cell lines. D–E, Modified THP-1 derivative cell lines were grown in liquid culture with 0 or 1 μg/mL Dox for 5 days. TRE-NRAS: a control cell line that retains endogenous NRASG12D; this cell line was induced with the TRIPZ-NRASG12V-GFP vector but not the Lenti-Dual gRNA-Cas9-p2a-Puro vector. E5: a cell line that was induced with both vectors but had incomplete knockout of endogenous NRAS and therefore retained expression of the endogenous NRASG12D allele in addition to the inducible allele. B11 and G6: cell lines with knockout of the endogenous NRASG12D allele and expression of the inducible NRASG12V allele. D, Western blot of NRAS. E, Live cells counts after 5 days in culture. Each bar represents the average of three biological replicates.

Figure 1.

Establishment of THP-1 subclones with inducible NRASG12V. A, Schema of Dox-inducible vector of NRASG12V. The CRISPR-gRNA-resistant variant is indicated (top). The Lenti-CRISPR vector used to delete endogenous NRASG12D alleles (bottom). Two gRNA were integrated in the Lenti-CRISPR vector. B, Pipeline of establishment Dox-inducible sub-clones. C, Schematic representating the state of the NRAS loci in the derivative cell lines. D–E, Modified THP-1 derivative cell lines were grown in liquid culture with 0 or 1 μg/mL Dox for 5 days. TRE-NRAS: a control cell line that retains endogenous NRASG12D; this cell line was induced with the TRIPZ-NRASG12V-GFP vector but not the Lenti-Dual gRNA-Cas9-p2a-Puro vector. E5: a cell line that was induced with both vectors but had incomplete knockout of endogenous NRAS and therefore retained expression of the endogenous NRASG12D allele in addition to the inducible allele. B11 and G6: cell lines with knockout of the endogenous NRASG12D allele and expression of the inducible NRASG12V allele. D, Western blot of NRAS. E, Live cells counts after 5 days in culture. Each bar represents the average of three biological replicates.

Close modal

We sequenced the genomic NRAS genes in B11, G6, and E5 cells (Supplementary Table S2; Supplementary Fig. S1E; and schemic diagram, Fig. 1C). E5 retained a functional coding region for one NRASWT and two NRASG12D alleles. In B11, one NRASG12D allele was unmodified. In G6, we detected deletion mutations in all NRAS genes. At the protein level, B11 and G6 cell lines expressed minimal NRAS levels in the absence of Dox, and NRAS expression was induced by Dox in these cell lines (Fig. 1D). In the absence of Dox treatment, B11 and G6 grow slowly, leading to very few cells after 5 days in culture, indicating an addiction to NRAS oncogene expression (Fig. 1E). In contrast, E5 expressed reduced but detectable NRAS protein levels in the absence of Dox, but these levels rose in the presence of Dox (Fig. 1D). B11 cells retain one copy of the endogenous G12D gene, but this allele is not expressed. In summary, in E5 cells, mutant NRAS protein is derived from both endogenous and Dox-inducible genes but expression of mutant NRAS in B11 and G6 clones is inducible by Dox.

Mutant NRAS levels correlate with dox dose

To assess the dose responsiveness of NRASG12V protein levels to Dox treatment, B11 and unmodified THP-1 parental cells were grown in varying concentrations of Dox for 96 hours. We performed RNA-seq to assess the expression of mutant NRAS genes in B11 cells grown with 0, 1, and 10 μg/mL Dox (Fig. 2A). Transcripts bearing the G12D mutation were only detected in parental THP-1 cells, confirming efficient knockout in the B11 derivative cell line. In contrast, B11 cells almost exclusively expressed the G12V mutation when grown in Dox, indicating Dox-responsiveness of the transgene. Among the three B11 samples grown in the absence of Dox, only one read (among 21 reads) contained the G12V mutation in only one replicate indicating minimal leakiness in the system. These data demonstrate that the B11 derivative cell line expresses mutant NRASG12V in a Dox-dose-dependent manner. The differences in mutant transcript levels between samples treated with 1 and 10 μg/mL of Dox are subtle (Fig. 2A; Supplementary Table S3).

Figure 2.

Oncogenic NRAS gene, protein, and activity levels are Dox dose-dependent. B11 and THP-1 parental cells were plated in liquid culture with 0, 0.1, 1, and 10 μg/mL Dox for 96 hours. A, RNA-seq reveals levels of mutant NRAS allele (endogenous G12D, or knocked-in G12V) in each sample. B, B11 and G6 cells were grown without Dox for 96 hours and then treated with the Dox dose indicated for 48 hours before being harvested and analyzed by Western blot analysis for NRAS and NRAS-effector, phosphorylated ERK (pERK), total ERK, phosphorylated AKT (pAKT), total AKT, and ß-actin. C, RAS-GTP level assay measures the amount of GTP-bound (active) Ras GTPase with the Raf1 protein-binding domain. D, Western blot analysis of B11 and G6 cells grown in the absence of Dox for 0, 24, 48, 72, and 96 hours.

Figure 2.

Oncogenic NRAS gene, protein, and activity levels are Dox dose-dependent. B11 and THP-1 parental cells were plated in liquid culture with 0, 0.1, 1, and 10 μg/mL Dox for 96 hours. A, RNA-seq reveals levels of mutant NRAS allele (endogenous G12D, or knocked-in G12V) in each sample. B, B11 and G6 cells were grown without Dox for 96 hours and then treated with the Dox dose indicated for 48 hours before being harvested and analyzed by Western blot analysis for NRAS and NRAS-effector, phosphorylated ERK (pERK), total ERK, phosphorylated AKT (pAKT), total AKT, and ß-actin. C, RAS-GTP level assay measures the amount of GTP-bound (active) Ras GTPase with the Raf1 protein-binding domain. D, Western blot analysis of B11 and G6 cells grown in the absence of Dox for 0, 24, 48, 72, and 96 hours.

Close modal

Next, NRAS protein levels were assessed by Western blot analysis. In the absence of Dox, NRAS expression was minimally detectable in B11 and G6 cells. However, NRAS was induced by Dox treatment as were levels of phosphorylated RAS-effectors, pERK and pAKT. The levels of these proteins were correlated to Dox dose (Fig. 2B). These data show that transcript levels do not accurately reflect protein expression (as has been previously reported; ref. 34) and subtle differences in transcript levels can lead to significant differences in protein level. To investigate whether these protein levels reflect levels of active RAS, we measured the amount of GTP-bound (active) RAS through its interaction with the RAF1 protein-binding domain (35, 36). This assay demonstrated that RAS activity level also correlated with Dox concentration (Fig. 2C). Furthermore, in the absence of Dox, B11 and G6 cells displayed progressive reduction in levels of phosphorylated RAS effectors, ERK and AKT, over time (Fig. 2D). Together, these data demonstrate that B11 and G6 cells express mutant NRAS in response to Dox and that the levels of the mutant NRAS protein correlates with Dox dose.

Moderate levels of oncogenic NRAS lead to maximal cell cycling

Next, we assessed the impact of NRASG12V levels on B11 cell growth. After 96 hours of liquid culture, cells grown in the absence of Dox showed minimal proliferation (Fig. 3A). Cells grown in 1 μg/mL of Dox showed maximal proliferation. In contrast, treatment with 10 μg/mL of Dox led to a reduction of cell numbers relative to 1 μg/mL of Dox. These data suggest that moderate levels of oncogenic NRAS activity correlate with optimal cell growth whereas high levels of NRAS activity led to suboptimal cell growth.

Figure 3.

Maximal cell cycle and proliferation is induced by moderate levels of oncogenic NRAS and inhibited by high levels. B11 and G6 cells were plated in liquid culture with 0, 1, and 10 μg/mL Dox. A, After 5 days of liquid culture, B11 cells were stained with Trypan blue, and live cells (those excluding Trypan blue) were counted. Each bar represents the average value from three biological replicates. B, Cells were labeled with antibodies to phospho-Retinoblastoma (pRb), Cyclin B1, and phospho-Histone 3 (pHis3), Ki67, and the DNA intercalator IdU. Cells were then assessed by CyTOF to determine cell cycle status as in ref. 30 and to identify proliferating cells. See Supplementary Table S9 for full list of P values. C, Senescence-associated β-galactosidase was measured in B11 and THP-1 (parental) cells using the SPIDER-βGal system in three biological replicates. D, RNA was extracted from B11 cells after 96 hours in culture and submitted for qPCR to amplify the senescence marker, TNFα. Data are representative of three biological replicates.

Figure 3.

Maximal cell cycle and proliferation is induced by moderate levels of oncogenic NRAS and inhibited by high levels. B11 and G6 cells were plated in liquid culture with 0, 1, and 10 μg/mL Dox. A, After 5 days of liquid culture, B11 cells were stained with Trypan blue, and live cells (those excluding Trypan blue) were counted. Each bar represents the average value from three biological replicates. B, Cells were labeled with antibodies to phospho-Retinoblastoma (pRb), Cyclin B1, and phospho-Histone 3 (pHis3), Ki67, and the DNA intercalator IdU. Cells were then assessed by CyTOF to determine cell cycle status as in ref. 30 and to identify proliferating cells. See Supplementary Table S9 for full list of P values. C, Senescence-associated β-galactosidase was measured in B11 and THP-1 (parental) cells using the SPIDER-βGal system in three biological replicates. D, RNA was extracted from B11 cells after 96 hours in culture and submitted for qPCR to amplify the senescence marker, TNFα. Data are representative of three biological replicates.

Close modal

To assess the impact of oncogenic NRAS levels on cell-cycle progression, THP-1 parental, B11, and G6 cells were grown in 0, 1, or 10 μg/mL of Dox and incubated for 96 hours. Cell-cycle assessments were performed by CyTOF as previously described (29,30). The absence of Dox led to a high frequency of cells in G1 phase, indicating G1-arrest in B11 and G6 cells depleted of oncogenic NRAS (Fig. 3B). Intermediate Dox concentration, which correlated with moderate levels of mutant NRAS, led to the highest rate of cycling (indicated by the highest percentage of S-phase cells). Ki67, an independent marker of proliferation, showed that intermediate NRASG12V levels correlated with maximal Ki67High cells as well. In contrast, high NRASG12V levels led to a reduction of S-phase percentage and an increase in G0 (quiescent cell) percentage. Notably, parental THP-1 cells treated with the highest Dox dose also displayed decreased S-phase percentage and percent of cells with high Ki67 levels (Fig. 3B), but these decreases were smaller than the decreases we observed in the modified cell lines. These data suggest that the magnitude of the effect of NRASG12V on S-phase reduction is less than what we observed in the modified cell lines. In addition, high NRASG12V levels led to increased markers of senescence as measured by SPIDER-βGal assay and TNFα (Fig. 3C and D; ref. 37). In contrast, no consistent difference in G0 cell-cycle phase or senescence marker were seen in Dox-treated parental THP-1 cells (Fig. 3C), indicating that the results we observed in B11 are attributable to induction of NRASG12V. Together, these data demonstrate that oncogenic NRAS activity is dose-dependent: loss of NRAS led to cell-cycle arrest, whereas moderate levels led to maximal cell-cycle progression and proliferation, and high levels impeded proliferation and possibly induced senescence.

Cells with moderate levels of oncogenic NRAS have increased apoptosis

To assess the impact of oncogenic NRAS levels on apoptosis, THP-1 parental, B11, and G6 cells were plated with Dox and incubated for 96 hours. Levels of proliferation marker, Ki67, and apoptosis mediators were assessed by CyTOF. B11 and G6 cells treated with of Dox consistently displayed maximal levels of apoptosis mediators, cleaved caspase 3, and cleaved PARP (Fig. 4). Control cells (0 μg/mL Dox) displayed the lowest levels apoptosis mediators, indicating that oncogenic NRAS leads to apoptosis, consistent with previous work that shows that rapidly proliferating cancer cells are poised to undergo apoptosis (38). Notably, the rapidly proliferating cellular subset (identified by high Ki67 staining, Fig. 4A) displayed the highest levels of cleaved caspase 3 and cleaved PARP. Levels of MCL1 did not change consistently with NRASG12V levels (Fig. 4B), suggesting that this survival mediator is not NRASG12V responsive. THP-1 cells did not replicate this pattern. These data show that moderate levels of NRASG12V oncoprotein induce maximal apoptosis but apoptosis is not enhanced by higher oncoprotein levels. In contrast, higher levels of NRASG12V are associated with a significant reduction of proliferation. Together, these data suggest the effect of NRASG12V dose on proliferation is likely the dominant determinant of cell fate in these cells.

Figure 4.

Expression of oncogenic NRAS induces markers of apoptosis in highly proliferative cells. B11 and G6 cells were plated in liquid culture with 0, 1, and 10 μg/mL of Dox. After 96 hours in culture, cells were labeled with antibodies to cleaved caspase 3, cleaved PARP, and MCL1 and analyzed by CyTOF. A, Representative samples are visualized by viSNE (56). viSNE is a dimensionality reduction visualization technique for CyTOF data. In a viSNE plot, cells with similar expression profiles are arranged next to each other in two-dimensional space. Each dot represents a single cell. The color of each dot reflects the relative amount of protein (indicated in the label at the top of each column of viSNE plots). B, The arcsinh ratio of expression (relative to the 0 μg/mL Dox control). The arcsinh ratio is the standard metric for comparison of CyTOF data (56, 57). Each sample was plated in 6 replicates. See Supplementary Table S9 for full list of P values.

Figure 4.

Expression of oncogenic NRAS induces markers of apoptosis in highly proliferative cells. B11 and G6 cells were plated in liquid culture with 0, 1, and 10 μg/mL of Dox. After 96 hours in culture, cells were labeled with antibodies to cleaved caspase 3, cleaved PARP, and MCL1 and analyzed by CyTOF. A, Representative samples are visualized by viSNE (56). viSNE is a dimensionality reduction visualization technique for CyTOF data. In a viSNE plot, cells with similar expression profiles are arranged next to each other in two-dimensional space. Each dot represents a single cell. The color of each dot reflects the relative amount of protein (indicated in the label at the top of each column of viSNE plots). B, The arcsinh ratio of expression (relative to the 0 μg/mL Dox control). The arcsinh ratio is the standard metric for comparison of CyTOF data (56, 57). Each sample was plated in 6 replicates. See Supplementary Table S9 for full list of P values.

Close modal

High levels of NRASG12V leads to maximal self-renewal capacity

Previous work demonstrated that oncogenic NRAS enforces self-renewal in HSCs and LSCs (10, 11). We used serial colony forming assays to assess the effect of oncogenic NRAS levels on self-renewal. THP-1 parental, B11, and G6 cells were plated in methylcellulose semisolid media (MethoCult H4435) for colony forming assays with Dox. Colonies were harvested after 7 days and replated in methylcellulose for a total of three generations of serial colony forming assays, a surrogate for self-renewal capacity. In these assays, high Dox concentration led to the maximal colony formation in B11 and G6 cells, but not in THP-1 (Fig. 5A). These data suggest that high levels of oncogenic NRAS enhance self-renewal beyond that of moderate NRASG12V levels. Our findings that cell-cycle progression and self-renewal are differentially sensitive to oncogenic NRAS dose are consistent with earlier work, demonstrating that these two effects of oncogenic NRAS are distinct and associated with differential levels of NRASG12V (10, 12).

Figure 5.

High levels of oncogenic NRAS lead to maximal self-renewal capacity. A, B11, G6, and THP-1 cells were plated in MethoCult colony forming assays with 0, 1, or 10 μg/mL of Dox. Colonies were harvested, dissociated, and re-plated in secondary and tertiary colony forming assays with the same concentration of Dox for the entire experiment. n = 3 replicates per cell line. To normalize colony counts across replicates, the number of colonies obtained at each condition was divided by the total number of colonies obtained by all conditions at that replicate. B, B11 cells were plated in liquid culture with 0, 1, and 10 μg/mL Dox. After 96 hours in culture, cells were labeled with antibodies to signaling intermediates and analyzed by CyTOF. Representative samples are visualized by viSNE (B and D) and levels of protein expression in each replicate is quantified (C and E). The arcsinh ratio of expression (relative to the 0 μg/mL Dox control) are displayed. The arcsinh ratio is the standard metric for comparison of CyTOF data (56, 57). Each sample was plated in 6 replicates. See Supplementary Table S9 for full list of P values.

Figure 5.

High levels of oncogenic NRAS lead to maximal self-renewal capacity. A, B11, G6, and THP-1 cells were plated in MethoCult colony forming assays with 0, 1, or 10 μg/mL of Dox. Colonies were harvested, dissociated, and re-plated in secondary and tertiary colony forming assays with the same concentration of Dox for the entire experiment. n = 3 replicates per cell line. To normalize colony counts across replicates, the number of colonies obtained at each condition was divided by the total number of colonies obtained by all conditions at that replicate. B, B11 cells were plated in liquid culture with 0, 1, and 10 μg/mL Dox. After 96 hours in culture, cells were labeled with antibodies to signaling intermediates and analyzed by CyTOF. Representative samples are visualized by viSNE (B and D) and levels of protein expression in each replicate is quantified (C and E). The arcsinh ratio of expression (relative to the 0 μg/mL Dox control) are displayed. The arcsinh ratio is the standard metric for comparison of CyTOF data (56, 57). Each sample was plated in 6 replicates. See Supplementary Table S9 for full list of P values.

Close modal

Gene expression profiling reveals differential effects of NRASG12V levels in AML

We performed whole transcriptomic RNA-seq on THP-1 and B11 cells grown with Dox. Notably, Dox did not significantly alter THP-1 gene expression; only 17 to 68 genes were differentially expressed between these treatment groups (Supplementary Table S4). In contrast, B11 cells treated with 0 μg/mL versus 1 μg/mL of Dox differentially expressed 2,887 unique genes (genes that were not differentially expressed by THP-1 cells, Supplementary Table S4). B11 cells treated with 1 μg/mL versus 10 μg/mL of Dox differentially expressed 875 unique genes. We performed GSEA to define the pathways that NRASG12V activates to drive proliferation and self-renewal in leukemia. B11 cells treated with 0 μg/mL versus 1 μg/mL of Dox displayed enrichment of well-known oncogene-induced functions such as apoptosis, cell-cycle progression, and leukemia signaling (representative results in Table 1, full analysis and data in Supplementary Tables S5–S8), consistent with the results presented in Figs. 3–5. Notably, NRASG12V expression was also strongly associated with gene sets of LSCs and self-renewal, confirming our recent findings that NRASG12V contributes self-renewal capacity to AML (12). Finally, we found that expression of NRASG12V was also associated with gene sets associated with metabolism, confirming the recent recognition of the important role of oncogenes in driving metabolic processes in AML (39). In contrast, loss of NRASG12V was associated with the expression of differentiation genes, consistent with our earlier observations (11, 12) and consistent with a central role of continued oncogene expression in leukemia maintenance and suppression of differentiation.

Table 1.

Gene sets enriched in samples with moderate NRASG12V.

Comparison of B11 samples treated with 1 μg/mL versus 0 μg/mL of Dox
FunctionPathwayNESNom-pFDR-q
Apoptosis GO_APOPTOTIC_SIGNALING_PATHWAY 1.64 0.00 0.03 
Cell cycling GO_NEGATIVE_REGULATION_OF_CELL_CYCLE_ARREST 1.8 0.00 0.01 
 TANG_SENESCENCE_TP53_TARGETS_DN 2.75 0.00 0.00 
 ZHANG_PROLIFERATING_vs._QUIESCENT 1.64 0.01 0.03 
 FISCHER_G2_M_CELL_CYCLE 3.07 0.00 0.00 
 BENPORATH_PROLIFERATION 2.97 0.00 0.00 
 MARKEY_RB1_ACUTE_LOF_UP 2.96 0.00 0.00 
 WHITFIELD_CELL_CYCLE_LITERATURE 2.95 0.00 0.00 
 REACTOME_CELL_CYCLE 2.94 0.00 0.00 
 WONG_EMBRYONIC_STEM_CELL_CORE 2.92 0.00 0.00 
 HALLMARK_MYC_TARGETS_V1 2.79 0.00 0.00 
 WU_APOPTOSIS_BY_CDKN1A_VIA_TP53 2.78 0.00 0.00 
 GO_MITOTIC_NUCLEAR_DIVISION 2.76 0.00 0.00 
 YU_MYC_TARGETS_UP 2.7 0.00 0.00 
 KEGG_CELL_CYCLE 2.67 0.00 0.00 
 REACTOME_S_PHASE 2.67 0.00 0.00 
 WHITFIELD_CELL_CYCLE_G2_M 2.66 0.00 0.00 
Leukemia signaling SEIDEN_ONCOGENESIS_BY_MET 1.97 0.00 0.00 
 WINTER_HYPOXIA_UP 1.89 0.00 0.00 
 GROSS_HYPOXIA_VIA_HIF1A_UP 1.89 0.00 0.00 
 KAMMINGA_EZH2_TARGETS 2.61 0.00 0.00 
 GO_MOTOR_ACTIVITY 1.68 0.00 0.02 
 LEI_MYB_TARGETS 1.99 0.00 0.00 
 LIU_CMYB_TARGETS_UP 1.64 0.00 0.01 
 DANG_MYC_TARGETS_UP 2.54 0.00 0.00 
 HALLMARK_MYC_TARGETS_V2 2.45 0.00 0.00 
Leukemia stem cells/self renewal GO_HEMATOPOIETIC_STEM_CELL_DIFFERENTIATION 1.81 0.00 0.01 
 GO_REGULATION_OF_STEM_CELL_DIFFERENTIATION 1.68 0.01 0.02 
 GROUP 1 (SELF RENEWING) 2.52 0.00 0.00 
 SOMERVAILLE LSC UP 2.07 0.00 0.00 
 RAMALHO_STEMNESS_UP 1.76 0.00 0.01 
 GO_STEM_CELL_PROLIFERATION 1.66 0.00 0.03 
 BHATTACHARYA_EMBRYONIC_STEM_CELL 2.28 0.00 0.00 
Metabolism HALLMARK_GLYCOLYSIS 1.86 0.00 0.00 
 HALLMARK_OXIDATIVE_PHOSPHORYLATION 1.95 0.00 0.00 
 GO_REGULATION_OF_OXIDATIVE_PHOSPHORYLATION 1.8 0.00 0.01 
 HALLMARK_MTORC1_SIGNALING 2.6 0.00 0.00 
 WIERENGA_STAT5A_TARGETS_GROUP1 1.85 0.00 0.00 
 GO_REGULATION_OF_MRNA_PROCESSING 1.69 0.00 0.02 
Gene sets enriched in samples without NRASG12V 
Function Pathway NES Nom-p FDR-q 
Differentiation HUANG_GATA2_TARGETS_UP −1.99 0.03 
 GROUP3 
 SOMERVAILLE LSC DN 0.02 0.02 
Comparison of B11 samples treated with 1 μg/mL versus 0 μg/mL of Dox
FunctionPathwayNESNom-pFDR-q
Apoptosis GO_APOPTOTIC_SIGNALING_PATHWAY 1.64 0.00 0.03 
Cell cycling GO_NEGATIVE_REGULATION_OF_CELL_CYCLE_ARREST 1.8 0.00 0.01 
 TANG_SENESCENCE_TP53_TARGETS_DN 2.75 0.00 0.00 
 ZHANG_PROLIFERATING_vs._QUIESCENT 1.64 0.01 0.03 
 FISCHER_G2_M_CELL_CYCLE 3.07 0.00 0.00 
 BENPORATH_PROLIFERATION 2.97 0.00 0.00 
 MARKEY_RB1_ACUTE_LOF_UP 2.96 0.00 0.00 
 WHITFIELD_CELL_CYCLE_LITERATURE 2.95 0.00 0.00 
 REACTOME_CELL_CYCLE 2.94 0.00 0.00 
 WONG_EMBRYONIC_STEM_CELL_CORE 2.92 0.00 0.00 
 HALLMARK_MYC_TARGETS_V1 2.79 0.00 0.00 
 WU_APOPTOSIS_BY_CDKN1A_VIA_TP53 2.78 0.00 0.00 
 GO_MITOTIC_NUCLEAR_DIVISION 2.76 0.00 0.00 
 YU_MYC_TARGETS_UP 2.7 0.00 0.00 
 KEGG_CELL_CYCLE 2.67 0.00 0.00 
 REACTOME_S_PHASE 2.67 0.00 0.00 
 WHITFIELD_CELL_CYCLE_G2_M 2.66 0.00 0.00 
Leukemia signaling SEIDEN_ONCOGENESIS_BY_MET 1.97 0.00 0.00 
 WINTER_HYPOXIA_UP 1.89 0.00 0.00 
 GROSS_HYPOXIA_VIA_HIF1A_UP 1.89 0.00 0.00 
 KAMMINGA_EZH2_TARGETS 2.61 0.00 0.00 
 GO_MOTOR_ACTIVITY 1.68 0.00 0.02 
 LEI_MYB_TARGETS 1.99 0.00 0.00 
 LIU_CMYB_TARGETS_UP 1.64 0.00 0.01 
 DANG_MYC_TARGETS_UP 2.54 0.00 0.00 
 HALLMARK_MYC_TARGETS_V2 2.45 0.00 0.00 
Leukemia stem cells/self renewal GO_HEMATOPOIETIC_STEM_CELL_DIFFERENTIATION 1.81 0.00 0.01 
 GO_REGULATION_OF_STEM_CELL_DIFFERENTIATION 1.68 0.01 0.02 
 GROUP 1 (SELF RENEWING) 2.52 0.00 0.00 
 SOMERVAILLE LSC UP 2.07 0.00 0.00 
 RAMALHO_STEMNESS_UP 1.76 0.00 0.01 
 GO_STEM_CELL_PROLIFERATION 1.66 0.00 0.03 
 BHATTACHARYA_EMBRYONIC_STEM_CELL 2.28 0.00 0.00 
Metabolism HALLMARK_GLYCOLYSIS 1.86 0.00 0.00 
 HALLMARK_OXIDATIVE_PHOSPHORYLATION 1.95 0.00 0.00 
 GO_REGULATION_OF_OXIDATIVE_PHOSPHORYLATION 1.8 0.00 0.01 
 HALLMARK_MTORC1_SIGNALING 2.6 0.00 0.00 
 WIERENGA_STAT5A_TARGETS_GROUP1 1.85 0.00 0.00 
 GO_REGULATION_OF_MRNA_PROCESSING 1.69 0.00 0.02 
Gene sets enriched in samples without NRASG12V 
Function Pathway NES Nom-p FDR-q 
Differentiation HUANG_GATA2_TARGETS_UP −1.99 0.03 
 GROUP3 
 SOMERVAILLE LSC DN 0.02 0.02 

Next, we used GSEA to assess the contribution of high levels of NRASG12V relative to moderate levels (Table 2). Notably, we found that high levels of NRASG12V were associated with enrichment of self-renewal and LSC genes, relative to moderate levels of NRASG12V. Likewise, gene sets of mitochondrial function, which is a hallmark of LSC function (40), were also enriched in high NRASG12V samples. Consistent with our SPIDER-βGal and TNFα assays (Fig. 3C and D), high levels of NRASG12V were associated with a senescence gene expression program. In contrast, samples with moderate NRASG12V levels were enriched in gene sets associated with inflammation and differentiation relative to samples with high NRASG12V levels. These functional differences did not correlate with significant changes in genes that encode cell surface proteins (Supplementary Fig. S2). Together, these data suggest that moderate levels of NRASG12V are required for maximal proliferation, but that higher levels of NRASG12V impede proliferation in favor of self-renewal. These data are consistent with prior work demonstrating that maximal proliferation and self-renewal capacity are distinct, mutually exclusive functions in normal tissues and in leukemia. However, a role for oncogene dose in mediating this dichotomy has not previously been established.

Table 2.

Gene sets enriched in B11 samples with moderate NRASG12V.

Comparison of B11 samples treated with 1 μg/mL versus 10 μg/mL of Dox
FunctionPathwayNESNom pFDR q
Inflammation HECKER_IFNB1_TARGETS −2.7 0.00 
 BOSCO_INTERFERON_INDUCED_ANTIVIRAL_MODULE −2.5 0.00 
 ALTEMEIER_RESPONSE_TO_LPS_WITH_MECHANICAL_VENT −2.5 0.00 
 HALLMARK_INTERFERON_ALPHA_RESPONSE −2.5 0.00 
 BROWNE_INTERFERON_RESPONSIVE_GENES −2.5 0.00 
 GO_RESPONSE_TO_TYPE_I_INTERFERON −2.5 0.00 
 ZHANG_INTERFERON_RESPONSE −2.4 0.00 
Differentiation JAATINEN_HEMATOPOIETIC_STEM_CELL_DN −2 0.00 
 GO_MYELOID_LEUKOCYTE_DIFFERENTIATION −1.9 0.00 0.01 
 GAL_LEUKEMIC_STEM_CELL_DN −1.9 0.00 0.01 
 NG ET AL DN −1.9 0.00 
 GO_MYELOID_CELL_DIFFERENTIATION −1.8 0.00 0.03 
 GROUP 3 (DIFFERENTIATED) −1.7 0.01 0.01 
 IVANOVA_HEMATOPOIESIS_MATURE_CELL −1.6 0.01 0.03 
Gene sets enriched in samples with high NRASG12V 
Function Pathway NES Nom p FDR q 
Mitochondrial GO_MITOCHONDRIAL_PROTEIN_COMPLEX 1.60 0.00 0.25 
 GO_MITOCHONDRIAL_MATRIX 1.90 0.00 0.10 
 REACTOME_MITOCHONDRIAL_TRANSLATION 1.90 0.00 0.09 
 GO_MITOCHONDRIAL_TRANSLATIONAL_TERMINATION 2.00 0.00 0.10 
 GO_MITOCHONDRIAL_TRANSLATION 2.00 0.00 0.08 
 GO_MITOCHONDRIAL_GENE_EXPRESSION 2.00 0.00 0.10 
 HALLMARK_OXIDATIVE_PHOSPHORYLATION 1.80 0.00 0.13 
Senscence GO_CELLULAR_SENESCENCE 1.40 0.04 0.48 
Stem cell IVANOVA_HEMATOPOIESIS_STEM_CELL_LONG_TERM 1.40 0.04 0.06 
 IVANOVA_HEMATOPOIESIS_EARLY_PROGENITOR 1.60 0.01 0.01 
 SOMERVAILLE LSC UP 1.70 0.00 0.01 
 NG LSC UP 1.40 0.04 0.05 
Comparison of B11 samples treated with 1 μg/mL versus 10 μg/mL of Dox
FunctionPathwayNESNom pFDR q
Inflammation HECKER_IFNB1_TARGETS −2.7 0.00 
 BOSCO_INTERFERON_INDUCED_ANTIVIRAL_MODULE −2.5 0.00 
 ALTEMEIER_RESPONSE_TO_LPS_WITH_MECHANICAL_VENT −2.5 0.00 
 HALLMARK_INTERFERON_ALPHA_RESPONSE −2.5 0.00 
 BROWNE_INTERFERON_RESPONSIVE_GENES −2.5 0.00 
 GO_RESPONSE_TO_TYPE_I_INTERFERON −2.5 0.00 
 ZHANG_INTERFERON_RESPONSE −2.4 0.00 
Differentiation JAATINEN_HEMATOPOIETIC_STEM_CELL_DN −2 0.00 
 GO_MYELOID_LEUKOCYTE_DIFFERENTIATION −1.9 0.00 0.01 
 GAL_LEUKEMIC_STEM_CELL_DN −1.9 0.00 0.01 
 NG ET AL DN −1.9 0.00 
 GO_MYELOID_CELL_DIFFERENTIATION −1.8 0.00 0.03 
 GROUP 3 (DIFFERENTIATED) −1.7 0.01 0.01 
 IVANOVA_HEMATOPOIESIS_MATURE_CELL −1.6 0.01 0.03 
Gene sets enriched in samples with high NRASG12V 
Function Pathway NES Nom p FDR q 
Mitochondrial GO_MITOCHONDRIAL_PROTEIN_COMPLEX 1.60 0.00 0.25 
 GO_MITOCHONDRIAL_MATRIX 1.90 0.00 0.10 
 REACTOME_MITOCHONDRIAL_TRANSLATION 1.90 0.00 0.09 
 GO_MITOCHONDRIAL_TRANSLATIONAL_TERMINATION 2.00 0.00 0.10 
 GO_MITOCHONDRIAL_TRANSLATION 2.00 0.00 0.08 
 GO_MITOCHONDRIAL_GENE_EXPRESSION 2.00 0.00 0.10 
 HALLMARK_OXIDATIVE_PHOSPHORYLATION 1.80 0.00 0.13 
Senscence GO_CELLULAR_SENESCENCE 1.40 0.04 0.48 
Stem cell IVANOVA_HEMATOPOIESIS_STEM_CELL_LONG_TERM 1.40 0.04 0.06 
 IVANOVA_HEMATOPOIESIS_EARLY_PROGENITOR 1.60 0.01 0.01 
 SOMERVAILLE LSC UP 1.70 0.00 0.01 
 NG LSC UP 1.40 0.04 0.05 

Effect of oncogenic NRAS dose on intracellular signaling

To assess the impact of oncogenic NRAS dose on the activation of downstream signaling pathways, we analyzed B11 and G6 cells grown with Dox. Levels of intracellular signaling proteins were assessed by CyTOF (Fig. 5BE; Supplementary Fig. S3). We identified canonical RAS-effectors whose expression levels correlated with NRASG12V levels including phospho-ERK1/2 (pERK), p4EBP1, pAKT, β-catenin, and SHP2. Notably, β-catenin is a well-recognized mediator of leukemia stemness (41–43). In contrast, we identified additional canonical RAS-effectors that were NRASG12V-dependent, but not dose-responsive. These molecules, pSTAT1, pSTAT5, and pP38, increased in response to NRASG12V expression, but their levels did not increase with increasing NRASG12V levels (Fig. 5D–E). Together, these data show that high levels NRASG12V expression drives dose-dependent increases in some but not all downstream mediators and suggest that the effects of these mediators on self-renewal and proliferation may be dose-dependent as well.

Using a genetically engineered AML cell line, we showed that the level of NRASG12V oncoprotein is a critical determinant of its biological effects on AML cells. We found that moderate levels of NRASG12V lead to maximal proliferation, cell-cycle progression, and markers of apoptosis and senescence whereas higher levels impede proliferation and increase self-renewal capacity. Varying NRASG12V levels induce different transcriptional programs. Activation of specific signal transduction pathways parallels NRASG12V activity levels and likely mediate the differential effects of oncogenic NRAS in leukemia. These data suggest that the functional contribution of oncogenic NRAS to leukemia behavior is determined by the level of oncogenic NRAS activity. These experiments were limited to a single cell line and genetically modified derivatives of that cell line. However, these results are consistent with recent work in an Mll-AF9/NRASG12V murine model of AML (12). These findings suggest that some intraleukemic phenotypic heterogeneity could be explained by varying levels of the NRAS oncoprotein.

Gene amplification is a well-known mechanism to activate oncogenes in leukemia and solid tumors (44). Most notably, MYC, EGFR, and negative regulators of TP53 are frequently activated by gene amplifications (44). RAS genes are most frequently mutated, but activation via amplification has been described in a variety of tumor types (15). In AML, copy-number variations, including gene duplications and amplifications, have been well described in many genes, including NRAS, and has been associated with uniparental disomy (9, 15, 16). In diseases with targetable mutations, such as chronic myelogenous leukemia, amplification of the disease-defining BCR-ABL1 fusion gene is a well-described feature of disease progression and mechanism of drug resistance (45). Indeed, we found that THP-1 cells harbor five copies of NRAS. Despite the well-recognized phenomenon of oncogene amplification, the impact of oncogene dose and oncoprotein level on cellular behavior has not been well-described. In mice, a heterozygous NRAS mutation enforces chronic myeloid neoplasms, whereas homozygous NRAS mutations lead to an aggressive myeloid neoplasm associated with loss of mature myeloid effector cells, enhanced growth, and cytokine independence (17). Furthermore, NRAS mutant AML was shown to express higher levels of NRAS transcript (17). These data show that higher NRAS gene dose and transcript levels contribute to leukemia behavior. Our previous work, using single-cell RNA-seq, revealed that oncogene transcript levels vary with functionally and transcriptionally distinct leukemia stem cell subpopulations, implicating oncogene dose as a determinant of distinct leukemic functions. In this study, we used a genetically engineered AML cell line to demonstrate that altering levels of activated NRAS oncoprotein directly impacts transcription and signaling pathway activation status, which regulate the functional repertoire in AML. These data show that oncoprotein level is a determinant of specific oncogene functions.

Several studies have shown that self-renewal and rapid proliferation are mutually exclusive functions in normal and malignant stem cells. Activated NRAS was shown to confer self-renewal capacity to HSCs. Using in vivo proliferation tracing, the capacity to establish hematopoiesis (i.e., self-renewal) was limited to the poorly proliferative HSC subset (10). Single-cell transcriptional profiling, coupled with functional assays confirmed these findings in HSCs (46) and LSCs (12). These studies reported the transcriptional features of these functionally distinct stem cell subsets. In our previous single-cell transcriptional study, we found that oncogenic NRAS transcript levels were significantly higher in the self-renewing subgroup, relative to the proliferative subgroup (11), but we did not demonstrate the causative role of NRASG12V levels in the functional capacity of these subsets. In this report, we show that NRASG12V-oncoprotein levels regulate whether a cell has enhanced proliferation or self-renewal capacity.

Previous studies have linked relative quiescence to self-renewal capacity (47–50). In contrast to bulk AML cells which are most notable for their high proliferative rates, LSCs are demonstrably poorly proliferative (50). The relative quiescence of LSCs renders them less sensitive to anti-proliferative therapy (such as standard chemotherapy; ref. 51). Indeed, we and others confirmed that LSCs are poorly proliferative and specifically express quiescent gene expression profiles at the single-cell level (12, 46). In contrast to the well-established role of activated NRAS in inducing rapid proliferation, our single-cell study revealed that high levels of NRASG12V transgene are associated with a poorly proliferative leukemia subset and quiescent gene expression. In this study, we show that high levels of NRASG12V induce more markers of quiescence than moderate NRASG12V levels and demonstrate that the well-known role of mitogenic oncogenes in promoting proliferation may be limited by oncogene dose.

Previous work has established metabolic alterations in AML and LSCs. Our transcriptional profiling revealed a role for NRASG12V levels in modulating leukemic metabolism. In our study, high NRASG12V levels led to an enrichment in gene sets of mitochondrial function and oxidative phosphorylation. In contrast to bulk leukemia cells that can induce both glycolysis and mitochondrial respiration to generate energy, LSCs demonstrate a reverse Warburg effect and are fully dependent on oxidative phosphorylation (40, 52). Chemoresistance in LSCs is dependent on mitochondrial energy production (53). The BCL2 inhibitor, venetoclax, targets LSCs by disrupting mitochondrial function and oxidative phosphorylation (40, 52, 54). Venetoclax is highly effective in the frontline treatment of AML, but has poor efficacy in RAS mutant AML (55). Together, these data suggest that mutant NRAS may facilitate a metabolic switch that confers self-renewal capacity but cannot be fully overcome by venetoclax. Future work could test whether targeting the NRASG12V-effectors whose levels correlate with NRASG12V protein levels might sensitize NRAS mutant AML to BCL2 inhibition.

In summary, our work defines a role for variable oncoprotein levels in modulating leukemia functions and elucidates the selective advantage conferred by oncogene amplification frequently seen in AML and other malignancies. Further work could test whether therapeutic targeting of NRAS-activated pathways may exert distinct effects on proliferation and self-renewal in AML. The modified AML cell lines we developed could be used in future drug screens designed to discriminate effects on self-renewal and proliferation in AML.

D.A. Largaespada reports grants from American Cancer Society during the conduct of the study; grants from Genentech, Inc., other support from Luminary Therapeutics, Inc., Recombinetics, Inc., Makana Therapeutics, Inc., ImmuSoft, Inc., Styx Biotechnologies, Inc., and NeoClone Biotechnologies International; personal fees from Bio-Techne, Inc., outside the submitted work. No disclosures were reported by the other authors.

M. Kurata: Conceptualization, formal analysis, investigation, methodology, writing–original draft, writing–review and editing. M.L. Antony: Formal analysis, investigation, methodology, writing–review and editing. K.E. Noble-Orcutt: Investigation, methodology, writing–review and editing. S.K. Rathe: Formal analysis, writing–review and editing. Y. Lee: Formal analysis. H. Furuno: Investigation, methodology. S. Ishibashi: Investigation, methodology. M. Ikeda: Investigation, methodology. K. Yamamoto: Investigation, methodology. M. Kitagawa: Investigation, methodology. D.A. Largaespada: Conceptualization, resources, supervision, funding acquisition, writing–review and editing. Z. Sachs: Conceptualization, resources, data curation, software, formal analysis, supervision, methodology, writing–original draft, writing–review and editing.

Dr. Kyle Williams and Mr. Rory Williams helped the analysis of active Ras Pull-Down assay. Funding for this project was provided by The Leukemia & Lymphoma Society (grant no. 7019-04), Grant-in-Aid for Scientific Research (C; 18K06953) from the Japan Society for the Promotion of Science, and Kawano Masanori Memorial Foundation for Promotion of Pediatrics. Z. Sachs was supported by the American Cancer Society, Frederick A. DeLuca Foundation, Mentored Research Scholar Grant (MRSG-16-195-01-DDC); the Lois and Richard King Assistant Professorship in Medicine at the University of Minnesota, the Clinical and Translational Science Institute at the University of Minnesota KL2 Career Development Award and K to R01 Award (NIH/NCATS ULI RR033183 and KL2 RR0333182); the American Cancer Society Institutional Research Grant at the University of Minnesota (124166-IRG-58-001-55-IRG12); the Masonic Cancer Center at the University of Minnesota Translational Working Group Award and Genetic Mechanisms of Cancer Award; the University of Minnesota Department of Medicine Women's Early Research Career Award; the division of Hematology, Oncology, and Transplantation, Department of Medicine; and the University of Minnesota Foundation donors. We extend our thanks to the many University of Minnesota resources involved in our project: the University of Minnesota Genomics Center provided services for Sanger and next-generation sequencing; the Flow Cytometry Resource and other services of the Masonic Cancer Center (which is supported by NIH P30 CA77598); the Mass Cytometry Shared Resource at the University of Minnesota which is supported by the Office of the Vice President for Research at the University of Minnesota, and the Minnesota Supercomputing Institute.

The publication costs of this article were defrayed in part by the payment of publication fees. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.

Note: Supplementary data for this article are available at Molecular Cancer Research Online (http://mcr.aacrjournals.org/).

1.
Cancer Genome Atlas Research Network
.
Integrated genomic characterization of pancreatic ductal adenocarcinoma
.
Cancer Cell
2017
;
32
:
185
203
.
2.
Cancer Genome Atlas Research Network
.
Genomic classification of cutaneous melanoma
.
Cell
2015
;
161
:
1681
96
.
3.
Cancer Genome Atlas Research Network
.
Comprehensive molecular profiling of lung adenocarcinoma
.
Nature
2014
;
511
:
543
50
.
4.
Hobbs
GA
,
Der
CJ
,
Rossman
KL
.
RAS isoforms and mutations in cancer at a glance
.
J Cell Sci
2016
;
129
:
1287
92
.
5.
Tyner
JW
,
Tognon
CE
,
Bottomly
D
,
Wilmot
B
,
Kurtz
SE
,
Savage
SL
, et al
.
Functional genomic landscape of acute myeloid leukaemia
.
Nature
2018
;
562
:
526
31
.
6.
McMahon
CM
,
Ferng
T
,
Canaani
J
,
Wang
ES
,
Morrissette
JJD
,
Eastburn
DJ
, et al
.
Clonal selection with RAS pathway activation mediates secondary clinical resistance to selective FLT3 inhibition in acute myeloid leukemia
.
Cancer Discov
2019
;
9
:
1050
63
.
7.
Ferng
TT
,
Terada
D
,
Ando
M
,
Tarver
TC
,
Chaudhary
F
,
Lin
KC
, et al
.
The irreversible FLT3 inhibitor FF-10101 is active against a diversity of FLT3 inhibitor resistance mechanisms
.
Mol Cancer Ther
2022
;
21
:
844
54
.
8.
Zhang
H
,
Savage
S
,
Schultz
AR
,
Bottomly
D
,
White
L
,
Segerdell
E
, et al
.
Clinical resistance to crenolanib in acute myeloid leukemia due to diverse molecular mechanisms
.
Nat Commun
2019
;
10
:
244
.
9.
Network CGAR
.
Genomic and epigenomic landscapes of adult de novo acute myeloid leukemia
.
N Engl J Med
2013
;
368
:
2059
74
.
10.
Li
Q
,
Bohin
N
,
Wen
T
,
Ng
V
,
Magee
J
,
Chen
SC
, et al
.
Oncogenic Nras has bimodal effects on stem cells that sustainably increase competitiveness
.
Nature
2013
;
504
:
143
7
.
11.
Sachs
Z
,
LaRue
RS
,
Nguyen
HT
,
Sachs
K
,
Noble
KE
,
Mohd Hassan
NA
, et al
.
NRASG12V oncogene facilitates self-renewal in a murine model of acute myelogenous leukemia
.
Blood
2014
;
124
:
3274
83
.
12.
Sachs
K
,
Sarver
AL
,
Noble-Orcutt
KE
,
LaRue
RS
,
Antony
ML
,
Chang
D
, et al
.
Single-cell gene expression analyses reveal distinct self-renewing and proliferating subsets in the leukemia stem cell compartment in acute myeloid leukemia
.
Cancer Res
2020
;
80
:
458
70
.
13.
Finney
RE
,
Bishop
JM
.
Predisposition to neoplastic transformation caused by gene replacement of H-ras1
.
Science
1993
;
260
:
1524
7
.
14.
Soh
J
,
Okumura
N
,
Lockwood
WW
,
Yamamoto
H
,
Shigematsu
H
,
Zhang
W
, et al
.
Oncogene mutations, copy number gains and mutant allele specific imbalance (MASI) frequently occur together in tumor cells
.
PLoS One
2009
;
4
:
e7464
.
15.
Dunbar
AJ
,
Gondek
LP
,
O'Keefe
CL
,
Makishima
H
,
Rataul
MS
,
Szpurka
H
, et al
.
250K single nucleotide polymorphism array karyotyping identifies acquired uniparental disomy and homozygous mutations, including novel missense substitutions of c-Cbl, in myeloid malignancies
.
Cancer Res
2008
;
68
:
10349
57
.
16.
Matsuda
K
,
Nakazawa
Y
,
Sakashita
K
,
Shiohara
M
,
Yamauchi
K
,
Koike
K
.
Acquisition of loss of the wild-type NRAS locus with aggressive disease progression in a patient with juvenile myelomonocytic leukemia and a heterozygous NRAS mutation
.
Haematologica
2007
;
92
:
1576
8
.
17.
Xu
J
,
Haigis
KM
,
Firestone
AJ
,
McNerney
ME
,
Li
Q
,
Davis
E
, et al
.
Dominant role of oncogene dosage and absence of tumor suppressor activity in Nras-driven hematopoietic transformation
.
Cancer Discov
2013
;
3
:
993
1001
.
18.
Wang
J
,
Liu
Y
,
Li
Z
,
Wang
Z
,
Tan
LX
,
Ryu
MJ
, et al
.
Endogenous oncogenic Nras mutation initiates hematopoietic malignancies in a dose- and cell type-dependent manner
.
Blood
2011
;
118
:
368
79
.
19.
Janssen
JW
,
Steenvoorden
AC
,
Lyons
J
,
Anger
B
,
Böhlke
JU
,
Bos
JL
, et al
.
RAS gene mutations in acute and chronic myelocytic leukemias, chronic myeloproliferative disorders, and myelodysplastic syndromes
.
Proc Natl Acad Sci U S A
1987
;
84
:
9228
32
.
20.
Kurata
M
,
Rathe
SK
,
Bailey
NJ
,
Aumann
NK
,
Jones
JM
,
Veldhuijzen
GW
, et al
.
Using genome-wide CRISPR library screening with library resistant DCK to find new sources of Ara-C drug resistance in AML
.
Sci Rep
2016
;
6
:
36199
.
21.
Kim
D
,
Langmead
B
,
Salzberg
SL
.
HISAT: a fast spliced aligner with low memory requirements
.
Nat Methods
2015
;
12
:
357
60
.
22.
Trapnell
C
,
Williams
BA
,
Pertea
G,
Mortazavi
A
,
Kwan
G
,
van Baren
MJ
, et al
.
Transcript assembly and quantification by RNA-seq reveals unannotated transcripts and isoform switching during cell differentiation
.
Nat Biotechnol
2010
;
28
:
511
5
.
23.
Kersey
PJ
,
Allen
JE
,
Allot
A
,
Barba
M
,
Boddu
S
,
Bolt
BJ
, et al
.
Ensembl Genomes 2018: an integrated omics infrastructure for non-vertebrate species
.
Nucleic Acids Res
2018
;
46
:
D802
-D8.
24.
Thorvaldsdóttir
H
,
Robinson
JT
,
Mesirov
JP
.
Integrative Genomics Viewer (IGV): high-performance genomics data visualization and exploration
.
Briefings Bioinf
2013
;
14
:
178
92
.
25.
Robinson
MD
,
McCarthy
DJ
,
Smyth
GK
.
edgeR: a Bioconductor package for differential expression analysis of digital gene expression data
.
Bioinformatics
2010
;
26
:
139
40
.
26.
Subramanian
A
,
Tamayo
P
,
Mootha
VK
,
Mukherjee
S
,
Ebert
BL
,
Gillette
MA
, et al
.
Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles
.
Proc Natl Acad Sci U S A
2005
;
102
:
15545
50
.
27.
Mootha
VK
,
Lindgren
CM
,
Eriksson
KF
,
Subramanian
A
,
Sihag
S
,
Lehar
J
, et al
.
PGC-1alpha-responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes
.
Nat Genet
2003
;
34
:
267
73
.
28.
Baughn
LB
,
Sachs
Z
,
Noble-Orcutt
KE
,
Mitra
A
,
Van Ness
BG
,
Linden
MA
.
Phenotypic and functional characterization of a bortezomib-resistant multiple myeloma cell line by flow and mass cytometry
.
Leuk Lymphoma
2017
;
58
:
1931
40
.
29.
Behbehani
GK
.
Cell cycle analysis by mass cytometry
.
Methods Mol Biol
2018
;
1686
:
105
24
.
30.
Behbehani
GK
,
Bendall
SC
,
Clutter
MR
,
Fantl
WJ
,
Nolan
GP
.
Single-cell mass cytometry adapted to measurements of the cell cycle
.
Cytometry A
2012
;
81
:
552
66
.
31.
Gao
J
,
Aksoy
BA
,
Dogrusoz
U
,
Dresdner
G
,
Gross
B
,
Sumer
SO
, et al
.
Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal
.
Sci Signal
2013
;
6
:
pl1
.
32.
Cerami
E
,
Gao
J
,
Dogrusoz
U
,
Gross
BE
,
Sumer
SO
,
Aksoy
BA
, et al
.
The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data
.
Cancer Discov
2012
;
2
:
401
4
.
33.
Cancer Genome Atlas Research Network
,
Ley
TJ
,
Miller
C
,
Ding
L
,
Raphael
BJ
,
Mungall
AJ
, et al
.
Genomic and epigenomic landscapes of adult de novo acute myeloid leukemia
.
N Engl J Med
2013
;
368
:
2059
74
.
34.
Potu
H
,
Peterson
LF
,
Kandarpa
M
,
Pal
A
,
Sun
H
,
Durham
A
, et al
.
Usp9x regulates Ets-1 ubiquitination and stability to control NRAS expression and tumorigenicity in melanoma
.
Nat Commun
2017
;
8
:
14449
.
35.
Williams
DA
,
Zheng
Y
,
Cancelas
JA
.
Rho GTPases and regulation of hematopoietic stem cell localization
.
Methods Enzymol
2008
;
439
:
365
93
.
36.
Taylor
SJ
,
Resnick
RJ
,
Shalloway
D
.
Nonradioactive determination of Ras-GTP levels using activated ras interaction assay
.
Methods Enzymol
2001
;
333
:
333
42
.
37.
Crescenzi
E
,
Pacifico
F
,
Lavorgna
A
,
De Palma
R
,
D'Aiuto
E
,
Palumbo
G
, et al
.
NF-κB-dependent cytokine secretion controls Fas expression on chemotherapy-induced premature senescent tumor cells
.
Oncogene
2011
;
30
:
2707
17
.
38.
Cox
AD
,
Der
CJ
.
The dark side of Ras: regulation of apoptosis
.
Oncogene
2003
;
22
:
8999
9006
.
39.
Jang
J
,
Lee
J
,
Jang
JH
,
Jung
CW
,
Park
S
.
Anti-leukemic effects of simvastatin on NRAS
.
Mol Biol Rep
2019
;
46
:
5859
66
.
40.
Lagadinou
ED
,
Sach
A
,
Callahan
K
,
Rossi
RM
,
Neering
SJ
,
Minhajuddin
M
, et al
.
BCL-2 inhibition targets oxidative phosphorylation and selectively eradicates quiescent human leukemia stem cells
.
Cell Stem Cell
2013
;
12
:
329
41
.
41.
Griffiths
EA
,
Golding
MC
,
Srivastava
P
,
Povinelli
BJ
,
James
SR
,
Ford
LA
, et al
.
Pharmacological targeting of β-catenin in normal karyotype acute myeloid leukemia blasts
.
Haematologica
2015
;
100
:
e49
52
.
42.
Luis
TC
,
Ichii
M
,
Brugman
MH
,
Kincade
P
,
Staal
FJ
.
Wnt signaling strength regulates normal hematopoiesis and its deregulation is involved in leukemia development
.
Leukemia
2012
;
26
:
414
21
.
43.
Wang
Y
,
Krivtsov
AV
,
Sinha
AU
,
North
TE
,
Goessling
W
,
Feng
Z
, et al
.
The Wnt/beta-catenin pathway is required for the development of leukemia stem cells in AML
.
Science
2010
;
327
:
1650
3
.
44.
Bagci
O
,
Kurtgöz
S
.
Amplification of cellular oncogenes in solid tumors
.
N Am J Med Sci
2015
;
7
:
341
6
.
45.
Chandran
RK
,
Geetha
N
,
Sakthivel
KM
,
Aswathy
CG
,
Gopinath
P
,
Raj
TVA
, et al
.
Genomic amplification of BCR-ABL1 fusion gene and its impact on the disease progression mechanism in patients with chronic myelogenous leukemia
.
Gene
2019
;
686
:
85
91
.
46.
Wilson
NK
,
Kent
DG
,
Buettner
F
,
Shehata
M
,
Macaulay
IC
,
Calero-Nieto
FJ
, et al
.
Combined single-cell functional and gene expression analysis resolves heterogeneity within stem cell populations
.
Cell Stem Cell
2015
;
16
:
712
24
.
47.
Hope
KJ
,
Jin
L
,
Dick
JE
.
Acute myeloid leukemia originates from a hierarchy of leukemic stem cell classes that differ in self-renewal capacity
.
Nat Immunol
2004
;
5
:
738
43
.
48.
Behbehani
GK
,
Samusik
N
,
Bjornson
ZB
,
Fantl
WJ
,
Medeiros
BC
,
Nolan
GP
.
Mass cytometric functional profiling of acute myeloid leukemia defines cell-cycle and immunophenotypic properties that correlate with known responses to therapy
.
Cancer Discov
2015
;
5
:
988
1003
.
49.
Guan
Y
,
Gerhard
B
,
Hogge
DE
.
Detection, isolation, and stimulation of quiescent primitive leukemic progenitor cells from patients with acute myeloid leukemia (AML)
.
Blood
2003
;
101
:
3142
9
.
50.
Ishikawa
F
,
Yoshida
S
,
Saito
Y
,
Hijikata
A
,
Kitamura
H
,
Tanaka
S
, et al
.
Chemotherapy-resistant human AML stem cells home to and engraft within the bone-marrow endosteal region
.
Nat Biotechnol
2007
;
25
:
1315
21
.
51.
Costello
RT
,
Mallet
F
,
Gaugler
B
,
Sainty
D
,
Arnoulet
C
,
Gastaut
JA
, et al
.
Human acute myeloid leukemia CD34+/CD38− progenitor cells have decreased sensitivity to chemotherapy and Fas-induced apoptosis, reduced immunogenicity, and impaired dendritic cell transformation capacities
.
Cancer Res
2000
;
60
:
4403
11
.
52.
Sriskanthadevan
S
,
Jeyaraju
DV
,
Chung
TE
,
Prabha
S
,
Xu
W
,
Skrtic
M
, et al
.
AML cells have low spare reserve capacity in their respiratory chain that renders them susceptible to oxidative metabolic stress
.
Blood
2015
;
125
:
2120
30
.
53.
Farge
T
,
Saland
E
,
de Toni
F
,
Aroua
N
,
Hosseini
M
,
Perry
R
, et al
.
Chemotherapy-resistant human acute myeloid leukemia cells are not enriched for leukemic stem cells but require oxidative metabolism
.
Cancer Discov
2017
;
7
:
716
35
.
54.
Pollyea
DA
,
Stevens
BM
,
Jones
CL
,
Winters
A
,
Pei
S
,
Minhajuddin
M
, et al
.
Venetoclax with azacitidine disrupts energy metabolism and targets leukemia stem cells in patients with acute myeloid leukemia
.
Nat Med
2018
;
24
:
1859
66
.
55.
DiNardo
CD
,
Tiong
IS
,
Quaglieri
A
,
MacRaild
S
,
Loghavi
S
,
Brown
FC
, et al
.
Molecular patterns of response and treatment failure after frontline venetoclax combinations in older patients with AML
.
Blood
2020
;
135
:
791
803
.
56.
Amir
eA
,
Davis
KL
,
Tadmor
MD
,
Simonds
EF
,
Levine
JH
,
Bendall
SC
, et al
.
viSNE enables visualization of high dimensional single-cell data and reveals phenotypic heterogeneity of leukemia
.
Nat Biotechnol
2013
;
31
:
545
52
.
57.
Bendall
SC
,
Simonds
EF
,
Qiu
P
,
Amir
eA
,
Krutzik
PO
,
Finck
R
, et al
.
Single-cell mass cytometry of differential immune and drug responses across a human hematopoietic continuum
.
Science
2011
;
332
:
687
96
.

Supplementary data