The KRAS-G12D mutation induces metabolic vulnerability in B-cell acute lymphoblastic leukemia

Abstract

Mutations in RAS pathway genes are highly prevalent in acute lymphoblastic leukemia (ALL). However, the effects of RAS mutations on ALL cell growth have not been experimentally characterized, and effective RAS-targeting therapies are being sought after. Here, we found that Reh ALL cells bearing the KRAS-G12D mutation showed increased proliferation rates in vitro but displayed severely compromised growth in mice. Exploring this divergence, proliferation assays with multiple ALL cell lines revealed that the KRAS-G12D rewired methionine and arginine metabolism. Isotope tracing results showed that KRAS-G12D promotes catabolism of methionine and arginine to support anabolism of polyamines and proline, respectively. Chemical inhibition of polyamine biosynthesis selectively killed KRAS-G12D B-ALL cells. Finally, chemically inhibiting AKT/mTOR signaling abrogated the altered amino acid metabolism and strongly promoted the in vivo growth of KRAS-G12D cells in B-ALL xenograft. Our study thus illustrates how hyperactivated AKT/mTOR signaling exerts distinct impacts on hematological malignancies vs. solid tumors.

Keywords: Biochemistry; Biological sciences; Molecular biology.

Graphical abstract

Figure 1

B-ALL cells expressing the KRAS-G12D mutant display compromised growth under nutrient-limited conditions (A) Growth curves of various Reh cells cultured in normal RPMI 1640 medium (as the “nutrient-proficient” condition). Data are shown as mean ± SD ∗∗∗: p < 0.005; two-tailed Student’s t-tests. Ctrl: control Reh cells, RASmt: Reh cells expressing KRAS-G12D. (B) The in vivo growth kinetics of various Reh xenograft cells. Cells were isolated from xenograft tibias and counted at different time points (day 7, 14, and 21) after tail-vein injection; at each time point, n = 3, the numbers of GFP⁺ Reh cells were determined by flow cytometry. Data are shown as the mean. ∗: p < 0.05, ∗∗∗: p < 0.005; two-tailed Student’s t-tests. (C) The tumor burden of various Reh xenografts (the percentage of GFP-labeled Reh cells in total bone marrow cells) at the indicated time points after tail-vein injection; at each time point, n = 3. Data are shown as the mean ± SD ∗∗: p < 0.01, ∗: p < 0.05; two-tailed Student’s t-tests. (D) Growth curves of various Reh cells cultured in RPMI 1640 medium containing low concentrations of nutrients (1 mM glucose and 30-fold diluted concentrations of amino acids); the “nutrient-limited” condition. Data are shown as the mean ± SD ∗∗∗: p < 0.005; two-tailed Student’s t-tests. (E) Growth curves of various BaF3 cells cultured in RPMI 1640 medium containing low concentrations of nutrients (1 mM glucose and 30-fold diluted concentrations of amino acids) and murine IL-3 (0.1 ng/mL); the “nutrient-limited” condition. Data are shown as the mean ± SD ∗∗∗: p < 0.005; two-tailed Student’s t-tests. Ctrl: control BaF3 cells, RASmt: BaF3 cells expressing KRAS-G12D.F. Growth curves of various Reh cells cultured in HPLM medium. Data are shown as mean ± SD ∗∗∗: p < 0.005; two-tailed Student’s t-tests. Ctrl: control Reh cells, RASmt: Reh cells expressing KRAS-G12D. (F) Growth curves of various Reh cells cultured in HPLM medium. Data are shown as mean ± s.d. ∗∗∗: p<0.005; two-tailed Student’s t-tests. Ctrl: control Reh cells, RASmt: Reh cells expressing KRAS-G12D.

Figure 2

The KRAS-G12D mutation sensitizes B-ALL cells to extracellular amino acids (A) The viabilities of various Reh cells grown in media containing different concentrations of glucose for 72 h. 10 mM is the standard glucose concentration in normal RPMI 1640 medium. Data are shown as the mean ± SD ns: p > 0.05, ∗: p < 0.05, ∗∗∗: p < 0.005; two-tailed Student’s t-tests. Ctrl: control Reh cells, RASmt: Reh cells expressing KRAS-G12D. (B and C) The viabilities of various Reh (B) or BaF3 (C) cells grown in media with different dilution ratios of a total 20 amino acids mixture for 72 h. BaF3 cells were cultured in media with a low concentration of IL-3 (0.1 ng/mL) for 72 h. Data are shown as the mean ± SD ∗∗: p < 0.01, ∗∗∗: p < 0.005; two-tailed Student’s t-tests. Ctrl: control cells, RASmt: cells expressing KRAS-G12D. (D) Growth curves for various Reh cells cultured in amino-acid-free RPMI 1640 medium. Data are shown as the mean ± SD. (E) Growth curves of various BaF3 cells cultured in amino-acid-free RPMI 1640 medium. Data are shown as the mean ± SD. (F) Apoptosis levels of various Reh cells with or without total amino acid starvation for the indicated times. The apoptosis levels were determined by Annexin V and PI staining; the numbers represent the proportions of Annexin V positive cells. See also Figure S1.

Figure 3

Growth of KRAS-G12D B-ALL cells can be decreased by Met and Arg deprivation (A) Heatmap indicating the fold changes (FC) for the intracellular levels of the indicated amino acids in various Reh cells grown in normal media. Ctrl: control Reh cells, RASmt: Reh cells expressing KRAS-G12D. (B) “Consumption levels” of various Reh cells grown in “normal condition” medium; the consumption level was determined as difference between each amino acid concentration in the medium with Reh cells and the concentration in the medium without cells after 24 h culture. All amino acids with increased consumption in RASmt cells are shown. Data are shown as the mean ± SD ns: p > 0.05, ∗: p < 0.05, ∗∗: p < 0.01, ∗∗∗: p < 0.005; two-tailed Student’s t-tests. (C) The relative concentrations of the indicated amino acids in mouse bone marrow matrix solutions, extracted from tibias and femurs of various xenografts, n = 3. The cell-free bone marrow matrix was extracted when the tumor burdens of xenografts are approximately equal (25%–35%). All amino acids with decreased levels in bone marrow of mice injected with RAS-mutant cells are shown. Data are shown as the mean ± SD ∗: p < 0.05, ∗∗: p < 0.01, ∗∗∗: p < 0.005; two-tailed Student’s t-tests. Blank: mice without Reh cells injection, Ctrl: mice injected with control Reh cells, RASmt: mice injected with RAS-mutant cells. (D) A Venn diagram indicating the overlap for results from Figures 3A–3C. (E) The viabilities of various Reh cells grown in medium with different concentrations of Met for 72 h. 100.7 μM is the standard concentration of Met in the normal RPMI 1640 medium. Data are shown as the mean ± SD ∗∗∗: p < 0.005; two-tailed Student’s t-tests. (F) The viabilities of various Reh cells grown in medium with different concentrations of Arg for 72 h. 1.1 mM is the standard concentration of Arg in normal RPMI 1640 medium. Data are shown as the mean ± SD ∗∗∗: p < 0.005; two-tailed Student’s t-tests. (G) The viabilities of the various BaF3 cells grown in media with different concentrations of Met for 72 h. BaF3 cells were cultured in media with a low concentration of IL-3 (0.1 ng/mL). 100.7 μM is the standard concentration of Met in normal RPMI 1640 medium. Data are shown as the mean ± SD ∗∗∗: p < 0.005; two-tailed Student’s t-tests. (H) The viabilities of various BaF3 cells grown in media with different concentrations of Arg for 72 h. BaF3 cells were cultured in media with a low concentration of IL-3 (0.1 ng/mL). 1.1 mM is the standard concentration of Arg in normal RPMI-1640 medium. Data are shown as the mean ± SD ∗∗∗: p < 0.005; two-tailed Student’s t-tests. Ctrl: control BaF3 cells, RASmt: BaF3 cells expressing KRAS-G12D. (I) The viabilities of various Reh cells grown in medium with different concentrations of Cys for 72 hr. 207.7 μM is the standard concentration of Cys in the normal RPMI 1640 medium. Data are shown as the mean ± s.d. ∗: p<0.05; two-tailed Student’s t-tests.See also Figure S2.

Figure 4

KRAS-G12D increased Met and Arg catabolism to support anabolism of polyamines and proline, respectively, in B-ALL cells (A) Immunoblotting with antibodies against monomethyl lysine (mme-K), di-methyl lysine (dme-K), and tri-methyl lysine (tme-K) of the methylation levels of total proteins in control (Ctrl) Reh cells or Reh cells expressing KRAS-G12D (RASmt). (B) Immunoblot showing methylation levels of H3 histones in Ctrl and RASmt Reh cells. (C) The intracellular levels of S-adenosyl-L-methionine (SAM; a downstream metabolite of methionine) in Ctrl and RASmt Reh cells. Data are shown as the mean ± SD ns: p > 0.05; two-tailed Student’s t-tests. (D) The intracellular levels of 5′-methylthioadenosin (MTA; another downstream metabolite of methionine) in Ctrl and RASmt Reh cells. Data are shown as the mean ± SD ∗∗∗: p < 0.005; two-tailed Student’s t-tests. (E) A methionine metabolism schematic related to the isotope tracing experiment using ¹³C₅-Met and assessing polyamine biosynthesis. (F) A heatmap indicating fold changes in the intracellular levels of labeled metabolites after ¹³C₅-Met incubation for 4 h. (G) A schematic for isotope tracing with ¹³C₆-Arg to assess proline biosynthesis. (H and I) The intracellular levels of labeled proline and putrescine in Reh cells (H) or BaF3 cells (I) after ¹³C₆-Arg incubation for 5h. Data are shown as the mean ± SD ∗: p < 0.05, ∗∗∗: p < 0.005, ns: p > 0.05; two-tailed Student’s t-tests. (J) A heatmap indicating fold changes in relative expression of key enzyme genes associated with Arg metabolism in indicated Reh cells according to the quantification by real-time PCR. (K) The viabilities of Reh cells pre-treated (or untreated) with the known arginase inhibitor BEC (1 mM) and grown in media with different concentrations of Arg for 72 h. 1.1 mM is the standard concentration of Arg in normal RPMI 1640 media. Data are shown as the mean ± SD ∗∗∗: p < 0.005 (RASmt vs. RASmt + BEC); two-tailed Student’s t-tests. See also Figure S3.

Figure 5

KRAS-G12D B-ALL cells are sensitive to killing by the polyamine biosynthesis inhibitor DFMO (A) The viabilities of various Reh cells supplied with a polyamine mixture (containing 1 μM putrescine, 1 μM spermidine, and 1 μM spermine) grown in media with different concentrations of Met for 72 h. 100.7 μM is the standard concentration of Met in normal RPMI 1640. Data are shown as the mean ± SD ∗∗: p < 0.01, ∗∗∗: p < 0.005 (RASmt vs. RASmt + polyamines); two-tailed Student’s t-tests. (B and C) Viabilities of Ctrl and RASmt Reh cells at increasing concentrations of the polyamine biosynthesis inhibitor DFMO (which targets the ODC1 enzyme) (B) and the IC₅₀ values (C). Data are shown as the mean ± SD ∗∗∗: p < 0.005; two-tailed Student’s t-tests. (D and E) Viabilities of Ctrl and RASmt BaF3 cells at increasing concentrations of the polyamine biosynthesis inhibitor DFMO (which targets the ODC1 enzyme) (D) and the IC₅₀ values (E). BaF3 cells were cultured in media with a low concentration of IL-3 (0.1 ng/mL). Data are shown as the mean ± SD ∗∗∗: p < 0.005; two-tailed Student’s t-tests.

Figure 6

KRAS-G12D B-ALL cells display activated mTOR signaling and chemical inhibition of mTOR rescues the growth defects of these cells in vivo (A) The phosphorylation levels of the indicated proteins (AKT, S6K1, S6, and GSK3) in Ctrl and RASmt Reh cells, implicating differential activity in the AKT/mTOR signaling pathways. (B) The protein levels of AMD1 in various Reh cells. Ctrl and RASmt cells were incubated with or without 500 nM AZD-8055 overnight and harvested for blotting. (C) The viabilities of Reh cells pre-treated (or untreated) with the mTOR inhibitor AZD-8055 for 12 h, and then grown in media with different dilution ratios of total 20 amino acids mixture for 72 h. Data are shown as the mean ± SD ns: p > 0.05, ∗∗∗: p < 0.005; two-tailed Student’s t-tests. Ctrl: control Reh cells, RASmt: Reh cells expressing KRAS-G12D. (D) The viabilities of Reh cells pre-treated (or untreated) with mTOR inhibitor AZD-8055 for 12 h, and then grown in media with the indicated dilution ratios of a Met/Arg mixture for 72 h. Data are shown as the mean ± SD ∗: p < 0.05, ∗∗: p < 0.01; two-tailed Student’s t-tests. (E) The cell numbers of various Reh cells in bone marrow xenografts (of tibias) on the 20^th day after tail-injection of 10⁷ Reh cells; mice were treated with the indicated dosage of AZD-8055 twice daily by oral gavage, n = 3. Data are shown as the mean ± SD ns: p > 0.05, ∗: p < 0.05; two-tailed Student’s t test. (F) A schematic showing of the KRAS-G12D-induced metabolic vulnerability mechanisms. KRAS-G12D could increase Arg and Met catabolism to support anabolism of Pro and polyamines, which led to reduced intracellular levels of Met and Arg and Met/Arg-related metabolic deficiencies. The ODC1 inhibitor DFMO inhibited polyamine biosynthesis and selectively killed KRAS-mutant cells. See also Figures S4–S6.