Synthetic Lethality of Wnt Pathway Activation and Asparaginase in Drug-Resistant Acute Leukemias

Abstract

Resistance to asparaginase, an antileukemic enzyme that depletes asparagine, is a common clinical problem. Using a genome-wide CRISPR/Cas9 screen, we found a synthetic lethal interaction between Wnt pathway activation and asparaginase in acute leukemias resistant to this enzyme. Wnt pathway activation induced asparaginase sensitivity in distinct treatment-resistant subtypes of acute leukemia, but not in normal hematopoietic progenitors. Sensitization to asparaginase was mediated by Wnt-dependent stabilization of proteins (Wnt/STOP), which inhibits glycogen synthase kinase 3 (GSK3)-dependent protein ubiquitination and proteasomal degradation, a catabolic source of asparagine. Inhibiting the alpha isoform of GSK3 phenocopied this effect, and pharmacologic GSK3α inhibition profoundly sensitized drug-resistant leukemias to asparaginase. Our findings provide a molecular rationale for activation of Wnt/STOP signaling to improve the therapeutic index of asparaginase.

Keywords: FBXW7; GSK3; Wnt signaling; acute leukemia; asparaginase; asparagine; drug resistance; proteasomal degradation; protein ubiquitination; synthetic lethality.

Figure 1.. Wnt Pathway Activation Sensitizes Leukemia Cells to Asparaginase

(A) CCRF-CEM cells were transduced with the GeCKO genome-wide guide RNA library in biologic duplicates, split into treatment with vehicle or asparaginase (10 U/L), and guide RNA representation was assessed after 5 days of treatment. (B) Significance of gene depletion in asparaginase-treated conditions, as assessed by robust ranking aggregation (RRA) score calculated using MAGeCK analysis. Note that microRNA genes are not shown. (C) CCRF-CEM cells were transduced with the indicated shRNAs, and knockdown efficiency was assessed by RT-PCR analysis. C_T values greater than 36 were defined as not detected (N.D.). (D) CCRF-CEM cells were transduced with the indicated shRNAs and subjected to Western blot analysis for active (nonphosphorylated) β-catenin or GAPDH. (E) CCRF-CEM cells transduced with a lentiviral TOPFlash-EGFP (7xTcf-EGFP) reporter of canonical Wnt/β-catenin driven transcription were transduced with the indicated shRNAs, and reporter-driven EGFP fluorescence was assessed. (F) The indicated cell lines were transduced with the indicated shRNAs and treated with the indicated doses of asparaginase. Relative viability was assessed after 8 days of treatment by counting viable cells. All cell counts were normalized to those in shLuc-transduced, no-asparaginase controls. (G) CCRF-CEM cells were transduced with the indicated shRNAs, treated with asparaginase (10 U/L) for 48 hr and caspase 3/7 activity assay was assessed. (H) CCRF-CEM cells were treated as indicated and the number of viable cells after 8 days of treatment was assessed. All cell counts were normalized to those in no-Wnt3a, no-asparaginase controls. All error bars represent SEM. See also Figures S1 and S2 and Tables S1 and S2.

Figure 2.. GSK3 Inhibition Sensitizes Distinct Acute Leukemia Subtypes, but not Normal Hematopoietic Progenitors, to Asparaginase-Induced Cytotoxicity

(A-B) T-ALL cells (A), AML cells (OCI-AML2, SKM-1, THP-1) or B-ALL cells (MHH-CALL2) (B) were treated with the GSK3 inhibitor CHIR99021 (CHIR, 1 μM) or vehicle control, together with the indicated doses of asparaginase for 8 days. Relative viability was assessed based on viable cell counts, all of which were normalized to those in vehicle-treated cells. Western blot analysis for the indicated proteins is shown as an inset. (C) Normal CD34⁺ human hematopoietic progenitor cells were treated with CHIR99021 (1 μM) or vehicle, together with the indicated doses of asparaginase, and viable cell counts were assessed after treatment for 4 days. Note that these normal hematopoietic progenitors could not be maintained for more than 4 days in culture. Western blot analysis for the indicated proteins is shown as an inset. (D) CCRF-CEM cells were treated with CHIR99021 (1 μM) or vehicle, and the indicated chemotherapeutic drugs for 8 days. All Western Blots shown indicate levels of active (nonphosphorylated) β-catenin or GAPDH after treatment with CHIR 99021 (1 μM) or vehicle. Viable cells were counted and results normalized to counts in no-CHIR, no-chemotherapy controls. Western blot analysis for the indicated proteins is shown as an inset. Statistical significance was calculated for the following doses of chemotherapy: 10 U/L asparaginase, 1 μM dexamethasone, 1 μM 6-mercaptopurine, 10 nM doxorubicin and 10 nM vincristine, respectively. Two-way ANOVA with Tukey post-hoc adjustment was performed for each cell line and included an interaction term between asparaginase dose and GSK3 inhibitor. The p value for the main effect of GSK3 inhibitor vs. vehicle is presented in each plot and all interaction terms were significant (p < 0.0001). Error bars represent SEM. See also Table S3.

Figure 3.. Wnt-induced sensitization to asparaginase is not mediated by β-catenin or mTOR activation.

(A) CCRF-CEM cells transduced with a constitutively active β-catenin were analyzed for expression of the indicated proteins by Western blot analysis (left), activity of a TOPFlash-EGFP reporter of β-catenin dependent transcriptional activity (mid), and transduction efficiency was assessed by immunostaining for β-catenin (right). Statistical significance was calculated using a two-sided Welch t-test. * p ≤ 0.05; ^** p ≤ 0.01; ^*** p ≤ 0.001; ^**** p ≤ 0.0001. n.s., p > 0.05. (B) CCRF-CEM cells transduced with the indicated constructs were treated with the indicated doses of asparaginase for 8 days, and the number of viable cells was counted. All cell counts were normalized to those in control-transduced, no-asparaginase cells. (C) CCRF-CEM cells transduced with the indicated shRNAs were analyzed by RT-PCR, and effects on TOPFlash-EGFP reporter activity were assessed. Statistical significance was calculated using a two-sided Welch t-test. * p ≤ 0.05; ^** p ≤ 0.01; ^*** p ≤ 0.001; ^**** p ≤ 0.0001. n.s., p > 0.05. (D) CCRF-CEM cells transduced with the indicated shRNAs were treated with the indicated doses of asparaginase for 8 days. Viability was assessed as in (B). (E) CCRF-CEM cells transduced with the indicated shRNAs were treated with vehicle, AZD2014 (100 nM), rapamycin (10 nM) or RAD001 (100 nM), and analyzed by Western blot analysis for the indicated proteins (top). Cells were treated with the indicated doses of asparaginase and viability was assessed as in (B). Note that mTOR inhibition did not rescue Wnt-induced sensitization to asparaginase. All error bars represent SEM.

Figure 4.. Wnt-Dependent Stabilization of Proteins Mediates Sensitization to Asparaginase.

(A) CCRF-CEM cells treated with vehicle or asparaginase (10 U/L) for 48 hr, and cell size was assessed by flow cytometry forward scatter height (FSC-H). (B) CCRF-CEM cells transduced with the indicated constructs were treated with asparaginase (10 U/L) for 48 hr, and cell size was assessed as in (A). Scatter plot (right) depicts results of individual biologic replicates, with horizontal bars indicating mean, and error bars indicating SEM. Differences between groups were analyzed using a one-way ANOVA with Dunnett’s adjustment for multiple comparisons. ^** p ≤ 0.01; ^*** p ≤ 0.001. (C) CCRF-CEM cells were treated with asparaginase (10 U/L) as well as either vehicle or Wnt3A (100 ng/ml) for 48 hr, and cell size was assessed as in− (A). Scatter plot depicts results of individual biologic replicates, with bars indicating mean +/− SEM. Difference between groups assessed by two-sided Welch t-test. ^**** p ≤ 0.0001. (D) CCRF-CEM cells transduced with the indicated shRNAs were incubated with a pulse of the methionine analog AHA, then released from AHA and treated with asparaginase (10 U/L) during the chase period. The degree of AHA label retention was assessed by flow cytometry. Error bars indicate SEM. (E-F) CCRF-CEM cells were transduced with the indicated constructs, and effects on abundance of Lys-48 ubiquitin chain proteins was assessed by Western blot analysis, in the absence (E) or presence (F) of an FBXW7 expression construct. (G) CCRF-CEM cells transduced with the indicated constructs were analyzed by Western blot (inset), treated with the indicated doses of asparaginase, and viability was assessed by Trypan blue staining after 8 days. Note that wild-type and mutant FBXW7 were Myc-tagged. Error bars represent SEM. (H-I) CCRF-CEM cells were transduced with the indicated constructs, the transduction efficiency (H) was assessed by immunostaining for the V5 tag, and their viability (I) upon treatment with the indicated doses of asparaginase was assessed as in (G). Error bars represent SEM. See also Figures S3, S4 and S5.

Figure 5.. GSK3α Inhibition Phenocopies Wnt-Induced Sensitization to Asparaginase and Promotes Leukemic Cell Death by Depleting Asparagine.

(A) CCRF-CEM cells were transduced with the indicated shRNAs and analyzed by Western blot (inset) for total GSK3 or GAPDH. Cells were then treated with the indicated doses of asparaginase, and viability was assessed after 8 days of treatment. All cell counts were normalized to those in shLuc-transduced, vehicle-treated controls. (B) CCRF-CEM cells were transduced with the indicated shRNAs, without or with a GSK3α expression construct that escapes shRNA targeting. Western blot analysis (inset) was performed for total GSK3 or GAPDH. Cells were then treated with the indicated doses of asparaginase and viability was assessed as in (A). (C) CCRF-CEM cells were treated with vehicle, the GSK3α-selective inhibitor BRD0705, or the GSK3β-selective inhibitor BRD3731, and analyzed by Western blot analysis (inset) for phospho-GSK3 or GAPDH. Cells were then treated with the indicated drugs for 8 days, and viability was assessed as in (A). (D) CCRF-CEM cells were treated with vehicle, asparaginase (10 U/L), the GSK3α inhibitor BRD0705 (100 nM) or both in combination for 12 hr. Culture media was collected and asparagine levels were quantified by UPLC. Fresh media (RPMI-1640) was used as a control for each UPLC run, and amino acid levels are normalized to those in stock RPMI-1640. Differences between groups were calculated using a two-sided Welch t-test. * p ≤ 0.05; ^** p ≤ 0.01; ^*** p ≤ 0.001; ^**** p ≤ 0.0001. n.s., p > 0.05. (E) CCRF-CEM cells were first transduced with shLuc control or Wnt-activating shRNAs targeting NKD2 or LGR6. Cells were then treated with asparaginase (10 U/L) or vehicle and grown in complete growth medium (RPMI-1640 + 10% FBS) or complete growth medium supplemented with 10x L-asparagine or 10x L-glutamine. Fifty percent of the media was removed every 12 hr and replaced with fresh growth medium, supplemented with the appropriate concentration of asparaginase or glutamine. Viability was assessed after 72 hr by counting viable cells. (F) Viability of cells in (E), normalized to viability in no-asparaginase controls. Difference between groups was analyzed using a two-sided Welch t-test. * p ≤ 0.05; ^** p ≤ 0.01; ^*** p ≤ 0.001; ^**** p ≤ 0.0001. n.s., p > 0.05. All error bars indicate SEM. See also Figures S6 and S7 and Table S4.

Figure 6.. Synthetic Lethality of GSK3α Inhibition and Asparaginase in Human Leukemia.

(A) Experimental schema. Human patient-derived leukemia xenografts were injected into NRG immunodeficient mice. Once engraftment of leukemia (>5% human leukemic cells in peripheral blood) was confirmed, mice were treated as indicated. (B) Body weights of mice injected with the chemotherapy-resistant T-ALL PDX in the experiment shown in (C). Error bars represent SEM. (C-E) Survival curves of mice injected with xenografts from patients with a primary asparaginase-resistant T-ALL (C), hypodiploid B-ALL (D), or chemotherapy-refractory MLL-rearranged B-ALL (E), and treated as indicated in (A). Treatment start and end points denoted by arrowheads on each graph.

Figure 7.. Proposed model.

Asparaginase structure is from (Yun et al., 2007).