Enhanced Vulnerability of LKB1-Deficient NSCLC to Disruption of ATP Pools and Redox Homeostasis by 8-Cl-Ado

Abstract

Loss-of-function somatic mutations of STK11, a tumor suppressor gene encoding LKB1 that contributes to the altered metabolic phenotype of cancer cells, is the second most common event in lung adenocarcinomas and often co-occurs with activating KRAS mutations. Tumor cells lacking LKB1 display an aggressive phenotype, with uncontrolled cell growth and higher energetic and redox stress due to its failure to balance ATP and NADPH levels in response to cellular stimulus. The identification of effective therapeutic regimens for patients with LKB1-deficient non-small cell lung cancer (NSCLC) remains a major clinical need. Here, we report that LKB1-deficient NSCLC tumor cells displayed reduced basal levels of ATP and to a lesser extent other nucleotides, and markedly enhanced sensitivity to 8-Cl-adenosine (8-Cl-Ado), an energy-depleting nucleoside analog. Treatment with 8-Cl-Ado depleted intracellular ATP levels, raised redox stress, and induced cell death leading to a compensatory suppression of mTOR signaling in LKB1-intact, but not LKB1-deficient, cells. Proteomic analysis revealed that the MAPK/MEK/ERK and PI3K/AKT pathways were activated in response to 8-Cl-Ado treatment and targeting these pathways enhanced the antitumor efficacy of 8-Cl-Ado. IMPLICATIONS: Together, our findings demonstrate that LKB1-deficient tumor cells are selectively sensitive to 8-Cl-Ado and suggest that therapeutic approaches targeting vulnerable energy stores combined with signaling pathway inhibitors merit further investigation for this patient population.

Figure 1.

LKB1-deficient NSCLC cells display increased sensitivity to 8-Cl-Ado. (A) Comparison of 8-Cl-Ado growth/survival inhibitory IC₅₀ values of individual NSCLC cell lines grouped by STK11 mutation status as indicated for 57 cell lines panel. Comparison of the growth/survival inhibitory IC₅₀ values of 8-Cl-Ado from a panel of 104 NSCLC cell lines screen delineated by (B) LKB1 deficiency or (C) LKB1 deficiency and/or KRAS mutation status. Statistical analysis was performed using a two-tailed non-parametric Mann-Whitney t-test. (D-J) Scatter plots showing the correlation between 8-Cl-Ado sensitivity and basal expression levels for indicated proteins for LKB1-deficient and proficient NSCLC cell lines tested. P values were calculated by Spearman correlation.

Figure 2.

LKB1 expression status determines 8-Cl-Ado–mediated cytotoxicity. (A) Immunoblot analysis for LKB1 protein expression in A549, H460 and Calu-6 isogenic pairs. (B) IC₅₀ values of 8-Cl-Ado for H460, A549 and Calu-6 isogenic pairs determined by Cell Titer Glo assay after 72 hours of treatment. Statistical analysis performed using a two-tailed unpaired t-test. (C) Flow cytometry analysis of annexin V/PI staining in LKB1 isogenic A549 cells treated with 8-Cl-Ado for 4 days at indicated doses. (D-F) Quantitation of percentage of late apoptotic population represented by annexin V and PI positive cells for triplicate experiments performed in (D) A549, (E) H460 and (F) Calu-6 LKB1 isogenic pairs. Statistical analysis performed using an ANOVA Sidak’s multiple comparisons test. Bars represent the Mean ± SEM.

Figure 3.

8-Cl-Ado–mediated DNA, RNA and protein synthesis impairment does not determine LKB1-dependent 8-Cl-Ado sensitivity. Macromolecules synthesis in NSCLC cells as measured by [3H]-thymidine, [3H]-uridine and [3H]-leucine incorporation after 72 hours of 8-Cl-Ado treatment at indicated concentration in H460 and A549 LKB1 isogenic pairs. Statistical analysis was performed using a two-tailed paired t-test. Data are presented as mean ± SEM. Bars represent the mean ± SEM.

Figure 4.

LKB1 expression status attenuates 8-Cl-Ado–mediated depletion of ATP and induction of ROS production. (A) Relative ATP levels in H460 and A549 LKB1-proficient vs LKB1-deficient A549 and H460 cell lines by HPLC analysis. (B) Relative ATP levels in LKB1 isogenic H460 and A549 LKB1 isogenic pairs treated with increasing concentrations of 8-Cl-Ado assessed by HPLC analysis. Values are percent of the average nucleotide levels compared to H460 or A549 control LKB1-proficient 0 hours untreated cells. (C) Intracellular quantification of 8-Cl-Ado’s cytotoxic metabolite, 8-Cl-ATP, in LKB1 isogenic H460 and A549 cells treated with increasing concentrations of 8-Cl-Ado. (D) Relative NTPs levels in H460 and A549 LKB1 isogenic pairs treated with 2μM 8-Cl-Ado for 6 hours assessed by HPLC analysis. Values are percent of the average nucleotide levels compared to 0 hours untreated cells. (E) HPLC ATP analysis in LKB1 isogenic H460 and A549 cells treated with 2μM 8-Cl-Ado, 10 mM metformin (Metf) or 20 mM 2-deoxyglucose (2DG). (F) Flow cytometry analysis of cellular ROS levels in LKB1 isogenic A549, H23, H2030, H460, and Calu6 cells after treatment with 5μM 8-Cl-Ado for 72 hours. ROS levels are graphed as fold change to average of all LKB1-deficient untreated cells. Statistical analysis was performed using a two-tailed paired t-test. Data are presented as mean ± SEM. Bars represent the mean ± SEM.

Figure 5.

8-Cl-Ado–mediated AMPK activation is LKB1 dependent. Immunoblot analysis of (A) phospho-AMPK and LKB1 in indicated LKB1 isogenic pairs; (B) phospho-ACC, LKB1 and phospho-p70S6K for H460 and A549 LKB1 isogenic pairs; and (C) phospho-ACC, LKB1, phospho-ACC, phospho-p70S6K and phospho- and total-S6 ribosomal protein levels in Calu-6 cells. Cells were treated with the indicated concentrations of 8-Cl-Ado for 24 h. Data shown in panel B and C for A549 and Calu-6 pairs respectively were generated by re-probing blots used in panel A. For presentation purpose loading controls from panel A have been repurposed in panel B and C for A549 and Calu-6 pairs, respectively. (D) IC₅₀ values and (E) combination indexes for 8-Cl-Ado and the AMPK inhibitor compound C (CmpC) at ED₂₅, ED₅₀, and ED₇₅. Statistical analysis was performed using an ANOVA Sidak’s multiple comparisons test. Bars represent the median with interquartile range.

Figure 6.

Basal protein expression patterns correlated with 8-Cl-Ado sensitivity. Protein markers as assessed by RPPA that significantly correlate with 8-Cl-Ado sensitivity/resistant in all NSCLC cell lines tested. Scatter plots showing the correlation between 8-Cl-Ado sensitivity and basal expression levels of (A-F) indicated proteins for LKB1-proficient and deficient NSCLC cell lines tested. P values were calculated by Spearman correlation.

Figure 7.

8-Cl-Ado treatment induces ERK1/2 and AKT activity and synergizes with MEK and PI3K inhibitors. (A) Heat map showing 8-Cl-Ado–mediated fold changes in the levels of proteins and phosphoproteins which showed significant differences at 24 hours of 8-Cl-Ado treatment measured by RPPA analysis only for LKB1-deficient cell lines assessed. (B) Immunoblot analysis of phospho-ERK1/2, total ERK1/2, phospho-AKT^(S473), total AKT, phospho-GSK3β^(S9), and total-GSK3β protein levels in LKB1 isogenic H460, A549, and Calu-6 cells treated with 8-Cl-Ado for 24 hours. (C) IC₅₀ values and (D) combination index values for 8-Cl-Ado and MEK inhibitor (trametinib). (E) IC₅₀ values and (F) combination index values for 8-Cl-Ado and PI3K inhibitor (A66) at ED₂₅, ED₅₀, and ED₇₅ in the cell lines assessed. Statistical analysis was performed using an ANOVA Sidak’s multiple comparisons test. Bars represent the median with interquartile range.