Aurora kinase targeting in lung cancer reduces KRAS-induced transformation

Abstract

Background: Activating mutations in KRAS are prevalent in lung cancer and have been causally linked to the oncogenic process. However, therapies targeted to oncogenic RAS have been ineffective to date and identification of KRAS targets that impinge on the oncogenic phenotype is warranted. Based on published studies showing that mitotic kinases Aurora A (AURKA) and B (AURKB) cooperate with oncogenic RAS to promote malignant transformation and that AURKA phosphorylates RAS effector pathway components, the aim of this study was to investigate whether AURKA and AURKB are KRAS targets in lung cancer and whether targeting these kinases might be therapeutically beneficial.

Methods: In order to determine whether oncogenic KRAS induces Aurora kinase expression, we used qPCR and western blotting in three different lung cell-based models of gain- or loss-of-function of KRAS. In order to determine the functional role of these kinases in KRAS-induced transformation, we generated KRAS-positive A549 and H358 cells with stable and inducible shRNA-mediated knockdown of AURKA or AURKB and evaluated transformation in vitro and tumor growth in vivo. In order to validate AURKA and/or AURKB as therapeutically relevant KRAS targets in lung cancer, we treated A549 and H358 cells, as well as two different lung cell based models of gain-of-function of KRAS with a dual Aurora kinase inhibitor and performed functional in vitro assays.

Results: We determined that KRAS positively regulates AURKA and AURKB expression. Furthermore, in KRAS-positive H358 and A549 cell lines, inducible knockdown of AURKA or AURKB, as well as treatment with a dual AURKA/AURKB inhibitor, decreased growth, viability, proliferation, transformation, and induced apoptosis in vitro. In addition, inducible shRNA-mediated knockdown of AURKA in A549 cells decreased tumor growth in vivo. More importantly, dual pharmacological inhibiton of AURKA and AURKB reduced growth, viability, transformation, and induced apoptosis in vitro in an oncogenic KRAS-dependent manner, indicating that Aurora kinase inhibition therapy can specifically target KRAS-transformed cells.

Conclusions: Our results support our hypothesis that Aurora kinases are important KRAS targets in lung cancer and suggest Aurora kinase inhibition as a novel approach for KRAS-induced lung cancer therapy.

Fig. 1

KRAS induces expression of AURKA and AURKB in lung cells. a Protein lysates of SALEB and SAKRAS cells were submitted to western blotting with the indicated antibodies. b H1703-TrexB and H1703-G12V lung cancer cells were treated with 2 μg/mL doxycycline (DOX) for 24 h where indicated to induce KRAS expression. Subsequently, protein lysates were prepared and submitted to western blotting with the indicated antibodies. TUBA) anti-α-tubulin. c A549 and H358 stable cells with inducible expression of 2 different shRNAs targeting KRAS (shKR#1 and shKR#2) or a non-targeting shRNA (shCtrl) were treated with 2 μg/mL doxycycline (DOX) for 5 days to induce shRNA expression where indicated. Protein lysates were prepared and submitted to western blotting with the indicated antibodies

Fig. 2

Dual pharmacological inhibition of AURKA and AURKB decreases the transformed phenotype of KRAS-positive lung cells. a A549 and H358 cells were treated with 0.1 % DMSO or increasing concentrations of AI II as indicated for 72 h and protein lysates were prepared and submitted to Western blotting with the indicated antibodies. b Growth curve analysis of A549 and H358 cells treated with the indicated concentrations of AI II compared to control-treated cells (0.1 % DMSO) for the indicated times. c A549 and H358 cells were plated for clonogenic assays as described in methods and treated for 21 days with either 0.1 % DMSO or different concentrations of AI II as indicated. Colonies formed were stained with crystal violet and counted. Images shown are representative of three independent experiments. d A549 and H358 cells were treated with 0.1 % DMSO or increasing concentrations of AI II as indicated for 72 h, stained for Annexin V and propidium iodide (PI) as described in methods and Annexin V positive cells were analyzed by flow cytometry. e A549 and H358 cells were treated with 0.1 % DMSO or increasing concentrations of AI II as indicated for 72 h, cells were stained with BrdU and PI as described in methods, and cell cycle analysis was performed by flow cytometry. f Anchorage-independent growth was evaluated by plating A549 and H358 in soft agar as described in methods. Cells were then treated for 21 days with either 0.1 % DMSO or different concentrations of AI II as indicated. Colonies formed were stained with MTT and counted. Images shown are representative of three independent experiments. In all cases, statistical significance was determined when appropriate by Student’s t-test (*p < 0.05, **p < 0.01) by comparing AI II-treated vs. DMSO-treated samples. Error bars represent average ± 1 s.d

Fig. 3

shRNA-mediated knockdown of AURKA or AURKB decreases the transformed phenotype of KRAS-positive lung cells. Unless otherwise indicated, A549 and H358 stable cells with inducible expression of 2 different shRNAs targeting AURKA (shAKA#1 and shAKA#2), AURKB (shAKB#1 and shAKB#2) or a non-targeting shRNA (shCtrl) were either treated with 2 μg/mL doxycycline (DOX) for 5 days to induce shRNA expression or left untreated (MOCK). a Protein lysates of doxycycline-treated (+) and untreated (−) cells were submitted to western blotting with the indicated antibodies. TUBA) anti-α-tubulin. b Growth curve analysis of the indicated cells. All cells were treated with 2 μg/mL doxycycline (DOX) for the indicated times. c The indicated cells were plated for clonogenic assays as described in methods and treated for 21 days with 2 μg/mL doxycycline (DOX). Colonies formed were stained with crystal violet and counted. Images shown are representative of three independent experiments. d The indicated treated (DOX) or untreated (MOCK) cells were stained with BrdU and propidium iodide (PI) as described in methods, and cell cycle analysis was performed by flow cytometry. e Protein lysates of A549 and H358 stable cells with inducible expression of 2 different shRNAs targeting AURKA (shAKA#1 and #2), AURKB (shAKB#1 and #2) or a non-targeting shRNA (shCtrl), treated (+) or not (-) with 2 μg/mL doxycycline (DOX) for 5 days, were submitted to western blotting with the indicated antibodies. C3) anti-caspase 3. f Anchorage-independent growth was evaluated by plating the indicated cells in soft agar as described in methods. Cells were then treated for 21 days with 2 μg/mL doxycycline (DOX) or left untreated (MOCK). Colonies formed were stained with MTT and counted. Images shown are representative of three independent experiments. In all cases, statistical significance was determined when appropriate by Student’s t-test (*p < 0.05, **p < 0.01) and the groups being compared are indicated by horizontal bars

Fig. 4

shRNA-mediated knockdown of AURKA reduces xenograft tumor growth. a A549 stable cells with inducible expression of a shRNA targeting AURKA (shAKA) or a non-targeting shRNA (shCtrl) were injected subcutaneously in nude mice (n = 9 per group). shRNA expression was induced by doxycycline (DOX) administration in the drinking water as described in methods (n = 5 per group). Alternatively, mice were left untreated (MOCK) to control for doxycycline-mediated effects (n = 4 per group). The graph shows tumor volume measurements, which were initiated 26 days after inoculation (day 0). Error bars represent average ± 1 s.d. b Representative images of the tumors at day 56 after inoculation. c A549 stable cells with inducible expression of a shRNA targeting AURKA (shAKA) or a non-targeting shRNA (shCtrl) were injected subcutaneously in nude mice (n = 12 per group). shRNA expression was induced by doxycycline (DOX) administration in the drinking water as described in methods (n = 6 per group). Tumor weights at day 56. Error bars represent average ± 1 s.d. In all cases, statistical significance was determined when appropriate by one-way analysis of variance (ANOVA) followed by Bonferroni’s multiple comparison test (a) or by Student's t-test (c) (*p < 0.05) and the significantly different comparisons are indicated by vertical (a) or horizontal (c) bars. d Protein lysates of xenograft tumors were submitted to western blotting with the indicated antibodies. ACTB) anti-β-actin

Fig. 5

AURKA or AURKB targeting decreases the transformed phenotype of lung cells in a KRAS-dependent manner. Primary immortalized human airway cells (SALEB) and their KRAS-transformed counterpart (SAKRAS) were treated with 0.1 % DMSO (DMSO) or 1uM AI II (AI II) as indicated. H1703-TrexB and H1703-G12V lung cancer cells were also treated with 0.1 % DMSO (DMSO) or 1 μM AI II (AI II) as indicated. To induce KRAS expression H1703-TrexB and H1703-G12V cells were simultaneously treated with 2 μg/mL doxycycline (DOX + DMSO or DOX + AI II) where indicated. a Growth curve analysis of cells. All drug treatments (DMSO, AI II, DOX + DMSO, DOX + AI II) were continued for 12 days. b Cells were plated for clonogenic assays as described in methods and treated for 21 days as indicated. Colonies formed were stained with crystal violet and counted. Images shown are representative of three independent experiments. c Anchorage-independent growth was evaluated by plating the indicated cells in soft agar as described in methods. Cells were then treated for 21 days as indicated. Colonies formed were stained with MTT and counted. Images shown are representative of three independent experiments. In all cases, statistical significance was determined when appropriate by Student’s t-test (*p < 0.05, **p < 0.01) and the groups being compared are indicated by horizontal bars