Phosphorylation of FADD by the kinase CK1α promotes KRASG12D-induced lung cancer

Abstract

Genomic amplification of the gene encoding and phosphorylation of the protein FADD (Fas-associated death domain) is associated with poor clinical outcome in lung cancer and in head and neck cancer. Activating mutations in the guanosine triphosphatase RAS promotes cell proliferation in various cancers. Increased abundance of phosphorylated FADD in patient-derived tumor samples predicts poor clinical outcome. Using immunohistochemistry analysis and in vivo imaging of conditional mouse models of KRAS(G12D)-driven lung cancer, we found that the deletion of the gene encoding FADD suppressed tumor growth, reduced the proliferative index of cells, and decreased the activation of downstream effectors of the RAS-MAPK (mitogen-activated protein kinase) pathway that promote the cell cycle, including retinoblastoma (RB) and cyclin D1. In mouse embryonic fibroblasts, the induction of mitosis upon activation of KRAS required FADD and the phosphorylation of FADD by CK1α (casein kinase 1α). Deleting the gene encoding CK1α in KRAS mutant mice abrogated the phosphorylation of FADD and suppressed lung cancer development. Phosphorylated FADD was most abundant during the G2/M phase of the cell cycle, and mass spectrometry revealed that phosphorylated FADD interacted with kinases that mediate the G2/M transition, including PLK1 (Polo-like kinase 1), AURKA (Aurora kinase A), and BUB1 (budding uninhibited by benzimidazoles 1). This interaction was decreased in cells treated with a CKI-7, a CK1α inhibitor. Therefore, as the kinase that phosphorylates FADD downstream of RAS, CK1α may be a therapeutic target for KRAS-driven lung cancer.

Fig. 1. A requirement for FADD in Kras-driven lung cancer

(A) Genetic strategy used to activate Kras^LSL-G12D, Rosa26^{LSL-Luciferase} and deplete Fadd:GFP expression in a lung-specific manner. (B) Representative bioluminescent images of K_Luc, and KF_Luc mice at 7 to 22 weeks after intranasal AdCre administration. (C) Average bioluminescence in Luc (n=7), F_Luc (n=7), K_Luc (n=25), and KF_Luc (n=34) mice at specified times. (D) Representative images of CT scans of lungs from Luc, FLuc, K_Luc, and KF_Luc animals at shown times, with H&E slide corresponding to that animal from week 18 or week 22 as indicated. Arrows, lesions. H, heart. (E) Tumor and vascular volumes in K_Luc (n=10) and KF_Luc (n=17) mice analyzed by CT at indicated times. Data are means ± SEM. *P=0.009, **P=0.018, unpaired Student’s t-tests. (F) Co-localization of bioluminescence imaging and CT imaging for lung tumors in Luc, KF_Luc and K_Luc mice. Inset, H&E of lesion in K_Luc mouse. (G) Survival plot of K_Luc (n=19) and KF_Luc (n=15) mice. Median survival time: KF_Luc mice, 51.4 weeks; K_Luc mice, 34 weeks. ***P=0.005, Wilcoxon log-rank test (95% CI).

Fig. 2. Fadd null lung tumors are less proliferative

(A) Representative histology images of lungs removed from mice 18 weeks after AdCre administration. Tissues were stained with hematoxylin and eosin (HE) or antibodies against cyclin D1, Ki-67, and FADD. Scale bars: 10x, 500 μm; 40x, 200 μm; 100x, 50 μm. (B) Average tumor area quantified from H&E-stained lung tissue sections from K_Luc (n=6) and KF_Luc (n=10) mice. *P=0.04, unpaired Student’s t-test. (C) Average percentage of positive Ki-67 stained cells in lung tissue sections from K_Luc and KF_Luc mice, 4 fields per 10 mice each. **P=2×10⁻⁵, unpaired Student’s t-test. (D) Representative Western blot for endogenous FADD, GFP (Fadd transgene), phosphorylated and total ERK1/2, phosphorylated RB, cyclin D1, cyclin B1 and β-Actin in lung tissue from the indicated mice. Blots are representative of 3 independent experiments.

Fig. 3. FADD and FADD phosphorylation are required for Kras-driven cell proliferation

(A) Western blotting for FADD, Cyclin D1 and β-Actin in lysates from Luc, K_Luc and KF_Luc MEFs. (B and C) Alamar blue proliferation assay in cultures of Luc, K_Luc and KF_Luc MEFs that were either (B) untreated or (C) treated with DMSO, 250 μM CKI-7, 10 μM lonafarnib, or 200 nM PD0325901. (D) Agar colony formation of K_Luc, KF_Luc, and FADD-reconstituted KF_Luc MEF cells treated as indicated for 2 weeks, concentrations as in (C). (E) Fluorescence-activated cell sorting (FACS) analysis of cell cycle distribution in K_Luc and KF_Luc MEFs 24 hours after the indicated treatment, concentrations as in (C). (F) Western blotting for the indicated proteins in K_Luc MEFs treated for 6 hours as in (C). Blots (A and F) and FACS data (E) are representative of 3 independent experiments. Data are means ± SEM from 3 independent experiments.

Fig. 4. FADD interacts with key mediators of G2/M transition

(A) Western blotting for the indicated proteins in H1975 cells after double thymidine block and release (G2/M). Cells were harvested at 6, 7, 8, and 9 hours after synchronization. AS, asynchronous cells. (B and C) Western blotting after FADD immunoprecipitation from A549 cells (B) and quantification of the BUB1-FADD interaction (C). Cells were asynchronous (AS) or treated with CKI-7, nocodozole (G2/M), or hydroxyurea (G1/S). BUB1 abundance was normalized to that in asynchronous cells. Data are means ± SEM from 3 independent experiments,*P = 0.005, unpaired Student’s t-test.

Fig. 5. CK1α mediates Kras-driven lung cancer

(A) Genetic strategy used to activate Kras^LSL-G12D, Rosa26^{LSL-Luciferase} and delete Csnk1a expression in a lung specific manner. Bioluminescence and/or CT imaging was performed at the indicated weeks after administration of Cre recombinase. (B) Bioluminescence of Luc (n=8), C_Luc (n=5), K_Luc (n=38), and KC_Luc (n=25) mice at the specified times. Data are means ± SEM. Pseudocolor scale is radiance as photons/second/cm2/steradian (p/sec/cm2/sr). (C) Representative images of CT scans of lungs from Luc, K_Luc, C_Luc, and KC_Luc mice, with corresponding H&E slide. Arrows, lesions. H, heart. (D) Average tumor and vascular volumes of K_Luc (n=8) and KC_Luc (n=7) mice analyzed by CT 18 weeks after AdCre administration. Data are means ± SEM; *P=7.47×10⁻⁶, unpaired Student’s t-test. (E) Representative histology for H&E or indicated proteins in lungs from mice 18 weeks after AdCre administration. Scale bars: 10x, 500 μm; 40x, 200 μm; 100x, 50 μm. (F) Representative Western blot for CK1α, FADD and β-actin, in lysates from KC_Luc MEFs treated with various concentrations of AdCre. Pfu, plaque-forming units (G and H) Representative Western blot (G) and quantification of phosphorylated fractions (H) for the indicated proteins in lysates of K_Luc and KC_Luc MEFs treated with AdCre or AdLuc. Data are means ± SEM from three independent experiments. **P =0.003, unpaired Student’s t-test.

Fig. 6. Model for the role of FADD in KRAS-mediated cell proliferation

We propose that FADD and CK1α are necessary in mediating mitogenic KRAS signaling in cancer cells. Inhibiting KRAS, MEK or CK1α, or deleting Csnk1a1 or Fadd decreased the abundance of phosphorylated FADD and decreased cell proliferation. We propose cells lacking FADD and CK1α fail to progress through G2/M because the interaction of FADD with key G2/M transition proteins like AURKA, PLK1 or BUB1 is lost.