Aurora kinase A interacts with H-Ras and potentiates Ras-MAPK signaling

Abstract

In cancer, upregulated Ras promotes cellular transformation and proliferation in part through activation of oncogenic Ras-MAPK signaling. While directly inhibiting Ras has proven challenging, new insights into Ras regulation through protein-protein interactions may offer unique opportunities for therapeutic intervention. Here we report the identification and validation of Aurora kinase A (Aurora A) as a novel Ras binding protein. We demonstrate that the kinase domain of Aurora A mediates the interaction with the N-terminal domain of H-Ras. Further more, the interaction of Aurora A and H-Ras exists in a protein complex with Raf-1. We show that binding of H-Ras to Raf-1 and subsequent MAPK signaling is enhanced by Aurora A, and requires active H-Ras. Thus, the functional linkage between Aurora A and the H-Ras/Raf-1 protein complex may provide a mechanism for Aurora A's oncogenic activity through direct activation of the Ras/MAPK pathway.

Keywords: Aurora A; MAPK; Raf; Ras; protein-protein interactions.

Figure 1. Detection of the Aurora A/H-Ras interaction

A. TR-FRET assay performed using lysates from HEK 293T cells in which GST H-Ras was co-expressed with Venus-Flag Aurora A or vector controls. TR-FRET signal calculated as X/Y*Z; Tb ex 340 nm; Tb em 486 nm (X); Venus em 520 nm (Y); Z = 10⁴). TR-FRET signals were recorded using an EnVision multilabel plate reader. Data shown are average signals with SD from duplicate samples. B. GST pull-down assay conducted after GST H-Ras complexes were isolated from HEK 293T cell lysates with co-expressed Venus-Flag Aurora or appropriate controls. The presence of Venus-Flag Aurora A in the GST H-Ras protein complex (GST PD) and protein expression levels in the cell lysate (Input) was detected by Western blotting using anti-Flag or anti-GST antibody, respectively. C. A Venus protein-fragment complementation (Venus PCA) assay was conducted in living HEK 293T cells co-expressing N-Venus Aurora A and C-Venus H-Ras or vector controls. Interaction between tagged proteins allowed reconstitution of fluorescent Venus protein. The percentage of Venus positive cells was quantified by fluorescence imaging and scoring from two independent experiments. The percentage represents the number of cells with positive interactions compared to the total number of cells (determined by Hoechst staining). Representative images: Venus (positive protein-protein interaction), Hoechst (nucleus), Merge (overlap of Venus and Hoechst signals). D. Co-immunoprecipitation assay performed using lysates from HEK 293T and MCF7 cells. The Aurora A/Ras interactions are shown in both directions with IP-Aurora A and IP-Ras.

Figure 2. Interactions between Aurora A/B and Ras proteins are mediated through conserved domains

A. Diagram of GST H-Ras protein domains and truncations used for deletion analysis: FL (amino acids 1-189), SI&II (amino acids 1-66), ΔSI (amino acids 36-189), ΔSI&II (amino acids 66-189). B. Characterization of the H-Ras protein domain responsible for binding to Aurora A. GST pull-down conducted from HEK 293T cells co-expressing GST H-Ras truncations and Venus-Flag Aurora A. Western blotting using anti-Flag or anti-GST antibody allowed detection of GST H-Ras peptides that were able to isolate full-length Aurora A. Full-length Aurora A/H-Ras protein binding was used as a positive control. C. Aurora A exists in protein complexes with H-, K-, or N-Ras. Binding of Aurora A as detected in GST pull-downs conducted from HEK 293T cells expressing GST H-Ras, GST K-Ras, or GST N-Ras and Venus-Flag Aurora A along with vector controls. D. Characterization of the H-Ras binding domain on Aurora A. Diagram of Aurora A protein domains and truncations used for deletion analysis: FL (amino acids 1-403), NK (amino acids 1-383), N (amino acids 1-130), K (amino acids 130-383), C (amino acids 383-403). E. GST pull-down conducted from HEK 293T cells co-expressing full-length GST H-Ras and Venus-Flag Aurora A truncations and analyzed by western blotting. Binding between full-length proteins served as a positive control. F. Aurora B interacts with H-Ras. Like Aurora A, Aurora B can be isolated in a protein complex with H-Ras. Binding of Aurora B as detected in GST pull-downs conducted from HEK 293T cells expressing GST H-Ras and Venus-Flag Aurora B along with vector controls was identified by western blotting using anti-Flag or anti-GST antibodies.

Figure 3. Aurora A potentiates ERK activation via H-Ras

A. Detection of the Aurora A/H-Ras interaction correlates with enhanced pERK. GST pull-down (described in Figure 1B) between GST H-Ras and Venus-Flag Aurora A with corresponding western blot analysis of cell lysate inputs to assess changes in pERK compared to total ERK 48-hours post-transfection in HEK 293T cells. B. Aurora A sustains pERK levels in MCF-7 breast cancer cells. MCF-7 cells were either untransfected, transfected with Venus-Flag Aurora A or GST H-Ras with appropriate controls, or transfected with GST H-Ras and Venus-Flag Aurora A. As detected by western blotting, changes in pERK induced by co-transfected plasmids was assessed after cells were stimulated with serum for 0, 5, 10, 45, and 90 minutes after 24-hours of serum starvation. A short exposure (SE) and longer exposure (LE) of pERK is shown. C. H-Ras activity is required for potentiation of pERK by Aurora A. GST pull-down comparing binding and signaling changes between co-expression of GST H-Ras (WT), GST H-Ras G12V activating mutant, or GST H-Ras S17N dominant negative mutant with Venus-Flag Aurora A in HEK 293T cells. Western blot analysis of inputs to assess changes in pERK compared to total ERK 48 hours post-transfection. D. Use of a pharmacological probe for the MAPK signaling pathway in HEK 293T cells co-expressing Aurora A and H-Ras alone or in combination. 24-hours post transfection, cells were treated with DMSO vehicle control (Veh.), serum starvation (S.S.) Sorafenib (Soraf.) or U0126 at 10 μm then subjected to a GST pull-down and western blot analysis. Western blotting was conducted using anti-Flag, anti-GST, anti-pMEK, anti-MEK, anti-pERK, and anti-ERK antibodies.

Figure 4. Aurora A forms a complex with H-Ras and Raf-1, acting through H-Ras to enhance ERK activation

A. Aurora A directly interacts with Raf-1. TR-FRET was performed using HEK 293T lysates in which GST Raf-1 and Venus-Flag Aurora A along with vector controls were co-expressed. TR-FRET signals were recorded using an EnVision multilabel plate reader. Data shown are average signals with SD from duplicate samples. B. Aurora A associates with Raf-1. GST pull-down (as described in Figure 1B) between GST Raf-1 and Venus-Flag Aurora A with corresponding western blot analysis of inputs to assess changes in pERK compared to total ERK 48-hours post-transfection in HEK 293T cells. C. Aurora A/H-Ras/Raf-1 interactions stabilize the protein signaling complex. GST pull-down comparing the ability of wild-type (H-Ras WT) or dominant negative (H-Ras S17N) H-Ras to isolate either co-expressed Aurora A, Raf-1, or both proteins. Western blot analysis demonstrates binding of Aurora A or Raf-1 to H-Ras and the induced effect on pERK. Since both epitope-tagged proteins resolve around the same size, anti-Aurora A and anti-Raf-1 antibodies were used instead of anti-Flag. GST-tagged H-Ras WT and H-Ras S17N were detected using anti-GST antibody. Changes in pERK were detected using anti-pERK antibody. D. Proposed model for the role of Aurora A in the Aurora A/H-Ras/Raf-1 oncogenic singaling complex.