Selective Aurora A-TPX2 Interaction Inhibitors Have In Vivo Efficacy as Targeted Antimitotic Agents

Abstract

Aurora A kinase, a cell division regulator, is frequently overexpressed in various cancers, provoking genome instability and resistance to antimitotic chemotherapy. Localization and enzymatic activity of Aurora A are regulated by its interaction with the spindle assembly factor TPX2. We have used fragment-based, structure-guided lead discovery to develop small molecule inhibitors of the Aurora A-TPX2 protein-protein interaction (PPI). Our lead compound, CAM2602, inhibits Aurora A:TPX2 interaction, binding Aurora A with 19 nM affinity. CAM2602 exhibits oral bioavailability, causes pharmacodynamic biomarker modulation, and arrests the growth of tumor xenografts. CAM2602 acts by a novel mechanism compared to ATP-competitive inhibitors and is highly specific to Aurora A over Aurora B. Consistent with our finding that Aurora A overexpression drives taxane resistance, these inhibitors synergize with paclitaxel to suppress the outgrowth of pancreatic cancer cells. Our results provide a blueprint for targeting the Aurora A-TPX2 PPI for cancer therapy and suggest a promising clinical utility for this mode of action.

Figure 1

Aurora A interactions and inhibition. (A) Complexes of Aurora A (gray) with different interacting molecules. From left to right, ATP-competitive clinical stage inhibitor alisertib (yellow carbons, PDB: 5ia0(43)), TPX2 peptide (orange carbons, PDB: 1ol5(24)), and N-Myc (orange carbons, PDB: 5g1x(44)). (B) Interaction of the N-terminal part of the TPX2 peptide with Aurora A, with key aromatic residues labeled in the two pockets on the N-lobe. (C) Complex of Aurora A with Aurkin A (yellow carbons, 5dpv), a low-affinity inhibitor of Aurora:TPX2 interaction.

Figure 2

Aurora A:TPX2 interaction inhibitor design. Overview of the fragment-based development of CAM2602 to inhibit the Aurora A:TPX2 protein–protein interaction. Chemical structures are shown for compounds 1–10 and CAM2602. Crystal structures of compounds 2 (PDB:8C1M), 3 (8C15), 4 (8C1D), 5 (8C1E), 6 (8C1F), 8 (8C1H), 9 (8C14), and 10 (8C1I) in complex with Aurora A are shown next to the chemical structures. The blue boxes on the chemical structures and corresponding blue arrows on the crystal structures highlight the key change(s) at each step. The “meta-tunnel” is marked in the structure of 5. The K_D values are obtained from a competitive FP assay or a direct ITC measurement.

Figure 3

CAM2602 characterization. (A) X-ray crystal structure of CAM2602 bound to Aurora A is shown (purple carbons; PDB: 8C1K) along with key interactions in the Tyr pocket (B) View from above the Tyr pocket with Aurora A as a molecular surface. (C) Overlay of compound 2 and CAM2602 shows remarkable preservation of the binding pose across the inhibitor development series. (D) Competition fluorescence polarization assay of CAM2602 competing with the TPX2 peptide. (E) Kinase panel results were obtained using compound 9. Red spheres indicate cross-reactivity with kinases in the phylogenetic tree with no observed reactivity for compound 9 in the panel of 97 human kinases; details in Figure S3. (F) Isothermal titration calorimetry titration of compound 8 to Aurora A (left) and to Aurora B (right). (G) Conservation of residues in the Tyr pocket between Aurora A (light gray, PDB: 8C1K) and Aurora B (pale blue, PDB: 4AF3) with residues lining the Tyr pocket shown as sticks and nonconserved residues in Aurora B colored red. The same residues are shown below the figure with red background for nonconserved ones. On the right is a zoomed-in view of CAM2602 binding to Aurora A, overlaid with the Aurora B structure (pale blue). The surfaces of additional methyl groups in Ile126 and Ile150 are displayed with surface dots, showing their proximity to CAM2602.

Figure 4

Cellular efficacy of the CAM2602 series. (A) High-content microscopy assays to evaluate mislocalization of Aurora A from the mitotic spindle or loss of P-Thr288 Aurora A in mitotic nuclei when treated with the inhibitor. HeLa cells were treated with titrations of the indicated compounds for 2 h before being fixed, stained for Aurora A, and analyzed using high-content microscopy to determine the percentage of observed mitotic cells at each concentration with spindle-displaced Aurora A (mislocalization). The indicated EC₅₀ values for each compound were calculated from the plots of assay scores against the compound concentration. (B) As in A but stained for dephosphorylated Thr288 Aurora A. (C) Viability assays in Jurkat cells. Jurkat cells were cultured for 72 h with titrations of the indicated compounds. Viability assays were performed following the treatment period and the data normalized to the vehicle-treated controls. GI₅₀ values were calculated from plots of the viability assay data. (D) Viability in HeLa cells, determined similarly to (C).

Figure 5

In vitro and in vivo characterization of CAM2602. (A) Jurkat cells were treated for 8 h with 20 μM CAM2602 or 14 nM alisertib and analyzed by flow cytometry for PH3-positive cells relative to vehicle controls. PH3-positive cells from each sample were assessed for a loss of P-Thr288 positivity. (B) Female NSG mice bearing solid Jurkat tumors (subcutaneous implantation, rear dorsum) were administered a single oral dose of either CAM2602 or vehicle. Tumor cells from 0, 8, or 12 h of treatment were analyzed by flow cytometry similarly to in vitro samples in panel A. (C) Pharmacokinetic analysis of CAM2602 or alisertib concentrations in tumor and plasma samples taken at 8 or 12 h after dosing with 200 and 30 mg/kg, respectively. (D) NSG mice bearing subcutaneous, solid tumor xenografts of Jurkat cells were dosed orally once per day with either vehicle, CAM2602, or alisertib, as indicated (n = 5). Tumor volumes were estimated periodically over the 26 days of dosing by calliper measurement. Error bars show the standard deviation from the mean.