Figure 4. DEER spectroscopy confirms that phosphorylation of AurA alters the DFG-In state.

(a) Background-corrected DEER spectra of unphosphorylated AurA bound to ADP (red), and phosphorylated AurA bound to either ADP (blue) or to Tpx2 (yellow). The inset shows an enlarged view of the spectra for the +ADP samples. (b) Population densities obtained by Tikhonov regularization for the data shown in (a), with prominent peaks in the distributions indicated. The increased sampling of distances beyond ~50 angstroms in the phosphorylated kinase bound to ADP is highlighted with darker blue shading. Data are from single representative experiments of two independent repeats. (c) Spin-spin distance distributions obtained by molecular dynamics simulations initiated from X-ray structures of AurA in either the DFG-Out state (purple) or the fully-active DFG-In state with both Tpx2 and phosphorylation (pink). The inset shows schematics of the spin-labeling scheme. (d) X-ray structure of SNS-314 bound to AurA highlighting interactions with the DFG motif, structured water molecules and the catalytic glutamate (E181) specific to the DFG-In state (PDB ID: 3D15). (e) DEER spectra (main panel) and distance distributions (inset) measured for unphosphorylated (red) and phosphorylated (blue) AurA bound to SNS-314. The distributions are vertically aligned with those shown in (b) to facilitate comparison. (f) Hypothesized energy landscape for AurA, highlighting the effect of phosphorylation on the DFG-In state.

Figure 4—figure supplement 1. Validation of spin-labeled AurA constructs used for EPR.

(a and b) Whole protein LC-MS spectra of (a) unphosphorylated AurA C290S C247A C393S L225C S284C and (b) phosphorylated AurA C290A C247A C393S L225C S284C homogeneously labeled with two MTSL spin labels. (c) Kinase assay data showing unlabeled and MTSL-labeled AurA are active and activated by Tpx2 binding to a similar extent. Data represent the average of three replicate experiments ± s.d.

Figure 4—figure supplement 2. DEER experiments with AMPPNP and peptide substrate.

(a) DEER spectra and distance distributions (inset) for unphosphorylated (red) and phosphorylated (blue) AurA bound to the non-hydrolyzable ATP-analog AMPPNP. The distribution for phosphorylated AurA bound to Tpx2 is shown as a yellow dotted line for comparison. The increased sampling of distances beyond ~50 angstroms in the phosphorylated kinase bound to AMPPNP is highlighted with darker blue shading (bottom). (b) DEER spectra and distance distributions (inset) for phosphorylated AurA bound to ADP alone (dark blue) or ADP and 10 mM kemptide substrate peptide (light blue).

Figure 4—figure supplement 3. DEER experiments with an alternate spin labeling site support a phosphorylation-driven structural transition in the DFG-In state.

DEER spectra (a) and distance distributions (b) obtained with unphosphorylated (red) and phosphorylated (blue) AurA labeled with MTSL at L225C and S283C and bound to SNS-314. The inset in a shows a schematic representation of the labeling scheme. The arrows in b highlight the increase in distance observed with phosphorylation.