An autoinhibitory tyrosine motif in the cell-cycle-regulated Nek7 kinase is released through binding of Nek9

Abstract

Mitosis is controlled by multiple protein kinases, many of which are abnormally expressed in human cancers. Nek2, Nek6, Nek7, and Nek9 are NIMA-related kinases essential for proper mitotic progression. We determined the atomic structure of Nek7 and discovered an autoinhibited conformation that suggests a regulatory mechanism not previously described in kinases. Additionally, Nek2 adopts the same conformation when bound to a drug-like molecule. In both structures, a tyrosine side chain points into the active site, interacts with the activation loop, and blocks the alphaC helix. Tyrosine mutants of Nek7 and the related kinase Nek6 are constitutively active. The activity of Nek6 and Nek7, but not the tyrosine mutant, is increased by interaction with the Nek9 noncatalytic C-terminal domain, suggesting a mechanism in which the tyrosine is released from its autoinhibitory position. The autoinhibitory conformation is common to three Neks and provides a potential target for selective kinase inhibitors.

Figure 1

The Crystal Structure of Nek7 Shows an Autoinhibited Conformation that Suggests a Regulatory Mechanism Not Previously Described in Kinases (A) Secondary structure and key residues mapped onto the protein sequence of Nek7. Disordered regions are marked with a dashed line, and ordered regions are marked with a solid line, black arrow, or white rectangle to indicate random coil, β strand, and α helix respectively. Lys63 and Glu82 are marked with red boxes, Leu86 and Leu111 with gray boxes, Tyr97 with a cyan box, and the DLG motif with a green box. (B) Overview of ADP-bound Nek7 structure in cartoon representation. The side chain of Tyr97 and the ADP ligand (magenta) are shown as spheres. The dashed box indicates the region magnified in (C). (C) Cartoon and stick representation of Tyr97 and its surrounding environment. The H-bond between Tyr97 and the backbone of Leu180 is shown as a dashed line.

Figure 2

Nek9-CTD or Mutations at Tyr97^Nek7 and Tyr108^Nek6 Increase Kinase Activity (A) In vitro kinase activity assay shows the relative activity of wild-type Nek7 (Nek7-WT) and the Tyr97-to-Ala mutant (Nek7-Y97A) in the presence or absence of GST-tagged Nek9 CTD (Nek9-CTD) or control GST. The upper panels show an autoradiograph (³²P) of a Coomassie blue-stained gel (CB) in a single assay. The histograms in the lower panel show the average of three independent in vitro kinase assay experiments, and the error bars show the standard deviation. (B) In vitro kinase activity assay shows the activity of Nek7 and Nek6 in the presence or absence of GST-tagged Nek9 CTD (Nek9-CTD) or control GST. The upper panels show an autoradiograph (³²P) of a CB-stained gel in a single assay. The histograms in the lower panel show the average of three independent in vitro kinase assay experiments, and the error bars show the standard deviation. (C) Shown is cell-cycle-specific kinase activity of Flag-tagged, wild-type Nek7 and Nek6 (Nek7-WT, Nek6-WT), Tyr97/Tyr108-to-Ala mutant (Nek7-Y97A, Nek6-Y108A), or untransfected control (-ve) immunoprecipitated from 293 cell lysates synchronized in G1/S or M phase. The upper panel shows an immunoblot of the kinase protein levels and the result of a single assay as an autoradiograph (³²P). The histograms in the lower panel show the average of three independent kinase assay experiments in which the amount of β-casein phosphorylation was normalized for the amount of protein immunoprecipitated. Error bars show standard deviation.

Figure 3

Comparison of the Nek7 Structure with that of an Active Kinase Suggests that Tyr97 Blocks the Active Position of the αC Helix (A) Superposition of Nek7 (orange) centered on Tyr97 with Plk1 in an active conformation (green, PDB code 2V5Q) shows the key residues that must move for Nek7 to adopt an active conformation. (B) Schematic representation of the differences between Nek7 (left, orange) and Plk1 (green, right) shown in a similar view to that in (A). The two secondary structure elements, strand β4 and helix αC, are shown as a rectangle and circle, respectively. Equivalent residues are shown as black lines, hydrophobic interactions are shown as transparent gray ellipses, and an electrostatic interaction that correctly positions the catalytic Lys residue in the active Plk1 structure is shown as a transparent purple star. (C) In vitro kinase activity assay shows the relative activity of wild-type Nek7 (Nek7-WT), the activating Y97A mutant (Nek7-Y97A), and a series of mutants designed to investigate the structural determinants for Tyr97 autoinhibition. The histograms show the average of four independent experiments, and the error bars show the standard deviation.

Figure 4

Crystal Structure of Nek2 Bound to a Drug-like Molecule that Induces the Tyr-Down Conformation (A) Nek2-CCT241950 structure shown in cartoon representation in the equivalent view to that shown in Figure 1B. CCT241950 and Tyr70^Nek2 are shown as spheres. The dashed box indicates the region magnified in (C). (B) Chemical structure of CCT241950 with potential H-bonds marked as dashed lines. (C) Cartoon and stick representation of Tyr70^Nek2 and its surrounding environment. The putative H-bonds that form a network between CCT241950 and the protein are shown as dashed lines.

Figure 5

Comparison of Nek7 and Nek2 Structures Reveals the Local Structural Differences between Tyr-Up and Tyr-Down Conformations (A) Superposition of Nek7 (orange) centered on Tyr97 with Nek2-ADP (yellow, PDB code 2W5A) and Nek2-CCT241950 (pink) shows the local structural differences between Tyr-down (Nek7, Nek2-CCT241950) and Tyr-up (Nek2-ADP) conformations. Nek7 residues are labeled in orange, and Nek2 residues are labeled in black. (B) Schematic showing the differences between Nek7 in the Tyr-down conformation, and Nek2 in the Tyr-up and Tyr-down conformations. H-bonds formed by the tyrosine side chain in the Tyr-down conformation are marked with a red dashed line. The tyrosine side chain does not form an H-bond with the protein in the Tyr-up conformation. (C) Surface representation of Nek7-ADP (left, orange), Nek2-ADP (center, yellow), and Nek2-CCT241950 (right, pink) viewed from above αC/β4. The Tyr-down conformation of Nek7-ADP does not create a surface cavity in the same position as that observed in the Nek2-CCT241950 structure. The surface generated by αC helix residues is colored a shade lighter than the surrounding surface.

Figure 6

Nek7 Encodes an Unusual NTE Motif that Is Required for Normal Kinase Activity (A) Sequence alignment of human Nek7, Nek6, and Nek2. Sequence conservation between Nek7 and Nek6 is highlighted in orange (identical), yellow (conservative substitution), and white (non-conserved). Residues that are identical in Nek2 are marked with asterisks. The ordered Nek7 NTE is marked with a gray box. Starting residues of Nek7 N-terminally truncated proteins are marked with black bars. Residues mutated in this study are marked with colored triangles: Tyr28^Nek7 and Leu31^Nek7, magenta; Asn33^Nek7, black; Tyr97^Nek7, cyan. (B) Stereo pair diagram depicts the unusual NTE motif in cartoon and sticks colored by conservation using the color scheme defined in (A). Residues mutated in this study are marked with colored triangles. (C) Sequence conservation between Nek7 and Nek6 mapped onto the surface of Nek7 using the color scheme defined in (A) in two views related by 180°. (D) In vitro kinase activity assay shows the relative activity of wild-type Nek7 (WT), the N-terminal truncations (Δ20, Δ30), and the NTE mutations N33D (ND) and Y28A/L31A (YA/LA). The upper panel shows the autoradiograph (³²P) of a CB-stained gel in a single assay. The histograms in the lower panel show the average of three independent in vitro kinase assay experiments, and the error bars show the standard deviation. (E) Anti-His₆ immunoblot of GST coprecipitation experiments using GST-tagged 9-CTD and GST control proteins to capture wild-type and mutant His₆-tagged Nek7 proteins.