Canonical protein kinase structure and motifs

Protein kinases (like PKA, shown here bound to ATP^1ATP) form a clamshell-like fold that clamps down on substrate. The N-terminal half of the protein (or N-lobe; orange in image) is predominantly β-sheet, while the C-lobe (blue) is mostly α-helical. Most kinase domains range in molecular weight from 30-40 kDa, and many kinase subfamilies have specific insertions. Most, but not all protein kinases evolved from a single common ancestor. Those that did not are known as "atypical" kinases, e.g. α- kinases, Rio, and lipid kinase relatives, such as TOR and its orthologs. Such atypical kinases are currently not represented in this database.

The kinase active site: A set of conserved motifs are required to coordinate 2 bound Mg²⁺ and the nucleotide required for activity. These motifs were first described by Hanks & Hunter, and often referred to by their "Hanks & Hunter subdomain" number. Note that some contacts are made indirectly, via water molecules (red in this image). Other conserved motifs mediate the actual reaction, while additional residues, sometimes distal from the active site, are involved in kinase activation and substrate interaction. Some kinase-fold proteins have naturally occurring substitutions in residues required for activity. Such proteins are known as pseudokinases, and may be involved in regulating the activity of other enzymes (including true kinases). In a few cases, new catalytic activities have been described for apparent pseudokinases.

The Gly-rich loop (green), also called the P-loop or Walker A motif, consists of the motif "GxGxxG". The center glycine is the only residue thought to be required for nucleotide-binding, as some kinases retain activity with substitutions at other positions. The Gly-loop is quite dynamic, and clamps down upon binding of nucleotide (see comparison between apo^1J3H and ATP-bound^1ATP PKA structures, and is often unresolved in crystal structures of apo kinases. While catalytically inactive pseudokinases often have degraded Gly-loops, both the Apicomplexan FIKK and WNG families appear to have lost the Gly-loop structure, but are still highly active kinases.

The "VAIK" motif sits in β-sheet 3 of the kinase N-lobe. The lysine (K) in the motif interacts with the β-phosphate of the bound nucleotide and is thought to facilitate catalysis through this coordination. Note that substitutions at this lysine disrupt nucleotide binding. Even substitution to arginine creates a pseudokinase, as in the well-characterized KSR family. Mutation of the Lys to an alaphatic (e.g. to Met) is a common kinase-dead allele. Note that such mutations block nucleotide binding, which can radically alter protein stability in the cell, and may therefore result in phenotypes that are independent of kinase activity. Remarkably, the well-conserved WNK (With-No-Lys) family of kinases have migrated their VAIK-Lys to β-sheet 2 and retain kinase activity in spite of a substitution that would otherwise render them pseudokinases. The WNK kinases are regulated by Na⁺ binding in the active site, The alanine in the motif is also important, as it points towards the purine ring. Substitution to larger residues disrupts nucleotide binding, though valine is often tolerated. Note that the WNG family require larger residues (e.g. Leu) at this position, likely due to their lack of Gly-loop.

The catalytic base of the active site is formed by the carboxylate of the Asp in the "HRDLKPEN" motif. In an activated kinase, the Asp held in place by the sidechain hydroxyl of the GT-motif Thr (which lies at the end of the activation loop). While there have been controversial reports of kinases with substitutions at this position (e.g. to Asn) retaining activity, this residue is usually thought to be absolutely required for phosphoryl transfer, and is an excellent site to mutate for a kinase-dead allele that does not affect nucleotide-binding. The HRD-Arg is conserved in kinases activated by phosphorylation on their activation loops, and recognizes the phosphorylated residue at this site to lock the activation loop in place (Ser/Thr or Tyr).

The aspartate of the DFG motif (green) is understood to be central to nucleotide binding and catalysis, as it coordinates both of the divalents in the active site. Any substitutions at this position result in a catalytically inactive protein. Intriguingly, the "DFG+1" residue has been implicated as a major determinant in Ser/Thr specificity; large bulky residues at this position (e.g. Phe) lead to preference for Ser, while smaller residues (Leu, Ser, etc.) lead to a preference for Thr as the phosphoacceptor.

There are several other well-conserved motifs in the kinase fold, though most are associated with stabilizing the overall structure rather than playing a catalytic or regulatory role. Examples include the DFW-helix (green), which forms the core of the C-lobe, and the APE-motif (purple) that anchors the C-terminal end of the activation loop to the C-lobe structure. Also notable it the GT-motif lies just N-terminal to the APE-motif and serves to stabilize the HRD-Asp for catalysis (see "HRD-motif" tab).