An interesting model of autoactivation has been recently suggested. A
number of kinases have been reported to form transient dimers in crystal
structures. Intriguingly, these kinases all contain additional domains increasing
local concentration and therefore the dimeric state of the inactive enzyme. In dimers
of these kinases, the activation segment domain-exchanged, thereby forming an
active kinase in a trans configuration, in which the phosphorylation sites are placed
in the active site of the interacting protomer. It has been suggested that this
conformation is important for autoactivation of kinases at non-consensus sites,
offering an explanation of how kinases can active on sites [42–44] (Fig. 4a).
Dimerization of the catalytic domain plays also a role for a number of
diverse kinases, but in contrast to the symmetric dimers reported for kinases
autophosphorylating at non-consensus sites, these kinases form asymmetric
dimers or heterodimers. The receptor tyrosine kinase EGF1R (epidermal growth
factor-1 receptor), for instance, forms homo- or heterodimers with its closely related
family members HER2, HER3 and HER4 including the inactive pseudokinase
HER3 [45, 46]. This activation model represents a refined mechanism of the
canonical ligand-induced receptor activation which has major implications for
our understanding of the mode of action of selective therapeutic antibodies and
kinase drugs [47–49]. Ligand binding to the extracellular domain of EGFR induces
large structural rearrangements and dimerization which orients the catalytic domains
Fig. 3 Architecture of the kinase activation segment. Shown is the phosphorylated activation
segment of PAK4 (PDB: 2CDZ). Highlighted are the activation segment structural elements
including the DFG motif (red), the activation loop region (green) and the P+1 loop (dark green).
The phosphorylation site is shown in ball and stick representation. The APE is indicated
Function, Structure and Topology of Protein Kinases
7
number of kinases have been reported to form transient dimers in crystal
structures. Intriguingly, these kinases all contain additional domains increasing
local concentration and therefore the dimeric state of the inactive enzyme. In dimers
of these kinases, the activation segment domain-exchanged, thereby forming an
active kinase in a trans configuration, in which the phosphorylation sites are placed
in the active site of the interacting protomer. It has been suggested that this
conformation is important for autoactivation of kinases at non-consensus sites,
offering an explanation of how kinases can active on sites [42–44] (Fig. 4a).
Dimerization of the catalytic domain plays also a role for a number of
diverse kinases, but in contrast to the symmetric dimers reported for kinases
autophosphorylating at non-consensus sites, these kinases form asymmetric
dimers or heterodimers. The receptor tyrosine kinase EGF1R (epidermal growth
factor-1 receptor), for instance, forms homo- or heterodimers with its closely related
family members HER2, HER3 and HER4 including the inactive pseudokinase
HER3 [45, 46]. This activation model represents a refined mechanism of the
canonical ligand-induced receptor activation which has major implications for
our understanding of the mode of action of selective therapeutic antibodies and
kinase drugs [47–49]. Ligand binding to the extracellular domain of EGFR induces
large structural rearrangements and dimerization which orients the catalytic domains
Fig. 3 Architecture of the kinase activation segment. Shown is the phosphorylated activation
segment of PAK4 (PDB: 2CDZ). Highlighted are the activation segment structural elements
including the DFG motif (red), the activation loop region (green) and the P+1 loop (dark green).
The phosphorylation site is shown in ball and stick representation. The APE is indicated
Function, Structure and Topology of Protein Kinases
7