Capturing Autoinhibited\nPDK1 Reveals the Linker’s\nRegulatory Role, Informing Innovative Inhibitor Design

PDK1 is crucial for PI3K/AKT/mTOR and Ras/MAPK cancer\nsignaling.\nIt phosphorylates AKT in a PIP<sub>3</sub>-dependent but S6K, SGK,\nand RSK kinases in a PIP<sub>3</sub>-independent manner. Unlike its\nsubstrates, its autoinhibited monomeric state has been unclear, likely\ndue to its low population time, and phosphorylation in the absence\nof PIP<sub>3</sub> has been puzzling too. Here, guided by experimental\ndata, we constructed models and performed all-atom molecular dynamics\nsimulations. In the autoinhibited PDK1 conformation that resembles\nautoinhibited AKT, binding of the linker between the kinase and PH\ndomains to the PIF-binding pocket promotes the formation of the Glu<sup>130</sup>-Lys<sup>111</sup> salt bridge and weakens the association\nof the kinase domain with the PH domain, shifting the population from\nthe autoinhibited state to states accessible to the membrane and its\nkinase substrates. The interaction of the substrates’ hydrophobic\nmotif and the PDK1 PIF-binding pocket facilitates the release of the\nautoinhibition even in <i>the absence of PIP</i><sub><i>3</i></sub>. Phosphorylation of the serine-rich motif within\nthe linker further attenuates the association of the PH domain with\nthe kinase domain. These suggest that while the monomeric autoinhibited\nstate is relatively stable, it can readily shift to its active, catalysis-prone\nstate to phosphorylate its diverse substrates. Our findings reveal\nthe PDK1 activation mechanism and discover the regulatory role of\nPDK1’s linker, which lead to two innovative linker-based inhibitor\nstrategies: (i) locking the autoinhibited PDK1 through optimization\nof the interactions of AKT inhibitors with the PH domain of PDK1 and\n(ii) analogs (small molecules or peptidomimetics) that mimic the linker\ninteractions with the PIF-binding pocket.