Lessons Learned from PKA: Assembly of Dynamic Macromolecular Switches

摘要

ABSTRACT My history with cAMP-dependent protein kinase (PKA) and NIGMS, from active site labeling to holoenzyme structures and tissue imaging, has been long and productive. My career has been guided by the fundamental principle that structure will reveal function with the ultimate goal being to elucidate how PKA signaling regulates biology and how it is altered in disease. Our tools include biochemistry, biophysics, and molecular biology to probe mechanisms as well as crystallography, cryoEM, molecular dynamics, and imaging to explore conforma- tional space and localization in cells. A hallmark of my laboratory has been to build interdisciplinary teams that reach across all of these scales. Although it has been over 30 years now since we solved that first protein kinase structure of the PKA catalytic (C) subunit, which has served ever since as the prototypical protein kinase, surprisingly we are still learning new things. PKA signaling in cells is mediated by full-length R2C2 holoenzymes that are targeted to discreet sites in the cell near dedicated substrates, and a major recent achievement was our solving the cryoEM structure of the compact full-length RII holoenzyme in 2020 where for the first time all of the domains could be visualized. During this next phase we will continue with our characterization of holoenzyme complexes focusing, in particular, on RIIβ, which is enriched in neurons and localizes to Golgi. In addition, however, we will build on two new discoveries that came from our work over the past three years. First is the discovery that Cβ subunits, a family of previously unexplored splice variants that account for ~50% of PKA signaling in neurons, are linked to a neurodegenerative phenotype that abolishes Sonic hedgehog (Shh) signaling. With imaging in human retina we then validated that Cβ is highly expressed in neurons, that it localizes differently than C, and that Cβ4/Cβ4ab are enriched at mitochondria. We are now characterizing the neuron-specific Cβ4 isoforms and the specific mutants that correlate with Shh signaling. Another new and potentially related discovery is that there is a functional PKI-like sequence embedded in the C-terminal tail of Smoothened, the GPCR that is associated with Shh signaling. A final discovery that the RI subunit undergoes liquid:liquid phase separation that contributes to cAMP buffering in cells opens another new frontier for non-canonical PKA signaling in cells. We find that RIβ also forms biomolecular condensates that are distinct from RIα, and we are now characterizing an RIβ mutant, RIβ(R335W), that is associated with dementia and autism. An essential part of our strategy is to use a multi-scale approach that includes not only biochemical characterizations and structure solutions but also high-resolution imaging in human tissues where we can hopefully correlate changes in localization and expression with pathogenic mutations in Cβ and RIβ. In parallel, we will build on our cryo-EM structure of the full length RIIβ holoenzyme where we hope to trap some of the domain dynamics that contribute to the highly allosteric and isoform-specific cAMP-mediated activation of each holoenzymes. With our exceptional team of collaborators we are poised to make rapid progress.