Finding. Protein kinase A is not the same object on every anchor. Park the type-II regulatory subunit on AKAP1 and you get phosphorylation-dependent control of mitochondrial morphology. Park it on MAP2 and you get straighter microtubules, wider bundles, and more dendritic arbor, without needing the kinase to phosphorylate, but needing the second cAMP pocket intact. A SpyCatcher-SpyTag swap is what makes those two sentences comparable in an otherwise intact cell.
Why this paper matters
AKAPs have been “where the kinase sits.” This paper says some of those seats are not for catalysis. The mitochondrial seat (AKAP1) still is. That split is the brief: organelle shape is an enzymatic PKA job; microtubule geometry can be a structural one.
The method is as important as the split. Site-defined plus/minus RII was the missing control.
What they actually measured
Replacement of RII anchoring on AKAP79, AKAP1, and MAP2; phospho-dependent membrane and mitochondrial-morphology effects; phospho-independent MAP2 microtubule/dendrite effects; CNB2 pocket requirement; native MAP2–RII recapitulation.
How to read the score
Mid-to-high seventies. Clean compartment logic, a real mito phenotype, a reusable tool. Confidence is medium. Score 76. Heuristic 55 undersold it.
What to do with it
If you study AKAP1, mitochondrial shape, or dendritic PKA, pull the replacement assays. Do not treat every RII blot as a kinase activity map. The directional implication is that anchored PKA can be an enzyme at mitochondria and a cAMP-gated structural brace on MAP2.