The ascent of AKAPs: from architectural elements to isotype selective PKA scaffolds

Abstract

The spatiotemporal organization of proteins within cells is essential for complex life. The scaffolding of protein kinase A (PKA) by A-kinase anchoring proteins (AKAPs) allows for the management of a multitude of simultaneous signal cascades within animal cells. AKAPs contain an 18-amino-acid-long amphipathic helix that anchors the docking and dimerization (d/d) domain of PKA regulatory subunits. There are four isoforms of the PKA regulatory subunit, RIα, RIβ, RIIα, and RIIβ, which fall into the two categories of type I and type II PKA (the α and β forms of each regulatory subunit share similar topology). There are >60 known AKAPs, and each one prefers to bind either type I or type II PKA. As PKA originated in the metazoan clade, the presence of 'AKAP' proteins in organisms that lack PKA suggests a PKA-independent function. These ancestral anchoring proteins localize to cilia and flagella across all eukaryotic kingdoms and bind docking and dimerization domain proteins that contain a d/d domain nearly identical to that of PKA, fused to other protein domains. In this work, we christen these PKA-independent scaffolds as docking and dimerization domain interacting proteins (DDIPs) and study the co-evolution of d/d proteins, PKA, AKAPs, and DDIPs. Phylogenetic analyses show that PKA and the d/d proteins exhibit telltale signs of divergent evolution. Driven by gene duplication and fusion events, both the structural folds and the primary sequences of these proteins have been highly conserved for over two billion years. The evolutionary path of AKAPs and DDIPs is strikingly different. Phylogenetic analyses reveal patterns of convergent evolution for these anchors. By tracing the evolution of anchoring domains on extant proteins, we reveal a paradigm of accretion of neutral mutations which eventually form a functional anchoring helix. Once this occurs, the mutation rate slows as this protein region becomes subject to selective pressures. Finally, I examine the features of AKAP helices which drive preference for either type I or type II PKA. Guided by bioinformatic analysis, we identified an F/Y, A motif at the N-terminus of all type I-selective AKAPs. We hypothesized that this "aromatic, alanine" motif controls AKAP preference. Using a total internal reflection fluorescence (TIRF)–fluorescence recovery after photobleaching (FRAP) assay, we measured the degree of R subunit binding of the characteristic type I and type II AKAPs, smAKAP and AKAP79. We then applied the "aromatic, alanine" motif to the helix of AKAP79, creating mutants with an "FA" on the N-terminus, an "AF" on the C-terminus, and a version with both terminal mutations known as "FAAF". Immunoprecipitation experiments with these mutants revealed that only the FAAF mutant gains the ability to pull down RI. Molecular dynamics simulations on AKAP79 and AKAP79FAAF reveal a rearrangement of contact points between the helix and the d/d domain which facilitate this preference switch. FRAP experiments confirm that AKAP79FAAF does switch its preference to RI in situ. The significance of this mutation on signaling was confirmed by a rescue of aberrant corticosterone release in an RI-biased Cushing's syndrome model. Finally, we performed a reciprocal experiment in which the 'aromatic, alanine' motif on smAKAP was replaced with two leucines to resemble AKAP79. This mutation was sufficient to impair RI binding by smAKAP in an immunoprecipitation, but did not confer RII binding measured by immunoprecipitation or RII overlay. Together these results clarify the evolutionary paths of these anchoring proteins and their d/d partners, and serve as a call to arms to identify and characterize new DDIPs. The analysis of AKAP preference will assist in the characterization of new AKAPs and reveals a new facet of what makes this binding interface such a powerful cellular tool.