It was once thought that proteins needed to have structures that were both ordered and stable, but this view was changed by the discovery that certain proteins contain regions that are disordered and flexible. In some cases these regions of intrinsic disorder help the protein to function by linking more stable regions that are active. However, in other proteins the disordered regions are themselves biologically active and can, for example, function as enzymes.

Protein kinase A is a family of enzymes that contains both ordered and disordered regions, with the ordered sections being involved in phosphorylation, a chemical process that is widely used for communication within cells. However, in order to initiate phosphorylation, these kinases must be anchored to a rigid substrate nearby, so a second group of proteins called AKAPs–which is short for A-kinase anchoring proteins–hold the kinases in place by binding to their disordered regions. These AKAPs also help the kinases to dock with other molecules involved in phosphorylation.

A full structural picture of how the kinases induce phosphorylation has yet to be obtained, partly because it is extremely difficult to determine the structure of the disordered regions within the kinases. Moreover, the AKAPs are also disordered, which makes it difficult to work out how the kinases are held in position.

Smith, Reichow et al. have used electron microscopy to reveal that the disordered region has two important roles: it determines how far away from the anchoring protein that the active region of the kinase can operate, and it influences how efficiently the kinase can bind to its target molecule in order to induce phosphorylation. Future challenges include investigating how the inherent flexibility of AKAP complexes contribute to the efficient phosphorylation of physiological targets.

Intrinsically disordered regions of proteins are widespread in nature, yet the mechanistic roles they play in biology are underappreciated. Such disordered segments can act simply to link functionally coupled structural domains or they can orchestrate enzymatic reactions through a variety of allosteric mechanisms (Dyson and Wright, 2005). The regulatory subunits of protein kinase A provide an example of this important phenomenon where functionally defined and structurally conserved domains are connected by intrinsically disordered regions of defined length with limited sequence identity (Scott et al., 1987). In this study, we show that this seemingly paradoxical amalgam of order and disorder permits fine-tuning of local protein phosphorylation events.

Phosphorylation of proteins is a universal means of intracellular communication that is tightly controlled within the spatial context of the cell. A variety of stimuli trigger these events, which are catalyzed by numerous protein kinases and reversed by phosphoprotein phosphatases (Hunter, 1995). A classic example is production of the second messenger cyclic AMP (cAMP), which stimulates a cAMP-dependent protein kinase (PKA) to phosphorylate a range of cellular targets (Taylor et al., 2012). The PKA holoenzyme is a tetramer composed of two regulatory subunits (R) and two autoinhibited catalytic subunits (PKAc). Binding of cAMP to each R subunit is believed to liberate active kinase and phosphorylation ensues. The local action of PKA is dictated by A-kinase anchoring proteins (AKAPs) that impose spatial constraint by tethering this kinase in proximity to substrates (Wong and Scott, 2004). AKAPs also organize higher-order macromolecular signaling complexes through their association with G-protein coupled receptors, GTPases and additional protein kinases. Likewise, AKAP-associated phosphatases and phosphodiesterases act to locally terminate these signals. While physiological roles for AKAPs that sequester enzymatic activity with ion channels, cytoskeletal components and regulatory enzymes have been well established, the structural mechanisms involved in these protein–protein interactions have been difficult to characterize.

Currently, structural details on PKA anchoring are limited because most AKAPs are large, intrinsically disordered macromolecules that lack recognizable structural domains. An exception is the crystal structure of the central domain of AKAP18γ that bears homology to bacterial 2H phosphoesterase domains (Gold et al., 2008). Likewise, high-resolution crystallographic structures of the catalytic subunit (PKAc) when free and in complex with the C-terminal autoinhibitory and cAMP binding domains of the type I or type II regulatory subunits of PKA (RI and RII) have provided details on the mechanisms of catalysis and autoinhibition (Knighton et al., 1991, 1992; Gold et al., 2006, 2008; Wu et al., 2007). Yet, despite decades of effort, a complete structural picture of the PKA holoenzyme is lacking. This is presumably due to the presence of long flexible intrinsically disordered linker regions within the R subunit that tether this complex. NMR spectroscopy and X-ray crystallographic studies show that the N-terminal domains of RI and RII homodimerize through a four-helix bundle docking and dimerization interface (D/D) (Newlon et al., 2001; Gold et al., 2006; Kinderman et al., 2006; Sarma et al., 2010). The D/D creates a high-affinity binding groove for a canonical amphipathic helix on each AKAP that forms the reciprocal binding surface (Gold et al., 2006).

Although the physiological consequences of anchored PKA phosphorylation events have been established in a variety of cellular contexts, we have yet to discern how the individual protein components are assembled and operate within AKAP complexes (Scott and Pawson, 2009). Even more elusive is a mechanistic role for intrinsically disordered domains within these macromolecular assemblies. For example, does internal flexibility within these signaling assemblies modulate other aspects of enzyme action in addition to sequestering PKA at specific regions of the cell? Here we have used electron microscopy (EM) to evaluate the topological arrangement of the fully assembled type IIα PKA holoenzyme when anchored to AKAP18γ. This analysis unveils a remarkable level of conformational plasticity that resides within the AKAP–PKA complex. In vitro and cell-based structure-function approaches reveal an unexpected functional requirement for intrinsically disordered regions within RIIα. These regions not only define a radius of action of the anchored catalytic subunit but also modulate the enzymatic efficiency of the kinase. Insights from these studies may be broadly applicable to the understanding of other intrinsically disordered proteins and anchoring complexes that organize enzymatic action in the cell.

Macromolecular assemblies were formed when purified human γ isoform of AKAP18 (AKAP18γ) was incubated with the PKA holoenzyme (PKA_holo), formed by the RIIα and PKAc subunits (Figure 1A, ‘Materials and methods’). Following separation by size-exclusion chromatography, fractions from the front of the elution peak (indicated by arrow in Figure 1B) were analyzed by SDS-PAGE (Figure 1C, left). The subunit composition was confirmed by immunoblotting (Figure 1C, mid). Native electrophoresis established that a majority of this material migrated as a single species with an apparent molecular weight in excess of 240 kD (Figure 1C, right). This molecular mass is consistent with a hetero-pentameric complex composed of a single AKAP18γ molecule anchored to an RIIα subunit dimer and two PKAc subunits.

Figure 1

Download asset Open asset

The structure of the AKAP18γ–PKA_holo complex was resolved by electron microscopy of negatively stained particles since these complexes are too small to be imaged in vitrified ice (Figure 1D). Single particle analysis revealed clusters of three densities, each approximately 60–100 Å in size resembling beads on a string (Figure 1D,E). Closer inspection of individual particles established that these tripartite structures adopted a range of conformations (Figure 1E). In solution, these complexes may adopt many more configurations and flattening on the carbon support of the EM grid likely captured only a subset of the possible topologies. Approximately 7,000 particles were selected from electron micrographs for structural analysis. Projection averages were derived by classifying particles of similar orientation and structural conformation using reference-free multivariate statistical analysis (van Heel et al., 1996), and iterative stable alignment and clustering procedures (Yang et al., 2012) (Figure 1F, ‘Materials and methods’). Class averages revealed a remarkable variety of configurations (Figure 1F; Video 1). These ranged from a tightly packed pseudo-symmetric triangular configuration (Figure 1F, left panel) to a fully extended linear configuration of the three densities (Figure 1F, right panel). In either case, the central density was consistently smaller (∼60 Å) than the two peripheral densities (85 × 100 Å). Similar size differences between the central and peripheral lobes were observed at the single particle level in raw micrographs (Figure 1D,E). Affinity-labeling of strep-tagged AKAP18γ with a 5 nm gold particle targeted the smaller central density, indicating that this element includes the anchoring protein (Figure 1D, inset).

Video 1

Download asset

Three-dimensional (3D) reconstructions for the triangular and linear configurations of the AKAP18γ–PKA_holo complex were determined at 35 Å resolution from a tilted-series dataset (Figure 2A, Figure 2—figure supplements 1 and 2). Negative staining in EM can cause flattening of particles, which might lead to apparent structural distortions. However, inspection of particles at various tilt angles showed that particles of linear and triangular conformations remained clearly distinguishable, even at high tilt angles (Figure 2—figure supplement 1). In both 3D reconstructions, a central mass with dimensions of 60 × 60 × 80 Å (corresponding to the site of AKAP18γ as demonstrated by affinity gold labeling, Figure 2A, black triangle and Figure 1D, inset) was flanked on either side by larger densities of 100 × 100 × 85 Å (Figure 2A). These flanking densities can each accommodate a sub-complex of RIIα and PKAc. In the triangular conformation, the two peripheral densities are oriented at a 100° angle with respect to the central density and exhibit an end-to-end length of ∼300 Å (Figure 2A, top). The end-to-end length increases to ∼385 Å in the extended linear configuration (Figure 2A, bottom). Back-projections calculated from the final 3D maps compare well with the experimental class averages (Figure 2—figure supplement 2). Moreover, this back-projection analysis demonstrated that when the triangular model is tilted completely on its edge (where it may appear more linear in projection) its dimensions (max length = 300 Å) are significantly smaller than the maximum end-to-end length obtained for the linear reconstruction (max length = 385 Å). Hence, we conclude that the linear and triangular conformations are structurally distinct.

Figure 2 with 2 supplements see all

Download asset Open asset

Pseudo-atomic models of the pentameric protein assembly in both configurations were constructed by fitting Protein Data Bank coordinates for regions of AKAP18γ and PKA_holo subunits (Knighton et al., 1991, 1992; Gold et al., 2006, 2008; Wu et al., 2007) (Figure 2A, ‘Materials and methods’). A model for AKAP18γ (residues 88–317) was derived by connecting the central domain (residues 88–290) (Gold et al., 2008) via a short linker to the PKA anchoring helix (AKAP_(helix), residues 301–317) (Gold et al., 2006) (Figure 2A, yellow , ‘Materials and methods’). When AKAP18γ is docked to residues 1–43 of RIIα (RII_(D/D)), the resulting sub-structure fits within the central density of maps for both the triangular and extended configurations (Figure 2A, yellow). This tallies with the single particle affinity-labeling studies that map AKAP18γ to this region (Figure 1D, inset). In a similar manner, the peripheral densities each accommodate a PKA sub-structure consisting of one catalytic subunit (Figure 2A, blue) in complex with the cAMP-binding domain of RIIα, residues 91–392 (Figure 2A, green) (Knighton et al., 1991; Wu et al., 2007). Finally these models were completed by connecting the central AKAP18γ–RII_(D/D) sub-complex to each RIIα-PKAc unit through a flexible chain corresponding to residues 44–90 of RIIα (Figure 2A, ‘Materials and methods’). The complete pseudo-atomic model of AKAP18γ–PKA_holo complex is presented in Figure 2B.

Our model of the anchored PKA complex implies that an intrinsically disordered flexible linker region within RIIα supports the array of conformations that were observed in the raw micrographs and the projection averages (Figures 1 and 2C). To quantitatively assess the distribution of conformations assumed by this complex, we measured the end-to-end distance between the two large peripheral densities of 223 individual particles (Figure 2D). This population of structures followed a Gaussian distribution (Figure 2D, green trace) with a mean particle length of 275 ± 65 Å (n = 223). We propose that conformational plasticity observed in these analyses is facilitated by this intrinsically disordered region between residues 44 and 90 of RIIα, a linker that connects the AKAP docking site (D/D domain) to the cAMP-responsive transduction domains. This notion is further substantiated by a primary sequence analysis of RIIα orthologs, showing that the linker regions are of similar length but exhibit low amino acid identity (Figure 3A).

Figure 3

Download asset Open asset

We reasoned that if the linker region in RIIα contributes to conformational flexibility of the holoenzyme, altering its length could affect the structure and function of the anchored kinase. This structural postulate was tested by producing modified AKAP18γ–PKA_holo complexes in which the linker region of RIIα was deleted (RIIα Δ44–86) or replaced with an extended sequence of 60 residues found in zebrafish (RIIα ZeChimera; Figure 3A,B). Formation of modified AKAP–PKA complexes (assembled as described above) was monitored by Coomassie blue staining and immunoblot detection of the component proteins (Figure 3C). The conformations of these modified AKAP18γ–PKA_holo complexes were analyzed by electron microscopy as previously described (Figure 3D–G). Single-particle EM and class averages of the assemblies formed with RIIα Δ44–86 yielded uniform complexes exclusively in a compact triangular configuration (Figure 3D,F). In contrast, complexes formed with RIIα ZeChimera resembled the array of conformations observed for the wild-type AKAP18γ–PKA_holo assembly (Figure 3E,G). Quantitative analysis of the differing linker lengths was assessed by measuring the radius of each particle, defined by the center of the AKAP18γ subunit to the distal end of each PKA subunit. Particle radii for the RIIα Δ44–86 complexes ranged in length from 55 to 125 Å, (mean value of 87 ± 13 Å, n = 142; Figure 3H) whereas the assemblies formed with the RIIα ZeChimera were extended with lengths ranging from 100 to 265 Å (mean value of 168 ± 33 Å, n = 296; Figure 3H). These latter measurements are similar to the parameters of the native complex, (160 ± 29 Å, n = 216; Figures 2D and 3D). Thus we conclude that a flexible linker in RIIα is responsible for the conformational plasticity of the AKAP18γ–PKA_holo assemblies.

One mechanistic ramification of our structural analyses is that flexibility within PKA_holo complex could permit precise orientation of the anchored catalytic subunit toward substrates. This would be particularly true for substrates that are physically associated with AKAPs. For example, the type 4 phosphodiesterase isoforms PDE4D3 and PDE4D5 associate with the central domain of AKAP18γ and are phosphorylated on two sites by PKA (Sette and Conti, 1996; Carlisle Michel et al., 2004; Stefan et al., 2007). We confirmed this protein–protein interaction upon co-expression of the components in HEK293 cells (Figure 4A). PDE4Ds co-precipitate with AKAP18γ and the RII subunits of PKA (Figure 4A, lane 1) but not with GFP controls (Figure 4A, lane 2). Accordingly, cAMP phosphodiesterase activity was enriched 3.27 ± 0.48-fold (n = 5) in AKAP18γ–GFP immune complexes as compared to GFP controls or samples treated with the PDE4 selective inhibitor rolipram (Figure 4B). These data allowed us to move to an in vitro system to study phosphorylation of PDE4D by AKAP18–PKA complexes.

Figure 4 with 1 supplement see all

Download asset Open asset

In vitro phosphorylation studies on AKAP18-associated PDE4D were performed in two phases. Our structural data predict that the C subunit of anchored PKA resides within ∼100 Å of its substrate PDE4D (shown schematically in Figure 4C). We reasoned that this tight configuration could permit cAMP-independent phosphorylation of the phosphodiesterase. Therefore, in the first phase, ‘basal’ phosphorylation of PDE4D was measured in the context of the AKAP18γ signaling complex (Figure 4D). Phosphate incorporation into PDE4D was increased 1.87 ± 0.14-fold (p<0.05) upon its tethering to AKAP18γ as assessed by autoradiography (Figure 4D, E). This suggests that the proximity to a substrate afforded by AKAP18γ augments the cAMP-independent action of PKA. This could explain why physiologically relevant PKA targets such as aquaporins or phospholamban, which require continual or instantaneous phosphorylation to fulfill their biological roles, have their own AKAP-associated pool of kinase (Henn et al., 2004; Lygren et al., 2007; Gold et al., 2012). Accordingly, a primary function of AKAPs may be to ensure basal phosphorylation of such tethered substrates.

In the second phase, experiments were conducted using higher-order complexes formed with wild-type RIIα, RIIα Δ44–86, or the RIIα ZeChimera (assembled as described above) to investigate whether manipulating the intrinsic flexibility of PKA altered phosphorylation of anchored PDE4D (Figure 4F–G). We measured cAMP-independent phosphorylation of PDE4D at 5 min, a time point that showed sub-maximal substrate phosphorylation (Figure 4F,G, Figure 4—figure supplement 1). Basal PDE4D phosphorylation was enhanced 1.97 ± 0.18-fold (n = 6, p<0.05) in complexes formed with RIIα Δ44–86 when compared to a wild-type complex (Figure 4G, bars 1 and 4). In contrast, extension of the linker region in the context of AKAP18γ–RIIα ZeChimera PKA_holo assembly had no effect as compared to wild type (Figure 4G, bars 1 and 7). Control experiments confirmed that addition of cAMP further augmented phosphorylation of PDE4D in all cases (Figure 4F, lanes 2, 5 and 8) and pretreatment with PKI inhibitor peptide abolished anchored kinase activity (Figure 4F, lanes 3, 6 and 9). These data show that the AKAP can be thought of as a catalyst that physically brings the reactants together, and the flexibility within the anchored PKA holoenzyme allows for the precise orientation of the enzyme and substrate. This mechanism may be particularly relevant for cAMP-independent phosphorylation events that are believed to represent approximately 30% of PKA action (Taylor et al., 2012). Therefore, this hitherto unexplained but critical component of cellular PKA activity may be accomplished by the persistent phosphorylation of substrates embedded in higher-order AKAP signaling assemblies.

To follow these in vitro studies, we tested our hypothesis that flexibility within the anchored PKA holoenzyme influences cAMP signaling in living cells. We generated a modified fluorescence resonance energy transfer (FRET) based PKA activity sensor using the A-kinase activity reporter (AKAR2) backbone (Zhang et al., 2005). Our modified sensor (AKAR-18_RBS) was constructed by fusing the PKA binding helix of AKAP18 (18_RBS) (Fraser et al., 1998; Gray et al., 1998) to the amino terminus of AKAR2 (Figure 5A). This genetically encoded reporter detects PKA phosphorylation in real-time by monitoring changes in the YFP/CFP emission ratio inside cells (Figure 5A). As a prelude to these studies AKAR-18_RBS association with wild-type RIIα or either of the modified RIIα constructs was confirmed by co-immunoprecipitation of each complex from HEK293 cells (Figure 5B). In parallel, immunoblot and confocal fluorescent imaging analyses confirmed that mCherry tagged versions of each RIIα form were expressed to equivalent levels (Figure 5C) and uniformly distributed in HEK293 cells (Figure 5—figure supplement 1A).

Figure 5 with 1 supplement see all

Download asset Open asset

Initial experiments evaluated basal FRET of the AKAR-18_RBS reporter using a modified Leica DMI 6000B microscope. The raw YFP/CFP emission ratio of AKAR-18_RBS was elevated in unstimulated cells expressing RIIα Δ44–86 as compared to RIIα wild type or the RIIα ZeChimera (Figure 5D). These data further develop the concept introduced in Figure 4 that the RIIα linker region influences basal phosphorylation of AKAP-associated substrates. Next, real-time changes in the YFP/CFP emission ratio of the AKAR-18_RBS reporter were monitored in cells expressing equivalent levels of the individual RIIα forms (Figure 5E–G). The β-adrenergic (β-AR) agonist isoproterenol (Iso) was administered after 100 s to initiate the cAMP response (Figure 5E–G). Notable increases in AKAR-18_RBS FRET were evident in cells expressing RIIα wild type or the RIIα ZeChimera (Figure 5E,G). In both the cases the normalized FRET ratio was maximal at 200 s and gradually declined over the remainder of the time course (Figure 5H, black and orange traces). This latter phenomenon may be attributed to either desensitization of the β-AR system or dephosphorylation of the reporter by phosphoprotein phosphatases (Bouvier et al., 1987; Pitcher et al., 1992; Violin et al., 2003). Importantly, cells expressing the more compact RIIα Δ44–86 form responded more rapidly and robustly to isoproterenol with a maximal FRET response at 200 s (Figure 5F,H, green trace). The RIIα Δ44–86 effect was abolished when control experiments were performed with a proline modified AKAR-18_RBS derivative that is unable to anchor PKA (Fraser et al., 1998; Gray et al., 1998; Figure 5—figure supplement 1). Other control experiments established that application of isoproterenol did not stimulate an AKAR-18_RBS derivative (T/A) where the threonine phospho-acceptor was replaced with alanine (Figure 5H, inset, red trace).

Scatter plot representation of the Iso stimulated FRET responses at 200 s reveal that AKAR-18_RBS-PKA complexes formed with RIIα or the RIIα ZeChimera responded with similar magnitudes (Figure 5I, black and orange). In contrast, the normalized FRET responses of the more compact and less flexible AKAR-18_RBS-PKA complexes formed with RIIα Δ44–86 were significantly higher (Figure 5I, green, p<0.001). Taken together, these cell-based studies indicate that removal of the flexible linker region in RIIα augments cAMP-responsive PKA phosphorylation of AKAR-18_RBS inside the cells. These findings complement the interpretation of our EM reconstructions and in vitro phosphorylation studies. These observations argue that compact and rigid AKAP–PKA assemblies enhance kinase action within the immediate vicinity of the substrate.

Local activation of PKA is believed to involve dissociation of the C subunits from the R subunit dimer. Yet, cAMP binds to the R subunits avidly (K_D 6–8 nM), and its rate of release is so slow that it is not readily apparent how the cAMP-binding sites turnover within the cell (Poppe et al., 2008). Additionally, there is a report that cAMP can activate PKA without C subunit release (Yang et al., 1995). Therefore, we evaluated whether cAMP stimulation altered the composition of PKA holoenzymes associated with the AKAR-18_RBS reporter. Interestingly, both the RII and C subunits of PKA were present in AKAR-18_RBS immune complexes, even after stimulation of cAMP production by isoproterenol (Figure 5J, lanes 1 and 2). The isoproterenol-stimulated activation of PKA was validated by immunoblot detection of phosphorylated substrates with a phospho-PKA site antibody (Figure 5J, bottom panel). Control experiments confirmed that PKA subunits did not associate with the proline-modified AKAR-18_RBS derivative, whereas the AKAR-18_RBS-T/A reporter retained the ability to anchor PKA (Figure 5J, lanes 3–6).

This AKAR-18_RBS reporter data raised the question of whether anchoring can stabilize the PKA holoenzyme even in the presence of cAMP. To further test this hypothesis, co-precipitation experiments were repeated with full-length AKAP18γ. Again the C subunit was retained within the AKAP complex upon isoproterenol stimulation (Figure 5K, lanes 1 and 2). This occurred regardless of whether the PKA holoenzymes were formed with wild-type RIIα, RIIα Δ44–86, or RIIα ZeChimera (Figure 5K, top panel). These results confirm that the increased agonist-stimulated activity of the AKAR-18_RBS–RIIα Δ44–86 complex shown in Figure 5H,I is not due to enhanced dissociation of the PKA catalytic subunit. A broader implication is that release of the C subunit from anchored PKA holoenyzmes may not be required for the phosphorylation of nearby substrates. Under this scenario, association with AKAPs limits and thereby defines the range of kinase action within the cell. This underscores the central role of AKAPs in the governance of cAMP responsive phosphorylation events.

A combination of steric effects and environmental factors may explain the enhanced catalytic efficiency of this condensed AKAP–PKA assembly inside cells. For example, the less flexible RIIα Δ44–86 dimer may constrain the PKAc subunits in a manner that minimizes the distance to their target substrate. Moreover, the more compact AKAP-PKAΔ44-86-substrate assembly may augment kinase activity by inhibiting the actions of signal termination enzymes. This could involve the steric exclusion of cellular protein phosphatases that catalyze the removal of phosphate or protection from phosphodiesterase activity that metabolizes cAMP. Deletion of the flexible linker between residues 44–86 of RIIα not only enhances basal phosphorylation of these anchored substrates, but also supplements the catalytic efficiency of the anchored PKA holoenzyme in this cellular context. However, this enhanced anchored kinase action may be an undesirable feature in vivo. Hence, we propose that the evolution of flexible linkers in RII may be a means to ensure bi-directional regulation of phosphorylation events by signal termination enzymes.

It is noteworthy that extending the number of residues in the RIIα ZeChimera linker had little effect on the overall structure of the AKAP-PKA complex, or on the activity of PKA monitored by our in vitro and cell-based assays. According to our structural analysis, this construct samples a range of radial conformations that are very similar to wild-type particles (Figure 3H). Moreover, this observation is consistent with theoretical analyses of random coiled chain models that demonstrate the average end-to-end length of random coil scales by the square root of the number of amino acids in the chain (average length ∼ N^0.5, where N is the number of residues) (Flory, 1975). This statement implies that the average length separating the ends of a random coil reaches a plateau, or maximum, that is only marginally altered upon the inclusion of more residues. Consequently extending the linker in the context of the RIIα ZeChimera may explain why this affords minimal benefit to the overall structure and dynamics of this seemingly more extended and malleable anchored enzyme complex. Hence, evolutionary constraint of the RIIα linker domain to 45–47 residues may ensure appropriate separation between the anchoring and transduction domains of anchored mammalian PKA holoenzymes.

Another key element of our model is that the array of conformations adopted by the anchored RII subunit dimer defines a ‘radius of action’ for the tethered catalytic subunit. Not only does this structural plasticity explain why an intact PKA holoenzyme structure has eluded X-ray crystallographers for over 20 years but also suggests that the ‘built in’ flexibility within this protein complex sets a dynamic range for basal and cAMP-dependent phosphorylation of nearby substrates. This latter concept could have profound implications for PKA and other kinases, where enzyme activity is constrained within anchored complexes or enzyme scaffolds. The conformational variability of the tethered complex could enhance kinase access to multiple phosphorylation sites on the same substrate, or, alternatively, orient each catalytic subunit toward distinct substrate proteins within the vicinity of the scaffolding site. Our structural and cellular characterization of higher-order AKAP assemblies emphasizes how static protein–protein interactions act in concert with the dynamic movement of sub-structures to orchestrate local phosphorylation events. We believe that these combined molecular approaches are broadly applicable for the investigation of related enzyme complexes that are spatially organized within the cell.

1. 1. Smith FD
   2. Reichow SL
   3. Esseltine JL
   4. Shi D
   5. Langeberg LK
   6. Scott JD
   7. Gonen T

   (2013) 3D EM reconstruction of the AKAP18-PKA complex in a bent conformation. (corresponding with PDB 3J4Q)

   ID EMD-5755. Publicly Available at Electron Microscopy Data Bank (EMDB).

   http://www.ebi.ac.uk/pdbe/entry/EMD-5755
2. 1. Smith FD
   2. Reichow SL
   3. Esseltine JL
   4. Shi D
   5. Langeberg LK
   6. Scott JD
   7. Gonen T

   (2013) 3D EM reconstruction of the AKAP18-PKA complex in a linear conformation. (corresponding with PDB 3J4R)

   ID EMD-5756. Publicly Available at Electron Microscopy Data Bank (EMDB).

   http://www.ebi.ac.uk/pdbe/entry/EMD-5756
3. 1. Smith FD
   2. Reichow SL
   3. Esseltine JL
   4. Shi D
   5. Langeberg LK
   6. Scott JD
   7. Gonen T

   (2013) Pseudo-atomic model of the AKAP18-PKA complex in a bent conformation derived from electron microscopy

   ID 3J4Q. Publicly Available at Protein Data Bank (PDB).

   http://www.rcsb.org/pdb/explore/explore.do?structureId=3J4Q
4. 1. Smith FD
   2. Reichow SL
   3. Esseltine JL
   4. Shi D
   5. Langeberg LK
   6. Scott JD
   7. Gonen T

   (2013) Pseudo-atomic model of the AKAP18-PKA complex in a linear conformation derived from electron microscopy

   ID 3J4R. Publicly Available at Protein Data Bank (PDB).

   http://www.rcsb.org/pdb/explore/explore.do?structureId=3J4R

1. 1. Bouvier M
   2. Leeb-Lundberg LM
   3. Benovic JL
   4. Caron MG
   5. Lefkowitz RJ

   (1987)

   Regulation of adrenergic receptor function by phosphorylation. II. Effects of agonist occupancy on phosphorylation of alpha 1- and beta 2- adrenergic receptors by protein kinase C and the cyclic AMP-dependent protein kinase

   J Biol Chem 262:3106–3113.

   • Google Scholar
2. 1. Carlisle Michel JJ
   2. Dodge KL
   3. Wong W
   4. Mayer NC
   5. Langeberg LK
   6. Scott JD

   (2004) PKA phosphorylation of PDE4D3 facilitates recruitment of the mAKAP signaling complex

   Biochem J 381:587–592.

   https://doi.org/10.1042/BJ20040846

   • Google Scholar
3. 1. Dyson HJ
   2. Wright PE

   (2005) Intrinsically unstructured proteins and their functions

   Nat Rev Mol Cell Biol 6:197–208.

   https://doi.org/10.1038/nrm1589

   • Google Scholar
4. 1. Emsley P
   2. Cowtan K

   (2004) Coot: model-building tools for molecular graphics

   Acta Crystallogr D Biol Crystallogr 60:2126–2132.

   https://doi.org/10.1107/S0907444904019158

   • Google Scholar
5. 1. Flory PJ

   (1975) Spatial configuration of macromolecular chains

   Science 188:1268–1276.

   https://doi.org/10.1126/science.188.4195.1268

   • Google Scholar
6. 1. Frank J
   2. Radermacher M
   3. Penczek P
   4. Zhu J
   5. Li Y
   6. Ladjadj M

   et al. (1996) SPIDER and WEB: processing and visualization of images in 3D electron microscopy and related fields

   J Struct Biol 116:190–199.

   https://doi.org/10.1006/jsbi.1996.0030

   • Google Scholar
7. 1. Fraser ID
   2. Tavalin SJ
   3. Lester LB
   4. Langeberg LK
   5. Westphal AM
   6. Dean RA

   et al. (1998) A novel lipid-anchored A-kinase anchoring protein facilitates cAMP- responsive membrane events

   EMBO J 17:2261–2272.

   https://doi.org/10.1093/emboj/17.8.2261

   • Google Scholar
8. 1. Gold MG
   2. Lygren B
   3. Dokurno P
   4. Hoshi N
   5. McConnachie G
   6. Tasken K

   et al. (2006) Molecular basis of AKAP specificity for PKA regulatory subunits

   Mol Cell 24:383–395.

   https://doi.org/10.1016/j.molcel.2006.09.006

   • Google Scholar
9. 1. Gold MG
   2. Reichow SL
   3. O’Neill SE
   4. Weisbrod CR
   5. Langeberg LK
   6. Bruce JE

   et al. (2012) AKAP2 anchors PKA with aquaporin-0 to support ocular lens transparency

   EMBO Mol Med 4:15–26.

   https://doi.org/10.1002/emmm.201100184

   • Google Scholar
10. 1. Gold MG
    2. Smith FD
    3. Scott JD
    4. Barford D

    (2008) AKAP18 contains a phosphoesterase domain that binds AMP

    J Mol Biol 375:1329–1343.

    https://doi.org/10.1016/j.jmb.2007.11.037

    • Google Scholar
11. 1. Gray PC
    2. Johnson BD
    3. Westenbroek RE
    4. Hays LG
    5. Yates JR
    6. Scheuer T

    et al. (1998) Primary structure and function of an A kinase anchoring protein associated with calcium channels

    Neuron 20:1017–1026.

    https://doi.org/10.1016/S0896-6273(00)80482-1

    • Google Scholar
12. 1. Grigorieff N

    (2007) FREALIGN: high-resolution refinement of single particle structures

    J Struct Biol 157:117–125.

    https://doi.org/10.1016/j.jsb.2006.05.004

    • Google Scholar
13. 1. Henn V
    2. Edemir B
    3. Stefan E
    4. Wiesner B
    5. Lorenz D
    6. Theilig F

    et al. (2004) Identification of a novel A-kinase anchoring protein 18 isoform and evidence for its role in the vasopressin-induced aquaporin-2 shuttle in renal principal cells

    J Biol Chem 279:26654–26665.

    https://doi.org/10.1074/jbc.M312835200

    • Google Scholar
14. 1. Hill EV
    2. Houslay MD
    3. Baillie GS

    (2005) Investigation of extracellular signal-regulated kinase 2 mitogen-activated protein kinase phosphorylation and regulation of activity of PDE4 cyclic adenosine monophosphate-specific phosphodiesterases

    Methods Mol Biol 307:225–237.

    https://doi.org/10.1385/1-59259-839-0:225

    • Google Scholar
15. 1. Hunter T

    (1995) Protein kinases and phosphatases: the yin and yang of protein phosphorylation and signaling

    Cell 80:225–236.

    https://doi.org/10.1016/0092-8674(95)90405-0

    • Google Scholar
16. 1. Kinderman FS
    2. Kim C
    3. von Daake S
    4. Ma Y
    5. Pham BQ
    6. Spraggon G

    et al. (2006) A dynamic mechanism for AKAP binding to RII isoforms of cAMP-dependent protein kinase

    Mol Cell 24:397–408.

    https://doi.org/10.1016/j.molcel.2006.09.015

    • Google Scholar
17. 1. Knighton DR
    2. Zheng J
    3. Ten Eyck LF
    4. Ashford VA
    5. Xuong N-H
    6. Taylor SS

    et al. (1992) Crystal structure of the catalytic subunit of cyclic adenosine monophosphate-dependent protein kinase

    Science 253:407–414.

    https://doi.org/10.1126/science.1862342

    • Google Scholar
18. 1. Knighton DR
    2. Zheng J
    3. Ten Eyck LF
    4. Zuong N-H
    5. Taylor SS
    6. Sowadski JM

    (1991) Structure of a peptide inhibitor bound to the catalytic subunit of cyclic adenosine monophosphate-dependent protein kinase

    Science 253:414–420.

    https://doi.org/10.1126/science.1862343

    • Google Scholar
19. 1. Lygren B
    2. Carlson CR
    3. Santamaria K
    4. Lissandron V
    5. McSorley T
    6. Litzenberg J

    et al. (2007) AKAP complex regulates Ca2+ re-uptake into heart sarcoplasmic reticulum

    EMBO Rep 8:1061–1067.

    https://doi.org/10.1038/sj.embor.7401081

    • Google Scholar
20. 1. Mindell JA
    2. Grigorieff N

    (2003) Accurate determination of local defocus and specimen tilt in electron microscopy

    J Struct Biol 142:334–347.

    https://doi.org/10.1016/S1047-8477(03)00069-8

    • Google Scholar
21. 1. Narayana N
    2. Cox S
    3. Shaltiel S
    4. Taylor SS
    5. Xuong N

    (1997) Crystal structure of a polyhistidine-tagged recombinant catalytic subunit of cAMP-dependent protein kinase complexed with the peptide inhibitor PKI(5-24) and adenosine

    Biochemistry 36:4438–4448.

    https://doi.org/10.1021/bi961947+

    • Google Scholar
22. 1. Newlon MG
    2. Roy M
    3. Morikis D
    4. Carr DW
    5. Westphal R
    6. Scott JD

    et al. (2001) A novel mechanism of PKA anchoring revealed by solution structures of anchoring complexes

    EMBO J 20:1651–1662.

    https://doi.org/10.1093/emboj/20.7.1651

    • Google Scholar
23. 1. Pettersen EF
    2. Goddard TD
    3. Huang CC
    4. Couch GS
    5. Greenblatt DM
    6. Meng EC

    et al. (2004) UCSF Chimera–a visualization system for exploratory research and analysis

    J Comput Chem 25:1605–1612.

    https://doi.org/10.1002/jcc.20084

    • Google Scholar
24. 1. Pitcher J
    2. Lohse MJ
    3. Codina J
    4. Caron MG
    5. Lefkowitz RJ

    (1992) Desensitization of the isolated beta 2-adrenergic receptor by beta-adrenergic receptor kinase, cAMP-dependent protein kinase, and protein kinase C occurs via distinct molecular mechanisms

    Biochemistry 31:3193–3197.

    https://doi.org/10.1021/bi00127a021

    • Google Scholar
25. 1. Poppe H
    2. Rybalkin SD
    3. Rehmann H
    4. Hinds TR
    5. Tang XB
    6. Christensen AE

    et al. (2008) Cyclic nucleotide analogs as probes of signaling pathways

    Nat Methods 5:277–278.

    https://doi.org/10.1038/nmeth0408-277

    • Google Scholar
26. 1. Sarma GN
    2. Kinderman FS
    3. Kim C
    4. von Daake S
    5. Chen L
    6. Wang BC

    et al. (2010) Structure of D-AKAP2:PKA RI complex: insights into AKAP specificity and selectivity

    Structure 18:155–166.

    https://doi.org/10.1016/j.str.2009.12.012

    • Google Scholar
27. 1. Schneider CA
    2. Rasband WS
    3. Eliceiri KW

    (2012) NIH Image to ImageJ: 25 years of image analysis

    Nat Methods 9:671–675.

    https://doi.org/10.1038/nmeth.2089

    • Google Scholar
28. 1. Scott JD
    2. Pawson T

    (2009) Cell signaling in space and time: where proteins come together and when they’re apart

    Science 326:1220–1224.

    https://doi.org/10.1126/science.1175668

    • Google Scholar
29. 1. Scott JD
    2. Zoller MJ
    3. Glaccum MB
    4. Uhler MD
    5. Helfman DM
    6. McKnight GS

    et al. (1987) The molecular cloning of a type II regulatory subunit of the cAMP-dependent protein kinase from rat skeletal muscle and mouse brain

    Proc Natl Acad Sci USA 84:5192–5196.

    https://doi.org/10.1073/pnas.84.15.5192

    • Google Scholar
30. 1. Sette C
    2. Conti M

    (1996) Phosphorylation and activation of a cAMP-specific phosphodiesterase by the cAMP-dependent protein kinase

    J Biol Chem 271:16526–16534.

    https://doi.org/10.1074/jbc.271.28.16526

    • Google Scholar
31. 1. Smith JM

    (1999) Ximdisp–A visualization tool to aid structure determination from electron microscope images

    J Struct Biol 125:223–228.

    https://doi.org/10.1006/jsbi.1998.4073

    • Google Scholar
32. 1. Stefan E
    2. Wiesner B
    3. Baillie GS
    4. Mollajew R
    5. Henn V
    6. Lorenz D

    et al. (2007) Compartmentalization of cAMP-dependent signaling by phosphodiesterase-4D is involved in the regulation of vasopressin-mediated water reabsorption in renal principal cells

    J Am Soc Nephrol 18:199–212.

    https://doi.org/10.1681/ASN.2006020132

    • Google Scholar
33. 1. Taylor SS
    2. Ilouz R
    3. Zhang P
    4. Kornev AP

    (2012) Assembly of allosteric macromolecular switches: lessons from PKA

    Nat Rev Mol Cell Biol 13:646–658.

    https://doi.org/10.1038/nrm3432

    • Google Scholar
34. 1. van Heel M
    2. Harauz G
    3. Orlova EV
    4. Schmidt R
    5. Schatz M

    (1996) A new generation of the IMAGIC image processing system

    J Struct Biol 116:17–24.

    https://doi.org/10.1006/jsbi.1996.0004

    • Google Scholar
35. 1. Violin JD
    2. Zhang J
    3. Tsien RY
    4. Newton AC

    (2003) A genetically encoded fluorescent reporter reveals oscillatory phosphorylation by protein kinase C

    J Cell Biol 161:899–909.

    https://doi.org/10.1083/jcb.200302125

    • Google Scholar
36. 1. Wong W
    2. Scott JD

    (2004) AKAP Signalling complexes: focal points in space and time

    Nature Rev Mol Cell Biol 5:959–971.

    https://doi.org/10.1038/nrm1527

    • Google Scholar
37. 1. Wu J
    2. Brown SH
    3. von Daake S
    4. Taylor SS

    (2007) PKA type IIalpha holoenzyme reveals a combinatorial strategy for isoform diversity

    Science 318:274–279.

    https://doi.org/10.1126/science.1146447

    • Google Scholar
38. 1. Yang S
    2. Fletcher WH
    3. Johnson DA

    (1995) Regulation of cAMP-dependent protein kinase: enzyme activation without dissociation

    Biochemistry 34:6267–6271.

    https://doi.org/10.1021/bi00019a002

    • Google Scholar
39. 1. Yang Z
    2. Fang J
    3. Chittuluru J
    4. Asturias FJ
    5. Penczek PA

    (2012) Iterative stable alignment and clustering of 2D transmission electron microscope images

    Structure 20:237–247.

    https://doi.org/10.1016/j.str.2011.12.007

    • Google Scholar
40. 1. Zhang J
    2. Hupfeld CJ
    3. Taylor SS
    4. Olefsky JM
    5. Tsien RY

    (2005) Insulin disrupts beta-adrenergic signalling to protein kinase A in adipocytes

    Nature 437:569–573.

    https://doi.org/10.1038/nature04140

    • Google Scholar

Author details

1. F Donelson Smith

   Department of Pharmacology, Howard Hughes Medical Institute, University of Washington, Seattle, United States

   Contribution

   FDS, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

   Contributed equally with

   Steve L Reichow

   Competing interests

   The authors declare that no competing interests exist.

2. Steve L Reichow

   Janelia Farm Research Campus, Howard Hughes Medical Institute, Ashburn, United States

   Contribution

   SLR, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

   Contributed equally with

   F Donelson Smith

   Competing interests

   The authors declare that no competing interests exist.

3. Jessica L Esseltine

   Department of Pharmacology, Howard Hughes Medical Institute, University of Washington, Seattle, United States

   Contribution

   JLE, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

   Competing interests

   The authors declare that no competing interests exist.

4. Dan Shi

   Janelia Farm Research Campus, Howard Hughes Medical Institute, Ashburn, United States

   Contribution

   DS, Conception and design, Acquisition of data

   Competing interests

   The authors declare that no competing interests exist.

5. Lorene K Langeberg

   Department of Pharmacology, Howard Hughes Medical Institute, University of Washington, Seattle, United States

   Contribution

   LKL, Conception and design, Drafting or revising the article

   Competing interests

   The authors declare that no competing interests exist.

6. John D Scott

   Department of Pharmacology, Howard Hughes Medical Institute, University of Washington, Seattle, United States

   Contribution

   JDS, Conception and design, Analysis and interpretation of data, Drafting or revising the article

   For correspondence

   scottjdw@u.washington.edu

   Competing interests

   The authors declare that no competing interests exist.

7. Tamir Gonen

   Janelia Farm Research Campus, Howard Hughes Medical Institute, Ashburn, United States

   Contribution

   TG, Conception and design, Analysis and interpretation of data, Drafting or revising the article

   For correspondence

   gonent@janelia.hhmi.org

   Competing interests

   The authors declare that no competing interests exist.

• 2,223

  views

• 335

  downloads

• 109

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.