Fig 1.
Crystal structures of the RIIβ holoenzyme showing active site phosphorylation in the presence of calcium.
(A) In the apo RIIβ2:C2 holoenzyme, the β4–β5 loop from one heterodimer pushes onto the C-tail of the opposite dimer, forcing the C-subunit into a closed conformation [19]. (B) By diffusing adenosine triphosphate and calcium (CaATP) into the crystals, we trapped ADP, pRIIβ, and two Ca2+ ions. (C) Reaction products trapped in the crystal.
Fig 2.
Steady-state kinetics of the catalytic subunit of PKA in the presence of Mg2+ versus Ca2+.
(A) Domain organization of the four isoforms of R-subunits. The inhibitor sequence contains a pseudosubstrate site in type I R-subunits, while type II R-subunits have a substrate site. (B) Steady-state kinetics (time course) of phosphoryl transfer on the substrate site of RIIβ (40 μM) by C-subunit in the presence of Mg2+ and Ca2+. Reactions were carried out both in the presence and absence of cAMP. C-subunit was first used at 5 nM for both MgATP and CaATP. However, because of the low efficiency of the C-subunit in the presence of calcium, the measurements fell in the range of low accuracy of the scintillation counter. Hence, 50-fold higher C-subunit concentration (250 nM) was used to measure phosphorylation in the presence of CaATP and cAMP (inset) to obtain accurate results. (C) Steady-state kinetic parameters obtained for phosphoryl transfer to the peptide Kemptide in the presence of Mg2+ versus Ca2+. (D) and (E) Time course and Michealis-Menten curves for phosphoryl transfer using Kemptide as substrate. Inset shows Kemptide phosphorylation in the presence of 50-fold more C-subunit with CaATP (as compared to MgATP). The data used to make this figure can be found in S1 Data.
Fig 3.
Single turnover and pre-steady-state kinetics for RIIβ holoenzyme phosphorylation in the presence of Mg2+ versus Ca2+.
(A) Holoenzyme phosphorylation can be followed by gel electrophoresis in which the pRIIβ is detected by antibodies. The C-subunit phosphorylates RIIβ using MgATP or CaATP in 30 min. Shown is a representative western blot in which the C-subunit was incubated with RIIβ alone, with RIIβ and MgATP, or with RIIβ and CaATP. The starting RIIβ sample was loaded on the last lane showing no phosphorylation before incubation with C-subunit and nucleotide. Also, no phosphorylation was observed when the C-subunit and RIIβ were incubated without nucleotide, suggesting that it is not ATP contamination of either protein sample that contributes to RIIβ phosphorylation. (B) Shown is a western blot in which the C-subunit was incubated with RIIβ alone, with RIIβ and AMP-PNP, or with RIIβ and ATP. Mg2+ or Ca2+ were added to the reaction buffer, which contains nucleotides. The phosphorylation status of RIIβ was monitored at 30 min, 1 d, and 2 d. The sample with AMP-PNP shows phosphorylation of RIIβ that increases over time. (C) Pre-steady-state kinetics of RIIβ holoenzyme in the presence of Mg2+ and Ca2+. Holoenzyme phosphorylation as followed by rapid quench analysis.
Fig 4.
Single turnover condition for phosphoryl transfer is achieved in the PKA RIIβ holoenzyme in the presence of Mg2+ and Ca2+.
(A) Formation of the RIIβ2:C2 holoenzymes leads to a special single turnover condition such that every catalytic subunit phosphorylates a stoichiometric 1:1 molecule of RIIβ. (B) Representative western blot for the single turnover seen in the RIIβ:C complex, with increasing ratios of RIIβ:C-subunit (in the absence of cAMP). Both Mg2+ and Ca2+ support the single turnover phosphoryl transfer as the holoenzymes forms at a 1:1 ratio of RIIβ:C-subunit. Addition of cAMP (1 mM) breaks the holoenzyme, and hence, the single turnover condition is lost. Band intensities for the western blots have been quantified using ImageJ software (see S4 Fig). (C) Comparison of single turnover seen in the RIIβ2:C2 holoenzyme as a function of time for MgATP versus CaATP as monitored by radio-labelled P32-ATP as a function of time. Both Mg and Ca support the single turnover condition, although calcium is slower than Mg in its phosphoryl transfer. The data used to make this figure can be found in S1 Data.
Fig 5.
Comparison of RIIβ:C heterodimer formation and dissociation in the presence of MgATP, CaATP, or EDTA/EGTA.
(A) Injection of 30 nM C-subunit over immobilized GST-RIIβ 102–416 showed complex formation under all conditions tested. The low dissociation phase (without cAMP) did not change significantly whether 10 mM Mg2+ or Ca2+ were used as cofactors, although Mg2+ seems to enhance the association of the complex. Holoenzyme dissociation was then induced by the injection of 100 nM cAMP in the respective buffer, resulting in a slower off-rate for holoenzyme complexes formed in the presence of Ca2+ compared to Mg2+. (B) The opposite effect can be observed for the P-site mutant GST-RIIβ 102–416 S112A, in which calcium enhances holoenzyme dissociation upon cAMP binding. Thus, the trapped phosphoryl group in combination with calcium is responsible for the decelerated dissociation. Moreover, both effects are further enhanced with increasing concentrations of cAMP for wild type (C) as well as for the S112A mutant (D). Error bars represent the standard deviation (SD) of two independent measurements. The data used to make this figure can be found in S1 Data.
Fig 6.
Influence of the inhibitor site phosphorylation on reassociation of the RIIβ holoenzyme.
SPR was used to monitor holoenzyme formation (RIIβ:C) and reassociation (pRIIβ:C) in the presence of 10 mM MgCl2 and 1 mM ATP. (A) Binding of PKA-C to immobilized nonphosphorylated GST-RIIβ 102–416 resulted in strong binding with extremely low off-rates, indicating a highly stable complex. (B) For the analysis of the reassociation of phosphorylated RIIβ with C-subunit, preformed holoenzyme (pRIIβ:C:MgADP) was immobilized on the chip surface, and the catalytic subunit was dissociated by the injection of 500 nM cGMP. cGMP instead of cAMP was used to easily release the cyclic nucleotide after holo-dissociation. cAMP has a much higher affinity to the cyclic nucleotide binding sites and cannot be removed even after extensive dialysis. After removal of cGMP, the reassociation of the holoenzyme was monitored again in the presence of 10 mM MgCl2 and 1 mM ATP. As can be seen in (C), association of the catalytic subunit was slowed down by a factor of almost 60 compared with unphosphorylated RIIβ. However, repeating the experiments in the absence of nucleotides/metals (EDTA) and in the presence of 10 mM MgCl2 and 1 mM ADP revealed only slight differences in kon. The data used to make this figure can be found in S1 Data.
Table 1.
SPR-derived rate and equilibrium binding constants of the interaction of PKA C-subunit with phospho- and nonphospho forms of RIIβ.
A series dilution of PKA-C (0.3–200 nM) was injected over the phospho- and nonphospho forms of immobilized GST-RIIβ 102–416 (80–120 resonance units [RUs]) in the presence of MgATP, MgADP, and chelators (EDTA) only. BIAevalution 4.1.1 (GE Healthcare) was utilized to determine the rate constants kon and koff using a global fit analysis assuming a 1:1 Langmuir binding.
Fig 7.
Endogenous RIIβ is constitutionally phosphorylated in 3T3 and HeLa cells.
pH gradient is shown horizontally, and molecular weight (MW) gradient is shown vertically. Left panel: The purified pRIIβ and RIIβ protein mixture, prepared in 2-D electrophoresis buffer, is shown in the last lane, displaying the successful separation and relative position of pRIIβ and RIIβ in a 2-D gel. Phosphorylated pRIIβ can be recognized by both of the phospho-PKA substrate antibody (Anti-RRXp(S/T)) and the RIIβ antibody (anti-RIIβ), while unphosphorylated RIIβ is only recognized by the RIIβ antibody (anti-RIIβ). The first and second two lanes show that RIIβ exists as phosphorylated pRIIβ in both untreated HeLa and HeLa cells. In 3T3 cells, one protein near the site of pRIIβ is recognized by phospho-PKA substrate antibody (Anti-RRXp(S/T); however, it is not recognized by the RIIβ antibody. Right panel: In HeLa and 3T3 cells treated with 50 μM forskolin and 50 μM IBMX for 15 min prior to adding the lysis buffer, the PKA RIIβ becomes completely dephosphorylated.
Fig 8.
Role of single turnover phosphorylation in the regulation of PKA signaling.
Type II holoenzymes of PKA are under an additional feedback regulation that is mediated by the phosphorylation of the R-subunits. Holoenzyme formation creates a single turnover condition in which the catalytic subunit phosphorylates the R-subunit. Increased cAMP levels dissociate the holoenzyme as the R-subunit binds cAMP or at least unleashes the inhibitor site from the active site cleft of the C-subunit. Reassociation of the inactive holoenzyme requires the concerted action of phosphodiesterases that hydrolyse cAMP and phosphatases that remove the phosphoryl group from the inhibitor site of RIIβ. A phosphorylated yet cAMP-free RIIβ is unable to bind efficiently to the catalytic subunit. This therefore creates a self-regulating cycle that contributes to the regulation of PKA signaling. Whether the R- and C-subunits fully dissociate under these single turnover conditions is not clear, but full dissociation is not necessary. The only necessary requirement for activation is that the inhibitor site of the RIIβ subunit be released, allowing access of the nearby tail of a channel or receptor.