Regulation of PKA Binding to AKAPs in the Heart
Alterations in Human Heart Failure
Background—cAMP-dependent protein kinase (PKA) regulates a broad range of cellular responses in the cardiac myocyte. Downstream regulation of the PKA pathway is mediated by a class of scaffolding proteins called A-kinase anchoring proteins (AKAPs), which sequester PKA to specific subcellular locations through binding to its regulatory subunit (R). However, the effect of RII autophosphorylation on AKAP binding and the degree of RII autophosphorylation in failing and nonfailing human hearts remains unknown.
Methods and Results—We investigated AKAP-RII binding by overlay analysis and surface plasmon resonance spectroscopy and measured RII autophosphorylation in human hearts by backphosphorylation. Binding of Ht31 peptide (representing the RII-binding region of AKAPs) to cardiac RII was increased ≈145% (P<0.01) for autophosphorylated RII relative to unphosphorylated control. By surface plasmon resonance, RII autophosphorylation significantly increased binding affinity to Ht31 by ≈200% (P<0.01). Baseline PKA-dependent phosphorylation of RII was significantly decreased ≈30% (P<0.05) in human hearts with dilated cardiomyopathy compared with nonfailing controls.
Conclusions—These results suggest that AKAP binding of PKA in the heart is regulated by RII autophosphorylation. Therefore AKAP targeting of PKA may be reduced in patients with end-stage heart failure. This mechanism may be responsible for the decreased cAMP-dependent phosphorylation of proteins in dilated cardiomyopathy that we and others have previously observed.
cAMP-dependent protein kinase (PKA) is a tetrameric holoenzyme consisting of 2 catalytic (C) and 2 regulatory (R) subunits.1 By substrate phosphorylation, PKA regulates a broad range of responses in the cardiomyocyte. Downstream regulation of the PKA pathway is mediated by A-kinase anchoring proteins (AKAPs), which sequester PKA to specific subcellular sites through binding to R.2 Although β-adrenoceptor density is reduced in failing human myocardium,3 evidence for defects distal to the receptor/adenylyl cyclase complex is rapidly accumulating. For example, even after stimulation by positive inotropic agents that act beyond the receptor, such as dibutyryl-cAMP and cAMP-phosphodiesterases inhibitors, a decreased positive inotropic response in failing hearts is still observed.4 Furthermore, protein phosphorylation mediated by β-adrenergic activation is decreased in the failing human heart5 6 as well as in animal models of cardiac hypertrophy.7 We previously found decreased baseline troponin-I (TnI) phosphorylation and decreased protein levels of the regulatory subunits of types I and II PKA (RI and RII) in myocardial samples from patients with end-stage dilated cardiomyopathy (DCM).5 Therefore the PKA signaling pathway appears to be downregulated at several levels beyond the β-receptor in human heart failure.
Type II regulatory subunits have an autophosphorylation site, serine 95, in the primary sequence, which is the only residue phosphorylated by C,8 whereas type I regulatory subunits have a pseudophosphorylation site.1 In vitro studies on cardiac RII indicate that autophosphorylation of RII by C at serine 95 controls activation of the kinase by decreasing affinity for C ≈10-fold.9 10 However, the degree of RII autophosphorylation in normal and failing human hearts has not been examined. Our aims were 2-fold: (1) to determine whether RII autophosphorylation affects RII binding to AKAPs and (2) to measure PKA-dependent RII phosphorylation in human hearts with DCM and in nonfailing (NF) human hearts.
We previously showed colocalization of RII with AKAP100 in ventricular myocytes.11 Interestingly, there were pockets of RII that did not colocalize with AKAP and regions of unoccupied AKAPs. In addition, previous studies showed that the KD for AKAP-RII interaction (≈10 nmol/L) is within physiological concentrations of most AKAPs.12 Together this suggests that a mechanism exists for regulating this interaction. However, the role of RII autophosphorylation in regulating AKAP binding of PKA has not yet been examined. This is the first study to determine the role of RII phosphorylation in regulating AKAP anchoring in the heart. We used the Ht31 peptide, which represents the RII binding domain of AKAPs,12 to measure binding affinities of dephosphorylated and autophosphorylated cardiac RII (RII-P) to Ht31 by both peptide overlay analysis and surface plasmon resonance (SPR) spectroscopy.
Characterization of RII Recombinant Protein
We generated purified recombinant RIIα subunits by cAMP-agarose affinity chromatography. The eluate from the cAMP-agarose column was separated by 10% SDS-PAGE followed by Coomassie blue staining. As shown in Figure 1⇓, bands of ≈57 and ≈55 kDa were identified on the Coomassie blue–stained gel for and RII-P and RII (lanes 4 and 5), respectively. The identity of RII was confirmed by 32P phosphorylation and immunoblot analysis as previously described11 (data not shown). Consistent with previous reports,13 we find that autophosphorylation of cardiac RII results in a decreased mobility in SDS gels. Because the RII autophosphorylation site is at a proteolytically sensitive hinge region,14 we first determined whether PKA phosphorylation affected proteolysis of RII. Phosphorylation by PKA had no effect on RII proteolysis in either bacterial lysate (Figure 1⇓, lanes 2 and 3) or purified protein (lanes 4 and 5), as determined by scanning densitometry of the RII band. Furthermore, the level of RII detected immunologically by anti–RII antibody did not change with RII phosphorylation (not shown).
Identification of AKAP100 by RII Overlay and Immunoblot Analysis
By using radioactively labeled RII in an overlay assay, we confirmed the presence of AKAP100 in the human heart, as previously reported in rat cardiac myocytes and myocyte H9C2 cell line.11 15 Consistent with previous findings from our laboratory, 32P-labeled RII identified an ≈82-kDa band in fresh human heart homogenates and a doublet of ≈85 to 82 kDa in rat cardiac myocytes (Figure 2A⇓). The binding was inhibited by incubation with 1 μmol/L Ht31 peptide (not shown). The anti-AKAP100 antibody recognized the same band at ≈82 kDa, with a proteolytic fragment at ≈56 kDa (Figure 2B⇓), although recent cloning studies suggest that the full-length protein may be larger.16 Thus AKAP100 is likely to be a major PKA II anchor in the human heart, consistent with our previous immunofluorescence studies in rat cardiac myocytes.11
Phosphorylation State of RII In Vivo
Because little is known about the phosphorylation state of RII in the intact heart, we performed PKA-dependent backphosphorylation in fresh human hearts and rat cardiomyocytes. Freshly isolated rat left ventricular cardiomyocytes were treated with isoproterenol (1 μmol/L) or were untreated (vehicle control). The phosphorylation of known PKA substrates such as TnI, phospholamban (PLB), and C-protein increased significantly with stimulation (Table 1⇓). However, PKA-dependent phosphorylation of RII did not increase in response to β-adrenergic stimulation (Figure 3⇓). The highly phosphorylated band at ≈40 kDa in both lanes is most likely autophosphorylated PKAcat, exogenously added as part of the assay (see also Figure 6A⇓). The lack of difference between basal and stimulated levels of RII phosphorylation (Table 1⇓) suggests that RII is probably constitutively phosphorylated in the heart. This is consistent with data from a freshly obtained, unfrozen, explanted human heart (cold ischemic time in cardioplegia 0.5 hours), which indicates a high level of in vivo phosphorylation (70 ±2%; mean±SD, n=3 determinations).
To investigate the importance of RII autophosphorylation in regulating AKAP-PKA interactions in the heart, we first used an Ht31 peptide overlay assay to measure binding affinities between Ht31 and the regulatory subunit. RII or RII-P was transferred to nitrocellulose, and the amount of bound Ht31B was determined by densitometric scanning. Ht31B recognized RII-P and RII at ≈57 and ≈55 kDa, respectively (Figure 4A⇓). At 70 ng recombinant protein, Ht31B binding was significantly increased 145 ±7% (P<0.01) for RII-P relative to control RII (Figure 4B⇓). At saturating amounts of RII protein (280 ng), no difference in Ht31B binding was detected. With high protein loading, however, a proteolytic product of RII (≈37 kDa) was visible by Ht31 overlay. Binding of Ht31B to this fragment was significantly greater for RII-P, relative to RII (not shown). No binding of Ht31B was observed in left ventricular homogenates of mouse (45 μg total protein/lane), rat (45 μg total protein), or human (30 μg total protein) by overlay analysis, suggesting that the protein was not sufficiently concentrated in the samples to be detected by this technique.
In a separate series of experiments, Ht31 overlays were performed under nondenaturing conditions on samples of bacterial lysate containing RII. This alternative approach was used to avoid possible alterations resulting from RII purification and denaturation in SDS gels. Aliquots of lysate (0.25 to 1.0 mg/mL) were spotted directly to nitrocellulose in triplicate, then probed with Ht31B. Under these conditions, binding of Ht31B was also significantly increased for RII-P relative to control RII (Figure 4C⇑).
Determination of Kinetic Parameters by SPR
To examine the effect of RII autophosphorylation on AKAP binding, ka, kd, and KD were determined for the interaction of RII and RII-P with Ht31. Injection of soluble Ht31 but not control Ht31P competitively removed RII bound to the active surface, indicating specificity of the measured RII-Ht31 interaction (Figure 5A⇓). Similarly, preincubation of RII with a 10-fold molar excess of soluble Ht31 competitively blocked the binding of RII to the active Ht31B surface, whereas preincubation with Ht31P did not (not shown). The level of RII-P bound after 1-minute contact time was significantly increased by 192 ±2% (mean±SD, n=5) relative to RII (Figure 5B⇓). The KD for RII-P–Ht31 interaction was significantly decreased 2-fold (n=3) compared with RII–Ht31 (Table 2⇓) as a result of increased association rates. These data are in agreement with previously obtained KD values of 11.2 nmol/L by SPR.17
RII Autophosphorylation in DCM
Because the results of experiments described above suggest that RII autophosphorylation may enhance the binding of PKA to AKAPs, we investigated the role of this mechanism in failing and NF human hearts. We measured baseline values of PKA-dependent RII phosphorylation in left ventricular homogenates of 4 human hearts with DCM and 4 NF control hearts (Figure 6A⇓). The highly phosphorylated band at ≈40 kDa in both lanes is most likely autophosphorylated PKAcat, exogenously added as part of the assay procedure. PKA-dependent RII phosphorylation was significantly decreased ≈32% (P<0.05) in DCM compared with NF (Figure 6B⇓). Because we previously found decreased RII protein in DCM,5 phosphorylation values were normalized to RII protein levels, as determined by immunoblot analysis of the same sample preparations used in the backphosphorylation assay.
The main findings are that RII autophosphorylation by C enhances its binding to the Ht31 peptide and that baseline PKA-dependent RII phosphorylation is significantly decreased in DCM compared with NF. These results demonstrate that RII autophosphorylation may increase local concentrations of C by increasing holoenzyme affinity for its anchoring protein. Therefore these results suggest that PKA anchoring to specific subcellular substrates may be decreased in failing cardiac myocytes.
We found that autophosphorylation of RII by PKA in vitro significantly increases its binding to Ht31. The Ht31 peptide is widely used to mimic AKAP-RII interactions18 or to inhibit them in vivo.15 19 Ht31 forms an amphipathic helix, which represents the essential RII binding domain of the full-length protein.12 On the basis of this and the fact that the proline substituted mutant peptide (Ht31P) has no effect in either Ht31 overlay, RII overlay, or SPR experiments, our results imply that RII autophosphorylation enhances binding to full-length AKAPs.
In freshly isolated rat left ventricular myocytes, PKA-dependent RII phosphorylation did not increase above baseline after β-stimulation. In addition, baseline PKA-dependent RII phosphorylation was ≈70% (sites phosphorylated in vivo) in the failing human heart. RII is thus likely to be constitutively phosphorylated in the heart. These results are in agreement with previous studies that suggest that bovine cardiac RII exists in vivo primarily in the autophosphorylated form.20 Although β-adrenergic stimulation of isolated myocytes had no significant effect on the phosphorylation state of RII, phosphorylation of other known PKA substrates such as PLB, TnI, and C-protein did significantly increase above baseline, indicating that the β-adrenergic pathway was intact in these myocytes and the isoproterenol treatment was effective.
Our results indicate that RII is constitutively phosphorylated by PKA in the heart and that this basal, steady-state phosphorylation may be decreased under pathological conditions. Although previous in vitro studies indicate that RII autophosphorylation occurs as an intramolecular event,9 10 recent evidence suggests that cAMP can induce catalytic activity of PKA without subunit dissociation.21 Furthermore, the presence of PKI (with substrate) in the cell may allow for catalytic subunit dissociation independent of cAMP binding, with subsequent phosphorylation of nearby RIIs.22 Thus decreased targeted holoenzyme in DCM (initially caused by decreased RII protein5 ) may result in decreased RII phosphorylation and thus a further decrease in the subcellular localization of the kinase. We observed decreased PKA phosphorylation of other proteins in DCM such as C-protein (reduced ≈30%) and TnI (reduced ≈30%) but not PLB, which is consistent with previous results.5 7 Arguing against overall downregulation of the β-adrenergic pathway in DCM, these results favor the idea of specific localized defects in DCM, mediated perhaps by decreased local availability of C as a result of altered AKAP anchoring in those hearts.
The regulatory subunit of PKA is responsible for AKAP binding and targeting of the holoenzyme. AKAP-mediated anchoring of PKA is functionally important to the extent that disruption of AKAP-RII interaction by Ht31 can result in loss of PKA catalytic activity and loss of PKA modulation of L-type channel activity.19 Therefore, in addition to decreased RII protein in DCM, decreased RII autophosphorylation may limit PKA targeting to functionally important regions in the failing human heart. Thus the reduced PKA-dependent phosphorylation of substrates that we observed in DCM may be due, at least in part, to altered AKAP-RII interaction in those hearts.
The use of tissue from failing and NF human hearts presents several potential limitations. All the failing hearts were receiving inotropic support (digoxin or dobutamine), which could potentially increase β-adrenergic tone. However, long-term elevated circulating catecholamines are believed to trigger downregulation of β-receptors.3 Furthermore, although brain death may be associated with massive catecholamine release in the donor hearts, circulating catecholamines are washed out by flushing and transport of the explanted hearts in ice-cold cardioplegia. Because transport time to the laboratory for donor hearts is longer than for failing hearts, any residual phosphatase activity would be expected to be greater in the donor hearts, resulting in decreased substrate phosphorylation, yet this is the opposite of what we observed.
In summary, we have shown by Ht31 overlay analysis and by SPR that RII autophosphorylation enhances RII binding to Ht31 peptide, the RII-binding domain of AKAPs. In addition, we found decreased RII autophosphorylation in DCM compared with NF. Overall, decreased autophosphorylation of RII may lead to a reduced amount of targeted RII in the failing human heart, with subsequent decreased local concentrations of C and decreased ability for phosphorylation of PKA substrates.
Human Myocardial Tissue
Failing human cardiac muscle was obtained from explanted hearts of heart transplantation recipients at the Cleveland Clinic Foundation. NF human heart tissue was obtained from unmatched organ donors through Life Banc of Northeast Ohio. Clinical characteristics of patients, transportation, and storage of the tissue are as previously described5 for NF (n=4) and DCM (n=4) hearts.
Isolation of Rat Ventricular Myocytes
All procedures involving animals followed the Guide for the Care and Use of Laboratory Animals (Institute of Laboratory Animal Resources, 1996). The Animal Care Facility of the Cleveland Clinic Foundation is accredited by the American Association for the Accreditation of Laboratory Care. Cardiac myocytes were isolated from the left ventricle of adult Sprague-Dawley rats purchased from Taconic Farms (Germantown, NY) at 20 to 30 weeks of age by collagenase digestion with the use of a modified Langendorff perfusion, according to methods previously described.23 This procedure was modified to include plating of the myocytes onto tissue-culture plates coated with laminin (4 μg/cm2) at a density of 1×105 cells/mL. After 2 hours of incubation in 95% O2/5% CO2 at 37°C to allow for attachment, the plates were washed to remove dead cells, and healthy cells were lysed with 20 mmol/L HEPES, pH 7.4, containing 5 mmol/L EDTA, 1 mmol/L DTT, 10% glycerol, 1 μmol/L leupeptin, 1 μmol/L pepstatin, PMSF, 0.1 μmol/L NaVO43−, 70 mmol/L NaF, and 2 nmol/L calyculin A. The viability of the isolated ventricular myocytes prepared in this manner was typically >90% as determined by rod-shaped morphology and lack of granulation or blebs.
Purification of Recombinant RII and Immunoblot Analysis
Purified recombinant RIIα was prepared as described.11 Briefly, the RIIα recombinant plasmid (RII-pET11 day), a gift of Dr John D. Scott (Vollum Institute, Portland, OR), was used to transform Escherichia coli bacteria strain BL21(DE3). RII recombinant protein was expressed under 2 mmol/L IPTG induction followed by cAMP-agarose affinity purification.24 Purified recombinant RII was separated on SDS-PAGE gels, analyzed by Coomassie blue staining, and immunoblotted with RII-specific antibodies. Immunoblot analysis for RII and AKAP in heart homogenates was as previously described5 with the use of polyclonal anti-RII antibody (Santa Cruz Biotechnology) or anti-AKAP100 antibody (Upstate Biotechnology).
Human and rat heart homogenates were backphosphorylated by PKA with modifications of Zakhary et al.5 In parallel reactions, maximal PKA-dependent phosphorylation was determined by pretreating with alkaline phosphatase (AP) to remove bound phosphate, then treating with PKAcat and [γ-32P]ATP to rephosphorylate only PKA sites. The concentration of sample protein and ATP was equal in both reactions. Dephosphorylation was initiated by addition of AP (1:100 enzyme/protein) and allowed to proceed for 20 minutes at 35°C. Here NaF was omitted from the first reaction, then added (70 mmol/L) after dephosphorylation to inactivate the phosphatase. Maximal phosphate incorporation was achieved after 20 minutes of AP pretreatment.
Ht31 Overlay Assay
All peptides were synthesized by the Molecular Biotechnology Core Facility at the Cleveland Clinic Foundation. The 23–amino acid inhibitory peptide Ht31, derived from the full-length human thyroid AKAP,12 is widely used to mimic or inhibit AKAP-RII interactions. It forms an amphipathic helix that binds RII with nanomolar affinity and competes with full-length AKAPs for RII binding. Proline substitution at the 2 isoleucine residues of Ht31 (Ht31P) disrupts the amphipathic helix, thereby rendering the peptide unable to bind RII. Peptides were biotinylated at the N-terminus (Ht31B, Ht31PB) to allow detection of RII-binding events.22 RII or RII-P was subjected to 10% SDS-PAGE, then transferred to nitrocellulose. For spot blots, aliquots (50 μL) were spotted directly onto nitrocellulose in triplicate. Blots were blocked for 1 hour with 3% fish gelatin (wt/vol), then probed with excess Ht31B (1 μmol/L). Fluorescence corresponding to RII-bound Ht31B was quantified by StormImager analysis with the use of a streptavidin-alkaline phosphatase conjugate. Fluorescent signal corresponding to bound Ht31B increased linearly with increasing RII protein loaded. Nonbiotinylated control peptides gave no signal.
RII Overlay Assay
Purified recombinant RII was labeled with [γ-32P]ATP by in vitro phosphorylation (30°C, 1 hour) with PKA catalytic subunits obtained from Sigma. An RII overlay assay was then performed11 to determine the interaction between the RII and endogenous AKAPs in the human heart. The RII binding proteins were detected with the use of a StormImager. Ht31 peptide, synthesized according to the published sequence,12 was used to compete with RII-AKAP binding in the RII overlay assay.
SPR was measured by a BIAcore 1000 instrument with a streptavidin (SA) sensor chip (Pharmacia Biosensor AB). Samples of RII or RII-P (analyte) were diluted in HBS buffer (10 mmol/L HEPES, pH 7.4, 150 mmol/L NaCl, 3 mmol/L EDTA, and 0.005% Surfactant P20). Samples were injected at 25°C at a flow rate of 10 μL/min over the active SA surface, on which the Ht31B (ligand) had been immobilized to ≈20 resonance units or over a control surface (control peptide, Ht31PB immobilized). Remaining SA sites were blocked with free biotin. Surfaces were not limited by mass transport effects, as determined by flow-rate variation experiments: increasing flow rate 3-fold (from 10 to 30 or from 30 to 90 μL/min) increased initial association rates by ≤5%. Binding surfaces were regenerated to remove bound analyte by injection of 10 mmol/L NaOH (1-minute contact). Regeneration conditions did not result in loss of SA surface binding capacity. At saturating analyte concentration, the maximum response (Rmax) was not significantly different between RII and RII-P. Kinetic constants were calculated by global fitting of the data to a 1:1 Langmuir binding model, after subtraction of control surface, with the use of BIA evaluation software, version 3.0, according to the pseudo–first-order rate equation where ka is the association rate, kd is the dissociation rate, C is the concentration of the injected analyte, and R is the response. Plots of dR/dt versus Rt have a slope of ks. When ks is plotted against C, the resulting slope is equal to ka. The dissociation rate, kd, was calculated by integrating the rate equation when C=0, yielding ln(Rt1/Rtn)=kd (tn−t1). The equilibrium binding constants were calculated from the equation Kd=kd/ka.
Paired and unpaired t tests were used to evaluate the effect of treatments on individual samples and to compare differences between DCM and NF, respectively. Differences at P<0.05 were considered significant. Results are expressed as mean±SEM unless otherwise indicated.
This work was supported by National Heart, Lung, and Blood Institute grants HL-56256 (M. Bond) and HL-49929 (C. Moravec), National Institutes of Health cardiovascular training grant HL-07653 (D. Zakhary), and by an American Heart Association Established Investigator Award and grant-in-aid to C. Moravec. We thank Dr Norman Ratliff, Department of Pathology, Drs Robert Stewart and Patrick McCarthy, Department of Cardiovascular Surgery, CCF, and Life Banc of Northeast Ohio for providing human tissue. We also thank Drs Tom Egelhoff, J.P. Jin, Mary Russell, and Jiacheng Yang for helpful discussions and Russell Desnoyer for expert technical assistance.
The Methods section of this article can be found at http://www.circulationaha.org
- Received June 18, 1999.
- Revision received September 16, 1999.
- Accepted October 11, 1999.
- Copyright © 2000 by American Heart Association
Dell’Acqua ML, Scott JD. Protein kinase A anchoring. J Biol Chem. 1997;272:12881–12884.
Feldman MD, Copelas L, Gwathmey JK, Phillips P, Warren SE, Schoen FJ, Grossman W, Morgan JP. Deficient production of cyclic AMP: pharmacologic evidence of an important cause of contractile dysfunction in patients with end-stage heart failure. Circulation. 1987;75:331–339.
Zakhary DR, Moravec CS, Stewart RW, Bond M. Protein kinase A–dependent troponin-I phosphorylation and PKA regulatory subunits are decreased in human dilated cardiomyopathy. Circulation. 1999;99:505–510.
Bodor G, Oakeley A, Allen P, Crimmins D, Ladenson J, Anderson P. Troponin I phosphorylation in the normal and failing adult human heart. Circulation. 1997;96:1495–1500.
McConnell BK, Moracec CS, Bond M. Troponin I phosphorylation and myofilament calcium sensitivity during decompensated cardiac hypertrophy. Am J Physiol. 1998;274:H385–H396.
Takio K, Smith S, Krebs E, Walsh K, Titani K. Primary structure of the regulatory subunit of type II cAMP-dependent protein kinase from bovine cardiac muscle. Proc Natl Acad Sci U S A. 1982;79:2544–2548.
Rangel AR, Rosen O. Effect of cAMP and ATP on the reassociation of phosphorylated and nonphosphorylated subunits of the cAMP-dependent protein kinase from bovine cardiac muscle. J Biol Chem. 1977;252:7140–7145.
Granot J, Mildvan A, Hiyama K, Kondo H, Kaiser E. Magnetic resonance studies of the effect of the regulatory subunit on metal and substrate binding to the catalytic subunit of bovine heart protein kinase. J Biol Chem. 1980;255:4569–4573.
Yang J, Drazba J, Ferguson D, Bond M. A-kinase anchoring protein 100 (AKAP100) is localized in multiple subcellular compartments in the adult rat heart. J Cell Biol. 1998;142:511–522.
Carr D, Hausken Z, Fraser I, Stofko HR, Scott J. Association of the type II cAMP-dependent protein kinase with a human thyroid RII-anchoring protein: cloning and characterization of the RII-binding domain. J Biol Chem. 1992;267:13376–13382.
Weldon SL, Mumby MC, Taylor SS. The regulatory subunit of neural cAMP-dependent protein kinase II represents a unique gene product. J Biol Chem. 1985;260:6440–6448.
Durgerian S, Taylor SS. The consequences of introducing an autophosphorylation site into the type I regulatory subunit of cAMP-dependent protein kinase. J Biol Chem. 1989;264:9807–9813.
McCartney S, Little B, Langeberg L, Scott J. Cloning and characterization of A-kinase anchor protein 100 (AKAP100): a protein that targets A-kinase to the sarcoplasmic reticulum. J Biol Chem. 1995;270:9327–9333.
Kapiloff MS, Schillace RV, Westphal AM, Scott JD. mAKAP: an A-kinase anchoring protein targeted to the nuclear membrane of differentiated myocytes. J Cell Sci.. 1999;112:2725–2736.
Burton K, Johnson B, Hausken Z, Westenbroek R, Idzerda R, Scheuer T, Scott J, Catterall W, McKnight G. Type II regulatory subunits are not required for the anchoring-dependent modulation of Ca2+ channel activity by cAMP-dependent protein kinase. Proc Natl Acad Sci U S A. 1997;94:11067–11072.
Hausken Z, Dell AM, Coghlan V, Scott J. Mutational analysis of the A-kinase anchoring protein (AKAP)-binding site on RII: classification of side chain determinants for anchoring and isoform selective association with AKAPs. J Biol Chem. 1996;271:29016–29022.
Rangel AR, Rosen O. Mechanism of self-phosphorylation of adenosine 3′:5′-monophosphate-dependent protein kinase from bovine cardiac muscle. J Biol Chem. 1976;251:7526–7529.
Johnson D, Leathers V, Martinez A, Walsh D, Fletcher W. Fluorescence resonance energy transfer within a heterochromatic cAMP-dependent protein kinase holoenzyme under equilibrium conditions: new insights into the conformational changes that result in cAMP-dependent activation. Biochemistry. 1993;32:6402–6410.
McConnell BK, Moravec CS, Morano I, Bond M. Troponin I phosphorylation in spontaneously hypertensive rat heart: effect of beta-adrenergic stimulation. Am J Physiol. 1997;273:H1440–H1451.
Dransfield DT, Bradford AJ, Smith J, Martin M, Roy C, Mangeat PH, Goldenring JR. Ezrin is a cyclic AMP-dependent protein kinase anchoring protein. EMBO J. 1997;16:35–43.