AKAP18: Bidirectional Regulator of Protein Phosphorylation" (2013)

University of Connecticut OpenCommons@UConn

Doctoral Dissertations University of Connecticut Graduate School

7-1-2013 AKAP18: idirB ectional Regulator Of Protein Phosphorylation Arpita Singh University of Connecticut Health Center, [email protected]

Follow this and additional works at: https://opencommons.uconn.edu/dissertations

Recommended Citation Singh, Arpita, "AKAP18: Bidirectional Regulator Of Protein Phosphorylation" (2013). Doctoral Dissertations. 193. https://opencommons.uconn.edu/dissertations/193

AKAP18: Bidirectional Regulator of Protein Phosphorylation

Arpita Singh, Ph.D.

University of Connecticut, 2013

(Dissertation Completed: May 2013)

The thesis work presented here focuses on AKAP18γ mediated multivalent complex formation and its oligomerization. AKAP18γ belongs to a family of structurally diverse but functionally analogous scaffolding proteins called A-Kinase Anchoring

Proteins (AKAPs). These scaffolds specifically target Protein Kinase A (PKA) holoenzyme to a distinct cellular compartment to regulate PKA activity and substrate specificity. Previously, the long isoforms of an AKAP called AKAP18 (AKAP18γ/δ) were established as important cardiac calcium regulating proteins due to their ability to mediate phosphorylation of a protein called Phospholamban (PLN). My dissertation work shows that AKAP18γ directly interacts with Inhibitior-1 (I-1) and enhances its phosphorylation under stimulatory conditions. The AKAP18γ-PKA-I-1 complex together regulates Protein Phosphatase 1 (PP1), also found to be associated with this complex.

PP1 is a negative regulator of cardiac function due to its ability to dephosphorylate PLN.

Based on these results, we propose a model in which AKAP18γ acts as multivalent complex organizing scaffold that may be targeted to the sarcoplasmic reticulum in cardiac myocytes for PLN phosphorylation regulation. This complex may be found in other cell types to regulate PP1 function alone as I-1, PP1 and AKAP18γ have been found in other tissue.

Arpita Singh- University of Connecticut, 2013

Moreover, I also discovered that AKAP18γ is an oligomerizing scaffold and the interaction is isoform specific. A shorter isoform of this gene, AKAP18α does not oligomerize and there is no hetero oligomerization between these isoforms (AKA18α and

AKAP18γ). Live cell microscopy shows that AKAP18γ can form heterogenous oligomers ranging from monomers to tetramers in cells. Furthermore, in order to understand the physiological implication of this oligomerization, we developed a computational model. The model was based on activation and release of PKA catalytic subunit, which requires PKA regulatory subunit phosphorylation. Presence of multiple

PKA holoezymes will potentiate release of multiple PKA catalytic subunits due to increased PKA regulatory subunit phosphorylation. When this possibility was tested against a generic membrane compartmentalized substrate, the computational model revealed increased PKA mediated phosphorylation of the substrate. These results are suggestive of a mechanism in which AKAP18γ oligomerizes to mediate enhanced substrate phosphorylation by PKA.

Based on the evidence presented in this thesis and work done by others, it appears that AKAP18γ not only regulates PLN phosphorylation but also its de-phosphorylation by regulating PP1 activity. Importantly, AKAP18γ can oligomerize which may be the potential mechanism by which AKAP18γ concentrates enzymes required for PLN regulation. The title of the thesis “AKAP18: bidirectional regulator of protein phosphorylation” reflects these properties of AKAP18γ.

AKAP18: Bidirectional Regulator of Protein Phosphorylation

Arpita Singh

B.Sc, Banaras Hindu University, 2005

M.Sc., Pondicherry University, 2007

A Dissertation

Submitted in Partial Fulfillment of the

Requirements for the Degree of

Doctor of Philosophy

at the

University of Connecticut

2013

Arpita Singh

2013

APPROVAL PAGE

Doctor of Philosophy Dissertation

AKAP18: Bidirectional Regulator of Protein Phosphorylation

Presented by

Arpita Singh, B.Sc., M.Sc.

Major Advisor Dr. Kimberly Dodge-Kafka

Associate Advisor Dr. Asis Das

Associate Advisor Dr. Bing Hao

Associate Advisor Dr. Betty A. Eipper

Associate Advisor Dr. Lixia Yue

University of Connecticut

2013

ACKNOWLEDGEMENTS

“I dedicate this thesis work to my grandparents, who were a source of motivation to me- their memories continue to inspire me”

iii

This thesis work is a culmination of five years of work, which I could not have finished without the support of these people whom I wish to acknowledge.

I owe my deepest gratitude to my supervisor, Dr. Kimberly Dodge-Kafka, for providing me the opportunity to work in her laboratory. She has been a great teacher, a guide for navigating through graduate school and extremely supportive through highs and lows. I thank her for her dedication towards ensuring my success.

I am extremely grateful to my PhD thesis committee members, Dr. Asis Das, Dr.

Betty Eipper, Dr. Bing Hao, and Dr. Lixia Yue, for their advice and guidance during various stages of my PhD. I am indebted to them for helping me develop qualities that are required to be a good scientist.

My current and previous lab mates, Dr. Max Vargas, Dr. Andrew Le, John

Redden, and Marc Riggati, also share the credit for my success. They have been extremely supportive both in and out of the lab. I am very thankful for the light moments we have shared and the tough ones we have endured together.

I would like to thank the departments of Molecular, Microbial and Structural

Biology and Calhoun Cardiology Center for extending their support and help. For the various seminars, journal clubs, retreats and holiday parties that I have attended and had a great experience at, I thank you. I also express my gratitude to Dr. Roger Thrall for his guidance and continued support. My special thanks to Dr. John Carson for helping me hone my presentation skills, to Dr. Ann Cowan for always being here to help with microscopy questions and to Dr. Chris Heinen for all his advice during the various department presentations and journal clubs.

I am indebted to UConn Health Center for inviting me to pursue my PhD studies at the institute. I also thank Dr. Barbara Kream and Dr. Lawrence A. Klobutcher who have been great help in understanding the nitty-gritty’s of graduate school. The graduate school office, the registrar’s office, the human resources office and the bursar’s office for helping me navigate through the formailities and requirements that needed to be fulfilled from time to time over the past five years.

My success at every stage of my life would be attributed to my family. My mother, who has been my pillar of strength, has made my success the mission of her life.

I cannot imagine coming this far without her. My father, who inculcated in me the love of biology and medicine, has been a role model for me. I thank my loving brother, who always had words of encouragement during the toughest times. I cannot find words to express my gratitude to my in-laws who have been very understanding and supportive through this journey. They truly have been my second set of parents. This thesis would have remained a dream had it not been for my husband who had motivated me to embark on this endeavor and has been constantly by my side from the start to the finish line. It is hard to imagine how things would have turned out without his constant encouragement and cheer.

A special thanks to all my friends, for sharing my joys and sorrows, for being there for me always. Finally, I would like to thank all the people who have touched my life, in ways big and small, and have encouraged and appreciated my efforts.

TABLE OF CONTENTS

APPROVAL PAGE ...... II ACKNOWLEDGEMENTS ...... III TABLE OF CONTENTS ...... VI 1. INTRODUCTION ...... 1 1.1. SCAFFOLDING PROTEINS IN CELL FUNCTION...... 1 1.2. AKAPS AND THEIR ROLE IN PKA SIGNAL INTEGRATION AND REGULATION...... 4 1.3. REGULATION OF PP1 FOR PKA SIGNAL PROPAGATION...... 13 1.4. REGULATION OF PP1 BY I-1 ...... 20 1.5. REGULATION OF PP1 BY AKAPS...... 23 1.6. IMPORTANCE OF PP1 AND I-1 IN DISEASE ...... 26 1.7. AKAP18 AS A MODEL AKAP...... 32 2. THE LARGE ISOFORMS OF AKAP18 MEDIATE THE PHOSPHORYLATION OF INHIBITOR-1 BY PKA AND THE INHIBITION OF PP1 ACTIVITY...... 38 2.1. ABSTRACT ...... 39 2.2. INTRODUCTION...... 39 2.3. MATERIALS AND METHODS ...... 41 2.3.1. Antibodies ...... 41 2.3.2. Expression Constructs ...... 42 2.4. EXPERIMENTAL PROCEDURES ...... 42 2.4.1. cAMP Agarose Purification...... 42 2.4.2. Pulldown Experiments...... 43 2.4.3. Immunoprecipitation ...... 43 2.4.4. Phosphatase Assay ...... 44 2.4.5. PKA activity assay ...... 45 2.4.6. Detection of I-1 phosphorylation...... 45 2.5. RESULTS...... 46 2.5.1. AKAP18 associates with Inhibitor-1 in vivo...... 46 2.5.2. I-1 Directly Binds a Long Isoform of AKAP18...... 50 2.5.3. Expression of AKAP18 enhances I-1 phosphorylation on Threonine 35 ..... 50 2.5.4. AKAP18 bind PP1 in vivo ...... 54 2.5.5. AKAP18 is the scaffold for a PKA, PP1 and I-1 signaling complex ...... 56 2.5.6. AKAP18 orchestrates the PKA-mediated inhibition of PP1 activity...... 58 2.6. DISCUSSION ...... 60

3. AKAP18γ OLIGOMERIZATION AUGMENTS PKA SIGNALING ...... 67 3.1. ABSTRACT ...... 68 3.2. INTRODUCTION...... 69 3.3. EXPERIMENTAL METHODS...... 71 3.3.1. Antibodies ...... 71 3.3.2. Construct design ...... 71 3.3.3. Recombinant Protein Purification...... 72 3.3.4. Cell transfection ...... 73 3.3.5. Pulldown Assays ...... 73 3.3.6. SPR Analysis...... 74 3.3.7. Photon Counting Histogram Analysis ...... 75 3.3.8. Confocal Imaging Number and Brightness Analysis...... 75 3.3.9. Computational Modeling...... 76 3.4. RESULTS...... 78 3.4.1. Direct cellular association of AKAP18γ with itself...... 78 3.4.2. Isoform specificity and binding affinity of AKAP18γ oligomers...... 80 3.4.3. Mapping the sites in AKAP18γ responsible for oligomerization...... 82 3.4.4. Mulitple oligomeric states of AKAP18γ can be detected in cells...... 84 3.4.5. Oligomerization of AKAP18γ functions to increase target phosphorylation. 86 3.5. DISCUSSION ...... 90 4. CONCLUSIONS AND FUTURE DIRECTIONS...... 96 5. IMPACT OF THESIS WORK ...... 107 6. REFERENCES...... 109

vii

FIGURE LIST

FIGURE 1.1: REPRESENTATIVE IMAGE SHOWING MECHANISMS OF DISTINCT SIGNALING COMPARTMENT FORMATION IN CELLS. 2 FIGURE 1.2: DIAGRAM SHOWING ACTIVATION OF CAMP INSIDE A CARDIAC MYOCYTE. 5 FIGURE 1.3: ACTIVATION AND DEACTIVATION OF PKA CATALYSIS. 7 FIGURE 1.4: REPRESENTATIVE IMAGE OF AN AKAP SHOWING ITS PROPERTIES AND FUNCTIONS. 8 FIGURE 1.5: AKAP COMPLEXES FOUND IN THE HEART. 11 FIGURE 1.6: DIAGRAM DEPICTING BINDING OF PP1 REGULATORY PROTEINS TO DIFFERENT DOMAINS ON PP1 SURFACE. 15 FIGURE 1.7: POSITIVE AND NEGATIVE REGULATION OF PP1 ACTIVITY BY SUBSTRATE-SPECIFYING GLYCOGEN TARGETING SUBUNITS. 19 FIGURE 1.8: REGULATION OF PP1 BY PKA PHOSPHORYLATED I-1. 22 FIGURE 1.9: PP1 REGULATION BY AKAP220. 25 FIGURE 1.10: REGULATION OF PHOSPHOLAMBAN (PLN) BY PP1. 27 FIGURE 1.11: REGULATION OF PP1 BY I-1 FOR NORMAL CARDIAC FUNCTION. 29 FIGURE 1.12: MISREGULATION OF PP1 DUE TO ALTERATION OF I-1 PHOSPHORYLATION. 30 FIGURE 1.13: AKAP18 ISOFORMS. 33 FIGURE 1.14: DIAGRAM SHOWING THE PREMISE OF HYPOTHESIS. 37 FIGURE 2.1: INHIBITOR-1 ASSOCIATES WITH PKA IN HEART CELLS. 47 FIGURE 2.2: AKAP18 ASSOCIATES WITH I-1 A. ISOLATED RAT HEART EXTRACT WAS INCUBATED WITH EITHER GST-I-1 OR CONTROL GST- PROTEIN. 49 FIGURE 2.3: DIRECT BINDING OF AKAP18G AND INHIBITOR-1. 51 FIGURE 2.4: AKAP18 ORCHESTRATES I-1 PHOSPHORYLATION BY LINKING PKA WITH THE INHIBITOR. 53 FIGURE 2.5: AKAP18 IS A PP1 ANCHORING PROTEIN 55

viii

FIGURE 2.6: AKAP18 SEQUESTERS PKA, PP1 AND I-1 TO A SPECIFIC SIGNALING COMPLEX. 57 FIGURE 2.7: PKA BINDING TO AKAP18 IS NECESSARY FOR PKA-INDUCED INHIBITION OF PP1. 59 SUPPLEMENTARY FIGURE 2.1: 63 SUPPLEMENTARY FIGURE 2.2: 64 SUPPLEMENTARY FIGURE 2.3: 65 SUPPLEMENTARY FIGURE 2.4: 66 FIGURE 3.1: AKAP18Γ INTERACTS WITH ITSELF IN CELLS AND DIRECTLY. 79 FIGURE 3.2: AKAP18Γ INTERACTS WITH ITSELF WITH A HIGH AFFINITY IN AN ISOFORM SPECIFIC MANNER. 81 FIGURE 3.3: MAPPING FOR DELINEATING SPECIFIC AKAP18Γ INTERACTION DOMAIN(S). 83 FIGURE 3.4: AKAP18Γ-EGFP TRANSFECTED VERO CELLS ANALYZED VIA PCH. 85 FIGURE 3.5: BRIGHTNESS ANALYSIS OF AKAP18Γ-EGFP IN VERO CELLS. 87 FIGURE 3.6: COMPARTMENTAL-DETERMINISTIC SIMULATION OF THE EFFECT OF AKAP OLIGOMERIZATION ON TARGET PHOSPHORYLATION 89 SUPPLEMENTARY FIGURE 3.1: BINDING OF R2 AND AKAP18 ISOFORMS 94 SUPPLEMENTARY TABLE 3.1: 95 FIGURE 4.1: GM 1-240 GST INTERACTS WITH PLN WITH HIGH AFFINITY. 98 FIGURE 4.2: I-1 GST INTERACTS WITH AKAP18Γ WITH HIGH AFFINITY. 100 FIGURE 4.3: I-1 PEPTIDES INTERACT WITH AKAP18Γ WITH HIGH AFFINITY. 101 FIGURE 4.4: PROPOSED AKAP18γ MEDIATED REGULATION OF I-1 . 104 FIGURE 5.1: AKAP18 COMPLEX BASED ON THESIS AND WORK BY OTHER LABS. 108

ix 1

1. INTRODUCTION

1.1. SCAFFOLDING PROTEINS IN CELL FUNCTION

The last two decades have been crucial in advancing our understanding of cellular signaling. The vaguely understood mechanisms by which cells relay specific signals inside the cytoplasm are now becoming clearer. One of the key drivers of this research has been the discovery of scaffolding proteins (1-3). These intracellular proteins form connections to transmit information from one signal point to the next by organizing signaling complexes that consist of specific proteins. This is important to ensure signal relay to precise proteins at the right time. Diffusion of molecules in the cytoplasm opposes this phenomenon due to randomness inherent to diffusing systems. To get around this problem, cells have evolved mechanisms to immobilize signaling proteins.

Signaling proteins can be co-localized by a few mechanisms. As shown in figure 1, the three major ways of doing this involves: co-localization of proteins to a specific cellular location, organelle formation and assembly of multi-molecular complexes by tethering proteins to cellular scaffolds (4).

Structurally, scaffolds are diverse proteins and their molecular architecture depends on the pathway they regulate. Even though the structures and origins of these proteins are diverse, they have a common converging function of signal integration, propagation and sometimes, termination. Since there is no strict definition of a scaffolding protein that is used by the scientific community, proteins can be defined as scaffolding proteins based on their functional properties. The four most important

Figure 1.1: Representative image showing mechanisms of distinct signaling compartment formation in cells.

Shown on left is a cell without signaling compartments. The hypothetical signaling pathway involves the yellow, light green and blue molecules represented as circles. In grey are the other nonspecific molecules in the cell. Shown on right, is a cell with specific signaling compartments. These compartments or micro-domains are structured in a cell by localizing specific proteins at a specific location in the cell. In this case, the yellow, light green and blue signaling molecules involved in a hypothetical signaling pathway can be compartmentalized by co-localization, by organelle formation or by binding to a scaffolding protein.

functions of a scaffold can be delineated as: 1) they act as molecular platforms for assembly of signaling molecules, 2) they localize signaling molecules at a distinct cellular location, 3) they coordinate negative and positive feedback signals to regulate signaling pathways and 4) they protect activated signaling molecules from getting inactivated. Most scaffolds will have at least two of the above described properties (5).

Sometimes, the scaffolds may themselves be subjected to different alterations, such as phosphorylations, sumoylations, and acetylations, to improve their efficiency of signal regulation. A well-understood example is phosphorylation of LAT (Linker for Activated

T cells) and SLP-76 (Leukocyte Specific Phosphoprotein) scaffold by a tyrosine kinase

Zap70 for T cell activation. The multiple phosphorylations of these scaffolds enable them to interact with several Src Homology 2 (SH2) domain containing proteins leading to activation of major signaling pathways involved in T cell activation (6, 7). Scaffolds are also capable of interacting with other scaffolding proteins to augment signal transmission, especially from one pathway to another. For example, a recent study showed that RACK1

(Receptor for Activated C Kinase), a PKC (Protein Kinase C) scaffold interacts with 14-

3-3ζ (8). The interaction between these two scaffolding proteins enables translocation of

RACK1 into the nucleus to activate expression of a transcription factor (8).

Another important mechanism that scaffolds may use to improve function is self- interaction, forming oligomers. The JNK (c-Jun N-terminal Kinase) associated scaffold

WDR62 (W-tryptophan D-aspartic acid Repeat domain 62) involved in mitosis regulation, dimerizes using a C-terminus motif downstream of the JNK binding site.

WDR62 mutants incapable of dimerization cannot efficiently bind JNK hence affecting

4 the signaling pathway (9). The C4b-binding protein is a heptameric scaffold found in immune cells for inhibiting the classical complement pathway and the lectin pathway

(10). Heptamer formation by C4b- binding protein increases its half-life in blood serum and helps it tether and circulate the anti blood coagulating S-protein in mammals. The extra-ordinary stability of this heptameric scaffold protein has made it an ideal candidate for drug delivery (11). Overall, scaffolds are recognized as important proteins that assist in maintaining fidelity and efficiency of signal transduction pathways in a cell.

1.2. AKAPS AND THEIR ROLE IN PKA SIGNAL INTEGRATION AND

REGULATION

One of the most characterized second messenger pathways involves cAMP signaling (shown in figure 1.2). For the heart, binding of norephinephrine to β- adrenergic (β-AR) receptors found on the cell membrane results in the activation of a small G-protein termed Gαs (12). Once active, the Gαs protein binds to the enzyme adenylyl cyclase (AC), thereby stimulating the production of cAMP (13). The major target of cAMP is the cAMP-dependent protein kinase A (PKA) (14). The PKA holoenzyme consists of two regulatory and two catalytic subunits (15, 16). Each regulatory subunit binds a single catalytic subunit. This gives rise to a tetrameric PKA holoenzyme (17, 18). Activation of PKA occurs when 2 molecules of cAMP bind to each of the regulatory subunits, resulting in dissociation of the catalytic

Figure 1.2: Diagram showing activation of cAMP inside a cardiac myocyte. cAMP activation in the cell requires activation of β-adrenergic receptors (represented as β-AR) upon binding of norepinephrine. This causes conformational changes in this ligand receptor leading to disassociation of the Gαs protein from the β -AR receptor protein bound trimeric G-protein complex intracellularly. Gαs traslocates to bind and activate the proximally localized adenylyl cyclases enzyme (represented as AC). Activated AC conducts breakdown of ATP to produce cAMP in the cell. The red to yellow gradient shown in the figure represents a cellular microdomain in which the cAMP production is localized. Specific cAMP microdomain formation is dependent on the ligand that leads to activation of AC in the cell. It has been found that different ligands give rise to different cAMP micro-domains for regulation of specific intra-cellular signaling proteins.

6 subunit (shown in figure 1.3). The now active catalytic subunit is free to phosphorylate target substrates. Inactivation of PKA occurs by hydrolysis of cAMP by enzymes called as Phosphodiesterases (PDEs) (19). Reduction of cAMP in cellular environment causes reassociation of PKA regulatory and catalytic subunit and hence PKA enzyme inactivation (20).

Mammals express 4 types of regulatory subunits, RIα, RIβ, RIIα, and RIIβ and 3 types of catalytic subunits (Cα, Cβ, Cγ) (21). All the isoforms of the catalytic subunit can interact with any of regulatory homodimers giving rise to various combinations. A

PKA tetramer consisting of R1 (α or β ) regulatory dimer is called as Type 1 PKA holoenzyme, whereas a PKA tetramer consisting of RII (α or β) regulatory dimer is called as Type II PKA holoenzyme (22). The tissue expression of these proteins is tightly regulated in order to achieve PKA function specificity (23).

Scaffolding proteins involved in regulation of Protein Kinase A (PKA) mediated phosphorylation are called as A-Kinase Anchoring Proteins (AKAPs). These scaffolds regulate the specificity and rapidity of substrate phosphorylation by tethering both the

PKA holoenzyme and its cellular target to a specific subcellular domain (21, 24).

AKAPs demonstrate three distinct properties to regulate substrate specificity (shown in figure 1.4); they must bind the holoenzyme of PKA, direct the kinase to a specific subcellular domain, and integrate multiple signaling enzymes that may be required for the regulation of the same or distinct signaling pathways. These properties are discussed in detail below (25).

Figure 1.3: Activation and deactivation of PKA catalysis.

Representative diagram shows the sequence of events involved in PKA holoenzyme activity regulation. PKA tetramer, also called as PKA holoenzyme consists of two regulatory subunits (shown as yellow triangles) that can dimerize. Each monomer of the regulatory dimer can bind to a single catalytic subunit, giving rise to the PKA tetramer (holoenzyme). The binding of the PKA catalytic subunit to the regulatory subunit keeps the catalytic subunit inactive. In order to activate the PKA catalytic subunit, cAMP is required. Under conditions of high cAMP, each regulatory subunit binds to two cAMP molecules (green circles). Binding of cAMP to the regulatory subunit causes release of catalytic subunit, which is free to catalyze phosphorylation of a Substrate (shown in mauve) using an ATP molecule. PKA catalytic subunit is deactivated by hydrolysis of cAMP by PDE. Loss of PKA regulatory subunit cAMP binding, promotes reassociation of PKA regulatory and catalytic subunits leading to reformation of the PKA holoenzyme and termination of PKA catalysis.

Figure 1.4: Representative image of an AKAP showing its properties and functions.

Three defining properties of an AKAP (in green): 1) Binds to the PKA holoenenzyme via the Regulatory subunit (shown in yellow) to localize Catalytic subunit (shown in red), 2) gets targeted to a specific cellular compartment (shown here an AKAP targeted to a membranous compartment), and 3) interacts with other proteins which might be involved in the signaling pathway (A and B proteins). Substrate phosphorylation can occur in 3 ways. Shown here are: Substrate 1 that is membrane associated and directly interacts with the AKAP for localization; Substrate 2 is bound to the AKAP directly and Substrate 3 present in the membrane but does not contribute to the targeting of the AKAP. Other AKAPs can get localized to a membrane via protein-protein interaction alone without interacting with the membrane (not shown here).

To be considered an AKAP, the scaffold must first bind PKA (26). This occurs via an amphipathic helix on the AKAP consisting of 14-18 amino acids that binds by hydrophobic interactions to the N-terminal dimerization/docking domain contained in the

RII dimer (27-29). While almost all AKAP PKA-binding sites may be modeled as an amphipathic helix motif, they share little primary sequence similarity, making identification of new AKAPs via BLAST or genomic searches unfeasible. Depending on the type of PKA an AKAP targets, AKAPs are defined as Type I, Type II or Dual

AKAPs. As their names suggest, Type I AKAPs bind to Type I PKA holoenzyme, Type

II AKAPs bind to Type II PKA holoenzyme, and Dual AKAPs are capable of binding to both types of PKA (30). Some well-characterized exclusive RII binding AKAPs are

AKAP79 and mAKAP (25, 31). Even though RII binding AKAPs are more common, there are AKAPs that exclusively bind to RI, such as integrin α4 and myosin 7a (32). The third category of AKAPs interacts with both RI dimers and RII dimers. Some examples are Ezrin (AKAP78), BIG-2, D-AKAP1 and D-AKAP2 (25, 33). It has been observed that AKAPs have a 10-100 fold higher affinity to bind to RII as compared to RI (30).

This difference in binding arises due to specific amino acid differences in the RI and RII

N-terminal helix that interacts with the AKAP amphipathic helix for docking of the dimer

(25).

Targeting of AKAPs at a sub-cellular location occurs by virtue of protein-protein, protein-lipid, or lipid-lipid interactions. The protein-protein interactions that mediate

AKAP localizations occur via a specific domain on the AKAP. For example, mAKAP has three spectrin like repeats that help localize this AKAP to the peri-nuclear region via its association with Nesprin-1α (34, 35). Yotiao interacts with its target protein, the

KCNQ1 (voltage gated potassium channel) channel, by a leucine zipper coiled-coil motif on the channel and specific N-terminal and C-terminal domains on Yotiao (36, 37). An example of protein-lipid interaction is exhibited by AKAP79, which directly interacts with phospholipids to be localized to the plasma membrane. Additional protein-protein interactions direct the AKAP to specific ion channels, such as the L-type Ca-channels or

M-type K-channels (38). Lastly, Gravin (AKAP250) is localized to the plasma membrane by virtue of N-terminal myristoylation along with presence of phospholipid binding motifs (39). These targeting mechanisms help an AKAP localize PKA to its substrate in specific cellular compartments.

Efficient compartmentalized cAMP signaling by AKAPs is mediated by their capacity to assemble multi-molecular complexes. AKAPs perform this function by assembling a multivalent complex that may consist of other kinases, ACs (Adenylyl

Cyclase(s), PPs (Protein Phosphatases), PDEs (Phosphodiesterases), and various other signaling proteins, depending on the requirement of the cell. The first AKAP-mediated protein complex was discovered for AKAP79. Yeast-2-hybrid analysis of the complex found that it associates not only with PKA, but also with PKC (Protein Kinase C) and

PP2B (Protein Phosphatase 2B, a calcium calmodulin dependent phosphatase) (40, 41).

Another example of a large complex assembly will be AKAP450. AKAP450 targets a complex consisting of PKA, PKC, polynucleotide kinase, PP2A, PP1, and PDE4D3 to centrosomes (32). Other examples are AKAP-Lbc, mAKAP, AKAP220, Gravin, and

WAVE. These multi-protein complexes are involved in signal integration of various pathways for a particular function and for regulation of substrate phosphorylation status.

Figure 1.5 shows some cardiac specific AKAP complexes.

Figure 1.5: AKAP complexes found in the heart.

Diagram represents few important AKAP complexes that have been detected in cardiac myocytes. Shown here is a cardiac myocyte with AKAP complexes associated with various cellular micro-domains. Example of plasma membrane channel associated AKAPs are Yotiao, AKAP79 and AKAP18α. AKAPs associated with cytoskeletal structures are Myospryn and AKAP Lbc. An example of cytoplasmic AKAP is D- AKAP2. Sarcoplasmic reticulum associated AKAPs are AKAP18γ and mAKAP. mAKAP also attaches to the nuclear envelope.

In cardiac myocytes, AKAP79, Yotiao and AKAP18α regulate function of ion channels located on the plasma membrane, whereas mAKAP and AKAP18δ have been shown to regulate calcium channels on the sarcoplasmic reticulum (42, 43). Other important AKAPs in a cardiac myocyte include AKAP-Lbc, myosryn, and D-AKAP2.

These are cytosolic AKAPs and have functions related to activation of transcription factors (AKAP Lbc), muscle contraction (myospryn) and endocytic cycling (D-AKAP2)

(44-46).

Some insight into the mechanism involved in the assembly of multi-molecular complexes is available to us. One hypothesis is that AKAPs can bind to multiple proteins by virtue of native unfolded regions in their protein structure (47-49). Another interesting mechanism is interaction of AKAPs with other scaffolding proteins. For example,

AKAP150 interacts with PSD-95 (Post Synaptic Density protein 95) to assemble the necessary molecules required for endocytosis of synaptic AMPA receptors (50). KSR-1

(Kinase Suppressor of Ras 1) scaffold, networks with AKAP-Lbc to mobilize cAMP responsive control of the ERK (Extracellular signal-Regulated Kinase) signaling pathway

(51). A third possible mechanism to construct multitude of AKAP complexes is AKAP oligomerization. AKAP-Lbc has been reported to form homo-dimers (52). This homo- dimerization is required for regulation of Rho-GEF activity of AKAP-Lbc. It was also found that AKAP79 dimerizes and this interaction is principally dependent on its C- terminus sequence. However, limited dimerization using the N-terminal domain was observed as well. It is thought that this dimerization increases the stoichiometry of phosphatase association with the AKAP (49). Another homo-oligmer forming AKAP is

Gravin (AKAP12) (47), although the function of Gravin oligomerizaiton is still unknown.

To summarize, AKAPs seem to adopt different mechanisms for efficient signal relay and execution. One of the mechanisms that they use is self-interaction. Discovery of the functional significance of AKAP oligomerization is an advancing field of research.

1.3. REGULATION OF PP1 FOR PKA SIGNAL PROPAGATION

Kinases regulate a plethora of cellular processes by phosphorylating various important substrates in a cell. Over 98% of the phosphorylations in a cell are on a serine/threonine residue while only around 2% of these phosphorylations occur at tyrosine residues (53). Hence, the role of serine/threonine phosphatases in regulating the phosphorylation states of proteins is critical. It is predicted that the human genome encodes for over 420 serine/threonine kinases but only about 40 serine/threonine phosphatases (54-57). Therefore, regulation of phosphatase specificity would be very important for achieving proper substrate dephosphorylation. Like kinases, the specificity of phosphatase activity in the context of its substrate is highly regulated. This is achieved by scaffolding proteins that anchor the phosphatase in a specific cellular domain and/or by endogenous phosphatase inhibitors that maintain the phosphatase in an inactive form when activity is not required (58, 59). The combination of accessory proteins regulating the phosphatase activity is not the same in all situations and this alteration in the combination of proteins regulates the substrate specificity of phosphatases.

Protein Phosphatase 1 is one the most well understood serine/threonine phosphatases in the mammalian system (60). It is a metallo-enzyme that requires di- valent cation for its activity (61). Mammals have three different genes that lead to

14 expression of four distinct gene products: PP1α, PP1β/δ, PP1γ1 and PP1γ2 (62). The human genome has an additional splicing event in PP1α gene product, giving rise to

PP1α1 and PP1α2 isoforms (63, 64). PP1 isoforms do not have any substrate distinction when examined under in vitro conditions. But, in cells, each isoform has a distinct substrate profile and a specific tissue distribution (65, 66). The molecular weight of PP1 isoforms is around 37kDa and the isoforms have a 90% amino acid sequence similarity.

The 10% difference arises due to differences in N- and C-terminus sequence of the isoforms, which also dictates specific interaction of each isoform with a specific regulatory protein (58, 67). There are over 200 bonafide or putative regulatory proteins known which can interact with 4 PP1 isoforms in mammalian cells (66). These regulatory proteins dictate the localization and substrate specificity of the PP1 isoforms.

PP1 Binding Proteins (PBPs) interact with PP1 with the help of specific amino acid motifs. Most PBPs have four to eight peptide domains for PP1 binding, whereas PP1 itself has over 30 non-.overlapping binding motifs on its surface for interacting with

PBPs (65). Shown in the figure 1.6 is the common structural motif, the RVxF motif that most PBPs contain (68). The RVxF motif interacts with a hydrophobic groove on the PP1 surface (shown by a green hexagon) not far from the PP1 catalytic site (red parallelogram) (69, 70). Binding of the RVxF motif alone to PP1 does not bring about any change in PP1 conformation, but it carries out the important function of bringing the

PBPs in close proximity to PP1 to allow for secondary interactions (71). Also shown in figure 1.6 are the secondary motifs such as the SILK motif and the MyPhoNE motif, found on multiple PBPs for PP1 regulation (68, 72). The functions of the secondary

Figure 1.6: Diagram depicting binding of PP1 regulatory proteins to different domains on PP1 surface.

Represented in light green is the PP1 enzyme with few of the known binding surfaces on the enzyme. PP1 binding protein (PBPs) A, B, C and D are represented here depicting the motifs they cover for binding to PP1. A: I-2 interaction with PP1 occurs via RVXF and two other motifs called as IDoHA and SILK, B: MyPT1 interaction with PP1 requires RVXF and MyPhone motif, C) Spinophilin interaction with PP1 is mediated by RVXF and SpiDoC, D) Spinophilin competes with I-2 for occupying the RVXF binding grove on PP1. This occurs when Spinophilin and I-2 bind to PP1 together.

16 binding motifs vary on a case-by-case basis. In the examples explained in figure 1.6, the

SILK motif which consists of consensus amino acid sequence K-[GS]-I-L-K, serves to bind I-2 (Inhibitor-2) to PP1 enzyme without causing conformational or catalysis changes to the latter (70, 73). On the other hand, the MyPhoNE motif (Myosin Phosphatase N- terminal Element) found in MyPT1 (Myosin Phosphatase Targeting subunit 1) contributes to PP1 substrate selection (72). MyPhoNe motif consists of the R-X-X-Q-

[VIL]-[KR]-X-[YW] consensus sequence (72). Other secondary binding motifs found in

PBPs consist of the SpiDoc (Spinophilin Docking site for C-terminal groove) found in

Spinophilin (74). In addition to RVXF and SILK motifs, I-2 also has the conserved

IDoHA motif (Inhibitor Docking site for Hydrophobic and Acidic groves) that binds to the PP1 active site (73). In summary, PBPs contain at least two domains that interact with PP1. The first, RVxF, provides a high affinity binding domain that allows for the initial interaction. Secondary domains, such as SILK and MyPhoNE, provide various, domain-dependent functions such as substrate definition and regulation of PP1 activity.

Since PP1 isoforms differ in their N- and C-terminal amino acid sequences, another important category of secondary binding motifs includes PBPs that specifically interact with either the PP1 N- and/or C-terminus. For example, MYPT1 binds to PP1β/δ

C-terminus (72). This governs PBP interaction in a PP1 isoform specific manner adding another level of specificity.

It is interesting that PBPs, when not bound to PP1, are either unstructured proteins altogether or have an unstructured domain that interacts with PP1 (75). A bioinformatics protein prediction analysis program called IUPRED predicts that about two-thirds of the total PBPs having the RVxF motif, are either unstructured or have unstructured domains

17 of at least 100 amino acids (75). A possible explanation for this phenomenon could be that having an unstructured protein might be crucial for formation of a large PP1 interaction surface to tightly regulate function of the PP1 holoenzyme. Unstructured

PBPs have an added advantage of intrinsic flexibility, enabling them to behave as scaffolds for other proteins to bind.

PP1 inhibition is an important facet in PP1 activity regulation. Potent inhibitory drugs such as microcystin-LR, nodularin-R and tautomycin have been well characterized

(76, 77). These molecules inhibit PP1 activity by binding to the PP1 active site. On the other hand, multiple endogenous PP1 inhibitors are PBPs and require binding via the

RVxF motif as well as secondary motif(s) to inhibit PP1. Examples of the well- characterized endogenous protein inhibitors of PP1 are DARPP32 (Dopamine and cAMP regulated PhosphoProtein 32), I-1 (Inhibitor-1) and I-2 (inhibitor-2). These inhibitors have the RVxF motif and secondary binding motifs to cause PP1 inhibition (60).

Additionally, some inhibitors, like DARPP-32 and I-1 require phosphorylation in order to inhibit PP1 (67). Other inhibitors, such as I-2, do not require a phosphorylation event to inhibit PP1, but inversely need a phosphorylation event to eliminate PP1 inhibition (78).

In both cases, binding of the inhibitor to the phosphatase is not dependent on inhibitor protein phosphorylation event.

In addition to being regulated at two different levels i.e by tissue specific expression and targeting and inhibition by PBPs, PP1 is also regulated by scaffold proteins that direct substrate specificity for this enzyme. Unlike PP1 inhibitors and targeting proteins, substrate-specifying proteins act in a more sophisticated manner. A number of substrate specifiers are targeting or inhibiting PBPs as well. Substrate

18 specifying PBPs can use positive or negative substrate selection mechanisms. Positive selection occurs when a substrate specifier protein increases PP1 activity towards a specific substrate and negative selection occurs when PP1 activity is selectively reduced for a subset of substrates (65). The glycogen interacting G-subunits (GM and GL) involved in PP1 targeting for glycogen metabolism regulation exemplify both positive and negative substrate selection (67, 71). As shown in figure 1.7, in the presence of insulin, these scaffolds inhibit PP1 activity towards glycogen phosphorylase (loss of activity), but promote PP1 mediated dephosphorylation of glycogen synthase (gain of activity) required for glycogen synthesis (58, 79, 80).

Another interesting example of a PP1 substrate specifier that was recently described is the Repo-Man protein (81). Repo-Man is a chromosomal bound protein that interacts with PP1. In this particular study, Repo-Man was discovered to restrict the activity of PP1 to two phosphorylation sites of histone H3, threonine-3 and threonine-11, although there are other phosphorylated amino acids on this protein (threonine-10 and serine-28) that are not dephosphorylated by Repo-Man-bound PP1. These specific de- phosphorylations are required for normal cell division. The molecular mechanism behind this phenomenon is still under investigation.

To summarize, PP1 isoforms have a tissue specific expression for their activity regulation. Importantly, the permutation and combination of the regulatory proteins

(subunits) with endogenous PP1 inhibitors and substrate specifiers makes PP1 activity highly specific within a cellular environment.

Figure 1.7: Positive and negative regulation of PP1 activity by substrate-specifying glycogen targeting subunits.

In the presence of Insulin, glycogen synthesis is promoted. PP1 is targeted to glycogen particles in tissue by glycogen targeting proteins called as GM and GL. Glycogen synthesizing enzyme called glycogen synthase needs to be dephosphorylated for it to be activated (blue arrows), whereas glycogen breakdown is inhibited by de-activating glycogen phosporylase by preventing its dehosphorylation by PP1. This exemplifies PP1 activity regulation by substrate specifying scaffolds GM and GL that direct its activity only to glycogen synthase and not to glycogen phosphorylase.

1.4. REGULATION OF PP1 BY I-1

Inhibitor-1 protein was discovered in 1976 and was the first endogenous PP1 inhibitor to be identified (82). Unlike PP1, which can be transported into the nucleus,

(83), I-1 is a cytoplasmic protein that is widely expressed in mammals including heart and skeletal muscle (84, 85). I-1 is only 171 amino acids long with a theoretical molecular weight of 19kDa (86, 87). I-1 expression is vertebrate specific and the N- terminus of the protein is highly conserved among species (58, 86). In humans, I-1 mRNA is alternatively spliced to give rise to three variants: full length I-1, I-1α and I-1β

(88). The specificity of each isoform to independent cellular processes is unknown at present.

Following the isolation of I-1 from rabbit skeletal muscle, it was found that nano- molar concentrations of purified PKA phosphorylated I-1 could inhibit PP1 whereas micro-molar concentration of unphosphorylated I-1 could not (89). The binding affinity of I-1 for PP1 is very low when I-1 is dephosphorylated. Phosphorylation by PKA dramatically increases the binding affinity, resulting in inhibition of phosphatase activity

(90, 91). Point mutation experiments on I-1 amino acid sequence found that I-1 phosphorylation by PKA at threonine-35 is absolutely required for I-1 mediated inhibition of PP1 (87). Therefore, cAMP dependent I-1 phosphorylation leads to a feed forward mechanism for enhancing substrate phosphorylation by mediating PP1 inhibition and by PKA mediated substrate phosphorylation. Inactivation of I-1 towards PP1 inhibition occurs via its dephosphorylation at the threonine-35 site by PP2A and PP2B in rat neonatal cardiac myocytes (92). Furthermore, a second phosphorylation event that

21 regulates I-1 activity towards PP1 is PKC mediated I-1 phosphorylation at serine-67 and threonine-74 in heart cells. These phosphorylation events cause decline in I-1 mediated

PP1 inhibition (93).

Since the T35 phosphorylated I-1/PP1 complex crystal structure has not been solved yet, our understanding of structural basis of I-1 mediated PP1 inhibition comes from NMR studies of this complex (94). Inability to perform X-ray crystallography of this complex can be primarily attributed to intrinsically disordered nature of the I-1 protein. NMR data (represented in figure 1.8) shows that unphosphorylated I-1 can interact with PP1 at a different site than the catalytic core (yellow rhombus in the PP1 enzyme in the figure). On the other hand, NMR study using T35 phosphorylated I-1/PP1 complex suggests that phosphorylation of I-1 pushes the I-1 protein inside the catalytic core of PP1, causing phosphatase inhibition (also shown in figure 1.8) (94). I-1 associates with PP1 with the help of a RVxF like motif, KIQF present in the N-terminus of the protein (amino acids 8-12). While this motif allows for the interaction of the two proteins, it does not play a role in the inhibition of the phosphatase by the inhibitor. Inhibition can only occur when I-1 is phosphorylated at T35 by PKA (87, 94, 95).

Figure 1.8: Regulation of PP1 by PKA phosphorylated I-1.

Figure represents the I-1 mediated PP1 regulation after PKA mediated phosphorylation of I-1. Shown on the left is PP1 enzyme (hexagonal dull orange) with its catalytic domain (yellow rhombus). Unphosphorylated I-1 (lavender polygon) can associated with PP1 but does not inhibit its activity. Upon PKA phosphorylation, I-1 goes through conformational changes that moves the I-1 phosphorylated region into the PP1 catalytic domain leading to PP1 inhibition. It has been found in cardiac myocytes that this process can be reversed by phosphatases called as PP2A and PP2B.

1.5. REGULATION OF PP1 BY AKAPS

Study of molecular mechanism of PP1 regulation by AKAPs is an emerging field.

Multiple AKAPs have been shown to interact with PP1 for regulation of substrate dephosphorylation. This is achieved by localizing the phosphatase in close proximity to its substrates, but work specifically focusing on molecular interactions of PP1 and AKAPs for PP1 regulation is in progress (96). One of the AKAPs that binds to PP1 and has been studied in this context is AKAP149. AKAP149 is a human AKAP that has been shown to play an important role by recruiting PP1 for nuclear envelope reassembly during mitotic exit. AKAP149 has the PP1 binding motif RVXF (K153GVLF157) and mutation of the binding motif disables this AKAP from binding to PP1 (97). Moreover, AKAP149 also inhibits PP1 activity against phosphorylase a in vitro but enables PP1 to dephosphorylate

B-type Lamin in vivo, suggesting that AKAP149 also acts as a substrate specifying scaffold for PP1 (98). Furthermore, serine phosphorylation around the RVXF motif of

AKAP149 itself is a mechanism for reversal of PP1 binding to the AKAP. The phosphorylation at S151 or S159 can occur by PKA or PKC bound to AKAP149 but the release of PP1 occurs only when one of the serine residues is phosphorylated by PKC associated with this complex. This suggests a model where PP1 is bound to the AKAP for a specific purpose and then released by activity of another kinase bound to the complex

(99).

Another interesting example of PP1 regulation by an AKAP is that of AKAP220.

AKAP220 not only has the consensus PP1 anchoring motif (KVxF), but also has C-

24 terminal motifs (2 or more,) that are involved in PP1 inhibition by binding to the PP1 active site [Figure 1.9]. Moreover, addition of RII dimer to AKAP220-PP1 reaction mixture enhances the ability of the AKAP to inhibit PP1 by four folds using phosphorylase a as a PP1 substrate. However, the molecular mechanism for RII-mediated inhibition of AKAP220-bound PP1 is not known. It is a very exclusive case of PP1 inhibition by an AKAP, as the PKA catalytic subunit does not participate in any phosphorylations that may inhibit the phosphatase. Instead, the authors hypothesize that binding of the regulatory subunit of PKA to the AKAP causes conformational changes in the AKAP that potentiate AKAP220 for PP1 inhibition (100).

A more recent discovery was the AKAP79 mediated inhibition of PP1 by our laboratory. Previously, AKAP79 was recognized for its PP2B (calcineurin) anchoring and inhibition (40). In a recent report, Le et al. showed that AKAP79 had a consensus PP1 binding motif in the first 44 amino acids. Furthermore, deletion of this domain did not attenuate PP1 binding to AKAP79 suggesting the presence of another PP1 binding domain. Mapping experiment showed that the second PP1 binding domain on AKAP79 exists between amino acids 150-250. Furthermore, this second PP1 binding domain not only bound the phosphatase but also inhibited its catalytic activity towards a generic substrate. On the other hand, AKAP79 bound PP1 was active in the context of a specific

PP1 substrate found in the brain, PSD95. Based on these observations, the study concluded that AKAP79 acts as a PP1 anchoring protein that defines the substrate selectivity of this phosphatase (101).

Figure 1.9: PP1 regulation by AKAP220.

AKAP220 has mutiple PP1 binding sites. The concensus PP1 binding motif of AKAP220 KVXF is located in the N-terminus of the protein but has no effect on PP1 inhibition. The C-terminus of the protein has multiple PP1 binding moifs (not completely defined yet) that can bind to PP1 and inhibit its activity (A). PP1 inhibition is greatly increased by AKAP220 upon RII dimer binding to the AKAP. RII binding causes conformational changes to the AKAP increasing its affinity for PP1 and hence increased inhibition. RII dimer does not alter KVXF motif mediated PP1-AKAP220 interaction.

1.6. IMPORTANCE OF PP1 AND I-1 IN DISEASE

One of the hallmarks of a failing heart is reduced phosphorylation of multiple

PKA targets which includes PLN (102, 103). PLN is a 6 kDA protein found in the SR membrane and has been widely recognized as a negative regulator of calcium uptake pump on the SR membrane, SERCA (Sarcoplasmic/Endoplasmic Reticulum Calcium

ATPase) (104). As illustrated in Figure 1.10A, in its dephosphorylated state PLN binds to SERCA, and reduces its affinity for calcium, ultimately resulting in less calcium being pumped across the membrane into the SR. Under β-adrenergic stimulation and the increase in cAMP and PKA activity (Figure 1.10B), PLN is phosphorylated.

Addtionally, PKA also phosphorylates I-1 to inhibit PP1 activity, allowing for enhanced

PLN phosphorylation. These two actions result in the loss of PLB/SERCA interaction and an increase in pump activity (105-107). After the stimulatory conditions have waned off, PLN is dephosphorylated by PP1 to restore contractility to basal state (108).

The reduction in PLN phosphorylation in both human and animal models of heart failure has been attributed to both upstream processes, such as reduced cAMP generation in the cardiac myocyte, and downstream processes, such as increased phosphatase activity (109-111). It was discovered that the increase in PP1 activity is regulated at the micro-domain level because while PLN was significantly dephosphorylated, other PP1 targets were not (37). Further work attributed the increased PP1 activity to a decrease in

I-1 expression (mRNA and protein level decreased by 60%) combined with decreased

Figure 1.10: Regulation of Phospholamban (PLN) by PP1.

A) Unphosphorylated PLN binds to SERCA pump and reduces its affinity for calcium leading to reduced calcium uptake into SR. B) β-adrenergic activation leads to PLN phosphorylation by PKA preventing its binding to the SR calcium uptake pump, SERCA. SERCA when not bound by PLN, has high affinity for calcium leading to efficient calium uptake inside SR. This is also promoted by simultaneous inhibition of PP1 by PKA phosphorylated I-1. These events are necessary for maintaining cardiac responses under which cardiac output has to be increased such as exercise.

28 percentage of total I-1 that was phosphorylated at T34 (about 80% reduction) in human hearts as well as animal models of heart failure (112-114).

Furthermore, it has also been identified that in cardiac myocytes, I-1 undergoes

PKC mediated phosphorylation at amino acids serine-67 and threonine-74. These phosphorylations prevent cAMP mediated I-1 phosphorylation and are not inhibitory to

PP1 activity. This is important under conditions where high levels of PKC activity attenuate the PKA effect due to lack of I-1 inhibition (93, 115). It is proposed that this might be the mechanism by which I-1 mediates the cross talk between the PKA and PKC pathways under normal conditions for regulation of cardiac function. [Figure 1.11].

Under pathological conditions, PKA is downregulated in cardiac myocytes whereas PKC is upregulated, causing a total reduction in I-1 phosphorylation at threonine-35 leading to loss of optimal PP1 inhibition (93, 112, 114, 116). Furthermore, PKC mediated I-1 phosphorylation also prevents I-1 from being accessible for PKA phosphorylation (93,

117). Together the PKA and PKC mediated I-1 phosphorylation changes lead to a high

PP1 activity in diseased hearts, which has been shown to be one of the reasons for heart failure (112, 118-121) [Figure 1.12]. Following this observation, various mice models of heart failure were developed in order to fully understand the role of I-1 in regulation of heart function and to explore the possibility of using I-1 as a novel drug target.

Figure 1.11: Regulation of PP1 by I-1 for normal cardiac function.

Pictorial representation of I-1 regulation by PKA and PKC has been shown. PKA mediated phosphorylation of I-1 helps in PP1 suppression to enhance cardiac function. PKC mediated phosphorylation of I-1 prevents I-1 mediated PP1 inhibition leading to depressed cardiac function. In a healthy heart, both PKA and PKC mediated phosphorylations maintain PP1 activity at optimum levels.

Figure 1.12: Misregulation of PP1 due to alteration of I-1 phosphorylation.

Diagram shows change in I-1 phoshorylation in a diseased heart. Decreased PKA activity in cardiac myocytes subjected to heart failure promoting conditions, causes reduction of PKA mediated I-1 phosphorylation leading to loss of PP1 inhibition. Decline in PP1 inhibition is further promoted by increased PKC activity in disease condition. PKC mediated phosphorylation of I-1 prevents I-1 from inhibiting PP1 and also makes I-1 unavailable for PKA phosphorylation. Together these changes in I-1 phosphorylation promote a significant increase in PP1 activity causing progression of heart disease into heart failure.

Early investigations revealed that heart specific overexpression of I-1c (N- terminus of I-1 having the PKA phosphorylation site was expressed with a threonine-35 to aspartate mutation for constitutive PP1 inhibition) in mice protected them against heart failure (122, 123). Recently it was also demonstrated that I-1c overexpression inhibits cardiomyocyte apoptosis induced by chronic beta agonist administration (124). These results suggested that targeting over-expression of I-1 could be a potential therapy for patients. However, contradictory to these results, another group of investigators have reported that I-1 knockout mice are protected against heart failure using a similar heart failure model (125). Further study demonstrated that conditional over-expression of I-1c on I-1 knockout background produced contractile dysfunction under administration of catecholamines and this effect could be reversed by turning off I-1c expression (126).

Importantly, I-1c overexpressing mice developed cardiomyopathies with age unlike normal aging heart (126, 127). The lack of concurrent results from these studies questions the therapeutic potential of I-1 (128).

The inconsistent results obtained by different groups requires further investigation of

I-1 mediated inhibition of PP1 in cardiac myocytes. Furthermore, even though the mechanistic details of I-1 inhibition of PP1 are known, it is not known which factors or proteins regulate these two proteins for their micro-domain localization (both these proteins are concentrated at SR, described above). The work presented in this thesis is an effort towards answering these questions.

1.7. AKAP18 AS A MODEL AKAP

AKAP18 (also called as AKAP7, based on gene name) was independently discovered in two separate laboratories (129, 130). Several studies before that had shown that injecting inhibitor peptides that disrupt AKAP-PKA binding attenuates membrane channel function suggestive of PKA being anchored at the cell membrane for regulation of L-type Ca channels in cardiac and skeletal muscle. The AKAP that was showing this effect was cloned and identified as AKAP15/18 (129, 130). Immediately after this other

AKAP18 isoforms were discovered. These are shown in Figure 1.13. It was found that alternative splicing leads to evolution of three isoforms: the originally discovered short isoform AKAP18α, the newly cloned short isoform AKAP18β, and a long isoform

AKAP18γ.

The two short isoforms are almost identical except that AKAP18β, a little longer than AKAP18α, had a 23 new amino acids inserted at its N-terminus. The N-terminal of

AKAP18γ was identical to the short isoforms except that it lacked the membrane targeting lipid modification motif that the short isoforms had. Moreover, AKAP18γ had an extended amino acid sequence between the identical N and C-terminus it shared with the shorter isoforms. Transfection studies using GFP tagged AKAP18 plasmids showed that the short isoforms were targeted to the membrane while AKAP18γ (long isoform) was targeted to the cytoplasm (131). A few years after these discoveries, another long isoform of AKAP18, called as AKAP18δ was discovered in the renal principal cells where it augments vasopressin-induced aquaporin 2 shuttling. AKAP18δ is the longest of the four AKAP18 isoforms with 356 amino acids while AKAP18γ has 326 amino

Figure 1.13: AKAP18 isoforms.

Linear diagram showing AKAP18 isoforms. AKAP18 isoforms are splice variants of AKAP18 gene and share significant protein sequence homology. The short splice variants are AKAP18α and AKAP18β. They are localized to the plasma membrane by virtue of post-translational lipid modifications. AKAP18α is directed to L-type Calcium Channel (LTCC) for channel phosphorylation. The larger isoforms are AKAP18γ and AKAP18δ. These are cytoplasmic and get localized to cellular domains via protein-protein interactions. The large isoforms have been shown to localize to the Sarcoplasmic Reticulum in cardiac myocytes and regulate Phospholamban (PLN) phosphorylation.

acids, followed by AKAP18β and AKAP18α, which have 104 and 81 amino acid residues respectively (132).

AKAP18α undergoes N-terminal myristoylation/palmitoylation in order to localize to the plasma membrane. At the plasma membrane, it interacts with L-type Ca channel in skeletal muscle and heart and it interacts with the Na-channel in the brain to modulate channel activity via PKA mediated phosphorylation of these channels. AKAP18β is only

23 amino acids longer than AKAP18α due to an insertion at its N-terminus, but both

AKAP18α and AKAP18β exhibit different expression patterns when transfected into

MDCK cells. AKAP18β is targeted to the apical plasma membrane in polarized MDCK cells whereas AKAP18α is localized to the basolateral and apical plasma membrane in these cells (130, 131). The physiological function of AKAP18β still needs to be defined.

AKAP18γ has a nuclear localization signal, which targets it to the nucleus in mouse oocytes. This is a particularly interesting report as the data showed species specific differences in AKAP18γ sequence leading to difference in PKA holoenzyme binding and also localization of the AKAP. Human AKAP18γ is predominantly a PKA type II binding AKAP and is localized in the cytoplasm via protein-protein interaction. In case of mouse oocytes, AKAP18γ has an additional RGD (Arginine-Glycine-Aspartic Acid) sequence which enables integrin binding. Also, AKAP18γ in mouse oocyte is a PKA type

I binding AKAP, which is attributed to predominant expression of PKA type I in these cells. The PKA binding domain of human pancreatic and mouse oocyte AKAP18γ is identical (133). AKAP18δ was discovered in the renal cells but later it was found to be important in regulating phosphorylation of PLN (134). A recent study elucidated that the

SR localization of PKA anchored to AKAP18δ in rats is carried out in humans and mice using AKAP18γ, suggesting that apart from tissue specific variations, there can be species specific variations in expression of AKAP18 isoforms (135). Interestingly, it has been shown that in INS-1E cell line, silencing of AKAP18α or AKAP18γ leads to corresponding decrease or increase in insulin secretion upon glucose sensitization. This is a unique example of an AKAP whose isoforms may have opposite effect on a physiological process (136).

Goals and summary of results of the thesis work:

In summary, previous work demonstrated PLN phosphorylation is the key mechanism used by the heart to enhance the speed of relaxation. By tethering PKA to

PLN in cardiac myocytes, AKAP18 mediates this phosphorylation event. However, in the failing heart, PLN is hypo-phosphorylated. This is correlated with a decrease in I-

1 phosphorylation by PKA and increased PP1 activity. I investigated if there is an

AKAP that anchors I-1 and regulates its PKA mediated phosphorylation required for

PP1 inhibition [Figure 1.14]. My thesis work demonstrates that AKAP18γ can bind to

I-1 and enhance its PKA mediated phosphorylation. Before my thesis work, there were no reports suggesting direct interaction between a PP1 inhibitor protein and an AKAP, even though PP1 has been found to interact with various AKAPs. Moreover, the

AKAP18γ-I-1 complex can inhibit PP1 in a PKA dependent manner. This is an important finding as AKAP18γ not only binds to I-1, it regulates its function of PP1 inhibition. These results suggest that AKAP18γ localizes I-1 to the SR for PP1 regulation. Furthermore, I also investigated the mechanism that may be involved in efficient PKA mediated PLN and I-1 phosphorylation by AKAP18γ. We found that

AKAP18γ oligomerizes. Oligomerization of AKAP18γ should enhance substrate phosphorylation impacting both PLN and I-1. Together, these findings show that

AKAP18γ organizes a multi-molecular complex to regulate PLN phosphorylation, by directly enhancing the associated PKA activity and by inhibiting PP1 via I-1. Hence,

AKAP18γ establishes itself as a bi-directional regulator of protein phosphorylation.

Figure 1.14: Diagram showing the premise of hypothesis.

AKAP18γ/δ is localized at the Sarcoplasmic Reticulum (SR) in cardiac myocytes to mediate PLN (phospholamban) phosphorylation. This AKAP is localized to the membranous compartment by virtue of its direct interaction with PLN. PP1 is also known to be localized at the SR for PLN dephosphorylation. I-1 has been shown to be concentrated at the SR for PP1 regulation. I-1 becomes a potent PP1 inhibitor after PKA mediated phosphorylation. We propose that the localized I-1 phosphorylation required for PP1 inhibition at the SR is regulated by an AKAP.

2. THE LARGE ISOFORMS OF AKAP18 MEDIATE THE

PHOSPHORYLATION OF INHIBITOR-1 BY PKA AND THE INHIBITION

OF PP1 ACTIVITY

Arpita Singh, John M. Redden, Michael S. Kapiloff, Kimberly L. Dodge-Kafka

AS, JMR and KLD-K: Pat and Jim Calhoun Center for Cardiology,

University of Connecticut Health Center, Farmington, CT 06030

MSK: Depts of Pediatrics and Medicine, Interdisciplinary Stem Cell Institute,

University of Miami Miller School of Medicine, Miami, FL 33101

Molecular Pharmacology, 2011, 79(3):533-40

Author Contributions: Majority of the experiments were designed and performed by AS

with help from JMR and KDK. AS, JMR, MSK and KDK helped write the manuscript.

2.1. ABSTRACT

Inhibitor-1 (I-1) is phosphorylated on threonine residue 35 (Thr-35) by the cAMP- dependent protein kinase (PKA), inducing the potent inhibition of the serine-threonine- specific protein phosphatase 1 (PP1). We now report that the formation of a signaling complex containing PKA and I-1 by the A-kinase anchoring protein, AKAP18, facilitates this regulation in cells. AKAP18 directly bound I-1, and AKAP18/I-1 complexes were isolated from both rat heart extract and transfected heterologous cells. Importantly, prevention of PKA binding to the AKAP18 scaffold decreased I-1 phosphorylation by

48% in cells. Moreover, the I-1 target PP1 was also associated with AKAP18 complexes.

The cAMP-mediated inhibition of phosphatase activity was contingent upon PKA binding to the scaffold. These observations reveal an additional level of complexity in

PP1 regulation due to its association with AKAP18 multi-molecular signaling complexes, and suggest that targeting of AKAP18 complexes may be an alternative method to alter phosphatase activity and modulate specific substrate dephosphorylation.

2.2. INTRODUCTION

Protein phosphorylation is a key regulator of cellular physiology that affects essentially all biological processes. Despite our vast knowledge of the consequences of protein phosphorylation, the molecular mechanisms that confer specificity to the enzymes controlling this post-translational modification are not well understood, especially for the handful of phosphatases that catalyze the dephosphorylation of protein substrates (57,

67). Seminal investigations have demonstrated that the cell has evolved multiple strategies that control the location, activity and substrate specificity of each phosphatase

(57, 67). Current research is focused on the contribution of individual multi-molecular signaling complexes to the specificity of intracellular signal transduction.

The first protein discovered to regulate phosphatase activity was the Protein

Phosphatase Inhibitor-1 (I-1) (91). I-1 is a 19 kDa, heat stable protein that was identified over 30 years ago as a regulator of glycogen metabolism (91). While I-1 has no known intrinsic catalytic activity, its binding reduces the activity of protein phosphatase 1 (PP1).

I-1 is highly expressed in the brain, where it plays a role in synaptic plasticity (137, 138).

Although I-1 is expressed at low levels in the adult myocardium, recent work suggests that regulation of PP1 by I-1 at the sarcoplasmic reticulum is required for normal cardiac function, as well as for the response of the heart to disease (139, 140). I-1 itself is phosphorylated by PKA, PKC and CDK5 (91, 139, 141, 142) and phosphorylation of I-1 at Thr-35 by PKA induces selective inhibition of PP1 (91, 140). This event is induced by

β-adrenergic stimulation in cardiac myocytes, and potentiates the phosphorylation of key

PKA substrates involved in excitation-contraction coupling by preventing their PP1- mediated dephosphorylation (140). As excessive PP1 activity resulting from a lack of I-1 function has been suggested to contribute to heart failure, a more complete understanding of the molecular regulation of this phosphatase may aid in the design of novel therapeutics to prevent or treat heart disease (93, 118, 122, 125).

The mechanisms conferring specificity on PKA phosphorylation have been of considerable interest in recent years, and research has shown that many PKA targets are co-localized with the kinase via the association with A-kinase anchoring proteins

(AKAP) (4, 42). These scaffold proteins function to enhance the kinetics and specificity of PKA phosphorylation by sequestering the kinase with its target substrates, allowing for

41 spatiotemporal control of cAMP signaling. Importantly, disruption of PKA binding to

AKAPs in the heart significantly decreases the ability of the kinase to phosphorylate many key proteins such as troponin I, the ryanodine receptor and phospholamban (143).

As I-1 is a target for PKA, we investigated whether an AKAP mediates this event.

We discovered that I-1 binds the large isoforms of AKAP18 in the heart. AKAP18 potentiated the phosphorylation of I-1 by PKA, and disruption of PKA binding to the scaffold significantly attenuated Thr-35 phosphorylation in HEK293 cells. Moreover,

PP1 was also associated with AKAP18 complexes, and the PKA-dependent inhibition of phosphatase activity required PKA-AKAP18 binding. These data imply that AKAP18 orchestrates the PKA-catalyzed phosphorylation of I-1 in the myocyte, providing an additional level of regulation of PP1 activity at the sarcoplasmic reticulum, where the large isoforms of the AKAP are enriched.

2.3. MATERIALS AND METHODS

2.3.1. Antibodies

The following antibodies were used for immunoblotting: rabbit polyclonal to inhibitor-1 (Abcam, 1:10,000), goat polyclonal to AKAP18 (received from Dr. John

Scott, University of Washington, 1:1000), rabbit polyclonal to GFP (Invitrogen, 1:500), rabbit polyclonal phospho-specific antibody against Threonine-35 of I-1 (Cell Signaling,

1:1000), mouse monoclonal to PP1 catalytic subunit (Santa Cruz Biotechnology, 1:1000), mouse monoclonal to myc tag (Santa Cruz Biotechnology, 1:1000), mouse monoclonal to

PKA RI subunit (BD Bioscience, 1:5000), mouse monoclonal to PKA-RII subunit (BD

Bioscience, 1:5000), and mouse monoclonal to PKA catalytic subunit (BD Bioscience

1:5000). The following antibodies were used for immunoprecipitations: rabbit polyclonal to inhibitor-1 (Abcam, 2ug), mouse monoclonal to PP1 catalytic subunit (Santa Cruz

Biotechnology, 5ug), mouse monoclonal to PP2A catalytic subunit (Santa Cruz

Biotechnology, 5ug), goat polyclonal to AKAP18 (received from Dr. John Scott,

University of Washington, 5ul), mouse monoclonal to myc tag (Santa Cruz

Biotechnology, 5ug), and mouse monoclonal to PKA-RII subunit (BD Biosciences, 2ug).

2.3.2. Expression Constructs

For I-1 constructs, human I-1 was PCR amplified using a cDNA, supplied by Dr.

Shirish Shenolikar, Duke University, and sub-cloned into the EcoRI-Xho sites of pGEX-

4T1 and pCDNA3. Full-length human AKAP18γ (supplied by Dr. John Scott, University of Washington) as well as the AKAP18γ deletion constructs were cloned into either the

EcoRI-BamHI sites of pEGFP-N1, or the EcoRI-HindIII sites of pET32a.

2.4. EXPERIMENTAL PROCEDURES

2.4.1. cAMP Agarose Purification

Rat heart extract was prepared by homogenization of a single rat heart (PelFreeze) in 10 ml HSE lysis buffer (20mM HEPES [pH 7.4], 150mM NaCl, 5mM EDTA,1%

Triton X-100, and protease inhibitors). Following centrifugation at 12,000g for 20 minutes at 4°C, 1 ml of supernatant (500ug) was incubated overnight with 60ul of Rp- cAMPs-agarose (Biolog, Bremen, Germany) in the presence and absence of 100mM cAMP (Sigma). The beads were washed twice with HSE lysis buffer, followed by a high

43 salt wash (20mM HEPES [pH 7.4], 600mM NaCl, 5mM EDTA). Bound proteins were analyzed by immunoblotting.

For cAMP agarose pulldowns from HEK293 cells, cells were transfected at 70% confluency in 60mm dishes using calcium phosphate. Cells were harvested and lysed 24 hours after transfection in 0.5ml of HSE buffer containing protease inhibitors. Cell lysate was added to 60ul of Rp-cAMPs-agarose and bound proteins were isolated as detailed above.

2.4.2. Pulldown Experiments

Heart extract (1ml) was prepared in HSE buffer as described above and incubated overnight with glutathione-resin pre-absorbed with bacterially expressed fusion proteins.

The beads were then washed as described and bound proteins were identified by immunoblotting. Alternatively, for pulldown experiments using purified proteins, GST fusion proteins were incubated with S-tagged purified AKAP18 fragments bound to S- protein agarose in HSE buffer lacking EDTA. After an overnight incubation, beads were washed 3X in the HSE buffer and associated proteins were identified by immunoblot.

2.4.3. Immunoprecipitation

For immunoprecipitation of AKAP18 constructs from heterologous cells, HEK293 cells were transfected at 70% confluency using the calcium phosphate method. 24 hours post transfection, cells were treated with 1uM Isoproterenol for 5 minutes. Cells were rinsed in PBS, and then harvested and lysed in 0.5ml of HSE buffer containing protease inhibitors plus phosphatase inhibitors (10uM sodium pyrophosphate, 50uM sodium

44 fluoride). Supernatants were incubated with 4ug of antibody and 20ul of protein G- agarose beads. Bound proteins were analyzed by immunoblotting.

For immunoprecipitations from rat heart, extract was prepared as described above.

Proteins were immunoprecitated from 1ml of extract using specific antibodies bound to protein G-agarose. After extensive washing with HSE buffer, the associated proteins were determined by immunoblot.

2.4.4. Phosphatase Assay

Phosphatase activity was measured as previously described using 32P-labeled histone as a substrate (144). Histone was radiolabeled in a reaction containing 250mM

MOPS (pH 7.4), 2.5mM MgAc, 100mM beta-mercaptoethanol, purified PKA catalytic subunit, 1mM ATP, 20mM histone and 1mCi [gamma-32P-ATP (6000 Ci/mmol)]. The reaction was terminated by the addition of 50% TCA and [32P]-histone was purified from free radionucleotide by centrifugation. After repeated washing, the [32P]-histone was suspended in 200ul of PP2A buffer (25mM Tris, [ph 7.4], 1mM DTT, and 10mM

MgCl2).

To measure phosphatase activity, immunoprecipitated complexes were washed twice in HSE buffer and again in PP2A buffer. The immunoprecipitates were incubated for 30 minutes at 30°C in 20ul of PP2A buffer containing 100,000 cpm of [32P]-histone in the presence and absence of inhibitors. Reactions were terminated by the addition of

100ul of 20% TCA followed by 10 min centrifugation. TCA supernatants containing

32 released P04 were measured using a scintillation counter.

2.4.5. PKA activity assay

PKA activity was measured according to the method of Corbin and Reimann (145).

Immunoprecipitated complexes were washed twice in HSE buffer and once in PKA assay buffer (50mM Tris, pH 7.5 and 5mM MgCl2). The beads were then incubated in assay buffer containing 30uM Kemptide 100uM ATP, 5uM γP32-ATP and 3uM cAMP. After incubating the samples at 30°C for 30 minutes, the reaction mixture was spotted onto phosphocellulose strips and washed five times in 75mM phosphoric acid and once in

95% ethanol. Filters were air dried and counted.

2.4.6. Detection of I-1 phosphorylation

For I-1 phosphorylation experiments, 24 hours post transfection, cells were stimulated with 1uM Isoproterenol for 5 minutes. Cells were rinsed in PBS and scraped into 0.5 ml of HSE buffer containing protease and phosphatase inhibitors (10mM sodium pyrophosphate, 50mM sodium fluoride). After centrifugation for 10 minutes at 13,000g, the clarified supernatant was added to 1ml of isopropanol, and total cellular protein was precipitated by incubation at -20°C for 2 hours followed by centrifugation at 13,000g for

30 minutes. Precipitated proteins were resuspended in 200ul of 2X SDS-PAGE sample buffer (0.5M Tris-HCl, [pH 6.8], 1% glycerol, 10% SDS, and 0.5% beta- mercaptoethanol) that was added to the cellular pellet. For immunoblot analysis using a phospho-specific antibody, 20ul of each isolated sample was subjected to SDS-PAGE.

For in vitro I-1 phosphorylation GST-fusion proteins were incubated with cellular extracts isolated from transfected HEK293 cells as described above. After an overnight incubation and washing of the beads with HSE buffer lacking EDTA, the complex was

incubated in 25ul of PKA assay buffer (25mM Tris [pH 7.4], 1mM DTT, 10mM MgCl2,

1mM ATP, and 3mM cAMP) for 30 minutes at 37°C. The bound proteins were then subjected to immunoblot analysis.

2.5. RESULTS

2.5.1. AKAP18 associates with Inhibitor-1 in vivo

The formation of signaling complexes by AKAPs that include PKA and individual

PKA substrates contributes to the efficiency and fidelity of cAMP signal transduction

(42). As I-1 is an important PKA target, we considered I-1 might associate with anchored PKA in the heart. To test this idea, we isolated PKA-AKAP complexes from rat heart extract by affinity chromatography using Rp-cAMPs-agarose that binds the PKA regulatory subunits. This compound is a competitive inhibitor of PKA, so it does not stimulate activation of the kinase, thus leaving the holoenzyme intact. I-1 was consistently detected in the Rp-cAMPs pulldowns, but not under conditions which excess cAMP was added to the extracts to block specific binding to the beads (Figure 2.1A). As a control, both the PKA regulatory subunit type I, type II, as well as the PKA catalytic subunit were found specifically in cAMP-agarose pulldowns. To confirm that the PKA holoenzyme containing the catalytic subunit associated with I-1, we used GST-tagged, purified I-1 to pulldown PKA from rat heart extract. As shown in Figure 2.1B, kinase

Figure 2.1: Inhibitor-1 associates with PKA in heart cells.

A. Rat heart extracts were incubated with Rp-cAMPs-agarose in the presence or absence of excess cAMP (50uM). I-1 in whole extract (Ex, 2% of total) and the cAMP-associated fraction of I-1, PKA-RI, PKA-RII, and PKA-catalytic subunit were determined by immunoblotting. n=3. B. Rat heart extract was incubated with either GST-I-1 (3µg) or control GST protein, and the bound proteins were assayed for PKA activity using Kemptide as a substrate. n=3, *p<0.008, **P<0.01. C. PKA-RI, PKA-RII, and PKA- catalytic subunit present in GST control and GST-I-1 pulldowns were detected by immunoblot. Extract is 5% of the total protein used for each pulldown.

48 activity was found specifically in GST-I-1 complexes, and not with GST alone. Kinase activity was specific for PKA, as the addition of the PKA specific inhibitor PKI inhibited

PKA activity in the pulldowns. Furthermore, PKA catalytic subunit as well as PKA regulatory type I and type II subunits were found in GST-I-1 pulldowns from rat heart extract. Importantly, the percentage of PKA-RII was enriched 20 fold over RI, suggesting this AKAP is a primarily RII binding AKAP (Figure 2.1C). In summary, these data show that I-1 associates with a PKA complex in the heart.

Next we sought to identify which AKAP might mediate this interaction between

PKA and Inhibitor-1 in the heart. To do this, we isolated proteins in rat heart extracts that associated with I-1 using GST-tagged Inhibitor-1 protein, and then blotted for some highly characterized AKAPs found in the heart. As shown in Figure 2.2A, a high molecular weight AKAP18 isoform was readily detected in GST-I-1 complexes, in addition to what are potentially lower molecular weight AKAP18 isoforms that were less prevalent. As a control, I-1 did not associate with mAKAP, another muscle enriched

AKAP involved in the induction of cardiac hypertrophy (Supplementary Figure 2.1)

(144). The molecular weight of the intense, 45kDa band implied that the major isoform pulled down by GST-I-1 was either AKAP18δ or AKAP18γ. Importantly, this interaction was not observed using GST alone, demonstrating the specificity of the interaction. The association of AKAP18 with I-1 was confirmed using an I-1 antibody to immunoprecipitate the scaffold from HEK293 cells expressing GFP-tagged AKAP18γ and I-1 (Figure 2.2B). Additionally, complexes containing PKA were purified from the

HEK293 cell extracts using Rp-cAMPs agarose. I-1 co-purified with the agarose only when co-expressed with the anchoring protein, mimicking the results found with the

Figure 2.2: AKAP18 associates with I-1 A. Isolated rat heart extract was incubated with either GST-I-1 or control GST-protein.

AKAP18 was detected by western blot (left) and total protein by Ponceau stain (right) n=3. B. HEK293 cells expressing GFP-AKAP18γ in the presence and absence of I-1 were used for immunoprecipitations with an I-1 antibody, and AKAP18γ association was determined by immunoblot using a GFP antibody. C. HEK293 cells were transfected with Inhibitor-1 in the presence and absence of GFP-tagged AKAP18γ. Isolated extracts were incubated with cAMP agarose and immunoblotting demonstrated association of I-1. Immunoblots of whole cell extracts (2.5% of total) are shown as controls for both B and C. n=3 for all panels.

endogenously expressed I-1 and AKAP18 in the heart (Figure 2.2C). Together, these data suggest that AKAP18, I-1, and PKA form a multi-molecular signaling complex in the heart as well as when expressed in heterologous HEK293 cells.

2.5.2. I-1 Directly Binds a Long Isoform of AKAP18

AKAP18γ is one of the larger protein isoforms encoded by the alternatively spliced

AKAP18 gene (schematic shown in Figure 2.3A). The smaller AKAP18α is N-terminally myristoylated and palmitoylated, allowing for attachment to the plasma membrane where it facilitates PKA phosphorylation of the L-type calcium channel (129). In contrast, the large isoforms of AKAP18 are targeted to the sarcoplasmic reticulum where they from a signaling complex with phospholamban (134). In order to define which isoform of

AKAP18 is the primary target for I-1 binding, we performed pulldown assays using bacterially expressed GST-tagged I-1 with these two isoforms of AKAP18. GST-I-1 incubated with either bacterially expressed, S-tagged AKAP18α or AKAP18γ was subjected to S-tag pulldown assay. GST-I-1 was found in the S-protein agarose only in the presence of the larger isoform AKAP18γ (Figure 2.3B). Importantly, these data also demonstrate that I-1 and the scaffolding protein can bind directly in a purified system.

2.5.3. Expression of AKAP18 enhances I-1 phosphorylation on Threonine 35

The role of an AKAP is to anchor the kinase in close proximity to its target in order to enhance PKA phosphorylation (42). Therefore, AKAP18 should enhance PKA- mediated phosphorylation of I-1 by assembling a signaling complex containing both I-1 and PKA. To test whether PKA binding to AKAP18 is important for I-1 phosphorylation, we first confirmed that a ternary complex consisting of AKAP18, PKA

Figure 2.3: Direct binding of AKAP18g and Inhibitor-1.

A. Schematic diagram depicting the known AKAP18 splice variants. PLN, phospholamban; LTCC, L-type calcium channel. B. GST-I-1 fusion protein (3µg) was incubated with S-tagged AKAP18α or S-tagged AKAP18γ fusion protein (3µg), or S- protein agarose alone. S-tagged pulldown assays were performed and association of GST-I-1 was determined by western blot. Total protein was detected by Ponceau stain.

and I-1 could be isolated from transfected HEK293 cells. Cell extracts were prepared from HEK293 cells expressing AKAP18γ or a C-terminal truncated AKAP18γ mutant lacking the PKA binding site (AKAP18γ 1-268). GST-I-1 was incubated with the isolated cell extracts, and PKA activity assays were performed on the pulldowns (Figure

2.4A). While I-1 bound both full length AKAP18γ and AKAP18γ 1-268 (Figure 2.4B, top panel), kinase activity was only detected with full length AKAP18γ (Figure 2.4A).

Kinase activity was blocked by the addition of PKI, demonstrating the specificity of the reaction to measure PKA activity assay. In support of this finding, PKA-RII was not detected in GST-I-1 pulldowns from cells expressing AKAP18γ 1-268 (Figure 2.4B, middle panel, GST-protein stain shown in Supplementary Figure 2.2A). These data show that AKAP18γ can serve to couple PKA to I-1 in a heterologous system.

Next, we determined if the associated kinase could phosphorylate I-1 at Thr-35.

Consistent with the results of the PKA activity assays, cAMP-dependent phosphorylation of I-1 was greatly enhanced in GST-I-1 pulldowns containing full-length AKAP18γ compared to those containing the AKAP18γ PKA-binding mutant (Figure 2.4C). In order to show that AKAP18 enhances the PKA phosphorylation of I-1 in response to G-protein coupled receptor stimulation, the AKAP18γ proteins were co-expressed with I-1 in

HEK293 cells. In the absence of stimulation, no I-1 phosphorylation was detected using the phospho-specific antibody. However, β-adrenergic stimulation with 1uM isoproterenol for 5 minutes rapidly enhanced I-1 phosphorylation when AKAP18γ was co-expressed (Figure 2.4D) in a PKA dependent manner (Supplementary Figure 2.2B).

Importantly, this stimulated increase was significantly attenuated in cells transfected with

I-1 and AKAP18γ lacking the PKA binding domain (42±5.7% of wildtype, Figure 2.4E).

Figure 2.4: AKAP18 orchestrates I-1 phosphorylation by linking PKA with the Inhibitor.

A. GST-I-1 fusion protein (3ug) was incubated with HEK293 cell lysate extracted from prepared from cells expressing AKAP18γ, or AKAP18γ 1-268. PKA activity was determined in each GST pulldown in the presence and absence of 50nM PKI. B. Pulldowns were performed as described in A and the association of the AKAP, as well as PKA-RII with GST-I-1 was demonstrated by immunoblot. C. GST-I-1 fusion protein (3ug) was incubated with HEK293 cell lysate as prepared in A, and PKA associated with the GST pulldowns was activated by the addition of cAMP. I-1 phosphorylation was detected by western blot using an antibody that specifically recognized the inhibitor when phosphorylated at Thr-35 (upper panel). Total I-1 in each pulldown is shown in the middle panel while protein expression of the different AKAP18 constructs is shown in the bottom panel. D. HEK293 cells expressing I-1 and either full-length AKAP18γ or AKAP18γ 1-268 were stimulated with 1mM Isoproterenol for 5 minutes. Whole cell lysate were analyzed by western blot using antibodies against phospho-I-1 Thr-35 (upper panel), I-1 (middle panel) and GFP for the AKAP18 constructs (lower panel). E. Densitometric quantification of I-1 phosphorylation shown in C normalized to isoproterenol, AKAP18γ sample. n=3 for each panel, *p<0.05.

Together, these data show that targeting of PKA to I-1 via the association with AKAP18 facilitates I-1 phosphorylation by the kinase.

2.5.4. AKAP18 bind PP1 in vivo

Because AKAPs often serve as scaffolds for multi-molecular signaling complexes that include both kinase and phosphatases, we considered that the I-1 target PP1 might also be present in AKAP18γ complexes. To test this hypothesis, protein complexes were immunoprecipitated from rat heart extract using an antibody that recognizes the catalytic subunit of PP1 (Figure 2.5A). A high molecular weight isoform of AKAP18 co- precipitated with the PP1 antibody, but not with IgG control, implying that PP1 and

AKAP18γ and/or AKAP18δ associate in vivo. In reciprocal experiments, STREP-tagged

AKAP18γ was used to pulldown PP1 from rat heart extract (Figure 2.5B). PP1 did not associate with STREP-beads alone, demonstrating the specificity of this interaction.

Furthermore, an okadaic acid-sensitive phosphatase activity was found in AKAP18 immunoprecipitates isolated from rat heart extract (Figure 2.5C). Taken together, these data show that PP1 and AKAP18 form a complex in the rat heart.

Next, we confirmed the PP1/AKAP18γ complexes could be reconstituted in transfected HEK293 cells. In support of our findings that PP1 and AKAP18 formed a complex in rat heart cell extract, PP1 also associated with AKAP18γ in transfected

HEK293 cells (Figure 2.5D). Importantly, this association was lost when cells not expressing the anchoring protein were immunoprecipitated with the AKAP18 antibody, demonstrating the specificity of the interaction. Importantly, we were unable to immunoprecipitate the scaffold with a PP2A catalytic subunit antibody (Figure 2.5E),

Figure 2.5: AKAP18 is a PP1 anchoring protein

A. Protein complexes were immunoprecipitated from rat heart extract with either an antibody specific to PP1 catalytic subunit or control IgG and detected using an AKAP18 antibody. Extract represents 2.5% of total input. B. Rat heart extracts were incubated with either STREP-tagged (ST) AKAP18γ fusion protein or control STREP protein. PP1 in the pulldowns and total extract (2.5% of total input) were detected with an antibody to PP1 catalytic subunit. C. Protein complexes were immunoprecipitated from rat heart extract with either an antibody specific to AKAP18 or control IgG and assayed for phosphatase activity in the presence or absence of okadaic acid (100nM). Amount of PP1 in each group is demonstrated by western blot analysis of the immunoprecipitate. D. Protein complexes were immunoprecipitated from lysates prepared from control HEK293 cells or cells expressing GFP-tagged AKAP18γ using an AKAP18 antibody. Precipitated PP1 was detected using a catalytic subunit antibody. E. Protein complexes were immunoprecipitated from lysates prepared from HEK293 cells expressing GFP-tagged AKAP18γ using either a PP1 or PP2A catalytic subunit antibody (designated as PP: Protein Phosphatase). AKAP18 in the immunoprecipitates and AKAP18, PP1 and PP2A in the total extract (2.5% of total) were detected by immunoblot. n=3 for all panels.

even though phosphatase activity was found in both PP2A and PP1 immunoprecipitations, (Supplementary Figure 2.2), suggesting that under these conditions, AKAP18γ selectively binds PP1. Taken together, our data demonstrate a complex consisting of AKAP18 and PP1 that can be isolated from both rat heart extract as well as from transfected HEK293 cells.

2.5.5. AKAP18 is the scaffold for a PKA, PP1 and I-1 signaling complex

The data presented here suggest that I-1, PKA and PP1 all associate with AKAP18 in both transfected cells and heart extract. One important implication of these data is that the anchoring protein may sequester these enzymes into the same signaling complex, and that compartmentation is dependent on AKAP18. To test this hypothesis, we investigated the importance of AKAP18γ expression for formation of the complex. HEK293 cells were transfected with I-1 in the presence and absence of AKAP18γ, while endogenous

PP1 catalytic and PKA regulatory subunit were used for these experiments. Each protein was immunoprecipitated, and association of the other three proteins was determined by western blot. As shown in Figure 2.6, the expression of AKAP18γ was required for the association of the PKA-RII with I-1 and PP1, as well as association of PP1 to both

AKAP18γ and the PKA-RII. Furthermore, I-1 was only found in PKA -RII immunocomplexes upon expression of AKAP18γ. While, I-1 and PP1 can associate in the absence of AKAP18γ, co-expression of the anchoring protein significantly increased the association of the two proteins. Controls demonstrating the specificity of each

Figure 2.6: AKAP18 sequesters PKA, PP1 and I-1 to a specific signaling complex.

Protein complexes were immunoprecipitated from lysates prepared from HEK293 cells transfected with I-1 plus and minus GFP-AKAP18g using antibodies to AKAP18, I-1, PP1 and PKA-RII. Proteins in the extracts (bottom panels) and immunoprecipitates (top panels) were detected with the respective antibodies. Total protein (5% of input) is shown below. n=3 for each.

58 immunoprecipitation are shown in Supplementary Figure 2.4. The data presented in

Figure 6 effectively argues the importance of AKAP18 for linking PKA to I-1 and PP1.

2.5.6. AKAP18 orchestrates the PKA-mediated inhibition of PP1 activity

Phosphorylation of I-1 by PKA induces potent inhibition of PP1 activity (91). Our data suggest AKAP18-bound PKA facilitates this phosphorylation, and hence, should enhance the ability of I-1 to inhibit the phosphatase. To study this potential functional consequence of AKAP18 complex formation, we first isolated AKAP18 complexes from rat heart extract, and determined the effect of cAMP on the associated endogenous PP1 activity. Activation of PKA in the complex by the addition of 3uM cAMP inhibited PP1 activity (95±4.67%) present in the AKAP18 immunoprecipitates (Figure 2.7A). This inhibition was attributed to the activation of PKA, as inhibition was reversed by inclusion of the PKA inhibitor PKI (50nM). These data demonstrate that the PKA-stimulated inhibition of AKAP18-bound PP1 is functional in native AKAP18 complexes isolated from rat heart.

Showing the necessity of PKA anchoring to AKAP18 for PP1 inhibition extended these results. HEK293 cells were transfected with I-1 and either full-length AKAP18γ or the AKAP18γ PKA binding mutant. After stimulation with 1uM isoproterenol for 5 minutes, complexes were isolated using an antibody that recognizes the anchoring protein, and the associated phosphatase activity was measured. As shown in Figure 2.7B,

PKA anchoring to the complex was required for effective inhibition of the associated

Figure 2.7: PKA binding to AKAP18 is necessary for PKA-induced inhibition of PP1.

A. Protein complexes were immunoprecipitated from rat heart extract with either an antibody specific to AKAP18 or control IgG and assayed for phosphatase activity in the presence and absence of CPT-cAMP (3uM) or PKI (50nM) B. Phosphatase activity associated with protein complexes immunoprecipitated using an antibody specific for myc-tag from lysates prepared from HEK293 cells expressing I-1 and either myc-tagged full-length AKAP18γ or AKAP18γ 1-268 was assayed. Endogenous PP1 was used. Cells were stimulated with 1uM isoproterenol for 5 minutes as indicated. AKAP18 proteins in the total lysates were detected with a myc antibody. n=3 for each panel, *p<0.05. C. Cell lysates were prepared as in B. PP1 immunocomplexes were isolated and association of AKAP18 and I-1 was determined by western blot. n=3 for each.

60 phosphatase activity, as the cAMP-induced inhibition of PP1 was lost when AKAP18 could not bind PKA. This result is not due to a loss of complex formation, as similar amounts of AKAP18 and I-1 were found in PP1 immunocomplexes (Figure 2.7C). Taken together, these data show that the binding of I-1, PP1 and PKA by AKAP18 permits the efficient regulation of PP1 activity by cAMP signaling in cells.

2.6. DISCUSSION

In this report, we identify the first interaction between Inhibitor-1 and an A-kinase anchoring protein, AKAP18. Functional AKAP18 complexes containing I-1, the target

PP1, and PKA were isolated from both rat heart extract and transfected HEK293 cells.

The functional significance of these interactions was demonstrated in two ways. First, complex formation facilitated the cellular phosphorylation of I-1, as disruption of PKA association with AKAP18 significantly decreased phosphorylation of I-1 at Thr-35.

Second, PKA binding to AKAP18 was required for isoproterenol-induced inhibition of the associated PP1 activity. This information establishes AKAP18 as a regulator of both

I-1 and PP1 activity and advances our understanding of phosphatase regulation.

PKA phosphorylation of I-1 at Thr-35 and the resulting inhibition of PP1 activity acts as a feed-forward mechanism to enhance PKA phosphorylation of downstream effectors that would otherwise be dephosphorylated by PP1 (123, 140). PKA anchoring to AKAP18 is now shown to play an integral role in this process, since disruption of PKA anchoring to AKAP18 prevented cAMP-mediated inhibition of the PP1 activity

61 associated with the complex. As a result, in cells such as cardiac myocytes that express

AKAP18γ and AKAP18δ, these complexes should be highly sensitive to cAMP signals.

Previous work has demonstrated the existence of multi-protein complexes containing I-1 and PP1. An I-1/PP1 complex mediated by interactions with the Growth arrest and DNA damage protein 34 (GADD34) was found in active squirrel brains (146).

GADD34 was shown to enhance the I-1-mediated inhibition of PP1 in order to regulate the phosphorylation of the transcription factor eIF-2a. While GADD34 is found in the heart, the role of this complex in the regulation of PP1 in this tissue is unknown (147).

An important difference between GADD34 and AKAP18 in the regulation of PP1 activity is that the AKAP also anchors PKA to the complex, enhancing I-1 phosphorylation. Thus, the ability of AKAP18 to sequester I-1 and PKA ensures spatial- temporal control of I-1 phosphorylation.

Several AKAPs have been previously implicated to bind to and regulate phosphatase activity. For example, the direct association between AKAP220 and PP1 inhibits phosphatase activity (100). Additionally, mAKAP binds to PP2A and mediates a cAMP-induced increase of phosphatase activity through PKA phosphorylation of PP2A-

B56d subunit (144). In contrast to these other complexes, PP1 bound to AKAP18 is active, and PKA phosphorylation of the phosphatase’s regulator results in its inhibition.

To date, this is the first identification of an AKAP-mediated regulation of I-1 phosphorylation.

In the heart, I-1 activity is spatially restricted and selectively regulates PP1 activity at the sarcoplasmic reticulum, where PP1 opposes phospholamban

62 phosphorylation (123, 140). In contrast to the smaller isoforms of AKAP18 (α and β) that are targeted to the plasma membrane, the long isoforms (γ and δ) bind directly to phospholamban, facilitating PKA phosphorylation of phospholamban (134). Our results imply that AKAP18 may facilitate the localization of I-1 to the sarcoplasmic reticulum, further promoting phospholamban phosphorylation. Future investigations in our laboratory will address how the intracellular targeting of I-1 by the long isoforms of

AKAP18 affects the kinetics and sensitivity of phospholamban phosphorylation.

2.7. SUPPLEMENTARY FIGURES

Supplementary Figure 2.1:

Isolated rat heart extract was incubated with either GST-I-1 or control GST-protein. mAKAP was detected by western blot (left) and total protein by Ponceau stain (right) n=3.

Supplementary Figure 2.2:

A. GST-I-1 fusion protein (3ug) was incubated with HEK293 cell lysate prepared from cells expressing AKAP18γ, or AKAP18γ 1-268. Protein stain corresponds to western blots shown in Figure 4B. B. HEK293 cells expressing I-1 and full-length AKAP18γ were stimulated with 1uM Isoproterenol for 5 minutes after a one hour preincubation with H-89 (10uM). Whole cell lysate were analyzed by western blot using antibodies against phospho-I-1 Thr-35 (upper panel), I-1 (middle panel) and GFP for AKAP18 (lower panel).

Supplementary Figure 2.3:

Protein complexes were immunoprecipitated from lysates prepared from HEK-293 cells expressing GFP-tagged AKAP18γ using either a PP1 or PP2A catalytic subunit antibody. Phosphatase activity in the immunoprecipitates was performed as described in the methods section. Immunoprecipitations correspond to western blots shown in Figure 5E.

A. B. C. 3500 1 !&$%" !#$%"3000 0.8 !$$%"2500 0.6 2000($%" 0.4 1500'$%" &$%" 75 1000 IP:AKAP18 0.2 #$%" 500 IB:AKAP18 Activity PKA (pmol/min/IP) 0 PP1activity (CPM/IP) $%" !" #" AKAP18 -GFP !" #"+ AKAP18!-GFP + AKAP18!-GFP + ! 75 75 75 IB:AKAP18 IB:AKAP18 IB:AKAP18

50 IB:RII 50 IB:RII 50 IB:RII

37 IB:PP1 37 IB:PP1 37 IB:PP1

IB:I-1 IB:I-1 IB:I-1 25 25 25

Supplementary Figure 2.4:

Protein complexes were immunoprecipitated from lysates prepared from HEK293 cells transfected with I-1 plus and minus GFP-AKAP18γ using antibodies to AKAP18, PKARII and PP1. A). AKAP18 was immunoprecipitated using an antibody specific for the anchoring protein followed by western blot analysis using an antibody that recognized the GFP tag on AKAP18. Proteins in the extracts are shown in the bottom panels. B). The regulatory subunit of PKA was immunoprecipitated and PKA activity was determined on the immunoprecipitate, demonstrating equal amounts of PKA was immunoprecipitated under each condition. Total protein in the extracts is shown in the bottom panels. C) Phosphatase assay performed on PP1 immunoprecipitates in the presence and absence of AKAP18. Similar amounts of phosphatase were immunoprecipitated under each condition. Total protein in the extracts is shown in the bottom panel. Total protein is 5% of input. n=3 for each.

3. AKAP18γ OLIGOMERIZATION AUGMENTS PKA SIGNALING

Arpita Singh1†, Marc Rigatti1†, Andrew V. Le1, Ion I. Moraru2, Kimberly L. Dodge-

Kafka1†

1Pat and Jim Calhoun Center for Cardiology and Department of Cell Biology, University

of Connecticut Health Center, Farmington, CT 06030, USA

2The Richard D. Berlin Center for Cell Analysis & Modeling, University of Connecticut

Health Center, Farmington, CT 06030, USA

Manuscript for publication

Author contributions: AS and KDK conceived the project, designed and performed biochemical analyses in cellular systems. MR and IIM performed all the microscopy and computational modeling. AVL conducted the SPR experiments. AS, MR, IIM and KDK

helped write the manuscript.

3.1. ABSTRACT

A-kinase anchoring proteins constitute a family of scaffolding proteins that contribute to the spatial temporal regulation of phosphorylation events mediated by the cAMP-dependent protein kinase A (PKA). In particular, AKAP18 is a family of alternatively spliced proteins that participates in cardiac calcium dynamics. Here, we demonstrate that AKAP18γ can oligomerize, as demonstrated by immunoprecipitation from transfected HEK293 cells and by direct protein-protein association. This appears to be an isoform specific phenomenon, as AKAP18α did not form homo-oligomers with itself or hetero oligomers with AKAP18γ. Surface plasmon resonance analysis suggests two sites of interactions between AKAP18γ and itself, with each site displaying high binding affinity and KD values in the nanomolar range. This raised the possibility of the existence of larger order complexes. Indeed, photon counting histogram (PCH) and

Numbers and Brightness analysis of AKAP18γ -EGFP expressed in Vero cells showed that AKAP18γ -EGFP are found in vivo as monomers, dimers, trimers and larger oligomers, with AKAP18γ dimers and trimers constituting the predominant isoforms. The functional significance of AKAP18γ oligomerization is to enhance substrate phosphorylation, as predicted by computational modeling of AKAP18γ -bound PKA dynamics in response to cAMP stimulation in a specific cell compartment. In conclusion, our study reveals that oligomerization of AKAP18γ is an important step in mediating effective substrate phosphorylation in cellular microdomains.

3.2. INTRODUCTION

Phosphorylation of target proteins is one of the key underlying mechanisms utilized by the cell to induce changes in physiology and function. Although much work has previously defined multiple kinase-substrate interactions, how specificity of these phosphorylation events is achieved is still under investigation. Recent work has suggested that the subcellular localization of signaling enzymes to discrete locations in the cell confers spatial temporal control over phosphorylation events (4, 148). This is facilitated by a category of scaffolding proteins, which tether the kinase to the same signaling complex as its substrates, allowing for increased speed and amplification of phosphorylation events (148, 149). While the physiological significance of many kinase binding scaffolds has been investigated, how their complex molecular architecture influences cellular processes is presently unclear.

An important mechanism for the transduction of cellular events is oligomer formation of signaling enzymes (150, 151). Whether occurring as a hetero or homo oligomer, this process often begins at the level of the receptor as a first step in transducing a signal from outside the cell to the intracellular environment (152, 153).

However, oligomer formation is not limited to upstream signaling events. A few recent studies have demonstrated that scaffolding proteins can oligomerize, influencing their signaling functions (154). For example, the yeast scaffold Ste5, which initiates the mitogen-activated protein (MAP) kinase cascade in the Saccharomyces cerevisiae pheromone response pathway, dimerizes upon pheromone treatment, to enhance MAP kinase signaling in the cell (155, 156). Similary, oligomerization of β-arrestin 1 keeps the

70 scaffold cytoplasmic, inhibiting its nuclear actions (154). Importantly, these examples demonstrate that oligomerization greatly impacts the function of the scaffold.

For cAMP signaling, A-kinase anchoring proteins (AKAPs) focus the actions of the cAMP-dependent protein kinase (PKA) to specific targets, allowing for microdomains of kinase activity (43, 149). Additionally, it is becoming more evident that

AKAPs have a more profound effect on the regulation of substrate phosphorylation than just binding PKA. The concept of holistic substrate phosphorylation (i.e. conditions in which substrate phosphorylation is promoted followed by dephosphorylation) regulation by AKAPs is backed by the finding that AKAPs can bind to Phosphodiesteraes, Protein

Phosphatases, and Adenyl Cyclases as well, allowing for localized fluctuations in cAMP concentrations (101, 157-160).

Our recent investigations focused on defining AKAP18-orchestrated cAMP microdomains. AKAP18 is a family of alternatively spliced isoforms that are known to play a role in cardiac calcium dynamics (129, 134, 161). Here, we reveal that AKAP18γ oligomerizes, as demonstrated both in vitro and in a cellular context, resulting in an enhancement of substrate phosphorylation. This self interaction is limited to the large isoforms of AKAP18, correlating with previous findings that AKAP18α and AKAP18γ are differentially targeted in the cell and display unique functions (131). In agreement with this finding, surface plasmon resonance analysis finds that AKAP18γ specifically interacts with itself and displays a low nano-molar binding affinity. Using Fluorescence

Correlation Microscopy and Numbers and Brightness analysis, we demonstrate that

AKAP18γ forms stable, higher order complexes in cells, with concurrent existence of monomers, dimers and trimers present. In order to understand the functional significance

71 of oligomerization in the context of substrate phosphorylation, we developed a mathematical model using the Virtual Cell Software. Computational analysis suggests that formation of higher order structures enhances the magnitude of substrate phosphorylation associated with AKAP complexes. Collectively, our work demonstrates that oligomerization of AKAP18 impacts phosphorylation events orchestrated by the AKAP and suggests these findings may be applied to other AKAP complexes.

3.3. EXPERIMENTAL METHODS

3.3.1. Antibodies

For western blot analyses, the following antibodies were used: mouse monoclonal

GFP (Santa Cruz, 1:1000), goat polyclonal dsRED (Santa Cruz, 1:1000), rabbit His tag

(Millipore, 1:10,000), and mouse GST tag (Santa Cruz, 1:10,000).

3.3.2. Construct design

AKAP18γ-pET32a, AKAP18γ-EGFP-N1 and AKAP18α-EGFP-N1 were used as previously described (129, 131, 162). AKAP18γ 1-150-EGFP and AKAP18γ 150-end-

EGFP were constructed by PCR amplification and sub-cloning into the EcoRI-BamHI sites of pEGFP-N1 (Clontech). AKAP18γ-pMAL was constructed by PCR amplification and sub-cloning into the BamHI-EcoRI sites of pMal-c5E (NEB). Deletion constructs were made by site-directed mutagenesis, placing stop codons in the appropriate

72 sequences. AKAP18γ-mCherry was constructed by PCR amplification and sub-cloning into the EcoRI-BamHI sites of pmCherry-C1 (Clontech).

3.3.3. Recombinant Protein Purification

For purification of recombinant AKAP18γ-His/S-tag, bacteria were grown to OD595

0.8 then induced with IPTG (1mM final concentration) for 3hrs. After centrifugation, the bacteria were lysed in binding buffer (20mM HEPES (pH 7.4), 0.5M NaCl, 5mM

Imidazole along with 0.5% triton-X, 0.25mg/ml lysozyme, 1mM DTT, 1mM EDTA, and

Protease Inhibitor Cocktail) for 2hrs. Supernatant was collected by centrifugation and protein purified by Ni-agarose pulldown overnight (0.5mL of 50% slurry; Ni-NTA solution, Qiagen). Ni-agarose was washed 3X in binding buffer and protein was eluted with binding buffer + 800mM Imidizole. Protein was concentrated with Amicon Ultra centrifugal filters (Millipore). Elution buffer was slowly exchanged with PBS while concentrating the protein in the centrifugal tubes. Protein quality and concentration was confirmed by running the protein on SDS-PAGE gel with BSA standards and staining with coommasie brilliant blue. The gel was scanned post staining to calculate the exact protein concentration using Image J software (NIH).

A similar protocol was followed for AKAP18γ-maltose binding protein (MBP) purification. The maltose binding protein tagged protein was lysed in buffer containing

Protease Inhibitor Cocktail, 20mM Tris-HCl (pH 7.4), 200mM NaCl, 1mM each of

EDTA, Na Azide, and DTT. Separated cell lysate was passed through Amylose resin packed column, followed by 5 washes with lysis buffer. The protein was eluted using

73 lysis buffer containing 10mM maltose. The protein was concentrated and analyzed as described above.

3.3.4. Cell transfection

For analysis of AKAP18γ binding to itself, HEK-293 cells at 50% confluency were transfected using calcium phosphate precipitation with pEGFP-AKAP18γ and either pmCherry-AKAP18γ or pmCherry control (10ug) overnight under 5% CO2 at 37°C.

Cells were harvested 18hrs later for cell extract.

Vero cells were used for all confocal studies. Briefly, 35mm dishes of Vero cells at ~70% confluency were transfected with 2ug plasmid DNA (pEGFP or AKAP18γ-

EGFP) in Opti-MEM (Invitrogen) serum free medium using 6ul PLUS reagent and 6ul lipofectamine. Cells were used for photon counting histogram analysis or number and brightness analysis 5-10hrs post transfection to achieve low yet detectable levels of the expressed protein construct.

3.3.5. Pulldown Assays

In-vitro: Purified His/S tagged AKAP18γ (5ug) was incubated with either

AKAP18γ-MBP) (5ug) or expressed MBP, and rocked overnight at 4°C. The next day, the Amylose resin was added for 4hrs to isolate the complex. The Amylose beads were washed 4X in lysis buffer (as described in maltose tagged protein purification above) for

10-15mins while rocking. The complex was then subjected to SDS-PAGE purification followed by western blot analysis.

From cell supernatant: Lysates prepared from AKAP18γ-EGFP transfected HEK

293 cells were incubated with either 5ug purified recombinant AKAP18α-His/S-tag pre- loaded on S-Agarose beads (Novagen), or with purified 5ug AKAP18γ-MBP pre-loaded on Amylose Agarose beads (Novagen). For control groups, either unrelated His/S tag protein or isolated MBP was used. The pulldowns were rocked overnight at 4°C, followed by 4 washes with their respective buffers for 10-15min while rocking. The isolated complex was then subjected to SDS-PAGE purification followed by western blot analysis.

3.3.6. SPR Analysis

SPR analysis was performed using a BIAcore T100. Either purified recombinant

AKAP18γ or AKAP18α were covalently immobilized using NHS (N- hydroxysuccinamide) and EDC [1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide]

(Biacore amine coupling kit) to the surface of a sensor chip (BIAcore type CM5). The amount of ligand bound (in resonance units RUs) was 150RU. Protein analytes

(AKAP18γ or AKAP18α) were diluted at increasing concentrations (12.5-200nM) in

HBS buffer (10mM Hepes (pH 7.4), 150mM NaCl and 0.005% Surfactant P20) and were injected over the sensor surface at a flow rate of 30ul/min for 300s. Post injection phase, dissociation was monitored in HBS buffer for 300s at the same flow rate. The surface was regenerated between injections using 10mM NaCl at a flow rate of 50ul/min for 30s.

Obtained sensograms were all processed by BIAcore T100 evaluation software.

3.3.7. Photon Counting Histogram Analysis

Photon counting of EGFP in solution and AKAP18γ-EGFP in live Vero cells was performed with a Zeiss LSM 510 ConFocor3 confocal microscope mounted on an

AxioVert 200M. Excitation was achieved at 488nm with a 30mW Argon laser at 0.5% power. Fluorescence emission was collected with a 40X water immersion objective

(Zeiss C-Apochromat 1.2W Korr UV-Vis-IR). Detection was performed with an avalanche photodiode (Zeiss). Photon-counting histograms (PCH) were generated using

LSM-FCS software using a bin time of 50us or 200us for EGFP in solution and

AKAP18γ-EGFP respectively. Both bin times are shorter than the translational diffusion time of their respective molecules. Fluorescence traces were integrated for 10s for each

PCH curve. Measurements were made 10 times per location at 5 different locations within the cytoplasm. PCH curves were fit with a nonlinear curve fitting routine using a one or two component model. EGFP in solution was fit with a single component model.

For AKAP18γ-EGFP brightness value of the first component was fixed to the measured brightness value of EGFP in solution. A single-photon three-dimensional Gaussian model was assumed for the excitation volume.

3.3.8. Confocal Imaging Number and Brightness Analysis

Confocal imaging of AKAP18γ-EGFP was performed on a Zeiss LSM780 microscope mounted on an inverted Axio Observer Z1. Excitation was achieved at

488nm with an Argon laser at 1% power. Emission was collected through a 40X water objective (Zeiss C-Apochromat 1.2W Korr UV-Vis-IR) and detected by a 32-channel

QUASAR GaAsp detector in photon counting mode. Images were collected at 512x512

76 pixels with a 3.15µs/pixel scan speed and 1.94s exposure time per frame. Images were saved as 8-bit and imported into SimFCS software (Laboratory for Fluorescence

Dynamics). Detailed description of the method and analysis has been previously described (163). Briefly, for each image series 100 frames were analyzed. Vero cells expressing monomeric EGFP were imaged under the same conditions as AKAP18γ-

EGFP. The brightness of EGFP was used as the brightness of a monomer (1.29), which allowed for the selection of oligomeric size based upon the average brightness.

3.3.9. Computational Modeling

A compartmental deterministic model of scaffold oligomerization was implemented using the Virtual Cell modeling software. The model is composed of 17 or 20 species depending on whether one or multiple oligomeric states of the scaffold species are present in the simulation, three compartments (plasma membrane, cytosol, and endoplasmic reticular membrane), and 13 reactions. The reaction network can be considered as being composed of four general modules: 1) production and degradation of cAMP, 2) the dynamic interactions of regulatory subunit, catalytic subunit and cAMP in the cytosol, 3) the dynamics of these same proteins and cAMP scaffolded on the ER membrane by AKAP, and 4) phosphorylation/dephosphorylation of target protein X on the ER membrane. For each oligomeric state represented in the different versions of the model, a simulation is run to equilibrium to determine the concentration of each species.

The equilibrium states are then used to initialize a simulation in which the amount of AC present on the plasma membrane is increased from 20/µm2 to 80/µm2 for a period of

500s, simulating transient activation of the β-AR. The resulting increase in cAMP leads to activation of scaffolded PKA and phosphorylation of the target protein X. The ability

77 of RIIα to self-regulate via autophosphorylation has been shown previously (164). We hypothesized that if a single catalytic subunit were to be activated within an oligomerized complex of AKAP and PKA, this may subsequently lead to activation of the remaining population of PKA within the complex. This hypothesis was implemented via the following equation (eq. 1), which adjusts for the probability of all catalytic subunits within the oligomeric complex (n= oligomer size) becoming active following activation of a single catalytic subunit in the complex:

(1) (Total AKAP bound PKA)(1-(Inactive PKA/ Total PKA) n) * n

This value representing the amount of catalytic subunits available to phosphorylate its target, X, is substituted for [E] in the Henri-Michaelis Menten irreversible reaction step for phosphorylation of X (eq 2).

(2) V= ([E] * kcat * [X]) / (Km + [X])

A separate model was created to investigate the phosphorylation dynamic of protein X in the presence of all forms of the oligomer as demonstrated via FCS/N&B analysis.

Concentrations of each oligomeric species were adjusted by tuning the forward and reverse rates of formation of each species.

3.4. RESULTS

3.4.1. Direct cellular association of AKAP18γ with itself.

In order to understand the molecular architecture of the AKAP18γ signaling complex and determine if the AKAP can be the basis for higher order structures, we first looked to see if purified, recombinant AKAP18γ could isolate an AKAP18γ complex from cells.

Lysate prepared from HEK293 cells transfected with AKAP18γ-EGFP was incubated with amylose beads charged with either AKAP18γ-maltose binding protein (MBP) or

MBP alone. As shown in Figure 3.1A, AKAP18γ-EGFP specifically associates with

AKAP18γ-MBP, suggesting the AKAP18γ signaling complex may contain multiple

AKAP18 scaffolds. To verify the existence of such complexes in cells, we co-expressed

AKAP18γ-EGFP with either AKAP18γ-mCherry or mCherry alone in HEK293 cells.

Immunoprecipitates from these cells isolated with a mCherry antibody contain

AKAP18γ-EGFP (Figure 3.1B), while AKAP18γ-EGFP does not co-precipitate with mCherry control. This coprecipitation could be due to multiple binding sites on other common binding partners from the AKAP complex. To determine if AKAP18γ can directly associate with itself, we conducted a pulldown assay using amylase beads charged with bacterially expressed AKAP18γ-MBP incubated with bacterially expressed

His/S-tagged AKAP18γ. Importantly, we detect a direct association between His/S- tagged AKAP18γ and AKAP18γ-MBP, but not MBP alone (Figure 3.1C). Based on these experiments, we conclude AKAP18γ can stably associate with itself.

Figure 3.1: AKAP18γ interacts with itself in cells and directly.

A) Recombinant purified full-length maltose binding protein (MBP) tagged AKAP18γ precharged on amylose agarose resin was used to conduct precipitation of AKAP18γ- EGFP from HEK cell lysates. HEK cells were transfected for over 24hours with AKAP18γ-EGFP before harvesting the cells for pulldown experiment. Anti-EGFP antibody was used for detecting protein interactions using western blot analysis. Purified MBP alone was used as control. B) AKAP18γ-EGFP and AKAP18γ-mCherry were co- transfected in HEK cells for over 24 hours before harvesting the cells to prepare a cell lysate. Cell lysate was subjected to an immuno-precipitation experiment using antibody against the mCherry tag. Anti EGFP antibody was used to identify protein-protein interaction using western blot analysis. C) Recombinant purified full length maltose binding protein (MBP) tagged AKAP18γ precharged on amylose agarose resin was used to conduct pulldown of recombinant purified full length His/S tagged AKAP18γ. Anti His antibody was used to detect the interaction using western blot analysis.

3.4.2. Isoform specificity and binding affinity of AKAP18γ oligomers.

The AKAP18 gene encodes at least 4 alternatively spliced isoforms: α, β, γ and δ (131).

The smaller isoforms, AKAP18α & AKAP18β are lipid modified in their N-terminal domain, directing their location to the plasma membrane (129). However, the larger isoforms, AKAP18γ & AKAP18δ require additional protein-protein interactions for their intracellular targeting (131). We investigated whether the shorter isoforms could also associate with themselves or other isoforms. We first performed pulldown experiments using amylose beads charged with AKAP18γ-MBP on cell lysate isolated from cells transfected with AKAP18α-EGFP. As shown in Figure 3.2A, we did not detect the presence of AKAP18α with either AKAP18γ-MBP or control MBP, suggesting there is no cross isoform interaction. To further confirm this observation, we conducted an inverse experiment. Lysate collected from HEK293 cells transfected with AKAP18γ-

EGFP was incubated with S-beads charged with S-tagged AKAP18α or S-beads alone.

As expected, we did not detect any interaction between the two isoforms (Figure 3.2B).

To further our understanding of this interaction, we explored the detailed binding kinetics of the AKAP18γ interaction by surface plasmon resonance (SPR). AKAP18γ was chemically conjugated on a CM5 sensor chip. Increasing concentrations of either

AKAP18γ (black) or AKAP18α (red) (12.5nM to 400nM) were then flowed over the sensor, and the binding dissociation constant was calculated. We detect a high affinity interaction between AKAP18γ and itself that is best fit to a heterogeneous analyte curve, suggesting multiple sites of interaction displaying binding affinities of 6nM and 0.13nM

(Figure 3.2C). However, no significant binding between AKAP18γ and AKAP18α is

Figure 3.2: AKAP18γ interacts with itself with a high affinity in an isoform specific manner.

A) Recombinant purified full-length maltose binding protein (MBP) tagged AKAP18γ precharged on amylose agarose resin was used to conduct precipitation of AKAP18α- EGFP from HEK cell lysates. HEK cells were transfected for over 24hours with AKAP18α-EGFP before harvesting the cells for pulldown experiment. Anti-EGFP antibody was used for detecting protein interactions using western blot analysis. B) Recombinant purified full length His/S tagged AKAP18α precharged on Ni-agarose resin was used to conduct precipitation of AKAP18γ-EGFP from HEK cell lysates. HEK cells were transfected for over 24hours with AKAP18γ-EGFP before harvesting the cells for pulldown experiment. Anti-EGFP antibody was used for detecting protein interactions using western blot analysis. C) Surface Plasmon Resonance (SPR) experiment showing robust interaction between AKAP18γ ligand and AKAP18γ analyte (blue). Significant interaction between AKAP18α ligand and AKAP18γ analyte was not detected (red). D) Surface Plasmon Resonance (SPR) experiment showing lack of binding between AKAP18γ ligand and AKAP18α analyte (blue). AKAP18α ligand and AKAP18α analyte did not show interaction as well (red).

82 detected. This was confirmed in reciprocal experiments, where we chemically conjugated AKAP18α to the sensor and no binding between AKAP18α and itself was detected (Figure 3.2D). These results support the hypothesis that AKAP18γ oligomerization is limited to AKAP18γ.

3.4.3. Mapping the sites in AKAP18γ responsible for oligomerization.

Following the detection of two sites of interaction between AKAP18γ molecules by our SPR analysis, we were interested in defining the sites that delineate the association.

To address this question, we synthesized deletion constructs of the AKAP18γ-MBP.

Amylose beads charged with these AKAP18γ deletion proteins were then incubated with

HEK293 cell lysate expressing either the first 150 amino acids of AKAP18γ (AKAP18γ

1-150-EGFP) or the last 182 amino acids (AKAP18γ 150-end-EGFP) (shown in Figure

3.3A). As shown in Figure 3.3B and 3.3C, not only does full length AKAP18γ-MBP interact with the two halves of AKAP18γ, but both the EGFP-tagged halves associate with all the AKAP18γ-MBP deletion constructs. Therefore, using this method, we were not able to delineate the specific amino acid binding domains on AKAP18γ that are responsible for this interaction.

Figure 3.3: Mapping for delineating specific AKAP18γ interaction domain(s).

A) Schematic representation of EGFP tagged AKAP18γ protein constructs that were separately expressed in HEK cells. B) Western blot showing interaction between different recombinant purified MBP tagged AKAP18γ deletion mutants and AKAP18γ 1-150 EGFP. AKAP18γ 1-150 EGFP was expressed in HEK cells for over 24 hours and cells were harvested to prepare cell lysates. Cell lysates were incubated with recombinant purified MBP tagged AKAP18γ deletion proteins precharged on amylose agarose resin. Antibody raised against EGFP was used to detect the interaction. C) Western blot showing interaction between different recombinant purified MBP tagged AKAP18γ deletion mutants and AKAP18γ 150-end EGFP. AKAP18γ 150-end EGFP was expressed in HEK cells for over 24 hours and cells were harvested to prepare cell lysates. Cell lysates were incubated with recombinant purified MBP tagged AKAP18γ deletion proteins precharged on amylose agarose resin. Antibody raised against EGFP was used to detect the interaction.

3.4.4. Mulitple oligomeric states of AKAP18γ can be detected in cells.

The existence of two binding sites for AKAP18γ to associate with itself suggests the potential to form not only dimers, but higher order oligomeric structures, independent of the participation of other AKAP signaling complex components. This possibility was investigated via photon counting histogram (PCH) analysis of AKAP18γ-EGFP expressed in live cells. Cells expressing low yet detectable amounts of AKAP18γ-EGFP were selected for analysis due to the dependence of signal variance on the inverse of the number of particles present in the confocal volume. For each cell chosen, five regions were selected from within the cytoplasm for analysis. The average PCH for all 50 measurements made within the cell is shown (Figure 4A). Fitting of the PCH with a dual- component model in which the first component has a fixed brightness value representing the brightness of monomeric EGFP (data not shown) reveals the presence of AKAP18γ-

EGFP oligomers ranging in size from monomer to tetramer (Figure 4B). The average concentration of each oligomeric state, as measured via PCH, indicated that the predominant species is a dimer (45nM). The monomer and trimer are both present at slightly more than half the concentration of the dimer (about 25nM). Although there is evidence of the formation of a tetramer, brightness values representing oligomers of this size were seen in the fit of only 2 measurements.

Figure 3.4: AKAP18γ-EGFP transfected Vero cells analyzed via PCH.

(A) Two-component fit of the average of 50 measurements in a single representative cell. (B) Average concentration of each oligomer. Component 1 is necessarily included in the fitting routine and functions to define the brightness of the monomeric state thus, fits of all measurements include the monomeric state. Concentrations of dimers, trimers and tetramers represent the average concentration of the second component detected by the fitting routine.

Further confirmation of the ability of AKAP18γ-EGFP to form oligomers in cells was obtained via number and brightness (N&B) analysis. This technique is similar to

PCH analysis in its ability to detect oligomerization via analysis of signal fluctuation, yet it offers the advantage of spatial resolution of the oligomeric states. A series of 100 confocal images of AKAP18γ-EGFP transfected live Vero cells (Figure 3.5A) was gathered. Number and brightness analysis of this series of images yields a map of both the intensity and brightness of all pixels within the region of interest selected for analysis

(Figure 3.5B& C). These maps are useful for determining regions of the image in which protein oligomerization or aggregation is occurring. From the brightness map it appears that a mix of different oligomers is present throughout the cell, and the distribution of each oligomeric state appears to show no significant difference in prevalence at different subcellular locations. Given the brightness value of EGFP measured under the same imaging conditions, we selected from the plot of brightness vs. intensity (Figure 3.5D) pixels with brightness values that are multiples the EGFP brightness and represent each oligomeric state (163). The monomer (blue), dimer (green) and higher order oligomers

(red) selected from the plot of brightness vs. intensity are displayed in the selection map

(Figure 3.5E). Taken together, these data demonstrate the formation of higher ordered

AKAP18γ complexes in cells, mostly in the dimer and trimer forms.

3.4.5. Oligomerization of AKAP18γ functions to increase target phosphorylation.

Preliminary studies were unsuccessful in using deletion constructs to map the binding sites and achieve selective disruption of the oligomeric complexes. Thus, we

Figure 3.5: Brightness analysis of AKAP18γ-EGFP in Vero cells.

(A) Confocal image of Vero cell transfected with AKAP18γ-EGFP. (B) Intensity map of the same image shows a region of higher intensity in the perinuclear region. (C) Brightness map of this image reveals an even distribution of oligomers of each size. (D) Brightness versus intensity plot displaying monomers, dimers, trimers, and some higher order oligomers. (E) Image of the cell displaying pixels with brightness corresponding to monomers (blue), dimers (green), trimers and higher order oligomers (red).

88 decided to test the possible functional significance of AKAP18γ oligomerization in silico.

We created a compartmental deterministic model that simulates the dynamics of the PKA

Regulatory (RIIα) and Catalytic subunits (Cα) in response to cAMP, both in the cytosol and on the Endoplasmic Reticulum (ER) membrane. We chose to model the dynamics of

AKAP18γ at the ER due to recent work demonstrating localization of the scaffold at the

ER in cardiac myocytes (134, 135). Included in the model is autophosphorylation of RIIα as this is documented as a key component for activation of Cα when complexed with RII

(164-168). In our model, it is assumed that the probability of activation of all Cα subunits within an oligomer increases as the total number of activated Cα subunits increases

(Equation 1, methods). Active Cα subunits were modeled to participate in the phosphorylation of a generic target protein X (X pX) existing on the ER membrane

(Equation 2, Methods). The time course of formation of pX is shown for separate simulations in which four different oligomer sizes were postulated (Figure 3.6A). Upon addition of adenylyl cyclase, the amount of pX increases in all cases. Removal of this stimulus leads to dephosphorylation of pX and return to baseline. The rate of formation of pX increases as the size of the oligomer increases, and the magnitude of change between baseline and maximum levels of pX appears to peak with the trimeric state

(Figure 3.6B). This analysis demonstrates the possible impact of AKAP oligomerization on the rate of PKA substrate phosphorylation.

Figure 3.6: Compartmental-deterministic simulation of the effect of AKAP oligomerization on target phosphorylation

(A) Time course of target protein X phosphorylation for each oligomeric state. An increase in cAMP concentration simulating β-AR stimulation is achieved by increasing the amount of active AC present on the plasma membrane from 20/um2 to 80/um2. This stimulation starts at 200s and lasts for 500s. (B) Change in [pX] from baseline to maximal phosphorylation relative to phosphorylation by the monomeric state.

3.5. DISCUSSION

Significant work has demonstrated the impact of scaffolding on the organization of signaling proteins into discrete focal points of subcellular enzyme activity (148).

However, information on how the molecular architecture of scaffolds influences phosphorylation events is currently lacking. Here, through the use of a combination of protein biochemistry, state-of-the-art cellular imaging and computational modeling, we deciphered that the scaffolding protein AKAP18γ oligomerizes in vitro and in cells.

Several species of the AKAP can be detected in cells, including monomers, dimers, and high order oligomers. Importantly, the cellular consequence of these structures is increased substrate phosphorylation. These findings represent a previously unappreciated mechanism involved in the configuration of AKAP complexes and the regulation of protein phosphorylation.

Oligomer formation is a common process to many signal transduction networks.

In fact, a search of the Brenda enzyme database (http://www.brenda.uni-koeln.de/) suggests that of the total 452 human enzymes for which the subunit composition is known, over 300 exist in mulitmeric states (150). Recent evidence suggests intracellular scaffolding proteins also oligomerize (154-156, 169). Several recent reports have demonstrated homo-oligomerization of AKAP5, AKAP12 and AKAP-Lbc (47, 49,

52). Dimer formation of these AKAPs is mediated by direct protein-protein interactions. A leucine zipper motif in the N-terminus of AKAP-Lbc dictates the interaction while multiple domains of AKAP5 are required for dimerization (49, 52).

AKAP18γ oligomerization is also via a direct interaction, but our mapping studies to define the binding motifs needed for oligomerization are inconclusive and suggest

91 multiple points of interaction. Interestingly, a crystal structure of the common central domain contained in the long AKAP18 isoforms (amino acids 76-292 of AKAP18γ) did not indicate oligomerizaiton of the AKAP (48). This structure lacks the N-terminus of

AKAP18γ, which is predicted to exist in an unordered arrangement, according to

FOLDindex©. Several AKAPs have a natively disorganized, unfolded configuration that is hypothesized to aid in the docking of a multitude of binding partners (47-49).

However, this unstructured flexibility also significantly hinders conventional structural analysis. Hence, this N-terminal domain of AKAP18γ may significantly contribute to its oligomerization. Other possible mechanisms include post-translational modification driven interactions, and coiled-coil interactions (170-172). Analysis of the AKAP18γ sequence using the PredictProtein web site as well as both MARCOIL and LOGICOIL web resources support these possibilities for AKAP18γ (Supplemental Table 3.1).

Testing these potential mechanisms for oligomerization will help in delineating the means of AKAP18γ complex formation.

Signaling proteins utilize oligomerization to enhance their known cellular functions. One of the better-studied examples is the receptor tyrosine kinase (RTK).

Dimerization of RTKs results in activation and autophosphorylation of tyrosine residues on the intracellular domain of the receptor (152, 173). This step initiates the recruitment of other signaling proteins and propagation of signals from the receptor to intracellular targets. Though often thought to function in a monomeric state, a number of G-protein coupled receptors (GPCRs) have been shown to form both homo- and hetero-dimers, often resulting in enhanced activity (153, 174).

The consequences of AKAP oligomerization are still unclear, but may impact multiple processes. Mass spectrometry analysis of the AKAP5 complex suggests that upon dimerization, AKAP5 associates with four calcineurin phosphatase heterodimers, and two calmodulin molecules (49). By increasing the concentration of the phosphatase in a localized area, it is suggested that AKAP5 can generate specific pools of second messenger-mediated events at the plasma membrane. How this change in stoichiometry affects substrate dephosphorylation or kinetics of enzyme activation has not been investigated. However, another study of the native AKAP5 complex in hippocampal neurons suggests this increase in stoichiometry is an in vitro experiment artifact, as no increase in stoichiometry was detected in vivo (175). Other processes oligomerization may affect include regulating the intrinsic enzymatic activity of an AKAP. AKAP-Lbc is a guanine nucleotide exchange factor (GEF) for RhoA (176). Disruption of AKAP-Lbc dimerization dramatically increased its GEF activity, suggesting oligomer formation acts to limit the basal activity of the AKAP (52). However, AKAP18γ does not display any intrinsic enzymatic activity, suggesting this mechanism would not be applicable.

To explore the possible functional significance of AKAP18γ oligomerization, we created a computational model of PKA scaffolding. Significant work has demonstrated that release of active C subunit from the PKA holoenzyme requires not only cAMP binding to the RII subunit, but also autophosphorylation of RII (164-168). This leads us to hypothesize that in the context of an AKAP18γ oligomer, release of a single Cα subunit can potentiate the release of additional Cα via phosphorylation of RIIα present in the oligomer. In the monomeric state, a single Cα can stimulate release of only one extra

Cα subunit, but, in an oligomeric state such as a tetramer, release of up to seven

93 additional Cα subunits is possible. This feed-forward loop ultimately increases substrate phosphorylation due to the increased probability of phosphorylation resulting from increased Cα release (Figure 3.7). This is simulated by defining activation of PKA in the

AKAP18α complex as a probability function in which the chance of activating all Cα subunits increases as the number of active Cα subunits within the complex increases (Eq.

1). The model demonstrates that oligomerization of AKAP18γ does indeed enhance target phosphorylation, increasing the magnitude of change in phosphorylation of substrate X (pX) with oligomer size up to a trimer. This pattern correlates well with our

PCH and N&B data, suggesting that the predominant AKAP18γ species is a dimer, with less monomer and trimer, and relatively no tetramer.

In conclusion, we have discovered an additional component of the AKAP18γ signaling complex and demonstrated oligomerizaiton of the AKAP impacts phosphorylation of PKA substrates. As several PKA targets of AKAP18γ have been identified, including PDE4D, the PP1 inhibitor I-1 and the cardiac calcium regulatory protein phospholamban, AKAP18γ oligomerization will likely effect phosphorylation of these endogenous substrates.

3.6. SUPPLEMENTARY FIGURES AND TABLE

Supplementary Figure 3.1: Binding of R2 and AKAP18 isoforms

R2 does not show significant difference in binding to AKAP18α (non-ologimerizing AKAP18) and AKAP18γ (oligomerizing AKAP18). 5uM purified RII 1-47 GST was incubated with 1uM each of either His/S-AKAP18α or His/S-AKAP18γ to conduct a pulldown experiment. The proteins were incubated together for a few hours in solution followed by addition of 20uL S-agarose beads and the pulldown was carried on for a few more hours. The precipitates were spun-down and thoroughly washed 3X using His protein purification buffer having protease inhibitors, followed by a high-salt wash (600mM NaCl in His purification buffer). Lamelli buffer was added to the precipitates and SDS-PAGE was performed followed by coomassie staining.

Supplementary Table 3.1:

Table showing the list of possible post-translational modifications AKAP18γ may undergo as a possible mechanism for oligomerization. Data extracted from predictprotein.org.

4. CONCLUSIONS AND FUTURE DIRECTIONS

The focal point of my thesis has been establishing AKAP18γ as a novel AKAP that regulates I-1 phosphorylation by localizing it to the AKAP18γ complex and oligomerizes to enhance substrate phosphorylation. The novelty of my thesis work is the discovery that AKAP18γ is involved in direct interaction with I-1 and importantly, enhances its phosphorylation by PKA. AKAP18γ can concomitantly interact with PKA,

PP1 and I-1 making an AKAP18γ complex and I found that PP1 can be inhibited by this complex in a PKA activation dependent manner. Moreover, the thesis work presented here provides a direct experimental evidence of enhancement in I-1 phosphorylation when bound to AKAP18γ. I also show that AKAP18γ oligomerizes in vitro and in cells and computational modeling shows that based on this oligomerization more PKA catalytic subunits can be released from the complex which improves phosphorylation of generic substrate in a specific cellular compartment.

Another unique facet of AKAP18γ complex is that PP1 does not directly interact with this AKAP. Immuno-precipitation and direct binding experiments confirm that

AKAP18γ is able to pulldown PP1 from rat heart extract but purified PP1 does not directly interact with purified AKAP18γ. Even though, it has been previously established that PP1 and I-1 can directly interact (87); in my experiments, purified I-1 could not mediate AKAP18γ and PP1 interaction suggesting that there is another PP1 binding partner that localizes PP1 to this complex. A glycogen binding scaffolding protein GM has been reported to be present in the SR membrane in order to localize PP1 required for

97 de-phosphorylation of PLN. Moreover, the intra-membrane domain of PLN directly interacts with intra-membrane domain of GM (177). Based on this evidence, we had proposed that GM might be the connecting link between AKAP18γ and PP1 interaction.

Using SPR to test protein-protein interaction, I found that GM 1-240 GST does not interact with AKAP18γ. On the other hand, SPR experiments also confirmed that the same domain of GM, GM 1-240 GST could interact with PLN with high affinity [Figure

4.1]. Based on these results, other domains of the GM protein should be tested for interaction with AKAP18γ. Also, mass spectrometric analysis of AKAP18γ immunoprecipitates from human heart homogenate or mouse heart homogenate should be conducted to identify the linking protein that mediates AKAP18γ and PP1 interaction.

The key finding that AKAP18γ can bind to I-1 opens up a new field of AKAP mediated regulation of this PP1 inhibitor which was previously unknown. PP1 is a highly regulated protein and it has been described in Chapter 1 how multiple subsets of

PP1 binding proteins participate and contribute to PP1 regulation. The discovery that an

AKAP can bind the PP1 endogenous inhibitor adds to our understanding of yet another layer of regulation. In an effort towards understanding the molecular basis of this interaction, I conducted mapping experiments to delineate a binding domain of interaction between AKAP18γ and I-1. For this, I synthesized deletion constructs of GFP tagged AKAP18γ that can be transfected into mammalian cell for expression followed by pulldown using purified recombinant I-1 GST. These experiments showed that AKAP18γ bound I-1 GST at multiple sites as all deletion constructs of AKAP18γ-GFP were able to bind to I-1 GST. I also attempted mapping the AKAP18γ binding site by expressing recombinant His tagged deletion constructs of AKAP18γ that can be purified from

Figure 4.1: GM 1-240 GST interacts with PLN with high affinity.

Kinetics data from SPR experiement showing association and disassociation curves for GM 1-240 GST (analyte) binding to PLN (ligand). Increasing concentrations of GM1-240 GST was used for kinetics analysis (25-400nM). The SPR program determined a KD of 9.71 E -09.

99 transformed E. coli. The purified deletion constructs of AKAP18γ when used for a pulldown experiment with I-1 GST yielded an inconclusive answer like the previous result

[data not shown]. A plausible reason for the failure of this mapping approach could be the highly disordered structure of C-and N-terminus of AKAP18γ and native conformation of I-1 protein. Following these attempts at mapping the AKAP18γ domain required for I-1 interaction, an SPR study was conducted to understand the molecular determinants of this interaction. SPR kinetics data revealed that the two proteins interacted with a high affinity and kinetics curves generated by this experiment best fit a two site interaction model of protein-protein interaction [Figure 4.2] suggesting there are two AKAP18γ domains that interact with I-1. In an inverse experiment, biotinylated I-1 peptides (1-40, 41-80, 81-120 and 121-end) were used for testing interaction with

AKAP18γ. To our surprise, I-1 C terminus half constituting peptides (81-120 and 121- end) did not show any binding to AKAP18γ whereas, both the N-terminus peptides (1-40 and 41-80) showed tight binding to AKAP18γ [Figure 4.3]. These results are suggestive of the fact that the PP1 inhibiting domain of I-1 is also involved in AKAP18γ interaction.

Future efforts focusing on delineating the I-1 motifs required for AKAP18γ and I-1 interaction will be useful in understanding how these two proteins together regulate PP1.

Peptide micro-array approach can be used for deciphering the motif(s) involved in

AKAP18γ and I-1 interaction.

AKAP18γ had been found to be expressed in multiple tissues and its importance in cardiac calcium regulation is indirectly established. The already existing evidence of

AKAP18δ being localized to SR membrane by virtue of protein-protein interaction with

100

Figure 4.2: I-1 GST interacts with AKAP18γ with high affinity.

Kinetics data from SPR experiement showing association and disassociation curves for I- 1 GST (analyte) binding to AKAP18γ (ligand). Increasing concentrations of I-1GST was used for kinetics analysis (12.5-200nM). The SPR program determined a two-site interaction model for I-1GST and AKAP18γ. The two KDs as calculated by the SPR program are: KD1= 8.86 E-07 and KD2= 6.51 E-07.

101

Figure 4.3: I-1 peptides interact with AKAP18γ with high affinity.

Kinetics data from SPR experiment showing association and disassociation curves for biotinylayed I-1 peptides (ligand) binding to AKAP18γ (analyte). Increasing concentrations of AKAP18γ was used for kinetics analysis (12.5-200nM) for each analysis. The KD for I-1 1-40 peptide (A) vs AKAP18γ determined by the SPR program was 1.07 E-08 and that for I-1 41-80 peptide (B) vs AKAP18γ was 2.39 E-09.

102

PLN supports this hypothesis. In this specific study, Lygren et al. had used rat cardiac myocytes as their model. Competing AKAP18δ interaction with PLN using peptides resulted in attenuation of PKA mediated phoshorylation of PLN in cardiac myocytes confirming the importance of AKAP18δ in rat cardiac myocytes (134). A recent publication dedicated to studying AKAP18 isoform distribution across species found that

AKAP18γ is the dominant cardiac cytoplasmic isoform in humans while AKAP18δ serves the same role in rat cardiac myocytes. This group also confirmed that AKAP18γ and AKAP18δ are localized to cardiac SR in mice which expresses both the isoforms

(135). Moreover, SPR experiment in our lab revealed that human AKAP18γ interacts with human PLN with a high affinity (KD in low nano-molar range). It is also known that in canine cardiac myocytes, about 78% of total I-1 is concentrated at the SR for PP1 regulation (121). Put together, data from my research and others suggests that an

AKAP18γ complex consisting of PKA, PP1 and I-1 may be localized in different pockets of the cell for specific PP1 regulation. As mentioned earlier, a well-characterized

AKAP18δ target is PLN (in rats) and hence we hypothesize that in humans and in mice,

AKAP18γ complex might exist at cardiac SR for regulation of PLN phosphorylation status in two ways. First, this complex directly mediates PLN phosphorylation. Second,

AKAP18γ prevents PLN de-phosphorylation by inhibiting PP1. Moreover PLN,

AKAP18γ, PP1 and I-1 are also expressed in skeletal muscle, brain and liver tissues as well, which suggests that this complex may be assembled in other tissue types. These evidence based predictions need to be confirmed as a next step study of this thesis work.

Another important discovery by our laboratory was AKAP18γ association with

PKC isoforms (178). PKC mediated phosphorylation of I-1 prevents I-1 mediated PP1

103 inhibition (115). Binding of PKC to this complex indicates existence of dual regulatory system for I-1. As shown in figure 4.4, PKA mediated phosphorylation of I-1 potentiates it for PP1 inhibition, while PKC mediated phosphorylation disables I-1 from inhibiting

PP1. My work provides evidence for enhanced PKA mediated phosphorylation of

AKAP18γ bound I-1 upon beta-adrenergic activation of cells. We believe AKAP18γ will also mediate the PKC-dependent phosphorylation of I-1 as well. Future work focused on exploring the physiological conditions that augment PKA and PKC mediated phosphorylation of AKAP18γ bound I-1 will help understand the importance of

AKAP18γ in mediating these phosphorylations ultimately for PP1 regulation.

Experimental evidence from our laboratory shows that AKAP18γ interacts not only with PKA, PP1, I-1 and PKC, but also with AC5 and PDE4D3. Even though, PLN is a substrate of AKAP18γ/δ complex and helps localize AKAP18γ/δ to the SR, one possibility that needs to be tested is if AKAP18γ complex is organized to regulate I-1 in other cell compartments. Furthermore, based on this data, it appears that AKAP18γ provides a scaffolding surface for activation and de-activation of PKA itself. It is known that PDE4D3 activation is mediated by PKA phosphorylation at its N-terminus and in the

UCR1 (Upstream Conserved Region 1) (179). PKA mediated phosphorylation of AC5 deactivates this enzyme (180). This suggests that upon activation PKA phosphorylates

PDE4D3 and AC5 to promote its own de-activation. It remains to be tested if this hypothesis is true and if these two PKA mediated phosphorylations of PDE4D and AC5 affects I-1 phosphorylation status.

Over the course of my work and others in the laboratory, we have accumulated evidence to suggest that AKAP18γ belongs to the family of AKAPs that form

104

Figure 4.4: Proposed AKAP18γ mediated regulation of I-1.

Shown in the diagram is the proposed model of I-1 regulation by tethered PKA and PKC based on cellular environment. PKC gets activated when there is an increase in Calcium (Ca2+) in the cell. Calcium gradient has been represented by the shaded blue heavy arrow. Darker shade reflects increase in calcium concentration. Light blue arrows show PKC- mediated phosphorylaton of I-1 and consequently no PP1 inhibition. PKA activation occurs due to increase in cAMP in the cell. cAMP gradient has been represented by a shaded red to yellow heavy arrow. Red shade indicates high cAMP concentration. Red arrows show PKA-mediated Phosphorylation of I-1, which inhibits PP1 activity.

105

multi-molecular complexes. Our lab has established that AKAP18γ interacts with PKC,

PDE4D, AC5, PP1 and I-1. My thesis work showed that AKAP18γ is required for this complex formation as in the absence of AKAP18γ, PP1 and I-1 were unable to interact with PKA. My work also provides evidence for the constitutive presence of PP1, I-1 and

PKA bound to AKAP18γ in cell extract. Future work involving understanding the conditions in which different components of the AKAP18γ are assembled and disassembled will be helpful in understanding the dynamics of the AKAP18γ complex.

More studies focused towards identifying separate pools of AKAP18γ complexes involving specific protein binding partners will also be useful in understanding whether the AKAP18γ complex can bind all the identified binding partners at the same time or

AKAP18γ regulates the activity of its binding partners in discrete AKAP18γ complexes.

Another finding of my thesis project is AKAP18γ oligomerization. I conceived the idea based on the premise that AKAP18γ would require a large binding surface for the reported protein associations to occur concurrently (described above). Binding affinity data available in our laboratory reports high affinity interaction between

AKAP18γ and I-1, AC5, PLN and PDE4D3 separately, with low disassociation rates for these interactions. This data is suggestive of stable interaction of these proteins with

AKAP18γ. Towards testing the AKAP18γ oligomerization hypothesis, with the help of pulldown experiments using over-expression systems, I have been able to show that

AKAP18γ can bind to itself. Moreover, to further the understanding, I collaborated with my labmates, Dr. Andrew Le and Marc Rigatti, to understand the molecular details of this process, as described in chapter 3. Our results suggest that oligomerization of AKAP18γ would cause recruitment of multiple PKA holo-enzymes, and under activated conditions

106 multiple PKA catalytic subunits would be released because of efficient PKA-regulatory subunit auto-phosphorylation. Multiple PKA catalytic subunits cause improved substrate phosphorylation. Future work will involve understanding the molecular basis of

AKAP18γ interaction with itself to develop reagents that will disrupt AKAP18γ oligomeriation in cells. It will be interesting to find out if disruption of AKAP18γ oligomerization affects PKA mediated phosphorylation of PKA substrates associated with AKAP18γ complex.

107

5. IMPACT OF THESIS WORK

The key discovery that AKAP18γ can directly regulate endogenous PP1 inhibitor protein I-1, was unknown before this thesis work. This study puts forth a new mechanism in which the endogenous PP1 inhibitor I-1 is localized and primed for PP1 inhibition as a part of a feed forward mechanism that promotes PKA signal propagation. Moreover, AKAP18γ was found to oligomerize which is indicative of a mechanism to enhance PKA mediated phosphorylation of multiple substrates by this

AKAP. AKAP18γ can interact with important PKA substrates PLN, PDE4D and AC5.

Together these discoveries point to a multi-molecular complex tethered by an AKAP18γ oligomer localized to SR for regulation of cardiac calcium cycling via PLN phosphorylation control.

PP1 is involved in the regulation of a plethora of cellular processes. A complete understanding of pathways and cellular domain specific PP1 regulation will contribute towards complete understanding of how cellular signaling pathways function.

Moreover, detailed understanding of PP1 regulation in the context of particular substrates will greatly help in the development of next generation drugs that focus on treating defective pathways and proteins specifically. The importance of pathway specific treatment is currently being realized in the treatment of cardiac disorders

[Figure 5.1].

108

Figure 5.1: AKAP18 complex based on thesis and work by other labs.

AKAP18γ oligomerizes making all ranges of structures between monomers and tetramers (shown as AKAP18γ/δ x 1-4). AKAP18γ directly interacts with its two substrates at the SR, PLN and I-1. Phosphorylation of PLN augments cardiac calcium cycling. Concomitant phosphorylation of I-1 enhances cardiac cycling by inhibiting PP1 and preserving PLN phosphorylation.

109

6. REFERENCES

1. Choi KY, Satterberg B, Lyons DM, Elion EA. Ste5 tethers multiple protein kinases in the MAP kinase cascade required for mating in S. cerevisiae. Cell.

1994;78(3):499-512.

2. Kornau HC, Schenker LT, Kennedy MB, Seeburg PH. Domain interaction between NMDA receptor subunits and the postsynaptic density protein PSD-95. Science.

1995;269(5231):1737-40.

3. Tsunoda S, Sierralta J, Sun Y, Bodner R, Suzuki E, Becker A, et al. A multivalent

PDZ-domain protein assembles signalling complexes in a G-protein-coupled cascade.

Nature. 1997;388(6639):243-9.

4. Scott JD, Pawson T. Cell signaling in space and time: where proteins come together and when they're apart. Science. 2009;326(5957):1220-4.

5. Shaw AS, Filbert EL. Scaffold proteins and immune-cell signalling. Nat Rev

Immunol. 2009;9(1):47-56.

6. Lin J, Weiss A. Identification of the minimal tyrosine residues required for linker for activation of T cell function. J Biol Chem. 2001;276(31):29588-95.

7. Zhang W, Sloan-Lancaster J, Kitchen J, Trible RP, Samelson LE. LAT: the ZAP-

70 tyrosine kinase substrate that links T cell receptor to cellular activation. Cell.

1998;92(1):83-92.

8. Neasta J, Kiely PA, He DY, Adams DR, O'Connor R, Ron D. Direct interaction between scaffolding proteins RACK1 and 14-3-3zeta regulates brain-derived neurotrophic factor (BDNF) transcription. J Biol Chem. 2012;287(1):322-36.

110

9. Cohen-Katsenelson K, Wasserman T, Darlyuk-Saadon I, Rabner A, Glaser F,

Aronheim A. Identification and analysis of a novel dimerization domain shared by various members of c-Jun N-terminal kinase (JNK) scaffold proteins. J Biol Chem.

2013;288(10):7294-304.

10. Hofmeyer T, Schmelz S, Degiacomi MT, Dal Peraro M, Daneschdar M, Scrima

A, et al. Arranged sevenfold: structural insights into the C-terminal oligomerization domain of human c4b-binding protein. J Mol Biol. 2013;425(8):1302-17.

11. Dervillez X, Huther A, Schuhmacher J, Griesinger C, Cohen JH, von Laer D, et al. Stable expression of soluble therapeutic peptides in eukaryotic cells by multimerisation: application to the HIV-1 fusion inhibitory peptide C46.

ChemMedChem. 2006;1(3):330-9.

12. Molinoff PB. Alpha- and beta-adrenergic receptor subtypes properties, distribution and regulation. Drugs. 1984;28 Suppl 2:1-15.

13. Simonds WF. G protein regulation of adenylate cyclase. Trends Pharmacol Sci.

1999;20(2):66-73.

14. Cytrynska M, Wojda I, Frajnt M, Jakubowicz T. PKA from Saccharomyces cerevisiae can be activated by cyclic AMP and cyclic GMP. Can J Microbiol.

1999;45(1):31-7.

15. Rubin CS, Erlichman J, Rosen OM. Molecular forms and subunit composition of a cyclic adenosine 3',5'-monophosphate-dependent protein kinase purified from bovine heart muscle. J Biol Chem. 1972;247(1):36-44.

111

16. Taylor SS, Buechler JA, Yonemoto W. cAMP-dependent protein kinase: framework for a diverse family of regulatory enzymes. Annu Rev Biochem.

1990;59:971-1005.

17. Kim C, Cheng CY, Saldanha SA, Taylor SS. PKA-I holoenzyme structure reveals a mechanism for cAMP-dependent activation. Cell. 2007;130(6):1032-43.

18. Kim C, Xuong NH, Taylor SS. Crystal structure of a complex between the catalytic and regulatory (RIalpha) subunits of PKA. Science. 2005;307(5710):690-6.

19. Butcher RW, Sutherland EW. Adenosine 3',5'-phosphate in biological materials. I.

Purification and properties of cyclic 3',5'-nucleotide phosphodiesterase and use of this enzyme to characterize adenosine 3',5'-phosphate in human urine. J Biol Chem.

1962;237:1244-50.

20. Moorthy BS, Gao Y, Anand GS. Phosphodiesterases catalyze hydrolysis of cAMP-bound to regulatory subunit of protein kinase A and mediate signal termination.

Mol Cell Proteomics. 2011;10(2):M110 002295.

21. Tasken K, Aandahl EM. Localized effects of cAMP mediated by distinct routes of protein kinase A. Physiol Rev. 2004;84(1):137-67.

22. Zoller MJ, Kerlavage AR, Taylor SS. Structural comparisons of cAMP-dependent protein kinases I and II from porcine skeletal muscle. J Biol Chem. 1979;254(7):2408-12.

23. Zaccolo M. cAMP signal transduction in the heart: understanding spatial control for the development of novel therapeutic strategies. Br J Pharmacol. 2009;158(1):50-60.

24. Colledge M, Scott JD. AKAPs: from structure to function. Trends Cell Biol.

1999;9(6):216-21.

112

25. Welch EJ, Jones BW, Scott JD. Networking with AKAPs: context-dependent regulation of anchored enzymes. Mol Interv. 2010;10(2):86-97.

26. Carr DW, Stofko-Hahn RE, Fraser ID, Bishop SM, Acott TS, Brennan RG, et al.

Interaction of the regulatory subunit (RII) of cAMP-dependent protein kinase with RII- anchoring proteins occurs through an amphipathic helix binding motif. J Biol Chem.

1991;266(22):14188-92.

27. Carr DW, Hausken ZE, Fraser ID, Stofko-Hahn RE, Scott JD. 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(19):13376-82.

28. Newlon MG, Roy M, Hausken ZE, Scott JD, Jennings PA. The A-kinase anchoring domain of type IIalpha cAMP-dependent protein kinase is highly helical. J

Biol Chem. 1997;272(38):23637-44.

29. Newlon MG, Roy M, Morikis D, Hausken ZE, Coghlan V, Scott JD, et al. The molecular basis for protein kinase A anchoring revealed by solution NMR. Nat Struct

Biol. 1999;6(3):222-7.

30. Herberg FW, Maleszka A, Eide T, Vossebein L, Tasken K. Analysis of A-kinase anchoring protein (AKAP) interaction with protein kinase A (PKA) regulatory subunits:

PKA isoform specificity in AKAP binding. J Mol Biol. 2000;298(2):329-39.

31. Carr DW, Stofko-Hahn RE, Fraser ID, Cone RD, Scott JD. Localization of the cAMP-dependent protein kinase to the postsynaptic densities by A-kinase anchoring proteins. Characterization of AKAP 79. J Biol Chem. 1992;267(24):16816-23.

113

32. Pidoux G, Tasken K. Specificity and spatial dynamics of protein kinase A signaling organized by A-kinase-anchoring proteins. J Mol Endocrinol. 2010;44(5):271-

84.

33. Huang LJ, Durick K, Weiner JA, Chun J, Taylor SS. Identification of a novel protein kinase A anchoring protein that binds both type I and type II regulatory subunits.

J Biol Chem. 1997;272(12):8057-64.

34. Michel JJ, Scott JD. AKAP mediated signal transduction. Annu Rev Pharmacol

Toxicol. 2002;42:235-57.

35. Pare GC, Easlick JL, Mislow JM, McNally EM, Kapiloff MS. Nesprin-1alpha contributes to the targeting of mAKAP to the cardiac myocyte nuclear envelope. Exp Cell

Res. 2005;303(2):388-99.

36. Chen L, Marquardt ML, Tester DJ, Sampson KJ, Ackerman MJ, Kass RS.

Mutation of an A-kinase-anchoring protein causes long-QT syndrome. Proc Natl Acad

Sci U S A. 2007;104(52):20990-5.

37. Marx SO, Kurokawa J, Reiken S, Motoike H, D'Armiento J, Marks AR, et al.

Requirement of a macromolecular signaling complex for beta adrenergic receptor modulation of the KCNQ1-KCNE1 potassium channel. Science. 2002;295(5554):496-9.

38. Altier C, Dubel SJ, Barrere C, Jarvis SE, Stotz SC, Scott JD, et al. AKAP79 modulation of L-type channels involves disruption of intramolecular interactions in the

CaV1.2 subunit. Channels (Austin). 2012;6(3):157-65.

39. Malbon CC, Tao J, Wang HY. AKAPs (A-kinase anchoring proteins) and molecules that compose their G-protein-coupled receptor signalling complexes. Biochem

J. 2004;379(Pt 1):1-9.

114

40. Kashishian A, Howard M, Loh C, Gallatin WM, Hoekstra MF, Lai Y. AKAP79 inhibits calcineurin through a site distinct from the immunophilin-binding region. J Biol

Chem. 1998;273(42):27412-9.

41. Klauck TM, Faux MC, Labudda K, Langeberg LK, Jaken S, Scott JD.

Coordination of three signaling enzymes by AKAP79, a mammalian scaffold protein.

Science. 1996;271(5255):1589-92.

42. Carnegie GK, Means CK, Scott JD. A-kinase anchoring proteins: from protein complexes to physiology and disease. IUBMB Life. 2009;61(4):394-406.

43. Diviani D, Dodge-Kafka KL, Li J, Kapiloff MS. A-kinase anchoring proteins: scaffolding proteins in the heart. Am J Physiol Heart Circ Physiol. 2011;301(5):H1742-

53.

44. Carnegie GK, Soughayer J, Smith FD, Pedroja BS, Zhang F, Diviani D, et al.

AKAP-Lbc mobilizes a cardiac hypertrophy signaling pathway. Mol Cell.

2008;32(2):169-79.

45. Reynolds JG, McCalmon SA, Tomczyk T, Naya FJ. Identification and mapping of protein kinase A binding sites in the costameric protein myospryn. Biochim Biophys

Acta. 2007;1773(6):891-902.

46. Eggers CT, Schafer JC, Goldenring JR, Taylor SS. D-AKAP2 interacts with Rab4 and Rab11 through its RGS domains and regulates transferrin receptor recycling. J Biol

Chem. 2009;284(47):32869-80.

47. Gao S, Wang HY, Malbon CC. AKAP5 and AKAP12 Form Homo-oligomers. J

Mol Signal. 2011;6:3.

115

48. Gold MG, Smith FD, Scott JD, Barford D. AKAP18 contains a phosphoesterase domain that binds AMP. J Mol Biol. 2008;375(5):1329-43.

49. Gold MG, Stengel F, Nygren PJ, Weisbrod CR, Bruce JE, Robinson CV, et al.

Architecture and dynamics of an A-kinase anchoring protein 79 (AKAP79) signaling complex. Proc Natl Acad Sci U S A. 2011;108(16):6426-31.

50. Bhattacharyya S, Biou V, Xu W, Schluter O, Malenka RC. A critical role for

PSD-95/AKAP interactions in endocytosis of synaptic AMPA receptors. Nat Neurosci.

2009;12(2):172-81.

51. Smith FD, Langeberg LK, Cellurale C, Pawson T, Morrison DK, Davis RJ, et al.

AKAP-Lbc enhances cyclic AMP control of the ERK1/2 cascade. Nat Cell Biol.

2010;12(12):1242-9.

52. Baisamy L, Jurisch N, Diviani D. Leucine zipper-mediated homo-oligomerization regulates the Rho-GEF activity of AKAP-Lbc. J Biol Chem. 2005;280(15):15405-12.

53. Olsen JV, Blagoev B, Gnad F, Macek B, Kumar C, Mortensen P, et al. Global, in vivo, and site-specific phosphorylation dynamics in signaling networks. Cell.

2006;127(3):635-48.

54. Peti W, Nairn AC, Page R. Structural basis for protein phosphatase 1 regulation and specificity. FEBS J. 2013;280(2):596-611.

55. Roy J, Cyert MS. Cracking the phosphatase code: docking interactions determine substrate specificity. Sci Signal. 2009;2(100):re9.

56. Shi Y. Serine/threonine phosphatases: mechanism through structure. Cell.

2009;139(3):468-84.

116

57. Virshup DM, Shenolikar S. From promiscuity to precision: protein phosphatases get a makeover. Mol Cell. 2009;33(5):537-45.

58. Ceulemans H, Bollen M. Functional diversity of protein phosphatase-1, a cellular economizer and reset button. Physiol Rev. 2004;84(1):1-39.

59. Eto M. Regulation of cellular protein phosphatase-1 (PP1) by phosphorylation of the CPI-17 family, C-kinase-activated PP1 inhibitors. J Biol Chem. 2009;284(51):35273-

60. Choy MS, Page R, Peti W. Regulation of protein phosphatase 1 by intrinsically disordered proteins. Biochem Soc Trans. 2012;40(5):969-74.

61. Egloff MP, Cohen PT, Reinemer P, Barford D. Crystal structure of the catalytic subunit of human protein phosphatase 1 and its complex with tungstate. J Mol Biol.

1995;254(5):942-59.

62. Colbran RJ, Carmody LC, Bauman PA, Wadzinski BE, Bass MA. Analysis of specific interactions of native protein phosphatase 1 isoforms with targeting subunits.

Methods Enzymol. 2003;366:156-75.

63. Durfee T, Becherer K, Chen PL, Yeh SH, Yang Y, Kilburn AE, et al. The retinoblastoma protein associates with the protein phosphatase type 1 catalytic subunit.

Genes Dev. 1993;7(4):555-69.

64. Yoshida K, Watanabe M, Kato H, Dutta A, Sugano S. BH-protocadherin-c, a member of the cadherin superfamily, interacts with protein phosphatase 1 alpha through its intracellular domain. FEBS Lett. 1999;460(1):93-8.

65. Bollen M, Peti W, Ragusa MJ, Beullens M. The extended PP1 toolkit: designed to create specificity. Trends Biochem Sci. 2010;35(8):450-8.

117

66. Watanabe T, Huang HB, Horiuchi A, da Cruze Silva EF, Hsieh-Wilson L, Allen

PB, et al. Protein phosphatase 1 regulation by inhibitors and targeting subunits. Proc Natl

Acad Sci U S A. 2001;98(6):3080-5.

67. Cohen PT. Protein phosphatase 1--targeted in many directions. J Cell Sci.

2002;115(Pt 2):241-56.

68. Hendrickx A, Beullens M, Ceulemans H, Den Abt T, Van Eynde A, Nicolaescu

E, et al. Docking motif-guided mapping of the interactome of protein phosphatase-1.

Chem Biol. 2009;16(4):365-71.

69. Egloff MP, Johnson DF, Moorhead G, Cohen PT, Cohen P, Barford D. Structural basis for the recognition of regulatory subunits by the catalytic subunit of protein phosphatase 1. EMBO J. 1997;16(8):1876-87.

70. Wakula P, Beullens M, Ceulemans H, Stalmans W, Bollen M. Degeneracy and function of the ubiquitous RVXF motif that mediates binding to protein phosphatase-1. J

Biol Chem. 2003;278(21):18817-23.

71. Bollen M. Combinatorial control of protein phosphatase-1. Trends Biochem Sci.

2001;26(7):426-31.

72. Terrak M, Kerff F, Langsetmo K, Tao T, Dominguez R. Structural basis of protein phosphatase 1 regulation. Nature. 2004;429(6993):780-4.

73. Hurley TD, Yang J, Zhang L, Goodwin KD, Zou Q, Cortese M, et al. Structural basis for regulation of protein phosphatase 1 by inhibitor-2. J Biol Chem.

2007;282(39):28874-83.

118

74. Ragusa MJ, Dancheck B, Critton DA, Nairn AC, Page R, Peti W. Spinophilin directs protein phosphatase 1 specificity by blocking substrate binding sites. Nat Struct

Mol Biol. 2010;17(4):459-64.

75. Dosztanyi Z, Csizmok V, Tompa P, Simon I. IUPred: web server for the prediction of intrinsically unstructured regions of proteins based on estimated energy content. Bioinformatics. 2005;21(16):3433-4.

76. Goldberg J, Huang HB, Kwon YG, Greengard P, Nairn AC, Kuriyan J. Three- dimensional structure of the catalytic subunit of protein serine/threonine phosphatase-1.

Nature. 1995;376(6543):745-53.

77. Kelker MS, Page R, Peti W. Crystal structures of protein phosphatase-1 bound to nodularin-R and tautomycin: a novel scaffold for structure-based drug design of serine/threonine phosphatase inhibitors. J Mol Biol. 2009;385(1):11-21.

78. Cohen P. The structure and regulation of protein phosphatases. Annu Rev

Biochem. 1989;58:453-508.

79. Aschenbach WG, Suzuki Y, Breeden K, Prats C, Hirshman MF, Dufresne SD, et al. The muscle-specific protein phosphatase PP1G/R(GL)(G(M))is essential for activation of glycogen synthase by exercise. J Biol Chem. 2001;276(43):39959-67.

80. Alemany S, Pelech S, Brierley CH, Cohen P. The protein phosphatases involved in cellular regulation. Evidence that dephosphorylation of glycogen phosphorylase and glycogen synthase in the glycogen and microsomal fractions of rat liver are catalysed by the same enzyme: protein phosphatase-1. Eur J Biochem. 1986;156(1):101-10.

119

81. Qian J, Lesage B, Beullens M, Van Eynde A, Bollen M. PP1/Repo-man dephosphorylates mitotic histone H3 at T3 and regulates chromosomal aurora B targeting. Curr Biol. 2011;21(9):766-73.

82. Udagedara I, Premaratne M, Rukhlenko ID, Hattori HT, Agrawal GP. Unified perfectly matched layer for finite-difference time-domain modeling of dispersive optical materials. Opt Express. 2009;17(23):21179-90.

83. Lesage B, Beullens M, Nuytten M, Van Eynde A, Keppens S, Himpens B, et al.

Interactor-mediated nuclear translocation and retention of protein phosphatase-1. J Biol

Chem. 2004;279(53):55978-84.

84. Abraham G, Rhodes DP, Banerjee AK. The 5' terminal structure of the methylated mRNA synthesized in vitro by vesicular stomatitis virus. Cell. 1975;5(1):51-

85. Aitken A, Cohen P. Isolation and characterisation of active fragments of protein phosphatase inhibitor-1 from rabbit skeletal muscle. FEBS Lett. 1982;147(1):54-8.

86. Elbrecht A, DiRenzo J, Smith RG, Shenolikar S. Molecular cloning of protein phosphatase inhibitor-1 and its expression in rat and rabbit tissues. J Biol Chem.

1990;265(23):13415-8.

87. Endo S, Zhou X, Connor J, Wang B, Shenolikar S. Multiple structural elements define the specificity of recombinant human inhibitor-1 as a protein phosphatase-1 inhibitor. Biochemistry. 1996;35(16):5220-8.

88. Liu HT, Lin TH, Kuo HC, Chen YC, Tsay HJ, Jeng HH, et al. Identification of the alternative splice products encoded by the human protein phosphatase inhibitor-1 gene. Biochem Biophys Res Commun. 2002;291(5):1293-6.

120

89. Nimmo GA, Cohen P. The regulation of glycogen metabolism. Purification and characterisation of protein phosphatase inhibitor-1 from rabbit skeletal muscle. Eur J

Biochem. 1978;87(2):341-51.

90. Foulkes JG, Strada SJ, Henderson PJ, Cohen P. A kinetic analysis of the effects of inhibitor-1 and inhibitor-2 on the activity of protein phosphatase-1. Eur J Biochem.

1983;132(2):309-13.

91. Huang FL, Glinsmann WH. Separation and characterization of two phosphrylase phosphatase inhibitors from rabbit skeletal muscle. Eur J Biochem. 1976;70(2):419-26.

92. El-Armouche A, Bednorz A, Pamminger T, Ditz D, Didie M, Dobrev D, et al.

Role of calcineurin and protein phosphatase-2A in the regulation of phosphatase inhibitor-1 in cardiac myocytes. Biochem Biophys Res Commun. 2006;346(3):700-6.

93. Braz JC, Gregory K, Pathak A, Zhao W, Sahin B, Klevitsky R, et al. PKC-alpha regulates cardiac contractility and propensity toward heart failure. Nat Med.

2004;10(3):248-54.

94. Huang YC, Chen YC, Tsay HJ, Chyan CL, Chen CY, Huang HB, et al. The effect of PKA-phosphorylation on the structure of inhibitor-1 studied by NMR spectroscopy. J

Biochem. 2010;147(2):273-8.

95. Oliver CJ, Shenolikar S. Physiologic importance of protein phosphatase inhibitors. Front Biosci. 1998;3:D961-72.

96. Gold MG, Barford D, Komander D. Lining the pockets of kinases and phosphatases. Curr Opin Struct Biol. 2006;16(6):693-701.

121

97. Steen RL, Martins SB, Tasken K, Collas P. Recruitment of protein phosphatase 1 to the nuclear envelope by A-kinase anchoring protein AKAP149 is a prerequisite for nuclear lamina assembly. J Cell Biol. 2000;150(6):1251-62.

98. Steen RL, Beullens M, Landsverk HB, Bollen M, Collas P. AKAP149 is a novel

PP1 specifier required to maintain nuclear envelope integrity in G1 phase. J Cell Sci.

2003;116(Pt 11):2237-46.

99. Kuntziger T, Rogne M, Folstad RL, Collas P. Association of PP1 with its regulatory subunit AKAP149 is regulated by serine phosphorylation flanking the RVXF motif of AKAP149. Biochemistry. 2006;45(18):5868-77.

100. Schillace RV, Voltz JW, Sim AT, Shenolikar S, Scott JD. Multiple interactions within the AKAP220 signaling complex contribute to protein phosphatase 1 regulation. J

Biol Chem. 2001;276(15):12128-34.

101. Le AV, Tavalin SJ, Dodge-Kafka KL. Identification of AKAP79 as a protein phosphatase 1 catalytic binding protein. Biochemistry. 2011;50(23):5279-91.

102. Bartel S, Stein B, Eschenhagen T, Mende U, Neumann J, Schmitz W, et al.

Protein phosphorylation in isolated trabeculae from nonfailing and failing human hearts.

Mol Cell Biochem. 1996;157(1-2):171-9.

103. El-Armouche A, Pohlmann L, Schlossarek S, Starbatty J, Yeh YH, Nattel S, et al.

Decreased phosphorylation levels of cardiac myosin-binding protein-C in human and experimental heart failure. J Mol Cell Cardiol. 2007;43(2):223-9.

104. MacLennan DH, Kranias EG. Phospholamban: a crucial regulator of cardiac contractility. Nat Rev Mol Cell Biol. 2003;4(7):566-77.

122

105. Chu G, Lester JW, Young KB, Luo W, Zhai J, Kranias EG. A single site (Ser16) phosphorylation in phospholamban is sufficient in mediating its maximal cardiac responses to beta -agonists. J Biol Chem. 2000;275(49):38938-43.

106. Luo W, Chu G, Sato Y, Zhou Z, Kadambi VJ, Kranias EG. Transgenic approaches to define the functional role of dual site phospholamban phosphorylation. J

Biol Chem. 1998;273(8):4734-9.

107. Simmerman HK, Jones LR. Phospholamban: protein structure, mechanism of action, and role in cardiac function. Physiol Rev. 1998;78(4):921-47.

108. Nicolaou P, Kranias EG. Role of PP1 in the regulation of Ca cycling in cardiac physiology and pathophysiology. Front Biosci. 2009;14:3571-85.

109. Bristow MR, Ginsburg R, Minobe W, Cubicciotti RS, Sageman WS, Lurie K, et al. Decreased catecholamine sensitivity and beta-adrenergic-receptor density in failing human hearts. N Engl J Med. 1982;307(4):205-11.

110. Neumann J, Eschenhagen T, Jones LR, Linck B, Schmitz W, Scholz H, et al.

Increased expression of cardiac phosphatases in patients with end-stage heart failure. J

Mol Cell Cardiol. 1997;29(1):265-72.

111. Yamada M, Ikeda Y, Yano M, Yoshimura K, Nishino S, Aoyama H, et al.

Inhibition of protein phosphatase 1 by inhibitor-2 gene delivery ameliorates heart failure progression in genetic cardiomyopathy. FASEB J. 2006;20(8):1197-9.

112. El-Armouche A, Pamminger T, Ditz D, Zolk O, Eschenhagen T. Decreased protein and phosphorylation level of the protein phosphatase inhibitor-1 in failing human hearts. Cardiovasc Res. 2004;61(1):87-93.

123

113. El-Armouche A, Rau T, Zolk O, Ditz D, Pamminger T, Zimmermann WH, et al.

Evidence for protein phosphatase inhibitor-1 playing an amplifier role in beta-adrenergic signaling in cardiac myocytes. FASEB J. 2003;17(3):437-9.

114. Gupta RC, Mishra S, Yang XP, Sabbah HN. Reduced inhibitor 1 and 2 activity is associated with increased protein phosphatase type 1 activity in left ventricular myocardium of one-kidney, one-clip hypertensive rats. Mol Cell Biochem. 2005;269(1-

2):49-57.

115. Rodriguez P, Mitton B, Nicolaou P, Chen G, Kranias EG. Phosphorylation of human inhibitor-1 at Ser67 and/or Thr75 attenuates stimulatory effects of protein kinase

A signaling in cardiac myocytes. Am J Physiol Heart Circ Physiol. 2007;293(1):H762-9.

116. Ferguson SS. Evolving concepts in G protein-coupled receptor endocytosis: the role in receptor desensitization and signaling. Pharmacol Rev. 2001;53(1):1-24.

117. Bowling N, Walsh RA, Song G, Estridge T, Sandusky GE, Fouts RL, et al.

Increased protein kinase C activity and expression of Ca2+-sensitive isoforms in the failing human heart. Circulation. 1999;99(3):384-91.

118. Carr AN, Schmidt AG, Suzuki Y, del Monte F, Sato Y, Lanner C, et al. Type 1 phosphatase, a negative regulator of cardiac function. Mol Cell Biol. 2002;22(12):4124-

35.

119. El-Armouche A, Boknik P, Eschenhagen T, Carrier L, Knaut M, Ravens U, et al.

Molecular determinants of altered Ca2+ handling in human chronic atrial fibrillation.

Circulation. 2006;114(7):670-80.

124

120. El-Armouche A, Gocht F, Jaeckel E, Wittkopper K, Peeck M, Eschenhagen T.

Long-term beta-adrenergic stimulation leads to downregulation of protein phosphatase inhibitor-1 in the heart. Eur J Heart Fail. 2007;9(11):1077-80.

121. Gupta RC, Mishra S, Rastogi S, Imai M, Habib O, Sabbah HN. Cardiac SR- coupled PP1 activity and expression are increased and inhibitor 1 protein expression is decreased in failing hearts. Am J Physiol Heart Circ Physiol. 2003;285(6):H2373-81.

122. Nicolaou P, Rodriquez P, Ren X, Zhou X, Qian J, Sadayappan S, et al. Inducible expression of active protein phosphatase-1 inhibitor-1 enhances basal cardiac function and protects against ischemic/reperfusion injury. Circ Res. 2009;104(8):1012-20.

123. Pathak A, del Monte F, Zhao W, Schultz JE, Lorenz JN, Bodi I, et al.

Enhancement of cardiac function and suppression of heart failure progression by inhibition of protein phosphatase 1. Circ Res. 2005;96(7):756-66.

124. Chen G, Zhou X, Florea S, Qian J, Cai W, Zhang Z, et al. Expression of active protein phosphatase 1 inhibitor-1 attenuates chronic beta-agonist-induced cardiac apoptosis. Basic Res Cardiol. 2010;105(5):573-81.

125. El-Armouche A, Wittkopper K, Degenhardt F, Weinberger F, Didie M,

Melnychenko I, et al. Phosphatase inhibitor-1-deficient mice are protected from catecholamine-induced arrhythmias and myocardial hypertrophy. Cardiovasc Res.

2008;80(3):396-406.

126. Wittkopper K, Fabritz L, Neef S, Ort KR, Grefe C, Unsold B, et al. Constitutively active phosphatase inhibitor-1 improves cardiac contractility in young mice but is deleterious after catecholaminergic stress and with aging. J Clin Invest. 2010;120(2):617-

26.

125

127. Leineweber K, Klapproth S, Beilfuss A, Silber RE, Heusch G, Philipp T, et al.

Unchanged G-protein-coupled receptor kinase activity in the aging human heart. J Am

Coll Cardiol. 2003;42(8):1487-92.

128. Wittkopper K, Eschenhagen T, El-Armouche A. Phosphatase-1-inhibitor-1: amplifier or attenuator of catecholaminergic stress? Basic Res Cardiol. 2010;105(5):569-

71.

129. Fraser ID, Tavalin SJ, Lester LB, Langeberg LK, Westphal AM, Dean RA, et al.

A novel lipid-anchored A-kinase Anchoring Protein facilitates cAMP-responsive membrane events. EMBO J. 1998;17(8):2261-72.

130. Gray PC, Tibbs VC, Catterall WA, Murphy BJ. Identification of a 15-kDa cAMP- dependent protein kinase-anchoring protein associated with skeletal muscle L-type calcium channels. J Biol Chem. 1997;272(10):6297-302.

131. Trotter KW, Fraser ID, Scott GK, Stutts MJ, Scott JD, Milgram SL. Alternative splicing regulates the subcellular localization of A-kinase anchoring protein 18 isoforms.

J Cell Biol. 1999;147(7):1481-92.

132. Henn V, Edemir B, Stefan E, Wiesner B, Lorenz D, Theilig F, et al. 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.

2004;279(25):26654-65.

133. Brown RL, August SL, Williams CJ, Moss SB. AKAP7gamma is a nuclear RI- binding AKAP. Biochem Biophys Res Commun. 2003;306(2):394-401.

126

134. Lygren B, Carlson CR, Santamaria K, Lissandron V, McSorley T, Litzenberg J, et al. AKAP complex regulates Ca2+ re-uptake into heart sarcoplasmic reticulum. EMBO

Rep. 2007;8(11):1061-7.

135. Johnson KR, Nicodemus-Johnson J, Carnegie GK, Danziger RS. Molecular evolution of A-kinase anchoring protein (AKAP)-7: implications in comparative PKA compartmentalization. BMC Evol Biol. 2012;12:125.

136. Josefsen K, Lee YC, Thams P, Efendic S, Nielsen JH. AKAP 18 alpha and gamma have opposing effects on insulin release in INS-1E cells. FEBS Lett.

2010;584(1):81-5.

137. Genous D, Haditsch U, Knobloch M, Michalon A, Storm D, Mansuy IM. Protein phosphatase 1 is a molecular constraint on learning and memory. Nature.

2002;418(6901):970-5.

138. Mansuy IM, Shenolikar S. Protein serine/threonine phosphatases in neuronal plasticity and disorders of learning and memory. Trends Neurosci. 2006;29(12):679-86.

139. Nguyen C, Nishi A, Kansy JW, Fernandex J, Hayashi K, Gillardon F, et al.

Regulation of protein phosphatase inhibitor-1 by cyclin-dependent kinase 5. J Biol Chem.

2007;282(22):16511-20.

140. Nicolaou P, Hajjar RJ, Kranias EG. Role of protein phosphoatae-1 inhibitor-1 in cardiac physiology and pathophysiology. J Mol Cell Cardiol. 2009;47(3):365-71.

141. Rodriquez P, Mitton B, Waggoner JR, Kranias EG. Identification of a novel phosphorylation site in protein phosphatase inhibitor-1 as as negative regulator of cardiac function. J Biol Chem. 2006;281(50):38599-608.

127

142. Sahin B, Shu H, Fernandez J, El-Armouche A, Molkentin JD, Nairn AC, et al.

Phosphorylation of protein phosphatase inhibitor-1 by protein kinase C. J Biol Chem.

2006;281(34):24322-35.

143. Mauban JR, O'Donnell M, Warrier S, Manni S, Bond M. AKAP-scaffolding proteins and regualtion of cardiac physiology. Physiology. 2009:78-87.

144. Dodge-Kafka KL, Bauman A, Mayer N, Henson E, Heredia L, Ahn J, et al. cAMP-stimulated protein phosphatse 2A activity associated with mAKAP signlaing complexes inhibits the phosphorylation and activity of the cAMP-specific phosphodiesterase PDE4D3. 2010. 2010;Submitted.

145. Corbin JD, Reimann EM. Assay of cyclic AMP-dependent protein kinases.

Methods Enzymol. 1974;38:287-90.

146. Connor JH, Weiser DC, Li S, Hallenbeck JM, Shenolikar S. Growth arrest and

DNA damage-inducible protein GADD34 assembles a novel signaling complex containing protein phosphatase 1 and inhibitor 1. Mol Cell Biol. 2001;21(20):6841-50.

PMCID: 99861.

147. Morton E, Macrae IM, McCabe C, Brown SM, White F. Identification of the growth arrest and DNA damage protein GADD34 in the normal human heart and demonstration of alterations in expression following myocardial ischaemia. Int J Cardiol.

2006;107(1):126-9.

148. Pawson CT, Scott JD. Signal integration through blending, bolstering and bifurcating of intracellular information. Nat Struct Mol Biol. 2010;17(6):653-8.

149. Logue JS, Scott JD. Organizing signal transduction through A-kinase anchoring proteins (AKAPs). FEBS J. 2010;277(21):4370-5. PMCID: 3014294.

128

150. Ali MH, Imperiali B. Protein oligomerization: how and why. Bioorg Med Chem.

2005;13(17):5013-20.

151. Marianayagam NJ, Sunde M, Matthews JM. The power of two: protein dimerization in biology. Trends Biochem Sci. 2004;29(11):618-25.

152. Burgess AW. EGFR family: structure physiology signalling and therapeutic targets. Growth Factors. 2008;26(5):263-74.

153. Tadagaki K, Jockers R, Kamal M. History and biological significance of GPCR heteromerization in the neuroendocrine system. Neuroendocrinology. 2012;95(3):223-31.

154. Pan CQ, Sudol M, Sheetz M, Low BC. Modularity and functional plasticity of scaffold proteins as p(l)acemakers in cell signaling. Cell Signal. 2012;24(11):2143-65.

155. Wang Y, Elion EA. Nuclear export and plasma membrane recruitment of the Ste5 scaffold are coordinated with oligomerization and association with signal transduction components. Mol Biol Cell. 2003;14(6):2543-58. PMCID: 194901.

156. Yablonski D, Marbach I, Levitzki A. Dimerization of Ste5, a mitogen-activated protein kinase cascade scaffold protein, is required for signal transduction. Proc Natl

Acad Sci U S A. 1996;93(24):13864-9. PMCID: 19450.

157. Dessauer CW. Adenylyl cyclase--A-kinase anchoring protein complexes: the next dimension in cAMP signaling. Mol Pharmacol. 2009;76(5):935-41. PMCID: 2774998.

158. Dodge KL, Carr DW, Yue C, Sanborn BM. A role for AKAP (A kinase anchoring protein) scaffolding in the loss of a cyclic adenosine 3',5'-monophosphate inhibitory response in late pregnant rat myometrium. Mol Endocrinol. 1999;13(12):1977-87.

129

159. Kapiloff MS, Piggott LA, Sadana R, Li J, Heredia LA, Henson E, et al. An adenylyl cyclase-mAKAPbeta signaling complex regulates cAMP levels in cardiac myocytes. J Biol Chem. 2009;284(35):23540-6. PMCID: 2749128.

160. Redden JM, Dodge-Kafka KL. AKAP phosphatase complexes in the heart. J

Cardiovasc Pharmacol. 2011;58(4):354-62. PMCID: 3158305.

161. Gray PC, Johnson BD, Westenbroek RE, Hays LG, Yates JR, 3rd, Scheuer T, et al. Primary structure and function of an A kinase anchoring protein associated with calcium channels. Neuron. 1998;20(5):1017-26.

162. Singh A, Redden JM, Kapiloff MS, Dodge-Kafka KL. The large isoforms of A- kinase anchoring protein 18 mediate the phosphorylation of inhibitor-1 by protein kinase

A and the inhibition of protein phosphatase 1 activity. Mol Pharmacol. 2011;79(3):533-

450.

163. Digman MA, Dalal R, Horwitz AF, Gratton E. Mapping the number of molecules and brightness in the laser scanning microscope. Biophys J. 2008;94(6):2320-32.

PMCID: 2257897.

164. Martin BR, Deerinck TJ, Ellisman MH, Taylor SS, Tsien RY. Isoform-specific

PKA dynamics revealed by dye-triggered aggregation and DAKAP1alpha-mediated localization in living cells. Chem Biol. 2007;14(9):1031-42.

165. Flockhart DA, Watterson DM, Corbin JD. Studies on functional domains of the regulatory subunit of bovine heart adenosine 3':5'-monophosphate-dependent protein kinase. J Biol Chem. 1980;255(10):4435-40.

166. Francis SH, Corbin JD. Structure and function of cyclic nucleotide-dependent protein kinases. Annu Rev Physiol. 1994;56:237-72.

130

167. Vigil D, Blumenthal DK, Brown S, Taylor SS, Trewhella J. Differential effects of substrate on type I and type II PKA holoenzyme dissociation. Biochemistry.

2004;43(19):5629-36.

168. Zelada A, Castilla R, Passeron S, Giasson L, Cantore ML. Interactions between regulatory and catalytic subunits of the Candida albicans cAMP-dependent protein kinase are modulated by autophosphorylation of the regulatory subunit. Biochim Biophys Acta.

2002;1542(1-3):73-81.

169. Hayashi MK, Ames HM, Hayashi Y. Tetrameric hub structure of postsynaptic scaffolding protein homer. J Neurosci. 2006;26(33):8492-501.

170. Hashimoto K, Nishi H, Bryant S, Panchenko AR. Caught in self-interaction: evolutionary and functional mechanisms of protein homooligomerization. Phys Biol.

2011;8(3):035007. PMCID: 3148176.

171. Mason JM, Arndt KM. Coiled coil domains: stability, specificity, and biological implications. Chembiochem. 2004;5(2):170-6.

172. Nishi H, Hashimoto K, Panchenko AR. Phosphorylation in protein-protein binding: effect on stability and function. Structure. 2011;19(12):1807-15. PMCID:

3240861.

173. De Meyts P. The insulin receptor: a prototype for dimeric, allosteric membrane receptors? Trends Biochem Sci. 2008;33(8):376-84.

174. Munoz LM, Lucas P, Holgado BL, Barroso R, Vega B, Rodriguez-Frade JM, et al. Receptor oligomerization: a pivotal mechanism for regulating chemokine function.

Pharmacol Ther. 2011;131(3):351-8.

131

175. Li H, Pink MD, Murphy JG, Stein A, Dell'Acqua ML, Hogan PG. Balanced interactions of calcineurin with AKAP79 regulate Ca2+-calcineurin-NFAT signaling. Nat

Struct Mol Biol. 2012;19(3):337-45. PMCID: 3294036.

176. Diviani D, Soderling J, Scott JD. AKAP-Lbc anchors protein kinase A and nucleates Galpha 12-selective Rho-mediated stress fiber formation. J Biol Chem.

2001;276(47):44247-57.

177. Berrebi-Bertrand I, Souchet M, Camelin JC, Laville MP, Calmels T, Bril A.

Biophysical interaction between phospholamban and protein phosphatase 1 regulatory subunit GM. FEBS Lett. 1998;439(3):224-30.

178. Redden JM, Le AV, Singh A, Federkiewicz K, Smith S, Dodge-Kafka KL.

Spatiotemporal regulation of PKC via interactions with AKAP7 isoforms. Biochem J.

2012;446(2):301-9.

179. MacKenzie SJ, Baillie GS, McPhee I, MacKenzie C, Seamons R, McSorley T, et al. Long PDE4 cAMP specific phosphodiesterases are activated by protein kinase A- mediated phosphorylation of a single serine residue in Upstream Conserved Region 1

(UCR1). Br J Pharmacol. 2002;136(3):421-33.

180. Iwami G, Kawabe J, Ebina T, Cannon PJ, Homcy CJ, Ishikawa Y. Regulation of adenylyl cyclase by protein kinase A. J Biol Chem. 1995;270(21):12481-4.