An Investigation of the Phospholamban-Serca Regulatory Interaction with Fluorescence Resonance Energy Transfer

Loyola University Chicago Loyola eCommons

Dissertations Theses and Dissertations

2012

Philip Adam Bidwell Loyola University Chicago

Follow this and additional works at: https://ecommons.luc.edu/luc_diss

Part of the Physiology Commons

Recommended Citation Bidwell, Philip Adam, "An Investigation of the Phospholamban-Serca Regulatory Interaction with Fluorescence Resonance Energy Transfer" (2012). Dissertations. 330. https://ecommons.luc.edu/luc_diss/330

This Dissertation is brought to you for free and open access by the Theses and Dissertations at Loyola eCommons. It has been accepted for inclusion in Dissertations by an authorized administrator of Loyola eCommons. For more information, please contact [email protected].

AN INVESTIGATION OF THE PHOSPHOLAMBAN-SERCA

REGULATORY INTERACTION WITH

FLUORESCENCE RESONANCE ENERGY TRANSFER

A DISSERTATION SUBMITTED TO

THE FACULTY OF THE GRADUATE SCHOOL

IN CANDIDACY FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY

PROGRAM OF CELL AND MOLECULAR PHYSIOLOGY

PHILIP A. BIDWELL

CHICAGO, IL

DECEMBER 2012

This dissertation is dedicated to my wife Christy for her continued love and support

ACKNOWLEDGEMENTS

First and foremost, I would like to thank my mentor, Dr. Seth Robia. He has provided amazing assistance guidance throughout my graduate career. He has always had an open door and has helped me progress as a scientist, on and off the bench. I could not have progressed to this point without his support and belief in me.

I would also like to thank the whole Robia Lab for continued assistance and encouragement. In particular, I would like to thank Dr. Zhanjia Hou. He has been instrumental in helping develop many of the FRET protocols have utilized. Also, I would like to thank Dan Blackwell for developing the adenoviral vectors that I have utilized.

I would like to thank Aleksey Zima and his lab for assistance as well as provided the myocytes used for this work. I would also like to thank Dr. Greg Mignery and his lab for assistance in developing our in cell activity assay.

Lastly, for their time, effort, and guidance, I would like to thank my committee,

Dr. Pieter de Tombe, Dr. Aleksey Zima, Dr. Edward Campbell, and Dr. J Michael

O’Donnell.

iii

TABLE OF CONTENTS

ACKNOWLEDGEMENTS iii

LIST OF FIGURES vi

LIST OF ABBREVIATIONS viii

CHAPTER I: INTRODUCTION 1

CHAPTER II: REVIEW OF RELATED LITERATURE 7 SERCA 7 Phospholamban 15 Mechanism of Phospholamban Inhibition 20 Potential Interacting Proteins 37 Disease and Therapy 38

CHAPTER III: MATERIALS AND METHODS 41 Acceptor Photobleaching FRET 41 E-FRET 44 Binding Curves and Fitting Analysis 45 Confocal Microscopy 48 Tissue Culture 49

CHAPTER IV: LIVE CELL CALCIUM UPTAKE ASSAY 52 Rationale 52 Protocol 53 Principle of Assay 54 Summary 56

CHAPTER V: STEADY-STATE DETECTION OF PLB-SERCA BINDING 57 Rationale 57 Mean FRET 57 Protein Dependence of FRET 60 E-FRET Rationale 63 Calcium Dose Effect 64 Thapsigargin Dose Effect 68 Phospholamban Antibody Effect 69

PLB-PLB FRET 71 Summary 73

CHAPTER VI: SERCA INTERACTION WITH TRUNCATED PHOSPHOLAMBAN 75 Rationale 75 Vector Construction 76 Simulation of tmPLB FRET 76 Calcium Effect 78 YFP-tmPLB Activity 78 Pentamer Affinity 80 Summary 80

CHAPTER VII: REAL-TIME PLB-SERCA FRET 82 Rationale 82 Characterization of Myocyte FRET 83 Single Calcium Transient FRET 85 FRET Response to Alternating Pacing 87 FRET Response to Calcium Changes in Heterologous Cells 89 Summary 91

CHAPTER VIII: DISCUSSION 92 Global vs Single Molecule Effects 92 Steady-State FRET 93 Alternative Binding Site 99 Proposed Model of PLB Inhibition 101 Effects of Pacing on Myocyte FRET 104 Future Directions 106 Concluding Remarks 109

BIBLIOGRAPHY 110

VITA 130

LIST OF FIGURES

Figure Page

1. Diagram of the basic calcium handling elements within cardiac myocytes 3

2. Simplified SERCA reaction diagram 10

3. Proposed SERCA structural transitions 13

4. Coupled equilibrium of PLB binding 18

5. Biochemical representation of SERCA activity 21

6. Models of PLB inhibition of SERCA 23

7. Predicted regulator binding locations 35

8. Acceptor Photobleaching FRET 43

9. Quantitative analysis of protein concentration dependence of FRET 46

10. Calicum uptake in live cells 55

11. Mean PLB-SERCA FRET 59

12. Protein dependence on FRET, acceptor photobleaching 62

13. Calcium effect on PLB-SERCA FRET, E-FRET 66

14. Calibration of calcium buffered solutions 67

15. Thapsigargin effect on PLB-SERCA FRET, E-FRET 70

16. PLB antibody (2D12) effect on PLB-SERCA FRET 72 vi

17. FRET between YFP-tmPLB and Cer-SERCA 77

18. Characterization of tmPLB 79

19. Adenoviral expression of Cer-SERCA and YFP-PLB in cardiac myocytes 84

20. Quantification of calcium and FRET changes in single transients 86

21. Quantification of calcium and FRET changes by alternating pacing 88

22. Time course of calcium effect on PLB-SERCA FRET 90

23. Summary of PLB binding 98

24. Models of PLB-SERCA binding 103

vii

LIST OF ABBREVIATIONS

Å angstrom

Ab antibody

AKAP A kinase anchoring protein

ATP adenosine-5'-triphosphate

AU arbitrary units

Ca2+ calcium

CamKII calcium/calmodulin kinase II

CFP cyan fluorescent protein

Da dalton

E FRET efficiency

EC50 half maximal effective concentration

EGTA ethylene glycol tetraacetic acid

EPR electron paramagnetic resonance

ER endoplasmic reticulum

FRET fluorescence resonance energy transfer

GFP green fluorescent protein

HAX-1 HS-1 associated protein X-1

viii

HRC histidine rich calcium binding protein

IP3 inositol 1,4,5-trisphosphate

KD dissociation constant

NKA Na, K-ATPase

NMR nuclear magnetic resonance

PKA protein kinase A

PLB phospholamban

PLM phospholemman

PP1 protein phosphatase-1

SERCA sarco/endoplasmic reticulum calcium ATPase

SLN sarcolipin

SR sarcoplasmic reticulum

Tg thapsigargin tmPLB transmbrane phospholamban

Vmax maximum rate

YFP yellow fluorescent protein

ABSTRACT

With Fluorescence Resonance Energy Transfer (FRET), we are able to detect

changes in the structure and affinity of the PLB-SERCA regulatory complex in live cells.

Using this approach, we have detected a high level of PLB-SERCA interaction even at

Ca2+ concentrations known to fully relieve PLB inhibition of SERCA, suggesting that

dissociation is not required for relief of inhibition. We also detect no real-time change

in PLB-SERCA binding over the course of a single Ca2+ transient in paced myocytes. The

effect of Ca2+ on the PLB-SERCA interaction is best described as a reduced affinity with

no change in the structure of the complex. Even though we do not observe a change in

the structure of the PLB-SERCA complex with changes in calcium concentrations, we currently hypothesize that the reduced affinity is produced by a transition of PLB into an alternative binding site. We propose that the transmembrane domain of PLB could

move independently of its cytosolic domain. Our fluorescent probe is attached to the

cytosolic domain of PLB, which would explain our inability to detect a change in the

position of the transmembrane domain. My current work attempts to address this by

measuring the FRET between SERCA and a truncated PLB lacking the cytosolic domain.

CHAPTER I

INTRODUCTION

The sarco/endoplasmic reticulum Ca2+ ATPase (SERCA) is a responsible for

maintaining the Ca2+ gradient across the membrane of intracellular vesicles of all cells

(MacLennan & Kranias, 2003). This process is of particular importance in muscle cells in

which Ca2+ triggers contraction by binding troponin C, allowing myosin-actin crossbridge

formation (Gordon, Homsher, & Regnier, 2000). SERCA is responsible for decreasing

cytosolic Ca2+ concentrations necessary for muscle relaxation as well as loading the SR with Ca2+ necessary for further release events (Bers, 2008). SERCA is believed to

function through a series of coordinated structure changes mediated through the

binding of Ca2+ and ATP and subsequent hydrolysis of ATP (MacLennan, Toyofuku, &

Lytton, 1992; Toyoshima & Inesi, 2004). The two primary generalized structural

conformations of SERCA are described as the Ca2+ bound state, E1, and Ca2+ unbound

state, E2. The continual beating of the heart requires almost constant SERCA activity,

meaning proper function and regulation is of the utmost importance in the survival of

the heart.

In cardiac tissue, phospholamban (PLB) is the primary regulator of SERCA

activity, and has been shown to bind to and inhibit the activity of SERCA (MacLennan &

Kranias, 2003). This inhibition is achieved by decreasing the apparent affinity of Ca2+ for 1

SERCA. In other words, higher concentrations of Ca2+ are required to the same levels of

activity that occur in the absence of PLB. Therefore, Ca2+ can overcome in the inhibition

of SERCA by PLB, but the mechanism for this relief of inhibition remains unclear. PLB is

a key target of β-adrenergic signaling with phosphorylation partially relieving the inhibition of SERCA. Its name is derived from the Greek words for “phosphate” and “to receive.”

A general depiction of Ca2+ handling in cardiac myocytes is depicted in Fig. 1

(Bers, 2008). Membrane depolarization causes the opening of L-type Ca2+ channels,

causing an influx of Ca2+ and resulting in a Ca2+-induced Ca2+ release from the sarcoplasmic reticulum (SR) through the ryanodine receptor. Ca2+ accumulates in the

cytosol, triggering muscle contraction. SERCA is responsible for pumping the

accumulated Ca2+ back into the SR, reducing cytosolic concentrations and loading the SR with Ca2+ for future release. PLB can inhibit SERCA, modulating Ca2+ movement.

Two basic models have been used to explain the mechanism by which Ca2+

overcomes the inhibition by PLB. For the sake of discussion, I will refer to them as the

“Dissociation Model” and the “Subunit Model.” The Dissociation Model suggests that

PLB stabilizes a structural conformation of SERCA, thereby preventing its transition to an

alternative conformation capable of binding Ca2+. Ca2+ essentially competes with PLB

for binding to SERCA. Therefore, PLB must be displaced from SERCA for Ca2+ to bind and

the pumping cycle to proceed. The Subunit Model suggests that PLB merely slows the

Figure 1: Diagram of the basic calcium handling elements within cardiac myocytes. Membrane depolarization opens L-Type calcium channels initiating a calcium induced calcium release by the ryanadine receptor (RyR), elevating cytosolic calcium concentrations. This cytosolic calcium is then pumped back into the SR against its concentration gradient via SERCA, the activity of which can be inhibited by PLB.

on-rate of Ca2+, allowing the possibility that PLB can remain bound to SERCA throughout

its reaction cycle. Both models appear to be incomplete and do not account all data in

field. The primary goal of this dissertation is to address these inconsistencies and

propose an alternative, more inclusive model which we have named the “Domain

Displacement Model.”

In this dissertation, I will propose that PLB has an alternative binding site on

SERCA and that the binding of Ca2+ displaces PLB to this site rather than preventing any

interaction. This model is supported through my research as well as evidence from the

Na, K-ATPase and its regulator phospholemman, a protein system highly homologous to

PLB-SERCA. This is also supported by previously published work by our lab showing that

phosphorylated PLB can still bind SERCA. This model challenges the general concept

that relief of inhibition requires dissociation, but is not in conflict with the previous

research in the field.

Much of our understanding of SERCA function comes from steady state

biochemical assays. However, the contracting heart is by no means a steady state

environment, especially with respect to Ca2+. Ca2+ concentration is oscillating on a beat- to-beat basis facilitating contraction and relaxation. One of the primary questions of this dissertation is how PLB interacts with SERCA over the course of a single Ca2+

transient. Can the elevated Ca2+ from a single transient overcome the PLB inhibition of

SERCA? Or can the effect of elevated Ca2+ integrate over time?

The biochemical steps of SERCA activity have been well understood since the

1970’s. With the more recent detailed structural information coming from crystallographic studies, SERCA has provided a unique venue in the study of the structure-function relationship of complex proteins. In this dissertation, I will address how PLB is able to interact with various structural conformations of SERCA.

This dissertation will first start with a review of literature describing SERCA and

PLB themselves as well our current understanding of the regulatory mechanism. I will then address the conflicting approaches used to assess the PLB-SERCA interaction. I will then discuss how disregulation of SERCA can be involved in disease to underscore the importance of this functional interaction.

In the next section I will discuss the materials and methods employed for this dissertation. This will be followed by a description of an in cell SERCA activity assay which we developed to test the functionality of our fusion proteins. The next section

addresses the steady-state nature of the PLB-SERCA interaction addressing how PLB

interacts with SERCA in various conformational steps. In chapter VI, I show my research

that directly tests our newly proposed model of PLB-SERCA interaction. In chapter VII, I

attempt to show the dynamic nature of the PLB-SERCA interaction by making real-time

measurements. Finally, I will discuss the broader implications of my research and how

my results integrate with the findings of others.

I would also like to note that the bulk of work shown in sections IV, V and VII was

published this past October in the Journal of Biological Chemistry (Bidwell, Blackwell,

Hou, Zima, & Robia, 2011).

CHAPTER II

REVIEW OF RELATED LITERATURE

SERCA

P-type Ion Pumps

SERCA belongs to the P-type pump family of ATPases (Kühlbrandt, 2004;

Wuytack, Raeymaekers, & Missiaen, 2002). P-type pumps utilize ATP hydrolysis to

undergo a series of structural transitions, producing a reversibly phosphorylated state,

from which the family name is derived. Most family members have 10 transmembrane

helices (M1-10), though the greatest variation between members occurs in the transmembrane region, which determines ion specificity.

The family also shares three cytosolic or headpiece domains, the most homologous region in the family (Kühlbrandt, 2004; Wuytack et al., 2002). They are commonly called the N, P, and A domains (Nucleotide, Phosphorylation, and Actuator).

The ATP binding site is located on the N-domain. A conserved lysine is an integral component of the nucleotide binding site. The gamma-phosphate of that ATP is then transferred to an aspartate in a conserved motif, DKTGTLT, on the P-domain. The P- domain is also the most highly conserved domain within the family. The P-domain is also homologous to a class of bacterial enzymes demonstrating its general biochemical role and importance. 7

SERCA Isoforms

In vertebrates, 3 distinct genes encode for SERCA that produce at least 10

different isoforms through alternative splicing (Periasamy & Kalyanasundaram, 2007;

Wuytack et al., 2002). SERCA2b is the “housekeeping” isoform found in all cells, at least

at some low level. SERCA2a is the predominant isoform found in cardiac muscle and slow twitch skeletal muscle. SERCA1a is expressed in adult fast twitch skeletal muscle and has been the most highly studied isoform because of its abundance in and ease of collection from animal tissue. There are 6 SERCA3 isoforms (a-f) found in various tissues, with a, d, and f found at low levels in cardiac tissue. SERCA2a and SERCA1a are

84% identical with the SERCA3 isoforms being roughly 75% identical to either

SERCA1a/2a. Because of the high degree of homology, all SERCA isoforms are believed to have almost identical structures. All isoforms are capable of being inhibited by thapsigargin. SERCA1a and SERCA2a have similar affinities for Ca2+, but SERCA1a has a

faster catalytic velocity. SERCA2b has a higher Ca2+ affinity than either SERCA1a/2a, but

lower catalytic rate.

Reaction Scheme

SERCA activity was first observed when isolated SR membrane was shown to

have ATPase and Ca2+ uptake ability from skeletal muscle (Ebashi & Lipmann, 1962), and

later from cardiac myocytes (Hasselbach & Makinose, 1962). Evidence of a

phosphorylated protein intermediate during the reaction process provides the first

insight into the mechanism of SERCA function. In 1966, Jardetzkey postulated the

9 simple but still accepted model of ion transport where ions bind with differential affinity on opposing sides of the membrane (high on the binding or uptake side and low on the release or discharge side). This is coupled to a phosphorylation dependent conformational change closing the binding pocket on one side and opening it on the other (Jardetzky, 1966). Landgraf and Inesi were the first to demonstrate a conformational change of SR protein in response to ATP through EPR spin labeling

(Landgraf & Inesi, 1969). Tanford further proposed that the change in binding pockets and affinities could occur simultaneously through a simple movement of membrane helices altering amino acid side chain interactions that facilitate ion interaction

(Tanford, 1982).

In 1970, Inesi first described a hallmark of SERCA activity, the dependence of

Ca2+ on the rate of ATPase activity. At low concentrations where little Ca2+ is available to be pumped across the membrane, SERCA activity is low. Increasing Ca2+ increases

SERCA activity up to maximal level. Along with Inesi, many other groups begin to characterize the stepwise reaction scheme describing the SERCA reaction cycle

(Carvalho, de Souza, & de Meis, 1976; Coan & Inesi, 1976; Hasselbach, 1978; Kanazawa

& Boyer, 1973; de Meis & Vianna, 1979), on which a general consensus is reached by

1979, as depicted in Fig. 2. All steps were shown to be reversible, providing the possibility of ATP production. These concepts are still accepted today. A Ca2+ to ATP

Figure 2: Simplified SERCA reaction diagram. One ATP and two Ca2+ ions bind on the cytosolic face of the E1 conformation of SERCA, which initiates ATP hydrolysis and phosphate transfer. The phosphorylated SERCA then readily transitions to the E2 conformation, changing the Ca2+ binding sites from high affinity and cytosolic facing to low affinity and lumen facing. Ca2+ is then released to the ER/SR lumen, along with the release of ADP and PI, producing the free E2 conformation.

stoichiometry of 2:1 was first proposed in 1963 (Hasselbach & Makinose, 1963), but not

clearly described until 1980 (Inesi, Kurzmack, Coan, & Lewis, 1980). In the same study,

Inesi et al also demonstrate the cooperative binding of Ca2+. Spin labeling detected a

structure change after binding of the first Ca2+ ion which is presumed to increase the

binding affinity of the second ion.

SERCA Structure

SERCA was first isolated from rabbit skeletal muscle in 1970 (MacLennan, 1970).

Structural information first came from proteolysis analyses where trypsin digestion of

SERCA produced various fragments and subfragments. ATP and Ca2+ binding could be

associated with these specific fragments. In 1980, partial amino acid sequences were determined for the cytosolic fragments (Allen, Trinnaman, & Green, 1980). In 1985,

SERCA1a was cloned for the first complete sequence (MacLennan, Brandl, Korczak, &

Green, 1985). The following year SERCA2a was cloned and sequenced (Brandl, Green,

Korczak, & MacLennan, 1986). From these sequences, the MacLennan group deduced

the topology and secondary structure of SERCA, predicting the helical transmembrane

domains and the globular cytosolic domains.

SERCA was first visualized with electron microscopy identifying 40 Å diameter

surface particles on skeletal muscle microsomes (Martonosi, 1968), and later identified

by freeze fracture (Baskin & Deamer, 1969; Deamer & Baskin, 1969). The first low

resolution crystal structure from X-ray diffraction (14 Å) showed a large cytosolic

headpiece attached to a narrow transmembrane region (Toyoshima, Sasabe, & Stokes,

1993). With enhancement of the resolution to 8 Å, the 10 transmembrane helices could be visualized and could be associated to the appropriate primary structure sequence

(Zhang, Toyoshima, Yonekura, Green, & Stokes, 1998). After developing conditions sufficient to grow crystals suitable for X-ray diffraction, a high resolution (2.6 Å)

structure was determined capable of resolving the 3 cytosolic domains (Toyoshima,

Nakasako, Nomura, & Ogawa, 2000). High resolution structures have subsequently

been determined under a variety of conditions and have been associated with

intermediate conformations of SERCA (Toyoshima & Inesi, 2004; Toyoshima & Nomura,

2002). Fig. 3 depicts a few of the proposed intermediate conformations showing

potential large movements of the cytosolic headpiece.

Structure Function Relationship

Integrating the enzymatic and structural data has been a fundamental goal

within the field. After cloning SERCA, the MacLennan lab performed extensive

mutagenesis studies investigating the role of individual residues (MacLennan, 1990).

Many loss-of-function mutants retained the capability of undergoing partial reactions,

enabling investigation into the particular steps affected by alteration of these residues.

This approach clearly identified Asp351 as the site of phosphoryl transfer (Maruyama et

al., 1989). This approach also lead to the preliminary identification of the Ca2+ binding

sites in the transmembrane region (Clarke et al., 1989; Vilsen, Andersen, Clarke, &

MacLennan, 1989) and of the ATP binding site (Clarke, Loo, & MacLennan, 1990). These

binding sites have been subsequently confirmed by high resolution X-ray crystallography

Figure 3: Proposed SERCA structural transitions. SERCA is believed to undergo conformational transitions upon binding of calcium and ATP, potentially producing large movements of the cytosolic headpiece. Images represent various crystal structures produced under conditions thought to be associated with reaction step intermediates.

(Toyoshima & Inesi, 2004; Toyoshima & Nomura, 2002; Toyoshima et al., 2000). These

studies demonstrate that ATP binding and hydrolysis occur in the cytosolic headpiece far

from the Ca2+ binding sites in the transmembrane region. Therefore the

transmembrane reorientation necessary for Ca2+ transport is linked to the required

chemical energy through long-range protein interactions.

The various high resolution crystal structures have shown many distinct

conformations of SERCA (Toyoshima & Inesi, 2004; Toyoshima & Nomura, 2002;

Toyoshima et al., 2000). The conditions under which these crystals were produced can

be associated with endpoints of the enzymatic half reactions. Therefore, these

conformations are thought to correspond to the various enzymatic states of the SERCA

reaction cycle (Fig. 3). Initial results suggested that the stepwise binding of Ca2+ and ATP induce large scale conformational changes in the cytosolic headpiece domains. These

conformational shifts in the headpiece were thought to produce the mechanical force

necessary to rearrange the transmembrane helices shifting the orientation of the Ca2+

binding sites. However, more recent crystal structures suggest that the structural changes are much smaller and more subtle than previously believed (Jensen, Sørensen,

Olesen, Møller, & Nissen, 2006; Sørensen, Møller, & Nissen, 2004). This concept of smaller structural movements is also supported by FRET distance measurements between probes attached to different cytosolic domains (Winters, Autry, Svensson, &

Thomas, 2008).

Phospholamban

Discovery

The existence of PLB was first observed with the PKA dependent stimulation of

the ATPase dependent Ca2+ uptake in cardiac SR vesicles (Kirchberger, Tada, Repke, &

Katz, 1972; Tada, Kirchberger, Repke, & Katz, 1974). These vesicles were then shown to be radiolabeled by P32 upon stimulation by PKA (Kirchberger, Tada, & Katz, 1974). This

radiolabeling was found to be targeted to a 22 kDa protein which was shown to be a

modulator of SERCA (Tada, Kirchberger, & Katz, 1975). This phosphoprotein was subsequently termed phospholamban coming from the Greek terms “to receive” and

“phosphate” (Kirchberger, Tada, & Katz, 1975).

This 22 kDa protein was shown to be capable of dissociating into lower

molecular weight components upon boiling or treatment with Triton-X (Jones, Besch,

Fleming, & Mcconnaughey, 1979; Lamers & Stinis, 1980; Le Peuch, Haiech, & Demaille,

1979; Will, Levchenko, Levitsky, Smirnov, & Wollenberger, 1978). Freezing of the

sample was shown to cause reassociation of the lower-molecular weight forms back to

the 22 kDa. These observations suggested that phospholamban could exist as an

oligomer. The lower molecular form was eventually determined to be a 6 kDa peptide

that could be independently phosphorylated by PKA (Bidlack & Shamoo, 1980; Bidlack,

Ambudkar, & Shamoo, 1982). After improved purification techniques, the high molecular weight form was shown to be a 30 kDa pentamer of composed of indistinguishable monomers which could each be independently phosphorylated (Fujii,

Kadoma, Tada, Toda, & Sakiyama, 1986; Wegener & Jones, 1984; Wegener, Simmerman,

Liepnieks, & Jones, 1986). The full sequence of PLB was determined after it was cloned by the Tada lab (Fujii et al., 1987).

Structure

PLB is a 52 amino acid integral membrane protein that is comprised of 3

domains, Ia (residues 1-20), Ib (residues 21-30), and II (residues 31-52) (Bhupathy, Babu,

& Periasamy, 2007). Proteolysis studies first suggested the existence of 2 domains, a

protease sensitive cytosolic domain (1a/b) and protease insensitive transmembrane

domain (II) (Wegener et al., 1986). Circular dichroism experiments further suggested

the existence a cytosolic (Ia) and transmembrane (II) alpha helices separated by and

unstructured linker or hinge domain (1b) (Simmerman, Lovelace, & Jones, 1989; Terzi,

Poteur, & Trifilieff, 1992). These observations were consistent with sequence analysis

proposing likely alpha helical and hydrophobic/hydrophilic regions (Fujii et al., 1986,

1987). These observations are also consistent with NMR structures of PLB (Traaseth et

al., 2009) and molecular dynamics simulations (Paterlini & Thomas, 2005).

A fundamental aspect of phospholamban is that it can exist in a monomer as

well as a pentamer (MacLennan & Kranias, 2003; Simmerman & Jones, 1998). The alpha

helical structure of the PLB transmembrane region generates heptad repeats. Because

this pattern, every 3 to 4 residues of the helix align, organizing leucines and isoleucines

along one side or face PLB. Mutagenesis targeting these residues destabilized pentamer

formation whereas mutations along the opposite face of PLB did not alter pentamer

stability. This suggests that the leucine zipper formation found within the PLB structure

is critical for pentamer assembly (Simmerman, Kobayashi, Autry, & Jones, 1996). This

conclusion has been supported by subsequent NMR studies (Verardi, Shi, Traaseth,

Walsh, & Veglia, 2011). Mutations at 3 cysteines within the transmembrane domain

(residues 36, 41, and 46) have been shown to destabilize PLB pentamer formation (Fujii,

Maruyama, Tada, & MacLennan, 1989), but disulfide linkages were not observed and are believed be unnecessary for pentamer stability (C. B. Karim, Stamm, Karim, Jones, &

Thomas, 1998; Simmerman, Collins, Theibert, Wegener, & Jones, 1986). Replacement of cysteines with an isosteric amino acid (α-amin-n-butyric) was shown to have no affect

on pentamer stability (C. B. Karim et al., 2001). This demonstrates that the steric

properties of the cysteines, not their reactivity, contribute pentamer stability.

Because PLB can exist as a pentamer, monomer, and monomer bound to SERCA,

PLB exists in a coupled equilibrium depicted in Fig. 4. This concept was first

demonstrated when mutations that make PLB more monomeric were shown have

enhanced inhibitory potential and are considered “superinhibitors” (Y Kimura, Asahi,

Kurzydlowski, Tada, & MacLennan, 1998; Y Kimura, Kurzydlowski, Tada, & MacLennan,

1997). These observations are also consistent with work in our lab showing that a

superinhibitory PLB mutant enhances FRET between PLB and SERCA while diminishing

PLB pentamer FRET to an equivalent degree (Kelly, Hou, Bossuyt, Bers, & Robia, 2008).

These studies suggest that themonomer is the active inhibitor of the pump, with the

pentamer being an inactive reservoir, incapable of interacting with SERCA. Through

Figure 4: Coupled equilibrium of PLB binding. PLB can exist as a pentamer, free monomer, and monomer bound to SERCA. PLB readily transitions between pentamer and free monomer (KD1) and between free monomer and monomer bound to SERCA (KD2). Therefore changes in pentamer binding will alter the binding of PLB to SERCA.

mass action, changes in pentamer binding will alter binding between PLB and SERCA, thereby altering the ability of PLB to inhibit SERCA.

X-ray crystal studies have suggested the possibility that the pentamer could

directly interact with SERCA (Stokes, Pomfret, Rice, Glaves, & Young, 2006). However,

this inconsistent with studies from our lab showing a 1:1 stoichiometry in PLB-SERCA

FRET (Kelly et al., 2008). Because this interaction could be an artifact of crystallization,

further study is needed for its validation.

Phosphorylation

PLB is the target of β-adrenergic signaling through PKA phosphorylation of Ser16

and CamKII phosphorylation of Thr17 (Fujii et al., 1989; Simmerman et al., 1986). Thr17

has also been shown be phosphorylated independently of Ser16 in a frequency

dependent manner likely through enhanced CamKII activity (Hagemann et al., 2000;

Mundiña-Weilenmann, Vittone, Ortale, de Cingolani, & Mattiazzi, 1996).

Phosphorylation of Ser16 and Thr17 can occur independently in vitro (Davis, Schwartz,

Samaha, & Kranias, 1983; Jackson & Colyer, 1996; Kranias, 1985). However, during in

vivo studies, phosphorylation of Ser16 appears to be a prerequisite for phosphorylation

of Thr17 through the β-adrenergic pathway (Chu et al., 2000; Luo et al., 1998). Protein

phosphatase-1 (PP1) has been shown to dephosporylate PLB at both sites which decreases SERCA activity (Kranias & Di Salvo, 1986; MacDougall, Jones, & Cohen, 1991;

Nicolaou, Hajjar, & Kranias, 2009).

These phosphorylations can independently relieve PLB inhibition of SERCA

thereby enhancing its activity by restoring the Ca2+ sensitivity of SERCA (MacLennan &

Kranias, 2003). In some early studies, it was suggested that the Ser16 and Thr17

phosphorylations have an additive effect on SERCA activity (Kranias, 1985; Le Peuch et

al., 1979; Raeymaekers, Hofmann, & Casteels, 1988), but these studies do not directly

assess phosphorylation efficiency of their experimental conditions. Further studies

demonstrating full phosphorylation show no additive effect suggesting that the

previously observed additive effect may have been due to incomplete phosphorylation

of PLB (Colyer & Wang, 1991; Jackson & Colyer, 1996).

Mechanism of PLB Inhibition of SERCA

SERCA activity increases with Ca2+ concentration up to a maximum level with a

Hill function relationship (Fig. 5, black). PLB is known to reduce the apparent Ca2+

affinity of SERCA, producing a rightward shift in the SERCA activity curve (Fig. 5, red).

Higher Ca2+ concentrations are then needed to achieve the same levels of activity observed in the absence of PLB. In other words, increases in Ca2+ concentration can

overcome the inhibition by PLB. Phosphorylation of PLB relieves the inhibition on SERCA by partially restoring Ca2+ sensitivity, represented as a leftward shift in the SERCA

activity curve (Fig. 5, blue). The process of this relief of inhibition is not fully understood

and is the focus of this dissertation.

SERCA is believed to exist in various conformational states. Early evidence comes from protein digestion studies, where changes in tertiary structure block or allow

Figure 5: Simulated example of SERCA activity assay. SERCA activity, which can be measured by ATPase activity or calcium uptake, increases with increasing cytosolic calcium concentration in a sigmoidal manner up to a maximal level (black). The presence of PLB causes a decreased apparent calcium affinity represented by a rightward shift in the activity curve (red). Higher cytosolic calcium concentrations are needed to achieve the same SERCA activity. Phosphorylation of PLB shifts this curve back to left (blue).

access to cut sites. This approach identified 2 primary SERCA conformations: the E1

Ca2+ bound state and the E2 Ca2+ unbound state. More recent X-ray structures have

been determined for SERCA under various conditions, corresponding to more conformational states. These conformational states are believed to correspond to

reaction cycle intermediates.

Dissociation Model

The prevailing view favors a model where PLB binds the E2 Ca2+ free state of

SERCA. Interaction with PLB stabilizes this conformation thereby blocking transition to

the E1 conformation by increasing the free energy barrier. Inability to transition to a

conformation capable of binding Ca2+ terminates the reaction cycle of SERCA. PLB must

then completely dissociate from SERCA for transition to occur and activity to be

restored. In this model, binding of PLB and Ca2+ to SERCA is mutually exclusive. I will refer to this model as the “Dissociation Model” which is depicted in Fig. 6A.

The Dissociation Model is well supported by chemical crosslinking studies from

Jones group. In an initial study, a mutant which introduced a Cys at residue 30 of PLB

(N30C) was coexpressed in insect vesicles (Jones, Cornea, & Chen, 2002). Addition of a homobifunctional thiol probe, bismaleimidohexane, was shown to chemically crosslink the introduced Cys of N30C to Cys318 of canine SERCA2a. This demonstrated that

residue 30 of PLB was in close proximity to residue 318 of SERCA. This chemical

crosslinkage could be disrupted through increases in Ca2+ concentration, exhibiting an

2+ EC50 of 130 nM. At high Ca concentrations, crosslinking was completely abolished.

A B

Figure 6: Models of PLB inhibition of SERCA. (A) The Dissociation Model suggests that PLB selectively binds the E2 calcium unbound state of SERCA preventing transition to the E1 conformation capable of binding calcium. PLB must then completely dissociate from SERCA for function to be restored. (B) The Subunit Model suggests that PLB binds all conformations of SERCA, and merely slows the binding of calcium and transition to the E1 conformation. Evidence for both models stems from assessment of the ability to bind the E1 conformation (orange boxes), produced by high calcium conditions, with different methods favoring different models.

Further studies using a similar approach showed that residue 27 of PLB was in

close proximity to residue 328 of SERCA, and that residues 45-52 of PLB were in close

proximity to residue 89 of SERCA. Crosslinking at these sites was also completely

2+ abolished by increases in Ca concentration, with EC50 values ranging from 100 to 900

nM (Chen, Akin, Stokes, & Jones, 2006; Chen, Stokes, & Jones, 2005).

In a recent study, the Jones group has generated a method to crosslink endogenous PLB and SERCA2a from isolated human cardiac tissue, generating results consistent with their previous studies (Akin, Jones, 2012). This is an advancement over

previous studies since measurements are made with endogenously expressed proteins

in the most physiologically relevant membrane environment. Any accessory and/or

anchoring proteins would also be intact within this system. Also, neither protein has

crosslinking sites introduced through a mutation that could alter proper function and

interaction.

These studies show the disruption of PLB-SERCA binding at multiple points of contact by elevated Ca2+ concentrations. While some residual level of crosslinking (5%

of control) can be observed even at saturating concentrations, the Jones group states

Ca2+ completely abolishes crosslinking. The Jones group then concludes that PLB is able

to compete with Ca2+ for its binding site on SERCA, and that complete dissociation of

PLB and SERCA is required for restoration of SERCA activity.

Their approach and conclusions are supported by additional studies exploring

the effects of PLB phosphorylation on PLB-SERCA crosslinking. Much like elevated Ca2+

concentration, phosphorylation of PLB is known to overcome SERCA inhibition by PLB.

Also like elevated Ca2+, they demonstrated that PKA phosphorylation of PLB prevented

crosslinking between PLB and SERCA (Chen, Akin, & Jones, 2007). This further supports

their previous conclusion that relief of PLB inhibition is accomplished through the

unbinding of PLB from SERCA.

The Dissociation Model is also supported by immunoprecipitation studies by the

MacLennan group. They first demonstrated that inhibitory ability of various mutants

(loss and gain of function) correlates with their ability to co-immunoprecipitate with

SERCA (Asahi, Kimura, Kurzydlowski, Tada, & MacLennan, 1999). Subsequently, they

demonstrated that Ca2+ can also disrupt the ability of PLB to co-immunoprecipitate with

SERCA (Asahi, McKenna, Kurzydlowski, Tada, & MacLennan, 2000). These studies

demonstrate that relief of inhibition correlates to loss of PLB-SERCA interaction, in

support of the Jones group conclusions.

Despite the clear and well characterized results of the crosslinking and

immunoprecipitation assays, the approaches and subsequent conclusions have limitations. While the crosslinking assay is highly sensitive and specific, the close proximity of targeted residues on PLB and SERCA are only assessed. Even subtle changes in the position and/or orientation of PLB in relation to SERCA could prevent the crosslinking. Small rotations in the transmembrane helices of SERCA could prevent access of the chemical crosslinkers to the target residues. These factors demonstrate how crosslinking could be abolished without significant dissociation of the PLB-SERCA

complex. Furthermore, even at saturating Ca2+ concentrations, a low level of crosslinking is observed in many cases. This observation is not in agreement with complete dissociation.

Despite agreement between crosslinking and immunoprecipitation on the effect of Ca2+ on the PLB-SERCA interaction, the observations from these approaches differ

under other conditions. A SERCA ligand, thapsigargin, as well as phosphorylation of PLB

have been shown to block the PLB-SERCA crosslinking (Chen et al., 2006, 2005), but

neither affects PLB-SERCA co-immunoprecipitation (Asahi et al., 2000).

The largest complication with immunoprecipitation is that the membrane is

perturbed by detergents prior to the detection of the interaction between PLB and

SERCA. Immunoprecipitation between membrane bound proteins like PLB and SERCA

will be highly susceptible to variations in experimental conditions. Furthermore,

antibodies targeted to PLB, such as those used to pull down the PLB-SERCA complex in

the co-immunoprecipitation experiments, have themselves been shown to block PLB-

SERCA crosslinking (Chen et al., 2007) and relieve PLB inhibition of SERCA (Cantilina,

Sagara, Inesi, & Jones, 1993; Y Kimura et al., 1991; Morris, Cheng, Colyer, & Wang, 1991;

Sham, Jones, & Morad, 1991). These factors may explain the inconsistent results

between the two approaches. These factors also argue why immunoprecipitation may

not be an effective strategy to investigate PLB-SERCA binding and therefore may not

lend the crosslinking studies sufficient corroborating support.

The Dissociation Model is described by the competitive binding of PLB and Ca2+

for SERCA. In competitive inhibition, Ca2+ binding (activity) can always be overcome by

further increases in the concentration of inhibitor. However, this is inconsistent with

studies measuring the dose response of PLB. The maximal inhibition of SERCA, indicated

by decreased apparent Ca2+ affinity, is achieved by a 2.6 fold increase of PLB expression

(Brittsan, Carr, Schmidt, & Kranias, 2000). Further overexpression of PLB does not

enhance the inhibition of SERCA, as would be predicted if PLB competed with Ca2+.

Furthermore, crosslinking is never completely abolished even at saturating

concentrations inconsistent with the Jones group conclusions and the concept of

competitive inhibition.

Subunit Model

In contrast to the Dissociation Model, the Subunit Model has been proposed

suggesting that PLB can remain bound to SERCA throughout its reaction cycle as

depicted in Fig. 6B. The decrease in apparent Ca2+ affinity is achieved by slowing a rate limiting step of the reaction cycle, likely the on-rate of Ca2+. This is consistent with the

observation that high Ca2+ can overcome inhibition by PLB.

Support for this model first comes from a study showing that a PLB antibody

enhances Ca2+ transport by SERCA in isolated cardiac vesicles (Cantilina et al., 1993).

This antibody produces an increased formation of the autophosphorylated SERCA

intermediate (E1-P) as would be expected with increased activity and turnover of the pump. However, this antibody was unable to alter the Ca2+ binding in the absence of

ATP at equilibrium. This suggests that SERCA inhibition by PLB is not achieved by

lowering the actual affinity of Ca2+ for the pump. The decrease in apparent Ca2+ affinity

in the presence of PLB is then achieved by a decreased turnover of the SERCA pump.

Decreased turnover is explained by a slowed transition from the E2 to E1 conformations

produced by a slower Ca2+ on-rate.

These equilibrium binding experiments were performed in the absence of ATP,

which has been viewed as a flaw. Since ATP is a critical and necessary for SERCA

function, it could then alter the Ca2+ affinity and control how PLB might affect this

equilibrium. The conclusion that PLB decreases the on-rate of Ca2+ has then been disputed on this assertion (Jones et al., 2002). However, no study has described equilibrium Ca2+ binding in the presence of ATP in the nearly 20 years since Cantilina et

al was published. This is likely due to the complexity of this seemingly simple

experiment. The presence both Ca2+ and ATP will induce SERCA cycling and Ca2+

transport, thereby preventing equilibrium from being achieved.

Equilibrium Ca2+ binding experiments have also been performed in using purified

proteins in a reconstituted membrane (Hughes, Starling, Sharma, East, & Lee, 1996).

Consistent with Cantilina et al, the results show no effect of PLB on Ca2+ binding at

equilibrium. However, in contrast to Cantilina et al, this study shows no change in the rate of the SERCA auto-phosphorylation, and suggests that PLB disrupts the off-rate of

Ca2+.

The Subunit Model is also supported by a study measuring FRET between PLB

and SERCA (Mueller, Karim, Negrashov, Kutchai, & Thomas, 2004). High levels of

energy transfer were detected between purified, fluorescently-labeled PLB and SERCA in

a reconstituted membrane, even at saturating Ca2+ concentrations. A small but significant increase in the distance between the fluorescent probes was detected by the elevation of Ca2+ concentration. This suggests that relief of inhibition occurs through a

structural rearrangement in the PLB-SERCA complex, rather than from complete

dissociation of PLB from SERCA.

One criticism of this study was the use of purified protein in a reconstituted

membrane. Being in such a non-physiological system could disrupt proper interaction

between PLB and SERCA, thereby producing anomalous results. The membrane

composition and high relative concentration of protein could produce nonspecific

binding. Also, the fluorescent probes themselves could induce aggregation of PLB and

SERCA. However, this argument neglects that the distance estimations between PLB

and SERCA are consistent with the predicted binding location. Furthermore, activity

assays demonstrate that elevated Ca2+ can overcome PLB inhibition in this system.

Since functional regulation occurs, it suggests that the PLB-SERCA binding detected are

specific and not anomalous. Also, if the fluorescent probes were inducing nonspecific

aggregation, this would reduce the ability of PLB to regulate SERCA, which is not

observed. Labeled and unlabeled PLB inhibit SERCA activity to the same degree.

Structure-Function Relationship of PLB inhibition

PLB is only expressed in cells expressing SERCA2a and not in cells expressing

SERCA1a (MacLennan & Kranias, 2003; Tada & Toyofuku, 1998). However, the

interaction between PLB and SERCA1a has been extensively studied despite not being

physiological binding partners. This interaction has been studied due to the wealth of

structural information of SERCA1a and the suspected high degree of structural and

functional similarities between the two isoforms (Asahi et al., 1999; Kühlbrandt, 2004;

Morth et al., 2007). Furthermore, PLB has been shown to inhibit SERCA1a and SERCA2a to the same extent (Toyofuku, Kurzydlowski, Tada, & MacLennan, 1993). PLB also shows

the ability to crosslink both isoforms in a similar manner (Chen et al., 2006, 2005; Jones

et al., 2002; Toyoshima et al., 2003). The study of the PLB-SERCA1a therefore aids in the

better modeling of the PLB-SERCA interaction. Despite these extensive similarities, our

lab has shown that PLB has a higher affinity for SERCA1a compared to SERCA2a (Hou &

Robia, 2010). While this observation does not discount the other similarities, it is an

important consideration when interpreting the results from SERCA1a.

The hallmark of PLB inhibition of SERCA is a reduced apparent affinity of Ca2+

(increased KCa as in Fig. 5) which is generally thought to occur through transmembrane

interactions of the two proteins (MacLennan & Kranias, 2003). This concept is

supported by studies showing that a truncated PLB lacking the cytosolic domain was

capable of reducing Ca2+ sensitivity of SERCA to the same degree as full length PLB

(Hughes et al., 1996; Y Kimura, Kurzydlowski, Tada, & MacLennan, 1996). Evidence for

the role of transmembrane interactions also comes from a homolog of PLB, sarcolipin

(SLN). The transmembrane regions of PLB and SLN have a high degree of similarity, but

SLN lacks a comparable cytosolic domain ( a Odermatt et al., 1997). Despite the lack of

cytosolic domain, SLN displays the ability to reduce the apparent affinity for Ca2+ of

SERCA like PLB (MacLennan, Asahi, & Tupling, 2003; a Odermatt et al., 1998).

While the role of the transmembrane region in the inhibition of SERCA is well described, the role of the cytosolic domain of PLB is less understood. PLB can inhibit both SERCA1a and SERCA2a, but is unable to inhibit SERCA3 (Kimura et al., 1996).

However, a truncated PLB lacking the cytosolic domain is able to inhibit SERCA3 indicating a critical role for the cytosolic domain. A region in the nucleotide binding domain of SERCA3 was identified that prevented regulation by PLB (Toyofuku et al.,

1993; Toyofuku, Kurzydlowski, Tada, & MacLennan, 1994). Specifically, when the

SERCA3 sequence of Gln-Gly-Glu-Gln-Leu401 was converted to the corresponding

SERCA2a sequence of Lys– Asp–Asp–Lys–Pro401, SERCA3 gained the ability to be

regulated by PLB. However, a specific role for this sequence has not been determined,

but these findings point to at least some regulatory role of the PLB cytosolic domain that

does not directly confer inhibition.

A direct inhibitory function of the cytosolic domain has been suggested and

investigated, though the findings have not been consistent. Many studies have

examined the potential effect in isolated cardiac vesicles. Early studies suggested that

phosphorylation of the cytosolic domain of PLB could enhance the maximal catalytic

rate (Vmax) of SERCA in isolated cardiac SR vesicles (Adunyah, Jones, & Dean, 1988).

However, later studies disputed these findings showing that phosphorylation of PLB had

no effect on Vmax in isolated cardiac vesicles ( a Mattiazzi, Hove-Madsen, & Bers, 1994;

M A Movsesian, 1992; Matthew A Movsesian, Colyer, Wang, & Krall, 1990). An isolated

peptide corresponding to the PLB cytosolic domain (residues 1-31) was shown to

2+ decrease the Vmax of SERCA with no impact on Ca affinity (Sasaki, Inui, Kimura, Kuzuya,

& Tada, 1992). These results were supported by studies showing that the removal of

the PLB cytosolic domain through tryptic digestion (Lu & Kirchberger, 1994; Lu, Xu, &

Kirchberger, 1993) caused an increase in the Vmax of SERCA in isolated cardiac vesicles.

Exposure of an antibody targeted to the cytosolic domain of PLB to isolated vesicles was

shown to decrease Vmax in one study (Antipenko, Spielman, Sassaroli, & Kirchberger,

1997) while having no effect on Vmax in another (Cantilina et al., 1993). Variations in

vesicle preparation were shown to impact detection of a Vmax change potentially

explaining the differing results (Antipenko, Spielman, & Kirchberger, 1999).

Rather than using isolated vesicles, many labs studied the role of the cytosolic

domain by expressing PLB an SERCA in heterologous cells or reconstituted them in a purified membrane. When SERCA was reconstituted into a membrane with and without

PLB, the expected change in KCa was observed with no change in the Vmax parameter

(Reddy et al., 1995; Reddy, Jones, Pace, & Stokes, 1996). These studies also observed no change in SERCA activity in response to the isolated peptide corresponding to residues

1-31 of PLB. These studies were also supported by an analogous study expressing PLB

and SERCA in HEK cells also showing no change in the Vmax parameter (A. Odermatt,

Kurzydlowski, & MacLennan, 1996). These studies have been disputed by another group

showing that the isolated PLB cytosolic domain can decrease Vmax in a reconstituted

system and that a truncation of PLB removing the cytosolic domain increases Vmax

(Hughes, East, & Lee, 1994; Hughes et al., 1996; A. P. Starling, Hughes, Sharma, East, &

Lee, 1995). They also demonstrate that experimental conditions used to reconstitute

PLB and SERCA have an effect on the observation of a Vmax change (A. P. Starling et al.,

1995), much like Antipenko et al describes with vesicle preparations (Antipenko et al.,

1999).

While phosphorylation of PLB is known to favor pentamer formation, mass

action equilibrium binding changes may not fully explain the relief of PLB inhibition

through phosphorylation. Phosphorylation of PLB has been shown to prevent chemical

crosslinking to SERCA (Chen et al., 2007), which is in agreement with the idea of mass action changes. However, our lab has shown using FRET that phosphomimetic mutants of PLB can interact with SERCA, though with a diminished affinity and altered structure compared to wild-type PLB (Hou, Kelly, & Robia, 2008). This structure change in the

regulatory complex might also be linked to relief of inhibition. EPR amd NMR studies have been used to show that phosphorylation of PLB introduces disorder into the alpha-

helix of the cytosolic domain (Metcalfe, Traaseth, & Veglia, 2005) and that this

phosphorylation does not alter the affinity of the transmembrane domain for SERCA

(Traaseth, Thomas, & Veglia, 2006). To explain the PLB-SERCA inhibitory mechanism as

well as the effect of phosphorylation on this mechanism, the Veglia group proposes an

allosteric model of PLB interaction (Traaseth et al., 2006; Zamoon, Nitu, Karim, Thomas,

& Veglia, 2005). They propose that PLB is in equilibrium between two states, an

ordered T state and a disordered R state. In both states, the transmembrane domain of

PLB is capable of interacting with SERCA, but only the T state can inhibit SERCA.

Unphosphorylated PLB favors the order T state and is therefore inhibitory, while

phosphorylated PLB favors the disordered R state and is unable to inhibit.

Predicted PLB Binding Location

A structure of a co-crystal between SERCA1a and PLB monomer has been

determined using cryo-electron microscopy, though not to sufficient resolution to determine the binding location of PLB (H. S. Young, Jones, & Stokes, 2001). Scanning

alanine mutagenesis of SERCA1a indentified several residues on helix M6 critical for PLB

inhibition suggesting that this helix is a likely point of contact (M Asahi et al., 1999). This observation was consistent with later molecular modeling positioning the transmembrane domain of PLB in close proximity to helix M6 (Hutter et al., 2002).

Using distance constraints from PLB-SERCA crosslinking studies, subsequent modeling predicts that the transmembrane domain of PLB is positioned in a groove comprised of helices M2, M4, and M6 of SERCA (Fig. 7A) (Toyoshima et al., 2003). This study utilized crosslinking results from both SERCA1 and SERCA2a since the distance constraints were compatible. The consistency in the crosslinking studies also demonstrates that PLB likely interacts with SERCA1a and SERCA2a in a similar manner.

A B

Figure 7: Predicted regulator binding locations. (A) Crosslinking studies predict PLB (red) to interact with SERCA in a binding pocket described by helices M2, M4, and M6 (blue). (B) Co-crystalization has provided evidence of that PLM (red) interacts with NKA with a more exterior helix M9 (green).

While the positioning of the transmembrane domain of PLB in relation to SERCA

is well supported, less is known about the binding location of the cytosolic domain.

Initial crosslinking studies predicted that Lys3 of PLB interacts with either Lys397 or Lys400

of SERCA (James, Inui, Tada, Chiesi, & Carafoli, 1989). This observation is supported by a

study showing that the sequence Lys– Asp–Asp–Lys–Pro–Val402 in SERCA is critical for

PLB regulation (Toyofuku et al., 1994). These results have been used to model the

position of the cytosolic domain. However, subsequent crosslinking studies have been

unable to provide corroborating evidence calling the initial study into question (Chen,

Stokes, Rice, & Jones, 2003).

Insights into the PLB-SERCA regulatory complex structure may also come from homologous proteins, Na, K-ATPase (NKA) and phospholemman (PLM). NKA is a P-type ion pump like SERCA and is thought to function in a similar manner (Kühlbrandt, 2004;

Wuytack et al., 2002). The X-ray crystal structure of NKA shares a high degree of resemblance to that of SERCA (Morth et al., 2007) further supporting their potential functional similarities. PLB and PLM are structurally homologous having a single transmembrane domain and a cytosolic domain capable of being phosphorylated. Much like PLB and SERCA, PLM inhibits NKA by reducing ion affinity (Bers & Despa, 2009).

Taken together, these observations suggest a high likelihood that PLB and PLM interact with their partners in a similar way. Unlike PLB and SERCA, a high resolution crystal structure has been determined for PLM and NKA (Shinoda, Ogawa, Cornelius, &

Toyoshima, 2009). Rather than showing interaction in the M2/M4/M6 binding groove

37 predicted for PLB, PLM interacts with the more exterior helix M9 of NKA (Fig. 7B). Due to the similarities between the proteins, this suggests the possibility of an alternative binding site on SERCA for PLB. However, this observation has not been supported through alternative approaches like crosslinking or mutagenesis, so it could be an artifact of crystallization. While unlikely due to the high degree of similarity, it is also possible that PLB and PLM confer inhibition through alternative means and binding sites.

Potential Interacting Proteins

While PLB and SERCA are a clear functional unit, one must also consider the possibility that additional proteins may interact with PLB and SERCA in a larger protein complex. Therefore, additional proteins may alter SERCA function and the ability of PLB to inhibit in a physiological setting. This has not been a heavily studied area, but there are several proteins of interest that have been identified. As with many proteins that can be phosphorylated by PKA, PLB has been found in complex with an A kinase anchoring protein (AKAP) and has been shown to affect the phosphorylation status of

PLB in vivo (Lygren et al., 2007; Manni, Mauban, Ward, & Bond, 2008). The anti- apoptotic HS-1 associated protein X-1 (HAX-1) has been shown to interact with SERCA and regulate its expression (Vafiadaki et al., 2009). HAX-1 overexpression has been shown to reduce contractility of cardiac myocytes, and this effect was abolished in the

PLB knockout mouse (Zhao et al., 2009). HAX-1 was identified as a binding partner of

PLB using the yeast two-hybrid method (Vafiadaki et al., 2007) and was shown to

destabilize PLB pentamer formation (Vafiadaki et al., 2009). Histidine rich Ca2+ binding

protein (HRC) is an SR Ca2+ buffering protein that has been shown to have a regulatory

role in Ca2+ release via triadin and the ryanadine receptor. Additionally, HRC has been

shown to interact with SERCA and alter Ca2+ sequestration (D. a Arvanitis et al., 2007;

Pritchard & Kranias, 2009). This dual role for HRC suggests a potential crosstalk role

between Ca2+ uptake and release.

Disease and Therapy

Physiological Principles of Disease

SERCA activity maintains SR Ca2+ load which is a primary determinant of Ca2+

release and contractility (Bers, 2008). Decreases in SERCA activity would then seemingly be associated with reduced contractility, a general characteristic of heart failure. In many observed cases, decreased SERCA activity has indeed been related with heart failure (Hasenfuss, 1998; Phillips et al., 1998; Wickenden et al., 1998). In most but not all cases, decreased SR Ca2+ content has been observed with heart failure. In many

instances, this decreased SR Ca2+ content has been connected to depressed SERCA

function. However, increased activity of the Na, Ca-exchanger has also been implicated

in reduced SR content when no change in SERCA function was observed (Bers, 2008;

Hobai & Rourke, 2000; Houser, Piacentino, Mattiello, Weisser, & Gaughan, 2000;

Pogwizd, 2000). Depressed SERCA function associated with heart failure can be

attributed to decreased expression of SERCA2a or enhanced inhibition from SERCA.

There are no mutations in SERCA2a known to cause a predisposition of heart failure (Kranias & Bers, 2007). In the only screen of SERCA2a mutations in heart failure patients, only 4 point mutations were found, all of which were conserved mutations resulting in no change in the amino acid sequence (Schmidt et al., 2003). This suggests that genetic polymorphisms are not likely to account for differences in SERCA activity in heart failure patients. Darier’s disease is characterized as a haploinsufficiency of the

SERCA2 gene resulting in a skin disorder (Brini & Carafoli, 2009; Dally, Corvazier,

Bredoux, Bobe, & Enouf, 2010). Interestingly however, adult Darier patients are not predisposed to heart failure suggesting a robust compensatory mechanism for cardiac

SERCA expression or sufficient expression from one allele. It has been speculated that the lack of genetic variation within SERCA2a suggests that even minor changes to

SERCA2a activity cannot be tolerated and are incompatible with life (Kranias & Bers,

2007).

While there are no known SERCA mutations associated with heart failure, there have been 3 important mutations identified within PLB associated with dilated cardiomyopathy. A substitution of Arg9 for Cys (R9C) has been shown to impair phosphorylation of PLB, thereby enhancing inhibition of SERCA (Brini & Carafoli, 2009;

Ha et al., 2011; Schmitt et al., 2003). A premature stop codon at Leu39 (L39stop) has been shown prevent proper insertion of PLB into the SR membrane thereby preventing its ability to regulate SERCA (Kranias & Bers, 2007). The deletion of Arg14 (R14del) is

thought to enhance the interaction with SERCA, thereby increasing the inhibition of

SERCA (Haghighi et al., 2006).

Therapeutic Targets

Since decreased contractility is a component of heart failure, enhancing SR load

through increased SERCA activity is a logical therapeutic path. This concept is supported

by studies showing that increasing SERCA2a expression levels through gene transfer in a

failing mouse heart model has be ability to restore contractility (Del Monte, Hajjar,

Harding, & Inesi, 2001). In fact this approach is currently in human clinical trials.

Decreasing the ability of PLB to inhibit SERCA is also a potential target for therapy. A phosphomimetic PLB mutant (S16E) which has a decreased ability to inhibit SERCA has been shown to be capable of relieving SERCA inhibition in heart failure (Hoshijima et al.,

2002). Also ablation of PLB through antisense DNA transfer has been effective in restoring contractility in a rat failing heart model (Tsuji et al., 2009). While PLB ablation or replacement with non-inhibitory mutants might provide an effective means of restoring contractility, deleterious effects might arise from an inability of SERCA function to respond to β-adrenergic stimulation. Therefore, development of designer PLB mutants that have a reduced ability to inhibit SERCA while retaining the ability to respond to phosphorylation would be a significant improvement of this approach.

CHAPTER III

MATERIALS AND METHODS

Overview of Fluorescence Resonance Energy Transfer

Fluorescence Resonance Energy Transfer (FRET) is a highly effective means of

detecting protein-protein interactions and is the predominant methodology I have utilized for this work. A thorough description of FRET can be found in Principles of

Fluorescence Spectroscopy 3rd Edition (Lakowicz, 2006), but will a brief description of

FRET will be provided here.

FRET utilizes the biophysical property of appropriately paired fluorescent

molecules (donor and acceptor) that causes the transfer of light energy when in close

proximity (<100 Å). This study uses Cerulean (CFP derivative) and YFP as the donor and

acceptor fluorophores, respectively. When Cerulean is excited, the energy can be

emitted from that fluorophore as a photon of blue light or transferred to YFP (exciting

YFP) causing the emission of a photon of yellow light. The percent of the donor

excitation energy that is transferred to the acceptor is known as a FRET Efficiency (E). E

is commonly determined through quantification of the donor and acceptor fluorescence

intensities. This work uses the acceptor-photobleaching and E-FRET methods to

measure FRET and both are discussed in detail below. When these fluorophores are

attached to proteins that interact (e.g., SERCA and PLB), the fluorophores can be (but

not always) in close enough proximity to undergo energy transfer. Observed energy

transfer therefore provides an indicator of interaction of the target molecules. E is highly sensitive to and can measure the distance between the fluorophores (inversely related to the distance to the power of 6). Therefore E is more than just an indicator of interaction since it provides some information about the orientation of the interacting proteins by producing a distance constraint.

Acceptor Photobleaching

Acceptor photobleaching experiments were performed with an inverted

microscope equipped with a 60X 1.49 NA objective, and a back-thinned CCD camera

(iXon 887, Andor Technology, Belfast, Northern Ireland). The detector was cooled to -

100 °C, using a recirculating liquid coolant system (Koolance, Inc., Auburn, WA). Image

acquisition and acceptor photobleaching was automated with custom software macros

in MetaMorph (Molecular Devices Corp, Downingtown, PA) that controlled motorized

excitation/emission filter wheels (Sutter Instrument Co., Novato, CA) with filters for

CFP/YFP/mCherry (Semrock, Rochester NY). The progressive photobleaching protocol

was as follows: 100 ms acquisition of CFP image, 40 ms acquisition of YFP image,

followed by 10 s exposure to YFP-selective photobleaching (504/12 nm excitation. FRET

efficiency was calculated from the fluorescence intensity of the CFP donor before and

after acceptor-selective photobleaching, according to the relationship: E = 1 -

(FPrebleach/FPostbleach). Representative pre and post bleach images are shown in Fig. 8A

43 A

Figure 8: Acceptor Photobleaching FRET. (A) CFP and YFP images pre and post YFP photobleaching. After photobleaching, the YFP fluorescence is abolished while CFP fluorescence has been enhanced. (B) Average normalized fluorescence intensity of CFP (blue) and YFP (green) during the course of YFP photobleaching. CFP intensity increases with decreasing YFP intensity.

44 displaying enhanced CFP fluorescence corresponding to abolished YFP fluorescence. Fig.

8B shows the time course of average fluorescent intensity of CFP increasing as average fluorescent intensity of YFP decreases.

E-FRET

I have also utilized E-FRET or “3-cube” FRET (Hou & Robia, 2010; Zal &

Gascoigne, 2004) rather than acceptor-photobleaching to improve on the speed to data collection. Measurements are consistent with the well established acceptor photobleaching method that our lab has previously used (Hou & Robia, 2010; Hou et al.,

2008; Kelly et al., 2008). For these measurements, we used a similar microscope setup described above for acceptor photobleaching. For each sample, automated acquisition of a field of 48 images was performed using a motorized stage (Prior, Rockland, MA) controlled by MetaMorph software. Focus was maintained by an optical feedback system (Perfect Focus System, Nikon). Images were obtained with a 40X 0.75 N.A. objective with 100 ms exposure for each channel: Cer, YFP, and “FRET” (Cer excitation/YFP emission). Fluorescence intensity was automatically quantified with a multi-wavelength cell scoring application in MetaMorph. Cell selection criteria included having a diameter of 35-75 pixels and having an average intensity of 100 counts above background. The data were transferred automatically to a spreadsheet for analysis. E-

FRET is a ratiometric FRET estimation calculated from

I − a(I ) − d(I ) E = DA AA DD IDA−a(IAA)+(G−d)(IDD)

where IAA is the intensity of fluorescence emission detected in the donor channel

(472/30 nm) with excitation of 427/10 nm; IAA is acceptor channel (542/27 nm) emission

with excitation of 504/12 nm; IDA is the “FRET” channel, with 542/27 nm emission and

excitation of 427/10 nm; a and d are cross-talk coefficients determined from acceptor-

only or donor-only samples, respectively. We obtained values for d of 0.82 for

mCerulean, and a value for a of 0.082 for YFP. G is the ratio of the sensitized emission

to the corresponding amount of donor recovery, which was 3.2 for this setup.

Binding Curves and Fitting Analysis

In addition the distance between the fluorescent probes, the percentage of

targets engaged in energy transfer determines the FRET efficiency of an individual cell.

Expression and affinity of the targets will dictate this percentage, causing FRET efficiency

to vary within a population of cells with heterogeneous protein expression. When the

FRET efficiency of individual cells is plotted against PLB protein concentration estimated by the YFP-PLB fluorescence intensity, FRET efficiency increases with protein expression up to a maximal level and can be well described by a hyperbolic function, y =

(FRETmax)x/(KD2 + x) An example scatter plot is shown in Fig. 9A along with its

corresponding hyperbolic fit in black. 300-800 cells are typically used to develop a single

binding curve. FRETmax describes the intrinsic FRET between SERCA and PLB where FRET

is limited only by the distance between the probes, providing structural information (Fig.

9A, blue). KD is the protein concentration at half maximal FRET which represents the dissociation constant of the regulatory complex providing an affinity index (Fig. 9A, red).

46 A

Figure 9: Quantitative analysis of protein concentration dependence of FRET. FRET efficiency varies within a population of cells, and can be related to protein concentration with a hyperbolic function. A representative scatter plot of individual cell values is shown along with corresponding fitting curve (A, B, black). (A) The FRET efficiency that the fitted curve is approaching is known as FRETmax (blue dashed line), the intrinsic FRET value that limited only by the distance between the fluorescent probes. The protein concentration at half maximal FRET efficiency is the KD value (red dashed lines). (B) The blue line corresponds to a sample with a decreased FRETmax, or increased distance. The red line corresponds to a sample with a reduced affinity.

The absolute values of FRET obtained through our methodology include some level of uncertainty. Due to their unstructured linkage with PLB and SERCA, the fluorescent proteins likely retain a high degree of flexibility with respect to the target proteins. The fluorescent proteins are also relatively large. In combination, these two factors limit the ability of our method to make precise structural conclusions. Presently, we are unable to determine specific concentrations from these fluorescence measurements. Therefore our concentration parameter is presented in arbitrary units

(AU) rather than molar units. Therefore, changes or shifts in these FRETmax and KD parameters provide more significant information than do their absolute values. The red and blue curves in Fig. 9B represent simulations of changes in affinity and structure, respectively. The right-shifted curve in red represents where data corresponding to reduced affinity should fall. The down-shifted curve in blue represents a distance increase corresponding to a structure change.

Non-specific FRET, or energy transfer between close but non-interacting targets, is a concern with FRET measurement, especially if the targets are constrained to the membrane. While we presently have no true negative control for PLB-SERCA FRET, we have estimated non-specific FRET through a competition assay. Addition of increasing amounts of unlabeled protein decreases FRET efficiency in our system to 4% (Hou &

Robia, 2010). Therefore, even at the highest expression levels, only 4% of the measured

FRET efficiency is accounted for by non-specific FRET.

Confocal Microscopy

Confocal Images

To verify proper expression of YFP-PLB and Cer-SERCA2a, high resolution confocal images were taken of infected myocytes. To visualize sarcolemma and t- tubules, cells were also stained with FM 4-64 (Invitrogen) according to manufactures instructions. Briefly, dye at a stock concentration of 1 mM was diluted to a concentration of 1 µM in PC-1 medium was used to bathe cells for approximately 10 min, after which cells were washed in additional PC-1 to remove excess dye. Images were obtained using an inverted Leica TCS SP5 confocal microscope with a 63x water immersion objective and an 8000 Hz resonant scanner. Cer, YFP, and FM 4-64 were sequentially excited using an argon laser with wavelengths of 458, 514, and 543 nm, respectively.

Confocal Scanning

Confocal scanning microscopy was used obtain simultaneous and high temporal resolution of FRET and Ca2+ concentration measurements in paced adult ventricular myocytes. As an index of FRET, the ratio of YFP/Cer fluorescence intensities was obtained with 458 excitation and detection of YFP (525-560 nm) and Cer (470-515 nm) using the Leica system described above. Cells were loaded with X-Rhod-1 AM as an indicator of Ca2+ concentration which was detected with 543 nm HeNe laser excitation and an emission range of 570-650 nm. Myocytes were perfused with PC-1 medium

during data collection. Experiments were conducted in x-t mode to study individual Ca2+

transients and x-y-t mode to study alternating periods of pacing and rest.

Tissue Culture

Heterologous Cell Culture and Transient Transfection

AAV-293 cells were cultured in complete DMEM growth medium (HyClone,

Waltham, MA) with 10% fetal bovine serum (heat inactivated) (HyClone, Waltham, MA),

2mM L-glutamine (CellGro, Manassas, VA). Cells were split every few days once a

confluency of 90-95% had been reached. Cells at approximately 70-80% confluency

were transfected according to manufacturer’s instructions, using MBS Mammalian

Transfection Kit (Stratagene, La Jolla, CA). 6% modified bovine serum was used instead

of the 10% suggested by the manufacturer during the three hour incubation of cells with

the CaPO4-DNA precipitate. Two days after transfection, cells were trypsinized and plated onto poly-D-lysine coated glass bottom dishes and incubated at 37 C for approximately 2 hours to permit cell attachment before data acquisition.

For Tg exposure, cells were washed 2 times with PBS, and then bathed in PBS containing the appropriate concentration of Tg, 10 pM – 100 uM. For Ca2+ experiments,

cells were permeabilized with 10 µg/ml saponin for 1 min. and then washed in bath

solution of composition 100 mM KCl, 5 mM NaCl, 2 mM MgCl2, 20 mM imidazole, 5 mM

EGTA, and 2 mM ATP (pH 7.0), with varying free Ca2+. Ca2+ concentration was calculated

with Maxchelator (Schoenmakers, Visser, Flik, & Theuvenet, 1992) and validated with a

Ca2+-sensitive electrode (Thermo Scientific, Waltham, MA).

Myocyte Culture and Adenoviral Infection

The Ca2+ phosphate transfection system used to express our fusion proteins in

AAV-293 is insufficient for expression in adult myocytes. We therefore developed adenoviral vectors of canine Cer-SERCA2a and canine YFP-PLB using the AdEasy system

(Stratagene, La Jolla, CA). Cardiac ventricular myocytes were isolated from adult New

Zealand White rabbits by a collaborating lab as previously described (Domeier, Blatter,

& Zima, 2009). All protocols were approved by the Loyola University Institutional

Animal Care and Use Committee. Myocytes were transferred to culture tubes and

washed with PC-1 medium (Lonza, Basel, Switzerland) containing penicillin and

streptomycin (Invitrogen).

Myocytes were cultured in dishes containing laminin coated glass coverslips and

adenoviruses were added at multiplicity of infection of 100. Sufficient expression of our

constructs required approximately 48 hours of culture, however viability of the

myocytes declines rapidly. To improve viability, myocytes were paced in culture using a

C-Pace EP pacer (IonOptix, Milton, MA) set to 10 volts with a frequency of 10 Hz and 5

ms pulse duration.

Calcium Indicator

Experiments required both AAV-293 and adult myocytes to be loaded with a red-

shifted fluorescent Ca2+ indicator, X-Rhod-1 AM (Invitrogen). The dye was resuspended

in Pluronic® F-127 *20% solution in DMSO (Invitrogen) to a concentration of 1 mM, and

sonicated briefly to aid in solubilization. This stock solution was diluted to 1 µM in

either PBS or PC-1, which was used to bathe the cells for 2 min. Cells were then washed gently 3x in either PBS or PC-1 and equilibrated for 20 min prior to imaging.

CHAPTER IV

LIVE CELL CALCIUM UPTAKE ASSAY

Rationale for Development of Assay

Much of the work done for this dissertation and other work in our lab utilizes

fluorescent protein fusions. These fusion proteins enable the detection of protein-

protein interactions with FRET in live cells, in real-time. While this methodology

provides significant advantages, it is not without drawbacks. How the attachment of

relatively large fluorescent proteins will affect our target proteins is a valid concern.

Therefore, assessing the functionality of our fusion proteins is a necessary step in

validating our overall approach.

Traditional SERCA activity assays such as ATP consumption and Ca2+ uptake are needed to quantify the activity of SERCA and the ability of PLB to inhibit this activity.

These assays are necessary to directly compare the activities of wild type SERCA and PLB to our fusion proteins. However, these assays can be difficult and time consuming.

Having an alternative method to rapid screen for function is therefore highly beneficial for our lab. To complement more direct quantification of our fusion protein activity that our lab has undertaken, we developed an indirect live cell Ca2+ uptake assay, the particulars of which are discussed below.

Protocol

AAV-293 cells are transfected with various constructs as discussed previously. 24 hours post-transfection, cells are plated on poly-D-lysine coated glass slides, and

allowed to attach for 2 hours also discussed previously. Ca2+ indicator loading is

described in the Materials and Methods section. Briefly, cells are washed 2 times with

PBS containing 2 mM Ca2+ and loaded with a red-shifted Ca2+ indicator, X-Rhod-1, AM.

Cells are washed an additional 3 times to remove excess dye, and cells are bathed in PBS

containing 2 mM Ca2+ for 20 min.

First, a 100 ms CFP or YFP is taken to assess transfection status of cells in a given field of view. Then a mCherry or red channel image is then taken every 5 secs for approximately 8 min (100 images). At the 1 min time point, a sufficient baseline X-Rhod

signal has been obtained, at which 100 µM ATP is added to the cells. 4 min later, 10 mM

thapsigargin is added, and data collection continues for approximately 3 more minutes.

The mean X-Rhod and std. error of 30-60 cells is plotted against time.

With typical transfection efficiency around 50-60%, there generally is a

heterogeneous population of transfected and untransfected cells in any field of view.

This is depicted in Fig. 10A showing CFP, YFP, and red channel images along with the

corresponding overlay. This field contains both transfected (T) and untransfected (UT)

cells. This allows simultaneous data collection from expressing and non-expressing cells,

thereby providing an internal control within the experiment.

Principle of Assay and Results

Activation of cell surface purinergic receptors through ATP stimulation induces

2+ an IP3 mediated Ca release in many cell types, including AAV-293 (Ralevic & Burnstock,

1998). Endogenous SERCA Ca2+ uptake is initially overwhelmed by this release event,

causing Ca2+ to accumulate in the cytosol. Desensitization of purinergic receptors and

2+ IP3 degradation halts Ca release, preventing further accumulation, and allowing the

endogenous SERCA to lower cytosolic Ca2+ to basal concentrations. If cells are loaded

with a fluorescent Ca2+ dye, a spike in fluorescence intensity following ATP stimulation

will be observed (Fig. 10B, black). In cells overexpressing our Cer-SERCA, we observed

that this spike was nearly abolished (Fig. 10B, red). We believe that enhanced SERCA activity due to overexpression of our fusion protein is able to keep pace with the Ca2+

release, thereby preventing Ca2+ accumulation.

Because we do not directly measure Ca2+ uptake activity, abolishing Ca2+

accumulation alone is not sufficient to demonstrate SERCA function. Overexpression of our fusion proteins (Cer-SERCA, YFP-PLB) could deleteriously affect the cell, perhaps causing a depletion of the ER Ca2+ stores that would prevent a Ca2+ release and

accumulation. There is a constant leak of Ca2+ from the ER, but basal SERCA Ca2+ uptake

maintains low cytosolic Ca2+ concentrations. Blocking SERCA activity will then cause

accumulation of cytosolic Ca2+. By applying the SERCA inhibitor thapsigargin (Tg)

following ATP stimulation, we are also able to assess ER Ca2+ content of these cells.

55 A

Figure 10: Ca2+ uptake in live cells. (A) Wide-field fluorescence images of AAV-293 cells transfected with Cer-SERCA and YFP-PLB, loaded with Ca2+ indicator X-rhod-1. This field contains examples of transfected (T) and untransfected (UT) cells. (B) Ca2+ transients were observed after the addition of extracellular ATP in untransfected cells (black), but not in cells expressing Cer-SERCA (red), suggesting Ca2+ uptake activity of Cer-SERCA. Ca2+ transients were partially restored in cells expressing both Cer-SERCA and YFP-PLB (blue). When compared with untransfected controls, Tg-releasable ER Ca2+ content was increased by Cer-SERCA. Cells expressing YFP-PLB (green) had reduced Ca2+ transients and reduced Ca2+ load, suggesting that exogenous PLB inhibited endogenous SERCA.

In both untransfected (Fig. 10B, black) and cells expressing exogenous SERCA

(Fig. 10B, red), we observed this Tg-mediated Ca2+ accumulation, demonstrating that

expression of our fusion protein does not deplete Ca2+ stores. In fact, cells

overexpressing Cer-SERCA show a greater Tg-mediated Ca2+ accumulation compared to

untransfected cells, suggesting a greater ER Ca2+ content. This provides further

evidence of enhanced Ca2+ uptake upon overexpression of our SERCA fusion protein.

Co-expression of Cer-SERCA with YFP-PLB partially restores the ATP-mediated

Ca2+ accumulation to control levels suggesting that YFP-PLB can functionally inhibit Cer-

SERCA (Fig. 10B, blue). Also, this further demonstrates the activity of Cer-SERCA, since

its known inhibitor can diminish its observed effect. Expression of YFP-PLB alone (Fig.

10B, green) showed smaller ATP and Tg-mediated Ca2+ accumulation compared to

untransfected cells, indicating YFP-PLB can regulate endogenous SERCA.

Summary

This assay provides a rapid and clear qualitative assessment of our fusion protein

activity. We observed that Cer-SERCA1a is active, which can be functionally regulated

by YFP-PLB. We also observed that Cer-SERCA2a is also active and can be functionally

regulated by YFP-PLB (not shown). While not used in this work, we have also observed

that a SERCA tagged with two fluorescent proteins (one of which is inserted in the

middle of SERCA) is also active and can be regulated by YFP-PLB (not shown). This

demonstrates that robustness of our system, wherein a complex containing 3

fluorescent proteins can remain functional.

CHAPTER V

STEADY-STATE DETECTION OF PLB-SERCA BINDING

Rationale

As discussed in the Literature Review, multiple models of PLB regulation have

been proposed, but no clear consensus has been achieved despite extensive research on

this interaction. Chemical crosslinking, immunoprecipitation, and FRET studies have

been used to investigate the steady-state PLB-SERCA interaction under various

conditions that favor particular conformations of SERCA. In addition to their strengths,

previous methods have significant limitations. This necessitates additional points of

view from new and alternative approaches. We have used FRET in intact cells to

quantitatively study the PLB-SERCA interaction, a method that has its own strengths and

weaknesses. By using this complementary method, we have sought to tease out some

nuances of the PLB-SERCA interaction and integrate the present data in the field into a

cohesive model.

Mean FRET

Thapsigargin (Tg) is a cell permeable SERCA inhibitor commonly used to block the

Ca2+ uptake of intact cells (Michelangeli & East, 2011). Tg has also been shown to

completely block chemical crosslinking between PLB and SERCA in isolate vesicles (Chen et al., 2006, 2005; Jones et al., 2002). We therefore initially hypothesized that Tg would 57

completely displace PLB from SERCA, and therefore completely abolish PLB-SERCA FRET

in intact AAV-293 cells. To test this hypothesis we exposed the cells transiently

expressing YFP-PLB and CFP-SERCA1a to either 1 µM Tg or vehicle control in PBS for 10

min. Using the acceptor photo-bleaching method, we then measured the average FRET

efficiency. We observed that the Tg decreased average FRET efficiency from 13.4

± 1.7% to 5.2 ± 1.1% (Fig. 11A). FRET between PLB and SERCA was decreased, but not

abolished.

In a parallel experiment, we examined the effect of Ca2+ on the average PLB-

SERCA FRET. As with Tg, we hypothesized that Ca2+ would completely displace PLB from

SERCA based upon chemical crosslinking studies, and therefore completely block PLB-

SERCA FRET in AAV-293 cells. Cells expressing YFP-PLB and CFP-SERCA1a were bathed in

10 µM ionomycin, a Ca2+ ionophore, for 10 min in PBS with or without 1.8 mM Ca2+.

FRET was then measured in these cells as before. FRET was not significantly decreased

by the presence of high Ca2+ (21.4 ± 2.2% at low Ca2+ and 18.8 ± 2.3% at high Ca2+) (Fig.

11C).

Antibodies targeting PLB have been previously shown to block the ability of PLB

to inhibit SERCA (Cantilina et al., 1993) as well as block PLB-SERCA crosslinking (Chen et

al., 2007). In an additional experiment, we examined the effect of rabbit anti-PLB

(2D12) on PLB-SERCA FRET in AAV-293 cells. Cells expressing YFP-PLB and CFP-SERCA1a

were permeabilized using 20 mg/ml saponin and then bathed in PBS with or without

A B

Figure 11: Mean PLB-SERCA FRET. The mean PLB-SERCA FRET is decreased by (A) Tg and (B) PLB antibody (2D12), but not by (C) elevated calcium concentration.

2D12 antibody. FRET was then measured in these cells using the acceptor-

photobleaching method. Average FRET was reduced by PLB antibody to 3.9 ± 1.2% from

a control of 9.7 ± 1.1% (Fig. 11B). FRET again was not completely abolished as was

initially hypothesized.

These average FRET results were inconsistent with the published crosslinking

studies which predicted we would observe a complete blockage of FRET with Ca2+, Tg, or

PLB antibody (Chen et al., 2007, 2006, 2005). We observed no significant reduction in

FRET with high Ca2+ concentrations, and only partially reduced FRET in the presence of

Tg and PLB antibody. The lack of the predicted effects could be due to several factors.

Our experimental conditions could have insufficiently saturated our system (incomplete

binding of Ca2+ and Tg to SERCA or 2D12 to PLB) thereby preventing the full dissociation

of PLB from SERCA. However, our initial hypothesis that PLB would completely

dissociated PLB from SERCA may have also been incorrect.

Protein Dependence of FRET

Using average FRET efficiency from a population of cells as an indicator of

protein-protein interaction is simple and straight forward, but is also very limiting. We

can detect binding and changes in binding, but cannot fully describe these changes. A

decrease in average FRET efficiency could be caused by unbinding or dissociation of a

portion of the proteins undergoing FRET indicating a decrease in affinity. Also, there could be an increase in the distance between our probes attached to our target proteins indicating a structure change. We decided further experiments should be analyzed

61 using a standard lab protocol showing the dependence of FRET efficiency of a cell on its protein expression level. This approach has been shown to distinguish between changes in affinity (KD2) and structure (FRETmax) of the PLB-SERCA regulatory complex (Ha et al.,

2011; Hou & Robia, 2010; Hou et al., 2008; Kelly et al., 2008) as well as the Na, K-

ATPase-phospholemman interaction (Song, Pallikkuth, Bossuyt, Bers, & Robia, 2011).

We are able to observe a clear trend when we plot FRET efficiency of many cells against their protein concentration, estimated by YFP fluorescence intensity (Fig. 9). In control cells, FRET increases with protein concentration up to a maximal level and is described by a hyberbolic function of the form y = (FRETmax)x/(KD2 + x). FRETmax and KD2 values are consistent with previously published work from this lab. Addition of 10 uM

Tg, as described above, causes a rightward shift of the fitted curve (Fig. 12A). KD2 is increased from 7.0 to 11.9 AU. We observed no change in the FRETmax parameter.

These two observations suggest that the decrease in average FRET was due to a decrease in the amount of PLB interacting with SERCA and not an increase in the distance between the two proteins. While there is reduced binding, there is not a complete abolishment in binding that we initially hypothesized based on the crosslinking experiments.

Using this same approach, we further analyzed the Ca2+ effect on PLB-SERCA binding. Unlike the mean FRET experiment, we utilized PBS containing 1 mM EGTA, along with 10 µM ionomycin, to further lower Ca2+ concentration. This condition

2+ produced a typical curve approaching a FRETmax of 30.6%. Addition of 2 mM Ca caused

62 A

Figure 12: Protein dependence on FRET, acceptor photobleaching. Both (A) Tg and (B) elevated calcium cause a rightward shift in the dependence curves, describing a reduced PLB-SERCA binding affinity.

a slight rightward shift of the data (Fig. 12B). The control and high Ca2+ datasets overlap, but are clearly resolvable with an increase of KD2 from 4.5 to 6.0 AU. As with

Tg, the Ca2+ effect on PLB-SERCA binding can be described as a decreased affinity,

partially reducing binding, but not completely abolishing it.

This is unlike the mean FRET result shown previously where no significant Ca2+

effect was observed. Several factors could explain this discrepancy. Use of EGTA could

have partially lowered the control KD2 value thereby producing a change that was not

present in the mean FRET experiment. Also, with the high degree of overlap in the

datasets and a relatively small change in KD2, about a 30% increase, the change may have been too subtle to observe with the mean FRET approach. This extra dimension provides an advantage of this binding curve FRET analysis of discerning very subtle changes.

E-FRET rationale

Although we observed clear results from the acceptor photobleaching approach,

we opted for the E-FRET method to improve the rate of data collection. A two condition

plot such as Fig. 12 would require a full day of data collection, whereas a comparable E-

FRET plot would take only a few minutes. Therefore it is much easier to compare data from many different conditions. With acceptor photobleaching, the time of Tg/ Ca2+

exposure would vary between 30-120 minutes. Due to the shorter time frame of data

collection, Tg or Ca2+ exposure could be more directly controlled, thereby reducing

variability within samples. Additionally, a typical E-FRET plot would have a much larger

sample size as an added benefit. Fewer coverslips and transfected cells are needed also

improving cost effectiveness of the assay as well.

E-FRET requires calculations with more variables and assumptions, thereby

producing more uncertainty in the absolute FRET efficiencies measured. However, we

observed consistent results with acceptor photobleaching and E-FRET for both the Tg

2+ and Ca effects. The FRETmax parameters were similar between the two approaches.

Because of the different microscope setups used, the KD2 values of the fitted curves,

measured in AU, are not the same between the approaches, but the magnitude of the

KD2 shifts are similar. Because my conclusions are based more on the relative changes

of these curves rather than the specific values, E-FRET was the most effective option for

further experimentation.

Calcium Dose Effect

Rather than using ionomycin, we opted to saponin-permeabilize our cells to

introduce internal solutions with EGTA buffered Ca2+ to more directly control Ca2+

concentration. Cells expressing Cer-SERCA1a and YFP-PLB were permeabilized with 10

µg/ml saponin for 1 min. and then washed in bath solution of composition 100 mM KCl,

5 mM NaCl, 2 mM MgCl2, 20 mM imidazole, 5 mM EGTA, and 2 mM ATP (pH 7.0), with

varying free Ca2+. Free Ca2+ concentration was calculated with Maxchelator.

After 2-3 min in bath solution, the automatic E-FRET acquisition protocol was performed for each Ca2+ concentration. The fluorescence intensities of cells were

quantified using the Metamorph multiwavelength application, from which FRET

efficiencies were generated as previously described. Five independent experiments

were performed.

Initially, all datasets were fit by a hyperbolic function of the form y =

(FRETmax)x/(KD2 + x), with all parameters independently fit. Because mean FRETmax was

not significantly different for low and high concentrations of Ca2+, we concluded that

2+ FRETmax did not change in response to changes in Ca . Therefore, this parameter was

shared in a subsequent global analysis, yielding a single FRETmax fit with independent KD2

values for each condition.

A typical plot of is shown in Fig. 13A for low (100 pM) and high (1 mM) Ca2+

concentrations with their corresponding fitted curves. The data points are pooled for

visualization, showing average data with error bars corresponding to standard error. A

similar Ca2+ effect was observed with cell permeabilization compared to ionomycin

treatment used previously producing an approximately 40% increase in KD2 value. The

2+ KD2 values generated from all data sets were then plotted against Ca concentration.

2+ Fig. 13B shows the dependence of KD2 on increasing Ca concentration. A modified Hill

2+ function shows a 41% increase in KD2 with Ca , with an EC50 of 410 nM and a Hill

coefficient of 1.1.

Calcium Electrode

Since the Maxchelator predicted free Ca2+ concentrations are just theoretical

estimations, they require further validation. Fig. 14A depicts the dose response of Ca2+

using solutions containing 1 mM EGTA, corresponding to the predicted Maxchelator

Figure 13: Calcium effect on PLB-SERCA FRET, E-FRET. (A) As with acceptor photobleaching, elevated calcium (red) causes a right-shifted curve compared to control (black) corresponding to a decreased PLB-SERCA affinity with no change in the FRETmax parameter (B) The calculated KD2 values from the example fitted curves in A are shown plotted against calcium concentration. KD2 appears to transition between two affinity states upon elevation of calcium.

67 A B

C D

Figure 14: Calibration of calcium buffered solutions. (A) The initial calcium response curve with an EGTA of 1 mM with the calcium concentration values measured estimated by MaxChelator. (B) The calcium calibration curve of our prepared solutions with 1 mM EGTA (red) compared to standard solutions (black). While in agreement at high calcium concentrations, our solutions deviated from the known standard solutions at lower concentrations. (C) The dose curve in A with calcium concentration adjusted using the voltage curve in B showing that the low calcium solutions were all clustered around the same concentration. The red curve shows the original fitted curve in A, and the blue and green curves show other potential curves that could also fit the data set. An estimate of the high affinity (low KD) state was therefore unable to verified using only 1 mM EGTA. (D) The calcium calibration curve of our prepared solutions with 5 mM EGTA (red) compared to standard solutions (black) showing agreement of the full range of calcium concentrations.

values. Using a Ca2+ sensitive electrode (Thermo Scientific), we compared the free Ca2+

concentration of our solutions, to a set of standard Ca2+ solutions (World Precision

Instruments). As shown in Fig. 14B, we observed good agreement with the standards at concentrations above 100 nM, but deviation from the standards at lower concentrations.

Our initial Ca2+ dependence plot is shown in Fig. 14C after correction with the

standard concentration curve. With the low Ca2+ data points clustered around 100nM,

we could not conclude that we were accurately quantifying the lowest possible KD2

value. The blue and green curves represent simulations that could also fit the data,

showing large differences in the potential low KD value. By increasing EGTA

concentration from 1 mM to 5 mM, our prepared solution were in good agreement with

the standards over the full range of concentrations (Fig. 14D). The experiments were

then independently repeated 5 times with the newly determined solutions. With the

increased EGTA concentration, a slightly larger increase in the KD2 change was observed,

but the overall trend remained the same. While it did not alter our final conclusions, we

further validated our methodology.

Thapsigargin Dose Effect

Using the same high-throughput E-FRET described for Ca2+, we quantified the Tg

effect over a range of Ca2+ concentrations, 10 pM to 100 uM, in AAV-293 cells

expressing YFP-PLB and Cer-SERCA1a (Fig. 15A). This dose response was repeated with

4 independent experiments. As with acceptor photobleaching, we observed a rightward

shift to the concentration dependence of PLB-SERCA FRET signifying an increased KD2

value, or a decreased affinity.

Initially, all datasets were fit by a hyperbolic function of the form y =

(FRETmax)x/(KD2 + x), with all parameters independently fit. Because mean FRETmax was

not significantly different for low and high concentrations of Tg, we concluded that

FRETmax did not change in response to changes in Tg. Therefore, this parameter was

shared in a subsequent global analysis, yielding a single FRETmax fit with independent KD2

values for each condition.

A typical plot of is shown in Fig. 15A for low (10 pM) and high (100 uM) Tg

concentrations with their corresponding fitted curves. Mean KD2 (+SE) values generated

from all data sets were then plotted against Tg concentration (Fig. 15B). A modified Hill

function shows a 78% increase in KD2 with Ca2+, with an EC50 of 350 nM and a Hill coefficient of 0.7.

PLB Antibody Effect

In a continuation of the mean FRET experiments, we also quantified the protein

concentration dependence of FRET in the presence of the PLB antibody 2D12. AAV-293

cells expressing Cer-SERCA1a and YFP-PLB were permeabilized with 50 µg/ml saponin

for 1 min. and then washed in bath solution of composition 100 mM KCl, 5 mM NaCl, 2

mM MgCl2, 20 mM imidazole, 1 mM EGTA, and 2 mM ATP (pH 7.0), containing 10 µg/ml

2D12 PLB antibody. The E-FRET protocol was performed at 5, 30, and 120 minutes on

the same cells in the presence or absence of antibody, to determine the protein

Figure 15: Tg effect on PLB-SERCA FRET, E-FRET. (A) As with acceptor photobleaching, elevated Tg (red) causes a right-shifted curve compared to control (black) corresponding to a decreased PLB-SERCA affinity with no change in the FRETmax parameter (B) The calculated KD2 values from the example fitted curves in A are shown plotted against TG concentration. KD2 appears to transition between two affinity states upon elevation of Tg.

dependence on FRET. At the 5 min. time point, no change in FRET was observed from

exposure to PLB antibody (not shown). PLB antibody greatly reduced PLB-SERCA FRET at

30 min (Fig. 16A), and to a greater extent at 120 min (Fig. 16B). At both time points,

there appears to be a reduction in the FRETmax parameter without a change in the KD2

value. More attempts are needed to determine significance.

Additionally, increased incubation time, higher antibody concentrations, and

more robust permeabilization could conceivably produce a larger effect by the PLB

antibody. Further tests are also needed to verify that a maximal effect has been

achieved before more explicit conclusions can be made. Attempts to generate a longer

time point (overnight) failed to produce usable data because of sample degradation.

Any future experiments should also include nonspecific IgG as control rather than no

antibody. However, this may suggest that the antibody may not fully dissociate PLB

from SERCA. The reduction in FRETmax would suggest an altered structure of the PLB-

SERCA complex.

PLB-PLB FRET

To assess PLB pentamer formation, we also quantified FRET in cells co-expressing

Cer-PLB and YFP-PLB using the E-FRET method. Cells were either exposed to 10 mM Tg

or had Ca2+ controlled to low (1 nM) and high (1 mM) concentrations with EGTA buffered solutions as in PLB-SERCA FRET experiments described earlier. Neither Tg

2+ exposure or changes in Ca concentrations affected the FRETmax or KD1 parameters,

(Bidwell et al., 2011) and FRETmax value of 54.2 ± 0.7 is consistent with previously

Figure 16: PLB antibody (2D12) effect on PLB-SERCA FRET. The protein dependence of FRET in the absence (black) and presence (red) of PLB antibody after (A) 30 min and (B) 120 min incubation times.

measured values (Hou et al., 2008; Kelly et al., 2008). This demonstrates that neither

our exposure of Tg or changes in Ca2+ concentrations alter pentamer structure or

affinity, which was expected since neither Tg or Ca2+ have been predicted to directly

interact with the PLB pentamer.

Since PLB exists in a coupled equilibrium (Fig. 4), this demonstrates that our

observed decreases in PLB-SERCA affinity by Ca2+ (Fig. 13) and Tg (Fig. 15) are not caused

my increases pentamer affinity. With no change in PLB-PLB FRET observed, this

demonstrates that our methods of Tg exposure and control of Ca2+ concentration do not

have an effect on fluorophore stability or membrane dynamics that could globally

decrease FRET efficiency or membrane protein affinity. Therefore, the absence of PLB-

PLB FRET changes provides a good control experiment validating our methodology and demonstrating that our PLB-SERCA effects are specific.

Summary

We observed a decrease in PLB-SERCA FRET in response to increases in Ca2+ and

Tg concentration. This decrease can be attributed to a reduced affinity, as observed by

an increased KD2 value. With no change in the FRETmax parameter, no change in the distance between the fluorescent probes was observed. However, some structural rearrangement decreasing the strength of the intermolecular contacts must account for the decrease in affinity. This structure change is likely just not accompanied by a change in the position of the fluorescent probes.

2+ We observed KD2 increases in response to increases in Ca and Tg concentrations up to a maximal level corresponding to a transition in the affinity of the

PLB-SERCA regulatory complex. We therefore believe that the Ca2+ dose response (Fig.

13B) shows PLB transitioning from binding the E2 (Ca2+ unbound) and the E1 (Ca2+ bound) conformations of SERCA. So even at saturating Ca2+ concentrations known to fully relieve PLB inhibition of SERCA, significant PLB-SERCA binding is still observed. This suggests that unbinding of PLB is not required for relief of PLB-SERCA inhibition.

CHAPTER VI

SERCA INTERACTION WITH TRUNCATED PHOSPHOLAMBAN

Rationale

To account for our data in the previous chapter as well as other data in the field,

we propose the Domain Displacement Model where high Ca2+ concentrations displace

the transmembrane domain of PLB to a second binding site on SERCA with no change in

the position of the cytosolic domain. To directly test this model, we need to assess

changes in position of the transmembrane which could not be accomplished in my

earlier studies. Since YFP is N-terminally linked to PLB in previous studies, we are only

able to confirm that the position of the cytosolic domain does not change in response to

Ca2+. If the transmembrane domain is displaced independently of the cytosolic domain,

the position of YFP would not change in response changes in Ca2+ concentration

To take this factor into account, we have sought to quantify FRET between

SERCA and a truncated PLB which lacks the cytosolic domain. An N-terminally linked YFP would then be contiguous with transmembrane domain. Therefore, if the transmembrane domain is displaced to an alternative binding site, we expect to see a change in our FRETmax parameter unlike with full-length PLB.

A concern with truncating PLB is that it may not then properly interact with

SERCA; therefore assessing its interaction would not be sufficient to test our hypothesis. 75

Previous studies have shown that truncated PLB corresponding to residues 25-52

(Hughes et al., 1996) and 28-52 (Y Kimura et al., 1996) are capable of fully inhibiting

SERCA activity. These results provide some feasibility of our approach

Vector Construction

The sequence corresponding to residues 28-52 of canine PLB was amplified with

PCR, the product of which included BglII and HindIII restriction sites. A double digestion and ligation were performed inserting amplified region of PLB into the eYFP-C1 vector

(Clontech) at the multiple cloning site, generating YFP-tmPLB. A complementary vector,

Cer-tmPLB, was produced by performing a double digest and ligation with YFP-tmPLB and pmCerulean-C1 (Clontech) using the BglII and HindIII restriction sites. Both vectors have a 7 residue spacer between the N-terminus of residues 28-52 of PLB and the fluorescent protein, coming from the remaining sequence in the multiple cloning site.

Simulation of tmPLB FRET

Using Pymol, we estimated the anticipated distance changes produced by YFP- tmPLB, compared to full-length PLB, using the PDB structures represented in Fig. 17A.

Since the positions of the fluorescent probes are unknown, some uncertainty in the distance measurements, the estimations were made by applying distance changes to the observed distance calculated from the FRETmax value generated by the steady-state experiments. I first measured the distance between the N-terminus and residue 28 of

77 A

Figure 17: FRET between YFP-tmPLB and Cer-SERCA. Simulated binding curves based on PDB structure measurements (A) and experimentally observed binding curves (B) with wtPLB shown shaded and dashed and tmPLB shown in solid. (A) At low calcium concentration (black), our model predicts that decrease in FRETmax with tmPLB with a further decrease at high calcium (red). (B) At low calcium, we observe the expected decrease in FRETmax (black), but do not observe the expected additional decrease at high calcium concentration (red).

PLB to account for the truncation. I then measured the distance between the predicted binding pocket of helices M2/M4/M6 and the exterior position of helix M9.

Calcium Effect

FRET was quantified by the E-FRET method while Ca2+ concentration was controlled as previously described in cells expressing Cer-SERCA1a and either YFP-PLB

(full-length) or YFP-tmPLB. As seen previously, saturating Ca2+ (1 mM) caused a rightward shift in the curve describing protein dependence of FRET compared to low

Ca2+ (< 1 nM) for YFP-PLB expressing cells, shown in Fig. 17B with the shaded, dashed lines. At low Ca2+ concentration in cells expressing YFP-tmPLB, the anticipated decrease

2+ in FRETmax to 20.1% was observed (Fig 17B, black). However, saturating Ca concentrations had no effect on the protein dependence of FRET in these cells (Fig. 17B, red). Neither the anticipated increase in KD2 or further decrease FRETmax was observed, suggesting that no change in the binding of YFP-tmPLB and Cer-SERCA1a occurred.

YFP-tmPLB Activity

To fully characterize the YFP-tmPLB that was used in the above experiment, we assessed its ability to inhibit SERCA activity using the Live Cell Calcium Uptake Assay described previously. Expression of Cer-SERCA1a alone greatly reduces the ATP mediated Ca2+ accumulation observed in untransfected cells. However, co-expression of

YFP-tmPLB does not attenuate the SERCA effect (Fig. 18B) as previously observed with

YFP-PLB (full-length) (Fig. 10). Interesting this suggests that YFP-tmPLB does not functionally regulate Cer-SERCA1a despite being able to interact.

79 A

Figure 18: Characterization of tmPLB. (A) tmPLB pentamer FRET has an increased FRETmax compared to wtPLB consistent with truncation o f the cytosolic domain. The KD1 parameter is the same for both tmPLB and wtPLB. (B) tmPLB does not restore the ATP dependent calcium accumulation in our in cell activity assay when co-expressed with Cer-SERCA.

Pentamer FRET

To further characterize YFP-tmPLB, we also assessed its ability to bind to itself by

measuring FRET between YFP-tmPLB and Cer-tmPLB using the E-FRET method described

previously. In cells expressing full length PLB tagged with YFP and Cer, we observed a

FRETmax of 53.1% (Fig. 18A, black), consistent with values obtained previously (Bidwell et

al., 2011). In cells expressing YFP-tmPLB and Cer-tmPLB, we observed an increased

FRETmax to 60.9% (Fig. 18A, red) consistent a decreased distanced produced with

removal of the cytosolic domain placing the fluorescent closer to together. The same

KD1 value was observed for both full length and truncated PLB, demonstrating that

truncation of PLB in YFP-tmPLB does not alter pentamer affinity.

Summary

2+ A change in the FRETmax parameter was not observed when saturating Ca was

added to cells expressing YFP-tmPLB and Cer-SERCA1a. Therefore, the movement of the

transmembrane domain to an alternative binding site was not observed as predicted.

YFP-tmPLB-Cer-SERCA FRET is completely unresponsive to changes in Ca2+ concentration

2+ with neither the FRETmax nor KD2 parameter changing upon addition of saturated Ca .

Also, YFP-tmPLB does not appear to functionally regulate SERCA activity. Taken

together, these observations suggest that YFP-tmPLB does not associate with SERCA as

does full-length PLB.

However, YFP-tmPLB does interact with SERCA since it expresses correctly in the

ER and undergoes FRET with Cer-SERCA1a. The FRETmax parameter is decreased to

20.1% compared to 28.2% in full length expressing cells. This decrease is consistent with a change in the position of YFP attached to a truncated PLB, as predicted by distance measurements of the PDB structure. Also, the observed KD1 parameter (pentamer

affinity) is the same for both full-length and truncated PLB. These observations suggest

that YFP-tmPLB retains some binding characteristics of full-length PLB despite inability

to regulate SERCA and unresponsiveness to changes of Ca2+ concentration.

Presently, we hypothesize that steric hindrance of the fluorescent protein in YFP-

tmPLB prohibits functional interaction with SERCA, thereby preventing its ability to

regulate SERCA and to respond to Ca2+. Changes to the YFP-tmPLB construct may then

allow proper interaction. The planned changes and further characterization of the

current YFP-tmPLB construct are reviewed in detail in the Discussion section.

CHAPTER VII

REAL-TIME PLB-SERCA FRET

Rationale

To this point, all investigation of the PLB-SERCA interaction has utilized steady-

state biochemical approaches. Because of the limitations of these studies, little is

understood about the kinetics and dynamics of this interaction. A fundamental

question that remains is how this interaction changes over-time in response to changes

in Ca2+ concentration. This question is all the more important considering Ca2+

concentrations change rapidly on a millisecond timescale in cardiac and slow twitch

skeletal muscle, the only PLB expressing tissues. In the case of cardiac muscle, these

changes are occurring continuously, on a beat to beat basis of the heart.

The range of Ca2+ concentration over which PLB alters the interaction with and

activity of SERCA corresponds to concentrations of a Ca2+ transient of cardiac myocyte

(100 nM – 1 µM). In other words, PLB interaction and inhibition of SERCA could theoretically vary over the course of a single Ca2+ transient. That is if changes in binding

and inhibition occur on a fast enough timescale, which is undetermined at the present

time due to limitations of past approaches.

Our FRET system describes a clear change in PLB-SERCA FRET in response to

elevated Ca2+. This change occurs in the same range of Ca2+ response to assays predicting complete dissociation, suggesting that the approaches are compatible and detect the same change in binding. The variations in the approaches produce a different observable endpoint, favoring different mechanistic models. A significant benefit of our system over all other methods is that we can detect changes in PLB-

SERCA binding in real time within intact cells. Therefore, our methodology has the potential to investigate the time domain of this interaction, no matter the explicit model of PLB-SERCA interaction.

Evidence of the timescale of PLB-SERCA binding comes from earlier work from our lab (Robia et al., 2007). Using a FRET recovery after photobleaching approach, rapid exchange between PLB and SERCA was observed with a time constant of 1.4 s. With this result in combination with our steady-state experiments where Ca2+ decreases PLB-

SERCA FRET, we initially hypothesized that a single Ca2+ transient would cause a small but detectable decrease in PLB-SERCA FRET. To test this hypothesis, we expressed our

fluorescent PLB and SERCA proteins in paced adult ventricular myocytes and

simultaneously measured FRET and Ca2+ concentration.

Characterization Myocyte FRET Detection

Myocyte culture and adenoviral infection are discussed earlier in the Materials and Methods section. Localization of YFP-PLB and Cer-SERCA2a was verified with

confocal microscopy. Images show adult ventricular myocytes expressing Cer-SERCA2a

Figure 19: Adenoviral expression of Cer-SERCA and YFP-PLB in cardiac myocytes. Confocal images of an isolated adult rabbit ventricular myocyte expressing Cer-SERCA and YFP-PLB. Both proteins were distributed in longitudinal streaks as well as striations. The striations were due to localization of both proteins at the Z-line, as indicated by counterstaining with the membrane dye FM 4-64. The overlay image shows the relative localization of Cer-SERCA, YFP-PLB, andFM4-64. Scale bar = 10 µM.

and YFP-PLB displaying a striated pattern of fluorescence as well as longitudinal streaks

(Fig. 19), suggesting localization in the junctional and transverse SR, respectively.

Counterstaining with the membrane dye FM 4-64 demonstrate that the t-tubule system

was largely intact after 2 days of culture. The overlay image shows that the striated

pattern of SERCA and PLB corresponds to the Z-lines of the sarcomeres. By measuring

fluorescence recovery after photobleaching, our lab has also shown that both PLB and

SERCA can diffuse across multiple sarcomeres during the course of several minutes

(Bidwell et al., 2011).

The protein dependence of FRET between PLB and SERCA in these cells was also

quantified using the E-FRET method used in the earlier steady-state experiments in AAV-

293 cells and described in the Materials and Methods section. FRET increased with

protein concentration towards a maximal value, consistent with the measurements

discussed previously in AAV-293 cells (Bidwell et al., 2011).

Single Calcium Transient FRET Measurements

Myocytes were infected with Cer-SERCA2a and YFP-PLB and loaded with Ca2+

indicator dye as discussed in Materials and Methods and then perfused with PC-1

medium and stimulated at a rate of .25 Hz. Confocal time course imaging was

performed in x-t mode (linescanning) also described in Materials and Methods. Fig. 20A shows the average of 21 con- transients obtained by line-scan mode measurements of the indicator X-rhod-1 (black trace) normalized to diastolic Ca2+. We obtained

simultaneous measurements of YFP and Cer fluorescence with 458 nm excitation. We

Figure 20: Quantification of calcium and FRET changes in single transients. (A) An average of 21 consecutive calcium transients quantified by X-rhod-1 fluorescence (black) and the corresponding PLB-SERCA FRET ratio (red) in a paced ventricular myocyte. (B) 5 µM isoproterenol increased the magnitude and decreased the width of the calcium transient.

87 did not detect a systole-to-diastole change in the normalized YFP/Cer ratio (Fig. 20A, red trace), suggesting that FRET from Cer-SERCA to YFP-PLB is not abolished during systole.

The β-adrenergic agonist isoproterenol (5 µM) increased the amplitude and decreased the duration of the Ca2+ transient (Fig. 20B), but FRET was still unaffected.

FRET Response to Pacing

Since individual Ca2+ transients were insufficient to produce an observable change in PLB-SERCA FRET we hypothesized that FRET would be abolished by a larger prolonged elevation of cytosolic Ca2+. To do so, we alternated between intervals of rest

(no stimulation) with intervals of 1-Hz pacing producing a rapid and progressive increase in average Ca2+. While not naturally occurring, this protocol shows that our detection method is capable of observing real-time changes in PLB-SERCA FRET in response to rapid changes in physiological Ca2+ concentration. Essentially, we use the cellular machinery of the myocytes rapidly alter Ca2+ concentration.

In response to Ca2+ changes of this protocol, we observed a modest reduction in the normalized YFP/Cer ratio with rapid pacing apparent in the absence (Fig. 21A) or presence (Fig. 21B) of 5 µM isoproterenol. Fig. 21C/D (black) shows the mean cytosolic

Ca2+ elevation for control (n = 6 cells) and after the addition of isoproterenol (n = 4 cells) normalized to resting Ca2+, and the corresponding mean YFP/Cer ratio is shown in red.

During pacing intervals, the mean cytosolic Ca2+ was significantly elevated when compared with rest, both before (p < 0.01) and after (p < 0.01) isoproterenol. FRET was also significantly decreased when compared with rest, both before (p < 0.05) and after

A B

C D

Figure 21: Quantification of calcium and FRET changes by alternating pacing. (A) Prolonged elevations of Ca2+ (black) were achieved with intervals of rapid (1 Hz) pacing after periods of rest (no stimulation). A modest decrease in FRET (red) was observed during rapid pacing. The blue lines represent the Ca2+ and FRET traces smoothed by 5 second adjacent averaging. (B) As in A, after 5 µM isoproterenol, a small decrease in FRET (red) was observed during rapid pacing. (C) Averaging multiple experiments showed that rapid pacing increased mean [Ca2+], with a larger increase observed after addition of 5 µM isoproterenol. (D) Averaging multiple experiments showed that rapid pacing decreased the mean YFP/Cer ratio, and this decrease was larger after addition of 5 µM isoproterenol. The first, second, and third pacing intervals are marked “a, b, c.”

(p < 0.01) isoproterenol. Overall, the data are consistent with a small decrease in FRET

during intervals of rapid pacing.

FRET Response to Calcium Changes in Heterologous Cells

With the time resolution of Fig. 21, we can estimate that the change in FRET

occurs within 2 seconds of elevation of the Ca2+. However, this is a much longer than

the time course of a single Ca2+ transient. We therefore cannot conclude if the

rest/pace protocol produced a change in FRET due to a higher Ca2+ concentration, a

more prolonged Ca2+ elevation, or both.

To test these possibilities, we set out to measure the time response of PLB-

SERCA FRET to elevations of Ca2+ in AAV-293. Cells expressing Cer-SERCA and YFP-PLB

were loaded with X-Rhod-1 as in the myocytes, and then bathed in PBS containing 1 mM

EGTA and ionomycin. Image acquisition of E-FRET and Ca2+ concentration every 10 s was started at this low concentration. After 1 min of acquisition to achieve baseline signals, PBS containing 2 mM Ca2+ added and acquisition continued. Fig. 22 shows that

increases in Ca2+ concentration appear to directly coincide with decreases in FRET signal,

suggesting a rapid response. However, time resolution of this approach is insufficient to

clearly determine the time course of the Ca2+ action. Further experimentation with

faster time resolution is needed, but indicates that this methodology clearly detects

changes in FRET in response to Ca2+ over time.

Figure 22: Time course of calcium effect on PLB-SERCA FRET. In AAV-293 cells bathed in 10 µM ionophore, addition of 1 mM Ca2+ to the bath solution increases intracellular Ca2+ concentrations estimated by average X-Rhod-1 fluorescence (black). Average PLB- SERCA FRET (red) decreases rapidly in response to the elevated Ca2+.

Summary

We observed no change in PLB-SERCA FRET in response to the individual Ca2+

transients of paced cardiac myocytes. Even with the enhancement of these transients

from the exposure to β-adrenergic stimulation, no change could be observed. However,

we did observe a small (2%) decrease in FRET in response to intervals of prolonged rapid

pacing and rest, which produced relatively higher and more prolonged elevations of

cytosolic Ca2+ compared to individual transients. While not a physiologic pacing pattern,

this protocol demonstrates the capability of our approach in detecting real-time changes in FRET in response to physiological changes in Ca2+ concentration. We must

now attempt to determine the specific time domain of Ca2+ action on PLB-SERCA FRET.

CHAPTER VIII

DISCUSSION

Global versus Single Molecule Effects

One of the ultimate goals of this research is to understand the mechanism of PLB

inhibition of SERCA with respect to how individual molecules interact and function.

However, our work, as well as the studies that have preceded it, is attempting to do so

by making global assessments of a large population of molecules. While we attempt to

favor the existence of a particular conformation of SERCA in our steady-state

measurements, we ultimately only achieve an equilibrium shift in a distribution of

several conformations. In our real-time FRET measurements, we are not observing

synchronous changes in all PLB-SERCA interactions, so at any given point in time we are

observing several interacting states. Therefore, any given measurement of the PLB-

SERCA interaction is a composite of many interacting molecules from which we attempt

to deduce information on the interaction of individual molecules. This is not a critique

of our approach given the inherent difficulty of making single molecule measurements,

but merely an important consideration necessary to put our results, and the results of

others, in the proper perspective.

Steady State FRET

We have used FRET to measure the interaction between fluorescently tagged

PLB and SERCA in cells under conditions that favor particular structural conformations of

SERCA. At low Ca2+ concentrations that favor the E2 SERCA state, we have observed that PLB interacts with SERCA in a protein dependent manner consistent with our

previous use of this approach (Ha et al., 2011; Hou & Robia, 2010; Hou et al., 2008; Kelly

et al., 2008; Song et al., 2011). This observation is also consistent with both the

Dissociation and Subunit Models of PLB-SERCA regulation, since both predict that PLB

interacts with the E2 conformation with high affinity.

When Ca2+ was elevated to saturating concentrations favoring the E1

conformation, we observed that PLB-SERCA FRET was decreased, but not completely

abolished (Fig. 13). The decreased FRET can be explained as a reduced affinity between

PLB and SERCA (increased KD2) without a change in the structure of the regulatory

complex (no FRETmax change). While binding is decreased at any given protein

concentration, significant levels of binding are detected at saturating concentrations of

Ca2+. This is inconsistent with the Dissociation Model where FRET would be predicted to be completely abolished. In theoretical terms, the mutually exclusive binding described by the Dissociation Model predicts a KD2 value that approaches infinity with increasing

2+ Ca concentrations. KD2 certainly would not achieve a maximal level as we observed.

Under physiological constraints, an increase in the KD2 value by a few orders of

magnitude at millimolar Ca2+ would effectively correspond to the Dissociation Model.

However, the much lower 40% increase in KD2 would not be consistent with complete dissociation of PLB and SERCA at physiological conditions Ca2+ concentrations.

One might argue detection of PLB-SERCA binding in the presence of Ca2+ is an artifact of our approach. The high concentrations of proteins in the membrane will place non-interacting proteins in close enough proximity to undergo energy transfer. It is certainly possible that unbound proteins will undergo FRET. However, in a competition assay using unlabeled PLB, we currently estimate FRET between unbound membrane proteins or “non-specific FRET” around at most 4%. This is clearly much lower than the 28% FRET observed with our assay. Furthermore, we have shown that the estimated distance between the probes can be specifically altered and that this distance is consistent with known structure of SERCA(Hou & Robia, 2010; Kelly et al.,

2008; Song et al., 2011). These observations suggest that FRET is not likely caused by aggregation of the fluorescent probes which has been suggested to skew FRET measurements (Chen et al., 2006).

These results are more in line with the Subunit Model and the previous FRET studies where high levels of PLB-SERCA binding are detected even in the presence of

Ca2+ (Mueller et al., 2004). Mueller et al observed only a minimal decrease in FRET in the presence of Ca2+ at the highest protein concentrations rather over the full range of concentration as was observed with our approach. This decrease in was consistent with a small structure change of the regulatory complex corresponding to approximately a 1

Å movement that was not detected with our approach. Unlike our results, Mueller et al

did not observe a change in PLB-SERCA affinity, but it is possible that their method did

not assess interaction at sufficiently low protein concentrations where the affinity effect

would be observed. Although we observed no change in our distance measurements, a

1 Å movement is within the error of our method. The larger size and rotational mobility

of the fluorescent proteins used in our method may mask the subtle change Mueller et

al observed using small, covalently attached fluorophores.

Just like Ca2+, the presence of the SERCA inhibitor Tg has been shown to

completely abolish PLB-SERCA crosslinking (Chen et al., 2006, 2005) but only slightly

diminish FRET in our system. Like Ca2+, we observed that Tg reduced PLB-SERCA affinity

(increased KD2) with no change in the structure of the complex (no change in FRETmax)

(Fig. 15). While assessing the ability of PLB and SERCA to interact in the presence of Tg

does not assist in the modeling the PLB-SERCA mechanism, it does aid in the validation

of our FRET approach. Because Tg is membrane permeable, its application is a more

straight forward protocol as compared to Ca2+ which requires saponin permeabilization

to introduce artificial cytosolic solutions. Our observation of a lack of complete PLB-

SERCA dissociation at elevated Ca2+ concentrations is likely not due to insufficient

permeabilization or inappropriate solutions since Tg produced a comparable effect.

Early proteolysis studies displayed no difference in the fragmentation patterns of

SERCA at low Ca2+ concentrations and in the presence of Tg (Sagara & Inesi, 1991). This observation was thought to indicate that SERCA bound by Tg was comparable to the E2

Ca2+ free conformation. Tg was then used to stabilize SERCA in what was assumed to be

the E2 conformation for some of the first high resolution crystallographic studies

(Danko, Yamasaki, Daiho, Suzuki, & Toyoshima, 2001; Toyoshima & Nomura, 2002).

However, current consensus suggests that SERCA bound by Tg is not representative of

the E2 conformation, and alternative crystal structures of the E2 conformation have been developed in the absence of Tg. Our Tg dose response curve (Fig. 15B), showing

two PLB affinity states, also suggests that SERCA bound by Tg is a structural

conformation distinct from the E2 Ca2+ free conformation. We have named this

conformation “E2-Tg” for its similarity to, yet distinct from, the E2 Ca2+ free

conformation. The E2 and E2-Tg conformations likely have similar cytosolic orientation

exposure the same cleavage sites, yielding the same banding pattern. They however

differ greatly in the transmembrane region where the strength of the PLB-SERCA

interaction is likely dictated.

Both the Ca2+ and Tg effects on our FRET measurements occur in the same range

of concentrations as the crosslinking approach. This similarity suggests that both approaches are likely detecting the same change in PLB-SERCA binding, but are just

detected by a different signal. This consistency helps validates our overall FRET

approach showing that the changes we observe are not unique to our system.

Because of the predicted large structure changes in the cytosolic headpiece

between the E1 and E2 conformations of SERCA, we thought it likely that this could

generate a large structural change in cytosolic region of the PLB regulatory complex.

However, we did not observe a change in the FRETmax parameter, indicating no change

97 in the position of the fluorescent probes attached to the N-terminus of both PLB and

SERCA, loci that would be thought to move apart. During the E2 to E1 transition, there could be a coordinated movement of both fluorescent probes that does not change the distance between them. On the other hand several studies have suggested that the movement of the SERCA headpiece is much smaller than initially predicted (Jensen et al., 2006; Sørensen et al., 2004; Winters et al., 2008). These more subtle changes could also explain the lack of a FRETmax change in our experiments

Under no conditions were we able to detect complete loss of PLB-SERCA FRET.

In particular, we observed a high degree of FRET between PLB and SERCA under conditions that favor particular conformation (E1 and E2-Tg) that have been shown to block chemical crosslinking. These crosslinking observations are the backbone of the

Dissociation Model. In contrast, our evidence suggests the high likelihood that PLB can bind all conformations of SERCA, although with varying affinities. This conclusion is summarized in Fig. 23 showing the coupled equilibrium between PLB pentamer, monomer, and monomer bound to various conformations of SERCA with independent dissociation constants.

While we have not performed an exhaustive investigation into all possible conformations, we have observed a high level of FRET even at saturating Ca2+ concentrations known to fully relieve PLB inhibition of SERCA. This suggests that PLB can remain bound throughout the reaction cycle SERCA, implying ability to bind all conformations. We therefore conclude that unbinding of PLB is not required for relief

Figure 23: Summary of PLB binding. PLB monomer can bind to itself forming a pentamer or bind to SERCA in any conformation, however doing so with varying affinity.

99 of PLB-SERCA inhibition. This is not without precedence since it phosphorylation of PLB has been suggested to relieve inhibition without causing dissociation of the regulatory complex.

Alternative Binding Site

In our Ca2+ dose curve, we detect two different modes PLB-SERCA binding. The change in binding states could describe a simple structural reorientation merely decreasing the strength of the intermolecular contacts of PLB at the same location on

SERCA. On the other hand, the detected change could be described by displacement of

PLB to an alternative binding site. Crosslinking by the Jones lab (Chen et al., 2007, 2006,

2005; Jones et al., 2002) and mutagenesis studies by the MacLennan lab (Y Kimura et al.,

1996; MacLennan, Kimura, & Toyofuku, 1998; Toyofuku et al., 1993) provide good evidence for a site of interaction that also likely facilitates inhibition. To this point, there has been no direct evidence of an alternative binding site for PLB on SERCA.

However, some evidence for an alternative site comes from X-ray crystallography studies predicting the binding of homologous proteins, NKA and PLM, at a more exterior location on helix M9 (Fig. 7) (Shinoda et al., 2009). While this provides sufficient evidence to hypothesize the existence of two distinct PLB binding sites on

SERCA, it certainly does not confirm it. This exterior binding site could be unique to the

PLM-NKA complex and could be the sole site of interaction. However, it seems unlikely that proteins with such similarities would have vastly different interactions and inhibitory mechanisms. The more likely explanation is that this detected interaction is

100

just an artifact of crystallization and does not describe the actual functional site of

interaction between NKA and PLM.

One could argue that interaction at the Ca2+ binding site is the only important

interaction. The displacement of PLB to an alternative binding site and the complete

displacement of PLB from SERCA are not significantly different, since both describe a movement out of the functional site. However, this difference would greatly alter the

dynamics of the system. Binding of PLB to an alternative site would enhance the on-rate

of PLB by keeping it in close proximity to the functional site. In other words, its effective

concentration would be much higher.

The Jones group has questioned our proposal of an alternative binding site for

PLB since their crosslinking studies have found no evidence of a second site (Akin &

Jones, 2012). However, they have yet to look for enhancement of crosslinking at an exterior site by Ca2+. Their evidence is the use of crosslinkers varying in distance from 0-

15 Å which have not shown differential binding. However, movement of PLB by 20 Å

places it just outside the predicted binding groove and our exterior site corresponds to a

movement of 40 Å.

The role of this potential binding site is unclear and could be merely serve as a

place holder to enhance on and off rates of PLB. The importance as a place holder

should not be underestimated, because variations in the ability to bind the alternative

site will likely have a tremendous impact on how PLB interacts with the functional site.

Changes at the alternative site would alter the equilibrium binding producing a different

101

stead-state interaction and inhibition. Also, there would be a significant impact on the dynamics of inhibition since the on and off rates at the functional site could be altered.

Besides affecting how PLB interacts at the primary functional site, the secondary site itself may have the ability to functionally regulate SERCA. Whatever the purpose of the

secondary site, the next step of the investigation is determining whether it exists,

followed by a characterization of the site. We could then attempt to alter the

interaction at the secondary site and then determine how the ability of PLB to regulate

SERCA is changed. If this secondary site is confirmed to exist, an intriguing question that

results is how binding at one site excludes binding at the other site, since a strict

PLB:SERCA stoichiometry of 1:1 is observed.

Proposed Model of PLB Inhibition

We have attempted to reconcile the findings of others to account for the

limitations of all approaches and develop a model that is consistent with all data in the

field. We propose that high Ca2+ displaces the transmembrane domain of PLB to an

alternative, lower affinity binding site on SERCA, with no change in the position of the

cytosolic domain. We have termed this the “Domain Displacement Model.” In other

words, binding to the E1 and E2 conformations of SERCA differs only in the position and

affinity of the transmembrane domain of PLB. This accounts for our data showing that

Ca2+ reduces the affinity of PLB for SERCA. This also explains why our data and FRET

measurements from Mueller et al detect no or only a small change in the position of the

cytosolic domain where the fluorophores are attached. Movement of the

102

transmembrane domain to a more exterior site would also explain why Jones sees no

crosslinking at high Ca2+. The differences between the models are summarized in Fig.

24.

We started to directly test this model by measuring FRET between SERCA and a truncated PLB lacking the cytosolic domain which would assess the position of the transmembrane domain of PLB in association with SERCA. Our YFP-tmPLB construct expresses normally and interacts with itself and SERCA at low Ca2+ concentrations as

anticipated. However, FRET between Cer-SERCA and YFP-tmPLB is unresponsive to Ca2+

and does not appear to regulate SERCA. So while YFP-tmPLB retains some of the

binding characteristics of full-length PLB, clear differences are apparent. We suspect

that steric barriers created by the membrane and SERCA may be preventing proper

localization, and therefore function, of YFP-tmPLB. Further experimentation is needed

to address these results including additional characterization of the YFP-tmPLB

construct as well as development of alternative constructs. While this construct did not

function as initially hypothesized, it does provide evidence for a non-functional PLB-

SERCA interaction. Assessment of how this interaction occurs without inhibition of

SERCA could also be critical to our understanding of the PLB-SERCA mechanism.

Earlier studies showing the ability of truncated PLB to inhibit SERCA were

performed at a high excess of PLB peptide (Hughes et al., 1996; Y Kimura et al., 1996),

with no assessment of its interaction or affinity. The conclusions from these studies that

103

Figure 24: Models of PLB-SERCA binding. The previously discussed Dissociation and Subunit Models are shown in comparison to our proposed Domain Displacement Model. We propose that elevated calcium concentration displaces the transmembrane domain of PLB to alternative SERCA binding site of lower affinity, with no change in the binding location of the cytosolic domain.

104 fully inhibit SERCA may have been incorrect, and even if the suspected steric issues with

YFP-tmPLB are corrected, full inhibition may not be achieved.

Effects of Pacing on Myocyte FRET

Irrespective of the mechanistic model of inhibition, it is critical to understand how the PLB-SERCA interaction changes over time. Because the ability of PLB to inhibit

SERCA varies with respect to Ca2+ concentration, the level of PLB inhibition will not be constant in cardiac myocytes. Not only is Ca2+ concentration perpetually changing over the course of individual Ca2+ transients; basal concentrations can also be altered by various stimuli. To date however, all measurements of the ability of PLB to bind to and regulate SERCA have been taken at equilibrium. So we have no comprehension of how rapidly the PLB-SERCA binding changes in response to changes in Ca2+ concentration.

Since Ca2+ concentration changes rapidly and constantly in the physiological environment of PLB and SERCA, we must assess the dynamics of the PLB-SERCA interaction to fully understand its physiologic function. Unlike previous studies, our

FRET approach is capable of assessing PLB-SERCA binding in real time. We are therefore uniquely poised to address this critical question.

We detected no change in PLB-SERCA FRET over the course of a single Ca2+ transient. Only by using a non-physiological rest/pace protocol that produced higher and more sustained Ca2+ increases were we able to observe changes in FRET. These observations were somewhat expected since our steady-state experiments predict at most a small 3% change in FRET in response to increases in Ca2+ concentration. These

105

observations could be explained in a few ways. It might indicate that there are two

significant pools of SERCA, PLB unbound (active) and PLB bound (inactive). This PLB

unbound and active pool of SERCA is sufficient to move the Ca2+ of a typical single

transient back into the SR. Only when Ca2+ concentrations are elevated during our rest/pace protocol, and additional SERCA activity is required, is PLB inhibition relieved

and a decrease in FRET observed. However, these distinct bound and unbound pools

are inconsistent with several studies that have suggested that SERCA is nearly 100%

bound by PLB under basal conditions in larger mammals (Waggoner et al., 2009). Our

observed pacing effects are more consistent with the scenario where dissociation of PLB

from SERCA is not required for the relief of inhibition. The SERCA activity necessary over

the course of a Ca2+ transient could then be achieved without a significant change in

PLB-SERCA interaction. Only when Ca2+ concentrations are elevated during our

rest/pace protocol is the PLB-SERCA affinity shifted sufficiently to generate a decrease in

FRET.

Lack of an observable change in FRET does not indicate an absence of change in

PLB-SERCA binding over the course of a single Ca2+ transient. Certainly, a change in

binding can occur that would not significantly alter FRET. As discussed previously

concerning an alternative binding site on SERCA, the position of the fluorescent protein

at the N-terminus of PLB does not allow detection of a change in the position of the

transmembrane domain. If applying our proposed Domain Displacement Model of PLB

interaction (Fig. 24), one could envision that the transmembrane domain of PLB

106 transitions rapidly between the two binding sites in response to changes in Ca2+, with no change in the position of the cytosolic domain. A YFP-tmPLB that could successfully produce the anticipated response to Ca2+ would be a useful tool to investigate potential changes in the position of the transmembrane domain in response to Ca2+ transients.

However, if successful, this type of experiment should be viewed with caution since the cytosolic domain could facilitate any transition. Its absence could certainly then affect the rates of transition.

Future Directions

While our proposed model of PLB inhibition of SERCA is a plausible explanation that could integrate our data with previously published research, we lack sufficient support for this model. We must first determine whether the secondary site exists and whether Ca2+ can cause transition to this binding site. The FRET experiments with the truncated PLB could be capable of showing both, but the initial experiments have not produced the anticipated results. We suspect that steric hindrance of the fluorescent protein may be preventing proper localization of tmPLB. If correct, tmPLB lacking YFP should then be able to properly interact and inhibit SERCA. This could be tested by developing an unlabeled peptide consisting of only residues 28-52 of PLB probing the ability of this peptide to inhibit SERCA. If unlabeled tmPLB is not functional, this would directly challenge previous observations that the transmembrane domain in isolation is fully capable of inhibiting SERCA (Hughes et al., 1996; Y Kimura et al., 1996). Though this seems unlikely since an analogous protein sarcolipin, which lacks a comparable

107 cytosolic domain, is able to inhibit SERCA activity (Bhupathy et al., 2007; Periasamy,

Bhupathy, & Babu, 2008).

Also, if FRET between YFP-tmPLB and Cer-SERCA is diminished by expression of unlabeled PLB or tmPLB (if functional), this would demonstrate binding competition.

This would indicate that the YFP-tmPLB interaction with SERCA is not anomalous and not an artifact of our approach. Furthermore, it could indicate that YFP-tmPLB preferentially interacts at the secondary, non-inhibitory site.

If steric hindrance of YFP is the contributing factor preventing ability of YFP- tmPLB to inhibit SERCA, modifications to the construct could relieve the steric block.

Specifically, the 7 amino acid linker between tmPLB and YFP could be lengthened so that

YFP could be positioned further away from the membrane. Use of the shortest linker that is able to allow proper inhibition of SERCA would be ideal. This would alleviate some potential variability in FRET measurements produced flexibility and movement of the fluorescent protein. In order to be versatile, the additional linker sequence should be comprised of simple amino acids such as glycines. Use of prolines could also be considered since they may provide some stability to the linker, thereby reducing flexibility and potential variability of FRET measurements.

Rather than developing PLB variants to investigate the potential secondary site, we could also make modified SERCA proteins. We could make targeted mutations of helix M9, in an attempt to disrupt binding at the proposed secondary site. Diminished binding at the secondary site would be observed as a further lowering of affinity at high

108

Ca2+ concentrations compared to non-mutated SERCA. Furthermore, if PLB exists in equilibrium between the two sites, decreased affinity at the secondary site could

enhance affinity at the primary functional site presenting as increased affinity at low

Ca2+ concentrations and enhanced inhibition by PLB.

SERCA3 could provide for an interesting model for the investigation of the PLB-

SERCA mechanism. While PLB has been clearly shown to have no effect on the activity

of SERCA3 in functional assays (Y Kimura et al., 1996; Toyofuku et al., 1994), the ability

of PLB to bind SERCA3 has not been determined. Measuring FRET between YFP-PLB

and Cer-SERCA3 might provide a unique perspective, potentially showing a non-

inhibitory binding interaction. Perhaps this interaction would be unresponsive to Ca2+

suggesting PLB can only bind the secondary, non-inhibitory site.

Our FRET results and proposed model would greatly benefit from support of alternative approaches. While crosslinking results to date have directly contrasted our findings, we believe this approach could be modified to show binding at a secondary site. Jones et al have not found evidence of a secondary binding site, but this could be attributed to the fact that they have yet to directly test the existence of an alternative binding site. All studies have targeted residues in the suspected binding pocket of

M2/M4/M6 on SERCA. Binding at a more exterior position such as M9 would go undetected. Therefore, use of untargeted reagents such as benzophenone and glutaraldehyde which bind carbon and amine groups may successfully crosslink PLB to

SERCA at a position outside the proposed binding pocket under saturating Ca2+

109

conditions. Additionally, SERCA mutants could be generated that introduce cysteines or

lysines on helix M9 to perform so that the site directed crosslinking protocols of the

Jones lab could be used.

Concluding Remarks

In this dissertation, I have shown our evidence that suggests that dissociation of

PLB from SERCA is not required for relief of inhibition. To account for this observation

as well as previous data from other labs, I have also described the Domain Displacement

Model of regulation. Furthermore, I have also demonstrated the ability to detect PLB-

SERCA binding in real-time in response to physiological changes of Ca2+ concentration.

Doing so, I have shown evidence that single Ca2+ transients of paced cardiac myocytes do not produce an observable change in PLB-SERCA binding. The results and methodology presented here have added to our understanding of PLB-SERCA regulation and continuation of this research is poised to answer critical questions in the field.

BIBLIOGRAPHY

Adunyah, S. E., Jones, L. R., & Dean, W. L. (1988). Structural and functional comparison of a 22 kDa protein from internal human platelet membranes with cardiac phospholamban. Biochimica et biophysica acta, 941(1), 63-70. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/2835982

Akin, B. L., & Jones, L. R. (2012). Characterizing Phospholamban to SERCA2a Binding Interactions in Human Cardiac Sarcoplasmic Reticulum Vesicles Using Chemical Cross-linking. The Journal of biological chemistry. doi:10.1074/jbc.M111.334987

Allen, G., Trinnaman, B. J., & Green, N. M. (1980). The primary structure of the calcium ion-transporting adenosine triphosphatase protein of rabbit skeletal sarcoplasmic reticulum. Peptides derived from digestion with cyanogen bromide, and the sequences of three long extramembranous segments. The Biochemical journal, 187(3), 591-616. Retrieved from http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1162443&tool=pmcen trez&rendertype=abstract

Antipenko, A. Y., Spielman, a. I., & Kirchberger, M. A. (1999). Kinetic Differences in the Phospholamban-Regulated Calcium Pump When Studied in Crude and Purified Cardiac Sarcoplasmic Reticulum Vesicles. Journal of Membrane Biology, 167(3), 257-265. doi:10.1007/s002329900490

Antipenko, A. Y., Spielman, A. I., Sassaroli, M., & Kirchberger, M. A. (1997). Comparison of the kinetic effects of phospholamban phosphorylation and anti-phospholamban monoclonal antibody on the calcium pump in purified cardiac sarcoplasmic reticulum membranes. Biochemistry, 36(42), 12903-10. doi:10.1021/bi971109v

Arvanitis, D. a, Vafiadaki, E., Fan, G.-C., Mitton, B., Gregory, K. N., Del Monte, F., Kontrogianni-Konstantopoulos, A., et al. (2007). Histidine-rich Ca-binding protein interacts with sarcoplasmic reticulum Ca-ATPase. American journal of physiology. Heart and circulatory physiology, 293(3), H1581-9. doi:10.1152/ajpheart.00278.2007

Asahi, M, Kimura, Y., Kurzydlowski, K., Tada, M., & MacLennan, D. H. (1999). Transmembrane helix M6 in sarco(endo)plasmic reticulum Ca(2+)-ATPase forms a 110

111

functional interaction site with phospholamban. Evidence for physical interactions at other sites. The Journal of biological chemistry, 274(46), 32855-62. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/10551848

Asahi, M, McKenna, E., Kurzydlowski, K., Tada, M., & MacLennan, D. H. (2000). Physical interactions between phospholamban and sarco(endo)plasmic reticulum Ca2+- ATPases are dissociated by elevated Ca2+, but not by phospholamban phosphorylation, vanadate, or thapsigargin, and are enhanced by ATP. The Journal of biological chemistry, 275(20), 15034-8. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/10809745

Baskin, R. J., & Deamer, D. W. (1969). Comparative ultrastructure and calcium transport in heart and skeletal muscle microsomes. The Journal of cell biology, 43(3), 610-7. Retrieved from http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2107794&tool=pmcen trez&rendertype=abstract

Bers, D. M. (2008). Calcium cycling and signaling in cardiac myocytes. Annual review of physiology, 70, 23-49. doi:10.1146/annurev.physiol.70.113006.100455

Bers, D. M., & Despa, S. (2009). Na/K-ATPase--an integral player in the adrenergic fight- or-flight response. Trends in cardiovascular medicine, 19(4), 111-8. Elsevier Inc. doi:10.1016/j.tcm.2009.07.001

Bhupathy, P., Babu, G. J., & Periasamy, M. (2007). Sarcolipin and phospholamban as regulators of cardiac sarcoplasmic reticulum Ca2+ ATPase. Journal of molecular and cellular cardiology, 42(5), 903-11. doi:10.1016/j.yjmcc.2007.03.738

Bidlack, J. M., & Shamoo, A. E. (1980). Adenosine 3’,5'-monophosphate-dependent phosphorylation of a 6000 and a 22,000 dalton protein from cardiac sarcoplasmic reticulum. Biochimica et biophysica acta, 632(2), 310-25. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/6251912

Bidlack, J. M., Ambudkar, I. S., & Shamoo, a E. (1982). Purification of phospholamban, a 22,000-dalton protein from cardiac sarcoplasmic reticulum that is specifically phosphorylated by cyclic AMP-dependent protein kinase. The Journal of biological chemistry, 257(8), 4501-6. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/6121803

Bidwell, P., Blackwell, D. J., Hou, Z., Zima, A. V., & Robia, S. L. (2011). Phospholamban binds with differential affinity to calcium pump conformers. The Journal of biological chemistry, 286(40), 35044-50. doi:10.1074/jbc.M111.266759

112

Brandl, C. J., Green, N. M., Korczak, B., & MacLennan, D. H. (1986). Two Ca2+ ATPase genes: homologies and mechanistic implications of deduced amino acid sequences. Cell, 44(4), 597-607. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/2936465

Brini, M., & Carafoli, E. (2009). Calcium pumps in health and disease. Physiological reviews, 89(4), 1341-78. Am Physiological Soc. doi:10.1152/physrev.00032.2008

Brittsan, a G., Carr, a N., Schmidt, A. G., & Kranias, E. G. (2000). Maximal inhibition of SERCA2 Ca(2+) affinity by phospholamban in transgenic hearts overexpressing a non-phosphorylatable form of phospholamban. The Journal of biological chemistry, 275(16), 12129-35. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/10766848

Cantilina, T., Sagara, Y., Inesi, G., & Jones, L. R. (1993). Comparative studies of cardiac and skeletal sarcoplasmic reticulum ATPases. Effect of a phospholamban antibody on enzyme activation by Ca2+. The Journal of biological chemistry, 268(23), 17018- 25. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/8349590

Carvalho, M. G., de Souza, D. G., & de Meis, L. (1976). On a possible mechanism of energy conservation in sarcoplasmic reticulum membrane. The Journal of biological chemistry, 251(12), 3629-36. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/932000

Chen, Z., Akin, B. L., & Jones, L. R. (2007). Mechanism of reversal of phospholamban inhibition of the cardiac Ca2+-ATPase by protein kinase A and by anti- phospholamban monoclonal antibody 2D12. The Journal of biological chemistry, 282(29), 20968-76. doi:10.1074/jbc.M703516200

Chen, Z., Akin, B. L., Stokes, D. L., & Jones, L. R. (2006). Cross-linking of C-terminal residues of phospholamban to the Ca2+ pump of cardiac sarcoplasmic reticulum to probe spatial and functional interactions within the transmembrane domain. The Journal of biological chemistry, 281(20), 14163-72. doi:10.1074/jbc.M601338200

Chen, Z., Stokes, D. L., & Jones, L. R. (2005). Role of leucine 31 of phospholamban in structural and functional interactions with the Ca2+ pump of cardiac sarcoplasmic reticulum. The Journal of biological chemistry, 280(11), 10530-9. doi:10.1074/jbc.M414007200

Chen, Z., Stokes, D. L., Rice, W. J., & Jones, L. R. (2003). Spatial and dynamic interactions between phospholamban and the canine cardiac Ca2+ pump revealed with use of

113

heterobifunctional cross-linking agents. The Journal of biological chemistry, 278(48), 48348-56. doi:10.1074/jbc.M309545200

Chu, G., Lester, J. W., Young, K. B., Luo, W., Zhai, J., & Kranias, E. G. (2000). A single site (Ser16) phosphorylation in phospholamban is sufficient in mediating its maximal cardiac responses to beta -agonists. The Journal of biological chemistry, 275(49), 38938-43. doi:10.1074/jbc.M004079200

Clarke, D. M., Loo, T. W., & MacLennan, D. H. (1990). Functional consequences of alterations to amino acids located in the nucleotide binding domain of the Ca2(+)- ATPase of sarcoplasmic reticulum. The Journal of biological chemistry, 265(36), 22223-7. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/2148317

Clarke, D. M., Maruyama, K., Loo, T. W., Leberer, E., Inesi, G., & MacLennan, D. H. (1989). Functional consequences of glutamate, aspartate, glutamine, and asparagine mutations in the stalk sector of the Ca2+-ATPase of sarcoplasmic reticulum. The Journal of biological chemistry, 264(19), 11246-51. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/2567733

Coan, C. R., & Inesi, G. (1976). Ca2+-dependent effect of acetylphosphate on spin- labeled sarcoplasmic reticulum. Biochemical and biophysical research communications, 71(4), 1283-8. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/184786

Colyer, J., & Wang, J. H. (1991). Dependence of cardiac sarcoplasmic reticulum calcium pump activity on the phosphorylation status of phospholamban. The Journal of biological chemistry, 266(26), 17486-93. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/1832673

Dally, S., Corvazier, E., Bredoux, R., Bobe, R., & Enouf, J. (2010). Multiple and diverse coexpression, location, and regulation of additional SERCA2 and SERCA3 isoforms in nonfailing and failing human heart. Journal of molecular and cellular cardiology, 48(4), 633-44. Elsevier Ltd. doi:10.1016/j.yjmcc.2009.11.012

Danko, S., Yamasaki, K., Daiho, T., Suzuki, H., & Toyoshima, C. (2001). Organization of cytoplasmic domains of sarcoplasmic reticulum Ca(2+)-ATPase in E(1)P and E(1)ATP states: a limited proteolysis study. FEBS letters, 505(1), 129-35. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/11557055

Davis, B. a, Schwartz, a, Samaha, F. J., & Kranias, E. G. (1983). Regulation of cardiac sarcoplasmic reticulum calcium transport by calcium-calmodulin-dependent

114

phosphorylation. The Journal of biological chemistry, 258(22), 13587-91. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/2982423

Deamer, D. W., & Baskin, R. J. (1969). Ultrastructure of sarcoplasmic reticulum preparations. The Journal of cell biology, 42(1), 296-307. Retrieved from http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2107567&tool=pmcen trez&rendertype=abstract

Del Monte, F., Hajjar, R. J., Harding, S. E., & Inesi, G. (2001). Overwhelming Evidence of the Beneficial Effects of SERCA Gene Transfer in Heart Failure Response. Circulation Research, 88(11), e66-e67. doi:10.1161/hh1101.092004

Domeier, T. L., Blatter, L. a, & Zima, A. V. (2009). Alteration of sarcoplasmic reticulum Ca2+ release termination by ryanodine receptor sensitization and in heart failure. The Journal of physiology, 587(Pt 21), 5197-209. doi:10.1113/jphysiol.2009.177576

Ebashi, S., & Lipmann, F. (1962). ADENOSINE TRIPHOSPHATE-LINKED CONCENTRATION OF CALCIUM IONS IN A PARTICULATE FRACTION OF RABBIT MUSCLE. The Journal of cell biology, 14(3), 389-400. Retrieved from http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2106118&tool=pmcen trez&rendertype=abstract

Fujii, J., Kadoma, M., Tada, M., Toda, H., & Sakiyama, F. (1986). Characterization of structural unit of phospholamban by amino acid sequencing and electrophoretic analysis. Biochemical and biophysical research communications, 138(3), 1044-50. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/3753485

Fujii, J., Maruyama, K., Tada, M., & MacLennan, D. H. (1989). Expression and site-specific mutagenesis of phospholamban. Studies of residues involved in phosphorylation and pentamer formation. The Journal of biological chemistry, 264(22), 12950-5. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/2502544

Fujii, J., Ueno, A., Kitano, K., Tanaka, S., Kadoma, M., & Tada, M. (1987). Complete complementary DNA-derived amino acid sequence of canine cardiac phospholamban. The Journal of clinical investigation, 79(1), 301-4. doi:10.1172/JCI112799

Gordon, a M., Homsher, E., & Regnier, M. (2000). Regulation of contraction in striated muscle. Physiological reviews, 80(2), 853-924. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/10747208

115

Ha, K. N., Masterson, L. R., Hou, Z., Verardi, R., Walsh, N., Veglia, G., & Robia, S. L. (2011). Lethal Arg9Cys phospholamban mutation hinders Ca2+-ATPase regulation and phosphorylation by protein kinase A. Proceedings of the National Academy of Sciences of the United States of America, 108(7), 2735-40. doi:10.1073/pnas.1013987108

Hagemann, D., Kuschel, M., Kuramochi, T., Zhu, W., Cheng, H. C., & Xiao, R. P. (2000). Frequency-encoding Thr17 phospholamban phosphorylation is independent of Ser16 phosphorylation in cardiac myocytes. The Journal of biological chemistry, 275(29), 22532-6. doi:10.1074/jbc.C000253200

Hasenfuss, G. (1998). Alterations of calcium-regulatory proteins in heart failure. Cardiovascular research, 37(2), 279-89. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/9614485

Hasselbach, W. (1978). The reversibility of the sarcoplasmic calcium pump. Biochimica et biophysica acta, 515(1), 23-53. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/147710

Hasselbach, W., & Makinose, M. (1962). [The calcium pump of the “relaxing granules” of muscle and its dependence on ATP-splitting.]. Biochemische Zeitschrift, 333, 518- 28. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/13712164

Hasselbach, W., & Makinose, M. (1963). [on the Mechanism of Calcium Transport Across the Membrane of the Sarcoplasmic Reticulum]. Biochemische Zeitschrift, 339, 94- 111. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/14095160

Hobai, I. A., & Rourke, B. O. (2000). Canine Pacing-Induced Heart Failure. Circulation Research, 690-698. doi:10.1161/01.RES.87.8.690

Hoshijima, M., Ikeda, Y., Iwanaga, Y., Minamisawa, S., Date, M.-o, Gu, Y., Iwatate, M., et al. (2002). Chronic suppression of heart-failure progression by a pseudophosphorylated mutant of phospholamban via in vivo cardiac rAAV gene delivery. Nature medicine, 8(8), 864-71. doi:10.1038/nm739

Hou, Z., & Robia, S. L. (2010). Relative affinity of calcium pump isoforms for phospholamban quantified by fluorescence resonance energy transfer. Journal of molecular biology, 402(1), 210-6. Elsevier Ltd. doi:10.1016/j.jmb.2010.07.023

Hou, Z., Kelly, E. M., & Robia, S. L. (2008). Phosphomimetic mutations increase phospholamban oligomerization and alter the structure of its regulatory complex.

116

The Journal of biological chemistry, 283(43), 28996-9003. doi:10.1074/jbc.M804782200

Houser, S. R., Piacentino, V., Mattiello, J., Weisser, J., & Gaughan, J. P. (2000). Functional properties of failing human ventricular myocytes. Trends in cardiovascular medicine, 10(3), 101-7. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/11427996

Hughes, G., East, J. M., & Lee, a G. (1994). The hydrophilic domain of phospholamban inhibits the Ca2+ transport step of the Ca(2+)-ATPase. The Biochemical journal, 303 ( Pt 2, 511-6. Retrieved from http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1137357&tool=pmcen trez&rendertype=abstract

Hughes, G., Starling, a P., Sharma, R. P., East, J. M., & Lee, a G. (1996). An investigation of the mechanism of inhibition of the Ca(2+)-ATPase by phospholamban. The Biochemical journal, 318 ( Pt 3, 973-9. Retrieved from http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1217713&tool=pmcen trez&rendertype=abstract

Hutter, M. C., Krebs, J., Meiler, J., Griesinger, C., Carafoli, E., & Helms, V. (2002). A structural model of the complex formed by phospholamban and the calcium pump of sarcoplasmic reticulum obtained by molecular mechanics. Chembiochem : a European journal of chemical biology, 3(12), 1200-8. doi:10.1002/1439- 7633(20021202)3:12<1200::AID-CBIC1200>3.0.CO;2-H

Inesi, G., Kurzmack, M., Coan, C. R., & Lewis, D. E. (1980). Cooperative calcium binding and ATPase activation in sarcoplasmic reticulum vesicles. The Journal of biological chemistry, 255(7), 3025-31. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/6244305

Jackson, W. A., & Colyer, J. (1996). Translation of Ser16 and Thr17 phosphorylation of phospholamban into Ca 2+-pump stimulation. The Biochemical journal, 316 ( Pt 1, 201-7. Retrieved from http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1217323&tool=pmcen trez&rendertype=abstract

James, P., Inui, M., Tada, M., Chiesi, M., & Carafoli, E. (1989). Nature and site of phospholamban regulation of the Ca2+ pump of sarcoplasmic reticulum. Nature, 342(6245), 90-92. Retrieved from http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt =Citation&list_uids=2530454

117

Jardetzky, O. (1966). Simple allosteric model for membrane pumps. Nature, 211(5052), 969-70. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/5968307

Jensen, A.-M. L., Sørensen, T. L.-M., Olesen, C., Møller, J. V., & Nissen, P. (2006). Modulatory and catalytic modes of ATP binding by the calcium pump. The EMBO journal, 25(11), 2305-14. doi:10.1038/sj.emboj.7601135

Jones, L. R., Besch, R., Fleming, W., & Mcconnaughey, M. (1979). Separation of Vesicles of Cardiac Sarcoplasmic Reticulum from Vesicles of Cardiac, 254(2), 530-539.

Jones, L. R., Cornea, R. L., & Chen, Z. (2002). Close proximity between residue 30 of phospholamban and cysteine 318 of the cardiac Ca2+ pump revealed by intermolecular thiol cross-linking. The Journal of biological chemistry, 277(31), 28319-29. doi:10.1074/jbc.M204085200

Kanazawa, T., & Boyer, P. D. (1973). Occurrence and characteristics of a rapid exchange of phosphate oxygens catalyzed by sarcoplasmic reticulum vesicles. The Journal of biological chemistry, 248(9), 3163-72. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/4267042

Karim, C. B., Paterlini, M. G., Reddy, L. G., Hunter, G. W., Barany, G., & Thomas, D. D. (2001). Role of cysteine residues in structural stability and function of a transmembrane helix bundle. The Journal of biological chemistry, 276(42), 38814-9. doi:10.1074/jbc.M104006200

Karim, C. B., Stamm, J. D., Karim, J., Jones, L. R., & Thomas, D. D. (1998). Cysteine reactivity and oligomeric structures of phospholamban and its mutants. Biochemistry, 37(35), 12074-81. doi:10.1021/bi980642n

Kelly, E. M., Hou, Z., Bossuyt, J., Bers, D. M., & Robia, S. L. (2008). Phospholamban oligomerization, quaternary structure, and sarco(endo)plasmic reticulum calcium ATPase binding measured by fluorescence resonance energy transfer in living cells. The Journal of biological chemistry, 283(18), 12202-11. doi:10.1074/jbc.M707590200

Kimura, Y, Asahi, M., Kurzydlowski, K., Tada, M., & MacLennan, D. H. (1998). Phospholamban domain Ib mutations influence functional interactions with the Ca2+-ATPase isoform of cardiac sarcoplasmic reticulum. The Journal of biological chemistry, 273(23), 14238-41. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/9603928

118

Kimura, Y, Inui, M., Kadoma, M., Kijima, Y., Sasaki, T., & Tada, M. (1991). Effects of monoclonal antibody against phospholamban on calcium pump ATPase of cardiac sarcoplasmic reticulum. Journal of molecular and cellular cardiology, 23(11), 1223- 30. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/1666413

Kimura, Y, Kurzydlowski, K., Tada, M., & MacLennan, D. H. (1996). Phospholamban regulates the Ca2+-ATPase through intramembrane interactions. The Journal of biological chemistry, 271(36), 21726-31. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/10551848

Kimura, Y, Kurzydlowski, K., Tada, M., & MacLennan, D. H. (1997). Phospholamban inhibitory function is activated by depolymerization. The Journal of biological chemistry, 272(24), 15061-4. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/9182523

Kirchberger, M. A., Tada, M., & Katz, a M. (1975). Phospholamban: a regulatory protein of the cardiac sarcoplasmic reticulum. Recent advances in studies on cardiac structure and metabolism, 5, 103-15. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/127351

Kirchberger, M. A., Tada, M., & Katz, A. M. (1974). Adenosine 3’:5'-monophosphate- dependent protein kinase-catalyzed phosphorylation reaction and its relationship to calcium transport in cardiac sarcoplasmic reticulum. The Journal of biological chemistry, 249(19), 6166-73. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/4371435

Kirchberger, M. A., Tada, M., Repke, D. I., & Katz, A. M. (1972). Cyclic adenosine 3’,5'- monophosphate-dependent protein kinase stimulation of calcium uptake by canine cardiac microsomes. Journal of molecular and cellular cardiology, 4(6), 673-80. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/4347149

Kranias, E. G. (1985). Regulation of Ca2+ transport by cyclic 3’,5'-AMP-dependent and calcium-calmodulin-dependent phosphorylation of cardiac sarcoplasmic reticulum. Biochimica et biophysica acta, 844(2), 193-9. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/2982423

Kranias, E. G., & Bers, D. M. (2007). Calcium and cardiomyopathies. Sub-cellular biochemistry, 45, 523-37. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/18193651

Kranias, E. G., & Di Salvo, J. (1986). A phospholamban protein phosphatase activity associated with cardiac sarcoplasmic reticulum. The Journal of biological chemistry,

119

261(22), 10029-32. ASBMB. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/12111240

Kühlbrandt, W. (2004). Biology, structure and mechanism of P-type ATPases. Nature reviews. Molecular cell biology, 5(4), 282-95. doi:10.1038/nrm1354

Lakowicz, J. R. (2006) Principles of Fluorescence Spectroscopy, Third Ed., Springer Publishing Company, New York, NY.

Lamers, J. M., & Stinis, J. T. (1980). Phosphorylation of low molecular weight proteins in purified preparations of rat heart sarcolemma and sarcoplasmic reticulum. Biochimica et biophysica acta, 624(2), 443-59. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/6251900

Landgraf, W. C., & Inesi, G. (1969). ATP dependent conformational change in “spin labelled” sarcoplasmic reticulum. Archives of biochemistry and biophysics, 130(1), 111-8. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/4305153

Le Peuch, C. J., Haiech, J., & Demaille, J. G. (1979). Concerted regulation of cardiac sarcoplasmic reticulum calcium transport by cyclic adenosine monophosphate dependent and calcium--calmodulin-dependent phosphorylations. Biochemistry, 18(23), 5150-7. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/227448

Lu, Y. Z., & Kirchberger, M. A. (1994). Effects of a nonionic detergent on calcium uptake by cardiac microsomes. Biochemistry, 33(17), 5056-62. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/8172881

Lu, Y. Z., Xu, Z. C., & Kirchberger, M. A. (1993). Evidence for an effect of phospholamban on the regulatory role of ATP in calcium uptake by the calcium pump of the cardiac sarcoplasmic reticulum. Biochemistry, 32(12), 3105-11. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/8384487

Luo, W., Chu, G., Sato, Y., Zhou, Z., Kadambi, V. J., & Kranias, E. G. (1998). Transgenic Approaches to Define the Functional Role of Dual Site Phospholamban Phosphorylation *. Biochemistry, 273(8), 4734-4739.

Lygren, B., Carlson, C. R., Santamaria, K., Lissandron, V., McSorley, T., Litzenberg, J., Lorenz, D., et al. (2007). AKAP complex regulates Ca2+ re-uptake into heart sarcoplasmic reticulum. EMBO reports, 8(11), 1061-7. doi:10.1038/sj.embor.7401081

120

MacDougall, L. K., Jones, L. R., & Cohen, P. (1991). Identification of the major protein phosphatases in mammalian cardiac muscle which dephosphorylate phospholamban. European journal of biochemistry / FEBS, 196(3), 725-34. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/1849481

MacLennan, D. H. (1970). Purification and properties of an adenosine triphosphatase from sarcoplasmic reticulum. The Journal of biological chemistry, 245(17), 4508-18. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/4250726

MacLennan, D. H. (1990). Molecular tools to elucidate problems in excitation- contraction coupling. Biophysical journal, 58(6), 1355-65. Elsevier. doi:10.1016/S0006-3495(90)82482-6

MacLennan, D. H., & Kranias, E. G. (2003). Phospholamban: a crucial regulator of cardiac contractility. Nature reviews. Molecular cell biology, 4(7), 566-77. doi:10.1038/nrm1151

MacLennan, D. H., Asahi, M., & Tupling, A. R. (2003). The regulation of SERCA-type pumps by phospholamban and sarcolipin. Annals of the New York Academy of Sciences, 986, 472-80. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/12763867

MacLennan, D. H., Brandl, C. J., Korczak, B., & Green, N. M. (1985). Amino-acid sequence of a Ca2+ + Mg2+-dependent ATPase from rabbit muscle sarcoplasmic reticulum, deduced from its complementary DNA sequence. Nature, 316(6030), 696-700. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/2993904

MacLennan, D. H., Kimura, Y., & Toyofuku, T. (1998). Sites of regulatory interaction between calcium ATPases and phospholamban. Annals of the New York Academy of Sciences, 853, 31-42. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/10603934

MacLennan, D. H., Toyofuku, T., & Lytton, J. (1992). Structure-function relationships in sarcoplasmic or endoplasmic reticulum type Ca2+ pumps. Annals of the New York Academy of Sciences, 671, 1-10. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/1337665

Manni, S., Mauban, J. H., Ward, C. W., & Bond, M. (2008). Phosphorylation of the cAMP- dependent protein kinase (PKA) regulatory subunit modulates PKA-AKAP interaction, substrate phosphorylation, and calcium signaling in cardiac cells. The Journal of biological chemistry, 283(35), 24145-54. doi:10.1074/jbc.M802278200

121

Martonosi, A. (1968). Sarcoplasmic reticulum. V. The structure of sarcoplasmic reticulum membranes. Biochimica et biophysica acta, 150(4), 694-704. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/18683494

Maruyama, K., Clarke, D. M., Fujii, J., Inesi, G., Loo, T. W., & MacLennan, D. H. (1989). Functional consequences of alterations to amino acids located in the catalytic center (isoleucine 348 to threonine 357) and nucleotide-binding domain of the Ca2+-ATPase of sarcoplasmic reticulum. The Journal of biological chemistry, 264(22), 13038-42. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/2526811

Mattiazzi, a, Hove-Madsen, L., & Bers, D. M. (1994). Protein kinase inhibitors reduce SR Ca transport in permeabilized cardiac myocytes. The American journal of physiology, 267(2 Pt 2), H812-20. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/8067437

Metcalfe, E. E., Traaseth, N. J., & Veglia, G. (2005). Serine 16 phosphorylation induces an order-to-disorder transition in monomeric phospholamban. Biochemistry, 44(11), 4386-96. doi:10.1021/bi047571e

Michelangeli, F., & East, J. M. (2011). A diversity of SERCA Ca2+ pump inhibitors. Biochemical Society transactions, 39(3), 789-97. doi:10.1042/BST0390789

Morris, G. L., Cheng, H. C., Colyer, J., & Wang, J. H. (1991). Phospholamban regulation of cardiac sarcoplasmic reticulum (Ca(2+)-Mg2+)-ATPase. Mechanism of regulation and site of monoclonal antibody interaction. The Journal of biological chemistry, 266(17), 11270-5. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/1710223

Morth, J. P., Pedersen, B. P., Toustrup-Jensen, M. S., Sørensen, T. L.-M., Petersen, J., Andersen, J. P., Vilsen, B., et al. (2007). Crystal structure of the sodium-potassium pump. Nature, 450(7172), 1043-9. doi:10.1038/nature06419

Movsesian, M A. (1992). Calcium uptake by sarcoplasmic reticulum and its modulation by cAMP-dependent phosphorylation in normal and failing human myocardium. Basic research in cardiology, 87 Suppl 1, 277-84. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/1323264

Movsesian, Matthew A, Colyer, J., Wang, J. H., & Krall, J. (1990). Phospholamban- mediated stimulation of Ca2+ uptake in sarcoplasmic reticulum from normal and failing hearts. The Journal of clinical investigation, 85(5), 1698-702. doi:10.1172/JCI114623

122

Mueller, B., Karim, C. B., Negrashov, I. V., Kutchai, H., & Thomas, D. D. (2004). Direct detection of phospholamban and sarcoplasmic reticulum Ca-ATPase interaction in membranes using fluorescence resonance energy transfer. Biochemistry, 43(27), 8754-65. American Chemical Society. doi:10.1021/bi049732k

Mundiña-Weilenmann, C., Vittone, L., Ortale, M., de Cingolani, G. C., & Mattiazzi, A. (1996). Immunodetection of phosphorylation sites gives new insights into the mechanisms underlying phospholamban phosphorylation in the intact heart. The Journal of biological chemistry, 271(52), 33561-7. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/8969222

Nicolaou, P., Hajjar, R. J., & Kranias, E. G. (2009). Role of protein phosphatase-1 inhibitor-1 in cardiac physiology and pathophysiology. Journal of molecular and cellular cardiology, 47(3), 365-71. Elsevier Inc. doi:10.1016/j.yjmcc.2009.05.010

Odermatt, a, Becker, S., Khanna, V. K., Kurzydlowski, K., Leisner, E., Pette, D., & MacLennan, D. H. (1998). Sarcolipin regulates the activity of SERCA1, the fast- twitch skeletal muscle sarcoplasmic reticulum Ca2+-ATPase. The Journal of biological chemistry, 273(20), 12360-9. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/9575189

Odermatt, a, Taschner, P. E., Scherer, S. W., Beatty, B., Khanna, V. K., Cornblath, D. R., Chaudhry, V., et al. (1997). Characterization of the gene encoding human sarcolipin (SLN), a proteolipid associated with SERCA1: absence of structural mutations in five patients with Brody disease. Genomics, 45(3), 541-53. doi:10.1006/geno.1997.4967

Odermatt, A., Kurzydlowski, K., & MacLennan, D. H. (1996). The vmax of the Ca2+- ATPase of cardiac sarcoplasmic reticulum (SERCA2a) is not altered by Ca2+/calmodulin-dependent phosphorylation or by interaction with phospholamban. The Journal of biological chemistry, 271(24), 14206-13. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/8662932

Paterlini, M. G., & Thomas, D. D. (2005). The alpha-helical propensity of the cytoplasmic domain of phospholamban: a molecular dynamics simulation of the effect of phosphorylation and mutation. Biophysical journal, 88(5), 3243-51. doi:10.1529/biophysj.104.054460

Periasamy, M., & Kalyanasundaram, A. (2007). SERCA pump isoforms: their role in calcium transport and disease. Muscle & nerve, 35(4), 430-42. doi:10.1002/mus.20745

123

Periasamy, M., Bhupathy, P., & Babu, G. J. (2008). Regulation of sarcoplasmic reticulum Ca2+ ATPase pump expression and its relevance to cardiac muscle physiology and pathology. Cardiovascular research, 77(2), 265-73. doi:10.1093/cvr/cvm056

Phillips, R. M., Narayan, P., Gómez, a M., Dilly, K., Jones, L. R., Lederer, W. J., & Altschuld, R. a. (1998). Sarcoplasmic reticulum in heart failure: central player or bystander? Cardiovascular research, 37(2), 346-51. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/9614491

Pogwizd, S. M. (2000). The online version of this article, along with updated information and services, is located on the World Wide Web at: http://circres.ahajournals.org/content/87/8/641. Circulation Research, 641-643. doi:10.1161/01.RES.87.8.641

Pritchard, T. J., & Kranias, E. G. (2009). Junctin and the histidine-rich Ca2+ binding protein: potential roles in heart failure and arrhythmogenesis. The Journal of physiology, 587(Pt 13), 3125-33. doi:10.1113/jphysiol.2009.172171

Raeymaekers, L., Hofmann, F., & Casteels, R. (1988). Cyclic GMP-dependent protein kinase phosphorylates phospholamban in isolated sarcoplasmic reticulum from cardiac and smooth muscle. The Biochemical journal, 252(1), 269-73. Retrieved from http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1149133&tool=pmcen trez&rendertype=abstract

Ralevic, V., & Burnstock, G. (1998). Receptors for purines and pyrimidines. Pharmacological reviews, 50(3), 413-92. Retrieved from http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3083132&tool=pmcen trez&rendertype=abstract

Reddy, L. G., Jones, L. R., Cala, S. E., O’Brian, J. J., Tatulian, S. A., & Stokes, D. L. (1995). Functional reconstitution of recombinant phospholamban with rabbit skeletal Ca(2+)-ATPase. The Journal of biological chemistry, 270(16), 9390-7. Retrieved from http://www.jbc.org/content/270/16/9390.short

Reddy, L. G., Jones, L. R., Pace, R. C., & Stokes, D. L. (1996). Purified, reconstituted cardiac Ca2+-ATPase is regulated by phospholamban but not by direct phosphorylation with Ca2+/calmodulin-dependent protein kinase. The Journal of biological chemistry, 271(25), 14964-70. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/8663079

124

Robia, S. L., Campbell, K. S., Kelly, E. M., Hou, Z., Winters, D. L., & Thomas, D. D. (2007). Förster transfer recovery reveals that phospholamban exchanges slowly from pentamers but rapidly from the SERCA regulatory complex. Circulation research, 101(11), 1123-9. doi:10.1161/CIRCRESAHA.107.159947

Sagara, Y., & Inesi, G. (1991). Inhibition of the sarcoplasmic reticulum Ca2+ transport ATPase by thapsigargin at subnanomolar concentrations. The Journal of biological chemistry, 266(21), 13503-6. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/1830305

Sasaki, T., Inui, M., Kimura, Y., Kuzuya, T., & Tada, M. (1992). Molecular mechanism of regulation of Ca2+ pump ATPase by phospholamban in cardiac sarcoplasmic reticulum. Effects of synthetic phospholamban peptides on Ca2+ pump ATPase. The Journal of biological chemistry, 267(3), 1674-9. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/1530941

Schmidt, A. G., Haghighi, K., Frank, B., Pater, L., Dorn, G. W., Walsh, R. A., & Kranias, E. G. (2003). Polymorphic SERCA2a variants do not account for inter-individual differences in phospholamban-SERCA2a interactions in human heart failure. Journal of molecular and cellular cardiology, 35(7), 867-70. doi:10.1016/S0022- 2828(03)00111-1

Schoenmakers, T. J., Visser, G. J., Flik, G., & Theuvenet, a P. (1992). CHELATOR: an improved method for computing metal ion concentrations in physiological solutions. BioTechniques, 12(6), 870-4, 876-9. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/1642895

Sham, J. S., Jones, L. R., & Morad, M. (1991). Phospholamban mediates the beta- adrenergic-enhanced Ca2+ uptake in mammalian ventricular myocytes. The American journal of physiology, 261(4 Pt 2), H1344-9. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/1681743

Shinoda, T., Ogawa, H., Cornelius, F., & Toyoshima, C. (2009). Crystal structure of the sodium-potassium pump at 2.4 A resolution. Nature, 459(7245), 446-50. Nature Publishing Group. doi:10.1038/nature07939

Simmerman, H. K., & Jones, L. R. (1998). Phospholamban: protein structure, mechanism of action, and role in cardiac function. Physiological reviews, 78(4), 921-47. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/9790566

Simmerman, H. K., Collins, J. H., Theibert, J. L., Wegener, A. D., & Jones, L. R. (1986). Sequence analysis of phospholamban. Identification of phosphorylation sites and

125

two major structural domains. The Journal of biological chemistry, 261(28), 13333- 41. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/3759968

Simmerman, H. K., Kobayashi, Y. M., Autry, J. M., & Jones, L. R. (1996). A leucine zipper stabilizes the pentameric membrane domain of phospholamban and forms a coiled-coil pore structure. The Journal of biological chemistry, 271(10), 5941-6. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/8621468

Simmerman, H. K., Lovelace, D. E., & Jones, L. R. (1989). Secondary structure of detergent-solubilized phospholamban, a phosphorylatable, oligomeric protein of cardiac sarcoplasmic reticulum. Biochimica et biophysica acta, 997(3), 322-9. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/2527565

Song, Q., Pallikkuth, S., Bossuyt, J., Bers, D. M., & Robia, S. L. (2011). Phosphomimetic Mutations Enhance Oligomerization of Phospholemman and Modulate Its Interaction with the Na/K-ATPase. The Journal of biological chemistry, 286(11), 9120-6. doi:10.1074/jbc.M110.198036

Starling, A. P., Hughes, G., Sharma, R. P., East, J. M., & Lee, A. G. (1995). The hydrophilic domain of phospholamban inhibits the Ca(2+)-ATPase--the importance of the method of assay. Biochemical and biophysical research communications, 215(3), 1067-70. doi:10.1006/bbrc.1995.2572

Stokes, D. L., Pomfret, A. J., Rice, W. J., Glaves, J. P., & Young, H. S. (2006). Interactions between Ca2+-ATPase and the pentameric form of phospholamban in two- dimensional co-crystals. Biophysical journal, 90(11), 4213-23. doi:10.1529/biophysj.105.079640

Sørensen, T. L.-M., Møller, J. V., & Nissen, P. (2004). Phosphoryl transfer and calcium ion occlusion in the calcium pump. Science (New York, N.Y.), 304(5677), 1672-5. doi:10.1126/science.1099366

Tada, M., & Toyofuku, T. (1998). Molecular regulation of phospholamban function and expression. Trends in cardiovascular medicine, 8(8), 330-40. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/14987547

Tada, M., Kirchberger, M. A., & Katz, A. M. (1975). Phosphorylation of a 22,000-dalton component of the cardiac sarcoplasmic reticulum by adenosine 3’:5'- monophosphate-dependent protein kinase. The Journal of biological chemistry, 250(7), 2640-7. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/235523

126

Tada, M., Kirchberger, M. A., Repke, D. I., & Katz, a M. (1974). The stimulation of calcium transport in cardiac sarcoplasmic reticulum by adenosine 3’:5'-monophosphate- dependent protein kinase. The Journal of biological chemistry, 249(19), 6174-80. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/199585

Tanford, C. (1982). Simple model for the chemical potential change of a transported ion in active transport. Proceedings of the National Academy of Sciences of the United States of America, 79(9), 2882-4. Retrieved from http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=346311&tool=pmcent rez&rendertype=abstract

Terzi, E., Poteur, L., & Trifilieff, E. (1992). Evidence for a phosphorylation-induced conformational change in phospholamban cytoplasmic domain by CD analysis. FEBS letters, 309(3), 413-6. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/1387620

Toyofuku, T., Kurzydlowski, K., Tada, M., & MacLennan, D. H. (1993). Identification of regions in the Ca(2+)-ATPase of sarcoplasmic reticulum that affect functional association with phospholamban. The Journal of biological chemistry, 268(4), 2809- 15. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/8428955

Toyofuku, T., Kurzydlowski, K., Tada, M., & MacLennan, D. H. (1994). Amino acids Lys- Asp-Asp-Lys-Pro-Val402 in the Ca(2+)-ATPase of cardiac sarcoplasmic reticulum are critical for functional association with phospholamban. The Journal of biological chemistry, 269(37), 22929-32. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/8083189

Toyoshima, C., & Inesi, G. (2004). Structural basis of ion pumping by Ca2+-ATPase of the sarcoplasmic reticulum. Annual review of biochemistry, 73, 269-92. doi:10.1146/annurev.biochem.73.011303.073700

Toyoshima, C., & Nomura, H. (2002). Structural changes in the calcium pump accompanying the dissociation of calcium. Nature, 418(6898), 605-11. doi:10.1038/nature00944

Toyoshima, C., Asahi, M., Sugita, Y., Khanna, R., Tsuda, T., & MacLennan, D. H. (2003). Modeling of the inhibitory interaction of phospholamban with the Ca2+ ATPase. Proceedings of the National Academy of Sciences of the United States of America, 100(2), 467-72. doi:10.1073/pnas.0237326100

127

Toyoshima, C., Nakasako, M., Nomura, H., & Ogawa, H. (2000). Crystal structure of the calcium pump of sarcoplasmic reticulum at 2.6 A resolution. Nature, 405(6787), 647-55. doi:10.1038/35015017

Toyoshima, C., Sasabe, H., & Stokes, D. L. (1993). Three-dimensional cryo-electron microscopy of the calcium ion pump in the sarcoplasmic reticulum membrane. Nature, 362(6419), 467-71. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/8385269

Traaseth, N. J., Shi, L., Verardi, R., Mullen, D. G., Barany, G., & Veglia, G. (2009). Structure and topology of monomeric phospholamban in lipid membranes determined by a hybrid solution and solid-state NMR approach. Proceedings of the National Academy of Sciences of the United States of America, 106(25), 10165-70. doi:10.1073/pnas.0904290106

Traaseth, N. J., Thomas, D. D., & Veglia, G. (2006). Effects of Ser16 phosphorylation on the allosteric transitions of phospholamban/Ca(2+)-ATPase complex. Journal of molecular biology, 358(4), 1041-50. doi:10.1016/j.jmb.2006.02.047

Tsuji, T., Del Monte, F., Yoshikawa, Y., Abe, T., Shimizu, J., Nakajima-Takenaka, C., Taniguchi, S., et al. (2009). Rescue of Ca2+ overload-induced left ventricular dysfunction by targeted ablation of phospholamban. American journal of physiology. Heart and circulatory physiology, 296(2), H310-7. doi:10.1152/ajpheart.00975.2008

Vafiadaki, E., Arvanitis, D. A., Pagakis, S. N., Papalouka, V., Sanoudou, D., Kontrogianni- Konstantopoulos, A., & Kranias, E. G. (2009). The anti-apoptotic protein HAX-1 interacts with SERCA2 and regulates its protein levels to promote cell survival. Molecular biology of the cell, 20(1), 306-18. Am Soc Cell Biol. doi:10.1091/mbc.E08- 06-0587

Vafiadaki, E., Sanoudou, D., Arvanitis, D. a, Catino, D. H., Kranias, E. G., & Kontrogianni- Konstantopoulos, A. (2007). Phospholamban interacts with HAX-1, a mitochondrial protein with anti-apoptotic function. Journal of molecular biology, 367(1), 65-79. doi:10.1016/j.jmb.2006.10.057

Verardi, R., Shi, L., Traaseth, N. J., Walsh, N., & Veglia, G. (2011). Structural topology of phospholamban pentamer in lipid bilayers by a hybrid solution and solid-state NMR method. Proceedings of the National Academy of Sciences of the United States of America, 108(22), 9101-6. doi:10.1073/pnas.1016535108

128

Vilsen, B., Andersen, J. P., Clarke, D. M., & MacLennan, D. H. (1989). Functional consequences of proline mutations in the cytoplasmic and transmembrane sectors of the Ca2(+)-ATPase of sarcoplasmic reticulum. The Journal of biological chemistry, 264(35), 21024-30. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/2531743

Waggoner, J. R., Ginsburg, K. S., Mitton, B., Haghighi, K., Robbins, J., Bers, D. M., & Kranias, E. G. (2009). Phospholamban overexpression in rabbit ventricular myocytes does not alter sarcoplasmic reticulum Ca transport. American journal of physiology. Heart and circulatory physiology, 296(3), H698-703. doi:10.1152/ajpheart.00272.2008

Wegener, A. D., & Jones, L. R. (1984). Phosphorylation-induced Mobility Shift in Phospholamban in Sodium Dodecyl Sulfate-Polyacrylamide Gels. Society, 259(3), 1834-1841.

Wegener, A. D., Simmerman, H. K., Liepnieks, J., & Jones, L. R. (1986). Proteolytic cleavage of phospholamban purified from canine cardiac sarcoplasmic reticulum vesicles. Generation of a low resolution model of phospholamban structure. The Journal of biological chemistry, 261(11), 5154-9. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/3007493

Wickenden, a D., Kaprielian, R., Kassiri, Z., Tsoporis, J. N., Tsushima, R., Fishman, G. I., & Backx, P. H. (1998). The role of action potential prolongation and altered intracellular calcium handling in the pathogenesis of heart failure. Cardiovascular research, 37(2), 312-23. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/9614488

Will, H., Levchenko, T. S., Levitsky, D. O., Smirnov, V. N., & Wollenberger, A. (1978). Partial characterization of protein kinase-catalyzed phosphorylation of low molecular weight proteins in purified preparations of pigeon heart sarcolemma and sarcoplasmic reticulum. Biochimica et biophysica acta, 543(2), 175-93. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/365242

Winters, D. L., Autry, J. M., Svensson, B., & Thomas, D. D. (2008). Interdomain fluorescence resonance energy transfer in SERCA probed by cyan-fluorescent protein fused to the actuator domain. Biochemistry, 47(14), 4246-56. doi:10.1021/bi702089j

Wuytack, F., Raeymaekers, L., & Missiaen, L. (2002). Molecular physiology of the SERCA and SPCA pumps. Cell calcium, 32(5-6), 279-305. doi:10.1016/S0143- 4160(02)00184-7

129

Young, H. S., Jones, L. R., & Stokes, D. L. (2001). Locating phospholamban in co-crystals with Ca(2+)-ATPase by cryoelectron microscopy. Biophysical journal, 81(2), 884-94. doi:10.1016/S0006-3495(01)75748-7

Zal, T., & Gascoigne, N. R. J. (2004). Photobleaching-corrected FRET efficiency imaging of live cells. Biophysical journal, 86(6), 3923-39. Elsevier. doi:10.1529/biophysj.103.022087

Zamoon, J., Nitu, F., Karim, C. B., Thomas, D. D., & Veglia, G. (2005). Mapping the interaction surface of a membrane protein: unveiling the conformational switch of phospholamban in calcium pump regulation. Proceedings of the National Academy of Sciences of the United States of America, 102(13), 4747-52. doi:10.1073/pnas.0406039102

Zhang, P., Toyoshima, C., Yonekura, K., Green, N. M., & Stokes, D. L. (1998). Structure of the calcium pump from sarcoplasmic reticulum at 8-A resolution. Nature, 392(6678), 835-9. doi:10.1038/33959

Zhao, W., Waggoner, J. R., Zhang, Z.-G., Lam, C. K., Han, P., Qian, J., Schroder, P. M., et al. (2009). The anti-apoptotic protein HAX-1 is a regulator of cardiac function. Proceedings of the National Academy of Sciences of the United States of America, 106(49), 20776-81. doi:10.1073/pnas.0906998106 de Meis, L., & Vianna, A. L. (1979). Energy interconversion by the Ca2+-dependent ATPase of the sarcoplasmic reticulum. Annual review of biochemistry, 48, 275-92. doi:10.1146/annurev.bi.48.070179.001423

VITA

The author, Philip Bidwell was born in Columbus, IN on November 19, 1982 to

Ronald and Elzbieta Bidwell. Philip is the youngest of five children: Dara, Kathy, Paul, and Stephen. Philip married Christine Joy Archer on August 29, 2009. Philip received a

B.S. in both Biology and Biochemistry from Indiana University in May of 2005. Following graduation, Philip worked as a Laboratory Technician in the lab of Maria Teresa Rizzo,

MD at the Methodist Research Institute in Indianapolis, IN.

In August of 2007, Philip joined the Department of Cell and Molecular Physiology at the Loyola University Medical Center, Maywood, IL. In the spring of 2008, Philip joined the lab of Seth L. Robia, Ph.D. and began the study of the regulatory interaction between phospholamban (PLB) and sarco/endoplasmic reticulum calcium ATPase

(SERCA). In the fall of 2011, Philip was awarded a predoctoral fellowship from the

American Heart Association (Midwest Affiliate).

Philip has accepted a postdoctoral position in the lab of Evangelia Kranias, Ph.D. at the University of Cincinnati to study the regulation of cardiac calcium homeostasis.

130