Molecular and Structural Requirements of the Β1l

MOLECULAR AND STRUCTURAL REQUIREMENTS OF THE

1L-ADRENOCEPTOR

Paul Klenowski BBiotech (Hons)

Submitted in [partial] fulfilment of the requirements for the degree of Doctor of Philosophy

Institute of Health and Biomedical Innovation Faculty of Health Queensland University of Technology [September 2012]

Keywords

Affinity, agonist, antagonist, 1L-adrenoceptor, 1H-adrenoceptor, 2-adrenoceptor,

3-adrenoceptor, -blocker, (-)-bupranolol, cardiostimulation, cardiovascular, catecholamines, (-)-CGP 12177, (-)-[3H]- CGP 12177, chimera, contractility, crystal structure, cyclic AMP, G-protein coupled receptor, heterologous, human atrial force, human heart failure, (-)-isoprenaline, L-748,337, ligand, low-affinity binding site, L- type Ca2+ channel, molecular modelling, non-conventional partial agonists, noradrenaline, pindolol, positive inotropic effect, potency, radioligand binding, recombinant, site-directed mutagenesis, transmembrane domain

Abstract

There are two binding sites on the β1-adrenoceptor (AR), β1H and β1L corresponding to high and low-affinity binding sites respectively, which can be activated to cause cardiostimulation. Some β-blockers typified by (-)-CGP 12177 and

(-)-pindolol block β1AR and β2ARs, but also activate β1LARs at higher concentrations than those required to cause blockade. However, in a report by Skeberdis et al., (2008) it was proposed that the positive inotropic effects of (-)-CGP

12177 are mediated through β3ARs. Consequently, experiments were performed using human atrial trabeculae, to investigate whether the β3AR selective antagonist

L-748,337 could antagonise the positive inotropic effects of the β3AR agonists SR 58611, BRL 37344 and (-)-CGP 12177. These experiments did not detect inotropic effects of SR 58611 (1 nM-10 M). The positive inotropic effects of BRL 37344 were antagonised by the 1AR and 2AR antagonist nadolol (200 nM) and 2AR antagonist ICI 118,551 (50 nM). Concurrent ICI 118,551 and CGP 20712A (β1AR selective antagonist) addition caused a greater shift of the curve of BRL 37344 than

ICI 118,551 alone. Furthermore, the 3AR-selective antagonist L-748,337 (1 M) did not affect the responses to BRL 37344. These results indicate that the effects of

BRL 37344 are mediated through 1AR and 2AR but not 3AR.

(-)-CGP 12177 (200 nM) caused stable increases in contractile force which were significantly reduced by the addition of (-)-bupranolol (1 M) but not affected by the addition of L-748,337 (1 M, P = 0.001, 1-way ANOVA). The results of trabeculae obtained from 4 patients revealed that L-748,337 did not affect the response to (-)-CGP 12177 (P = 0.12), inconsistent with mediation through 3AR. In contrast, (-)-bupranolol reduced the response by 91 ± 16%, n = 4, P = 0.002, consistent with mediation through 1LAR, as observed before on human atrium

(Kaumann et al., 1996a) and recombinant 1AR (Joseph et al., 2004b).

Studies have demonstrated that β2AR does not form a corresponding low- affinity binding site (Baker et al., 2002), therefore it was hypothesised that

heterologous β1- β2AR amino acids are responsible for the formation of β1LAR. This study investigated whether heterologous amino acids of the fifth transmembrane domain (TMDV) of β1AR and β2ARs contribute to β1LAR. β1ARs, β2ARs and mutant

β1ARs containing all (β1/β2TMDVAR) or single amino acids of TMDV of the β2AR were prepared and stably expressed in Chinese Hamster Ovary cells. Concentration- effect curves for cyclic AMP accumulation were carried out for (-)-CGP 12177 or (-)-isoprenaline in the absence and presence of (-)-bupranolol.

The potencies (pEC50) of (-)-CGP 12177 and affinities (pKB) of (-)-bupranolol versus (-)-CGP 12177 were:

Cell Line pEC50 (-)-CGP 12177 n pKB (-)-bupranolol n

β2AR 9.10 ± 0.18 14 9.31 ± 0.16 7

β1(V230I)AR 9.08 ± 0.07 10 7.64 ± 0.12 8

β1/β2TMDVAR 8.90 ± 0.10 15 8.06 ± 0.17 8

β1(R222Q)AR 8.10 ± 0.10 6 7.33 ± 0.23 5

β1AR 7.96 ± 0.09 14 7.20 ± 0.16 8

The potency of (-)-CGP 12177 was higher at β2AR than at β1AR, consistent with activation through a low-affinity site at the β1AR (β1LAR). The presence of valine at position 230 in β1AR accounted for the lower potency of (-)-CGP 12177.

The affinity of (-)-bupranolol was lower at β1AR compared to β2AR. The presence of valine 230 in β1AR accounted in part for the lower affinity. In conclusion, TMDV and valine 230 of the β1AR contribute in part to the low-affinity binding site of

β1AR.

iii

Table of Contents

Keywords ...... i Abstract ...... ii Table of Contents ...... iv List of Figures ...... vi List of Tables ...... ix List of Abbreviations ...... x Statement of Original Authorship ...... xiii Publications and Conference Abstracts ...... xiv Acknowledgements ...... xv CHAPTER 1: INTRODUCTION ...... 1 CHAPTER 2: LITERATURE REVIEW ...... 5 2.1 G-Protein Coupled Receptors ...... 5 2.2 -Adrenoceptors ...... 6 2.2.1 1AR Gene and Protein Structure ...... 7

2.3 Structural Components of the 1AR ...... 9 2.3.1 N-Terminus and Extracellular Loops ...... 9 2.3.2 Transmembrane Domain Regions and 1AR Ligand Binding Pocket ...... 10 2.3.3 Intracellular Loops and G-Protein Binding ...... 13 2.4 GPCR Signalling and Activation ...... 15 2.4.1 1AR Activation ...... 16 2.4.2 1AR Signalling ...... 21 2.5 AR Signalling in the Human Heart ...... 23 2.5.1 The Role of1AR in Heart Failure ...... 27

2.6 1LAR Pharmacology ...... 31 2.6.1 Mechanistic Studies Into the Pharmacology of 1LAR ...... 32 2.6.2 The Putative β4-Adrenoceptor: A Working Hypothesis ...... 33 2.6.3 Proposal of a Low-Affinity Form of the β1-Adrenoceptor ...... 35 2.6.4 Pharmacological Differences Between the Putative β4-Adrenoceptor and Low- Affinity β1-Adrenoceptor ...... 37 2.6.5 The Underlying Role of β1-Adrenoceptors in Putative β4-Adrenoceptor Pharmacology ...... 38 2.6.6 Structural Implications of High and Low-Affinity Binding of Non-Conventional Partial Agonists to β1-Adrenoceptors ...... 39 2.6.7 Functional and Clinical Implication of the β1L-Adrenoceptor ...... 41 2.7 Summary and Implications ...... 43 CHAPTER 3: MATERIALS AND METHODS ...... 45 3.1 Research Design ...... 45 3.2 materials and Methods ...... 47 3.2.1 General Reagents ...... 47 3.2.2 General Methods ...... 50 3.2.3 Radioligand Binding Experiments ...... 58 3.2.4 Cyclic AMP Enzyme Immunoassays ...... 62 3.2.5 Tissue Bath Experiments ...... 67

CHAPTER 4: HUMAN ATRIAL 1L-ADRENOCEPTOR BUT NOT 3-ADRENOCEPTOR ACTIVATION INCREASES FORCE AND Ca2+ CURRENT AT PHYSIOLOGICAL TEMPERATURE ...... 71 4.1 Introduction ...... 71 4.2 Results ...... 74 4.2.1 Patients ...... 74 4.2.2 Antagonism of the Inotropic Effects of BRL 37344 by the 1AR/2AR Antagonist Nadolol, 2AR Antagonist ICI 118,551 but not 3AR Antagonist L-748,337 in Atrial Trabeculae...... 75 4.2.3 The 3AR Agonist SR 58611 does not Increase Contractile Force...... 78 4.2.4 Stable (-)-CGP 12177-Evoked Contractions are Decreased by (-)-Bupranolol but not by L-748,337...... 79 4.2.5 L-748,337 does not Modify the Contractile Potency of (-)-CGP 12177...... 81 4.2.6 ICa-L Responses to SR 58611, BRL 37344, and (-)-CGP 12177 at 24°C...... 82 4.2.7 BRL 37344 and SR 58611 do not Increase ICa-L at 37°C...... 84 4.2.8 Antagonism of (-)-CGP 12177-Evoked ICa-L Increases by (-)-Bupranolol but not by L-748,337 at 37C ...... 84 4.2.9 SR 58611 does not Increase Atrial Force at 24C...... 86 4.2.10 (-)-CGP 12177-Evoked Increases in Atrial Force at 24°C are Antagonised by (-)-Bupranolol but not by L-748,337...... 86 4.3 Discussion ...... 88 CHAPTER 5: CONTRIBUTION OF TRANSMEMBRANE DOMAIN V OF THE 1-ADRENOCEPTOR TO 1L-ADRENOCEPTOR PHARMACOLOGY ...... 95 5.1 Introduction ...... 95 5.2 Results ...... 97 5.2.1 Characterisation and Validation of Wild-type and Mutant βARs Expressed in CHOAA8 Cells...... 97 5.2.2 Cyclic AMP Responses to (-)-CGP 12177 and (-)-Isoprenaline at 1AR, 2AR and 1/2TMDVAR in CHOAA8 cells...... 109 5.2.3 Cyclic AMP Responses to (-)-CGP 12177 and (-)-Isoprenaline at Heterologous Point Mutations of 1-2AR TMDV...... 119 5.2.4 Cyclic AMP Responses to (-)-CGP 12177 and (-)-Isoprenaline at Homologous Serine Residues of 1AR, 2AR TMDV...... 125 5.3 Discussion ...... 129 CHAPTER 6: PHARMACOLOGICAL ANALYSIS OF 5-[3-(TERT-BUTYLAMINO))2- HYDROXYPROPOXY]1,3-DIHYDRO-2H-BENZIMIDAZOL-2-ONE ...... 137 6.1 Introduction ...... 137 6.2 Results ...... 139 6.2.1 Competition Binding ...... 139 6.2.2 Cyclic AMP Experiments ...... 139 6.3 Discussion ...... 142 CHAPTER 7: GENERAL DISCUSSION AND FUTURE DIRECTIONS ...... 145 BIBLIOGRAPHY ...... 151 APPENDICES ...... 179 Appendix A Table of Chemical Structures ...... 179 Appendix B British Journal of Pharmacology Publication ...... 181 Appendix C Molecular Modelling ...... 198

List of Figures

Figure 2.1. Two-dimensional diagram of the human 1AR adapted from the turkey 1AR-m23 model...... 8

Figure 2.2. The ligand binding pocket of β1AR-m23, displaying the amino acids interacting with cyanopindolol...... 11

Figure 2.3. The ligand binding pockets of 2AR with carazolol, 1AR with cyanopindolol and 1AR with isoprenaline...... 17

Figure 2.4. Comparison of the 1AR-m23 structures with either cyanopindolol or isoprenaline...... 18

Figure 2.5. Comparison of the inactive carazolol bound β2AR with the Nb80 stabilised β2AR and the active β2AR-Gs structure...... 20 Figure 2.6. Schematic representation of main βAR signal transduction pathways in cardiac sympathetic nervous transmission...... 24 Figure 2.7. The effects of the βAR-Gs-protein-adenylyl cyclase-cAMP-protein kinase A pathway in human heart...... 26 Figure 2.8. The correlation between the potencies (pD2) and affinities (pKa) of partial agonists on isolated heart preparations...... 32 Figure 2.9. The biphasic activity of CGP 12177...... 36 Figure 2.10. Polygraph traces comparing the effects (-)-CGP 12177 on left atrium from β2AR knockout and β1AR/β2AR double knockout mice...... 39 Figure 2.11. Chemical structures of (-)-CGP 12177 and (-)-pindolol...... 43 Figure 3.1. Structural considerations of -blockers with different agonist activity at

1LAR...... 46

Figure 3.2. The 1/2TMDVAR chimera...... 53 Figure 3.3. The cAMP EIA assay plate...... 64 Figure 3.4. The cAMP EIA reaction...... 65

Figure 4.1. Mediation of the positive inotropic effects of BRL 37344 through 1-and 2ARs but not 3ARs in human atrial trabeculae...... 75

Figure 4.2. Mediation of the positive inotropic effects of BRL 37344 through 1-and 2ARs in human atrial trabeculae...... 76 Figure 4.3. Lack of inotropic effects of BRL 37344 at 24°C in human atrial trabeculae...... 76 Figure 4.4. Lack of positive inotropic effects of SR 58611 at 37 ºC in human atrial trabeculae...... 78 Figure 4.5. Lack of positive inotropic effects of SR 58611 at 24ºC in human atrial trabeculae...... 79 Figure 4.6. Mediation of the positive inotropic effects of (-)-CGP 12177 through β1LAR but not β3AR...... 80 Figure 4.7. A comparison of the contractile force obtained in the presence of 200 nM (-)-CGP 12177 and after incubation of L-748,337, (-)-bupranolol or control...... 81

Figure 4.8. L-748,337 does not change the inotropic potency of (-)-CGP 12177 in the presence of nadolol (200 nM) and IBMX (10 M)...... 82

Figure 4.9. L-748,337(1 µM) prevents small ICa-L responses to SR 58611 (1 µM) and BRL 37344 (1 µM) but not to (-)-CGP 12177 (1 µM) at 24ºC in the presence of nadolol (200 nM)...... 83 Figure 4.10. SR 58611 (1 µM) and BRL 37344 (1 µM), in the presence of nadolol (200 nM) failed to increase ICa-L in the absence and presence of IBMX (10 µM) at 37ºC...... 84

Figure 4.11. (-)-CGP 12177 (1 µM) causes small increases of ICa-L at 37ºC, compared to (-)-isoprenaline (1 µM), in the presence but not absence of IBMX (10 µM) that are prevented by (-)-bupranolol (10 µM) but not by L-748,337 (1µM) ICI 118,551 (50 nM) or nadolol (200 nM)...... 85 Figure 4.12. Inotropic effects of (-)-CGP 12177 at 24ºC...... 87 Figure 4.13. (-)-CGP 12177 causes positive inotropic effects in the presence of both (-)-propranolol (200 nM) and IBMX (10 µM) through β1LAR but not β3AR...... 87 Figure 5.1. Saturation binding studies performed using (-)-[3H]-CGP 12177 on purified membranes from CHOAA8 cells expressing wild-type β1AR and 1/2TMDVAR...... 99 Figure 5.2. Saturation binding performed using (-)-[3H]-CGP 12177 on purified membranes from CHOAA8 cells expressing β1(S228A)AR and β1(S229A)AR ...... 100 Figure 5.3. Competition binding between CGP 20712A or ICI 118,551 graph A, 3 (-)-bupranolol or bisoprolol graph B and (-)-[ H]-CGP 12177 at 1AR...... 103 Figure 5.4. Competition binding between CGP 20712A or ICI 118,551 graph A, 3 (-)-bupranolol or bisoprolol graph B and (-)-[ H]-CGP 12177 at 2AR...... 104 Figure 5.5. Competition binding between CGP 20712A or ICI 118,551 graph A, (-)-bupranolol or bisoprolol graph B and (-)-[3H]-CGP 12177 at

1/2TMDVAR ...... 105 Figure 5.6. Competition binding between CGP 20712A or ICI 118,551 at 3 1(V230I)AR graph A, 1(V230A)AR graph B and (-)-[ H]-CGP 12177...... 106 Figure 5.7. Competition binding between CGP 20712A or ICI 118,551 at 3 1(R222Q)AR graph A, 1(S228A)AR graph B and (-)-[ H]-CGP 12177...... 107 Figure 5.8. Competition binding between CGP 20712A or ICI 118,551 at 3 1(S229A)AR graph A, 1(S232A)AR graph B and (-)-[ H]-CGP 12177...... 108 Figure 5.9. Concentration-effect curves for cAMP accumulation in CHOAA8 cells

stably expressing 1AR in response to (-)-CGP 12177 and (-)-isoprenaline...... 111 Figure 5.10. Concentration-relationship (A) and concentration-effect curve (B) for

cAMP accumulation in CHOAA8 cells stably expressing 2AR in response to (-)-CGP 12177 and (-)-isoprenaline...... 112 Figure 5.11. Concentration-effect curves for cAMP accumulation in CHOAA8 cells

stably expressing 2AR at high density in response to (-)-CGP 12177 and (-)-isoprenaline...... 113 Figure 5.12. Concentration-effect curves for cAMP accumulation in CHOAA8 cells

stably expressing 1/2TMDVAR in response to (-)-CGP 12177 and (-)- isoprenaline...... 114

vii

Figure 5.13. Concentration-effect curves for cAMP accumulation in CHOAA8 cells

stably expressing 1AR in response to (-)-CGP 12177 and (-)-isoprenaline in the absence and presence of 100 nM (-)-bupranolol...... 116 Figure 5.14. Concentration-effect curve for cAMP accumulation in CHOAA8 cells

stably expressing 2AR in response to (-)-CGP 12177 in the absence and presence of 100 nM (-)-bupranolol...... 117 Figure 5.15. Concentration-effect curves for cAMP accumulation in CHOAA8 cells

stably expressing 1/2TMDVAR in response to (-)-CGP 12177 and (-)- isoprenaline in the absence and presence of 100 nM (-)-bupranolol...... 118

Figure 5.16. Homology model of the 1AR ...... 119 Figure 5.17. Concentration-effect curves for cAMP accumulation in CHOAA8 cells

stably expressing 1(R222Q)AR in response to (-)-CGP 12177 and (-)- isoprenaline in the absence and presence of 100 nM (-)-bupranolol...... 121 Figure 5.18. Concentration-effect curves for cAMP accumulation in CHOAA8 cells

stably expressing 1(V230I)AR in response to (-)-CGP 12177 and (-)- isoprenaline in the absence and presence of 100 nM (-)-bupranolol...... 122 Figure 5.19. Concentration-effect curves for cAMP accumulation in CHOAA8 cells

stably expressing 1(V230A)AR in response to (-)-CGP 12177 and (-)- isoprenaline in the absence and presence of 100 nM (-)-bupranolol...... 124 Figure 5.20. Concentration effect curves for cAMP accumulation in CHOAA8 cells

stably expressing 1(S228A)AR in response to (-)-CGP 12177 and (-)- isoprenaline in the absence and presence of 100 nM (-)-bupranolol...... 126 Figure 5.21. Concentration-effect curves for cAMP accumulation in CHOAA8 cells

stably expressing 1(S229A)AR in response to (-)-CGP 12177 and (-)- isoprenaline in the absence and presence of 100 nM (-)-bupranolol...... 127 Figure 5.22. Concentration-effect curves for cAMP accumulation in CHOAA8 cells

stably expressing 1(S232A)AR in response to (-)-CGP 12177 and (-)- isoprenaline in the absence and presence of 10 nM (-)-bupranolol...... 128

Figure 5.23. Overlay of TMDV regions of the 1AR homology model, inactive β2AR- T4L bound to carazolol and nanobody stabilised active β2AR ...... 133 Figure 6.1. Two possible alignments of the common features of noradrenaline and (-)- CGP 12177...... 138 Figure 6.2. Chemical structure of (-)-CGP 12177 and its structural isomer 5-[3-(tert- butylamino))2-hydroxypropoxy]1,3-dihydro-2H-benzimidazol-2-one...... 138 Figure 6.3. Concentration-effect relationship for cAMP accumulation in CHOAA8

cells stably expressing 1AR in response to 5-[3-(tert-butylamino))2- hydroxypropoxy]1,3-dihydro-2H-benzimidazol-2-one...... 140 Figure 6.4. Concentration-effect curves for cAMP accumulation in CHOAA8 cells

stably expressing 1AR in response to (-)-isoprenaline in the absence and presence of 1 M 5-[3-(tert-butylamino))2-hydroxypropoxy]1,3-dihydro- 2H-benzimidazol-2-one...... 141 Figure 6.5. Concentration-effect curves for cAMP accumulation in CHOAA8 cells

stably expressing 1AR in response to (-)-CGP 12177 in the absence and presence of 10 M 5-[3-(tert-butylamino))2-hydroxypropoxy]1,3-dihydro- 2H-benzimidazol-2-one...... 141

viii

List of Tables

Table 3.1 Primers Used to Generate 1AR Mutations ...... 53 Table 4.1 Patient Demographics ...... 74 Table 4.2 Effects of Antagonists on the Inotropic Potency of BRL 37344 in the Presence of IBMX (10 M) ...... 77 3 Table 5.1 Affinity Values of (-)-[ H]-CGP 12177 (pKD) and Receptor Densities (Bmax) for ARs Stably Expressed in CHOAA8 Cells ...... 98 Table 5.2 Affinity Values (pKi) of AR Antagonists used in Competition Binding Experiments ...... 102

Table 5.3 Potencies (pEC50) of (-)-Isoprenaline (1HAR), (-)-CGP 12177 (1LAR) and (-)-Bupranolol Affinity Values (pKB) for ARs Expressed in CHOAA8 Cells ...... 110 Table 5.4 Analogous AR Amino Acids at Different AR Structures That Interact at the TMDV-TMDVI Interface ...... 132

List of Abbreviations

AC adenylyl cyclase

ATP adenosine 5-triphosphate

MEM Alpha Modified minimal essential medium

ASCEPT Australasian Society of Clinical and Experimental Pharmacologists and Toxicologists

1AR 1-adrenoceptor

2AR 2-adrenoceptor

3AR 3-adrenoceptor

1LAR low-affinity 1-adrenoceptor

1HAR high-affinity 1-adrenoceptor

1/2TMDVAR 1-adrenoceptor containing transmembrane domain V of the

2-adrenoceptor

β1AR-m23 modified turkey β1-adrenoceptor

β2AR-T4L modified human β2-adrenoceptor

BRL 37344 (RR+SS)[4-[2-[[2-(3-chlorophenyl)-2-hydroxy-ethyl]amino] propyl]phenoxy]acetic acid

CAMKII calmodulin-dependent protein kinase cAMP cyclic AMP

CGP 12177 4-[3-(tert-butylamino))2-hydroxypropoxy]1,3-dihydro-2H- benzimidazol-2-one

CGP 20712A 2-hydroxy-S-[2-[[2-hydroxy-3-[4-[methyl-4-(trifluoromethyl)- 1H-imidazol-2-yl]phenoxy]propyl]amino]ethoxy]-benzamide

CHOAA8 Chinese Hamster Ovary cell line that expresses the tetracycline-controlled transactivator

DMEM Dulbecco’s modified Eagle’s medium

DRY motif aspartic acid-arginine-tyrosine motif

ECL/EL extracellular loop

EIA enzyme immunoassay

GDP guanosine 5-diphosphate

GPCR G-protein coupled receptor

GTP guanosine 5-triphosphate

Gi inhibitory G protein for adenylyl cyclase

Gs stimulatory G protein for adenylyl cyclase

IBMX 3-isobutyl-1-methylxanthine

2+ ICa-L L-type Ca current

ICL intracellular loop

ICI 118,551 1-[2,3-dihydro-7-methyl-1H-inden-4-yl]oxy-3-[(1- methylethyl) amino-2-butanol)]

L-748,337 N-(3-[3-[2-(4-benzenesulphonylamino phenyl)ethylamino]-2- hydroxypropoxy]benzyl acetamide

PDE phosphodiesterase enzyme

SR 58611 ethyl{(7S)-7-[(2R)-2-(3-chlorophenyl)-2-hydroxyethylamino]- 5,6,7,8-tetrahydronaphtyl2-yloxy} acetate hydrochloride pEC50 -log molar concentration of agonist required to produce half maximal response

PI3K phosphatidylinositol 3-kinase

PKA cAMP-dependent protein kinase

PKB protein kinase B pKB -log molar concentration of antagonist which at equilibrium would occupy 50% of receptors in the absence of agonist pKD -log molar equilibrium dissociation constant for a radioligand pKi -log molar binding inhibition constant

RyR2 sarcoplasmic reticulum ryanodine channel

TMD transmembrane domain

S.E.M standard error of the mean

xii

Statement of Original Authorship

The work contained in this thesis has not been previously submitted to meet requirements for an award at this or any other higher education institution. To the best of my knowledge and belief, the thesis contains no material previously published or written by another person except where due reference is made.

Signature: ______

Date: ______

xiii

Publications and Conference Abstracts

Christ T, Molenaar P, Klenowski PM, Ravens U, Kaumann AJ (2011). Human atrial beta(1L)-adrenoceptor but not beta-adrenoceptor activation increases force and Ca(2+) current at physiological temperature. Br J Pharmacol 162(4): 823-839.

Klenowski P, Semmler ABT, Chee K, Iconomou M, Molenaar P (2010).

Contribution of transmembrane domain V amino acids to β1L-adrenoceptor activity and affinity. ASCEPT and Molecular Pharmacology of G-Protein Coupled Receptors Meeting Melbourne.

Molenaar P, Klenowski P, Semmler AB, Chee K, Iconomou M, Tugiono N, Kiriazis

H, Du XJ, Ravens U, Christ T, Kaumann A (2010). The highs (1H) and lows (1L) of human heart 1-adrenoceptors (AR). ASCEPT and Carney Symposium Christchurch.

Molenaar P, Klenowski P, Kaumann AJ (2009). (-)-CGP12177 increases contractile force through β1L-adrenoceptors but not through β3-adrenoceptors in human right atrial myocardium. ASCEPT Sydney.

Klenowski P, Semmler A, Mälzer A, Wei M, Kaumann A, Molenaar P (2008)

Transmembrane spanning domain 5 (TMD5) of the 1-adrenoceptor (1AR) is required for the formation of the ‘low affinity state’. Molecular Pharmacology of G- Protein Coupled Receptors Meeting Sydney.

xiv

Acknowledgements

This work was supported by an APA/IHBI postgraduate student award and a NHMRC grant awarded to Assoc. Prof. Peter Molenaar.

Firstly, I would like to thank my supervisors, Assoc. Prof. Peter Molenaar and Dr Annalese Semmler for their guidance, support and encouragement throughout my candidature. I would also like to thank the past and present members of the Molecular Human Cardiac Pharmacology Group, and in particular, Kelly Chee for her hardwork and assistance with experiments.

I am most grateful to the heart surgeons of The Prince Charles Hospital for providing right atrial appendages, and to Prof. Malcolm West and Assoc. Prof. Ian Yang of the School of Medicine, University of Queensland, to carry out experiments in the in vitro human heart laboratory.

I would also like to acknowledge the lab’s collaborators Prof. David Fairlie (UQ), Dr Andrew Lucke (UQ), Dr Robert Reid (UQ), Dr Torsten Christ (Dresden University) and Prof. Alberto Kaumann (University of Cambridge) for their valuable contribution to this research project.

My thanks also to the members of IHBI for their help and friendly attitude, and to Prof. Ross Young for use of IHBI facilities to carry out molecular pharmacology experiments

Lastly, but by no means least, I would like to thank my family and friends (particularly my Mother, Father and Brother) for their love, encouragement and prayers. Your support, generosity, comfort and guidance has given me the strength to complete this great challenge.

Chapter 1: Introduction

The 1-adrenoceptor (1AR) is activated by (-)-noradrenaline and blocked by all clinically used -blockers. Some -blockers, typified by (-)-CGP 12177 and

(-)-pindolol not only block the 1AR, but also activate it at higher concentrations (~2-3 orders of magnitude) than those required to block it (reviewed Kaumann et al., 2008). To accommodate these findings, it was hypothesised that -blockers such as

(-)-CGP 12177 and (-)-pindolol bind to the 1AR at two different sites, one that blocks (-)-noradrenaline from activating the receptor, the 1H site and another which activates the receptor, the 1L site. Other -blockers typically block 1H- with ~2-3 orders of magnitude higher affinity than 1LAR (Kaumann et al., 2008).

The physiological manifestations of the 1LAR were originally observed in the heart where (-)-pindolol and related indoleamines, (-)-CGP 12177 and other -blockers such as oxprenolol and (-)-alprenolol caused cardiostimulant effects resistant to blockade by propranolol and bisoprolol but which were blocked by moderately low (1 M) concentrations of (-)-bupranolol (reviewed Kaumann et al.,

2008). 1LAR mediated agonist effects were subsequently observed in adipose (Granneman et al., 1992; Sennitt et al., 1998) and smooth muscle tissues (Oostendorp et al., 2000). Despite relative resistance of the cardiostimulant effects to (-)-CGP 12177 and like drugs to blockade by -blockers, notably complete resistance in the presence of 200 nM (-)-propranolol (Kaumann, 1996), a concentration that causes >60-fold rightward shift of noradrenaline effects in heart (Gille et al., 1985), a

1AR mechanism was authenticated in murine cardiac tissues where cardiostimulant effects were observed in tissues from 2AR knockout mice but not 1-/2AR knockout mice (Kaumann et al., 2001). Consistent findings were observed at recombinant 1AR where (-)-CGP 12177 exhibited antagonist properties at low concentrations, but agonist effects at higher concentrations (Pak et al., 1996), and other antagonists displayed a characteristic 2-3 orders of magnitude difference of

Chapter 1 1

affinity against noradrenaline/isoprenaline versus (-)-CGP 12177/(-)-pindolol (reviewed Kaumann et al., 2008).

A recent report claimed that (-)-CGP 12177 increased L-type Ca2+ current

(ICa-L) and contractile force via activation of 3-adrenoceptors (3AR) (Skeberdis et al., 2008). As observed previously with the non-selective phosphodiesterase (PDE) inhibitor 3-isobutyl-1-methylxanthine (IBMX) (Kaumann et al., 1997c) and PDE3- selective inhibitor cilostamide (Kaumann et al., 2007), Skeberdis et al., (2008) observed a marked increase of the contractile responses to (-)-CGP 12177 in the presence of IBMX, in human atrium. However, they interpreted the effects of

(-)-CGP 12177 to be mediated through activation of 3AR. Skeberdis et al., (2008) based the interpretation of their results from human atrial myocytes in which (-)-CGP

12177 increased ICa-L, an effect blocked by the 3AR-selective antagonist L-748,337 (Candelore et al., 1999). However, they did not verify with L-748,337 their claim that the contractile effects of (-)-CGP 12177 are mediated through 3AR (Skeberdis et al., 2008).

The interpretation of Skeberdis et al., (2008) challenged the fundamental hypothesis used as a basis for this PhD project, that the agonist properties of (-)-CGP 12177 and other non-conventional partial agonists are mediated through

1LAR (Joseph et al., 2003; Kaumann et al., 2007; Kaumann et al., 2008; Sarsero et al., 2003). It has been shown previously in human atrial trabeculae that L-748,337 failed to antagonise the positive inotropic effects of the non-conventional partial agonist (-)-pindolol mediated through 1LAR (Joseph et al., 2003). Therefore experiments were performed during the undertaking of this PhD project to investigate whether L-748,337 could antagonise the inotropic effects of (-)-CGP 12177 on human atrial preparations under the conditions of Skeberdis et al., (2008), using IBMX. In addition, concentration-effect curves to (-)-CGP 12177 in the absence and presence of (-)-bupranolol were determined. Evidence from these experiments ruled out an involvement of 3AR and confirmed previous work on human atrial myocardium (Kaumann, 1996; Sarsero et al., 2003), and recombinant

2 Chapter 1

1AR (Joseph et al., 2004b), that showed that the inotropic effects of (-)-CGP 12177 are mediated through 1LAR.

Having confirmed the original hypothesis of this project, the remaining work focused on identifying the molecular and structural determinants of the 1LAR. The molecular features of the 1AR that differentiate between the 1H- and 1LAR forms are poorly understood. Mutagenesis studies have provided conflicting results with respect to amino acids required for 1LAR activity. Amino acids critical at 1HAR have been investigated to determine their involvement in 1LAR pharmacology (Baker et al., 2008). This has revealed that Asp138 in transmembrane domain III (TMDIII) and Asn363 (TMDVI) are required for binding and agonist activity at both

1H- and 1LAR, whilst Ser228 and Ser229 of TMDV contribute to agonist

[(-)-CGP 12177] potency at 1LAR (Baker et al., 2008). It is proposed that an interaction between a nitrogen atom of the indole group of (-)-pindolol and one of the

TMDV serines may contribute to 1LAR activity (Baker et al., 2008; Warne et al.,

2008). The serine residues are also important for catecholamine binding to 1HAR, leading to the suggestion that the 1L- and 1H binding sites must overlap (Baker et al., 2008). Interestingly, the proposed involvement of Asp138 at 1LAR is in contrast to that reported previously (Joseph et al., 2004a), which proposed an involvement of

Asp138 in binding of (-)-CGP 12177 at 1HAR but not 1LAR.

The TMDV region of the 1AR is a critical anchor point for the catechol groups of catecholamines (Rasmussen et al., 2007; Sato et al., 1999; Strader et al.,

1989b; Sugimoto et al., 2002; Warne et al., 2008). The turkey 1AR crystal structure reveals that the catechol hydroxyl groups of (-)-isoprenaline are likely to interact with Ser211 and Ser215 in TMDV. For antagonist activity at 1HAR, cyanopindolol appears to bind in a similar pocket to that occupied by catecholamines such as (-)-isoprenaline, to interdict catecholamine agonist activity (Warne et al., 2011; Warne et al., 2008). The benzimidazolone and indole groups of (-)-CGP 12177 and

(-)-pindolol play a critical role in 1LAR agonist activity (Kaumann et al., 2008), however the binding partners of these groups to the 1LAR are still speculative.

Chapter 1 3

The interactions formed upon ligand binding to the 1AR also influences intramolecular TMD interactions. These effects can cause movement of the TMD regions and may contribute to the stabilisation of inactive and active receptor conformations (Rasmussen et al., 2011a; Rasmussen et al., 2011b; Warne et al., 2011). Consequently, amino acids within TMDV that are remote to the binding pocket may also contribute to the formation of 1LAR via steric/stabilising interactions. Because the 2AR does not form a corresponding low-affinity state (Baker et al., 2002), heterologous amino acids within TMDV were targeted to determine their involvement in binding at 1AR, and their potential involvement in

1LAR activity. Initially, a conservative approach was adopted that examined TMDV as a whole entity. It was hypothesised that if TMDV contributes to 1LAR agonist activity, a chimeric 1AR which incorporates the 2AR TMDV would not display characteristic 1LAR pharmacology. Site-directed mutagenesis was used to generate a chimeric 1AR that contained the TMDV region of 2AR. The functionality of this receptor was tested using radioligand binding studies and cAMP enzyme immunoassays (EIA). Based on these data and results obtained from molecular modelling, individual amino acids within this region were identified and investigated using the above mentioned techniques, to determine their involvement at the 1LAR.

4 Chapter 1

Chapter 2: Literature Review

2.1 G-PROTEIN COUPLED RECEPTORS

GPCRs have a conserved structure that consists of seven helical transmembrane spanning domains, an extracellular N terminus, an intracellular C terminus and three interhelical loops on either side of the membrane. This structure provides a link that allows extracellular molecules to activate intracellular signal transduction pathways without crossing the plasma membrane (Oldham et al., 2008). GPCRs are activated by a diverse range of endogenous ligands including neurotransmitters, inflammatory mediators, hormones, odorants, peptides and enzymes. Upon activation, GPCRs undergo a conformational change that promotes the transfer of GDP for GTP to the -subunit of heterotrimeric GTP binding proteins (G-proteins). This causes the dissociation of the coupled G-protein which splits into , and  subunits. The individual subunits then interact with intracellular proteins, leading to subsequent downstream signalling and modulation of various cellular responses involved in autonomic nervous system transmission, cell homeostasis and cell growth (Deupi et al., 2007; Oldham et al., 2008).

The GPCR superfamily consists of approximately 800 members which have been broadly classified into 5 subfamilies (A,B,C,D and E) based on sequence homology and function (Kobilka, 2007). The rhodopsin-like class A subfamily is the largest comprising approximately 85% of all GPCRs. The development of compounds that block or activate GPCRs has provided an important field of research and therapeutic strategy. While nearly 30% of all commercially available medicines mediate their therapeutic effects through GCPRs, there remains an enormous scope for further research considering these drugs are targeted against a small percentage of the receptor family (Landry et al., 2008; Leach et al., 2007; Overington et al., 2006).

Chapter 2 5

2.2 -ADRENOCEPTORS

Adrenoceptors (AR) are members of the class A GPCR subfamily. Their expression and activation is mediated by the biogenic amines, noradrenaline and adrenaline. Since the subdivision of adrenoceptors into - and - (Ahlquist, 1948), the study of βARs has remained an active field of research. Early investigations into the physiological and pharmacological properties of βARs using a series of structurally related catecholamines resulted in further subdivision into β1AR, β2AR

(Lands et al., 1967a; Lands et al., 1967b) and β3AR (Emorine et al., 1989).

In humans all three members of the βAR subtype are widely and heterogeneously distributed throughout the body in cardiac, skeletal and smooth muscle, neuronal, endocrine and adipose tissues (Brodde et al., 1984; Carstairs et al., 1985; Golf et al., 1985; Heitz et al., 1983; Kaumann et al., 1987; Langin et al., 1991; Lemoine et al., 1988; Liggett et al., 1988; Minneman et al., 1979). Multiple βARs subtypes often contribute to the overall βAR activity within a specific region (Carlsson et al., 1972). Upon activation all three receptors can cause intracellular levels of cAMP to increase via the protein kinase A (PKA)/cAMP signalling pathway (Galandrin et al., 2006; Lefkowitz et al., 2002), although studies have also revealed regulation of other intracellular effectors and signalling pathways including phosphatidylinositol 3-kinase (Leblais et al., 2004), c-Src (Luttrell et al., 1999; Schmitt et al., 2002), Erk 1/2 (Gerhardt et al., 1999; Lindquist et al., 2000; Sato et al., 2008; Soeder et al., 1999), p38 mitogen activated protein kinase (Cao et al., 2000; Gong et al., 2000; Sato et al., 2008; Zheng et al., 2000), Ca2+/calmodulin- dependent protein kinase II pathway (Wang et al., 2004) and the JAK/STAT pathway (Westphal et al., 2008). In the heart, the effects of the sympathetic nervous system are mainly mediated through activation of 1AR (Gilman, 1987; Katz et al.,

1975). Activation of 1AR in heart results in an elevation of cAMP levels, increased PKA activity and phosphorylation of target proteins including L-type Ca2+ channels, phospholamban, troponin I and C-protein which enhance contractility and hastens relaxation (Kaumann et al., 1999; Kaumann et al., 1996b; Molenaar et al., 2000; Molenaar et al., 2007b). Increased cAMP in the sinoatrial node hastens the action potential to cause an increase in heart rate and conduction (Gilman, 1987; Katz et al., 1975).

6 Chapter 2

2.2.1 1AR Gene and Protein Structure

The gene encoding human β1AR is localised on chromosome 10 (Yang-Feng et al., 1990) and was first cloned in 1987 (Frielle et al., 1987). The 2.4 kbp sequence contains a short 5’ untranslated region of 86 bp, a single exon of 1434 bp and an 3’ non-coding region of ~900 bp (Frielle et al., 1987). Within the exon, there are two common naturally occurring variations observed at positions 49 (Ser/Gly) (Borjesson et al., 2000; Maqbool et al., 1999) and 389 (Arg/Gly) (Maqbool et al., 1999; Mason et al., 1999; Tesson et al., 1999). There are a further 24 single nucleotide polymorphisms (SNPs) reported in the population, 11 of which are non-synonymous (Brodde, 2008).

The human β1AR exon encodes a 477 amino acid protein (Figure 2.1). There is a high degree of amino acid sequence homology between the receptor subtypes with

β1- and β2ARs sharing 48.9% homology, while β3AR exhibits 50.7% and 45.5% with

β1AR and β2AR respectively (Emorine et al., 1989). β1AR and β2AR exhibit the highest degree of homology within the TMD regions (71%) (Frielle et al., 1988). The

β1AR protein exhibits the typical GPCR structure containing seven hydrophobic transmembrane -helices connected by hydrophilic extracellular and intracellular loops (Figure 2.1). Within the protein structure, important amino acids involved in ligand binding and G-protein coupling have been identified (Sato et al., 1999; Strader et al., 1989b; Strader et al., 1988; Strader et al., 1987b). This work, performed initially in the β2AR has since been confirmed in the homologous β1AR amino acid residues (Baker et al., 2008).

Chapter 2 7

Figure 2.1. Two-dimensional diagram of the human 1AR adapted from the turkey 1AR-m23 model.

8 Chapter 2

2.3 STRUCTURAL COMPONENTS OF THE 1AR

The β1AR exhibits the typical GPCR structure and provides a mechanism for signal transduction across the plasma membrane (Figure 2.1). In order to facilitate how the β1AR protein structure determines functionality, the important structural features and amino acids will be discussed.

2.3.1 N-Terminus and Extracellular Loops

The major roles of the β1AR N-terminus are in regulation of receptor levels as well as receptor trafficking to the plasma membrane. The β1AR is glycosylated on Asn15 (Frielle et al., 1987), a site that is implicated in receptor trafficking and dimerisation (He et al., 2002). Regulation of cell surface expression is also thought to occur through constitutive and agonist induced cleavage of the N-terminus at 2 sites, Arg31-Leu32 and Pro52-Leu53 (Hakalahti et al., 2010).

The recent publication of the human β2AR-T4 lysozyme fusion protein (β2AR-

T4L) and the modified turkey 1AR (1AR-m23) crystal structures has provided insights into the structural and functional roles of the extracellular loops (ECLs). In both β1AR and β2AR, ECL1 and ECL3 are short loops connecting TMDII and III and TMDVI and VII. ECL2 which connects TMDIV and V is much larger and is more exposed to the extracellular side (Figure 2.1). It is stabilised by two disulfide bonds and a sodium ion (Warne et al., 2008). Furthermore, ECL2 appears to provide an insight into the observed specificity of βAR ligands. ECL2 of the β2AR forms a short -helix that exists as a partial opening over the ligand binding pocket on the extracellular surface (Cherezov et al., 2007). This is in contrast to the corresponding loop in rhodopsin which is a -sheet that is buried in the transmembrane region and covers the ligand binding pocket. The opening created by ECL2 appears to act as an entrance to the binding pocket and is thought to be involved in the initial recognition of ligands and their binding kinetics (Avlani et al., 2007; Klco et al., 2005; Shi et al.,

2004; Warne et al., 2008). ECL2 in both β1AR and β2AR also forms part of the ligand binding site, by forming stabilising interactions with bound ligands. At β2AR, an aromatic interaction is formed between the inverse agonist carazolol and Phe193

Chapter 2 9

of ECL2 (Bokoch et al., 2010), while at the β1AR, Thr203 forms a hydrogen bond with cyanopindolol and Phe201 forms a van der Waals interaction (Warne et al.,

2008). While the backbone structure of the three β1AR ECLs are very similar to

β2AR (Warne et al., 2008), there are 22 amino acid differences between ECL2 and

ECL3 of β1AR compared to β2AR (Bokoch et al., 2010). It is proposed that the difference in amino acid side chains results in spatial and charge distribution differences that may contribute to the sub-type selectivity of β1AR and β2AR ligands (Bokoch et al., 2010).

2.3.2 Transmembrane Domain Regions and 1AR Ligand Binding Pocket The βAR ligand binding pocket is formed by the cylindrical arrangement of the seven hydrophobic transmembrane -helices within the plasma membrane. The specific interactions that occur between catecholamines and amino acid side chains within the TMD regions are well established. In the 1980s the 2AR served as the prototype GPCR for investigating binding mechanisms of endogenous ligands and blockers directly at the receptor. Site-directed mutagenesis and molecular modelling techniques were implemented, resulting in the identification of amino acids involved in catecholamine binding at the 2AR (Sato et al., 1999; Strader et al., 1989b; Strader et al., 1988; Strader et al., 1987b). These results suggested that the main + contact points between catecholamines and β2AR were the bioamine NH3 group with Asp113 in TMDIII, hydroxyl groups of the catechol ring with Ser203, Ser204 and Ser207 in TMDV, the aromatic ring with Phe290 in TMDVI and the chiral -hydroxyl with Asn293 in TMDVI. The recent publication of ARs in complex with antagonists and agonists, has provided further characterisation of the 1AR and

2AR binding sites. This structural information has provided a more authoritative view of the involvement of amino acids identified from mutagenesis studies and afforded a greater understanding of specific βAR ligand interactions.

At the 1AR-m23 crystal structure containing the antagonist cyanopindolol, a binding pocket is formed by 15 amino acid side chains from four different TMDs (III, IV, V and VII) and ECL2 (Figure 2.2). In comparison, the antagonist binding pocket of β2AR defined by carazolol is highly conserved considering there is only

10 Chapter 2

two amino acid differences within an 8 Å distance of the ligand (Thr164 in TMDIV and Tyr308 in TMDVII of β2AR, corresponding to Val172 and Phe325 of β1AR).

Additionally, the recent publication of agonist bound structures of β1AR-m23 has afforded comparison to the cyanopindolol bound structure. The overall structures and binding pockets are very similar (Warne et al., 2011), a result that was expected considering that the introduced point mutations stabilise the receptor preferentially in an inactive state.

Figure 2.2. The ligand binding pocket of β1AR-m23, displaying the amino acids (aquamarine, grey) interacting with cyanopindolol (yellow). H, helix, EL, extracellular loop. Figure from Warne et al., (2008).

The interactions between GPCRs and the ligands that bind to them result in conformational changes to the receptor. Agonists stabilise conformations that promote G-protein coupling leading to increased signal transduction. Antagonists stabilise conformations that reduce the level of G-protein coupling, while inverse agonists reduce basal and agonist independent levels of signal transduction. As a consequence, the amino acids within the binding pocket not only interact with ligands but also play a role in the functionality of the receptor by promoting conformational changes that affect signalling. Here the amino acids critical for ligand binding and their role with respect to βAR functionality will be discussed.

Chapter 2 11

Studies have demonstrated that in the 2AR, Asp79 within TMDII is required for agonist binding (Chung et al., 1988; Sato et al., 1999; Strader et al., 1988; Strader et al., 1987b; Sugimoto et al., 2002) and mutation of the equivalent Asp104 of the

β1AR does not affect agonist or antagonist affinity, but significantly alters Gs- protein coupling and cAMP signalling (Baker et al., 2008). Within TMDIII of β2AR, the side chain of Asp113 interacts with the positively charged protonated amine group of βAR ligands (Strader et al., 1988). Alteration of this residue dramatically decreases agonist potency and antagonist affinity. The homologous residue in the

β1AR, Asp138, is also essential for the binding of antagonists and the βAR agonist isoprenaline (Baker et al., 2008; Joseph et al., 2004b).

There are three conserved serine residues within TMDV of both β1AR and

β2AR which serve as an anchor point for the catechol groups of catecholamines. It was proposed that the hydroxyl side chains of Ser203 and Ser204 at β2AR interact with the meta-hydroxyl and Ser207 with the para-hydroxyl of (-)-isoprenaline (Sato et al., 1999; Strader et al., 1989b). The removal of Ser203 also reduced the affinity of antagonists containing a nitrogen in the heterocyclic ring (Liapakis et al., 2000).

Mutation of the analogous β1AR residues, Ser228, Ser229 and Ser232 resulted in reduced affinity and potency of isoprenaline (Baker et al., 2008). These residues are also involved in antagonist binding at the β1AR, although this result was antagonist dependent (Baker et al., 2008).

The Phe290 within TMDVI of the β2AR appears to form non-polar interactions with the aromatic ring of catecholamines (Swaminath et al., 2005). This residue is part of a cluster of highly conserved aromatic residues which surround the Proline- kink (Pro288) in TMDVI (Shi et al., 2002). It is proposed that these residues form a molecular switch known as the “rotamer toggle switch”, and that the interactions between the aromatic catechol ring of catecholamines, Phe290 and possibly other residues within the switch, alter the angle of the bend around the Pro288, resulting in movement of the cytoplasmic end of TMDVI upon receptor activation (Shi et al.,

2002). Analysis of the equivalent β1AR residue Phe341, suggests no significant

12 Chapter 2

functional role and variable effects with respect to antagonist binding (Baker et al., 2008).

The TMDVI residue Asn293 in β2AR appears to interact with the -hydroxyl group of catecholamines. Replacing Asn293 with leucine, caused a reduction in the affinity and potency of (-)-isoprenaline but not (+)-isoprenaline (Wieland et al., 1996). These results provided a rationale for the observed stereoselectivity for the (-)-enantiomer of catecholamines (Wieland et al., 1996). The affinity and potency of

(-)-isoprenaline was also reduced when the homologous β1AR residue Asn344 was changed to alanine (Baker et al., 2008). Mutagenesis studies have also demonstrated a functional role for the β1AR Lys324 residue in TMDVI. It is proposed that upon agonist binding, an interaction with Lys324 causes movement of TMDVI, inducing a GTP sensitive form of the receptor (Zeitoun et al., 2006). Mutation of the Lys324 to alanine, results in decreased potency of full and partial agonists (Zeitoun et al.,

2006). Finally, Asp363 within TMDVII of β1AR appears to interact with both agonists and antagonists since its mutation significantly reduces antagonist affinity and agonist potency (Baker et al., 2008).

2.3.3 Intracellular Loops and G-Protein Binding

The final structural components of the β1AR include the intracellular loops (ICLs) and the C-terminus. In βARs, ICL2 and ICL3 appear to play a role in receptor activation and the recognition of binding G-proteins (Dixon et al., 1987; Hausdorff et al., 1990; O'Dowd et al., 1988; Strader et al., 1987a). These receptor components also interact with -arrestins, which can regulate expression levels in both an agonist dependent and independent manner, by initiating receptor internalisation (Marion et al., 2006). While the underlying mechanism behind formation of GPCR-G-protein ternary complex has not been fully elucidated, evidence for direct contact of G- proteins within this region of the receptor does exist (Chung et al., 1988; Hausdorff et al., 1990; O'Dowd et al., 1988; Rasmussen et al., 2011b; Strader et al., 1987a).

Studies have identified a highly conserved triplet of amino acids located at the boundary between TMDIII and ICL2 in class A GPCRs. In the β2AR, the aspartic

Chapter 2 13

acid-arginine-tyrosine (DRY) motif has been implicated in the formation of an “ionic lock” that stabilises the receptor in an inactive conformation. It is proposed that the non-covalent intramolecular interactions supporting the formation of the “ionic lock” occur between Arg131 on the cytoplasmic end of TMDIII, the adjacent Asp130 and

Glu268 at the cytoplasmic end of TMDVI at 2AR (Ballesteros et al., 2001; Rasmussen et al., 1999). The mutation of these residues disrupts the interactions between TMDIII and TMDVI resulting in the β2AR becoming constitutively active (Ballesteros et al., 2001). Furthermore, agonists induce conformational changes at

β2AR that lead to disruption of the “ionic lock” interactions (Yao et al., 2006). The

1AR-m23 structure reveals that Tyr149 in ICL2 is close enough to interact with the

Asp138 (homologous to β2AR Asp130) via hydrogen bond formation, suggesting that this residue may also be involved in the formation of the ionic lock at the β1AR.

Elucidation of the contact points between GPCRs and G-proteins has proved difficult. One of the challenges researchers have faced is that altering areas of interest using mutagenesis has led to a reduction in agonist affinity at the receptor, thus reducing G-protein coupling efficiency. To alleviate this problem, studies using peptides and G-proteins have been performed to investigate these interactions (Wu et al., 2000). A drawback of this method could be that binding determinants that promote high-affinity G-protein binding may only be exposed when the receptor is in an agonist-induced active conformation. Studies have provided evidence to support the contact of G-proteins to ICL2 of the M1-muscarinic acetylcholine receptor (Moro et al., 1993), and ICL3 of the 2AR (Taylor et al., 1996), M2 and M3 muscarinic acetylcholine receptor (Wu et al., 2000) and the β2AR (Cheung et al., 1992; Hausdorff et al., 1990; Strader et al., 1987a). Removal of residues 222-229 by deletion from the β2AR resulted in complete loss of coupling to adenylyl cyclase (Strader et al., 1987a), and the corresponding peptide was able to activate the Gs- protein (Okamoto et al., 1991). A mutant β2AR containing a seven amino acid deletion at the carboxyl terminus of its ICL3 exhibited a 50% reduction in adenylyl cyclase activity when exposed to agonist compared with the wild-type receptor, while agonist affinity was not affected (Hausdorff et al., 1990). Furthermore, mutation of the hydrophobic residues at the amino acid terminus of ICL3 significantly reduced isoprenaline mediated adenylyl cyclase activity, suggesting that

14 Chapter 2

the chemical composition of amino acid side chains within this region is critical for G-protein coupling (Cheung et al., 1992).

The recent publication of the β2AR-Gs-protein structure has provided further insights into β2AR-G-protein interactions. Although this structure displays the contact interface between the β2AR intracellular regions of TMDV, TMDVI and ICL2 with the Gs subunit, no direct interaction with G and G subunits was observed (Rasmussen et al., 2011b). This was somewhat surprising considering that the heterotrimeric G-protein is required for efficient coupling (Hekman et al., 1987; Ratnala et al., 2009). Nevertheless, G appeared to play an indirect role in coupling by stabilising the amino end of the -helix of Gs(Rasmussen et al., 2011b). A network of non-polar interactions is formed along the β2AR-Gs interface. Additionally, rearrangement of the cytoplasmic ends TMDV, TMDVI and ICL2 is observed, that accommodates the 5-helix of Gs subunit, causing a rearrangement of the interactions formed by the DRY motif(Rasmussen et al., 2011b).

2.4 GPCR SIGNALLING AND ACTIVATION

Over the last two decades, the understanding of GPCR activation and signalling has advanced considerably. This new knowledge has led to the realisation that the mechanisms underlying these processes are more complex than first thought. Numerous studies have explored these concepts using βARs. Here, the information pertaining to our current understanding, along with challenges facing researchers will be presented.

GPCR activation involves the binding of an agonist to the receptor, resulting in the formation of a ternary complex consisting of a ligand, the receptor and a G-protein (De Lean et al., 1980; Ross et al., 1977). This leads to the exchange of GDP for GTP to the G-protein -subunit, causing dissociation of the complex and of the heterotrimeric G-protein into the GTP- subunit and  subunits. Both GTP-G and G interact with intracellular effector proteins that regulate signalling pathways. In order to determine the function of GPCRs at a molecular level, a fundamental

Chapter 2 15

understanding of how agonist binding results in receptor activation is required. Early studies proposed a two state kinetic model whereby agonist binding induces a conformational change that converts the receptor from an inactive state R to an active state R* (Bond et al., 1995; Chidiac et al., 1994; Gotze et al., 1994; Lefkowitz et al., 1993; Mewes et al., 1993; Monod et al., 1965; Samama et al., 1993; Samama et al., 1994; Schutz et al., 1992). The equilibrium between R and R* determines the level of basal activity (Bond et al., 1995). A full agonist shifts the equilibrium to the high- affinity GTP binding R* state, while inverse agonists and antagonists stabilise the low-affinity GTP binding R state (Bond et al., 1995; De Lean et al., 1980). Partial agonists and neutral antagonists have affinity for both R and R* (Bond et al., 1995). The efficacy of ligands is determined by their ability to shift the equilibrium between the two states. In the case of βARs, a two state model cannot be applied because different agonists have varying efficacies, resulting in diverse functional outcomes. In order to provide a mechanistic explanation of this scenario, it has been proposed that at the β2AR, agonist binding initiates a series of intermediate conformations, ultimately leading to an active conformation with varying affinity towards one or more G-proteins. In addition, this process is further complicated because full and partial agonists appear to stabilise distinct conformational states based on their interactions with the ligand binding pocket and their effect on conformational switches within the receptor (Kobilka et al., 2007).

2.4.1 1AR Activation Until recently, the conformational changes resulting in βAR activation were predicted from models of the inactive βAR crystal structures containing antagonists. However the recent publication of 2 active structures of rhodopsin (Choe et al., 2011; Standfuss et al., 2011), the active adenosine A2 receptor structure (Xu et al., 2011), the active state β2AR-Gs-protein ternary complex (Rasmussen et al., 2011b) and βAR crystal structures in complex with agonists (Rasmussen et al., 2011a; Rasmussen et al., 2011b; Warne et al., 2011), has afforded a greater understanding of the agonist induced conformational changes at these receptors.

16 Chapter 2

The binding pocket of the β1AR is revealed by the position of cyanopindolol (Figure 2.2, Warne et al., 2008). These interactions occur with surrounding hydrophobic residues and the polar residues, Asp121, Asn329, Tyr333 at the “front” and Ser211, Ser215 and Asn310 at the “back” of the binding pocket (Figure 2.3). The residues in the β1AR that interact with cyanopindolol are identical to those that make contact with carazolol in the β2AR binding pocket (Cherezov et al., 2007; Warne et al., 2008). Furthermore, comparison of β1AR and β2AR antagonist structures with

β1AR structures containing bound agonists reveals very little difference with respect to ligand binding (Warne et al., 2011). Additionally, the 1AR-m23 structures containing full and partial agonists are almost identical, with an exception being that full agonists appear to induce a rotamer conformational change of Ser215 in addition to Ser212 (Figure 2.3). It is proposed that the additional hydrogen bonding and stronger helice-helice interactions observed in full agonist binding causes contraction of the ligand binding pocket, while reduced hydrogen bond formation and weaker TMD helical interactions observed with partial agonists do not (Warne et al., 2011).

Figure 2.3. The ligand binding pockets of 2AR with carazolol (a, yellow), 1AR with cyanopindolol (b, yellow) and 1AR with isoprenaline (c, yellow) (Warne et al., 2011). Interacting amino acids are shown in green. Models were generated from 1AR-m23 crystallised with the antagonist cyanopindolol (Warne et al., 2008) and agonist isoprenaline (Warne et al., 2011) and β2AR-T4L crystallised with the inverse agonist carazolol (Cherezov et al., 2007).

There appears to be three structurally significant differences in the ligand pocket at β1AR when full agonists are bound compared to antagonists. These include the rotamer conformation changes in the side chains of Ser212 and Ser215, and contraction of the binding pocket, measured between TMDV and TMDVII (Figures 2.3 and 2.4, Warne et al., 2011). These changes alter the configuration of the binding

Chapter 2 17

pocket to allow the formation of additional hydrogen bonds. Furthermore, they may contribute to disruption of the interhelical interactions between TMDIV (Val172) and TMDV (Ser215) and strengthening of the TMDV (Ser212) - TMDVI (Asn310) interface (Figure 2.3, Warne et al., 2011). The rotamer changes of the TMDV serines are not present in the 1AR-m23 antagonist structure containing cyanopindolol (Figure 2.3), reducing the potential number of hydrogen bonds. A reduced number of hydrogen bonds are also predicted to occur between the headgroup of the partial agonist salbutamol and TMDV (Warne et al., 2011). Further stabilisation of the ligand binding pocket is observed via the formation of additional hydrogen bonds between Asp121 in TMDIII and the -hydroxyl of (-)-isoprenaline, strengthening the interactions of TMDIII and TMDVII (Warne et al., 2011). Dobutamine lacks a -hydroxyl group, reducing the potential number of hydrogen bonds to TMDIII. It is proposed that disruption of the TMDIII-TMDVII interface for dobutamine and the TMDV-TMDVI interface for salbutamol, in addition to reduced number of hydrogen bonds, may contribute to the decreased efficacy of these ligands. The increased network of hydrogen bonds and stabilisation of TMDIII-TMDVI and TMDV-

TMDVI interfaces, predicted from the 1AR-m23-isoprenaline bound structure, results in a 1 Å contraction of the binding pocket measured from the distance between TMDV and TMDVII (Figure 2.4).

Figure 2.4. Comparison of the 1AR-m23 structures with either cyanopindolol (grey) or isoprenaline (orange). The figure shows the positions of the ligands and the interactions formed between TMDV and TMDVII. Isoprenaline causes a 1 Å contraction of the binding pocket compared to cyanopindolol. Figure from Warne et al., (2011).

18 Chapter 2

While comparison of the βAR structures containing antagonists and agonists have provided important structural information relating to the binding pocket, it has afforded little structural insight into how agonist binding facilitates receptor activation. The fact that even agonist bound structures represent inactive receptor conformations, demonstrates the need for formation of the ternary complex to complete βAR activation (Yao et al., 2009).

Researchers have faced a difficult task in crystallising a GPCR ternary complex due to their instability in detergents (Rasmussen et al., 2011a). Recently however, the publication of a nanobody stabilised active state of the β2AR

(Rasmussen et al., 2011a) and the β2AR-Gs complex structure (Rasmussen et al., 2011b) has provided an insight into the structural changes that occur as a result of βAR activation. The Nb80 nanobody is a camelid antibody fragment that displays

G-protein like properties towards β2AR (Rasmussen et al., 2011a). Crystallisation of the active β2AR state complex was achieved using β2AR-T4L (complex used to obtain inactive β2AR carazolol structure), Nb80 and the high-affinity agonist BI- 167107. This structure has provided the first clues about the rearrangement of TMD regions, and interactions with the intracellular loops that occur in order to facilitate G-protein binding.

The major structural differences between the inactive β2AR-T4L, β2AR-Nb80 and β2AR-Gs occur at the cytoplasmic end of the receptor and involve a large outward movement of TMDV and TMDVI and an inward movement of TMDIII and TMDVII (Figure 2.5, Rasmussen et al., 2011a; Rasmussen et al., 2011b). This helical rearrangement generates a hydrophobic pocket which is occupied by Nb80. The TMD movements are facilitated by the interacting Nb80, which appears to break the hydrogen bonds of the ionic lock residues Arg130, Arg131 and Glu368, and form its own interaction with Arg131 (Rasmussen et al., 2011a). In the β2AR-Gs structure, similar helix movements are observed and the ionic lock Arg131 residue interacts with a tyrosine residue of the Gs subunit (Rasmussen et al., 2011b). An inward movement of TMDIII and TMDVII by 4 Å and 2.5 Å respectively, in addition to an outward 6 Å movement of TMDV is also observed, similar to the β2AR-Nb80

Chapter 2 19

structure (Figure 2.5, Okamoto et al., 1991; Rasmussen et al., 2011a; Rasmussen et al., 2011b).

Figure 2.5. Comparison of the inactive carazolol bound β2AR (β2AR-Cz) with the Nb80 stabilised β2AR (β2AR-Nb80, orange) and the active β2AR-Gs structure (β2AR-Gs, green). This cytoplasmic view depicts the large outward movements of TMDVI in the active structures β2AR-Nb80 and β2AR-Gs relative to the inactive β2AR-Cz. The smaller TMDIII, TMDV and TMDVII movements are also shown. Figures from Rasmussen et al., 2011a (left) and Rasmussen et al., 2011b (right).

The BI-167107-β2AR-T4L interactions are relatively similar to the interactions observed in the 1AR-m23-isoprenaline bound structure and also the β2AR-carazolol bound structure. The biggest difference within the β2AR active and inactive ligand binding pocket is the inward movement (2.1 Å) of TMDV (Rasmussen et al., 2011a).

A contraction of the TMDV region is also observed in the 1AR-m23-isoprenaline structure (Figure 2.3, Warne et al., 2011). At the β2AR, the agonist BI-167107 interacts with TMDV Ser203 and Ser204, similarly to the interactions formed between isoprenaline and TMDV Ser211 and Ser215 at the β1AR. Additionally, the

Ser212 in β1AR and Ser204 in β2AR do not interact directly with the ligand, but form an additional hydrogen bond with Asn310 on TMDVI in β1AR and Asn293 in β2AR, which in turn interacts with Tyr308 in ECL3. In contrast, carazolol forms only a single hydrogen bond with Ser203 at β2AR, similarly to cyanopindolol at β1AR, which forms a hydrogen bond with the analogous residue (Ser211). These results suggest that the increased stabilisation of the binding pocket via interactions between ligands and the TMDV serines, in addition to the formation of helice-helice

20 Chapter 2

interactions when full agonists are bound to βARs, are critical in stabilising a high- affinity G-protein coupling state.

While these crystal structures provide a fascinating insight into βAR-ligand interactions, how these relatively subtle changes are propagated into the large structural changes observed in the cytoplasmic region of the receptor is still speculative. The development of ligands that will allow the manipulation of this process is further complicated by the fact that different agonists are likely to induce and/or stabilise different conformations at different GPCRs, leading to varying structural outcomes. Kinetic studies also support the formation of early intermediate conformations upon agonist binding, followed by larger and slower conformational changes that involve intramolecular switches and TMD interactions (Kobilka et al., 2007; Liapakis et al., 2004; Swaminath et al., 2005). The available βAR crystal structures represent a single static view of this process. Another consideration is that βARs display basal activity. This suggests that the tertiary structure of the binding pocket is highly dynamic and can transition between inactive (R) and active (R*) states in the absence of a ligand. Furthermore, studies demonstrating that constitutively active mutants can be generated by a single point mutation, shows that disruption of a single interaction may be adequate to cause receptor activation. This level of sensitivity and structural complexity displayed by GPCRs, suggests that a dynamic view of receptor activation is required in order to understand how intermediate conformations are formed by rearrangement of amino acid packing and TMD movements.

2.4.2 1AR Signalling Recent advances in GPCR research have revealed the complex nature of GPCR signalling. It is now well established that the classical lock and key mechanism cannot be applied. Studies have demonstrated that GPCR activation often results in a multitude of signalling outcomes and may involve coupling to multiple G-proteins, signalling via G and/or G and even G-protein independent pathways (reviewed in Audet et al., 2008; Hoffmann et al., 2008; Kenakin, 2007; Seifert et al., 2009). Consequently, the classification of GPCR ligands has also become increasingly

Chapter 2 21

difficult. Ligands once classed as simple agonists, have been shown to possess varied efficacy or even inhibit one or more signalling pathways controlled by a particular GPCR (Kenakin, 2001). This phenomenon, often referred to a ligand bias or biased agonism, may in the future be exploited as a therapeutic strategy to develop drugs that activate clinically favourable pathways while inhibiting detrimental pathways. The multiple signalling pathways that are activated by βARs have been extensively reviewed (Audet et al., 2008; Hoffmann et al., 2008; Kenakin, 2007; Seifert et al., 2009). This section will address functional outcomes resulting from βAR activation with particular emphasis on β1AR signalling and function in the human heart.

β1ARs are expressed in a variety of mammalian tissues. Tissues that contain high levels of β1AR expression include the brain, kidney, adipose tissue and the heart. In the brain, β1ARs are localised in various regions and are the dominantly expressed βAR subtype (Gibbs et al., 2005; Machida et al., 1990; Minneman et al.,

1979; Paschalis et al., 2009). β1ARs are found predominantly within neurons of the brain where they play a role in neuronal plasticity, memory and learning, and memory retrieval processes (Gibbs et al., 2005; Murchison et al., 2004; Ramos et al.,

2005). β1ARs are also involved in regulating the secretion of melatonin from the pineal gland and have also been implicated in mood change (Leonard, 1997).

β1ARs in the kidney act to increase renin release from the juxtaglomerular apparatus (Kurtz et al., 1999; Torretti, 1982). High levels of β1AR expression are also found in brown (Bukowiecki et al., 1980; Levin et al., 1986; Rothwell et al., 1985) and white adipose tissues (Germack et al., 2000; Lacasa et al., 1986; Mauriege et al., 1988). β1AR activation in white adipose tissue controls lipolysis (Berlan et al.,

1978), while in brown adipose tissue β1AR activation has been implicated in heat production (Lafontan et al., 1993). In the heart, β1AR expression has been extensively characterised in a variety of mammalian species including rodent (Juberg et al., 1985; Ota et al., 1993), guinea pig (Molenaar et al., 1987; Summers et al., 1987), feline (Lemoine et al., 1991; Molenaar et al., 1985), ferret (Lowe et al., 2002), canine (Molenaar et al., 1988), pig (Sillence et al., 2005) and human (Buxton et al., 1987; Molenaar et al., 1989; Summers et al., 1989). Activation of cardiac

β1AR and β2ARs mediates positive chronotropic, inotropic, lusitropic and

22 Chapter 2

dromotropic effects (Gille et al., 1985; Hall et al., 1990; Katz et al., 1975; Kaumann et al., 1989; Kaumann et al., 1996b; Molenaar et al., 2000; Molenaar et al., 2002;

Molenaar et al., 2007b). Chronic β1AR stimulation in the human heart is thought to result in receptor desensitisation and cause changes in heart pathophysiology that are associated with the progression of heart failure (reviewed in Bristow, 2000; Brodde et al., 1995).

2.5 AR SIGNALLING IN THE HUMAN HEART

In the human heart, β1AR and β2AR coexist (Ablad et al., 1974) with β1AR representing ~60-70% and ~70-80% of the total βAR population in atria and ventricles respectively (reviewed in Brodde, 1991; Brodde et al., 1999). The endogenous catecholamines adrenaline and noradrenaline released from adrenal glands and sympathetic nerves respectively, bind to β1AR and β2AR and stabilise a receptor conformation that facilitates binding of the heterotrimeric Gs-protein (Birnbaumer et al., 1987; Gilman, 1987).

Formation of the βAR-Gs ternary complex results in the transfer of GDP for GTP from the Gs subunit (Asano et al., 1984; Brandt et al., 1986). The mechanism underlying GDP/GTP transfer has not been elucidated. GTP binding results in the dissociation of Gs from the  subunits, allowing the Gs subunit to activate the membrane bound adenylyl cyclase, resulting in the conversion of ATP to cAMP (Figure 2.6, reviewed in Levitzki, 1988). Adenylyl cyclase is more tightly coupled to

β2AR than β1AR even though β1AR is dominantly expressed (Brodde et al., 1984; Waelbroeck et al., 1983). This variance has been reported in right atrium (Brodde et al., 1984; Gille et al., 1985; Lemoine et al., 1988), right ventricle (Bristow et al., 1989; Kaumann et al., 1989; Kaumann et al., 1987; Lemoine et al., 1988) and recombinant βARs (Green et al., 1992; Levy et al., 1993). A proline-rich region in ICL3 appears to contribute in part to the decreased adenylyl cyclase coupling efficiency of β1AR compared to β2AR (Green et al., 1994), but the physiological role underlying this anomaly is still unclear. βAR mediated increases in intracellular cAMP activates PKA which phosphorylates numerous proteins involved in cardiac excitation-contraction including, L-type calcium channels, phospholamban, troponin

Chapter 2 23

I and C protein (Kaumann et al., 1999; Kaumann et al., 1996b; Molenaar et al.,

2000; Molenaar et al., 2007b). Additionally, acute β1AR activation causes PKA- independent phosphorylation of Thr17-phospholamban via the calmodulin-dependent protein kinase (CAMKII) pathway. Dephosphorylation of Thr17-phospholamban is catalysed by type 1 phosphatase which in turn, is inhibited by PKA-dependent phosphorylation of protein phosphatase inhibitor 1 (Kaumann et al., 1999). Activation of this pathway appears to evoke myocyte apoptosis in animal models (Wang et al., 2004; Zaugg et al., 2000; Zhu et al., 2003).

Figure 2.6. Schematic representation of main βAR signal transduction pathways in cardiac sympathetic nervous transmission. Pre-junctional regulation of noradrenaline is facilitated by β2AR and αARs. Noradrenaline binds to post-junctional β1ARs that are coupled to the Gsα-cAMP pathway. Cyclic AMP activates cAMP dependant protein kinase leading to the phosphorylation of several proteins involved in increasing force of contraction and hastening of relaxation. βARs are also coupled to other pathways including the β2AR Giα signalling pathway, which causes antiapoptotic effects through phosphatidylinositol 3-kinase (PI3K) and protein kinase B (PKB) signalling (Schluter et al., 1998). Additionally, Ca2+ increases mediated by 2+ β1AR stimulate Ca calmodulin-dependent protein kinase (CAMKII) in animals (Machida et al., 2005). Figure adapted from Molenaar et al., (2005).

There is evidence for β3AR expression in human heart (Moniotte et al., 2001a; Moniotte et al., 2001b), however their functional role remains controversial. Studies have reported conflicting results with respect to contractile function (Gauthier et al., 1996; Kaumann et al., 2008; Molenaar et al., 1997; Rozec et al., 2006; Skeberdis et al., 2008). Gauthier et al., (1996), reported that β3AR agonists produced negative

24 Chapter 2

inotropic effects on human ventricle obtained from endomyocardial biopsies. This effect could be due to release of nitric oxide (NO) from endothelial and/or endocardial cells, considering the negative inotropic effects of the β3AR agonist BRL 37344 were prevented by the NO antagonist L-NMA (Napp et al., 2009). In other studies, β3AR agonists were devoid of cardiodepressant or cardiostimulant effects in human ventricular trabeculae (Kaumann et al., 2008; Molenaar et al., 1997). In contrast, increased contractile force in human atrium and increases in L-type calcium current (ICa-L) reportedly from β3AR activation have been observed (Skeberdis et al., 2008).

In human cardiomyocytes, cardiac excitation-contraction coupling is modulated by activation of β1AR and β2AR via the Gs-PKA-cAMP pathway (Figure 2.7, Kaumann et al., 1997c; Molenaar et al., 2007a). Increased contractility results from Ca2+ influx through L-type calcium channels, causing the release of larger amounts of Ca2+ from the sarcoplasmic reticulum ryanodine channel (RyR2) and Na+/Ca2+ exchanger, which then diffuses through the myocyte and interacts with troponin C on myofilaments to initiate contraction (Kaumann et al., 1997c; Molenaar et al., 2007a). This part of the contraction mechanism is called systole. The diffusion of Ca2+ away from the myofilaments and re-uptake in the sarcoplasmic reticulum by Ca2+ATPase (and to lesser extent the Na+/Ca2+ exchanger) and extrusion from the myocyte causes relaxation and is known as diastole (Bers, 2002; Blayney et al., 2009; Molenaar et al., 2007a). The quantity of Ca2+ released from the sarcoplasmic reticulum through RyR2 plays a critical role in determining the magnitude of systole. The duration of contraction and length of diastole is determined by the rate of uptake of Ca2+ into the sarcoplasmic reticulum.

Chapter 2 25

Figure 2.7. The effects of the βAR-Gs-protein-adenylyl cyclase-cAMP-protein kinase A pathway in human heart on systole (left panels) and diastole (right panels). Activation of 1-and 2ARs during systole causes PKA-dependent phosphorylation of L-type Ca2+ channels which increases Ca2+ influx leading to a larger release of Ca2+ from the sarcoplasmic reticulum (SR) through ryanodine channels (RyR2). Ca2+ binds to troponin C initiating actin–myosin cross-bridge cycling. 2AR activation causes higher cAMP production compared to 1AR due to tighter coupling to AC. Activation of 1-and 2ARs causes hastening of relaxation by PKA-catalysed phosphorylation of Ser16-phospholamban and troponin I. Phosphorylation of Thr17- phospholamban is caused by calcium calmodulin kinase II. AR, adrenoceptor; PDE, phosphodiesterase; PLB, phospholamban; Ac, adenylyl cyclase; TNC, troponin C; TNI-P, phosphorylated troponin I. Figure from Molenaar et al., (2007).

26 Chapter 2

In the human heart both β1AR and β2AR regulate systole and duration of diastole. This occurs due to PKA dependent phosphorylation of L-type calcium channels causing increased influx of Ca2+ (Trautwein et al., 1990), increased sensitivity of phosphorylated RyR2 to Ca2+ (Blayney et al., 2009; Xiao et al., 2007), increased sarcoplasmic reticulum Ca2+ stores (Blayney et al., 2009; Li et al., 2000) caused by PKA-dependent phosphorylation of phospholamban, which reduces the inhibitory effect of phospholamban on the IIA sarcoplasmic reticulum Ca2+-ATPase pump (Kaumann et al., 1999; Molenaar et al., 2000; Molenaar et al., 2007b), phosphorylation of troponin I which causes reduced affinity of Ca2+ for troponin C (Zhang et al., 1995), resulting in hastening of relaxation (reviewed in Molenaar et al., 2007a) and increased sensitivity to luminal Ca2+ concentrations (Figure 2.7, Blayney et al., 2009; Xiao et al., 2007). PKA-dependent phosphorylation of RyR2 increases the sensitivity to Ca2+ induced activation and increases the probability of the channel opening (Blayney et al., 2009; Marx et al., 2000). Hyperphosphorylation of RyR2 is observed in heart failure and causes destabilisation and desensitisation of the receptor resulting in increased Ca2+ leak from the sarcoplasmic reticulum (Blayney et al., 2009). This appears to lead to impaired contractile function and is also associated with cardiac arrhythmias (Marx et al., 2000). In the human heart, it is not known whether PKA-dependent phosphorylation of the RyR2 occurs as a result of βAR activation. βAR subtype activation of the RyR2 requires investigation, because it may provide a therapeutic strategy for the management of arrhythmias in heart failure.

2.5.1 The Role of1AR in Heart Failure Human heart failure is a complex clinical condition that is characterised by reduced cardiac output, fluid retention and elevated venous pressure (Katz, 2003). In the failing human heart, activation of 1ARs occurs as a physiological response to facilitate adequate hemodynamic function and perfusion pressure in organs. During heart failure, chronic activation of 1AR leads to hypertrophy and cardiac remodelling (Cohn et al., 1984; Engelhardt et al., 2004). A decrease in the sensitivity of 1AR to catecholamines caused by receptor down-regulation and decreased 1AR density is observed in human heart failure (Bristow et al., 1986; Brodde, 1991).

Chapter 2 27

The desensitisation and internalisation of 1AR, like many GPCRs, appears to involve a three step process that includes, receptor phosphorylation, interactions with scaffolding proteins such as postsynaptic density 95/disc large/zonula occuldens-1 (PDZ) proteins and -arrestins, receptor endocytosis and degradation mediated by clathrin-coated pits (reviewed in Hanyaloglu et al., 2008). The 1AR is phosphorylated at consensus sites within ICL3 and the C-terminal tail by PKA, protein kinase C and G-protein receptor kinases (Friedman et al., 2002; Hausdorff et al., 1991; January et al., 1998; January et al., 1997; Seibold et al., 1998; Seibold et al., 2000; Tran et al., 2004) and interacts with the scaffolding proteins PSD 95, MAGI 2 and -arrestin 1 and 2 resulting in receptor internalisation and recycling (Gage et al., 2005; Gardner et al., 2007; He et al., 2006; Xiang et al., 2002; Xu et al., 2001). Receptor expression and receptor occupancy appear to be the main contributing factors that regulate these processes (He et al., 2006). Because heart failure patients exhibit an increase in the level of plasma noradrenaline (Cohn et al., 1984), it would appear that the processes involved in receptor desensitisation and internalisation are upregulated (He et al., 2006; Hu et al., 2000; Xiang et al., 2002; Xu et al., 2001).

The current therapeutic strategy employed to manage heart failure involves administering compounds that block the effects of catecholamines at βARs. However, until the mid 1990s βARs agonists (noradrenaline, adrenaline, dobutamine and dopamine) were commonly prescribed to patients with chronic heart failure on the premise that increased sympathetic nervous system activity would increase blood pressure and maintain cardiac output (Molenaar et al., 2005). It later became apparent that βAR agonists and other positive inotropic agents had detrimental effects on the heart resulting in increased mortality (Packer, 1992; Packer et al., 1991; Persson et al., 1996; The Xamoterol in Severe Heart Failure Study Group, 1990). Nevertheless, AR agonists still have a role in the management of human heart failure to provide inotropic support for the failing heart when it is decompensated and when other reparative measures have failed (Molenaar et al., 2005). In practice, they are more often used as a ‘bridge to transplantation’ in end- stage heart failure.

28 Chapter 2

The introduction of propranolol into clinical practice in 1965 by Sir James Black for the treatment of coronary artery disease provided the basis for the mechanism of β-blockade as a therapeutic strategy (Black et al., 1965). This clinical practice was also pioneered by Finn Waagstein and colleagues in 1975 (Waagstein et al., 1975) who introduced β-blockers as a treatment for congestive heart failure. The principle of β-blockade was based on decreasing the energy demand of the heart in order to provide a more favourable hemodynamic state (Waagstein et al., 1975). This hypothesis was initially met with uncertainty and scepticism, especially considering the report that short term administration of β-blockers was detrimental to patients with ventricular dysfunction (Epstein et al., 1966). It wasn’t until the mid 1990s when large scale clinical trials revealed the benefits of β-blockers and established their use as a means to clinically manage heart failure.

The pharmacological effects of β-blockers differ significantly and hence have evolved since their introduction into clinical practice. The first clinically used β-blockers such as propranolol were introduced as a treatment for angina and were non-selective towards β1AR and β2ARs (Black et al., 1965). The second generation of β-blockers developed in the 1970s and 80s selectively antagonised β1ARs and included compounds such as metoprolol and bisoprolol (Ablad et al., 1975; Schliep et al., 1984). These compounds, mainly used to treat hypertension, aimed at selectively blocking the β1AR in order to ameliorate the vascular and pulmonary side effects associated with non-selective 1-2AR-blockers. This hypothesis was never conclusively proved (Newton et al., 1999; Newton et al., 1996). The third class of β-blockers developed in the 1980s included the compounds carvedilol and bucindolol and had additional properties in that they not only blocked both β1ARs and β2ARs, but they also demonstrated antioxidant activity and vasodilating properties through + blockade of α1-adrenoceptors or activation of K channels (Bristow et al., 1992; Bristow et al., 1998; Gilbert et al., 1996; Gilbert et al., 1990).

The therapeutic effects of second and third generation β-blockers have been extensively tested in large scale clinical trials (Colucci et al., 1996; Packer et al.,

Chapter 2 29

1996; Packer et al., 2001; Poole-Wilson et al., 2003; The Beta-Blocker Evaluation of Survival Trial Investigators, 2001; The CIBIS-II Investigators, 1999; The MERIT- HF Investigators, 1999; Waagstein et al., 1993). Results of these trials have reported reduced morbidity and mortality with the β-blockers metoprolol, bisoprolol and carvedilol. Interestingly however, not all trials produced survival benefits (The Beta- Blocker Evaluation of Survival Trial Investigators, 2001; The Xamoterol in Severe Heart Failure Study Group, 1990). This appears due to the intrinsic sympathomimetic activity of some β-blockers, an effect which is clinically unfavourable in heart failure (Packer, 1998; The Xamoterol in Severe Heart Failure Study Group, 1990).

Although the results of clinical trials demonstrate that some β-blockers produce survival benefits in heart failure patients, considerable inter-individual variability exists (Lechat et al., 1997). Studies have begun to explore genetic variation as a potential mechanism underlying inter-individual responses to β-blockers (Chen et al., 2007; Liggett et al., 2006; Mialet Perez et al., 2003; Terra et al., 2005). These studies have implicated the 389 (Arg/Gly) polymorphism as a contributing factor to the effectiveness of β-blocker treatment. Results from these studies demonstrate that patients carrying the Arg389Arg variant exhibit significantly greater improvement in left ventricular ejection fraction (LVEF) compared to Gly389Gly patients after treatment with carvedilol and metoprolol (Chen et al., 2007; Mialet Perez et al., 2003; Terra et al., 2005). The observation that Arg389 βARs have higher basal activity than Gly389 βARs in recombinant systems (Joseph et al., 2004c; Mason et al., 1999), has led to the suggestion that catecholamines produce larger effects through Arg389 βARs, which leads to more effective β-blockade (Chen et al., 2007).

Bucindolol, another β-blocker which like carvedilol is non-selective with respect to β1ARs and β2ARs and causes vasodilatation, produced no survival benefits in clinical studies (Poole-Wilson et al., 2003). However a follow-up study did uncover a significant survival benefit in Arg389Arg patients, although no improvement in LVEF was observed (Liggett et al., 2006). A closer examination of the pharmacology of bucindolol could uncover the reason for it’s ineffectiveness compared to carvedilol. One possible difference could be that it can activate the low- affinity site of the β1AR. This property, previously observed with drugs such as

30 Chapter 2

(-)-pindolol and (-)-CGP 12177 (4-(3-tertiarybutylamino-2-hydroxypropoxy)- benzimidazole-2-on hydrochloride), causes cardiostimulant effects in humans and arrhythmias in animal models (Kaumann et al., 1996a; Molenaar et al., 1997).

2.6 1LAR PHARMACOLOGY

Results that demonstrated cardiostimulant effects of β-blockers at concentrations that were higher than those that caused blockade were first reported by Kaumann et al. (1973). Kaumann initially set out to test the affinity of a series of β-blockers for βARs, and found that some of these compounds had intrinsic agonist activity in tissue preparations of feline and guinea pig hearts. Because these β-blockers produced cardiostimulant effects that were less than the maximal effect for the agonist (-)-isoprenaline (a reference agonist used to measure maximal βAR effects), they were initially considered to be partial agonists. This was proposed based on receptor theory which stated by definition that a full agonist could produce a maximal effect at a concentration that occupies a small percentage of receptors, while a partial agonist is required to saturate the receptor population in order to produce it’s maximal effect (Stephenson, 1956). Therefore, one could expect the KP (equilibrium dissociation constant of a partial agonist) of a partial agonist to be the same as the concentration required to produce a 50% maximal effect (EC50). This was the case for the β-blockers pronethalol and practolol, however pindolol was atypical because it did not exhibit the classic behaviour of a partial agonist (Kaumann et al., 1973a).

The half maximal cardiostimulant effects of pindolol occurred at concentrations that were significantly greater (no less than one order of magnitude) than the KB (equilibrium dissociation constant of an antagonist, measured from competitive blockade of the effects of an agonist) recorded against (-)-isoprenaline (Kaumann et al., 1973a). The dissociation between concentrations of pindolol required for blockade of the cardiostimulant effects of (-)-isoprenaline and it’s own cardiostimulant effects prompted the re-classification of β-blockers as either conventional or non-conventional partial agonists (Figure 2.8, Kaumann, 1989). A

Chapter 2 31

non-conventional partial agonist causes activation only at high receptor occupancy (Kaumann et al., 1973b)

Figure 2.8. The correlation between the potencies (pD2) and affinities (pKa) of partial agonists on isolated heart preparations. The open symbols represent conventional partial agonists acting via β1ARs (○), β2ARs (∆) or most probably both (□). The closed symbols represent non-conventional partial agonists such as pindolol and indoleamines (●), CGP 12177 (▲) and alprenolol (■) acting via atypical βARs. Graph from Kaumann, 1989.

2.6.1 Mechanistic Studies Into the Pharmacology of 1LAR The dissociation between blockade and stimulation raised questions with regard to the mechanism through which pindolol, other indoleamines and chemically similar compounds elicited cardiostimulation. The first question was whether these effects were mediated through βARs. Using both enantiomers, (-)-pindolol and (+)-pindolol on guinea pig atrium, it was found that (-)-pindolol produced biphasic activity, measured on a concentration-effect curve (Walter et al., 1984). The curve

32 Chapter 2

consisted of a high-potency (H) component (pEC50 = 9.2) and a low-potency (L) component (pEC50 = 6.1). The addition of the β1AR selective antagonist (-)-bisoprolol into the system effectively blocked the H-component but had no effect on the L-component. Meanwhile, the L-component was blocked with moderate potency by the non-selective βAR antagonist (-)-bupranolol (Morris et al., 1981;

Walter et al., 1984). The β2AR selective antagonist ICI 118,551 had no effect on either component. The concentration-effect curve produced by (+)-pindolol was monophasic (pEC50 = 7.6), and the cardiostimulant effects were blocked by ICI

118,551, suggesting that these effects were mediated through β2AR (Walter et al., 1984). It was concluded from these experiments that at low concentrations

(-)-pindolol mediated antagonism of (-)-isoprenaline through a high-affinity β1AR population, while at high concentrations, (-)-pindolol caused cardiostimulation through a low-affinity receptor population (Walter et al., 1984). At this time the βAR subtype which (-)-pindolol had low-affinity for remained contentious.

Further studies analysing the behaviour of non-conventional partial agonists provided evidence to suggest that the heart may express a third βAR (Kaumann, 1989). Properties of this putative cardiac βAR were it’s activation by non- conventional partial agonists at high concentrations, inability of propranolol and bisoprolol to block stimulant effects and blockade by moderately low (1 μM) concentrations of (-)-bupranolol (Kaumann, 1989; Walter et al., 1984).

2.6.2 The Putative β4-Adrenoceptor: A Working Hypothesis Studies into the dissociation between stimulation and blockade by the hydrophilic βAR antagonist CGP 12177 (Staehelin et al., 1983) on cardiac tissue were subsequently performed (Arch et al., 1993; Kaumann, 1983; Kaumann, 1989; Pak et al., 1996). Kaumann, (1996) reported that high concentrations of (-)-CGP 12177 (60-600 nM) caused positive inotropic effects on right atrial appendages (human right atrial heart muscle) obtained from patients with coronary artery disease undergoing surgery (Kaumann, 1996). These effects were resistant to blockade by (-)-propranolol (200 nM), a concentration that causes a >60 fold rightward shift to the concentration-effect curve of (-)-noradrenaline at β1AR (Gille et al., 1985), and

Chapter 2 33

blocked with moderate potency by (-)-bupranolol (pKB = 7.3, Kaumann, 1996, compared to 9.1 and 9.7 for β1AR and β2AR, Lemoine et al., 1983). Results from this study led Kaumann to conclude that cardiostimulation of (-)-CGP 12177 was mediated through a third cardiac βAR population which was possibly related to the

β3AR cloned by Emorine et al., (1989). This assumption was based on:

 the inotropic effects of (-)-CGP 12177 measured in right atrial tissue were

propranolol (a high-affinity β1- and β2AR blocker) resistant, but were blocked with moderate affinity by bupranolol.

 (-)-bupranolol blocked (-)-CGP 12177 (pKB = 7.3) at β1AR in human right

atrial tissue similarly to transfected human β3AR (pKB = 7.7) (Blin et al.,

1993; Blin et al., 1994) and β3ARs from rat adipose tissue (pKB = 7.3-7.5) (Langin et al., 1991).

 The potency of (-)-CGP 12177 on human right atrial appendages causing

agonist effects (pEC50 = 7.3) was similar to the binding affinities of

(-)-CGP 12177 measured at cloned recombinant β3AR (pKI = 7.3) (Blin et

al., 1993) and in β3AR from human omental fat (Revelli et al., 1993).

Shortly after, Kaumann and Molenaar published results of a study comparing similarities and differences between the third cardiac βAR and colonic β3AR from rat (Kaumann et al., 1996a). In these experiments (-)-CGP 12177 caused positive inotropic effects in rat left and right atrial tissue preparations. These effects were resistant to blockade by 200 nM (-)-propranolol and 3 μM ICI 118,551 (β2AR selective antagonist) and blocked by 1 μM (-)-bupranolol (pKB = 6.4-6.8), 3 μM CGP

20712A (pKB = 6.3-6.4) (β1AR selective antagonist) and 6.6 μM SR 59230A (β3AR selective antagonist) (pKB = 5.1-5.4). (-)-CGP 12177 also displayed partial agonist activity via β3AR (pKA = 7.3-7.5), causing colonic relaxation effects which were maximal at about approximately 55% of those cause by (-)-isoprenaline. These effects were resistant to blockade by 200 nM (-)-propranolol, 3 μM CGP 20712A and

3 μM ICI 118,551. However, 2 μM (-)-propranolol (pKB = 6.0), 1 μM (-)-bupranolol

(pKB = 6.4), and 3 μM SR 59230A (pKB = 6.3) antagonised the relaxant effects of

(-)-CGP 12177. The results differ in that the β1AR selective agonist CGP 20712A blocked the effects of (-)-CGP 12177 in atria, but did not cause blockade of effects in the colon. Also, the β3AR selective antagonist SR 59230A effectively blocked the

34 Chapter 2

colonic relaxation effects of (-)-CGP 12177 but displayed decreased affinity in antagonising the positive inotropic effects of (-)-CGP 12177 in right and left atria. This pharmacology suggested that the activity mediated by (-)-CGP 12177 occurred through different receptors. Furthermore, micromolar concentrations of the following

β3AR selective agonists; BRL 37344, ZD 2079, CL 316242 and SR 58611A were potent colonic relaxants, but did not cause cardiostimulation or affect cardiostimulation caused by (-)-CGP 12177. Due to the distinct differences in pharmacology displayed by (-)-CGP 12177 and β3AR selective agonists and antagonists, Kaumann and Molenaar concluded that the third cardiac βAR was different to the colonic β3AR.

Because the third cardiac βAR exhibited significant differences to the colonic

β3AR and had distinctively different pharmacology compared to β1AR and β2AR, the term “putative β4-adrenoceptor” was coined to differentiate it from the β3AR, thus avoiding confusion (Molenaar et al., 1997). Further studies followed, that aimed to identify this novel receptor encoded gene and further characterise its pharmacological properties.

2.6.3 Proposal of a Low-Affinity Form of the β1-Adrenoceptor Pak and Fishman et al. (1996) were the first to propose that the hypothetical

βAR was in fact a low-affinity form of the β1AR, in what has become a seminal study in the field. Using (±)-CGP 12177 on human or rat β1AR stably transfected into Chinese Hamster Fibroblast Cells (CHW), Pak and Fishmann reported its ability to antagonise increases in cAMP levels produced by (-)-isoprenaline at low concentrations, while higher concentrations (approximately 2 log units) caused concentration-dependent increases in cAMP (Figure 2.9). (±)-CGP 12177 also increased cAMP in Baby Hamster Kidney (BHK) and Chinese Hamster Ovary

(CHO) cells expressing human β1AR (pEC50 = ~7.7). In general, (±)-CGP 12177 behaved as a partial agonist with respect to isoprenaline, however in cell lines containing high receptor densities (~ 1 pmol/mg protein) it behaved as a full agonist.

The agonist effects of (±)-CGP 12177 were proportional to β1AR receptor densities, and were antagonised by the β1AR selective antagonist CGP 20712A (pKB ~ 7.7).

Chapter 2 35

The β2AR selective antagonist ICI 118,551 blocked (±)-CGP 12177 agonists effects with 80-fold lower potency, which suggested that the agonist activity was being mediated through β1AR (Pak et al., 1996). (±)-CGP 12177 behaved as a weak partial agonist in Chinese Hamster cells expressing β2AR.

Figure 2.9. The biphasic activity of CGP 12177 is represented in the above concentration-effect curve. Chinese Hamster Fibroblast cells expressing human β1ARs were exposed to 2 nM isoprenaline in the presence of increasing concentrations of (±)-CGP 12177. Figure from Pak et al., 1996.

Competition binding studies were employed to further study the binding sites responsible for the effects of (±)-CGP 12177. Using membrane preparations 3 expressing β1AR, (-)-[ H]-CGP 12177 labelled high-affinity β1AR binding sites with pKD = 9.7. (±)-CGP 12177 competed with high-affinity (pKi = 9.5) for 90% of 125 binding sites occupied by (-)-[ I]-cyanopindolol, and with low-affinity (pKi = 7.7) for the remaining 10% of binding sites. The low-affinity of (±)-[3H]-CGP 12177 at

β1AR was similar in its potency, leading Pak and Fishman to conclude that the antagonist effects could be mediated through a high-affinity β1AR receptor population, while the agonist effects were mediated through a low-affinity β1AR receptor population coupled to Gs protein. This conclusion prompted subsequent

36 Chapter 2

recombinant βAR studies (Baker, 2005a; Baker, 2005b; Baker et al., 2003; Joseph et al., 2004b; Konkar et al., 2000b), and studies in βAR knockout mice (Kaumann et al., 2001; Kaumann et al., 1998; Konkar et al., 2000a; Konkar et al., 2000b).

2.6.4 Pharmacological Differences Between the Putative β4-Adrenoceptor and Low-Affinity β1-Adrenoceptor In 1998, the contribution of Sarsero and colleagues resulted in the development 3 of a radioligand binding assay using (-)-[ H]-CGP 12177 to study the putative β4AR (Sarsero et al., 1998). The use of this assay on rat atrial membranes demonstrated the 3 stereoselectivity of catecholamines for the putative β4AR. (-)-[ H]-CGP 12177 was 3 used to label β1AR, β2AR and putative β4AR. The observed binding of (-)-[ H]-CGP

12177 was biphasic, consisting of a high-affinity binding site (pKD = 9.4,

β1AR/β2AR) and low-affinity binding site (pKD = 7.6, β4AR). Binding in the presence of 500 nM (-)-propranolol revealed that 35% of total specific sites were labelled with (-)-[3H]-CGP 12177 at high-affinity , while 65% were labelled with low-affinity (Sarsero et al., 1998). The biphasic binding of (-)-[3H]-CGP 12177 was also observed in ventricular membranes with pKD = 9.0 (β1AR/β2AR) and pKD = 7.3 3 (β4AR) (Sarsero et al., 1999). The ratio of (-)-[ H]-CGP 12177 binding at high and low-affinity sites was different to that reported by Pak and Fishmann where low- affinity binding represented only 10% of the total binding sites (Pak et al., 1996).

The low-affinity binding of (-)-CGP 12177 observed by Pak and Fishmann also caused an increase in cAMP accumulation, which was attributed to stimulation of a guanidine nucleotide sensitive form of the β1AR, coupled to the Gs protein which comprised 10% of the total β1AR population (Pak et al., 1996). However, in rat atrium the putative β4AR represented approximately 35% of the total binding sites and appeared insensitive to guanidine nucleotides (Sarsero et al., 1998). These inconsistencies, required further investigation to determine if the agonist properties of (±)-CGP 12177 observed through the low-affinity form of recombinant β1AR (Pak et al., 1996), were indeed the same as the putative β4AR in rat atria (Sarsero et al., 1998).

Chapter 2 37

2.6.5 The Underlying Role of β1-Adrenoceptors in Putative β4-Adrenoceptor Pharmacology

Direct evidence supporting the contribution of β1AR to the effects caused by non-conventional partial agonists such as (-)-CGP 12177 was provided from experiments with recombinant β1AR and tissues from genetically engineered βAR knockout mice (Kaumann et al., 2001; Kaumann et al., 1998; Konkar et al., 2000a; Konkar et al., 2000b). Experiments using brown adipose tissue membranes from wild-type mice demonstrated that (-)-CGP 12177 stimulated adenylyl cyclase activity with a high-affinity component and low-affinity component (Konkar et al., 2000a). The low-affinity activity of (-)-CGP 12177 was absent in brown adipose tissue membranes from β3AR knockout mice, while tissues from β1AR knockout mice lacked the high-affinity activity. In addition, the low-affinity activity of (-)-CGP 12177 was resistant to blockade by 1 M CGP 20712A, while CGP 20712A caused blockade of the high-affinity component with a pKB of 7.4, consistent with the value obtained from recombinant β1AR (Konkar et al., 2000a; Konkar et al., 2000b). Taken together, these data suggested that (-)-CGP 12177 increased adenylyl cyclase activity with low-affinity via β3AR and with high-affinity via β1AR (Konkar et al., 2000a).

Experiments using β3AR knockout mice (Kaumann et al., 1998), β2AR knockout mice and β1AR/ β2AR knockout mice (Kaumann et al., 2001) confirmed that the previously observed cardiostimulation caused by (-)-CGP 12177 was mediated through β1AR. (-)-CGP 12177 elicited cardiostimulant effects in atria from

β2AR knockout mice and β3AR knockout mice but these effects were not present in the atria from β1AR/β2AR knockout mice (Kaumann et al., 2001; Kaumann et al.,

1998) (Figure 2.10). The results indicate that the pharmacology of the putative β4AR in human heart was not due to a novel receptor gene, as previously proposed (Kaumann et al., 1998). Furthermore, these results clearly demonstrated the obligatory requirement of β1AR in the mediation of cardiostimulation effects by (-)-CGP 12177 (Kaumann et al., 2001).

38 Chapter 2

Figure 2.10. Polygraph traces comparing the effects (-)-CGP 12177 (1 μM/L), on left atrium from β2AR knockout (β2AR KO) and β1AR/β2AR (β1AR/β2AR KO) double knockout mice set up in the same tissue bath. Increased force of contraction was observed in the β2AR KO atria but not the β1AR/β2AR KO atria after addition of (-)-CGP 12177 (1 μM/L). IBMX (a phosphodiesterase inhibitor) and dibutyryl cAMP caused increased force of contraction in both atria. Data from Kaumann et al., 2001

These observations were further characterised using CHO cells stably expressing rat and human β1AR at physiological densities (Konkar et al., 2000b). (-)-CGP 12177 increased adenylyl cyclase activity in CHO cells expressing rat or human β1AR with pEC50 values of 7.83 and 7.90 respectively (Konkar et al., 2000b). These effects were relatively resistant to blockade by propranolol, while

CGP 20712A caused blockade of (-)-CGP 12177 activity with pKB values of 7.35 and 7.38 at rat and human β1AR respectively. Furthermore, higher concentrations of CGP 20712A were required to block the effects of (-)-CGP 12177 compared to isoprenaline, providing further evidence for the proposal of two distinct sites within the β1AR (Konkar et al., 2000b).

2.6.6 Structural Implications of High and Low-Affinity Binding of Non- Conventional Partial Agonists to β1-Adrenoceptors

The potential for the β1AR to exist in high and low-affinity states was first endorsed by Pak and Fishmann et al., 1996, and became an accepted view among other researchers (Baker, 2005a; Baker et al., 2003; Kompa et al., 1999; Lowe et al., 2002) following further investigation (Kaumann et al., 2001; Kaumann et al., 1998;

Konkar et al., 2000a; Konkar et al., 2000b). This concept implies that β1AR can form two separate receptor conformations, a high-affinity conformation that is formed

Chapter 2 39

when (-)-CGP 12177 binds to the β1H site and acts as an antagonist to block the effects of catecholamines, and a low-affinity conformation that causes cardiostimulant effects, and is formed when (-)-CGP 12177 binds to the β1L site. Saturation and kinetic binding studies demonstrated that (-)-[3H]-CGP 12177 binds to two distinct sites in both heart (Sarsero et al., 2003) and recombinant β1AR

(Joseph et al., 2004b). It was postulated that the binding of (-)-CGP 12177 to β1HAR caused a conformational change in the receptor that inhibited the binding of catecholamines. In contrast, the binding of (-)-CGP 12177 to β1LAR causes its own distinct conformational change, that resulted in the observed agonist activity of non- conventional partial agonists (Kaumann et al., 2008). The high-affinity (β1HAR) and low-affinity (β1LAR) agonist states of the β1AR should not be confused with the high (R*) and low-affinity (R) GTP binding states, proposed from the ternary complex models (Bond et al., 1995; De Lean et al., 1980).

Elucidating the important structural components involved in agonist and antagonist binding to βARs has emerged as a critical area of research. The binding partners of catecholamines at β1HAR are now well established, however, limited structural information exists on the contact points between non-conventional partial agonists and β1LAR. A study that investigated the role of amino acids involved in catecholamine binding at β1AR revealed a role for Asp138 and Asn363 at both 1H- 3 and 1LAR (Baker et al., 2008). A complete loss of specific binding of (-)-[ H]-CGP 12177 was observed when the Asp138 residue was mutated to alanine, glutamate, histidine, asparagine or serine. In addition, no agonist responses were observed for (-)-isoprenaline and (-)-CGP 12177 when Asp138 was changed to alanine. Similarly, specific binding of (-)-[3H]-CGP 12177 and agonist activity of (-)-CGP 12177 was abolished when the Asn363 residue was mutated to alanine (Baker et al., 2008).

Interestingly, the proposed involvement of Asp138 in the low-affinity agonist binding of (-)-CGP 12177 is in marked contrast to that proposed previously, where the potency of (-)-CGP 12177 was reduced by only 5-fold and the affinity of (-)-bupranolol was reduced by only 0.4 log units when Asp138 was changed to glutamate (Joseph et al., 2004a). Furthermore, high-affinity antagonist binding of

40 Chapter 2

(-)-CGP 12177 (pKD = 9.4) measured at β1AR was reduced (pKD = 7.6) when Asp138 was mutated to glutamate. The potency of (-)-isoprenaline for increases in cAMP was decreased 500,000 fold at Glu138-1AR compared to Asp138-1AR and (-)-bupranolol antagonism (1 μM) was abolished (Joseph et al., 2004a). Based on these data, it was concluded that Asp138 contributes to ligand affinity and activity at

β1HAR not but β1LAR (Joseph et al., 2004a).

The common human 1AR polymorphic amino acids Arg389Gly differentially affect (-)-isoprenaline and (-)-CGP 12177 agonist responses at recombinant 1AR

(Joseph et al., 2004c). At Gly389-1ARs, the maximum cAMP response of

(-)-isoprenaline was reduced by 97% compared to Arg389-1ARs, while for (-)-CGP 12177 it was reduced by only 46%, making (-)-CGP 12177 a full agonist at Gly389-

1ARs (Joseph et al., 2004c). These studies have provided evidence to suggest that

(-)-isoprenaline through 1HAR and (-)-CGP 12177 through 1LAR, have different binding partners and stabilise different 1AR conformations that result varying levels of efficacy.

2.6.7 Functional and Clinical Implication of the β1L-Adrenoceptor

Cardiostimulation by non-conventional partial agonists through β1LAR has been observed in rat atrium (Kaumann et al., 1997b) and human heart (Kaumann, 1996). These effects are mediated through Gs-protein-cAMP-PKA pathway (Kaumann et al., 1997b; Sarsero et al., 2003), and have been shown to result in arrhythmias in murine ventricular myocytes (Lowe et al., 1998), presumably through

β1LAR (Kaumann et al., 2008). The activity of non-conventional partial agonists was reduced in the end stages of heart failure in both rat and human (Kompa et al., 1999; Sarsero et al., 2003), a result that was consistent with reduced binding density of

β1AR observed in end stage heart failure (Bristow et al., 1986; Brodde, 1991).

These observations may demonstrate why the effectiveness of β-blockers used in the treatment of heart failure is varied. A classic example of this is bucindolol, which unlike other clinically used β-blockers metoprolol (The MERIT-HF

Chapter 2 41

Investigators, 1999), bisoprolol (The CIBIS-II Investigators, 1999) and carvedilol (Packer et al., 1996; Packer et al., 2001) does not provide an overall benefit to heart failure patients (The Beta-Blocker Evaluation of Survival Trial Investigators, 2001).

It has been postulated that bucindolol may cause positive inotropic effects via β1LAR (Kaumann et al., 2008). Bucindolol has been shown to elicit cardiostimulation in human myocardium preparations in a propranolol insensitive manner (Bundkirchen et al., 2002). It has also been suggested that the partial agonist activity of bucindolol may cause arrhythmias via β1LAR (Kaumann et al., 2008). However the effects of chronic activation of β1LARs and the clinical mechanisms underlying the ineffectiveness of bucindolol in human heart failure have not been investigated.

Although 1H- and 1L- are separate states of 1AR, they can both cause cardiostimulant effects through activation of the Gs-protein-cAMP-PKA pathway.

Consequently, increased activity of 1LARs could be expected to contribute to the progression of heart failure. It is also possible that chronic activation of 1LAR may lead to cardiac dysfunction and heart failure (Molenaar et al., 2011; Tugiono et al., 2010), forming the basis of the development of β-blockers that effectively block the

β1LAR to be used in combination with current β1HAR β-blockers for treating heart failure.

It has been proposed that the unique pharmacology of (-)-CGP 12177, (-)-pindolol and derivatives is conferred by the presence of the benzimidazolone [(-)-CGP 12177] and indole [(-)-pindolol] groups. A common feature of these substituents is the common nitrogen atom at position 1 (Figure 2.11) which may be critical for agonist activity. Modifying this atom by steric hindrance or by changing the polarity may result in compounds that bind to β1LAR with reduced agonist effects. While these assertions are yet to be investigated, it is believed that identifying the molecular features that confer agonist activity of non-conventional partial agonists will lead to the development of novel compounds that block 1LAR with higher affinity (pK ~ 9) than those currently available (current pK  7).

42 Chapter 2

3 O N O N H 4 H 3 4 H N OH OH O 2 2 N 5 N 5 H 6 1 H 6 1 (-)-CGP 12177 (-)-Pindolol

Figure 2.11. Chemical structures of (-)-CGP 12177 and (-)-pindolol show the common nitrogen at position 1.

2.7 SUMMARY AND IMPLICATIONS

The literature review has provided a detailed insight into the structure and function of 1AR, the role of 1AR in human heart failure, the atypical pharmacology of 1AR and the existence of 1H- and 1LAR forms. Clearly demonstrated is the advancement and implementation of research methodology in the field of AR research that has significantly developed our current knowledge of AR activation and signalling. The research has revealed the complexity of these mechanisms and raised additional questions that require further investigation in order to understand the intricacies of these processes.

The literature review has demonstrated a clear knowledge gap with respect to

1AR pharmacology. The identification and subsequent characterisation of non- conventional partial agonists has been investigated for more than 30 years, however the existence of two forms of pharmacology at 1AR produced by a single ligand has not been conclusively explained. In the present study, the following research hypotheses and aims were formed:

The hypotheses of the project are that:

 Non-conserved amino acids of 1- and 2AR in TMDV confer agonist

activity of (-)-CGP 12177 at the low-affinity site of the 1AR.

Chapter 2 43

 Modification of non-conserved amino acids in TMDV of the 1AR will

alter the activity of (-)-CGP 12177 at the 1AR.

 The benzimidazolone group of (-)-CGP 12177 and indole group of

(-)-pindolol are critical for agonist activity at 1LAR. Structural

modification will lead to the development of 1LAR antagonists.

The aims to address the hypotheses were to:

 determine the effect of heterologous 1AR - 2AR TMDV amino acids on

agonist activity following substitution into the 1AR.

 determine if the agonist properties of (-)-CGP 12177 are conferred by the substitution of the benzimidazolone group by investigating the properties of the structural isomer 5-[3-(tert-butylamino))2-hydroxypropoxy]1,3-

dihydro-2H-benzimidazol-2-one at recombinant 1AR.

The long term aim being:-

 To develop a -blocker that can block 1LAR with higher affinity (pK ~ 9) than those currently available (current pK  7) for potential therapeutic use in heart failure.

44 Chapter 2

Chapter 3: Materials and Methods

3.1 RESEARCH DESIGN

Currently, the molecular attributes that distinguish β1LAR and β1HAR are not fully understood. Because the β2AR does not form a corresponding low-affinity binding site (Baker et al., 2002), it was hypothesised that TMDV amino acids of

1AR that are heterologous with respect to 2AR, might be critical for 1LAR and that their absence in 2AR prevents the existence of a corresponding “2LAR”.

Preliminary experiments targeted a broad region of the β1AR (TMDV) to determine whether this region is involved in the formation of β1LAR. This region was hypothesised as being a critical region based on pharmacological observations including:

1, The differing agonist activity of a series of compounds at 1LAR where, (-)-CGP 12177 > (-)-pindolol >>> (-)-propranolol (Figure 3.1).

2. The assumption that the agonist activity is conferred through the indole group (blue) of pindolol and benzimidazolone group of (-)-CGP 12177 (blue) and not (-O.CH2.CHOH.CH2.NH.CH(CH3)x, (where x = 2 or 3) since (-)-propranolol has no intrinsic activity while pindolol and (-)-CGP 12177 have agonist activity (Figure 3.1).

3. An assumption of the binding site of the position 1 nitrogens of the indole group of pindolol and benzimidazolone group of (-)-CGP 12177, predicted from the interaction of the catechol group of catecholamines with serines residues within

TMDV of 1ARs.

Chapter 3 45

O-CH2-CHOH-CH2-NH-CH(CH3)2

(-)-Propranolol

O-CH2-CHOH-CH2-NH-CH(CH3)2

(-)-Pindolol N O-CH2-CHOH-CH2-NH-C(CH3)3

N (-)-CGP 12177 O N

Figure 3.1. Structural considerations of -blockers with different agonist activity at

1LAR.

The generation of a 1/2TMDVAR chimera and subsequent analysis demonstrating altered intrinsic activity of (-)-CGP 12177, prompted further characterisation of this region using molecular modelling. A homology model of

1AR based on the crystal structure of β2AR-T4 lysozyme fusion protein identified two amino acid residues (Arg222, Val230) that could act alone or in combination to contribute to ligand binding/efficacy and confer the intrinsic activity of

(-)-CGP 12177 at 1LAR. It was proposed that by altering these residues to the corresponding 2AR residues (Arg222-1AR to Gln222 and Val230-1AR to Ile230), would affect the agonist activity of (-)-CGP 12177.

These two strategies have provided information about the 1LAR ligand binding site and have identified a region of the 1AR that contributes to the formation of 1LAR.

A modification of the original research design was implemented to include experiments that confirmed that the positive inotropic effects of (-)-CGP 12177 are mediated through 1LAR but not 3AR. The requirement for these experiments was evoked by a report published during the course of planned studies, which claimed that (-)-CGP 12177 increased ICa-L and contractile force through activation of 3AR

46 Chapter 3

(Skeberdis et al., 2008). This interpretation was based on results from human atrial myocytes in which (-)-CGP 12177 increased ICa-L, an effect that was reversed by the

3AR-selective antagonist L-748,337 (Candelore et al., 1999). The claim that the contractile effects of (-)-CGP 12177 were mediated through 3AR was not verified with L-748,337.

3.2 MATERIALS AND METHODS

3.2.1 General Reagents Molecular Biology General molecular biology reagents were purchased from the following vendors: Qiagen Spin Miniprep Kit, Qiagen Qiaquick Gel Extraction Kit, and Qiagen Plasmid Purification Kit from Qiagen (Doncaster, Australia); Macherey-Nagel Plasmid Mini Kit from Macherey-Nagel (Clayton, Australia); Expression vectors pBI-L and pTRE2hyg from Clonetech (Mountain View, CA); DNA polymerase (Klenow fragment), and T4 DNA ligase from Invitrogen (Mulgrave, Australia); Restriction enzymes HindIII, XbaI, EcoRV and Thermosensitive Alkaline Phosphatase from Promega (Alexandria, Australia); Stratagene QuikChange™ multi site-directed mutagenesis kit distributed by Integrated Sciences (Chatswood, Australia); Oligonucleotides used in mutagenesis polymerase chain reactions (PCR) and DNA sequencing were synthesised by Sigma (Castle Hill, Australia).

Cell Culture General cell culture reagents were purchased from the following vendors: CHOAA8 cell line from Clonetech (Mountain View, CA); Alpha Modified minimal essential medium from Lonza (Mt Waverly, Australia); Dulbecco’s modified Eagle’s medium from Lonza or Invitrogen (Mulgrave, Australia); 10% heat inactivated tetracycline approved foetal calf serum (FCS) from Scientifix (Clayton, Australia); 1% penicillin/streptomycin, 1% L-Glutamine, 1% GlutaMAX-I, G418, hygromycin and trypsin/EDTA (0.25% trypsin (w/v), 0.913 mM EDTA) from Invitrogen (Mulgrave, Australia); doxycycline from Scientifix (Clayton, Australia); FuGene 6 transfection reagent from Roche Diagnostics (Castle Hill, Australia); sterile 1.5 ml

Chapter 3 47

eppendorf tubes from Quantum Scientific (Murarrie, Australia); 3 mm cloning discs and 1X cell freezing media-DMSO from Sigma Aldrich (Castle Hill, Australia).

Drugs Drugs were purchased from the following vendors: BRL 37344 [(RR + SS)[4- [2-[[2-(3-chlorophenyl)-2-hydroxy-ethyl]amino] propyl]phenoxy]acetic acid], (-)-CGP 12177 [(7)-4-(3-tertiarybutylamino-2-hydroxypropoxy) benzimidazol-2- one], IBMX (3-isobutyl-1-methylxanthine), (-)-isoprenaline hydrochloride, (-)-nadolol, ICI 118,551 [1-[2,3-dihydro-7-methyl-1H-inden-4-yl] oxy-3-[(1- methylethyl) amino-2-butanol)], CGP 20712A [(2-hydroxy-S-[2-[[2-hydroxy-3-[4- [methyl-4-(trifluoromethyl)-1H-imidazol-2-yl]phenoxy]propyl]amino]ethoxy] benzamide) were from Sigma Aldrich (Castle Hill, Australia); bisoprolol was from Merck KGaA (Darmstadt, Germany); (-)-[3H]-CGP 12177 (specific activity 30-60 Ci/mmol) from Perkin Elmer (Melbourne, Australia); (-)-bupranolol was a gift Dr Klaus Sandrock (Sanol-Schwarz, Monheim, Germany); L-748,337 [N- (3- [3-[2-(4- benzenesulphonylaminophenyl)ethylamino]-2- hydroxyl-propoxy]benzyl acetamide was a gift from Prof. Alberto Kaumann (University of Cambridge). See Appendix A for list of chemical structures.

Buffers If a pH value is stated, the baseline value was obtained using a labChem-TPS pH electrode (TPS Pty. Ltd. Brisbane, Australia), before addition of either 1 M HCl or 2 M NaOH to obtain the desired pH:

Phosphate-buffered saline (PBS): 137 mM NaCl, 10 mM sodium phosphate buffer, 2.7 mM KCl (pH 7.4).

Tris-acetate-EDTA (TAE): 40 mM Tris-acetate, 1 mM EDTA (pH 8.0).

Agarose gel loading dye: 0.09% (w/v) bromophenol blue, 0.09% (w/v) xylene cyanol FF, 60% (v/v) glycerol, 60 mM EDTA, 1 M Tris (pH 7.6).

Tris-Incubation buffer: 5 mM EGTA, 1 mM EDTA, 4 mM MgCl2, 1 mM ascorbic acid, 0.5 mM phenyl-methyl-sulfonyl fluoride, 50 mM Tris HCl, (pH 7.7).

48 Chapter 3

Modified Krebs buffer: 125 mM Na+, 5 mM K+, 2.25 mM Ca2+, 0.5 mM Mg2+, 98.5 − 2- - 2- mM Cl , 0.5 mM SO4 , 29 mM HCO3 , 1 mM HPO4 , 0.04 mM EDTA, and equilibrated with 95% O2/5% CO2

Cyclic AMP assay buffer: 0.05 M acetate, 0.02% (w/v) Bovine Serum Albumin, 0.01% (w/v) preservative (pH 5.8).

Cyclic AMP lysis buffer 1B: 0.25% dodecyltrimethylammonium bromide.

Cyclic AMP wash buffer: 0.01 M phosphate containing 0.05% (w/v) Tween 20 (pH 7.5)

Bacterial Culture Bacterial culture was performed in Luria-Bertani (LB) broth [1% (w/v) Bacto- Tryptone 0.5% (w/v) Bacto-yeast extract and 0.5% (w/v) NaCl, pH 7.0]. Construct

DNA (1AR, 2AR or modified 1AR gene cloned into pCDNA3.1, pBI-L or pTRE2hyg) was transformed into Library Efficiency DH5™ competent cells from Invitrogen (Mulgrave, Australia). Ligation reactions were transformed into One Shot® MAX Efficiency® DH5™ competent cells from Invitrogen (Mulgrave, Australia). Mutagenesis reactions were transformed into XL-10 Gold ultra-competent cells (provided in the Stratagene QuikChange™ multi site-directed mutagenesis kit). Transformation and growth conditions are detailed in the General Methods section.

Cyclic AMP assays Cyclic AMP assays were performed using reagents supplied in the GE Healthcare Biotrak cAMP enzyme-immunoassay kit (GE Healthcare Bio-sciences Pty. Ltd., Rydalmere, Australia) following the manufacturer’s instructions (RPN225). Kit reagents were equilibrated to room temperature prior to use. The cAMP standard was prepared by addition of 2 ml lysis 1B buffer. The final solution contained cAMP at a concentration of 32 pmol/ml. The concentrations of cAMP required for the standard curve were obtained by diluting the cAMP solution in lysis buffer 1B. The anti-cAMP serum was prepared by adding 11 ml of lysis buffer 2B (components of lysis buffer 2B not provided by the manufacturer). The cAMP- horseradish peroxidase conjugate was made up in 11 ml of diluted assay buffer.

Chapter 3 49

3.2.2 General Methods

It is acknowledged that the generation of 1AR and 2AR recombinant clones and 1/2TMDVAR chimera were performed by Dr Annalese Semmler. The

1(V230I)AR, 1(R222Q)AR, 1(S228A)AR, 1(S229A)AR and 1(S232A)AR were generated by Kelly Chee and Dr Annalese Semmler. I performed the recombinant techniques associated with the generation of 1(V230A)AR and subcloned the

1(S232A)AR modified DNA fragment from pCDNA3.1 into pTRE2hyg.

Molecular Biology

Generating 1AR and 2AR recombinant clones

A 1 wild-type construct (1pBI-L) was created as described by Kaumann et al., (2007). In addition two plasmids containing either a full-length ~1.3 Kb wild type 1AR or a full-length ~1.2 Kb wild type 2AR cloned into the HindIII/XbaI site of pCDNA3.1, were kindly donated by Professor Roger Summer’s laboratory

(Monash University, Australia). It is noted that the 1AR gene used in both constructs (pBI-L and pTRE2hyg) contained the Arg389 genotype.

The cDNA fragments were subcloned into pBI-L and/or pTRE2hyg cloning vectors (Clonetech, Australia). The 1- or AR construct DNA was transformed into Library Efficiency DH5™ competent cells by standard heat shock procedure. Briefly, the cells were thawed on ice and 50 l aliquots of cells were mixed with ~1 l construct DNA (10 ng) in ice cold 1.5 ml eppendorf tubes and incubated on ice for a further 30 min. The cells were then heat shocked for 20 sec in a 42oC water bath and chilled on ice for 2 min. LB broth (950 l) was added to transformed cells which were then incubated at 225 rpm for 1 hr at 37oC. The transformed culture (2.5-100 l) was plated on 1.5% LB ampicillin agar plates and incubated overnight at 37oC. Single colonies were selected and inoculated into 5 ml LB broth supplemented with ampicillin (100 g/ml) and incubated overnight at 37oC with shaking at 225 rpm. The overnight cultures were minipreped using the Qiagen Spin Miniprep Kit or the Macherey-Nagel Plasmid Mini Kit using the manufacturer’s instructions. Agarose loading dye was added to 1 l of the samples and the DNA was visualised

50 Chapter 3

by gel electrophoresis with 0.8% (w/v) TAE agarose gels containing ethidium bromide (0.5 g/ml). The gels were run in TAE buffer at 100 V for 40-60 min.

The wild type 1AR or wild type 2AR gene fragments were then excised from pCDNA3.1 as HindIII/XbaI fragments and were separated by gel electrophoresis as detailed above. The gene fragments were then excised from the agarose gel and purified using the Qiagen Qiaquick Gel Extraction Kit as per the manufacturer’s protocol. Prior to ligation, the purified gene inserts were treated with DNA polymerase (Klenow fragment) to create blunt ends. Reactions contained 10 l purified DNA insert (0.5-1 g), 1.5 l of 10 X Klenow reaction buffer, 0.2 l acetylated BSA (10 mg/ml), 0.8 l dNTP (1 mM), 10 l DNA polymerase (Klenow fragment, 0.5 U/l) and made up to a total of 15 l with nuclease free H2O. Reactions were incubated at room temperature for 15 min. The vector DNA (pBI-L or pTRE2hyg) was linearised using EcoRV (Promega, Australia) and dephosphorylated with Thermosensitive Alkaline Phosphatase (TSAP, Promega, Australia) following the manufacturer’s protocol. All DNA fragments were electrophoresed and purified as above. DNA concentrations were determined by spectrophotometric analysis at λ=260 nm.

Ligation of the blunt-ended genes into the EcoRV site of the respective cloning vectors was performed using T4 DNA ligase following the manufacturer’s instructions. Briefly, purified DNA inserts were ligated to the pBI-L or pTRE2hyg vector at molar ratios of 3:1 and 2:1 (DNA insert : Vector DNA). Ligation reactions were prepared using 3 l of 5 X ligase reaction buffer, 2 l vector DNA (50 ng), 1 l of T4 DNA ligase (5 U/l), required molar equivalent of DNA insert and made to a o total of 15 l with H2O. Reactions were incubated overnight at 14 C and transformed into One Shot® MAX Efficiency® DH5™ competent cells by heat shock. This involved the addition of the ligation reaction (5 l) to a 50 l vial of One Shot® cells, which was incubated on ice for 30 min and subsequently heat shocked for 30 sec at 42oC. The transformed cells were then incubated in 250 l of SOC medium (2% (w/v) Bacto Tryptone, 0.5% yeast extract, 10 mM NaCl, 2.5 mM KCl, 10 mM o MgCl2, 10 mM MgSO4 and 20 mM glucose) and shaken at 225 rpm for 1 hr at 37 C.

Chapter 3 51

The transformed cultures (100-200 l) were plated on 1.5% LB ampicillin agar plates and incubated overnight at 37oC. Single colonies were selected and inoculated into ~1 ml LB broth supplemented with ampicillin (100 g/ml) and incubated overnight at 37oC with shaking (225 rpm). The overnight cultures were minipreped and visualised by gel electrophoresis as detailed previously. Restriction mapping of the purified construct DNA was performed to identify clones containing the insert in the correct orientation.

The 1ARpTRE2hyg, 1ARpBI-L, 2ARpBI-L and 2ARpTRE2hyg clones were sequence verified by di-deoxy sequencing using the ABI BigDye Terminator kit at either the Australian Genome Research Facility (AGRF) or the Griffith University DNA Sequencing Facility (GUDSF). A sequence reaction consisted of 2 l BigDye terminator mix, 3.2 pmol of specific sequencing primer, 400-600 ng of construct DNA and was made up to 20 l with nuclease free H2O. Following an initial denaturation period of 96oC for 1 min, cycling conditions of 90oC for 10 sec, 50oC for 5 sec and 60oC for 4 min were repeated for 30 cycles. The sequencing products were precipitated in a 1.5 ml eppendorf tube containing 5 l of 125 mM EDTA, pH 8.0. The tubes were vortexed and quickly spun, then 60 l of 100% ethanol was added and the tubes were vortexed and respun. The samples were incubated at room temperature for 15 min and then centrifuged at 16,000 X g. The supernatant was removed and the pellet was washed with 250 l of 70% ethanol and subsequently centrifuged at 16,000 X g for 5 min. The supernatant was removed and the pellets were left to air dry.

Generating 1/2TMDVAR chimera recombinant clone

Using the donated 1pCDNA3.1 clone, the 1/2TMDVAR chimera was created by 3 successive rounds of multi site-directed mutagenesis (using the QuikChange™ multi site-directed mutagenesis kit following the manufacturer’s instructions) to change the 5 amino acids that are not conserved between the 1 and 2 TMDV (RVCAL to QIVVS, Figure 3.2). All reagents were provided in the kit unless otherwise stated. To obtain the desired substitution, primers were designed (using the Quikchange primer design program www.agilent.com/genomics/qcpd) which

52 Chapter 3

incorporated single base pair changes, which would then enable this change to be incorporated in the wild-type 1AR construct during the PCR process (Table 3.1).

Figure 3.2. The 1/2TMDVAR chimera was prepared by substitution of the 5 heterologous amino acids of 2AR TMDV into 1AR.

Table 3.1

Primers Used to Generate 1AR Mutations Primer Sequence (5’ to 3’)

1/2TMDV 1 GACTTCGTCACCAACCAGGCCTACGCCATCGCC

1/2TMDV 2 ATCGCCTCGTCCATAGTCTCCTTCTACGTGC

1/2TMDV 3 CCTTCTACGTGCCCCTGGTCATCATGGCCTTCGTGTACC

1/2TMDV 4 CCCTGGTCATCATGGTCTTCGTGTACCTGCG

1/2TMDV 5 ATCATGGTCTTCGTGTACTCGCGGGTGTTCC

R222Q GACTTCGTCACCAACCAGGCCTACGCCATCGCC

V230I GCCATCGCCTCGTCCATAGTCTCCTTCTACG

V230A CATCGCCTCGTCCGCAGTCTCCTTCTACG

S228A CTACGCCATCGCCGCGTCCGTAGTCTC

S229A CGCCATCGCCTCGGCCGTAGTCTCCTT

S232A CCTCGTCCGTAGTCGCCTTCTACGTGCCC *Bold underlined letters indicate changed nucleotides

Chapter 3 53

The PCRs were performed in a PTC-200 Peltier thermal cycler DNA Engine (Bresatec, South Australia). The PCRs (25 l) contained 2.5 l 10 X Quikchange Multi reaction buffer, 0.75 l Quikchange solution, 1 l dNTP mix, 0.8 l DNA template (100 ng), 1 l of specific mutagenesis primer (100 ng) (Table 3.1), 1 l

Quikchange multienzyme blend and 17.75 l H20. PCR amplification was performed using an initial denaturation period of 95oC for 1 min, followed by 30 cycles of 95oC for 1 min, 55oC for 1 min and 65oC for 14 min. The amplification products were treated with 1 l DpnI (10 U/ml) and incubated for 1 hr at 37oC to digest parental template.

The mutagenesis reactions were then transformed into XL-10 Gold ultra- competent cells by heat shock. For each reaction, 45 l of cells were aliquoted into a pre-chilled Falcon 2059 tube (Integrated Sciences, Australia), followed by addition of 2 l of -mercaptoethanol mix. Cells were incubated for 10 min on ice and gently swirled every 2 min. The DpnI treated DNA (4 l) was added to the cells which were gently swirled and then incubated for 30 min on ice. The cells were heat shocked in a 42oC water bath for 30 sec then incubated on ice for 2 min. The transformed cells were incubated in 500 l of pre-warmed (42oC) NZY+ broth (1% (w/v) casein hydrolysate, 0.5% (w/v) Bacto-yeast extract, 0.5% (w/v) NaCl, 12.5 mM MgCl2, o 12.5 mM MgSO4, 20 mM glucose, pH 7.5) and shaken at 250 rpm for 1 hr at 37 C. The transformation cultures (10-200 l) were plated on 1.5% LB ampicillin agar plates and incubated overnight at 37oC. Single colonies were selected and inoculated into 2 ml LB broth supplemented with ampicillin (100 g/ml) and incubated overnight at 37oC with shaking (225 rpm). The overnight cultures were minipreped and visualised by gel electrophoresis as detailed earlier. Following each round of site-directed mutagenesis, the nucleotide changes were confirmed by sequence analysis as described earlier. The chimeric gene was then subcloned from pCDNA3.1 into pBI-L and confirmed by sequence analysis as detailed above.

54 Chapter 3

Generating 1(V230I)AR, 1(V230A)AR, 1(R222Q)AR, 1(S228A)AR, 1(S229A)AR and 1(S232A)AR recombinant clones

The 1(V230I)AR, 1(V230A)AR, 1(R222Q)AR, 1(S228A)AR,

1(S229A)AR and 1(S232A)AR clones were created using site-directed mutagenesis as detailed previously. The first clone contained a change to the valine at amino acid position 230 into the corresponding isoleucine found in 2AR

(1(V230I)AR). The second clone contained a valine to alanine change at amino acid position 230 (1(V230A)AR), the third clone had a substitution of arginine to glutamine at amino acid position 222 (1(R2222Q)AR), whilst the conserved serines at positions 228, 229, and 232 were changed to alanine in the clones 1(S228A)AR,

1(S229A)AR and 1(S232A)AR. All changes were confirmed by di-deoxy sequencing as detailed earlier and the modified genes were then subcloned from pCDNA3.1 into pTRE2hyg as detailed above and confirmed by sequence analysis.

Cell Culture Cell maintenance All cell culture procedures were performed in a PS2 laminar flow hood under sterile conditions using sterile equipment. The CHOAA8 cell line (Chinese Hamster Ovary cell line that expresses the tetracycline-controlled transactivator) was used in transfections. Cells were grown on sterile 10 cm2 plates (Corning®, distributed by Sigma Aldrich, Castle Hill Australia) in growth medium [Alpha Modified minimal essential medium (MEM), supplemented with 10% heat inactivated tetracycline approved foetal calf serum (FCS), 1% penicillin/streptomycin and 1% L-Glutamine o or 1% GlutaMAX-I] at 37 C in an atmosphere containing 5% CO2. For CHOAA8 cells stably expressing receptors, G418 (100 g/ml) and hygromycin (400 g/ml) were added to the growth medium. Doxycycline (1 mg/ml) was only used during the selection of the stable transfectants to inhibit gene transcription.

CHOAA8 cells were removed from plates every 2 to 4 days (depending on cell growth) by the addition of 1.5 ml trypsin/EDTA (0.25% trypsin (w/v), 0.913 mM EDTA). Cells were grown until 70-95% confluency was reached. At this point, the media was discarded and cells washed twice with 5 ml MEM. The washed cells

Chapter 3 55

were detached by adding 1.5 ml of trypsin/EDTA followed by incubation for 4 min at 37oC. The reaction was halted by addition of 5 ml MEM containing 10% FCS. The cell suspension was pelleted by centrifugation for 5 min at 100 X g. The supernatant was removed and the cell pellet resuspended in fresh media.

Cell counting Cells were trypsinised and resuspended in fresh growth medium. The cell suspension was diluted (1:1) in PBS and 10 l transferred to a haemocytometer. The suspension dispersed by capillary action between the haemocytometer and the cover slip. Counting was performed using established protocols (Freshney, 1994).

Transient transfection of CHOAA8 cells Prior to transfection, suitably purified DNA was prepared using the Qiagen Plasmid Purification Kit following the manufacturer’s instructions. DNA concentrations were determined by spectrophotometric analysis at λ=260 nm. CHOAA8 cells for transfection were grown in antibiotic free growth medium. Transfection into CHOAA8 cells was performed in 6-well plates, with cells seeded at a density of 1x105 cells/well in Dulbecco’s modified Eagle’s medium (DMEM) and cultured overnight to achieve ~50% confluence.

Before transfection, the FuGene 6 transfection reagent was warmed to room temperature. The transfection mix was prepared in sterile 1.5 ml eppendorf tubes using DMEM, Fugene 6 transfection reagent and purified construct DNA in a total volume of 100 l. The ratios of reagent (l) : DNA (g) used were 3:1, 3:2 and 6:1. Initially, the required volume of Fugene 6 transfection reagent was added to each tube containing DMEM and incubated for 5 min at room temperature. The construct DNA was then added to each tube to form the transfection complex. The complex was incubated for 60 min at room temperature. The cells were removed from the incubator and each complex was added to the designated well in the plate, then the cells were returned to the incubator. At 24 and 48 hrs post transfection, cells were trypsinised by adding 0.5 ml of trypsin/EDTA followed by incubation for 2.5 min at 37oC. Cells from each well were pooled and pelleted at 100 X g for 5 min and

56 Chapter 3

resuspended in 1 ml of ice cold Tris-Incubation buffer. Cells were stored at -80oC and subsequently analysed by radioligand binding.

Generation of CHOAA8 cells stably expressing human 1AR, 2AR or mutant ARs To generate stable transfectants, CHOAA8 cell transfection was performed using the aforementioned protocol. At 48 hrs post transfection, each well was washed twice with 2 ml of MEM. The cells were lifted by adding 0.5 ml of trypsin/EDTA followed by incubation for 2.5 min at 37oC. The reaction was then halted with 2 ml of MEM containing 10% FCS. The cells from each well were pelleted by centrifugation at 100 X g for 5 min. The supernatant was removed and cell pellets were resuspended in 1 ml of growth medium containing G418 (100 g/ml), doxycycline (1 mg/ml) and hygromycin (400 g/ml). The 10 cm2 plates were seeded with 500 l, 100 l, 50 l or 10 l of the appropriate cell suspension and 10 ml of fresh growth medium was added to each plate. Different volumes of cell suspension were used to ensure good separation of colonies. A media change was performed every 3 days and doxycycline was replenished every 2 days.

Stable hygromycin-resistant colonies formed over a 2-4 week period. Between 30 and 80 stable cell clones were selected from each transfection. Microscopy was used to identify well spaced single colonies. These colonies were selected using 3 mm cloning discs soaked in trypsin/EDTA. A disc was placed over a single colony and incubated for 10 min at 37oC. Each disc was then removed and placed into a single well of a 24 well plate containing 1 ml of growth medium supplemented with G418 (100 g/ml), doxycycline (1 mg/ml) and hygromycin (400 g/ml). Colony collection was confirmed by microscopy. The 24 well plate containing selected clones was cultured at 37oC. The cells received a media change every 3 days and fresh doxycycline every 2 days.

When the transfected cells had reached ~80% confluency they were trypsinised, transferred to 10 cm2 plates containing 10 ml of growth medium supplemented with G418 (100 g/ml), doxycycline (1 mg/ml) and hygromycin (400

Chapter 3 57

g/ml), cultured at 37oC and allowed to grow to ~80% confluency. The cells were then trypsinised and either resuspended in 1 ml of cell freezing media-DMSO 1X, o stored overnight at -80 C in a cyrovessel and then transferred to liquid N2 for storage, or passaged onto new 10 cm2 plates containing fresh growth media, supplemented with G418 (100 g/ml) and hygromycin (400 g/ml) at 37oC. Doxycycline, which inhibits gene transcription in the pTRE2hyg Tet-off expression system was discontinued at this point. Transient and stable transfections of 1(V230I)AR,

1(V230A)AR, 1(R222Q)AR were performed by myself, while 1AR,

2AR,1(S228A)AR, 1(S229A)AR and 1(S232A)AR transfections were performed by Kelly Chee and Dr Annalese Semmler.

3.2.3 Radioligand Binding Experiments Cell membrane preparation Cells were trypsinised by the addition of 1.5 ml trypsin/EDTA when 70-95% confluency was reached. The resulting cell pellets were resuspended in 3 ml of ice cold Tris-Incubation buffer. The cell suspension was pelleted by centrifugation for 10 min at 200 X g at 4oC. The supernatant was removed and the cells were resuspended in fresh buffer and re-centrifuged. The supernatant was again removed and each pellet was resuspended in 1.5 ml of Tris-Incubation buffer, then spilt into 2 X 750 l aliquots and stored at -80oC.

The cells were disrupted using a Kinematica Polytron PT 2100 (7 mm probe, 1 x 3 sec, speed setting 10). Cell membranes were than isolated by centrifugation for 15 min at 51,500 X g at 4°C. Fresh Tris-incubation buffer (3 ml) was added to membrane pellets which were than homogenised by Polytron (settings as above). The volume of cell suspension was made up to between 7-20 ml depending on the type of binding assay.

Single point binding Single point binding was used as a screening tool to estimate the receptor density of clones from transfection. Each homogenate was incubated for 2 hr at 37°C

58 Chapter 3

in a final volume of 0.25 ml containing a final concentration of 2 nM 3 (-)-[ H]-CGP 12177 (specific activity 30-60 Ci/mmol) for 1AR (pBI-L and pTRE2hyg), 2AR (pBI-L and pTRE2hyg), 1/2TMDVAR, 1(V230I)AR,

1(V230A)AR, 1(R222Q)AR and 1(S232A)AR, 10 nM for 1(S229A)AR and

70 nM for 1(S228A)AR. Non-specific binding was measured in the presence of 200 μM (-)-isoprenaline. Bound radioligand was isolated by filtration through Whatman GF/B paper and radioactivity counted in 10 ml Ultima Gold™ scintillation cocktail (Perkin Elmer, Melbourne, Australia) using a Tri-Carb 2800TR scintillation counter (Perkin Elmer, Melbourne, Australia).

Saturation binding To identify cell lines expressing receptors to approximate densities observed for 1AR and 2AR in human heart (~70 and ~30 fmol/mg protein respectively, reviewed in Brodde, 1991), saturation binding experiments were carried out with 3 0.02 nM – 10 nM (-)-[ H]-CGP 12177 for cell lines 1AR (pBI-L and pTRE2hyg),

2AR (pBI-L and pTRE2hyg), 1/2TMDVAR, 1(V230I)AR, 1(V230A)AR,

1(R222Q)AR and 1(S232A)AR. The concentration ranges for 1(S228A)AR and 3 1(S229A)AR were 0.5 nM – 80 nM and 0.1 nM – 40 nM (-)-[ H]-CGP 12177 respectively. All saturation experiments were performed in the presence of 0.1 mM GTP (Roche Diagnostics, Castle Hill, Australia). Non-specific binding was determined in the presence of 200 μM (-)-isoprenaline. Bound radioligand was isolated and radioactivity counted as detailed above.

Competition binding Competition binding experiments between known AR ligands and (-)-[3H]-CGP 12177 were performed to generate a pharmacological profile of CHOAA8 cell membranes expressing wild-type or mutant ARs. Cell membranes containing 1AR (pBI-L), 2AR (pBI-L), 1/2TMDVAR, 1(V230I)AR, 3 1(V230A)AR, 1(R222Q)AR were labelled with 0.5 nM (-)-[ H]-CGP 12177, whilst 1(S228A)AR, 1(S229A)AR and 1(S232A)AR were labelled with 10 nM, 2.5 nM and 1 nM respectively, in the absence and presence of the competing agents

Chapter 3 59

ICI 118, 551 (0.1 nM-100 M), CGP 20712A (1 aM-100 M). Competition binding experiments using (-)-bupranolol (10 pM-10 M) or bisoprolol (0.2 nM-100 M) were carried out for 1AR, 2AR and 1/2TMDVAR. Bound radioligand was isolated and radioactivity counted as above. Assays were carried out in triplicate and replicated at least 4 times. Results are presented as mean ± s.e. mean.

The amount of protein was determined for all samples used in radioligand binding experiments using the method of Bradford (Bradford, 1976).

Analysis of single point binding data In order to screen large numbers of transfection clones, a single concentration of radioligand was used to estimate the receptor density using the equation below:

Where:

X = the concentration of radioligand (nM)

Y = specific binding, disintegrations per minute (dpm)

*Bmax = the asymptotic value of Y, representing the maximum number of binding sites (dpm)

KD = the equilibrium dissociation constant representing the radioligand concentration at which half of the binding sites are occupied (nM)

3 *The KD of (-)-[ H]-CGP 12177 for 1AR (10.15 ± 0.04, n =11) was used to estimate the Bmax at mutant 1ARs. Bmax (dpm) was converted to mol/mg protein following protein determination.

Analysis of saturation binding data All saturation binding curves were analysed using nonlinear regression to determine the equilibrium dissociation constant (KD) of the radioligand and Bmax. Iterative curve fitting was performed with GraphPad Prism 5 (Graphpad Software,

60 Chapter 3

San Diego, CA) using a one-site binding, nonlinear regression model (one-site binding hyperbola) represented by the equation:

Where:

X = the concentration of radioligand (nM)

Y = specific binding, disintegrations per minute (dpm)

Bmax = the asymptotic value of Y, representing the maximum number of binding sites (dpm)

KD = the equilibrium dissociation constant, representing the radioligand concentration at which half of the binding sites are occupied (M)

Analysis of competition binding data Competition binding curves were analysed using nonlinear regression to determine the concentration of ligand required to compete for half of the specific radioligand binding (Ki). The iterative curve fitting was performed with Prism 5 (Graphpad) using a one-site binding, nonlinear regression model (one-site competition) represented by the equation:

110

Where:

X = logarithm of the drug concentration

Y = specific binding (dpm)

Top = value of Y for the maximal curve asymptote, representing no competition with the radioligand

Bottom = value of Y for the minimal curve asymptote, representing the maximum competition with the radioligand

IC50 = concentration of ligand required to compete for half of the specific radioligand binding

Chapter 3 61

The presence of two binding sites was observed for some competition binding curves. For these curves, iterative curve fitting was performed with Prism 5 (Graphpad) using a two-site binding, nonlinear regression model (two-site competition) represented by the equation:

1 1 1 1101 1102

Where:

X = logarithm of the drug concentration

Y = specific binding (dpm)

Top = value of Y for the maximal curve asymptote, representing no competition with the radioligand

Bottom = value of Y for the minimal curve asymptote, representing the maximum competition with the radioligand

Fraction 1 = the fraction of binding sites that have affinity described by logIC501.

The remaining binding sites have affinity described by logIC502

For this equation the assumption was made that (-)-[3H]-CGP 12177 has the same affinity for both sites.

The IC50 values were used to determine Ki values using the Cheng & Prusoff equation (Cheng et al., 1973). The –log Ki (pKi) was used as a measure of ligand affinity for the receptor.

IC K Ligand 1 K

3.2.4 Cyclic AMP Enzyme Immunoassays Stably transfected CHOAA8 cells were assayed as described in Kaumann et al., 2007. Briefly, wild-type, chimeric or single amino acid altered cell lines were trypsinised and counted as described previously. Cells were seeded at a density

62 Chapter 3

between 4x105 and 8x105/well in sterile 96-well flat bottom tissue culture plates (Sarstedt, Technology Park, Australia), containing MEM supplemented with 10% FCS, 1% penicillin/streptomycin, 1% L-glutamine, G418 (100 g/ml), and hygromycin (400 g/ml) and grown overnight to ~80% confluency at 37oC. Cyclic AMP activity was measured using a modification of the method described by Joseph et al., (2004b). Briefly, cells were washed with warm (37°C) PBS and pre-incubated in DMEM in the presence and absence of 10 nM or 100 nM (-)-bupranolol for 20 min at 37°C, 5% CO2. The media was then changed to assay media [DMEM containing 25 mM N-2-hydroxyethylpiperazine-N′-2- ethanesulfonic acid, 1 mM 3- isobutyl-1-methylxanthine (IBMX) and 0.2 mM ascorbic acid in the presence or absence of 10 nM or 100 nM (-)-bupranolol] and cells were incubated for a further 30 min at 37°C. The cells were then exposed to varying concentrations of the agonist

(-)-isoprenaline (1HAR, 3 pM – 300 M) or the non-conventional partial agonist

(-)-CGP 12177 (1LAR, 3 pM – 300 M) made up in assay media (without

(-)-bupranolol) and incubated for 20 min at 37°C, 5% CO2. Three wells on each plate were treated with forskolin alone.

After agonist treatment, the assay media was removed by aspiration and cells were lysed by addition of lysis buffer 1B (200 l/well). Cells were agitated for 10 min after the addition of lysis buffer 1B. The 96-well cAMP EIA assay plate coated with donkey anti-rabbit IgG was setup on ice as illustrated in Figure 3.3. Non- specific binding (NSB) wells contained 100 l of lysis buffer 1B and 100 l lysis buffer 2B. The zero point (0) of the standard curve contained 100 l of lysis buffer 1B, whilst 100 l of each standard was added to the appropriate wells to generate the standard curve. For the sample wells, 50 l of lysed cells were added to the appropriate well and the final well volume was made up to 100 l by addition of lysis buffer 1B. This dilution was required to obtain cAMP levels that fitted within the standard curve. After all the samples were transferred, 100 l of Rabbit anti- cAMP antiserum was added to all wells except the blank and NSB. The plate was then incubated for 2 hrs at 4°C on ice. Cyclic AMP-horseradish peroxidase conjugate (50 l) was then added to all wells except the blank. The plate was incubated for a further 60 min at 4°C on ice. The contents of all wells were then removed and each

Chapter 3 63

well was washed four times with 400 l of wash buffer. Finally, 150 l of TMB substrate (3,3’,5,5’-tetramethylbenzidine/hydrogen peroxide) was added to all wells and the plate was covered with aluminium foil and mixed on a microplate shaker for 30 mins at room temperature. Addition of the TMB substrate is the final step of the EIA reaction (Figure 3.4), resulting in the development of a blue colour which is proportional to the amount of cAMP in each well. This can be quantified by reading the optical density of the blue colour at 630 nm in a microplate reader. The reaction was halted by the addition of 100 l of 1 M sulphuric acid into each well and the optical density at 450 nm determined using a plate reader. Assays were carried out in triplicate and replicated at least 5 times to give stable quantitative readings for analysis. Results are presented as mean ± s.e. mean.

Figure 3.3. The cAMP EIA assay plate illustrating the position of the substrate blank (B), non-specific binding (NSB), standard curve (0 – 3200 fmol, cAMP), sample (S) and forskolin (F).

64 Chapter 3

Figure 3.4. The cAMP EIA reaction involving the Donkey anti-rabbit IgG coated plate, the Rabbit anti-cAMP antiserum and competition for binding to the cAMP antibody between cAMP and the cAMP peroxidase conjugate. Addition of the TMB substrate results of the formation of a blue colour which can be quantified by measuring the optical density at 630 nm.

Cyclic AMP experiments were also performed using the compound 5-[3- (tert-butylamino))2-hydroxypropoxy]1,3-dihydro-2H-benzimidazol-2-one (structural isomer of (-)-CGP 12177, synthesised by Dr Robert Reid, IMB, University of Queensland). A concentration-effect relationship (200 pM – 60 mM) was generated to determine the potency of the compound for increases in cAMP using CHOAA8 cells expressing 1AR. Cells were prepared as above and exposed to 5-[3-(tert- butylamino))2-hydroxypropoxy]1,3-dihydro-2H-benzimidazol-2-one for 20 min at

37°C, 5% CO2. In other experiments, cells were pre-incubated for 20 min at 37°C,

5% CO2 in DMEM in the presence of 1 M of 5-[3-(tert-butylamino))2- hydroxypropoxy]1,3-dihydro-2H-benzimidazol-2-one. The cells were then exposed to varying concentrations of (-)-isoprenaline (1HAR) or (-)-CGP 12177 (1LAR) for

20 min at 37°C, 5% CO2, to determine antagonist affinity. Cells from all experiments were lysed and assayed as detailed above.

Analysis of cAMP data The amount of cAMP was determined by calculating the average optical density (OD) for each set of replicate wells. The percent of cAMP bound (%B/B0) for each standard and sample was determined using the following relationship:

Chapter 3 65

– %/ 100

A standard curve was generated by plotting %B/B0 (y-axis) against fmol cAMP standard per well (x-axis).

The cAMP dose response curves were analysed using nonlinear regression to determine the concentration of ligand required to produce half maximal effect

(EC50). Iterative curve fitting was performed with Prism 5 (Graphpad) using a nonlinear regression model (log(agonist) vs. response – variable slope) represented by the equation:

1 10∗

Where:

X = logarithm of the drug concentration

Y = amount of cAMP (pmol/mg protein)

Bottom = value of Y for the minimal curve asymptote, representing basal cAMP

Top = value of Y for the maximal curve asymptote, representing maximum agonist response

LogEC50 = logarithm of the concentration of agonist required to produce half maximal response

For cAMP experiments performed in the absence and presence of (-)-bupranolol, the pKB of (-)-bupranolol was determined using the Gaddum equation (Gaddum, 1957).

log 1 log

The concentration ratio 1 /

Where KB is the equilibrium dissociation constant of (-)-bupranolol for 1HAR or

1LAR.

66 Chapter 3

Statistical comparsion of two means was performed with Student’s t-tests, while comparsion of more than two means was performed with 1-way ANOVA using Prism 5 (Instat, Graphpad).

3.2.5 Tissue Bath Experiments Contractility studies The experiments providing data for Figures 4.1 to 4.3, 4.6, 4.7 and 4.8 (see section 4.2) were performed at The Prince Charles Hospital, Brisbane, using right atrial appendages from patients undergoing coronary artery bypass graft surgery. After excision, the appendages were immediately placed in modified ice-cold Krebs buffer equilibrated with 95% O2/5% CO2. Trabeculae were dissected and set up, on occasion, in pairs to contract at 1 Hz in an apparatus with a 50 ml organ bath in Krebs buffer supplemented with 15 mM Na+, 5 mM fumarate, 5 mM pyruvate, 5 mM L-glutamate, 10 mM glucose at 37°C. The tissues were attached to Swema SG4-45 strain gauge transducers (Swema, Stockholm, Sweden) and force recorded on a Watanabe polygraph (Graphtec Corporation, Yokohama, Japan). The tissues were driven with square-wave pulses of 5 ms duration and just over threshold voltage. After determination of a length–tension curve, the length of each trabeculum was set to obtain 50% of the resting tension associated with maximum developed force.

Contractility experiments were performed using the conditions of Skeberdis et a1., (2008). Experiments were carried out in the presence of nadolol (200 nM), a concentration expected to block both 1HAR and 2AR but hardly 3AR [pKD = 6.2,

(Baker, 2005a)] or 1LAR [pKB = 6.2, (Joseph et al., 2004b)], and IBMX (10 M) to boost the response to (-)-CGP 12177 (Kaumann et al., 2007; Kaumann et al., 1997c).

Experiments providing data for Figures 4.4, 4.5, 4.12 and 4.13 (see section 4.2) were performed by collaborators, Dr Torsten Christ and Prof. Alberto Kaumann at the Dresden University of Technology and were carried out in Tyrode’s solution + + 2+ 2+ − 2- 4- (mM): Na 171.12, K 5.4, Ca 1.8, Mg 1.05, Cl 137.8, HCO3 22, HPO 0.42,

EDTA 0.04, and equilibrated with 95% O2/5% CO2.

Chapter 3 67

Cumulative concentration-effect curves Cumulative concentration-effect curves for (-)-CGP 12177, BRL 37344 and SR 58611 were carried out in the absence and presence of L-748,337 (1 μM). L-748,337 was pre-incubated 15 min after the addition of nadolol (200 nM) to the tissues, followed by the administration of IBMX (10 M) 30 min later, and finally the curve for an agonist begun 15 min later. At least 2 trabeculae from each patient were used and curves for (-)-CGP 12177 in the absence and presence of L-748,337 were time-matched. In other experiments concentration-effect curves for BRL 37344 were carried out in the absence or presence of nadolol (200 nM), ICI 118,551 (50 nM) or CGP 20712A (300 nM) or the combination of ICI 118,551 and CGP 20712A, in the presence of IBMX (10 μM).

Kinetic experiments Kinetic experiments were used to investigate the influence of L-748,337 on the contractile responses to (-)-CGP 12177. When 4 or more atrial trabeculae were available, a time-matched procedure was used comprising 4 experimental groups (as shown in the representative experiment, see Section 4.2.4, Figure 4.6). Trabeculae were incubated with IBMX (10 M) which remained in contact with the trabeculae for the remainder of the experiment. By the 30th min, 3 groups of trabeculae were incubated with (-)-CGP 12177 (200 nM). Twenty minutes later, upon establishment of the cardiostimulant effect of (-)-CGP 12177 one group received L-748,337 (1 M), another received (-)-bupranolol (1 M), while the third group was used as a time-matched control to establish the time-course of the cardiostimulant effect of (-)-CGP 12177.

All tissue bath experiments were concluded by the administration of a receptor- saturating concentration of (-)-isoprenaline (200 μM) and, after an equilibrium response to (-)-isoprenaline was established, by raising the Ca2+ concentration to 9.25 mM (total bath concentration).

68 Chapter 3

Measurements of ICa-L Experiments of Figures 4.9 to 4.11 were performed at Dresden University of Technology, Dresden by Dr Torsten Christ and Prof. Alberto Kaumann (Christ et al., 2011, Appendix B). Human atrial myocytes were enzymatically dissociated as described (Christ et al., 2001). Myocytes were stored at room temperature until use in a solution containing (mM): K+ 130, Mg2+ 3, Cl- 46, taurine 20, glutamic acid 50, EGTA 0.5, HEPES 10 and glucose 10, pH 7.4. The single electrode patch clamp + technique was used to measure ICa-L at 37°C. Holding potential was -80 mV. K currents were blocked by replacing K+ with Cs+. The external perfusing solution contained (mM): tetraethylammonium 120, Cs+ 10, Ca2+ 2, Mg2+ 1, Cl- 136, HEPES 10 and glucose 10 with pH adjusted with CsOH. The pipette solution contained (mM): Cs+ 110, Ca2+ 3, Mg2+ 4, Cl- 26, methanesulphonate- 90, HEPES 10, ATP 4, Tris-GTP 0.4 and EGTA 10 with a calculated free Ca2+ concentration of 60 nM (EQCAL, Biosoft, Cambridge, UK) and pH 7.2, adjusted using CsOH. Current amplitude was determined as the difference between peak inward current and current at the end of the 200 ms depolarizing step to +10 mV from a holding potential of -80 mV. Myocytes were exposed only once to the different agonists in the absence or presence of IBMX (10 µM), nadolol (200 nM), L-748,337 (1 µM), ICI 118,551 (50 nM) or (-)-bupranolol (10 µM).

Analysis of contractility experiments

Results from kinetic experiments are expressed as mN force. The –logEC50M values for BRL 37344 and (-)-CGP 12177 were calculated from the corresponding concentration effect curves. Data comparisons were made with ANOVA and non- paired or paired t-tests as appropriate using Prism 5 (Graphpad). Expected concentration ratios (CR) of BRL 37344 in the presence and absence of an antagonist concentration [(B)] were calculated using the equation:

CR 1/

Where KB is the equilibrium dissociation constant.

Results are presented as mean ± s.e. mean, where n values refer to the number to trabeculae.

Chapter 3 69

Chapter 4: Human Atrial 1L-Adrenoceptor but not 3-Adrenoceptor Activation Increases Force and Ca2+ Current at Physiological Temperature

4.1 INTRODUCTION

There is evidence for the expression of 3-adrenoceptors (3AR) in human heart (Moniotte et al., 2001a; Moniotte et al., 2001b) but the relevance to contractile function is controversial. Agonists for 3AR have been reported to cause cardiodepressant effects on human ventricle obtained from endomyocardial biopsies (Gauthier et al., 1996; Rozec et al., 2006), possibly by release of nitric oxide from endothelial and/or endocardial cells. In contrast, 3AR agonists did not produce cardiostimulant nor depressant effects in human ventricular trabeculae (Kaumann et al., 2008; Molenaar et al., 1997). However, it was recently reported that the negative inotropic effect of BRL 37344 was prevented by the NO antagonist L-NMA (Napp et al., 2009). Previous reports in human atrium show that 3AR agonists were devoid of cardiostimulant or cardiodepressant effects in the presence of (-)-propranolol

(200 nM) (Kaumann et al., 1997c), a concentration that blocks human atrial β1AR and 2AR (Gille et al., 1985), but is not expected to block 3AR (Cohen et al., 1999;

Hoffmann et al., 2004). Modest cardiostimulant effects of additional 3AR agonists (Sennitt et al., 1998), including BRL 37344 (Arch et al., 1993; Pott et al., 2003), were antagonised by (-)-propranolol (200-300 nM), consistent with mediation through 1AR and 2AR (Gille et al., 1985) but not 3AR (Cohen et al., 1999; Hoffmann et al., 2004). Taken together, published information is inconsistent with a cardiostimulant or cardiodepressant role of human atrial 3AR.

Some -blockers with high-affinity for both 1AR and 2AR cause cardiostimulation at concentrations considerably greater than those that significantly antagonise the effects of catecholamines. Their agonist effects are smaller and more resistant to antagonism by -blockers [e.g. propranolol, nadolol, (Kaumann et al.,

Chapter 4 71

2008)] than the effects of catecholamines. Evidence from cardiac and recombinant receptors suggests that these non-conventional partial agonists induce their agonist effects through a 1AR site (1LAR) for which they have lower affinity than for the site (1HAR) through which they antagonise the effects of catecholamines (reviewed Kaumann et al., 2008). The hydrophilic compound CGP 12177, introduced as a high- affinity AR radioligand by Staehelin et al., (1983) and soon thereafter discovered to exert cardiostimulant properties (Kaumann, 1983), has since been extensively used as an experimental non-conventional partial agonist (Kaumann et al., 2008). The cardiostimulant effects of (-)-CGP 12177, mediated through 1LAR, are antagonised by (-)-bupranolol (Kaumann et al., 2008). Although CGP 12177 does bind to 3AR (Cohen et al., 1999; Hoffmann et al., 2004), it is only a weak lipolytic partial agonist in murine adipocytes (Preitner et al., 1998) through 3AR but hardly in human adipocytes (Tavernier et al., 1996), and there is no evidence that its cardiostimulant effects are mediated through 3AR (Cohen et al., 1999; Kaumann et al., 2008).

However, when human 3AR are overexpressed into murine heart they mediate cardiostimulation (Kohout et al., 2001).

Surprisingly, Skeberdis et al., (2008) recently postulated that 3ARs mediate increases in human atrial contractility. They demonstrated in human atrial myocytes at room temperature that the agonists SR 58611A, BRL 37344 and CGP 12177 2+ increase L-type Ca current (ICa-L) and that these responses were reversed with the

3AR-selective antagonist L-748,337 (Candelore et al., 1999). They reported agonist-evoked increases in contractility of atrial tissues at 37ºC. The modest human atrial responses to CGP 12177 (Skeberdis et al., 2008) were enhanced by the non- selective phosphodiesterase (PDE) inhibitor 3-isobutyl-1-methylxanthine (IBMX) (Kaumann et al., 1997c) and PDE3 inhibitor cilostamide (Kaumann et al., 2007). Skeberdis et al., (2008) also observed marked increases of the contractile responses to CGP 12177 and BRL 37344 in the presence of IBMX in human atrium at 37°C. Importantly, however, they did not investigate whether these effects could be antagonised with L-748,337, as one would expect from an interaction through 3AR.

72 Chapter 4

Activation of β3AR has been proposed as a therapeutic target for the treatment of overactive bladder disorder (Yamaguchi, 2002), obesity (Sawa et al., 2006), type II diabetes mellitus (Sawa et al., 2006), disorders of anxiety and depression

(Stemmelin et al., 2008). However, if human atrial 3ARs mediate cardiostimulation, this could be associated with tachycardia and potentially deleterious arrhythmias, including atrial fibrillation, thus limiting or even excluding the use of 3AR agonists in humans. Further research is therefore needed to confirm whether or not 3ARs mediate increases in human atrial contractility.

The interpretation by Skeberdis et al., (2008) also challenges the previously reported working hypothesis of this research project, that the positive inotropic effects of (-)-CGP 12177 and other non-conventional partial agonists on human atrial myocardium are mediated through 1LAR (Joseph et al., 2003; Kaumann et al., 2008; Sarsero et al., 2003). Previous studies have found that in human atrial trabeculae, L- 748,337 failed to antagonise the positive inotropic effects of the non-conventional partial agonist (-)-pindolol, mediated through 1LAR, as confirmed with recombinant

1AR (Joseph et al., 2003). The aim of this study was to therefore investigate whether L-748,337 could antagonise the inotropic effects of the β3AR agonists SR 58611, BRL 37344 and (-)-CGP 12177 on human atrial preparations in the presence of IBMX at 37°C. However, inotropic effects with SR 58611 were not observed. To investigate whether different receptors could be activated by BRL 37344, SR 58611 and (-)-CGP 12177 as a function of temperature, experiments were performed to compare their ICa-L responses at 24°C and 37°C. To investigate whether the ICa-L responses to the agonists at 24°C couple to atrial contractility, force experiments were also carried out at 24°C with SR 58611 and (-)-CGP 12177.

Evidence from this study indicates that activation of human atrial 3AR does not affect contractility. BRL 37344 increases force through β1AR/2AR. A 3AR- mediated increase in ICa-L by BRL 37344 and SR 58611 is restricted to low non- physiological temperatures. (-)-CGP 12177 increases force and ICa-L through 1LAR.

Chapter 4 73

4.2 RESULTS

4.2.1 Patients Right atrial appendages were obtained from patients undergoing coronary artery bypass graft surgery at The Prince Charles Hospital, Brisbane (The Prince Charles Hospital Ethics Committee EC27133; QUT human ethics committee 0800000066) and Dresden University of Technology (Medical Faculty Ethics committee document EK790799), who had provided written informed consent. Patient demographics are outlined in Table 4.1.

Table 4.1 Patient Demographics n 123 Gender, m/f 89/34 Age, yr 68±1 BMI, kg/m2 27±1 CAD, n 74 AVD/MVD, n 29 CAD+AVD/MVD, n 20 Hypertension, n 95 Diabetes, n 32 Hyperlipidemia, n 60 LVEF, % 52±2 Cardiovascular medication, n Digitalis 4

ACE inhibitors/AT1-blockers 72 -blockers 92 Metoprolol 62 Bisoprolol 25 Carvedilol 4 Nebivolol 1 Calcium channel blockers 25 Diuretics 13 Nitrates 22 Lipid-lowering drugs 55

CAD = coronary artery disease; AVD = aortic valve disease; MVD = mitral valve disease; LVEF = left ventricular ejection fraction; ACE = angiotensin-converting enzyme; AT = Angiotensin receptor.

74 Chapter 4

4.2.2 Antagonism of the Inotropic Effects of BRL 37344 by the 1AR/2AR Antagonist Nadolol, 2AR Antagonist ICI 118,551 but not 3AR Antagonist L-748,337 in Atrial Trabeculae. The experiments were carried out in the presence of the non-selective PDE inhibitor IBMX to reduce cAMP hydrolysis and enhance inotropic responses. BRL 37344 caused concentration-dependent increases in contractile force (Figures

4.1 and 4.2). The 1AR and 2AR antagonist nadolol (200 nM) and 2AR antagonist ICI 118,551 (50 nM) produced a rightward shift of the curve for BRL 37344 (Figures

4.1 and 4.2) by 1.7 log units and 0.9 log units (Table 4.2) respectively. The 3AR antagonist L-748,337 (1 M) caused a 0.3 log unit shift of the curve for BRL 37344 (Figure 4.1 and Table 4.2). L-748,337 in the presence of nadolol did not cause further antagonism of the effects of BRL 37344 (Figure 4.1). The 1AR antagonist CGP 20712A tended to cause a rightward shift of the curve for BRL 37344 (Figure 4.2) but it did not reach statistical significance (Table 4.2). Concurrent ICI 118,551 and CGP 20712A caused a greater shift of the curve of BRL 37344 than ICI 118,551 alone (Figure 4.2, Table 4.2). At 24°C BRL 37344 did not increase contractile force (Figure 4.3). These results indicate that the effects of BRL 37344 are mediated through 1AR and 2AR but not 3AR.

Figure 4.1. Mediation of the positive inotropic effects of BRL 37344 through 1-and 2ARs but not 3ARs in human atrial trabeculae. Cumulative concentration-effect curves were carried out in the presence of IBMX (10 mM) for BRL 37344 in the absence and presence of nadolol (200 nM, 1HAR/2AR antagonist), L-748,337 (1 M, 3AR antagonist) or nadolol + L-748,337. (-)-Isoprenaline (200 M), administered after a steady state response to 200 M BRL 37344, increased force (% over IBMX) by 284 ± 42.7%. Nadolol was incubated for 15 min, followed by the additional administration of L-748,337 for 30 min, followed by the administration of IBMX (10 mM) for 15 min before a curve was begun. Numbers between parentheses represent trabeculae/patient. ISO, (-)-isoprenaline.

Chapter 4 75

Figure 4.2. Mediation of the positive inotropic effects of BRL 37344 through 1-and 2ARs in human atrial trabeculae. Curves for BRL 37344 were also determined in the absence and presence of ICI 118,551 (50 nM, 2AR-selective antagonist, ICI) or CGP 20712A (300 nM, 1HAR-selective antagonist) or co-administration of ICI 118,551 + CGP 20712A. (-)-Isoprenaline (200 M), administered after a steady state response to 60 M BRL 37344, increased force (% over IBMX) by 312 ± 63.5%. Antagonists were incubated for at least 45 min before a curve for an agonist was started. IBMX (10 mM) was administered 15 min before a curve was begun. Numbers between parentheses represent trabeculae/patient. ISO, (-)-isoprenaline.

Figure 4.3. Lack of inotropic effects of BRL 37344 at 24°C. At 24°C (-)-isoprenaline and Ca2+ (8 mM) increased force (mN) from a basal of 5.2 ± 0.9 to 10.8 ± 1.5 and 13.3 ± 2.1, respectively (n = 28/4). At 37°C (-)-isoprenaline and Ca2+ increased force (mN) from a basal level of 7.6 ± 0.6 to 12.7 ± 0.7 and 13.5 ± 0.7, respectively (n=86/17).

76 Chapter 4

Table 4.2 Effects of Antagonists on the Inotropic Potency of BRL 37344 in the Presence of IBMX (10 M)

n -LogEC50 (M) Log CR Observed Expected Control (no antagonist) 32/17 6.82 ± 0.11 - –

a CGP 20712A (300 nM) 5/4 6.44 ± 0.10 0.38 2.28 (1H)

a ICI 118,551 (50 nM) 4/4 5.96 ± 0.18** 0.86 1.72 (2)

CGP 20712A + 5/4 4.61 ± 0.28** 2.21 – (1H + 2)

ICI 118,551

b L-748,337 (1 M) 22/8 6.49 ± 0.11** 0.33 0.55 (1H)

b 0.77 (2)

c  1.67 (3)

b 2.55 (3)

d Nadolol (200 nM) 8/6 5.13 ± 0.10*** 1.69 1.02 (1H)

e 0.64 (1H)

e  1.91 (2)

Nadolol + L-748,337 10/6 5.10 ± 0.09*** 1.72 – n = Trabeculae/patients CR, concentration ratios ** P < 0.01, *** P < 0.001 compared with control a Kaumann et al., 1987 b Candelore et al., 1999 c Wuest et al., 2009 d Joseph et al., 2004b e Baker, 2005a

Chapter 4 77

4.2.3 The 3AR Agonist SR 58611 does not Increase Contractile Force.

Inotropic results with SR 58611 are shown in Figures 4.4 and 4.5. The 3AR agonist SR 58611 (1 nM-10 M) did not produce positive inotropic effects in the absence or presence of nadolol or L-748,337 at 37°C and 24°C (Figures 4.4 and 4.5) High SR 58611 concentrations tended to reduce force in the presence of nadolol but the effects of the highest concentration used (10-5 M) were not significant (P = 0.14 IBMX vs. nadolol; P = 0.12 IBMX + L748337 vs. IBMX + nadolol + L748337).

Figure 4.4. Lack of positive inotropic effects of SR 58611 at 37 ºC in human atrial trabeculae. Cumulative concentration-effect curves were carried out for SR 58611 in the absence and presence of nadolol (200 nM, β1HAR/β2AR antagonist) or L-748,337 (1 μM, β3AR antagonist) or nadolol + L-748,337, each curve in the presence of IBMX (10 μM). (-)-Isoprenaline (200 µM), administered after a steady state response to 10-5 M of SR 58611, increased force (% of IBMX) by 162 ± 13.9 (% of IBMX, n = 30). Numbers between parentheses represent trabeculae/patient. At 37°C (-)-isoprenaline and Ca2+ increased force (mN) from a basal value of 5.5 ± 1.0 to 10.6 ± 1.5 and 11.3 ± 1.5, respectively (n=28/4).

78 Chapter 4

Figure 4.5. Lack of positive inotropic effects of SR 58611 at 24ºC in human atrial trabeculae. Curves for SR 58611 were carried out in the absence or presence of nadolol, nadolol + L-748,337, or nadolol + IBMX or nadolol + L-748,337 + IBMX. The increase in force by (-)-isoprenaline (% over basal) in the absence and presence of IBMX was 767 ± 152%, n = 14/4 and 198 ± 24%, n = 30/4, respectively. Numbers between parentheses represent trabeculae/patient. At 24°C (-)-isoprenaline and Ca2+ (8 mM) increased force (mN) from a basal value of 2.6 ± 0.5 to 6.3 ± 0.8 and 6.8 ± 0.7, respectively (n = 28/4).

4.2.4 Stable (-)-CGP 12177-Evoked Contractions are Decreased by (-)-Bupranolol but not by L-748,337.

Human right atrial trabeculae were incubated with nadolol to block β1HAR and

β2AR, and IBMX was used to enhance the contractile effects of (-)-CGP 12177 (Skeberdis et al., 2008). Figure 4.6 illustrates results from a representative experiment showing the blunting effect of (-)-bupranolol but not L-748,337 on contractile effects of (-)-CGP 12177 and Figure 4.7 shows statistical results. IBMX (10 M) slightly increased trabecular force after 30 min (basal 4.13 ± 0.49 mN, IBMX 5.92 ± 0.79 mN, n = 24 trabeculae, four patients, P < 0.0001). The effect of IBMX remained stable for 1 hr but faded by the 90th min of exposure. (-)-CGP 12177 (200 nM) caused a stable increase in force which was significantly reduced by the addition of (-)-bupranolol (1 M) but not affected by the addition of L-748,337 (1 M), (P = 0.001, 1-way ANOVA, Figure 4.7). Results of trabeculae obtained from 4 patients demonstrate that L-748,337 did not affect the response to (-)-CGP 12177

(P = 0.12), inconsistent with mediation through 3AR. In contrast, (-)-bupranolol

Chapter 4 79

reduced the response by 91 ± 16%, n = 4, P = 0.002 (Figure 4.7), consistent with mediation through 1LAR, as observed before on human atrium (Kaumann et al.,

1996a) and recombinant 1AR (Joseph et al., 2004b).

Figure 4.6. Mediation of the positive inotropic effects of (-)-CGP 12177 through β1LAR but not β3AR. (-)-Bupranolol, but not L-748,337, reduces the increase in contractile force elicited by (-)-CGP 12177 (200 nM) in the presence of IBMX (10 M) and nadolol (200 nM). Representative and time-matched experiments in 6 trabeculae from the right atrial appendage of a 51 year old male patient undergoing coronary artery bypass grafting depicting the time course of the effects of IBMX (top tracing), of (-)-CGP 12177 (second tracing), lack of antagonism by L-748,337 (1 M) (third and fourth tracings) and blunting effect of (-)-bupranolol (1 M) (bottom two tracings). Experiments were concluded by the addition of 200 M (-)-isoprenaline (Iso) and 7 mM Ca2+ (final bath concentration 9.25 mM). Arrows indicate the addition of drugs. Bars indicate forces and time scale.

80 Chapter 4

Figure 4.7. Statistical data from 24 trabeculae of 4 patients. Shown is mean data corresponding to the four different experimental conditions used, as shown in Figure 4.6. A comparison of the contractile force obtained in the presence of 200 nM (-)-CGP 12177 and after incubation of L-748,337, (-)-bupranolol or control was carried out by 1-way ANOVA which indicated differences in cardiodepression (P = 0.001). Post-hoc tests revealed that (-)-bupranolol caused a reduction in contractile force, while there was no difference between control (marginal fade of (-)-CGP 12177 contractile force over 50 min corresponding to the time of L-748,337 exposure) and L-748,337 (statistics indicated on the Figure).

4.2.5 L-748,337 does not Modify the Contractile Potency of (-)-CGP 12177. Conceivably, the lack of antagonism by L-748,337 of the inotropic response to

(-)-CGP 12177 could be due to persistant binding of (-)-CGP 12177 to 3AR.

Therefore experiments were performed to investigate whether the 3AR occupancy caused by a 30 min preincubation of the atrial trabeculae with L-748,337 could prevent the inotropic effects of (-)-CGP 12177. However, cumulative concentration- effect curves for (-)-CGP 12177 in the presence of nadolol (200 nM) and IBMX (10 μM) were not shifted to the right by pre-incubation with L-748,337 (1 M) (Figure

4.8), inconsistent with mediation through 3AR. The -logEC50M values of (-)-CGP 12177 in the absence and presence of L-748,337 were 7.21 ± 0.09 and 7.41 ± 0.13 respectively (data from 25 trabeculae from 8 patients, P = 0.2).

Chapter 4 81

Figure 4.8. L-748,337 does not change the inotropic potency of (-)-CGP 12177 in the presence of nadolol (200 nM) and IBMX (10 M). Concentration-effect curves for (-)-CGP 12177 in the absence and presence of L-748,337 (1 M). Dotted concentration-effect curve shows a theoretical curve for (-)-CGP 12177 in the presence of L-748,337, if L-748,337 antagonised the effects of (-)-CGP 12177 at 3AR with pKB = 7.65. -LogEC50 values for (-)-CGP 12177 in the absence (control) and presence for L-748,337 were 7.21 ± 0.09 (12 trabeculae) and 7.41 ± 0.13 (11 trabeculae), respectively, from 8 patients. (-)-Isoprenaline (200 µM) increased force by 511 ± 103% over IBMX, n = 23/8. Numbers between parentheses represent trabeculae/patients. ISO, (-)-isoprenaline.

4.2.6 ICa-L Responses to SR 58611, BRL 37344, and (-)-CGP 12177 at 24°C.

ICa-L responses to the agonists in the presence of nadolol (200 nM) at 24C are shown in Figure 4.9. Similar to the findings of Skeberdis et al., (2008), SR 58611 and BRL 37344 increased ICa-L and the responses were prevented by L-748,337, consistent with mediation through 3AR. The responses to BRL 37344 (1 μM) were: 29.4 ± 9.1% (range –10.7 to 94.7%), n = 14/9 myocytes/patients (P = 0.01, paired t-test). The responses to SR 58611 (1 μM) were (% over basal): 21.1 ± 10.0% (range: from -4.8 to 217.6%), n = 22/8 (P = 0.034).

(-)-CGP 12177 (1 μM) also increased ICa-L (Figure 4.9). This effect was not observed when 1LAR was blocked with (-)-bupranolol. However, L-748,337 did not

82 Chapter 4

prevent the response to (-)-CGP 12177. These results therefore indicate mediation of the response to (-)-CGP 12177 through 1LAR but not 3AR. The response to (-)-CGP 12177 was 34.5 ± 10.8% (-3.9 to 151.9%) n = 18/7 (P = 0.002). For comparison, the response to (-)-isoprenaline (1 μM) was 289.6 ± 57.9% (90.0 to 705.0%) n = 10/5 (P = 0.00002).

Figure 4.9. L-748,337(1 µM, L) prevents small ICa-L responses to SR 58611 (1 µM, SR) and BRL 37344 (1 µM, BRL) but not to (-)-CGP 12177 (1 µM, CGP) at 24ºC in the presence of nadolol (200 nM, N). The responses to (-)-CGP 12177 were prevented by (-)-bupranolol (10 M, Bu). The ICa-L response to (-)-isoprenaline (1 µM, ISO) in the absence of nadolol is shown for comparison. Numbers in columns represent myocytes/patients. *P < 0.05 compared to control. C = control.

Chapter 4 83

4.2.7 BRL 37344 and SR 58611 do not Increase ICa-L at 37°C.

BRL 37344 and SR 58611 failed to elicit ICa-L responses at 37C in the absence or presence of IBMX (Figure 4.10). (-)-Isoprenaline (1 μM) in the absence of IBMX increased ICa-L to 344 ± 34.3 (160-530%) over basal (n=13/9 P = 0.00002).

Figure 4.10. SR 58611 (1 µM, SR) and BRL 37344 (1 µM, BRL), in the presence of nadolol (200 nM, N) failed to increase ICa-L in the absence and presence of IBMX (10 µM) at 37ºC. The ICa-L response to (-)-isoprenaline (1 µM, ISO) in the absence of IBMX is shown for comparison. Numbers in columns represent myocytes/patients. *P < 0.05 compared to control (C).

4.2.8 Antagonism of (-)-CGP 12177-Evoked ICa-L Increases by (-)-Bupranolol but not by L-748,337 at 37C

The ICa-L responses to (-)-CGP 12177 at 37C are shown in Figure 4.11.

(-)-CGP 12177 (1 µM) did not affect ICa-L while (-)-isoprenaline caused a 4-fold increase. However, in the presence of IBMX, which caused a small increase of ICa-L,

(-)-CGP 12177 elicited a further increase in ICa-L. The response to IBMX (% over basal) was 21.8 ± 9.4% (-12.5 to 131.3%) n = 20/9 (P = 0.037) and that to (-)-CGP 12177 (as % of IBMX) was 31.4 ± 10.0% (-14.2 to 124.1%) (P = 0.015). The response to (-)-CGP 12177 was resistant to blockade by ICI 118,551 and

84 Chapter 4

L-748,337, but prevented by (-)-bupranolol. During blockade of 1HAR and 2AR with nadolol (200 nM) (-)-CGP 12177 caused an increase of ICa-L in the presence of IBMX that was prevented by (-)-bupranolol but resistant to blockade by L-748,337. The response to IBMX in the presence of nadolol (200 nM) was 9.3 ± 4.8% over basal (-16.4 to 39.1%) n = 13/5 (P = 0.04). The response to (-)-CGP 12177 in the presence of both nadolol and IBMX (as % over IBMX) was 16.8 ± 8.2% (0 to 103.1%) (P = 0.027). The responses to (-)-CGP 12177 in the presence of IBMX were not different in the absence and presence of nadolol (P = 0.3). Responses to (-)-CGP 12177 in the presence of IBMX did not significantly differ from the corresponding responses in the presence of ICI 118,551 or L-748,337. Taken together, these results are consistent with mediation of the ICa-L responses to

(-)-CGP 12177 through 1LAR, but not 1HAR, 2AR or 3AR.

Figure 4.11. (-)-CGP 12177 (1 µM, CGP) causes small increases of ICa-L at 37ºC, compared to (-)-isoprenaline (1 µM, ISO), in the presence but not absence of IBMX (10 µM) that are prevented by (-)-bupranolol (10 µM, BU) but not by L-748,337 (1 µM, L), ICI 118,551 (50 nM, ICI) or nadolol (200 nM, N). *P < 0.05 compared to the absence of IBMX. #P < 0.05 compared to IBMX alone. Numbers in columns represent myocytes/patients. One arrow indicates agonist addition, two arrows indicate addition of IBMX followed by addition of (-)-CGP 12177. C = control.

Chapter 4 85

4.2.9 SR 58611 does not Increase Atrial Force at 24C.

To investigate whether the small SR 58611-evoked increase of ICa-L, inhibitable by L-748,337, is inotropically relevant, the effects of SR 58611 at 24C in the presence of nadolol (200 nM) were investigated, in the absence and presence of IBMX. SR 58611 (1 nM-10 M) failed to increase contractility under these conditions (Figure 4.5).

4.2.10 (-)-CGP 12177-Evoked Increases in Atrial Force at 24°C are Antagonised by (-)-Bupranolol but not by L-748,337. Human right atrial trabeculae were used to investigate whether the increase in

ICa-L observed with (-)-CGP 12177 in the presence of IBMX at 24°C is also translated into a positive inotropic effect (Figure 4.12). Nadolol (200 nM) decreased contractile force, probably as an inverse agonist. Since the positive inotropic effects to (-)-CGP 12177 tend to fade (Kaumann et al., 2007; Kaumann et al., 1996a) due to phosphodiesterase activity, and to amplify possible force responses (Kaumann et al., 2007; Kaumann et al., 1997c), IBMX (10 μM) was administered. IBMX increased force until a plateau ensued, on which cumulatively increasing (-)-CGP 12177 concentrations were administered. (-)-CGP 12177 caused concentration-dependent decreases of force at low concentrations, as a weak inverse agonist, followed by marginal increases in force at high concentrations. Pre-incubation with L-748,337 (1 μM) for 30 min before the IBMX administration did not affect the cardiodepressant or cardiostimulant effects of (-)-CGP 12177 (Figure 4.12).

The full inverse agonist (-)-propranolol (Chidiac et al., 1994) abolishes the inverse agonist activity of (-)-CGP 12177 at ICa-L in the presence of IBMX in murine ventricular myocytes (Freestone et al., 1999). Therefore, to prevent the inverse agonist activity and uncover agonist activity of (-)-CGP 12177 (-)-propranolol was used. As expected from a robust inverse agonist, (-)-propranolol decreased contractile force significantly more than nadolol (Figure 4.13). (-)-CGP 12177 produced positive inotropic effects in the presence of both IBMX and (-)-propranolol. L-748,337 did not antagonise the effects of (-)-CGP 12177 but (-)-bupranolol (1 M) caused a log unit rightward shift of the concentration-effect

86 Chapter 4

curve for (-)-CGP 12177 (Figure 4.13), consistent with mediation through 1LAR, but not 3AR.

Figure 4.12. Inotropic effects of (-)-CGP 12177 at 24ºC. (-)-CGP 12177 is a moderate inverse agonist in the presence of both nadolol (200 nM) and IBMX (10 μM). Antagonists were incubated for 45 min, followed by an additional administration of IBMX for 15 min before a concentration-effect curve for (-)-CGP 12177 was carried out. Numbers between parentheses represent trabeculae/patients. (-)-Isoprenaline (200 µM), administered after a steady state response to the highest agonist concentration had been observed, increased force (% over IBMX) by 340 ± 79%, n = 15/6.

Figure 4.13. (-)-CGP 12177 causes positive inotropic effects in the presence of both (-)-propranolol (200 nM) and IBMX (10 µM) through β1LAR but not β3AR. The effects of (-)-CGP 12177 were resistant to blockade by L-748,337 (1 μM) but antagonised by (-)-bupranolol (1 μM). Antagonists were incubated for 45 min, followed by an additional administration of IBMX for 15 min before a concentration- effect curve for (-)-CGP 12177 was carried out. Numbers between parentheses represent trabeculae/patients. (-)-Isoprenaline (200 µM), administered after a steady state response to the highest agonist concentration had been observed, increased force (% over IBMX) by 947 ± 202%, n = 19/9.

Chapter 4 87

4.3 DISCUSSION

In this study, evidence is provided against the hypothesis that 3AR activation enhances human atrial contractility. The 3AR agonist SR 58611 did not modify atrial force. The inotropic effects of BRL 37344 were not mediated through 3AR but through 1AR/2AR. BRL 37344 and SR 58611 caused small increases of ICa-L through 3AR at 24C but not at 37C. The SR 58611-evoked increase of ICa-L at

24C was uncoupled from contractility. (-)-CGP 12177 increased force and ICa-L at both 24°C and 37°C through 1LAR but not 3AR.

Unlikelihood of positive inotropy through human atrial 3AR SR 58611 causes colon relaxation (Bianchetti et al., 1990; Kaumann et al.,

1996a) and lipolytic effects (Langin et al., 1991) through native 3AR, as well as marked activation of adenylyl cyclase through recombinant 3AR (Blin et al., 1994). However Skeberdis et al., (2008), reported a marginal increase in atrial force with SR 58611 in the absence of IBMX, but did not investigate the effect in the presence of IBMX. Experiments presented in this thesis did not detect inotropic effects of SR 58611 (1 nM-10 M) despite the presence of IBMX (Figures 4.4 and 4.5), which should have boosted the inotropic responses mediated through a cAMP-dependent pathway following AR activation. The absence of increases of human atrial force is in agreement with the failure of SR 58611 to cause cardiostimulation in guinea-pig (Bianchetti et al., 1990) and rat (Kaumann et al., 1996a).

The work presented in this thesis confirms the findings (Skeberdis et al., 2008) that BRL 37344 increased atrial force in the presence of IBMX and nadolol.

However, the 3AR-selective antagonist L-748,337 did not affect the responses to

BRL 37344 (Figure 4.1), inconsistent with mediation through 3AR.

The positive inotropic effects of BRL 37344 in human atrial myocardium are antagonised by propranolol (Arch et al., 1993; Pott et al., 2003) and it has been suggested that BRL 37344 acts through both 1AR and 2AR, plausibly mainly

88 Chapter 4

through the latter receptor (Arch et al., 1993). Nadolol caused a rightward shift of the concentration-effect curve for BRL 37344 (Figure 4.1), indicating mediation through

1AR/2AR. The inotropic effects of BRL 37344 shown by Skeberdis et al., (2008) in the presence of nadolol, are consistent with the agonist surmounting the blockade of 1AR and 2AR by nadolol. The affinity estimate (dissociation equilibrium constant Ki) of nadolol of 22 nM from recombinant 1AR, expressed at physiological density (Joseph et al., 2004b), allows the prediction of a one log unit rightward shift with the nadolol concentration of 200 nM used by Skeberdis et al., (2008). As expected, the experiments shown in Figure 4.1 demonstrate that the effects of BRL 37344 are surmountably antagonised by 200 nM nadolol. However, the nadolol- induced shift of the concentration-effect curve was 0.6 log unit greater than the expected one log unit shift (Table 4.2). This could be related to preferential activation of 2AR by BRL 37344 in human atrium (Arch et al., 1993), in line with the 20-fold higher affinity of BRL 37344 for recombinant 2AR compared to 1AR (Sennitt et al., 1998). Affinity estimates from recombinant receptors, expressed at high density, reveal a 23-fold higher affinity of nadolol for 2AR than for 1AR (Baker, 2005a), in line with the greater than one log unit shift of the concentration- effect curve for BRL 37344. A preferential activation of 2AR by BRL 37344 is also supported by the greater antagonism of its inotropic effects by 2AR-selective

ICI 118,551 compared to 1AR-selective CGP 20712A in human atrial trabeculae (Figure 4.2). The greater antagonism by concurrent ICI 118,551 + CGP 20712A than by ICI 118,551 alone (Figure 4.2 and Table 4.2), suggests that BRL 37344 activates

2AR at low concentrations and 1AR at high concentrations.

The effects of both (-)-CGP 12177 and 3AR agonists are relatively resistant to blockade by (-)-propranolol (Hoffmann et al., 2004; Kaumann et al., 2008). It has been previously reported for human atrium that 6-60 M of the 3AR agonists BRL 37344, SR 58611, CL 316243 and ZD 2079 are devoid of cardiostimulant and cardiodepressant effects and do not antagonise the positive inotropic effects of (-)-CGP 12177 in the presence of (-)-propranolol (200 nM) (Kaumann et al., 1997a). The lack of antagonism of the effects of (-)-CGP 12177 by BRL 37344 and

Chapter 4 89

additional 3AR agonists is inconsistent with the conclusion of Skeberdis et al.,

(2008) that (-)-CGP 12177 can cause cardiostimulation through 3AR.

Additional evidence against a 3AR-mediated positive inotropic effect of

(-)-CGP 12177 is obtained from human heart failure where ventricular 3AR expression is increased (Moniotte et al., 2001b). In contrast, binding of (-)-[3H]-CGP

12177 to both human atrial 1HAR and 1LAR sites and the inotropic potency of (-)-CGP 12177 are decreased in heart failure (Sarsero et al., 2003).

1LAR but not 3AR mediates the positive inotropic effects and ICa-L responses to (-)-CGP 12177 in human atrium at 37°C

The 3AR-selective blocker L-748,337 failed to antagonise the positive inotropic responses and ICa-L responses to (-)-CGP 12177 which is inconsistent with mediation through 3AR (Figures 4.8 and 4.11). This together with the finding that

(-)-bupranolol blocks both the increases in ICa-L (Figure 4.11) and contractile force by (-)-CGP 12177 (Figure 4.6) in human atrial myocardium, suggests mediation through

1LAR.

Affinity estimates for the L-748,337-3AR complex (-logKL = pKL) are available from recombinant human 3AR, transfected into CHO cells (pKL = 8.40,

Candelore et al., 1999) and from functional human 3AR of detrusor muscle (pKL =

7.65, Wuest et al., 2009). Knowing the KL of L-748,337, the concentration-ratio (CR) of equieffective (-)-CGP 12177 concentrations in the presence and absence of

L-748,337 can be calculated from CR=1+[L-748,337]/KL, where KL is the equilibrium dissociation constant of L-748,337 for the 3AR. Using 22 nM (from the pKL = 7.65) as the equilibrium dissociation constant (Wuest et al., 2009) and assuming that L-748,337 and (-)-CGP 12177 compete for the same 3AR population, 1 M L-748,337 would cause a CR of 46.5, equivalent to a 1.67 log unit rightward shift of the concentration-effect curve for (-)-CGP 12177 (dotted line of Figure 4.8). However, 1 M L-748,337 did not change the concentration-effect curve and potency (-logEC50M) of (-)-CGP 12177, ruling out mediation through 3AR.

90 Chapter 4

Moreover, L-748,337 did not significantly reduce the effects of (-)-CGP 12177 in the kinetic experiments (Figure 4.6). Instead, (-)-bupranolol antagonised the effects of (-)-CGP 12177, as previously observed through human atrial (Kaumann, 1996;

Kaumann et al., 2008) and recombinant 1AR (Joseph et al., 2004b), indicating mediation through 1LAR (Kaumann et al., 2008). Furthermore, (-)-bupranolol also prevented the small increase in ICa-L caused by (-)-CGP 12177, as expected from mediation through 1LAR (Figue 4.11). The consistency of moderate inotropic and 2+ 2+ ICa-L responses to (-)-CGP 12177 suggests that the Ca entry through the L-type Ca channel may have released more Ca2+ from the sarcoplasmic reticulum through Ca2+- sensitive ryanodine RyR2 channels (Bers, 2002), thereby contributing to the increase in contractile force produced by (-)-CGP 12177. These results agree with the persistence of cardiostimulant effects of (-)-CGP 12177 and blockade by

(-)-bupranolol in 3AR knockout mice (Kaumann et al., 1998).

(-)-CGP 12177-evoked increases in ICa-L are mediated through 1LAR at both 24°C and 37°C Skeberdis et al., (2008) demonstrated in human atrial cardiomyocytes at room temperature (19-25°C) an increase in ICa-L by BRL 3744, SR 58611 and (-)-CGP 12177 which was antagonised by 1 M L-748,337, suggesting mediation through 3AR. This study confirmed that the 3 agonists increased ICa-L at 24°C and that L-748,337 antagonised the effects of BRL 37344 and SR 58610 (Figure 4.9).

However, it was found that the (-)-CGP 12177-evoked ICa-L responses were resistant to blockade by L-748,337 but antagonised by (-)-bupranolol (Figures 4.9 and 4.11), ruling out an involvement of 3AR, but indicating mediation through 1LAR. The reason for the discrepancy between these results and those of Skeberdis et al., (2008) is unclear. Inspection of Figure 2 of Skeberdis et al., (2008), which compares the blockade by L-748,337 of the ICa-L responses to BRL 37344 and (-)-CGP 12177, reveals that the onset kinetic of the response to BRL 37344 after washout of L-748,337 was conspicuously slower than the onset kinetic of BRL 37344 before the L-748,337 exposure. The slow onset response to BRL 37344 after washout of L-748,337 is presumably due to the time it takes for the antagonist to dissociate from the 3AR. One would expect a similar delay of the onset kinetics of (-)-CGP 12177 after washout of L-748,337. Surprisingly, however, for unknown reasons the onset

Chapter 4 91

kinetics of (-)-CGP 12177, before and after L-748,337, were virtually identical. Unfortunately, these puzzling kinetic differences between BRL 37344 and (-)-CGP 12177 and the antagonism by L-748,337 were not discussed (Skeberdis et al., 2008).

Electrophysiology experiments did not detect ICa-L responses to (-)-CGP 12177 at 37°C but uncovered small but significant increases of ICa-L in the presence of

IBMX (Figure 4.11). L-748,337 (1 µM) did not prevent these ICa-L responses to

(-)-CGP 12177, rejecting an involvement of 3AR but (-)-bupranolol blocked the responses, confirming mediation through 1LAR (Figure 4.11). These results suggest that at physiological temperature phosphodiesterases hydrolyse sufficient cAMP, generated through 1LAR stimulation, to blunt PKA-catalysed phosphorylation of the L-type Ca2+ channel. A previous study reports that inhibition of phosphodiesterases with IBMX facilitates the appearances of small (-)-CGP 12177-evoked ICa-L responses (equivalent to 19% of the response to (-)-isoprenaline) on murine ventricular cardiomyocytes (Freestone et al., 1999).

Human atrial 2AR mediate robust increases in ICa-L at room temperature (Skeberdis et al., 1997) and there is evidence for agonist effects of racemic

CGP 12177, mediated through recombinant 2AR (Pak et al., 1996) and

(-)-CGP 12177 in human 2AR overexpressed in the hearts of TG4 mice (Heubach et al., 2003b). However, the 2AR-selective antagonist ICI 118,551 failed to antagonise the (-)-CGP 12177-evoked ICa-L response, therefore ruling out a contribution of 2AR under the conditions at 37ºC.

The 3AR-mediated ICa-L responses to BRL 37344 and SR 58611 at 24ºC are uncoupled from the contractile machinery Experiments from this study confirmed that at 24ºC, BRL 37344 and SR 58611 caused a β3AR-mediated increase of ICa-L in myocytes at 24ºC (Figure 4.9). In contrast to Skeberdis et al., (2008), this β3AR-mediated increase of ICa-L was surprisingly small compared to the effect of non-selective βAR-stimulation with (-)-isoprenaline (Figure 4.9). It should be noted that the size of the agonist effects on

92 Chapter 4

ICa-L reported by Skeberdis et al., (2008) differ widely between 136  21% and 45  6% increase over basal, of which only the latter estimate was confirmed. Therefore experiments were performed to investigate whether the small increases in ICa-L are translated into contractile responses at 24°C. The results with atrial trabeculae at 24ºC in the presence of nadolol failed to uncover any inotropic effect of BRL 37344 and SR 58611 (1 nM-10 M, Figures 4.3 and 4.5), not even in the presence of the phosphodiesterase inhibitor IBMX. The lack of inotropic response to BRL 37344 and

SR 58611 could be related to the small magnitude of the ICa-L response at 24ºC (10% and 7%, respectively, of the (-)-isoprenaline response), or to an inherent inability of the 3AR to produce the coupling messages necessary to activate the contractile machinery. For either situation, the enhanced activation of ICa-L through 3AR appears to be unable to stimulate Ca2+-induced Ca2+ release from the RyR2 channels of the sarcoplasmic reticulum (Bers, 2002; Fabiato, 1983), thereby preventing an increase in atrial contractility.

Relationship between ICa-L responses and inotropic responses to (-)-CGP12177 at 24ºC

This study has demonstrated that at 37ºC the small ICa-L responses correlate with the moderate inotropic responses to (-)-CGP 12177, both mediated through

1LAR under the conditions of PDE inhibition with IBMX. The inotropic relevance of the 1LAR-mediated increase in ICa-L at 24ºC is however not clear. (-)-CGP 12177 only produced marginal increases of contractile force in the presence of nadolol (Figure 4.12). Experiments attempted to uncover (-)-CGP 12177-evoked increases in force by inhibiting phosphodiesterase with IBMX. However, low (-)-CGP 12177 concentrations in the presence of both nadolol and IBMX caused concentration- dependent negative inotropic effects, higher concentrations tended to increase contractile force (Figure 4.12). These cardiodepressant effects and marginal cardiostimulant effects of (-)-CGP 12177 were not significantly modified by

L-748,337 and appear therefore unrelated to 3AR (Figure 4.12). This study attributes the negative inotropic effects of (-)-CGP 12177 to inverse agonism via

1AR, as previously observed on murine ventricular ICa-L at room temperature. Under this situation (-)-CGP 12177 was an inverse agonist in the presence of IBMX, but increased ICa-L (Freestone et al., 1999) in the presence of the efficacious inverse

Chapter 4 93

agonist (-)-propranolol (Chidiac et al., 1994). It is suspected that the marginal positive inotropic responses to higher (-)-CGP 12177 concentrations were probably mediated through 1LAR. As previously observed with murine ICa-L responses to (-)-CGP 12177 (Freestone et al., 1999), (-)-propranolol abolished the negative inotropic effects and facilitated positive inotropic effects of (-)-CGP 12177. These positive inotropic effects of (-)-CGP 12177 in the presence of (-)-propranolol and IBMX were antagonised by (-)-bupranolol but not L-748,337, consistent with mediation of ICa-L responses to (-)-CGP 12177 through 1LAR but not 3AR (Figure 4.13).

Although the small ICa-L responses to (-)-CGP 12177 through 1LAR and

SR 58611 through 3AR at 24ºC were not significantly different (P = 0.39), only activation of the 1LAR, but not 3AR, enhanced contractility. However there is a caveat, unlike the ICa-L responses to the two agonists, which could be elicited in the absence of IBMX, a special condition containing not only IBMX but also (-)-propranolol (to prevent (-)-CGP 12177-evoked inverse agonism) was necessary to demonstrate small concentration-dependent increases in atrial contractility with

(-)-CGP 12177 (Figure 4.13). Nevertheless at 24ºC, 1LAR but not 3AR, appear to activate mechanisms leading to enhanced contractility. Thus, excitation-contraction coupling was receptor-dependent but not temperature-dependent.

94 Chapter 4

Chapter 5: Contribution of Transmembrane Domain V of the 1-Adrenoceptor to 1L-Adrenoceptor Pharmacology

5.1 INTRODUCTION

The 1-adrenoceptor (1AR) is activated by (-)-noradrenaline and blocked by all clinically used -blockers. Some -blockers, typified by (-)-CGP 12177 and

(-)-pindolol not only block the 1AR, but also activate it at higher concentrations (~2-3 orders of magnitude) than those required to cause blockade (reviewed in Kaumann et al., 2008). The results presented in the preceding section, in addition to evidence from other studies, demonstrates that -blockers such as (-)-CGP 12177 and (-)-pindolol bind to the 1AR at two different sites, one that blocks

(-)-noradrenaline from activating the receptor, the 1H site, and another which activates the receptor, the 1L site (Kaumann et al., 2008). Other -blockers typically block 1H- with ~2-3 orders of magnitude higher affinity than 1LAR (Kaumann et al., 2008).

The atypical pharmacology of some -blockers at 1AR was originally observed in the heart where (-)-pindolol and related indoleamines, (-)-CGP 12177 and other -blockers such as oxprenolol and (-)-alprenolol caused cardiostimulant effects resistant to blockade by propranolol and bisoprolol but which were blocked by moderately low (1 M) concentrations of (-)-bupranolol (reviewed in Kaumann et al., 2008). Because the effects of (-)-CGP 12177 were resistant to 200 nM (-)-propranolol (Kaumann, 1996), a concentration that causes >60-fold rightward shift of noradrenaline effects in heart (Gille et al., 1985), the AR subtype through which these effects were mediated was not directly evident. It wasn’t until experiments were performed using genetically engineered AR knockout mice that a

1AR mechanism was authenticated. The stimulant effects of (-)-CGP 12177 were observed in murine cardiac and adipose tissues from 3AR knockout mice, 2AR knockout mice but not 1-/2AR knockout mice (Kaumann et al., 2001; Kaumann et

Chapter 5 95

al., 1998; Konkar et al., 2000a; Konkar et al., 2000b). Consistent findings were observed at recombinant 1AR where (-)-CGP 12177 exhibited antagonist properties at low concentrations, but agonist effects at higher concentrations (Pak et al., 1996) and other antagonists displayed a characteristic 2-3 orders of magnitude difference of affinity against noradrenaline/isoprenaline versus (-)-CGP 12177/(-)-pindolol (reviewed in Kaumann et al., 2008). These findings are consistent with results of the previous section and confirm that the cardiostimulant effects of (-)-CGP 12177 are mediated through 1LAR.

Although studies have successfully characterised the pharmacological properties of (-)-CGP 12177 and other non-conventional partial agonists, the molecular features that contribute to the existence of two forms of pharmacology at

1AR have not been conclusively identified. A model of the molecular properties of the 1LAR binding site has been proposed from mutagenesis studies combined with molecular modelling (Baker et al., 2008). Key findings were that the amino acids, Asp138 (TMDIII) and Asn363 (TMDVI) were critical for binding and agonist activity at 1AR, and Ser228 and Ser229 of TMDV contribute to agonist

[(-)-CGP 12177] potency at 1LAR. These amino acids are also important for catecholamine binding to 1HAR leading to the conclusion that the agonist binding site for (-)-CGP 12177 (1LAR) must overlap with the agonist binding site for catecholamines (1HAR) (Baker et al., 2008).

Additionally, binding sites have been found that allow a distinction between

1HAR and 1LAR. Asp138-1AR was found to have less of an effect on affinity and potency at 1LAR compared to 1HAR (Joseph et al., 2004a). The common human

1AR polymorphic amino acid Arg389Gly differentially affects (-)-isoprenaline and

(-)-CGP 12177 agonist responses at 1HAR and 1LAR respectively at recombinant

1AR (Joseph et al., 2004c). At Gly389-1ARs, the maximum cAMP response of

(-)-isoprenaline was reduced by 97% compared to Arg389-1ARs, while for (-)-CGP 12177 it was reduced by only 46%, making (-)-CGP 12177 a full agonist at Gly389-

1ARs (Joseph et al., 2004c).

96 Chapter 5

To further elucidate the molecular requirements of the 1LAR, this study focused on investigating amino acids of transmembrane domain V (TMDV). TMDV plays a critical role in binding at 1HAR, by binding the catechol hydroxyl groups of catecholamines through interactions with serine amino acids 228 and 232 (Liapakis et al., 2000; Rasmussen et al., 2007; Sato et al., 1999; Strader et al., 1989b;

Sugimoto et al., 2002; Warne et al., 2008). Furthermore, Ser211 of the turkey 1AR which corresponds with human Ser228 (TMDV), binds the nitrogen atom of the indole cyanopindolol at 1HAR (Warne et al., 2008). If 1HAR and 1LAR are distinct, separate binding sites, it is possible that TMDV amino acids could differentially contribute to binding and receptor activation at 1HAR and 1LAR.

Because the 2AR does not form a corresponding low-affinity binding site (Baker et al., 2002), it was hypothesised that TMDV amino acids of 1AR that are heterologous with respect to 2AR might be critical for 1LAR and that their absence in 2AR prevents the existence of a corresponding 2LAR.

Results from this study indicate that TMDV and valine 230 of the 1AR contribute in part to the formation of the 1LAR binding site. The involvement of Val230 may result from its influence on the movement of TMDV during activation.

5.2 RESULTS

5.2.1 Characterisation and Validation of Wild-type and Mutant βARs Expressed in CHOAA8 Cells. Radioligand binding experiments with (-)-[3H]-CGP 12177 were used to determine receptor densities of wild-type and mutant ARs in order to identify cell lines expressing physiological receptor densities. For this purpose, radioligand binding was carried out at the high-affinity binding site corresponding to 1HAR.

i) Saturation binding experiments

3 (-)-[ H]-CGP 12177 binding to wild-type 1HAR and 2AR was saturable and of high-affinity (Table 5.1). Cell lines with receptor densities in the ‘physiological’

Chapter 5 97

range expected for 1AR and 2AR, to reflect human heart receptor densities (~70 and ~30 fmol/mg protein respectively) were used for subsequent cAMP experiments. Cell lines expressing high (97 fmol/mg protein) and very high (1292 fmol/mg protein) 2AR densities were also used for cAMP experiments.

Table 5.1 3 Affinity Values of (-)-[ H]-CGP 12177 (pKD) and Receptor Densities (Bmax) for ARs Stably Expressed in CHOAA8 Cells

3 Cell Line pKD (-)-[ H]-CGP12177 Bmax fmol/mg protein n

1AR 10.15 ± 0.04 81 ± 11 14

2AR 9.78 ± 0.09* 27 ± 3 5

 9.88 ± 0.05* 97 ± 13 5

 9.90 ± 0.07* 1292 ± 167 5

1/2TMDVAR 10.21 ± 0.05 101 ± 14 11

1(R222Q)AR 10.21 ± 0.06 96 ± 11 8

1(V230I)AR 10.19 ± 0.04 91 ± 14 8

1(V230A)AR 10.15 ± 0.04 100 ± 5 9

1(S228A)AR 7.86 ± 0.06** 234 ± 13 7

1(S229A)AR 8.78 ± 0.07** 106 ± 8 10

1(S232A)AR 9.71 ± 0.14** 51 ± 6 9 * P < 0.05, Student’s t-test ** P < 0.001, Student’s t-test

3 The affinity of (-)-[ H]-CGP 12177 at mutant 1/2TMDVAR1(R222Q)AR,

1(V230I)AR, 1(V230A)AR did not differ from 1AR (P > 0.05, Figure 5.1, Table 5.1). Site-directed mutation of serine residues 228, 229 and 232 to alanine reduced the affinity of (-)-[3H]-CGP 12177 (P < 0.001, Figure 5.2, Table 5.1). Cell lines with heterologous (1 vs 2) mutant receptor densities ranging from (91-102 fmol/mg protein) and serine to alanine mutations (51 – 234 fmol/mg protein) were used for subsequent experiments.

98 Chapter 5

Wild-type 1AR 100

50 dpm

0 0 2 4 6 8 10 (-)-[3H] CGP 12177

1/2TMDVAR

250

200

150

dpm 100

0 0 2 4 6 8 10 12 14 (-)-[3H] CGP 12177

Figure 5.1. Saturation binding studies performed using (-)-[3H]-CGP 12177 on purified membranes from CHOAA8 cells stably expressing wild-type β1AR (A) and 1/2TMDVAR (B). Graphs represent specific binding from a single representative experiment and demonstrate unchanged affinity of (-)-[3H]-CGP 12177 at 1/2TMDVAR compared to β1AR.

Chapter 5 99

1(S228A)AR 100

50 dpm

0 0 20 40 60 80 (-)-[3H] CGP 12177

1(S229A)AR 100

50 dpm

0 0 10 20 30 40 (-)-[3H] CGP 12177

Figure 5.2. Saturation binding studies performed using (-)-[3H]-CGP 12177 on purified membranes from CHOAA8 cells expressing β1(S228A)AR (A) and β1(S229A)AR (B). Graphs represent specific binding from a single representative 3 experiment and demonstrate reduced affinity of (-)-[ H]-CGP 12177 at β1(S228A)AR and β1(S229A)AR compared to β1AR.

100 Chapter 5

ii) Competition binding experiments Competition binding experiments between known AR ligands and (-)-[3H]-CGP 12177 were carried out to further characterise wild type and mutant

ARs. The selective 1AR antagonists CGP 20712A and bisoprolol and the selective 3 2AR antagonist ICI 118,551 competed monophasically with (-)-[ H]-CGP 12177 at wild type 1AR and 2AR with affinities consistent with known values at 1AR and

2AR (Figures 5.3 and 5.4, Table 5.2). (-)-Bupranolol was 7.9 fold selective for

2AR.

The affinities of CGP 20712A, bisoprolol, ICI 118,551 and (-)-bupranolol at the 1/2TMDVAR chimera were intermediate between those at 1AR and 2AR, being closer to their 1AR affinities, except for bisoprolol (1AR 7.88, 2AR 6.64,

1/2TMDVAR 7.18, Figure 5.5, Table 5.2).

Competition binding curves for CGP 20712A at 1(V230I)AR and

1(V230A)AR were monophasic with an affinity for CGP 20712A of 8.54 and 8.01 respectively (Figures 5.6 to 5.8, Table 5.2). On the other hand, competition binding curves for CGP 20712A, but not ICI 118,551 were biphasic for 1(R222Q)AR,

1(S228A)AR, 1(S229A)AR and 1(S232A)AR (Figures 5.7 and 5.8) with affinities for two sites (pKi ~ 14.7 – 15.1 and 8.9 – 10.8, Table 5.2). The affinity of ICI

118,551 at these mutant receptors did not differ significantly from that of 1AR (P >

0.05, 1-way ANOVA), but were significantly different to 2AR (P < 0.001, 1-way ANOVA, Figures 5.6 to 5.8, Table 5.2).

Chapter 5 101

Table 5.2 Affinity Values (pKi) of AR Antagonists used in Competition Binding Experiments

Cell Line pKi CGP20712A % binding n pKi ICI 118,551 n pKi bupranolol n pKi bisoprolol n

1AR 8.15 ± 0.07 4 6.93 ± 0.17 5 8.77 ± 0.05 4 7.88 ± 0.10 4

2AR low density 5.63 ± 0.13 4 9.05 ± 0.14 4 9.67 ± 0.09 4 6.64 ± 0.07 4

1/2TMDVAR 7.76 ± 0.09 4 6.82 ± 0.02** 4 8.59 ± 0.04** 4 7.18 ± 0.04* 4

1(R222Q)AR 8.88 ± 0.12 70 7.21 ± 0.03** 4 4  14.68 ± 0.44 30

1(V230I)AR 8.54 ± 0.04* 4 7.05 ± 0.05** 4

1(V230A)AR 8.01 ± 0.08 5 6.76 ± 0.12** 5

1(S228A)AR 8.98 ± 0.14 40 7.00 ± 0.04** 5 6  14.95 ± 0.39 60

1(S229A)AR 10.84 ± 0.69 35 6.65 ± 0.04** 4 5  15.81 ± 0.47 65

1(S232A)AR 9.33 ± 0.14 45 7.38 ± 0.02** 5 5  15.11 ± 0.24 55 * P < 0.05, pKi of CGP20712A at 1(V230I)AR compared with 1AR ** P < 0.001, pKi of ICI 118,551 at all mutant cell lines compared with 2AR ** P < 0.001, pKi of bupranolol at 1/2TMDVAR compared with 2AR, * P < 0.01, pKi of bupranolol at 1/2TMDVAR compared with 1ARand2AR

102 Chapter 5

Wild-type 1AR

100 CGP 20712A ICI 118,551 80

20 H] (-)-CGP 12177 Bound 12177 H] (-)-CGP 3 0 10 9 8 7 6 5 4 % [ -log [competitor] M

Wild-type 1AR

100 (-)-bupranolol bisoprolol 80

20 H] (-)-CGP Bound 12177 3 0 % [ 11 10 9 8 7 6 5 4 -log[competitor] M

Figure 5.3. Competition binding between CGP 20712A (●) or ICI 118,551 (○) graph A, (-)-bupranolol (●) or bisoprolol (○) graph B and (-)-[3H]-CGP 12177. CGP 20712A (pKi = 8.15), ICI 118,551 (pKi = 6.93), (-)-bupranolol (pKi = 8.77) 3 and bisoprolol (pKi = 7.88) inhibited binding of (-)-[ H]-CGP 12177 to 1AR. Data points show mean ± S.E.M. of 4-5 experiments.

Chapter 5 103

Wild-type 2AR

100 CGP 20712A ICI 118,551 80

20 H] (-)-CGP 12177 Bound 12177 H] (-)-CGP 3 0 % [ 10 9 8 7 6 5 4 -log [competitor] M

Wild-type 2AR

100 (-)-bupranolol bisoprolol 80

20 H] (-)-CGP 12177 Bound 12177 (-)-CGP H] 3 0 % [ 11 10 9 8 7 6 5 4 -log[competitor] M

Figure 5.4. Competition binding between CGP 20712A (●) or ICI 118,551 (○) graph A, (-)-bupranolol (●) or bisoprolol (○) graph B and (-)-[3H]-CGP 12177. CGP 20712A (pKi = 5.63), ICI 118,551 (pKi = 9.05), (-)-bupranolol (pKi = 9.67) 3 and bisoprolol (pKi = 6.64) inhibited binding of (-)-[ H]-CGP 12177 to 2AR. Data points show mean ± S.E.M. of 4 experiments.

104 Chapter 5

1/2TMDVAR

100 CGP 20712A ICI 118,551 80

20 H] (-)-CGP Bound 12177 3 0 % [ 10 9 8 7 6 5 4 -log[competitor] M

1/2TMDVAR

100 (-)-bupranolol bisoprolol 80

20 H] (-)-CGP 12177 Bound 12177 H] (-)-CGP 3 0 % [ 10 9 8 7 6 5 4 -log[competitor] M

Figure 5.5. Competition binding between CGP 20712A (●) or ICI 118,551 (○) graph A, (-)-bupranolol (●) or bisoprolol (○) graph B and (-)-[3H]-CGP 12177. CGP 20712A (pKi = 7.76), ICI 118,551 (pKi = 6.82), (-)-bupranolol (pKi = 8.59) 3 and bisoprolol (pKi = 7.18) inhibited binding of (-)-[ H]-CGP 12177 to 1/2TMDVAR through the 1H site. Data points show mean ± S.E.M. of 4 experiments.

Chapter 5 105

1(V230I)AR

100 CGP 20712A ICI 118,551 80

H] (-)-CGP 12177 Bound 12177 H] (-)-CGP 20 3 % [ 0 11 10 9 8 7 6 5 4 -log [competitor] M

1(V230A)AR

100 CGP 20712A ICI 118,551 80

H] (-)-CGP 12177 Bound 12177 H] (-)-CGP 20 3 % [ 0 10 9 8 7 6 5 4 3 -log [competitor] M

Figure 5.6. Competition binding between CGP 20712A (●) or ICI 118,551 (○) at 3 1(V230I)AR graph A, 1(V230A)AR graph B and (-)-[ H]-CGP 12177. CGP 20712A (pKi = 8.54, 1(V230I)AR and pKi = 8.01, 1(V230A)AR) and ICI 118,551 (pKi = 7.05, 1(V230I)AR and pKi = 6.76, 1(V230A)AR) inhibited 3 binding of (-)-[ H]-CGP 12177 through the 1H site with similar affinity to 1AR. Data points show mean ± S.E.M. of 4-5 experiments.

106 Chapter 5

1(R222Q)AR

CGP 20712A 100 ICI 118,551

H] (-)-CGP 12177 Bound 12177 H] (-)-CGP 20 3 % [ 0 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 -log [competitor] M

1(S228A)AR

100 CGP 20712A ICI 118,551

H] (-)-CGP 12177 Bound 12177 (-)-CGP H] 20 3 % [ 0 18 17 16 15 14 13 12 11 10 9 8 7 6 5 -log [competitor] M

Figure 5.7. Competition binding between CGP 20712A (●) or ICI 118,551 (○) at 3 1(R222Q)AR graph A, 1(S228A)AR graph B and (-)-[ H]-CGP 12177. Competition binding curves for CGP 20712A were biphasic against 3 (-)-[ H]-CGP 12177 with affinities of 14.68 and 8.88 for 1(R222Q)AR and 14.95 3 and 8.98 for 1(S228A)AR. ICI 118,551 inhibited binding of (-)-[ H]-CGP 12177 with a pKi = 7.21 for 1(R222Q)AR and a pKi = 7.00 for 1(S228A)AR). Data points show mean ± S.E.M. of 4-6 experiments.

Chapter 5 107

1(S229A)AR

CGP 20712A 100 ICI 118,551

H] (-)-CGP Bound 12177 20 3 % [ 0 18 17 16 15 14 13 12 11 10 9 8 7 6 5 -log [competitor] M

1(S232A)AR

CGP 20712A 100 ICI 118,551

H] (-)-CGP 12177 Bound 12177 (-)-CGP H] 20 3 % [ 0 18 17 16 15 14 13 12 11 10 9 8 7 6 5 -log [competitor] M

Figure 5.8. Competition binding between CGP 20712A (●) or ICI 118,551 (○) at 3 1(S229A)AR graph A, 1(S232A)AR graph B and (-)-[ H]-CGP 12177. Competition binding curves for CGP 20712A were biphasic against 3 (-)-[ H]-CGP 12177 with affinities of 15.81 and 10.84 for 1(S229A)AR and 15.11 3 and 9.33 for 1(S232A)AR. ICI 118,551 inhibited binding of (-)-[ H]-CGP 12177 with a pKi = 6.65 for 1(S229A)AR and a pKi = 7.38 for 1(S232A)AR). Data points show mean ± S.E.M. of 4-5 experiments.

108 Chapter 5

5.2.2 Cyclic AMP Responses to (-)-CGP 12177 and (-)-Isoprenaline at 1AR, 2AR and 1/2TMDVAR in CHOAA8 cells.

(-)-CGP 12177 caused an increase in cAMP at 1AR with a potency of

7.96 ± 0.09 (n = 14), corresponding to activation of 1LAR (Figure 5.9, Table 5.3).

(-)-CGP 12177 had no effect at 2AR when expressed at levels similar to that found in human heart (27 fmol/mg protein), but caused slight increases in cAMP when expressed at higher levels (97 fmol/mg protein) (Figures 5.10 and 5.11). The potency of (-)-CGP 12177 at 2AR, 9.24 ± 0.14 (n = 5) was consistent with earlier descriptions as a partial agonist through activation of a site on the 2AR that is analogous with 1HAR but not 1LAR (Baker et al., 2002). The potencies for

(-)-isoprenaline were similar for 1AR and 2AR (97 ± 13 fmol/mg protein, high density) (P > 0.05, 1-way ANOVA, Table 5.3, Figures 5.9 to 5.11).

Chapter 5 109

Table 5.3

Potencies (pEC50) of (-)-Isoprenaline (1HAR), (-)-CGP 12177 (1LAR) and (-)-Bupranolol Affinity Values (pKB) for ARs Expressed in CHOAA8 Cells

1HAR 1LAR

Cell Line pEC50 (-)- Forskolin % n pKB (-)- n pEC50 (-)-CGP Forskolin n pKB (-)- n

1AR 8.63 ± 0.08 210 ± 75.7 11 9.52 ± 0.28 5 7.96 ± 0.09 83.7 ± 23.5 14 7.20 ± 0.16 8

2AR low density 7.91 ± 0.22 206 ± 28.2 5

2AR high density 8.27 ± 0.09 160.6 ± 18.9 5 9.24 ± 0.14 8 ± 2.0 5

2AR very high 8.67 ± 0.12 292.9 ± 82.2 9 9.31 ± 0.16 7 density

1/2TMDVAR 9.21 ± 0.19* 62.0 ± 11.5 11 9.08 ± 0.22 6 8.90 ± 0.10*** 38.6 ± 8.6 15 8.06 ± 0.17** 8

1(R222QAR) 8.92 ± 0.08 724.0 ± 240.1 7 9.36 ± 0.08 6 8.10 ± 0.10 283.9 ± 59.9 6 7.33 ± 0.23 5

1(V230I)AR 10.08 ± 0.19*** 141.7 ± 54.8 9 9.36 ± 0.28 9 9.08 ± 0.07*** 84.4 ± 22.4 10 7.64 ± 0.12* 8

1(V230A)AR 8.70 ± 0.15 789.3 ± 144.9 7 8.84 ± 0.26 7 7.59 ± 0.08* 201.8 ± 38.4 8 7.28 ± 0.19 5

1(S228A)AR 6.61 ± 0.11*** 3065.0 ± 5 8.86 ± 0.20 5 6.06 ± 0.11*** 510.1±114.7 7 7.60 ± 0.20 7

886.8

1(S229A)AR 8.55 ± 0.06 77.2 ± 14.8 6 9.09 ± 0.18 6 7.53 ± 0.15* 69.3 ± 16.3 6 7.76 ± 0.11* 6

1(S232A)AR 7.41± 0.09*** 125.2 ± 12.0 6 9.78 ± 0.14 6 8.68 ± 0.21** 24.0 ± 5.3 5 8.16 ± 0.15** 5 Forskolin (1 µM) was used as a receptor independent reference point, at a submaximal concentration. * P < 0.05 ** P < 0.01, *** P < 0.001 compared with 1AR

110 Chapter 5

1AR

pEC50 7.96  0.09, n=14

30 Basal

cyclic AMP cyclic 20 pmol/mg protein pmol/mg

11 10 9 8 7 6 5 -Log [(-)-CGP 12177] M

1AR

pEC50 8.63  0.08, n=11

200

150

100 Basal cyclic AMP cyclic 50 pmol/mg protein

11 10 9 8 7 6 -Log [(-)-Isoprenaline] M

Figure 5.9. Concentration-effect curves for cAMP accumulation in CHOAA8 cells stably expressing 1AR (81 ± 11 fmol/mg protein) in response to (-)-CGP 12177 (A) and (-)-isoprenaline (B). Data points show mean ± S.E.M. of 11-14 experiments performed in triplicate.

Chapter 5 111

Low 2AR n=5 40

30 Basal 20 cyclic AMP cyclic pmol/mg protein

11 10 9 8 7 6 5 -Log [(-)-CGP 12177] M

Low 2AR

pEC50 7.91  0.22, n=5 120 100 80 60 40 Basal cyclic AMP cyclic

pmol/mg protein pmol/mg 20 0

11 10 9 8 7 6 -Log [(-)-Isoprenaline] M

Figure 5.10. Concentration-relationship (A) and concentration-effect curve (B) for cAMP accumulation in CHOAA8 cells stably expressing 2AR (27 ± 3 fmol/mg protein) in response to (-)-CGP 12177 and (-)-isoprenaline. Data points show mean ± S.E.M. of 5 experiments performed in triplicate.

112 Chapter 5

High 2AR pEC50 9.24  0.14, n=5 40

30 Basal

20 cyclic AMP pmol/mg protein pmol/mg

12 11 10 9 8 7 6 5 -Log [(-)-CGP 12177] M

High 2AR

pEC50 8.27 ± 0.09, n=5 250

200

150

100 cyclic AMP 50 Basal pmol/mg protein pmol/mg

12 11 10 9 8 7 6 -Log [(-)-Isoprenaline] M

Figure 5.11. Concentration-effect curves for cAMP accumulation in CHOAA8 cells stably expressing 2AR (97 ± 13 fmol/mg protein) in response to (-)-CGP 12177 (A) and (-)-isoprenaline (B). (-)-CGP 12177 increased cAMP through activation of a site on the 2AR that is analogous with 1HAR but not 1LAR (Baker et al., 2002). Data points show mean ± S.E.M. of 5 experiments performed in triplicate.

Chapter 5 113

In order to determine whether heterologous amino acids between 1AR- and

2AR TMDV contribute to 1LAR, cAMP experiments were performed with the

1/2TMDVAR chimera. (-)-CGP 12177 caused an increase in cAMP at 1/2TMDVAR with pEC50 8.90 ± 0.10 (Figure 5.12) which was not different to its potency at 2AR

(pEC50 9.24, P > 0.05, 1-way ANOVA) but was different to 1AR (pEC50 7.96

P < 0.001, 1-way ANOVA). This suggests that 1/2TMDVAR has lost at least in part, the ability to form the low-affinity binding site. A

1/2TMDVAR

pEC50 8.90 ± 0.10, n=15

20 Basal

cyclic AMP 10 pmol/mg protein

11 10 9 8 7 6 5 -Log [(-)-CGP 12177] M

1/2TMDVAR

pEC50 9.21 ± 0.19, n=11 60 50 40

30 Basal 20 cyclic AMP cyclic

pmol/mg protein pmol/mg 10 0

11 10 9 8 7 6 -Log [(-)-Isoprenaline] M

Figure 5.12. Concentration-effect curves for cAMP accumulation in CHOAA8 cells stably expressing 1/2TMDVAR (101 ± 14 fmol/mg protein) in response to (-)-CGP 12177 (A) and (-)-isoprenaline (B). Data points show mean ± S.E.M. of 11-14 experiments performed in triplicate.

114 Chapter 5

Receptor characterisation using agonist potency alone may not be entirely reliable because it is determined by the efficiency of G-protein coupling and other factors, therefore the antagonist affinity of (-)-bupranolol at 1AR, 2AR and

1/2TMDVAR was determined.

The affinity of (-)-bupranolol determined against (-)-CGP 12177 at 1LAR was

7.20 ± 0.16 (n = 8) (Figure 5.13), in line with its affinity at 1LAR in human right atrium (pKB 7.3, Kaumann, 1996) and 9.31 at 2AR (Figure 5.14). The affinity of

(-)-bupranolol at 1/2TMDVAR was 8.06 ± 0.17 (n = 8) (Figure 5.15), which was significantly higher than at 1AR (P < 0.01, Student’s t-test).

The potency of (-)-isoprenaline was increased (~ ½ log unit) at 1/2TMDVAR compared to 1AR, however the affinity of (-)-bupranolol vs (-)-isoprenaline was reduced (~ ½ log unit, Figure 5.15). Whilst it could be argued that the basis of increased potency was due to more efficient coupling of both 1LAR and 1HAR, the effects on the affinity of the antagonist (-)-bupranolol were in opposite directions.

Collectively, these data suggest that the low-affinity binding site of 1LAR has been modified by the substitution of 2AR TMDV amino acids into 1AR.

Chapter 5 115

1AR 60 50 40 30

20 Basal cyclic AMP

pmol/mg protein pmol/mg 10 0

11 10 9 8 7 6 5 -Log [(-)-CGP12177] M

(-)-CGP 12177

(-)-CGP 12177 + (-)-Bupranolol pKB 7.20  0.16 n=8

1AR 300 250 200 150 100 cyclic AMP cyclic Basal

pmol/mg protein pmol/mg 50 0

11 10 9 8 7 6 5 4 3 -Log [(-)-Isoprenaline] M (-)-Isoprenaline

(-)-Isoprenaline + (-)-Bupranolol pKB = 9.52  0.28, n=5

Figure 5.13. Concentration-effect curves for cAMP accumulation in CHOAA8 cells stably expressing 1AR in response to (-)-CGP 12177 (A) and (-)-isoprenaline (B) in the absence (●) or presence (○) of 100 nM (-)-bupranolol. Data points show mean ± S.E.M. of 5-8 experiments performed in triplicate.

116 Chapter 5

2AR 60 50 40 30 20 Basal cyclic AMP cyclic

pmol/mg protein 10 0

12 11 10 9 8 7 6 5 4 -Log [(-)-CGP12177] M (-)-CGP 12177

(-)-CGP12177 + (-)-Bupranolol pKB = 9.31  0.16 n=7

Figure 5.14. Concentration-effect curve for cAMP accumulation in CHOAA8 cells stably expressing 2AR (1292 ± 167 fmol/mg protein) in response to (-)-CGP 12177 in the absence (●) or presence (○) of 100 nM (-)-bupranolol. Data points show mean ± S.E.M. of 7 experiments performed in triplicate.

Chapter 5 117

1/2TMDVAR 60 50 40 30 Basal 20 cyclic AMP cyclic

pmol/mg protein pmol/mg 10 0

11 10 9 8 7 6 5 -Log [(-)-CGP 12177] M (-)-CGP 12177

(-)-CGP 12177 + (-)-Bupranolol pKB = 8.06  0.17, n=8

1/2TMDVAR 80 70 60 50

40 Basal 30

cyclic AMP cyclic 20

pmol/mg protein 10 0

11 10 9 8 7 6 5 4 -Log [(-)-Isoprenaline] M (-)-Isoprenaline

(-)-Isoprenaline + (-)-Bupranolol pKB = 9.08  0.22, n=6

Figure 5.15. Concentration-effect curves for cAMP accumulation in CHOAA8 cells stably expressing 1/2TMDVAR in response to (-)-CGP 12177 (A) and (-)-isoprenaline (B) in the absence (●) or presence (○) of 100 nM (-)-bupranolol. Data points show mean ± S.E.M. of 6-8 experiments performed in triplicate.

118 Chapter 5

5.2.3 Cyclic AMP Responses to (-)-CGP 12177 and (-)-Isoprenaline at Heterologous Point Mutations of 1-2AR TMDV. Molecular Modelling Experiments were performed to determine the individual contribution of the heterologous amino acids in 1AR- 2AR TMDV to the 1LAR. There are 5 heterologous amino acids in TMDV of 1AR and 2AR (Figure 3.2). Arginine 222 (Arg222) and valine 230 (Val230) were investigated on the basis of molecular models provided by Dr Andrew Lucke and Prof David Fairlie (IMB, University of

Queensland). Homology models of the 1AR and 1/2TMDVAR were generated based on the human β2AR-T4 lysozyme fusion protein structure containing carazolol (Figure 5.16). The binding pocket, demonstrated by the position of carazolol in the top half of the structure (above the highlighted proline at position 236), showed that the only residues that would appear to interact with the binding site from TMDV would be those located above Pro236. However in the analysis of this model, docking software suggested that residues Arg222 and Val230 of β1AR do not interact with carazolol or docked conformations of (-)-CGP 12177 in this inactive state.

Figure 5.16. Homology model of the 1AR. Side chains of residues in TMDV of β1AR and the corresponding β1/β2TMDVAR residues are represented by green and red sticks respectively. Carazolol (grey sticks) is shown in the corresponding orientation from the 2AR crystal structure. The purple highlight is the position of a helix kink proline (model provided by Dr Lucke and Prof Fairlie).

Chapter 5 119

The hypothesis was made that in an active state, the TMDV region may rotate inwards, through the helix kink proline. This rotation could potentially bring Arg222 and Val230 into contact with the binding site and influence binding in this region. Consequences of the proposed rotation of TMDV in an active state (not modelled) would mean that the replacement of β1AR Val230 with isoleucine in the

1/2TMDVAR, would introduce a larger hydrophobic group and result in the binding site of 1/2TMDVAR changing shape, and becoming more constricted relative to

β1AR.

Experimental Data

Substitution of glutamine for arginine at position 222 (1(R222Q)AR) had no effect on the potency of (-)-CGP 12177 at 1LAR or antagonist affinity of

(-)-bupranolol at 1HAR and 1LAR (P > 0.05, Figure 5.17, Table 5.3). It did however cause a small increase in potency (~ 0.3 log unit) of (-)-isoprenaline at

1HAR (P = 0.03, Student’s t-test). In cell lines with similar levels of receptors [1AR

81 ± 11 fmol/mg protein, n = 14; 1(R222Q)AR 96 ± 11 fmol/mg protein, n = 8, P > 0.05, Student’s t-test], agonist stimulated cAMP responses were much greater in

1(R222Q)AR cell lines [increase in cAMP (pmol/mg protein), (-)-CGP 12177,

1AR 27.85 ± 4.67, n = 14, 1(R222Q)AR 148.38 ± 12.41, n = 6, P < 0.0001;

Student’s t-test, (-)-isoprenaline, 1AR 129.22 ± 28.32, n = 11, 1(R222Q)AR 700.55 ± 160.90, n = 7, P < 0.001, Student’s t-test, Figure 5.17].

Substitution of isoleucine for valine at position 230 (1(V230I)AR) caused an increase in potency of both (-)-CGP 12177 at 1LAR and (-)-isoprenaline at 1HAR (P < 0.001, Figure 5.18, Table 5.3), but increased the affinity of (-)-bupranolol

(1(V230I)AR 7.64, 1AR 7.20, P < 0.05, Student’s t-test) only against (-)-CGP

12177 and not against (-)-isoprenaline (1(V230I)AR 9.36, 1AR 9.52, P > 0.05, Student’s t-test, Figure 5.17). Thus valine at position 230 appears to exert an inhibitory effect, by reducing the potency and affinity of compounds at 1LAR. The inhibitory effects are however not restricted to 1LAR, since it also contributes to

(-)-isoprenaline potency at 1HAR.

120 Chapter 5

R222Q)AR 200

150

100

cyclic AMP cyclic 50 Basal pmol/mg protein

11 10 9 8 7 6 5 -Log [(-)-CGP 12177] M (-)-CGP 12177

(-)-CGP 12177 + (-)-Bupranolol pKB 7.33  0.23. n=5

R222Q)AR 800

600

400

cyclic AMP cyclic 200 Basal pmol/mg protein pmol/mg

13 12 11 10 9 8 7 6 5 4 -Log [(-)-Isoprenaline] M (-)-Isoprenaline

(-)-Isoprenaline + (-)-Bupranolol pKB = 9.36  0.08 n=6

Figure 5.17. Concentration-effect curves for cAMP accumulation in CHOAA8 cells stably expressing 1(R222Q)AR (96 ± 11 fmol/mg protein) in response to (-)-CGP 12177 (A) and (-)-isoprenaline (B) in the absence (●) or presence (○) of 100 nM (-)-bupranolol. Data points show mean ± S.E.M. of 5-6 experiments performed in triplicate.

Chapter 5 121

V230I)AR 50

20 Basal cyclic AMP 10 pmol/mg protein

11 10 9 8 7 6 5 -Log [(-)-CGP 12177] M (-)-CGP 12177

(-)-CGP 12177 + (-)-Bupranolol pKB 7.64  0.12, n=8

V230I)AR 50

20 Basal

cyclic AMP cyclic 10 pmol/mg protein pmol/mg

13 12 11 10 9 8 7 6 5 4 -Log [(-)-Isoprenaline] M (-)-Isoprenaline

(-)-Isoprenaline + (-)-Bupranolol pKB = 9.36  0.28, n=9

Figure 5.18. Concentration-effect curves for cAMP accumulation in CHOAA8 cells stably expressing 1(V230I)AR (91 ± 14 fmol/mg protein) in response to (-)-CGP 12177 (A) and (-)-isoprenaline (B) in the absence (●) or presence (○) of 100 nM (-)-bupranolol. Data points show mean ± S.E.M. of 8-9 experiments performed in triplicate.

122 Chapter 5

To elucidate whether the effects caused by the amino acid change at position 230 were the result of the size of the isoleucine residue, the Val230 was changed to alanine (1(V230A)AR). The potency of (-)-CGP 12177 was reduced (7.59 vs 7.96, P = 0.01, Student’s t-test) while there was no effect on the affinity of (-)-bupranolol vs (-)-CGP 12177 (Figure 5.19, Table 5.3). Nor was there any effect on the potency of (-)-isoprenaline or the affinity of (-)-bupranolol vs (-)-isoprenaline (P > 0.05, Student’s t-test, Figure 5.19, Table 5.3).

Chapter 5 123

1(V230A)AR 60

50 40

20 cyclic AMP cyclic Basal

pmol/mg protein pmol/mg 10

12 11 10 9 8 7 6 5 -Log [(-)-CGP12177] M (-)-CGP 12177

(-)-CGP 12177 + (-)-Bupranolol pKB 7.28  0.19, n=5

1(V230A)AR 400

300

200

cyclic AMP 100 pmol/mg protein pmol/mg Basal

13 12 11 10 9 8 7 6 5 4 -Log [(-)-Isoprenaline] M (-)-Isoprenaline

(-)-Isoprenaline + (-)-Bupranolol pKB = 8.84 0.26 n=7

Figure 5.19. Concentration-effect curves for cAMP accumulation in CHOAA8 cells stably expressing 1(V230A)AR (100 ± 5 fmol/mg protein) in response to (-)-CGP 12177 (A) and (-)-isoprenaline (B) in the absence (●) or presence (○) of 100 nM (-)-bupranolol. Data points show mean ± S.E.M. of 5-7 experiments performed in triplicate.

124 Chapter 5

5.2.4 Cyclic AMP Responses to (-)-CGP 12177 and (-)-Isoprenaline at Homologous Serine Residues of 1AR, 2AR TMDV. Valine 230 lies between the serine groups, Ser228, Ser229 and Ser232 which are critical for binding catecholamines at 1HAR (Liapakis et al., 2000; Rasmussen et al., 2007; Sato et al., 1999; Strader et al., 1989b; Sugimoto et al., 2002; Warne et al., 2008). Ser228 and Ser229 also contribute to (-)-CGP 12177 agonist potency at

1LAR (Baker et al., 2008). Ser228, Ser229 and Ser232 are homologous with respect to 1AR and 2AR and are therefore not the critical amino acids that determine the ability of 1AR to form 1LAR and restrict 2AR from forming an analogous site. However because of their proximity to Val230, their affect on (-)-CGP 12177 agonist potency and the affinity of (-)-bupranolol was explored. The experimental conditions used to investigate the serines included some differences to those used previously by

Baker et al., (2008). Differences included using the Arg389-1AR genotype and measuring cAMP accumulation as a surrogate of 1AR activation. The 1AR construct used in the Baker et al., (2008) study contained the Gly389 polymorphism and 1AR activation was quantitated by measuring cAMP response element (CRE) gene transcription.

Mutation of Ser228 to alanine (1(S228A)AR) caused a decrease in (-)-CGP 12177 agonist potency (P < 0.001, 1-way ANOVA, Figure 5.20) while altering Ser229 to alanine (1(S229A)AR) caused a trend to a reduction in potency (not significant, Figure 5.21, Table 5.3). Neither mutation affected the affinity of

(-)-bupranolol. Mutation of Ser232 to alanine (1(S232A)AR) caused an increase in (-)-CGP 12177 potency and affinity of (-)-bupranolol (Figure 5.22).

Mutation of Ser228 and Ser232, but not Ser229 to alanine caused a reduction in potency of (-)-isoprenaline (P < 0.001, 1-way ANOVA), but none of the alanine substitutions affected the affinity of (-)-bupranolol (P > 0.05, Student’s t-test, Figures 5.20 to 5.22, Table 5.3).

Chapter 5 125

1(S228A)AR

150

100

50 cyclic AMP cyclic Basal pmol/mg protein pmol/mg

10 9 8 7 6 5 4 3 -Log [CGP12177] M (-)-CGP 12177

(-)-CGP 12177 + (-)-Bupranolol pKB 7.60  0.20, n=7

1(S228A)AR

800

600

400

cyclic AMP 200 Basal pmol/mg protein

10 9 8 7 6 5 4 3 -Log [(-)-Isoprenaline] M (-)-Isoprenaline

(-)-Isoprenaline + (-)-Bupranolol pKB 8.86  0.20, n=5

Figure 5.20. Concentration-effect curves for cAMP accumulation in CHOAA8 cells stably expressing 1(S228A)AR (234 ± 13 fmol/mg protein) in response to (-)-CGP 12177 (A) and (-)-isoprenaline (B) in the absence (●) or presence (○) of 100 nM (-)-bupranolol. Data points show mean ± S.E.M. of 5-7 experiments performed in triplicate.

126 Chapter 5

1(S229A)AR

15 Basal 10 cyclic AMP cyclic 5 pmol/mg protein pmol/mg

10 9 8 7 6 5 4 -Log [(-)-CGP12177] M (-)-CGP 12177

(-)-CGP 12177 + (-)-Bupranolol pKB 7.76  0.11, n=6

1(S229A)AR

20 Basal 10 cyclic AMP cyclic pmol/mg protein pmol/mg

11 10 9 8 7 6 5 4 -Log [(-)-Isoprenaline] M (-)-Isoprenaline

(-)-Isoprenaline + (-)-Bupranolol pKB 9.09  0.18, n=6

Figure 5.21. Concentration-effect curves for cAMP accumulation in CHOAA8 cells stably expressing 1(S229A)AR (106 ± 8 fmol/mg protein) in response to (-)-CGP 12177 (A) and (-)-isoprenaline (B) in the absence (●) or presence (○) of 100 nM (-)-bupranolol. Data points show mean ± S.E.M. of 6 experiments performed in triplicate.

Chapter 5 127

1(S232A)AR 25

15 Basal 10

cyclic AMP cyclic 5 pmol/mg protein pmol/mg

11 10 9 8 7 6 5 4 -Log [(-)-CGP12177] M (-)-CGP 12177

(-)-CGP 12177 + (-)-Bupranolol pKB 8.16  0.15, n=5

1(S232A)AR 70 60 50 40 30

20 Basal cyclic AMP cyclic

pmol/mg protein pmol/mg 10 0

10 9 8 7 6 5 4 3 -Log [(-)-Isoprenaline] M (-)-Isoprenaline

(-)-Isoprenaline + (-)-Bupranolol pKB 9.78  0.14, n=6

Figure 5.22. Concentration-effect curves for cAMP accumulation in CHOAA8 cells stably expressing 1(S232A)AR (51 ± 6 fmol/mg protein) in response to (-)-CGP 12177 (A) and (-)-isoprenaline (B) in the absence (●) or presence (○) of 10 nM (-)-bupranolol. Data points show mean ± S.E.M. of 5-6 experiments performed in triplicate.

128 Chapter 5

5.3 DISCUSSION

(-)-CGP 12177 and some clinically used -blockers such as (-)-pindolol block

1AR and 2AR at low concentrations but have the ability to activate 1AR, but not

2AR at higher (~2 orders of magnitude) concentrations than those that cause blockade (Kaumann et al., 2008). To accommodate these observations, it has been proposed that the 1AR has two separate binding sites; one that is activated by catecholamines (site 1 or the classic “catecholamine” site, (Baker et al., 2008) or high-affinity site (1HAR), (Kaumann et al., 2008) and another site activated by (-)-CGP 12177, (-)-pindolol and other compounds (site 2, (Baker et al., 2008) or low- affinity site (1LAR), (Kaumann et al., 2008). An intriguing property revealed by

(-)-CGP 12177 is the ability of the 1AR but not 2AR to form the low-affinity binding site (Baker et al., 2002; Kaumann et al., 2008; Pak et al., 1996). The current study exploited this finding in an attempt to determine which amino acids of the

1AR are responsible for the formation of 1LAR, by replacing 1AR amino acids with heterologous 2AR amino acids. By doing so, it was thought that a chimeric

1AR that incorporates heterologous 2AR amino acids corresponding to critical amino acids of 1AR that are responsible for 1LAR, will lose the ability to form a low-affinity binding site, 1LAR.

There are a few leads in terms of identification of the binding partners of

(-)-CGP 12177 to 1LAR. Baker et al., (2008) concluded that there must be overlap of the binding site for (-)-CGP 12177 and catecholamines at 1AR. Common amino acids include Asp138 (TMDIII), Asn363 (TMDVII), Ser228 and Ser229 (TMDV) (Baker et al., 2008). It was proposed that a nitrogen of the benzimidazolone group of (-)-CGP 12177 could face the serine residues of TMDV and bind with one of the serines. The serine residues of the 1AR are not heterologous with respect to the

2AR, and therefore are unlikely to differentiate between the 1AR and 2AR in terms of the ability to form the low-affinity binding site of the 1AR (Kaumann et al., 2008). For these reasons, the initial focus of this study was on heterologous amino acids of TMDV and whether any of these amino acids contribute to the formation of 1LAR.

Chapter 5 129

There are five heterologous amino acids in TMDV with respect to 1AR and

2AR which formed the basis of the chimera 1/2TMDVAR (Figure 3.2.). Both the potency of (-)-CGP 12177 for increases in cAMP and the affinity of (-)-bupranolol determined against (-)-CGP 12177, were increased and were significantly different to values at the 1AR. These effects could be attributed to a large extent by substitution of Val230 to isoleucine, which caused an increase in the potency and affinity of compounds at 1LAR. These data suggest that heterologous amino acids of TMDV at least contribute in part and are necessary for the 1LAR binding site and that Val230 is primarily responsible.

From the homology model of the 1AR containing carazolol (Figure 5.16), the Val230 appears on the opposite face of TMDV, away from the binding site and into the membrane. Therefore, it would appear unlikely that Val230 makes direct contact with interacting ligands. The β1AR-m23 crystal structures in complex with full and partial agonists, has demonstrated that non-interacting residues may also contribute to ligand affinity and efficacy (Warne et al., 2011). This occurs via ligand induced movement of the TMD helices resulting in the disruption and/or formation of interhelical TMD interactions (Warne et al., 2011). Furthermore, comparison of the inactive and active AR crystal structures demonstrates that receptor activation induces a contraction of the ligand binding pocket and additional movement of the TMD helices at the cytoplasmic ends (Rasmussen et al., 2011a; Rasmussen et al., 2011b). It is therefore possible that TMD residues and Val230, which are remote to the binding pocket, could affect ligand efficacy and receptor activation by influencing TMD movements relative to each other during the activation process.

Substitution of Val230 for isoleucine introduces a larger hydrophobic residue at this position. This may indirectly affect the movement of TMDV and reduce the formation of 1LAR. Conversely, the presence of isoleucine at position 205 in the

2AR may in part prevent the formation of an analogous 2LAR, by inhibiting movement of TMDV. Substitution of Val230 to alanine [1(V230A)AR] caused a reduction in (-)-CGP 12177 potency but did not affect (-)-bupranolol affinity at

1LAR.

130 Chapter 5

The results of this study suggest that the serine residues, Ser228 and possibly

Ser229 are involved with the 1LAR binding site, but unlike Val230, do not determine its existence. It is well established that the TMDV region plays a critical role in ligand binding at 1HAR and 2AR (Liapakis et al., 2000; Rasmussen et al., 2007; Rasmussen et al., 2011b; Sato et al., 1999; Strader et al., 1989b; Sugimoto et al., 2002; Warne et al., 2008). The recent publication of AR crystal structures has provided insight into the molecular interactions formed between the headgroup of AR ligands and the conserved serine residues in TMDV. Hydrogen bonds appear to form between the catechol groups of isoprenaline and Ser211 and Ser215 (analogous to Ser228 and Ser232 in human β1AR) in the β1AR-m23 crystal structure (Warne et al., 2011). In addition, an agonist induced rotamer conformation change in the side chain of Ser212, allows the side chain to be positioned to form a hydrogen bond with Asn310 in TMDVI, providing additional stabilisation of the ligand binding pocket (Warne et al., 2011). The rotamer conformation change of Ser212 (analogous to

Ser229 in human β1AR) and hydrogen bond formation to Asn310 are absent in the

β1AR-m23 structure containing cyanopindolol (Warne et al., 2008).

Mutation of Ser228 to Ala228 caused ~ 2 log unit reduction in (-)-CGP 12177 potency and there was a trend (not significant) towards a reduction in (-)-CGP 12177 agonist potency for Ala229. The magnitude of the reduction of (-)-CGP 12177 potency at Ala228 was similar to the reduction of (-)-isoprenaline potency at 1HAR (~ 2 log units). The latter is attributed to disruption of hydrogen bonding between Ser228 and the meta-hydroxyl of isoprenaline (Baker et al., 2008) and adrenaline [by analogy with human 2AR (Liapakis et al., 2000)]. It may be inferred that a disruption of hydrogen bonding also occurred with (-)-CGP 12177 and Ala228, resulting in reduced activity at 1LAR.

Substitution of heterologous Arg222 to glutamine did not affect the 1LAR site since the potency of (-)-CGP 12177 and affinity of (-)-bupranolol were unchanged compared to 1AR. Substitution of glutamine increased efficacy of 1(R222Q)AR compared to 1AR, for agonist responses at both 1HAR and 1LAR. The 1AR homology model indicated that Arg222 formed a hydrogen bond to the backbone

Chapter 5 131

Ala348 in TMDVI. This interaction appears to hold TMDV and TMDVI apart and possibly stabilises the inactive form of AR. This interaction was also observed between the analogous residues in the inactive β1AR-m23 crystal structure (see Table 5.4 and Figure 5.23).

A comparison of the inactive β2AR-T4L crystal structure in complex with carazolol and the nanobody stabilised active state β2AR crystal structure reveals an important difference in the rotamer conformations of AR TMDV Gln197

(analogous to β1AR Arg222). The inactive β2AR-T4L structure indicated no hydrogen bond between the side chain of Gln197 and TMDVI backbone Val297

(Cherezov et al., 2007), whereas the active state β2AR structure indicated Gln197 in a different rotamer conformation that allowed a potential hydrogen bond to the backbone of Val297 (Rasmussen et al., 2011a), which formed in conjunction with a helical movement that contracted the β2AR binding site.

The AR Arg222Gln substitution potentially disrupts the hydrogen bond between Arg222 and Ala348. This may destabilise the inactive AR conformation by allowing TMD helices V and VI to move closer together, producing a more active like conformation and increasing the potential for activation relative to the AR.

This observation supports the hypothesis that ligand binding to the β1AR influences amino acids interactions that are remote to the binding pocket. Furthermore, the ligand induced manipulation of these interactions can affect structural processes which inhibit or facilitate βAR activation.

Table 5.4 Analogous AR Amino Acids at Different AR Structures That Interact at the TMDV- TMDVI Interface

1AR homology Equivalent1AR-m23 Equivalentβ2AR-T4L Equivalent active

model residues residues residues state β2AR residues

TMDV Arg222 Arg205 Gln197 Gln197 TMDVI Ala348 Val314 Val297 Val297

132 Chapter 5

Figure 5.23. Overlay of TMDV regions of the AR homology model (green), inactive β2AR-T4L bound to carazolol (cyan) and nanobody stabilised active β2AR (yellow). Potential hydrogen bonds shown in dashed yellow lines. The carazolol bound inactive AR structure indicated no hydrogen bonding interaction between AR Gln197 sidechain and the backbone carbonyl of AR Val297. Agonist bound active AR structure shows Gln197 sidechain making a potential hydrogen bond to the backbone carbonyl of AR Val297. AR Arg222 sidechain makes a potential hydrogen bond to the corresponding backbone carbonyl oxygen of Ala348 in an inactive state (overlay model provided by Dr Lucke and Prof Fairlie).

CGP 20712A was able to identify an extremely high-affinity binding site with pKi ~ 15, in a number of 1AR TMDV mutants (R222Q, S228A, S229A, S232A) in competition binding experiments with (-)-[3H]-CGP 12177. The site was not detected by ICI 118,551. CGP 20712A does not detect a corresponding high-affinity binding site in wild type 2AR where glutamine is present. The significance of the binding site remains to be determined by further experimentation.

An important property of 1LAR experimental pharmacology was that agonist responses to (-)-CGP 12177, (-)-pindolol and (-)-cyanopindolol were unaffected by -blockers such as 200 nM (-)-propranolol and 2 M nadolol (Joseph et al., 2004b; Joseph et al., 2003; Sarsero et al., 2003). Indeed many functional studies, frequently in heart, have been carried out utilising a basal condition of -blockade. Another key

Chapter 5 133

observation made in cell lines expressing 1ARs (Pak et al., 1996) and in in vitro human heart experiments (Joseph et al., 2003; Kaumann, 1996; Sarsero et al., 2003), was that low concentrations of (-)-CGP 12177 and (-)-pindolol blocked the receptor and then activated it at higher concentrations. This raises the question whether the receptor can accommodate two molecules simultaneously, one that blocks 1HAR (200 nM (-)-propranolol, 200 nM nadolol, < 0.5 nM (-)-CGP 12177) and another that activates 1LAR (> 200 nM (-)-CGP 12177).

Results provided from the study by Baker et al., (2008) indicate that there is some overlap of 1HAR and 1LAR binding sites. In this instance, it is unlikely that

1AR can accommodate 2 molecules simultaneously. However in another study, Asp138 (TMDIII), which interacts with the -hydroxyl and protonated amine groups of 1AR ligands (Warne et al., 2011; Warne et al., 2008),was shown to be critical for binding to the 1HAR site but not the 1LAR site (Joseph et al., 2004a). This finding suggests that Asp138 is an obligatory binding partner for 1HAR but not

1LAR, and that there is a distinction between the two sites. A possible explanation for this discrepancy may arise from the different genetic variants used in each study. Baker et al., (2008) reported using the Gly389 variant, whilst the Joseph et al., (2004a) study used the Arg389 variant. Previous reports have demonstrated the reduced coupling efficiency to Gs protein of Gly389 receptors compared to Arg389 receptors (Joseph et al., 2004c; Mason et al., 1999; Small et al., 2003). Joseph et al.,

(2004c) found that in CHO cells expressing 1ARs at physiological densities, the maximum cAMP response caused by (-)-isoprenaline was reduced 97%, at Gly389 receptors compared to Arg389 receptors. Furthermore, the maximum (-)-CGP 12177 response at Gly389 receptors was 54% of the observed response at Arg389 receptors. The reduced coupling efficiency of the Gly389 variant, in addition to the reduced binding affinity caused by altering Asp138, may have diminished further, the activity of 1AR in the Baker et al., (2008) study. This claim requires further investigation.

Preliminary modelling data provided by our collaborators (Dr Andrew Lucke and Prof David Fairlie, UQ) suggests that 1HAR and 1LAR could exist as distinct binding sites (see Appendix C). This model has predicted the 1LAR binding site

134 Chapter 5

based on the assumption that the 1HAR site is first occupied by (-)-CGP 12177. This forms the bottom part of the potential second 1LAR site which is occupied by a second (-)-CGP 12177 molecule that interacts with the first (-)-CGP 12177 molecule and also with amino acids in extracellular loop 2 (see Appendix C). The validity of this model requires further investigation. This will involve targeting the extracellular loop 2 amino acid residues that are predicted to interact with the second (-)-CGP 12177 molecule.

Chapter 5 135

Chapter 6: Pharmacological Analysis of 5-[3- (tert-butylamino))2- hydroxypropoxy]1,3-dihydro-2H- benzimidazol-2-one

6.1 INTRODUCTION

Activation of 1AR can occur through two binding sites, 1HAR and 1LAR, resulting in activation of the Gs-protein-cyclic-AMP-protein kinase A pathway

(Kaumann et al., 2008). In human heart failure, chronic activation of 1AR leads to disease progression and cardiac remodelling (Bristow et al., 1986; Brodde, 1991;

Cohn et al., 1984). -Blockers antagonise the effects of catecholamines at 1AR through the 1H site and are used for the management of human heart failure. However some clinically used -blockers such as pindolol and bucindolol do not produce overall survival benefits in heart failure patients (Cruickshank, 2007). Studies have demonstrated that pindolol and bucindolol possess intrinsic sympathomimetic activity, mediated via activation of 1LAR (Bundkirchen et al., 2002; Kaumann et al., 2008; Pak et al., 1996). Consequently, it can be argued that activation of 1LAR could contribute to the progression of heart failure. Recent evidence supports this hypothesis and has revealed that chronic activation of 1LAR could lead to cardiac remodelling, dysfunction and heart failure (Molenaar et al., 2011; Tugiono et al., 2010). Therefore, the development of β-blockers that effectively block β1LAR could improve blockade of 1AR compared to blockade of

1HAR alone. The ability to antagonise the stimulant effects of β1LAR may provide a novel therapeutic approach for the management of heart failure. This strategy will involve developing β1LAR blockers with decreased agonist activity for β1LAR, while retaining affinity for β1LAR.

It has been proposed that the unique pharmacology of -blockers that activate

1LAR including (-)-CGP 12177, (-)-pindolol and derivatives, is conferred by the presence of the benzimidazolone [(-)-CGP 12177] and indole [(-)-pindolol] groups

Chapter 6 137

(Kaumann et al., 2008). Inspection of the 2D structures shows that (-)-CGP 12177 and (-)-pindolol could create two separate conformational states of 1AR, one at

1HAR in which (-)-CGP 12177 and (-)-pindolol block the effects of (-)-noradrenaline but are unable to activate the receptor because the benzimidazolone/indole groups, in particular the polar nitrogen groups which are not optimally positioned to activate the receptor (Figure 6.1C). Another conformation is formed (1LAR) in which the nitrogens of (-)-CGP 12177 are aligned with the catechol group enabling activation of the receptor (Figure 6.1D). In order to test this hypothesis, the structural isomer of (-)-CGP 12177, 5-[3-(tert-butylamino))2- hydroxypropoxy]1,3-dihydro-2H-benzimidazol-2-one (Figure 6.2) was synthesised. In this orientation, it was hypothesised that the positions of the heteroatoms would mimic that of (-)-noradrenaline to establish favourable hydrogen bonding with

1HAR but not 1LAR.

A O B NH OH OH CH HN O NH 3 HO CH H C 3 3 NH2 HO CGP 12177 Noradrenaline C O D O NH OH OH NHOH OH CH3 HN CH3 HHON O NH HO O NH

CH3 CH3 H C OH H3C NH2 3 HO

H2N Figure 6.1. Two possible alignments (C,D) of the common features of noradrenaline and (-)-CGP 12177. It can be noted that the heteroatoms on the aromatic rings (catechol, blue and urea, red) cannot be aligned without twisting the sidechains from alignment.

O NH OH H O N OH CH HN O NH 3 * * NH CH3 O NH H3C CH3

H3C (-)-CGP 12177 CH3 4-[3-(tert-butylamino)-2-hydroxypropoxy]-1,3-dihydro-2H-benzimidazol-2-one 5-[3-(tert-butylamino)-2-hydroxypropoxy]-1,3-dihydro-2H-benzimidazol-2-one

Figure 6.2. Chemical structure of (-)-CGP 12177 and its structural isomer 5-[3-(tert- butylamino))2-hydroxypropoxy]1,3-dihydro-2H-benzimidazol-2-one.

138 Chapter 6

The pharmacological properties of 5-[3-(tert-butylamino))2- hydroxypropoxy]1,3-dihydro-2H-benzimidazol-2-one at 1HAR and 1LAR were analysed using radioligand binding and cAMP experiments. The agonist activity of 5-[3-(tert-butylamino))2-hydroxypropoxy]1,3-dihydro-2H-benzimidazol-2-one at

1LAR was determined using cAMP experiments. The affinity of 5-[3-(tert- butylamino))2-hydroxypropoxy]1,3-dihydro-2H-benzimidazol-2-one was determined 3 against (-)-[ H]-CGP 12177 (1HAR), (-)-isoprenaline (1HAR) and (-)-CGP 12177

(1LAR).

6.2 RESULTS

6.2.1 Competition Binding A competition binding experiment was performed using CHOAA8 cell membranes expressing 1AR. Membranes were labelled with a final concentration of 0.5 nM (-)-[3H]-CGP 12177 in the absence or presence of 5-[3-(tert-butylamino))2- hydroxypropoxy]1,3-dihydro-2H-benzimidazol-2-one at final concentrations of 1 nM, 100 nM and 1 M. At concentrations of 1 nM and 100 nM, 5-[3-(tert- butylamino))2-hydroxypropoxy]1,3-dihydro-2H-benzimidazol-2-one did not 3 compete with (-)-[ H]-CGP 12177 at 1HAR, while 1 M 5-[3-(tert-butylamino))2- hydroxypropoxy]1,3-dihydro-2H-benzimidazol-2-one caused ~10 % inhibition of 3 (-)-[ H]-CGP 12177 binding to 1HAR (data not shown).

6.2.2 Cyclic AMP Experiments In order to determine the agonist properties of 5-[3-(tert-butylamino))2- hydroxypropoxy]1,3-dihydro-2H-benzimidazol-2-one at 1AR, a concentration- effect experiment (200 pM – 60 mM) was performed using CHOAA8 cells expressing 1AR. Concentrations within this range did not cause an increase in cAMP (Figure 6.3).

Chapter 6 139

 1AR 10

8 Basal 6

4 cyclic AMP cyclic 2 (pmol/mg protein) 0

16 15 14 13 12 11 10 9 8 7 6 5 4 3 -Log [Compound] M

Figure 6.3. Concentration-effect relationship for cAMP accumulation in CHOAA8 cells stably expressing 1AR in response to 5-[3-(tert-butylamino))2- hydroxypropoxy]1,3-dihydro-2H-benzimidazol-2-one (n = 1).

The affinity of 5-[3-(tert-butylamino))2-hydroxypropoxy]1,3-dihydro-2H- benzimidazol-2-one (1 M) was measured against (-)-isoprenaline (1HAR) and

(-)-CGP 12177 (1LAR). A ~0.3 log unit rightward shift in the concentration-effect curve of (-)-isoprenaline was observed (Figure 6.4). At a concentration of 10 M, 5-[3-(tert-butylamino))2-hydroxypropoxy]1,3-dihydro-2H-benzimidazol-2-one did not antagonise the effects of (-)-CGP 12177 (Figure 6.5).

140 Chapter 6

 1AR 50

cAMP 20

10 (pmol/mg protein) (pmol/mg 0

12 1197510 8 6 4 [-(-)Isoprenaline]M

Figure 6.4. Concentration-effect curves for cAMP accumulation in CHOAA8 cells stably expressing 1AR in response to (-)-isoprenaline in the absence (●) and presence (○) of 1 M 5-[3-(tert-butylamino))2-hydroxypropoxy]1,3-dihydro-2H- benzimidazol-2-one (n = 1).

 1AR 15

10 Basal

5 cyclic AMP (pmol/mg protein) (pmol/mg 0

11 10 9 8 7 6 5 4 [(-)-CGP12177]M

Figure 6.5. Concentration-effect curves for cAMP accumulation in CHOAA8 cells stably expressing 1AR in response to (-)-CGP 12177 in the absence (●) and presence (○) of 10 M 5-[3-(tert-butylamino))2-hydroxypropoxy]1,3-dihydro-2H- benzimidazol-2-one (n = 1).

Chapter 6 141

6.3 DISCUSSION

Human heart failure is characterised by increased sympathetic nervous system activity, noradrenaline spillover and activation of 1ARs (Bristow et al., 1986;

Brodde, 1991; Cohn et al., 1984). Chronic activation of 1ARs causes progression of human heart failure resulting in remodelling, worsening haemodynamic function, morbidity and mortality (Bristow, 2000; Brodde et al., 1995). Primarily, research has focused on 1HAR and has established its involvement in human heart failure

(Bristow, 2000; Brodde, 2008). Until recently, chronic activation of 1LAR in heart failure was unknown. Recent studies have been performed in mice to study the effects of chronic infusion of (-)-CGP 12177 (0.01 mg – 100 mg/kg/24 hours via osmotic minipump) for 2 or 4 weeks (Molenaar et al., 2011; Tugiono et al., 2010). Results revealed a dose-dependent increase in heart rate and cardiac contraction measured as fractional shortening which was maintained for 4 weeks (Molenaar et al., 2011; Tugiono et al., 2010). In mice with trans-aortic constriction for a period of

8 weeks, chronic activation of β1LAR with (-)-CGP 12177 during weeks 5 to 8 caused a more severe cardiac hypertrophy, interstitial fibrosis and inflammation compared to trans-aortic constriction alone, indicating more severe myocardial remodelling (Molenaar et al., 2011; Tugiono et al., 2010). These results, in addition to those demonstrating the activation of 1LAR mediates cardiostimulant effects through the Gs-protein-cyclic-AMP-protein kinase A pathway in human atrium

(Kaumann et al., 2008), suggest that increased 1LAR activity could be potentially harmful and cause progression of heart failure.

The aim of the experiments in this section was to obtain preliminary data pertaining to the chemical features of ligands that activate 1LAR, with the long term aim being to develop a -blocker for use in heart failure that can block 1LAR with higher affinity (pK ~ 9) than those currently available (current pK  7). The structural isomer of (-)-CGP 12177, 5-[3-(tert-butylamino))2-hydroxypropoxy]1,3- dihydro-2H-benzimidazol-2-one was synthesised and tested to determine if the heteroatoms (urea) would mimic those of (-)-noradrenaline (catechol) (Figure 6.1). It was hypothesised that this compound would interact favourably with 1HAR but not

1LAR due to twisting of the backbone and repositioning of the heteroatoms.

142 Chapter 6

Analysis of results obtained from competition binding and cAMP experiments using 5-[3-(tert-butylamino))2-hydroxypropoxy]1,3-dihydro-2H-benzimidazol-2-one at 1AR demonstrated that this compound was inactive at 1HAR and 1LAR. These results indicate that substitution of the benzimidazolone group is critical for the binding and activity of (-)-CGP 12177 at 1LAR.

To elucidate why the benzimidazolone substitution determines the activity of

(-)-CGP 12177 at 1AR, one must consider the effect on the ligand-receptor interactions caused by modifying the substitution. Common interactions between ligands and β1AR include hydrogen bond formation between the aromatic head group and TMDV and TMDVI, in addition to interactions between the -hydroxyl group and protonated amine with TMDIII and TMDVII (Warne et al., 2011). The strength of the interactions formed within these regions appears to contribute to the efficacy of the ligand (Warne et al., 2011). From the turkey β1AR-m23 crystal structure containing (-)-isoprenaline, the main receptor-ligand interactions occur between the aromatic hydroxyl groups and Ser211 and Ser215 in TMDV and Asn310 in TMDVI and the -hydroxyl group and protonated amine which interact with Asn329 in TMDVII and Asp121 in TMDIII (analogous to Ser228, Ser232, Asn344, Asn363 and

Asp138 in human β1AR) (Warne et al., 2011). Substitution of analogous human amino acids to alanine in the β1AR reveals that Ser228 and Asn344 are involved in

(-)-CGP 12177 binding at β1AR and that Asp138 and Asn363 are required for binding and agonist activity (Baker et al., 2008). A model of the binding of

(-)-CGP 12177 to β1AR proposed by Baker et al., (2008) suggests that a nitrogen of the benzimidazolone group of (-)-CGP 12177 could face the serine residues of TMDV and bind with one of the serines. A similar interaction is observed between the indole nitrogen of cyanopindolol and Ser212 in the β1AR-m23 crystal structure (Warne et al., 2008).

For 5-[3-(tert-butylamino))2-hydroxypropoxy]1,3-dihydro-2H-benzimidazol-2- one, the secondary amine and -hydroxyl group are further away from the benzimidazolone group and the angle between the heteroatoms and the backbone is altered compared to (-)-CGP 12177. These changes may have disrupted the

Chapter 6 143

formation of interactions between the secondary amine and -hydroxyl with Asp138. Furthermore, the altered position of the heteroatoms could be expected to affect how they align with TMDV. The loss of β1AR specific activity produced by substitution of the benzimidazolone group in 5-[3-(tert-butylamino))2-hydroxypropoxy]1,3- dihydro-2H-benzimidazol-2-one is most likely due to disruption of the predicted Asp138 and/or TMDV interactions. This hypothesis is supported by previous studies which demonstrate the requirement of these interactions for binding and activity at the 1AR (Baker et al., 2008; Strader et al., 1989a; Strader et al., 1987b; Warne et al., 2011).

This proposal, taken together with results reported by Baker et al., (2008), suggest that Asp138 in TMDIII, Asn363 in TMDVII and interactions between the benzimidazolone group and TMDV are required for (-)-CGP 12177 binding and activation of 1LAR. The loss of one of these interactions has a catastrophic effect on the ability of (-)-CGP 12177 to bind and activate β1AR. Preserving these key interactions while altering the activity of (-)-CGP 12177 at 1AR, could be achieved by using compounds that target the effect of substituents at the common nitrogen atom of the indole group of (-)-pindolol, and the benzimidazolone group of (-)-CGP 12177 at position 1 (Figure 2.10). In theory, the modified substituents at this position would alter the interactions formed with TMDV. This principle is endorsed by β1AR- m23 crystal structures containing full and partial agonists, which reveal that interactions formed with TMDV, are a key determinant of ligand efficacy (Warne et al., 2011).

144 Chapter 6

Chapter 7: General Discussion and Future Directions

The present study provides strong evidence against the hypothesis that 3AR activation increases human atrial contractility. Furthermore, the results are in agreement with previous studies which claim that (-)-CGP 12177 and other non- conventional partial agonists cause increases in atrial contractility through activation of 1LAR. In addition, analysis of the heterologous 1AR, 2AR TMDV amino acids has identified Val230 as a contributing residue to the 1LAR binding site.

Results from Chapter 4 demonstrate that increases in human atrial force by

BRL 37344 and (-)-CGP 12177 are mediated through 2AR>1AR and 1LAR respectively. The 3AR agonist SR 58611 did not increase atrial force, in agreement with the lack of cardiostimulation in human heart by several 3AR-selective agonists reported in the literature. Small 3AR-mediated increases in ICa-L by BRL 37344 and SR58611 become apparent only at low, non-physiological temperatures but appear uncoupled from contractility. The inotropic and ICa-L responses to (-)-CGP 12177 are mediated through the low-affinity site 1LAR of the β1AR at both 24ºC and 37ºC.

The lack of human atrial responses through 3AR, suggests that therapeutically beneficial 3AR agonists do not pose a risk to cardiac function.

The results presented in Chapter 5 show that substitution of the heterologous

2AR TMDV amino acids into 1AR, disrupts formation of 1LAR. The potency of (-)-CGP 12177 for increases in cAMP and the affinity of (-)-bupranolol determined against (-)-CGP 12177, were increased significantly at the 1/2TMDVAR chimera compared to 1AR. Subsequent analysis of individual amino acids within this region revealed that these effects were largely attributable to substitution of Val230 to isoleucine. The effects of this substitution may arise from the larger isoleucine side chain interfering with the movement of TMDV during receptor activation. Considering these data, it can be concluded that amino acids of TMDV contribute

Chapter 7 145

and are required for the 1LAR binding site and that Val230 is primarily responsible.

Conversely, the presence of I205 in 2AR may play an inhibitory role in the formation of an analogous 2LAR.

Using cAMP experiments to analyse CHO cell lines expressing wild-type and mutant ARs in the physiological range for 1AR and 2AR found in human heart, provided a robust method of measuring AR activation. Previous studies have reported that overexpression of ARs and other GPCRs in both in vitro and in vivo systems can alter their pharmacology (Baker et al., 2002; Heubach et al., 2003a; Kaumann et al., 2008; Kohout et al., 2001; Pak et al., 1996; Rohrer et al., 1998; Wellner-Kienitz et al., 2001). Consequently, the use of cell lines containing physiological receptor densities was a critical part of the research design. This experimentation is time consuming in that it involved selecting and screening a large number of clones for each cell line. A large number of clones were selected (between 30-80 clones per cell line) to ensure that a clone expressing receptors at physiological densities could be identified from each stable transfection. Clone selection was optimised during the course of the study by changing the expression vector from pBI-L to pTRE2hyg. The main drawback of pBI-L was the requirement for co- transfection with a linear selection marker for hygromycin. This method of transfection produced a high number of false positive clones containing only the linear marker. The use of pTRE2hyg, which contains the hygromycin resistance gene, abolished false positives and ensured expression of the gene of interest.

Large numbers of clones required screening using single point radioligand binding experiments. This process could be optimised by developing a single point assay in a 96-well plate format, in order to facilitate a high-throughput screening approach. This would involve seeding a 96-well plate with cells from each colony in the first passage after transfer to 10 cm2 plates. Cells would need to be seeded and grown overnight to achieve ~80% confluency. The 96-well plate could then be used for a single point binding experiment. These experiments would require a 96-well plate cell harvester and a microplate scintillation counter. This format would increase the screening number compared to the method described in this thesis.

146 Chapter 7

Using cAMP assays to measure AR activation requires considerably more time compared to other methods, including those that use gene reporter systems. Previous studies have measured cAMP response element (CRE)-dependent secreted placental alkaline phosphatise (SPAP) reporter activity to quantify AR activation using transiently transfected cell lines (Baker, 2005b; Baker et al., 2003; Baker et al., 2002; Baker et al., 2008). Although gene reporter assays are quicker to perform compared to cAMP assays, which require stably transfected cell lines, there are some limitations. Assays that measure CRE-dependent SPAP reporter activity, require incubation periods of 4-6 hr with agonists for reporter gene expression (Hill et al., 2001). This can lead to receptor densensitisation and alter potency values (Hill et al., 2001). Furthermore, reporter expression is the last step in the Gs-protein-cyclic- AMP-protein kinase A pathway. This increases the potential for external factors to contribute to the magnitude of the end signal (Hill et al., 2001). In comparison, cAMP assays measure increases in cAMP. The conversion of ATP to cAMP mediated by adenylyl cyclase is an initial step in the Gs-protein-cyclic-AMP- protein kinase A pathway. This minimises the potential for interference to the cAMP measurement. Additionally, the incubation period with agonist used for cAMP assays is 20 min, which significantly reduces the possibility of receptor densensitisation compared to assays for CRE-dependent SPAP reporter activity. Nevertheless, the use of CRE-dependent SPAP reporter assays in 24- or 96-well plate format could provide a useful tool to screen large numbers stable tranfectants, similarly to a 96-well plate single point binding experiment.

The recent publication of the 1AR-m23 crystal structure has revealed that the binding pocket (1HAR) of cyanopindolol is formed by 15 amino acid side chains from four different TMDs (III, IV, V and VII) and ECL2. Heterologous amino acids in other regions of this 1AR crystal structure that are in close proximity to the binding pocket could be targeted alone and in combination with TMDV, in order to further characterise the 1LAR binding site. The outlined time constraints limited the potential for this study to investigate other regions of the 1AR and determine their role in 1LAR pharmacology. The implementation of the discussed optimised

Chapter 7 147

procedures, in addition to further optimisation may help to increase the productivity of other studies aimed at investigating other regions of the 1AR.

There is also the potential to extend this study to the 2AR in order to investigate the 1LAR binding site. This research strategy would involve incorporating heterologous 1AR TMD residues into the 2AR. In this instance, the research aim would be to generate a mutant 2AR that displayed pharmacology analogous to 1LAR. Because the 2AR does not form an analogous 2LAR, generating a 2AR that displayed low-affinity agonist activity for (-)-CGP 12177, could uncover amino acids that are critical for 1LAR.

Considering the recent advances in methods used to solve the crystal structures of ARs, the possibility exists where a more complete description of the 1LAR binding site may be provided by 1AR crystal structures that contain two ligands bound to one receptor. This scenario will rely fundamentally on whether the 1AR can accommodate two ligands simultaneously. The implementation of molecular modelling has begun to uncover structural information related to this question. A putative molecular model has identified a potential secondary low-affinity site that could reflect the 1LAR binding site. The secondary site is formed when the 1HAR site is first occupied by (-)-CGP 12177. This allows a second (-)-CGP 12177 molecule to bind and interact with the first (-)-CGP 12177 molecule and amino acids within ECL2. Further experimentation is required to investigate the interactions that are predicted to occur between the second (-)-CGP 12177 molecule and amino acids within ECL2.

This research project was implemented with a long term aim to develop a - blocker that can block 1LAR with higher affinity (pK ~ 9) than those currently available (current pK  7) for therapeutic use in heart failure. The potential for a - blocker of this nature to provide clinical benefits, has recently been endorsed by evidence from studies in mice demonstrating that cardiac 1LAR provides functional support to the heart and that chronic activation could promote the progression of

148 Chapter 7

heart failure. In order to obtain preliminary data regarding the chemical properties of compounds that activate 1LAR, the functionality of the structural isomer of (-)-CGP 12177, 5-[3-(tert-butylamino))2-hydroxypropoxy]1,3-dihydro-2H- benzimidazol-2-one at 1AR was determined (Chapter 6). This compound exhibited no activity at 1LAR, hence it appears that substitution of the benzimidazolone group of (-)-CGP 12177 is critical for agonist activity, possibly due to interactions formed between the benzimidazolone group and TMDV. Additional experiments are required using compounds with modified substituents at the common nitrogen atom of the indole group of (-)-pindolol and pindolol derivatives and the benzimidazolone group ((-)-CGP 12177) at position 1 to determine their effect on the TMDV interactions. Chemical manipulation of the common nitrogen atom could lead to the development of compounds with reduced agonist potency for 1LAR, while retaining affinity for 1LAR.

Chapter 7 149

Bibliography

Ablad B, Borg KO, Carlsson E, Ek L, Johnson G, Malmfors T, et al. (1975). A survey of the pharmacological properties of metoprolol in animals and man. Acta Pharmacol Toxicol (Copenh) 36(Suppl 5): 7-23.

Ablad B, Carlsson B, Carlsson E, Dahlof C, Ek L, Hultberg E (1974). Cardiac effects of beta-adrenergic receptor antagonists. Adv Cardiol 12(0): 290-302.

Ahlquist RP (1948). A study of the adrenotropic receptors. Amer. J. Physiol. 153: 586-600.

Arch JR, Kaumann AJ (1993). Beta 3 and atypical beta-adrenoceptors. Med Res Rev 13(6): 663-729.

Asano T, Pedersen SE, Scott CW, Ross EM (1984). Reconstitution of catecholamine- stimulated binding of guanosine 5'-O-(3-thiotriphosphate) to the stimulatory GTP- binding protein of adenylate cyclase. Biochemistry 23(23): 5460-5467.

Audet M, Bouvier M (2008). Insights into signaling from the beta2-adrenergic receptor structure. Nat Chem Biol 4(7): 397-403.

Avlani VA, Gregory KJ, Morton CJ, Parker MW, Sexton PM, Christopoulos A (2007). Critical role for the second extracellular loop in the binding of both orthosteric and allosteric G protein-coupled receptor ligands. J Biol Chem 282(35): 25677-25686.

Baker JG (2005a). The selectivity of beta-adrenoceptor antagonists at the human beta1, beta2 and beta3 adrenoceptors. Br J Pharmacol 144(3): 317-322.

Baker JG (2005b). Site of action of beta-ligands at the human beta1-adrenoceptor. J Pharmacol Exp Ther 313(3): 1163-1171.

Baker JG, Hall IP, Hill SJ (2003). Agonist actions of "beta-blockers" provide evidence for two agonist activation sites or conformations of the human beta1- adrenoceptor. Mol Pharmacol 63(6): 1312-1321.

Baker JG, Hall IP, Hill SJ (2002). Pharmacological characterization of CGP 12177 at the human beta(2)-adrenoceptor. Br J Pharmacol 137(3): 400-408.

Bibliography 151

Baker JG, Proudman RG, Hawley NC, Fischer PM, Hill SJ (2008). Role of key transmembrane residues in agonist and antagonist actions at the two conformations of the human beta1-adrenoceptor. Mol Pharmacol 74(5): 1246-1260.

Ballesteros JA, Jensen AD, Liapakis G, Rasmussen SG, Shi L, Gether U, et al. (2001). Activation of the beta 2-adrenergic receptor involves disruption of an ionic lock between the cytoplasmic ends of transmembrane segments 3 and 6. J Biol Chem 276(31): 29171-29177.

Berlan M, Dang Tran L (1978). [The role of beta and alpha adrenergic receptors in the lipolytic effect of catecholamines on dog adipocytes (author's transl)]. J Physiol (Paris) 74(6): 601-608.

Bers DM (2002). Cardiac excitation-contraction coupling. Nature 415(6868): 198- 205.

Bianchetti A, Manara L (1990). In vitro inhibition of intestinal motility by phenylethanolaminotetralines: evidence of atypical beta-adrenoceptors in rat colon. Br J Pharmacol 100(4): 831-839.

Birnbaumer L, Codina J, Mattera R, Yatani A, Scherer N, Toro MJ, et al. (1987). Signal transduction by G proteins. Kidney Int Suppl 23: S14-42.

Black JW, Duncan WA, Shanks RG (1965). Comparison of some properties of pronethalol and propranolol. Br J Pharmacol Chemother 25(3): 577-591.

Blayney LM, Lai FA (2009). Ryanodine receptor-mediated arrhythmias and sudden cardiac death. Pharmacol Ther 123(2): 151-177.

Blin N, Camoin L, Maigret B, Strosberg AD (1993). Structural and conformational features determining selective signal transduction in the beta 3-adrenergic receptor. Mol Pharmacol 44(6): 1094-1104.

Blin N, Nahmias C, Drumare MF, Strosberg AD (1994). Mediation of most atypical effects by species homologues of the beta 3-adrenoceptor. Br J Pharmacol 112(3): 911-919.

Bokoch MP, Zou Y, Rasmussen SG, Liu CW, Nygaard R, Rosenbaum DM, et al. (2010). Ligand-specific regulation of the extracellular surface of a G-protein-coupled receptor. Nature 463(7277): 108-112.

152 Bibliography

Bond RA, Leff P, Johnson TD, Milano CA, Rockman HA, McMinn TR, et al. (1995). Physiological effects of inverse agonists in transgenic mice with myocardial overexpression of the beta 2-adrenoceptor. Nature 374(6519): 272-276.

Borjesson M, Magnusson Y, Hjalmarson A, Andersson B (2000). A novel polymorphism in the gene coding for the beta(1)-adrenergic receptor associated with survival in patients with heart failure. Eur Heart J 21(22): 1853-1858.

Bradford MM (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72: 248-254.

Brandt DR, Ross EM (1986). Catecholamine-stimulated GTPase cycle. Multiple sites of regulation by beta-adrenergic receptor and Mg2+ studied in reconstituted receptor- Gs vesicles. J Biol Chem 261(4): 1656-1664.

Bristow MR (2000). beta-adrenergic receptor blockade in chronic heart failure. Circulation 101(5): 558-569.

Bristow MR, Ginsburg R, Umans V, Fowler M, Minobe W, Rasmussen R, et al. (1986). Beta 1- and beta 2-adrenergic-receptor subpopulations in nonfailing and failing human ventricular myocardium: coupling of both receptor subtypes to muscle contraction and selective beta 1-receptor down-regulation in heart failure. Circ Res 59(3): 297-309.

Bristow MR, Hershberger RE, Port JD, Minobe W, Rasmussen R (1989). Beta 1- and beta 2-adrenergic receptor-mediated adenylate cyclase stimulation in nonfailing and failing human ventricular myocardium. Mol Pharmacol 35(3): 295-303.

Bristow MR, Larrabee P, Minobe W, Roden R, Skerl L, Klein J, et al. (1992). Receptor pharmacology of carvedilol in the human heart. J Cardiovasc Pharmacol 19 Suppl 1: S68-80.

Bristow MR, Roden RL, Lowes BD, Gilbert EM, Eichhorn EJ (1998). The role of third-generation beta-blocking agents in chronic heart failure. Clin Cardiol 21(12 Suppl 1): I3-13.

Brodde OE (2008). Beta-1 and beta-2 adrenoceptor polymorphisms: functional importance, impact on cardiovascular diseases and drug responses. Pharmacol Ther 117(1): 1-29.

Brodde OE (1991). Beta 1- and beta 2-adrenoceptors in the human heart: properties, function, and alterations in chronic heart failure. Pharmacol Rev 43(2): 203-242.

Bibliography 153

Brodde OE, Michel MC (1999). Adrenergic and muscarinic receptors in the human heart. Pharmacol Rev 51(4): 651-690.

Brodde OE, Michel MC, Zerkowski HR (1995). Signal transduction mechanisms controlling cardiac contractility and their alterations in chronic heart failure. Cardiovasc Res 30(4): 570-584.

Brodde OE, O'Hara N, Zerkowski HR, Rohm N (1984). Human cardiac beta- adrenoceptors: both beta 1- and beta 2-adrenoceptors are functionally coupled to the adenylate cyclase in right atrium. J Cardiovasc Pharmacol 6(6): 1184-1191.

Bukowiecki L, Follea N, Paradis A, Collet A (1980). Stereospecific stimulation of brown adipocyte respiration by catecholamines via beta 1-adrenoreceptors. Am J Physiol 238(6): E552-563.

Bundkirchen A, Brixius K, Bolck B, Schwinger RH (2002). Bucindolol exerts agonistic activity on the propranolol-insensitive state of beta1-adrenoceptors in human myocardium. J Pharmacol Exp Ther 300(3): 794-801.

Buxton BF, Jones CR, Molenaar P, Summers RJ (1987). Characterization and autoradiographic localization of beta-adrenoceptor subtypes in human cardiac tissues. Br J Pharmacol 92(2): 299-310.

Candelore MR, Deng L, Tota L, Guan XM, Amend A, Liu Y, et al. (1999). Potent and selective human beta(3)-adrenergic receptor antagonists. J Pharmacol Exp Ther 290(2): 649-655.

Cao W, Luttrell LM, Medvedev AV, Pierce KL, Daniel KW, Dixon TM, et al. (2000). Direct binding of activated c-Src to the beta 3-adrenergic receptor is required for MAP kinase activation. J Biol Chem 275(49): 38131-38134.

Carlsson E, Ablad B, Brandstrom A, Carlsson B (1972). Differentiated blockade of the chronotropic effects of various adrenergic stimuli in the cat heart. Life Sci 11: 953-958.

Carstairs JR, Nimmo AJ, Barnes PJ (1985). Autoradiographic visualization of beta- adrenoceptor subtypes in human lung. Am Rev Respir Dis 132(3): 541-547.

Chen L, Meyers D, Javorsky G, Burstow D, Lolekha P, Lucas M, et al. (2007). Arg389Gly-beta1-adrenergic receptors determine improvement in left ventricular systolic function in nonischemic cardiomyopathy patients with heart failure after chronic treatment with carvedilol. Pharmacogenet Genomics 17(11): 941-949.

154 Bibliography

Cheng Y, Prusoff WH (1973). Relationship between the inhibition constant (K1) and the concentration of inhibitor which causes 50 per cent inhibition (I50) of an enzymatic reaction. Biochem Pharmacol 22(23): 3099-3108.

Cherezov V, Rosenbaum DM, Hanson MA, Rasmussen SG, Thian FS, Kobilka TS, et al. (2007). High-resolution crystal structure of an engineered human beta2- adrenergic G protein-coupled receptor. Science 318(5854): 1258-1265.

Cheung AH, Huang RR, Strader CD (1992). Involvement of specific hydrophobic, but not hydrophilic, amino acids in the third intracellular loop of the beta-adrenergic receptor in the activation of Gs. Mol Pharmacol 41(6): 1061-1065.

Chidiac P, Hebert TE, Valiquette M, Dennis M, Bouvier M (1994). Inverse agonist activity of beta-adrenergic antagonists. Mol Pharmacol 45(3): 490-499.

Choe HW, Kim YJ, Park JH, Morizumi T, Pai EF, Krauss N, et al. (2011). Crystal structure of metarhodopsin II. Nature 471(7340): 651-655.

Christ T, Wettwer E, Dobrev D, Adolph E, Knaut M, Wallukat G, et al. (2001). Autoantibodies against the beta1 adrenoceptor from patients with dilated cardiomyopathy prolong action potential duration and enhance contractility in isolated cardiomyocytes. J Mol Cell Cardiol 33(8): 1515-1525.

Chung FZ, Wang CD, Potter PC, Venter JC, Fraser CM (1988). Site-directed mutagenesis and continuous expression of human beta-adrenergic receptors. Identification of a conserved aspartate residue involved in agonist binding and receptor activation. J Biol Chem 263(9): 4052-4055.

Cohen ML, Bloomquist W, Kriauciunas A, Shuker A, Calligaro D (1999). Aryl propanolamines: comparison of activity at human beta3 receptors, rat beta3 receptors and rat atrial receptors mediating tachycardia. Br J Pharmacol 126(4): 1018-1024.

Cohn JN, Levine TB, Olivari MT, Garberg V, Lura D, Francis GS, et al. (1984). Plasma norepinephrine as a guide to prognosis in patients with chronic congestive heart failure. N Engl J Med 311(13): 819-823.

Colucci WS, Packer M, Bristow MR, Gilbert EM, Cohn JN, Fowler MB, et al. (1996). Carvedilol inhibits clinical progression in patients with mild symptoms of

Bibliography 155

heart failure. US Carvedilol Heart Failure Study Group. Circulation 94(11): 2800- 2806.

Cruickshank JM (2007). Are we misunderstanding beta-blockers. Int J Cardiol 120(1): 10-27.

De Lean A, Stadel JM, Lefkowitz RJ (1980). A ternary complex model explains the agonist-specific binding properties of the adenylate cyclase-coupled beta-adrenergic receptor. J Biol Chem 255(15): 7108-7117.

Deupi X, Kobilka B (2007). Activation of G protein-coupled receptors. Adv Protein Chem 74: 137-166.

Dixon RA, Sigal IS, Rands E, Register RB, Candelore MR, Blake AD, et al. (1987). Ligand binding to the beta-adrenergic receptor involves its rhodopsin-like core. Nature 326(6108): 73-77.

Emorine LJ, Marullo S, Briend-Sutren MM, Patey G, Tate K, Delavier-Klutchko C, et al. (1989). Molecular characterization of the human beta 3-adrenergic receptor. Science 245(4922): 1118-1121.

Engelhardt S, Hein L, Dyachenkow V, Kranias EG, Isenberg G, Lohse MJ (2004). Altered calcium handling is critically involved in the cardiotoxic effects of chronic beta-adrenergic stimulation. Circulation 109(9): 1154-1160.

Epstein SE, Braunwald E (1966). The effect of beta adrenergic blockade on patterns of urinary sodium excretion. Studies in normal subjects and in patients with heart disease. Ann Intern Med 65(1): 20-27.

Fabiato A (1983). Calcium-induced release of calcium from the cardiac sarcoplasmic reticulum. Am J Physiol 245(1): C1-14.

Freestone NS, Heubach JF, Wettwer E, Ravens U, Brown D, Kaumann AJ (1999). Beta4-adrenoceptors are more effective than beta1-adrenoceptors in mediating arrhythmic Ca2+ transients in mouse ventricular myocytes. Naunyn Schmiedebergs Arch Pharmacol 360(4): 445-456.

Freshney RI (1994). Culture of animal cells: a manual of basic technique. 3rd edn. Wiley-Liss, Inc: New York.

Friedman J, Babu B, Clark RB (2002). Beta(2)-adrenergic receptor lacking the cyclic AMP-dependent protein kinase consensus sites fully activates extracellular signal-

156 Bibliography

regulated kinase 1/2 in human embryonic kidney 293 cells: lack of evidence for G(s)/G(i) switching. Mol Pharmacol 62(5): 1094-1102.

Frielle T, Collins S, Daniel KW, Caron MG, Lefkowitz RJ, Kobilka BK (1987). Cloning of the cDNA for the human beta 1-adrenergic receptor. Proc Natl Acad Sci U S A 84(22): 7920-7924.

Frielle T, Daniel KW, Caron MG, Lefkowitz RJ (1988). Structural basis of beta- adrenergic receptor subtype specificity studied with chimeric beta 1/beta 2- adrenergic receptors. Proc Natl Acad Sci U S A 85(24): 9494-9498.

Gaddum JH (1957). Theories of drug antagonism. Pharmacol Rev 9(2): 211-218.

Gage RM, Matveeva EA, Whiteheart SW, von Zastrow M (2005). Type I PDZ ligands are sufficient to promote rapid recycling of G Protein-coupled receptors independent of binding to N-ethylmaleimide-sensitive factor. J Biol Chem 280(5): 3305-3313.

Galandrin S, Bouvier M (2006). Distinct signaling profiles of beta1 and beta2 adrenergic receptor ligands toward adenylyl cyclase and mitogen-activated protein kinase reveals the pluridimensionality of efficacy. Mol Pharmacol 70(5): 1575-1584.

Gardner LA, Naren AP, Bahouth SW (2007). Assembly of an SAP97-AKAP79- cAMP-dependent protein kinase scaffold at the type 1 PSD-95/DLG/ZO1 motif of the human beta(1)-adrenergic receptor generates a receptosome involved in receptor recycling and networking. J Biol Chem 282(7): 5085-5099.

Gauthier C, Tavernier G, Charpentier F, Langin D, Le Marec H (1996). Functional beta3-adrenoceptor in the human heart. J Clin Invest 98(2): 556-562.

Gerhardt CC, Gros J, Strosberg AD, Issad T (1999). Stimulation of the extracellular signal-regulated kinase 1/2 pathway by human beta-3 adrenergic receptor: new pharmacological profile and mechanism of activation. Mol Pharmacol 55(2): 255- 262.

Germack R, Starzec A, Perret GY (2000). Regulation of beta 1- and beta 3- adrenergic agonist-stimulated lipolytic response in hyperthyroid and hypothyroid rat white adipocytes. Br J Pharmacol 129(3): 448-456.

Gibbs ME, Summers RJ (2005). Contrasting roles for beta1, beta2 and beta3- adrenoceptors in memory formation in the chick. Neuroscience 131(1): 31-42.

Bibliography 157

Gilbert EM, Abraham WT, Olsen S, Hattler B, White M, Mealy P, et al. (1996). Comparative hemodynamic, left ventricular functional, and antiadrenergic effects of chronic treatment with metoprolol versus carvedilol in the failing heart. Circulation 94(11): 2817-2825.

Gilbert EM, Anderson, J.L., Deitchman, D., Yanowitz, FG., O’Connell, J.B.,, Renlund DG, Bartholomew, M., Mealey, P.C., Larrabee, P., Bristow, M.R. (1990). Chronic b-blocker-vasodilator therapy improves cardiac function in idiopathic dilated cardiomyopathy: a double-blind, randomized study of bucindolol versus placebo. Am J Med. 88: 223–229.

Gille E, Lemoine H, Ehle B, Kaumann AJ (1985). The affinity of (-)-propranolol for beta 1- and beta 2-adrenoceptors of human heart. Differential antagonism of the positive inotropic effects and adenylate cyclase stimulation by (-)-noradrenaline and (-)-adrenaline. Naunyn Schmiedebergs Arch Pharmacol 331(1): 60-70.

Gilman AG (1987). G proteins: transducers of receptor-generated signals. Annu Rev Biochem 56: 615-649.

Golf S, Lovstad R, Hansson V (1985). Beta-adrenoceptor density and relative number of beta-adrenoceptor subtypes in biopsies from human right atrial, left ventricular, and right ventricular myocard. Cardiovasc Res 19(10): 636-641.

Gong H, Adamson DL, Ranu HK, Koch WJ, Heubach JF, Ravens U, et al. (2000). The effect of Gi-protein inactivation on basal, and beta(1)- and beta(2)AR-stimulated contraction of myocytes from transgenic mice overexpressing the beta(2)- adrenoceptor. Br J Pharmacol 131(3): 594-600.

Gotze K, Jakobs KH (1994). Unoccupied beta-adrenoceptor-induced adenylyl cyclase stimulation in turkey erythrocyte membranes. Eur J Pharmacol 268(2): 151- 158.

Granneman JG, Lahners KN (1992). Differential adrenergic regulation of beta 1- and beta 3-adrenoreceptor messenger ribonucleic acids in adipose tissues. Endocrinology 130(1): 109-114.

Green SA, Holt BD, Liggett SB (1992). Beta 1- and beta 2-adrenergic receptors display subtype-selective coupling to Gs. Mol Pharmacol 41(5): 889-893.

Green SA, Liggett SB (1994). A proline-rich region of the third intracellular loop imparts phenotypic beta 1-versus beta 2-adrenergic receptor coupling and sequestration. J Biol Chem 269(42): 26215-26219.

158 Bibliography

Hakalahti AE, Vierimaa MM, Lilja MK, Kumpula EP, Tuusa JT, Petaja-Repo UE (2010). Human beta1-adrenergic receptor is subject to constitutive and regulated N- terminal cleavage. J Biol Chem 285(37): 28850-28861.

Hall JA, Kaumann AJ, Brown MJ (1990). Selective beta 1-adrenoceptor blockade enhances positive inotropic responses to endogenous catecholamines mediated through beta 2-adrenoceptors in human atrial myocardium. Circ Res 66(6): 1610- 1623.

Hanyaloglu AC, von Zastrow M (2008). Regulation of GPCRs by endocytic membrane trafficking and its potential implications. Annu Rev Pharmacol Toxicol 48: 537-568.

Hausdorff WP, Campbell PT, Ostrowski J, Yu SS, Caron MG, Lefkowitz RJ (1991). A small region of the beta-adrenergic receptor is selectively involved in its rapid regulation. Proc Natl Acad Sci U S A 88(8): 2979-2983.

Hausdorff WP, Hnatowich M, O'Dowd BF, Caron MG, Lefkowitz RJ (1990). A mutation of the beta 2-adrenergic receptor impairs agonist activation of adenylyl cyclase without affecting high affinity agonist binding. Distinct molecular determinants of the receptor are involved in physical coupling to and functional activation of Gs. J Biol Chem 265(3): 1388-1393.

He J, Bellini M, Inuzuka H, Xu J, Xiong Y, Yang X, et al. (2006). Proteomic analysis of beta1-adrenergic receptor interactions with PDZ scaffold proteins. J Biol Chem 281(5): 2820-2827.

He J, Xu J, Castleberry AM, Lau AG, Hall RA (2002). Glycosylation of beta(1)- adrenergic receptors regulates receptor surface expression and dimerization. Biochem Biophys Res Commun 297(3): 565-572.

Heitz A, Schwartz J, Velly J (1983). Beta-adrenoceptors of the human myocardium: determination of beta 1 and beta 2 subtypes by radioligand binding. Br J Pharmacol 80(4): 711-717.

Hekman M, Holzhofer A, Gierschik P, Im MJ, Jakobs KH, Pfeuffer T, et al. (1987). Regulation of signal transfer from beta 1-adrenoceptor to adenylate cyclase by beta gamma subunits in a reconstituted system. Eur J Biochem 169(2): 431-439.

Heubach JF, Blaschke M, Harding SE, Ravens U, Kaumann AJ (2003a). Cardiostimulant and cardiodepressant effects through overexpressed human beta2- adrenoceptors in murine heart: regional differences and functional role of beta1- adrenoceptors. Naunyn-Schmiedeberg's archives of pharmacology 367(4): 380-390.

Bibliography 159

Heubach JF, Blaschke M, Harding SE, Ravens U, Kaumann AJ (2003b). Cardiostimulant and cardiodepressant effects through overexpressed human beta2- adrenoceptors in murine heart: regional differences and functional role of beta1- adrenoceptors. Naunyn Schmiedebergs Arch Pharmacol 367(4): 380-390.

Hill SJ, Baker JG, Rees S (2001). Reporter-gene systems for the study of G-protein- coupled receptors. Curr Opin Pharmacol 1(5): 526-532.

Hoffmann C, Leitz MR, Oberdorf-Maass S, Lohse MJ, Klotz KN (2004). Comparative pharmacology of human beta-adrenergic receptor subtypes-- characterization of stably transfected receptors in CHO cells. Naunyn Schmiedebergs Arch Pharmacol 369(2): 151-159.

Hoffmann C, Zurn A, Bunemann M, Lohse MJ (2008). Conformational changes in G-protein-coupled receptors-the quest for functionally selective conformations is open. Br J Pharmacol 153 Suppl 1: S358-366.

Hu LA, Tang Y, Miller WE, Cong M, Lau AG, Lefkowitz RJ, et al. (2000). beta 1- adrenergic receptor association with PSD-95. Inhibition of receptor internalization and facilitation of beta 1-adrenergic receptor interaction with N-methyl-D-aspartate receptors. J Biol Chem 275(49): 38659-38666.

January B, Seibold A, Allal C, Whaley BS, Knoll BJ, Moore RH, et al. (1998). Salmeterol-induced desensitization, internalization and phosphorylation of the human beta2-adrenoceptor. Br J Pharmacol 123(4): 701-711.

January B, Seibold A, Whaley B, Hipkin RW, Lin D, Schonbrunn A, et al. (1997). beta2-adrenergic receptor desensitization, internalization, and phosphorylation in response to full and partial agonists. J Biol Chem 272(38): 23871-23879.

Joseph SS, Colledge WH, Kaumann AJ (2004a). Aspartate138 is required for the high-affinity ligand binding site but not for the low-affinity binding site of the beta1- adrenoceptor. Naunyn Schmiedebergs Arch Pharmacol 370(3): 223-226.

Joseph SS, Lynham JA, Colledge WH, Kaumann AJ (2004b). Binding of (-)-[3H]- CGP12177 at two sites in recombinant human beta 1-adrenoceptors and interaction with beta-blockers. Naunyn Schmiedebergs Arch Pharmacol 369(5): 525-532.

Joseph SS, Lynham JA, Grace AA, Colledge WH, Kaumann AJ (2004c). Markedly reduced effects of (-)-isoprenaline but not of (-)-CGP12177 and unchanged affinity of beta-blockers at Gly389-beta1-adrenoceptors compared to Arg389-beta1- adrenoceptors. Br J Pharmacol 142(1): 51-56.

160 Bibliography

Joseph SS, Lynham JA, Molenaar P, Grace AA, Colledge WH, Kaumann AJ (2003). Intrinsic sympathomimetic activity of (-)-pindolol mediated through a (-)- propranolol-resistant site of the beta1-adrenoceptor in human atrium and recombinant receptors. Naunyn Schmiedebergs Arch Pharmacol 368(6): 496-503.

Juberg EN, Minneman KP, Abel PW (1985). Beta 1- and beta 2-adrenoceptor binding and functional response in right and left atria of rat heart. Naunyn Schmiedebergs Arch Pharmacol 330(3): 193-202.

Katz AM (2003). Heart failure: a hemodynamic disorder complicated by maladaptive proliferative responses. J Cell Mol Med 7(1): 1-10.

Katz AM, Tada M, Kirchberger MA (1975). Control of calcium transport in the myocardium by the cyclic AMP-Protein kinase system. Adv Cyclic Nucleotide Res 5: 453-472.

Kaumann A, Bartel S, Molenaar P, Sanders L, Burrell K, Vetter D, et al. (1999). Activation of beta2-adrenergic receptors hastens relaxation and mediates phosphorylation of phospholamban, troponin I, and C-protein in ventricular myocardium from patients with terminal heart failure. Circulation 99(1): 65-72.

Kaumann A, Lynham J, Sarsero D, Molenaar P (1997a). The atypical cardiostimulant β-adrenoceptor is distinct from the β3-adrenoceptor and coupled to a cyclic AMP dependent pathway in rat and human myocardium. Br J Pharmacol 120: 102P.

Kaumann A, Semmler AB, Molenaar P (2007). The effects of both noradrenaline and CGP12177, mediated through human beta1 -adrenoceptors, are reduced by PDE3 in human atrium but PDE4 in CHO cells. Naunyn Schmiedebergs Arch Pharmacol 375(2): 123-131.

Kaumann AJ (1983). [Cardiac beta receptors--experimental aspects]. Z Kardiol 72(2): 63-82.

Kaumann AJ (1996). (-)-CGP 12177-induced increase of human atrial contraction through a putative third beta-adrenoceptor. Br J Pharmacol 117(1): 93-98.

Kaumann AJ (1989). Is there a third heart beta-adrenoceptor? Trends Pharmacol Sci 10(8): 316-320.

Kaumann AJ, Birnbaumer L (1973a). Proceedings: Adrenergic receptors in heart muscle: similarity of apparent affinities of beta-blockers for receptors mediating adenyl-cyclase activity, inotropic and chronotropic effects of catecholamines. Acta Physiol Lat Am 23(6): 619-620.

Bibliography 161

Kaumann AJ, Birnbaumer L (1973b). Proceedings: Adrenergic receptors in heart muscle: similarity of apparent affinities of beta-blockers for receptors mediating adenyl-cyclase activity, inotropic and chronotropic effects of catecholamines. Acta Physiol Lat Am 23(6): 619-620.

Kaumann AJ, Engelhardt S, Hein L, Molenaar P, Lohse M (2001). Abolition of (-)- CGP 12177-evoked cardiostimulation in double beta1/beta2-adrenoceptor knockout mice. Obligatory role of beta1-adrenoceptors for putative beta4-adrenoceptor pharmacology. Naunyn Schmiedebergs Arch Pharmacol 363(1): 87-93.

Kaumann AJ, Hall JA, Murray KJ, Wells FC, Brown MJ (1989). A comparison of the effects of adrenaline and noradrenaline on human heart: the role of beta 1- and beta 2-adrenoceptors in the stimulation of adenylate cyclase and contractile force. Eur Heart J 10 Suppl B: 29-37.

Kaumann AJ, Lemoine H (1987). Beta 2-adrenoceptor-mediated positive inotropic effect of adrenaline in human ventricular myocardium. Quantitative discrepancies with binding and adenylate cyclase stimulation. Naunyn Schmiedebergs Arch Pharmacol 335(4): 403-411.

Kaumann AJ, Lynham JA (1997b). Stimulation of cyclic AMP-dependent protein kinase in rat atria by (-)-CGP 12177 through an atypical beta-adrenoceptor. Br J Pharmacol 120(7): 1187-1189.

Kaumann AJ, Molenaar P (1996a). Differences between the third cardiac beta- adrenoceptor and the colonic beta 3-adrenoceptor in the rat. Br J Pharmacol 118(8): 2085-2098.

Kaumann AJ, Molenaar P (2008). The low-affinity site of the beta1-adrenoceptor and its relevance to cardiovascular pharmacology. Pharmacol Ther 118(3): 303-336.

Kaumann AJ, Molenaar P (1997c). Modulation of human cardiac function through 4 beta-adrenoceptor populations. Naunyn Schmiedebergs Arch Pharmacol 355(6): 667-681.

Kaumann AJ, Preitner F, Sarsero D, Molenaar P, Revelli JP, Giacobino JP (1998). (- )-CGP 12177 causes cardiostimulation and binds to cardiac putative beta 4- adrenoceptors in both wild-type and beta 3-adrenoceptor knockout mice. Mol Pharmacol 53(4): 670-675.

Kaumann AJ, Sanders L, Lynham JA, Bartel S, Kuschel M, Karczewski P, et al. (1996b). Beta 2-adrenoceptor activation by zinterol causes protein phosphorylation,

162 Bibliography

contractile effects and relaxant effects through a cAMP pathway in human atrium. Mol Cell Biochem 163-164: 113-123.

Kenakin T (2007). Functional selectivity through protean and biased agonism: who steers the ship? Mol Pharmacol 72(6): 1393-1401.

Kenakin T (2001). Inverse, protean, and ligand-selective agonism: matters of receptor conformation. FASEB J 15(3): 598-611.

Klco JM, Wiegand CB, Narzinski K, Baranski TJ (2005). Essential role for the second extracellular loop in C5a receptor activation. Nat Struct Mol Biol 12(4): 320- 326.

Kobilka BK (2007). G protein coupled receptor structure and activation. Biochim Biophys Acta 1768(4): 794-807.

Kobilka BK, Deupi X (2007). Conformational complexity of G-protein-coupled receptors. Trends Pharmacol Sci 28(8): 397-406.

Kohout TA, Takaoka H, McDonald PH, Perry SJ, Mao L, Lefkowitz RJ, et al. (2001). Augmentation of cardiac contractility mediated by the human beta(3)- adrenergic receptor overexpressed in the hearts of transgenic mice. Circulation 104(20): 2485-2491.

Kompa AR, Summers RJ (1999). Desensitization and resensitization of beta 1- and putative beta 4-adrenoceptor mediated responses occur in parallel in a rat model of cardiac failure. Br J Pharmacol 128(7): 1399-1406.

Konkar AA, Zhai Y, Granneman JG (2000a). beta1-adrenergic receptors mediate beta3-adrenergic-independent effects of CGP 12177 in brown adipose tissue. Mol Pharmacol 57(2): 252-258.

Konkar AA, Zhu Z, Granneman JG (2000b). Aryloxypropanolamine and catecholamine ligand interactions with the beta(1)-adrenergic receptor: evidence for interaction with distinct conformations of beta(1)-adrenergic receptors. J Pharmacol Exp Ther 294(3): 923-932.

Kurtz A, Wagner C (1999). Cellular control of renin secretion. J Exp Biol 202(Pt 3): 219-225.

Lacasa D, Mauriege P, Lafontan M, Berlan M, Giudicelli Y (1986). A reliable assay for beta-adrenoceptors in intact isolated human fat cells with a hydrophilic radioligand, [3H]CGP-12177. J Lipid Res 27(4): 368-376.

Bibliography 163

Lafontan M, Berlan M (1993). Fat cell adrenergic receptors and the control of white and brown fat cell function. J Lipid Res 34(7): 1057-1091.

Landry Y, Gies JP (2008). Drugs and their molecular targets: an updated overview. Fundam Clin Pharmacol 22(1): 1-18.

Lands AM, Arnold A, McAuliff JP, Luduena FP, Brown TG, Jr. (1967a). Differentiation of receptor systems activated by sympathomimetic amines. Nature 214(5088): 597-598.

Lands AM, Luduena FP, Buzzo HJ (1967b). Differentiation of receptors responsive to isoproterenol. Life Sci 6(21): 2241-2249.

Langin D, Portillo MP, Saulnier-Blache JS, Lafontan M (1991). Coexistence of three beta-adrenoceptor subtypes in white fat cells of various mammalian species. Eur J Pharmacol 199(3): 291-301.

Leach K, Sexton PM, Christopoulos A (2007). Allosteric GPCR modulators: taking advantage of permissive receptor pharmacology. Trends Pharmacol Sci 28(8): 382- 389.

Leblais V, Jo SH, Chakir K, Maltsev V, Zheng M, Crow MT, et al. (2004). Phosphatidylinositol 3-kinase offsets cAMP-mediated positive inotropic effect via inhibiting Ca2+ influx in cardiomyocytes. Circ Res 95(12): 1183-1190.

Lechat P, Escolano S, Golmard JL, Lardoux H, Witchitz S, Henneman JA, et al. (1997). Prognostic value of bisoprolol-induced hemodynamic effects in heart failure during the Cardiac Insufficiency BIsoprolol Study (CIBIS). Circulation 96(7): 2197- 2205.

Lefkowitz RJ, Cotecchia S, Samama P, Costa T (1993). Constitutive activity of receptors coupled to guanine nucleotide regulatory proteins. Trends Pharmacol Sci 14(8): 303-307.

Lefkowitz RJ, Pierce KL, Luttrell LM (2002). Dancing with different partners: protein kinase a phosphorylation of seven membrane-spanning receptors regulates their G protein-coupling specificity. Mol Pharmacol 62(5): 971-974.

Lemoine H, Kaumann AJ (1991). Regional differences of beta 1- and beta 2- adrenoceptor-mediated functions in feline heart. A beta 2-adrenoceptor-mediated positive inotropic effect possibly unrelated to cyclic AMP. Naunyn Schmiedebergs Arch Pharmacol 344(1): 56-69.

164 Bibliography

Lemoine H, Schonell H, Kaumann AJ (1988). Contribution of beta 1- and beta 2- adrenoceptors of human atrium and ventricle to the effects of noradrenaline and adrenaline as assessed with (-)-atenolol. Br J Pharmacol 95(1): 55-66.

Leonard BE (1997). The role of noradrenaline in depression: a review. J Psychopharmacol 11(4 Suppl): S39-47.

Levin BE, Sullivan AC (1986). Beta-1 receptor is the predominant beta- adrenoreceptor on rat brown adipose tissue. J Pharmacol Exp Ther 236(3): 681-688.

Levitzki A (1988). From epinephrine to cyclic AMP. Science 241(4867): 800-806.

Levy FO, Zhu X, Kaumann AJ, Birnbaumer L (1993). Efficacy of beta 1-adrenergic receptors is lower than that of beta 2-adrenergic receptors. Proc Natl Acad Sci U S A 90(22): 10798-10802.

Li L, Desantiago J, Chu G, Kranias EG, Bers DM (2000). Phosphorylation of phospholamban and troponin I in beta-adrenergic-induced acceleration of cardiac relaxation. Am J Physiol Heart Circ Physiol 278(3): H769-779.

Liapakis G, Ballesteros JA, Papachristou S, Chan WC, Chen X, Javitch JA (2000). The forgotten serine. A critical role for Ser-2035.42 in ligand binding to and activation of the beta 2-adrenergic receptor. J Biol Chem 275(48): 37779-37788.

Liapakis G, Chan WC, Papadokostaki M, Javitch JA (2004). Synergistic contributions of the functional groups of epinephrine to its affinity and efficacy at the beta2 adrenergic receptor. Mol Pharmacol 65(5): 1181-1190.

Liggett SB, Mialet-Perez J, Thaneemit-Chen S, Weber SA, Greene SM, Hodne D, et al. (2006). A polymorphism within a conserved beta(1)-adrenergic receptor motif alters cardiac function and beta-blocker response in human heart failure. Proc Natl Acad Sci U S A 103(30): 11288-11293.

Liggett SB, Shah SD, Cryer PE (1988). Characterization of beta-adrenergic receptors of human skeletal muscle obtained by needle biopsy. Am J Physiol 254(6 Pt 1): E795-798.

Lindquist JM, Fredriksson JM, Rehnmark S, Cannon B, Nedergaard J (2000). Beta 3- and alpha1-adrenergic Erk1/2 activation is Src- but not Gi-mediated in Brown adipocytes. J Biol Chem 275(30): 22670-22677.

Bibliography 165

Lowe MD, Grace AA, Vandenberg JI, Kaumann AJ (1998). Action potential shortening through the putative beta4-adrenoceptor in ferret ventricle: comparison with beta1- and beta2-adrenoceptor-mediated effects. Br J Pharmacol 124(7): 1341- 1344.

Lowe MD, Lynham JA, Grace AA, Kaumann AJ (2002). Comparison of the affinity of beta-blockers for two states of the beta 1-adrenoceptor in ferret ventricular myocardium. Br J Pharmacol 135(2): 451-461.

Luttrell LM, Ferguson SS, Daaka Y, Miller WE, Maudsley S, Della Rocca GJ, et al. (1999). Beta-arrestin-dependent formation of beta2 adrenergic receptor-Src protein kinase complexes. Science 283(5402): 655-661.

Machida CA, Bunzow JR, Searles RP, Van Tol H, Tester B, Neve KA, et al. (1990). Molecular cloning and expression of the rat beta 1-adrenergic receptor gene. J Biol Chem 265(22): 12960-12965.

Machida K, Inoue H, Matsumoto K, Tsuda M, Fukuyama S, Koto H, et al. (2005). Activation of PI3K-Akt pathway mediates antiapoptotic effects of beta-adrenergic agonist in airway eosinophils. Am J Physiol Lung Cell Mol Physiol 288(5): L860- 867.

Maqbool A, Hall AS, Ball SG, Balmforth AJ (1999). Common polymorphisms of beta1-adrenoceptor: identification and rapid screening assay. Lancet 353(9156): 897.

Marion S, Oakley RH, Kim KM, Caron MG, Barak LS (2006). A beta-arrestin binding determinant common to the second intracellular loops of rhodopsin family G protein-coupled receptors. J Biol Chem 281(5): 2932-2938.

Marx SO, Reiken S, Hisamatsu Y, Jayaraman T, Burkhoff D, Rosemblit N, et al. (2000). PKA phosphorylation dissociates FKBP12.6 from the calcium release channel (ryanodine receptor): defective regulation in failing hearts. Cell 101(4): 365- 376.

Mason DA, Moore JD, Green SA, Liggett SB (1999). A gain-of-function polymorphism in a G-protein coupling domain of the human beta1-adrenergic receptor. J Biol Chem 274(18): 12670-12674.

Mauriege P, De Pergola G, Berlan M, Lafontan M (1988). Human fat cell beta- adrenergic receptors: beta-agonist-dependent lipolytic responses and characterization of beta-adrenergic binding sites on human fat cell membranes with highly selective beta 1-antagonists. J Lipid Res 29(5): 587-601.

166 Bibliography

Mewes T, Dutz S, Ravens U, Jakobs KH (1993). Activation of calcium currents in cardiac myocytes by empty beta-adrenoceptors. Circulation 88(6): 2916-2922.

Mialet Perez J, Rathz DA, Petrashevskaya NN, Hahn HS, Wagoner LE, Schwartz A, et al. (2003). Beta 1-adrenergic receptor polymorphisms confer differential function and predisposition to heart failure. Nat Med 9(10): 1300-1305.

Minneman KP, Hedberg A, Molinoff PB (1979). Comparison of beta adrenergic receptor subtypes in mammalian tissues. J Pharmacol Exp Ther 211(3): 502-508.

Molenaar P, Bartel S, Cochrane A, Vetter D, Jalali H, Pohlner P, et al. (2000). Both beta(2)- and beta(1)-adrenergic receptors mediate hastened relaxation and phosphorylation of phospholamban and troponin I in ventricular myocardium of Fallot infants, consistent with selective coupling of beta(2)-adrenergic receptors to G(s)-protein. Circulation 102(15): 1814-1821.

Molenaar P, Chen L, Semmler AB, Parsonage WA, Kaumann AJ (2007a). Human heart beta-adrenoceptors: beta1-adrenoceptor diversification through 'affinity states' and polymorphism. Clin Exp Pharmacol Physiol 34(10): 1020-1028.

Molenaar P, Jones CR, McMartin LR, Summers RJ (1988). Autoradiographic localization and densitometric analysis of beta-1 and beta-2 adrenoceptors in the canine left anterior descending coronary artery. J Pharmacol Exp Ther 246(1): 384- 393.

Molenaar P, Klenowski P, Semmler AB, Chee K, Kaumann AJ, Kiriazis H, et al. (2011). Molecular determinants and functional consequences of the β1L- adrenoceptor in the heart. In: ASCEPT. Perth.

Molenaar P, McPherson GA, Malta E, Raper C (1985). Comparison of beta- adrenoceptor populations in cat and guinea-pig left atria. J Pharm Pharmacol 37(4): 257-260.

Molenaar P, Parsonage WA (2005). Fundamental considerations of beta- adrenoceptor subtypes in human heart failure. Trends Pharmacol Sci 26(7): 368-375.

Molenaar P, Rabnott G, Yang I, Fong KM, Savarimuthu SM, Li L, et al. (2002). Conservation of the cardiostimulant effects of (-)-norepinephrine across Ser49Gly and Gly389Arg beta(1)-adrenergic receptor polymorphisms in human right atrium in vitro. J Am Coll Cardiol 40(7): 1275-1282.

Molenaar P, Russell FD, Shimada T, Summers RJ (1989). Function, characterization and autoradiographic localization and quantitation of beta-adrenoceptors in cardiac tissues. Clin Exp Pharmacol Physiol 16(6): 529-533.

Bibliography 167

Molenaar P, Sarsero D, Kaumann AJ (1997). Proposal for the interaction of non- conventional partial agonists and catecholamines with the 'putative beta 4- adrenoceptor' in mammalian heart. Clin Exp Pharmacol Physiol 24(9-10): 647-656.

Molenaar P, Savarimuthu SM, Sarsero D, Chen L, Semmler AB, Carle A, et al. (2007b). (-)-Adrenaline elicits positive inotropic, lusitropic, and biochemical effects through beta2 -adrenoceptors in human atrial myocardium from nonfailing and failing hearts, consistent with Gs coupling but not with Gi coupling. Naunyn Schmiedebergs Arch Pharmacol 375(1): 11-28.

Molenaar P, Summers RJ (1987). Characterization of beta-1 and beta-2 adrenoceptors in guinea pig atrium: functional and receptor binding studies. J Pharmacol Exp Ther 241(3): 1041-1047.

Moniotte S, Kobzik L, Feron O, Trochu JN, Gauthier C, Balligand JL (2001a). Upregulation of beta(3)-adrenoceptors and altered contractile response to inotropic amines in human failing myocardium. Circulation 103(12): 1649-1655.

Moniotte S, Vaerman JL, Kockx MM, Larrouy D, Langin D, Noirhomme P, et al. (2001b). Real-time RT-PCR for the detection of beta-adrenoceptor messenger RNAs in small human endomyocardial biopsies. J Mol Cell Cardiol 33(12): 2121-2133.

Monod J, Wyman J, Changeux JP (1965). On the Nature of Allosteric Transitions: A Plausible Model. J Mol Biol 12: 88-118.

Moro O, Lameh J, Sadee W (1993). Serine- and threonine-rich domain regulates internalization of muscarinic cholinergic receptors. J Biol Chem 268(10): 6862-6865.

Morris TH, Sandrock K, Kaumann AJ (1981). 3H-(-)-Bupranolol, a new beta- adrenoceptor radioligand: characterization of its binding to kitten heart beta- adrenoceptors. Naunyn-Schmiedeberg's archives of pharmacology 317(1): 19-25.

Murchison CF, Zhang XY, Zhang WP, Ouyang M, Lee A, Thomas SA (2004). A distinct role for norepinephrine in memory retrieval. Cell 117(1): 131-143.

Napp A, Brixius K, Pott C, Ziskoven C, Boelck B, Mehlhorn U, et al. (2009). Effects of the beta3-adrenergic agonist BRL 37344 on endothelial nitric oxide synthase phosphorylation and force of contraction in human failing myocardium. J Card Fail 15(1): 57-67.

Newton GE, Azevedo ER, Parker JD (1999). Inotropic and sympathetic responses to the intracoronary infusion of a beta2-receptor agonist: a human in vivo study. Circulation 99(18): 2402-2407.

168 Bibliography

Newton GE, Parker JD (1996). Acute effects of beta 1-selective and nonselective beta-adrenergic receptor blockade on cardiac sympathetic activity in congestive heart failure. Circulation 94(3): 353-358.

O'Dowd BF, Hnatowich M, Regan JW, Leader WM, Caron MG, Lefkowitz RJ (1988). Site-directed mutagenesis of the cytoplasmic domains of the human beta 2- adrenergic receptor. Localization of regions involved in G protein-receptor coupling. The Journal of biological chemistry 263(31): 15985-15992.

Okamoto T, Murayama Y, Hayashi Y, Inagaki M, Ogata E, Nishimoto I (1991). Identification of a Gs activator region of the beta 2-adrenergic receptor that is autoregulated via protein kinase A-dependent phosphorylation. Cell 67(4): 723-730.

Oldham WM, Hamm HE (2008). Heterotrimeric G protein activation by G-protein- coupled receptors. Nat Rev Mol Cell Biol 9(1): 60-71.

Oostendorp J, Preitner F, Moffatt J, Jimenez M, Giacobino JP, Molenaar P, et al. (2000). Contribution of beta-adrenoceptor subtypes to relaxation of colon and oesophagus and pacemaker activity of ureter in wildtype and beta(3)-adrenoceptor knockout mice. Br J Pharmacol 130(4): 747-758.

Ota A, Matsui H, Asakura M, Nagatsu T (1993). Distribution of beta 1- and beta 2- adrenoceptor subtypes in various mouse tissues. Neurosci Lett 160(1): 96-100.

Overington JP, Al-Lazikani B, Hopkins AL (2006). How many drug targets are there? Nat Rev Drug Discov 5(12): 993-996.

Packer M (1998). Beta-blockade in heart failure. Basic concepts and clinical results. Am J Hypertens 11(1 Pt 2): 23S-37S.

Packer M (1992). The neurohormonal hypothesis: a theory to explain the mechanism of disease progression in heart failure. J Am Coll Cardiol 20(1): 248-254.

Packer M, Bristow MR, Cohn JN, Colucci WS, Fowler MB, Gilbert EM, et al. (1996). The effect of carvedilol on morbidity and mortality in patients with chronic heart failure. U.S. Carvedilol Heart Failure Study Group. N Engl J Med 334(21): 1349-1355.

Packer M, Carver JR, Rodeheffer RJ, Ivanhoe RJ, DiBianco R, Zeldis SM, et al. (1991). Effect of oral milrinone on mortality in severe chronic heart failure. The PROMISE Study Research Group. N Engl J Med 325(21): 1468-1475.

Bibliography 169

Packer M, Coats AJ, Fowler MB, Katus HA, Krum H, Mohacsi P, et al. (2001). Effect of carvedilol on survival in severe chronic heart failure. N Engl J Med 344(22): 1651-1658.

Pak MD, Fishman PH (1996). Anomalous behavior of CGP 12177A on beta 1- adrenergic receptors. J Recept Signal Transduct Res 16(1-2): 1-23.

Paschalis A, Churchill L, Marina N, Kasymov V, Gourine A, Ackland G (2009). beta1-Adrenoceptor distribution in the rat brain: an immunohistochemical study. Neurosci Lett 458(2): 84-88.

Persson H, Eriksson SV, Erhardt L (1996). Effects of beta receptor antagonists on left ventricular function in patients with clinical evidence of heart failure after myocardial infarction. A double-blind comparison of metoprolol and xamoterol. Echocardiographic results from the Metoprolol and Xamoterol Infarction Study (MEXIS). Eur Heart J 17(5): 741-749.

Poole-Wilson PA, Swedberg K, Cleland JG, Di Lenarda A, Hanrath P, Komajda M, et al. (2003). Comparison of carvedilol and metoprolol on clinical outcomes in patients with chronic heart failure in the Carvedilol Or Metoprolol European Trial (COMET): randomised controlled trial. Lancet 362(9377): 7-13.

Pott C, Brixius K, Bundkirchen A, Bolck B, Bloch W, Steinritz D, et al. (2003). The preferential beta3-adrenoceptor agonist BRL 37344 increases force via beta1-/beta2- adrenoceptors and induces endothelial nitric oxide synthase via beta3-adrenoceptors in human atrial myocardium. Br J Pharmacol 138(3): 521-529.

Preitner F, Muzzin P, Revelli JP, Seydoux J, Galitzky J, Berlan M, et al. (1998). Metabolic response to various beta-adrenoceptor agonists in beta3-adrenoceptor knockout mice: evidence for a new beta-adrenergic receptor in brown adipose tissue. Br J Pharmacol 124(8): 1684-1688.

Ramos BP, Colgan L, Nou E, Ovadia S, Wilson SR, Arnsten AF (2005). The beta-1 adrenergic antagonist, betaxolol, improves working memory performance in rats and monkeys. Biol Psychiatry 58(11): 894-900.

Rasmussen SG, Choi HJ, Fung JJ, Pardon E, Casarosa P, Chae PS, et al. (2011a). Structure of a nanobody-stabilized active state of the beta(2) adrenoceptor. Nature 469(7329): 175-180.

Rasmussen SG, Choi HJ, Rosenbaum DM, Kobilka TS, Thian FS, Edwards PC, et al. (2007). Crystal structure of the human beta2 adrenergic G-protein-coupled receptor. Nature 450(7168): 383-387.

170 Bibliography

Rasmussen SG, Devree BT, Zou Y, Kruse AC, Chung KY, Kobilka TS, et al. (2011b). Crystal structure of the beta(2) adrenergic receptor-Gs protein complex. Nature.

Rasmussen SG, Jensen AD, Liapakis G, Ghanouni P, Javitch JA, Gether U (1999). Mutation of a highly conserved aspartic acid in the beta2 adrenergic receptor: constitutive activation, structural instability, and conformational rearrangement of transmembrane segment 6. Mol Pharmacol 56(1): 175-184.

Ratnala VR, Kobilka B (2009). Understanding the ligand-receptor-G protein ternary complex for GPCR drug discovery. Methods Mol Biol 552: 67-77.

Revelli JP, Muzzin P, Paoloni A, Moinat M, Giacobino JP (1993). Expression of the beta 3-adrenergic receptor in human white adipose tissue. J Mol Endocrinol 10(2): 193-197.

Rohrer DK, Kobilka BK (1998). G protein-coupled receptors: functional and mechanistic insights through altered gene expression. Physiol Rev 78(1): 35-52.

Ross EM, Maguire ME, Sturgill TW, Biltonen RL, Gilman AG (1977). Relationship between the beta-adrenergic receptor and adenylate cyclase. J Biol Chem 252(16): 5761-5775.

Rothwell NJ, Stock MJ, Sudera DK (1985). Beta-adrenoreceptors in rat brown adipose tissue: proportions of beta 1- and beta 2-subtypes. Am J Physiol 248(4 Pt 1): E397-402.

Rozec B, Gauthier C (2006). beta3-adrenoceptors in the cardiovascular system: putative roles in human pathologies. Pharmacol Ther 111(3): 652-673.

Samama P, Cotecchia S, Costa T, Lefkowitz RJ (1993). A mutation-induced activated state of the beta 2-adrenergic receptor. Extending the ternary complex model. J Biol Chem 268(7): 4625-4636.

Samama P, Pei G, Costa T, Cotecchia S, Lefkowitz RJ (1994). Negative antagonists promote an inactive conformation of the beta 2-adrenergic receptor. Mol Pharmacol 45(3): 390-394.

Sarsero D, Molenaar P, Kaumann AJ (1998). Validity of (-)-[3H]-CGP 12177A as a radioligand for the 'putative beta4-adrenoceptor' in rat atrium. Br J Pharmacol 123(3): 371-380.

Bibliography 171

Sarsero D, Molenaar P, Kaumann AJ, Freestone NS (1999). Putative beta 4- adrenoceptors in rat ventricle mediate increases in contractile force and cell Ca2+: comparison with atrial receptors and relationship to (-)-[3H]-CGP 12177 binding. Br J Pharmacol 128(7): 1445-1460.

Sarsero D, Russell FD, Lynham JA, Rabnott G, Yang I, Fong KM, et al. (2003). (-)- CGP 12177 increases contractile force and hastens relaxation of human myocardial preparations through a propranolol-resistant state of the beta 1-adrenoceptor. Naunyn Schmiedebergs Arch Pharmacol 367(1): 10-21.

Sato M, Hutchinson DS, Evans BA, Summers RJ (2008). The beta3-adrenoceptor agonist 4-[[(Hexylamino)carbonyl]amino]-N-[4-[2-[[(2S)-2-hydroxy-3-(4- hydroxypheno xy)propyl]amino]ethyl]-phenyl]-benzenesulfonamide (L755507) and antagonist (S)-N-[4-[2-[[3-[3-(acetamidomethyl)phenoxy]-2-hydroxypropyl]amino]- ethyl] phenyl]benzenesulfonamide (L748337) activate different signaling pathways in Chinese hamster ovary-K1 cells stably expressing the human beta3-adrenoceptor. Mol Pharmacol 74(5): 1417-1428.

Sato T, Kobayashi H, Nagao T, Kurose H (1999). Ser203 as well as Ser204 and Ser207 in fifth transmembrane domain of the human beta2-adrenoceptor contributes to agonist binding and receptor activation. Br J Pharmacol 128(2): 272-274.

Sawa M, Harada H (2006). Recent developments in the design of orally bioavailable beta3-adrenergic receptor agonists. Curr Med Chem 13(1): 25-37.

Schliep HJ, Harting J (1984). Beta 1-selectivity of bisoprolol, a new beta- adrenoceptor antagonist, in anesthetized dogs and guinea pigs. J Cardiovasc Pharmacol 6(6): 1156-1160.

Schluter KD, Goldberg Y, Taimor G, Schafer M, Piper HM (1998). Role of phosphatidylinositol 3-kinase activation in the hypertrophic growth of adult ventricular cardiomyocytes. Cardiovasc Res 40(1): 174-181.

Schmitt JM, Stork PJ (2002). PKA phosphorylation of Src mediates cAMP's inhibition of cell growth via Rap1. Mol Cell 9(1): 85-94.

Schutz W, Freissmuth M (1992). Reverse intrinsic activity of antagonists on G protein-coupled receptors. Trends Pharmacol Sci 13(10): 376-380.

Seibold A, January BG, Friedman J, Hipkin RW, Clark RB (1998). Desensitization of beta2-adrenergic receptors with mutations of the proposed G protein-coupled receptor kinase phosphorylation sites. J Biol Chem 273(13): 7637-7642.

172 Bibliography

Seibold A, Williams B, Huang ZF, Friedman J, Moore RH, Knoll BJ, et al. (2000). Localization of the sites mediating desensitization of the beta(2)-adrenergic receptor by the GRK pathway. Mol Pharmacol 58(5): 1162-1173.

Seifert R, Dove S (2009). Functional selectivity of GPCR ligand stereoisomers: new pharmacological opportunities. Mol Pharmacol 75(1): 13-18.

Sennitt MV, Kaumann AJ, Molenaar P, Beeley LJ, Young PW, Kelly J, et al. (1998). The contribution of classical (beta1/2-) and atypical beta-adrenoceptors to the stimulation of human white adipocyte lipolysis and right atrial appendage contraction by novel beta3-adrenoceptor agonists of differing selectivities. J Pharmacol Exp Ther 285(3): 1084-1095.

Shi L, Javitch JA (2004). The second extracellular loop of the dopamine D2 receptor lines the binding-site crevice. Proc Natl Acad Sci U S A 101(2): 440-445.

Shi L, Liapakis G, Xu R, Guarnieri F, Ballesteros JA, Javitch JA (2002). Beta2 adrenergic receptor activation. Modulation of the proline kink in transmembrane 6 by a rotamer toggle switch. J Biol Chem 277(43): 40989-40996.

Sillence MN, Hooper J, Zhou GH, Liu Q, Munn KJ (2005). Characterization of porcine beta1- and beta2-adrenergic receptors in heart, skeletal muscle, and adipose tissue, and the identification of an atypical beta-adrenergic binding site. J Anim Sci 83(10): 2339-2348.

Skeberdis VA, Gendviliene V, Zablockaite D, Treinys R, Macianskiene R, Bogdelis A, et al. (2008). beta3-adrenergic receptor activation increases human atrial tissue contractility and stimulates the L-type Ca2+ current. J Clin Invest 118(9): 3219- 3227.

Skeberdis VA, Jurevicius J, Fischmeister R (1997). Beta-2 adrenergic activation of L-type Ca++ current in cardiac myocytes. J Pharmacol Exp Ther 283(2): 452-461.

Small KM, McGraw DW, Liggett SB (2003). Pharmacology and physiology of human adrenergic receptor polymorphisms. Annual review of pharmacology and toxicology 43: 381-411.

Soeder KJ, Snedden SK, Cao W, Della Rocca GJ, Daniel KW, Luttrell LM, et al. (1999). The beta3-adrenergic receptor activates mitogen-activated protein kinase in adipocytes through a Gi-dependent mechanism. J Biol Chem 274(17): 12017-12022.

Staehelin M, Simons P, Jaeggi K, Wigger N (1983). CGP-12177. A hydrophilic beta- adrenergic receptor radioligand reveals high affinity binding of agonists to intact cells. J Biol Chem 258(6): 3496-3502.

Bibliography 173

Standfuss J, Edwards PC, D'Antona A, Fransen M, Xie G, Oprian DD, et al. (2011). The structural basis of agonist-induced activation in constitutively active rhodopsin. Nature 471(7340): 656-660.

Stemmelin J, Cohen C, Terranova JP, Lopez-Grancha M, Pichat P, Bergis O, et al. (2008). Stimulation of the beta3-Adrenoceptor as a novel treatment strategy for anxiety and depressive disorders. Neuropsychopharmacology 33(3): 574-587.

Stephenson RP (1956). A modification of receptor theory. Br J Pharmacol Chemother 11(4): 379-393.

Strader CD, Candelore MR, Hill WS, Dixon RA, Sigal IS (1989a). A single amino acid substitution in the beta-adrenergic receptor promotes partial agonist activity from antagonists. J Biol Chem 264(28): 16470-16477.

Strader CD, Candelore MR, Hill WS, Sigal IS, Dixon RA (1989b). Identification of two serine residues involved in agonist activation of the beta-adrenergic receptor. J Biol Chem 264(23): 13572-13578.

Strader CD, Dixon RA, Cheung AH, Candelore MR, Blake AD, Sigal IS (1987a). Mutations that uncouple the beta-adrenergic receptor from Gs and increase agonist affinity. J Biol Chem 262(34): 16439-16443.

Strader CD, Sigal IS, Candelore MR, Rands E, Hill WS, Dixon RA (1988). Conserved aspartic acid residues 79 and 113 of the beta-adrenergic receptor have different roles in receptor function. J Biol Chem 263(21): 10267-10271.

Strader CD, Sigal IS, Register RB, Candelore MR, Rands E, Dixon RA (1987b). Identification of residues required for ligand binding to the beta-adrenergic receptor. Proc Natl Acad Sci U S A 84(13): 4384-4388.

Sugimoto Y, Fujisawa R, Tanimura R, Lattion AL, Cotecchia S, Tsujimoto G, et al. (2002). Beta(1)-selective agonist (-)-1-(3,4-dimethoxyphenetylamino)-3-(3,4- dihydroxy)-2-propanol [(-)-RO363] differentially interacts with key amino acids responsible for beta(1)-selective binding in resting and active states. J Pharmacol Exp Ther 301(1): 51-58.

Summers RJ, Molenaar P, Stephenson JA, Jones CR (1987). Autoradiographic localization of receptors in the mammalian cardiovascular system. Clin Exp Pharmacol Physiol 14(5): 437-447.

174 Bibliography

Summers RJ, Molnaar P, Russell F, Elnatan J, Jones CR, Buxton BF, et al. (1989). Coexistence and localization of beta 1- and beta 2-adrenoceptors in the human heart. Eur Heart J 10 Suppl B: 11-21.

Swaminath G, Deupi X, Lee TW, Zhu W, Thian FS, Kobilka TS, et al. (2005). Probing the beta2 adrenoceptor binding site with catechol reveals differences in binding and activation by agonists and partial agonists. J Biol Chem 280(23): 22165- 22171.

Tavernier G, Barbe P, Galitzky J, Berlan M, Caput D, Lafontan M, et al. (1996). Expression of beta3-adrenoceptors with low lipolytic action in human subcutaneous white adipocytes. J Lipid Res 37(1): 87-97.

Taylor JM, Jacob-Mosier GG, Lawton RG, VanDort M, Neubig RR (1996). Receptor and membrane interaction sites on Gbeta. A receptor-derived peptide binds to the carboxyl terminus. J Biol Chem 271(7): 3336-3339.

Terra SG, Hamilton KK, Pauly DF, Lee CR, Patterson JH, Adams KF, et al. (2005). Beta1-adrenergic receptor polymorphisms and left ventricular remodeling changes in response to beta-blocker therapy. Pharmacogenet Genomics 15(4): 227-234.

Tesson F, Charron P, Peuchmaurd M, Nicaud V, Cambien F, Tiret L, et al. (1999). Characterization of a unique genetic variant in the beta1-adrenoceptor gene and evaluation of its role in idiopathic dilated cardiomyopathy. CARDIGENE Group. J Mol Cell Cardiol 31(5): 1025-1032.

The Beta-Blocker Evaluation of Survival Trial Investigators (2001). A trial of the beta-blocker bucindolol in patients with advanced chronic heart failure. N Engl J Med 344(22): 1659-1667.

The CIBIS-II Investigators (1999). The Cardiac Insufficiency Bisoprolol Study II (CIBIS-II): a randomised trial. Lancet 353(9146): 9-13.

The MERIT-HF Investigators (1999). Effect of metoprolol CR/XL in chronic heart failure: Metoprolol CR/XL Randomised Intervention Trial in Congestive Heart Failure (MERIT-HF). Lancet 353(9169): 2001-2007.

The Xamoterol in Severe Heart Failure Study Group (1990). Xamoterol in severe heart failure. The Xamoterol in Severe Heart Failure Study Group. Lancet 336(8706): 1-6.

Torretti J (1982). Sympathetic control of renin release. Annu Rev Pharmacol Toxicol 22: 167-192.

Bibliography 175

Tran TM, Friedman J, Qunaibi E, Baameur F, Moore RH, Clark RB (2004). Characterization of agonist stimulation of cAMP-dependent protein kinase and G protein-coupled receptor kinase phosphorylation of the beta2-adrenergic receptor using phosphoserine-specific antibodies. Mol Pharmacol 65(1): 196-206.

Trautwein W, Hescheler J (1990). Regulation of cardiac L-type calcium current by phosphorylation and G proteins. Annu Rev Physiol 52: 257-274.

Tugiono N, Kiriazis H, Gao X, Xu Q, Jennings N, Molenaar P, et al. (2010). Role of low affinity beta1-adrenergic receptors in normal and diseased hearts. In: Cardiac Society of Australia and New Zealand Annual Scientific Meeting and International Society for Heart Research Australasian Section Annual Scientific Meeting. Adelaide.

Waagstein F, Bristow MR, Swedberg K, Camerini F, Fowler MB, Silver MA, et al. (1993). Beneficial effects of metoprolol in idiopathic dilated cardiomyopathy. Metoprolol in Dilated Cardiomyopathy (MDC) Trial Study Group. Lancet 342(8885): 1441-1446.

Waagstein F, Hjalmarson A, Varnauskas E, Wallentin I (1975). Effect of chronic beta-adrenergic receptor blockade in congestive cardiomyopathy. Br Heart J 37(10): 1022-1036.

Waelbroeck M, Taton G, Delhaye M, Chatelain P, Camus JC, Pochet R, et al. (1983). The human heart beta-adrenergic receptors. II. Coupling of beta 2-adrenergic receptors with the adenylate cyclase system. Mol Pharmacol 24(2): 174-182.

Walter M, Lemoine H, Kaumann AJ (1984). Stimulant and blocking effects of optical isomers of pindolol on the sinoatrial node and trachea of guinea pig. Role of beta-adrenoceptor subtypes in the dissociation between blockade and stimulation. Naunyn Schmiedebergs Arch Pharmacol 327(2): 159-175.

Wang W, Zhu W, Wang S, Yang D, Crow MT, Xiao RP, et al. (2004). Sustained beta1-adrenergic stimulation modulates cardiac contractility by Ca2+/calmodulin kinase signaling pathway. Circ Res 95(8): 798-806.

Warne T, Moukhametzianov R, Baker JG, Nehme R, Edwards PC, Leslie AG, et al. (2011). The structural basis for agonist and partial agonist action on a beta(1)- adrenergic receptor. Nature 469(7329): 241-244.

Warne T, Serrano-Vega MJ, Baker JG, Moukhametzianov R, Edwards PC, Henderson R, et al. (2008). Structure of a beta1-adrenergic G-protein-coupled receptor. Nature 454(7203): 486-491.

176 Bibliography

Wellner-Kienitz MC, Bender K, Pott L (2001). Overexpression of beta 1 and beta 2 adrenergic receptors in rat atrial myocytes. Differential coupling to G protein-gated inward rectifier K(+) channels via G(s) and G(i)/o. The Journal of biological chemistry 276(40): 37347-37354.

Westphal S, Perwitz N, Iwen KA, Kraus D, Schick R, Fasshauer M, et al. (2008). Expression of ATRAP in adipocytes and negative regulation by beta-adrenergic stimulation of JAK/STAT. Horm Metab Res 40(3): 165-171.

Wieland K, Zuurmond HM, Krasel C, Ijzerman AP, Lohse MJ (1996). Involvement of Asn-293 in stereospecific agonist recognition and in activation of the beta 2- adrenergic receptor. Proc Natl Acad Sci U S A 93(17): 9276-9281.

Wu G, Bogatkevich GS, Mukhin YV, Benovic JL, Hildebrandt JD, Lanier SM (2000). Identification of Gbetagamma binding sites in the third intracellular loop of the M(3)-muscarinic receptor and their role in receptor regulation. J Biol Chem 275(12): 9026-9034.

Wuest M, Eichhorn B, Grimm MO, Wirth MP, Ravens U, Kaumann AJ (2009). Catecholamines relax detrusor through beta 2-adrenoceptors in mouse and beta 3- adrenoceptors in man. J Pharmacol Exp Ther 328(1): 213-222.

Xiang Y, Devic E, Kobilka B (2002). The PDZ binding motif of the beta 1 adrenergic receptor modulates receptor trafficking and signaling in cardiac myocytes. J Biol Chem 277(37): 33783-33790.

Xiao B, Tian X, Xie W, Jones PP, Cai S, Wang X, et al. (2007). Functional consequence of protein kinase A-dependent phosphorylation of the cardiac ryanodine receptor: sensitization of store overload-induced Ca2+ release. J Biol Chem 282(41): 30256-30264.

Xu F, Wu H, Katritch V, Han GW, Jacobson KA, Gao ZG, et al. (2011). Structure of an agonist-bound human A2A adenosine receptor. Science 332(6027): 322-327.

Xu J, Paquet M, Lau AG, Wood JD, Ross CA, Hall RA (2001). beta 1-adrenergic receptor association with the synaptic scaffolding protein membrane-associated guanylate kinase inverted-2 (MAGI-2). Differential regulation of receptor internalization by MAGI-2 and PSD-95. J Biol Chem 276(44): 41310-41317.

Yamaguchi O (2002). Beta3-adrenoceptors in human detrusor muscle. Urology 59(5 Suppl 1): 25-29.

Bibliography 177

Yang-Feng TL, Xue FY, Zhong WW, Cotecchia S, Frielle T, Caron MG, et al. (1990). Chromosomal organization of adrenergic receptor genes. Proc Natl Acad Sci U S A 87(4): 1516-1520.

Yao X, Parnot C, Deupi X, Ratnala VR, Swaminath G, Farrens D, et al. (2006). Coupling ligand structure to specific conformational switches in the beta2- adrenoceptor. Nat Chem Biol 2(8): 417-422.

Yao XJ, Velez Ruiz G, Whorton MR, Rasmussen SG, DeVree BT, Deupi X, et al. (2009). The effect of ligand efficacy on the formation and stability of a GPCR-G protein complex. Proc Natl Acad Sci U S A 106(23): 9501-9506.

Zaugg M, Xu W, Lucchinetti E, Shafiq SA, Jamali NZ, Siddiqui MA (2000). Beta- adrenergic receptor subtypes differentially affect apoptosis in adult rat ventricular myocytes. Circulation 102(3): 344-350.

Zeitoun O, Santos NM, Gardner LA, White SW, Bahouth SW (2006). Mutagenesis within helix 6 of the human beta1-adrenergic receptor identifies Lysine324 as a residue involved in imparting the high-affinity binding state of agonists. Mol Pharmacol 70(3): 838-850.

Zhang R, Zhao J, Mandveno A, Potter JD (1995). Cardiac troponin I phosphorylation increases the rate of cardiac muscle relaxation. Circ Res 76(6): 1028-1035.

Zheng M, Zhang SJ, Zhu WZ, Ziman B, Kobilka BK, Xiao RP (2000). beta 2- adrenergic receptor-induced p38 MAPK activation is mediated by protein kinase A rather than by Gi or gbeta gamma in adult mouse cardiomyocytes. J Biol Chem 275(51): 40635-40640.

Zhu WZ, Wang SQ, Chakir K, Yang D, Zhang T, Brown JH, et al. (2003). Linkage of beta1-adrenergic stimulation to apoptotic heart cell death through protein kinase A-independent activation of Ca2+/calmodulin kinase II. J Clin Invest 111(5): 617- 625.

178 Bibliography

Appendices

Appendix A Table of Chemical Structures

Chemical Name Chemical Structure

(±)-Bisoprolol

BRL 37344

(-)-Bupranolol

(-)-CGP 12177

Appendices 179

Chemical Name Chemical Structure

(±)-CGP20712A

ICI 118,551

L-748,337

IMBX

(-)-Isoprenaline

(±)-Nadolol

180 Appendices

Appendix B British Journal of Pharmacology Publication

Appendices 181

Appendix C Molecular Modelling

Based on sequence alignment of 1ARwith the first crystal structure of 2AR, it was hypothesised that although TMDV residues were required for agonist binding and subsequent activity at the high-affinity binding site for both 1AR and 2AR; the differing 1AR TMDV residues vs 2AR were unlikely to directly influence ligand binding at either the high or low-affinity binding sites. This was ‘virtual’ tested at the high-affinity binding site by building homology models of 1AR and 1/2TMDVAR and docking selected ligands (including(-)-CGP 12177) into the high-affinity binding site. We examined the models to determine whether a second low-affinity binding pocket could be identified close to the high-affinity binding pocket, and whether (-)- CGP 12177 which is a known agonist at the low-affinity site would dock appropriately. If this hypothesis was correct; neither the high or low-affinity binding sites should require the 1AR and 2AR TMDV differing residues for affinity or be in close proximity to docked ligands.

Insertion of 2AR TMDV into the 1/2TMDVAR model revealed that the TMDV mutated residues do not directly influence the high-affinity binding pocket (Figure B1). Of the five mutated residues, only Arg222Gln and Val230Ile are above the Pro236 kink in TMDV and close to the high-affinity binding pocket. 1AR Arg222 was solvent exposed at the extracellular end of TMDV and, as observed in

Turkey 1AR crystal structures, the arginine side chain was found to be within H- bonding distance to the backbone carbonyl group of Ala346 (Arg222 NH…O

Ala346; 1.6Å) on TMDV. By contrast, 1/2TMDVAR Gln222 did not make this H- bond (NH…O 3.7Å). The second residue of interest, 1AR Val230Ile on TMDV, was found to project on the opposite face of the TMDV helix away from the ligand- binding site and into the membrane (Figure B1).

198 Appendices

Figure B1. Superposition of homology models for 1AR and 1/2TMDVAR, showing TMDV mutated residues (magenta sticks), docked (-)-CGP 12177 (green sticks). Expanded view shows (-)-CGP 12177 docked in high-affinity binding site with key residues labelled, potential H-bonds to Ser228 and Asp138 highlighted as dashed yellow lines.

A comparison of the physiochemical properties by Heliquest of TMDV of

1AR and 1/2TMDVAR found the 1AR TMDV to have less hydrophobicity (H =

0.941) relative to 1/2TMDVAR (H = 0.923) but conversely it had greater hydrophobic moment (H = 0.161 vs 0.140). These findings are consistent with

1AR TMDV having a single charged residue Arg222 at the extracellular N-terminal of TMDV relative to the neutral Gln222 of 1/2TMDVAR and that it lies on the more polar face of TMDV, (Figure B2). Generally, the mutations found in 1/2TMDVAR replace smaller 1AR hydrophobic residues with larger hydrophobic residues e.g. V230I, C238V and A241V. Critically an exception to this was L245S which introduced a polar residue towards the C-terminal end of TMDV and importantly disturbed the hydrophobic face of 1/2TMDVAR TMDV relative to 1AR, (Figure

B2). The smaller hydrophobic residues of 1AR may help facilitate movement of TMDV against TMDVI during activation and introduction of larger hydrophobic groups and a polar residue in 1/2TMDVAR causes unfavourable contacts to be made with TMDVI during activation reducing overall activity relative to 1AR.

Appendices 199

Figure B2: Helical wheel representation of TMDV of 1AR (left) and 1/2TMDVAR (right) residues coloured as follows, small hydrophobic (grey), large hydrophobic (yellow), neutral (pink), positive (blue) and proline (green). N and C-terminal residues indicated by red letters and hydrophobic moment (H) represented by direction and size of arrow.

To determine if Glide in XP mode would correctly dock ligands into the high- affinity binding site; cross docking of dobutamine back into 1AR homology was performed with one H-bond constraint to Asp138, the top docked conformer had a heavy atom RMSD of 0.45Å relative to the original ligand position, values ≤ 2Å are considered acceptable. (-)-CGP 12177 was then all docked into 1AR and

1/2TMDVAR homology models, the top docked conformations in both structures were very similar to the conformer shown in Figure B1. Shown is the conformer of (- )-CGP 12177 that has the β-hydroxy and charged protonated amine groups in an orientation that most closely resembles the bound agonists found in the turkey 1AR crystal structures. H-bonds are formed between the β-hydroxy and charged protonated amine groups and Asp138 and also between one benzimidazol-2-one-NH and Ser228. Other top docked conformations not shown, make H-bonds between the (-)-CGP12177 β-hydroxy-OH and the carbonyl group of Asn363 and also (-)-CGP 12177 benzimidazol-2-one carbonyl group accepts a H-bond from Asn344. Top docked conformers of (-)-CGP12177 also indicated close hydrophobic contacts between the benzimidazol-2-one ring of (-)-CGP12177 and residues Val139 and Phe341.

200 Appendices

1AR is known to form a second low-affinity binding site that upon introduction of increased concentrations of (-)-CGP 12177 that can bind to this binding site 1AR becomes partially activated. Upon docking of (-)-CGP 12177 into the high-affinity binding pocket it was apparent that there was still a substantial potential binding groove above the docked (-)-CGP 12177 close to ECL2. To further explore this binding site as the potential low-affinity binding site the top docked (-)- CGP 12177 binding pose was submitted to binding site detection analysis using SiteMap. Sitemap generates a series of descriptors of the potential binding sites and outputs a variety of measures that have been validated against submicromolar binding sites. Sitemap found a potential binding site above the docked conformation of (-)-CGP 12177 and gave a Sitescore of 1.06 (Sitescore>1 = <µM), drugability score Dscore of 1.08 (Dscore>1 = drugable target) exposure score of 0.42 (<0.49 desirable), and an enclosure score of 0.8 (> 0.78 desirable) all were favourable for the presence of a second potential binding pocket (Figure B3).

Figure B3. Left panel shows 1AR with (-)-CGP 12177 docked into the high-affinity site with potential second binding site illustrated with blue spheres (PocketPicker; darker spheres represent greater buriedness), TMDVI and TMDVII clipped for clarity. Right panel shows the two top ranked IFD conformations of (-)-CGP 12177 as magenta and pink sticks respectively. Potential H-bonds shown in dashed yellow lines.

A second (-)-CGP 12177 ligand was docked in to the binding site identified by SiteMap using both Glide in XP mode and the IFD protocol; results of docking were

Appendices 201

compared and ranked via Glide XPscores. It was found that IFD produced docked conformations with better Glide XPscores. Shown are two of the top docked conformations of the second (-)-CGP 12177 in the second potential low-affinity binding site identified by SiteMap (Figure B3). Both structures docked in the second binding site make a series of H-bonds; firstly, in both conformers the charged amine interacts with ECL2 Asp217 and the -hydroxy-OH H-bonds to the -hydroxy of the first (-)-CGP 1277 in the high-affinity site. The magenta coloured conformer (shown as sticks, Figure B3) has the heterocyclic benzimidazol-2-one ring pointing directly towards the benzimidazol-2-one ring of the first (-)-CGP 12177 and forms an H-bond between its carbonyl group and the benzimidazol-2-one NH of the first (-)-CGP 12177. The magenta conformer forms a second H-bond between its benzimidazol-2- one NH and ECL2 Thr220-OH. The second (-)-CGP 12177 conformer shown in pink sticks donates and accepts H-bonds via the benzimidazol-2-one ring to the ECL2 backbone amides of Phe218 and Thr220 respectively.

202 Appendices