DISCOVERING THE MOLECULAR AND CELLULAR MECHANISMS UNDERLYING FENFLURAMINE-INDUCED CARDIOPULMONARY SIDE EFFECTS
Vincent Setola
Dissertation Advisor: Bryan L. Roth, MD, PhD
Department of Biochemistry Case School of Medicine Cleveland, OH
August 12, 2005
Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy
CASE WESTERN RESERVE UNIVERSITY
SCHOOL OF GRADUATE STUDIES
We hereby approve the dissertation of
______
candidate for the Ph.D. degree *.
(signed)______(chair of the committee)
______
______
______
______
______
(date) ______
*We also certify that written approval has been obtained for any proprietary material contained therein. TABLE OF CONTENTS
List of Tables v
List of Figures vi
Acknowledgements viii
List of Abbreviations xii
Abstract xvi
CHAPTER 1: INTRODUCTION
1.1 G Protein-Coupled Receptors: Overview 1
1.1.1 G Protein-Coupled Receptor Signal Transduction 1
1.1.2 G Protein-Coupled Receptor Topology/Structure 2
1.2 Serotonin (5-HT) Receptors 8
1.2.1 5-HT2 Receptors 11
1.2.1.1 5-HT2B Receptors 14
1.2.1.2 Tissue Distribution of 5-HT2B Receptors 14
1.2.1.3 Signal Transduction of 5-HT2B Receptors 15
1.2.1.4 Biochemical Consequences of 5-HT2B Receptor 19
Activation
1.3 The Rise of Fenfluramine 22
1.4 The Fall of Fenfluramine 29
1.5 Biological Activity of Fenfluramine: Generalities of Amphetamine Action 34
1.6 Evidence Linking the 5-HT2B Receptor to VHD 37
i CHAPTER 2: MATERIALS AND METHODS
2.1 Materials 40
2.1.1 Chemicals 40
2.1.2 Transfection Reagents, Cell Culture, and Transfection 41
2.1.3 cDNA Constructs 43
2.1.3.1 Construction of pUniversal-Signal 43
2.1.3.2 Sub-cloning of Human 5-HT2 Receptors 47
2.1.3.3 Generation of Mutant 5-HT2 Receptors 50
2.2 Methods 53
2.2.1 Radioligand Binding Assay 53
2.2.2 Functional Assays 58
2.2.2.1 Inositol Phosphate Accumulation Assay 58
2.2.2.2 [3H]Thymidine Deoxyribose Incorporation Assay 60
2.2.2.3 Activated Mitogen-Activated Protein Kinase Assay 60
2.2.2.4 Molecular Modeling, Ligand Docking Simulations, and 61
Molecular Dynamics Simulations
CHAPTER 3: ESTABLISHING THAT FENFLURAMINE CAUSES MITOSIS IN
PRIMARY CULTURES OF VALVULAR INTERSTITIAL CELLS VIA
ACTIVATION OF 5-HT2B RECEPTORS
3.1 Introduction and Rationale 65
3.2 Results 69
ii 3.2.1 Screening the Receptorome Reveals the 5-HT2B Receptor as a 69
Molecular Target for MDMA and MDA
3.2.2 Valvulopathic Drugs Induce Prolonged Mitogenic Responses 82
in Human Heart Valve Interstitial Cells
3.2.3 Valvulopathic Drugs Induce a Mitogenic Marker— 87
Phosphorylation of Mitogen-Activated Protein Kinase—
in Human Heart Valve Interstitial Cells
3.3 Discussion 87
CHAPTER 4: MOLECULAR DETERMINANTS FOR THE INTERACTION OF
THE ANOREXIGEN NORFENFLURAMINE WITH THE 5-HT2B RECEPTOR
4.1 Introduction and Rationale 91
4.2 Results 93
4.2.1 Effect of Point Mutations on Ligand Affinity 93
4.2.2 Modeling, Ligand Docking Simulations, and Molecular 105
Dynamics Simulations of Ligand Binding to 5-HT2B Receptors
4.2.3 Effect of Point Mutations on Agonist Potency and Efficacy 124
4.3 Discussion 130
CHAPTER 5: IMPLICATIONS AND FUTURE DIRECTIONS
5.1 Summary 142
5.2 High-Throughput Screening Efforts to Identify Potential 143
Valvulopathogens Among Current and Future Pharmacotherapies
iii 5.3 Design of “Second Generation” Phenylisopropylamine Anorexigens 147
5.4 Conclusion 153
BIBLIOGRAPHY 157
iv LIST OF TABLES
1.1 Clinical studies of the efficacy of appetite suppressants 23
2.1 PCR primer sequences used to amplify and sub-clone human 5-HT2 receptor 48
cDNA
2.2 Sequence of sense primers used for site-directed mutagenesis of 5-HT2 receptors 51
3.1 MDMA, MDA, and other valvulopathic drugs bind to recombinant human 74
5-HT2B receptors
3.2 MDMA and MDA, similar to other valvulopathic drugs, activate human 79
5-HT2B receptors in vitro
4.1 Affinity constants (Ki’s) for SNF at wild type and mutant 5-HT2 receptors 97
4.2 Affinity constants (Ki’s) for other 5-HT2B receptor agonist ligand sat wild type 101
and mutant 5-HT2 receptors
4.3 Affinity constants (Ki’s) for SNF and congeners at wild type and V2.53L 5-HT2B 115
receptors and wild type 5-HT2C and 5-HT2A receptors
4.4 Potency (EC50) and relative efficacy (Emax) values for SNF and RNF at wild type 125
and mutant 5-HT2 receptors
v LIST OF FIGURES
1.1 G protein-coupled receptor signal transduction: the phospholipase C pathway 3
1.2 Schematic representation of a plasma membrane GPCR 5
1.3 Chemical structures of the indolamine 5-hydroxytryptamine (5-HT, serotonin), 9
and the phenylisopropylamines 3,4-methylenedioxymethamphetamine
(MDMA, “Ecstasy”), 3,4-methylenedioxyamphetamine (MDA),
fenfluramine, norfenfluramine, and phentermine
1.4 Signal transduction pathways modulated by 5-HT2B receptors 17
1.5 Amphetamine actions on biogenic amine reuptake, storage, and release 32
2.1 Schematic of the plasma membrane protein expression vector pUniversal-Signal 44
2.2 Sequences of the 5-HT2 receptors sub-cloned into pUniversal-Signal 54
3.1 Large-scale screening of the receptorome reveals that MDMA preferentially 70
interacts with the human 5-HT2B receptor
3.2 MDMA and MDA potently activate 5-HT2B receptors in vitro 77
3.3 MDMA and MDA induce mitogenesis in human heart valve interstitial cells 83
in vitro
4.1 3-D molecular model of the human 5-HT2B receptor showing putative ligand 94
binding residues that are non-conserved among 5-HT2 family receptors
4.2 Representative competition binding isotherms for SNF at wild type and mutant 99
5-HT2 receptors
4.3 Competition binding isotherms for several 5-HT2B receptor agonist ligands at 103
wild type and V2.53L 5-HT2B receptors
vi 4.4A 3-D molecular models showing the results of ligand docking simulations that 107
are consistent with conserved features of biogenic amine ligand binding
4.4B Representative energy-minimized structures from ten rounds of computer- 109
simulated annealing of solution 1 (A,B) and solution 2 (C,D) after
insertion of the V2.53L mutation
4.5 Representative energy-minimized structure from ten rounds of computer- 112
simulated annealing of solution 1 after insertion of the V2.53I mutation
4.6 Representative energy-minimized structure from ten rounds of computer- 118
simulated annealing of solution 1 bearing the V2.53L mutation after
addition or removal of SNF α-carbon substituents
4.7 Competition binding isotherms for SNF at wild type and V2.53A 5-HT2B 122
receptors
4.8 Concentration-response isotherms for agonist-stimulated inositol phosphate 127
accumulation
vii ACKNOWLEDGEMENTS
I wish to acknowledge the contributions of some of those who have made my work at Case possible and pleasurable:
I am profoundly grateful to my advisor, Dr. Bryan Roth, for his outstanding mentorship. Since joining the Rothlab in 2001, I have learned from Bryan’s guidance and from his example an enormous, invaluable amount about the practice and the art of scientific research and analytical thought. Whatever scientific successes I may achieve in the future will be in no small part the result of Bryan’s involvement in my training. I will remember him with fondness and gratitude.
I am thankful to my colleagues and friends in the Rothlab, whose talent and motivation have been inspirational to me. I sincerely hope for chances to work with this gifted group of individuals in the future.
I deeply appreciate the intellectual and personal support of my fellow graduate students in the Rothlab.
Since my first day in the lab, Douglas Sheffler has been helpful in so many ways.
I am grateful to him for sharing his many talents, his kindness, and his friendship
during my years at Case. I am also thankful to him for being an “uncle” to
Maxime, Madeleine, and Maurice.
viii I am thankful to my former bay-mate, fellow former graduate student, and friend
Zongqi Xia for countless hours of intriguing conversation. Whatever the topic—
science, politics, economics, art, cuisine, culture—he never failed to provide
unique, thought-provoking perspectives and ideas.
Ryan Strachan and Atheir Abbas joined the Rothlab towards the end of my tenure.
I thank them for being supportive colleagues and friends.
Je suis profondemment reconnaissant pour l’amitié et le soutien de Duna
Massillon, Fredéric Bone, et Nadia Rachdaoui. Ils sont parmi les premiers amis que j’ai rencontré à Cleveland et ils sont devenus une « famille » loin de la mienne. J’apprécie egalement l’amitie de France David, Anne-Laure Bulteau, Laurent Chavatte, Michele Le
Moing, et Hossein Izeml : les soirées, les anniversaires, les week-end de ski, les sorties sont de tres beaux souvenirs.
Wesley Kroeze’s friendship and guidance have made me a better scientist and a better person. Durinsg the best and worst of times, Wes has celebrated with me, stood by me, and counseled me. Wes’s personal and professional stories have inspired me to work hard, to strive for excellence, to make time to enjoy life, and to be a good friend. In
addition, Wes has shared with me one of his life’s loves: the great sport of sailboat
racing. During our adventures on Ensigns—in Lake Erie and off the Gulf and Atlantic
Coasts of Florida—Wes taught me so much about sailing and about competition. For everything Wes has given me, shared with me, and taught me, I am profoundly grateful.
ix I have the distinct honor to call Anushree Bhatnagar my dear friend. We met as graduate students in the Rothlab, where we frequently discussed our science and our lives. Anushree and I soon became very close friends. Over the years, our friendship has blossomed into one the likes of which I have never before experienced. Anushree has taught me so much about hard work, integrity, friendship, family, and faith. I have had the privilege of watching Anushree be a graduate student, a post-doc, a friend, a spouse, a sibling, and a parent. In all of these roles, Anushree has provided me with an example of that for which all human beings should strive. Having Anushree in my life—having her as a dear friend—is one of my richest blessings. I am profoundly grateful to Anushree for her friendship and her many gifts to me.
I met Marilyn Davies towards the end of my time in Cleveland. Soon after
Marilyn joined the Rothlab she and I became friends. When I found myself in need of a place to stay, Marilyn and her husband Bill opened their home to me. Since that time, I have gotten to know the Davies to an extent that I have never gotten to know any other of my friends. I have experienced first-hand their generosity, their commitment to each other and to their family and friends, their integrity. During my last and most difficult months at Case, Marilyn and Bill have supported me to an extent that surpasses that typical of friendship by leaps and bounds. The Davies (Marilyn and Bill, their son Chad, and their daughter Ashley) have been to me like a second family. The evenings I have spent with the Davies—at the kitchen table, by the fire, watching Pitt games (go
Panthers!)—are my fondest memories of Cleveland and among the greatest times of my life. The Davies’ love and support the have not only helped me to complete my studies,
x they have made me a better person. I have been blessed to be a part of the Davies’ lives and I am eternally grateful to them for being such an important part of mine.
To my family:
Mom, Dad, John, and Anthony –
I will not find the words to express what your love and support have meant to me.
Everything good I have accomplished in life has been motivated and facilitated by you.
Every time I have stumbled, you have picked me up and guided me back to my path.
Every day of my life has been made better by sharing it with you. You are the best part of me. Thank you for your love, for your example, for your advice, for your support, for being the most wonderful people I will ever know.
I pray that I will be able to make the most of the opportunities your love and your hard work have made possible. I am eternally grateful to you for all that is and will be good in my life. I dedicate this and every other accomplishment in my life to you. With you in my life, I know there will be many, many more to come…
I love you with all my heart and soul.
xi LIST OF ABBREVIATIONS
5-HT 5-hydroxytryptamine
5-HIAA 5-hydroxyindole acetic acid
7TM seven transmembrane
AMBER assisted model building with energy refinement
AMPH amphetamines
ATP adenosine triphosphate
BMI body mass index
BP blood pressure
BT behavioral therapy
BW723C86 α-methyl-5-(2-thienylmethoxy)-1H-indole-3-ethanamine
BSA bovine serum albumin
C chlorphentermine cDNA cloned DNA
CMV cytomegalovirus cNOS constitutive nitric oxide synthase
CSA computer-simulated annealing
CSF cerebrospinal fluid
DAM dexamphetamine
DA dopamine
DAG diacylglycerol
xii DI desimipramine
DMEM Dulbecco’s modified essential medium
DNA deoxyribonucleic acid
DP diethylproprion e1,2,3 first,second,third extracellular loop
F fenfluramine
(F)BG (fasting) blood glucose
FBS fetal bovine serum
FDA Food and Drug Administration
Fen-Phen fenfluramine and phentermine
FFA free fatty acids
FLX fluoxetine
GDP guanosine diphosphate
GPCR G protein-coupled receptor
GTP guanosine triphosphate
HbA1C Hemoglobin A1C
HEK human embryonic kidney
HVA homovanillic acid iNOS inducible nitric oxide synthase i1,2,3 first,second,third intracellular loop
IP3 inositol-1,4,5-trisphosphate
IRES internal ribosome entry site
xiii IVS intervening sequence
JCV JC virus
LB Luria Bertani
LSC liquid scintillation counting
LSD lysergic acid diethylamine
M mazindol
MAPK mitogen-activated protein kinase
MD molecular dynamics
MDA 3,4-methylenedioxyamphetamine
MDMA 3,4-methylenedioxymethamphetamine (“Ecstasy”)
NCBI National Center for Biotechnology Information
NET norepinephrine transporter
NIDA National Institute on Drug Abuse
NF norfenfluramine
NOS nitric oxide synthase
P phentermine
PCR polymerase chain reaction
PDB protein data bank
PDGFR platelet-derived growth factor receptor
PDZ postsynaptic density 95/disc large/zona occludens-1
xiv PI3K phosphatidylinositol 3-kinase
PIP2 phosphatidylinositol-4,5-bisphosphate
PKC protein kinase C
PLA2 phospholipase A2
PLCβ phospholipase Cβ
PML progressive multifocal leukoencephalopathy
RDF receptor description file
RNF R-(-)-norfenfluramine
SB206,553 5-methyl-1-(3-pyridylcarbamoyl)-1,2,3,5-tetrahydropyrrolo[2,3-
f]indole
SDS sodium dodecylsulfate
SNF S-(+)-norfenfluramine
TG triglycerides vdW van der Waals’
VIC human heart valve interstitial cells
VMAT vesicular monoamine transporter
xv Discovering the Cellular and Molecular Mechanisms Underlying Fenfluramine-Induced Cariopulmonary Side Effects
Abstract
by
VINCENT SETOLA
G protein-coupled receptors (GPCRs) are plasma membrane proteins that act as
sensors for extracellular stimuli (e.g., photons, neurohumoral moduluators, odorants and tastants, lipids). The cognate stimulus for a GPCR induces a conformational change that
catalyzes, via heterotrimeric G proteins, a series of biochemical reactions inside the cell
(i.e., a response); the responses can lead to phenomena such as differentiation,
morphological changes, chemotaxis, and mitosis. G protein-coupled receptors are the
targets of, in addition to endogenous neurohumoral molecules, more than 30% of
currently marketed pharmaceutical drugs. For instance, the effective appetite suppressant
fenfluramine reduces appetite via activation of serotonin (5-hydroxytryptamine, 5-HT)
2C (5-HT2C) receptors. Fenfluramine was, despite its efficacy as an anorexigen,
withdrawn from market due to its association with life-threatening cardiovascular side
effects [i.e., valvular heart disease (VHD) and pulmonary hypertension (PH)]. The work presented in this thesis describes efforts to identify the cellular and molecular mechanisms responsible for fenfluramine-induced VHD. Evidence is presented that both fenfluramine and, to a greater extent, its major in vivo metabolite norfenfluramine, activate mitogenic 5-HT2B receptors, which are expressed in human heart valve
interstitial cells (VICs). Both fenfluramine and norfenfluramine induce mitogenic
responses from VICs in vitro that are abrogated by a 5-HT2B receptor antagonist. Further,
xvi data implicating van der Waals’ (vdW) interactions between the α-methyl group of S-(+)- norfenfluramine and a valine in the second transmembrane helix of the receptor (V2.53) in the compound’s valvulopathic—but not anorexigenic—effects are described. The implication of the findings described herein on preventing drug-associated cardiopulmonary side effects and developing safe, effective anorexigens are also discussed.
xvii CHAPTER 1: INTRODUCTION
1.1 G Protein-Coupled Receptors: Overview
1.1.1 G Protein-Coupled Receptor Signal Transduction
G protein-coupled receptors (GPCRs) are proteins embedded in the plasma
membrane that act as sensors for a wide variety of extracellular stimuli (e.g., photons, neurohumoral modulators such as biogenic amines and peptides, odorants and tastants, and lipids, to name a few). Interactions of these extracellular stimuli with their cognate receptor(s) lead to conformational changes that enable interactions between the receptor and large heterotrimeric guanine nucleotide-binding proteins (G proteins). [For a review
of the role of G proteins in GPCR signal transduction, see (Gilman 1987; Casey and
Gilman 1988).] The interaction of the G protein with a GPCR stimulates the α subunit of
the former to release its constitutively-bound guanosine diphosphate (GDP) and to bind
guanosine triphosphate (GTP). Upon binding GTP, the α subunit dissociates from the β and γ subunits, which themselves remain associated. The dissociated α and βγ subunits then modulate the activity of plasma membrane-bound effector molecules such as nucleotide cyclases (e.g., adenylate and guanylate cyclases) and phospholipases (e.g., phospholipase C). The activity of these effector enzymes generates intracellular second messengers that modulate biochemical processes inside the cell. For example, stimulation of phospholipase Cβ (PLCβ) by Gαq leads to the hydrolysis of the plasma membrane
consitutent phosphatidylinositol-4,5-bisphosphate (PIP2) and generation of inositol-1,4,5-
1 trisphosphate (IP3) and diacylglycerol (DAG) (Hughes and Putney 1988) (Figure 1.1).
Inositol-1,4,5-trisphosphate, which is hydrophilic, stimulates the release of calcium stores
from the endoplasmic reticulum (Patterson, Boehning et al. 2004; Taylor, da Fonseca et
al. 2004). Diacylglycerol, which is hydrophobic, recruits to the plasma membrane and
activates protein kinase C (PKC) (Goni and Alonso 1999; Parker and Murray-Rust 2004).
Increased cytoplasmic calcium and activated PKC regulate the activity of many
intracellular proteins involved in diverse functions (e.g., gene transcription, apoptosis/anti-apoptosis, mitosis) (Catt and Balla 1989; Putney and Ribeiro 2000;
Ventura and Maioli 2001).
1.1.2 G Protein-Coupled Receptor Topology/Structure
G protein-coupled receptors are heptathelical proteins, having seven plasma membrane-spanning α-helices connected by intracellular and extracellular loops (Figure
1.2). For this reason, GPCRs are also called 7TM proteins/receptors. The first 7TM protein for which structural data were available was bacteriorhodopsin, the only plasma membrane protein expressed in certain halophilic bacteria. Unwin and Henderson
predicted, based on electron diffraction by halophilic bacteria membranes, that
bacteriorhodopsin contains seven membrane-spanning helices, each of which is
perpendicular to and embedded in the membrane (Unwin and Henderson 1975). The
complete amino acid sequence was later deduced biochemically (Ovchinnikov, Abdulaev
et al. 1977; Khorana, Gerber et al. 1979) and, based on secondary structure prediction
algorithms, was consistent with the predictions of Unwin and Henderson. Soon after, the
2 Figure 1.1: GPCR signal transduction: the phospholipase C pathway. The stimulus, in this case an agonist (such as a biogenic amine), binds to its cognate GPCR. A conformational change occurs that enables association between the intracellular portion of the GPCR and a heterotrimeric G protein. This association stimulates the α subunit of the G protein to release GDP and bind GTP, which, in turn, leads to its dissociation from the βγ subunits. The α subunit stimulates the activity of phospholipase Cβ, which hydrolyzes the membrane constituent phosphatidylinositol-4,5-bisphosphate (PIP2) to
yield inositol-1,4,5-trisphophate (IP3) and diacylglycerol (DAG). These compounds
stimulate the release of intracellular calcium stores from the endoplasmic reticulum (IP3) and activity of protein kinase C (calcium and DAG).
3 A Stimulus (Agonist) Phospholipase Cβ GPCR (second messenger-producing enzyme) Extracellular
Intracellular Gα GDP Gβ Gγ GTP Heterotrimeric G protein B Stimulus (Agonist) Phospholipase Cβ GPCR (second messenger-producing enzyme) Extracellular
Intracellular
Gα GTP Gβ Gγ GDP
Heterotrimeric G protein C Stimulus (Agonist) Phospholipase Cβ GPCR (second messenger-producing enzyme) Extracellular
DAG
Intracellular Second Gα Messengers
GTP
IP3 Gβ Gγ GDP
Heterotrimeric G protein
4 Figure 1.2: Schematic representation of a plasma membrane G protein-coupled receptor (GPCR). The α-helical membrane-spanning domains are shown as long, colored cylinders (numbered 1 to 7) connected by alternating intracellular (i1, i2, and i3) and extracellular (e1, e2, and e3) loops. The length of the intracellular and extracellular loops varies widely among GPCRs. However, i3 and e2 are typically the longest intracellular and extracellular loops, respectively.
5
e2 + NH3 e1 e3
Extracellular 5 6 7
4 3 2 1
Intracellular
i2 i1
- i3 COO
6 amino acid sequence of the retinal GPCR bovine rhodopsin was deduced biochemically
and the structure predicted to be similar to that of bacteriorhodopsin (Hargrave,
McDowell et al. 1983). Among the 7TM proteins that are receptors (GPCRs) for neurohumoral ligands (e.g., epinephrine), the first to be cloned was the hamster β2-
adrenergic receptor (Dixon, Kobilka et al. 1986)—a biogenic amine (catecholamine)
receptor—followed very soon thereafter by the cloning of other biogenic amine receptors
(Frielle, Collins et al. 1987; Fukuda, Kubo et al. 1987; Kobilka, Dixon et al. 1987;
Cotecchia, Schwinn et al. 1988; Regan, Kobilka et al. 1988; Lomasney, Lorenz et al.
1990; Schwinn, Lomasney et al. 1990; Lomasney, Cotecchia et al. 1991). All of the cloned biogenic amine GPCRs shared sequence homology with the rhodopsins: they were predicted, based on the amino acid sequence, to contain seven hydrophobic α-helices.
Thus, the seven transmembrane (7TM) motif emerged as a hallmark structural feature of
GPCRs (Figure 1.2).
In the 1990’s, global models for the arrangement of α-helices in GPCRs were proposed based on rhodopsin structural data, the primary sequence of many rhodopsin- like proteins, and/or theoretical considerations (Baldwin 1993; Baldwin 1994; Baldwin,
Schertler et al. 1997). These models were validated by the 2.8-Å X-ray diffraction data obtained from bovine rhodopsin crystals reported by Palczewski and colleagues
(Palczewski, Kumasaka et al. 2000). The published coordinates, which match very closely the predictions that preceded them, have since been used as a template for GPCR molecular model building and have led to a better understanding of GPCR-ligand interactions and of the dynamic processes that affect GPCR function (Ballesteros, Shi et
7 al. 2001; Stenkamp, Filipek et al. 2002; Archer, Maigret et al. 2003). Receptors for the indoleamine neurohumoral modulator serotonin, particularly serotonin (5- hydroxytryptamine, 5-HT) 5-HT2A receptors (see below), are examples for which rhodopsin-based homology modeling has led to a deeper understanding of the molecular determinants underlying ligand binding and receptor activation (Choudhary, Sachs et al.
1995; Sealfon, Chi et al. 1995; Almaula, Ebersole et al. 1996; Kristiansen, Kroeze et al.
2000; Shapiro, Kristiansen et al. 2000; Shapiro, Kristiansen et al. 2002).
1.2 Serotonin (5-HT) Receptors
Serotonin (5-hydroxytryptamine, 5-HT, Figure 1.3) is a biogenic amine that is involved in diverse biological processes. In the central nervous system, serotonin has been found to play role in the regulation of mood, perception, cognition, appetite, aggression, and anxiety (Bradley 1984; Roth 1994). In the periphery, serotonin has a well-characterized role in smooth muscle tone and platelet aggregation (Roth 1994;
Hoyer, Hannon et al. 2002; Kroeze, Kristiansen et al. 2002). The biological effects of serotonin are mediated through no fewer than 15 receptors belonging to seven families:
5-HT1, 5-HT2, 5-HT3, 5-HT4, 5-HT5, 5-HT6, and 5-HT7 (Kroeze, Kristiansen et al. 2002).
All but the 5-HT3 family, the members of which are ligand-gated cation channels
(Leonard 1992), are GPCRs. These receptors are the target of many therapeutic drugs, as well as drugs of abuse (Roth, Craigo et al. 1994; Nichols 2004). For example, drugs that activate 5-HT1 receptors exhibit anti-migraine efficacy (e.g., sumatriptan, methysergide)
(Diener 1994). Drugs that activate 5-HT4 receptors (e.g., cisapride) are useful in treating
8 Figure 1.3: Chemical structures of the indolamine 5-hydroxytryptamine (5-HT, serotonin), and the phenylisopropylamines 3,4-methylenedioxymethamphetamine
(MDMA, “Ecstasy”), 3,4-methylenedioxyamphetmaine (MDA), fenfluramine, norfenfluramine, phentermine, amphetamine, and tranylcypromine. All compounds share a benzene ring with a 2-aminoethyl group. The pheylisopropylamines bear a methyl group on the chiral α-carbon.
9
HO H NH N 3 NH3
CH3
5-Hydroxytryptamine (5-HT) Amphetamine Serotonin
H2 F3C N F3C NH3 CH2CH3
CH3 CH3
Fenfluramine Norfenfluramine
F3C NH3 F3C NH3
H2C
CH3 α-Desmethyl-norfenfluramine α-Ethyl-norfenfluramine
H2 N NH O O 3 CH3
CH3 CH3 O O
3,4-Methylenedioxymethamphetamine 3,4-Methylenedioxyamphetamine MDMA MDA
NH3
H3C CH3
Phentermine
10 gastroesophageal reflux disease (Dumuis, Sebben et al. 1989); drugs that block the 5-HT4 receptor (e.g., piboserod) may be effective in treating cardiac arrhythmia (Kaumann and
Sanders 1994). Blockade of 5-HT6 receptors has been shown to enhance cognition and memory, likely through modulation of cholinergic neurotransmission (Rogers and Hagan
2001; Woolley, Bentley et al. 2001; Leng, Ouagazzal et al. 2003; Lindner, Hodges et al.
2003). Also, some anti-psychotic medications (e.g., haloperidol, clozapine) block 5-HT6 and 5-HT7 receptors, suggesting a role for these receptors in schizophrenia (Roth, Craigo et al. 1994). However, of all the 5-HT receptors, those most widely targeted by
pharmacotherapies are the 5-HT2 receptors.
1.2.1 5-HT2 Receptors
The 5-HT2 receptor family comprises three GPCRs, namely the 5-HT2A, 5-HT2B, and 5-HT2C receptors. As mentioned above, these receptors transduce the presence of extracellular serotonin predominantly via the Gαq-PLCβ pathway, leading to increases in
cytosolic calcium and activation of protein kinase C (PKC) (Roth and Chuang 1987;
Roth, Willins et al. 1998) (Figure 1.2). However, these receptors can transduce signals
through alternate pathways, depending on the cells in which they are expressed and the
level of expression. For instance, in addition to coupling to the Gαq-PLCβ pathway, there
is evidence that the 5-HT2A and 5-HT2C receptors can inhibit adenylate cyclase via Gαi/o
(Garnovskaya, Nebigil et al. 1995; Lucaites, Nelson et al. 1996). The 5-HT2B and 5-HT2C receptors expressed at low levels can activate adenylate cyclase, presumably through Gαs
(Lucaites, Nelson et al. 1996). There is also evidence that 5-HT2 receptors can generate
11 arachadonic acid release via activation of phospholipase A2 (PLA2) (Berg, Maayani et al.
1996; Berg, Maayani et al. 1998; Berg, Maayani et al. 1998; Tournois, Mutel et al. 1998).
As can be inferred from the preceding, signal transduction of 5-HT2 receptors is complex.
While in most cells activation of the receptors leads to increases in cytosolic levels of the
second messengers IP3 and calcium, the receptors can modulate other second messengers
as well. These cell type differences in second messenger generation likely contribute to
the different 5-HT2 receptor-mediated physiological responses exhibited by different
tissues.
The 5-HT2 receptors have emerged as “stars” among 5-HT receptors in terms of their utility as targets for pharmacotherapies. Specifically, drugs active at 5-HT2 receptors
are effective psychiatric medications. For instance, 5-HT2 receptor antagonists exhibit
anti-depressant-like effects (Marek, Li et al. 1989; Fontaine 1993; Palvimaki, Roth et al.
1996). Also, 5-HT2 receptor expression is decreased by long-term treatment with anti-
depressants (Peroutka and Snyder 1980; Peroutka and Snyder 1980) and some anxiolytics
(Robinson, Alms et al. 1989). Many anti-psychotic medications, particularly atypical anti-psychotics, are potent 5-HT2 receptor antagonists (Meltzer 1989; Canton, Verriele et
al. 1994; Roth, Sheffler et al. 2004). The mood altering effects of these psychiatric drugs
are largely a function of their activity at the 5-HT2A receptor (Reviewed in Roth, Willins
et al. 1998).
Very recently, a novel clinical use for 5-HT2A receptor-blocking drugs (e.g.,
atypical antipsychotics) was reported. Elphick et al. presented evidence that the human
12 polyoma virus JCV, the causative agent of the demyelinating disease progressive
multifocal leukoencephalopathy (PML), infects oligodendrocytes in a 5-HT2A receptor- dependent fashion (Elphick, Querbes et al. 2004). Blocking 5-HT2A receptors with
selective antagonists and antibodies, as well as “knocking down” 5-HT2A receptor expression with anti-sense oligonucleotides, significantly decreased JCV infection of 5-
HT2A receptor-expressing glial and HeLa cells. Further, JCV virus co-localized with
intracellular 5-HT2A receptors, strongly suggesting that the virus uses the 5-HT2A receptor
to enter target cells. Thus, atypical anti-psychotics may represent a novel prophylactic for
patients at risk for PML (i.e., immunocompromised patients).
In addition to being effective treatments for psychiatric conditions, some drugs with activity at 5-HT2 receptors are potent appetite suppressors. A primary example of
this is the now-banned serotonergic, anorectic drug fenfluramine (Figure 1.3).
Fenfluramine is an amphetamine-like compound that causes non-exocytotic 5-HT release
(Fuxe, Farnebo et al. 1975; Garattini, Buczko et al. 1975; Carboni and Di Chiara 1989;
Schwartz, Hernandez et al. 1989; Series, Cowen et al. 1994); both it and its major in vivo
metabolite norfenfluramine also activate 5-HT2C receptors (Rothman, Baumann et al.
2000). This activity is believed to underlie fenfluramine’s anorectic actions (Vickers,
Dourish et al. 2001; Heisler, Cowley et al. 2002). However, as will be described in detail
in section 1.3, fenfluramine was voluntarily withdrawn from the US marketplace due to
potentially fatal cardiopulmonary side effects, likely resulting from its activation of 5-
HT2B receptors (Fitzgerald, Burn et al. 2000; Rothman, Baumann et al. 2000).
13 1.2.1.1 5-HT2B Receptors
1.2.1.2 Tissue Distribution of 5-HT2B Receptors
The 5-HT2B receptor (formerly the 5-HT2F receptor) was cloned from rat cDNA in
1992 and shown to have a pharmacology identical to that of the receptor responsible for
5-HT-stimulated contraction of rat stomach fundus (Foguet, Hoyer et al. 1992; Kursar,
Nelson et al. 1992; Wainscott, Cohen et al. 1993). In rodents, peripheral 5-HT2B receptor
expression has been demonstrated in both the endothelial and smooth muscle
compartments of the vasculature (Choi and Maroteaux 1996; Watts, Yang et al. 2001;
Villazon, Padin et al. 2002; Watts and Thompson 2004), in cardiac myocytes (Loric,
Launay et al. 1992; Choi and Maroteaux 1996; Choi, Ward et al. 1997; Nebigil, Etienne
et al. 2003; Bush, Fielitz et al. 2004; Jaffre, Callebert et al. 2004), in the smooth muscle
and nervous tissue of the gut (Loric, Launay et al. 1992; Choi and Maroteaux 1996; Cox and Cohen 1996; Fiorica-Howells, Maroteaux et al. 2000; Enguix, Sanchez et al. 2003), and in osteoblasts (Bliziotes, Eshleman et al. 2001). In the rodent central nervous system,
5-HT2B receptor expression has been observed in the dorsal raphé nucleus (Bonaventure,
Guo et al. 2002), in the cerebellum, lateral septum, dorsal hypothalamus, and medial amygdala (Choi and Maroteaux 1996; Duxon, Flanigan et al. 1997), and in astrocytes
(Hirst, Cheung et al. 1998), as well as in developing enteric neurons (Fiorica-Howells,
Maroteaux et al. 2000). The 5-HT2B receptor has also been detected in rodent cochlea
(Oh, Drescher et al. 1999). Thus, the major sites of 5-HT2B receptor expression in rodents
are the cardiovascular compartment, smooth muscles of the gut, and the central nervous
system.
14 A grossly similar expression pattern is evident in human tissues. That is,
peripheral expression of 5-HT2B receptors has been demonstrated in the vasculature
(Ullmer, Schmuck et al. 1995; Ullmer, Boddeke et al. 1996; Ishida, Kawashima et al.
1998); in the heart, particularly in the valves (Fitzgerald, Burn et al. 2000), and in the gut,
namely in the smooth muscles of the colon (Borman, Tilford et al. 2002) and in the
pancreas, liver, spleen, and kidneys (Bonhaus, Bach et al. 1995). Recently, 5-HT2B receptor expression has also been shown in immature dendritic cells (Idzko, Panther et al.
2004) and in the ciliary body (Chidlow, Hiscott et al. 2004). In the central nervous system, 5-HT2B receptor expression has been observed in the brain (Kursar, Nelson et al.
1994; Bonhaus, Bach et al. 1995; Kong, Peng et al. 2002), in the spinal cord (Helton,
Thor et al. 1994), and in the meninges (Schmuck, Ullmer et al. 1996). Thus, the distribution of 5-HT2B receptor expression in human, as well as in rodent, tissues implies
important roles for the receptor in cardiovascular, digestive, and neurological processes.
Further, the similarity in 5-HT2B receptor tissue distribution between rodents and humans
imply similar physiological roles for the receptor.
1.2.1.3: Signal Transduction of 5-HT2B Receptors
The intracellular signaling of 5-HT2B receptors has been studied in great detail. In
cell lines expressing cloned 5-HT2B receptors (Wainscott, Cohen et al. 1993; Loric,
Maroteaux et al. 1995; Lucaites, Nelson et al. 1996; Schmuck, Ullmer et al. 1996;
Bonhaus, Flippin et al. 1999; Bymaster, Nelson et al. 1999; Fitzgerald, Burn et al. 2000;
Manivet, Mouillet-Richard et al. 2000; Rothman, Baumann et al. 2000; Cussac, Newman-
15 Tancredi et al. 2002; Manivet, Schneider et al. 2002; Newman-Tancredi, Cussac et al.
2002; Alberts, Chio et al. 2003; Blanpain, Le Poul et al. 2003; Schaerlinger, Hickel et al.
2003; Setola, Hufeisen et al. 2003), as well as in cell lines bearing endogenous 5-HT2B receptors (Kellermann, Loric et al. 1996; Ullmer, Boddeke et al. 1996), activation of the receptor has been shown to stimulate phospholipase C-mediated inositol 1,4,5- trisphosphate production (Figure 1.4). In addition to its ability to stimulate phospholipase
C, activation of endogenous 5-HT2B receptors in murine 1C11* cells was shown to
stimulate phospholipase A2, resulting in arachadonic acid release (Tournois, Mutel et al.
1998). Elevations in intracellular calcium concentration following activation of
exogenous (Bonhaus, Flippin et al. 1999; Porter, Benwell et al. 1999; Jerman, Brough et
al. 2001; Porter, Malcolm et al. 2001; Blanpain, Le Poul et al. 2003) and endogenous
(Cox and Cohen 1995; Ullmer, Boddeke et al. 1996; Ishida, Kawashima et al. 1998;
Sanden, Thorlin et al. 2000) 5-HT2B receptors have also been demonstrated (Figure 1.4).
While this effect in most cellular and tissular contexts is likely due to inositol 1,4,5-
trisphosphate-mediated release of endoplasmic reticulum calcium stores, activation of 5-
HT2B receptors on stomach fundus appears to modulate cytoplasmic calcium levels not
through release of intracellular, ryanodine-sensitive stores but rather through voltage-
gated, nitrendipine-sensitive channels (Cox and Cohen 1995). Also, Ullmer et al. showed
that activation of pulmonary artery endothelial cell 5-HT2B receptors causes release of
ryanodine-sensitive calcium stores independently of IP3 (Ullmer, Boddeke et al. 1996).
Signal transduction through 5-HT2B receptors can also involve the second
messenger nitric oxide (Figure 1.4). Manivet and colleagues showed, both in murine
16 Figure 1.4: Signal transduction pathways modulated by 5-HT2B receptors. The
various pathways regulating cell proliferaction/survival shown to be affected by 5-HT2B receptors are schematized. The pathways can be divided into three categories: second messenger generation, mitogenesis, and anti—apoptosis. The second messenger- generating pathway involves 5-HT2B receptor-mediated activation of phospholipase Cβ,
which leads to hydrolysis of membrane phosphatidylinositol-4,5-bisphosphate into
inositol-1,4,5-trisphosphate (IP3) and diacylglycerol (DAG). IP3, which is hydrophilic,
activates ligand-gated Ca++ channels on the membrane of the endoplasmic reticulum,
causing release of intracellular Ca++ stores. DAG, which is hydrophobic, recruits to the
plasma membrane and, along with the increased intracellular Ca++, acitvates protein
kinase C (PKC). Interactions between the 5-HT2B receptor with nitric oxide synthase
stimulate the latter to generate the second messenger nitric oxide. The mitogenic pathway
has two branches, both of which start with 5-HT2B receptor-mediated activation of Src.
The first branch involves Src-mediated transactivation of the platelet-derived growth
factor receptor, which leads to activation of MAPK and, subsequently, cyclin D1/cdk4.
The second branch involves Src-dependent activation of cyclin E/cdk2 via a MAPK- independent mechanism. Both branches lead to phosphorylation of retinoblastoma protein, which then dissociates from the transcription factor E2F. E2F then translocates to the nucleus where it stiumlates transcription of genes involved in DNA synthesis. The
anti-apoptotic pathway involves the down-regulation of the mitochondrial permeablility
factors Bax and ANT-1 via 5-HT2B receptor-mediated activation of MAPK and PI3K,
respectively. Down-regulation of Bax and ANT-1 lead to decreased cytoplasmic levels of
cytochrome c, thereby reducing cytochrome c-mediated activation of caspase 9 activity.
17
5-HT2B Receptor PDGFR NO Extracellular PLCβ
DAG Intracellular Gα P NO 13 Gαq PKC P NOS PKC P GTP P GTP Ca++ Gβ Gγ P P IP3 Ca++ Target protein Ca++ Ca++ phosphorylation MAPK PI3K Ca++ Src
Ca++ Ca++ Ca++ Ca++ Ca++ Ca++ Ca++ CytC CytC ++ Ca++ ++ ++ ++ Ca ++ Ca Ca Ca ++ ++ Ca ++ Ca Cyclin Ca ++ ++ Ca ++ Cdk4 Cdk2 Cyclin Ca Ca Ca D1 E Bax Ant-1 Endoplasmic Reticulum
CytC CytC CytC Rb E2F P CytC P P CytC
Nucleus CytC CytC CytC CytC CytC Go S Mitochondrion
18 1C11* cells that endogenously express 5-HT2B receptors and in mouse fibroblasts
cNOS). The iNOS- and cNOS-stimulating activities of the 5-HT2B receptor were shown
to depend on the receptor’s carboxy-terminal, consensus type I target sequence for PDZ
domain-containing proteins (Manivet, Mouillet-Richard et al. 2000). In addition, the
activation of iNOS, but not that of cNOS, involved a heterotrimeric G protein, Gα13, as
antibodies against Gα13, but not those against Gαq, Gαi, or Gαs, eliminated the agonist-
induced response (Manivet, Mouillet-Richard et al. 2000).
1.2.1.4: Biochemical Consequences of 5-HT2B Receptor Activation
The results of several studies show that in many cell types and physiological
contexts, mitosis represents a “final common pathway.” Activation of endogenous
(Fiorica-Howells, Maroteaux et al. 2000; Setola, Hufeisen et al. 2003; Bhasin, Kernick et
al. 2004) and exogenous (Launay, Birraux et al. 1996; Fitzgerald, Burn et al. 2000;
Nebigil, Launay et al. 2000; Nebigil, Etienne et al. 2003) 5-HT2B receptors leads to
activation of the mitogen-activated protein kinases (MAPK) Erk 1/2 (p42/44). In mouse fibroblasts stably expressing 5-HT2B receptors, but not in non-transfected cells, Launay
and colleagues showed a rapid and transient activation of the proto-oncogene product
p21Ras in response to 5-HT (Launay, Birraux et al. 1996). The 5-HT-stimulated p21Ras activity led to activation of MAPK, which was completely abolished by antibodies
Ras against Gαq, Gβγ, and p21 (Launay, Birraux et al. 1996). Furthermore, the 5-HT2 receptor inverse agonist ritanserin, which completely blocks the 5-HT-stimulated activation of p21Ras, also abrogated both the constitutive and agonist-stimulated
19 mitogenic response in mouse fibroblasts expressing the 5-HT2B receptor (Launay, Birraux
et al. 1996). Nebigil et al. found, in the same cells, that activation of the 5-HT2B receptor
leads to activation of the oncogenic, cytoplasmic tyrosine kinase Src, which in turn
induces both Cyclin E and Cyclin D1 via separate pathways (Nebigil, Launay et al.
2000). Serotonin 2B receptor-mediated induction of Cyclin D1 occurs throug Src-
mediated activation of the platelet-derived growth factor receptor-tyrosine kinase
(PDGFR), which in turn activates MAPK (Nebigil, Launay et al. 2000). The induction of
Cyclin E, in contrast, while sensitive to inhibitors of Src, is insensitive to inhibitors of
PDGFR and MAPK (Nebigil, Launay et al. 2000). The activated cyclins
hyperphosphorylate retinoblastoma protein (pRb) (Nebigil, Launay et al. 2000), leading
to disinhibition of the transcription factor E2F and transactivation of DNA replication-
associated genes. In the developing mouse myocardium, as in mouse fibroblasts, the
mitogenic signaling of the 5-HT2B receptor appears to involve transactivation of a growth
factor receptor tyrosine kinase, namely the epidermal growth factor receptor ErbB-2,
since the expression of this protein is reduced in mutant mice lacking the receptor
(Nebigil, Choi et al. 2000).
The 5-HT2B receptor has also been shown to display anti-apoptotic activity. For instance, in developing mouse embryos, treatment with the 5-HT2 receptor inverse
agonist ritanserin, at a time when 5-HT2B receptors are the main 5-HT2 receptors
expressed, induced apoptosis in the cephalic region, the heart, and the neural tube (Choi,
Ward et al. 1997). In mouse myocardial fibroblasts from wild type mice, but not in those obtained from mutant mice lacking 5-HT2B receptors, treatment with 5-HT greatly
20 reduced the apoptosis induced by serum deprivation (Nebigil, Etienne et al. 2003). The
anti-apoptotic activity of 5-HT2B receptors was shown to be dependent on both the
MAPK and PI3 kinase (PI3K) pathways and was due to 5-HT-induced down-regulation
of the mitochondrial permeability regulators Bax and ANT-1, thereby decreasing
cytochrome C release (Nebigil, Etienne et al. 2003). The importance of the 5-HT2B receptor in cell proliferation and survival during myocardial development is further underscored by the embryonic and neonatal lethality of targeted disruption of 5-HT2B receptor expression. In mice bearing such a mutation, histological examination of the heart revealed decreased cell number in the ventricular trabeculae, as well as abnormal sarcomeric organization in the sub-epicardial layer (Nebigil, Choi et al. 2000).
Thus, the studies reviewed in this section collectively demonstrate the complex nature of 5-HT2B receptor signaling. In addition to modulating the activity of the second
messenger-producing phospholipases C and A2 and NOS, 5-HT2B receptors also regulate
the activity of intraceullar protein kinases such as MAPK, PI3K, and Src. These effects
are transduced through the heterotrimeric G proteins Gαq and Gα13, small G proteins
(p21Ras), protein-protein interactions with effector molecules such as nitric oxide synthase
(via the carboxy-terminal PDZ motif), and cross-talk with growth factor receptors. In
addition to regulating cell function (e.g., smooth muscle cell contraction and relaxation in
the gut and blood vessels), major consequences of 5-HT2B receptor signaling are mitotic
and anti-apoptotic responses that play an important role during development, particularly
in the heart. Also, as will be revealed in the sections that follow, the proliferative/anti-
apoptotic signals shown to emanate from 5-HT2B receptors are consistent with the
21 receptor’s involvement in drug-induced fibroplasia in the valves of the heart (VHD) and in the wall of the pulmonary artery (PH).
1.3 The Rise of Fenfluramine
In 1973, the United States Food and Drug Administration (FDA) first approved the amphetamine derivative fenfluramine [trade name: Pondimin®; chemical name: N- ethyl-α-methyl-3-(trifluoromethyl)-benzeneethanamine (Figure 1.3)] for use as an anorexigen, or appetite suppressant. Twenty-three years later, the FDA approved the (+) rotamer, dexfenfluramine, for use as an anorexigen. However, prior to its introduction into the US marketplace, the drug had a long history of clinical trial as an adjunct to calorie restriction and other methods of weight reduction (Table 1). While the results of such studies vary widely, likely due mainly to differences in protocols, dosing, study duration, and patient medical and genetic backgrounds, a general trend is apparent: in the first three to six months, patients taking fenfluramine lose more weight (between 0.3 kg/mo and 4.6 kg/mo) than those taking placebo (Ranquin and Brems 1977; Stunkard,
Craighead et al. 1980; Weintraub, Sriwatanakul et al. 1983; Weintraub, Hasday et al.
1984; Finer, Craddock et al. 1988; Guy-Grand, Apfelbaum et al. 1989; Finer 1992;
Mathus-Vliegen 1993; Stewart, Stein et al. 1993; O'Connor, Richman et al. 1995; Van
Gaal, Vansant et al. 1995; Breum, Moller et al. 1996; Ditschuneit, Flechtner-Mors et al.
1996; Pedrinola, Sztejnsznajd et al. 1996; Chow, Ko et al. 1997). Also, despite an apparent reduction in efficacy after six months, patients who stopped taking fenfluramine regained more weight than patients taking placebo (Stunkard, Craighead et al. 1980;
22 Table 1.1: Clinical studies of the efficacy of appetite suppressants. Weight reduction from several studies is indicated, as are the duration of drug therapy, the drugs studied, and whether a placebo group was included. Other effects reported in the studies are included in the Notes column.
23 Author (Year) Duration Drug(s) (N) Placebo, N Weight Reduction Notes F more effective than DI in reducing bingeing and vomiting Blouin (1988) ? F (?), DI (?) ? as well as psychological symptoms in bulimics Bulimic patients consumed fewer calories 2 hrs. after F dose Robinson (1985) 2 hrs. F (15) 15 No change than placebo group Decrease in calories consumed; decrease in CSF 5-HIAA, Shoulson (1975) 8 ds. F (7) 7 0.64 kg upon cross-over increase in CSF HVA Decrease in BP; F group gained 1.08 kg when switched to Brodbin (1967) 1.5 mos. F (20) 20 5.01 kg vs. 0.64 kg placebo for 1.5 mos. after F Lele (1972) 1.5 mos. F (21) 21 2.4 kg vs. 0.1 kg Sainani (1973) 1.5 mos. F (23) 27 2.70 kg vs. 0.76 kg Placebo group had taken fenfluramine for 1.5 mos. prior to Sainani (1973) 1.5 mos. F (27) 23 2.59 kg vs. 0.17-kg gain being switched to placebo. Prader-Willi syndrome patients; improvement in food- Selikowitz (1990) 1.5 mos. F (14) 14 0.9 kg vs. 0.1-kg gain related behavior observed in F group FBG and post-prandial BG, TG, chol. lower and HDL higher Salmela (1981) 1.75 mos. F (?) ? ? in F group than in placebo group; fasting insulin unchanged Increase glucose tolerance and insulin response, decreased Ranquin (1977) 2 mos. F (15) 15 10.18 kg vs. 0.89 kg serum chol. and TG in F Follows (1971) 2 mos. F (10), DP (7) - 1.6 kg, 1.5 kg F decreased BP and pulse, DP increased BP and pulse Weintraub (1983) 2.5 mos. F (?) ? 5.9 kg vs. 3.3 kg Munro (1966) 3 mos. F (25) 25 4.2 kg vs. 0.2-kg gain F group gained 0.5 kg when switched to placebo for 3 mos. Traherne (1965) 3 mos. F (29) 29 3.18 kg vs. 1.04 kg after F After 12-mo. follow-up, F/BT group gained 10.7 kg, F group Stunkard (1980) 6 mos. F (32), F/BT (25) BT (23) 14.5 kg, 15.3 kg vs. 10 kg gained 8.2 kg, BT group gained 1.9 kg. Adverse effects less frequent with F/P than with either drug Weintraub (1984) 6 mos. F (?), P (?), F+P (?) ? 7.5 kg, 10 kg, 8.4 kg vs. 4.4 kg alone, despite halved doses of F and P in the combination Also looked at intermittent F, which was less effective than Steel (1973) 9 mos. F (31), P (32) - 11.9 kg vs. 11.8 kg continuous F Weight loss for >200 ng/ml, 100-199 ng/ml, <100 ng/ml Innes (1977) 5 mos. F (?) - 9 kg, 5 kg, 2 kg plasma fenfluramine, respectively Widhalm (1976) 18 mos. F (1) - 26 kg Prader-Willi patient BG, insulin, chol., TG did not change; BP decreased; NE Andersson (1991) 4 ds DF (?) ? No change in weight and renin decreased Ditschuneit (1993) 1 wk. DF (9) 9 Slight reduction for both groups Lower BP, insulin, and FFA in the DF group Massand (1997) 3 wks.-5 mos. DF (6) - 0.9-10 kg Bulimic patients; DF group experienced decreased bingeing Russell (1988) 1.5 mos. DF (21) 21 No change and purging/vomiting Fahy (1993) 2 mos. DF (20) 23 No change Abnormal eating improved equally in both groups Binge eating decreased 3 times faster in DF group; return to Stunkard (1996) 2 mos. DF (12) 16 0.07 kg vs. 0.2-kg gain baseline at 4-mo. follow-up Van Gaal (1995) 3 mos. DF (?) ? 16 kg vs. 12.8 kg Chow (1997) 3 mos. DF (?) ? 1.2 kg/m2 vs. 0.1 kg/m2 Significant decreases in FBG and HbA1c Stewart (1993) 3 mos. DF (?) ? 3.8 kg vs. 0.3-kg gain FBG, HbA1c, TG reduced in the DF group, increased in the
24 placebo group Schizophrenics on neuroleptics; no deterioration in mental Goodall (1988) 3 mos. DF (9) 7 5.4 kg vs. 2.8 kg. state Finer (1988) 3 mos. DF (17) 17 2.9 kg vs. 0.7-kg gain No change in FBG, HDL, TG; decreased BP for DF group; Enzi (1988) 3 mos. DF (64) 69 8.1 kg vs. 3.5 kg reduced serum chol. in both groups Finer (1988) 3/6 mos. DF (29) - 5.7 kg and 7.0 kg Breum (1994) 3.75 mos. DF (43), EC (38) - 6.9 kg vs. 8.3 kg (NS) BP significantly reduced for both groups Patients gained 2.6 kg in the 2-mo. interimc; patients gained Ditschuneit (1996) 6 mos. DF (25) - 3.7 kgc, 4.8 kgd 0.8 kg in the 2-mo. interimd Breum (1996) 6 mos. DF (16) 13 12.8 kg vs. 13.8 kg (NS) Trp/LNAA and Tyr/LNAA reduced for both DF and placebo DF group had decreased TG and fasting insulin and O’Connor (1995) 6 mos. DF (27) 24 9.7 kg vs. 4.9 kg increased HDL (females); placebo group worsened Finer (1992) 6.5 mos. DF (?) ? 5.8 kg vs. 2.9 kg Pedrinola (1996) 8 mos. DF + Flx (20) Flx (13) 13.4 kg vs. 6.2 kg. HDL increased, serum lipid decreased DF group gained 3.24 kg 2 mos. after study termination, vs. Mathus-Vleigen 12 mos. DF (?) ? 12.8 kg vs. 8.6 kg 0.82 kg for the placebo group; health risk factors improved (1993) for both groups 14.2 kg vs. 4.92 kga Decrease in systolic BPa,b, BGa, insulina, TGa; increase in Ditschuneit (1996) 12 mos. DF (12a/16b) 10a/6b 11.1 kg vs. 2.6 kgb HDLa Guy-Grand (1989) 12 mos. DF (295) 268 9.82 kg vs. 7.15 35/40/30% REW vs. 25/25/20% Andersen (1992) 4/6/12 mos. DF (17) 13 REW Greater dietary compliance and decreased anxiety about (1994) 14 mos. DF (?) ? eating behaviors in DF group Redmon (1999) 6/12 mos. F+P (13/8) 13/8 9.6/8.1 kg vs. 2.7/2.5 kg Significiant decreases in HbA1c, FPG, TG Weintraub (1992) 8.5 mos. F+P (58) 54 14.2 kg vs. 4.6 kg BP decreased for both groups Siginificant decreases in FBG, total chol., TG, BP, insulin Spitz (1993) 12/24 mos. F+P (298) - 8.4 kg/m2 and 7.6 kg/m2 (fast.) Wadden (1998) 12/24 mos. F+P (21) - 17.1% and 13.9% F (?), DA (?), P (?), C 4 kg, 2.5 kg, 2.5 kg, 1.5 kg, 1.5 Munro (1973) 3 mos. ? (?), M (?), DP (?) kg, 1.5 kg vs. 0 kg BED patients, 12 (75%) of whom were binge-free after the Devlin (1999) 5 mos. P+Flx (16) - 8.6 kg active phase Campbell (1977) 6 mos. P (34) 32 5.3 kg vs. 1.4 kg No significant changes in diabetic control, BP, serum chol. Campbell (1977) 6 mos. P (14) - 6.5 kg Munro (1968) 10 mos. P (?) ? 13 kg vs. 5 kg F = fenfluramine (racemic); DF = dexfenfluramine; P = phentermine; Flx = fluoxetine; EC = ephedrine and caffeine; DA = dexamphetamine; C = chlorphentermine; M = mazindol; DP = diethylproprion; BT = behavioral therapy; DI = desipramine; BMI = body mass index; (F)BG = (fasting) blood glucose; TG = triglycerides; BP = blood pressure; HbA1c = hemoglobin A1c; FFA = free fatty acids; NE = norepinephrine; aindividuals with upper body obesity; bindividuals with lower body obesity; cpatients taking DF 2 mos. after terminating a 12-mo. study during which they had taken DF; dpatients taking DF 2 mos. after terminating a 12-mo study during which they had taken placebo; 5-HIAA = 5-hydroxyindole acetic acid; HVA = homovanillic acid; CSF = cerebrospinal fluid
25 Mathus-Vliegen 1993; Ditschuneit, Flechtner-Mors et al. 1996). This latter observation
shows that fenfluramine continues to have a weight-reducing effect even beyond six
months treatment.
In addition to its contributory effects on weight loss, fenfluramine exhibited, in
many clinical trials, efficacy in improving metabolic and cardiovascular risk factors
associated with obesity, such as blood levels of glucose, insulin, norepinephrine,
glycosylated hemoglobin (HbA1c), free fatty acids, cholesterol, triglycerides, and high-
density lipoproteins, and both systolic and diastolic blood pressure (Mathus-Vliegen
1993; O'Connor, Richman et al. 1995; Ditschuneit, Flechtner-Mors et al. 1996; Pedrinola,
Sztejnsznajd et al. 1996). While some of these latter effects were likely secondary to the
weight loss induced by the drug, some studies reported fenfluramine-induced reductions
in blood pressure, as well as decreases in plasma levels of norepinephrine, free fatty acids, and insulin, occurring within one week of the treatment (Andersson, Zimmermann et al. 1991; Ditschuneit, Flechtner-Mors et al. 1993). This early onset suggested that fenfluramine can modify these parameters independent of its effects on weight loss.
Along with its salutary effects on the biochemical parameters enumerated above, fenfluramine also appeared to have beneficial effects on food-associated attitudes and
behaviors. For instance, a few studies reported a fenfluramine-induced relief of the
psychological symptoms of bulimia nervosa (Blouin, Blouin et al. 1988; Russell,
Checkley et al. 1988), improvement in the food-related behaviors associated with Prader-
Willi syndrome (Selikowitz, Sunman et al. 1990), and reduction in anxiety about eating
26 behaviors in obese individuals (1994). These limited studies suggested that, in addition to
their ability to modulate food intake and metabolic and cardiovascular processes,
fenfluramine could also normalize disordered, food-centered affects.
Another amphetamine derivative, phentermine [trade names: Adipex®, Fastin®,
Ionamin®; chemical name: α,α-dimethyl-benzeneethanamine (Figure 1.3)], had been in
use as an appetite suppressant since its approval by the FDA in 1959. However, there are fewer clinical studies of its effectiveness in weight loss programs. Munro and colleagues
(Munro, MacCuish et al. 1968) studied the effectiveness of phentermine during a 36- week trial and reported that the therapy resulted in a 13% reduction of initial body weight, compared to a 5.2% reduction for placebo.
In 1984, researchers at Rochester University first reported the short-term use of fenfluramine and phentermine in combination (Weintraub, Hasday et al. 1984). The rationale was that both drugs, equally effective in terms of weight loss, have contrasting side effect profiles: fenfluramine causes drowsiness, diarrhea, and dry mouth, while phentermine causes insomnia and irritability. By combining the two, Weintraub et al. showed that half as much of each drug was as effective as a full dose of either and better tolerated in terms of side effects. Then in 1992, the same group studied the efficacy of long-term fenfluramine and phentermine combination therapy (Weintraub, Sundaresan et al. 1992). In the first phase, a double-blind 34-week study, patients with a mean ± SEM ideal body weight of 154% ± 12% received either placebo or fenfluramine (60 mg) plus phentermine (15 mg) daily as an adjunct to behavior modification, calorie restriction, and
27 exercise. After 34 weeks, the patients receiving the drug combination lost, on average,
14.2 ± 0.9 kg; the placebo group lost, on average, 4.6 ± 0.8 kg. This represents an almost
three-fold additional, statistically significant weight loss compared to diet and placebo. In
addition, using visual analog scales, patients taking the drug combination reported it to be
“not bothersome” to a similar extent as those taking placebo (17.4 ± 0.3 vs 13.5 ± 0.2).
Many patients did, however, complain of dry mouth. Thus, the combination of fenfluramine and phentermine was three times more effective vis-à-vis weight loss and tolerated only slightly worse than placebo, demonstrating that combining the two poorly tolerated drugs reduces the side effects of both without compromising overall effectiveness.
In the second phase, all participants from the first phase received the fenfluramine-phentermine combination, either daily or intermittently, for a period of 70 weeks (Weintraub, Sundaresan et al. 1992). During the entire, two-year study (phases 1 and 2), patients lost an average of 10.8 ± 0.8 kg, demonstrating the sustained, long-term effectiveness of the drug combination. With respect to side effects, dry mouth continued to be the major complaint (41%); however, only 20 of the original 121 participants dropped out due to adverse effects, demonstrating that most patients tolerated well the fenfluramine-phentermine combination over the course of two years. Later phases of the study followed patients for more than 3years and many maintained a weight loss greater
than or equal to 10% of their original body weight (Weintraub, Sundaresan et al. 1992).
28 These findings precipitated the widespread use of the fenfluramine-phentermine combination, which came to be known as “Fen-Phen.” The dual administration of the two anorexigens had never been approved by the FDA, nor had the long-term safety of the therapy ever been established. Nevertheless, weight loss clinics specializing in Fen-Phen therapy were established throughout the country. In 1996, the FDA narrowly approved the more potent, less adverse-effect-prone (+) stereoisomer of fenfluramine, dexfenfluramine, for less-than-one-year use in the treatment of obesity. The New York
Times reported that in 1996, 18 million prescriptions had been written for fenfluramine alone or in combination with phentermine, and that about 6 million Americans took the drug.
1.4 The Fall of Fenfluramine
Fenfluramine’s time at the top of the diet drug market ended in 1997 with a report in the New England Journal of Medicine from researchers at the Mayo Clinic. Connolly and colleagues (Connolly, Crary et al. 1997) presented 24 cases of newly-documented heart valve abnormalities in patients taking Fen-Phen for a duration of 1 to 28 months.
Specifically, the authors discovered, upon echocardiography, unusual valvular morphology and regurgitation affecting the valves on both sides of the heart, though in all patients at least one left-sided valve was affected. In those cases requiring surgical intervention, the abnormal valves displayed glistening white leaflets and chordae, and diffuse thickening. They noted that the histopathology was identical to the valvular heart disease (VHD) associated with carcinoid syndrome and ergotamine use. Microscopic
29 examination of the thick, white lesions revealed proliferative myofibroblasts in an abundant extracellular matrix containing CD3-positive T-cells (Connolly, Crary et al.
1997; Steffee, Singh et al. 1999).
In light of these findings, the FDA obtained echocardiographic data from several
US health care facilities. An uncontrolled survey of 284 patients taking fenfluramine in some form for a median of 14 months (1997) study revealed VHD among 34% of the patients, a much larger prevalence than in the general population (Singh, Evans et al.
1999). Based on these data, the FDA recommended the voluntary withdrawal of fenfluramine and dexfenfluramine from the US marketplace. In September of 1997,
Wyeth (a subsidiary of American Home Products), the manufacturer of fenfluramine, complied with the FDA’s recommendation. Since that time, retrospective studies have addressed the incidence of VHD among fenfluramine users, with widely disparate findings (Loke, Derry et al. 2002). In five case-controlled studies of fenfluramine users finding significantly greater aortic and/or mitral valve regurgitation in the treated group compared to controls, the lowest incidence of VHD was 11 cases per 9765 patients
(0.1%) treated for an average of 1 month; the highest incidence was 25.2% among patients treated for an average duration of 20 months (Khan, Herzog et al. 1998;
Weissman, Tighe et al. 1998; Burger, Sherman et al. 1999; Gardin, Schumacher et al.
2000; Jollis, Landolfo et al. 2000). Thus, several controlled studies reproduced the findings of the FDA and suggested that long-term fenfluramine use was associated with a significantly increased risk for developing VHD, though potentially less than that initially suggested by the FDA survey.
30 Fenfluramine use has also been linked with pulmonary hypertension (PH). Brenot and colleagues found that 20% of 73 PH patients had a history of fenfluramine use
(Brenot, Herve et al. 1993). This report precipitated a controlled study of 95 PH patients and 355 controls, which revealed a 23-fold increased risk for developing PH after 3 months of fenfluramine use (Abenhaim, Moride et al. 1996). Connolly et al. also found eight cases of newly-documented PH—a higher frequency (33%) than observed in the general population (Connolly, Crary et al. 1997). Like the lesions associated with VHD, pulmonary arteries from PH patients typically manifest hyperplasia in both the intima
(endothelial layer) and the media (the layer containing smooth muscle and elastic tissue)
(Mark, Patalas et al. 1997; Strother, Fedullo et al. 1999; Tomita and Zhao 2002). Thus,
VHD and PH are similar from a histopathological standpoint in that proliferative plaques that compromise tissue integrity and function occur in both conditions.
Findings from Dr. Luc Maroteaux’s laboratory implicate activation of 5-HT2B receptors by the fenfluramine metabolite norfenfluramine in PH (Launay, Herve et al.
2002). In this regard, mice exposed to chronic hypoxia develop PH. Treatment with S-
(+)-norfenfluramine exacerbates hypoxia-induced PH. Interestingly, mice lacking 5-HT2B receptors do not develop PH following exposure to hypoxia, whether or not they are treated with S-(+)-norfenfluramine. Examination of the pulmonary arteries of mice with hypoxia- and hypoxia+S-(+)-norfenfluramine-induced PH reveals smooth muscle cell proliferation. This observation, coupled with 1) the mitogenic activity of 5-HT2B receptors described above, and 2) the expression of 5-HT2B receptors in pulmonary artery tissue provide strong evidence that fenfluramine-associated PH is due, at least in part, to
31 Figure 1.5: Amphetamine actions on biogenic amine reuptake, storage, and release.
A. The plasma membrane monoamine transporter (MAT), a monoamine-Na+ symporter,
+ transports released monoamines into the cytoplasm of the cell using the Na gradient for energy. Cytoplasmic monoamines, either newly-synthesized or those transported back into the cell, are stocked into exocytic vesicles through the vesicular monoamine transporter (VMAT). The lumen is acidic due to vesicular H+-ATPase pumps, and the
VMAT, a monoamine-H+ antiporter, uses the H+ gradient as energy for stocking the monoamines. B. Amphetamines (AMPH) act as competitive inhibitors of MAT, thereby blocking monoamine reuptake. AMPH attains the cytoplasm via MAT, where it then inhibits VMAT. The build-up of monoamines in the cytoplasm then causes MAT to work in reverse, releasing monoamines in a non-exocytic fashion.
32 H+ A H+-ATPase
H+ Exocytic H+ H+ + Vesicle H
Synthesized Monoamines VMAT
H+ Na+
Intracellular
MAT
+ Na+ Extracellular Na Na+ Released Na+ Na+ Monoamines Na+ Na+
H+ B H+-ATPase Exocytic Vesicle
H+ H+ H+ H+
AMPH VMAT Synthesized + Monoamines H Na+
Intracellular
MAT
Extracellular + + Na Na AMPH Na+ Na+ Na+ Na+ Na+ Released Monoamines
33 activation of pulmonary artery 5-HT2B receptors by the fenfluramine metabolite norfenfluramine.
1.5 Biological Activity of Fenfluramine: Generalities of Amphetamine Action
Both fenfluramine and phentermine are derivatives of amphetamine [chemical name: α-methyl-benzeneethanamine (Figure 1.3)]. Amphetamine-like compounds exert many of their characteristic biological effects by causing release of biogenic amines from neurons and other cells capable of synthesizing them, stocking them, releasing them, and taking them up from the extracellular milieu (see Jones, Gainetdinov et al. 1998 for review). Drugs related to amphetamine elicit biogenic amine release from cells in a multi- step process (Figure 1.5) that first involves transport of the amphetamine into the cell via biogenic amine-sodium chloride symporters known as monoamine transporters (MATs).
The MATs normally transport biogenic amines back into the cells that release them, using dissipation of the sodium gradient across the plasma membrane for energy.
Amphetamine-like compounds also act as substrates for these MATs and, through them, enter the cytoplasm of cells that synthesize, stock, and release biogenic amines.
The second step of amphetamine-mediated biogenic amine release recruits another type of monoamine transporter called vesicular monoamine transporters
(VMATs). Vesicular monoamine transporters are located on the surface of the exocytic vesicles into which biogenic amines are stocked after synthesis and prior to release. The membranes of these exocytic vesicles also contain proton-ATPase pumps that generate a
34 low pH in the lumen using ATP hydrolysis for energy. The VMATs, which are biogenic amine-proton antiporters, derive the energy needed for the stocking of biogenic amines into the exocytic vesicle from the luminal-cytoplasmic pH gradient established by the proton-ATPase pumps. Amphetamine-like compounds in the cytoplasm of biogenic amine-synthesizing cells inhibit VMATs, resulting in the accumulation of biogenic amines in the cytoplasm.
As the cytoplasmic concentration of biogenic amines increases, the plasma membrane symporter that normally transports them from the extracellular space to the cytoplasm, begins transporting them down their concentration gradient and out of the cell. This last step results in the non-exocytic, plasma membrane potential-independent release of biogenic amine stores from the cells that synthesize them. In the case of amphetamine, at typical doses, it causes the release of dopamine (DA) via its cognate
transporter (DAT); at higher doses, amphetamine also causes the release of
norepinephrine (NE) by way of its cognate transporter (NET); at very high doses,
amphetamine can elicit serotonin 5-hydroxytryptamine, 5-HT) release through its cognate
transporters (SERT) (see Fleckenstein, Gibb et al. 2000 for review). The biogenic amine
releasing activity of phentermine is similar to that of amphetamine; however, given its
low potential for abuse (a phenomenon that implicates DA), it probably releases DA to a
lesser extent than amphetamine. Fenfluramine, in contrast, given its higher affinity for
SERT than for NET and DAT, more selectively releases 5-HT (Setola, Hufeisen et al.
2003).
35 A second mechanism through which amphetamine-like compounds exert their
biological actions is by directly modulating the activity of cellular proteins sensitive to
biogenic amines. Amphetamine congeners are very similar in structure to endogenous
biogenic amines and, as such, it is likely that the molecular “targets” of endogenous
biogenic amines can be modulated by amphetamines. Indeed, some amphetamine-like
compounds inhibit biogenic amine-catabolizing enzymes (i.e., monoamine oxidases)
(Zychlinski and Montgomery 1984; Zychlinski and Montgomery 1985; Ulus, Maher et al.
2000), albeit with low affinity (Kilpatrick, Traut et al. 2001; Nandigama, Newton-Vinson
et al. 2002). Furthermore, there is some evidence that amphetamine-like compounds can
modulate ion channel activity as well. In this regard, fenfluramine has been shown to
block potassium channel currents in rat tissues in a manner that is only partially sensitive
to non-selective 5-HT receptor antagonists (Hu, Wang et al. 1998; Belohlavkova, Simak
et al. 2001). In addition, amphetamine-like compounds can also directly activate biogenic amine receptors. For example, fenfluramine exhibits affinity for, and agonist activity at, a
subset of 5-HT receptors, particularly members of the 5-HT2 receptor family (Rothman,
Baumann et al. 2000). Indeed, it is by activation of hypothalamic 5-HT2C receptors that
fenfluramine is thought to exert its anorectic actions (Vickers, Dourish et al. 2001;
Heisler, Cowley et al. 2002). A post-synaptic, direct 5-HT receptor agonist effect of fenfluramine on food intake is also supported by the finding that fenfluramine-induced anorexia is not altered by selective serotonin re-uptake inhibitors or depletion of brain 5-
HT (Gibson, Kennedy et al. 1993; Pedrinola, Sztejnsznajd et al. 1996; McCann, Yuan et al. 1997). Thus, in addition to eliciting biogenic amine release (i.e., acting as indirect biogenic amine receptor agonists), amphetamine-like compounds can also directly
36 modulate the activity of some proteins that are sensitive in biogenic amines-mediated
signaling (e.g., monoamine oxidases, post-synaptic receptors, and ion channels).
1.6 Evidence Linking the Serotonin (5-hydroxytryptamine, 5-HT) 5-HT2B Receptor
to VHD
Since 1997, researchers have sought to understand why fenfluramine and other
drugs cause the fibrotic heart valve and pulmonary artery lesions described above. Such
lesions are very frequent in patients with serotonin-secreting tumors (carcinoid
syndrome) and in migraine headache sufferers taking serotonergic ergot drugs (Bana,
MacNeal et al. 1974; Hauck, Edwards et al. 1990; Hendrikx, Van Dorpe et al. 1996;
Pritchett, Morrison et al. 2002; Serratrice, Disdier et al. 2002). Given the serotonin-
releasing properties of fenfluramine (Fuxe, Farnebo et al. 1975; Garattini, Buczko et al.
1975; Carboni and Di Chiara 1989; Schwartz, Hernandez et al. 1989; Series, Cowen et al.
1994), the involvement of serotonin, its receptors, and/or its transporter in fenfluramine- induced VHD seemed likely. Indeed, many suspected that fenfluramine-induced increases in plasma 5-HT levels were responsible for the cardiopulmonary effects of the drug (Fishman 1999). However, other drugs, such as lithium (a treatment for bipolar disorder) and monoamine oxidase inhibitors (antidepressant medications) increase plasma
5-HT levels about two-fold, and are not associated with VHD (Artigas, Sarrias et al.
1989; Celada, Sarrias et al. 1990; Celada, Perez et al. 1992). Furthermore, long-term treatment with fenfluramine has been shown to reduce plasma 5-HT levels (Raleigh,
Brammer et al. 1986; Celada, Martin et al. 1994; Redmon, Raatz et al. 1997; Rothman,
37 Redmon et al. 2000). These findings argued against the role of increased plasma 5-HT in fenfluramine-associated VHD. Thus, investigators began searching for other mechanisms underlying fenfluramine’s cardiopulmonary side effects.
In 2000, Dr. Roth’s group profiled the in vitro pharmacology of the known VHD- associated drugs fenfluramine, ergotamine, and methysergide, and of the in vivo metabolites norfenfluramine (fenfluramine metabolite) and methylergonovine
(methysergide metabolite), as well as that of several serotonergic drugs not associated with valvular heart disease (Rothman, Baumann et al. 2000). This study revealed that the
VHD-associated compounds norfenfluramine, ergotamine, and methylergonovine were all potent agonists at recombinant serotonin 5-HT2B receptors; in contrast, none of the
non-VHD-associated drugs displayed affinity for or efficacy at these receptors. Around the same time, a group at DuPont independently reported that norfenfluramine induced a mitogenic marker (i.e., phosphorylation of mitogen-associated protein kinase) in HEK-
293 cells via activation of recombinant 5-HT2B receptors, the high-level expression of which they measured in both human and porcine heart valves (Fitzgerald, Burn et al.
2000). The implication of these studies was that VHD-associated drugs and/or their metabolites activate 5-HT2B receptors on heart valve interstitial cells, leading to the
formation of proliferative foci and subsequent changes (e.g., increased extracellular matrix deposition and leukocyte infiltration) that compromised tissue function. This hypothesis seemed reasonable, given the well-established cross-talk between 5-HT2B receptors and mitotic pathway proteins, and the 5-HT2B receptor’s role in cell
proliferation in the developing heart (see above). However, direct proof that heart valve
38 interstitial cells exhibit mitogenic responses to VHD-associated drugs, and that those
responses are sensitive to pharmacological blockade of 5-HT2B receptors, was lacking.
In the chapters that follow, evidence will be presented establishing that activation
of 5-HT2B receptors by fenfluramine and norfenfluramine elicits mitogenic responses
from heart valve interstitial cells in primary culture. This response is reminiscent of the
fibrotic lesions observed in drug- and carcinoid-associated VHD. In addition, a similar
mitotic response is described for drugs chosen solely based on their agonist properties at
recombinant 5-HT2B receptors [i.e., 3,4-methylenedioxymethamphetamine (MDMA,
“Ecstasy”) and its in vivo metabolite 3,4-methylenedioxyamphetamine (MDA)] (Figure
1.3). These findings suggest that drugs with the potential to induce VHD can be identified by commonly-performed functional screens, which are typically automated and high- throughput. Evidence is also presented implicating stabilizing van der Waals’ interactions between a valine in transmembrane helix 2 and the α-methyl group of the active S-(+)- fenfluramine metabolite S-(+)-norfenfluramine in the valvulopathogen’s selectively high affinity for and efficacy at 5-HT2B receptors compared to 5-HT2C and 5-HT2A receptors.
These findings suggest that fenfluramine analogs bearing α-carbon substituents that
hinder van der Waals’ interactions with the TM2 valine may retain activity at anorectic 5-
HT2C receptors without triggering 5-HT2B-mediated cardiopulmonary side effects. The
implications of the findings presented herein for drug screening efforts, side effect
prediction, and molecular modeling-based drug design will also be discussed.
39 CHAPTER 2: MATERIALS AND METHODS
2.1 Materials
2.1.1 Chemicals
Serotonin (5-hydroxytryptamine, 5-HT) creatinine sulfate, lysergic acid
diethylamide (LSD), 5-methyl-1-(3-pyridylcarbamoyl)-1,2,3,5-tetrahydropyrrolo[2,3- f]indole (SB-206553), and α-methyl-5-(2-thienylmethoxy)-1H-indole-3-ethanamine (BW
723C86) were purchased from Sigma (St. Louis, MO). Optically pure and racemic preparations of fenfluramine, norfenfluramine (SNF and RNF), racemic MDMA, and
racemic MDA were furnished by Dr. Richard B. Rothman (NIDA, Baltimore, MD) and
Dr. Richard A. Glennon (Virginia Commonwealth University, Richmond, VA). [3H]Myo- inositol (21.0 Ci/mmol), [6-3H]thymidine (10 Ci/mmol), and [3H]lysergic acid
diethylamide (76.4 Ci/mmol) were obtained from Perkin Elmer Life Sciences (Boston,
MA). 3a70B liquid scintillation cocktail was from Research Products International (Elk
Grove Village, IL). Ecoscint A liquid scintillation cocktail was procured from National
Diagnostics (Atlanta, GA). All buffers, inorganic salts, ethidium bromide, and
polyethyleneimine were acquired from Sigma (St. Louis, MO).
40 2.1.2 Transfection Reagents, Cell Culture, and Transfection
Lipofectamine 2000 transfection reagent was purchased from Invitrogen
(Carlsbad, CA). Dulbecco’s Modified Essential Medium (DMEM), OPTI-MEM (reduced serum medium), inositol-free DMEM, F12 nutrient mixture, fetal bovine serum (FBS), dialyzed fetal bovine serum (5% dialyzed FBS), trypsin, sodium pyruvate, penicillin, and streptomycin were procured from Life Technologies/Gibco BRL (Gaithersburg, MD).
Cell culture dishes (12-well and 10-cm2) and flasks (75 cm2) were obtained from
Corning/Costar (Acton, MA)
2 HEK-293 and HEK-293T cells were maintained at 37°C, 5% CO2 in 75-cm flasks with 15 ml DMEM containing 100 mM sodium pyruvate, 100 U/ml penicillin, 100
µg/ml streptomycin, and 10% fetal bovine serum (DMEM, 10% FBS). At confluence, cells were trypsinized, harvested, and split 1:5 into four 10-cm2 dishes containing 10 ml
DMEM, 10% FBS (for transfection) and one 75-cm2 flask containing 15 ml DMEM, 10%
FBS (for cell maintenance). One day after seeding into 10-cm2 dishes (i.e. at ~90%
confluence), the medium was removed and replaced with 10 ml OPTI-MEM. Each 10-
cm2 dish was transfected with 24 µg of receptor plasmid using Lipofectamine 2000
exactly as specified by the manufacturer. Twenty-four hours after transfection, cells were
processed for either radioligand binding assays or inositol phosphate accumulation assays
as described below.
41 Primary cultures of human heart valve interstitial cells were prepared by and obtained from Dr. Ivan Vesely and Dr. K. Jane Grande-Allen at the Lerner Research
Institute of the Cleveland Clinic Foundation. Heart valves were obtained from donor
hearts deemed unsuitable for transplantation, or from hearts that were removed from
transplant recipients at the Cleveland Clinic Foundation. All Cleveland Clinic patients
who have tissue surgically removed have authorized its subsequent use for research
purposes (protocols approved by the CCF IRB 2378). To remove the cells from the
tissue, the specimens were placed into sterile containers, immersed in a solution of
collagenase II (2 mg/ml; Worthington Biochemicals, Freehold, NJ) in serum-free
Dulbecco’s Modified Eagle Medium, then agitated in an incubated shaker (140 rpm, 20
min, 37°C). After return to the sterile flow hood, all surfaces were rubbed with a sterile
cotton swab to remove the endothelial cells. The valve specimens were then finely minced and then digested with collagenase III (1 mg/ml; Worthington Biochemicals,
Freehold, NJ) in an incubated shaker (4 hours, 140 rpm, 37°C). Each resulting cell suspension was filtered through sterile, 70-µm filters to remove debris, and the filtered
cells were pelleted, then resuspended in DMEM/F12 medium (1:1, containing low
glucose with HEPES, pH 7.4) supplemented with 10% fetal bovine serum and 1%
antibiotic-antimycotic solution (Invitrogen, Carlsbad, CA) (VIC medium). The resulting
2 culture of heart valve interstitial cells (VIC) was maintained at 37°C, 5% CO2 in 75-cm flasks containing VIC medium. At confluence, the cells were split 1:5 into four poly-L- lysine (Sigma, St. Louis, MO)-coated 24-well plates (for [3H]thymidine deoxyribose
incorporation and activate mitogen-activated protein kinase assays) and one 75-cm2 flask
(for maintenance).
42 2.1.3 cDNA Constructs
2.1.3.1 Construction of pUniversal-Signal
Dr. Wesely Kroeze in Dr. Bryan Roth’s laboratory constructed a pIRES-neo
(Clontech, Palo Alto, CA)-based, plasma membrane protein expression vector
(pUniversal-Signal) using a strategy first applied to the β2-adrenergic receptor (Guan,
Kobilka et al. 1992). The principle involves appending to a cDNA sequence an N- terminal, cleavable signal sequence to that enhances plasma membrane targeting.
Towards this end, the multi-cloning site of the bicistronic expression vector pIRES-neo was digested by incubating several µg of the circular plasmid DNA with EcoR V (2 U) and EcoR I (2 U), BSA (100 µg/ml final), 3-5 U of calf intestinal phosphatase (to reduce re-circularization of the digested vector in ligation reactions), and an appropriate buffer as recommended by the manufacturer of the enzymes (New England Biolabs, Cambridge,
MA). Sense and anti-sense oligonucleotides containing the following elements (5’ to 3’) were ordered from Invitrogen (Carlsbad, CA): a 5’ EcoR V recognition sequence; a consensus Kozak sequence (CACCATG) to enhance translation (Kozak 1994); a cleavable, N-terminal, hemagglutinin plasma membrane targeting sequence (to enhance targeting of the translated receptor to the plasma membrane); a FLAG epitope tag (to facilitate immunodetection); an adenosine (to maintain the reading frame starting at the
ATG of the cleavable targeting sequence); a Not I site; and an EcoR I site (Figure 2.1).
The sense and anti-sense oligonucleotides were annealed in vitro, then digested with
EcoR V and EcoR I as described above for the vector save for the omission of calf
43 Figure 2.1: Schematic of the plasma membrane protein expression vector
pUniversal-Signal. The vector is based on Clontech’s bicistronic pIRES-neo expression
vector. A hemagglutinin cleavable signal sequence and a FLAG epitope tag were inserted
into the EcoR I-EcoR V site of pIRES-neo such that they are in frame with an open reading frame inserted into the Not I site. neoR, neomycin resistance gene; ampR,
ampicillin resistance gene; CMV promoter, cytomegalovirus promoter; IVS, intervening
sequence; IRES, internal ribosome entry site.
44
CMV promoter amp-R
EcoRV (915) Kozak-start-signal-FLAG
NotI (996)
EcoRI (1003) pUNIV-SIG IVS 5322 bp
IRES
neo-R
HEMAGGLUTININ SIGNAL SEQUENECE------ECORVKOZAK M K T I I A L S Y I F C GATATC-ACC-ATG-AAG-ACG-ATC-ATC-GCC-CTG-AGC-TAC-ATC-TTC-TGC-
------> FLAG EPITOPE TAG------> L V F A * D Y K D D D D A NOTI CTG-GTA-TTC-GCC-GAC-TAC-AAG-GAC-GAT-GAT-GAC-GCC-AGC-GGC- ECORI CGC-GAA-TTC
45 intestinal phosphatase. The digested, double-stranded oligonucleotide (insert) and the
digested, phosphatase-treated vector were then were purified from restriction
endonuclease digestion reactions by electrophoresis through a 0.7% agarose (AquaPor
LE, National Diagnostics, Atlanta, GA) gel containing 0.5 µg/ml ethidium, excision of
the portions of the gel containing the desired fragment under UV illumination, and
extraction of the DNA from the excised agarose using a Qiagen (Valencia, CA) Gel
Extraction Kit exactly as indicated by the manufacturer. The insert and vector were eluted from the purification columns using 50 µl of nuclease-free water (supplied with the Gel
Extraction Kit), then ligated together at a vector:insert ratio of 1:3 (~ 1 µl digested, phosphatased, gel purified, eluted vector:~ 3 µl digested, gel purified, eluted insert) in reactions using a Roche (Indianapolis, IN) Rapid Ligation Kit exactly as directed by the
manufacturer. After a 5-min incubation, 1-2 µl of the rapid ligation reaction were used to
transform OneShot TOP10 (Invitrogen, Carlsbad, CA) ultra-competent E. coli according to the manufacturer’s protocol. Transformants were selected by overnight growth on LB agar (Fisher Chemical, Fairlawn, NJ) 10-cm2 plates (Corning/Costar, Acton, MA)
containing 100 µg/ml ampicillin (Roche, Indianapolis, IN). Individual transformed clones were picked using a sterile pipette tip, inoculated into 3-ml sterile LB broth (Fisher
Chemical, Fairlawn, NJ) containing 100 µg/ml ampicillin in sterile, 15-ml Falcon tubes
(Corning, Acton, MA), and grown overnight at 37°C with shaking (250 rpm). Plasmids
were isolated from 2.5 ml of the overnight cultures using Promega (Madison, WI) SV
Mini Prep kit exactly as recommended by the manufacturer. Automated DNA sequencing
(Cleveland Genomics, Cleveland, OH) of the isolated plasmids was performed to identify those containing the entire desired insert. The remaining 500 µl of a 3-ml overnight
46 culture that gave rise to the desired plasmid was used to inoculate 1 L of LB broth containing 100 µg/ml ampicillin, which was then incubated overnight at 37°C with shaking (250 rpm) and processed the next day to isolate mg quantities of pUniversal-
Signal using a Promega (Madison, WI) Wizard Midi Prep Kit as instructed by the product insert. To prepare the vector for the cloning of the three 5-HT2 receptor cDNAs,
pUniversal-Signal (~100 ng in 1 µl) was digested in a reaction containing Not I (2 U),
BSA (10 µg/ml), calf intestinal phosphatase (3-5 U), and an appropriate buffer (all from
New England Biolabs, Cambridge, MA) according to the instructions of the enzymes’ manufacturer, then gel purified as described above.
2.1.3.2 Sub-clonining of human 5-HT2 receptors
The human 5-HT2A, 5-HT2B, and 5-HT2C (the non-edited, INI isofom) receptors
were amplified by polymerase chain reaction (PCR) from plasmids containing each cDNA (Rothman, Baumann et al. 2000). The forward and reverse PCR primers, made-to- order by Invitrogen (Carlsbad, CA), used to amplify each receptor cDNA contained a 5’
Not I recognition sequence (linker) followed by the 5’-most and 3’-most nucleotides of
the sequence to be amplified, respectively (Table 2.1). The amplicons were purified from
the PCR reactions using Qiagen’s (Valencia, CA) PCR Purification Kit exactly as
indicated by the manufacturer. PCR-purified amplicons eluted in 40 µl distilled water
were digested with Not I by the addition of 2 U restriction enzyme, 5 µl 10 X buffer, and
5 µl 1-mg/µl BSA (100 µg/µl final) (New England Biolabs, Cambridge, MA) followed
by incubation at 37°C for 1 hour. The digested amplicons were purified from restriction
47 Table 2.1: PCR primer sequences used to amplify and sub-clone human 5-HT2 receptor cDNA. Each primer contained a 5’ Not I recognition sequence (underlined) followed by 18 bp of sequence corresponding to the first (forward) or last (reverse) 6 codons of the receptor to be amplified.
48
NCBI Accession Receptor Forward PCR Primer Reverse PCR Primer Number 5’-GCGGCCGC-ATG- 5’-GCGGCCGC-TCA- 5-HT2A GAT-ATT-CTT-TGT- CAC-ACA-GCT-CAC- NM_000621 GAA-3’ CTT-3’ 5’-GCGGCCGC-ATG- 5’-GCGGCCGC-CTA- 5-HT2B GCT-CTC-TCT-TAC- TAC-ATA-ACT-AAC- NM_000867 AGA-3’ TCG-3’ 5’-GCGGCCGC-ATG- 5’-GCGGCCGC-TCA- 5-HT2C GTG-AAC-CTG-AGG- CAC-ACT-GCT-AAT- NM_000868 AAT-3’ CCT-3’
49 endonuclease digestion reactions by agarose gel electrophoresis and gel extraction as
described above. Amplicons were eluted in 50 µl distilled water and used in ligation
reactions (as described above) with Not I-digested, calf intestinal phosphatase-treated
pUniversal-Signal. Ligation reactions were used to transform OneShot TOP10
(Invitrogen, Carlsbad, CA) ultra-competent E. coli, which were then selected for clones containing the desired construct, amplified, and subjected to plasmid preparatory procedures as described above.
The digested product was isolated by gel electrophoresis and gel extraction as described above. The digested, purified vector and insert were added, at a ratio of ~1:3
(by volume), to a reaction containing T4 rapid ligation ligase and rapid ligation buffer according to the manufacturer’s recommendations (Roche, Indianapolis, IN). After a 5- min incubation, 1-2 µl of the rapid ligation reaction were used to transform One Shot
TOP10 (Invitrogen, Carlsbad, CA) ultra-competent E. coli exactly as described by the manufacturer. As above, isolated plasmids were subjected to automated sequencing
(Cleveland Genomics, Cleveland, OH) to verify the sequence of the insert and the absence of PCR-induced mutations.
2.1.3.3 Generation of Mutant 5-HT2 Receptors
Site-directed mutagenesis was carried out on wild type 5-HT2 receptor cDNA cloned into pUniversal-Signal using Stratagene’s (La Jolla, CA) QuikChange kit exactly
as indicated by the manufacturer. The primer sequences for the forward (sense)
50 Table 2.2: Sequence of sense primers used for site-directed mutagenesis of 5-HT2 receptors. The sequence of the sense primers comprised six codons upstream of the codon to be mutagenized, the desired mutant codon [the nuclotide(s) mutagenized are underlined], and 17-20 bp of 3’ sense sequence (chosen so as to end in at least one C or
G).
51
Receptor Sense Primer 5-HT2A L2.53V 5’-GCC-ATA-GCT-GAT-ATG-CTG-GTG- GGT-TTC-CTT-GTC-ATG-CCC-3’ 5-HT2B A1.35P 5’-GGA-AAT-AAA-CTG-CAC-TGG-CCA-GCT-CTT-CTG-ATA-CTC-ATG-G-3’ 5-HT2B A1.35S 5’-GGA-AAT-AAA-CTG-CAC-TGG-TCA-GCT-CTT-CTG-ATA-CTC-ATG-G-3’ 5-HT2B L1.38S 5’-CTG-CAC-TGG-GCA-GCT-CTT-TCG-ATA-CTC-ATG-GTG-ATA-ATA-CC-3’ 5-HT2B I1.39T 5’-CAC-TGG-GCA-GCT-CTT-CTG-ACA- CTC-ATG-GTG-ATA-ATA-CCC-3’ 5-HT2B V1.42I 5’- GCT-CTT-CTG-ATA-CTC-ATG-ATA-ATA-ATA-CCC-ACA-ATT-GG-3’ 5-HT2BV2.53L 5’-GCG-GTG-GCT-GAT-TTG-CTG-CTT-GGA-TTG-TTT-GTG-ATG-CC-3’ 5-HT2B V2.53I 5’-GCG-GTG-GCT-GAT-TTG-CTG-ATT-GGA-TTG-TTT-GTG-ATG-CC-3’ 5-HT2B L3.29V 5’-GTT-CTA-TGT-CCT-GCC-TGG-GTA-TTT-CTT-GAC-GTT-CTC-TTT-TC-3’ 5-HT2B M5.39V 5’-GAA-CGT-TTT-GGC-GAT-TTC-GTG-CTC-TTT-GGC-TCA-CTG-GC-3’ 5-HT2B E7.36N 5’-ACT-CTC-CAA-ATG-CTC-CTG-AAC-ATA-TTT-GTG-TGG-ATA-GGC-3’ 5-HT2B S7.45C 5’-GTG-TGG-ATA-GGC-TAT-GTT-TGC-TCA-GGA-GTG-AAT-CCT-TTG-G-3’ 5-HT2C V2.53L 5’-GCC-ATT-GCT-GAT-ATG-CTA-CTG-GGA-CTA-CTT-GTC-ATG-CCC-3’
52 mutagenesis primers, which were made-to-order by Invitrogen (Carlsbad, CA), are listed in Table 2.2. The reverse (anti-sense) mutagenesis primer sequences were the reverse complement of the sense mutagenesis primer sequences. The presence of the desired mutation and the absence of PCR-induced mutations was verified by automated sequencing of the entire mutagenized cDNA insert (Figure 2.2).
2.2 Methods
2.2.1 Radioligand Binding Assays
Twenty-four hours after transfection, the medium was removed and replaced with
DMEM containing 5% dialyzed FBS (Gibco BRL, Gaithersburg, MD). After 24 hr, the cells were incubated overnight in serum-free DMEM. The next day, the cells were harvested by scraping, pelleted, and resuspended in lysis buffer (50 mM Tris-HCl, pH
6.9). The membranes were then pelleted by centrifugation and, after removal of the supernatant, the membrane fraction was frozen at -80°C (if not used immediately).
Radioligand binding assays were set up in 24 wells of a 96-well plate (1 ml/well capacity) as follows: 25 µl of 10 µM [3H]LSD (Perkin Elmer Life Sciences, Boston,
MA), 25 µl of membrane pellet that had been resuspended in 700 µl ice-cold binding buffer (50 mM Tris-HCl, 5 mM MgCl2, 0.5 mM EDTA, pH 7.4), 25 µl vehicle (binding buffer) or 10X test compound (dissolved in binding buffer) at various concentrations spanning seven orders of magnitude (2 wells/concentration), and 200 ml binding buffer, such that the final [3H]LSD and test compound concentrations were 1X. Reactions were
53 Figure 2.2: Sequences of the 5-HT2 receptors sub-cloned into pUniversal-Signal-
FLAG. The orientation and full length sequence of the 5-HT2A (A), 5-HT2B (B), and 5-
HT2C (C) wild type and mutant receptor cDNA sub-cloned into pUniversal-Signal-FLAG was verified by automated sequencing and found to be in accord with the NCBI database sequences (Table 2.1). Amino acid numbers are indicated above the mutated codons
(wild type amino acid, blue; mutant amino acid, red), under which the codon sequence is shown. The nucleotide(s) subjected to site-directed mutagenesis are underlined.
54 A
M D I L C E E N T S L S S T T N S L M Q L N ATG-GAT-ATT-CTT-TGT-GAA-GAA-AAT-ACT-TCT-TTG-AGC-TCA-ACT-ACG-AAC-TCC-CTA-ATG-CAA-TTA-AAT-
D D T R L Y S N D F N S G E A N T S D A F N GAT-GAC-ACC-AGG-CTC-TAC-AGT-AAT-GAC-TTT-AAC-TCT-GGA-GAA-GCT-AAC-ACT-TCT-GAT-GCA-TTT-AAC-
W T V D S E N R T N L S C E G C L S P S C L TGG-ACA-GTC-GAC-TCT-GAA-AAT-CGA-ACC-AAC-CTT-TCC-TGT-GAA-GGG-TGC-CTC-TCA-CCG-TCG-TGT-CTC-
S L L H L Q E K N W S A L L T A V V I I L T TCC-TTA-CTT-CAT-CTC-CAG-GAA-AAA-AAC-TGG-TCT-GCT-TTA-CTG-ACA-GCC-GTA-GTG-ATT-ATT-CTA-ACT-
I A G N I L V I M A V S L E K K L Q N A T N ATT-GCT-GGA-AAC-ATA-CTC-GTC-ATC-ATG-GCA-GTG-TCC-CTA-GAG-AAA-AAG-CTG-CAG-AAT-GCC-ACC-AAC-
123 Y F L M S L A I A D M L L G F L V M P V S M TAT-TTC-CTG-ATG-TCA-CTT-GCC-ATA-GCT-GAT-ATG-CTG-CTG-GGT-TTC-CTT-GTC-ATG-CCC-GTG-TCC-ATG V GTG
L T I L Y G Y R W P L P S K L C A V W I Y L TTA-ACC-ATC-CTG-TAT-GGG-TAC-CGG-TGG-CCT-CTG-CCG-AGC-AAG-CTT-TGT-GCA-GTC-TGG-ATT-TAC-CTG-
D V L F S T A S I M H L C A I S L D R Y V A GAC-GTG-CTC-TTC-TCC-ACG-GCC-TCC-ATC-ATG-CAC-CTC-TGC-GCC-ATC-TCG-CTG-GAC-CGC-TAC-GTC-GCC-
I Q N P I H H S R F N S R T K A F L K I I A ATC-CAG-AAT-CCC-ATC-CAC-CAC-AGC-CGC-TTC-AAC-TCC-AGA-ACT-AAG-GCA-TTT-CTG-AAA-ATC-ATT-GCT-
V W T I S V G I S M P I P V F G L Q D D S K GTT-TGG-ACC-ATA-TCA-GTA-GGT-ATA-TCC-ATG-CCA-ATA-CCA-GTC-TTT-GGG-CTA-CAG-GAC-GAT-TCG-AAG-
V F K E G S C L L A D D N F V L I G S F V S GTC-TTT-AAG-GAG-GGG-AGT-TGC-TTA-CTC-GCC-GAT-GAT-AAC-TTT-GTC-CTG-ATC-GGC-TCT-TTT-GTG-TCA-
F F I P L T I M V I T Y F L T I K S L Q K E TTT-TTC-ATT-CCC-TTA-ACC-ATC-ATG-GTG-ATC-ACC-TAC-TTT-CTA-ACT-ATC-AAG-TCA-CTC-CAG-AAA-GAA-
A T L C V S D L G T R A K L A S F S F L P Q GCT-ACT-TTG-TGT-GTA-AGT-GAT-CTT-GGC-ACA-CGG-GCC-AAA-TTA-GCT-TCT-TTC-AGC-TTC-CTC-CCT-CAG-
S S L S S E K L F Q R S I H R E P G S Y T G AGT-TCT-TTG-TCT-TCA-GAA-AAG-CTC-TTC-CAG-CGG-TCG-ATC-CAT-AGG-GAG-CCA-GGG-TCC-TAC-ACA-GGC-
R R T M Q S I S N E Q K A C K V L G I V F F AGG-AGG-ACT-ATG-CAG-TCC-ATC-AGC-AAT-GAG-CAA-AAG-GCA-TGC-AAG-GTG-CTG-GGC-ATC-GTC-TTC-TTC-
L F V V M W C P F F I T N I M A V I C K E S CTG-TTT-GTG-GTG-ATG-TGG-TGC-CCT-TTC-TTC-ATC-ACA-AAC-ATC-ATG-GCC-GTC-ATC-TGC-AAA-GAG-TCC-
C N E D V I G A L L N V F V W I G Y L S S A TGC-AAT-GAG-GAT-GTC-ATT-GGG-GCC-CTG-CTC-AAT-GTG-TTT-GTT-TGG-ATC-GGT-TAT-CTC-TCT-TCA-GCA-
V N P L V Y T L F N K T Y R S A F S R Y I Q GTC-AAC-CCA-CTA-GTC-TAC-ACA-CTG-TTC-AAC-AAG-ACC-TAT-AGG-TCA-GCC-TTT-TCA-CGG-TAT-ATT-CAG-
C Q Y K E N K K P L Q L I L V N T I P A L A TGT-CAG-TAC-AAG-GAA-AAC-AAA-AAA-CCA-TTG-CAG-TTA-ATT-TTA-GTG-AAC-ACA-ATA-CCG-GCT-TTG-GCC-
Y K S S Q L Q M G Q K K N S K Q D A K T T D TAC-AAG-TCT-AGC-CAA-CTT-CAA-ATG-GGA-CAA-AAA-AAG-AAT-TCA-AAG-CAA-GAT-GCC-AAG-ACA-ACA-GAT-
D D C S M V A L G K Q H S E E A S K D N S D AAT-GAC-TGC-TCA-ATG-GTT-GCT-CTA-GGA-AAG-CAG-CAT-TCT-GAA-GAG-GCT-TCT-AAA-GAC-AAT-AGC-GAC-
G V N E K V S C V - GGA-GTG-AAT-GAA-AAG-GTG-AGC-TGT-GTG-TGA
55
B M A L S Y R V S E L Q S T I P E H I L Q S T ATG-GCT-CTC-TCT-TAC-AGA-GTG-TCT-GAA-CTT-CAA-AGC-ACA-ATT-CCT-GAG-CAC-ATT-TTG-CAG-AGC-ACC-
F V H V I S S N W S G L Q T E S I P E E M K TTT-GTT-CAC-GTT-ATC-TCT-TCT-AAC-TGG-TCT-GGA-TTA-CAG-ACA-GAA-TCA-ATA-CCA-GAG-GAA-ATG-AAA-
057 060 061 064 Q I V E E Q G N K L H W A A L L I L M V I I CAG-ATT-GTT-GAG-GAA-CAG-GGA-AAT-AAA-CTG-CAC-TGG-GCA-GCT-CTT-CTG-ATA-CTC-ATG-GTG-ATA-ATA- P S T I CCA TCG ACA ATA S TCA
P T I G G N T L V I L A V S L E K K L Q Y A CCC-ACA-ATT-GGT-GGA-AAT-ACC-CTT-GTT-ATT-CTG-GCT-GTT-TCA-CTG-GAG-AAG-AAG-CTG-CAG-TAT-GCT-
103 T N Y F L M S L A V A D L L V G L F V M P I ACT-AAT-TAC-TTT-CTA-ATG-TCC-TTG-GCG-GTG-GCT-GAT-TTG-CTG-GTT-GGA-TTG-TTT-GTG-ATG-CCA-ATT- L CTT I ATT 132 A L L T I M F E A M W P L P L V L C P A W L GCC-CTC-TTG-ACA-ATA-ATG-TTT-GAG-GCT-ATG-TGG-CCC-CTC-CCA-CTT-GTT-CTA-TGT-CCT-GCC-TGG-TTA- V GTA
F L D V L F S T A S I M H L C A I S V D R Y TTT-CTT-GAC-GTT-CTC-TTT-TCA-ACC-GCA-TCC-ATC-ATG-CAT-CTC-TGT-GCC-ATT-TCA-GTG-GAT-CGT-TAC-
I A I K K P I Q A N Q Y N S R A T A F I K I ATA-GCC-ATC-AAA-AAG-CCA-ATC-CAG-GCC-AAT-CAA-TAT-AAC-TCA-CGG-GCT-ACA-GCA-TTC-ATC-AAG-ATT-
T V V W L I S I G I A I P V P I K G I E T D ACA-GTG-GTG-TGG-TTA-ATT-TCA-ATA-GGC-ATT-GCC-ATT-CCA-GTC-CCT-ATT-AAA-GGG-ATA-GAG-ACT-GAT-
218 V D N P N N I T C V L T K E R F G D F M L F GTG-GAC-AAC-CCA-AAC-AAT-ATC-ACT-TGT-GTG-CTG-ACA-AAG-GAA-CGT-TTT-GGC-GAT-TTC-ATG-CTC-TTT- V GTG
G S L A A F F T P L A I M I V T Y F L T I H GGC-TCA-CTG-GCT-GCC-TTC-TTC-ACA-CCT-CTT-GCA-ATT-ATG-ATT-GTC-ACC-TAC-TTT-CTC-ACT-ATC-CAT-
A L Q K K A Y L V K N K P P Q R L T W L T V GCT-TTA-CAG-AAG-AAG-GCT-TAC-TTA-GTC-AAA-AAC-AAG-CCA-CCT-CAA-CGC-CTA-ACA-TGG-TTG-ACT-GTG-
S T V F Q R D E T P C S S P E K V A M L D G TCT-ACA-GTT-TTC-CAA-AGG-GAT-GAA-ACA-CCT-TGC-TCG-TCA-CCG-GAA-AAG-GTG-GCA-ATG-CTG-GAT-GGT-
S R K D K A L P N S G D E T L M R R T S T I TCT-CGA-AAG-GAC-AAG-GCT-CTG-CCC-AAC-TCA-GGT-GAT-GAA-ACA-CTT-ATG-CGA-AGA-ACA-TCC-ACA-ATT-
G K K S V Q T I S N E Q R A S K V L G I V F GGG-AAA-AAG-TCA-GTG-CAG-ACC-ATT-TCC-AAC-GAA-CAG-AGA-GCC-TCA-AAG-GTC-CTA-GGG-ATT-GTG-TTT-
F L F L L M W C P F F I T N I T L V L C D S TTC-CTC-TTT-TTG-CTT-ATG-TGG-TGT-CCC-TTC-TTT-ATT-ACA-AAT-ATA-ACT-TTA-GTT-TTA-TGT-GAT-TCC-
363 372 C N Q T T L Q M L L E I F V W I G Y V S S G TGT-AAC-CAA-ACT-ACT-CTC-CAA-ATG-CTC-CTG-GAG-ATA-TTT-GTG-TGG-ATA-GGC-TAT-GTT-TCC-TCA-GGA- N C AAC TGC
V N P L V Y T L F N K T F R D A F G R Y I T GTG-AAT-CCT-TTG-GTC-TAC-ACC-CTC-TTC-AAT-AAG-ACA-TTT-CGG-GAT-GCA-TTT-GGC-CGA-TAT-ATC-ACC-
C N Y R A T K S V K T L R K R S S K I Y F R TGC-AAT-TAC-CGG-GCC-ACA-AAG-TCA-GTA-AAA-ACT-CTC-AGA-AAA-CGC-TCC-AGT-AAG-ATC-TAC-TTC-CGG-
N P M A E N S K F F K K H G I R N G I N P A AAT-CCA-ATG-GCA-GAG-AAC-TCT-AAG-TTT-TTC-AAG-AAA-CAT-GGA-ATT-CGA-AAT-GGG-ATT-AAC-CCT-GCC-
M Y Q S P M R L R S S T I Q S S S I I L L D ATG-TAC-CAG-AGT-CCA-ATG-AGG-CTC-CGA-AGT-TCA-ACC-ATT-CAG-TCT-TCA-TCA-ATC-ATT-CTA-CTA-GAT-
T L L L T E N E G D K T E E R V S Y V - ACG-CTT-CTC-CTC-ACT-GAA-AAT-GAA-GGT-GAC-AAA-ACT-GAA-GAG-CGA-GTT-AGT-TAT-GTA-TAG
56 C
M V N L R N A V H S F L V H L I G L L V W Q ATG-GTG-AAC-CTG-AGG-AAT-GCG-GTG-CAT-TCA-TTC-CTT-GTG-CAC-CTA-ATT-GGC-CTA-TTG-GTT-TGG-CAA-
C D I S V S P V A A I V T D I F N T S D G G TGT-GAT-ATT-TCT-GTG-AGC-CCA-GTA-GCA-GCT-ATA-GTA-ACT-GAC-ATT-TTC-AAT-ACC-TCC-GAT-GGT-GGA-
R F K F P D G V Q N W P A L S I V I I I I M CGC-TTC-AAA-TTC-CCA-GAC-GGG-GTA-CAA-AAC-TGG-CCA-GCA-CTT-TCA-ATC-GTC-ATC-ATA-ATA-ATC-ATG-
T I G G N I L V I M A V S M E K K L H N A T ACA-ATA-GGT-GGC-AAC-ATC-CTT-GTG-ATC-ATG-GCA-GTA-AGC-ATG-GAA-AAG-AAA-CTG-CAC-AAT-GCC-ACC-
102 N Y F L M S L A I A D M L V G L L V M P L S AAT-TAC-TTC-TTA-ATG-TCC-CTA-GCC-ATT-GCT-GAT-ATG-CTA-GTG-GGA-CTA-CTT-GTC-ATG-CCC-CTG-TCT- L CTG
L L A I L Y D Y V W P L P R Y L C P V W I S CTC-CTG-GCA-ATC-CTT-TAT-GAT-TAT-GTC-TGG-CCA-CTA-CCT-AGA-TAT-TTG-TGC-CCC-GTC-TGG-ATT-TCT-
L D V L F S T A S I M H L C A I S L D R Y V TTA-GAT-GTT-TTA-TTT-TCA-ACA-GCG-TCC-ATC-ATG-CAC-CTC-TGC-GCT-ATA-TCG-CTG-GAT-CGG-TAT-GTA-
A I R N P I E H S R F N S R T K A I M K I A GCA-ATA-CGT-AAT-CCT-ATT-GAG-CAT-AGC-CGT-TTC-AAT-TCG-CGG-ACT-AAG-GCC-ATC-ATG-AAG-ATT-GCT-
I V W A I S I G V S V P I P V I G L R D E E ATT-GTT-TGG-GCA-ATT-TCT-ATA-GGT-GTA-TCA-GTT-CCT-ATC-CCT-GTG-ATT-GGA-CTG-AGG-GAC-GAA-GAA-
K V F V N N T T C V L N D P N F V L I G S F AAG-GTG-TTC-GTG-AAC-AAC-ACG-ACG-TGC-GTG-CTC-AAC-GAC-CCA-AAT-TTC-GTT-CTT-ATT-GGG-TCC-TTC-
V A F F I P L T I M V I T Y C L T I Y V L R GTA-GCT-TTC-TTC-ATA-CCG-CTG-ACG-ATT-ATG-GTG-ATT-ACG-TAT-TGC-CTG-ACC-ATC-TAC-GTT-CTG-CGC-
R Q A L M L L H G H T E E P P G L S L D F L CGA-CAA-GCT-TTG-ATG-TTA-CTG-CAC-GGC-CAC-ACC-GAG-GAA-CCG-CCT-GGA-CTA-AGT-CTG-GAT-TTC-CTG-
K C C K R N T A E E E N S A N P N Q D Q N A AAG-TGC-TGC-AAG-AGG-AAT-ACG-GCC-GAG-GAA-GAG-AAC-TCT-GCA-AAC-CCT-AAC-CAA-GAC-CAG-AAC-GCA-
R R R K K K E R R P R G T M Q A I N N E R K CGC-CGA-AGA-AAG-AAG-AAG-GAG-AGA-CGT-CCT-AGG-GGC-ACC-ATG-CAG-GCT-ATC-AAC-AAT-GAA-AGA-AAA-
A S K V L G I V F F V F L I M W C P F F I T GCT-TCG-AAA-GTC-CTT-GGG-ATT-GTT-TTC-TTT-GTG-TTT-CTG-ATC-ATG-TGG-TGC-CCA-TTT-TTC-ATT-ACC-
N I L S V L C E K S C N Q K L M E K L L N V AAT-ATT-CTG-TCT-GTT-CTT-TGT-GAG-AAG-TCC-TGT-AAC-CAA-AAG-CTC-ATG-GAA-AAG-CTT-CTG-AAT-GTG-
F V W I G Y V C S G I N P L V Y T L F N K I TTT-GTT-TGG-ATT-GGC-TAT-GTT-TGT-TCA-GGA-ATC-AAT-CCT-CTG-GTG-TAT-ACT-CTG-TTC-AAC-AAA-ATT-
Y R R A F S N Y L R C N Y K V E K K P P V R TAC-CGA-AGG-GCA-TTC-TCC-AAC-TAT-TTG-CGT-TGC-AAT-TAT-AAG-GTA-GAG-AAA-AAG-CCT-CCT-GTC-AGG-
Q I P R V A A T A L S G R E L N V N I Y R H CAG-ATT-CCA-AGA-GTT-GCC-GCC-ACT-GCT-TTG-TCT-GGG-AGG-GAG-CTT-AAT-GTT-AAC-ATT-TAT-CGG-CAT-
T N E P V I E K A S D N E P G I E M Q V E N ACC-AAT-GAA-CCG-GTG-ATC-GAG-AAA-GCC-AGT-GAC-AAT-GAG-CCC-GGT-ATA-GAG-ATG-CAA-GTT-GAG-AAT-
L E L P V N P S S V V S E R I S S V - TTA-GAG-TTA-CCA-GTA-AAT-CCC-TCC-AGT-GTG-GTT-AGC-GAA-AGG-ATT-AGC-AGT-GTG-TGA
57 equilibrated in the dark for 1 hr at room temperature (~22°C), and then receptor-ligand
complexes were harvested by vacuum filtration onto Whatman (Clifton, NJ) GF/C filters
(pre-soaked in 0.3% polyethyleneimine), then washed three times with ice-cold 50 mM
Tris-HCl, pH 6.9 using a Brandel 24-well harvester. The filters were dried overnight, and each was added to a 6-ml vial, into which 4 ml EcoScint liquid scintillation cocktail
(National Diagnostics, Atlanta, GA) was then added. The total [3H]LSD binding present
on each filter was quantified by liquid scintillation counting using a Wallac LSC
instrument. The log KD (for LSD) or log Ki (for all other compounds) and Bmax values
were determined using Prism 4.0 (GraphPad, San Diego, CA) by fitting the average total
[3H]LSD binding (in DPM) from several independent experiments, plotted as a function
of the log [compound], to a homologous (for LSD) or heterologous (for all other
compounds) competition model that takes into account ligand depletion and shares the
log KD or log Ki among all data sets.
2.2.2 Functional Assays
2.2.2.1 Inositol Phosphate Accumulation Assay
Twenty-four hours after transfection, cells in 10-cm plates were harvested by trypsinization and resuspended in 25 ml DMEM, 5% dialyzed FBS and seeded (1 ml cell
suspension/well) in a 24-well, poly-L-lysine-coated plate. The next day, the medium was
removed and replaced with BME (Gibco BRL, Gaithersburg, MD) containing 1 µCi/ml
[3H]myo-inositol (Perkin Elmer Life Sciences, Boston, MA) (500 µl/well). Twenty-four
58 hours later, the medium was replaced (1 ml/well) with 1X Hank’s Balanced Salt Solution
(Gibco BRL, Gaithersburg, MD) containing 25 mM sodium bicarbonate, 11 mM glucose,
and 10 mM LiCl (IP buffer). Compounds (5-HT, SNF, and RNF) were diluted in IP
buffer to 100X and added (10 µl/well) to the cells. After a 1 hr incubation at 37°C, the
buffer was removed and replaced with 1 ml 1 mM formic acid to extract the cytosolic
fraction. The [3H]inositol phosphates accumulated in the cytosol were isolated by loading the cytosolic fraction onto columns containing 1 ml AG 1-X8 formate form anion exchange resin (bead size: 75-150 µM; mesh: 100-200) (Bio-Rad, Hercules, CA), washing the columns twice with water (first with 2 ml/column, then with 10 ml/column),
then washing with 10 ml wash buffer (5 mM sodium borate, 50 mM sodium
formate)/column, and finally eluting with 10 ml elution buffer (100 mM formic acid, 200
mM sodium formate)/column into 30-ml vials containing 10 ml 3a70B liquid scintillation cocktail (Research Products International, Mount Prospect, IL). [3H]Inositol phosphate
accumulation in each sample was quantified by liquid scintillation counting using a
Wallac LSC instrument. The log EC50 and Emax values were determined using Prism 4.0
(GraphPad, San Diego, CA) by fitting the average total [3H]inositol phosphate from
several independent experiments, plotted as a function of the log [agonist], to a 3-
parameter logistic equation (sigmoidal dose-response) and sharing the log EC50 and Emax among all data sets.
59 2.2.2.2 [3H]Thymidine Deoxyribose Incorporation Assay
Sub-confluent VIC seeded in 24-well clusters were incubated overnight in serum- free DMEM (Invitrogen). Cells were then treated over the course of three days with
various concentrations of test agents. Twelve hours before the end of the treatment
period, cells were pulsed with 2-5 µCi/ml [3H]thymidine deoxyribose (PerkinElmer Life
Sciences, Boston, MA). After treatment, the medium was removed and the cells were
washed thoroughly with 500 µl ice-cold phosphate-buffered saline, pH 7.4. To extract
genomic DNA, 500 µl ice-cold 10% trichloroacetic acid was added to each well and the
cells were incubated for 30 min at 4°C. The cells were again washed thoroughly with 500
µl ice-cold phosphate-buffered saline, pH 7.4, and then lysed in 500 µl 0.5 N NaOH.
After neutralization with 14.4 ml glacial acetic acid, samples were assayed for
[3H]thymidine deoxyribose incorporation by adding the supernatant from each well to a
6-ml tube containing 4 ml EcoScint liquid scintillation cocktail (National Diagnostics,
Atlanta, GA) and liquid scintillation counting on a Wallac LSC instrument. Values are
reported as the mean ± S.E.M. of three independent experiments, each measured in
triplicate.
2.2.2.3 Activated Mitogen-Activated Protein Kinase Assays
Human heart valve interstitial cells seeded in 24-well plates (Corning/Costar,
Acton, MA) were incubated overnight first in DMEM containing 5% dialyzed fetal bovine serum and then in serum-free DMEM. Cells were treated over the course of 15
60 min with 10 µM fenfluramine, norfenfluramine, MDMA, MDA, or SB-206553. After
treatment, the medium in each well was replaced with 200 µl of 1x Laemmli sample
buffer and collected. Samples were resolved on 10% SDS-polyacrylamide gels and
electroblotted onto nitrocellulose membranes (Bio-Rad, Hercules, CA). The membranes
were probed for phospho-Erk 1/2MAPK immunoreactivity using a 1:1,000 dilution of
polyclonal primary rabbit anti-phospho-ERK antibody (Cell Signaling Inc., Beverly, MA) and a 1:1,000 dilution of horseradish peroxidase-conjugated goat anti-rabbit IgG
secondary antibody (Vector Laboratories, Burlingame, CA) according to the
manufacturer's recommendations. Immunoreactivity was revealed using LumiLight
horseradish peroxidase substrate (Roche, Indianapolis, IN) and imaged on a Kodak
Digital Science Image Station 440CF (Eastman Kodak, Rochester, NY). Densitometric
analysis was performed using Scion Image software (Scion Corporation, Frederick, MD).
Samples were similarly analyzed for total Erk 1/2MAPK immunoreactivity, and the
resulting values were used to correct phospho-Erk 1/2MAPK measurements for slight
differences in sample protein content. Values are reported as the mean ± S.E.M. of
duplicate determinations and are representative of three independent experiments.
2.2.2.4 Molecular Modeling, Ligand Docking Simulations, and Molecular Dynamics
Simulations
Transmembrane helix residues will be referred to using the convention described
by Ballesteros and Weinstein (Ballesteros and Weinstein 1994). The convention uses an
absolutely conserved residue in each helix as a reference, numbered X.50, where X is the
61 helix number. Residues N-terminal to X.50 are numbered X.49, X.48, etc.; residues C-
terminal to X.50 are numbered X.51, X.52, etc. So, for example, the absolutely conserved
arginine in transmembrane helix 3 of the 5-HT2B receptor (R153) is identified as R3.50 according to the Ballesteros and Weinstein convention. The N-terminal adjacent residue
(an aspartic acid) is identified as D3.49, and the C-terminal adjacent residue (a tyrosine) is identified as Y3.51.
Three-dimensional models of the human 5-HT2B, 5-HT2A, and 5-HT2C receptors
were constructed using a 3-D model of the rat 5-HT2A receptor (Shapiro, Kristiansen et
al. 2002) as a template. The raw sequence of each receptor was first fit onto the template
using the DeepView (GlaxoSmithKline, Research Triangle Park, NC) modeling
program’s Magic Fit function, then up-loaded to the Swiss Model server
(http://swissmodel.expasy.org//SWISS-MODEL.html). The returned PDB files were
loaded into Sybyl 6.91 molecular modeling and molecular dynamics software (Tripos,
Inc., St. Louis, MO) and, after atom typing and calculation of charges, energy minimized
using the Powell method under an AMBER FF99 force field with an 8-Å non-bonded
cut-off and a distance-dependent dielectric constant set to 4. The local minimum was identified using a Wolfe line search for successive iterations that differed by less than
0.05 kcal/mol·Å. To generate a reference structure for ligand docking simulations, 5-HT
was manually placed in the interhelical space of the energy-minimized 5-HT2B receptor according to 3 conserved features of 5-HT binding the receptor: 1) the protonated amine was 2 to 3 Å away from D3.32 (reviewed in Roth, Willins et al. 1998), and 2) the 5’-OH of the ligand was 2 to 3 Å away from S3.36 (reviewed in Roth, Willins et al. 1998;
62 Manivet, Schneider et al. 2002), and 3) the aromatic moiety was near the “aromatic box”
delineated by W3.28, W6.48, F6.51, and F6.52 (Roth, Shoham et al. 1997; Manivet,
Schneider et al. 2002). To open up the binding pocket, receptor side chains within 6 Å of
the manually-docked 5-HT were energy minimized as above. The interatomic distances
in the resulting structure were in good agreement with previous descriptions of biogenic amine ligand binding to 5-HT2-family receptors (Roth, Shoham et al. 1997; Kristiansen,
Kroeze et al. 2000; Shapiro, Kristiansen et al. 2000; Manivet, Schneider et al. 2002). The
ligand was then extracted from the minimized receptor, which was used as the receptor
description file (RDF) for the simulated docking of SNF and RNF by the FLEXX module
of Sybyl6.91 (Tripos, Inc., St. Louis, MO). The ligand binding site was defined as those
residues within 6 Å of D3.32. Only those solutions that were consistent with the
conserved features described above were retained for consideration.
To explore the possible effects of a mutation on SNF or RNF binding, the
mutation was computationally introduced into selected ligand docking simulation
solutions and 10 rounds of computer-simulated annealing (CSA) were performed to
explore possible side chain orientations relative to the bound ligand. Docked solutions
bearing the mutation to be studied were heated to and held at 700 K for 1 ps, then cooled
(0.5 K/fs) to 200 K during a 1-ps interval using an exponential temperature-vs-time
ramping method. During the simulations, the distance between the protonated amine
group of the ligand and the nearest carboxylate oxygen of D3.32, the ligand atoms, and
the α-carbon backbone of the receptor, and all atoms greater than 8 Å away from the
mutated side chain were constrained. For molecular dynamics calculations during CSA, a
63 AMBER FF99 force field was applied with an 8-Å non-bonded cut-off and a distance- dependent dielectric constant set to 4. The resulting structure from each round of CSA was examined vis-à-vis the orientation of the mutated residue’s side chain relative to the ligand.
64 CHAPTER 3: ESTABLISHING THAT FENFLURAMINE CAUSES MITOSIS IN
PRIMARY CULTURES OF VALVULAR INTERSTITIAL CELLS VIA
ACTIVATION OF 5-HT2B RECEPTORS
3.1 Introduction and Rationale
As described in the introductory chapter, in the highly effective appetite
suppressant fenfluramine (Pondimin®), a component of the drug combination "Fen-
Phen", and the optically pure S-(+) isomer dexfenfluramine (Redux®) were voluntarily
removed the from the marketplace at the urging of the US FDA because of their association with heart valve fibroplasia and dysfunction, a condition known as valvular
heart disease (VHD). Since then, several independent echocardiographic studies of
patients who received long-term fenfluramine therapy revealed an increased prevalence
of VHD (Connolly, Crary et al. 1997; Jick, Vasilakis et al. 1998; Weissman, Tighe et al.
1998; Weissman 2001). Histopathological examination of resected valves has revealed proliferative foci containing interstitial cells and increased levels of extracellular matrix
(Steffee, Singh et al. 1999). Identical pathology has been seen in valves resected from
persons undergoing long-term administration of certain ergot derivatives (e.g.,
ergotamine and methysergide) and from those suffering from carcinoid syndrome
(Steffee, Singh et al. 1999).
65 Recently, Dr. Roth and colleagues proposed that drugs (and/or their metabolites)
associated with VHD will preferentially bind with relatively high affinity to a single,
proximal receptor, channel, or transporter, whereas similar medications (e.g., fluoxetine,
phentermine) not associated with VHD would not (Rothman, Baumann et al. 2000). To
identify the receptor targeted by VHD-associated drugs, Rothman et al. screened fenfluramine, methysergide, and their in vivo metabolites (norfenfluramine and methysergide, respectively)—as well as medications not associated with VHD (i.e.,
fluoxetine, phentermine)—at a limited number of cloned receptors, transporters, and ion
channels. The screening effort revealed that VHD-associated compounds potently
activated 5-HT2B receptors and the non-VHD-associated drugs exhibited no agonist
activity at 5-HT2B receptors (Rothman, Baumann et al. 2000). The 5-HT2B receptor seemed a plausible candidate for mediating drug-induced VHD, given its relatively high expression in heart valves (Fitzgerald, Burn et al. 2000; Rothman, Baumann et al. 2000;
Roy, Brand et al. 2000). Also, given the proliferative nature of drug-induced VHD lesions and the well-established mitogenic consequences of 5-HT2B receptor activation in
cell lines as well as in heart tissue (Launay, Birraux et al. 1996; Nebigil, Launay et al.
2000), a link between 5-HT2B receptor activation and drug-induced VHD seemed highly
likely. In this regard, Fitzgerald and colleagues contemporaneously reported that the
fenfluramine metabolite norfenfluramine was an agonists at cloned 5-HT2B receptors—
the expression of which they observed in human heart valves—and independently
suggested that activation by norfenfluramine of mitogenic 5-HT2B receptors in heart
valves was responsible for fenfluramine-induced VHD (Fitzgerald, Burn et al. 2000).
However, direct evidence that fenfluramine causes heart valve cell proliferation via
66 activation of 5-HT2B receptors (i.e., the “5-HT2B receptor hypothesis of VHD”) has been
lacking.
In order to directly test the hypothesis that fenfluramine and/or norfenfluramine induce mitosis in human heart valve interstitial cells—the cells affected in VHD—we
measured vehicle- and drug-stimulated DNA synthesis in the presence and absence of a
5-HT2B receptor-selective antagonist (SB-206553) (Kennett, Wood et al. 1996). Based on
the “5-HT2B receptor hypothesis of VHD,” we anticipated that fenfluramine and
norfenfluramine, by virtue of their 5-HT2B receptor agonist properties in second- messenger accumulation assays, would both induce mitosis in human heart valve interstitial cells in the absence, but not in the presence, of SB-206553. Further, because norfenfluramine is a more efficacious 5-HT2B receptor agonist in vitro, we expected that
it would cause a more robust mitotic response in human heart valve interstitial cells that
would fenfluramine.
The “5-HT2B receptor hypothesis for VHD” also predicts that any drug that
activates 5-HT2B receptors will induce heart valve interstitial cell proliferation. Thus, we
sought to identify in a panel of amphetamine-like compounds previously-unidentified 5-
HT2B receptor agonists. Recently, Dr. Roth and colleagues pioneered the use of large-
scale screening of psychoactive drugs at a large panel of cloned receptors (i.e.,
"receptorome") to identify novel molecular targets and unidentified mechanisms of
67 action. In this way, the κ-opioid receptor was identified as the site of action of the novel hallucinogen salvinorin A (Roth, Baner et al. 2002; Sheffler and Roth 2003). A similar receptorome screen of the “club drug” 3,4-methylenedioxymethamphetamine (MDMA,
"Ecstasy") was performed to identify novel molecular targets responsible for the drug’s actions in humans. Interestingly, these studies revealed that MDMA, like the VHD-
associated drugs (e.g. fenfluramine, methysergide), exhibit relatively high affinity for 5-
HT2B receptors. Further study showed that MDMA and one of its in vivo metabolites, 3,4-
methylenedioxyamphetamine (MDA), were both agonists at 5-HT2B receptors. Thus, the
“5-HT2B receptor hypothesis of VHD” predicts that MDMA and MDA—two drugs not
previously associated with VHD—would elicit mitogenic responses from heart valve
interstitial cells via activation of 5-HT2B receptors.
In this chapter, we will present data showing that fenfluramine and its main metabolite norfenfluramine induce mitogenic responses in primary cultures of heart valve interstitial cells. Further, we will show that the mitogenic effect of fenfluramine and norfenfluramine on heart valve interstitial cells (VICs) in primary culture requires activation of endogenous 5-HT2B receptors. Because heart valve interstitial cells are the
cells affected in drug-induced VHD, and because mitotic activity underlies the
fibroproliferative lesions that give rise to VHD, the data we will present represent the
best evidence to date supporting the “5-HT2B receptor hypothesis of VHD.”
68 We will also present data demonstrating that MDMA and MDA induce 5-HT2B receptor-mediated mitogenic responses in primary cultures of VICs that are similar in magnitude to those induced by fenfluramine and norfenfluramine. Because MDMA and
MDA were selected for testing based only on their agonist activity at cloned 5-HT2B receptors, the data validate the use of cell lines expressing cloned 5-HT2B receptors to
identify compounds with the potential to induce VHD. Since cell lines expressing cloned
receptors are easier to obtain and work with than primary cultures, the data have
prompted large-scale screens of current and investigational pharmacotherapies aimed at
preventing VHD outbreaks similar to that associated with fenfluramine. As such, the findings have significant public health importance.
3.2 Results
3.2.1 Screening the Receptorome Reveals the 5-HT2B Receptor As a Molecular
Target for MDMA and MDA.
To discover novel molecular targets for the effects of psychoactive compounds
(see Roth, Baner et al. 2002 for a recent example), the “club drugs” MDMA and MDA
were screened at a large number of cloned (mostly human) neurotransmitter and hormone
receptors, ion channels, and transporters. Surprisingly, MDMA exhibited preferentially
high affinity for the 5-HT2B receptor (Figure 3.1A), a receptor previously implicated in
fenfluramine-induced VHD (Fitzgerald, Burn et al. 2000; Rothman, Baumann et al.
2000). As shown in Figure 3.1A and Table 3.1, other valvulopathic drugs are also
69 Figure. 3.1: Large-scale screening of the receptorome reveals that MDMA
preferentially interacts with the human 5-HT2B serotonin receptor. Top, Ki values for
various drugs screened at a large number of mainly human recombinant receptors, ion
channels, or transporters using the resources of the National Institute of Mental Health
Psychoactive Drug Screening Program. For these studies, test compounds were initially
screened at 10 µM. When greater than 50% inhibition of radioligand specific binding was
obtained, Ki values were determined in quadruplicate. A. Three-dimensional mesh plot of the data was made in which the Ki values were color-coded. The red arrow indicates that
MDMA has preferentially high affinity for h5-HT2B receptors. B. Representative
3 isotherms showing [ H]LSD displacement from h5-HT2B receptors expressed in HEK-
293 cells, the nonlinear regression of which was used to determine IC50 values. Ki values
were calculated using the Cheng-Prusoff approximation. C. Ki values for MDMA in bar
chart format; the arrow shows the Ki value for the h5-HT2B receptor; Ki values >10,000
nM are set to zero for clarity. Red arrow, Ki value for MDMA.
70
71
72
73 Table 3.1: MDMA, MDA, and other valvulopathic drugs bind to human
recombinant 5-HT2B receptors. To obtain pKi values, data from at least three
3 independent experiments in which [ H]LSD was displaced from 5-HT2B receptor-
containing HEK-293 cell membranes by various concentrations of test ligand were fit to a
heterologous competition model of radioligand binding (GraphPad Prism 4.0, San Diego,
CA). The model assumes one class of receptor sites and accounts for ligand depletion.
The pKi was fit directly and is reported with the associated S.E. For ease of reference, the
Ki is reported in parentheses. Non-specific radioligand binding never exceeded 20% of the total radioligand binding.
74
pK ± SE Drug i (Ki, nM) 5.9 ± 0.2 Fenfluramine (4000) 7.83 ± 0.06 Norfenfluramine (15) 7.82 ± 0.06 Dihydroergotamine (15) 7.8 ± 0.1 Pergolide (10) 6.30 ± 0.08 MDMA (500) 6.8 ± 0.1 MDA (100)
75
characterized by preferentially high affinities for 5-HT2B receptors (Figure 3.1A, Table
3.1). For instance, we discovered that two commonly prescribed medications exhibit high
affinity at 5-HT2B receptors: pergolide, a drug used in treating Parkinson's disease, and
dihydroergotamine, a drug used in treating migraine headaches (Figure 3.1B).
Importantly, both pergolide and dihydroergotamine have been shown to induce
fenfluramine-like VHD in humans (Hauck, Edwards et al. 1990; Hendrikx, Van Dorpe et al. 1996; Pritchett, Morrison et al. 2002).
We subsequently examined the abilities of MDMA and MDA to activate cloned,
human 5-HT2B receptors transiently expressed in HEK-293 cells. These studies identified
MDA as a more potent and efficacious agonist than MDMA (Figure 3.2A, Tables 3.1 and
3.2). In this regard, it was previously reported that the N-dealkylated metabolites of drugs
known to induce VHD (e.g., norfenfluramine and methylergonovine) were also more
potent and efficacious 5-HT2B receptor agonists than their respective parent compounds
(Table 3.2) (Rothman, Baumann et al. 2000). MDMA shares with VHD-associated drugs
the characteristic that N-dealkylation yields a more potent, more efficacious 5-HT2B receptor agonist. Thus, N-dealkylation in vivo of VHD-associated drugs gives rise to metabolites that are more potent 5-HT2B receptor agonists; according to the “5-HT2B receptor hypothesis of VHD,” the metabolites are, by virtue of their being more potent 5-
HT2B receptor agonists, more likely than the parent compounds to contribute to
valvulopathic activity. Importantly, the EC50 values for activating phosphoinositide
hydrolysis at 5-HT2B receptors for MDMA (2,000 nM) and MDA (190 nM) are nearly
identical to the plasma concentrations found in humans after a single recreational dose
76 Figure. 3.2: MDMA and MDA potently activate h5-HT2B-serotonin receptors in vitro.
Concentration-dependent stimulation of phosphatidylinositide hydrolysis via activation of
h5-HT2B receptors expressed in HEK-293 cells by fenfluramine, norfenfluramine, (R,S)-
MDMA, (R)-MDMA, (S)-MDMA, (R,S)-MDA, (R)-MDA, or (S)-MDA was assayed as
described in Chapter 2 (Material and Methods). Data represent mean ± S.E.M. for n = 3
separate experiments of percentage stimulation of [3H]IP accumulation relative to the full agonist 5-HT.
77
78 Table 3.2: MDMA and MDA, similar to other valvulopathic drugs, activate human
5-HT2B receptors in vitro. To obtain pEC50 and Emax values, data from at least three independent experiments in which the drug conentration-dependent accumulation of inositol phosphate in HEK-293 cells transiently expressing 5-HT2B receptors was
measured and fit to a sigmoidal concentration-response model (GraphPad Prism 4.0, San
Diego, CA). The pEC50 and Emax (in DPM) were fit directly. In each experiment, the
maximal response to 5-HT was used to normalize the Emax for each drug, such that Emax is
relative to the maximum response to 5-HT. The value ± S.E.M are reported.
*Significantly different (P < 0.05) from the other enantiomer by two-tailed t test.
Significantly different from racemate (P < 0.05) by two-tailed t test.
79 pEC ± SEM E Drug 50 max (EC50, nM) (% 5-HT) 9.0 ± 0.1 5-HT 1.00 ± 0.06 (1) 6.4 ± 0.2 Fenfluramine 0.13 ± 0.02 (400) 7.2 ± 0.1 Norfenfluramine 0.96 ± 0.03 (60) 7.52 ± 0.09 Dihydroergotamine 0.73 ± 0.02 (30) 7.27 ± 0.09 Pergolide 1.12 ± 0.04 (53) 5.8 ± 0.1 S-(+)-MDMA 0.32 ± 0.02 (2000) 6.0 ± 0.2* R-(-)-MDMA 0.27 ± 0.02 (1000) 5.2 ± 0.2* S-(+)-MDMA 0.38 ± 0.03 (6000) 6.73 ± 0.05 R,S-MDA 0.80 ± 0.02 (190) 6.83 ± 0.05 R-(-)-MDA 0.76 ± 0.02 (150) 6.9 ± 0.1 S-(-)-MDA 0.81 ± 0.04 (1000) 6.8 ± 0.1 Methysergide 0.18 ± 0.02 (150) 9.2 ± 0.1 Methylergonovine 0.40 ± 0.02 (0.8)
80 (150 mg) of MDMA in humans. For instance, after a single 150-mg dose of MDMA, de
la Torre et al. (de la Torre, Farre et al. 2000) reported a peak plasma concentration (Cmax)
for MDMA of 2,000 nM and a Cmax for MDA of 150 nM. Table 3.2 also demonstrates
that fenfluramine, norfenfluramine, pergolide, and dihydroergotamine, all of which have
been demonstrated to induce VHD of the in humans (Hauck, Edwards et al. 1990;
Hendrikx, Van Dorpe et al. 1996; Pritchett, Morrison et al. 2002), are also potent 5-HT2B
agonists.
The more potently anorexigenic S-(+) stereoisomer of fenfluramine,
dexfenfluramine (Redux®), has also been linked to VHD and PH. As such, we evaluated
optically pure preparations of MDMA and MDA for potency and efficacy at human 5-
HT2B receptors. We found no significant difference in efficacy between the R-(-) and S-
(+) stereoisomers of either MDMA or MDA; with respect to potency, the S-(+)
stereoisomer of MDMA was slightly more potent than the R-(-) stereoisomer, whereas the
R-(-) and S-(+) stereoisomers of MDA exhibited no statistically significant difference in
potency (Figure 3.2B and Table 3.2). Thus, both stereoisomers of MDMA and MDA are equi-potent, equi-efficacious 5-HT2B receptor agonists and, according to the “5-HT2B receptor hypothesis of VHD,” equally likely to induce mitogenic responses from VIC in primary culture.
81 3.2.2 Valvulopathic Drugs Induce Mitogenic Responses in Human Heart Valve
Interstitial Cells.
Because much of the evidence implicating 5-HT2B receptor activation in drug-
induced VHD is inferential, we set out to directly test the mitogenic activity of valvulopathic drugs using primary cultures of human heart valve interstitial cells (VICs).
To do so, we evaluated the abilities of selected VHD-associated drugs to affect markers
associated with mitosis (i.e., DNA replication and Erk 1/2 phosphorylation). For these
studies, we incubated serum-starved VICs for 48 hours with fenfluramine,
norfenfluramine, MDMA, MDA, SB-206553 (a 5-HT2B/2C receptor-selective antagonist),
or 5-HT and measured [3H]thymidine incorporation into genomic DNA. The VHD-
associated drugs fenfluramine and norfenfluramine, as well as MDMA, MDA, and 5-HT
each induced statistically significant stimulation of [3H]thymidine deoxyribose incorporation in VICs compared to vehicle (Figure 3.3A). The 5-HT2B/2C receptor-
selective antagonist SB-206553, however, actually reduced [3H]thymidine deoxyribose
incorporation into genomic DNA (Figure 3.3A). The DNA replication elicited by each drug was abrogated by prior treatment with the 5-HT2B/2C receptor-selective antagonist
SB-206553, demonstrating that the response was caused by 5-HT2B/2C receptor activation
(Figure 3.3B). Since heart valve cells express little or no 5-HT2C receptors (Fitzgerald,
Burn et al. 2000; Roy, Brand et al. 2000), it is likely, then, that the stimulatory effects of the drugs on DNA replication are due mainly to activation of 5-HT2B receptors.
82 Figure. 3.3: MDMA and MDA induce mitogenesis in human heart valve interstitial cells
in vitro. A. Stimulation of [3H]thymidine deoxyribose incorporation in VICs treated for
48 h with either vehicle (V), 5-HT, the 5-HT2B/2C receptor-selective antagonist SB-
206553 (SB), fenfluramine (F), norfenfluramine (NF), MDMA (X), or MDA (M) reveals a prolonged mitogenic response that is blocked by pretreatment with SB (B). All drugs
used at 10 µM except SB-206553, which was used at 1 µM. *, P < 0.05; **, P < 0.01;
***, P < 0.001, significant difference from vehicle-treated cells by two-tailed t test. C.
Immunoblot analysis of MAPK phosphorylation in VICs treated for 10 min with
norfenfluramine, MDMA, MDA, SB-206553, or 5-HT reveals a short-term mitogenic
response (i.e., increase in percent of total cellular MAPK phosphorylated; see Chapter 2:
Materials and Methods for details) after exposure to 5-HT2B receptor agonists. *, P <
0.05; **, P < 0.01, significant difference from vehicle-treated cells by two-tailed t test.
83
84
85
86 3.2.3 Valvulopathic Drugs Induce a Mitogenic Marker—Phosphorylation of
Mitogen-Activated Protein Kinase—in Human Heart Valve Interstitial Cells.
Immunoblot analysis of vehicle- and drug-treated VIC lysates revealed that short- term (10-min) treatment of serum-starved cells with either norfenfluamine, MDMA,
MDA, or 5-HT induced an increase (statistically significant for all drugs but MDMA) in mitogen-activated protein kinase (MAPK) phosphorylation, an early marker of mitosis, compared to vehicle-treated cells (Figure 3.4). The 5-HT2B/2C receptor antagonist SB-
206553, which caused a statistically significant decrease in [3H]thymidine deoxyribose incorporation, had no effect on basal MAPK phosphorylation compared to vehicle-treated
cells, suggesting that 5-HT2B receptors regulate basal VIC mitogenesis via a MAPK- independent mechanism. In terms of drug-stimulated MAPK phosphorylation, however, the responses to norfenfluramine, MDMA, MDA, and 5-HT were consistent with the results of the [3H]thymidine deoxyribose incorporation assays that reflected mitogenic activity. That fenfluramine did not significantly stimulate MAPK phosphorylation may be due to its relatively low efficacy and/or the low signal-to-noise ratio for immunoblot analysis of MAPK phosphorylation in VICs.
3.3 Discussion
A major finding of the present study is that fenfluramine and its in vivo metabolite norfenfluramine induce mitogenesis in primary cultures of heart valve interstitial cells via
87 activation of 5-HT2B receptors. The demonstration that fenfluramine and norfenfluramine
exert a direct mitogenic effect on heart valve interstitial cells that is blocked by the 5-
HT2B receptor-selective antagonist SB-206553 constitutes the most compelling evidence
to date in favor of the “5-HT2B receptor hypothesis of VHD.” Because VICs are the cells
affected in drug-induced VHD, they represent the most relevant system for in vitro assays
of a drug’s potential to induce VHD. However, given the difficulty in obtaining human
hearts and establishing primary cultures, which senesce after about eight passages, their
utility for high-throughput compound screens is limited. Thus, to identify potential
valvulopathogens from large compound libraries, a more readily available system (e.g., a
cell line expressing human 5-HT2B receptors that generate easily measurable responses
upon activation) is needed.
Another major finding described herein is that two commonly abused drugs,
MDMA and MDA, selected solely based on their agonist properties at cloned 5-HT2B receptors, induce mitosis in human heart valve interstitial cells in vitro via activation of 5-
HT2B receptors. Further, we demonstrate that pergolide, which was recently shown to
induce VHD (Pritchett, Morrison et al. 2002), and dihydroergotamine, which has long
been known to induce VHD (Hauck, Edwards et al. 1990; Hendrikx, Van Dorpe et al.
1996), both bind to and activate cloned 5-HT2B receptors. Our data suggest, then, that
cells expressing 5-HT2B receptors represent a reliable and readily available system for
identifying drugs that will induce heart valve interstitial cell mitosis in vitro and, as such, will likely induce VHD in vivo. In future studies, we plan to screen commercially available chemical libraries using HEK-293 cells expressing cloned, human 5-HT2B
88 receptors for previously unidentified 5-HT2B receptor agonists. We expect these drugs to
have mitogenic effects on VICs and, as such, be associated with VHD risk in humans.
In addition to expressing 5-HT2B receptors, VICs also express 5-HT2A receptors
(Fitzgerald, Burn et al. 2000). Thus, it is conceivable that the mitogenic responses we measured in VICs treated with fenfluramine, norfenfluramine, MDMA, MDA, and 5-HT are due to activation of 5-HT2A receptors, since these drugs exhibit some agonist activity
at 5-HT2A receptors (Nash, Roth et al. 1994; Rothman, Baumann et al. 2000). Indeed, activation of 5-HT2A receptors has been implicated in the mitogenic effect of serotonergic
agonists on sheep aortic valve interstitial cells (Xu, Jian et al. 2002). However, there are
two reasons to suspect that 5-HT2A receptor agonism is not a major mechanism for drug-
induced VIC mitogenesis: 1) the in vitro pharmacology of the inhibitors used by Xu et al. in their studies of sheep aortic valve interstitial cells has not been well-characterized, precluding the conclusion that 5-HT2B receptor activity contributes little to drug-
stimulated mitosis, and 2) genetic ablation of 5-HT2A receptor expression has no effect on
-/- cardiac development (J. Gingrich, personal communication), while 5-HT2B mice have
significantly fewer cells in the heart (Nebigil, Choi et al. 2000; Nebigil, Hickel et al.
2001), suggesting that 5-HT2A receptors are less important than 5-HT2B receptors for
cardiac cell proliferation. In addition, activity at 5-HT2C receptors likely contributed little
to the drug-stimulated mitosis we measured in primary cultures of human heart valve
interstitial cells because the cells express barely detectable levels of the receptor
(Fitzgerald, Burn et al. 2000). Taken together, these results imply that activation of
mitogenic pathways by 5-HT2A receptors is inessential for cardiac cell proliferation and
89
that 5-HT2B receptor activation is most likely responsible for the mitogenic responses
induced by valvulopathic drugs. Another finding consistent with activation of 5-HT2B receptors as the mechanism behind fenfluramine-induced VHD is the observation that 5-
HT2B receptor activation leads to mitogenic responses in several cellular contexts
(Fitzgerald, Burn et al. 2000; Nebigil, Choi et al. 2000; Nebigil, Launay et al. 2000),
clearly linking the 5-HT2B receptor to the regulation of mitosis.
Of equal importance, recent data have implicated the 5-HT2B receptor in the
pathogenesis of pulmonary hypertension, a severe and frequently fatal illness (Launay,
Herve et al. 2002). Importantly, in this regard, fenfluramine use increases the risk of
developing pulmonary hypertension (Abenhaim, Moride et al. 1996). Thus, these data further highlight the necessity of screening current and potential pharmacotherapies for agonist potencies and efficacies at human 5-HT2B receptors.
90 CHAPTER 4: MOLECULAR DETERMINANTS FOR THE INTERACTION OF
THE ANOREXIGEN NORFENFLURAMINE WITH THE 5-HT2B RECEPTOR
4.1 Introduction and Rationale
As detailed in the introductory chapter, the neurohumoral modulator serotonin (5-
hydroxytryptamine, 5-HT) is a biogenic amine that regulates a broad spectrum of
processes in both the central nervous system and in the periphery via no fewer than 15
plasma membrane receptors divided into seven families (Kroeze, Kristiansen et al. 2002;
Kroeze, Sheffler et al. 2003). All but one family belong to the Class A, rhodopsin-like, G
protein-coupled receptor superfamily. The 5-HT2 family receptors, comprising the 5-
HT2A, 5-HT2B, and 5-HT2C subtypes, represents one of the best-characterized groups of 5-
HT receptors (Roth, Willins et al. 1998; Kroeze, Kristiansen et al. 2002). These receptors are the major targets/sites of action of atypical antipsychotic medications (Meltzer 1989;
Roth, Sheffler et al. 2004), most hallucinogens (Glennon, Titeler et al. 1984; Roth, Baner
et al. 2002; Nichols 2004), and some appetite suppressants (Moses and Wurtman 1984;
Neill and Cooper 1989; Vickers, Dourish et al. 2001).
As mentioned earlier, the appetite suppressant fenfluramine, which is very effective in the treatment of obesity (Weintraub, Sundaresan et al. 1992), was withdrawn
from the US marketplace due to its association with valvular heart disease (VHD)
(Connolly, Crary et al. 1997). Fenfluramine was also known to be associated with
pulmonary hypertension (PH) (Pouwels, Smeets et al. 1990; Brenot, Herve et al. 1993;
91 Connolly, Crary et al. 1997; Simonneau, Fartoukh et al. 1998). Fenfluramine-induced
VHD and PH likely result from the activation, by the fenfluramine metabolite
norfenfluramine, of mitogenic 5-HT2B receptors on heart valve and pulmonary artery
interstitial cells (Fitzgerald, Burn et al. 2000; Rothman, Baumann et al. 2000; Launay,
Herve et al. 2002; Setola, Hufeisen et al. 2003), leading to the formation of proliferative,
fibromyxoid plaques that compromise tissue integrity and function (Steffee, Singh et al.
1999; Tomita and Zhao 2002). Fenfluramine-induced anorexia, in contrast, appears to be
mediated through activation, by the fenfluramine metabolite norfenfluramine, of
hypothalamic 5-HT2C receptors (Vickers, Dourish et al. 2001; Heisler, Cowley et al.
2002). Thus, novel anorexigens that activate 5-HT2C receptors and are devoid of 5-HT2B receptor activity are expected to be safe and effective treatments for obesity; the elucidation of 5-HT2B receptor-unique SNF-receptor intermolecular interactions could
facilitate the rational design of such 5-HT2C receptor-selective agents.
In this chapter, we present evidence that residue 2.53 (see Chapter 2, section 2.2.4
for an explanation of residue nomenclature) in 5-HT2 receptors plays an important role in the sub-type selective in vitro pharmacology of S-(+)-norfenfluramine (SNF). Site- directed mutagenesis, SNF and NF congener binding studies, molecular modeling, and ligand docking and MD simulations all suggest that both terminal methyl groups of
V2.53 in the 5-HT2B receptor form stabilizing van der Waals’ (vdW) interactions with the
α-methyl group of SNF. In addition, we provide functional data demonstrating that these
interactions are also important for SNF-mediated activation of the 5-HT2B receptor. We
92 also present evidence that the role of residue 2.53 on SNF binding to, and activation of, the 5-HT2B receptor is unique among 5-HT2 receptors.
4.2 Results
4.2.1 Effect of Point Mutations on Ligand Affinity
In order to identify residues involved in the sub-type selective binding of the
anorexigen S-(+)-norfenfluramine (SNF) to 5-HT2 family receptors, we constructed every
possible 5-HT2B-to-5-HT2C/2A receptor mutant residing near a putative ligand binding
pocket. To do so, we first generated a molecular model of the human 5-HT2B receptor by
homology model building using our bovine rhodopsin-based (Palczewski, Kumasaka et
al. 2000), refined 3-D homology model of the rat 5-HT2A receptor (Shapiro, Kristiansen
et al. 2002) as a template. To verify that our rhodopsin-based model was similar in
predicted structure to the published co-ordinates of bovine rhodopsin (Palczewski,
Kumasaka et al. 2000), we superimposed our 5-HT2B receptor homology model onto
1HZXA (the bovine rhodopsin PDB file). All absolutely conserved transmembrane helix
residues aligned and were oriented very similarly relative to each other (data not shown).
We next identified putative ligand binding residues in the 5-HT2B receptor (i.e., those
residues in the extracellular half of the transmembrane helices having their side chains
oriented into the interhelical space) that were not conserved in linear alignments of the 5-
HT2 receptor transmembrane helices (Figure 4.1). We then measured the affinity constant
(Ki) of SNF at wild type and point mutant 5-HT2B receptors in which one non-conserved,
93 Figure 4.1: 3-D molecular model of the human 5-HT2B receptor transmembrane
helices showing putative ligand binding residues that are non-conserved among 5-
HT2 family receptors. Side chains shown by mutagenesis and radioligand binding assays
to be involved in SNF binding are labeled in green; those shown not to affect SNF binding are labeled in white. Atom color code: carbon (gray); nitrogen (blue); oxygen
(red); hydrogen (cyan); sulfur (yellow). The α-carbons backbone of the seven helices is shown in purple. (See Materials and Methods for details regarding generation of the model.)
94
95 putative ligand binding residue was mutated to its 5-HT2C and/or 5-HT2A receptor analog.
As shown in Table 4.1 and Figure 4.2A, only the V2.53L and the M5.39V mutations
caused a decrease in SNF affinity: the V2.53L (5-HT2B/2C-to-5-HT2A) mutation decreased
SNF affinity 17-fold, while the M5.39V (5-HT2B-to-5-HT2C/2A) mutation led to a ten-fold
decrease in ligand affinity. The two mutations in tandem decreased SNF affinity 37-fold
(Table 4.1). Because V2.53 was predicted by our model to reside in the putative binding
pocket, we chose to investigate further its potential, direct role in the 5-HT2B receptor- selective binding of SNF.
In the 5-HT2C receptor, residue 2.53 is a valine. However, the V2.53L mutation in
the 5-HT2C receptor only caused a 9-fold decrease in SNF affinity (i.e., half as large a
decrease as that caused by the mutation in the 5-HT2B receptor) (Figure 4.2B and Table
4.1). The inverse L2.53V mutation in the 5-HT2A receptor had no effect on SNF affinity
(Figure 4.2C and Table 4.1). Thus, the role of residue 2.53 in SNF binding to 5-HT2 receptors is sub-type selective, being most important in the 5-HT2B receptor, of moderate importance in the 5-HT2C receptor, and not at all important in the 5-HT2A receptor. This
mirrors the rank order potency of SNF at the 5-HT2 family receptors (i.e., Ki = 22 nM,
170 nM, and 1,900 nM, respectively).
To ensure that the apparent decrease in SNF affinity due to the V2.53L mutation
3 did not result from altered affinity for the radioligand ([ H]LSD), we measured the Ki of
LSD at wild type and V2.53L 5-HT2B receptors. As shown in Table 4.2 and Figure
4.3A,B, the V2.53L mutation had little effect on LSD affinity. Indeed, none of the
96 Table 4.1: Affinity constants (Ki’s) for SNF binding to wild type and mutant 5-HT2 receptors. p values were obtained from F-tests comparing curve fits of competition
a binding isotherms for two receptors: p<0.05 compared the appropriate wild type 5-HT2
b c receptor; p<0.05 compared to the V2.53L 5-HT2B receptor; p<0.05 compared to the
M5.39V 5-HT2B receptor
97
pK S-NF ± SE Receptor i (Ki S-NF, nM) 7.66 ± 0.02 5-HT WT 2B (22) 7.81 ± 0.07a A1.35P (15) 7.77 ± 0.07 A1.35S (17) 7.76 ± 0.06 L1.38S (17) 7.84 ± 0.07a I1.39T (14) 7.82 ± 0.09 V1.42I (15) 6.43 ± 0.04a V2.53L (370) 7.45 ± 0.04a V2.53I (35) 5.48 ± 0.08a V2.53A (3,300) 7.65 ± 0.07 L3.29I (22) 6.65 ± 0.06a M5.39V (220) 6.09 ± 0.08a,b,c V2.53L, M5.39V (810) 7.80 ± 0.06 E7.36N (16) 7.65 ± 0.05 S7.45C (22) 5.73 ± 0.04 5-HT WT 2A (1900) 5.74 ± 0.04 5 HT L2.53V 2A (1800) 6.77 ± 0.05 5-HT WT 2C (170) 5.8 ± 0.1a 5-HT V2.53L 2C (2000)
98 Figure 4.2: Representative competition binding isotherms for SNF at wild type and
3 mutant 5-HT2 receptors. The percent total binding of ~1 nM [ H]LSD remaining in the
presence of the indicated concentration of SNF is shown for WT, V2.53L, and M5.39V
5-HT2B receptors (A); WT and V2.53L 5-HT2C receptors (B); WT and L2.53V 5-HT2A receptors (C); WT and V2.53I 5-HT2B receptors (D). Data are presented as the mean ± SE
of at least three independent experiments measured in duplicate. To obtain log Ki values from these data, the data were fit to a heterologous competition model of radioligand binding to one class of receptor sites that takes ligand depletion into account (GraphPad
Prism 4.0). Non-specific radioligand binding never exceeded 20% of the total radioligand binding.
99
120 WT2B-SNF 100 V2.53L2B-SNF M5.39V2B-SNF 80 60
40 (% remaining) (% remaining) H]LSD sp. sp. binding binding
3 20 [ A 0 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 log [SNF]
120 WT2C-SNF 100 V2.53L2C-SNF
80
60
40
bind. sp. H]LSD (% remaining) 3 [ 20 B 0 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 log [SNF]
120 WT2A-SNF 100 L2.53V2A-SNF 80
60
40 H]LSD sp. bind. (% remaining) 3 [ 20 C 0 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 log [SNF]
100 Table 4.2: Affinity constants (Ki’s) for other 5-HT2B receptor agonist ligands at wild
* type and V2.53 mutant 5-HT2B receptors. p<0.05 from F-tests comparing curve fits of competition binding isotherms for wild type and mutant 5-HT2B receptors. Average Bmax values in DPM, fit directly and adjusted for radioligand specific activity and protein content, were as follows: WT, 3.7 ± 0.7 pmol/mg; V2.53L, 9 ± 2 pmol/mg; V2.53A, 7 ±
1 pmol/mg; V2.53I, 2.4 ± 0.2 pmol/mg.
101
WT V2.53L V2.53A V2.53I Drug pKi pKi pKi pKi (Ki, nM) (Ki, nM) (Ki, nM) (Ki, nM) 7.3 ± 0.1 7.17 ± 0.08 6.5 ± 0.2* 6.2 ± 0.4* 5-HT (50) (68) (300) (600) 7.8 ± 0.1 7.3 ± 0.1* 6.4 ± 0.2* 6.7 ± 0.1* α-methyl-5-HT (20) (50) (400) (200) 7.63 ± 0.08 6.9 ± 0.2* 7.3 ± 0.2 7.2 ± 0.1* DHE (23) (120) (50) (60) 7.68 ± 0.09 7.5 ± 0.1 6.6 ± 0.1* 7.3 ± 0.1* Pergolide (21) (30) (200) (50) 8.59 ± 0.05 8.34 ± 0.04* 8.41 ± 0.05 8.61 ± 0.05 LSD (2.6) (4.6) (3.9) (2.4)
102 Figure 4.3: Competition binding isotherms for several 5-HT2B receptor agonist
ligands at wild type and V2.53L 5-HT2B receptors. The percent total binding of ~1 nM
[3H]LSD remaining in the presence of 5-HT (A), α-methyl-5-HT (B), pergolide (C), and
dihydroergotamine (DHE, E) is indicated for WT and V2.53L 5-HT2B receptors. Data are
presented as the mean ± SE of at least three independent experiments measured in
duplicate. To obtain log Ki values from these data, the data were fit to a heterologous
competition model of radioligand binding to one class of receptor sites that takes ligand
depletion into account (GraphPad Prism 4.0). Non-specific radioligand binding never exceeded 20% of the total radioligand binding. The V2.53A and V2.53I 5-HT2B receptors were assayed and analyzed simultaneously (see Table 4.2). Note: Because our molecular model and molecular dynamics simulations were not consistent with M5.39 directly participating in ligand binding (see below), the affinities of the compounds at the
M5.39V 5-HT2B receptor were not measured.
103
120 125 V2.53L2B-LSD V2.53L2B 100 WT2B-LSD 100 WT2B 80 75 60 50 40 H]LSD sp. bind. remaining) (% H]LSD sp. bind. remaining) (%
3 3 25 [ 20 [ A B 0 0 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 log [LSD] log [5-HT]
120 120 V2.53L2B V2.53L2B 100 WT2B 100 WT2B 80 80 60 60
40 40 (% remaining) (% remaining) (% H]LSD sp. bind. H]LSD sp. bind.
[3 20 [3 20 C D 0 0 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 log [α -methyl-5-HT] log [pergolide]
125 V2.53L2B 100 WT2B
75
50
H]LSD sp. bind. remaining) (% 3 25 [ E 0 -12 -11 -10 -9 -8 -7 -6 -5 -4 log [DHE]
104 mutations studied herein altered LSD affinity by more than two-fold (Figure 4.3, Table
4.2, and data not shown), suggesting that none of the mutations dramatically alters
receptor folding/topology. We also examined the effect of the V2.53L mutation on four
other 5-HT2B receptor agonists: 5-HT, α-methyl-5-HT, DHE, and pergolide (Figure
4.3A,B and Table 4.2). The V2.53L mutation had less or no effect on the binding of the
other agonists assayed, the largest effect being a five-fold decrease in affinity for
dihydroergotamine (Figure 4.3B and Table 4.2). The preceding observations demonstrate
that the V2.53L mutation markedly and uniquely affects SNF binding to 5-HT2B receptors.
4.2.2 Modeling, Ligand Docking Simulations, and Molecular Dynamics Simulations of Ligand Binding to 5-HT2B Receptors
To investigate the possible atomic interactions by which V2.53 in the 5-HT2B receptor contributes to high-affinity SNF binding, we performed ligand docking simulations. Each of the 30 solutions generated was inspected to determine whether 1) the protonated amine nitrogen of the ligand was close enough to D3.32 to form the conserved, anchoring ionic interaction, and 2) whether the aromatic group of the ligand was near one or more of the conserved aromatic residues known to stabilize biogenic amine ligand binding (Choudhary, Craigo et al. 1993; Roth, Shoham et al. 1997; Manivet,
Schneider et al. 2002). Only 16 of the 30 docked structures met the first criterion (i.e., the protonated amine nitrogen of the ligand was less than 3.0 Å away from one of the carboxylate oxygens of D3.32). Of these, only two placed the ligand close enough to
105 V2.53 for the latter to have an affect on SNF affinity, as was indicated by our mutant
receptor binding data (Figure 4.4A). The lower energy solution (solution 1, total score = -
10.58 kcal/mol) placed the α-methyl group carbon of SNF 3.73 Å and 3.76 Å away from the terminal γ-methyl group carbons of V2.53 (Figure 4.4A). The other solution (solution
2, total score = -10.48 kcal/mol) placed the α-methyl group carbon of SNF 3.68 Å and
4.57 Å away from the terminal γ-methyl group carbons of V2.53 (Figure 4.4B).
To determine which of the two solutions was best able to explain our mutagenesis data, we computationally introduced the V2.53L mutation into each solution, and then did ten rounds of computer-simulated annealing to examine several possible orientations of the V2.53L side chain relative to the ligand (Figure 4.4B). For the solution 1-based
MD simulations, the average distances ± SEM between the terminal δ-methyl group carbons of V2.53L and the α-methyl group carbon of SNF in the 10 lowest energy structures were 3.35 ± 0.05 Å and 4.6 ± 0.1 Å (Figure 4.4B). We compared the V2.53L terminal δ-methyl group-ligand α-methyl group intercarbon distances to those between the terminal γ-methyl group carbons of V2.53 and the α-methyl group carbon of SNF in solution 1 (i.e., 3.73 Å and 3.76 Å). Assuming favorable van der Waals’ (vdW) methyl group-methyl group interactions occur at intercarbon distances of 3-4 Å, our MD simulations suggested that the net effect of the V2.53L mutation resulted from the loss of a stabilizing vdW interaction between the 2.53 side chain and the α-methyl group of the ligand. For solution 2-based MD simulations, the V2.53L terminal δ-methyl group-ligand
α-methyl group intercarbon distances ± SEM were 3.57 ± 0.05 Å and 4.71 ± 0.09 Å
(Figure 4.4B). Comparing these receptor side chain-ligand intercarbon distances to those
106 Figure 4.4A: 3-D molecular models showing the results of ligand docking simulations that are consistent with conserved features of biogenic amine ligand binding. Simulated docking of SNF to the model shown in Figure 3 was performed using the FLEXX module of Sybyl 6.91 software. The solutions consistent with conserved features of biogenic amine ligand binding are shown. A (solution 1), C (solution 2):
Predicted orientation of SNF relative to conserved residues important for biogenic amine ligand binding. B (solution 1), D (solution 2): Space filling representation of predicted docking of SNF relative to V2.53, in which the terminal methyl groups of V2.53 and the
α-methyl group of SNF are indicated by arrowheads and an arrow, respectively. Atom color code: carbon (white); nitrogen (blue); oxygen (red); hydrogen (cyan); sulfur
(yellow). The α-carbon backbone of the helices is shown in purple.
107
108 Figure 4.4B: Representative energy-minimized structures from ten rounds of computer-simulated annealing of solution 1 (A,B) and solution 2 (C,D) after insertion of the V2.53L mutation. The results predict, in both cases, one stabilizing van der Waals’ interaction between the α-methyl group of SNF (arrows) and the terminal δ- methyl groups of V2.53L (arrowheads). Atom color code: carbon (white); nitrogen
(blue); oxygen (red); hydrogen (blue); sulfur (yellow). The α-carbon backbone of the helices is shown in purple.
109
110 between the terminal γ-methyl group carbons of V2.53 and the α-methyl group carbon of
SNF in solution 2 (3.58 Å and 4.57 Å) revealed very little difference. Accordingly,
solution 2 did not predict a change in the number of stabilizing vdW interactions between
the 2.53 side chain and the α-methyl group of SNF upon mutation of V2.53 to leucine.
Thus, vis-à-vis the V2.53L mutation, solution 1 best modeled our experimental
observations.
As for M5.39, in none of the 30 solutions was the residue predicted to be close enough to SNF to directly influence ligand binding. Furthermore, in none of the CSA simulations was the M5.39 side chain predicted to be within less than 4 Å of bound SNF.
Nevertheless, our data demonstrate a role for M5.39 in SNF binding, which our current model either indicates is indirect or does not explain.
To further test the predictive power of our model, we performed identical MD simulations and analyses after introducing a conservative V2.53I mutation into solution 1
(Figure 4.5). The average distances ± SEM between the terminal γ- and δ-methyl group carbons of V2.53I and the α-methyl group carbon of SNF in the ten lowest energy structures were 3.91 ± 0.05 Å and 3.8 ± 0.1 Å, respectively. Comparing these values to those for the terminal γ-methyl group carbons of V2.53 and the α-methyl group carbons of SNF in solution 1 (3.73 Å and 3.76 Å) revealed very little difference. Thus, solution 1 predicted that the conservative V2.53I mutation would not alter vdW interactions with the α-methyl group of SNF. Based on these predictions, we made the V2.53I mutant 5-
HT2B receptor and measured its affinity for SNF, which we expected to be similar to that
111 Figure 4.5: Representative energy-minimized structure from ten rounds of computer-simulated annealing of solution 1 after insertion of the V2.53I mutation.
The result predicts two stabilizing van der Waals’ interaction between the α-methyl group of SNF (arrow) and the terminal methyl groups of V2.53I (arrowheads). (A) Stick representation; (B) space filling representation. Atom color code: carbon (cyan); nitrogen
(blue); oxygen (red); hydrogen (white); sulfur (yellow). The α-carbons backbone of the seven helices is shown in purple.
112
113 of the wild type receptor. As predicted by our model, the affinity of the V2.53I mutant receptor for SNF was not altered by the mutation (Table 4.1). Solution 1, therefore, best modeled the effects of both the V2.53L and the V2.53I mutations. Notably, the V2.53I mutation did decrease 5-HT and α-methyl-5-HT affinity by approximately ten-fold, while the affinity for DHE, pergolide, and LSD (the radioligand) were affected less than three- fold by the mutation (Figure 4.3C and Table 4.2).
Assuming that the R enantiomer of norfenfluramine (RNF) binds in the same orientation as SNF, our model suggested that the α-H of RNF would appose V2.53 and that the α-methyl group would project “down” into the interhelical space, away from
V2.53. Because hydrogen atoms, due to their smaller atomic radius, are less efficient than bulkier methyl groups at forming attractive vdW interactions, and because the α-H of
RNF would be more than 4 Å away from V2.53, our model predicted that RNF would bind to the wild type 5-HT2B receptor with a lower affinity than SNF. As shown in Table
4.3, RNF displayed a 3-fold decrease in affinity compared to SNF at the wild type 5-
HT2B receptor. Furthermore, because the R-(-) α-methyl group of RNF would be more than 4 Å away from V2.53 and therefore not likely to form stabilizing vdW interactions, our model predicted that the V2.53L mutation would have a smaller effect on RNF affinity than it did on SNF affinity. As shown in Table 4.3, the affinity for RNF was decreased only 3-fold by the V2.53L mutation, compared to a 17-fold decrease in SNF affinity due to the mutation.
114 Table 4.3: Affinity constants (Ki’s) for SNF and congeners at wild type and V2.53L
5-HT2B receptors and wild type 5-HT2C and 5-HT2A receptors. p values were obtained
from F-tests comparing curve fits of competition binding isotherms for two receptors:
ap<0.05 compared to S-(+)-NF; bp<0.05 compared to R-(-)-NF; cp<0.05 compared to α- desmethyl-NF; dp<0.05 compared to WT.
115
WT V2.53L WT2C WT2A Drug pKi pKi pKi pKi (Ki, nM) (Ki, nM) (Ki, nM) (Ki, nM) 7.76 ± 0.05 6.64 ± 0.07d 7.16 ± 0.07 5.74 ± 0.08 SNF (17) (230) (69) (1,800) 7.21 ± 0.04 6.84 ± 0.07d 7.07 ± 0.08 5.85 ± 0.07 RNF (62) (140) (85) (1,400) α-desmethyl- 7.26 ± 0.06a 6.58 ± 0.05b,d 7.06 ± 0.07 5.87 ± 0.06 NF (55) (260) (87) (1,300) 6.11 ± 0.08a,b,c 6.44 ± 0.08a,b,c,d 6.0 ± 0.1 5.0 ± 0.1a,b,c α-ethyl-NF (780) (360) (1,000)a,b,c (10,000)
116 We also measured the affinity of α-desmethyl-NF at the wild type and V2.53L
mutant 5-HT2B receptors. As shown in Table 4.3, α-desmethyl-NF, like RNF, exhibited a
3-fold reduction in affinity at the 5-HT2B receptor compared to SNF. The affinity for α-
desmethyl-NF was reduced 5-fold by the V2.53L mutation (Table 4.3). This effect was
very similar to the reduction in RNF affinity caused by the V2.53L mutation, suggesting
that the decrease in RNF affinity due to the V2.53L mutation did not result from altered
interactions with the α-methyl group. Our model, however, suggested that the V2.53L
side chain forms one vdW interaction with the α-methyl group of SNF. As such, we
would have expected the V2.53L 5-HT2B receptor to have lower affinity for α-desmethyl-
NF than for SNF.
To investigate potential interactions between the V2.53L and α-desmethyl-NF, we
replaced the α-methyl group of SNF with a hydrogen atom in our V2.53L 5-HT2B receptor model, then performed computer-simulated annealing to explore several possible orientations of the V2.53L side chain relative to α-desmethyl-NF. The results of our MD simulations predicted that in the absence of an S-(+) α-methyl group, at least one of the terminal δ-methyl groups of V2.53L was close enough to the α-carbon of α-desmethyl-
NF to form a stabilizing vdW interaction (Figure 4.6A,B). With SNF docked, however, the V2.53L side chain was predicted to be oriented such that only one of the terminal δ- methyl groups apposed the ligand. Such an orientation accommodates the S-(+) α-methyl group (Figure 4.4B). Thus, our MD simulations provided a plausible explanation for the lack of difference in SNF and α-desmethyl-NF affinity at the V2.53L 5-HT2B receptor:
removal of the S-(+)-methyl group, with which one of the V2.53L terminal δ-methyl
117 Figure 4.6: Representative energy-minimized structure from ten rounds of
computer-simulated annealing of solution 1 bearing the V2.53L after addition or
removal of SNF α-carbon substituents. (A,B) The result predicts a stabilizing van der
Waals’ interaction between a terminal δ-methyl group of V2.53L (arrowheads) and the α- carbon of α-desmethyl-NF (arrow). (C,D) The result predicts a stabilizing van der Waals’ interaction between a terminal δ-methyl group of V2.53L (arrowheads) and the α-carbon-
proximal methylene group of the α-ethyl substituent of α-ethyl-NF; the terminal ethyl
group of the α-ethyl substituent points ‘down’ away from V2.53L. Atom color code:
carbon (cyan); nitrogen (blue); oxygen (red); hydrogen (white); sulfur (yellow). The α- carbons backbone of the seven helices is shown in purple.
118
119 groups forms a stabilizing vdW interaction, allows V2.53L to project into the binding
pocket, permitting a stabilizing vdW interaction between a V2.53L terminal δ-methyl
group and the α-carbon of α-desmethyl-NF.
Given our mutagenesis data and modeling predicting that the α-methyl group of
SNF is within 4 Å of V2.53, we hypothesized that an α-carbon substituent bulkier than a
methyl group (i.e., an ethyl group) would reduce NF binding to the wild type 5-HT2B receptor. As shown in Table 4.3, α-R,S-(±)-ethyl-NF exhibited a 46-fold lower affinity for the wild type 5-HT2B receptor than did SNF. According to our model, the reduction in
affinity due to the α-ethyl substitution was most likely due to 1) steric hindrance between
the α-ethyl group of the ligand and V2.53, or 2) the adoption of a different binding
orientation due to non-tolerance of the α-ethyl group by V2.53. In either case, our results
with α-R,S-(±)-ethyl-NF corroborated our model implicating interactions between the α-
methyl group of SNF and the γ-methyl groups of V2.53 in the ligand’s high-affinity
binding to 5-HT2B receptors. At the V2.53L 5-HT2B receptor, α-R,S-(±)-ethyl-NF affinity
was disrupted less than two-fold compared to SNF. Initially, because leucine is bulkier than valine, we had predicted the V2.53L mutant 5-HT2B receptor to be more sensitive to
the bulkier α-carbon substituent of α-R,S-(±)-ethyl-NF.
To explore possible rationales for the lack of difference between SNF and α-R,S-
(±)-ethyl-NF affinity at the V2.53L 5-HT2B receptor, we replaced the α-methyl group of
SNF with an ethyl group in our model, and then performed computer-simulated annealing
120 to explore several possible orientations of the V2.53L side chain relative to the α-ethyl group. As depicted in Figure 4.6C,D, our MD simulations predicted the V2.53L side chain and the α-ethyl group of the ligand to be oriented such that one V2.53L terminal δ- methyl group could form a favorable vdW interaction with the α-carbon-proximal methylene group of the α-ethyl substituent; the terminal methyl group of the α-ethyl substituent projected ‘down’ into the binding pocket, away from the V2.53L side chain.
Thus, our MD simulations suggested that for SNF, α-desmethyl-NF, and α-R,S-(±)-ethyl-
NF binding to the V2.53L 5-HT2B receptor, the V2.53L side chain forms one stabilizing vdW interaction with each, consistent with the compounds having similar affinity (Table
4.3).
If, indeed, the decrease in SNF affinity in the V2.53L mutant was due to the loss of one stabilizing vdW interaction between the 2.53 side chain and the α-methyl group of
SNF—and not steric hindrance due to the bulkiness of the V2.53L side chain—we expected that a V2.53A mutation would result in the loss of two stabilizing vdW interactions and, thus, an even larger decrease in SNF affinity than that caused by the
V2.53L mutation. To test this prediction, we made the V2.53A mutant receptor and measured its affinity for SNF. As shown in Figure 4.7 and Table 4.1, and as predicted by our model, the V2.53A mutation caused a dramatic 190-fold decrease in SNF affinity.
The affinities of other agonists were altered much less by the V2.53A mutation (Table
4.2), suggesting a SNF-specific effect. These observations strongly suggest that the effect of the V2.53L mutation on SNF affinity was most likely due to a decrease in vdW stabilization rather than steric hindrance.
121 Figure 4.7: Competition binding isotherms for SNF at wild type and V2.53A 5-HT2B receptors. The percent total binding of ~1 nM [3H]LSD remaining in the presence of the
indicated concentration of SNF is shown. Data are presented as the mean ± SE of at least
three independent experiments measured in duplicate. To obtain log Ki values from these data, the data were fit to a heterologous competition model of radioligand binding to one class of receptor sites that takes ligand depletion into account (GraphPad Prism 4.0).
Non-specific radioligand binding was never exceeded 20% of the total radioligand binding.
122
150 V2.53A2B-SNF 125 WT2B-SNF 100
75
50 H]LSD sp. bind. (% remaining) 3 [ 25
0 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 log [drug]
123 Having established the importance of interactions between V2.53 and the α-
methyl group of SNF in ligand binding to 5-HT2B receptors, we sought to determine
whether the α-methyl group of SNF contributes to the ligand’s binding to 5-HT2C and 5-
HT2A receptors. To do so, we measured the affinity of the SNF analogs RNF, α- desmethyl-NF and α-R,S-(±)-ethyl-NF at the two receptors. As shown in Table 4.3, the affinities of α-desmethyl-NF at the 5-HT2C and 5-HT2A receptors were not different than
SNF and RNF affinities, suggesting that the α-methyl group of SNF does not contribute
significantly to ligand binding to those receptors. In contrast, NF analogs lacking an S-(+)
α-methyl group (i.e., RNF and α-desmethyl-NF) exhibited a 3-to-4-fold reduction in
binding affinity at wild type 5-HT2B receptors, reflecting productive receptor-S-(+) α-
methyl group interactions. The affinities of the 5-HT2C and 5-HT2A receptors for α-R,S-
(±)-ethyl-NF were reduced 14-fold and 5-fold, respectively, compared to SNF; the wild
type 5-HT2B receptor exhibited a dramatic 46-fold decrease in affinity for the compound compared to SNF. Thus, a bulkier α-carbon substituent has the most dramatic effect (~50-
fold) on NF binding to 5-HT2B receptors, a much lesser (~10-fold) effect on NF binding
to 5-HT2C receptors, and a modest (5-fold) effect on NF binding to 5-HT2A receptors.
4.2.3 Effect of Point Mutations on Agonist Potency and Efficacy
For each of the wild type and point mutant 5-HT2B receptors, we also measured SNF
potency (EC50) and relative efficacy compared to the full agonist 5-HT (Emax) (Table 4.4
and Figure 4.8). As was true for ligand affinity, agonist potency was decreased only by
the V2.53L and M5.39V mutations (Table 4.4 and Figure 4.8A-C). In contrast to its
124 effect on Table 4.4: Potency (EC50) and relative efficacy (Emax) values for SNF and
RNF at wild type and point mutant 5-HT2 receptors. Emax is reported as % maximum
response to 5-HT. p values were obtained from F-tests comparing curve fits of concentration-response isotherms for two receptors: ap<0.05 compared to the appropriate
b c WT 5-HT2 receptor; p<0.05 compared to V2.53L 5-HT2B receptor; p<0.05 compared to
the M5.39V 5-HT2B receptor. ND = not determined.
125
pEC SNF ± SE pEC RNF ± SE SNF E ± SE RNF E ± SE Receptor 50 50 max max (EC50 SNF, nM) (EC50 RNF, nM) (% 5-HT max.) (% 5-HT max.) 7.66 ± 0.07 6.62 ± 0.07 5-HT WT 75 ± 3 81 ± 4 2B (22) (240) 7.7 ± 0.1 A1.35P ND ND ND (20) 7.69 ± 0.06 A1.35S ND ND ND (20) 7.51 ± 0.08 L1.38S ND ND ND (30) 7.50 ± 0.09 I1.39T ND ND ND (31) 7.44 ± 0.08 V1.42I ND ND ND (36) 7.09 ± 0.07a 6.99 ± 0.07a V2.53L 101.7 ± 0.6a 28.2 ± 0.5a (82) (100) 7.7 ± 0.1 7.4 ± 0.1a V2.53I 87 ± 4a 73 ± 4 (20) (40) 7.92 ± 0.03a L3.29I ND ND ND (12) 7.19 ± 0.09a 6.7 ± 0.1 M5.39V 70 ± 3 74 ± 5 (64) (200) V2.53L, 6.49 ± 0.05a,b,c 6.80 ± 0.08b 73 ± 4b 18 ± 1a,b,c M5.39V (320) (160) 7.6 ± 0.1 E7.36N ND ND ND (20) 7.8 ± 0.2 S7.45C ND ND ND (10) 6.7 ± 0.2 6.6 ± 0.2 5-HT WT 75 ± 3 69 ± 5 2A (200) (300) 5.3 ± 0.1a 5.07 ± 0.09a 5 HT L2.53V 70 ± 3 68 ± 5 2A (5000) (8500) 7.2 ± 0.2 6.7 ± 0.1 5-HT WT 93 ± 6 89 ± 6 2C (70) (200) 7.2 ± 0.1 6.97 ± 0.8a 5-HT V2.53L 85 ± 5 82 ± 4 2C (70) (100)
126 Figure 4.8: Concentration-response isotherms for agonist-stimulated inositol
phosphate accumulation. Agonist-, concentration-dependent activation of WT (A),
V2.53L (B), M5.39V (C), V2.53I (D), and V2.53L,M5.39V (E) 5-HT2B receptors
transiently expressed in HEK293 cells is shown. Data represent the mean ± SE of at least three independent experiments measured in duplicate. Average baseline and maximal
3 drug-stimulated [ H]IP accumulation (in DPM) in wild type 5-HT2B receptor-expressing
cells were: 2,700 ± 400 (baseline), 13,000 ± 3,000 (5-HT), 10,000 ± 2,000 (SNF), and
11,000 ± 2,000 (RNF); corresponding values for V2.53L 5-HT2B receptor-expressing
cells were: 900 ± 100 (baseline), 13,000 ± 5,000 (5-HT), 14,000 ± 5,000 (SNF), and
4,000 ± 2,000. To obtain log EC50 and Emax values from concentration-response data, the
data were fit to a three parameter logistic concentration-response model (GraphPad Prism
4.0). Emax is expressed as percent of the maximum response to 5-HT.
127
120 120 WT2B-5-HT V2.53L2B-5-HT 100 A WT2B-SNF 100 B V2.53L2B-SNF WT2B-RNF 80 80 V2.53L2B-RNF
60 60 accumulation accumulation x 40 x 40 (% 5-HTmax.) (% 5-HTmax.) H]IP
20 H]IP 3 20 3 [ [ 0 0 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 log [drug] log [drug]
120 M5.39V2B-5-HT 120 100 C M5.39V2B-SNF V2.53I2B-5-HT 100 D V2.53I2B-SNF 80 M5.39V2B-RNF 80 V2.53I2B-RNF 60 60 accumulation
x 40 accumulation
x 40 (% 5-HT max.) (% 5-HTmax.) H]IP 20 3 [ H]IP 20 3 0 [ -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 0 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 log [drug] log [drug]
120 103,218 5-HT 100 E 103,218 SNF 80 103,218 RNF 60 accumulation
x 40 (% 5-HT max.)
H]IP 20 3 [ 0 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 log [drug]
128 SNF affinity at 5-HT2B receptors, the V2.53L mutation in the 5-HT2C receptor did not
alter the agonist’s potency (Table 4.4). This is in line with our finding that SNF affinity
was less sensitive to the V2.53L mutation in the 5-HT2C receptor than in the 5-HT2B receptor. The L2.53V mutation in the 5-HT2A receptor caused a dramatic 20-fold
decrease in the potency of SNF (Table 4.4). Such a change is contrary to what would
have been predicted based on valine in position 2.53 contributing to SNF potency at 5-
HT2 receptors and suggests that residue L2.53 plays a different role in SNF potency at the
5-HT2A receptor than does V2.53 in the 5-HT2B and 5-HT2C receptors.
Only the V2.53L mutation in the 5-HT2B receptor altered the agonist relative efficacy: SNF efficacy was increased 36% by the mutation (Table 4.4 and Figure
4.8A,B). In terms of RNF potency, the V2.53L mutation in both the 5-HT2B and 5-HT2C receptors increased potency to a similar extent: approximately 2-fold (Table 4.4 and
Figure 4.8A,B). The L2.53V mutation in the 5-HT2A receptor decreased RNF potency 30-
fold (Table 4.4), an effect similar to that observed for SNF potency. With respect to RNF
relative efficacy, as with SNF relative efficacy, only the V2.53L mutation in the 5-HT2B receptor affected the parameter, decreasing it a dramatic 65% (Table 4.4 and Figure
4.8A,B). However, while without effect on agonist efficacy itself, the M5.39V mutation in tandem with the V2.53L mutation eliminated the latter mutation’s effect on SNF efficacy and increased its effect on RNF efficacy ~1.2-fold (Figure 4.8E and Table 4.4).
129 4.3 Discussion
The main finding reported herein is that V2.53 in the 5-HT2B receptor contributes
to the high-affinity binding of SNF through vdW interactions between both V2.53
terminal γ-methyl groups and the α-methyl group of the ligand. Of the 11 5-HT2B-to-5-
HT2C/2A mutations tested, only the V2.53L (5-HT2B/2C-to-5-HT2A) and M5.39V (5-HT2B- to-5-HT2C/2A) mutations resulted in substantial changes in SNF affinity (Figure 4.2 and
Table 4.1). Residue 2.53 in the 5-HT2C receptor is also a valine; however, the V2.53L
mutation in the 5-HT2C receptor diminished SNF binding only half as much as it did in
the 5-HT2B receptor (Figure 4.2 and Table 4.1). The inverse L2.53V mutation in the 5-
HT2A receptor did not affect SNF affinity (Figure 4.2 and Table 4.1), suggesting that
interactions between V2.53 and SNF are more important in 5-HT2B receptors than in 5-
HT2C receptors, and that L2.53 plays no role in SNF binding to the 5-HT2A receptor.
Given the rank order affinity of SNF for 5-HT2 receptors (5-HT2B>5-HT2C>5-HT2A), our
data suggest that interactions between V2.53 in the 5-HT2B receptor and the α-methyl
group of the ligand contribute to sub-type selectivity.
To ascertain whether the V2.53L mutation affected the binding of other agonists,
we measured the affinity of two tryptamines (5-HT and α-methyl-5-HT) and three
ergolines (dihydroergotamine, pergolide, and LSD) at 5-HT2B receptors. The data
presented in Table 4.2 demonstrate that 2.53 mutations in the 5-HT2B receptor did not
globally perturb agonist binding. Notably, LSD (i.e., the radioligand) affinity was altered
130 less than 2-fold. Taken together, the preceding results demonstrate that the effect of the
V2.53L mutation is selective for SNF.
To explore how V2.53 might contribute to SNF affinity at 5-HT2B receptors, we
performed ligand docking simulations. Of the 30 solutions generated, only two were both
consistent with conserved features of biogenic amine ligand binding and predicted V2.53
to be close enough to SNF for the residue to directly affect ligand binding (i.e., within 4
Å) (Figure 4.4). In both of these, the α-methyl group of the ligand was predicted to be
near both (solution 1) or one (solution 2) of the V2.53 terminal γ-methyl groups. After
introduction of the V2.53L mutation into each solution and computer-simulated
annealing to explore possible side chain orientations relative to the α-methyl group of
SNF, both solutions predicted only one vdW interaction (Figure 4.4). For solution 2, then,
V2.53 and V2.53L were predicted to interact similarly with the ligand, each contributing
one stabilizing vdW interaction. For solution 1, V2.53 was predicted to contribute two
stabilizing vdW interactions and V2.53L was predicted to contribute only one.
Calculating the SNF binding energy using the Nernst equation (∆G = -2.303RTlogKi) for
each experiment and averaging the results gives a value of -10.3 ± 0.1 kcal/mol for the
wild type and -8.66 ± 0.08 kcal/mol for the V2.53L mutant. Thus, the energetic cost (i.e.,
the ∆∆G) of the V2.53L mutation on SNF binding was +1.6 ± 0.2 kcal/mol. Others have measured the contribution of methyl group-methyl group interactions to ligand binding
energy, with values ranging from +0.4 kcal/mol to +1.4 kcal/mol (Bigler, Lu et al. 1993;
Huang, Lu et al. 1995; Morton, Baase et al. 1995; Faergeman, Sigurskjold et al. 1996; Lu,
Apostol et al. 1997; Oue, Okamoto et al. 1997; Kawaguchi and Kuramitsu 1998). The
131 value calculated from our data (+1.6 ± 0.2 kcal/mol) is at the upper end of that range and is thus consistent with the V2.53L mutation causing the loss of a 2.53 methyl group-SNF
α-methyl group vdW interaction, a prediction of our modeling and MD simulations. In addition, the ∆∆G for the V2.53A mutation reveales a +3.0 ± 0.2-kcal/mol difference in
SNF binding energy compared to the wild type. Thus, removal of the second stabilizing methyl group-methyl group vdW interaction costs [+3.0 ± 0.2 kcal/mol – (+1.6 ± 0.2 kcal/mol)] = +1.4 ± 0.3 kcal/mol, again consistent with reported interaction energies and further supporting our model predicting two stabilizing vdW interactions between V2.53 and the α-methyl group of SNF.
To further test our modeling and MD simulations, we studied the effect of a
V2.53I mutation on 2.53-SNF interactions. Computer-simulated annealing predicted that
V2.53I would, like V2.53, form two stabilizing vdW interactions with the α-methyl group of SNF (Figure 4.5). Thus, we predicted that the mutation would not affect SNF affinity.
As shown in Table 4.1, the V2.53I 5-HT2B receptor’s affinity for SNF was not significantly different from that of the wild type, in agreement with our modeling and
MD simulations. This result validates our computational approach and substantiates the prediction that the α-methyl group of SNF interacts with both terminal methyl groups of
V2.53.
Given the apparent importance of the α-methyl group-V2.53 interaction for SNF binding to 5-HT2B receptors, we measured the affinity of the wild type and V2.53L mutant receptors for NF analogs bearing α-carbon “mutations.” Assuming that RNF
132 binds in the same orientation as SNF, our model predicts that the α-methyl group of RNF
will project “down” towards the cytoplasmic face of the receptor and the α-H will appose
V2.53. We reasoned that both the increased distance of the α-methyl group from V2.53
and the smaller vdW radius of the apposing α-H would render RNF much less sensitive to
the V2.53L mutation. As shown in Table 4.3, RNF affinity was indeed decreased only 3-
fold by the V2.53L mutation, compared to a 17-fold effect on SNF affinity. Similarly, α-
desmethyl-NF also exhibited only a 3-fold decreased affinity at V2.53L 5-HT2B receptors
compared to its affinity at the wild type receptor (Table 4.3). Thus, the 5-HT2B receptor-
selective effect of the V2.53L mutation on NF binding requires the ligand to bear an α- methyl group in the S-(+) configuration; the R-(-) α-methyl group does not appear to contribute to NF binding to the wild type 5-HT2B receptor—an observation that is
consistent with the similar affinities of α-desmethyl-NF, which lacks an α-methyl group,
and RNF. We also measured the affinity of α-R,S-(±)-ethyl-NF for wild type 5-HT2B receptors. We hypothesized that, if the α-methyl group of SNF is within 4 Å of either terminal γ-methyl group of V2.53, an α-ethyl substituent would be poorly tolerated due to steric hindrance. As shown in Table 4.3, the 5-HT2B receptor exhibited a 46-fold decreased affinity for α-R,S-(±)-ethyl-NF compared to SNF, further supporting our model wherein V2.53 of the 5-HT2B receptor is within 4 Å of the α-methyl group of SNF—close
enough for both terminal γ-methyl groups to form stabilizing vdW interactions with the
α-methyl group of ligand. In addition, the 5-HT2C and 5-HT2A receptors displayed
considerably smaller reductions in α-R,S-(±)-ethyl-NF affinity—14-fold and 5-fold,
respectively (Table 4.3). One possible explanation for the reduced sensitivity of the 5-
HT2C and 5-HT2A receptors to α-carbon ‘mutations’ in the ligands is that the ligands may
133 bind in a different orientation to the 5-HT2C and 5-HT2A receptors than does SNF to the
5-HT2B receptor.
Our studies of NF congener binding to the V2.53L 5-HT2B are also consistent
with interactions between V2.53 and the α-methyl group of SNF. For instance, our
modeling and MD simulations predicted the formation of one stabilizing vdW interaction
between V2.53L and SNF (compared to two such interactions between V2.53 and the α-
methyl group of SNF). At first glance, one would have expected “removal” of the α-
methyl group (i.e., RNF or α-desmethyl-NF) to eliminate the remaining stabilizing vdW
interaction, thereby decreasing affinity. However, we observed that RNF and α- desmethyl-NF bound to the V2.53L 5-HT2B receptor with affinities similar to that of
SNF. In accord with the previous observations, our modeling and MD simulations suggested that, absent an S-(+) α-methyl group, both V2.53L terminal δ-methyl groups
can project into the binding pocket and form at least one stabilizing vdW interaction with the α-carbon of α-desmethyl-NF. Thus, our MD simulations provided a plausible
molecular rationale for the observation that SNF, RNF, and α-desmethyl-NF displayed
similar affinities at the V2.53L 5-HT2B receptor. Our modeling and MD simulations also
provided an explanation for the similar affinities of α-R,S-(±)-ethyl-NF and SNF for the
V2.53L 5-HT2B receptor. With α-R,S-(±)-ethyl-NF bound in the pocket, our MD simulations predicted that the V2.53L side chain orients such that one of the V2.53L terminal δ-methyl groups can form a stabilizing vdW interaction; the other terminal δ- methyl group projects away from the ligand, leaving room for the bulkier α-carbon substituent.
134 A much simpler explanation for the effect of the V2.53L mutation on SNF
binding to the 5-HT2B receptor is that the mutation results in steric hindrance with some
group(s) on the ligand. However, the observations that 1) the bulky V2.53I side chain did
not affect SNF binding, and 2) generating a larger region of bulk tolerance via a V2.53A
mutation resulted in a much larger perturbation of SNF binding than did the V2.53L
mutation (i.e., 46-fold vs. 17-fold) argue against the simpler explanation. However, the fact that the wild type 5-HT2B receptor displays smaller decreases in affinity for RNF and
α-desmethyl-NF than that caused by the V2.53L mutation on SNF affinity suggests the
existence of additional factors. For example, it is possible that the S-(+) α-carbon substitutent causes SNF to bind in a different orientation to the 5-HT2B receptor than do
SNF analogs without an S-(+) α-carbon substituent.
A secondary finding of this study is that M5.39 also participates in SNF binding.
We attempted to explain the role of M5.39 in SNF binding, as we had for V2.53, using
modeling, ligand docking simulations, and MD simulations. In none of the computer-
simulated annealing results, however, was M5.39 predicted to be less than 10 Å away
from bound SNF. A possible explanation for our M5.39 data that is consistent with our
model is that the residue stabilizes, through hydrophobic and/or hydrogen bonding
interactions, extracellular loop 2 (e2) in the binding pocket in such a way that the loop
favorably interacts with bound SNF. Indeed, the analogous dopamine D2 receptor residue
(V5.39) along with e2 likely plays a role in the binding of some ligands (Shi and Javitch
2004). Alternatively, given its position at the extracellular surface of the receptor, M5.39
might participate in initial receptor-ligand interactions that guide SNF “down” into the
135 binding pocket. It is also possible that the M5.39 mutation causes a subtle conformational
change in transmembrane helix 5 that brings the residue close enough to the ligand
binding pocket to have a direct effect on SNF binding.
We also determined SNF potency and efficacy at wild type and mutant 5-HT2 receptors (Figure 4.8 and Table 4.4). SNF potency, like its binding affinity, was only decreased at V2.53L and M5.39V mutant 5-HT2B receptors (Figure 4.8 and Table 4.4).
The changes in SNF potency at 5-HT2B receptors caused by the V2.53L and M5.39V
mutations were much less dramatic than the effect of either mutation on SNF affinity at
5-HT2B receptors (i.e., 3-to-4-fold vs. 10-to-20-fold, respectively, compared to the wild
type value). A similar result has been reported by Manivet et al. (Manivet, Schneider et
al. 2002), who observed 10-to-15-fold decreases in 5-HT affinity (KD) due to either a
W6.48A or a F6.52A mutation, though neither mutation altered 5-HT potency. Thus, the finding that a mutation can affect ligand affinity without a concomitant effect on potency is not unprecidented. One possible explanation for the larger decrease in affinity than potency due to the V2.53L and M5.39V is that the mutations hinder SNF-induced conformational changes that selectively affect 5-HT2B receptor coupling to G proteins not
involved in inositol phosphate accumulation. In this regard, work from Dr. Luc
Maroteaux’s group (Manivet, Mouillet-Richard et al. 2000; Deraet, Manivet et al. 2005)
has demonstrated that the 5-HT2B receptor couples to both Gαq and Gα13, and that Gαq, but not Gα13, mediates inositol phosphate accumulation. Thus, it is possible that the
V2.53L and M5.39V mutations have a greater effect on SNF-induced 5-HT2B receptor
coupling to non-Gαq G proteins than they do on coupling to Gαq G proteins. This could
136 be directly addressed by introducing into the V2.53L and M5.39V mutant 5-HT2B receptors the recently described R393X mutation, which abolishes Gq coupling (and inositol phosphate accumulation) (Blanpain, Le Poul et al. 2003; Deraet, Manivet et al.
2005) but not G13 coupling (stimulation of nitric oxide synthase, activation of MAPK, and [3H]thymidine incorporation) (Deraet, Manivet et al. 2005). Since the R393X
mutation does not affect SNF affinity per se, if the V2.53L and/or M5.39V mutations
were to have a similar effect on SNF potency at activating MAPK, for example, as they
do on SNF affinity (i.e. 10-to-20 fold), such a result would suggest that the mutations do,
indeed, preferentially affect SNF-induced activation of non-Gαq G proteins. This would
suggest a molecular basis for the observation that, despite causing greater-than-10-fold
decreases in affinity, the V2.53L and M5.39V mutations only cause subtle, 3-4-fold
decreases in SNF potency.
We observed a slightly different result for the effect of the V2.53L and M5.39V
mutations on RNF potency at 5-HT2B receptors. The V2.53L mutation increased RNF
potency at the 5-HT2B receptor 2.5-fold compared to the wild type value, even though the mutation caused a 4-fold decrease in RNF affinity relative to the agonist’s affinity at wild
type receptors. The V2.53I mutation, which did not affect SNF or RNF affinity at 5-HT2B receptors, nor did the mutation affect SNF potency at 5-HT2B receptors, caused a 6-fold
increase in RNF potency at 5-HT2B receptors. Again, there is a discrepancy between the
effect of V2.53 mutations on agonist binding affinity and potency at 5-HT2B receptors. As
speculated above, the discrepancy may stem from the V2.53 mutations differentially
affecting 5-HT2B receptor coupling to G proteins. For example, for SNF, the V2.53L
137 mutation may reduce agonist-induced 5-HT2B receptor coupling to Gαq 4-fold, while for
RNF, the mutation may increase agonist-induced 5-HT2B receptor-Gαq coupling 2-fold.
However, given the reduction in SNF and RNF affinity at 5-HT2B receptors due to the
V2.53L mutation, overall NF-induced 5-HT2B coupling to non-Gαq G proteins would be
expected to be reduced. Again, further experiments are required to ascertain whether the
V2.53L mutation alters constitutive or agonist-induced coupling to any or all G proteins.
As for agonist relative efficacy, a unique role for V2.53 in 5-HT2B receptor activation became apparent (Figure 4.8 and Table 4.4). That is, both SNF and RNF Emax were altered only in the 5-HT2B receptor, and only by the V2.53L. The V2.53I mutation did not affect NF relative efficacy at 5-HT2B receptors, suggesting that leucine at position
2.53 in the 5-HT2B receptor plays a unique role in NF-mediated receptor activation.
Specifically, the V2.53L mutation, despite reducing SNF potency at 5-HT2B receptors 4-
fold compared to the wild type value, increases SNF Emax at 5-HT2B receptors by 36%
compared to the wild type value. Conversely, for RNF, the V2.53L mutation increased
the agonist’s potency at 5-HT2B receptors about 2-fold compared to the wild type value, but decreased the agonist’s relative efficacy at 5-HT2B receptors by 65% compared to the
wild type value. How the mutation might simultaneously change potency and efficacy in opposite directions is not readily obvious in terms of NF-stimulated 5-HT2B receptor-G protein coupling. Perhaps the V2.53L mutation-induced decrease in SNF potency at 5-
HT2B receptors slows the acute desensitization (i.e., phosphorylation of the receptor by G protein-coupled receptor kinases, binding of arrestins, internalization) kinetics allowing more G proteins to be activated per SNF-bound receptor leading to increased
138 phospholipase Cβ activation. In contrast, the V2.53L mutation, which increases RNF
potency at 5-HT2B receptors, speeds up the acute desensitization kinetics thereby
reducing the number of G proteins stimulated per RNF-bound receptor. Studies of the
NF-mediated inositol phosphate accumulation in cells expressing desensitization-
deficient 5-HT2B receptors bearing the V2.53L mutation would be instructive. In this
regard, Gray et al. (Gray, Compton-Toth et al. 2003) recently reported the importance of
two serines in the rat 5-HT2A receptor—one in the i2 loop and one in the C-terminus—for
agonist-induced desensitization. By analogy, alanine scanning mutagenesis of i2 and C-
terminal serines in the 5-HT2B receptor might generate a desensitization-deficient
receptor. Such a tool would allow us to investigate whether the V2.53L mutation alters
NF Emax at 5-HT2B receptors via an effect on desensitization kinetics.
The M5.39V mutation, which also affected NF affinity and potency at 5-HT2B receptors, did not affect NF relative efficacy at 5-HT2B receptors. This finding
underscores the unique effect of the V2.53 mutations on NF-mediated activation of 5-
HT2B receptors. Unconstrained MD simulations in which the helices are allowed to move
relative to one another could yield insights into why V2.53 mutations—and not the
M5.39V mutation—affect NF potency and efficacy. For instance, a comparison of MD
simulations with apo-5-HT2B receptor and SNF-bound 5-HT2B receptor may reveal V2.53
interactions that, in the presence of ligand, contribute to helix movements that expose
intracellular receptor surfaces that contribute to G protein activation. Similar MD
simulations with a V2.53L 5-HT2B receptor could shed light on how the mutation might
139 affect such helical movements, leading to a decrease in SNF potency and/or SNF-induced
5-HT2B receptor-G protein coupling.
The V2.53L mutation in the 5-HT2A receptor caused a dramatic ~25-fold decrease
in potency of both SNF and RNF, though neither SNF nor RNF affinity at 5-HT2A receptors was affected by the mutation. NF relative efficacy at 5-HT2A receptors was also
not affected by the V2.53L mutation. Thus, if the observed effects of the V2.53L
mutation on NF potency at 5-HT2B receptors were due to altered 5-HT2B receptor-G
protein coupling and desensitization, as was speculated above to account for
discrepancies among affinity, potency, and relative efficacy, the effect of the V2.53L
mutation on NF potency at 5-HT2A receptors would likely be due to a different
mechanism. If, for example, the V2.53L mutation-induced decrease in NF potency at 5-
HT2A receptors were due to a decrease in 5-HT2A receptor-G protein coupling, the
mechanism proposed above to account for the mutations effects on NF potency at 5-HT2B receptors would suggest that Emax would be increased (due to retarded desensitization kinetics). The data show this is not the case. Therefore, the mechanism by which the
V2.53L mutation alters NF potency at 5-HT2A and 5-HT2B receptors is likely different.
In conclusion, we have discovered that residue 2.53—a valine—plays a role in
SNF binding to, and activation of, the 5-HT2B receptor that is unique among 5-HT2 family receptors. Our studies suggest the existence of two stabilizing vdW interactions between the terminal methyl groups of V2.53 and the α-methyl group of SNF in the ligand-bound receptor. These interactions also have consequences on receptor function,
140 since SNF potency and efficacy are both affected when these interactions are perturbed.
Thus, vdW interactions between V2.53 and the α-methyl group of SNF contribute to the sub-type selective pharmacology. Targeting these interactions via ligand ‘mutations’
(e..g., by synthesizing SNF analogs lacking, or bearing bulkier, α-carbon substituents in
the S-(+) orientation) may result in a more selective 5-HT2C agonist. Indeed, as we show herein, NF congeners either lacking an α-methyl group or bearing a bulkier α-ethyl subsitutent exhibit reductions in 5-HT2B—but not 5-HT2C or 5-HT2A—receptor affinity.
Similar ligand-receptor interactions that are unique to the 5-HT2B receptor among 5-HT2 receptors are likely to exist for SNF-induced activation as well, and a better understanding of these could guide drug design efforts towards an effective appetite suppressant devoid of VHD- and PH-inducing potential.
141 CHAPTER 5: IMPLICATIONS AND FUTURE DIRECTIONS
5.1 Summary
The work presented herein (Chapters 3 and 4) describes a potential mechanism for
the valvulopathic side effects of serotonergic medications at the cellular and molecular
levels. At the cellular level, the data presented in Chapter 3 show that valvular interstitial
cells in primary culture exhibit mitogenic responses to VHD-associated drugs that are
consistent with the in vivo effects of the drugs (Chapter 3). That is, damaged valves
resected from patients undergoing VHD-associated drug therapy are characterized by
proliferative interstitial cell plaques that compromise valve function (Connolly, Crary et
al. 1997; Steffee, Singh et al. 1999). At the molecular level, the data presented in Chapter
3 demonstrate that the mitogenic responses induced in primary cultures of heart valve
interstitial cells by VHD-associated drugs require 5-HT2B receptor activation. The
aforementioned observations are the most conclusive evidence to date supporting the 5-
HT2B receptor hypothesis for drug-induced VHD. The data presented in Chapter 4
provide a higher resolution picture of anorexigen binding to valvulopathic 5-HT2B receptors, implicating van der Waals’ interactions between the 5-HT2B receptor and the α- methyl group of the active fenfluramine metabolite S-(+)-norfenfluramine as contributing to its valvulopathic, but not anorectic, effects. The findings presented in this thesis, therefore, suggest some avenues of investigation that could reduce the incidence of drug- induced VHD and also lead to the design of safer anorectic agents.
142 5.2 High-Throughput Screening Efforts to Identify Potential Valvulopathogens
Among Current and Future Pharmacotherapies
Using primary cultures of VICs—a novel in vitro system—the data presented in
Chapter 3 show that valvulopathic drugs induce mitogenic responses in the cells affected
in VHD. Such an observation is compelling evidence in favor of the “5-HT2B receptor hypothesis of VHD.” Nevertheless, VICs in primary culture grow slowly and senesce after eight or nine passages. Thus, these cells are inconvenient for high-throughput screening efforts to identify potential valvulopathogens. This difficulty is ameliorated by the demonstration that agonist activity, as measured by second messenger (i.e., inositol phosphate) accumulation in HEK-293 cells expressing cloned 5-HT2B receptors, is a
reliable predictor of a drug’s ability to induce heart valve interstitial cell mitosis in vitro
and in vivo (Chapter 3). Thus, the data from Chapter 3 suggest that individual drugs—as
well as chemical classes—with fenfluramine-like mitogenic activity on heart valve
interstitial cells can be identified from large compound libraries via measurements of
second messenger accumulation in cell lines expressing cloned 5-HT2B receptors.
Because cell lines and recombinant DNA techniques are widely available, many research groups already have in-hand the technology required for such large-scale screening efforts. Indeed, we have been informed that 5-HT2B receptor screening is now routinely
used during pre-clinical drug development. The results of such screening efforts will
facilitate the identification of currently used drugs that might induce VHD, as well as
“VHD high risk” chemical structural classes.
143 For instance, several groups have shown both fenfluramine (Chapter 3) and one of
its major, active metabolites, norfenfluramine (Porter, Benwell et al. 1999; Fitzgerald,
Burn et al. 2000; Rothman, Baumann et al. 2000), to be 5-HT2B receptor agonists. The
data presented in Chapter 3 demonstrate that both 3,4-methylenedioxymethamphetamine
(MDMA, “Ecstasy”) and one of its major, active metabolites, 3,4-
methylenedioxyamphetamine (MDA), are 5-HT2B receptor agonists. Porter et al. (1999)
also reported the 5-HT2B agonist activity of 2,5-dimethoxy-4-iodoamphetamine (DOI)
and 2,5-dimethoxy-4-bromoamphetamine (DOB). All of these compounds belong to the
phenylisopropylamine structural class (e.g., amphetamine derivatives). Nelson and
colleagues (Nelson, Lucaites et al. 1999) examined the affinities of several
hallucinogenic 2,5-dimethoxy-phenylisopropylamines at human 5-HT2B receptors. Of the
17 compounds tested in the aforementioned study, 15 exhibited sub-micromolar 5-HT2B receptor affinities; seven of these displayed Ki values between 20 nM and 50 nM. Thus,
there is ample evidence in the literature suggesting that phenylisopropylamines as a
structural class are likely to bind to and activate 5-HT2B receptors. Combined with the
findings presented in Chapter 3, these observations suggest that phenylisoproylamines are
likely to induce VHD in humans and, as such, are one “VHD high risk” structural class.
The data presented in the preceding chapters, as well as others’ published data
(see below), identify ergolines as another structural class likely to bind to and activate 5-
HT2B receptors. For instance, the data in Chapters 3 and 4 demonstrate that
dihydroergotamine and pergolide have sub-micromolar 5-HT2B receptor affinities and agonist potencies. Both dihydroergotamine and pergolide have been linked to VHD
144 (Creutzig 1992; Pritchett, Morrison et al. 2002). Ergotamine and methysergide, both of
which are associated with VHD (Bana, MacNeal et al. 1974; Hauck, Edwards et al. 1990;
Hendrikx, Van Dorpe et al. 1996), were shown by Rothman and colleagues (Rothman,
Baumann et al. 2000) to exhibit high affinity and potent agonist activity at 5-HT2B receptors. In addition, both bromocriptine and cabergoline have been shown to bind to 5-
HT2B receptors with sub-micromolar affinity (Millan, Maiofiss et al. 2002). Both bromocriptine and cabergoline induce VHD (Serratrice, Disdier et al. 2002; Horvath,
Fross et al. 2004). Thus, several clinical reports establish an association between ergoline therapy and VHD; in vitro studies demonstrate that the VHD-associated ergolines bind to and/or activate 5-HT2B receptors. Taken together, the aforementioned findings and the
data presented in Chapters 3 and 4 suggest that ergolines are likely to exhibit high affinity
and potent agonist activity at 5-HT2B receptors and, as such, are a “VHD high risk”
structural class.
In light of the findings described in Chapter 3, efforts to identify current
pharmaceuticals that bind to and/or activate cloned 5-HT2B receptors are underway. The
Prestwick Chemical library is a structurally diverse set of 880 small molecules, more than
85% of which are currently-marketed drugs. The results of the screening effort have have identified many compounds that bind to and/or activate 5-HT2B receptors. A subset of the
5-HT2B receptor ligands in the Prestwick Library are ergolines. Of the ten ergolines in the
library (lisuride, nicergoline, dihydroergotamine, pergolide, metergoline,
methylergometrine, lysergol, dihydroergocristine, dihydroergotoxine, and ergocryptine-
α), eight exhibit 5-HT2B receptor affinity (Ki < 10 µM) (only metergoline and
145 methylergometrine did not). The preliminary functional screens, which are currently
~50% complete, have thus far covered all of the ergolines in the library except for
lysergol; only two (lisuride and nicergoline) did not exhibit agonist activity. Thus, data
from an on-going screen of the Prestwick Chemical Library support the prediction that
ergolines as a class tend to bind to and activate 5-HT2B receptors and, as such, represent
“VHD high risk” compounds.
In summary, the data presented in the previous chapters strongly underscore the
need to assess the 5-HT2B receptor agonist activity of those compounds that are currently
or destined to be used by humans. The most physiologically relevant way to do so, absent
an established model system for drug-induced VHD, would be to expose primary cultures
of human heart valve interstitial cells (i.e., the cells affected in VHD) to the compounds
under study and identify those that induce mitogenesis (i.e., the response that gives rise to
VHD lesions). However, because of their limited availability, human heart valve interstitial cells are an inconvenient system. The data presented in Chapter 3 show that measuring second messenger accumulation in cell lines expressing recombinant human 5-
HT2B receptors is a reliable and convenient alternative. Large-scale screening efforts
using 5-HT2B receptor-expressing cell lines can identify not only single compounds that
are likely to cause VHD in humans, but also structural classes that tend to activate 5-
HT2B receptors and, as such, should be avoided in the design and synthesis of novel
pharmaceuticals. Such efforts would likely reduce future occurrences of drug-induced
VHD.
146 5.3 Design of “Second Generation” Phenylisopropylamine Anorexigens
Having established a link between VHD-associated drugs, heart valve interstitial
cells, 5-HT2B receptors, and pathogenic mitogenic responses in Chapter 3, the work
described in Chapter 4 attempted a higher-resolution understanding of the interactions
between valvulopathogens and the 5-HT2B receptor. To do so, the fenfluramine
metabolite S-(+)-norfenfluramine, which is a more potent 5-HT2B receptor agonist than
fenfluramine (Chapter 3; (Rothman, Baumann et al. 2000), was chosen as a ligand for
three reasons: 1) the metabolite, via its potent activation of 5-HT2B receptors, has been
linked to the fibroproliferative lesions that underlie VHD and PH; 2) the compound is a very effective appetite suppressant, likely due in large part to its potent activation of hypothalamic 5-HT2C receptors; and 3) the drug has low affinity and potency at 5-HT2A receptors. Thus, given the high degree of sequence homology among 5-HT2 family receptors, it was reasoned that mutating each non-conserved, putative ligand binding residue in the 5-HT2B receptor to its 5-HT2C receptor and/or 5-HT2A receptor analog
would reveal 5-HT2B receptor residues important for the high-affinity binding of S-(+)-
norfenfluramine that do not contribute significantly to the ligand’s affinity at 5-HT2C and/or 5-HT2A receptors. The approach led to the identification of valine 2.53 as
contributing to the rank order affinity of SNF at 5-HT2 receptors (i.e., 5-HT2B > 5-HT2C >
5-HT2A). Mutation of V2.53 to leucine (a 5-HT2B/2C-to-5-HT2A mutation) in the 5-HT2B receptor caused a 17-fold decrease in ligand affinity, but only a nine-fold decrease in
SNF affinity at the 5-HT2C receptor; the inverse L2.53V mutation in the 5-HT2A receptor had no effect on SNF affinity. Through molecular modeling, ligand docking simulations,
147 and molecular dynamics simulations, a plausible model to explain the results of mutant 5-
HT2B receptor SNF binding studies was obtained: both γ-methyl groups of V2.53 form
stabilizing van der Waals’ interactions with the α-methyl group of bound SNF. Such a
model, then, would predict that a V2.53A mutation in the 5-HT2B receptor should
diminish SNF affinity more than does the V2.53L mutation, since the V2.53A mutation
would eliminate two potential van der Waals’ interactions and the V2.53L mutation
would eliminate only one.
To directly test the effect of eliminating both putative 2.53 γ-methyl group-SNF
α-methyl group van der Waals’ interactions, the affinity of SNF at the V2.53A 5-HT2B receptor was measured. As shown in Chapter 4, the V2.53A mutation decreased SNF binding 150-fold compared to the compound’s binding at the wild type 5-HT2B receptor
(i.e., 3,300 nM vs 22 nM). Given a range of +0.4 kcal/mol to +1.4 kcal/mol for methyl
group-methyl group van der Waals’ interaction energies, the decrease in SNF binding energy (∆∆G) due to the V2.53A mutation (+3.0 ± 0.2 kcal/mol) is consistent with the
loss of two van der Waals’ interactions due to the mutation. The effect of the V2.53A
mutation suggests that the V2.53L mutation does not impair SNF binding due to steric
hindrance, since replacing V2.53 with both a smaller (alanine) and larger (leucine) hydrophobic amino acid impaired ligand binding. However, replacing V2.53 with the larger hydrophobic amino acid isoleucine did not affect ligand affinity. Molecular dynamics simulations predicted that I2.53, like V2.53, can form two van der Waals’
interactions with the α-methyl group of SNF. Thus, the experimental data reinforce the
148 model predicting bivalent hydrophobic interactions between the terminal γ-methyl groups
of V2.53 and the α-methyl group of SNF.
The predictions of the model were subjected to further empirical testing using NF
analogs bearing α-carbon “mutations.” Dr. Richard Glennon and Dr. Malgorzata Dukat
(Virginia Commonwealth University) prepared α-desmethyl-NF and α-ethyl-NF. Based
on the model suggesting that V2.53-SNF α-methyl group interactions are important for
SNF binding, α-desmethyl-NF would be predicted to exhibit a decreased affinity
compared to SNF at wild type 5-HT2B receptors. In line with the predictions of the model,
α-desmethyl-NF displayed a 3-fold decreased affinity for 5-HT2B receptors compared to
NF (55 nM vs. 17 nM). The R-(-) enantiomer of NF (RNF) exhibits an affinity similar to
that of α-desmethyl-NF at 5-HT2B receptors (62 nM). Assuming RNF and α-desmethyl-
NF bind to 5-HT2B receptors in the same orientation, the preceding observations suggest
that the R-(-) α-methyl group does not engage in productive receptor-ligand interactions.
At the 5-HT2C receptor, SNF affinity was not significantly different than α-desmethyl-NF
affinity (69 nM vs. 87 nM, respectively); the same was true of SNF and α-desmethyl-NF
affinities at the 5-HT2A receptor (1,800 nM and 1,300 nM, respectively). Thus, the α- desmethyl-NF-wild type receptor binding data agree with the SNF-receptor mutant data
(Chapter 4) demonstrating that V2.53-ligand interactions contribute more to 5-HT2B receptors binding than to 5-HT2C and 5-HT2A receptor binding.
Since the molecular modeling and SNF-V2.53A 5-HT2B receptor binding data
predict that removal of the α-methyl group of SNF should lead to a ~150-fold decrease in
149 affinity, the α-desmethyl-NF affinity constant data are, in terms of the magnitude of the
predicted change, in discord with the model. The discrepancy may indicate that the
underlying assumption, that α-desmethyl-NF binds to the 5-HT2B receptor in the same
orientation as does SNF, may not be true. However, that α-desmethyl-NF would assume a
grossly different orientation than SNF in the 5-HT2B receptor binding pocket is
improbable, given the likelihood that canonical TM 6-aromatic and TM3-cationic
receptor-ligand interactions remain intact. Thus, the ligand “mutant” binding data and
molecular modeling suggest that α-desmethyl-NF might adopt a subtly different binding
orientation that energetically compensates for the loss of 5-HT2B receptor-ligand α- methyl group interactions. For instance, in the absence of the S-(+) α-methyl group, the ligand might be able to bind slightly higher so as to form stronger interactions with TM6 and TM3 than are possible in the presence of an S-(+) α-methyl group. Therefore, removal of the α-methyl group from SNF, which leads to a modest 3-fold decrease in the ligand’s affinity only at 5-HT2B receptors, will not yield an inert 5-HT2B receptor ligand
without further modifications. Through additional molecular modeling and molecular
dynamics simulations, it may be possible to identify receptor-α-desmethyl-NF
interactions that explain why α-desmethyl-NF affinity at 5-HT2B receptors is only slightly
decreased compared to that of SNF. Further ligand modifications to disrupt or prevent
those interactions may be possible, and could lead to an even more selective 5-HT2C receptor agonist.
In light of the model and the data suggesting that V2.53 in the 5-HT2B receptor is
within 4 Å of the α-methyl group of SNF, an NF analog bearing a bulkier hydrophobic α-
150 carbon substituent should be more poorly tolerated due to steric hindrance. Therefore, Dr.
Richard Glennon and Dr. Malgorzata Dukat (Virginia Commonwealth University) synthesized and furnished α-ethyl-NF for testing. The affinity of α-ethyl-NF at the wild
type 5-HT2B receptor displayed a 46-fold decrease compared to SNF (780 nM vs. 17 nM).
At wild type 5-HT2C receptors, α-ethyl-NF exhibited only a 14-fold decrease in affinity
(1,000 nM vs. 69 nM); at wild type 5-HT2A receptors, the affinity of the compound was reduced only 6-fold (10,000 nM vs. 1800 nM). Thus, as was true for removing the α- methyl group from SNF, replacing the α-methyl group of NF with a bulkier ethyl group selectively perturbed binding to 5-HT2B receptors (46-fold compared to SNF); binding to
5-HT2C and 5-HT2A receptors was much less perturbed (14-fold and 6-fold, respectively,
compared to SNF). The preceding data suggest, then, that perturbing 2.53-NF analog
interactions with bulky hydrophobic α-carbon substituents might lead to more dramatic,
5-HT2B receptor-selective changes in NF analog affinity than would removal of the α- methyl group from SNF.
In Chapter 4, evidence is presented that residue V2.53 in the 5-HT2B receptor plays a role in NF-induced activation that is unique among the 5-HT2 receptors.
Specifically, the V2.53L mutation causes a 4-fold decrease in SNF potency at the 5-HT2B receptor. Residue 2.53 in the 5-HT2C receptor is also valine; however, the V2.53L
mutation in the 5-HT2C receptor, which decreases SNF affinity 9-fold, has no effect on
agonist potency. Furthermore, the relative efficacy (Emax) of SNF was increased 35% by
the V2.53L mutation in the 5-HT2B receptor; the V2.53L mutation had no effect on SNF
Emax at the 5-HT2C receptor. As for RNF, the V2.53L mutation in the 5-HT2B receptor,
151 while increasing potency about 2-fold, reduces Emax a dramatic 65%. The V2.53L
mutation in the 5-HT2C receptor increases RNF potency two-fold, as it does for the 5-
HT2B receptor; agonist efficacy at the 5-HT2C receptor is unaffected by the mutation.
Thus, the V2.53L mutation only alters SNF potency and efficacy at 5-HT2B receptors and
not 5-HT2C receptors; RNF efficacy is also only affected by the mutation in 5-HT2B receptors and not in 5-HT2C receptors. The agonist activity of the two NF analogs at 5-
HT2B and 5-HT2C receptors remains to be determined. However, in terms of binding
affinity, α-desmethyl-NF behaves like RNF; as such, it is conceivable that the two
compounds might behave similarly in agonist functional assays. If so, the modest 3-fold
decrease observed in α-desmethyl-NF affinity relative to SNF at 5-HT2B receptors
coupled with a possible 60% to 70% decrease in Emax would render α-desmethyl-NF a
weaker agonist than SNF at 5-HT2B receptors—one that potentially retains SNF-like
agonist properties at 5-HT2C receptors. Such a result would imply that α-desmethyl-NF would be an equally efficacious anorexigen as SNF, but safer in terms of VHD risk. As for α-ethyl-NF, like SNF, it may exhibit, despite its decreased affinity for 5-HT2C receptors compared to SNF, agonist character at 5-HT2C receptors similar to that of SNF,
while displaying reduced potency relative to SNF at 5-HT2B receptors. Thus, in future
experiments, it will be important to ascertain the 5-HT2C receptor agonist activity in
functional assays of the NF analogs described herein.
In summary, the data discussed in Chapter 4 suggest that it may be possible to
chemically modify SNF so as to render it inert at 5-HT2B receptors while retaining 5-
HT2C receptor agonist character. The computational simulations, ligand binding and
152 functional data collectively suggest that SNF analogs designed, based on the model
presented in Chapter 4, to eliminate (e.g., α-desmethyl-NF) or perturb (e.g., α-ethyl-NF)
5-HT2B receptor-α-carbon substituent interactions are likely to exhibit 5-HT2B receptor- selective reductions in binding, potency, and efficacy. In addition, the results presented in
Chapter 4 underscore the utility of molecular modeling and molecular dynamics to interpret the results of, and guide further mutagenesis experiments. The combined computational and experimental approaches led to a better understanding of the molecular determinants of NF and NF analog binding to wild type and mutant 5-HT2B receptors. Thus, the identification, via a combination of computational and empirical methods, of 5-HT2B receptor-unique 2.53-SNF α-methyl group interactions bodes well
for the feasibility of discovering additional determinants of SNF’s 5-HT2 receptor sub-
type-selective pharmacology. Such studies could hasten the development of “second
generation” phenylisopropylamine anorexigens devoid of VHD-inducing potential.
5.4 Conclusion
In conclusion, the studies that are the subject of this thesis resulted in the
following novel findings:
1) Activation of 5-HT2B receptors on human heart valve interstitial cells in primary
culture leads to a mitogenic response. This demonstrates—for the first time—that
fenfluramine and norfenfluramine, both of which are associated with proliferative
heart valve lesions in humans, induces heart valve interstitial cell proliferation via
153 activation of 5-HT2B receptors. This finding is the strongest evidence to date
linking 5-HT2B receptor activation to drug induced VHD. Additionally, the data
presented herein suggest that, in cases where the benefits of 5-HT2B receptor
agonist therapy might outweigh the cardiopulmonary risks (e.g., treatment of
Parkinson’s patients with pergolide), 5-HT2B receptor-selective antagonists might
be a prophylactic adjunct.
2) Activation of recombinant human 5-HT2B receptors ectopically expressed in cell
lines normally devoid of 5-HT2B receptors by a drug is a reliable predictor of the
drug’s likelihood to induce human heart valve cell proliferation—the
pathophysiological response that gives rise to VHD lesions. For example, data are
presented in this thesis that demonstrate that both MDMA and MDA activate
recombinant 5-HT2B receptors expressed in HEK-293 cells and induce mitosis in
human heart valve interstitial cells in vitro. Whether MDMA users are at
increased risk for VHD is not yet known. Also, data presented in this thesis
demonstrate that the VHD-associated anti-Parkinsonian medication pergolide
activates recombinant 5-HT2B receptors expressed in HEK-293 cells. Therefore,
by measuring drug-induced 5-HT2B receptor activation in cell lines expressing
recombinant receptors, heart valve interstitial cell mitogens can be identified;
because such compounds are likely to induce VHD in humans, they could then be
excluded from further use or development. Since recombinant systems are readily
available to most laboratories, whereas primary cultures of human heart valve
154 interstitial cells are not, the data presented herein, which have been published, are
expected to facilitate screening efforts to identify currently used and
investigational pharmacotherapies for the potential to induce VHD. As such, the
results presented in this thesis have significant public health implications.
3) The α-methyl group of SNF contributes significantly to its valvulopathic, but not
its anorexigenic, actions. Specifically, the data presented in this thesis provide
convincing evidence that stabilizing hydrophobic interactions between V2.53 of
the 5-HT2B receptor and the α-methyl group of SNF explain why SNF is a better
agonist at 5-HT2B receptors that at 5-HT2C and 5-HT2A receptors. Consistent with
this, NF analogs bearing α-carbon “mutations” exhibit impairmed binding to 5-
HT2B receptors with comparatively little or no change in 5-HT2C and 5-HT2A
receptor affinities. Thus, the data presented herein suggest that α-carbon
modifications to SNF might result in a 5-HT2C receptor agonist devoid of 5-HT2B
receptor agonist activity. Such a compound is expected to be a safe and effective
appetite suppressant, which would be of great clinical utility.
The work described in the preceding chapters is the beginning of a story that will, it is hoped, inspire future studies aimed at reducing the exposure of humans to 5-HT2B receptor agonists and designing 5-HT2C receptor agonists that do not activate 5-HT2B receptors. The result of such studies would very likely reduce the incidence of VHD and
PH due to drug use and lead to safer, effective appetite-suppressants. It is the sincerest
155 hope of this thesis’ author that the work described herein will in some small way contribute to the attainment of these outcomes.
156 BIBLIOGRAPHY
(1994). "Dexfenfluramine influences dietary compliance and eating behaviors of obese subjects." Nutr Rev 52(2 Pt 1): 65-8. (1997). "Cardiac valvulopathy associated with exposure to fenfluramine or dexfenfluramine: U.S. Department of Health and Human Services interim public health recommendations, November 1997." MMWR Morb Mortal Wkly Rep 46(45): 1061-6. Abenhaim, L., Y. Moride, et al. (1996). "Appetite-suppressant drugs and the risk of primary pulmonary hypertension. International Primary Pulmonary Hypertension Study Group." N Engl J Med 335(9): 609-16. Alberts, G. L., C. L. Chio, et al. (2003). "A unique phenotype of 5-HT2C, agonist- induced GTPgamma35S binding, transferable to 5-HT2A and 5-HT2B, upon swapping intracellular regions." Br J Pharmacol 138(3): 427-34. Almaula, N., B. J. Ebersole, et al. (1996). "Mapping the binding site pocket of the serotonin 5-Hydroxytryptamine2A receptor. Ser3.36(159) provides a second interaction site for the protonated amine of serotonin but not of lysergic acid diethylamide or bufotenin." J Biol Chem 271(25): 14672-5. Andersson, B., M. E. Zimmermann, et al. (1991). "Haemodynamic, metabolic and endocrine effects of short-term dexfenfluramine treatment in young, obese women." Eur J Clin Pharmacol 40(3): 249-54. Archer, E., B. Maigret, et al. (2003). "Rhodopsin crystal: new template yielding realistic models of G-protein-coupled receptors?" Trends Pharmacol Sci 24(1): 36-40. Artigas, F., M. J. Sarrias, et al. (1989). "Increased plasma free serotonin but unchanged platelet serotonin in bipolar patients treated chronically with lithium." Psychopharmacology (Berl) 99(3): 328-32. Baldwin, J. M. (1993). "The probable arrangement of the helices in G protein-coupled receptors." Embo J 12(4): 1693-703. Baldwin, J. M. (1994). "Structure and function of receptors coupled to G proteins." Curr Opin Cell Biol 6(2): 180-90. Baldwin, J. M., G. F. Schertler, et al. (1997). "An alpha-carbon template for the transmembrane helices in the rhodopsin family of G-protein-coupled receptors." J Mol Biol 272(1): 144-64. Ballesteros, J. A., L. Shi, et al. (2001). "Structural mimicry in G protein-coupled receptors: implications of the high-resolution structure of rhodopsin for structure- function analysis of rhodopsin-like receptors." Mol Pharmacol 60(1): 1-19. Ballesteros, J. A. and H. Weinstein (1994). "Integrated methods for the construction of three dimensional models and computational probing of structure-function relations in G-protein coupled receptors." Methods Neurosci 25: 366-428. Bana, D. S., P. S. MacNeal, et al. (1974). "Cardiac murmurs and endocardial fibrosis associated with methysergide therapy." Am Heart J 88(5): 640-55. Belohlavkova, S., J. Simak, et al. (2001). "Fenfluramine-induced pulmonary vasoconstriction: role of serotonin receptors and potassium channels." J Appl Physiol 91(2): 755-61.
157 Berg, K. A., S. Maayani, et al. (1996). "5-hydroxytryptamine2C receptor activation inhibits 5-hydroxytryptamine1B-like receptor function via arachidonic acid metabolism." Mol Pharmacol 50(4): 1017-23. Berg, K. A., S. Maayani, et al. (1998). "Pleiotropic behavior of 5-HT2A and 5-HT2C receptor agonists." Ann N Y Acad Sci 861: 104-10. Berg, K. A., S. Maayani, et al. (1998). "Effector pathway-dependent relative efficacy at serotonin type 2A and 2C receptors: evidence for agonist-directed trafficking of receptor stimulus." Mol Pharmacol 54(1): 94-104. Bhasin, N., E. Kernick, et al. (2004). "Differential regulation of chondrogenic differentiation by the serotonin2B receptor and retinoic acid in the embryonic mouse hindlimb." Dev Dyn 230(2): 201-9. Bigler, T. L., W. Lu, et al. (1993). "Binding of amino acid side chains to preformed cavities: interaction of serine proteinases with turkey ovomucoid third domains with coded and noncoded P1 residues." Protein Sci 2(5): 786-99. Blanpain, C., E. Le Poul, et al. (2003). "Serotonin 5-HT(2B) receptor loss of function mutation in a patient with fenfluramine-associated primary pulmonary hypertension." Cardiovasc Res 60(3): 518-28. Bliziotes, M. M., A. J. Eshleman, et al. (2001). "Neurotransmitter action in osteoblasts: expression of a functional system for serotonin receptor activation and reuptake." Bone 29(5): 477-86. Blouin, A. G., J. H. Blouin, et al. (1988). "Treatment of bulimia with fenfluramine and desipramine." J Clin Psychopharmacol 8(4): 261-9. Bonaventure, P., H. Guo, et al. (2002). "Nuclei and subnuclei gene expression profiling in mammalian brain." Brain Res 943(1): 38-47. Bonhaus, D. W., C. Bach, et al. (1995). "The pharmacology and distribution of human 5- hydroxytryptamine2B (5-HT2B) receptor gene products: comparison with 5- HT2A and 5-HT2C receptors." Br J Pharmacol 115(4): 622-8. Bonhaus, D. W., L. A. Flippin, et al. (1999). "RS-127445: a selective, high affinity, orally bioavailable 5-HT2B receptor antagonist." Br J Pharmacol 127(5): 1075-82. Borman, R. A., N. S. Tilford, et al. (2002). "5-HT(2B) receptors play a key role in mediating the excitatory effects of 5-HT in human colon in vitro." Br J Pharmacol 135(5): 1144-51. Bradley, P. B. (1984). "The characterization of 5-HT receptors. Introduction." Neuropharmacology 23(12B): 1465-6. Brenot, F., P. Herve, et al. (1993). "Primary pulmonary hypertension and fenfluramine use." Br Heart J 70(6): 537-41. Breum, L., S. E. Moller, et al. (1996). "Long-term effect of dexfenfluramine on amino acid profiles and food selection in obese patients during weight loss." Int J Obes Relat Metab Disord 20(2): 147-53. Burger, A. J., H. B. Sherman, et al. (1999). "Low prevalence of valvular heart disease in 226 phentermine-fenfluramine protocol subjects prospectively followed for up to 30 months." J Am Coll Cardiol 34(4): 1153-8. Bush, E., J. Fielitz, et al. (2004). "A small molecular activator of cardiac hypertrophy uncovered in a chemical screen for modifiers of the calcineurin signaling pathway." Proc Natl Acad Sci U S A 101(9): 2870-5.
158 Bymaster, F. P., D. L. Nelson, et al. (1999). "Antagonism by olanzapine of dopamine D1, serotonin2, muscarinic, histamine H1 and alpha 1-adrenergic receptors in vitro." Schizophr Res 37(1): 107-22. Canton, H., L. Verriele, et al. (1994). "Competitive antagonism of serotonin (5-HT)2C and 5-HT2A receptor-mediated phosphoinositide (PI) turnover by clozapine in the rat: a comparison to other antipsychotics." Neurosci Lett 181(1-2): 65-8. Carboni, E. and G. Di Chiara (1989). "Serotonin release estimated by transcortical dialysis in freely-moving rats." Neuroscience 32(3): 637-45. Casey, P. J. and A. G. Gilman (1988). "G protein involvement in receptor-effector coupling." J Biol Chem 263(6): 2577-80. Catt, K. J. and T. Balla (1989). "Phosphoinositide metabolism and hormone action." Annu Rev Med 40: 487-509. Celada, P., F. Martin, et al. (1994). "Effects of chronic treatment with dexfenfluramine on serotonin in rat blood, brain and lung tissue." Life Sci 55(15): 1237-43. Celada, P., J. Perez, et al. (1992). "Monoamine oxidase inhibitors phenelzine and brofaromine increase plasma serotonin and decrease 5-hydroxyindoleacetic acid in patients with major depression: relationship to clinical improvement." J Clin Psychopharmacol 12(5): 309-15. Celada, P., M. J. Sarrias, et al. (1990). "Serotonin and 5-hydroxyindoleacetic acid in plasma. Potential use as peripheral measures of MAO-A activity." J Neural Transm Suppl 32: 149-54. Chidlow, G., P. S. Hiscott, et al. (2004). "Expression of serotonin receptor mRNAs in human ciliary body: a polymerase chain reaction study." Graefes Arch Clin Exp Ophthalmol 242(3): 259-64. Choi, D. S. and L. Maroteaux (1996). "Immunohistochemical localisation of the serotonin 5-HT2B receptor in mouse gut, cardiovascular system, and brain." FEBS Lett 391(1-2): 45-51. Choi, D. S., S. J. Ward, et al. (1997). "5-HT2B receptor-mediated serotonin morphogenetic functions in mouse cranial neural crest and myocardiac cells." Development 124(9): 1745-55. Choudhary, M. S., S. Craigo, et al. (1993). "A single point mutation (Phe340-->Leu340) of a conserved phenylalanine abolishes 4-[125I]iodo-(2,5- dimethoxy)phenylisopropylamine and [3H]mesulergine but not [3H]ketanserin binding to 5-hydroxytryptamine2 receptors." Mol Pharmacol 43(5): 755-61. Choudhary, M. S., N. Sachs, et al. (1995). "Differential ergoline and ergopeptine binding to 5-hydroxytryptamine2A receptors: ergolines require an aromatic residue at position 340 for high affinity binding." Mol Pharmacol 47(3): 450-7. Chow, C. C., G. T. Ko, et al. (1997). "Dexfenfluramine in obese Chinese NIDDM patients. A placebo-controlled investigation of the effects on body weight, glycemic control, and cardiovascular risk factors." Diabetes Care 20(7): 1122-7. Connolly, H. M., J. L. Crary, et al. (1997). "Valvular heart disease associated with fenfluramine-phentermine." N Engl J Med 337(9): 581-8. Cotecchia, S., D. A. Schwinn, et al. (1988). "Molecular cloning and expression of the cDNA for the hamster alpha 1-adrenergic receptor." Proc Natl Acad Sci U S A 85(19): 7159-63.
159 Cox, D. A. and M. L. Cohen (1995). "5-Hydroxytryptamine2B receptor signaling in rat stomach fundus: role of voltage-dependent calcium channels, intracellular calcium release and protein kinase C." J Pharmacol Exp Ther 272(1): 143-50. Cox, D. A. and M. L. Cohen (1996). "5-HT2B receptor signaling in the rat stomach fundus: dependence on calcium influx, calcium release and protein kinase C." Behav Brain Res 73(1-2): 289-92. Creutzig, A. (1992). "[Ergotamine-induced heart valve fibrosis and coronary microangiopathy?]." Dtsch Med Wochenschr 117(45): 1736. Cussac, D., A. Newman-Tancredi, et al. (2002). "Characterization of phospholipase C activity at h5-HT2C compared with h5-HT2B receptors: influence of novel ligands upon membrane-bound levels of [3H]phosphatidylinositols." Naunyn Schmiedebergs Arch Pharmacol 365(3): 242-52. de la Torre, R., M. Farre, et al. (2000). "Non-linear pharmacokinetics of MDMA ('ecstasy') in humans." Br J Clin Pharmacol 49(2): 104-9. Deraet, M., P. Manivet, et al. (2005). "The natural mutation encoding a C terminus- truncated 5-hydroxytryptamine 2B receptor is a gain of proliferative functions." Mol Pharmacol 67(4): 983-91. Diener, H. C. (1994). "A review of current treatments for migraine." Eur Neurol 34 Suppl 2: 18-25. Ditschuneit, H. H., M. Flechtner-Mors, et al. (1996). "The effect of withdrawal and re- introduction of dexfenfluramine in the treatment of obesity." Int J Obes Relat Metab Disord 20(3): 280-2. Ditschuneit, H. H., M. Flechtner-Mors, et al. (1996). "The effects of dexfenfluramine on weight loss and cardiovascular risk factors in female patients with upper and lower body obesity." J Cardiovasc Risk 3(4): 397-403. Ditschuneit, H. H., M. Flechtner-Mors, et al. (1993). "Endocrine and metabolic effects of dexfenfluramine in patients with android obesity." Horm Metab Res 25(11): 573- 8. Dixon, R. A., B. K. Kobilka, et al. (1986). "Cloning of the gene and cDNA for mammalian beta-adrenergic receptor and homology with rhodopsin." Nature 321(6065): 75-9. Dumuis, A., M. Sebben, et al. (1989). "The gastrointestinal prokinetic benzamide derivatives are agonists at the non-classical 5-HT receptor (5-HT4) positively coupled to adenylate cyclase in neurons." Naunyn Schmiedebergs Arch Pharmacol 340(4): 403-10. Duxon, M. S., T. P. Flanigan, et al. (1997). "Evidence for expression of the 5- hydroxytryptamine-2B receptor protein in the rat central nervous system." Neuroscience 76(2): 323-9. Elphick, G. F., W. Querbes, et al. (2004). "The Human Polyomavirus, JCV, Uses Serotonin Receptors to Infect Cells." Science 306(5700): 1380-1383. Enguix, M. J., L. Sanchez, et al. (2003). "Differential regulation of rat peripheral 5- HT(2A) and 5-HT(2B) receptor systems: influence of drug treatment." Naunyn Schmiedebergs Arch Pharmacol 368(2): 79-90. Faergeman, N. J., B. W. Sigurskjold, et al. (1996). "Thermodynamics of ligand binding to acyl-coenzyme A binding protein studied by titration calorimetry." Biochemistry 35(45): 14118-26.
160 Finer, N. (1992). "Body weight evolution during dexfenfluramine treatment after initial weight control." Int J Obes Relat Metab Disord 16 Suppl 3: S25-9. Finer, N., D. Craddock, et al. (1988). "Effect of 6 months therapy with dexfenfluramine in obese patients: studies in the United Kingdom." Clin Neuropharmacol 11 Suppl 1: S179-86. Fiorica-Howells, E., L. Maroteaux, et al. (2000). "Serotonin and the 5-HT(2B) receptor in the development of enteric neurons." J Neurosci 20(1): 294-305. Fishman, A. P. (1999). "Aminorex to fen/phen: an epidemic foretold." Circulation 99(1): 156-61. Fitzgerald, L. W., T. C. Burn, et al. (2000). "Possible role of valvular serotonin 5-HT(2B) receptors in the cardiopathy associated with fenfluramine." Mol Pharmacol 57(1): 75-81. Fleckenstein, A. E., J. W. Gibb, et al. (2000). "Differential effects of stimulants on monoaminergic transporters: Pharmacological consequences and implications for neurotoxicity." European Journal of Pharmacology 406(1): 1-13. Foguet, M., D. Hoyer, et al. (1992). "Cloning and functional characterization of the rat stomach fundus serotonin receptor." Embo J 11(9): 3481-7. Fontaine, R. (1993). "Novel serotonergic mechanisms and clinical experience with nefazodone." Clin Neuropharmacol 16 Suppl 3: S45-50. Frielle, T., S. Collins, et al. (1987). "Cloning of the cDNA for the human beta 1- adrenergic receptor." Proc Natl Acad Sci U S A 84(22): 7920-4. Fukuda, K., T. Kubo, et al. (1987). "Molecular distinction between muscarinic acetylcholine receptor subtypes." Nature 327(6123): 623-5. Fuxe, K., L. O. Farnebo, et al. (1975). "On the in vivo and in vitro actions of fenfluramine and its derivatives on central monoamine neurons, especially 5- hydroxytryptamine neurons, and their relation to the anorectic activity of fenfluramine." Postgrad Med J 51 Suppl 1: 35-45. Garattini, S., W. Buczko, et al. (1975). "The mechanism of action of fenfluramine." Postgrad Med J 51 Suppl 1: 27-35. Gardin, J. M., D. Schumacher, et al. (2000). "Valvular abnormalities and cardiovascular status following exposure to dexfenfluramine or phentermine/fenfluramine." Jama 283(13): 1703-9. Garnovskaya, M. N., C. G. Nebigil, et al. (1995). "5-Hydroxytryptamine2A receptors expressed in rat renal mesangial cells inhibit cyclic AMP accumulation." Mol Pharmacol 48(2): 230-7. Gibson, E. L., A. J. Kennedy, et al. (1993). "d-Fenfluramine- and d-norfenfluramine- induced hypophagia: differential mechanisms and involvement of postsynaptic 5- HT receptors." Eur J Pharmacol 242(1): 83-90. Gilman, A. G. (1987). "G proteins: transducers of receptor-generated signals." Annu Rev Biochem 56: 615-49. Glennon, R. A., M. Titeler, et al. (1984). "Evidence for 5-HT2 involvement in the mechanism of action of hallucinogenic agents." Life Sci 35(25): 2505-11. Goni, F. M. and A. Alonso (1999). "Structure and functional properties of diacylglycerols in membranes." Prog Lipid Res 38(1): 1-48.
161 Gray, J. A., B. A. Compton-Toth, et al. (2003). "Identification of two serine residues essential for agonist-induced 5-HT2A receptor desensitization." Biochemistry 42(36): 10853-62. Guan, X. M., T. S. Kobilka, et al. (1992). "Enhancement of membrane insertion and function in a type IIIb membrane protein following introduction of a cleavable signal peptide." J Biol Chem 267(31): 21995-8. Guy-Grand, B., M. Apfelbaum, et al. (1989). "International trial of long-term dexfenfluramine in obesity." Lancet 2(8672): 1142-5. Hargrave, P. A., J. H. McDowell, et al. (1983). "The structure of bovine rhodopsin." Biophys Struct Mech 9(4): 235-44. Hauck, A. J., W. D. Edwards, et al. (1990). "Mitral and aortic valve disease associated with ergotamine therapy for migraine. Report of two cases and review of literature." Arch Pathol Lab Med 114(1): 62-4. Heisler, L. K., M. A. Cowley, et al. (2002). "Activation of central melanocortin pathways by fenfluramine." Science 297(5581): 609-11. Helton, L. A., K. B. Thor, et al. (1994). "5-hydroxytryptamine2A, 5- hydroxytryptamine2B, and 5-hydroxytryptamine2C receptor mRNA expression in the spinal cord of rat, cat, monkey and human." Neuroreport 5(18): 2617-20. Hendrikx, M., J. Van Dorpe, et al. (1996). "Aortic and mitral valve disease induced by ergotamine therapy for migraine: a case report and review of the literature." J Heart Valve Dis 5(2): 235-7. Hirst, W. D., N. Y. Cheung, et al. (1998). "Cultured astrocytes express messenger RNA for multiple serotonin receptor subtypes, without functional coupling of 5-HT1 receptor subtypes to adenylyl cyclase." Brain Res Mol Brain Res 61(1-2): 90-9. Horvath, J., R. D. Fross, et al. (2004). "Severe multivalvular heart disease: a new complication of the ergot derivative dopamine agonists." Mov Disord 19(6): 656- 62. Hoyer, D., J. P. Hannon, et al. (2002). "Molecular, pharmacological and functional diversity of 5-HT receptors." Pharmacol Biochem Behav 71(4): 533-54. Hu, S., S. Wang, et al. (1998). "Inhibition of delayed rectifier K+ channels by dexfenfluramine (Redux)." J Pharmacol Exp Ther 287(2): 480-6. Huang, K., W. Lu, et al. (1995). "Water molecules participate in proteinase-inhibitor interactions: crystal structures of Leu18, Ala18, and Gly18 variants of turkey ovomucoid inhibitor third domain complexed with Streptomyces griseus proteinase B." Protein Sci 4(10): 1985-97. Hughes, A. R. and J. W. Putney, Jr. (1988). "Metabolism and functions of inositol phosphates." Biofactors 1(2): 117-21. Idzko, M., E. Panther, et al. (2004). "The serotoninergic receptors of human dendritic cells: identification and coupling to cytokine release." J Immunol 172(10): 6011- 9. Ishida, T., S. Kawashima, et al. (1998). "Nitric oxide is produced via 5-HT1B and 5- HT2B receptor activation in human coronary artery endothelial cells." Kobe J Med Sci 44(2): 51-63. Jaffre, F., J. Callebert, et al. (2004). "Involvement of the serotonin 5-HT2B receptor in cardiac hypertrophy linked to sympathetic stimulation: control of interleukin-6,
162 interleukin-1beta, and tumor necrosis factor-alpha cytokine production by ventricular fibroblasts." Circulation 110(8): 969-74. Jerman, J. C., S. J. Brough, et al. (2001). "Pharmacological characterisation of human 5- HT2 receptor subtypes." Eur J Pharmacol 414(1): 23-30. Jick, H., C. Vasilakis, et al. (1998). "A population-based study of appetite-suppressant drugs and the risk of cardiac-valve regurgitation." N Engl J Med 339(11): 719-24. Jollis, J. G., C. K. Landolfo, et al. (2000). "Fenfluramine and phentermine and cardiovascular findings: effect of treatment duration on prevalence of valve abnormalities." Circulation 101(17): 2071-7. Jones, S. R., R. R. Gainetdinov, et al. (1998). "Mechanisms of Amphetamine Action Revealed in Mice Lacking the Dopamine Transporter." J. Neurosci. 18(6): 1979- 1986. Kaumann, A. J. and L. Sanders (1994). "5-Hydroxytryptamine causes rate-dependent arrhythmias through 5-HT4 receptors in human atrium: facilitation by chronic beta-adrenoceptor blockade." Naunyn Schmiedebergs Arch Pharmacol 349(4): 331-7. Kawaguchi, S. and S. Kuramitsu (1998). "Thermodynamics and molecular simulation analysis of hydrophobic substrate recognition by aminotransferases." J Biol Chem 273(29): 18353-64. Kellermann, O., S. Loric, et al. (1996). "Sequential onset of three 5-HT receptors during the 5-hydroxytryptaminergic differentiation of the murine 1C11 cell line." Br J Pharmacol 118(5): 1161-70. Kennett, G. A., M. D. Wood, et al. (1996). "In vitro and in vivo profile of SB 206553, a potent 5-HT2C/5-HT2B receptor antagonist with anxiolytic-like properties." Br J Pharmacol 117(3): 427-434. Khan, M. A., C. A. Herzog, et al. (1998). "The prevalence of cardiac valvular insufficiency assessed by transthoracic echocardiography in obese patients treated with appetite-suppressant drugs." N Engl J Med 339(11): 713-8. Khorana, H. G., G. E. Gerber, et al. (1979). "Amino acid sequence of bacteriorhodopsin." Proc Natl Acad Sci U S A 76(10): 5046-50. Kilpatrick, I. C., M. Traut, et al. (2001). "Monoamine oxidase inhibition is unlikely to be relevant to the risks associated with phentermine and fenfluramine: a comparison with their abilities to evoke monoamine release." Int J Obes Relat Metab Disord 25(10): 1454-8. Kobilka, B. K., R. A. Dixon, et al. (1987). "cDNA for the human beta 2-adrenergic receptor: a protein with multiple membrane-spanning domains and encoded by a gene whose chromosomal location is shared with that of the receptor for platelet- derived growth factor." Proc Natl Acad Sci U S A 84(1): 46-50. Kong, E. K., L. Peng, et al. (2002). "Up-regulation of 5-HT2B receptor density and receptor-mediated glycogenolysis in mouse astrocytes by long-term fluoxetine administration." Neurochem Res 27(1-2): 113-20. Kozak, M. (1994). "Features in the 5' non-coding sequences of rabbit alpha and beta- globin mRNAs that affect translational efficiency." J Mol Biol 235(1): 95-110. Kristiansen, K., W. K. Kroeze, et al. (2000). "A highly conserved aspartic acid (Asp-155) anchors the terminal amine moiety of tryptamines and is involved in membrane
163 targeting of the 5-HT(2A) serotonin receptor but does not participate in activation via a "salt-bridge disruption" mechanism." J Pharmacol Exp Ther 293(3): 735-46. Kroeze, W. K., K. Kristiansen, et al. (2002). "Molecular biology of serotonin receptors structure and function at the molecular level." Curr Top Med Chem 2(6): 507-28. Kroeze, W. K., D. J. Sheffler, et al. (2003). "G-protein-coupled receptors at a glance." J Cell Sci 116(Pt 24): 4867-9. Kursar, J. D., D. L. Nelson, et al. (1994). "Molecular cloning, functional expression, and mRNA tissue distribution of the human 5-hydroxytryptamine2B receptor." Mol Pharmacol 46(2): 227-34. Kursar, J. D., D. L. Nelson, et al. (1992). "Molecular cloning, functional expression, and pharmacological characterization of a novel serotonin receptor (5- hydroxytryptamine2F) from rat stomach fundus." Mol Pharmacol 42(4): 549-57. Launay, J. M., G. Birraux, et al. (1996). "Ras involvement in signal transduction by the serotonin 5-HT2B receptor." J Biol Chem 271(6): 3141-7. Launay, J. M., P. Herve, et al. (2002). "Function of the serotonin 5-hydroxytryptamine 2B receptor in pulmonary hypertension." Nat Med 8(10): 1129-35. Leng, A., A. Ouagazzal, et al. (2003). "Effect of the 5-HT6 receptor antagonists Ro04- 6790 and Ro65-7199 on latent inhibition and prepulse inhibition in the rat: comparison to clozapine." Pharmacol Biochem Behav 75(2): 281-8. Leonard, B. E. (1992). "Sub-types of serotonin receptors: biochemical changes and pharmacological consequences." Int Clin Psychopharmacol 7(1): 13-21. Lindner, M. D., D. B. Hodges, Jr., et al. (2003). "An assessment of the effects of serotonin 6 (5-HT6) receptor antagonists in rodent models of learning." J Pharmacol Exp Ther 307(2): 682-91. Loke, Y. K., S. Derry, et al. (2002). "Appetite suppressants and valvular heart disease - a systematic review." BMC Clin Pharmacol 2(1): 6. Lomasney, J. W., S. Cotecchia, et al. (1991). "Molecular cloning and expression of the cDNA for the alpha 1A-adrenergic receptor. The gene for which is located on human chromosome 5." J Biol Chem 266(10): 6365-9. Lomasney, J. W., W. Lorenz, et al. (1990). "Expansion of the alpha 2-adrenergic receptor family: cloning and characterization of a human alpha 2-adrenergic receptor subtype, the gene for which is located on chromosome 2." Proc Natl Acad Sci U S A 87(13): 5094-8. Loric, S., J. M. Launay, et al. (1992). "New mouse 5-HT2-like receptor. Expression in brain, heart and intestine." FEBS Lett 312(2-3): 203-7. Loric, S., L. Maroteaux, et al. (1995). "Functional serotonin-2B receptors are expressed by a teratocarcinoma-derived cell line during serotoninergic differentiation." Mol Pharmacol 47(3): 458-66. Lu, W., I. Apostol, et al. (1997). "Binding of amino acid side-chains to S1 cavities of serine proteinases." J Mol Biol 266(2): 441-61. Lucaites, V. L., D. L. Nelson, et al. (1996). "Receptor subtype and density determine the coupling repertoire of the 5-HT2 receptor subfamily." Life Sci 59(13): 1081-95. Manivet, P., S. Mouillet-Richard, et al. (2000). "PDZ-dependent activation of nitric-oxide synthases by the serotonin 2B receptor." J Biol Chem 275(13): 9324-31.
164 Manivet, P., B. Schneider, et al. (2002). "The serotonin binding site of human and murine 5-HT2B receptors: molecular modeling and site-directed mutagenesis." J Biol Chem 277(19): 17170-8. Marek, G. J., A. A. Li, et al. (1989). "Evidence for involvement of 5-hydroxytryptamine1 receptors in antidepressant-like drug effects on differential-reinforcement-of-low- rate 72-second behavior." J Pharmacol Exp Ther 250(1): 60-71. Mark, E. J., E. D. Patalas, et al. (1997). "Fatal pulmonary hypertension associated with short-term use of fenfluramine and phentermine." N Engl J Med 337(9): 602-6. Mathus-Vliegen, E. M. (1993). "Prolonged surveillance of dexfenfluramine in severe obesity." Neth J Med 43(5-6): 246-53. McCann, U. D., J. Yuan, et al. (1997). "Selective serotonin reuptake inhibitors dissociate fenfluramine's anorectic and neurotoxic effects: importance of dose, species and drug." J Pharmacol Exp Ther 281(3): 1487-98. Meltzer, H. Y. (1989). "Clinical studies on the mechanism of action of clozapine: the dopamine-serotonin hypothesis of schizophrenia." Psychopharmacology (Berl) 99 Suppl: S18-27. Millan, M. J., L. Maiofiss, et al. (2002). "Differential Actions of Antiparkinson Agents at Multiple Classes of Monoaminergic Receptor. I. A Multivariate Analysis of the Binding Profiles of 14 Drugs at 21 Native and Cloned Human Receptor Subtypes." J Pharmacol Exp Ther 303(2): 791-804. Morton, A., W. A. Baase, et al. (1995). "Energetic origins of specificity of ligand binding in an interior nonpolar cavity of T4 lysozyme." Biochemistry 34(27): 8564-75. Moses, P. L. and R. J. Wurtman (1984). "The ability of certain anorexic drugs to suppress food consumption depends on the nutrient composition of the test diet." Life Sci 35(12): 1297-300. Munro, J. F., A. C. MacCuish, et al. (1968). "Comparison of continuous and intermittent anorectic therapy in obesity." Br Med J 1(5588): 352-4. Nandigama, R. K., P. Newton-Vinson, et al. (2002). "Phentermine inhibition of recombinant human liver monoamine oxidases A and B." Biochem Pharmacol 63(5): 865-9. Nash, J. F., B. L. Roth, et al. (1994). "Effect of the R(-) and S(+) isomers of MDA and MDMA on phosphatidyl inositol turnover in cultured cells expressing 5-HT2A or 5-HT2C receptors." Neurosci Lett 177(1-2): 111-5. Nebigil, C. G., D. S. Choi, et al. (2000). "Serotonin 2B receptor is required for heart development." Proc Natl Acad Sci U S A 97(17): 9508-13. Nebigil, C. G., N. Etienne, et al. (2003). "Serotonin is a novel survival factor of cardiomyocytes: mitochondria as a target of 5-HT2B receptor signaling." Faseb J 17(10): 1373-5. Nebigil, C. G., P. Hickel, et al. (2001). "Ablation of serotonin 5-HT(2B) receptors in mice leads to abnormal cardiac structure and function." Circulation 103(24): 2973-9. Nebigil, C. G., J. M. Launay, et al. (2000). "5-hydroxytryptamine 2B receptor regulates cell-cycle progression: cross-talk with tyrosine kinase pathways." Proc Natl Acad Sci U S A 97(6): 2591-6. Neill, J. C. and S. J. Cooper (1989). "Evidence that d-fenfluramine anorexia is mediated by 5-HT1 receptors." Psychopharmacology (Berl) 97(2): 213-8.
165 Nelson, D. L., V. L. Lucaites, et al. (1999). "Comparisons of hallucinogenic phenylisopropylamine binding affinities at cloned human 5-HT2A, -HT(2B) and 5-HT2C receptors." Naunyn Schmiedebergs Arch Pharmacol 359(1): 1-6. Newman-Tancredi, A., D. Cussac, et al. (2002). "Differential actions of antiparkinson agents at multiple classes of monoaminergic receptor. III. Agonist and antagonist properties at serotonin, 5-HT(1) and 5-HT(2), receptor subtypes." J Pharmacol Exp Ther 303(2): 815-22. Nichols, D. E. (2004). "Hallucinogens." Pharmacol Ther 101(2): 131-81. O'Connor, H. T., R. M. Richman, et al. (1995). "Dexfenfluramine treatment of obesity: a double blind trial with post trial follow up." Int J Obes Relat Metab Disord 19(3): 181-9. Oh, C. K., M. J. Drescher, et al. (1999). "Selective expression of serotonin receptor transcripts in the mammalian cochlea and its subdivisions." Brain Res Mol Brain Res 70(1): 135-40. Oue, S., A. Okamoto, et al. (1997). "Paracoccus denitrificans aromatic amino acid aminotransferase: a model enzyme for the study of dual substrate recognition mechanism." J Biochem (Tokyo) 121(1): 161-71. Ovchinnikov, Y. A., N. G. Abdulaev, et al. (1977). "Recent findings in the structure- functional characteristics of bacteriorhodopsin." FEBS Lett 84(1): 1-4. Palczewski, K., T. Kumasaka, et al. (2000). "Crystal structure of rhodopsin: A G protein- coupled receptor." Science 289(5480): 739-45. Palvimaki, E. P., B. L. Roth, et al. (1996). "Interactions of selective serotonin reuptake inhibitors with the serotonin 5-HT2c receptor." Psychopharmacology (Berl) 126(3): 234-40. Parker, P. J. and J. Murray-Rust (2004). "PKC at a glance." J Cell Sci 117(Pt 2): 131-2. Patterson, R. L., D. Boehning, et al. (2004). "Inositol 1,4,5-trisphosphate receptors as signal integrators." Annu Rev Biochem 73: 437-65. Pedrinola, F., C. Sztejnsznajd, et al. (1996). "The addition of dexfenfluramine to fluoxetine in the treatment of obesity: a randomized clinical trial." Obes Res 4(6): 549-54. Peroutka, S. J. and S. H. Snyder (1980). "Long-term antidepressant treatment decreases spiroperidol-labeled serotonin receptor binding." Science 210(4465): 88-90. Peroutka, S. J. and S. H. Snyder (1980). "Regulation of serotonin2 (5-HT2) receptors labeled with [3H]spiroperidol by chronic treatment with the antidepressant amitriptyline." J Pharmacol Exp Ther 215(3): 582-7. Porter, R. H., K. R. Benwell, et al. (1999). "Functional characterization of agonists at recombinant human 5-HT2A, 5-HT2B and 5-HT2C receptors in CHO-K1 cells." Br J Pharmacol 128(1): 13-20. Porter, R. H., C. S. Malcolm, et al. (2001). "Agonist-induced functional desensitization of recombinant human 5-HT2 receptors expressed in CHO-K1 cells." Biochem Pharmacol 62(4): 431-8. Pouwels, H. M., J. L. Smeets, et al. (1990). "Pulmonary hypertension and fenfluramine." Eur Respir J 3(5): 606-7. Pritchett, A. M., J. F. Morrison, et al. (2002). "Valvular heart disease in patients taking pergolide." Mayo Clin Proc 77(12): 1280-6.
166 Putney, J. W., Jr. and C. M. Ribeiro (2000). "Signaling pathways between the plasma membrane and endoplasmic reticulum calcium stores." Cell Mol Life Sci 57(8-9): 1272-86. Raleigh, M. J., G. L. Brammer, et al. (1986). "Effects of chronic fenfluramine on blood serotonin, cerebrospinal fluid metabolites, and behavior in monkeys." Psychopharmacology (Berl) 90(4): 503-8. Ranquin, R. and H. M. Brems (1977). "Metabolic effects and anti-obesity action of fenfluramine in prolonged-acting form." Curr Med Res Opin 5(4): 341-6. Redmon, B., S. Raatz, et al. (1997). "Valvular heart disease associated with fenfluramine- phentermine." N Engl J Med 337(24): 1773-4; author reply 1775. Regan, J. W., T. S. Kobilka, et al. (1988). "Cloning and expression of a human kidney cDNA for an alpha 2-adrenergic receptor subtype." Proc Natl Acad Sci U S A 85(17): 6301-5. Robinson, D. S., D. R. Alms, et al. (1989). "Serotonergic anxiolytics and treatment of depression." Psychopathology 22 Suppl 1: 27-36. Rogers, D. C. and J. J. Hagan (2001). "5-HT6 receptor antagonists enhance retention of a water maze task in the rat." Psychopharmacology (Berl) 158(2): 114-9. Roth, B. L. (1994). "Multiple serotonin receptors: clinical and experimental aspects." Ann Clin Psychiatry 6(2): 67-78. Roth, B. L., K. Baner, et al. (2002). "Salvinorin A: a potent naturally occurring nonnitrogenous kappa opioid selective agonist." Proc Natl Acad Sci U S A 99(18): 11934-9. Roth, B. L. and D. M. Chuang (1987). "Multiple mechanisms of serotonergic signal transduction." Life Sci 41(9): 1051-64. Roth, B. L., S. C. Craigo, et al. (1994). "Binding of typical and atypical antipsychotic agents to 5-hydroxytryptamine-6 and 5-hydroxytryptamine-7 receptors." J Pharmacol Exp Ther 268(3): 1403-10. Roth, B. L., D. J. Sheffler, et al. (2004). "Magic shotguns versus magic bullets: selectively non-selective drugs for mood disorders and schizophrenia." Nat Rev Drug Discov 3(4): 353-9. Roth, B. L., M. Shoham, et al. (1997). "Identification of conserved aromatic residues essential for agonist binding and second messenger production at 5- hydroxytryptamine2A receptors." Mol Pharmacol 52(2): 259-66. Roth, B. L., D. L. Willins, et al. (1998). "5-Hydroxytryptamine2-family receptors (5- hydroxytryptamine2A, 5-hydroxytryptamine2B, 5-hydroxytryptamine2C): where structure meets function." Pharmacol Ther 79(3): 231-57. Rothman, R. B., M. H. Baumann, et al. (2000). "Evidence for possible involvement of 5- HT(2B) receptors in the cardiac valvulopathy associated with fenfluramine and other serotonergic medications." Circulation 102(23): 2836-41. Rothman, R. B., J. B. Redmon, et al. (2000). "Chronic treatment with phentermine combined with fenfluramine lowers plasma serotonin." Am J Cardiol 85(7): 913- 5, A10. Roy, A., N. J. Brand, et al. (2000). "Expression of 5-hydroxytryptamine receptor subtype messenger RNA in interstitial cells from human heart valves." J Heart Valve Dis 9(2): 256-60; discussion 260-1.
167 Russell, G. F., S. A. Checkley, et al. (1988). "A controlled trial of d-fenfluramine in bulimia nervosa." Clin Neuropharmacol 11 Suppl 1: S146-59. Sanden, N., T. Thorlin, et al. (2000). "5-Hydroxytryptamine2B receptors stimulate Ca2+ increases in cultured astrocytes from three different brain regions." Neurochem Int 36(4-5): 427-34. Schaerlinger, B., P. Hickel, et al. (2003). "Agonist actions of dihydroergotamine at 5- HT2B and 5-HT2C receptors and their possible relevance to antimigraine efficacy." Br J Pharmacol 140(2): 277-84. Schmuck, K., C. Ullmer, et al. (1996). "Activation of meningeal 5-HT2B receptors: an early step in the generation of migraine headache?" Eur J Neurosci 8(5): 959-67. Schwartz, D., L. Hernandez, et al. (1989). "Fenfluramine administered systemically or locally increases extracellular serotonin in the lateral hypothalamus as measured by microdialysis." Brain Res 482(2): 261-70. Schwinn, D. A., J. W. Lomasney, et al. (1990). "Molecular cloning and expression of the cDNA for a novel alpha 1-adrenergic receptor subtype." J Biol Chem 265(14): 8183-9. Sealfon, S. C., L. Chi, et al. (1995). "Related contribution of specific helix 2 and 7 residues to conformational activation of the serotonin 5-HT2A receptor." J Biol Chem 270(28): 16683-8. Selikowitz, M., J. Sunman, et al. (1990). "Fenfluramine in Prader-Willi syndrome: a double blind, placebo controlled trial." Arch Dis Child 65(1): 112-4. Series, H. G., P. J. Cowen, et al. (1994). "p-Chloroamphetamine (PCA), 3,4- methylenedioxy-methamphetamine (MDMA) and d-fenfluramine pretreatment attenuates d-fenfluramine-evoked release of 5-HT in vivo." Psychopharmacology (Berl) 116(4): 508-14. Serratrice, J., P. Disdier, et al. (2002). "Fibrotic valvular heart disease subsequent to bromocriptine treatment." Cardiol Rev 10(6): 334-6. Setola, V., S. J. Hufeisen, et al. (2003). "3,4-methylenedioxymethamphetamine (MDMA, "Ecstasy") induces fenfluramine-like proliferative actions on human cardiac valvular interstitial cells in vitro." Mol Pharmacol 63(6): 1223-9. Shapiro, D. A., K. Kristiansen, et al. (2000). "Differential modes of agonist binding to 5- hydroxytryptamine(2A) serotonin receptors revealed by mutation and molecular modeling of conserved residues in transmembrane region 5." Mol Pharmacol 58(5): 877-86. Shapiro, D. A., K. Kristiansen, et al. (2002). "Evidence for a model of agonist-induced activation of 5-hydroxytryptamine 2A serotonin receptors that involves the disruption of a strong ionic interaction between helices 3 and 6." J Biol Chem 277(13): 11441-9. Sheffler, D. J. and B. L. Roth (2003). "Salvinorin A: the "magic mint" hallucinogen finds a molecular target in the kappa opioid receptor." Trends Pharmacol Sci 24(3): 107-9. Shi, L. and J. A. Javitch (2004). "The second extracellular loop of the dopamine D2 receptor lines the binding-site crevice." Proc Natl Acad Sci U S A 101(2): 440-5. Simonneau, G., M. Fartoukh, et al. (1998). "Primary pulmonary hypertension associated with the use of fenfluramine derivatives." Chest 114(3 Suppl): 195S-199S.
168 Singh, J. P., J. C. Evans, et al. (1999). "Prevalence and clinical determinants of mitral, tricuspid, and aortic regurgitation (the Framingham Heart Study)." Am J Cardiol 83(6): 897-902. Steffee, C. H., H. K. Singh, et al. (1999). "Histologic changes in three explanted native cardiac valves following use of fenfluramines." Cardiovasc Pathol 8(5): 245-53. Stenkamp, R. E., S. Filipek, et al. (2002). "Crystal structure of rhodopsin: a template for cone visual pigments and other G protein-coupled receptors." Biochim Biophys Acta 1565(2): 168-82. Stewart, G. O., G. R. Stein, et al. (1993). "Dexfenfluramine in type II diabetes: effect on weight and diabetes control." Med J Aust 158(3): 167-9. Strother, J., P. Fedullo, et al. (1999). "Complex vascular lesions at autopsy in a patient with phentermine-fenfluramine use and rapidly progressing pulmonary hypertension." Arch Pathol Lab Med 123(6): 539-40. Stunkard, A. J., L. W. Craighead, et al. (1980). "Controlled trial of behaviour therapy, pharmacotherapy, and their combination in the treatment of obesity." Lancet 2(8203): 1045-7. Taylor, C. W., P. C. da Fonseca, et al. (2004). "IP(3) receptors: the search for structure." Trends Biochem Sci 29(4): 210-9. Tomita, T. and Q. Zhao (2002). "Autopsy findings of heart and lungs in a patient with primary pulmonary hypertension associated with use of fenfluramine and phentermine." Chest 121(2): 649-52. Tournois, C., V. Mutel, et al. (1998). "Cross-talk between 5-hydroxytryptamine receptors in a serotonergic cell line. Involvement of arachidonic acid metabolism." J Biol Chem 273(28): 17498-503. Ullmer, C., H. G. Boddeke, et al. (1996). "5-HT2B receptor-mediated calcium release from ryanodine-sensitive intracellular stores in human pulmonary artery endothelial cells." Br J Pharmacol 117(6): 1081-8. Ullmer, C., K. Schmuck, et al. (1995). "Expression of serotonin receptor mRNAs in blood vessels." FEBS Lett 370(3): 215-21. Ulus, I. H., T. J. Maher, et al. (2000). "Characterization of phentermine and related compounds as monoamine oxidase (MAO) inhibitors." Biochem Pharmacol 59(12): 1611-21. Unwin, P. N. and R. Henderson (1975). "Molecular structure determination by electron microscopy of unstained crystalline specimens." J Mol Biol 94(3): 425-40. Van Gaal, L. F., G. A. Vansant, et al. (1995). "Effects of dexfenfluramine on resting metabolic rate and thermogenesis in premenopausal obese women during therapeutic weight reduction." Metabolism 44(2 Suppl 2): 42-5. Ventura, C. and M. Maioli (2001). "Protein kinase C control of gene expression." Crit Rev Eukaryot Gene Expr 11(1-3): 243-67. Vickers, S. P., C. T. Dourish, et al. (2001). "Evidence that hypophagia induced by d- fenfluramine and d-norfenfluramine in the rat is mediated by 5-HT2C receptors." Neuropharmacology 41(2): 200-9. Villazon, M., J. F. Padin, et al. (2002). "Functional characterization of serotonin receptors in rat isolated aorta." Biol Pharm Bull 25(5): 584-90.
169 Wainscott, D. B., M. L. Cohen, et al. (1993). "Pharmacological characteristics of the newly cloned rat 5-hydroxytryptamine2F receptor." Mol Pharmacol 43(3): 419- 26. Watts, S. W. and J. M. Thompson (2004). "Characterization of the contractile 5- hydroxytryptamine receptor in the renal artery of the normotensive rat." J Pharmacol Exp Ther 309(1): 165-72. Watts, S. W., P. Yang, et al. (2001). "Activation of Erk mitogen-activated protein kinase proteins by vascular serotonin receptors." J Cardiovasc Pharmacol 38(4): 539-51. Weintraub, M., J. D. Hasday, et al. (1984). "A double-blind clinical trial in weight control. Use of fenfluramine and phentermine alone and in combination." Arch Intern Med 144(6): 1143-8. Weintraub, M., K. Sriwatanakul, et al. (1983). "Extended-release fenfluramine: patient acceptance and efficacy of evening dosing." Clin Pharmacol Ther 33(5): 621-7. Weintraub, M., P. R. Sundaresan, et al. (1992). "Long-term weight control study. VI. Individual participant response patterns." Clin Pharmacol Ther 51(5): 619-33. Weintraub, M., P. R. Sundaresan, et al. (1992). "Long-term weight control study. I (weeks 0 to 34). The enhancement of behavior modification, caloric restriction, and exercise by fenfluramine plus phentermine versus placebo." Clin Pharmacol Ther 51(5): 586-94. Weintraub, M., P. R. Sundaresan, et al. (1992). "Long-term weight control study. II (weeks 34 to 104). An open-label study of continuous fenfluramine plus phentermine versus targeted intermittent medication as adjuncts to behavior modification, caloric restriction, and exercise." Clin Pharmacol Ther 51(5): 595- 601. Weintraub, M., P. R. Sundaresan, et al. (1992). "Long-term weight control study. III (weeks 104 to 156). An open-label study of dose adjustment of fenfluramine and phentermine." Clin Pharmacol Ther 51(5): 602-7. Weissman, N. J. (2001). "Appetite suppressants and valvular heart disease." Am J Med Sci 321(4): 285-91. Weissman, N. J., J. F. Tighe, Jr., et al. (1998). "An assessment of heart-valve abnormalities in obese patients taking dexfenfluramine, sustained-release dexfenfluramine, or placebo. Sustained-Release Dexfenfluramine Study Group." N Engl J Med 339(11): 725-32. Woolley, M. L., J. C. Bentley, et al. (2001). "A role for 5-ht6 receptors in retention of spatial learning in the Morris water maze." Neuropharmacology 41(2): 210-9. Xu, J., B. Jian, et al. (2002). "Serotonin mechanisms in heart valve disease II: the 5-HT2 receptor and its signaling pathway in aortic valve interstitial cells." Am J Pathol 161(6): 2209-18. Zychlinski, L. and M. R. Montgomery (1984). "Alterations in monoamine oxidase activity after single and repeated exposure of rats to chlorphentermine." Toxicol Lett 22(2): 133-8. Zychlinski, L. and M. R. Montgomery (1985). "Chlorphentermine-induced biochemical alterations in mitochondrial membranes." J Pharmacol Exp Ther 233(1): 39-44.
170