Growth Hormone Secretagogues in Anti-Doping

Home , Capromorelin, Ibutamoren, Tabimorelin

POMPEU FABRA UNIVERSITY

Department of Experimental and Health Sciences

ASSESSMENT OF GROWTH HORMONE SECRETAGOGUES IN ANTI-DOPING

PhD thesis

Armand Pinyot Comelles

Neurosciences Research Programme

Municipal Institute of Medical Research

IMIM-Institut Hospital del Mar d’investigacions Mèdiques

Barcelona, July 2012

POMPEU FABRA UNIVERSITY

Department of Experimental and Health Sciences

Doctoral Programme: Health and Life Sciences

ASSESSMENT OF GROWTH HORMONE SECRETAGOGUES IN ANTI-DOPING

Memòria presentada per Armand Pinyot Comelles per a optar al títol de Doctor per la Universitat Pompeu Fabra. Aquesta tesis ha estat realitzada sota la codirecció del Dr. Jordi Segura Noguera i el Dr. Ricardo Gutiérrez Gallego, en el grup d’Investigació en Bioanàlisi i Serveis Analítics, programa de Neurociències de l’IMIM-Institut Hospital del Mar d’Investigacions Mèdiques. Programa de Doctorat en Ciències de la Salut i de la Vida de la Universitat Pompeu Fabra.

Dr Jordi Segura Dr Ricardo Gutiérrez Armand Pinyot Noguera Gallego Comelles Director de tesis Director de tesis Doctorand

Barcelona, Juliol 2012

Agraïments / Acknowledgements

Em disposo a presentar la meva tesi. Tot i el temps que fa que somio en poder pronunciar aquestes paraules, ara que va de debò, resulta que són completament falses. Efectivament, la tesi va al meu nom però seria impossible haver-la realitzat sense un llarg reguitzell de gent que m’ha proporcionat el seu suport, ja sigui científic, moral o simplement estant al meu costat un dia en el que no han sortit bé els experiments. Creieu-me si us dic que podria escriure diverses pàgines d’agraïments però resulta que la tesi tracta un altre tema i així, en aquest petit apartat, m’agradaria fer esment a algunes de les moltes persones de les que us parlo.

En primer lloc, d’aquesta tesi no n’existiria ni una paraula si no fos gràcies al Dr. Jordi Segura. Vull agrair-li profundament que em donés l’oportunitat d’entrar de la seva mà al món de la recerca científica, gràcies per la direcció de la tesi, per les correccions, pels consells i critiques sempre constructives. És un plaer poder treballar amb algú del seu talent i experiència. Espero de tot cor haver complert amb les expectatives dipositades en mi a l’inici del doctorat.

Gràcies al Dr. Ricardo Gutiérrez per la codirecció. El Dr. Ricardo Gutiérrez, ha estat per mi un inestimable guia del fosc túnel en el que es pot convertir la recerca després d’una temporada de mals resultats. He aprés tantes coses d’ell que és complicat triar-ne una de sola però la més encertada seria el rigor, ja sigui experimental com literari. Moltes gràcies per les incomptables hores dedicades a les correccions provisionals, a la ingent quantitat de e-mails intercanviats per tal que aquesta tesi sigui el que és. Degut a l’elevat nombre d’hores conviscudes, puc afirmar que a part d’un formidable director de tesi, ha resultat una gran persona i a més, un bon amic.

Voldria també agrair als altres components del grup GHS. Al Dr. Zoran Nikolovski per creure que jo podria portar a bon port aquest projecte quan encara ni tant sols havia acabat la carrera. Tot i que hem discutit més que no pas treballat, sé que m’emporto un amic. Sóc conscient que el teu afany per treure’m de polleguera, qüestionar tot el que deia o feia, era senzillament per fer-me créixer com a persona i com a científic. Agrair també al Dr. Jaume Bosch per les incomptables hores de xerrameca científica, per donar-me tot el seu suport en infinitat de situacions i per il·luminar-me amb el seu coneixement en termes de ciència, de la vida en general i dels mots encreuats en particular.

Fent memòria de la gran quantitat d’hores passades dins les parets de l’IMIM, em venen munts d’imatges d’un grup d’amics que s’ho passava de conya mentre treballava. Crec que vam aconseguir crear un grup capaç de fer més esport de forma conjunta que cadascú per separat, confeccionant moments memorables dels quals m’aprofitaré per anar agraint individualment. Gràcies Josep, Quim i Raúl pel moment èpic de pujar tots en el segon calaix d’un podi mentre al primer lloc hi havia un nen de sis anys. Gràcies de forma individual al Quim per haver de baixar cames ajudeu- me del cim de l’Aneto. Gràcies al Toni per la més gran demostració de superació després de pujar i baixar del Mont Perdut amb unes hamburgueses als talons. Gràcies a la Mito i la Civit per demostrar que és pràcticament impossible controlar un caiac si aquest està del revés. Gràcies a la Beth per les lliçons magistrals de basquet i sobretot per aquelles pilotes de voleibol xutades amb ràbia. Perdó a la Carmen per la infinitat de pilotades que va rebre, juro que de forma totalment involuntària, en diversos partits de voleibol, bàsquet o futbol, en aquest apartat també es pot afegir l’Ester, que en un partit de bàsquet va quedar grogui després un xoc, tothom sospita que voluntari, amb el Gerard. Ah, el Gerard, mestre de mestres, gran mag del caos, l’il·lusionista de S-ants (si, si, així, separat), gràcies per l’esport de més alt risc que és intentar desxifrar algun dels teus correus. A tots vosaltres, esportistes d’elit frustrats, mil gràcies.

Em trobo en la obligació moral d’agrair aquesta tesi a tres grans amics, al Juanjo, al Pep i al Salva. Segurament per motius completament diferents l’un de l’altre però que sense la suma dels tres aquesta tesi no hauria estat possible.

En aquests anys a l’IMIM, a part d’una gran quantitat d’amics, he trobat també algú amb qui compartir molt més que esports accidentats (que també). Algú amb qui compartir el dia a dia i els dies que vindran. Gràcies Cris, per fer que tot el que m’envolta tingui sentit.

Per acabar, m’agradaria agrair a la meva família, als meus pares, al meu germà, la Clàudia, la Laura i el Marc i la Cris pel suport que sempre he rebut i per ser uns referents modèlics. M’agradaria dedicar-vos aquesta tesi ja que és, en la major part, vostra.

Abstract

The presented doctoral thesis describes de development of a new analytical method in anti-doping. At present, the detection of recombinant growth hormone abuse is relatively settled, and therefore it is anticipated that growth hormone secretagogues (GHS) will replace GH, as they can induce similar anabolic effects. The aim of this thesis is the detection of growth hormone secretagogues (GHS) from urine samples.

GHS presents high structural diversity rendering very difficult a single conventional chemical determination for all. Thus, in this approach the single feature shared by all GHS, interaction with the specific receptor GHS-R1a, is targeted using a method based on a competitive assay against a labelled ligand. By monitoring the signal of the labelled ligand at the end of the test, the presence of any GHS in the sample can be determined.

A threshold value below which a sample is considered suspicious with low probability of being false positive was set up. The developed methodology was tested with 8 different GHS, and also samples from an administration assay of one of the GHS. Variables such as age, gender or physical exercise do not interfere with the method, which is currently under validation. Resum

La present tesis doctoral desenvolupa una nova metodologia analítica antidopatge. Degut a que actualment la detecció de l'abús d'hormona del creixement recombinant està relativament resolta, es tem que els secretagogs de la hormona del creixement (GHS) siguin el relleu natural ja que es creu provoquen efectes anabòlics semblants. L'objectiu de la tesi és la detecció de secretagogs de l'hormona del creixement (GHS) a partir de mostres d’orina.

Els GHS presenten gran diversitat estructural que fa difícil una única determinació química convencional. Per això s'ha optat per utilitzar la seva propietat comú d'interaccionar amb el receptor específic GHS-R1a, mitjançant un mètode basat en un assaig competitiu front un lligand marcat. Monitorant la senyal del lligand marcat al final de l’assaig es pot determinar la presència de qualsevol GHS a la mostra.

S'ha determinat per l’assaig un valor de tall de la senyal del lligand marcat per sota de la qual una mostra és considerada sospitosa amb una baixa probabilitat d'obtenir un fals positiu. El mètode desenvolupat ha funcionat amb els 8 GHS estudiats, incloses mostres reals procedents d'un assaig d’administració d’un d’ells. Variables com l'edat, gènere o exercici físic no provoquen interferències en el mètode, que actualment es troba en fase de validació. Abbreviations

125I Iodine 125 Ala Alanine AMP Adenosine monophosphate Bmax Receptor density cAMP Cyclic adenosine monophosphate CHO Chinese hamster ovary cell CS-A Chorionic somatomammotropin-A CS-B Chorionic somatomammotropin-B CS-L Chorionic somatomammotropin-like CV Coefficient of variation Cy5 Cyanine D Concentration of labelled ligand Da Daltons DELFIA Dissociation-enhanced lanthanide fluorescent immunoassay DMEM Dulbecco's modified Eagle medium DMSO Dimethyl sulfoxide EPO Erythropoietin FAM Carboxyfluorescein FBS Foetal bovine serum FITC Fluorescein isothiocyanate fmol Femtomol FRET Fluorescence resonance energy transfer GHBP Growth hormone binding protein GHIH Growth hormone inhibiting hormone GHRH Growth hormone releasing hormone GHRH-R Growth hormone releasing hormone receptor GHRP GH-releasing peptide GHS Growth hormone secretagogue GHS-R1a Growth hormone secretagogue variant 1a GHS-R1b Growth hormone secretagogue variant 1b Gly Glycine Glu Glutamic acid GOAT Ghrelin O-acyltransferase h Hour HEK Human embryonic kidney hGH Human growth hormone hGH-N Growth hormone-normal hGH-V Growth hormone-variant His Histidine HPLC High-performance liquid chromatography IC50 Half maximal inhibitory concentration IGFBP Insulin-like growth factor binding protein IGF-I Insulin-like growth factor I Kd Equilibrium dissociation constant of the labelled drug. kDa Kilodalton KeV Kiloelectronvolt kg Kilogram Ki Equilibrium dissociation constant of the unlabelled drug. L Litre LC Liquid chromatography LOD Limit of detection Lys Lysine m Minute M Molar MS Mass spectrometry Met Methionine mCi Millicurie min Minute ml Millilitre mM Millimolar mOsmol Milliosmole MPC Minimal positive concentration mRNA Messenger RNA nM Nanomolar nm Nanometre NF-kB Nuclear factor kappa-light-chain-enhancer of B cells. NMR Nuclear magnetic resonance pg Picogram Phe Phenylalanine P-III-NP Procolagen III N-terminal peptide P-III-P Procolagen III peptide pM Picomolar rhGH Recombinant human growth hormone rpm Revolutions per minute RSB Relative specific binding s Seconds SD Standard deviation Ser Serine SIRF Somatotropin release-inhibiting factor SPE Solid phase extraction SSTR Somatostatin receptor Trp Tryptophan Tyr Tyrosine USADA United States Anti-Doping Agency WADA World Anti-Doping Agency WOO Window of opportunity λ Wavelength µg Microgram µl Microliter µM Micromolar

Thesis structure

This manuscript is structured in six main chapters. The first chapter includes the Introduction, covering the background information on human growth hormone secretagogues and its role in the anti-doping matter. The second chapter comprises the Objectives, including a list of the main targets of the work. The third chapter states the initial development phase of the project leading to the experimental competition binding protocol described in chapter 4 and the final method described in chapter 5. Chapter 4 and 5 are based on the cited published material and adapted to the thesis format in order to facilitate the reading. Finally, in chapter 6 an overall discussion is detailed. Each chapter contains a list of cited references, figures and table numeration, so as to maintain the published artwork structure in chapters 4 and 5. An appendix is included containing the mentioned publications derived from the work.

Contents

1 Chapter 1: INTRODUCTION ...... 1 1.1 Growth hormone ...... 2 1.2 Growth hormone biological functions ...... 5 1.3 Growth hormone secretion ...... 7 1.4 Growth hormone secretagogues ...... 8 1.5 Growth hormone secretagogue receptor ...... 11 1.6 Endogenous GHS: Ghrelin ...... 12 1.7 Biological functions of ghrelin ...... 17 1.8 Synthetic growth hormone secretagogues ...... 20 1.9 hGH and GHS for doping purposes ...... 22 1.10 rhGH anti-doping detection ...... 24 1.11 GHS anti-doping detection ...... 28 1.12 Proposal of a GHS detection test ...... 29 1.13 References ...... 31

2 Chapter 2: OBJECTIVES ...... 49 2.1 Main objective ...... 50 2.2 Secondary objectives ...... 50

3 Chapter 3: METHODOLOGICAL DEVELOPMENTS ...... 51 3.1 Principle of competitive binding assays ...... 52 3.2 Competitive binding in GHS detection ...... 56 3.2.1 General considerations ...... 56 3.2.2 Initial competitive binding methodology ...... 57 3.2.3 Labelled ligand ...... 58 3.2.3.1 [125I-His9]-ghrelin ...... 59 3.2.4 Receptor source ...... 61 3.2.4.1 Corroboration of receptor functionality ...... 61 3.2.4.2 Commercial membranes containing GHS-R1a ...... 63 3.2.4.3 Whole cell binding ...... 64 3.2.4.4 Cell disruption procedure optimization ...... 65 3.2.5 Sample treatment ...... 67 3.3 Concluding statements ...... 67 3.4 References ...... 68

4 Chapter 4: ON THE USE OF CELLS OR MEMBRANES FOR RECEPTOR BINDING: GROWTH HORMONE SECRETAGOGUES ...... 71 4.1 Abstract ...... 72 4.2 Introduction ...... 73 4.3 Material and methods ...... 77 4.3.1 Chemicals ...... 77 4.3.2 Cells ...... 78 4.3.3 GHSR-1a activity ...... 80 4.3.4 Binding assay ...... 81 4.3.5 Whole cell preparation ...... 81 4.3.6 Membrane preparation ...... 82 4.3.7 Saturation binding ...... 82 4.3.8 Competition binding ...... 83 4.4 Results ...... 84 4.4.1 GHSR-1a activity ...... 85 4.4.2 Radioligand receptor assays ...... 85 4.4.2.1 Membrane preparation ...... 85 4.4.2.2 Receptor stability over time ...... 87 4.4.2.3 Steady-state binding versus experimental conditions ...... 88 4.4.3 Comparison of isolated membranes and intact cells ...... 90 125 4.4.3.1 Saturation binding of [ I-His9]-ghrelin to GHS-R1a ...... 90 4.4.3.2 Competition of different agonists for GHS-R1a with radiolabeled ghrelin ...... 93 4.5 Discussion ...... 95 4.6 Acknowledgments ...... 101 4.7 References ...... 101

5 Chapter 5: GROWTH HORMONE SECRETAGOGUES: OUT OF COMPETITION ...... 107 5.1 Abstract ...... 108 5.2 Introduction ...... 110 5.3 Material and methods ...... 112 5.3.1 Urine purification ...... 113 5.3.2 Calibrators ...... 114 5.3.3 Samples ...... 115 5.4 Results ...... 115 5.4.1 Sample purification ...... 116 5.4.1.1 Desalting ...... 116 5.4.1.2 Selective secretagogue extraction ...... 117 5.4.2 Assay accuracy ...... 117 5.4.3 Gender and age effect ...... 118 5.4.4 Exercise effect ...... 119 5.4.5 Threshold value ...... 120 5.4.6 Excretion study ...... 122 5.4.7 Linearity of the quantifiable range of GHRP-2 spiked urine samples ...... 123 5.5 Discussion ...... 126 5.6 Acknowledgments ...... 133 5.7 References ...... 134

6 Chapter 6: DISCUSSION ...... 141 6.1 The test ...... 142 6.1.1 Need for a test to detect GHS ...... 142 6.1.2 Material of the test ...... 147 6.1.2.1 The radioligand ...... 148 6.1.2.2 Receptor ...... 153 6.1.3 Methodology and assay requirements ...... 158 6.1.4 Summary ...... 162 6.2 Future improvements and alternative methodologies ...... 167 6.2.1 Future improvements ...... 167 6.2.2 Alternative methodologies ...... 171 6.2.2.1 Receptor-based methodologies ...... 171 6.2.2.2 Non receptor-based methodologies ...... 177 6.3 References ...... 180

7 Chapter 7: CONCLUSIONS ...... 193

8 Chapter 8: APENDIX ...... 197

1 Chapter 1:

INTRODUCTION

1.1 Growth hormone

Human growth hormone (hGH) is a single-chain protein that is synthesized, stored, and secreted by the somatotroph cells of the anterior lobe of the pituitary gland [1].

This protein results from the partial transcription of the hGH gene cluster located in the long arm of chromosome 17, which is approximately 66,500 bases long. This cluster is composed of five related genes: hGH-N (growth hormone- normal) gene, CS-L (chorionic somatomammotropin-like) gene, CS-A (chorionic somatomammotropin-A) gene, hGH-V (growth hormone-variant) gene, and the CS-B (chorionic somatomammotropin-B). hGH-N is only expressed in the pituitary and hGH-V only in the placenta [2-4].

The hGH-N gene (formed by five introns and four exons) encodes a 22 kDa isoform with the presence of the five introns, and the 20 kDa isoform that occurs through the mRNA alternative splicing of intron 4. Although this hormone is present in a rather heterogeneous mixture in the body, a clear consensus exists towards that the 22 kDa isoform, consisting of 191 amino acids, is the predominant variant both in the pituitary gland and in circulation, while the 20 kDa form represents only de 5-10 % of the total growth hormone in the pituitary [3;5]. Once in circulation, approximately 50% of hGH forms a complex with the hGH binding protein

(GHBP), the proteolytic product of GH receptor, increasing the protein half-life (from 7 minutes for the free hGH versus 27 minutes the bound version [6]). These complexes between hGH and GHBP are thought to act as a hGH repository in blood during the interpulse interval [7] to provide a sustained concentration of hGH and thus partially compensate for the pulsatile secretion from the pituitary. Proteolytic cleavage of the 22-kDa protein allegedly originates hGH variants of 5 and 17 kDa [8;9] although attempts to reproduce these findings have not been successful [10]. A list of hGH variants and their concentrations in blood circulation is provided in table 1.

Growth hormone is released from the anterior pituitary gland in a pulsatile manner resulting in very stable low basal levels that are abruptly interrupted by bursts of secretion [11]. These secretory bursts are stimulated by many factors. Fasting and sleep markedly amplify the intensity of hGH peaks as well as the frequency [11]. Exercise also increases the hGH secretion and a direct relationship to the body temperature was established [12]. Growth hormone is negatively correlated with age and body fat percentage, and also accounts for gender differences, where women have larger daytime hGH serum concentrations and greater 24-hour hGH secretion [13] than men.

Table 1. Percentages of the different hGH variants in blood circulation. Reproduced with a copyright license obtained from Rightslink®, (Baumann, 2009).

Approximate mean distribution of pituitary hGH isoforms in human blood 15–30 min after a secretory pulse. Monomeric hGH 22 kDa hGH Free 22% Bound to high affinity GHBP 21%

Bound to low affinity GHBP (α2-macroglobulin) 2% Total 22 kDa hGH 45% 20 kDa hGH Free 2% Bound to high affinity GHBP 0.5%

Bound to low affinity GHBP (α2-macroglobulin) 2% Total 20 kDa hGH 5% Acidic hGH (desamido-, acylated and glycosylated hGH) Total acidic forms (bound fractions unknown) 5% Dimeric hGH 22 kDa hGH Dimers Non-covalent dimers / Disulphide dimers 14% / 6% Total 22 kDa GH dimers (bound fractions unknown) 20% 20 kDa hGH Dimers Non-covalent dimers / Disulphide dimers 3% / 2% Total 20 kDa hGH (bound fractions unknown) 5% Acidic hGH Dimers (desamido-, acylated and glycosylated hGH) Non-covalent dimers / Disulphide dimers 1.5% / 0.5% Total acidic hGH dimers (bound fractions unknown) 2% Oligomeric hGH (trimer-pentamer) 22 kDa hGH Oligomers Non-covalent oligomers 7% Disulphide oligomers 3% Total 22 kDa hGH oligomers (bound fractions unknown) 10% 20 kDa hGH Oligomers Non-covalent oligomers / Disulphide oligomers 1% / 0.5% Total 20 kDa hGH oligomers (bound fractions unknown) 2% Acidic hGH Oligomers (desamido-, acylated and glycosylated hGH) Non-covalent oligomers / Disulphide oligomers 1% / 0.5% Total acidic hGH oligomers (bound fractions unknown) 2% Fragments (12, 16 and 30 KDa immunoreactive species) variable

1.2 Growth hormone biological functions

Growth hormone exerts its physiological effect both indirectly and directly; the former occurs through stimulation of insulin- like growth factor I (IGF-I) production and its serum binding proteins (IGFBP-1 to IGFBP-6) [1]. It was even thought that growth hormone mainly acted in the liver stimulating the production of IGF-I but new evidences support the concept that growth hormone may act directly on different tissues as liver, muscle, skin, lung, bone, or cartilage to induce metabolic changes. Moreover, hGH can also stimulate IGF-I in an autocrine or paracrine fashion. Together, hGH and IGF-I then act to promote growth [14]. Most of the anabolic effects of hGH are related to the increase in protein synthesis, skeletal muscle growth, chondroblasts and osteoblasts proliferation, and synthesis of type-I collagen, and linear growth are mediated by IGF-I [1;15]. Otherwise, hGH directly stimulates lipolysis, glucose uptake, and expression of hepatocyte and other growth factors [1]. Both hGH and IGF-I have auto-inhibitory effects upon their own release [16].

It has been concluded that hGH is essential for a proper embryonic development as hGH knockout mice suffer from enteric nervous system abnormalities that are severe enough to compromise the support of life. Despite this fact, foetal progress is possible with only small amounts of this hormone

5 as observed with deficient hGH production or impaired hGH assimilation life [17].

Figure 1: Biological functions of hGH. After secretion, hGH acts directly over several tissues but it also acts indirectly mainly due to the secretion of IGF-I (but also via P-III-P, osteocalcine, etc.). High levels of hGH and IGF-I produce an inhibitory feedback that reduces the secretion of the former.

Growth hormone deficiency in childhood onset is characterized by short stature and subnormal growth rates for the corresponding bone age, clinical appearance of immature facial features and increased peritruncal fat, abnormal hGH secretion in response to defined stimuli, and low levels of hGH-dependent growth factors such as IGF-I and IGFBP-3 [18;19]. In adults, growth hormone deficiency often develops after removal of pituitary tumour, following trauma to the head, or in association with other autoimmune endocrinopathies [20]. In both, children and adults, deficiency of hGH affects body composition, skeletal mineralization, cardiac function, lipoprotein distribution, and quality of life [21]. A schematic overview of major hGH biological functions is included in Figure 1.

1.3 Growth hormone secretion

Until 1977 it was thought that growth hormone release was only controlled by two hypothalamic mechanisms: stimulated by GH-releasing hormone (GHRH) and inhibited by somatostatin also known as growth hormone-inhibiting hormone (GHIH) or somatotropin releases-inhibiting factor (SIRF). These peptide hormones act through the receptors on the somatotrophs cell surface; GHRH binds to the GHRH receptor (GHRH-R) and somatostatin to the somatostatin

7 receptor (SSTR) [22]. Following several publications by Bowers and his coworkers [23-25], an additional release mechanism was proposed. In these papers, the authors described the synthesis of opioid peptide derivates in an attempt to develop new compounds that would be more potent and less addictive but, finally, they discovered that a synthetic Met5-enkephalin peptide (Tyr-dTrp-Gly-Phe-Met-

NH2) had in vitro hGH-releasing activity on the pituitary of rats. Additionally, they observed that opiate peptides and analogues (like morphine) that are well established to release hGH in vivo had no activity on pituitary cells in vitro, indicating that these compounds have an hypothalamic site of action while the Met5-enkephalin appeared to act directly in the pituitary.

1.4 Growth hormone secretagogues

Based on the structure of these new GH-releasing compounds that Bowers and coworkers identified, as well as the subsequent modelling studies, other compounds were developed leading to the nowadays known growth hormone secretagogue (GHS) family, which involves molecules that are apparently totally unrelated from a structural perspective; peptidic as GH-releasing peptide 6 (GHRP-6) or hexarelin and non-peptidic like the L-692,429 or the potent MK-0677.

All of them share the capability to release growth hormone from the cultured pituitary cells [26-28]. Structures of some growth hormone secretagogues (GHS) can be observed in figure 2.

Figure 2. Structures of some GHS. MF, molecular formula; MM, molecular mass.

Subsequently, researchers investigated the mechanisms of GHS action observing that, while GHRH induced the hGH secretion throughout an increase of cAMP, the newly developed secretagogues triggered the intracellular Ca2+ levels in order to achieve the hGH releasing activity.

Figure 3: Two different hGH secretion pathways in a somatotroph cell. While the GHRH/Somatostatin way promotes the hGH secretion via the modulation of cAMP, GHS acts through the mobilization of intracellular calcium via the GHS-R1a.

At this point, the fact that some synthetic compounds were able to promote the secretion of the growth hormone from the somatotroph cells using the calcium mobilization entailed the 10 presence of an hitherto unknown specific receptor, different from the GHRH-R, in these cells and, in assumption that it was not an orphan receptor, an endogenous ligand for this new receptor should exist. See figure 3.

1.5 Growth hormone secretagogue receptor

In 1996, the receptor for the second GH-releasing mechanism was cloned [26;29]. In their study the authors concluded that two different classes of the GHS receptor are encoded by the same gene. These two classes of the growth hormone receptor were named GHS-R1a and GHS-R1b. The GHS-R1a variant appeared to be a 366 amino acid polypeptide composed of 7 transmembrane domains while the 1b variant was a 289 amino acid long polypeptide with only five predicted transmembrane domains. Both variants of the same GHS receptor belong to the family of G protein- coupled receptors. The authors also showed that the 1a variant is active under the secretagogue addition (MK-0677 100nM; GHRP-6 1µM) but not the 1b variant. For the latter until this day no function has been identified. Finally, it was also corroborated that the 1a receptor is only activated by growth hormone secretagogues but not GHRH, confirming the new growth hormone secretion pathway.

GHS-R1a is mainly expressed in the pituitary gland [30] but is also expressed, at a lower level, in a few other tissues such as: thyroid, pancreas, spleen, myocardium, and adrenal gland. On the other hand, the 1b variant is widespread, entailing a yet unknown physiological significance [31].

1.6 Endogenous GHS: Ghrelin

It was not until 1999 that the endogenous ligand for the GHS- R1a receptor was discovered in the rat stomach [32]. Using a stable CHO cell line, expressing the GHS-R1a, and monitoring the intracellular calcium levels when a rat tissue extract was added to the media, the authors observed that the higher mobilization of calcium was found in the stomach sample. Purifying this extract they finally isolated a 28 amino- acid long peptide that was able to provoke the calcium mobilization on its own and they named this peptide “ghrelin”. The etymology is also unique, because “ghre” is the root of the word “grow”, but ghrelin evidently also connotes GH release. Moreover, other studies elucidate that ghrelin is mainly produced in the stomach [32] specifically by the enteroendocrine X/A-like cells [33] but it has also been found in other tissues [34;35].

As stated, ghrelin resulted to be a 28 amino-acid long peptide (GSSFLSPEHQKAQQRKESKKPPAKLQPR), but when

Kojima and coworkers synthesized the same sequence and compared the calcium mobilization due to the presence of purified or synthetic ghrelin observed that the synthetic ghrelin produced no response. Meticulous analytical studies led to the conclusion that serine 3 of the endogenous ghrelin has a unique post-translational modification consisting of an eight-carbon fatty-acid chain that is essential for its interaction with the GHS-R1a [32] (See figure 4). Nuclear magnetic resonance (NMR) analysis was used to determine the tertiary structure of ghrelin, which exhibits random coil behaviour in aqueous solution [36] but new evidences indicates the presence of a possible α-helical region between Glu8 and Lys20 [37].

Figure 4: Primary structure of human ghrelin. In Ser3, the unique octanoyl modification can be observed.

In the bloodstream, ghrelin also occurs in as the non-acylated isoform (des-octanoyl-ghrelin) being its concentration much higher [38;39] than that of ghrelin. Posterior, in 2007, it was confirmed that the des-octanoyl-ghrelin is also capable to activate the GHS-R1a but it requires much higher concentrations than ghrelin (about 1000 times higher) [40].

Ghrelin is highly conserved across several species [41]. The ghrelin gene is located on chromosome 3p25-26 comprising five exons and four introns. An alternative splicing process in the transcript, from exon 2 to exon 5, produces the main form of human ghrelin mRNA. The translation of the main mRNA gives the prepro-ghrelin, a 117 amino acid peptide, which after the cleavage of the signal peptide becomes the pro- ghrelin. This pro-ghrelin is finally modified by the processing protease PC1/3 obtaining as a result the 28 amino acid long peptide.

In humans, it also coexists as two minor transcripts, apart of the mentioned mRNA, one without the codon for glutamine 14 and another one that lacks the exon 4 due to another alternative splicing [42;43]. The lack of the codon for glutamine 14 develops in a des-Q14-ghrelin that resulted to retain the same activity and potency as ghrelin [44]. On the other side, the exclusion of exon 4 from the transcript results in a peptide sequence that still retain the full coding sequence for mature ghrelin thus, the exon-4 deleted prepro-ghrelin will produce the normal pro-ghrelin which is likely to lead to the

14 production of ghrelin although this is not yet clearly defined. A schematic representation of the transcription, translation and post-translation modifications of ghrelin gene can be seen in figure 5.

Is worth mentioning that, together with ghrelin, another peptide is produced due to a different post-translational modification of the prepro-ghrelin molecule, a 23 amino acid long peptide called obestatin [45].

As previously commented, the octanoyl residue in serine 3 is essential for almost all the biological function of ghrelin but it was not until 2008 that different research groups reported how this acylation occurs [46;47]. They reported on the identification and characterization of the human enzyme named GOAT (ghrelin O-acyltransferase), which – as ghrelin – is mainly produced in stomach and pancreas. GOAT is a member of membrane-bound acyltransferases and its function is to incorporate the octanoyl residue to pro-ghrelin, previous to the PC1/3 modification [46;47]. Thus far GOAT has not been able to produce acylation in any other substrate, being specific for ghrelin, resulting in a unique and complex hGH releasing mechanism [48;49].

Not all pro-ghrelin molecules are modified by the GOAT, producing the previously mentioned des-octanoyl-ghrelin. The biological function of this des-octanoyl-ghrelin is still not clearly defined but some studies have been done [38;50] and

15 some opposite effects have been described for ghrelin and des-octanoyl-ghrelin [51].

Figure 5: Main transcription, translation and post-translational modifications of the ghrelin gene. Different alternative splicing pathways produce three prepro-ghrelin peptides containing the signal peptide and

16 ghrelin and obestatin sequences. After posttranslational cleavage and processing by acyltransferase GOAT and protease PC1/3, the mature form of ghrelin is produced.

Further investigations on how ghrelin interacts with the receptor revealed that the entire sequence of ghrelin is not necessary for binding and activation of the receptor. The four N-terminal residues together with acylation in Ser 3 constitute the minimally required sequence [37;52]. On the other side, it was also described how receptor behaves during and after activation, indicating that it suffers desensitization and internalization [53].

1.7 Biological functions of ghrelin

Alongside the hGH releasing activity, ghrelin resulted to perform several other functions.

A notable function of ghrelin in the hypothalamus is its appetite-stimulating effect. In animals and in humans, ghrelin administration increases appetite, stimulates food intake and body weight gain. Ghrelin stimulates appetite by acting on the hypothalamic arcuate nucleus, a region known to control food intake. Furthermore, circulating ghrelin levels are increased by fasting and decreased by feeding, suggesting a role in meal initiation [54-56]. Apart from regulating food consumption, ghrelin is somehow involved in lipid metabolism.

Some publications conclude that after binding of ghrelin in hypothalamus, ghrelin receptor activates AMP-activated protein kinase (AMPK), and then it suppresses lipid synthesis [57;58] but, on the other hand, other publications show that chronic delivery of ghrelin leads to body weight gain in rodents, not only through an increased appetite but also by promoting fat storage in white adipose tissue [59;60].

Numerous studies in the last decade suggest that ghrelin has an important role in regulating β-cells function and glucose homeostasis [61-63]. Ghrelin is also expressed in islet α- and ε-cells in the pancreas, where it has also been shown to stimulate insulin secretion [64-66].

Ghrelin also modulates gastric acid secretion and stimulates gastric motility by inducing the migrating motor complex and accelerating gastric voiding. Moreover, ghrelin exerts a gastroprotective effect against stress, ethanol, and cysteamine-induced ulcers [67]. Furthermore, ghrelin has also diverse cardiovascular effects. It inhibits apoptosis of cardiomyocytes and endothelial cells in vitro. Ghrelin might counterbalance inflammation of the cardiovascular system by inhibiting the nuclear factor kappa-light-chain-enhancer of B cells (NF-kB) activation in human endothelial cells. Administration of ghrelin decreases mean arterial pressure without changing the heart rate and it improves cardiac contractility and left ventricular function in chronic heart failure and reduces infarct size [68].

Ghrelin also participates in the regulation of the reproductive function inhibiting testosterone secretion in testis. It may participate in the regulation of gonadotropin secretion and as such influence the timing of puberty [69]. In addition to these functions, ghrelin also modulates cell proliferation in various cell types [70].

Despite the vast number of different situations in which ghrelin appears to be involved, the ablation of ghrelin causes no obvious phenotypes. Ghrelin knockout mice showed normal size, growth rate, food intake and body composition without any pathological changes [71-73]. It is worth to mention that GHS-R knockout mice also presented normal growth and development with no changes in appetite and body composition. These mice lacking the GHS-R do not show the typical rise of GH when ghrelin is administrated, fact that confirms that this receptor is indeed the primary biological relevant ghrelin receptor [74].

In summary, ghrelin appears to be involved in many different physiological processes. However, because of its fairly recent discovery the role of ghrelin in some of these mechanisms has not been fully elucidated.

Meanwhile, obestatin was initially supposed to suppress food intake and promote the digestive motility, being these effects antagonists of ghrelin, and to be the natural ligand of the GPR39 receptor [45;75] but afterwards, these effects have not been reproduced. Neither obestatin resulted to be the 19 natural ligand for such receptor [76] nor ghrelin-antagonist effects were reproduced [77;78]. As such, it is not clear today what the biological function of obestatin is but some studies reported unexpected central actions on drinking and anxiogenic behaviour [79-81].

1.8 Synthetic growth hormone secretagogues

Encouraged by the potential of ghrelin as a therapeutic target, pharmaceutical companies are developing and testing both peptide and non-peptide compounds with pro- or anti-ghrelin activity. However, due to the multifunctional nature of ghrelin, both agonists and antagonists carry considerable potential to cause undesirable effects independent from the targeted primary indication. Interestingly, the concern of the pharmaceutical industry in these secretagogue compounds started prior to the discovery of ghrelin or its natural receptor. Seven years after the discovery of a met-enkephalin with in vitro growth hormone secretion effect, Bowers and coworkers presented GHRP-6, a hexapeptide that was able to elicit a dose-dependent release of hGH in vitro and in vivo in a variety of animal species that was accompanied by an increase of body weight [24;25]. Some years later, two analogues of GHRP-6 were developed [82;83], which were slightly more potent than GHRP-6 in vitro and threefold more

20 effective in vivo. These new compounds, GHRP-1 and GHRP-2, conserve four of the six amino acid of GHRP-6.

With the publication of GHRP-6, Mediolanum Farmaceutici synthesized Hexarelin that presented an enhanced potency and chemical stability with lower toxicity than GHRP-6. Hexarelin resulted to contain a methylation in one of the amino acids of GHRP-6. In a parallel way, Merck screened its compound library for GH releasing stimuli and some nonpeptidic benzolactam molecules displayed a modest GHS activity. Developing these hits further finally the compound named as L-163 191 and posteriorly renamed as MK-0677 or Ibutamoren was synthesized. It was this compound that enabled the characterization and cloning of the GHS-R1a by introducing a radioisotopic label (32S) in the structure. MK- 0677 resulted to be a high-affinity compound, unaffected by peptidases and presented and excellent bioavailability in canine [26;84-86]. Apart from anything else, Merck synthesized ghrelin analogues with the aim of establishing the structural elements necessary to conserve its biological activity [52].

Once the binding and activation of the GHS-R1a mechanism was elucidated, other pharmaceutical industries started their investigation in order to obtain a ghrelin analogue, antagonist or reverse agonist. Already in 2007, more than 50 ghrelin analogues with promising clinical capabilities were reported [28] and this number keeps growing steadily.

Nowadays, only GHRP-2 is commercially available (Kaken Pharmaceutical, Tokyo, Japan) but, may others compounds are being evaluated in human studies indicating that is only matter of time to have a diverse range of GHS available [87- 92].

1.9 hGH and GHS for doping purposes

Already in the ancient Olympic games, athletes have experimented with natural or synthetic compounds in order to improve their results. Records of special diets studied by athletes are reported to be as early as 668 BC [93] but with the advent of modern pharmacology in the 19th century, many athletes began to experiment with cocktails of synthetic drugs to improve strength and/or overcome fatigue.

Growth hormone extraction and purification from human pituitary glands in 1956 lead to the discovery of its anabolic effects. It was shown to promote growth in hypopituitary animals and was soon used to treat children with hypopituitarism [94;95].

The first record of hGH misuse by athletes was in 1982, when Dan Duchaine, considered an expert bodybuilder, published the book “Underground Steroid handbook” [96]. In this book Duchaine explains the hGH benefits for athletes who claimed

22 for an extra muscle gain. Cadaveric hGH was the only source of the hormone until the appearance in 1987 of the recombinant version of growth hormone (rhGH) [97]. The relationship of cadaveric hGH and Creutzfelt-Jacob disease led to its withdrawal from the market in 1985 although supplies of pituitary-derived hGH continue to be available on the black market. Thus, when rhGH appeared, it made the usage of this hormone easier and safer.

The most famous case of rhGH abuse in professional athletics came to light in 1988 following Ben Jonson’s amazing win in the 100 m final at the Olympic Games in Seoul. His was subsequently disqualified when stanazolol was detected in his urine and he finally admitted that he had taken rhGH in addition to anabolic steroids [98;99].

Although initially advocated for strength disciplines, endurance athletes are also attracted to rhGH’s lipolytic action and reduced fat mass and in 1988 a large quantity of rhGH was found in a team car at the Tour de France [100]. Furthermore, the rhGH black market is no longer limited to elite athletes as nowadays a simple Internet search brings hundreds of sites where nutritional supplements containing claimed hGH (however often fake) could be bought.

1.10 rhGH anti-doping detection

Until the 2004 Olympic Games in Athens, human growth hormone was considered to be undetectable and it is because of that that the prevalence of rhGH misuse in sport was not easy to estimate. The short half-life of hGH, its fluctuating secretion pattern and its low excretion ratio made the detection of exogenous GH a difficult task for scientists. The amino acid sequence of the recombinant molecule is identical to the major 22 kDa isoform secreted by the pituitary gland. In most of the cases, there is no possibility of using a post-transcription or post-translational modification of the molecule to find out the difference between the recombinant and the natural forms.

Detection tests based on quantifying the total amount of the hormone in urine had some clear limitations, the large influence of the process of renal excretion on the concentration measured in urine [101] and the lack of discrimination and specificity of the result, made these tests less promising than a blood test.

Two main strategies were followed to detect rhGH doping using blood as the biological fluid leading to the indirect and the direct approach.

The indirect approach is based on the capacity to determine the endogenous variability of several hGH dependent factors

24 as IGF-I, IGFBPs or some markers of the bone turnover obtaining a database of normal ranges of the concentration of these factors. This may lead to establishment of cut-off levels of so-called abnormal values outside the normal constellation of measure and correlate these numbers individually or in combination with the administration of exogenous growth hormone. The advantage of an indirect approach to target rhGH misuse is certainly that these biological factors are less variable than hGH itself.

This approach was investigated in mid 1990’s by an international group of endocrinologists [102-104] lead by Prof. Sonksen. Scientists focused on the variation of hGH secondary elements during or after exercise, considering IGF- I, IGFBP-3, P-III-P, or osteocalcine as good biological markers. These markers showed slight but significant changes after acute exercise what, together with the fact that the inter-individual variability in the reaction to rhGH administration, made this approach vulnerable in front of a court as an absolute proof of doping. When the project was reported in 1999, IOC (International Olympic Committee) recognized that further research was needed to ensure that the test worked in non-Caucasian ethnic groups and the test was not affected by injury. Finally, IOC rejected funding further research and the project was discontinued for a while.

Nowadays, several research groups have taken over the project conserving its philosophy, focussing specifically on

IGF-I and P-III-P, including several extra factors in order to finally achieve the initial objective [102;103;105-109]. Apart from the global database, this approximation could lead to an individual biological follow-up (endocrine biological passport), where serial values of each marker in a given subject may be compared with a statically developed individual cut-off.

On the other hand, the Strassburger-Bidlingmaier group in Munich developed a detection method of rhGH doping [110- 112] based on the so-called direct approach. This methodology is based on the fact that circulating endogenous hGH is actually a group of isoforms with a fairly stable ratio between the components (Table 1). Contrary to the endogenous hGH, the rhGH (recombinant human GH) variant consists of one single species, which is exclusively represented by the native 22-kDa form. Of all the hGH variants that have been described, the two best characterized are the 22- and 20-kDa forms, which are transcribed at a surprisingly constant rate of approximately 9 to 1 (22 vs. 20 kDa) independently of gender, age, exercise and even certain pathologies [113-115].

When the recombinant form is injected in the body, this increase in concentration, for a period of time, yields a proportion of the 22 kDa form in comparison with the total circulating isoforms different from the endogenous situation. Strassburger and co-workers developed the Rec and Pit immunoassays where the first (Rec) is based on a

26 combination of antibodies that are expected to preferentially recognize mainly the unchanged 22-kDa isoform whereas another combination is more sensitive to the presence of pituitary hGH isoforms (Pit). This ratio is known as the Rec/Pit ratio and a threshold limit is established in order to determine if a biological sample is considered negative or positive.

Alternatively, other researchers in Spain and Japan, and a decade later also in Germany developed an antibody specific to the 20-kDa [116-118] for medical purposes, which lead to the development of the 22-kDa:20 kDa ratio. This alternative ratio is nowadays still under development by the Japanese group.

The rhGH isoform test was initially performed during the Athens Olympic Games in 2004 and also in Turin 2006 but yielded no adverse analytical findings, probably due to low window of opportunity of less than 24 h and the fact that most cheats stop administration during competition. During the Beijing Olympic Games in 2008, tests were performed out of competition in order to prevent the athlete from clearing the system beforehand. Despite the 471 analyses for hGH during the Beijing Games, no positive results for this substance were found. Otherwise, the analyses have reported some positives in routine tests, Terry Newton, a British professional rugby player; Patrick Sinkewitz, a German cyclist from the Farnese Vini team or Andrus Veerpalu, double Olympic champion in cross-country skiing from Estonia, are among the 8 doping

27 cases reported up to may 2012 (O Barroso, personal communication).

1.11 GHS anti-doping detection

Nowadays that rhGH doping can be efficiently addressed albeit with a short timeframe (up to 24 h after administration), scientist face up to another challenge, the GHS detection. A single dose of one of those multiple GHS options, could lead to an important rise of circulating endogenous hGH, and depending on the GHS potency and bioavailability it could even be much higher than a single dose of rhGH [89;119- 123].

An artificial rise of rhGH using either endogenous ghrelin or any synthetic secretagogue may interfere significantly in the direct approach for the rhGH detection. The fact that the circulating hGH contains all its isoforms could invalidate the Rec/Pit ratio [90]. Considering this same idea, Okano and coworkers tested the rhGH detection methodology after a GHS administration [123]. This study concludes that after a GHRP- 2 injection, the Rec/Pit ratio is not altered. They also show that if prior to the GHS injection rhGH is administrated, the ratio is also normalized. In conclusion, according to Okano et al., GHS have the capacity to induce same physiological effects as rhGH going unnoticed by actual anti-doping tests or

28 even act as a masking agent for rhGH. These results indicate that as GHS may become an alternative to rhGH for doping purposes, and as such an analytical test is needed for GHS.

Recent works reveal that secretagogues could be misused nowadays by athletes, as some GHS have been detected in nutritional supplements or identified in products confiscated in the black market [124-126]. Thus, the GHS detection test might be introduced as soon as possible as these products are apparently being misused.

1.12 Proposal of a GHS detection test

A direct detection (maybe based on LC-MS) of a growth hormone secretagogue in a biological sample might not be difficult as secretagogues, except ghrelin, are synthetic compounds and the mere presence should be enough for considering an adverse analytical finding. Actually, some efforts have already been made in this direction to detect GHRP-2 in urine [127;128]. The main problem is the existence of a large number and still growing range of different molecules that might be misused and the detection of an unknown is unfeasible. The different nature (peptidic and non-peptidic) of different secretagogues might difficult the detection task and the appearance of new compounds should be associated to an up-to-date database and its analytical

29 methodology. This mentioned database not only should contain the possible GHS but all its metabolites that are not always available. Despite the fact that techniques as LC-MS are fast, sensitive and reliable, an analysis of high number of samples examining for a huge number of potential targets could be very time consuming.

The work described in this thesis is focused on an alternative technique, a universal detection test applicable to urine samples that should detect any possible secretagogue present in such biological fluid even if this secretagogue is unknown. The aim of this project might be accomplished bearing in mind the only property that all growth hormone secretagogue should have, its interaction with the GHS-R1a.

Considering a cell culture expressing the GHS receptor and a labelled ligand for this receptor, the eventual presence of another non-labelled ligand should provoke the competition of both ligand populations for the receptor. Quantification of the bound labelled ligand might be inversely related to the non- labelled ligand concentration. This competition ought to be independent of the nature of the non-labelled ligand making this methodology universal for all GHS.

Such methodology might only indicate the presence of any GHS-R1a but not unveil the particular secretagogue. So, the proposed test is initially headed to be a screening step and might be followed by a confirmation analysis.

Thus, this thesis is focused on the development of such methodology, optimizing the test for urine samples and eventually, analyse a batch of real dosed human urines for determining its suitability as a possible usable test.

1.13 References

1. Le RD, Bondy C, Yakar S, Liu JL, Butler A. The somatomedin hypothesis: 2001. Endocr Rev 2001; 22(1):53-74.

2. Kopchick JJ, Andry JM. Growth hormone (GH), GH receptor, and signal transduction. Mol Genet Metab 2000; 71(1-2):293-314.

3. Baumann G. Growth hormone heterogeneity in human pituitary and plasma. Horm Res 1999; 51 Suppl 1:2-6.

4. Lacroix MC, Guibourdenche J, Frendo JL, Muller F, Evain-Brion D. Human placental growth hormone--a review. Placenta 2002; 23 Suppl A:S87-S94.

5. Baumann GP. Growth hormone isoforms. Growth Horm IGF Res 2009; 19(4):333-340.

6. Laursen T. Clinical pharmacological aspects of growth hormone administration. Growth Horm IGF Res 2004; 14(1):16-44.

7. Frystyk J, Andreasen CM, Fisker S. Determination of free growth hormone. J Clin Endocrinol Metab 2008; 93(8):3008-3014.

8. Singh R, Agrawal JK, Kumar V, Raman RD. Growth hormone and diabetic retinopathy. Indian J Ophthalmol 1983; 31 Suppl:1049-1052.

9. Sinha YN, Jacobsen BP. Human growth hormone (hGH)-(44-191), a reportedly diabetogenic fragment of hGH, circulates in human blood: measurement by radioimmunoassay. J Clin Endocrinol Metab 1994; 78(6):1411-1418.

10. Such-Sanmartin G, Bosch J, Segura J, Gutierrez- Gallego R. Generation of 5 and 17 kDa human growth hormone fragments through limited proteolysis. Growth Factors 2009; 27(5):255-264.

11. Veldhuis JD, Bowers CY. Human GH pulsatility: an ensemble property regulated by age and gender. J Endocrinol Invest 2003; 26(9):799-813.

12. Jorgensen JO, Krag M, Kanaley J, Moller J, Hansen TK, Moller N et al. Exercise, hormones, and body temperature. regulation and action of GH during exercise. J Endocrinol Invest 2003; 26(9):838-842.

13. Stokes K. Growth hormone responses to sub-maximal and sprint exercise. Growth Horm IGF Res 2003; 13(5):225-238.

14. Green H, Morikawa M, Nixon T. A dual effector theory of growth-hormone action. Differentiation 1985; 29(3):195-198.

15. Casanueva FF. Physiology of growth hormone secretion and action. Endocrinol Metab Clin North Am 1992; 21(3):483-517.

16. Giustina A, Veldhuis JD. Pathophysiology of the neuroregulation of growth hormone secretion in experimental animals and the human. Endocr Rev 1998; 19(6):717-797.

17. Hagl CI, Holland-Cunz S, Schafer KH. What do knockout models teach us about the enteric nervous system? Eur J Pediatr Surg 2003; 13(3):170-175.

18. Argente J, Abusrewil SA, Bona G, Chiarelli F, Kelnar CJ, Skordis N. Isolated growth hormone deficiency in children and adolescents. J Pediatr Endocrinol Metab 2001; 14 Suppl 2:1003-1008.

19. Root AW, Diamond FB, Jr. Regulation and clinical assessment of growth hormone secretion. Endocrine 2000; 12(2):137-145.

20. Stavrou S, Kleinberg DL. Diagnosis and management of growth hormone deficiency in adults. Endocrinol Metab Clin North Am 2001; 30(3):545-563.

21. Drake WM, Howell SJ, Monson JP, Shalet SM. Optimizing gh therapy in adults and children. Endocr Rev 2001; 22(4):425-450.

22. Whitfield GK, Jurutka PW, Haussler CA, Haussler MR. Steroid hormone receptors: evolution, ligands, and molecular basis of biologic function. J Cell Biochem 1999; Suppl 32-33:110-122.

23. Bowers CY, Chang J, Momany FA, Folkers K. Effect of enkephalins and enkephalins analogs on release of pituitary hormones in vitro. Mol Endocrinol 1977;287- 292.

24. Bowers CY, Momany F, Reynolds GA, Chang D, Hong A, Chang K. Structure-activity relationships of a synthetic pentapeptide that specifically releases growth hormone in vitro. Endocrinology 1980; 106(3):663-667.

25. Bowers CY, Momany FA, Reynolds GA, Hong A. On the in vitro and in vivo activity of a new synthetic hexapeptide that acts on the pituitary to specifically release growth hormone. Endocrinology 1984; 114(5):1537-1545.

26. Patchett AA, Nargund RP, Tata JR, Chen MH, Barakat KJ, Johnston DB et al. Design and biological activities of L-163,191 (MK-0677): a potent, orally active growth hormone secretagogue. Proc Natl Acad Sci U S A 1995; 92(15):7001-7005.

27. Guerlavais V, Boeglin D, Mousseaux D, Oiry C, Heitz A, Deghenghi R et al. New active series of growth hormone secretagogues. J Med Chem 2003; 46(7):1191-1203.

28. Moulin A, Ryan J, Martinez J, Fehrentz JA. Recent developments in ghrelin receptor ligands. ChemMedChem 2007; 2(9):1242-1259.

29. Howard AD, Feighner SD, Cully DF, Arena JP, Liberator PA, Rosenblum CI et al. A receptor in pituitary and hypothalamus that functions in growth hormone release. Science 1996; 273(5277):974-977.

30. Muccioli G, Ghe C, Ghigo MC, Papotti M, Arvat E, Boghen MF et al. Specific receptors for synthetic GH secretagogues in the human brain and pituitary gland. J Endocrinol 1998; 157(1):99-106.

31. Papotti M, Ghe C, Cassoni P, Catapano F, Deghenghi R, Ghigo E et al. Growth hormone secretagogue binding sites in peripheral human tissues. J Clin Endocrinol Metab 2000; 85(10):3803-3807.

32. Kojima M, Hosoda H, Date Y, Nakazato M, Matsuo H, Kangawa K. Ghrelin is a growth-hormone-releasing acylated peptide from stomach. Nature 1999; 402(6762):656-660.

33. Date Y, Kojima M, Hosoda H, Sawaguchi A, Mondal MS, Suganuma T et al. Ghrelin, a novel growth hormone-releasing acylated peptide, is synthesized in a distinct endocrine cell type in the gastrointestinal tracts of rats and humans. Endocrinology 2000; 141(11):4255-4261.

34. Hattori N, Saito T, Yagyu T, Jiang BH, Kitagawa K, Inagaki C. GH, GH receptor, GH secretagogue receptor, and ghrelin expression in human T cells, B cells, and neutrophils. J Clin Endocrinol Metab 2001; 86(9):4284-4291.

35. Volante M, Fulcheri E, Allia E, Cerrato M, Pucci A, Papotti M. Ghrelin expression in fetal, infant, and adult human lung. J Histochem Cytochem 2002; 50(8):1013- 1021.

36. Silva Elipe MV, Bednarek MA, Gao YD. 1H NMR structural analysis of human ghrelin and its six truncated analogs. Biopolymers 2001; 59(7):489-501.

37. Martin-Pastor M, De CA, Alvarez CJ, Diaz-Hernandez MD, Jimenez-Barbero J, Casanueva FF et al. Interaction between ghrelin and the ghrelin receptor (GHS-R1a), a NMR study using living cells. Bioorg Med Chem 2010; 18(4):1583-1590.

38. Yoshimoto A, Mori K, Sugawara A, Mukoyama M, Yahata K, Suganami T et al. Plasma ghrelin and desacyl ghrelin concentrations in renal failure. J Am Soc Nephrol 2002; 13(11):2748-2752.

39. Hosoda H, Kojima M, Mizushima T, Shimizu S, Kangawa K. Structural divergence of human ghrelin. Identification of multiple ghrelin-derived molecules produced by post-translational processing. J Biol Chem 2003; 278(1):64-70.

40. Gauna C, van de Zande B, van KA, Themmen AP, van der Lely AJ, Delhanty PJ. Unacylated ghrelin is not a functional antagonist but a full agonist of the type 1a growth hormone secretagogue receptor (GHS-R). Mol Cell Endocrinol 2007; 274(1-2):30-34.

41. van der Lely AJ, Tschop M, Heiman ML, Ghigo E. Biological, physiological, pathophysiological, and

pharmacological aspects of ghrelin. Endocr Rev 2004; 25(3):426-457.

42. Seim I, Collet C, Herington AC, Chopin LK. Revised genomic structure of the human ghrelin gene and identification of novel exons, alternative splice variants and natural antisense transcripts. BMC Genomics 2007; 8:298.

43. Gualillo O, Lago F, Casanueva FF, Dieguez C. One ancestor, several peptides post-translational modifications of preproghrelin generate several peptides with antithetical effects. Mol Cell Endocrinol 2006; 256(1-2):1-8.

44. Sato T, Nakamura Y, Shiimura Y, Ohgusu H, Kangawa K, Kojima M. Structure, regulation and function of ghrelin. J Biochem 2012; 151(2):119-128.

45. Zhang JV, Ren PG, Avsian-Kretchmer O, Luo CW, Rauch R, Klein C et al. Obestatin, a peptide encoded by the ghrelin gene, opposes ghrelin's effects on food intake. Science 2005; 310(5750):996-999.

46. Gutierrez JA, Solenberg PJ, Perkins DR, Willency JA, Knierman MD, Jin Z et al. Ghrelin octanoylation mediated by an orphan lipid transferase. Proc Natl Acad Sci U S A 2008; 105(17):6320-6325.

47. Yang J, Brown MS, Liang G, Grishin NV, Goldstein JL. Identification of the acyltransferase that octanoylates ghrelin, an appetite-stimulating peptide hormone. Cell 2008; 132(3):387-396.

48. Kang K, Schmahl J, Lee JM, Garcia K, Patil K, Chen A et al. Mouse ghrelin-O-acyltransferase (GOAT) plays a critical role in bile acid reabsorption. FASEB J 2012; 26(1):259-271.

49. Shlimun A, Unniappan S. Ghrelin o-acyl transferase: bridging ghrelin and energy homeostasis. Int J Pept 2011; 2011:217957. 36

50. Stengel A, Goebel M, Wang L, Tache Y. Ghrelin, desacyl ghrelin and nesfatin-1 in gastric X/A-like cells: role as regulators of food intake and body weight. Peptides 2010; 31(2):357-369.

51. Gauna C, Delhanty PJ, Hofland LJ, Janssen JA, Broglio F, Ross RJ et al. Ghrelin stimulates, whereas des-octanoyl ghrelin inhibits, glucose output by primary hepatocytes. J Clin Endocrinol Metab 2005; 90(2):1055-1060.

52. Bednarek MA, Feighner SD, Pong SS, McKee KK, Hreniuk DL, Silva MV et al. Structure-function studies on the new growth hormone-releasing peptide, ghrelin: minimal sequence of ghrelin necessary for activation of growth hormone secretagogue receptor 1a. J Med Chem 2000; 43(23):4370-4376.

53. Camina JP, Carreira MC, El MS, Llorens-Cortes C, Smith RG, Casanueva FF. Desensitization and endocytosis mechanisms of ghrelin-activated growth hormone secretagogue receptor 1a. Endocrinology 2004; 145(2):930-940.

54. Ariyasu H, Takaya K, Tagami T, Ogawa Y, Hosoda K, Akamizu T et al. Stomach is a major source of circulating ghrelin, and feeding state determines plasma ghrelin-like immunoreactivity levels in humans. J Clin Endocrinol Metab 2001; 86(10):4753-4758.

55. Cummings DE, Purnell JQ, Frayo RS, Schmidova K, Wisse BE, Weigle DS. A preprandial rise in plasma ghrelin levels suggests a role in meal initiation in humans. Diabetes 2001; 50(8):1714-1719.

56. Callahan HS, Cummings DE, Pepe MS, Breen PA, Matthys CC, Weigle DS. Postprandial suppression of plasma ghrelin level is proportional to ingested caloric load but does not predict intermeal interval in humans. J Clin Endocrinol Metab 2004; 89(3):1319-1324.

57. Lopez M, Lage R, Saha AK, Perez-Tilve D, Vazquez MJ, Varela L et al. Hypothalamic fatty acid metabolism mediates the orexigenic action of ghrelin. Cell Metab 2008; 7(5):389-399.

58. Kola B, Hubina E, Tucci SA, Kirkham TC, Garcia EA, Mitchell SE et al. Cannabinoids and ghrelin have both central and peripheral metabolic and cardiac effects via AMP-activated protein kinase. J Biol Chem 2005; 280(26):25196-25201.

59. Davies JS, Kotokorpi P, Eccles SR, Barnes SK, Tokarczuk PF, Allen SK et al. Ghrelin induces abdominal obesity via GHS-R-dependent lipid retention. Mol Endocrinol 2009; 23(6):914-924.

60. Tschop M, Smiley DL, Heiman ML. Ghrelin induces adiposity in rodents. Nature 2000; 407(6806):908-913.

61. Sangiao-Alvarellos S, Cordido F. Effect of ghrelin on glucose-insulin homeostasis: therapeutic implications. Int J Pept 2010; 2010.

62. Iwakura H, Ariyasu H, Li Y, Kanamoto N, Bando M, Yamada G et al. A mouse model of ghrelinoma exhibited activated growth hormone-insulin-like growth factor I axis and glucose intolerance. Am J Physiol Endocrinol Metab 2009; 297(3):E802-E811.

63. Bando M, Iwakura H, Ariyasu H, Hosoda H, Yamada G, Hosoda K et al. Transgenic overexpression of intra- islet ghrelin does not affect insulin secretion or glucose metabolism in vivo. Am J Physiol Endocrinol Metab 2011.

64. Date Y, Nakazato M, Hashiguchi S, Dezaki K, Mondal MS, Hosoda H et al. Ghrelin is present in pancreatic alpha-cells of humans and rats and stimulates insulin secretion. Diabetes 2002; 51(1):124-129.

65. Prado CL, Pugh-Bernard AE, Elghazi L, Sosa-Pineda B, Sussel L. Ghrelin cells replace insulin-producing 38

beta cells in two mouse models of pancreas development. Proc Natl Acad Sci U S A 2004; 101(9):2924-2929.

66. Kojima M, Kangawa K. Ghrelin: more than endogenous growth hormone secretagogue. Ann N Y Acad Sci 2010; 1200:140-148.

67. Peeters TL. Ghrelin: a new player in the control of gastrointestinal functions. Gut 2005; 54(11):1638-1649.

68. Isgaard J, Johansson I. Ghrelin and GHS on cardiovascular applications/functions. J Endocrinol Invest 2005; 28(9):838-842.

69. Garcia MC, Lopez M, Alvarez CV, Casanueva F, Tena- Sempere M, Dieguez C. Role of ghrelin in reproduction. Reproduction 2007; 133(3):531-540.

70. Korbonits M, Goldstone AP, Gueorguiev M, Grossman AB. Ghrelin--a hormone with multiple functions. Front Neuroendocrinol 2004; 25(1):27-68.

71. Sun Y, Ahmed S, Smith RG. Deletion of ghrelin impairs neither growth nor appetite. Mol Cell Biol 2003; 23(22):7973-7981.

72. Sato T, Kurokawa M, Nakashima Y, Ida T, Takahashi T, Fukue Y et al. Ghrelin deficiency does not influence feeding performance. Regul Pept 2008; 145(1-3):7-11.

73. Wortley KE, Anderson KD, Garcia K, Murray JD, Malinova L, Liu R et al. Genetic deletion of ghrelin does not decrease food intake but influences metabolic fuel preference. Proc Natl Acad Sci U S A 2004; 101(21):8227-8232.

74. Sun Y, Wang P, Zheng H, Smith RG. Ghrelin stimulation of growth hormone release and appetite is mediated through the growth hormone secretagogue receptor. Proc Natl Acad Sci U S A 2004; 101(13):4679-4684. 39

75. Nogueiras R, Pfluger P, Tovar S, Arnold M, Mitchell S, Morris A et al. Effects of obestatin on energy balance and growth hormone secretion in rodents. Endocrinology 2007; 148(1):21-26.

76. Holst B, Egerod KL, Schild E, Vickers SP, Cheetham S, Gerlach LO et al. GPR39 signaling is stimulated by zinc ions but not by obestatin. Endocrinology 2007; 148(1):13-20.

77. Gourcerol G, Tache Y. Obestatin--a ghrelin-associated peptide that does not hold its promise to suppress food intake and motility. Neurogastroenterol Motil 2007; 19(3):161-165.

78. De SB, Thijs T, Peeters TL, Depoortere I. Effect of peripheral obestatin on gastric emptying and intestinal contractility in rodents. Neurogastroenterol Motil 2007; 19(3):211-217.

79. Samson WK, White MM, Price C, Ferguson AV. Obestatin acts in brain to inhibit thirst. Am J Physiol Regul Integr Comp Physiol 2007; 292(1):R637-R643.

80. Samson WK, Yosten GL, Chang JK, Ferguson AV, White MM. Obestatin inhibits vasopressin secretion: evidence for a physiological action in the control of fluid homeostasis. J Endocrinol 2008; 196(3):559-564.

81. Tschop M, Castaneda TR, Joost HG, Thone-Reineke C, Ortmann S, Klaus S et al. Physiology: does gut hormone PYY3-36 decrease food intake in rodents? Nature 2004; 430(6996):1.

82. Bowers CY. GH releasing peptides--structure and kinetics. J Pediatr Endocrinol 1993; 6(1):21-31.

83. Akman MS, Girard M, O'Brien LF, Ho AK, Chik CL. Mechanisms of action of a second generation growth hormone-releasing peptide (Ala-His-D-beta Nal-Ala- Trp-D-Phe-Lys-NH2) in rat anterior pituitary cells. Endocrinology 1993; 132(3):1286-1291. 40

84. Dean DC, Nargund RP, Pong SS, Chaung LY, Griffin P, Melillo DG et al. Development of a high specific activity sulfur-35-labeled sulfonamide radioligand that allowed the identification of a new growth hormone secretagogue receptor. J Med Chem 1996; 39(9):1767-1770.

85. Jacks T, Smith R, Judith F, Schleim K, Frazier E, Chen H et al. MK-0677, a potent, novel, orally active growth hormone (GH) secretagogue: GH, insulin-like growth factor I, and other hormonal responses in beagles. Endocrinology 1996; 137(12):5284-5289.

86. Leung KH, Cohn DA, Miller RR, Doss GA, Stearns RA, Simpson RE et al. Pharmacokinetics and disposition of L-692,429. A novel nonpeptidyl growth hormone secretagogue in preclinical species. Drug Metab Dispos 1996; 24(7):753-760.

87. Frieboes RM, Antonijevic IA, Held K, Murck H, Pollmacher T, Uhr M et al. Hexarelin decreases slow- wave sleep and stimulates the secretion of GH, ACTH, cortisol and prolactin during sleep in healthy volunteers. Psychoneuroendocrinology 2004; 29(7):851-860.

88. Pandya N, DeMott-Friberg R, Bowers CY, Barkan AL, Jaffe CA. Growth hormone (GH)-releasing peptide-6 requires endogenous hypothalamic GH-releasing hormone for maximal GH stimulation. J Clin Endocrinol Metab 1998; 83(4):1186-1189.

89. Piccoli F, Degen L, MacLean C, Peter S, Baselgia L, Larsen F et al. Pharmacokinetics and pharmacodynamic effects of an oral ghrelin agonist in healthy subjects. J Clin Endocrinol Metab 2007; 92(5):1814-1820.

90. Svensson J, Boguszewski CL, Shibata F, Carlsson B, Carlsson LM, Bengtsson BA. The effect of treatment with the oral growth hormone (GH) secretagogue MK-

677 on GH isoforms. Growth Horm IGF Res 2003; 13(1):1-7.

91. Svensson J, Monson JP, Vetter T, Hansen TK, Savine R, Kann P et al. Oral administration of the growth hormone secretagogue NN703 in adult patients with growth hormone deficiency. Clin Endocrinol (Oxf) 2003; 58(5):572-580.

92. Zdravkovic M, Olsen AK, Christiansen T, Schulz R, Taub ME, Thomsen MS et al. A clinical study investigating the pharmacokinetic interaction between NN703 (tabimorelin), a potential inhibitor of CYP3A4 activity, and midazolam, a CYP3A4 substrate. Eur J Clin Pharmacol 2003; 58(10):683-688.

93. Yesalis C.E., Bahrke M.S. History of doping in sport. In: Yesalis C.E., Bahrke M.S., editors. Performance- Enhancing Substances in Sport and Exerise. 2002. p. 1-20.

94. LI CH, PAPKOFF H. Preparation and properties of growth hormone from human and monkey pituitary glands. Science 1956; 124(3235):1293-1294.

95. RABEN MS. Treatment of a pituitary dwarf with human growth hormone. J Clin Endocrinol Metab 1958; 18(8):901-903.

96. Duchaine D. Underground Steroid Handbook. Santa Monica: 1982.

97. Salomon F, Cuneo RC, Hesp R, Sonksen PH. The effects of treatment with recombinant human growth hormone on body composition and metabolism in adults with growth hormone deficiency. N Engl J Med 1989; 321(26):1797-1803.

98. Buzzini SR. Abuse of growth hormone among young athletes. Pediatr Clin North Am 2007; 54(4):823-43, xiii.

99. Mackay. How HGH cheats hit the end of the line. Guardian. In press 2004.

100. Abt S. Tour de France steadfast in Ouster of Festina Team. New York Times. In press 1998.

101. Saugy M, Cardis C, Rivier L. Urinary human growth hormone II. Measurements in competition tests: more the expression of the renal function than the endocrinological status. In: Donkie M, Geyer W, Gotzmann A, editors. Recent advances in doping analysis. Köln, Germany: 1995. p. 213-222.

102. Longobardi S, Keay N, Ehrnborg C, Cittadini A, Rosen T, Dall R et al. Growth hormone (GH) effects on bone and collagen turnover in healthy adults and its potential as a marker of GH abuse in sports: a double blind, placebo-controlled study. The GH-2000 Study Group. J Clin Endocrinol Metab 2000; 85(4):1505-1512.

103. Wallace JD, Cuneo RC, Lundberg PA, Rosen T, Jorgensen JO, Longobardi S et al. Responses of markers of bone and collagen turnover to exercise, growth hormone (GH) administration, and GH withdrawal in trained adult males. J Clin Endocrinol Metab 2000; 85(1):124-133.

104. Wallace JD, Cuneo RC, Baxter R, Orskov H, Keay N, Pentecost C et al. Responses of the growth hormone (GH) and insulin-like growth factor axis to exercise, GH administration, and GH withdrawal in trained adult males: a potential test for GH abuse in sport. J Clin Endocrinol Metab 1999; 84(10):3591-3601.

105. Powrie JK, Bassett EE, Rosen T, Jorgensen JO, Napoli R, Sacca L et al. Detection of growth hormone abuse in sport. Growth Horm IGF Res 2007; 17(3):220-226.

106. Such-Sanmartin G, Bosch J, Segura J, Gutierrez- Gallego R. Growth hormone abuse and biological

passport: is mannan-binding lectin a complementary candidate? Clin J Sport Med 2011; 21(5):441-443.

107. Sonksen P. The International Olympic Committee (IOC) and GH-2000. Growth Horm IGF Res 2009; 19(4):341-345.

108. Holt RI, Erotokritou-Mulligan I, McHugh C, Bassett EE, Bartlett C, Fityan A et al. The GH-2004 project: the response of IGF1 and type III pro-collagen to the administration of exogenous GH in non-Caucasian amateur athletes. Eur J Endocrinol 2010; 163(1):45-54.

109. Bosch J, Ueki M, Such-Sanmartin G, Segura J, Gutierrez-Gallego R. Tracking growth hormone abuse in sport: A comparison of distinct isoform-based assays. Anal Chim Acta 2012.

110. Wallace JD, Cuneo RC, Bidlingmaier M, Lundberg PA, Carlsson L, Boguszewski CL et al. Changes in non-22- kilodalton (kDa) isoforms of growth hormone (GH) after administration of 22-kDa recombinant human GH in trained adult males. J Clin Endocrinol Metab 2001; 86(4):1731-1737.

111. Bidlingmaier M, Wu Z, Strasburger CJ. Test method: GH. Baillieres Best Pract Res Clin Endocrinol Metab 2000; 14(1):99-109.

112. Wu Z, Bidlingmaier M, Dall R, Strasburger CJ. Detection of doping with human growth hormone. Lancet 1999; 353(9156):895.

113. Ishikawa M, Yokoya S, Tachibana K, Hasegawa Y, Yasuda T, Tokuhiro E et al. Serum levels of 20- kilodalton human growth hormone (GH) are parallel those of 22-kilodalton human GH in normal and short children. J Clin Endocrinol Metab 1999; 84(1):98-104.

114. Radetti G, Buzi F, Tonini G, Bellone J, Pagani S, Bozzola M. Growth hormone (GH) isoforms following

acute 22-kDa GH injection: is it useful to detect GH abuse? Int J Sports Med 2004; 25(3):205-208.

115. Tsushima T, Katoh Y, Miyachi Y, Chihara K, Teramoto A, Irie M et al. Serum concentration of 20K human growth hormone (20K hGH) measured by a specific enzyme-linked immunosorbent assay. Study Group of 20K hGH. J Clin Endocrinol Metab 1999; 84(1):317- 322.

116. Mellado M, Rodriguez-Frade JM, Kremer L, Martinez- Alonso C. Characterization of monoclonal antibodies specific for the human growth hormone 22K and 20K isoforms. J Clin Endocrinol Metab 1996; 81(4):1613- 1618.

117. Hashimoto Y, Ikeda I, Ikeda M, Takahashi Y, Hosaka M, Uchida H et al. Construction of a specific and sensitive sandwich enzyme immunoassay for 20 kDa human growth hormone. J Immunol Methods 1998; 221(1-2):77-85.

118. He C, Wu M. Detection of doping with recombinant human growth hormone. Bioanalysis 2009; 1(5):953- 965.

119. Murphy MG, Weiss S, McClung M, Schnitzer T, Cerchio K, Connor J et al. Effect of alendronate and MK-677 (a growth hormone secretagogue), individually and in combination, on markers of bone turnover and bone mineral density in postmenopausal osteoporotic women. J Clin Endocrinol Metab 2001; 86(3):1116- 1125.

120. Copinschi G, Van OA, L'Hermite-Baleriaux M, Mendel CM, Caufriez A, Leproult R et al. Effects of a 7-day treatment with a novel, orally active, growth hormone (GH) secretagogue, MK-677, on 24-hour GH profiles, insulin-like growth factor I, and adrenocortical function in normal young men. J Clin Endocrinol Metab 1996; 81(8):2776-2782.

121. Huhn WC, Hartman ML, Pezzoli SS, Thorner MO. Twenty-four-hour growth hormone (GH)-releasing peptide (GHRP) infusion enhances pulsatile GH secretion and specifically attenuates the response to a subsequent GHRP bolus. J Clin Endocrinol Metab 1993; 76(5):1202-1208.

122. Khojasteh-Bakht SC, O'donnell JP, Fouda HG, Potchoiba MJ. Metabolism, pharmacokinetics, tissue distribution, and excretion of [14C]CP-424391 in rats. Drug Metab Dispos 2005; 33(1):190-199.

123. Okano M, Nishitani Y, Sato M, Ikekita A, Kageyama S. Influence of intravenous administration of growth hormone releasing peptide-2 (GHRP-2) on detection of growth hormone doping: growth hormone isoform profiles in Japanese male subjects. Drug Test Anal 2010; 2(11-12):548-556.

124. Thomas A, Kohler M, Mester J, Geyer H, Schanzer W, Petrou M et al. Identification of the growth-hormone- releasing peptide-2 (GHRP-2) in a nutritional supplement. Drug Test Anal 2010; 2(3):144-148.

125. Kohler M, Thomas A, Geyer H, Petrou M, Schanzer W, Thevis M. Confiscated black market products and nutritional supplements with non-approved ingredients analyzed in the Cologne Doping Control Laboratory 2009. Drug Test Anal 2010; 2(11-12):533-537.

126. Henninge J, Pepaj M, Hullstein I, Hemmersbach P. Identification of CJC-1295, a growth-hormone- releasing peptide, in an unknown pharmaceutical preparation. Drug Test Anal 2010; 2(11-12):647-650.

127. Okano M, Sato M, Ikekita A, Kageyama S. Determination of growth hormone secretagogue pralmorelin (GHRP-2) and its metabolite in human urine by liquid chromatography/electrospray ionization tandem mass spectrometry. Rapid Commun Mass Spectrom 2010; 24(14):2046-2056.

128. Thevis M, Wilkens F, Geyer H, Schanzer W. Determination of therapeutics with growth-hormone secretagogue activity in human urine for doping control purposes. Rapid Commun Mass Spectrom 2006; 20(22):3393-3402.

2 Chapter 2:

OBJECTIVES

2.1 Main objective

The main objective is to develop the methodology for the detection of growth hormone secretagogues (GHS) for anti- doping purposes.

Such methodology should fulfil two conditions:

• It must be capable to detect the presence of any GHS, independently of their chemical structure or even if the drug is unknown.

• It should be highly sensitive as concentrations of GHS in biological fluids are foreseen to be in the picomolar range.

2.2 Secondary objectives

To demonstrate that the test can be successfully applied on urine samples, the most used biological matrix for anti-doping testing.

Evaluate the effect of several relevant variables (age, gender, exercise practice…) on the methodology.

To study samples obtained after a real administration of GHS to humans.

3 Chapter 3:

METHODOLOGICAL DEVELOPMENTS

3.1 Principle of competitive binding assays

The competitive binding methodology was chosen to allow addressing any GHS present in a biological fluid. This methodology consists in a simultaneous incubation of two agonists that compete for the binding site of the receptor. At least one of these agonists should be known, present at a constant concentration and it must carry a label in order to monitor and evaluate its final response after incubation. As both agonists compete for the receptor site, the final concentration of the labelled ligand will be inversely proportional to the concentration of the non-labelled agonist. This methodology is mainly used to:

• Determine whether a drug candidate binds to the receptor, permitting to screen thousands of compounds to identify drugs that specifically bind to a known receptor. • Investigate the interaction of low affinity drugs with receptors. Low affinity agonists may not be used as the labelled ligand due to the fast dissociation rate constant yielding them better as the unlabelled competitors. • Demonstrate that a certain drug binds the receptor with the expected potency. This could be crucial in order to demonstrate the presence of a specific receptor in

certain tissue or even elucidate a hitherto unknown receptor. • Determine the number of receptor molecules present at the surface and the affinity of the interaction using the same compound as the labelled and unlabelled ligand, known as the “homologous competitive binding curves”.

Generally, a fixed concentration of labelled ligand is incubated together with increasing concentrations of unlabelled agonist. When the incubation media contains no competing ligand, receptor sites are occupied by the labelled agonist obtaining the highest possible value in the final counting; when the sample contains unlabelled ligand that may displace the labelled analogue, a lower counting is obtained. A general view of the commented methodology can bee seen in figure 1.

A competitive binding assay results in a plot known as the competition curve, where the percentage of total binding is usually plotted against the logarithm of the concentration of unlabelled ligand present in the sample (see figure 1). The top of the curve is a plateau at a value equal to labelled ligand binding in the absence (or concentrations too low to produce a significant displacement) of the competing unlabelled drug; this value is known as total binding. The bottom of the curve is a plateau equal to nonspecific binding, where high concentration of unlabelled agonist provokes the

53 displacement of all labelled ligand from the specific site of the receptor. Sometimes, data is normalized from 100% (total binding) and 0% (non-specific binding); range from 0 to 100% is called specific binding.

Fig 1: Schematic view of a competitive binding methodology. The final radioactivity depends on the concentration of the unlabelled ligand present during the incubation. 54

The concentration of unlabelled drug that results in a residual binding of 50% is called IC50 (Inhibitory concentration 50%).

In other words, the IC50 is the concentration of unlabelled drug that blocks half the specific binding.

The value of IC50 is determined by three factors:

• The equilibrium dissociation constant, expressed in M,

of the unlabelled ligand (Ki). This is the affinity of the

unlabelled drug for the receptor. If the Ki is low, the

affinity is high and therefore, the IC50 would be low. • The equilibrium dissociation constant, expressed in M,

of the labelled drug (Kd). It is the affinity of the labelled ligand for the receptor. It takes more unlabelled drug to

compete with a tightly bound labelled ligand (low Kd)

than with a loosely bound ligand (high Kd). • The concentration of labelled ligand (D) expressed in M. Under non-saturation conditions, the use of a high concentration of labelled drug will take a larger amount of unlabelled ligand to compete for the receptor site.

Ki, Kd, D and IC50 are related in the equation of Cheng and Pruosoff [1],

IC K = 50 i D 1+ Kd

The IC50 value could vary depending on the experimental conditions while Kd and Ki are a defined constant of each ligand.

Considering all these conditions, an unlabelled drug may displace the labelled ligand depending on the concentration and affinity of the labelled agonist and, of course, the concentration of the unlabelled material in the biological fluid.

3.2 Competitive binding in GHS detection

3.2.1 General considerations

In order to render appropriate the competitive binding methodology to the detection of GHS in a biological fluid, several considerations have to be taken into account.

In competitive binding, if it’s known which unlabelled ligand is present, binding results could be interpolated to a standard competition curve for that particular compound to determine its concentration (an example can be seen in Chapter 5). However, as the identity of the putative GHS in the biological fluid will presumably be unknown, the proposed technique could determine the presence but not its nature nor concentration.

As commented, the affinity (Ki) of each unlabelled GHS will define the IC50 of such secretagogue. As the IC50 is the concentration of the unlabelled GHS that block 50% of the specific binding, the detection for each compound will basically depend on the affinity and the concentration of each GHS. For example, a sample containing a high affinity secretagogue (low Ki) will require less concentration of GHS than another with a low affinity (high Ki).

The competitive binding methodology requires in-depth optimization depending on the purpose.

3.2.2 Initial competitive binding methodology

Virtually all protocols for radioactive binding are very similar, consisting in three clearly differentiated steps: incubation, separation and counting. From these three steps, the separation (where the labelled ligand that has bound to the receptor is isolated from the ligand excess) is usually the most complicated. This separation is usually done by a simple filtration of the sample, because unbound ligands are not retained in the filter while membranes or cells do [2-6].

The free/bound ligand separation was optimised as a function of the results obtained (vide infra). Initially, samples, after being incubated, were filtered under vacuum over pre-soaked

GF/C Whatman filters with 0.5% polyethyleneimine and washed with 5 ml ice-cold wash buffer (50 mM Tris, pH 7.4). Filters were finally disposed into an Eppendorf-type tube and counted in the gamma-counter.

After experiencing problems with this procedure, an alternative separation protocol was put together (see 3.2.5.3 and Chapter 4) that, based on centrifugation, permitted reducing significantly the disparity of values between the replicates. This alternative also resulted to be cleaner and produced less radioactive waste material.

3.2.3 Labelled ligand

When considering which secretagogue could be used as the labelled ligand, several options where contemplated based on the published literature. Basically three different options where considered, ghrelin, MK-0677 and Hexarelin. Among these options, only ghrelin could be purchased both unlabelled as well as carrying different detectable groups, for example: fluorescence, radioactivity or biotinylation (for potential conjugation to avidin).

The choice was made to label with radioactive iodine (125I) ghrelin due to the intrinsic high sensitivity. This ghrelin is identical to the endogenous ghrelin but carries a 125-iodine

58 atom in the His 9. This labelling was described not to modify the affinity of ghrelin for the receptor [5;7].

3.2.3.1 [125I-His9]-ghrelin

Unlabelled ghrelin was iodinated in-house at the facilities of the Institute of Advanced Technology (IAT, Barcelona, Spain). Labelling of ghrelin was achieved using the chloramine-T method as this chemistry addresses free tyrosyl or histidyl groups available [8] and was already successfully used for ghrelin [7]. Considering the amino acid sequence of ghrelin it was anticipated that only one iodine atom per ghrelin molecule could be incorporated, as only a histidine in position 9 is available while no tyrosine residues are present.

Ghrelin iodination was carried out at room temperature by mixing 2 µl of peptide suspended in 40 µl of 50 mM phosphate buffer at pH 8.2 with 50 µg (2 µl) of chloramine-T in the same phosphate buffer and 2 µl of radioactive sodium iodine (Na125I) (Amersham Biosciences, Barcelona, Spain), corresponding to approximately 0.5 mCi. The reaction time was fixed to 2 minutes and stopped by adding 2 µl (100 µg in the same phosphate buffer) of sodium metabisulphite.

Afterwards, the radiolabelled peptide was purified over a C4 HPLC column (100 C4-tracer, 5 µm, 25x0.46 Kromasil) (BC Aplicaciones analíticas, Esplugues de Llobregat, Spain) using

59 trifluoro acetic acid (TFA) 0.1% in milli-Q water (buffer A) and TFA 0.085% in methanol (buffer B) with a gradient from 10% to 90% buffer B over 15 minutes and back total run time 25 minutes to initial conditions. The corresponding chromatogram can be seen in figure 2. The purified products were dried under nitrogen steam and reconstituted in the desired buffer.

Figure 2: Purification of radioactive labelled ghrelin. A) UV detection. The peak corresponding to ghrelin can be observed at around minute 11. B) Radioactivity detection of 125I. Peaks at 5 minutes are attributed to free iodine while peak at around minute 9.5 to ghrelin-bound 125I. 50% of iodine binding is calculated. Different elution time of ghrelin is ascribed to the delay between detectors.

However, despite some successful iodination reactions, the methodology resulted difficult to optimize and afforded lower specific activity than commercially available material rendering the in-house prepared material less sensitive. Thus,

60 eventually, the commercially available radioactive ghrelin from PerkinElmer (Boston, MA, USA) was employed.

3.2.4 Receptor source

HEK-293 cells stably expressing the GHS-R1a where obtained from Dr. R. Smith (Baylor College of Medicine, Houston, TX, USA) through Dr. F. Casanueva (University of Santiago de Compostela, Spain).

The morphological cell properties did not change noticeably during the entire period in which the experimental work was conducted and which are detailed in Chapter 4.

Initially, cells were disrupted following indications described by Guerlavais and coworkers [4]. The procedure to obtain membranes required optimization ending-up in the methodology mentioned below in 3.2.5.2 and in Chapter 4.

3.2.4.1 Corroboration of receptor functionality

To determine the presence of the receptor and its functionality, the intracellular calcium levels were monitored when incubated together with ghrelin. The calcium mobilization was checked using both fluorescence 61 microscopy and flow cytometry. Results obtained by using the fluorescence microscopy technique are described in Chapter 4.

Alike the fluorescence microscopy, flow cytometry required the incorporation of a calcium indicator into the cell cytosol and this was achieved as described in Chapter 4 for the former methodology. After the indicator incorporation, cells were detached from the plate surface with HB. Analyses were performed using a BD LSR Flow Cytometer (Beckton Dickinson, San Diego, USA). After monitoring baseline for the calcium-induced fluorescence during approximately 30 seconds, cell aspiration was stopped, the agonist added, and cell aspiration resumed and continued till 210 seconds.

Flow cytometry resulted to be a much faster methodology than fluorescence microscopy for the verification of the presence and functionality of GHS-R1a. Figure 3 shows the response for four different secretagogues injected to a final concentration of 30 nM. GHS-R1a transfected cells displayed a rapid and transient mobilization of intracellular Ca2+ resulting in a clear-cut fluorescence. All three types of secretagogues produced a similar response indicating that the concentration employed was beyond the saturation point of the receptor (figure 3). In contrast, at the concentration employed no mobilization of intracellular Ca2+ was observed for des-octanoyl ghrelin.

Figure 3: Flow cytometry results. All GHS except for des-octanoyl ghrelin provoked an increment of the intracellular calcium.

Taking into account these results and those from the fluorescence microscopy, it could be corroborated that cultured HEK293 cells expressed the receptor in its functional state. However, remained the possibility of altered receptor functionality due to the membrane-obtaining protocol.

3.2.4.2 Commercial membranes containing GHS-R1a

For the GHS-R1a throughout the years diverse commercial alternatives to our cell culture have emerged and we tested several. Membranes containing GHS-R1a were purchased 63 from Euroscreen (Gosselies, Belgium) and from PerkinElmer (Boston, MA, USA), those membranes were obtained from the disruption of CHO and HEK293 cells overexpressing the GHS-R1a respectively. Incubating 1 µg of protein per sample and adding radioactive ghrelin, the percentage value of radioactivity bound to these membranes (4 and 11% for Perkin Elmer and Euroscreen, respectively) was very similar to that obtained with our membranes (3%) when the same amount of protein was tested.

3.2.4.3 Whole cell binding

When using intact cells fairly high specific binding values were obtained (about 30%) compared to those obtained using membranes, irrespective of the origin of the membranes. These results indicated that the major part of the receptor molecules were damaged during the cell disruption process.

In this case, the free-vs.-bound fraction separation was not carried out by filtration because intact cells can be pelleted through centrifugation (detailed procedure is included in Chapter 4). It is worth mentioning that this new separation technique diminished the disparity of results also because the incubation, separation and both, initial and final, countings of radioactivity takes place in the same reaction vial to minimise loss of material and production of radioactive waste. In

64 contrast, when the separation step is to be carried out by filtration, the handling of the contaminated filters from the vacuum mechanism and transfer to the counting tube results in a major variation as well as inter-sample contamination. In contrast, when centrifugation is used for separation, these contamination problems are avoided.

Considering all benefits that the separation by centrifugation offered, this application was further explored in order to determine whether it could be also applied using membranes. Several literature citations [9] describe the pelleting of membranes (even after freeze and thaw disruption) and this separation procedure resulted also appropriate for the separation step while using membranes (vide infra).

3.2.4.4 Cell disruption procedure optimization

As mentioned above, receptors were damaged during the membrane obtaining protocol yielding low specific binding values. Thus, the freeze and thaw protocol was evaluated and finally changed for a less stringent methodology.

The procedure proposed by Guerlavais et al. [4] consisted in 3 freeze and thaw cycles. After monitoring the specific binding, it was concluded that each cycle diminished the binding in a progressive manner loosing the major part in the initial cycle. Bearing in mind the high specific binding of intact 65 cells, it seemed unwise insisting in using such protocol as only one freeze and thaw cycle led to a substantial loss in sensitivity.

Other approaches were evaluated to obtain membranes without any (or low) receptor damage. As such, a protocol employing osmotic shock was developed. Cells were incubated in deionized water for 10 minutes, centrifuged and resuspended in binding buffer. This protocol yielded good results that were practically identical to those with intact cells. However, this approximation was finally dismissed because resuspending the pellets with binding buffer was rather difficult and laborious due to a filament formation in the sample and this ended with mostly as an inhomogeneous suspension that entailed high results disparity.

Finally, another approach based on ultrasonication was applied and optimized (see Chapter 4). This new disruption process yielded high binding values, even higher than using intact cells (up to 40% specific binding), standardizing by cell number.

From all these cell-disrupting procedures two excellent receptor supports were selected for further evaluation: cells in their intact form or membranes obtained from those cells from the sonication procedure. All the experimental work to determine which platform would be the optimal for our purposes was compiled and published in article presented in Chapter 4. 66

3.2.5 Sample treatment

Throughout the initial phase aimed at optimization of the receptor binding methodology, samples consisted in spiked aliquots into binding buffer. After all aspects of receptor source, labelled ligand source, and overall procedure were adjusted, urine samples were incorporated into the assay.

Initial attempts to introduce a urine sample into the incubation mixture produced extremely low binding values even though no unlabelled competitor was added. Further investigations permitted concluding that salt excess inhibits the proper binding conditions (see Chapter 4). This drawback was avoided by introducing a solid phase extraction step using Oasis HLB cartridges that theoretically may retain acid, basic and neutral compounds (see Chapter 5).

3.3 Concluding statements

A fully functional receptor binding methodology was optimized using a receptor source produced in-house together with a commercial radioactive ligand. As mentioned, two excellent receptor platforms were available (membranes and intact cells) and all differences and particularities of each support were examined, evaluated and discussed so as to determine

67 which better suits to the needs in an anti-doping protocol. This experimental work and discussion was published and it is exposed in Chapter 4. Furthermore, this procedure has been applied in analysing blank and spiked biological samples as well as samples obtained from a GHS excretion study. Exercise effect on endogenous ghrelin excretion and limits of detection for several secretagogues were determined and the methodology evaluated as a putative GHS anti- doping test. Results were also published and can be seen in Chapter 5.

3.4 References

1. Cheng Y, Prusoff WH. Relationship between the inhibition constant (K1) and the concentration of inhibitor which causes 50 per cent inhibition (I50) of an enzymatic reaction. Biochem Pharmacol 1973; 22(23):3099-3108.

2. McKee KK, Palyha OC, Feighner SD, Hreniuk DL, Tan CP, Phillips MS et al. Molecular analysis of rat pituitary and hypothalamic growth hormone secretagogue receptors. Mol Endocrinol 1997; 11(4):415-423.

3. Papotti M, Ghe C, Cassoni P, Catapano F, Deghenghi R, Ghigo E et al. Growth hormone secretagogue binding sites in peripheral human tissues. J Clin Endocrinol Metab 2000; 85(10):3803-3807.

4. Guerlavais V, Boeglin D, Mousseaux D, Oiry C, Heitz A, Deghenghi R et al. New active series of growth

hormone secretagogues. J Med Chem 2003; 46(7):1191-1203.

5. Moulin A, Demange L, Berge G, Gagne D, Ryan J, Mousseaux D et al. Toward potent ghrelin receptor ligands based on trisubstituted 1,2,4-triazole structure. 2. Synthesis and pharmacological in vitro and in vivo evaluations. J Med Chem 2007; 50(23):5790-5806.

6. Nagamine J, Nagata R, Seki H, Nomura-Akimaru N, Ueki Y, Kumagai K et al. Pharmacological profile of a new orally active growth hormone secretagogue, SM- 130686. J Endocrinol 2001; 171(3):481-489.

7. Katugampola SD, Pallikaros Z, Davenport AP. [125I- His(9)]-ghrelin, a novel radioligand for localizing GHS orphan receptors in human and rat tissue: upregulation of receptors with athersclerosis. Br J Pharmacol 2001; 134(1):143-149.

8. HUNTER WM, GREENWOOD FC. Preparation of iodine-131 labelled human growth hormone of high specific activity. Nature 1962; 194:495-496.

9. Gauna C, van de Zande B, van KA, Themmen AP, van der Lely AJ, Delhanty PJ. Unacylated ghrelin is not a functional antagonist but a full agonist of the type 1a growth hormone secretagogue receptor (GHS-R). Mol Cell Endocrinol 2007; 274(1-2):30-34.

4 Chapter 4:

ON THE USE OF CELLS OR MEMBRANES FOR RECEPTOR BINDING: GROWTH HORMONE SECRETAGOGUES

This chapter has been published as:

Pinyot A, Nikolovski Z, Bosch J, Segura J, Gutierrez-Gallego R. On the use of cells or membranes for receptor binding: growth hormone secretagogues. Anal Biochem 2010; 399(2):174-181.

Analytical Biochemistry 399 (2010) 174–181

Contents lists available at ScienceDirect Analytical Biochemistry

journal homepage: www.elsevier.com/locate/yabio

On the use of cells or membranes for receptor binding: Growth hormone secretagogues

A. Pinyot a, Z. Nikolovski a, J. Bosch a, J. Segura a,b,*, R. Gutiérrez-Gallego a,b a Bioanalysis Group, Neuropsychopharmacology Program, Municipal Institute for Medical Research–Hospital del Mar, 08003 Barcelona, Spain b Department of Experimental and Health Sciences, Pompeu Fabra University, 08002 Barcelona, Spain

4.1 Abstract

Receptor binding techniques have been widely used in different biochemical applications, with isolated membranes being the most used receptor preparation in this type of assays. In this study, intact cells were compared with isolated membranes as receptor support for radioligand receptor binding assay. The growth hormone secretagogue receptor 1a (GHSR-1a) expressed in human embryonic kidney 293 (HEK293) cells was used as a model of G-protein-coupled 72 receptors. Differences between using intact cells in suspension and using isolated membranes were evaluated for different aspects of the receptor binding assay: total binding variations while both receptor preparations remain on ice, modifications in incubation conditions, saturation, and competition using different agonists. Intact cells are more prone to variability. Although under optimized settings both preparations were equivalent, the Kd value for intact cells was three times higher than that using isolated membranes. However, no significant differences were observed in competition assays obtaining practically identical Ki values for all ligands tested. For the GHSR-1a, isolated membranes are the better choice if particular incubation conditions are required (less variability), whereas intact cells yield easy, fast, and physiological conditions for receptor binding assays.

Keywords: Receptor binding. Intact cell. Isolated membranes. Growth hormone secretagogues.

4.2 Introduction

Over the past few decades, a variety of ligand–receptor binding techniques have been used extensively to identify natural receptors, often G protein coupled, to characterize

73 their cellular distribution and study the affinity for the natural ligand and also its pharmacological action. Furthermore, estimates of relative receptor concentration and relative affinities toward agonists, antagonists, and binding modulators can be obtained, and the intracellular effect of each ligand can be monitored. From the initial approaches that employed tissue homogenates to the actual membrane preparation from transfected cells that overexpress the target receptor, the technique has found many applications such as affinity and functionality evaluation of synthetic ligands [1;2], locating target receptors [3;4], unmasking the metabolism of biological processes in neuroscience and endocrinology [5;6], and estimating the biochemical mechanism of drug action in toxicology [7].

To monitor the interaction of a ligand with its receptor, most ligands require labeling, with radioisotopes and fluorescent substituents being the most used. Although both provide the sensitivity required, the former is most akin to the natural ligand because it does not perturb the structural identity. Receptor–ligand assays distinguish three well-defined steps: (i) incubation of the receptor with (radio)ligand (variants may also include nonlabeled endogenous or synthetic ligands, displacing agents, and interaction modulators), (ii) separation of receptor bound from the free ligand, and (iii) quantification of the bound ligand.

Binding studies require the use of receptors in a conformation preserving binding capacity. Different receptor preparations have been used in radioligand receptor assays, including membrane preparation from tissues or cell lines, intact cells, solubilized receptors, receptors in tissue slices, and even living animals. Isolated membranes are definitely the most widely used, mainly because it is easier to disrupt intact cells than to preserve them. There are several protocols published to obtain membranes using different techniques (e.g., sonication, cell lysis by osmotic shock [8], solubilisation of the receptor [9], freeze–thaw [10]). Even using cultured cells, where tissue disaggregation is not needed, cell disruption could result in (partial) loss of the active receptor population or ligand affinity could change to a different extent for different ligands [11;12].

As an alternative to membrane preparations, intact cells may be used. Cells offer particular advantages, especially if biological events cascading from the initial ligand–receptor interaction are to be evaluated. However, depending on the cell type, mild to drastic changes in physiology and morphology may occur as a consequence of environmental changes such as the buffer in which the binding experiments are performed. In addition, when using living entities, processes such as receptor endocytosis and receptor desensitization [13] could affect the results. Extrapolation of binding results obtained with membranes to the intact cell, and from there to the receptor’s natural environment, is done 75 regularly, but rarely are these assumptions substantiated through scientific evidence.

To compare the extents to which intact cells and isolated membranes behave similarly in a radioligand receptor assay, the growth hormone secretagogue receptor 1a (GHSR-1a) was selected. Indeed, three decades ago, when the natural GHSR ligand and receptor were unknown, Bowers and coworkers [14;15] synthesized opioid peptide derivates in an attempt to develop new compounds that would be more potent and less addictive and discovered that some of these peptides had growth hormone-releasing activity. Based on these structures and subsequent modeling studies, other compounds were developed, leading to functional molecules unrelated from a structural perspective (Fig. 1) [16;17]. In 1996, the GHSR was cloned and was located mainly in the anterior pituitary [18]. The GHSR-1a is a typical receptor with seven transmembrane domains and belongs to the family of G-protein-coupled receptors [18]. In 1999, the endogenous ligand (ghrelin) for this orphan receptor was discovered in the rat stomach [19]. This is an example of reverse pharmacology, that is, starting with the synthesis of functional molecules and ending with the discovery of the natural ligand via the discovery of the natural receptor. Ghrelin is a 28-amino-acid peptide that has a unique posttranslational modification in the form of an eight-carbon fatty acid chain linked to Ser3 that is essential for its interaction with the GHSR-1a. Des-octanoyl ghrelin exists in larger amounts but requires high 76 concentrations to interact with the GHSR-1a [20]. Nowadays, this receptor is a prime target of many pharmaceutical companies as over time the spectrum of biological activities in which ghrelin seems to be involved increases [21]. Today, more than 55 different promising agonists (both peptidic and nonpeptidic) have been synthesized and are pipelined to the pharmaceutical industry [22].

In the current study, we have used a selection of representative growth hormone secretagogues and compared the interaction with the receptor using intact human embryonic kidney 293 (HEK293) cells expressing the GHSR- 1a or isolated membranes obtained from the same cells as a model to evaluate the affinity of different agonists in both receptor preparations. The outcome is some insight into the comparative use of intact cells and membranes for receptor binding studies.

4.3 Material and methods

4.3.1 Chemicals

Ghrelin was purchased from RayBiotech (Norcross, GA, USA). Growth hormone-releasing peptide 2 (GHRP-2, pralmorelin), alexamorelin, ipamorelin, tabimorelin

77 hemisulfate, and sermorelin (ligand for growth hormone- releasing hormone [GHRH] receptor and used as a negative control) were kindly supplied by Shinji Kageyama (Tokyo Laboratory, Anti-Doping Center, Mitsubishi Chemical Medience, Japan). MK-0677, hexarelin, and GHRP-6 were kindly supplied by Giampiero Muccioli (University of Torino, Italy). Des-octanoyl ghrelin was obtained from the Department of Experimental and Health Sciences (Pompeu Fabra University, Barcelona, Spain). Radiolabeled ghrelin 125 ([ I-His9]-ghrelin) was purchased from PerkinElmer (Boston, MA, USA). All other chemicals were of the highest grade commercially available.

4.3.2 Cells

HEK293 cells stably expressing the GHSR-1a were obtained from R. Smith (Baylor College of Medicine, Houston, TX, USA) through F. Casaneuva (University of Santiago de Compostela, Spain), and wild-type HEK293 cells were obtained from C. Fillat (CRG, Barcelona, Spain). HEK293 GHSR-1a and HEK293 wild-type cells were cultured in 100- mm dishes in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS), penicillin (100 U/ml), streptomycin (100 U/ml), and 2 mM glutamine (Invitrogen, Paisley, UK).

Fig. 1. Structures of all secretagogues included in this study. Des-octanoyl ghrelin is identical to ghrelin except for the absence of the octanoyl modification on serine-3. MF, molecular formula; MM, molecular mass.

Cells stably expressing the GHSR-1a were subjected to G418 (Invitrogen) selection (800 µg/ml). Cells were grown under a

79 humidified atmosphere of 95% air and 5% CO2 at 37 ºC up to 70–80% confluence.

4.3.3 GHSR-1a activity

For fluorescent microscopy, 30,000 cells were seeded on a circular coverslip in a 4-well plate for 2 days. Cells were washed using isotonic solution (140 nM NaCl, 10 nM Hepes,

5 nM KCl, 5 nM glucose, 1.2 nM CaCl, and 0.5 nM MgCl2, pH 7.4, 300 mOsmol/L) and then incubated with 5 µM fluo-3 (Invitrogen) for 40 min at room temperature. After a second wash, a single coverslip was mounted in a perfusion chamber that was previously sealed with grease. Isotonic perfusion was stopped, cells were put in contact with ghrelin (10 nM in isotonic solution) for 10 min, and then isotonic perfusion was added again to wash off ghrelin. Results were obtained using an Olympus IX70 inverted microscope (Hamburg, Germany) with a 20 x objective (Olympus). Light (excitation λ = 490 nm, emission λ = 520 nm), which was directed toward the cells in the field of view by a 595DR dichromatic mirror (Omega Optical, Brattleboro, VT, USA). Fluorescent images were collected by a digital charge-coupled device (CCD) camera using AquaCosmos software (version 2.5.0.0, Hamamatsu Photonics, Japan). Once the images were collected, using AquaCosmos software, the light emission analysis was

80 performed for each individual cell and finally all values were integrated.

4.3.4 Binding assay

Cultured cells were trypsinized and pelleted by centrifugation at 3000g for 10 min and frozen at -80 ºC in 90% FBS and 10% dimethyl sulfoxide (DMSO, Sigma–Aldrich, St. Louis, MO, USA) at a concentration of 2 x 106 cells/ml for radioligand receptor binding experiments.

4.3.5 Whole cell preparation

One aliquot of 2 x 106 cells was thawed and centrifuged at room temperature for 10 min at 3000g, and the cell pellet was resuspended in 2 ml of freshly prepared binding buffer (25 mM Hepes, 5 mM MgCl2, 1 mM CaCl2, 2.5 mM ethylenediaminetetraacetic acid (EDTA), and 0.4% bovine serum albumin (BSA), pH 7.4).

4.3.6 Membrane preparation

After testing several membrane preparation protocols, based on either freeze–thaw described by Guerlavais and coworkers [10] or other less aggressive techniques (e.g., osmotic shock), the optimized protocol was as follows. One aliquot of 2 x 106 cells was thawed and centrifuged at room temperature for 10 min at 3000g. The cell pellet was resuspended in 1 ml of homogenization buffer (50 mM Tris– HCl and 10% sucrose, pH 7.4) and sonicated for 5 min. Disrupted cells were centrifuged again at room temperature for 10 min at 3000g, and membrane pellet was resuspended in 2 ml of freshly prepared binding buffer (vide supra).

4.3.7 Saturation binding

Aliquots of 50,000 cells or membranes (obtained from 50,000 cells) were incubated in triplicate together with unlabeled ghrelin (at a concentration of 1 µM) and increasing 125 concentrations of [ I-His9]-ghrelin in a final volume of 300 µl. 125 Prior to the addition to the cells or membranes, the [ I-His9]- ghrelin aliquots were counted in a 1470 Wizard gamma scintillation counter (PerkinElmer) to obtain the exact value of radioactivity present in the experiment. Then samples were incubated for 40 min at 25 ºC under continuous agitation. The 82 reaction was stopped by rapid centrifugation at 4 ºC for 5 min at 16,000g. The cell pellet was rinsed once with ice-cold 50 mM Tris–HCl (pH 7.4) with the intention of washing the radiolabeled ligand present in the tube surface and centrifuged again at 4 ºC for 3 min at 16,000g to prevent the pellet from resuspending. Finally, the pellet was measured in the gamma counter. Results were analyzed with GraphPad Prism 5 software (San Diego, CA, USA). Due to the fact that a high fraction of the added radioligand concentration is in the bound form, Kd and Bmax values cannot be calculated using the classical mathematical equations; therefore, they were calculated considering binding depletion equations [23], where Kd is the concentration of ligand that occupies 50% of the binding sites and Bmax is the density of the receptor in the tissue being studied.

4.3.8 Competition binding

A total of 50,000 cells, or membranes obtained from 50,000 cells, were used in competition binding as described in the saturation binding experiment. However, in this case a fixed 125 concentration of [ I-His9]-ghrelin (15 pM) in each sample was used. Each experiment consisted of a triplicate incubation of a particular concentration of unlabeled competitor (from 1 x 10-12 to 1 x 10-5 M). A triplicate with no

83 unlabeled competitor was also included as a control. IC50 and

Ki for each competitor were also calculated considering binding depletion equations [23]. IC50 is defined as the concentration of an unlabeled drug required to inhibit specific binding of the radioligand by 50%. IC50 values are transformed to inhibition constants (Ki) using the binding depletion equations.

All competition and saturation binding experiments were performed in triplicate at least three times.

4.4 Results

The binding of different growth hormone secretagogues to the GHSR-1a was evaluated for 10 different ligands and two receptor preparations. The recombinant expression and functionality of the receptor was established, the membrane preparation was optimized, the stability of the latter and other factors affecting the binding assay were evaluated, and finally the binding with intact cells or membrane preparations was assessed.

4.4.1 GHSR-1a activity

To corroborate the presence and functionality of the GHSR- 1a in the transfected HEK293 cells, fluorescence microscopy was used to monitor the intracellular calcium mobilization as a consequence of the interaction between ghrelin (10 nM) and the receptor. Wildtype HEK293 cells were used as a control.

The addition of ghrelin to HEK293 GHSR-1a cells (Fig. 2, continuous trace) results in intracellular calcium release, with maximum fluorescence observed in less than 30 s. Within 2 min, the response returned to the basal level. The absence of calcium release in the wild-type HEK293 cells is depicted in the dotted trace in Fig. 2.

4.4.2 Radioligand receptor assays

4.4.2.1 Membrane preparation

Initially, membranes from GHSR-1a HEK293 cells were obtained employing the freeze–thaw protocol described by Guerlavais and coworkers [10], but the specific binding of isolated membranes using this protocol was approximately 30% lower than the specific binding using intact cells.

Fig. 2. Intracellular calcium mobilization analyzed by fluorescent microscopy. Effects due to the addition of 10 nM ghrelin in GHSR-1a HEK293 cells (continuous line) and in wild-type (wt) HEK293 cells (dotted line) are presented.

An alternative, less aggressive procedure was attempted. Cells were lysed using an osmotic shock by resuspending the cells with deionized water (MilliQ). With this protocol, the specific binding increased substantially, reaching values that were approximately 50% higher than those for the intact cells. This approach, although very effective in maintaining the receptor population, yielded a suspension that was hard to homogenize and, thus, difficult to reproduce. The protocol was further elaborated, and cells were eventually lysed by sonication in a Tris–HCl buffer containing sucrose as 86 described in Materials and methods. With this approach, the 50% higher binding with respect to intact cells was preserved and homogeneous suspensions could be produced.

4.4.2.2 Receptor stability over time

Usually, radioligand binding assays include a high number of samples that are processed batchwise for logistic reasons. Thus, most protocols include storage of the receptor preparation on ice for a period of time, and such incubation might modify the specific binding. To evaluate the extent of this phenomenon on isolated membranes and in intact cells, binding studies where preparations were kept on ice for different time intervals prior to the incubation were performed.

As can be observed in Fig. 3, incubation on ice for short periods of time (<40 min) resulted in a 20% decrease of total binding obtained using membranes with respect to a null incubation, whereas in the case of intact cells the decrease was 10% (see inset in figure). Following this period, membrane preparations were relatively stable and binding levels returned to 80–100% of the initial binding. A similar pattern was observed for intact cells, and the total binding reached values up to 150% (Fig. 3). As such, regardless of the receptor matrix employed, a control sample in each batch

87 appears to be essential to compare or compensate for the effects of different experimental conditions.

4.4.2.3 Steady-state binding versus experimental conditions

With intact cells, steady-state binding might be expected to be less stable than using isolated membranes because of their higher dependence on changes in environmental conditions. Therefore, the effects of changes in the standard incubation medium (binding buffer, pH 7.4) and in the standard incubation conditions (40 min at 25 ºC) were evaluated. In general, any change in the whole cell assay led to higher variability on the total binding values (Fig. 4). For example, changing the incubation temperature from 25 to 4 ºC and maintaining the incubation time at 40 min resulted in a rise in total binding for the cell assay but not for the membrane assay.

Maintaining the incubation at 4 ºC but prolonging the incubation for 120 min resulted in no net change for either membranes or cells. In contrast, if the temperature was elevated from 25 ºC to physiological temperature (37 ºC) and the incubation time was kept at 40 min, a substantial decrease in binding was observed for both preparations, although both variability and magnitude were more

88 pronounced with membranes. These effects could be avoided, in turn, by shortening the incubation time to 20 min (Fig. 4).

Fig. 3. Total binding percentage (y axis) versus time maintaining intact cells () and isolated membranes () on ice (x axis) before proceeding with the incubation. All points in the figure are depicted with error bars that represent the standard errors of the mean. The principal figure corresponds to the whole period of time studied. The inset corresponds to the period of time from 0 to 40 min.

When the pH of the incubation was lowered from 7.4 to 7.1, an increase in binding that reached approximately 50% for cells and 20% for membranes could be visualized (Fig. 4). Increasing the pH from 7.4 to 8.0 appears to negatively affect both membranes and cells (where in the latter the large fluctuation may have levelled out a potential effect). The negative effect of elevated salt concentrations was corroborated through the incubation with 100 mM NaCl that caused a substantial decrease (~60–80%) of total binding in both assays.

4.4.3 Comparison of isolated membranes and intact cells

125 4.4.3.1 Saturation binding of [ I-His9]-ghrelin to GHS- R1a

To evaluate the number of functional receptors in each preparation as well as the effects of using living entities, saturation binding was performed. Experiments using increasing concentrations of radiolabeled ligand revealed evidence of specific saturable receptors in both cells and membranes, as can be observed from the specific binding isotherms in Fig. 5.

Fig. 4. Binding as a result of different incubation conditions. Binding values under standard conditions (incubation at 25 ºC for 40 min at pH 7.4) are represented as 100%, and all modifications are referenced to the standard conditions. The modifications with respect to the standard conditions are represented on the x axis. Relative total binding (y axis) is represented using intact cells (A, ) and isolated membranes (B, ). The nonspecific binding is the binding obtained in the presence of GHRP-6 (1 µM) prior to the incubation and is represented in columns. The standard errors of the mean are represented with error bars in all points.

The similar values for Bmax in both cases suggests a similar number of active receptor molecules in both preparations and

91 indicates that both procedures affect the membrane-bound protein in a similar fashion. However, under nonsaturation conditions, the specific binding with intact cells was lower than that with membranes, resulting in a higher Kd value for the cell assay.

125 Fig. 5. Saturation binding experiments of [ I-His9]-ghrelin to GHSR-1a using intact cells () and isolated membranes (). Receptor density in membranes is calculated from initial cells prior to disruption. Standard errors of the mean are represented with error bars in all points.

4.4.3.2 Competition of different agonists for GHS-R1a with radiolabeled ghrelin

The binding of distinct secretagogues to cells and membranes under identical conditions was studied. Nine different nonlabeled secretagogues (both peptidic and nonpeptidic) were employed in a competition setting against 125 [ I-His9]-ghrelin (Fig. 6). The Ki values for all secretagogues were determined for both assay settings, and the data are presented in the table in Fig. 6. As can be observed in the lower right panel of this figure, sermorelin, a GHRH receptor 125 agonist used as a negative control, did not displace [ I-His9]- ghrelin even at the highest concentration. All GHSR-1a 125 agonists employed did displace [ I-His9]-ghrelin from the receptor, and nearly identical curves were obtained with both intact cells and membranes obtained through the optimized protocol (vide supra).

Fig. 6. Competition curves, showing the effect of each competitor 125 concentration (x axis) on the inhibition of [ I-His9]-ghrelin binding, using intact cells () and isolated membranes (). Standard errors of the mean are represented with error bars in all points. All structures are shown in

Fig. 1 except sermorelin (amino acid sequence: H2N- YADAIFTNSYRKVLGQLSARKLLQDIMSR- COOH).

4.5 Discussion

Although receptor binding protocols have been described extensively in the literature [24-26], the vast majority of reports ignore the fact that the receptor should be at the appropriate surface density and with its functional three- dimensional configuration. The protocol used to handle the cells is a critical step that potentially affects the receptor conformation and, as such, its functionality. In our experiments, the preparation of membranes as described by one protocol [10] seemed to result in a partial loss or inaccessibility of the receptor as the specific binding using isolated membranes became lower than that using intact cells. In consequence, a simple but effective protocol needed to be established for our system and resulted in specific binding in the membrane assay higher than that with intact cells (Fig. 5). This higher binding could indicate that with intact cells, receptor endocytosis, receptor desensitization, and/or other biological processes take place, resulting in a lower copy number of active receptor molecules at the surface.

Different conditions have been described in radioligand receptor experiments [3;4;13;20] to evaluate different agonists, antagonists, and ligand modulators. The results of these different conditions are to be always compared with the binding obtained in standard conditions. In these kinds of experiments, it is important that the total binding of test

95 samples remains stable within the experimental time course, especially if not all samples can be processed at once. Maintaining either whole cells or membranes on ice prior to the incubation with the radiolabeled ligand diminished the variability, but still some changes were observed. Surprisingly, both cells and membranes showed an initial decrease with storage time that was later reverted to yield relative binding of up to 150% with intact cells. Following 24 h on ice, the binding decreased again, and it remains to be investigated whether beyond that time the membrane preparations finally stabilize or whether the decline is indicative of a progressive deterioration. In any case, a parallel behavior was observed for both membranes and cells, eliminating potential stability arguments to favor one preparation over the other. In fact, because cells have their biological structure intact, it seems predictable that any change in the incubation medium, incubation time, or temperature should modify the equilibrium in a more significant manner than it does for membranes. This was indeed the case and was corroborated using steady-state stability experiments.

When the incubation temperature was decreased from 25 to 4 ºC in intact cells, a significant increase in binding was observed without increasing the variability of the assay. The same conditions did not change either the absolute binding level or the variability using isolated membranes. It could well be that at 4 ºC biological processes such as internalization and desensitization of the receptors are relented or even 96 completely inhibited [27]. Alternatively, cells could present a more dynamic environment in which the lower temperature restricts the dynamism and alters the association– dissociation rates, resulting in a rise of total binding in intact cells. However, prolonging the incubation for 120 min under identical conditions results in reduced binding with respect to the 40-min incubation for both receptor preparations, and the total binding results are similar to those in standard conditions. Notwithstanding the absolute binding level, the variability of results in the triplicate measurements in intact cells increased. This might indicate that intact cells could require longer incubation times when incubated at low temperatures so as to reach the steady state. At physiological temperature (37 ºC) and 40-min incubation, the total binding decreased in both preparations. Whether this points toward a more rapid action mechanism in the secretagogue–GHSR-1a interaction, where absolute binding levels are less relevant, remains to be clarified because most binding experiments are performed at nonphysiological temperatures. Nevertheless, using incubation times of 20 min and keeping the physiological temperature did not alter the total binding with respect to the control under standard conditions. Thus, it can be assumed that the equilibrium situation is distorted at prolonged incubations and that the dissociation rates prevail over the association rates.

The influence of the pH changes on the stability of the steady state binding was also tested. A rise in pH from 7.4 to 8.0 97 decreased the total binding in both membranes and intact cells, and the variability of the latter measurements was increased. If, in contrast, the pH is decreased from 7.4 to 7.1, both preparations exhibit an increase in total binding, and this effect could be related to the extracellular part of the receptor (common exposure in both preparations). Although changes in pH of the medium can modify the ligand affinity [28] and give different total binding values, the effect on the variability of the results is an important aspect to consider; although incubation at pH 7.1 is not far from physiological conditions and yields approximately 150% binding, the large variation (±9.76%) renders this experimental condition less attractive because it seriously hampers interassay comparisons. Eventually, the addition of NaCl (up to 100 mM) to the binding buffer induced a substantial decrease in the total binding value, probably because of the allosteric modulation by the monovalent ions [29;30].

In summary, any change in the incubation time, temperature, or pH affects the stability of the whole cell binding assay more than that of the membrane binding assay. Consequently, if the given experiment needs to be under particular conditions, isolated membranes could be a better choice. However, considering that for intact cells no membrane preparation process is required (achieving a faster assay, less receptor conformation change, and consequently less receptor loss), this approach could be an appropriate alternative if conditions are optimized, especially at low temperature and with short 98 incubation times. The saturation experiment yielded practically identical Bmax values, suggesting that only a small part of the receptor was lost due to the cell lysis. These results clearly indicate that sonication is less aggressive than the freeze–thaw protocol to disrupt the membranes. Yet with a similar number of receptor molecules, saturation curves yielded Kd values that were higher for intact cells than for membranes. This could be attributable to any of the biological processes that take place in cells such as internalization and desensitization of the receptor [13] or to effector proteins and/or small molecule cofactors (e.g., Ca2+, free amino acids, guanine nucleotides) [31]. Because these processes may affect the association and dissociation constants, they may also affect the steady state, resulting in different total binding values.

In spite of the difference in Kd values, the obtained Ki values from competition experiments were practically identical in both assays and no differences in intraassay variability were 125 present. All nonlabeled ligands displaced [ I-His9]-ghrelin in the concentration range described in the literature [22]. One difference was for sermorelin, which is a GHRH receptor agonist, and it should not interact with the GHSR-1a [32]. These results indicate that in competition experiments, both isolated membranes and intact cells are appropriate, but isolated cells should be more representative of the native state of the receptor. Furthermore, because no membrane

99 preparation protocol is needed using intact cells, the whole radioligand receptor assay becomes faster and easier.

In conclusion, it can be assumed that both methodologies are applicable for a receptor binding assay. Nevertheless, each methodology has its own particularities. The fact that no membrane lysis procedure is needed in the intact cell binding assay permits avoiding loss of receptor, speeds up the protocol, and works under more natural-like conditions. Intact cell assays also resulted less variable to short-term storage on ice prior to the binding assay, resulting in lower intraassay variability in the case of prolonged experiments (e.g., testing of different agonists, modulators, and antagonists). In contrast, isolated membranes should be more applicable to total binding experiments when changes in the medium or incubation conditions are required because membranes tolerate these changes without modifying the intraassay dispersion in a significant manner. As such, both methodologies seem to deliver the same answers, but one should not forget to optimize the protocol for obtaining membranes in each particular case, and this is where intact cells allow a faster, easier, and more reliable comparison.

100

4.6 Acknowledgments

The authors thank the following colleagues from the Biomedical Research Park of Barcelona (PRBB) for their support in the application of different methodologies employed: M. Valverde, C. Fillat, N. Andreu, and O. Fornas. We acknowledge F. Casanueva and R. Smith for HEK293 cells expressing GHSR-1a and acknowledge S. Kageyama and G. Muccioli for the different secretagogues. This work was carried out with the financial support of the U.S. Anti- Doping Agency (USADA), the Ministerio de Ciencia e Innovación (DEP2009- 09717), and the Generalitat de Catalunya (2009SGR00492).

4.7 References

1. Houston D, Costanzi S, Jacobson KA, Harden TK. Development of selective high affinity antagonists, agonists, and radioligands for the P2Y1 receptor. Comb Chem High Throughput Screen 2008; 11(6):410-419.

2. Sweetnam PM, Caldwell L, Lancaster J, Bauer C, Jr., McMillan B, Kinnier WJ et al. The role of receptor binding in drug discovery. J Nat Prod 1993; 56(4):441- 455.

3. Katugampola SD, Pallikaros Z, Davenport AP. [125I- His(9)]-ghrelin, a novel radioligand for localizing GHS

101

orphan receptors in human and rat tissue: upregulation of receptors with athersclerosis. Br J Pharmacol 2001; 134(1):143-149.

4. Muccioli G, Ghe C, Ghigo MC, Papotti M, Arvat E, Boghen MF et al. Specific receptors for synthetic GH secretagogues in the human brain and pituitary gland. J Endocrinol 1998; 157(1):99-106.

5. Frey KA, Albin RL. Receptor binding techniques. In: Crawley C, Gerfen C, McKay R, Rogawski M, Sibley D, Skolnick P, editors. Current protocols in Neuroscience. 1997. p. 1.4.1-1.4.14.

6. Pang SF, Li L, Ayre EA, Pang CS, Lee PP, Xu RK et al. Neuroendocrinology of melatonin in reproduction: recent developments. J Chem Neuroanat 1998; 14(3- 4):157-166.

7. Kathmann M, Flau K, Redmer A, Trankle C, Schlicker E. Cannabidiol is an allosteric modulator at mu- and delta-opioid receptors. Naunyn Schmiedebergs Arch Pharmacol 2006; 372(5):354-361.

8. Yu W, Yuan B, Deng X, He L, Youyi Z, Qide H. The preparation of HEK293 alpha1A or HEK293 alpha1B cell membrane stationary phase and the chromatographic affinity study of ligands of alpha1 adrenoceptor. Anal Biochem 2005; 339(2):198-205.

9. Pomes A, Pong SS, Schaeffer JM. Solubilization and characterization of a growth hormone secretagogue receptor from porcine anterior pituitary membranes. Biochem Biophys Res Commun 1996; 225(3):939-945.

10. Guerlavais V, Boeglin D, Mousseaux D, Oiry C, Heitz A, Deghenghi R et al. New active series of growth hormone secretagogues. J Med Chem 2003; 46(7):1191-1203.

11. Hulme EC, Buckley NJ. Receptor preparations for binding studies. In: Hulme EC, editor. Receptor-ligand 102

interactions: a practical approach. New York: Oxford university press; 1992.

12. Qume M. Overview of ligand-receptor binding techniques. Methods Mol Biol 1999; 106:3-23.

13. Camina JP, Carreira MC, El MS, Llorens-Cortes C, Smith RG, Casanueva FF. Desensitization and endocytosis mechanisms of ghrelin-activated growth hormone secretagogue receptor 1a. Endocrinology 2004; 145(2):930-940.

14. Bowers CY, Momany F, Reynolds GA, Chang D, Hong A, Chang K. Structure-activity relationships of a synthetic pentapeptide that specifically releases growth hormone in vitro. Endocrinology 1980; 106(3):663-667.

15. Bowers CY, Momany FA, Reynolds GA, Hong A. On the in vitro and in vivo activity of a new synthetic hexapeptide that acts on the pituitary to specifically release growth hormone. Endocrinology 1984; 114(5):1537-1545.

16. Smith RG, Cheng K, Schoen WR, Pong SS, Hickey G, Jacks T et al. A nonpeptidyl growth hormone secretagogue. Science 1993; 260(5114):1640-1643.

17. Smith RG, van der Ploeg LH, Howard AD, Feighner SD, Cheng K, Hickey GJ et al. Peptidomimetic regulation of growth hormone secretion. Endocr Rev 1997; 18(5):621-645.

18. Howard AD, Feighner SD, Cully DF, Arena JP, Liberator PA, Rosenblum CI et al. A receptor in pituitary and hypothalamus that functions in growth hormone release. Science 1996; 273(5277):974-977.

19. Kojima M, Hosoda H, Date Y, Nakazato M, Matsuo H, Kangawa K. Ghrelin is a growth-hormone-releasing acylated peptide from stomach. Nature 1999; 402(6762):656-660.

103

20. Gauna C, van de Zande B, van KA, Themmen AP, van der Lely AJ, Delhanty PJ. Unacylated ghrelin is not a functional antagonist but a full agonist of the type 1a growth hormone secretagogue receptor (GHS-R). Mol Cell Endocrinol 2007; 274(1-2):30-34.

21. Korbonits M, Goldstone AP, Gueorguiev M, Grossman AB. Ghrelin--a hormone with multiple functions. Front Neuroendocrinol 2004; 25(1):27-68.

22. Moulin A, Ryan J, Martinez J, Fehrentz JA. Recent developments in ghrelin receptor ligands. ChemMedChem 2007; 2(9):1242-1259.

23. Swillens S. Interpretation of binding curves obtained with high receptor concentrations: practical aid for computer analysis. Mol Pharmacol 1995; 47(6):1197- 1203.

24. Bigott-Hennkens HM, Dannoon S, Lewis MR, Jurisson SS. In vitro receptor binding assays: general methods and considerations. Q J Nucl Med Mol Imaging 2008; 52(3):245-253.

25. Bylund DB, Toews ML. Radioligand binding methods: practical guide and tips. Am J Physiol 1993; 265(5 Pt 1):L421-L429.

26. DeupreeJ.D., Bylund DB. Basic principles and techniques for receptor binding. Tocris Rev. 18. 2002. p. 43.

27. Zoon KC, Arnheiter H. Studies of the interferon receptors. Pharmacol Ther 1984; 24(2):259-278.

28. Gillard M, Chatelain P. Changes in pH differently affect the binding properties of histamine H1 receptor antagonists. Eur J Pharmacol 2006; 530(3):205-214.

29. Clark EA, Hill SJ. Differential effect of sodium ions and guanine nucleotides on the binding of thioperamide and clobenpropit to histamine H3-receptors in rat 104

cerebral cortical membranes. Br J Pharmacol 1995; 114(2):357-362.

30. Horstman DA, Brandon S, Wilson AL, Guyer CA, Cragoe EJ, Jr., Limbird LE. An aspartate conserved among G-protein receptors confers allosteric regulation of alpha 2-adrenergic receptors by sodium. J Biol Chem 1990; 265(35):21590-21595.

31. Kobilka B. Allosteric activation of the CaR by L-amino acids. Proc Natl Acad Sci U S A 2000; 97(9):4419- 4420.

32. Prakash A, Goa KL. Sermorelin: a review of its use in the diagnosis and treatment of children with idiopathic growth hormone deficiency. BioDrugs 1999; 12(2):139- 157.

105

106

5 Chapter 5:

GROWTH HORMONE SECRETAGOGUES: OUT OF COMPETITION

This chapter has been published as

Pinyot A, Nikolovski Z, Bosch J, Such-Sanmartin G, Kageyama S, Segura J et al. Growth hormone secretagogues: out of competition. Anal Bioanal Chem 2012; 402(3):1101-1108.

107

Anal Bioanal Chem DOI 10.1007/s00216-011-5544-8

ORIGINAL PAPER

Growth hormone secretagogues: out of competition

A. Pinyot a, Z. Nikolovski a, J. Bosch a, Gerard Such- Sanmartín a, Shinji Kageyama b, J. Segura a,c, R. Gutiérrez- Gallego a,c a Bioanalysis Group, Neuropsychopharmacology Program, Municipal Institute for Medical Research–Hospital del Mar, 08003 Barcelona, Spain b Anti-Doping Laboratory, Mitsubishi Medience Corporation, 3-30-1 Shimura, Itabashi-ku, JP-Tokyo 174 – 8555, Japan c Department of Experimental and Health Sciences, Pompeu Fabra University, 08002 Barcelona, Spain

5.1 Abstract

Growth hormone secretagogues (GHS) constitute a new GH deficiency treatment increasing exponentially in number and improved potency and bioavailability over the last decade. The growth hormone releasing activity makes these compounds attractive for the artificial improvement of the human sports skills, now that recombinant human growth hormone (rhGH) administration is effectively detected. The 108

GHS family is extremely diverse both in number and chemical heterogeneity and keeps growing continuously. In this paper, a general screening test is proposed. To develop a universal method, the single common property of growth hormone secretagogues has been targeted: their capacity to bind to the GHS receptor 1a (GHS-R1a). Pretreated urine samples have been tested in a competition assay where eventually the GHS presence detached a radiolabelled ligand from the receptor in a dose-dependent manner. Blank urine samples were processed to determine potential age, gender and exercise effects, and to define a threshold beyond which a specimen is considered positive. Samples from a growth hormone releasing peptide 2 (GHRP-2) excretion study corroborated the screening assay applicability with a detection window of approximately 4.5 h, and results were confirmed by comparison with a dedicated LC–MS quantification of the intact compound.

Keywords Receptor binding. Growth hormone secretagogues. Anti-doping.

109

5.2 Introduction

Human growth hormone (hGH) was first extracted and purified from pituitary glands in 1956 [1] and in less than a decade became a therapeutic treatment to promote growth in children with hypopituitarism and to increase vitality or sense of well-being in women with adult hypopituitarism [2]. Due to the anabolic effects of growth hormone, elite athletes found an alternative use as a doping agent, first ever documented in 1982 [3], long before the first recombinant version of hGH (rhGH) became available in 1987.

After the growth hormone–releasing hormone (GHRH) pathway, the growth hormone–secretagogue (GHS) pathway was unveiled through the discovery of the GHS receptor 1a (GHS-R1a) [4;5] first and ghrelin [6] later on. The latter discovery resulted from the finding that some GH-releasing compounds did not act through GHRH [4;5]. Given this reverse pharmacological development most synthetic compounds with ghrelin-like properties are totally unrelated from the structural point of view. While ghrelin is a 28-amino acid long peptide, other GHS are both of peptide and nonpeptide nature, representing one of the most heterogeneous families known to target a single receptor [7]. Thus far only a single compound (growth hormone releasing peptide 2, GHRP-2, Kaken Pharmaceutical, Tokyo, Japan) has been commercialized, but the entire family of molecules for GHS-

110

R1a contains at least 55 structures [7], and this list is steadily growing [8]. Despite the limited number of commercialized substances thus far, reports of black market substances and even nutritional supplements containing GHRPs are frequent [9-11].

As with growth hormone, GHS are susceptible to be misused in sport. Administration leads to elevated hGH levels or could be combined with rhGH as stimulating masking agent [12]. As such, GHS are included in the prohibited list from the World Anti-Doping Agency [13] under the growth hormone releasing factors. Recently, a detection test for GH was implemented which is based on the ratio between the recombinant 22-kDa isoform and endogenous pituitary hGH isoforms (Rec/Pit ratio) [14], and subsequently several athletes were proven rhGH offenders. It is therefore to be expected that, if not already, cheating athletes will seek alternative treatments.

As such, the development of a practical detection test for GHS seems to be essential in order to determine the presence of GHS in a biological sample. Some efforts were made addressing a single secretagogue present [15;16] and recently also on a mixture of eight GHS that were spiked in [17]. The direct detection approach is feasible (e.g. based on LC–MS), but the swiftly growing GHS family, with hitherto unknown structural features, renders a single-entity based approach inefficient.

111

A logical strategy in the development of a general purpose protocol should include the single component that all GHS share: the interaction with GHS-R1a [18]. This strategy is common in the pharmaceutical industry, and nowadays cells overexpressing a wide range of different receptors and ligands, tagged with different labels, are available for screening purposes or development of new quantification methodologies [19].

In this context we have recently established the optimal experimental conditions for the interaction between GHSR1a, using either intact cells or extracted cell membranes, with a variety of different secretagogues [20]. In this paper we center on the evaluation of true biological specimen and the effect of different human variables. Finally, a screening test for anti-doping purposes is proposed where the exogenous secretagogue, present in a urine sample, shall displace a radiolabelled ligand from the receptor.

5.3 Material and methods

Chemicals, cell culture, membranes preparations, competition binding and data analysis have already been published [20]. Briefly, HEK293 cells stably expressing the growth hormone secretagogue receptor type 1a (GHSR-1a) were a kind gift from Roy G. Smith (Baylor College of Medicine, Houston, TX,

112

USA) through F. Casanueva (University of Santiago de Compostela, Spain). Cells were cultured in 100-mm dishes in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% of fetal bovine serum (FBS), penicillin (100 U/ml), streptomycin (100 U/ml) and 2 mM glutamine (Invitrogen, Paisley, UK). Cells were subjected to G418 (Invitrogen) selection (800 µg/ml). HEK293 cells were grown under a humidified atmosphere of 95% air and 5% CO2 at 37 °C up to 70–80% confluence. Cultured cells were trypsinized and pelleted by centrifugation at 3000g for 10 min and frozen at −80 °C in 90% FBS 10% dimethyl sulfoxide (DMSO, Sigma- Aldrich, St Louis, MO, USA) at a concentration of 2×106 cells/ml. Membranes were prepared as follows: one aliquot of 2×106 cells was thawed and centrifuged at room temperature for 10 min at 3000g. The cell pellet was resuspended in 1 ml of homogenization buffer (50 mM Tris–HCl and 10% sucrose, pH 7.4) and sonicated for 5 min. Disrupted cells were centrifuged again at room temperature for 10 min at 3000g, and membrane pellet was resuspended in the desired volume.

5.3.1 Urine purification

Urine samples (stored at −20 °C) were thawed and centrifuged for 15 min at 4000×g. Urine samples at desired

113 concentrations were prepared by addition of secretagogues at known concentration. To desalt, 2.5 ml was loaded in a 3- ml Oasis solid phase extraction (SPE) cartridge (Waters, Mildford, USA) after conditioning with 2.5 ml of Milli-Q water. Columns were washed with 12.5 ml Milli-Q water and eluted with 2.5 ml methanol. After drying in a N2-evaporator TurboVap LV (Caliper, Hopkinton, USA) they were reconstituted in 1 ml freshly prepared binding buffer and sonicated for 3 min. Secretagogues were extracted by incubation with membranes from 2.5×106 cells (in 200 µl) for 40 min under continuous agitation at room temperature (200– 300 rpm). Samples were centrifuged for 10 min at 4000×g at 4 °C and pellets reconstituted with 0.5 ml acetic acid (0.2 M) and sonicated for 3 min to detach the secretagogue. Finally, samples were centrifuged for 10 min at 4000×g at 4 °C; the supernatant was dried in a Speed-Vac DNA 120 (Thermo Fisher Scientific, Waltham, USA) and reconstituted in 150 µl of binding buffer.

5.3.2 Calibrators

Two calibrators without the addition of urine were included in each competition binding experiment, a blank calibrator (no GHS) as the maximum binding and a saturation calibrator (7.5 µM GHRP-2) as the minimum to give the maximum

114 possible binding (100% of relative specific binding, RSB) and minimal possible binding (0% RSB), respectively. All sample binding values were calculated using these limits.

5.3.3 Samples

Urine samples were collected, frozen immediately and kept at −20 °C till analysis. For the evaluation of the exercise effect, samples were collected prior to and after the marathon race of Barcelona (March 7th 2010) and a training session of the basketball team of Sant Cugat del Valles (Barcelona, Spain). Urine samples from a pralmorelin dihydrochloride (GHRP-2) excretion study were collected by the Anti-doping Laboratory in Tokyo (Tokyo, Japan). These were of a single administration of 100 mg GHRP-2, orally, to a single subject and of a single administration of 100 µg GHRP-2, intravenously, to nine subjects as described by Okano et al. [16].

5.4 Results

To enable accurate comparison for each processed sample, the result is expressed in RSB. An RSB proximal to 100% shall be considered a negative sample, and values near 0% 115 shall be considered positive (do not displace or displace strongly the radioactive bound ligand, respectively). A threshold value is proposed later in the manuscript.

5.4.1 Sample purification

5.4.1.1 Desalting

The effect of the salt concentration in urine was observed in a competition experiment in the presence of a raw urine sample [20]. Four different urine samples without treatment gave an apparent RSB of 24.3% with a CV of 10.7%. The detaching of ~75% of the radiolabelled ligand in the absence of a competing reaction imposed a desalting step. The same four samples were retested after SPE obtaining this time an RSB value of 101.4% with a CV of 5.6%.

Spiked urine samples (75 nM of GHRP-2), reconstituted after SPE, with the same initial volume, gave higher RSB values than the aqueous countersample (RSB aqueous sample: 11.9%; RSB urine sample: 18.6%). This difference was initially attributed to the loss of secretagogue due to the desalting step, but when urine samples were reconstituted in a volume 10-fold lower than the initial, the resultant RSB became even higher (61.4%) with a CV of 57.8%. This

116 indicated that urine contains an interfering agent that is also concentrated along with the true secretagogue making a secondary purification step mandatory.

5.4.1.2 Selective secretagogue extraction

To isolate the GHS from the desalted sample, an extraction step using an excess of receptor molecules was applied. This step allowed a nearly 17-fold concentration of the initial 2.5- ml sample (reconstitution in 150 µl) with an RSB value similar to that obtained when incubating the same initial concentration of GHRP-2 in an aqueous matrix.

5.4.2 Assay accuracy

To evaluate the intra-assay and inter-assay variability, a blank urine sample was tested six times using the same assay. A different blank urine sample was tested in six separate experiments. Identical repetitions were carried out using the same urine samples spiked with GHRP-2 at 22.5 nM (see Table 1).

117

Table 1 Intra-assay and inter-assay accuracy

RSB (%) CV (%)

Blank 100.0 3.4 Calibrators (water) Spiked 31.2 8.8

Blank 101.0 4.2 Intra-assay (urine) Spiked 30.1 10.7

Blank 90.1 1.1 Inter-assay (urine) Spiked 29.7 8.3

5.4.3 Gender and age effect

Once the reproducibility of the established purification protocol was confirmed, urine samples obtained from Caucasian population without a secretagogue administration or ingestion were collected in order to evaluate the potential matrix effect due to the different composition of each individual urine.

In the samples from different gender, no differences were observed as mean values of 95.0% (n=27) and 92.3% RSB (n=7) with a CV of 6.2% and 5.6% were obtained for men and women, respectively. Thus, the 33 urine samples from men and women were screened for age effects (from 19 to 60 years, median 30 years). No variation in the binding 118 percentage with respect to the control was observed as a function of age, and the mean value of 94.5% with a CV of 6.1% (see Fig. 1a) was established.

5.4.4 Exercise effect

In order to determine the exercise effect on the competition assay, two different sports were evaluated. First, a prolonged and low intensity exercise (marathon) and second an intermittent and high intensity exercise such as basketball training.

For the marathon test, urine samples from nine athletes (eight males, one female) were collected before and after the race. Urine samples before exercise yielded a mean value of 92.9% RSB with a CV of 6.3%, while the postexercise samples yielded 91.2% RSB with a CV of 7.7% (see Fig. 1b).

Similar results were obtained when testing urine samples from a male basketball (n=11) team before and after a training session. Mean values were 98.5% and 98.7% RSB with CVs of 5.0% and 3.5% for pre-exercise and postexercise, respectively (see Fig. 1c).

At the individual level no trend was observed as a result of exercise for the competition assay. Thus, these results 119 indicate that the urine composition does not affect the results more than the intrinsic accuracy (Table 1) of the methodology.

Fig. 1 Evaluation of different urine samples. a) Urine sample vs. age. b) Pre-marathon and post-marathon samples. Lines connect same volunteer. c) Pre-basketball and post-basketball training samples. Lines connect same volunteer.

5.4.5 Threshold value

To establish a threshold value for the presence of a competing substance, blank urine samples were processed to evaluate the matrix effect evaluated yielding a mean value of 94.8% RSB (n=54) with a CV of 6.3%. Taking these results into account, and considering a one-tailed probability of 1/1000 (3.09 SD), the threshold was set at 76.3% RSB. 120

Lower values should be indicative of GHS presence and considered as a positive sample.

By spiking urine samples at different secretagogue concentrations, a competition curve for the GHS was setup. By interpolating the RSB threshold value into this curve, the minimal positive concentration (MPC) was calculated for the most common secretagogues (see Table 2). The MPC evidently depends on the affinity for the receptor. As such, the sensitivity is proportional to the affinity of the GHS, but as the biological effect is also related to this parameter, potential abuse of the lower affinity GHS shall require higher dosage.

Table 2 Minimal positive concentrations (MPC) for GHS in urine

GHS MPC (M)

MK-0677 1.5E-10

Ghrelin 3.0E-09

GHRP-2 3.5E-09

GHRP-6 9.7E-09

Hexarelin 9.9E-09

Tabimorelin 3.9E-07

Ipamorelin 7.2E-07

Alexamorelin 10.0E-07

121

5.4.6 Excretion study

In order to determine the sensitivity of the established protocol, samples from an excretion study were employed. Urine samples from a group of human mongoloid volunteers dosed intravenously with 100 µg of pralmorelin dihydrochloride (GHRP-2) were collected. Samples were collected prior to the administration as well as at 1.5, 4.5, 7, 10, 13, 20.5 and 31 h post-administration. Samples from nine volunteers were analyzed; the results are depicted in Fig. 2. For the sake of completion, analysis of the samples was also performed using LC–MS, and these results were published recently [16].

As can be seen in the Fig. 2, urine samples obtained prior to the secretagogue administration yielded an RSB value proximal to the hundred percent (mean 101.7% RSB, CV of 6.9%).

All urine samples collected immediately after the intravenous 125 GHS injection (1.5 h) displace to a large extent [ I-His9]- ghrelin (Fig. 2), obtaining a value well below the established threshold (dashed line in Fig. 2) determining a positive sample. The basal state was recovered after 4.5–7 h and remained unaltered at least after 31 h postadministration (data not shown). Preliminary results from an oral administration study (100 mg) yielded similar results (49.7%

122

RSB at 4.5 h post-ingestion), and this route of administration is further evaluated as part of a different study.

Fig. 2 RSB profile in human urine after a 100-µg intravenous dose of GHRP-2 dihydrochloride. Values below the threshold (76.3%; dotted line) are considered positive. Error bars indicate the range (nine subjects) of maximum and minimum values.

5.4.7 Linearity of the quantifiable range of GHRP-2 spiked urine samples

The objective of the screening test is to establish the presence/absence of exogenous GHS, but prior knowledge of

123 the GHS might permit to determine the approximate concentration. The sigmoidal shape of the competition curve allows only samples containing a GHS concentration proximal the EC50 to be reliably quantified. In order to ensure that concentration of a given secretagogue in a sample would not introduce a bias in the assay results through the purification procedure, a batch of samples spiked at different concentrations was tested. Seventeen urine samples with concentrations ranging from values that should slightly detach to samples that almost completely displace the radiolabel were evaluated.

The sigmoidal shape of the curve renders the termini with minimal slope less reliable for estimating concentrations. A minimum RSB quantifiable value was established at 5%. For GHRP-2, the 5% RSB is achieved at ~200 ng/ml (~267 nM) in the incubation media. Logit transformation renders the calibration curve linear (Fig. 3) but does not necessarily improve the indicated reliability.

124

Fig. 3 Linearity for GHRP-2. Logarithm of the urinary GHRP-2 concentration is represented in x-axis and the Logit values of the obtained RSB (ln RSB/(1−RSB)) in y-axis.

Considering the linearity of the quantifiable range, the urine samples from the excretion study were quantified using a urine–GHRP-2 control curve and compared with the results obtained from LC–MS for the same sample set [16] (see Fig. 4) yielding very similar numbers. For example, Okano et al. found 154.9 ng/ml (207.3 nM) and 16.3 ng/ml (21.8 nM) at 1.5 and 4.5 h post-injection, respectively, for volunteer 6. The competition assay yielded 149.3 ng/ml (199.9 nM) and 13.5 ng/ml (18.1 nM) for the same time points. For volunteer 1, Okano et al. found 7.9 ng/ml (10.6 nM) at 1.5 h post-injection and <0.05 ng/ml at 4.5 h, while the competition assay gave 19.5 ng/ml (26.1 nM) and 4.4 ng/ml (5.9 nM), respectively,

125 suggesting that possibly the competition assay suffers less from background noise or also interacts with metabolites, enabling analysis of lower concentrations.

Fig. 4 Comparison between different detection methods for GHRP-2 excretion. On the right the LCMS analysis of the excretion of GHRP-2 in urine is depicted (13). On the left the equivalent curve obtained from the competition assay is shown. For the transformation of the RSB into concentration, the calibration curve for GHRP-2 in urine was used. Error bars indicate maximum and minimum both in a and b.

5.5 Discussion

Competition binding is widely used by researchers for locating orphan receptors [21-23] or its tissue distribution. It is also used by the pharmaceutical industry for developing new, improved synthetic ligands [7;24;25]. Thus, the employed

126 methodology is known to be reliable and trustworthy. However, given the stringent anti-doping analytical requirements and the structural diversity of GHS, their interaction to GHS-R1a was carefully studied, and universal incubation conditions were established [20]. Here, we have tested the assay performance on human samples, evaluated biological variables and finally applied the method to excretion study specimens.

As in any assay it is of utmost importance to minimize all possible nonspecific contributions to the final read-out. These matrix effects could be due to any component (i.e. salts or proteins) or to the specific effect of the endogenous ligand for the GHS-R1a, ghrelin, secreted in urine at very low levels.

Endogenous ghrelin coexists together with a nonacylated form that interacts with the receptor only at very high concentrations [26] and has no GH-releasing capacity [27]. Plasma ghrelin and non-acylated ghrelin levels (100–200 fmol/ml) increase before meals to 300 fmol/ml [28-30] and decrease post-prandially, indicating an action in energy homeostasis in parallel with the growth hormone secretion [31]. Of these concentrations, about 10% corresponds to ghrelin, and secretion into urine does not exceed 10 fmol/ml [30]. Considering these values and the minimal concentration considered to be positive (MPC) for ghrelin, the basal values of the endogenous secretagogue should not interfere in the competition binding experiments. This is in agreement with

127 the absence of competition for all natural urine specimens that did not originate from the excretion study.

Despite that endogenous ghrelin does not interfere with the assay, other matrix effects imposed sample pretreatment. Finally a two-step protocol, a desalting process followed by selective secretagogue extraction, was established. The first one is in concord with the literature where it is described that elevated salt concentrations alter in a significant manner the binding equilibrium of the G-coupled protein receptors [32], especially GHS-R1a [20]. The second interfering agent(s) impeded concentrating urine sample, as volume reduction 125 consequently led to increased [ I-His9]-ghrelin displacements. The use of a 5.0×104 -fold GHS-R1a excess allowed extracting both GHS and interfering agent, but only the former was eluted. Despite that the interfering agent was not identified, it can be eliminated, and by using a protocol based on the same principle as the screening test ensures that all potential secretagogues will be captured warranting the universal applicability.

With these two parameters addressed, all posterior blank samples (intra-assay and inter-assay) yielded fairly low coefficients of variation (ca 6%) that indicate that salts and other interfering agents were efficiently removed. From a comparison between spiked samples and the spiked calibrator in aqueous matrix, not submitted to the purification procedure, it could be concluded that the employed

128 purification methodology is very suitable for the screening assay.

In the next step true biological variables were addressed. No differences were observed when comparing the RSB results between men and women and both compared with the blank calibrator. This indicates that no gender differential compound is present at the level where it would be picked up by the assay or, alternatively, it has been discarded during the purification. Similarly, the age effect was evaluated and found to be negligible. Samples of both genders in a wide age range (19–60 years) were tested covering pretty much the active period of any athletic discipline. Due to ethical issues, this study did not include adolescent samples, and it remains to be evaluated whether in this age range a specific matrix effect might occur.

Ghrelin is apparently not associated with GH regulation during exercise [33;34], and it is shown that ghrelin is even suppressed due to the exercise practice [35]. Nevertheless, as the screening test should be employed as anti-doping method, and considering the relationship between exercise practice and GH release [36;37], we tested two representative disciplines: on one hand a long lasting (several hours), continuous and less emphatic sport such as marathon and on the other hand strenuous and intermittent basketball training. No differences were observed between sports, both at the pre-exercise and post-exercise levels, and no

129 differences could be appreciated between exercise and sedentary samples. Even though the number of disciplines evaluated here is exiguous, the results are in agreement with earlier papers and indicate that the proposed test is appropriate for testing urine samples from athletes independently from the daily training or the sport nature.

Ultimately, implementation of a screening assay for GHS requires a threshold RSB value to discriminate negative from positive samples. After analyzing a high number (>50) of blank urine samples obtained under different conditions, a mean RSB value together with its variability was set. Then, the false-positive rate was set to 1 in 1000, which is fairly stringent for a screening assay, yielding a cut-off value of 76.3%. In establishing the threshold only Caucasoid people were included, but the excretion study samples were from Mongoloid individuals and the pre-treatment samples behaved identical to the basal population. Other ethnicities should be targeted in the future, but initial results from two different ethnic groups hold promise for a global application.

One fundamental part of any potential method remains the applicability to true samples. For this purpose samples from an excretion study were processed. Samples were processed and quantified using LC–MS affording the concentration of most samples [16]. Therefore, expectations were to obtain positive values in some cases. From a total of 45 processed samples pertaining to the pharmacokinetic profile, all nine

130 volunteers yielded a positive result at 1.5 h post-injection, and in some cases the positive result could be confirmed for the 4.5-h sample. None was positive after 7.5 h either by the LC– MS approach or the competition assay, indicating a detection window of about 4.5 h for both techniques for the intact compound. The slightly higher concentrations encountered in the competition assay at the 7.5-h time point suggest that the GHRP-2 metabolites could have some interaction with the receptor. However, this remains to be investigated with the pure substance, but in any case the interaction is expected to be fairly weak as the WOO is not extended. In contrast, in the LC–MS approach knowledge of the GHRP-2 metabolism allowed specific screening for one major metabolite, namely the 3 N-terminal amino acids indicated as AA-3. This metabolite was detected up to 24 h after administration [16]. It is evident that with prior knowledge of the GHS metabolism, the LC–MS approach offers the advantage of screening for degradation products that significantly enhance the WOO. It remains to be seen what will happen to the excretion profile upon prolonged administration with similar, lower or higher dosage as frequently occurs in non-clinical regimes, but expectations are that it may only improve the analytical capacity. Furthermore, this detection window might also depend on the administration way, the stability of the secretagogue, excretion ratio or the bioavailability. In this context, the single specimen derived from an oral administration of 100 mg of pralmorelin dihydrochloride was

131 also tested and detected 4.5 h after ingestion. No further time points were available in this preliminary study, but the similarity with the intravenous study indicates that, at least for pralmorelin, no major differences shall exist.

This competition-binding assay is qualitative and not quantitative. However, when the identity of the secretagogue is known, the concentration can be determined. In this context it is relevant to determine how the purification procedure addresses each GHS; e.g. even though the SPE procedure for extracting out the hydrophobic material should not discriminate, not all shall behave equally. Furthermore, the purification procedure might be concentration dependent, and this could lead to underestimation of the true concentration. Whereas this could only produce false negatives, it is relevant to address this point. In this study we established the linearity of the quantifiable range for GHRP-2 knowing that at either end of the curve the validity is compromised. Nevertheless, when comparing the concentrations of GHRP-2 obtained by testing the samples from the excretion study with data from LC–MS analysis [16], it could be concluded that this quantification was similar and concentration induced discrimination occurs. Obviously the prerequisite of knowing which secretagogue is present in the biological sample invalidates this test as a quantification method when this information is unknown.

132

In summary, we have established a general purpose screening procedure to address GHS in an anti-doping setting. Following an initial study focused on the GHSR1a [20], we have targeted biological samples addressing variables such as age, gender and exercise and found negligible effects on the outcome. The protocol was successfully applied on the first-ever commercially available GHS. Considering the large number of samples, a universal screening procedure for GHS appears to be extremely useful in order to address this class of compounds. The confirmation of the presence of secretagogues, and which secretagogue, could ultimately come from additional analysis using mass spectrometry based approaches. As the main virtue of the test is the use of GHS-R1a, target of all growth hormone secretagogues and the competition against the labeled ligand for the receptor, this screening test is applicable for current GHS but already anticipates future pharmaceutical developments in the field.

5.6 Acknowledgments

The authors thank M. Okano of Mitsubishi Chemical Medience Corporation (Tokyo, Japan) for sharing excretion study samples, and F. Casanueva (Santiago de Compostela, Spain) and R.G. Smith (Houston, USA) for GHSR-1a-HEK293

133 cells. The authors acknowledge the financial support from the USADA, the MICINN (DEP2009-11454), WADA (10B11RG), IMIM Foundation and the Catalan Government (DIUE 2009SGR492).

5.7 References

1. LI CH, PAPKOFF H. Preparation and properties of growth hormone from human and monkey pituitary glands. Science 1956; 124(3235):1293-1294.

2. RABEN MS. Treatment of a pituitary dwarf with human growth hormone. J Clin Endocrinol Metab 1958; 18(8):901-903.

3. Duchaine D. Underground Steroid Handbook. Santa Monica: 1982.

4. Bowers CY, Momany F, Reynolds GA, Chang D, Hong A, Chang K. Structure-activity relationships of a synthetic pentapeptide that specifically releases growth hormone in vitro. Endocrinology 1980; 106(3):663-667.

5. Bowers CY, Momany FA, Reynolds GA, Hong A. On the in vitro and in vivo activity of a new synthetic hexapeptide that acts on the pituitary to specifically release growth hormone. Endocrinology 1984; 114(5):1537-1545.

6. Kojima M, Hosoda H, Date Y, Nakazato M, Matsuo H, Kangawa K. Ghrelin is a growth-hormone-releasing acylated peptide from stomach. Nature 1999; 402(6762):656-660.

134

7. Moulin A, Ryan J, Martinez J, Fehrentz JA. Recent developments in ghrelin receptor ligands. ChemMedChem 2007; 2(9):1242-1259.

8. Perdona E, Faggioni F, Buson A, Sabbatini FM, Corti C, Corsi M. Pharmacological characterization of the ghrelin receptor antagonist, GSK1614343 in rat RC- 4B/C cells natively expressing GHS type 1a receptors. Eur J Pharmacol 2011; 650(1):178-183.

9. Henninge J, Pepaj M, Hullstein I, Hemmersbach P. Identification of CJC-1295, a growth-hormone- releasing peptide, in an unknown pharmaceutical preparation. Drug Test Anal 2010; 2(11-12):647-650.

10. Kohler M, Thomas A, Geyer H, Petrou M, Schanzer W, Thevis M. Confiscated black market products and nutritional supplements with non-approved ingredients analyzed in the Cologne Doping Control Laboratory 2009. Drug Test Anal 2010; 2(11-12):533-537.

11. Thomas A, Kohler M, Mester J, Geyer H, Schanzer W, Petrou M et al. Identification of the growth-hormone- releasing peptide-2 (GHRP-2) in a nutritional supplement. Drug Test Anal 2010; 2(3):144-148.

12. Okano M, Nishitani Y, Sato M, Ikekita A, Kageyama S. Influence of intravenous administration of growth hormone releasing peptide-2 (GHRP-2) on detection of growth hormone doping: growth hormone isoform profiles in Japanese male subjects. Drug Test Anal 2010; 2(11-12):548-556.

13. WADA and List Comitee (2011). The 2011 Prohibited list. International standard. The World Anti-Doping Agency [ 2012 Available from: URL:http://www.wada- ama.org/Documents/World_Anti- Doping_Program/WADP-Prohibited- list/2012/WADA_Prohibited_List_2012_EN.pdf

135

14. Bidlingmaier M, Suhr J, Ernst A, Wu Z, Keller A, Strasburger CJ et al. High-sensitivity chemiluminescence immunoassays for detection of growth hormone doping in sports. Clin Chem 2009; 55(3):445-453.

15. Constanzer ML, Chavez-Eng CM, Matuszewski BK. Determination of a novel growth hormone secretagogue (MK-677) in human plasma at picogram levels by liquid chromatography with atmospheric pressure chemical ionization tandem mass spectrometry. J Chromatogr B Biomed Sci Appl 1997; 693(1):131-137.

16. Okano M, Sato M, Ikekita A, Kageyama S. Determination of growth hormone secretagogue pralmorelin (GHRP-2) and its metabolite in human urine by liquid chromatography/electrospray ionization tandem mass spectrometry. Rapid Commun Mass Spectrom 2010; 24(14):2046-2056.

17. Thomas A, Hoppner S, Geyer H, Schanzer W, Petrou M, Kwiatkowska D et al. Determination of growth hormone releasing peptides (GHRP) and their major metabolites in human urine for doping controls by means of liquid chromatography mass spectrometry. Anal Bioanal Chem 2011; 401(2):507-516.

18. Camina JP, Carreira MC, El MS, Llorens-Cortes C, Smith RG, Casanueva FF. Desensitization and endocytosis mechanisms of ghrelin-activated growth hormone secretagogue receptor 1a. Endocrinology 2004; 145(2):930-940.

19. Leyris JP, Roux T, Trinquet E, Verdie P, Fehrentz JA, Oueslati N et al. Homogeneous time-resolved fluorescence-based assay to screen for ligands targeting the growth hormone secretagogue receptor type 1a. Anal Biochem 2011; 408(2):253-262.

136

20. Pinyot A, Nikolovski Z, Bosch J, Segura J, Gutierrez- Gallego R. On the use of cells or membranes for receptor binding: growth hormone secretagogues. Anal Biochem 2010; 399(2):174-181.

21. Deghenghi R, Papotti M, Ghigo E, Muccioli G. Cortistatin, but not somatostatin, binds to growth hormone secretagogue (GHS) receptors of human pituitary gland. J Endocrinol Invest 2001; 24(1):RC1- RC3.

22. Howard AD, Feighner SD, Cully DF, Arena JP, Liberator PA, Rosenblum CI et al. A receptor in pituitary and hypothalamus that functions in growth hormone release. Science 1996; 273(5277):974-977.

23. Katugampola SD, Pallikaros Z, Davenport AP. [125I- His(9)]-ghrelin, a novel radioligand for localizing GHS orphan receptors in human and rat tissue: upregulation of receptors with athersclerosis. Br J Pharmacol 2001; 134(1):143-149.

24. Lu Z, Tata JR, Cheng K, Wei L, Chan WW, Butler B et al. Highly potent growth hormone secretagogues. Bioorg Med Chem Lett 2007; 17(13):3657-3659.

25. Torsello A, Ghe' C, Bresciani E, Catapano F, Ghigo E, Deghenghi R et al. Short ghrelin peptides neither displace ghrelin binding in vitro nor stimulate GH release in vivo. Endocrinology 2002; 143(5):1968- 1971.

26. Gauna C, van de Zande B, van KA, Themmen AP, van der Lely AJ, Delhanty PJ. Unacylated ghrelin is not a functional antagonist but a full agonist of the type 1a growth hormone secretagogue receptor (GHS-R). Mol Cell Endocrinol 2007; 274(1-2):30-34.

27. Broglio F, Benso A, Gottero C, Prodam F, Gauna C, Filtri L et al. Non-acylated ghrelin does not possess the pituitaric and pancreatic endocrine activity of acylated

137

ghrelin in humans. J Endocrinol Invest 2003; 26(3):192-196.

28. Cummings DE, Purnell JQ, Frayo RS, Schmidova K, Wisse BE, Weigle DS. A preprandial rise in plasma ghrelin levels suggests a role in meal initiation in humans. Diabetes 2001; 50(8):1714-1719.

29. Tschop M, Wawarta R, Riepl RL, Friedrich S, Bidlingmaier M, Landgraf R et al. Post-prandial decrease of circulating human ghrelin levels. J Endocrinol Invest 2001; 24(6):RC19-RC21.

30. Yoshimoto A, Mori K, Sugawara A, Mukoyama M, Yahata K, Suganami T et al. Plasma ghrelin and desacyl ghrelin concentrations in renal failure. J Am Soc Nephrol 2002; 13(11):2748-2752.

31. Ariyasu H, Takaya K, Tagami T, Ogawa Y, Hosoda K, Akamizu T et al. Stomach is a major source of circulating ghrelin, and feeding state determines plasma ghrelin-like immunoreactivity levels in humans. J Clin Endocrinol Metab 2001; 86(10):4753-4758.

32. Horstman DA, Brandon S, Wilson AL, Guyer CA, Cragoe EJ, Jr., Limbird LE. An aspartate conserved among G-protein receptors confers allosteric regulation of alpha 2-adrenergic receptors by sodium. J Biol Chem 1990; 265(35):21590-21595.

33. Dall R, Kanaley J, Hansen TK, Moller N, Christiansen JS, Hosoda H et al. Plasma ghrelin levels during exercise in healthy subjects and in growth hormone- deficient patients. Eur J Endocrinol 2002; 147(1):65- 70.

34. Kraemer RR, Durand RJ, Acevedo EO, Johnson LG, Kraemer GR, Hebert EP et al. Rigorous running increases growth hormone and insulin-like growth factor-I without altering ghrelin. Exp Biol Med (Maywood ) 2004; 229(3):240-246.

138

35. Broom DR, Stensel DJ, Bishop NC, Burns SF, Miyashita M. Exercise-induced suppression of acylated ghrelin in humans. J Appl Physiol 2007; 102(6):2165- 2171.

36. Kraemer RR, Kilgore JL, Kraemer GR, Castracane VD. Growth hormone, IGF-I, and testosterone responses to resistive exercise. Med Sci Sports Exerc 1992; 24(12):1346-1352.

37. Pritzlaff CJ, Wideman L, Weltman JY, Abbott RD, Gutgesell ME, Hartman ML et al. Impact of acute exercise intensity on pulsatile growth hormone release in men. J Appl Physiol 1999; 87(2):498-504.

139

140

6 Chapter 6:

DISCUSSION

141

It is important to discuss several aspects of the method developed: the need for a test to detect GHS, the developed methodology itself, and alternative approaches. The discussion is relevant for the applicability of the proposed test in anti-doping laboratories around the world.

6.1 The test

6.1.1 Need for a test to detect GHS

The main argument for the development of any methodology to detect GHS in biological samples for anti-doping purposes is because it is included in the prohibited list from The World Anti-Doping Agency [1], under the section “S2. Peptide hormones, growth factors and related substances” in which it is stated explicitly: “substances and their releasing factors are prohibited”. Thus, all growth hormone secretagogues and other releasing factors are included.

It has been a long time since rhGH appeared on the list of prohibited substances. Nowadays, with the detection method for rhGH in place [2-6] the prevalence of rhGH abuse starts to be perceived: the number of results during the first period of testing has indicated at least 10 recent adverse findings. Its use will probably decrease in favour of other undetectable

142 substances such as the, GHS, that could be seen as the natural rhGH replacement.

Apart from the sake of fair play in sport competition, the administration of GHS might be potentially dangerous as these compounds are relatively new, many not thoroughly tested and even endogenous ghrelin is still under investigation due to the numerous roles in which it is apparently involved. Ghrelin takes part in several biological functions and consequently, the related GHS could promote also unexpected reactions. Moreover, in the study of Copinschi et al. [7], the authors demonstrated that the long- term administrations of MK-677 effects are quite different than the ones produced by a single dose (higher circulating hGH levels and higher pulse frequency after a 7 day treatment). Thus, the presumable side effects of a dosage will also depend on the administration regime. Moreover, a second receptor variant (GHS-R1b) is expressed due to an alternative splicing of the same cluster gene [8] and it has been reported to be more widespread than GHS-R1a. Thusfar, no physiological relevant role has been ascribed to GHS-R1b. Thus, the administration of any GHS could entail unexpected side effects if such compound would act also on the receptor variant.

Already in 1993, Huhn et al. [9] demonstrated the hGH releasing capacity of the first synthesized GHRP. In that study, authors administrated 1 µg/kg/h of GHRP-6 for 24

143 hours to several healthy human subjects. They concluded that the hGH secretion rate was enhanced 8-fold, doubling the number of detectable hGH pulses, and 4- to 6-fold increases in individual hGH pulse area comparing to the control group to which a saline volume was constantly administrated. In 1996, Copinschi et al. [7] orally dosed several healthy men just before bedtime (coinciding with the natural GH peaks) with 5 mg of MK-677 during 7 days. Despite the fact that after the first day of administration the value of circulating hGH resulted very similar to the control (13 and 20 ng/ml of hGH for control and dosed, respectively), after 7 days of treatment values up to 50 ng/ml of hGH where observed for the dosed group while the control remained unaltered. Recently, Okano et al. [10], demonstrated that with a single intravenous dose of 100 µg of GHRP-2 a peak of circulating hGH with absolute values between 50 and 120 ng/ml is achieved, while a subcutaneous injection of 0.04 mg/kg of rhGH itself provoked a rising of only up to 20 ng/ml. They also demonstrated that this tremendous increase in circulating hGH after GHRP-2 administration did not alter the Rec/Pit ratio in the test presently used to detect hGH abuse. Even more, they demonstrated that the altered Rec/Pit ratio value provoked by an administration of rhGH could be normalised through a GHRP-2 injection. Despite that this masking effect has only been described for GHRP-2, it is easily anticipated that all other secretagogues will produce such effect depending on their affinity for the GHS-R1a. In

144 addition to these three examples, several other studies [11- 17] appear to corroborate that GHS could be used to increase the circulating hGH levels without administrating rhGH. It is to be expected that those other GHS administrations do not modify the Rec/Pit ratio and that GHS could be used therefore as masking agents for rhGH abuse.

Some anti-doping laboratories have also targeted GHSs aiming at developing of a detection test for such substances. In 2006, Thevis and coworkers [18] developed a procedure based on liquid chromatography and mass spectrometry (LC- MS) in order to quantify the concentration of an MK-0677 analogue in a spiked urine sample. In 2007 the same authors [19] reviewed the potential of GHS releasing peptides (GHRH and some GHS) as candidates for abuse in elite sports due to their clinical and pharmacological properties. In 2010, Okano et al. [20] also quantified the concentration of GHRP-2 and its metabolite (AA-3), again using LC-MS, present in urine samples after the administration of this secretagogue. Thomas et al. [21] conducted another study where several GHSs were considered. No administration was conducted here but authors analysed urines fortified with varying amounts of GHRP-1, GHRP-2, GHRP-4, GHRP-5, GHRP-6, Alexamorelin, Hexarelin and Ipamorelin. In addition, they evaluated a urine from an excretion study where GHRP-2 was administrated thus reproducing Okano´s study. All these tests, however, are capable of addressing the compounds

145 only if their identity is known leaving unknowns to escape detection.

The discussion about the suitability of a detection test for misuse of GHS in sport appears relevant if the recent discoveries in the black market are taken into account. In 2010 Kohler et al. [22] analysed several vials confiscated by the Anti-Doping Authority of the Netherlands and found recombinant erythropoietin, IGF-I, and Hexarelin in individual vials. They also analysed two nutritional supplements, Hemogex and Hemotropin that claim containing GHRP-2. Hemogex, which is presented as a drinking solution, resulted to contain 9 mg of GHRP-2 per vial, which is close to the tested pharmacological amounts [23;24]. Hemotropin, which is presented in tablets, contained only 50 µg of GHRP-2 per tablet, a very low concentration for obtaining any benefit, considering the metabolic fate of orally ingested peptides. More recently, Goebels from the Australian Drug Testing Laboratory performed similar analyses finding GHRP-2 and GHRP-6 with the highest frequency but also Hexarelin, Ipamorelin, MK-0677, Capromorelin or Tabimorelin in distinct specimens. Samples were obtained via Internet (n = 63) and from the Australian Customs and Border Protection Service (n = 155). Similar to the early days when GH abuse could be suspected only, these findings apparently demonstrate that GHS are currently been misused.

146

In summary, the scientific community together with official organizations have invested many resources and time to develop a hGH detection test and now that it is in place and successful the threat of bypassing this test through GHS is real. As such, a GHS detection test appears to be essential.

6.1.2 Material of the test

The procedure described in this dissertation to detect GHS is new in many ways. Not only does it address this largely ignored family of compounds, it also entails the use of reagents that are not regularly employed in an anti-doping laboratory, noticeably the radioligand or the cellular membranes that harbour the receptor.

The use of radioactivity in an anti-doping laboratory is not very extended but apparently this trend is about to change. Nowadays, an alternative for the detection of hGH misuse is about to be implemented during the Olympic Games of London 2012. Such methodology, developed by the GH-2000 project [25], is based on the measurement of the GH- sensitive biomarkers Insulin-like growth factor I (IGF-I) and the amino-terminal pro-peptide of type III collagen (P-III-NP), both of which rise in response to exogenous GH administration [26]. Due to the required sensitivity, at least the

147 methodology for P-III-NP quantification uses an antibody carrying a 125-iodine label [25].

6.1.2.1 The radioligand

In order to quantify the amount of ligand that is displaced from the receptor site, a radioactively labelled ghrelin is used.

If radioactive labelling is the choice, several radioisotopes can be employed in biochemistry and the choice depends on the isotope characteristics, such as the half-life and specific activity, as well as the methodology being used. A resuming table with the characteristics of several radioisotopes is presented in Table 1.

Table 1: Physical data of te more commonly used in biochemistry radioisotopes [27].

Emission type, penetrance (Energy Isotope emission) Half-life

3H Beta, very low penetrance (18,6 KeV) 12.28 years

14C Beta, low penetrance (156 KeV) 5730 years

35S Beta, low penetrance (167 KeV) 87.5 days

32P Beta, high penetrance (1709 KeV) 14.29 days

125I Gamma, very high penetrance (35KeV) 59.4 days

148

The most commonly labelled GHS is [125I-His9]-ghrelin, which was first synthesized by Katugampola and coworkers in 2001 [28] with the aim to localize the GHS-R1a in other tissues apart from the hypothalamus and pituitary. They demonstrated that this radioligand bound to the receptor with high affinity and it resulted to be specific, saturable, and reversible, characteristics of a ligand-receptor interaction. Authors also demonstrated that radiolabelled ghrelin and unlabelled ghrelin competed for the same receptor site indicating that the iodine did not participate in the interaction. This feature is of particular relevance to establish a proper competition experiment. For the iodination, the authors used the chloramine-T method that permits incorporating iodine to a tyrosine or histidine rendering only one iodine atom at histidine at position 9 in the case of ghrelin.

The [125I-His9]-ghrelin has become widely used in research; e.g. the synthesis and evaluation of new or existing secretagogues or dynamics of the receptor during the signal transduction [29-35]. Nowadays, iodinated ghrelin can be purchased from different companies such as Bachem, Millipore, or Perkin-Elmer. Otherwise, ghrelin and hexarelin compounds have been modified in order to incorporate a tyrosine residue as the iodination process in this amino acid is easier yielding 125I-Tyr-Ala-Hexarelin and the 125I-Tyr4- ghrelin molecules that were also used as competitors in receptor binding studies [36-38].

149

Not all labelled secretagogues have been radiolabelled with iodine isotope. The first GHS that was radiolabelled was MK- 0677 even prior to the discovery of ghrelin. MK-0677 was labelled substituting the 32S with its 35S isotope and this compound was useful for the isolation and cloning of the GHS-R1a [8] and used in further experiments [39-42].

The initial consideration of in-house production of the radioligand was driven by the commercial availability. Of all cited labelled GHS, only non-modified ghrelin was commercially available and therefore the iodination by the chloramine-T methodology was chosen. As commented in Chapter 3, during the course of this thesis several ghrelin iodination reactions were carried out but despite the fact that the chloramine-T methodology is well documented in literature, the obtained material did not always fulfilled the expectations, as discussed below.

Firstly, the radioactive material presented high adhesion to all sorts of tube and tip surfaces. This non-specific binding implied a loss of 60-80% of the radioactivity that remained in the vial (irrespective of this being a plastic of glass tube). Such effect was partially resolved by silanizing tips and tubes but this additional step was time-consuming and not fully reproducible. The lack of reproducibility was attributed to the resuspension buffer that initially consisted of only deionized water and greatly improved when a buffer containing 0.25% BSA and 0.2% citric acid was employed.

150

Moreover, the labelling procedure was miniaturised, and without performing an extensive optimization, a high variability in the labelling efficiency was observed. This fact, together with some difficulties to further purify the labelled from the non-labelled peptide, made that the different batches presented a high variability in the specific activity. Some labelled ligand batches presented affinity for the receptor with specific binding values of only 3%. At that stage the failing was attributed to the labelled material albeit later results and achievements in protocol optimization disclosed that also receptor disruption due to the membrane obtaining protocol could have contributed to the low specific binding values.

Obviously, the use of radioactivity requires dedicated equipment (HPLC, column, dryer) and was conducted at the Institute of Advanced Technology (IAT, PRBB, Barcelona, Spain). Labelling procedures were conducted with 10 mCi Na125I per experiment and yielded reduced specific binding to the receptor. Given this important drawback and that institutional restrictions on the maximum amount of radioisotope to be manipulate per annum (linked to the permits granted by the national nuclear security council) limited the number of labelling procedures, it was decided to employ commercial [125I-His9]-ghrelin. In retrospect, it would be interesting to evaluate a labelling procedure now that the protocol is fully optimized and revaluate its suitability.

151

Considering the principle of the proposed assay, where an unlabelled secretagogue displaces a radioactive ligand in competition for the same receptor site, one could reason that a low affinity radioligand would be ideal for the assay as low concentrations of non-labelled ligand could displace it and therefore, thus lowering the limit of detection. This was addressed through the use of an iodinated truncated ghrelin peptide containing amino acids 1 to 14. Truncated ghrelin peptide (1-14) was described to present practically 40-times lower affinity for the receptor [43] and therefore would represent an ideal competitor as very low concentrations of non-labelled ligand could compete. However, in our hands, the [125I-His9]-ghrelin-1-14 presented no affinity for the receptor, even at high concentrations and without competitor present. Further publications based on competition binding studies demonstrated that the affinity was much lower [37] than initially proposed. Moreover, contrary to the argument that a low affinity labelled ligand may be ideal in a competition-binding assay as it could be easily detached from the receptor, published data present ghrelin and MK-0677 as the two most used radioligands for the GHS-R1a and these compounds result to have the highest affinity for the receptor. The fact is that eventually a high-affinity ligand is preferred for several reasons; a lower concentration of radioligand can be used in the assay resulting in a lower level of nonspecific binding, and with a higher affinity, a slower dissociation rate is usually presented which provides for a more convenient

152 assay [44]. Thus, the use of labelled truncated ghrelin peptides was abandoned due to the preliminary results and low expectations offered.

6.1.2.2 Receptor

The use of membrane bound receptors in anti-doping analysis also represents a novelty for the non-specialized laboratories. The use of these receptors is rather imperative for the correct development of the presented assay in order to maintain the claim of universal use. GHS are purified and detected by the interaction with the receptor, and this aspect is crucial for the assay. Otherwise, the receptor support (e.g. cells, membranes...) could be selected so as to facilitate the assay procedure, the receptor production, storage or even distribution.

In chapter 4, the comparative use of intact cells versus isolated membranes is described and discussed rendering the conclusion was that it depends on the biological function to be monitored. For example, it is impossible to use isolated membranes to monitor the intracellular calcium mobilization. In this study it was concluded that, using an optimized membrane obtaining methodology, isolated membranes offered a higher specific binding value than intact cells. As

153 this circumstance favoured the accuracy of the GHS detection assay, this receptor support was eventually selected.

The GHS receptor 1a was expressed in transfected Human embryonic kidney (HEK-293) cells that were obtained from Dr. Roy Smith from the Baylor College of Medicine in Houston through Dr. Casanueva from the University of Santiago de Compostela, two scientific groups with high relevance in the GHS field [8;31;39;42;45-55]. Cells were grown under conditions mentioned in chapter 4 presenting no particular problems. Nevertheless, when incubating isolated membranes with radioactive ligand initially no specific binding was obtained. This result was eventually attributed to the membrane obtaining procedure used at that stage. Initially, to isolate membranes from cultured cells, a reproduction of a procedure presented by Guerlavais et al. [29] was employed. This methodology was based on three freeze and thaw cycles with the aim of disrupting the cell membrane. As commented in chapter 3, this procedure provoked that GHS-R1a molecules were unable to bind to the ligand and when used in a binding experiment, and not more than a 3% of the initial radioactivity was obtained. Binding experiments revealed that specific binding decreased with each freeze and thaw cycle being indistinguishable from non-specific binding after three cycles. After several optimization steps a fully functional methodology, based on sonication to disrupt the cells, was developed obtaining specific binding values up to the 40% of the initial radioactivity. When using intact cells the value of 154 specific binding never exceeded 30% and, as commented in chapter 4, this was attributed to the likely endocytosis and desensitization of the receptor when the cell metabolism is intact [31]. This endocytosis of a receptor may result in the incorporation of part of the radioisotope inside the cell contributing to the noise level, and this metabolic procedure affects the law of mass action in which the binding model is based on [44]. The law of mass action assumes:

• All receptors are equally accessible to ligands. • Receptors are either free or bound to ligand. No partial bindings are considered. • Binding does not alter the ligand or receptor. • Binding is reversible.

In the intact cell scenario, the ligand promotes the receptor internalization and desensitization partially breaking the law of mass action. This fact does not imply that intact cells cannot be used in such experiments but a comparison with an assay of different characteristics (isolated membranes) is more complicated. It is worth mentioning that the study which concludes that the GHS-R1a can be desensitized or endocitized was carried out by Dr. Casanueva’s group [31] using the same HEK293 cells as we employed for the development of the assay.

The mentioned specific binding values obviously depend on the receptor density in the incubation media. With the aim of standardization, cultured cells were counted in a cell counter 155 in order to aliquot these to 5 million cells per aliquot. Independently whether such cells were destined to be used in an intact cell binding or isolated membranes assay the initial cell count allowed to dilute the cells or the membranes into the same final volume and therefore obtaining the same incubation receptor density.

As happens for the radiolabeled ghrelin, the GHS receptor is also commercially available. Due to the initial drawbacks with the in-house cultured cells, commercial membranes from two different suppliers where tested. Such membranes presented rather low specific binding (4-11%) compared with the specific binding obtained later by us using intact cells or isolated membranes (higher than 30 %) from our cultures and this can be attributed to several factors: lower receptor expression, receptor breakdown due to disruption protocol or lower total protein used in the assay as these parameters are difficult to standardise amongst different preparations. Despite the lower specific binding of these commercial membranes, they allowed determining the failure of the initial in-house [125I- His9]-ghrelin bathes.

Another receptor alternative would be a different cell line. Despite that HEK 293 cells are very easy to grow and have been widely used in cell biology research, other cell lines over-expressing the GHS-R1a have been described such as Chinese hamster ovary (CHO) [31;34;40], COS [55], LLC PK- 1 [29;32]. Any other cell line expressing the receptor at a

156 similar density could be an appropriate receptor alternative for the proposed assay if other parameters such as nonspecific binding, receptor-ligand affinity, receptor survival during the cell lysis, etc. are not altered. In essence, the use of another production system might comply with the current rules of the world anti-doping agency with respect to the application in screening and confirmation procedures. However, thus far no study has been performed on the comparative properties of these cell lines and therefore a switch in membrane origin is not recommended.

All mentioned membranes obtaining protocols, aimed to produce a membrane suspension without altering the receptors. One alternative procedure would be to solubilize the receptor and remove all other compounds. Pomés and coworkers [41] achieved to solubilize the GHS-R1a from porcine anterior pituitary cells by adding a detergent to the cell suspension. However, authors reported great difficulties in order to maintain the receptor functionality and concluded that the ligand presence was required so as to maintain the conformation of the receptor protein while solubilizing. The difficulties encountered in the solublization, added to the absence of assay improvement, and the additional time- consuming steps (filtration or ultracentrifugation for separating the bound from the free radioligand) discouraged this approximation.

157

A cell culture is not frequently, (not to say never), available in an anti-doping laboratory and therefore, the use of commercial membranes could be implemented. Otherwise, during the time course of this thesis, cell aliquots have been stored in an -80 ºC freezer for long periods of time (more than 16 months) obtaining minimal divergence in results. Thus, cells could be cultured at a central location and distributed to other laboratories providing a guarantee of crucial chemicals, often found to be the bottleneck in new procedures that rely on commercial products. Obviously, this approximation requires further study in order to corroborate the receptor stability in storage or sending conditions.

6.1.3 Methodology and assay requirements

The methodology employed suffered several changes as a result of the different steps that were optimized. The major modification was focused on the separation of the free/bound fraction where initially a filtrating procedure was applied and exchanged later for a centrifugation process. Even though literature advocates filtration for this particular application [29;32;38-40], in the present situation this methodology resulted quite inappropriate. First of all, manipulation of the filter led to contamination between samples as the radioactive contamination was evident in the tweezers required to

158 introduce the filter into the counting tube. A decontamination step for each single sample resulted unattainable without substantial delays and loss in throughput capacity. Furthermore, each filter had to be washed with at least 5 ml of washing buffer and therefore, this volume had to be treated as radioactive waste. Another, smaller but not less important, impediment was that filters had to be positioned in the same position in the counting tube as, depending on its relative localization with respect to the gamma counter’s photomultiplier, the reader yielded different values. All these drawbacks motivated a methodology variation and after using intact cells as the receptor platform, in which separation is performed by centrifugation, no further filtration was used. Thus, after the centrifugation step, the supernatant (unbound material) was aspired and only 300 µl per tube were added to wash the remaining radioactivity from the tube surface. This new process diminished significantly the variability among samples because no pellet manipulation was required and the final radioactivity is counted in the same tube where the incubation and separation took place. Moreover, as the pellet is located always in the same position, the gamma counter results became more accurate and reproducible.

Apart from the methodological protocol, several assay requirements were taken into account. As commented in Chapter 5, the final results of an evaluated sample are given as relative specific binding (RSB). Therefore, in order to calculate such percentage two different calibrators are 159 required, on one hand the blank calibrator (without addition of any competitor, giving the maximum binding) and on the other the saturation calibrator (containing a saturation concentration of competitor, yielding only the nonspecific binding). These calibrators are used to set to 100% and 0% of RSB, respectively. In the presented work the saturation of the saturation calibrator was achieved by the addition of 7.5 mM of GHRP-2 although depending on the availability, any other secretagogue could be used provided the concentration is adjusted according to the affinity of the ligand. Despite that cell aliquots are standardized in cell number, the use of the RSB percentage avoided large variation in the number of receptors present in the incubation media, as the blank calibrator could compensate for this variation.

Despite that the methodology is highly sensitive, a purification process must be applied to the urine sample in order to avoid interferences. The major aim of using receptors for determining the presence of GHS is the universality that it brings although any non-specific purification process could distort the situation. Thus, one ideal purification step that can be applied is to use the receptors also for this purpose. In the purification step an excess of receptors is used to ascertain capturing as much of the secretagogue present in the sample and after a detaching procedure, the eluting media is evaporated and reconstituted in a minimal volume to increase the final GHS concentration. Optimum purification was obtained with 50-fold receptor excess, as compared with the 160 final competition-binding step, is used, as mentioned in chapter 5.

Another assay requirement is the desalting process that must be applied to the sample. It is known that the addition of high concentrations of NaCl (up to 100 mM) can cause an allosteric modulation of the ligand-receptor interaction [56;57] and therefore, result in a substantial decrease of the final binding. This could provoke false positives in the anti-doping assay. As such, desalting is mandatory and this has to be done prior to the purification step because salts will interfere also in the purification when incubating with excess receptor. In chapter 4 we reported that for samples with high salt content, the resulting specific binding is only between 25-35% of the normal specific binding when salts are not present. For sample desalting, several methodologies were tested but the solid phase extraction was chosen because it is relatively fast, or high throughput capacity, and also permits to concentrate the analyte as it is eluted in methanol and the solvent evaporated under nitrogen. Furthermore, this procedure is already common practice for many other anti-doping protocols. Solid phase extraction was carried out using Oasis HLB cartridges (Waters, Mildford, USA) that consists in a hydrophilic-lipophilic-balanced reverse phase sorbent that is supposed to retain acids, bases and neutrals. All tested GHS but ghrelin, were retained efficiently by the cartridge and eluted when methanol was added. Despite that the GHS structures are diverse, an important feature of GHS bioactivity 161 is the presence of basic amino groups [55], which also play a role in retention by SPE. Thus, theoretically, this type of procedure should retain any secretagogue. Even though, periodical control experiments with newly developed GHS should be done in order to confirm that no secretagogue escapes the SPE purification.

6.1.4 Summary

Taking the results altogether, it can be concluded that the methodology is generally applicable for detection of growth hormone secretagogues in biological samples.

To substantiate this claim samples from an administration study were evaluated and the secretagogue could be quantified by comparison with the calibration curve obtained for this same secretagogue. Following comparison with the original study by Okano et al. [20] who employed HPLC-MS for the detection and quantification, the conclusion is that the presented methodology is as sensitive as an HPLC-based technique, at least for intact GHRP-2. It should be mentioned that it has been impossible to evaluate if the presented minimal positive concentration (MPC) of GHS other than GHRP-2 suits the human excretion in an administration study. Thus far limited information has been released on this matter and on only few secretagogues. The MPC as determined for

162 each secretagogue with the technique here presented depends to a large extend on the affinity of each secretagogue and its capacity to compete with the radioligand for the receptor site, while the limit of detection (LOD) of HPLC-based methodology may only depend on the purification process. However, one should certainly bear in mind that the dose, administration route, bioavailability of the secretagogue, metabolism, and excretion ratio will principally determine the concentration of every given secretagogue that ultimately reaches the urine. Thomas et al. [21], also using LC-MS, showed similar limit of detection for two pure reference standards of secretagogues (GHRP-2 and Alexamorelin) that display different affinity for the GHS receptor. However, for the athlete to obtain similar physiological response the dose will be higher for the lower affinity GHS than for the higher affinity GHS and therefore most probably higher concentrations will result in urine. In practice, similar limits of detection for HPLC will represent different pharmacodynamic detection between the two secretagogues. The receptor assay, however, will react similar, as the different affinities of each GHS for the receptor will be probably counteracted by the differences in urinary concentrations. The only secretagogue that is not properly detected is ghrelin itself as it is apparently retained in the desalting procedure. Rather than being a drawback, this is beneficial as if an adverse finding is detected it cannot be attributed to an elevation of endogenous ghrelin

163 concentration in circulation. Even more, ghrelin is not likely to have much penetrance in the black market, as the pharmaceutical industry has never considered this molecule a true candidate for GH related therapy.

Another issue to bear in mind is the window of opportunity (WOO). Thus far, for the single compound that has found commercial application GHRP-2, the WOO was established to 4,5 hours post administration, both for the LC-MS approach as well as the competition binding assay employing excretion study samples. In that study Okano screened for GHRP-2 as well as for several metabolites. One of these, AA- 3, could be detected up to 24 hours post administration. Later on Thomas and coworkers [21] repeated a pilot study from Okano with an oral administration of 10 mg GHRP-2 and they were also able to detect the metabolites for 20 hours post administration. Considering the metabolites for each GHS, the capacity of the competition assay to detect them will depend upon the affinity of the metabolite to bind to the receptor. As discussed in chapter 5, apparently the AA-3 metabolite of GHRP-2 does not appear to have affinity for the receptor. Our approach will therefore not be sensitive to inactive metabolites and therefore our approach may be more specific but less sensitive as compared with metabolites- related approaches.

Regarding costs, the competition assay is, despite using affinity-based reagents that are not usual in anti-doping

164 laboratories, fairly cheap. For the fully optimized procedure and considering that the method consists in several different procedures the estimated cost is approximately 5 € per sample (20% is receptor production, another 20% corresponds to the radioligand, and the 60% remaining is destined to the SPE column). Despite the radioligand costs around 800€, there is sufficient material to spike run approximately 850 samples at a level of 20000 cpm, the current quantity of [125I-His9]-ghrelin in the assays. For the cell culture, the culture medium costs around 0,05€/ml (DMEM, FBS, Penicillin/Streptomycin, L-Glutamine and G-418) and bearing in mind culture dishes for maintenance and amplification, 50 million cells could be harvested investing less than 20€. The farmost expensive material is the SPE column which costs 3€ per column and it is a mandatory expense per sample. In any case, the estimated costs are well within the standard range of expenses for anti-doping purposes.

Regarding the number of steps and the duration of the protocol, the assay could be divided in two different sections, -1- the sample purification, and -2- the competition-binding assay.

For the desalting process, samples were eluted from de SPE cartridge in methanol and dried in a Turbo Vap that typically lasted 120 min. After extraction of secretagogues from the desalted sample using excess receptor, the GHS were

165 detached from the receptor with an acetic acid solution, which also required drying using a Speed Vac, typically during 2-3 hours. Once more, the instrumental set-up will certainly influence the throughput of the procedure, i.e. all steps are prone to be automated and robotic handling certainly boosts both speed and capacity. However, for manual processing, as is custom in most anti-doping procedures, up to 18 samples can be handled simultaneously, with an approximate elaboration time of 5 hours.

The duration of the competition-binding procedure depends on the number of samples that can be processed simultaneously. In this case the centrifugational separation capacity is certainly a limiting factor. When employing a standard 24-tube centrifuge, a batch containing 6 samples and 2 calibrators can be analysed in triplicate in approximately 70 minutes. With further optimisation and careful planning up to three batches can be analysed in approximately 3 hours. Thus, if only one sample is analysed, such analysis could be performed within 4 hours, and the same time would be employed to analyse up to 6 samples. With careful planning a maximum of 18 samples (and the corresponding calibrators) could be analysed in 8 hours.

166

6.2 Future improvements and alternative methodologies

6.2.1 Future improvements

The established protocol is fully functional and can be implemented for anti-doping purposes. Moreover, it is currently undergoing validation and with these data the procedure might be further endorsed by the governing bodies. However, several aspects of the procedure could be further optimised or changed otherwise to meliorate the protocol. In the following section several points are addressed that could be improved.

For instance, all blank samples used to determine the threshold limit were collected in at random daytime anticipating that the ghrelin fluctuation due to food intake does not play a role [58-61]. It has been reported that plasma ghrelin levels in patients with anorexia nervosa are markedly elevated (500 pg/ml the control group and 1050 pg/ml the anorectic patients) [62] and these elevated ghrelin levels return toward normality after refeeding. It has to be considered that ghrelin levels detailed in most studies refer to the sum of des-octanoyl-ghrelin and ghrelin, and that they coexist in an approximate 9 to 1 ratio, respectively. Only the latter has the ability to bind to the receptor. Moreover, the

167 excretion ratio in urine for endogenous ghrelin is about 60% of circulating [63]. Bearing in mind that the urine concentration of ghrelin presented by Yoshimoto et al [63] is about 10 pM (33,7 pg/ml) and that the minimal positive concentration for ghrelin as presented in Chapter 5 is about 10 nM (33,7 ng/ml) it is expected that ghrelin, even under extreme circumstances, shall not interfere. In addition, we have verified that ghrelin itself is not extracted by the solid phase step of the procedure. Nevertheless, in order to fully validate this aspect samples collected before and after food ingestion should be tested in order to corroborate that endogenous ghrelin do not interfere with the test.

Moreover, the threshold value where a sample is considered positive or negative has been established by testing blank urine samples collected from Caucasian people and other human ethnics should be evaluated. In this context, in the application of the protocol to the samples from the GHRP-2 excretion study carried out in Mongoloid individuals, all the pre administration samples gave a relative specific binding (RSB) values that were in line with those obtained for Caucasians indicating that ethnicity does not seem to play a role.

It also seems imperative to test for GHS in human samples after administration of other secretagogues so as to determinate the windows of opportunity for secretagogues different than GHRP-2. This objective, however, is extremely

168 difficult to accomplish as collecting urine samples from a real excretion study are subject to clinical permits that are hardly granted to non-approved medications. Obviously, pharmaceutical developments are accompanied by human studies as part of the development but biological samples from these studies are seldom shared. Thus, dedicated study for each secretagogue must be considered once the product is on the market, be it as a pharmaceutical or in the form of a dietary supplement. Nowadays, only GHRP-2 is commercially available as an injectable drug (Kaken Pharmaceutical, Tokyo, Japan) and several other secretagogues are supposed to be commercialized in the near future.

One other aspect that needs careful consideration is the metabolic fate of the known molecules, and more importantly, how does the competition assay address these metabolites. As with the alleged activity of the truncated ghrelin peptides, if a metabolite maintains the capacity to bind to the receptor the assay will still be functional. Additionally, if the metabolite has a longer half-life than the original molecule, this may be beneficial for the assay. In such scenario the competition assay could display enhanced sensitivity as all molecules contribute to the observed displacement, in contrast to an LC- MS based approach where each molecule (GHS or metabolites) give rise to individual signals. However, if a metabolite looses the ability to interact with the receptor the contrary would be true.

169

Another aspect that could be improved is the biological fluid that is used to determine the presence of the GHS. The most suitable biological fluid to target is urine as it is of lower molecular complexity than serum/plasma, it is easily obtained and the extraction procedure is not invasive. On the down side, the excretion ratio is known to be rather low for most drugs because a part could be excreted via faeces or metabolized. Thus, it could be interesting to develop a purification protocol to extract the secretagogue from plasma or serum as this could enhance the window of detection. This hypothetic purification step should be based, once more, on the principle of universal applicability not to exclude any GHS. The use of receptors for GHS extraction may be, as in urine, apparently an optimal approach but the desalting process is anticipated to constitute a major hurdle due to cartridge blockage. An alternative for plasma purification has already been published [64] aiming to detect MK-0677 by liquid chromatography. This plasma purification was done by a liquid-liquid extraction from the aqueous into an organic layer but despite is the fact that it is effective for MK-0677, it must be confirmed whether other secretagogues can also be purified by this methodology.

The confirmation procedure, mandatory in an anti-doping procedure, also needs to be addressed. The initial approach was to use a competition binding for screening purposes and use of a different methodology, e.g. a LC-MS approach, to confirm the adverse finding. However, the use of a different 170 methodology is not imperative as illustrated for example by the analysis of recombinant erythropoietin. That protocol is based on an isoelectric focusing procedure followed by a double blotting assay and a result is given depending on an IEF pattern disclosed by an anti-erythropoietin antibody interaction [65;66]. The method fully relies on the specificity of the monoclonal antibody (clone AE7A5) and it is considered a critical reagent that shall not be changed. In that particular case no confirmatory procedure different from the described for screening is required. A parallelism with the GHS detection procedure presented here would be the receptor- ligand interaction that should be considered as specific as the antigen-antibody interaction for EPO. Alternatively, and for the sake of additional corroboration, another cell line, equally over-expressing the GHS-R1a, could be used for the confirmatory procedures as mentioned (vide supra).

6.2.2 Alternative methodologies

6.2.2.1 Receptor-based methodologies

A desirable modification could be the use of a non-radioactive label for ghrelin. Such alternative would maintain practically most of process described here but with a different detection principle, not requiring the extra measures of working with 171 radiolabeled molecules. Initially, radioactive labelling was chosen due to the sensibility. Bearing in mind this premises the only alternative that could provide a similar sensibility is chemiluminescence. Chemiluminescence is the emission of energy due to a chemical reaction with limited emission of heat and in biochemistry several chemiluminescent probes are used. Acridinium salts have proven to be useful chemiluminescent labels but they are light sensitive [67-69]. In the presence of alkaline hydrogen peroxide, acridinium- labelled proteins can be stimulated to emit light without the use of a catalyst and the protein can be detected by measuring at 430 nm in a standard luminometer.

The main advantage of using a chemiluminescent label compared with the radioactivity is the lower toxicity for men, absence of the need for special handling as well as the lack for special nuclear permission for the laboratory. Moreover, luminometers are instruments that are nowadays present in most anti-doping laboratories. During the course of the thesis, several labelling attempts using acridinium were made but products did not bind to the receptor. The lack of biological activity of the ghrelin-acridinium compound was attributed probably to the steric interference of the acridinium group, as it is a relatively bulky molecule compared to ghrelin (632 Da and 3370 Da respectively). Ghrelin-acridinium was submitted to different enzyme digestions in order to determine the position of the acridinium group in ghrelin demonstrating that the acridinium group was introduced in the amino acid stretch 172

21 and 24. Notwithstanding the distance from the allegedly relevant N-terminus, the acridinium label appeared to interfere in the affinity with the receptor. In addition to the loss of biological activity, other methodological drawbacks where encountered when using this chemiluminescent tagging procedure. In case of the radioactive drug labelling, the sediment formed prior the final counting does not require reconstitution as the radioactivity is detected anyway. In contrast, as the chemiluminescent label requires from a chemical reaction employing peroxidase it seems advisable to homogenize the sediment to reduce potential variability depending on the homogenization success.

Alternatively, biotinylation could be employed for the labelling of ghrelin. Biotinylated-ghrelin has already been used to determine the location of the GHSR-1 [70] in rat brain but, as it was used in excess (1M), an affinity comparison with the non-labelled ghrelin was not possible. The comparison of affinities between the labelled and non-labelled ligand is rather important as only high affinity ligands should be used as competitors. The position of biotinylation is also important, as with the usual 1-biotin-ghrelin, disruption of binding is expected. Thus, further research should be performed in order to determine the suitability of other labelling methodologies. In this context one should certainly consider the incorporation of additional amino acids, or modifying the sequence of ghrelin as recently described by Phoenix Pharmaceuticals (Porcine Ghrelin[-Biotinyl-Lis29]) 173

Ghrelin is also available carrying a fluorescent label, being this label rhodamine, fluorescein isothiocyanate (FITC), carboxyfluorescein (FAM), or cyanine (Cy5) group. In general, fluorescent molecules, mainly antibodies, are widely used to study the role in different metabolism mechanisms but are rarely used in a competition assay where a high sensitivity is required. Despite the different fluorescent ghrelin products, few publications have been encountered detailing its use. McGirr and coworkers have recently designed and synthesized a 1-18 truncated ghrelin analogue modified as Ser3 and fluorescein isothiocyanate was coupled at AA-18 [71]. Such fluorescein-modified ghrelin (1-18) was used for several immunohistochemical experiments presenting an affinity nearly 40 times lower than intact ghrelin. Depending on the desired methodology a lower affinity might not be seen as a handicap but a careful evaluation of the lower activity and potentially lower sensitivity is required.

Despite that fluorescence might not be an appropriate alternative to radioactive labelling, due to the intrinsically lower sensitivity, it can be used to monitor the response of an intact cell to the receptor-ligand interaction as shown in Chapter 3 and 4 by the use of fluorescent microscopy or flow cytometry. Both methodologies revealed the expected fluorescence when intracellular calcium was mobilized due to the GHS interplay with the cells transfected with GHS-R1a. These results confirmed both techniques as a potential alternative/support to the presented receptor-ligand 174 methodology. Fluorescent microscopy resulted to be a very laborious and time-consuming technique that required skilled personnel and expensive, dedicated, equipment. Moreover, the analysis of a single sample containing transfected and non-transfected cells required almost a days work. On the other hand, flow cytometry resulted to be faster as several samples could be processed within an hour. As mentioned in Chapter 3, several GHS at 30 nM (receptor saturation conditions) were tested and all of them presented significant calcium mobilization except for des-octanoyl-ghrelin. Obviously, no fluorescence rise was observed when non- transfected cells were used. Despite these promising results, no urine samples were analysed and no detection limits where established as the development was driven in the direction of the protocol that has been eventually described. Nevertheless, flow cytometry is a promising alternative that requires a dedicated study in order to determine its applicability to the present field.

The ligand labelling or the intracellular response monitoring are only two ways in which the fluorescence can be used to evaluate binding. Techniques as the dissociation-enhanced lanthanide fluorescent immunoassay (DELFIA) or the fluorescence resonance energy transfer (FRET) have been successfully used to define the interaction between ligands and G-protein-coupled receptors [72-77]. In DELFIA, the ligand is labelled with a lanthanide chelate (generally Europium), which absorbs the excitation light and transfers it 175 to the lanthanide that will produce a specific emission spectrum. In FRET, both ligand and receptor must carry a fluorescent group that acts as donor and acceptor. The donor group will absorb an exciting light and transfer the quantisized energy to the acceptor that will emit at a particular wavelength. This energy transfer can only occur when both fluorescent groups are close enough in space. However, these two fluorescent-based techniques have been used to develop several binding assays for the receptors other than GHSR-1a. What has been published is a Tag-lite approximation which combines homogeneous time resolved fluorescence detection with a covalent labelling called SNAP- tag resulting in a combination of FRET and DELFIA approximations [33]. In that study, the receptor (GHS-R1a) is fused with the SNAP-tag enzyme and transiently expressed at the cell surface. The SNAP-GHS-R1a construct was then covalently labelled with a terbium cryptate (Lumi4-Tb). On binding of a red fluorescent derivative of ghrelin (red ghrelin), FRET occurred between the Lumi4-Tb and red-ghrelin. When the affinity of the red-ghrelin molecule was compared to that of intact ghrelin in a competition assay against a radioactive 125 9 ligand ([ I-His ]-ghrelin) it could be shown that the Ki of red- ghrelin was nearly equivalent to the non-labelled and labelled ghrelin (19 nM and 12nM, respectively) indicating that the fluorophore on ghrelin molecule does not interfere in its binding capacity. This is remarkable in view of our findings with the acridinium ester modification in the N-terminal region

176 and the nearly complete loss of receptor affinity. Finally, non- labelled ghrelin and MK-0677 were tested as competitors in the radioactive and in the Tag-lite assays obtaining similar results for both methodologies despite using different concentration for [125I-His9]-ghrelin (1 nM) and red-ghrelin (3 nM). This technique could be apparently a valuable alternative to the radioactive labelling; only slight differences are obtained in the matter of sensitivity. In contrast, it should be mentioned that the Tag-lite methodology may be much more time-consuming as the cell manipulation procedures are far more laborious. Moreover, cell transfection procedure is presented as a transient transfection with the SNAP-GHSR- 1a plasmid indicating that each cell batch could present a different receptor expression and therefore introduce some bias in the assay. In conclusion, this Tag-lite approximation could be appropriate as an alternative for the radioactive methodology, provided a stable expression system can be set-up,

6.2.2.2 Non receptor-based methodologies

The proposed methodology was conceived as a screening test prior to a subsequent confirmation step based on an LC- MS procedure. As presented in Chapter 5, an LC-MS approach provided practically equal sensitivity than the radioligand competition technique when the studied 177 compound was intact GHRP-2, but when comparing limits of detection presented for low-affinity secretagogues [21] major differences appear. The detection limit of an LC-MS assay does not depend on the affinity of the secretagogue for the receptor and it may also detect the metabolites irrespective of the affinity for the receptor. Actually, limits of detection for GHRP-2, GHRP-1, GHRP-6, GHRP-5, GHRP-4, Alexamorelin, Hexarelin, Ipamorelin and MK-0677 have been published being on the order of nanograms per millilitre [21]. In addition to the high sensitivity, independence from the receptor affinity and the possibility to detect the metabolites, the LC-MS methodology could also present other benefits as the potentially higher throughput of the final analysis (autosamplers with large capacity). Otherwise, as commented earlier, it may also present some disadvantages. In order to monitor the presence of the secretagogue or a metabolite, prior information on the elution time, fragmentation pattern, etc. should be available. In a recent presentation at the annual anti-doping workshop, scientists from the German Sport University of Cologne presented an elegant study on the potential metabolism of six growth hormone releasing peptides through incubation with fresh human blood. They found that trough the action of three different enzymes (amidase, aminoexopeptidase, carboxyexopaptidase), up to 28 different metabolites can be detected [78] following in-vitro incubation. Thus, in a scenario where low amounts of possible analytes to be detected, the LC-MS technique would

178 be rather optimal but, considering the large secretagogue family, the fact that is still growing [17;79-85], and that many possible metabolites have to be taken into account, renders the approximation less plausible. It is worth to mention that the sensitivity of the LC-MS methodology depends to a large extend on the efficacy of the purification process prior to the HPLC separation. While a solid phase extraction procedure yields a limit of detection between 0.3 and 0.5 ng/ml [20;21] a purification step based on a precipitation of proteins gives a limit of detection of 5 ng/ml [86]. The purification step or steps to implement will depend on the biologic substrate, e.g. urine or plasma.

In conclusion, any methodological approximation, being a receptor-based or a non receptor-based technique, will present advantages and disadvantages. In this thesis the former, as a generally applicable screening procedure for GHS abuse is presented; it is an alternative that is claimed to be functional, of high sensitivity and universal for all GHS, currently available or hitherto undiscovered. Obviously, it may require some optimization and further investigations, but general basis have been set.

179

6.3 References

1. World Anti-Doping Agency and List Comitee (2011). The 2011 Prohibited list. International standard. The World Anti-Doping Agency [ 2012 Available from: URL:http://www.wada- ama.org/Documents/World_Anti- Doping_Program/WADP-Prohibited- list/2012/WADA_Prohibited_List_2012_EN.pdf

2. Wallace JD, Cuneo RC, Bidlingmaier M, Lundberg PA, Carlsson L, Boguszewski CL et al. Changes in non-22- kilodalton (kDa) isoforms of growth hormone (GH) after administration of 22-kDa recombinant human GH in trained adult males. J Clin Endocrinol Metab 2001; 86(4):1731-1737.

3. Bidlingmaier M, Wu Z, Strasburger CJ. Test method: GH. Baillieres Best Pract Res Clin Endocrinol Metab 2000; 14(1):99-109.

4. Wu Z, Bidlingmaier M, Dall R, Strasburger CJ. Detection of doping with human growth hormone. Lancet 1999; 353(9156):895.

5. World Anti-Doping Agency. International standard for laboratories. The World Anti-Doping Agency [ 2009 Available from: URL:http://www.wada- ama.org/rtecontent/document/International_Standard_f or_Laboratories_v6_0_January_2009.pdf

6. World Anti-Doping Agency. hGH isoform differential immunoassays. The World Anti-Doping Agency [ 2012 Available from: URL:http://www.wada- ama.org/Documents/Science_Medicine/Anti-Doping- Labs-Guidelines/WADA_SC- Guidelines_hGH%20Differential%20Immunoassays_E N_June10.pdf

180

7. Copinschi G, Van OA, L'Hermite-Baleriaux M, Mendel CM, Caufriez A, Leproult R et al. Effects of a 7-day treatment with a novel, orally active, growth hormone (GH) secretagogue, MK-677, on 24-hour GH profiles, insulin-like growth factor I, and adrenocortical function in normal young men. J Clin Endocrinol Metab 1996; 81(8):2776-2782.

8. Howard AD, Feighner SD, Cully DF, Arena JP, Liberator PA, Rosenblum CI et al. A receptor in pituitary and hypothalamus that functions in growth hormone release. Science 1996; 273(5277):974-977.

9. Huhn WC, Hartman ML, Pezzoli SS, Thorner MO. Twenty-four-hour growth hormone (GH)-releasing peptide (GHRP) infusion enhances pulsatile GH secretion and specifically attenuates the response to a subsequent GHRP bolus. J Clin Endocrinol Metab 1993; 76(5):1202-1208.

10. Okano M, Nishitani Y, Sato M, Ikekita A, Kageyama S. Influence of intravenous administration of growth hormone releasing peptide-2 (GHRP-2) on detection of growth hormone doping: growth hormone isoform profiles in Japanese male subjects. Drug Test Anal 2010; 2(11-12):548-556.

11. Patchett AA, Nargund RP, Tata JR, Chen MH, Barakat KJ, Johnston DB et al. Design and biological activities of L-163,191 (MK-0677): a potent, orally active growth hormone secretagogue. Proc Natl Acad Sci U S A 1995; 92(15):7001-7005.

12. Murphy MG, Weiss S, McClung M, Schnitzer T, Cerchio K, Connor J et al. Effect of alendronate and MK-677 (a growth hormone secretagogue), individually and in combination, on markers of bone turnover and bone mineral density in postmenopausal osteoporotic women. J Clin Endocrinol Metab 2001; 86(3):1116- 1125.

181

13. Khojasteh-Bakht SC, O'donnell JP, Fouda HG, Potchoiba MJ. Metabolism, pharmacokinetics, tissue distribution, and excretion of [14C]CP-424391 in rats. Drug Metab Dispos 2005; 33(1):190-199.

14. Piccoli F, Degen L, MacLean C, Peter S, Baselgia L, Larsen F et al. Pharmacokinetics and pharmacodynamic effects of an oral ghrelin agonist in healthy subjects. J Clin Endocrinol Metab 2007; 92(5):1814-1820.

15. Broglio F, Benso A, Gottero C, Muccioli G, Deghenghi R, Ghigo E et al. Endocrine activities of alexamorelin (Ala-His-d-2-methyl-Trp-Ala-Trp-d-Phe-Lys-NH2), a synthetic GH secretagogue, in humans. Eur J Endocrinol 2000; 143(3):419-425.

16. Pandya N, DeMott-Friberg R, Bowers CY, Barkan AL, Jaffe CA. Growth hormone (GH)-releasing peptide-6 requires endogenous hypothalamic GH-releasing hormone for maximal GH stimulation. J Clin Endocrinol Metab 1998; 83(4):1186-1189.

17. Garcia JM, Polvino WJ. Pharmacodynamic hormonal effects of anamorelin, a novel oral ghrelin mimetic and growth hormone secretagogue in healthy volunteers. Growth Horm IGF Res 2009; 19(3):267-273.

18. Thevis M, Wilkens F, Geyer H, Schanzer W. Determination of therapeutics with growth-hormone secretagogue activity in human urine for doping control purposes. Rapid Commun Mass Spectrom 2006; 20(22):3393-3402.

19. Thevis M, Schanzer W. Emerging drugs--potential for misuse in sport and doping control detection strategies. Mini Rev Med Chem 2007; 7(5):531-537.

20. Okano M, Sato M, Ikekita A, Kageyama S. Determination of growth hormone secretagogue pralmorelin (GHRP-2) and its metabolite in human

182

urine by liquid chromatography/electrospray ionization tandem mass spectrometry. Rapid Commun Mass Spectrom 2010; 24(14):2046-2056.

21. Thomas A, Hoppner S, Geyer H, Schanzer W, Petrou M, Kwiatkowska D et al. Determination of growth hormone releasing peptides (GHRP) and their major metabolites in human urine for doping controls by means of liquid chromatography mass spectrometry. Anal Bioanal Chem 2011; 401(2):507-516.

22. Kohler M, Thomas A, Geyer H, Petrou M, Schanzer W, Thevis M. Confiscated black market products and nutritional supplements with non-approved ingredients analyzed in the Cologne Doping Control Laboratory 2009. Drug Test Anal 2010; 2(11-12):533-537.

23. Argente J, Garcia-Segura LM, Pozo J, Chowen JA. Growth hormone-releasing peptides: clinical and basic aspects. Horm Res 1996; 46(4-5):155-159.

24. Mericq V, Cassorla F, Bowers CY, Avila A, Gonen B, Merriam GR. Changes in appetite and body weight in response to long-term oral administration of the ghrelin agonist GHRP-2 in growth hormone deficient children. J Pediatr Endocrinol Metab 2003; 16(7):981-985.

25. Erotokritou-Mulligan I, Guha N, Stow M, Bassett EE, Bartlett C, Cowan DA et al. The development of decision limits for the implementation of the GH-2000 detection methodology using current commercial insulin-like growth factor-I and amino-terminal pro- peptide of type III collagen assays. Growth Horm IGF Res 2012; 22(2):53-58.

26. Powrie JK, Bassett EE, Rosen T, Jorgensen JO, Napoli R, Sacca L et al. Detection of growth hormone abuse in sport. Growth Horm IGF Res 2007; 17(3):220-226.

183

27. Yale University. Safety information and specific handling precautions for radionuclides H-3, C-14, S-35, P-32 and I-125. Yale University, Office of Enviromental Health and safety, New Haven, Connecticut [ 2012 Available from: URL:http://www.yale.edu/ehs/Documents/rad/Specific Handling.pdf

28. Katugampola SD, Pallikaros Z, Davenport AP. [125I- His(9)]-ghrelin, a novel radioligand for localizing GHS orphan receptors in human and rat tissue: upregulation of receptors with athersclerosis. Br J Pharmacol 2001; 134(1):143-149.

29. Guerlavais V, Boeglin D, Mousseaux D, Oiry C, Heitz A, Deghenghi R et al. New active series of growth hormone secretagogues. J Med Chem 2003; 46(7):1191-1203.

30. Gauna C, van de Zande B, van KA, Themmen AP, van der Lely AJ, Delhanty PJ. Unacylated ghrelin is not a functional antagonist but a full agonist of the type 1a growth hormone secretagogue receptor (GHS-R). Mol Cell Endocrinol 2007; 274(1-2):30-34.

31. Camina JP, Carreira MC, El MS, Llorens-Cortes C, Smith RG, Casanueva FF. Desensitization and endocytosis mechanisms of ghrelin-activated growth hormone secretagogue receptor 1a. Endocrinology 2004; 145(2):930-940.

32. Moulin A, Demange L, Berge G, Gagne D, Ryan J, Mousseaux D et al. Toward potent ghrelin receptor ligands based on trisubstituted 1,2,4-triazole structure. 2. Synthesis and pharmacological in vitro and in vivo evaluations. J Med Chem 2007; 50(23):5790-5806.

33. Leyris JP, Roux T, Trinquet E, Verdie P, Fehrentz JA, Oueslati N et al. Homogeneous time-resolved fluorescence-based assay to screen for ligands

184

targeting the growth hormone secretagogue receptor type 1a. Anal Biochem 2011; 408(2):253-262.

34. Holst B, Brandt E, Bach A, Heding A, Schwartz TW. Nonpeptide and peptide growth hormone secretagogues act both as ghrelin receptor agonist and as positive or negative allosteric modulators of ghrelin signaling. Mol Endocrinol 2005; 19(9):2400- 2411.

35. Hernandez AS, Cheng PT, Musial CM, Swartz SG, George RJ, Grover G et al. Discovery, synthesis, and structure-activity studies of tetrazole based growth hormone secretagogues. Bioorg Med Chem Lett 2007; 17(21):5928-5933.

36. Muccioli G, Ghe C, Ghigo MC, Papotti M, Arvat E, Boghen MF et al. Specific receptors for synthetic GH secretagogues in the human brain and pituitary gland. J Endocrinol 1998; 157(1):99-106.

37. Torsello A, Ghe' C, Bresciani E, Catapano F, Ghigo E, Deghenghi R et al. Short ghrelin peptides neither displace ghrelin binding in vitro nor stimulate GH release in vivo. Endocrinology 2002; 143(5):1968-1971.

38. Papotti M, Ghe C, Cassoni P, Catapano F, Deghenghi R, Ghigo E et al. Growth hormone secretagogue binding sites in peripheral human tissues. J Clin Endocrinol Metab 2000; 85(10):3803-3807.

39. McKee KK, Palyha OC, Feighner SD, Hreniuk DL, Tan CP, Phillips MS et al. Molecular analysis of rat pituitary and hypothalamic growth hormone secretagogue receptors. Mol Endocrinol 1997; 11(4):415-423.

40. Nagamine J, Nagata R, Seki H, Nomura-Akimaru N, Ueki Y, Kumagai K et al. Pharmacological profile of a new orally active growth hormone secretagogue, SM- 130686. J Endocrinol 2001; 171(3):481-489.

185

41. Pomes A, Pong SS, Schaeffer JM. Solubilization and characterization of a growth hormone secretagogue receptor from porcine anterior pituitary membranes. Biochem Biophys Res Commun 1996; 225(3):939-945.

42. Smith RG, Pong SS, Hickey G, Jacks T, Cheng K, Leonard R et al. Modulation of pulsatile GH release through a novel receptor in hypothalamus and pituitary gland. Recent Prog Horm Res 1996; 51:261-285.

43. Bednarek MA, Feighner SD, Pong SS, McKee KK, Hreniuk DL, Silva MV et al. Structure-function studies on the new growth hormone-releasing peptide, ghrelin: minimal sequence of ghrelin necessary for activation of growth hormone secretagogue receptor 1a. J Med Chem 2000; 43(23):4370-4376.

44. Bylund DB, Toews ML. Radioligand binding methods: practical guide and tips. Am J Physiol 1993; 265(5 Pt 1):L421-L429.

45. Casanueva FF. Physiology of growth hormone secretion and action. Endocrinol Metab Clin North Am 1992; 21(3):483-517.

46. Sun Y, Ahmed S, Smith RG. Deletion of ghrelin impairs neither growth nor appetite. Mol Cell Biol 2003; 23(22):7973-7981.

47. Gualillo O, Lago F, Casanueva FF, Dieguez C. One ancestor, several peptides post-translational modifications of preproghrelin generate several peptides with antithetical effects. Mol Cell Endocrinol 2006; 256(1-2):1-8.

48. Garcia MC, Lopez M, Alvarez CV, Casanueva F, Tena- Sempere M, Dieguez C. Role of ghrelin in reproduction. Reproduction 2007; 133(3):531-540.

49. Dean DC, Nargund RP, Pong SS, Chaung LY, Griffin P, Melillo DG et al. Development of a high specific activity sulfur-35-labeled sulfonamide radioligand that allowed 186

the identification of a new growth hormone secretagogue receptor. J Med Chem 1996; 39(9):1767-1770.

50. Jacks T, Smith R, Judith F, Schleim K, Frazier E, Chen H et al. MK-0677, a potent, novel, orally active growth hormone (GH) secretagogue: GH, insulin-like growth factor I, and other hormonal responses in beagles. Endocrinology 1996; 137(12):5284-5289.

51. Martin-Pastor M, De CA, Alvarez CJ, Diaz-Hernandez MD, Jimenez-Barbero J, Casanueva FF et al. Interaction between ghrelin and the ghrelin receptor (GHS-R1a), a NMR study using living cells. Bioorg Med Chem 2010; 18(4):1583-1590.

52. Sun Y, Wang P, Zheng H, Smith RG. Ghrelin stimulation of growth hormone release and appetite is mediated through the growth hormone secretagogue receptor. Proc Natl Acad Sci U S A 2004; 101(13):4679-4684.

53. Smith RG, Cheng K, Schoen WR, Pong SS, Hickey G, Jacks T et al. A nonpeptidyl growth hormone secretagogue. Science 1993; 260(5114):1640-1643.

54. Smith RG, van der Ploeg LH, Howard AD, Feighner SD, Cheng K, Hickey GJ et al. Peptidomimetic regulation of growth hormone secretion. Endocr Rev 1997; 18(5):621-645.

55. Feighner SD, Howard AD, Prendergast K, Palyha OC, Hreniuk DL, Nargund R et al. Structural requirements for the activation of the human growth hormone secretagogue receptor by peptide and nonpeptide secretagogues. Mol Endocrinol 1998; 12(1):137-145.

56. Horstman DA, Brandon S, Wilson AL, Guyer CA, Cragoe EJ, Jr., Limbird LE. An aspartate conserved among G-protein receptors confers allosteric regulation

187

of alpha 2-adrenergic receptors by sodium. J Biol Chem 1990; 265(35):21590-21595.

57. Clark EA, Hill SJ. Differential effect of sodium ions and guanine nucleotides on the binding of thioperamide and clobenpropit to histamine H3-receptors in rat cerebral cortical membranes. Br J Pharmacol 1995; 114(2):357-362.

58. Cummings DE, Purnell JQ, Frayo RS, Schmidova K, Wisse BE, Weigle DS. A preprandial rise in plasma ghrelin levels suggests a role in meal initiation in humans. Diabetes 2001; 50(8):1714-1719.

59. Tschop M, Wawarta R, Riepl RL, Friedrich S, Bidlingmaier M, Landgraf R et al. Post-prandial decrease of circulating human ghrelin levels. J Endocrinol Invest 2001; 24(6):RC19-RC21.

60. Gualillo O, Caminos JE, Nogueiras R, Seoane LM, Arvat E, Ghigo E et al. Effect of food restriction on ghrelin in normal-cycling female rats and in pregnancy. Obes Res 2002; 10(7):682-687.

61. Toshinai K, Mondal MS, Nakazato M, Date Y, Murakami N, Kojima M et al. Upregulation of Ghrelin expression in the stomach upon fasting, insulin- induced hypoglycemia, and leptin administration. Biochem Biophys Res Commun 2001; 281(5):1220- 1225.

62. Otto B, Cuntz U, Fruehauf E, Wawarta R, Folwaczny C, Riepl RL et al. Weight gain decreases elevated plasma ghrelin concentrations of patients with anorexia nervosa. Eur J Endocrinol 2001; 145(5):669-673.

63. Yoshimoto A, Mori K, Sugawara A, Mukoyama M, Yahata K, Suganami T et al. Plasma ghrelin and desacyl ghrelin concentrations in renal failure. J Am Soc Nephrol 2002; 13(11):2748-2752.

188

64. Constanzer ML, Chavez-Eng CM, Matuszewski BK. Determination of a novel growth hormone secretagogue (MK-677) in human plasma at picogram levels by liquid chromatography with atmospheric pressure chemical ionization tandem mass spectrometry. J Chromatogr B Biomed Sci Appl 1997; 693(1):131-137.

65. Catlin D, Nissen-Lie G, Howe C, Pascual JA, Lasne F, Saugy M. Harmonization of the method for the identification of epoetin alfa and beta (rEPO) and darboepoetin alfa (NESP) by IEF-Double blotting and chemiluminescent detection. 2012. WADA. Ref Type: Online Source

66. Lasne F, Thioulouse J, Martin L, de CJ. Detection of recombinant human erythropoietin in urine for doping analysis: interpretation of isoelectric profiles by discriminant analysis. Electrophoresis 2007; 28(12):1875-1881.

67. Miller SA, Morton MS, Turkes A. Chemiluminescence immunoassay for progesterone in plasma incorporating acridinium ester labelled antigen. Ann Clin Biochem 1988; 25 ( Pt 1):27-34.

68. Weeks I, Beheshti I, McCapra F, Campbell AK, Woodhead JS. Acridinium esters as high-specific- activity labels in immunoassay. Clin Chem 1983; 29(8):1474-1479.

69. Weeks I, Sturgess M, Brown RC, Woodhead JS. Immunoassays using acridinium esters. Methods Enzymol 1986; 133:366-387.

70. Abizaid A, Liu ZW, Andrews ZB, Shanabrough M, Borok E, Elsworth JD et al. Ghrelin modulates the activity and synaptic input organization of midbrain dopamine neurons while promoting appetite. J Clin Invest 2006; 116(12):3229-3239.

189

71. McGirr R, McFarland MS, McTavish J, Luyt LG, Dhanvantari S. Design and characterization of a fluorescent ghrelin analog for imaging the growth hormone secretagogue receptor 1a. Regul Pept 2011; 172(1-3):69-76.

72. Inglese J, Samama P, Patel S, Burbaum J, Stroke IL, Appell KC. Chemokine receptor-ligand interactions measured using time-resolved fluorescence. Biochemistry 1998; 37(8):2372-2377.

73. Gao X, Hsu CK, Heinz LJ, Morin J, Shi Y, Shukla NK et al. Europium-labeled melanin-concentrating hormone analogues: ligands for measuring binding to melanin-concentrating hormone receptors 1 and 2. Anal Biochem 2004; 328(2):187-195.

74. Martikkala E, Lehmusto M, Lilja M, Rozwandowicz- Jansen A, Lunden J, Tomohiro T et al. Cell-based beta2-adrenergic receptor-ligand binding assay using synthesized europium-labeled ligands and time- resolved fluorescence. Anal Biochem 2009; 392(2):103-109.

75. Ilien B, Franchet C, Bernard P, Morisset S, Weill CO, Bourguignon JJ et al. Fluorescence resonance energy transfer to probe human M1 muscarinic receptor structure and drug binding properties. J Neurochem 2003; 85(3):768-778.

76. Tahtaoui C, Guillier F, Klotz P, Galzi JL, Hibert M, Ilien B. On the use of nonfluorescent dye labeled ligands in FRET-based receptor binding studies. J Med Chem 2005; 48(24):7847-7859.

77. Hachet-Haas M, Balabanian K, Rohmer F, Pons F, Franchet C, Lecat S et al. Small neutralizing molecules to inhibit actions of the chemokine CXCL12. J Biol Chem 2008; 283(34):23189-23199.

190

78. Metabolism of Ghrowth Hormone Releasing Peptides (GHRP). 2012.

79. Maletinska L, Pychova M, Holubova M, Blechova M, Demianova Z, Elbert T et al. Characterization of new stable ghrelin analogs with prolonged orexigenic potency. J Pharmacol Exp Ther 2012; 340(3):781-786.

80. Costantini VJ, Vicentini E, Sabbatini FM, Valerio E, Lepore S, Tessari M et al. GSK1614343, a novel ghrelin receptor antagonist, produces an unexpected increase of food intake and body weight in rodents and dogs. Neuroendocrinology 2011; 94(2):158-168.

81. Potter R, Horti AG, Ravert HT, Holt DP, Finley P, Scheffel U et al. Synthesis and in vivo evaluation of (S)-6-(4-fluorophenoxy)-3-((1-[11C]methylpiperidin-3- yl)methyl)-2-o-tolylquinazol in-4(3H)-one, a potential PET tracer for growth hormone secretagogue receptor (GHSR). Bioorg Med Chem 2011; 19(7):2368-2372.

82. Borel AG, Jones TM, Barbuch RJ, Jackson DA, Kulanthaivel P, Mattiuz E et al. Metabolism of LY654322, a growth hormone secretagogue, to an unusual diimidazopyridine metabolite. Drug Metab Dispos 2011; 39(5):740-749.

83. Tomita D, Yamatsugu K, Kanai M, Shibasaki M. Enantioselective synthesis of SM-130686 based on the development of asymmetric Cu(I)F catalysis to access 2-oxindoles containing a tetrasubstituted carbon. J Am Chem Soc 2009; 131(20):6946-6948.

84. Davulcu AH, McLeod DD, Li J, Katipally K, Littke A, Doubleday W et al. Process research and development for a tetrazole-based growth hormone secretagogue (GHS) pharmaceutical development candidate. J Org Chem 2009; 74(11):4068-4079.

85. MacLean CM, Casanova AT, Baselgia-Jeker L, Neave N, Larsen F, Skillern L et al. Effect of food on the

191

pharmacokinetics and pharmacodynamics of an oral ghrelin agonist (ARD-07) in healthy subjects. J Clin Pharmacol 2009; 49(5):553-559.

86. Gil J, Cabrales A, Reyes O, Morera V, Betancourt L, Sanchez A et al. Development and validation of a bioanalytical LC-MS method for the quantification of GHRP-6 in human plasma. J Pharm Biomed Anal 2012; 60:19-25.

192

7 Chapter 7:

CONCLUSIONS

193

1) A protocol for GHS detection has been successfully developed and applied to human urine samples. The methodology is based on two different steps: a sample pre- treatment and a competitive receptor-binding assay.

2) For the sample pre-treatment step the critical aspects are:

• Urine samples need to be desalted for the proper interaction of the putative GHS with the receptor at the subsequent binding assay step. • The desalted sample requires further purification to eliminate interfering agents. The method proposed takes advantage of the specificity afforded by purification through an excess of the same receptor preparation used for the binding assay.

3) For the receptor-binding assay the critical aspects are:

• Intact HEK293 engineered cells or isolated membranes thereof can be used as the GHS-R1a scaffold. Intact cells provide a faster protocol that work under more natural-like conditions, whereas isolated membranes deliver higher specific binding and better tolerate changes in the incubation medium. • A highly sensitive labelled ligand is needed. Radioactive labelling has been shown to offer enough sensitivity to detect GHS in real samples. Any potential future change towards a non-radioactive ligand will require reaching at least the same sensitivity.

194

4) A threshold value of a minimum of 76,3% of relative specific binding (RSB) for the labelled ligand in the binding assay has been established to differentiate any potential GHS containing sample from blank urine samples. Such value is considered to entail a probability if 1 in 1000 to obtain a false positive. This probability is acceptable for a screening method given the anti-doping protocols that includes an independent confirmatory procedure.

5) Different variables such as age, gender or exercise have no effect on the threshold limit.

6) The developed methodology has been tested with 8 different structurally diverse GHS and has shown functionality for all of them.

7) The approach has been successfully applied to samples obtained from administration studies with the only commercially existing GHS (GHRP-2: pralmorelin). The concentrations obtained and the time of detection for the unaltered compound resulted to be similar to those obtained with a recent GHRP-2 targeted LC-MS methodology.

195

196

8 Chapter 8:

APENDIX

197

216