Thesis Tom Verduyckt - Groep Biomedische Wetenschappen KU

KATHOLIEKE UNIVERSITEIT LEUVEN FACULTEIT FARMACEUTISCHE WETENSCHAPPEN

RADIOLABELLED THIOFLAVIN-T DERIVATIVES FOR IN VIVO DIAGNOSIS OF PERIPHERAL AMYLOIDOSIS AND ALZHEIMER’S DISEASE

PROEFSCHRIFT TER VERKRIJGING VAN DE GRAAD VAN DOCTOR IN DE FARMACEUTISCHE WETENSCHAPPEN

door

TOM VERDUYCKT

LEUVEN 2004

Auditorium Kasteel van Arenberg Kasteelpark Arenberg 1

Heverlee

17 september 2004, 17 uur

Promotor: Professor Dr. A. Verbruggen Co-promotor: Professor Dr. G. Bormans

Faculteit Farmaceutische Wetenschappen Laboratorium voor Radiofarmaceutische chemie UZ Gasthuisberg Herestraat 49 B-3000 Leuven

In Alzheimer’s disease the mind dies first: names, dates, places – the interior scrapbook of an entire life – fade into mists of nonrecognition. Matt Clark

DANKWOORD

There is a light at the end of the tunnel. Wat vier jaar geleden begon en toen een schier onmogelijke opdracht leek, staat nu gedrukt in deze thesis. Vier jaar is niets in een mensenleven, maar vier jaar kunnen een mens wel veel leren. Het leven van een wetenschappelijk onderzoeker is mijn inziens sterk vergelijkbaar met een zware Alpenrit tijdens de Tour de France. Steeds opnieuw doemen er nieuwe bergen op die smeken om beklommen te worden en steeds opnieuw is er die genoegdoening als een poging om de top te halen ook daadwerkelijk lukt. Veel hangt op zo’n moment af van je eigen instelling, maar het overwinnen van die toppen, en nog veel meer, het doorworstelen van de dalen, is het werk van velen. Een dankwoord is dan ook de ideale manier om aan de hele ploeg duidelijk te maken dat zij allemaal onmisbare schakels zijn geweest in het onderzoeksproces dat tot dit resultaat heeft geleid… Eerst en vooral wil ik uiteraard mijn promotor Prof. A. Verbruggen bedanken voor het toekennen van dit doctoraatsonderzoek aan iemand die als ‘gezondheidszorger’ eigenlijk gevormd was om de officina in te gaan. Bedankt voor het intrigerende onderwerp en voor de talloze kansen om mezelf wetenschappelijk bij te scholen (zowel binnen het labo als daarbuiten op congressen en symposia). Bedankt voor het nalezen en bijsturen van de thesistekst op plaatsen waar het nodig was… Vervolgens wil ik ook graag mijn co-promotor Prof. G. Bormans bedanken voor alle hulp met ongeacht welk toestel, gaande van simpele rekenmachientjes over HPLC en LC-MS toestellen, het cyclotron en zelfs een telefoonaansluiting. Bedankt ook voor het mede verschaffen van wetenschappelijk inzicht en kennis en het kritisch nalezen van de thesistekst. Uiteraard hoeft er geen verder uitleg bij het wetenschappelijk en minder wetenschappeliljk genot dat het laatste SNM congres in Philadelphia ons bracht… Een speciaal woordje van dank is weggelegd voor de leden van de examencommissie omwille van de tijd die zij hebben vrijgemaakt om dit lijvige werk kritisch na te lezen en hun opmerkingen door te geven om de tekst verder te verfijnen: Prof. P. Augustijns, Prof. A. Verbruggen, Prof. G. Bormans, Prof. F. Van Leuven, Prof. K. Van Laere en Prof. D. Guilloteau (Université François Rabellais, Tours, France). Woorden schieten te kort om Tjibbe en Jan te bedanken omwille van hun onmisbare hulp bij de chemische synthese van talloze verbindingen. Bedankt Tjibbe voor het opnemen van de NMR spectra, voor de hulp bij de interpretatie ervan, voor de discussies over al dan niet mogelijke reacties en de bereidheid om steeds mee te zoeken naar een mogelijke oplossing. Jan,

bedankt voor de zeer uitgesproken hulp bij de synthese van een aantal precursoren, liganden en eindproducten. Zonder jouw JC producten zou deze thesis een heel pak dunner zijn. Er allebei blijven voor gaan zou ik zeggen… Een fijne werkomgeving is een heel belangrijke schakel in het uitvoeren van een onderzoekswerk. Het Laboratorium voor Radiofarmaceutische Chemie heeft in de loop van de voorbije vier jaar een aantal medewerkers zien verdwijnen, maar over het algemeen is de groep jaar na jaar groter geworden. Met het risico sommigen te vergeten, wil ik mijn welgemeende dank uitdrukken aan Nancy, voor de hulp in het uitbreiden van mijn praktische (scheikunde) kennis; Kristin, die haar kennis over de merkingen met technetium-99m regelmatig wist door te spelen in mijn richting; Bert, voor de oeverloze discussies over ongeacht welk onderwerp (de boog kan niet altijd gespannen staan) en de hulp bij de merkingen; Marva, voor de hulp en het toelaten van mijn experimenten in het sowieso al bijzonder drukke PET schema; Christelle, voor de hulp bij de biodistributiestudies en het praktische merkingswerk; Dirk, thank you for your aid with the preparation of the technetium tricarbonyl precursor and the subsequent labelling; Kristof, voor de talrijke bereidingen van tricarbonyl kitten waar ik telkens een deeltje van mocht afsnoepen; Mireille, Kim, Sofie, Satish en Humphrey voor het aangename gezelschap; Davy en Dieter, voor de toffe momenten, zowel in als buiten het labo, of op een congres ergens ten velde (zoals in Spa bijvoorbeeld). Een buitengewone merci ook voor Dominique, ‘my partner in crime’. Allebei werken aan hetzelfde onderwerp en veelal moeten werken met dezelfde experimentele opstellingen kan voor wrevel zorgen, maar heeft het mijn inziens nooit gedaan. Bedankt Dominique voor al het gezaag dat je hebt moeten aanhoren toen ik er pas bijkwam, voor al de kilometers die je met mij binnen het UZ hebt afgelegd om me de weg te wijzen, voor al de hulp met tal van toestellen (waaronder vaak een laptop ressorteerde), voor het overbrengen van de kennis opgedaan bij Prof. H. F. Kung in Philadelphia, … Bedankt voor het gezelschap, de toffe momenten en de wederzijdse sfeer in het algemeen. Mijn thesis is mede door jouw ervaring tot stand gekomen en dus deels ook jouw werk. Merci. Verder wil ik graag Lynn bedanken voor het nalezen van mijn thesis en het aanbrengen van een aantal wijzigingen in de Engelse taal. Prof. W. Van Paesschen wil ik bedanken voor de hulp en de uitleg bij het voorbereiden van mijn doctoraatsseminarie over epilepsie. Prof. R. Vandenberghe bedank ik voor de nodige uitleg over de ziekte van Alzheimer bij de voorbereiding van mijn IWT verdediging. Prof. L. Mortelmans wil ik bedanken om mij de kans te geven mij wetenschappelijk te ontplooien binnen de dienst Nucleaire Geneeskunde van het UZ Gasthuisberg waar radiofarmacie een onderdeel van is. Francine (secretariaat Nucleaire Geneeskunde) wil ik bedanken voor allerlei hulp bij verzendopdrachten en typografie.

Prof. F. Van Leuven en Peter Borghgraef van het Centrum voor Menselijke Erfelijkheid ben ik uitermate erkentelijk voor het schenken van de transgene Alzheimer muizen en voor het uitvoeren van de bindingstesten op hersencoupes van deze Alzheimer muizen en op humane AD hersencoupes. I am indebted to Prof. H. F. Kung and his co-workers at the University of Pennsylvania for the execution of the affinity tests on human AD brain homogenates and the donation of 125I-TZDM together with the necessary knowledge to perform the affinity studies on fibrillar amyloid β. Ik ben ook veel dank verschuldigd aan het instituut voor de aanmoediging van Innovatie door Wetenschap en Technologie in Vlaanderen (IWT) voor de financiële steun die het mogelijk maakte dit project tot een goed einde te brengen. Last but not least, een onmetelijk grote ‘dank u’ voor mijn ouders die mij alle mogelijke kansen gegeven hebben om mij intellectueel en spiritueel te ontwikkelen, ondanks het feit dat de drang om muzikant of een iets minder intellectueel iemand te worden soms toch wel zeer groot was/is. Mijn vrouw Elke verdient ongetwijfeld het meeste lof omdat zij vier jaar lang al het gezaag en de problemen heeft moeten aanhoren die wetenschappelijk onderzoek nu eenmaal met zich schijnt mee te brengen. Over ongeveer twee jaar zal zij zelf haar doctoraatsproefschrift moeten verdedigen, en ik kan alleen maar hopen dat zij op dat moment evenveel aan mij zal hebben als ik nu aan haar. In goede en kwade dagen: vandaag is een fantastisch goede dag. Ik draag deze thesis op aan een aantal mensen die mij gemaakt hebben tot wie ik ben, maar er spijtig genoeg niet meer bij kunnen zijn (ons Moeke, Bert, onze Va, nonkel Luc en ons Moe) alsook aan een aantal mensen die altijd in mij zijn blijven geloven en nog steeds geloven, ondanks alles (mijn ouders, mijn echtgenote Elke, de papie, mijn broer, de ‘fritsteekbende’, Kris en Mr. Rossenbacker). Uit de grond van mijn hart, allemaal bedankt voor alles. Tom

LIST OF PUBLICATIONS

Publications in journals with peer review

Verduyckt, T.; Kieffer, D.; Huyghe, D.; Cleynhens, B.; Verbeke, K.; Verbruggen, A.; Bormans, G. Identity confirmation of Sestamibi and

99m

99m

Tc-MAG3,

99m

Tc-

Tc-ECD using radio-LC-MS. J Pharm Biomed Anal, 2003;

32: 669-678

Publications in journals without peer review

Verduyckt, T.; Cleynhens, B.; Verbruggen, A.; Bormans, G. Structure identification of

99m

Tc-MAG3-precomplex using radio-LC-MS. In: Nicolini, M.;

Mazzi, U.; eds. Technetium and Rhenium in Chemistry and Nuclear Medicine 6. Padova, Italy: SGE Editoriali; 2002: 243-245

Bormans, G.; Verbeke, K.; Vanderghinste, D.; Verduyckt, T.; Huyghe, D.; Kieffer,

D.;

Verbruggen,

A.

Radio-LC-MS

analysis

of

99m

Tc-labelled

radiopharmaceuticals. In: Nicolini, M.; Mazzi, U.; eds. Technetium and Rhenium in Chemistry and Nuclear Medicine 6. Padova, Italy: SGE Editoriali; 2002: 259-261

Bormans, G.; Vanderghinste, D.; Verduyckt, T.; Huyghe, D.; Kieffer, D.; Verbeke, K.; Verbruggen, A. Structure confirmation of Tc-99m-labeled complexes with radio-LC-MS. J Nuc Med, 2002; 43: 377P-378P (Suppl. S)

Contents DANKWOORD LIST OF PUBLICATIONS CONTENTS LIST OF ABBREVIATIONS 1. PERIPHERAL AMYLOIDOSIS AND ALZHEIMER’S DISEASE: INTRODUCTION

1

1.1. Introduction

1

1.2. Amyloid and amyloidosis

1

1.3. Peripheral amyloidosis

5

1.3.1. Amyloid deposits and pathogenesis of amyloidosis

5

1.3.2. Diagnosis of peripheral amyloidosis

8

1.3.3. Treatment of amyloidosis: novel strategies, potential therapies

14

1.4. Alzheimer’s disease

16

1.4.1. The pathogenesis of Alzheimer’s disease

18

1.4.2. Diagnosis of Alzheimer’s disease

31

1.4.3. Treatment of Alzheimer’s disease: current and future therapies

36

1.5. Aims of this study

39

2. DEVELOPMENT OF THIOFLAVIN-T DERIVATIVES FOR SPECT STUDIES

45

2.1. Single Photon Emission (Computed) Tomography (SPE(C)T)

45

2.2. Materials and methods

49

2.2.1. Synthesis of a benzothiazole-BAT conjugate as a precursor for a neutral

99m

2.2.2. Synthesis of BTA-BCL conjugates as precursors for negatively charged

99m

derivatives

Tc-BTA 53

2.2.3. Deprotection and labelling with 2.2.4. Analysis of

Tc-BTA derivative51

99m

Tc

99m

Tc-labelled reaction mixtures with RP-HPLC

59 61

2.2.5. Analysis of 99mTc-labelled compounds with radio-LC-MS

62

2.2.6. Partition coefficient determination

63

2.3. Results and discussion

64

2.3.1. Synthesis 2.3.2. Deprotection, labelling with

64 99m

Tc and RP-HPLC analysis

71

Contents 2.3.3. Radio-LC-MS analysis of the labelled compounds

77

2.3.4. Partition coefficients

84

2.4. Conclusion

85

3. DEVELOPMENT OF THIOFLAVIN-T DERIVATIVES LABELLED WITH A POSITRON EMITTER

87

3.1. Positron emission tomography (PET)

87

3.2. Materials and methods

90

3.2.1. Synthesis of neutral thioflavin-T derivatives for diagnosis of AD

91

18

99

11

100

3.2.2. Synthesis of 2-(4’-[ F]fluorophenyl)-1,3-benzothiazole (3.19) 3.2.3. Synthesis of 2-(4’-[ C]methylaminophenyl)-7-hydroxy-1,3-benzothiazole (3.20) 11

3.2.4. Synthesis of 2-(4’-[ C]methylaminophenyl)-5-hydroxy-1,3-benzothiazole (3.21) 11

102

3.2.5. Synthesis of 2-(4’-aminophenyl)-1-N-[ C]methyl-1,3-benzimidazole (3.22)

102

3.2.6. Analysis of PET tracers with radio-LC-MS

102

3.3. Results and discussion

103

3.3.1. Synthesis

103

3.3.2. Synthesis of 2-(4’-[18F]fluorophenyl)-1,3-benzothiazole

110

11

112

11

3.3.4. Synthesis of 2-(4’-[ C]methylaminophenyl)-5-hydroxy-1,3-benzothiazole (3.21)

114

3.3.5. Synthesis of 2-(4’-aminophenyl)-1-N-[11C]methyl-1,3-benzimidazole (3.22)

115

3.3.3. Synthesis of 2-(4’-[ C]methylaminophenyl)-7-hydroxy-1,3-benzothiazole (3.20)

3.4. Conclusion

117

4. BIODISTRIBUTION STUDIES OF THE RADIOLABELLED THIOFLAVIN-T DERIVATIVES

119

4.1. Introduction

119

4.2. Materials and methods

122

4.2.1. Preparation of radiolabelled thioflavin-T derivatives

122

4.2.2. Tissue distribution in normal mice

122

4.3. Results and discussion

123

4.3.1. Derivatives labelled with technetium-99m

123

4.3.2. Biodistribution studies of phenylbenzothiazoles labelled with carbon-11 or fluorine-18

131

4.4. Conclusion

137

Contents 5. IN VITRO STUDIES TO ASSESS THE AFFINITY OF THIOFLAVIN-T DERIVATIVES FOR AMYLOID β

139

5.1. Introduction

139

5.2. Materials and methods

141

5.2.1. Preparation of technetium-99m labelled thioflavin-T derivatives

141

5.2.2. Synthesis of the non-radioactive analogues of the tracer agents labelled with carbon-11 or fluorine-18

141

5.2.3. Synthesis of a cationic thioflavin-T derivative intended for diagnosis of PA 5.2.4. Determination of affinity of

Tc-labelled compounds for amyloid β

99m

5.2.5. Determination of the affinity of authentic non-radioactive compounds for amyloid β

142 144 146

5.2.6. Determination of the affinity of authentic non-radiolabelled derivatives for post mortem human AD brain homogenates

147

5.2.7. Determination of in vitro affinity for plaques in mouse and human AD brain sections

149

5.3. Results and discussion

151

5.3.1. Synthesis of 2-(4’-dimethylaminophenyl)-6-hydroxy-N-methyl-1,3-benzothiazole (5.4)

151

5.3.2. Affinity tests: set-up

152

5.3.3. Affinity of the authentic non-radiolabelled derivatives 3.2, 3.8, 3.16 and 3.18 for post mortem human brain homogenates

158

5.3.4. In vitro affinity of 3.2, 3.8 and 3.16 for plaques in mouse and human AD brain sections

159

5.4. Conclusion

165

6. GENERAL CONCLUSION

167

7. SUMMARY

171

LIST OF ABBREVIATIONS α2 M AA Aβ Aβ2M AβPP AD AH, AL Aλ, Aκ AP APLP Apo ATTR β 2M BACE BAM BAT BBB BCL BOC BTA CERAD CPHPC cpm cps CSF CT CTF Da DDNP DIEA DMF DMSA DMSO EC ECD EDCI.HCl EDDA EDTA EEG ESI FAD FCT FDG GAG HMPAO HOBT

α2-microglobulin amyloid A protein amyloid β amyloid of β2-microglobulin origin amyloid β precursor protein Alzheimer's disease amyloid of immunoglobulin heavy/light chain amyloid of lambda/kappa light chain amyloid P component amyloid precursor-like protein apolipoprotein amyloid of transthyretin origin β2-microglobulin beta-site AβPP cleaving enzyme bioactive molecule Bis-aminoethanethiol blood-brain-barrier bifunctional chelating ligand t-butoxycarbonyl benzothiazole-aniline Consortium to Establish a Registry for Alzheimer's Disease R-1-[6-[R-2-carboxy-pyrrolidin-1-yl]-6-oxohexanoyl]pyrrolidine-2-carboxylic acid counts per minute counts per second cerebrospinal fluid computed tomography carboxy-terminal fragment Dalton 1,1-dicyano-2-[6-(dimethylamino)naphtalen-2-yl]propene N,N-diisopropylethylamine N,N-dimethylformamide dimercaptosuccinic acid dimethyl sulphoxide electron capture L,L-ethylcysteinate dimer 1-(3-dimethylaminopropyl)-3-ethyl carbodiimide hydrochloride ethylenediamine-N,N'-diacetic acid ethylenediaminetetraacetic acid electroencephalogram electrospray ionization familial Alzheimer's disease fetal calf serum 2-[18F]fluoro-2-deoxy-D-glucose glycosaminoglycan hexamethylpropyleneamine oxime hydroxybenzothriazole

HYNIC i.v. IBOX ICD ID ID/g IDA IMPY kBq keV KPI LC-MS mAβ MAG3 MBq MeV Mp MPLC MRI MS NFT NINCDSADRDA NMP NMR NSAID NTF p.i. PA PBS(T) PBT PET PHF PPA PS1, PS2 RP-HPLC Rt SAA SAP SPECT TEA TFA TLC TP TT TTR TZDM UV

hydrazinonicotinic acid intravenous iodobenzoxazole International Classification of Diseases injected dose injected dose per gram iminodiacetic acid iodo-imidazo[1,2-a]pyridine kilobecquerel kilo electron volt Kunitz-type of serine protease inhibitor liquid chromatography coupled to mass spectrometry human anti-Aβ monoclonal antibodies mercaptoacetyltriglycine megabecquerel mega electron volt melting point medium pressure liquid chromatography magnetic resonance imaging mass spectrometry neurofibrillary tangle National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer's Disease and Related Disorders Association N-methyl pyrrolidone nuclear magnetic resonance non-steroidal anti-inflammatory drug amino-terminal fragment post injection peripheral amyloidosis phosphate buffered saline (containing 0.05 % Tween® 20) phenylbenzothiazole positron emission tomography paired helical fragments polyphosphoric acid presenilin 1, presenilin 2 reversed phase high pressure liquid chromatography retention time serum amyloid A protein serum amyloid P component single photon emission computed tomography triethylamine trifluoroacetic acid thin layer chromatography time point thioflavin-T transthyretin thiazoledimethyl ultraviolet

1

CHAPTER 1

1. PERIPHERAL AMYLOIDOSIS AND ALZHEIMER’S DISEASE: INTRODUCTION

1.1. Introduction Alzheimer’s disease (AD) and peripheral amyloidosis (PA) are two severe illnesses of which, although the underlying pathology is the same, the outcome is completely different. In both cases deposits of commonly present proteins (in a β-pleated sheet structure1) cause severe tissue damage that will ultimately lead to the death of the patient. In the case of AD, the damage occurs in the brain, especially in cortical regions, while in PA damage can arise throughout the entire body or can be more localized in a single organ. Specific mechanisms, which are still not entirely understood at this moment, lead to the formation of insoluble amyloid deposits that will disrupt the normal tissue structure and spark off a disease that has no proper cure, even to date…

1.2. Amyloid and amyloidosis Although the term amyloid was already introduced by Virchow in 18531, it has taken a long time to thoroughly understand the importance of amyloid deposits in the symptomatic picture of a diversity of diseases. The ‘waxy, eosinophilic’ tissue deposits that Virchow believed to be of polysaccharide composition and which could be stained with iodine and sulphuric acid2,3 led him to characterize them as amyloid (starch-like; amylum = Latin word for starch; starch is commonly visualized by the addition of iodine). Due to further research with constantly improving techniques, it was proven that these deposits have a proteinaceous nature instead of the polysaccharide

2

Chapter 1

composition Virchow ascribed to them. The inconsistency between the proteinaceous nature and the (wrong) name, referring to a polysaccharide-like composition, brought Glenner (who was the first to purify the amyloid fibril protein4) to the term β-fibrilloses1. The latter refers to the β-pleated sheet structure of amyloid, a structure that is normally not present in mammalian tissue in this almost pure conformational state. The manifestation of amyloid depositions was therefore the first described pathologic process that manifested itself by a specific and unique protein conformation. Today, the terms amyloid and amyloidosis have been commonly accepted and are probably more common than the term β-fibrilloses.

Amyloidosis is a generic term that describes a wide spectrum of diseases characterized by the deposition of insoluble fibrillar proteins or amyloid in different organs5. These fibrils are long twisted filaments, 6-8 nm wide, and they are mainly composed of low molecular weight proteins that are normally soluble under physiologic conditions. Misfolding of these extracellular proteins plays a prominent role in amyloidosis6. In this case, the normal physiological process of protein folding is replaced by a dynamic process which generates insoluble, toxic protein aggregates that are deposited in tissue. Until now, 21 different proteins have been recognized as causative agents of amyloid disease7. Although these proteins all have a different structure and different functions within the human body, they share a common feature, namely the possibility to form morphologically indistinguishable amyloid fibrils. This common feature, i.e. the possibility to acquire more than one conformation, has led to the term chameleon proteins8. Hallmarks of these amyloid fibrils are: a predominant β-pleated sheet conformation, apple green birefringence under polarized light after Congo red staining, a very poor solubility, fibrillar appearance when viewed under electron microscopy and cross beta X-ray diffraction appearance9. The structure of the mature fibrils can be compared to steel cables with 3-6 filaments wrapped around one another10. It is also this β-pleated sheet structure that is responsible for these proteins’ relative resistance to dissolution in physiologic solvents and to normal proteolytic digestion, besides the tinctorial and optical properties

Peripheral amyloidosis and Alzheimer’s disease: introduction

3

revealed by Congo red staining. The low solubility of these proteins will lead to deposition and accumulation in tissue followed by pressure atrophy and subsequent death, resulting from interference with normal physiologic processes of affected vital organs. At first, classification of amyloidosis was based upon the localization or appearance of the amyloid proteins11. Primary amyloid was amyloid that appeared de novo without preceding disorder (e.g. plasma cell disorder), secondary amyloid was a complication of a previously existing disorder (e.g. rheumatoid arthritis), familial amyloid (e.g. familial Mediterranean fever) and isolated amyloid in which a single organ system is involved (e.g. AD). Due to the lack of techniques, scientists were not able to obtain information on the amino acid sequence of the proteins that formed the amyloid fibrils or their precursors. Due to the progress in the field of protein purification and analysis, the knowledge became available to divide cases of amyloidosis according to the protein that formed the base of the amyloid deposits. As is already obvious from the above description, nomenclature and classification of a case of amyloidosis is now possible by referring to the fibril protein that makes up the amyloid deposits12. Furthermore, due to the presence of an amyloid precursor protein which differs in different types of amyloidosis, this precursor protein should also be noted. This precursor protein is the protein of which the actual amyloid fibril protein is thought to be derived. With the exception of the β protein precursor of AD, all these precursor proteins can be demonstrated in serum and the amyloid proteins are mostly fragments (incomplete degradation products) of their precursor. In case of different subgroups (for example for immunoglobulins), mutants and the prohormone procalcitonin, this protein variant is also noted. The clinical manifestation of a specific case of amyloidosis is the last issue that can be mentioned in the description and classification of amyloidosis.

4

Chapter 1

Table 1. Classification of different types of amyloid and amyloidosis. Adapted from: Cohen, A. S., Bull Rheum Dis, 1991; 40. (AA = amyloid A protein, SAA = serum amyloid A protein, L/H = immunoglobulin light/heavy chain) Amyloid protein

Protein precursor

Protein type/variant

Clinical diagnosis

ApoSAA

Reactive (secondary) amyloidosis Familial Mediterranean fever Muckle-Wells’ syndrome

AL

κ, λ e.g. κIII

Aκ, Aλ e.g. AκIII

Idiopathic (primary) amyloidosis, associated with myeloma or macroglobulinaemia

AH

IgG 1 (γ1)

Aλ1

ATTR

Transthyretin

e.g. Met 30

AA

e.g. Met III TTR or Ile 122 AApoAI

apoAI

Arg 26

AGel

Gelsolin

Asn 187 (15)

ACys

Cystatin C

Gln 68

Aβ

β protein precursor

Gln 618 (22)

Aβ2M

β2-microglobulin

AScr

Scrapie protein

(Pro)calcitonin

AANF

Atrial natriuretic factor Islet amyloid polypeptide

AIAPP

Familial amyloid polyneuropathy, Iowa Familial amyloidosis, Finnish Hereditary cerebral haemorrhage with amyloidosis, Icelandic Alzheimer’s disease Down syndrome Hereditary cerebral haemorrhage with amyloidosis, Dutch Associated with chronic dialysis

Scrapie protein 27-30 e.g. Leu 102

ACal

Familial amyloid polyneuropathy, Portuguese Familial amyloid cardiomyopathy, Danish Systemic senile amyloidosis

(Pro)calcitonin

Creutzfeldt-Jakob disease, etc.

Gerstmann-Straüssler-Scheinker syndrome In medullary carcinomas of the thyroid Isolated atrial amyloid In islets of Langerhans Diabetes type II, insulinoma

Peripheral amyloidosis and Alzheimer’s disease: introduction

5

1.3. Peripheral amyloidosis

1.3.1. Amyloid deposits and pathogenesis of amyloidosis

As can be seen in Table 1, the classification does not include several other proteins that are often present in amyloid deposits. Since these proteins are not fibrillar and are not characterized by a β-sheet formation12, they are not listed in the classification. However, they do present an important part of the amyloid deposits. The most prominent of these so-called amyloid associated proteins or chaperones5 is amyloid P component (AP)9 and its identical serum counterpart (SAP). AP is composed of a pair of pentagonally shaped subunits and probably serves as a scaffold for amyloid fibril formation. AP is distinct from amyloid, it does not bind Congo red and has no fibrillar structure or cross β

pattern.

It

is

nonetheless

identifiable

in

all

forms

of

amyloid.

Glycosaminoglycans (GAGs) are also often associated with amyloid. These GAGs are mostly highly sulphated and it is postulated that these negatively charged molecules affect the protein product or precursor processing. Other chaperones that have been described are α1-antichymotrypsin, apolipoprotein E (apoE), apoJ, complement components, vitronectin and extracellular matrix proteins5. The question is of course if these proteins are just innocent bystanders or whether their presence is related to the mechanism of amyloidogenesis. For AP and apoE, scientists are in favour of the latter hypothesis because they are present in several types of fibrillar deposits but absent in non-fibrillar lesions. In vitro studies with amyloid β (Aβ) (1-40 and 142, see further on) have shown that apoE (especially apoE4), apoJ and α1antichymotrypsin have the ability to modulate the formation of amyloid-like fibrils.

Although the pathogenic process is not yet entirely clear, different observations have been described6. The formation of the predominantly antiparallel β-sheet secondary structure is a pathologic process that closely relates to physiologic protein folding. In the cytoplasm newly synthesized

6

Chapter 1

polypeptides undergo different kinds of post-translational modifications which will direct them towards their function. According to the ‘folding energy landscape theory’13 this is a funnel-like pathway in which the conformational intermediates progressively merge into a final species. In the case of amyloid protein, different possible pathways can be followed. At a minimum of energy, comparable to that reached by the native protein, these proteins can acquire alternative and relatively stable misfolded states, which are highly prone to aggregation. This misfolding can occur in different ways6: - the amyloid protein can have an intrinsic propensity to assume a pathologic conformation which becomes evident with aging (e.g. normal transthyretin in patients with senile systemic amyloidosis) - the serum protein concentrations can reach very high levels (e.g. β2microglobulin in long-time hemodialysis patients) - mutations can lead to the replacement of a single amino acid in the protein which may lead to a higher propensity to aggregate as in hereditary amyloidosis - proteolytic remodelling of the precursor protein as in AD (see further on). Besides the possible intrinsic amyloidogenic potential of the pathogenic protein, other factors may act synergistically in amyloid deposition. For instance, the protein precursor must reach a critical local concentration to trigger the fibril formation, and this can be enhanced by environmental factors and interaction with extracellular matrices.

A

small

proportion

of

immunoglobulin

light

chains

can

be

6

amyloidogenic . AL amyloidosis only occurs in 12 to 15 % of patients with myeloma. The λ isotype and the VλVI variability subgroup are related to amyloidogenicity because of structural features. During the immune response the variable domains of the light chains V(L) often mutate and some of these physiologic mutations can lead to destabilization of the domain and can favour an aggregation-prone state14,15. Transthyretin and lysozyme are prototypical proteins of hereditary amyloidosis. Approximately 80 different mutations have been reported in transthyretin and only a few are not associated with amyloidosis, while a couple is thought to have a protective effect on amyloid

Peripheral amyloidosis and Alzheimer’s disease: introduction

7

deposition. For lysozyme, four pathogenic variants have been reported. Furthermore, studies of cystatin C, immunoglobulin light chains and gelsolin have confirmed the property shared by these amyloidogenic variants, namely a native conformation that is thermodynamically less stable than that of their normal counterpart. Although destabilization is necessary, it is probably not sufficient to confer an amyloidogenic propensity onto a protein. Other structural features are required for the formation of fibrils and the role of charged residues in modulating the aggregation process by means of repulsive forces has recently been highlighted16. Furthermore, a gelsolin mutant which is partially unfolded, renders the protein susceptible to the attack of proteases and provokes the release of highly amyloidogenic polypeptides.

Al these findings provide support for the mechanism in which amyloidogenic and normal counterparts are synthesized and secreted as native proteins, but the system of intracellular quality control appears to be incapable of recognizing and thus removing dangerous mutants. Once secreted, an equilibrium is reached between fully folded and partially folded forms, but the fluctuation in the concentrations of the two forms is much greater than would be expected. A lot of factors influence the threedimensional structure of a protein such as low pH, oxidation, increased temperature, limited proteolysis, metal ions and osmolytes. All these factors can shift the equilibrium towards the partially folded amyloidogenic state. The above-mentioned GAGs and SAP may have an identical effect by hastening the integration of a soluble polypeptide into a more stable fibril. Moreover, proteins that need a conformational plasticity for their function (e.g. apoE which partially unfolds as lipids are released, but refolds when lipids are taken up) seem to favour the formation of amyloid at the same time.

8

Chapter 1

1.3.2. Diagnosis of peripheral amyloidosis

1.3.2.1. Clinical diagnosis of amyloidosis

Due to the extensive range of proteins that can be responsible for the symptoms accompanying peripheral amyloidosis, it is very difficult to diagnose this illness. Furthermore, the affected organs can be very diverse (going from liver, heart, kidneys to joints and even skin) and the underlying clinical features vary widely and are non-specific. They depend entirely on which organ has amyloid deposits and the extent of the deposit9. This makes the clinical diagnosis a challenge. The symptoms are very unspecific and in a lot of cases the search for amyloid deposits is overlooked because of symptoms that are in the first place suggestive of other diseases. For example, patients with nephritic syndrome of unexplained origin, hepatosplenomegaly in combination with certain chronic inflammatory diseases including rheumatic disease (which is a rather frequently seen symptom), should be suspected of having amyloid deposits.

Reactive (AA) (secondary) amyloid is classically associated with chronic inflammatory and infectious disease. Renal involvement is common in all forms of systemic amyloid, hepatic enlargement is also a common feature in case of hepatic amyloid, the gastrointestinal tract commonly shows effects from amyloid deposits, the amyloid infected heart will primarily show congestive heart failure and is common in the AL form; the respiratory tract, the endocrine glands, the skin, the musculoskeletal system, the nervous system and the haematological system can all be laden with an amyloid burden and this again will lead to symptoms that the clinician will not directly consider to be amyloidosis. For peripheral amyloidosis of the AA, Aλ, Aκ… type, the prevalence is only 0.8 %, as was seen in an unselected biopsied Malaysian patient population17. On the other hand, the incidence of β2M amyloidosis18 increases progressively with years on dialysis reaching 80% in patients dialysed over 15 years19 because the dialysis procedure cannot remove the β2-microglobulin from the blood. Also in AD patients and patients

Peripheral amyloidosis and Alzheimer’s disease: introduction

9

with type II diabetes (amyloid deposits in the islets of Langerhans) the localized deposits are widespread and they become universal with age increasing over 60 years20. The latter cases (AD, β2M, type II diabetes) are more easily detected than the AA, Aλ, Aκ… types because of their incidence. This myriad of possible symptoms has nevertheless lead to a few noninvasive techniques to detect the amyloid deposits21. The Congo red uptake method of Bennhold22 and the use of antiserums to the amyloid AA fibrils are two examples. The latter method was meant to detect elevated levels of serum amyloid A protein (SAA), but it has been found that SAA has a more acute-phase nature and therefore its diagnostic usefulness is limited, just as is the diagnostic value of the Bennhold method. A lot of patients with acquired systemic amyloidosis have unusual Bence-Jones proteins in their urine, so electrophoresis and immunoelectrophoresis would be useful, but it is not clear if their presence is invariably related to amyloidosis.

1.3.2.2. Histological diagnosis of amyloidosis

Due to the specific properties of amyloid (apple green birefringence when stained with Congo red and examined under polarized light) the most evident way of diagnosing amyloidosis would be by taking biopsy samples and staining them with Congo red to visualize the presence of amyloid21. However, biopsies are invasive and therefore less convenient for the patient. This together with the fact that biopsies should be taken from an organ that is affected by amyloid, which is not always clear from the clinical symptoms the patient describes, also renders histological diagnosis a rather difficult issue. Furthermore, a biopsy mostly is a very small tissue sample and there is always a rather large chance that a sample has been taken from a place in a specific organ that has no amyloid deposits. This leads to false negative results and because of the severe outcome of the disease, false negative results should be avoided as much as possible. When a biopsy specimen cannot be obtained from the organ that is suspected of amyloid infiltration, rectal biopsy might be a good alternative. In generalized amyloidosis these rectal specimens are positive in 75 to 85 % of the cases.

10

Chapter 1 The first step to undertake on a biopsy is staining of the amyloid protein

with Congo red. Once the presence of amyloid is determined, the next step is the use of immunostaining to ascertain the nature of the protein that leads to the amyloid deposits. The latter is necessary because unequivocal identification of the deposited amyloidogenic protein is essential to avoid misdiagnosis and inappropriate treatment6, to offer the patient a favourable prognosis and to offer genetic counselling in case of hereditary amyloidosis. Ever since the method of Congo red staining was first described23, several modifications have been proposed to improve the specificity, sensitivity and reliability of the staining method. The alkaline-alcoholic method of Puchtler et al.24 is usually used today. Other dyes that have been used for staining are fluorescent products like thioflavin-T (TT) or thioflavin-S and metachromatic stains such as crystal violet, methyl violet and toluidine blue. These dyes are used on a large scale but they are less reliable than Congo red for routine purposes. Recently, the stilbene derivative X-34 has been described by Klunk and co-workers25. This is a Congo red derivative in which the diazo groups are replaced by carbon double bonds and which also has good properties for staining amyloid. Figure 1 shows the structure of Congo red, thioflavin-T and X-34. Another recent technique is the use of Congo red fluorescence26, which seems to be the most sensitive method27 for the direct detection of amyloid because it enhances the recognition of tissue-bound Congo red due to its property as a fluorochrome. This technique can be used on Congo red stained sections visualized under ultraviolet (UV) light.

The use of commercially available antibodies against most known amyloid fibrils (AA, AL (λ or κ), ATTR, Aβ2M, Aβ) makes it possible to determine the protein that forms the amyloid fibrils. This allows classifying the amyloidosis but this immunohistochemical approach cannot stand alone because it provides no proof of the presence of amyloid fibrils which in the end will be the cause of structural and functional tissue damage28. Therefore, tissue staining with Congo red and visualization of the subsequent applegreen birefringence under polarized light is a necessity to conclude that amyloid fibrils are present. It is also possible to detect amyloid fibrils with

Peripheral amyloidosis and Alzheimer’s disease: introduction

11

electron microscopy, which is claimed to be more sensitive than light microscopy, but it is not always evident to identify the amyloid fibrils ultrastructurally and the same is true for immunohistochemical staining. Therefore, electron microscopy cannot stand on its own to confirm the diagnosis of amyloidosis22.

H 2N

NH2

N N

N N HO3S

SO3H

Congo red

HO

CH CH

CH

CH

OH

HOOC

COOH X-34 CH3 N N H 3C

S

CH3 CH3

Thioflavin-T

Figure 1. Structure of Congo red, X-34 and thioflavin-T

For these reasons, neither a clinical diagnosis nor a histological diagnosis of amyloidosis can stand on their own because of the shortages of both methods. Each method on its own generates data which might lead to a correct diagnosis, but only a combination of both methods will provide the clinician with sufficient data to guarantee with relative certainty that he is making the correct diagnosis. Additional tests which are less invasive and highly specific are therefore most welcome to aid the clinician in diagnosing amyloidosis.

12

Chapter 1

1.3.2.3. 123I-SAP scintigraphy

As mentioned in 1.3.1, the amyloid P component is identifiable in all forms of amyloidosis. Immunochemistry and electrophoresis29 have shown that AP is identical to a normal serum globular 9.5 S α1-glycoprotein as described by Haupt and co-workers1,30,31 and that its binding to amyloid proteins is apparently calcium dependent. SAP (the precursor of AP) is a normally occurring serum constituent and is therefore unrelated to the presence or absence of amyloidosis of any type. One of the useful properties of SAP is that it specifically binds to amyloid fibrils in vivo and in vitro32. Because of this property, SAP, when injected, will bind to amyloid deposits if they are present in the patient’s body and this means that radiolabelled SAP might be a useful tracer for visualization of amyloid deposits.

Studies have been undertaken with 99 % pure SAP which was isolated from serum (heated at 56 °C for 30 minutes) and from an accredited donor in the U.K. This isolated SAP was labelled with iodine-123, a frequently used single photon emission computed tomography (SPECT) radionuclide. It is prepared by bombarding xenon-124 with protons in a cyclotron and yields 123

Cs following a (p,2n) reaction (Figure 2).

124

1

1

1

= 6 min, EC = 2 h, EC = 13 h, EC Xe (p,2n)123 Cs ⎯T⎯ ⎯ ⎯⎯→123 Xe ⎯T⎯ ⎯ ⎯ ⎯→123 I ⎯T⎯ ⎯ ⎯ ⎯→123Te 2

2

2

Figure 2. Production and radioactive decay of 123I (EC = electron capture)

123

I decays by electron capture with a half-life of 13 h and emits γ-rays

of 159 keV which is ideal for external detection with a gamma camera. The physical characteristics of

123

I are nearly ideal for in vivo visualization, but the

high cost of its production make it a less favourable radionuclide in nuclear medicine. Nevertheless, despite the high cost, body imaging.

123

I is usable for in vivo whole

Peripheral amyloidosis and Alzheimer’s disease: introduction 123

13

I-SAP was evaluated in healthy subjects and patients with

histological proven amyloid deposits33,34,35. After intravenous (i.v.) injection, the

123

I-SAP activity is initially confined to the blood pool but the rapid

decomposition of

123

I-SAP leads to excretion of radioactive fragments in the

urine. Within 48 hours, 45 % of the injected dose was cleared from the body in healthy and control subjects. All patients with histologically proved amyloidosis showed rapid uptake of activity into one or more sites (viscera, bone marrow, carpal region, skin, tongue), producing characteristic images. In these patients with systemic amyloidosis the initial 6-hour plasma clearance and whole-body retention of

123

I-SAP was increased. Patients with massive

amyloidosis of liver and/or spleen concentrated up to 95 percent of the injected dose in those areas 30 minutes post injection (p.i.). This indicates that 123

I-SAP that is bound to the amyloid deposits is degraded more slowly than

the residual

123

I-SAP in the circulation. In some patients who died of their

disease, on different time points after imaging, the scintigraphic images were correlated to the histological findings and a precise correlation was found between the histological distribution of amyloid at autopsy and the previous in vivo images with

123

I-SAP. An interesting finding was the non-homogenous

uptake of the tracer (particularly in affected livers and spleens) despite the homogeneity on CT scanning. This may reflect a patchy deposition of the amyloid fibrils and is supportive for the above-mentioned issue of unreliability of needle biopsy sampling for histological monitoring of amyloid. Serial studies also showed that

123

I-SAP can be used for the follow-up of patients with

amyloidosis.

Although these results are seemingly positive and suggest the usefulness of

123

I-SAP as a possible diagnostic tool, there are also a lot of

drawbacks to this method and this is probably one of the reasons why it has not entered clinical routine on a global scale. A very important drawback is the availability of clinical grade pure SAP combined with the fact that the use of human proteins for injection includes certain risks. The high cost and suboptimal availability of

123

I is another drawback of this method. The latter

problem was partially solved by the preparation of technetium-99m labelled SAP36. For use in nuclear medicine,

99m

Tc is much more interesting than

123

I

14

Chapter 1

(see further on). Apart from its nearly ideal physical characteristics (half-life of 6.02 h, γ-ray of 140 keV, almost pure γ-ray emission), it is also continuously available at a low cost from a dose to the patient37.

99m

99m

Tc- generator and delivers a low radiation

Tc-SAP showed similar features as

123

I-SAP, but the

problem of SAP availability and use remains, although (rapid) SAP isolation methods have been developed38,39. All these facts have led to the use of

123

I-

SAP in only a few specialized centres over the world. The lack of the commercial availability of 123I-SAP and the low amounts of SAP available from isolation out of human serum, has stimulated the development of other tracer agents for visualization of amyloid fibrils in systemic

amyloidosis.

diphosphonate

42

and

99m

67

Ga-citrate,

99m

Tc-pyrophosphate40,41

Tc(V)-dimercaptosuccinic acid (

99m

or

-

Tc(V)-DMSA) are

examples of tracers that have been used but proved to be rather nonspecific43. Aprile and co-workers44 investigated the possible usefulness of 99m

Tc-aprotinin for visualization of amyloid deposits based on the finding that

some proteinase inhibitors (e.g. α1-antichymotrypsin) can be present in the deposits. Aprotinin is a non-specific serine protease inhibitor which is extracted from bovine lung.

99m

Tc-aprotinin is known for its accumulation in

kidneys, so the entire splanchnic area will show a high background radiation. Therefore, this tracer can only be used in the extra-abdominal area. Clinical studies43,44, however, showed the possible usefulness of

99m

Tc-aprotinin for

visualization of amyloid deposits in heart and other extra-abdominal regions. Its use is very limited due to the renal excretion which makes it impossible to visualize amyloid deposits in kidneys and liver, which are two organs that often are affected in this disease.

99m

Tc-aprotinin and

123

I-SAP, both proteins,

show very limited passage over the blood-brain-barrier (BBB), making these tracers useless for visualization of amyloid in brain.

1.3.3. Treatment of amyloidosis: novel strategies, potential therapies

Until recently it was believed that amyloid deposition is an irreversible process, but the turnover of the deposits is relatively slow and the rate of fibril

Peripheral amyloidosis and Alzheimer’s disease: introduction

15

formation usually exceeds that of mobilization10. The result is a relentless accumulation of the fibrils and an unavoidable disruption of tissue structure and function. A logical start for treatment is an attempt to reduce the supply of the fibrils, namely of their precursor proteins, and to support or even replace the function of the compromised organ. In reactive systemic AA amyloidosis the use of anti-inflammatory drugs and immunosuppressants reduces the production of the acute phase reactant SAA, while in AL monoclonal immunoglobulin light chain amyloidosis the use of chemotherapy targeted towards the underlying clonal B-cell dyscrasia might be useful and successful45. Orthotopic liver transplantation normally leads to a stop in mutant ATTR production in these hereditary forms. Newer concepts in therapy are45: - stabilization of the precursor protein; structural analogues of diclofenac can be used to prevent TTR fibril formation in case of the already more than 80 mutants of TTR that have been described. - inhibition of fibrillogenesis; novel sugar analogues of N-acetylglucosamine inhibit the binding between heparan sulphate or heparan sulphate proteoglycan perlecan with amyloid precursor proteins and thus interfere with the interaction of the amyloid fibrils with GAGs. The GAG mimetic agent Fibrillex™ is currently being tested in phase II/III clinical trials in patients with AA amyloidosis. - inhibition of aggregation; compounds are sought in libraries using screening assays that can measure amyloid fibril formation through their binding of dyes such as TT. Most research up-to-date has been focused on AD and none of these agents has yet been studied in man. - immunotherapy; the human body does not react with an immune response towards

the

amyloid

deposits,

but

antibodies

aimed

at

common

conformational epitopes expressed on amyloid fibrils might lead to clearance of the deposits. -

SAP

targeting;

hexanoyl]pyrrolidine-2-carboxylic

R-1-[6-[R-2-carboxy-pyrrolidin-1-yl]-6-oxoacid

(CPHPC),

a

compound

recently

discovered through screening assays has the ability to cross-link opposing pairs of SAP molecules face to face, thereby occluding the binding face of the

16

Chapter 1

SAP molecule. Clinical studies are ongoing and data achieved to date are promising.

1.4. Alzheimer’s disease As already mentioned before, AD is a form of amyloidosis but in this case the amyloid deposits are localized in one single organ, namely the brain and more precisely in the brain cortex. The research area of this disease has become of major importance due to its severity and its worldwide increasing prevalence as a consequence of the increasing lifespan. Alzheimer’s disease is a progressive, neurodegenerative disorder with specific characteristics. Alois Alzheimer was the first (in 1907) to describe a form of dementia in a 55-year old female of which post mortem staining of the brain - with at that time newly available silver stains - revealed the presence of tangled fibres and clusters of degenerating nerve endings46. Furthermore, Alzheimer described progressive memory impairment, disordered cognitive functions, altered behaviour (paranoia, delusions) and decline in language function, all features that are still identifiable in Alzheimer patients today47. This collection of symptoms came to bear his name and is nowadays still a topic of extensive research.

AD is the most frequent form of dementia and after cardiovascular diseases and cancer the third on the list of most frequent causes of death48. The incidence of AD increases exponentially with age and there is no levelling off in very old age (at least up to the age of 90)49. If mild cases are included, the prevalence is as high as 10.3 % in a western population over 65 years of age and increases to almost 47 % for people over 85 years of age. In the 85+ group, the disease is most prominent in females (ratio 2.8/1 over 75 years of age)50. By 2050 the number of people affected with AD will be doubled and this will lead to a boom in the cost of medical care51. It should be noted that taking care of an ill person and taking care of a demented person are two completely different issues. Although people with dementia can have an almost normal physical state, they are robbed of their specific human, cognitive functions: memory, reasoning, abstraction and speech. Or, as John

Peripheral amyloidosis and Alzheimer’s disease: introduction

17

Hughlings Jackson stated: ‘…there is no person, but only a living creature’48. The devastating consequences of AD, not only for the patient, but also for his family and relatives, give rise to frustration and a lot of tension because of the twilight zone in which the AD patient resides. The urge to obtain as much relevant information as possible on the disease and its causes and to find a cure for this dehumanizing illness is until now of major importance.

The aetiology of AD is complex but the disease can be divided into an early-onset form (onset 60 year)52. Up-to-date, three genes have been linked to the earlyonset form of AD. These genes are: the β-amyloid precursor protein (AβPP) on chromosome 21, presenilin 1 (PS1) on chromosome 14 and presenilin 2 (PS2) on chromosome 153. Missense mutations in AβPP were confirmed in only about two dozen families worldwide, so the major part of the early-onset AD is thought to be related to mutations in PS1 and PS2. The inheritance is autosomal dominant and estimates of the prevalence in inherited forms of AD ranged from 5-10 % to even 50 % and more47. Despite the uncertainty about the percentage of inherited AD (familial AD, FAD), it has become clear that the distinction between the histological phenotype of the early-onset forms and the late-onset forms is very difficult. The same is true for the clinical manifestations in FAD and the sporadic cases of AD (although certain families show very distinct clinical signs like myoclonus, seizures…). This phenotypic similarity strongly suggests that information about the mechanism of the FAD forms caused by mutations in the AβPP and presenilin genes, is likely to be directly relevant to the pathogenesis of the common, apparently non-familial forms. For late-onset AD, an association with the inheritance of the ApoE allele ε4 has been described54. The presence of this ε4 allele seems to be predominantly a risk factor for patients with AD onset between 60 and 70 years. Emphasis should be put on the fact that apoE4 is a risk factor for and not an invariant cause of AD. Reports of humans homozygous for the ε4 allele showing no Alzheimer symptoms (even in their ninth decade of life and beyond) and humans with AD that do not harbour ε4 alleles prove this statement. Research is still ongoing, but an AD linked locus on chromosome

18

Chapter 1

12 representing alterations in or near the gene encoding α2-microglobulin (α2M) has been reported55; also, a major locus for late-onset FAD on chromosome 10q was recently discovered56. Over the next decades a much larger portion of AD will probably be shown to have genetic determinants.

The histological identifiable hallmarks of AD are: 1) intraneuronal, cytoplasmatic deposits of neurofibrillary tangles (NFT), 2) extracellular amyloid deposits called neuritic plaques, 3) cerebrovascular amyloidosis and 4) synaptic loss52. The clinical hallmarks are by far the more difficult to observe and include progressive memory loss and decrease of cognitive functions. These clinical symptoms are mostly the first to be detected, but they require the use of neuropsychological tests and depend on the experience of the practicing clinician (see further on). All in all it can be stated that AD is a very complex disease with a lot of different characteristics which make it often hard to diagnose because the clinical diagnosis is only a probable diagnosis and the histological diagnosis can only be obtained after death. All these features support the necessity for a tracer agent that is non-invasive and that can be used to visualize the plaques during life.

1.4.1. The pathogenesis of Alzheimer’s disease

A lot of scientists share the opinion that the Aβ protein has no particular physiological function and that, despite its presence in the plasma and cerebrospinal fluid (CSF) of healthy persons57, its mere presence just leads to amyloid deposits over the years due to accumulation of the protein and conversion to its fibrillar form. Aβ was thought to be a toxic waste product58. Since Kamenetz and co-workers59 published their findings about the possible neurological effect of Aβ, this will surely lead to new understandings and hopefully even better drug design and therapeutics. In their research work these authors showed that neuronal activity modulated the formation and secretion of Aβ peptides in hippocampal slice neurons that overexpress AβPP. Furthermore, Aβ selectively depressed excitatory synaptic transmission onto neurons that overexpress AβPP, as well as nearby neurons that do not

Peripheral amyloidosis and Alzheimer’s disease: introduction

19

and this depression depends on N-methyl-D-aspartate receptor activity and can be reversed by blockade of neuronal activity. These findings indicate that excessive Aβ could contribute to cognitive decline during early AD. The observation that Aβ is also secreted from healthy neurons in response to activity and that Aβ, in turn, downregulates excitatory synaptic transmission, is of major importance, not only for the understanding of AD pathogenesis, but also for rational drug design. These findings shine a new light on the physiological function of Aβ and on the possible cause of AD.

As stated before, most amyloid proteins are derived from precursor proteins and this is also the case for Aβ. The AβPP undergoes sequential proteolytic cleavage and compromises a heterogeneous group of ubiquitously expressed polypeptides. The heterogeneity arises from alternative splicing (which yields 3 major isoforms of 695, 751 and 770 residues) and a variety of post-translational modifications (addition of N- and O-linked sugars, sulphation and phosphorylation)60,61,62,63. The 695-residue isoform is expressed to a high level in neuronal cells, while it has a low abundance in non-neuronal cells; the opposite is true for the AβPP splice forms that contain 751 and 770 amino acids. This difference in length is due to the presence of an exon that codes for a 56-amino acid motif that is homologous to the Kunitz-type of serine protease inhibitors (KPI). This indicates one potential function for the longer forms when present in human platelets, namely inhibitors of factor XIa, a serine protease in the coagulation cascade. AβPP is a member of a larger gene family, called the amyloid precursor-like proteins (APLPs)64,65, which have substantial homology in the large ectodomain and particularly in the cytoplasmic tail, but they are largely divergent in the Aβ region.

AβPP is a single transmembrane protein that is cotranslationally translocated into the endoplasmatic reticulum of the cell via its signal peptide and then post-translationally modified through the secretory pathway. Its halflife is relatively short (~45-60 minutes)63. During the trafficking through the secretory pathway, AβPP can undergo a variety of proteolytic cleavages to release secreted derivatives into vesicle lumens and the extracellular space.

20

Chapter 1

Figure 3 shows the 770 amino acid form with its principal metabolic derivatives.

1 18

289

TM

671

770

COOH

NH2 KPI

Aβ α−secretase

18

γ−secretase 711 or 713

687

C83

APPs-α β−secretase 18

p3

671

APPs-β

γ−secretase 711 or 713

C99 Aβ

Figure 3. Schematic diagram of AβPP and its principal metabolic derivatives. Regions of interest (at their correct relative position): the 17-residue signal peptide at the NH2 terminus (box with vertical lines), two alternatively spliced exons of 56 and 19 amino acids inserted at residue 289 (the first is the abovementioned KPI domain), the single transmembrane domain (TM) at amino acids 700-723 (vertical dotted black lines), the white dot in the upper line indicates the cutting place of the α-secretase, and the different proteolytic fragments as described in the text. Figure adapted from: Selkoe, D. J. Physiol Rev, 2001; 81.

The amyloid β protein fragment includes 28 residues just outside the membrane plus the first 12-14 residues of the transmembrane domain. The first proteolytic cleavage identified is that made by α-secretase and this occurs at 12 amino acids from the amino-terminal of the single transmembrane domain of AβPP66,67. This results in the release of a large soluble ectodomain fragment called α-APPs. This α-APPs is released into the lumen/extracellular space and a 83-amino acid COOH-terminal fragment (CTF) is retained in the membrane. AβPP molecules that are not cleaved by

Peripheral amyloidosis and Alzheimer’s disease: introduction

21

α-secretase can be cleaved by the so-called β-secretase. This enzyme cuts 16 residues NH2-terminal to the α-cleavage site. The ectodomain derivative formed as a consequence of β-secretase cleavage is β-APPs and this ectodomain is slightly smaller than α-APPs68. The CTF residue that is retained has 99 amino acids and begins at amino acid 1 of the Aβ protein. Due to the fact that α-secretase cuts the AβPP in the middle of the Aβ sequence, this enzyme will not lead to the production of Aβ. As mentioned in the first part of 1.4.1, it was assumed that Aβ generation is a pathological event because the cleavage of the C99 fragment by γ-secretase occurs in the middle of the transmembrane domain. It was thought that membrane injury is necessary to allow a soluble protein to reach the cleavage site. Sensitive Aβ antibodies revealed secreted Aβ released from cells under complete normal cellular circumstances69,70,71,72. The p3 fragment (formed by consecutive cleaving of AβPP by α-secretase and γ-secretase, see Figure 3) was discovered at the same time73,70. These findings both indicated that the production of Aβ is indeed a normal metabolic event and the peptide can be detected in CSF and plasma of healthy subjects throughout life71,72. This is also what Kamenetz and co-workers saw during their research59. To date, it is not completely clear where α- and β-secretase exert their acitivity, but a substantial portion of αAPPs is generated by interaction of α-secretase on membrane inserted AβPP. This can also occur during intracellular trafficking of AβPP; β-secretase can occur partially late in the secretory trafficking of AβPP. After formation of the C99 or C83 fragments by β- and α-secretase respectively, these intracellular fragments can be cleaved by the so-called γsecretase. This enzyme cleaves the AβPP as shown in Figure 3 at residue 711 or 713. The consecutive splicing of AβPP by β- and γ-secretase can therefore lead to Aβ fragments with 40 or 42 amino acids. This probably happens to a large extent during the internalization and endosomal processing of AβPP74,75. The major part of the proteolytically formed Aβ1-40 and Aβ1-42 is probably destined for secretion, a hypothesis that is supported by steady-state levels of Aβ in human CSF76 and plasma77.

22

Chapter 1 All the above-mentioned proteolytic systems finally lead to the Aβ

protein which will become a substantial part of the amyloid deposits. The question about the nature of these secretases has remained unanswered until the last few years. Due to highly sophisticated techniques and progress, researchers have been able to elucidate the function and the structure of these enzymes. It appeared that β-secretase is a novel transmembrane aspartic protease called beta-site AβPP cleaving enzyme (or BACE)78. Five groups employing three different approaches realised the identification of the same enzyme which makes a strong case for BACE being β-secretase. The active site of aspartic proteases of the pepsin family is contained in the BACE enzyme. BACE is an unusual aspartic protease because it has a C-terminal transmembrane domain and is membrane bound. The active site motifs are located in the luminal domain, so the active site of BACE and the β-secretase cleavage

site

of

AβPP

are

in

correct

topological

orientation

for

endoproteolysis. BACE is expressed in all tissues but higher concentrations are found in neurons; it is localized within acidic intracellular compartments, βsecretase cleavage products are increased by BACE overexpression, antisense inhibition of BACE decreases β-secretase cleavage, purified forms of BACE cleave AβPP substrates in vitro and BACE has an acidic pH optimum and is not inhibited by pepstatin (a common aspartic protease inhibitor). All these issues match one-to-one with those of β-secretase78. A homologous gene, BACE-278, was identified soon after BACE and they share 64 % amino acid similarity and have a C-terminal transmembrane domain. Due to the fact that they have only about 40 % similarity to other aspartic proteases in the pepsin family, they probably constitute a novel family. Because the BACE-2 gene is located on chromosome 21 within the critical Down’s syndrome region, it was thought to be involved in the AD pathology of Down’s syndrome, but the fact that BACE-2 cuts more efficiently within the Aβ domain and can therefore limit the production of pathogenic Aβ species, makes this hypothesis questionable. Furthermore, recent research indicated that homozygous knock-out of the BACE gene (also referred to as BACE-1) did not allow any Aβ generation, which is consistent with the findings

Peripheral amyloidosis and Alzheimer’s disease: introduction

23

mentioned above that BACE-2 does not contribute to the amyloidogenic processing of AβPP79. While the structure determination of the β-secretase went relatively straightforward, the structure of the γ-secretase proved to be more difficult to unravel. Recently, De Strooper and co-workers published their findings on the γ-secretase activity80. Their research indicates that γ-secretase is a multiprotein complex consisting of presenilin (PS), nicastrin (Nct), Aph-1 and Pen-2 and all four proteins are necessary for the full proteolytic activity of γsecretase.

In

general, γ-secretase cleaves the hydrophobic integral

membrane domain of its substrates, which leads to the release of protein fragments at the extracellular and at the cytoplasmatic side of the membrane. The presenilins probably provide the active core of the proteins (which was already presumed earlier because deficiency of PS 1 leads to an inhibition of normal AβPP cleavage81) and as mentioned before, there are two mammalian homologs, namely PS1 and PS2. These proteins span the cellular membrane several times and two aspartate residues (Asp257 and Asp385) located in transmembrane domains 6 and 7 respectively, are essential for the catalytic activity of the protease. Stabilisation of the presenilins is accompanied by a proteolytic cleavage performed by an unknown presenilinase. This leads to a so-called amino-terminal fragment (NTF) and a CTF which each deliver one aspartyl residue to the catalytic site of the enzyme.

With the aid of antibodies against the presenilin fragments a second member of the complex was purified, nicastrin82. This protein binds relatively well to both NTF as well as CTF of the presenilins. Nct needs PS to leave the endoplasmatic reticulum and to reach the cell surface; suppression of Nct expression with siRNA results in decreased steady-state levels of PS, indicating that Nct is one of the stabilizing factors of the PS fragments. Combination of PS and Nct is nevertheless not sufficient to increase the γsecretase activity in most cell lines. With special techniques two other proteins were discovered, namely Pen-2, a small, hairpin-like membrane protein and Aph-1, a multimembrane spanning protein that like PS is needed to correctly

24

Chapter 1

transport Nct to the cell surface83. Downregulation of Pen-2 or Aph-1 in cell cultures using siRNA leads to a decline in γ-secretase activity comparable to what was demonstrated with Nct and PS80. These four proteins all are necessary for γ-secretase activity because only overexpression of all four proteins leads to enhanced γ-secretase activity, not the overexpression of one or two proteins. This complex of proteins assembled to form γ-secretase cleaves quite a broad range of substrates and it appears that this secretase almost by default cleaves any type I integral membrane protein with an ectodomain that is shorter than a certain number of amino acid residues. This is an important issue concerning drugs aimed at this γ-secretase. It is clear that blockage of this secretase will also affect functions thereof that are necessary in other cleavage pathways as for example in Notch-signalling84. Mice with inactivated PS1 show a lethal partial Notch signalling deficiency. Inactivation of PS2 renders largely normal mice, and double-deficient mice show a severe Notch deficient phenotype with early embryonic death. Further research is therefore necessary to unravel the secrets of this specialized enzyme complex. Figure 4 shows a schematic representation of the enzyme complex which forms Aβ from AβPP via proteolytic processing by β- and γsecretase. All these data explain the formation of amyloid β and the way it is generated from its precursor protein AβPP. The working mechanism of γsecretase is as such that cleavage by this secretase can lead to Aβ1-40 as well as Aβ1-42. Research has revealed that there are numerous Aβ species with extensive amino and carboxyl terminal heterogeneity as was seen from soluble Aβ secreted by cells, soluble Aβ in CSF and insoluble Aβ isolated from AD brain85. More than 60 to 70 % is made up from Aβ1-40 although some Aβ142

is also present (approximately 15 %) and minor amounts of other peptides

(Aβ1-28, Aβ1-33, Aβ1-34, Aβ3-34, Aβ1-37, Aβ1-38, Aβ1-39).

Peripheral amyloidosis and Alzheimer’s disease: introduction

25

Aggregation Amyloid β plaques Aβ

Deposition

APP

Neurofibrillary tangles

Secretion

Nicastrin Presenilin A

1 2 3 4 5 6

D

7 8

1 2

D

β-secretase

1 2 3

PEN 2 NTF

4 5 6 7

APH-1

CTF

APP-ICD

nuclear signaling

Figure 4. Schematic drawing of the γ-secretase complex. Figure adapted from: Haass, C. EMBO Journal, 2004; 23.

The Aβ1-40 is the major species produced but it is the two amino acid longer Aβ1-42 that will lead to amyloid aggregation and deposition in the brain. This seems strange in view of the relatively low concentration of Aβ1-42 which is more prone to aggregation than Aβ1-40, but as can be seen in Table 2, most, if not all, forms of FAD will lead to an increase of Aβ1-42.

The neuritic plaque is the microscopic focus of extracellular amyloid deposition and is associated with axonal and dendritic injury and generally found in large numbers in the limbic and association cortices. The Aβ protein is principally deposited in a filamentous form, namely star-shaped masses of amyloid fibrils47. The dendritic injury is visible by dilated and tortuous neurites which contain enlarged lysosomes, numerous mitochondria and paired helical fragments that are generally indistinguishable from those that comprise the NFT’s.

26

Chapter 1

Table 2. Effects of some AβPP and PS mutants on Aβ1-40 and Aβ1-42. Table adapted from: Golde, T. E. et al. Biochim Biophys Acta, 2000; 1502. Model Transgenic mouse brain

Mutation Study PS1 Borchelt et al. (1997) PS2 Oyama et al. (1998)

Result Increase Aβ42/Aβ40 ratio by 50 % Increase Aβ42

Transfected cells

AAPNL

Increase Aβ total 5-6 fold

APP717 APP716 PS1 PS2

Citron et al. (1992); Cai et al. (1997) Suzuki et al. (1994) Eckman et al. (1997) Borchelt et al. (1996) Tomita et al. (1997)

Increase Aβ42/Aβ40 ratio by 60 % Increase Aβ42 Increase Aβ42/Aβ40 ratio 30-73 % Increase Aβ42

Plasma

APPNL APP717 APP716 PS1 PS2

Scheuner et al. (1996) Scheuner et al. (1996) Eckman et al. (1997) Scheuner et al. (1996) Scheuner et al. (1996)

Increase Aβ total 2-3 fold Increase Aβ42 2-fold Increase Aβ42 Increase Aβ42 by 100 % Increase Aβ42 by 50 %

Fibroblasts

APPNL

Citron et al. (1994); Johnston et al. (1994) Scheuner et al. (1996) Scheuner et al. (1996)

Increase Aβ42 total

PS1 PS2

Increase Aβ42 by 350 % Increase Aβ42 by 300 %

Plaques are also associated with microglia that express surface antigens for activation (like CD45 and HLA-DR) and they are surrounded by reactive astrocytes displaying abundant glial filaments. These astrocytes often ring the outside of the plaque while the microglia are found within and adjacent to the central amyloid core of the plaque. It is not exactly known how long it takes to develop such a plaque, but it takes perhaps many months or years. Although the plaques predominantly contain Aβ1−42 , the shorter Aβ1−40 is most frequently colocalized in the plaque. The cross-sectional diameter of neuritic plaques in microscopic brain sections varies widely from 10 to >120 µm. The density and degree of compaction of the fibrils also shows great variation among plaques.

Immunohistochemical staining revealed besides neuritic plaques far more Aβ deposits than had been visualized by the use of classical silver impregnation methods (Bielschowsky staining and Bodian stains). Using the

Peripheral amyloidosis and Alzheimer’s disease: introduction

27

most sensitive silver staining methods (modified Bielschowsky stain) one could also recognize many Aβ deposits that lacked the compacted, fibrillar appearance of the classical neuritic plaques. Furthermore, these deposits were also found in brain regions that are not clearly implied in the symptomathology of AD (thalamus, caudate, putamen, cerebellum) and they had no clear fibrillar, compact centre. These non-fibrillar, amorphousappearing plaques showed also little or no detectable neuritic dystrophy. Because these plaques were also present in the limbic and association cortices, researches came to the concept of ‘diffuse’ plaques or ‘preamyloid deposits’ because these diffuse plaques might represent precursor lesions of neuritic

plaques86,87,88.

These

diffuse

plaques

show

little

or

no

Aβ1−40 immunoreactivity (in contrast to neuritic plaques). The best support for this hypothesis were immunohistochemical studies of patients with Down’s syndrome89 or trisomy 21, in which there are three copies of chromosome 21 that bear the AβPP gene. These patients often display diffuse deposits as early as their teenage years, but they do not show neuritic/glial plaques until some twenty years later when abundant neurofibrillary tangles appear in limbic and association cortex.

Another hallmark in the brain of AD patients are neurofibrillary tangles (NFT), composed of hyperphosphorylated tau (τ). A lot of neurons in the typically affected AD brain regions (entorrhinal cortex, hippocampus, parahippocampal gyrus, amygdala, frontal, temporal, parietal and occipital association cortices and certain subcortical nuclei projecting to these regions) contain large, non-membrane-bound bundles of abnormal fibres that occupy much of the perinuclear cytoplasm47. These fibres consist of pairs of approximately 10 nm filaments wound into helices (paired helical fragments or PHF) with a helical period of ~160 nm as revealed by electron microscopy90. Immunocytochemical and biochemical analyses of these NFT’s suggested that they are composed of the microtubule-associated protein τ. By isolation of a subset of PHF and digestion with harsh proteases, τ proteins were released which migrate at a higher molecular weight on electrophoresis than normal τ prepared from tangle-free human or animal brains. This slower migration was

28

Chapter 1

shown to result from increased phosphorylation of tau. Gamblin and coworkers found a link between amyloid and neurofibrillary tangles in AD91 (although other researchers report the opposite of Gamblin92). They report that tau is proteolyzed in vitro in neurons treated with Aβ1-42 peptide by multiple caspases at a highly conserved aspartate residue (Asp421) in its Cterminus. Tau is rapidly cleaved at Asp421 in Aβ1-42 treated neurons and proteolysis appears to precede the nuclear events of apoptosis. Furthermore, the caspase cleavage of tau yields a truncated protein that lacks its C-terminal 20 amino acids and this renders the protein more prone to aggregation than wild-type tau. Members of the caspase family of cysteine proteases are activated in apoptotic neurons in AD and they play a critical role in Aβ induced apoptosis. So, accumulation of Aβ is followed by neuronal damage which triggers caspases that can cleave tau protein and lead to pathological filament assembly and/or inhibits disassembly (and thus stabilizes filaments). Certain kinases are also activated by Aβ and this leads to phosphorylation of tau at multiple sites (including Ser422) at the caspase cleavage site. The question arises if this has an influence on the sensitivity of tau for caspase cleavage. Tau protein kinase II, for example, plays a role in tau aggregation and hyperphosphorilation by direct interaction of soluble human recombinant tau with Aβ1-42 in vitro93.

Although a lot of the above-mentioned research was done very recently, it is still possible to implement these data in the amyloid cascade hypothesis (Figure 5), that was already proposed in 1991 by Hardy and Higgins94, but is still up-to-date.

Peripheral amyloidosis and Alzheimer’s disease: introduction

29

Lipid bilayer

AβPP N

IL-1 and IL-6

C

Aβ

Secretases: β

α γ

Endosomal-lysosomal pathway

Secretory pathway

Secreted AβPPβ

Secreted AβPPα Aβ

Aβ−associated proteins ACT, (S)AP, apoE, … fAβ

Activated microglia

KCai2+

free radicals, H2O2 KCai2+ Neuronal

Neurofibrillary

degeneration

tangles

Figure 5. The amyloid cascade hypothesis. AβPP can be processed via two different pathways: a) the secretory pathway in which AβPP is cleaved by αand γ-secretase; b) the endosomal-lysosomal pathway where AβPP is cleaved by β- and γ-secretase leads to the formation of Aβ that aggregates to form fibrillar amyloid β (fAβ) that in turn will lead to neurofibrillary tangles and neuronal degradation. IL-1 is interleukin 1; IL-6 is interleukin 6; ACT is α1antichymotrypsin; (S)AP is (serum) amyloid P component. Figure adapted from: Hardy, J. A. & Higgins, G. A. Science, 1992; 256.

Besides the formation of amyloid β plaques and NFT’s in AD, there are also different neurotransmitters that are affected during the disease. The first neurotransmitter abnormality documented in AD was related to the cholinergic system, but it soon became clear that this was accompanied by the loss of neurons using other neurotransmitters (e.g. glutamate, somatostatin, corticotropin-releasing factor, serotonin …)53. These deficits are also included in the pathogenesis described in Figure 6. Although these neurotransmitters and their receptors are also potential points of contact to develop tracer agents to visualize early AD, the focus of this study was aimed at tracer agents with binding affinity for amyloid β plaques.

30

Chapter 1

A hypothetical sequence of the pathogenetic steps of familial forms of Alzheimer’s disease Missense mutations in APP, PS1 and PS2 genes Altered proteolysis of APP Increased production of Aβ42 Progressive accumulation and aggregation of Aβ42 in brain interstitial fluid Deposition of aggregated Aβ42 as diffuse plaques (in association with proteoglycans and other amyloid-promoting substrates) Aggregation of Aβ40 onto diffuse Aβ42 plaques Accrual of certain plaque-associated proteins (complement c1q, etc.) “Inflammatory” response: -microglial activation and cytokine release - astrocytosis and acute phase protein release Progressive neuritic injury within amyloid plaques and elsewhere in the neuropil Disruption of neuronal metabolic and ionic homeostasis; oxidative injury Altered kinase/phosphatase activities J hyperphosphorylated tau J PHF formation Widespread neuronal/neuritic dysfunction and death in hippocampus and cerebral cortex with progressive neurotransmitter deficits DEMENTIA

Figure 6. A hypothetical sequence of the pathogenetic steps of FAD. Figure adapted from: Selkoe, D. J. Phys Rev, 2001; 81.

Peripheral amyloidosis and Alzheimer’s disease: introduction

31

1.4.2. Diagnosis of Alzheimer’s disease

1.4.2.1. Clinical diagnosis

Although much has been revealed about the complexity of AD and its pathogenesis, the exact relation between the different hallmarks of this disease is still not clear. AD is a progressive, neurodegenerative disease and the first symptom that can be detected by the physician is a gradual onset of progressive memory loss95. It therefore cannot be excluded that the clinician misses these subtle initial changes and that the disease remains unrecognized96. Motor, sensory and linguistic abilities are relatively preserved in early stages (as well as cognitive and social skills97) with impairment in memory and orientation functions. This early memory loss is most significant in the short-term memory while the long-term memory is mostly maintained for a longer time period. This hiatus in short-term memory leads to problems in learning new things. Moreover, the patient does not remember what he has done most recently (e.g. two minutes ago) while he still can tell what he did 20 years ago. Other manifestations are disinhibition, agitation, irritability, delusions, impaired drawing and route finding. These findings become more severe when the disease progresses. A lot of manifestations like behavioural and cognitive features of AD and progression of intellectual deterioration are variable. Some patients have periods of arrested decline intermingled with periods of extremely rapid deterioration. Most patients though, have a rather steadily progressive disease with survival of 5 to 12 years after clinical diagnosis of the disease. In the later stages all intellectual capacities are lost and other difficulties arise like problems with movement and even urinary and faecal incontinence. Mutism may be observed as well as extrapyramidal rigidity and spasticity. The cause of death is often a rather trite, incidental illness like pneumonia or systemic infection, which are diseases that can normally be cured rather easily.

Because AD cannot yet be definitively diagnosed with non-invasive methods (as mentioned before), the clinical diagnosis can only be a probable

32

Chapter 1

diagnosis. Of course, different approaches have been undertaken to yield criteria that will lead to a probable diagnosis that has a high degree of certainty, but unfortunately, it does not provide a 100 % certainty. A working group was established to reach that goal, namely the National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer’s disease and Related Disorders Association (or NINCDS-ADRDA). This working group drafts criteria that divide the diagnosis in probable, possible or definitive AD98. These criteria are depicted in Table 3 and are compatible with those of DSM-IV (Diagnostic and Statistical Manual-IV) and the International Classification of Diseases (ICD)99. However, the criteria of NINCDS-ADRDA are more detailed for the diagnosis of the disease.

These NINCDS-ADRDA criteria proved valid for clinical and research purposes. As stated by Tierney and co-workers99 the overall accuracy rate between clinical and neuropathologic classification of AD was high and ranged from 81 to 88 % across the 9 neuropathologic criteria they used in that study for probable AD. This confirms the quality of the criteria and, furthermore, there are a few features that render the diagnosis of probable AD uncertain or unlikely, namely sudden apoplectic onset; focal neurological findings such as hemiparesis, sensory loss, visual field deficits and incoordination early in the course of the illness; seizures or gait disturbances at the onset or very early in the course of the illness. The dementia should be diagnosed by clinical examination and should be documented by the minimental state test100, Blessed dementia scale101 or similar examinations102,103 and should be confirmed by neuropsychological tests104.

Peripheral amyloidosis and Alzheimer’s disease: introduction

33

Table 3. NINCDS-ADRDA criteria for diagnosis of Alzheimer’s disease Definite Alzheimer’s disease Clinical diagnosis for probable Alzheimer’s disease Histopathological evidence for Alzheimer’s disease

Probable Alzheimer’s disease Dementia* established by clinical examination and documented by mental status questionnaire or similar examination Dementia confirmed by neuropsychological testing Deficits in two or more areas of cognition Progressive worsening of memory and other cognitive functions No disturbance of consciousness Onset between ages 40 and 90, most often after 65 Absence of systemic disorders or other brain diseases that could account for the progressive deficits in memory and cognition

Possible Alzheimer’s disease Atypical onset, presentation, or progression of a dementia syndrome without a known aetiology A systemic or other brain disease capable of producing dementia but not thought to be the cause of the dementia that is present There is a gradually progressive decline in a single intellectual function in the absence of any other identifiable cause

* Dementia is defined as an acquired, global and progressive decline of memory and other cognitive functions in the absence of impaired consciousness

The use of imaging techniques can aid in differentiating between treatable and non-treatable causes of dementia. Treatable causes of dementia are, for example, tumours, brain abscesses or strokes105. Computed tomography or CT is the most widely available technique. Specific lesions in the brain can be visualized with CT, but the question is what the diagnostic

34

Chapter 1

value is of CT for AD. Certainly in the early stages of the disease it is not useful. In later stages it might be helpful to visualize brain atrophy which is linked to progressive AD, but this partially implicates that there are previous CT images of the same patient needed for comparison. Another technique is MRI or magnetic resonance imaging106. MRI allows examining the patient without the aid of ionizing radiation, can produce accurate measurements of the brain and CSF volume and can discriminate between grey and white matter in the brain. Nevertheless, the same question arises as with CT. Thus, it seems that AD cannot be diagnosed on CT and/or MRI images alone. Other imaging techniques are SPECT and positron emission tomography (PET), both requiring a radiolabelled tracer. Routinely used SPECT and PET tracers like

99m

Tc-exametazine (99mTc-HMPAO or

oxime) and

18

99m

Tc-hexamethylpropyleneamine

F-FDG (2-[18F]fluoro-2-deoxy-D-glucose)107 respectively, can be

useful in the diagnosis of Alzheimer’s disease because they visualize cerebral blood flow and regional brain glucose metabolism, respectively. The intra- and interindividual variability of the changes in cerebral blood flow and glucose metabolism nevertheless hamper the use of these tracers for visualization of AD108. Therefore, it will be useful if a PET or SPECT tracer becomes available that can specifically bind to amyloid plaques or other hallmarks that are specific for AD. The electroencephalogram (EEG) and lumbar punction are two other means proposed by the NINCDS-ADRDA criteria, but they suffer from the same problem as FDG-PET and

99m

Tc-HMPAO-SPECT, namely the

large inter- and intra-individually variability. Up-to-date, the use of histological data is a necessity to obtain a 100 % certainty for the diagnosis of AD.

1.4.2.2. Neuropathological diagnosis The silver stains Alois Alzheimer used at the beginning of the 20th century were new at that time, but they are still useful to date, albeit in a modified way. As became clear through years of research, extracellular neuritic plaques, intracellular neurofibrillary tangles and amyloid deposits in cerebral blood vessel walls, are three histological landmarks of Alzheimer’s disease. Although preamyloid or diffuse plaques are thought to be precursors of neuritic plaques, they are no useful histological markers for AD because they

Peripheral amyloidosis and Alzheimer’s disease: introduction

35

can be found in the brain of normal aging persons and in brain areas commonly not affected in AD. Furthermore, as mentioned earlier, these diffuse plaques cannot be stained with Congo red (see 1.2). Both neuritic plaques and diffuse plaques can, however, be stained immunohistochemically with antibodies109 directed to the amyloid β protein or with the modified Bielschowsky technique (silver stain). The neuritic plaques can also be stained with Congo red and thioflavin S110, two stains that have high affinity for the β-pleated sheets present in amyloid β. The NFT’s can be visualized with the aid of immunohistochemistry using tau antibodies and also with the modified Bielschowsky technique or thioflavin S, which provides evidence for a broader affinity range of thioflavin S than only for β-pleated sheets. Yamamoto and Hirano nevertheless111 showed better sensitivity and specificity for the modified Bielschowsky stain than for thioflavin S when used for NFT staining. The amyloid β deposits that are present in arterioles and small

arteries

in

the

cortex

(cerebrovascular)

or

leptomeninges

(meningovascular) can also be stained with Congo red or thioflavin S112.

Because of the specific brain regions where these deposits are found, a few minimum microscopic criteria were proposed105. Three regions of neocortex should be sampled (frontal, temporal and parietal lobes) together with the amygdala, the hippocampal formation, the basal ganglia, the substantia nigra, the cerebellar cortex and the spinal cord. Other obvious causes of organic dementia should be excluded such as chronic subdural haematoma,

neoplasm,

Pick’s

disease

and

multi-infarct

dementia.

Furthermore, criteria were determined with the number of NFT’s and plaques that should be present in a microscopic field of 1 mm² with a microscopic magnification of 200x.

It is clear that when immunohistochemistry is used for post mortem visualization of amyloid β plaques, care should be taken, for this technique cannot differentiate between neuritic and diffuse plaques113. The consortium to establish a registry for Alzheimer’s disease (CERAD) assesses senile plaques as well as NFT’s (although one has to be careful as to the techniques

36

Chapter 1

used to identify NFT’s) as a neuropathological criterion for the diagnosis of AD114. Semi-quantitative neuropathological data and clinical data, together with age, result in a definitive, probable or possible diagnosis of AD according to CERAD criteria. A combination of NINCDS-ADRDA criteria for the clinical diagnosis with CERAD criteria will lead to 100 % certainty in the diagnosis of AD113.

1.4.3. Treatment of Alzheimer’s disease: current and future therapies

In view of the devastating nature of AD, an effective therapy has been a long awaited goal. The mental regression of an AD patient is hard to bear, for the patient as well as for his family. The AD patient’s body, although physically often in a rather fine shape, becomes an empty shell and the care for this patient exceeds the means and hopes of family members who find this care very laborious and time consuming. Uptake of the patient in specialized caring facilities is a gigantic step to take and is fairly expensive. It is, however, commonly the only possibility left and the increasing number of AD patients will lead to a great need for those facilities, and of course for an effective cure.

Drugs currently used in the therapeutic scheme of AD patients are all symptomatic and none of them actually cures the disease. Until now, no cure has been found, but scientists have made extensive progression throughout the last five years, which will hopefully lead to an effective cure. Donepezil ®

®

®

(Aricept ), galantamine (Reminyl ) and rivastigmine (Exelon ) are all cholinesterase inhibitors which lead to an increase of acetylcholine in the brain and therefore can aid the patient when he suffers from a decrease of this neurotransmitter as a consequence of AD115. These products stabilize the memory decline for a variable period (typically 6 to 12 months), but all patients reach a stage in which these products are no longer effective and the AD deploys itself completely. The clinical usefulness of these drugs is doubtful, but to date these are the only products that can influence the progress of the disease in an early stage. Moreover, there are side effects caused by the cholinergic effect of the drugs which can confuse the patient even more.

Peripheral amyloidosis and Alzheimer’s disease: introduction

37

Because AD is often accompanied by symptoms of depression and anxiety, anti-depressants and anxiolytics are frequently added to the therapeutic scheme of the patient. Multi-drug therapy is not a recommendable issue, but in case of AD, it is often necessary. Evidently, the compliance of a patient that is extremely forgetful due to his disease is an additional problem in AD treatment. Studies with a benzothiazole analogue (Sabeluzole) have also been performed116. The expected effect of sabeluzole in AD is slowing of the neuronal degeneration, thereby reaching a positive impact on cognition. The properties ascribed to sabeluzole are: potentiation of the action of nerve growth factor; acceleration of axonal transport; promotion of neuritic outgrowth in neuroblastoma cells, hippocampal neurons and dorsal root ganglion neurons; prevention of tangle formation and increased cell survival in hippocampal cell cultures exposed to a hyperstimulating medium.

Recently, due to the increasing knowledge about the pathogenesis of the disease, drugs have been designed that are hoped to have a much more curative effect than the ones used to date. One example is the use of induction of Aβ immunity117,118,119. Injecting transgenic AD mice with the purified Aβ protein leads to the development of antibodies against the Aβ protein. Studies in AD mice showed that this immunization therapy not only inhibited the generation of new plaques, but could even lead to the removal of already existing plaques117. DeMattos et al.120 showed that the peripheral administration of a monoclonal antibody (m266) which is directed against the central domain (residues 13-28) of Aβ, can reduce the cerebral Aβ burden in mice by acting as a peripheral sink and shifting the central nervous systemplasma Aβ equilibrium. Active immunization with synthetic intact Aβ42 or conjugated fragments of Aβ42 has also been evaluated in humans while techniques for passive immunization with human anti-Aβ monoclonal antibodies (mAβ) are approaching clinical development. Phase II clinical trials of the active immunization study had to be halted because some patients developed fatal aseptic meningo-encephalitis53. It may seem strange that antibodies are able to pass the BBB but that is what seems to happen in this case. Zhang et al.121 also prepared conjugates that can bind to specific

38

Chapter 1

domains in Aβ and these conjugates might also be useful to break β-sheets. Removal of SAP is another option as mentioned already in 1.3.3, but when this has to be done in the brain, the drug should be able to cross the BBB. Blocking of β- and/or γ-secretase seems to be the most favourable option, but as is clear from the pathogenesis of AD, the biological function of BACE-1 has not yet been determined and the question arises which unwanted consequences blocking BACE-1 will have. However, in BACE-1 knock-out mice, there is no generation of any Aβ and these mice appear to be healthy without

perceivable

neurological

or

behavioural

abnormalities122,123.

Nevertheless, further research has to be performed in this field. As to blockage of γ-secretase, caution is also needed because of its function in Notch processing which is of great importance for cell survival. However, extensive medicinal screening already led to compounds like N-[N-(3,5difluorophenacetyl)-L-alanyl]-S-phenylglycine t-butyl ester (DAPT), which caused an acute, dose-dependent reduction of Aβ in brain and plasma when given orally to APP transgenic mice124. The brain levels of the C99 and C83 fragments increased, confirming that DAPT is a good γ-secretase inhibitor. Further studies are needed to see what the effect will be of continuous administration of DAPT, especially with regard to Notch signalling.

On the other hand, Weggen et al. reported that administration of certain non-steroidal anti-inflammatory drugs (NSAID’s) in the appropriate doses, leads to a reduction in Aβ1-42 formation independently of their influence on cyclooxygenase activity125. Doses used to obtain this effect are very high and further research is necessary to obtain more accurate data on this matter. In epidemiological studies it was observed that chronic use of NSAID’s had some protecting effect against AD, associated with a significantly lower risk of developing AD126. Due to the increased insight in the pathology, hopes are currently high to find a cure for this devastating disease in the near future.

Peripheral amyloidosis and Alzheimer’s disease: introduction

39

1.5. Aims of this study The definitive diagnosis of both amyloidosis and AD requires the aid of clinical as well as histological data. While biopsies can be obtained from PA patients, it has to be stated that this technique has a lot of drawbacks. It is invasive, there is a high risk for haemorrhage, it takes some time to process the biopsied material and there is a risk of false negative results due to the sampling technique and according sampling error. On the other hand, noninvasive techniques like the use of

123

I-SAP are very specific tools that can

only be used in specialized centres and that also bear potential risk factors due to the human origin of the protein used (i.e. the danger of viral contamination of the sample). All these drawbacks make the search for a full synthetic radiolabelled tracer that can visualize amyloid deposits in peripheral tissue more than worthy.

The diagnosis of AD is an even more complex case. Visualization of the amyloid deposits in the brain is at this time only possible after the patient’s death. Moreover, these techniques are expensive and the benefit for the patient is zero.

There has been made a lot of effort in the last decade in the search for a radiolabelled tracer that can be used for in vivo visualization of amyloid deposits in either peripheral tissue or in the brain. The amyloid cascade hypothesis, the modified/elaborated hypothesis of Selkoe (Figures 4-5) and the demonstration of deposition of amyloid due to the formation of secondary protein structures prone to aggregation in systemic amyloidosis, indicate that amyloid deposits constitute a central and probably early event in the pathogenesis of both PA and AD. Furthermore, the presence of these chameleon protein deposits is in both diseases a main hallmark. Therefore, a tracer agent that specifically binds to these deposits, will provide an important tool for the non-invasive in vivo diagnosis of these illnesses.

40

Chapter 1 One has to take into account a few important factors that can

jeopardize the biological features of such a tracer agent. Primarily, with respect to AD, a difficult issue to resolve is the fact that a radiolabelled tracer useful for in vivo visualization of amyloid β, has to be able to cross the BBB. Since the BBB is a difficult barrier to cross, this is not an easily resolvable issue. Secondly, chemical modification required for the labelling of molecules with a known affinity for amyloid β, may decrease this affinity. A last requirement is that the radioactive isotope used for the labelling of the molecule, should have ideal properties for in vivo visualization. As mentioned before, γ-rays, which easily pass through tissue, can be detected outside the body with the aid of a SPECT or PET camera (see further on).

In case of radiolabelled tracers for the in vivo visualization of amyloid deposits in peripheral tissue, the affinity for amyloid deposits and the emission of γ-rays remain extremely important. Passage through the BBB is not necessary, but high uptake in the abdominal region should be avoided to limit background noise.

A few molecules have already been mentioned throughout this introduction that display in vitro affinity for amyloid deposits. The structures of Congo red and thioflavin-T are depicted in Figure 1, together with X-34 (which was only synthesized and described in 200025, while derivatives of thioflavin-T only entered literature during 2001). Derivatives of Congo red were reported before 2000 (some structures of the molecules mentioned down below are depicted in Figure 7). The group of Lansbury127,128 reported two technetium99m labelled derivatives of both Congo red and chrysamine G (Figure 7, A). Due to the presence of polar groups and the high molecular mass, these molecules show a limited brain uptake and can therefore only be useful for the diagnosis of PA129. The authors proposed to coinject BBB penetration enhancers with these molecules130. Mathis and co-workers131 reported an 125Ilabeled derivative of chrysamine G in 1997 (Figure 7, B). Dezutter et al.132,133 also reported a technetium-99m labelled derivative of chrysamine-G (Figure 7, C) which showed in vitro affinity for amyloid deposits but was not able to cross

Peripheral amyloidosis and Alzheimer’s disease: introduction

41

the BBB, probably because of its high molecular mass of 844 Da134. An additional problem with these Congo red and chrysamine G derivatives is the presence of functional groups that can be ionized at physiological pH and thereby inhibit the BBB passage. The chrysamine G derivatives also bear two diazo groups which are believed to be potentially carcinogenic135 after oral intake. Although these radiolabelled tracers are injected intravenously they show extensive hepatobiliar excretion to the intestines. All these factors led to the introduction of X-34 in 2000. In X-34 the chemical structure is changed from two central rings to one ring and the diazo groups are replaced with carbon double bonds as shown in Figure 1. Although X-34 can be used to visualize amyloid plaques in vitro, its in vivo properties are not so favourable because of the limited brain uptake. Besides, synthesis of derivatives of X-34 is difficult, which makes these tracer agents less attractive. The groups of Klunk and co-workers136 and Kung and co-workers137 were the first to introduce the thioflavin-T derivatives. Zhuang et al. described two thioflavin-T derivatives including 125I-TZDM (Figure 7, D) whereas Klunk et al. described the evaluation of carbon-11 labelled 6-Me-BTA-1 (BTA = benzothiazole-aniline) (Figure 7, E). While the two thioflavin-T derivatives of Kung showed low brain uptake, the BTA analogue of thioflavin-T by Klunk showed high uptake and fast wash-out in normal rodent brain. Furthermore, all these tracers show good in vitro affinity for Aβ fibrils and the BTA analogue was shown to stain neuritic plaques as well as NFT’s depending on the concentration of BTA. On the basis of these promising results, we decided to explore other benzothiazoles for in vivo visualization of amyloid deposits. It was also clear from the above-mentioned results of Lansbury, Dezutter, Klunk and Kung that these benzothiazole derivatives showed more potential than styrylbenzenes (Figure 7, F) and derivatives of Congo red and Chrysamine-G. Furthermore, synthesis of these benzothiazoles is more convenient and straight forward138,139,140,141,142.

In the last three years the groups of Klunk and Kung have reported extensively about radiolabelled tracers for in vivo detection of amyloid β

42

Chapter 1

plaques in the brain of AD patients. Klunk and co-workers published the first clinical data in 2003 with the carbon-11 labelled compound (N-[11CH3]-2-(4’methylaminophenyl)-6-OH-benzothiazole, also referred to as Pittsburgh compound-B (Figure 7, G)143,144,145,146,147,148,149,150. Kung and co-workers on the other hand151, reported derivatives of thioflavin-T with an imidazo[1,2a]pyridine152,153,154 structure (iodo-imidazo[1,2-a]pyridine (IMPY), Figure 7, H), a benzoxazole structure155 (iodo-benzoxazole (IBOX), Figure 7, I), and with a benzofuran ring instead of the benzothiazole ring156. More recently, they also reported dimethylamino-fluorene157 (Figure 7, J) and stilbene158,159 (Figure 7, K)

derivatives

and

made

a

comparison

between

iodine

labelled

benzothiazoles, imidazo[1,2-a]pyridine, stilbenes and fluorenes of which the imidazo[1,2-a]pyridine showed the most promising characteristics for future in vivo visualization of amyloid β in AD brain160. Barrio and co-workers reported in 1996161 a fluorophore called DDNP (1,1-dicyano-2-[6-(dimethylamino)naphtalen-2-yl]propene),

a

structure

analogue of the NSAID naproxen. A few years later, they suggested the possible use of this fluorophore, when labelled with fluorine-18, for visualization of amyloid β plaques162. Although the reported biological background of [18F]-FDDNP (Figure 7, L) was rather incomplete, this group already published results of a study with this compound in patients163,164,165.

It has been shown that FDDNP has affinity for binding domains on amyloid β fibrils which are different from these of the benzothiazoles. Some structurally related NSAID’s like naproxen and ibuprofen where shown to share the Aβ binding domains of [18F]-FDDNP166. Recently it has also been shown that FDDNP binds to prion plaques in vitro167, suggesting that it has affinity for β-pleated sheets in general. Even more recently Suemoto et al.168 have described acridine orange analogues called BF-009 and BF-108 (Figure 7, M). BF-108 labelled amyloid plaques in APP23 transgenic mice after intravenous injection. This group also evaluated styrylbenzoxazole derivatives in vivo169,170. Fluorine-18 labelled BF-

Peripheral amyloidosis and Alzheimer’s disease: introduction

43

168 (Figure 7, N) had high in vitro binding affinity for Aβ fibrils and also showed in vivo affinity in two different transgenic mouse models. From these results, it can be stated that the search for a useful in vivo tracer is still fully ongoing.

On the basis of the information in recent literature, we decided to synthesize and evaluate derivatives of thioflavin-T because of: -

its low molecular mass

-

the possibility to obtain neutral derivatives

-

the relatively accessible chemistry

-

its affinity for amyloid β (which increases when the positive charge is removed)

-

the pronounced fluorescence

-

the fact that some uncharged derivatives had already shown the ability to cross the BBB

The importance of AD as a disease syndrome is enormous, but this cannot minimize the importance of systemic amyloidosis and it is very quaint that there is much less interest from a research point of view for tracers for PA than for AD. The aim of our study therefore, was the development of tracer agents for the in vivo visualization of AD as well as PA. The research started from thioflavin-T and used this as the backbone to synthesize its derivatives that could be labelled with technetium-99m (chapter 2), carbon-11 or fluorine18 (chapter 3). In case of tracers for PA, enhancement of urinary excretion was strived after by using charged derivatives. In case of agents for AD, uncharged, lipophilic tracers that would be able to cross the BBB, were the focal point. The biological evaluation of these tracers was done in normal mice (chapter 4) while their in vitro binding characteristics were obtained through different binding studies (chapter 5).

44

Chapter 1

tBu

HOOC

H

N

COOH

N

N HO

N

S

HO

N N

A

OH

N

C N

G

N

(Klunk 143- 150,

CH 3

3)-6-O H-BTA-1)

Tc

( Lan sbury12 7,128)

C

N

C

C

N

N tBu

N

tBu

NMe2

N

I

H (Kung 152-1 54, IMPY)

tBu 125

I

N

N HO

11

N-(11CH

NMe2

N

N

N

B (Klunk 131 )

HOOC

O

I

OH

I (Kung 155, IBOX)

COOH

O

N Tc

CH2 O

HOOC

S S

J (Kung157 , 7- iodo -2-N,N-dimethylaminofluorene)

COOH

NH

HO

NMe2

I

S

O

N

N N

I

OH

N

NMe2

C (Dezut ter132 ,13 3)

K (Kung 158,159 , m-I-Stilbene)

N NC

NMe2

CN

S

125 I

D (Kung 137,

CH3

125I-TZDM) 18

F

N

N NHMe

L (Barrio162-165 , 18F[ FDDNP])

S

H 3C

E (Klunk 13 6, 6-CH3-BTA-1) H 3C R 2O

OR 2

H 3C

R1

HOOC

N

N

F

N

M (Suemoto 168, BF-108)

CH3

COOH R1 BSB ISB IMSB

Br I I

R2 N

H H Me

18

F

O N

O

H CH3

N (Suemoto169 -170, [1 8F]BF-168)

F (Kung137 , styrylbenzenes)

Figure 7. Chemical structures of reported potential tracers for visualization of amyloid

β

plaques

iodo(methyl)styrylbenzene)

(BSB

=

bromostyrylbenzene,

I(M)SB

=

45

CHAPTER 2

2. DEVELOPMENT OF THIOFLAVIN-T DERIVATIVES FOR SPECT STUDIES

2.1. Single

Photon

Emission

(Computed)

Tomography

(SPE(C)T) SPECT is a technique for producing cross-sectional images of the in vivo distribution of a radiotracer within the body by external detection of individual γ−rays (= single photons) emitted during the decay of the radionuclide. These γ−rays arise from transition of a nucleus from an excited state to either a lower energy level or its ground state. The transition from a higher to a lower nuclear energy level results in the emission of a γ−ray with an energy corresponding to the difference in energy between the subsequent nuclear energy levels. The γ−rays of sufficient energy show a high penetration through tissue because of their non-particulate nature, and can be detected outside the body with the aid of a gamma camera. Typical radionuclides used in SPECT are mainly technetium-99m and iodine-123 and to a lesser extent thallium-201, gallium-67, indium-111, chromium-51 and iodine-131.

In view of its favourable properties, technetium-99m is the most commonly used radionuclide for clinical SPECT studies.

99m

Tc has a short

half-life of 6.02 hours which is long enough to perform the labelling procedure and to allow nuclear imaging. Its g-ray has a nearly ideal energy of 140 keV and no particulate radiation (α- or β-rays) is emitted during the decay of 99mTc. The latter, as well as the short half-life, result in a low radiation dose to the patient.

99m

Tc is readily available from a

low cost. In these generators Al2O3 and the decay of

99

99

99

Mo/99mTc generator at a relatively

MoO42- is adsorbed on a column packed with

Mo yields pertechnetate (99mTcO4-). The monovalent

46 anion

Chapter 2 99m

TcO4- can be eluted from the Al2O3 column with a saline solution,

whereas the divalent molybdate stays firmly adsorbed on the column. The resulting saline solution, containing

99m

TcO4-, can be used as such (e.g. for

visualization of the thyroid) or serves as a source of technetium to label a variety of molecules. The molecule to be labelled needs to have a suitable combination of donor atoms which allow the efficient chelation of technetium. As the combination of technetium with this Tc chelating system may be quite bulky, its presence may alter the biological behaviour of the compound to be labelled especially for low molecular weight molecules.

Most bioactive molecules that are potentially useful for imaging do not carry such a technetium binding ligand, nor do they have enough donor atoms (N, O, S, P) in a suitable configuration to form a stable complex with technetium-99m. In that case a chelating ligand has to be introduced in the biocompound to allow labelling with technetium-99m. This metal binding bifunctional chelating ligand (BCL) is in general a structure which, on the one hand, forms a stable complex with 99mTc and, on the other hand, also contains a functional group that can be used to covalently bind the BCL to the bioactive compound. The chelating group may by separated from this functional group by a spacer. Although the introduction of such a BCL may lead to changes in affinity of the bioactive compound for its receptor or target, this technique has long been used to enable researchers to obtain technetium-99m labelled agents for different purposes. A lot of Tc-tracers routinely used in the clinic are nevertheless composed of a chelating ligand as such (e.g.

99m

Tc-L,L-

ethylcysteinate dimer (99mTc-ECD)) and only some recently developed radiopharmaceuticals (e.g.

99m

Tc-HYNIC-annexin-V171) are composed of

bioactive compounds covalently bound to a BCL.

Labelling of such BCL linked bioactive compounds with

99m

Tc can be

done in different ways: a direct or exchange labelling procedure using the BCL-bioactive molecule (BCL-BAM) conjugate and a procedure in which the BCL is labelled first with technetium-99m and is then coupled to the bioactive compound. The latter is called the preformed chelate approach and has the advantages that the bioactive compound is not subject to harsh conditions

Development of thioflavin-T derivatives for SPECT studies

47

that are often required for labelling (alkaline or acidic pH, heating at 100 °C…) and that there is no non-specific binding of technetium-99m to the BAM. On the other hand, this is a rather complex and time consuming technique and it is very hard to convert it into an instant labelling kit formulation.

Direct or exchange labelling allows one-pot labelling. Direct labelling is performed by simply adding pertechnetate to the labelling vial after which Tc is bound by the BCL. Exchange labelling requires the presence of a chelating ligand such as tartrate which forms a weak intermediary complex with technetium-99m during the deprotection of the BCL, in this way preventing the formation of colloidal technetium. After deprotection of the BCL, an exchange of technetium will take place from the weak 99mTc-tartrate complex to the BCL, yielding the 99mTc-labelled BCL-BAM (Figure 8).

O

O O

NH

O COOH

NH +

S

HN

O CO-BAM CO S-benzoyl-MAG3-BAM

H

OH

H

OH

O

NH

NH

COO

TcO 4-

99m

Sn

2+

+ S

COOH

HN

CO

H

Tc

O

OOC O

H OH O CO-BAM COOH

H

HO

H COOH

Tartric acid

T=100 °C O

O O

COOH

N O N +

Tc S

N

O CO-BAM 99m Tc-MAG3-BAM

H

OH

H

OH COOH

O

NH

O COO

NH +

SH

HN

O CO

Tc

H

O

H

OH

OOC O

H

HO

H

COOH

COOH

BAM

Figure 8. Example of the exchange labelling of mercaptoacetyltriglycine-BAM (MAG3-BAM)

Most widely used BCL’s are tetraligands. They have four donor atoms able to bind to Tc in a suitable structural arrangement allowing the formation of three to four 5- or 6-rings upon complexation with Tc. Examples of such tetraligand

systems

are

mercaptotriamides

(SN3,

e.g.

mercaptoacetyltripeptides), diamide dithiols (DADT; Figure 9 a), mono amine

48

Chapter 2

mono amide (MAMA; Figure 9 b) and bis-aminoethanethiol (BAT; Figure 9 c)172,173. The first two yield negatively charged

99m

Tc-complexes, whereas the

MAMA and BAT ligands yield neutral 99mTc-complexes174.

O

* O

NH

HN

SH

HS

a

O

O

HO

NH

HN

SH

*

HS

HN *

NH SH

b

O

O

NH

HN

SH

HS

c

HO

d

*

O

Figure 9. Bifunctional chelating agents: a: DADT; b: MAMA; c: BAT; d: mercaptoacetyl-L-asp-gly. * indicates the site of coupling with a bioactive compound.

In the present study, a TT derivative was prepared because of the favourable characteristics (see 1.5) of TT and the assumption that the addition of a bulky BCL to TT would still yield a tracer agent with a favourable molecular

mass.

Therefore,

TT

was

coupled

to

a

(protected)

mercaptoacetyltripeptide (S-benzyl mercaptoacetyl-L-asp(tBu)-gly175 (Figure 9, d) to obtain a negatively charged such a

99m

99m

Tc-complex. It was hypothesized that

Tc-labelled TT derivative could be most appropriate for detection of

PA in view of its more polar characteristics which would favour renal excretion and decrease the hepatobiliary excretion. On the other hand, a conjugate of benzothiazole with a BAT ligand was prepared as well, in order to obtain an uncharged

99m

Tc-labelled thioflavin derivative. The latter was supposed to

have potential as a tracer agent for in vivo visualization of amyloid plaques in the brain in AD.

In addition, two other BCL approaches were used to develop charged 99m

Tc-labelled BTA derivatives for detection of PA. In one of these tracer

agents BTA was coupled to a hydrazinonicotinic acid (HYNIC) ligand (see further on) because of the well known polar characteristics of the

99m

Tc-

Development of thioflavin-T derivatives for SPECT studies

49

labelled HYNIC complexes176,177,178. The HYNIC approach for labelling with technetium-99m requires a so-called co-ligand (tricine or ethylenediamineN,N’-diacetic acid = EDDA) to obtain sufficient donor atoms for the stable chelation of technetium.

For the labelling of the fourth tracer we used the relatively new technetium-tricarbonyl179 method. In this approach, first a precursor is formed by reaction of

99m

99m

Tc(CO)3(OH2)3

TcO4-, CO, Na2CO3, NaBH4 and

NaKtartrate. This precursor is then reacted with a BTA derivatized with a side chain containing an iminodiacetic acid (IDA) moiety. In this way, a negatively charged 99mTc-BTA derivative is formed.

In this chapter the synthesis of 4 derivatives of thioflavin-T, the deprotection of the technetium binding groups, the labelling procedure and purification methods are described.

2.2. Materials and methods All reagents and solvents used in synthesis were obtained from Acros Organics (Geel, Belgium), Aldrich, Fluka or Sigma (Sigma-Aldrich, Bornem, Belgium) and were used without further purification. pH values of nonradioactive solutions were measured with a P600 pH-meter (Consort, Turnhout, Belgium) provided with a glass electrode. pH determination of radioactive solutions was done with paper pH-strips (pH strips pH 0-14, Merck, Darmstadt, Germany). Evaporation of organic solvents under reduced pressure was done with a Büchi Rotavapor® (Büchi, Flawil, Switzerland). MgSO4 was used as a drying agent. Purification of reaction mixtures was done by column chromatography with silica gel (particle size between 0.04 and 0.063 mm (230-400 Mesh) (MN Kieselgel 60M, Macherey-Nagel, Düren, Germany) as the stationary phase or by medium pressure liquid chromatography (MPLC). Thin layer chromatography (TLC) was done on precoated

silica

TLC

plates

(DC-Alufolien-Kieselgel,

Fluka,

Buchs,

Switzerland). The structure of the synthesized products was confirmed with

50 1

Chapter 2

H-nuclear magnetic resonance (NMR) spectroscopy on a Gemini 200 Mhz

spectrometer (Varian, Palo Alto, CA, USA). Chemical shifts are reported as δvalues (parts per million) relative to tetramethylsilane (δ=0). Coupling constants are reported in hertz. Splitting patterns are defined by s (singlet), br s (broad singlet), d (doublet), dd (double doublet), t (triplet) en m (multiplet). Peak assignment was based on theoretical data from ChemWindow® 6.5 Spectroscopy (Bio-Rad laboratories, Sadtler division, PA, USA), on literature values and, if possible, on comparison with values from previous compounds in the synthetic chain. Melting points were determined with an IA9100 digital melting point apparatus (Electrothermal, Essex, UK) in open capillaries and are reported uncorrected.

Generator eluate containing

99m

Tc in the form of pertechnetate was

obtained from an Ultratechnekow generator (Tyco Healthcare, Petten, The Netherlands).

Quantitative determination of radioactivity in samples was done in an automatic gamma counter coupled to a multi-channel analyzer (Wallac 1480 Wizard® 3”, Wallac, Turku, Finland).

Combined liquid chromatography and mass spectrometry (LC-MS) was performed on a system consisting of a Waters Alliance 2690 separation module (Waters, Milford, MA, USA) coupled to a reverse phase Xterra™ MS C18 3.5 µm column (2.1 mm x 50 mm) (Waters). The eluent was a mixture of acetonitrile and 0.05 M ammonium acetate at a flow rate of 0.3 ml/min. The eluate was first analyzed by a UV-spectrometer (Waters 2487 Dual wavelength absorbance detector), subsequently by radiometric detection consisting of a 3-inch NaI(Tl)-crystal coupled to a single channel analyzer (The Nucleus, Oak Ridge, TE, USA) and ultimately by a time-of-flight mass spectrometer (LCT, Micromass, Manchester, England) equipped with an electrospray ionization (ESI) source. Acquisition and processing of data was performed with Masslynx software (version 3.5).

Development of thioflavin-T derivatives for SPECT studies

51

Reversed phase high pressure liquid chromatography (RP-HPLC) runs were performed with a system consisting of a TSP SpectraSeries P4000 quaternary pump (Thermo Separation Products, San Jose, CA, USA) connected with a TSP SpectraSeries UV 100 detector (Thermo Separation Products) and a 3-inch NaI(Tl) crystal connected with a Medi-Lab Select SC7II single channel analyzer (Medi-Lab Select, Mechelen, Belgium).

Gas chromatography was performed with a DI 200 gas chromatograph (Delsi Instruments, Suresnes, France) with a Porapak® QS 80/100 column (Alltech) of 180 cm x 0.25 inch.

2.2.1. Synthesis of a benzothiazole-BAT conjugate as a precursor for a neutral 99mTc-BTA derivative

2.2.1.1. Synthesis of S,S'-bis-triphenylmethyl-N-(t-butoxycarbonyl(BOC))-N’(acetic acid)-1,2-ethylenedicysteamine

S,S'-Bis-triphenylmethyl-N-(BOC)-N’-(acetic acid ethyl ester)-1,2ethylenedicysteamine (2.1)

A solution of S,S'-bis-triphenylmethyl-N-(BOC)-1,2-ethylenedicysteamine (3.1 g, 4 mmol), N,N-diisopropylethylamine (DIEA) (2 ml, 11.2 mmol) and ethyl bromoacetate (2 ml, 18 mmol) in 50 ml of CH2Cl2 was stirred at room temperature overnight. The solution was extracted with water (2 x 50 ml) and brine (1 x 50 ml), dried on MgSO4, filtered and evaporated. Purification of the residue

was

done

by

column

chromatography

with

hexane/ethyl

acetate/triethylamine (90:5:5) as eluent and yielded 1.3 g of 2.1 as a colourless oil (1.5 mmol, yield 37.5 %). 1

H-NMR (CDCl3): δ 1.3 (3H, t, OCH2CH3), δ 1.38 (9H, s, C(CH3)3), δ 2.27-2.58

(8H, m, NBocCH2CH2NCH2COOEt +2x SCH2CH2N), δ 2.90-3.04 (6H, m, CH2NCH2COOEt + CH2N-Boc), δ 4.1 (2H, q, OCH2CH3), δ 7.1-7.4 (30H, m, ArH). +

Mass: [M+H] 851 (calculated: 851)

52

Chapter 2

S,S'-Bis-triphenylmethyl-N-(BOC)-N’-(acetic acid)-1,2-ethylenedicysteamine (2.2)

To a dispersion of 2.1 (1.3 g, 1.5 mmol) in 10 ml water/EtOH (50:50, v/v) 1.5 ml of a saturated NaOH solution was added. The mixture was refluxed overnight. Ethanol was evaporated from the mixture under reduced pressure and 20 ml of water was added. Citric acid 0.1 M was then added to adjust the pH to 5. The water layer was extracted twice with dichloromethane (50 ml), the organic layer was isolated and washed with brine, dried on MgSO4, filtered and evaporated under reduced pressure. The residue was purified with column chromatography with a gradient starting with 100 % dichloromethane and then gradually adding methanol. The purified product 2.2 was obtained as a white foam, (700 mg, 0.75 mmol). Yield 50%. 1

H-NMR

(CDCl3):

δ

1.35

NBocCH2CH2NCH2COOH

(9H, +2x

s,

δ

C(CH3)3),

SCH2CH2N),

δ

2.25-2.32 2.81-2.97

(8H, (6H,

m, m,

CH2NCH2COOH + CH2NBoc), δ 7.16-7.42 (30H, m, ArH). +

Mass: [M+H] 823 (calculated: 823) Melting point (Mp): 70.9-72.3 °C

2.2.1.2. Synthesis

of

S,S'-bis-triphenylmethyl-N-(BOC)-N’-(2-(4’

acetamidophenyl)-1,3-benzothiazole)-1,2-ethylenedicysteamine

2-(4’-Aminophenyl)-1,3-benzothiazole (2.3)

2-Aminothiophenol (6.634 g, 53 mmol) and p-aminobenzoic acid (6.857 g, 50 mmol) were added to 80 g of PPA and the mixture was stirred at 180 °C for 4 h. The reaction mixture was allowed to cool down to room temperature and poured into a 10 % (m/v) Na2CO3 solution. The white precipitate formed was filtered off and dried in a vacuum oven. Crystallization from methanol yielded 7.24 g (32 mmol, 64 %) of 2.3 as light yellow crystals. 1

H-NMR (DMSO, 200 MHz): δ 5.8 (2H, s, NH2); δ 6.68 (2H, d, 3’-H 5’-H); δ

7.32 (1H, t, 6-H); δ 7.45 (1H, t, 5-H); δ 7.76 (2H, d, 2’-H 6’-H); δ 7.89 (1H, d, 4H); δ 7.98 (1H, d, 7-H).

Development of thioflavin-T derivatives for SPECT studies

53

-

Mass: [M-H] 225 (calculated: 225) Mp: 147.5-150 °C

S,S'-Bis-triphenylmethyl-N-(BOC)-N’-[2-(4’-acetamidophenyl)-1,3benzothiazole]-1,2-ethylenedicysteamine (2.4)

To a solution of 2.3 (68 mg, 0.3 mmol) in 5 ml dichloromethane, 1-(3dimethylaminopropyl)-3-ethyl carbodiimide hydrochloride (EDCI.HCl) (19 mg, 0.1 mmol), hydroxybenzotriazole (HOBT) (13.5 mg, 0.1 mmol) and 2.2 (82 mg, 88 µmol) were added and the mixture was stirred at room temperature overnight. The last three reagents were added for a second time in the same quantities and the mixture was stirred further over the weekend. The precipitate formed was filtered off and the organic solvent was evaporated under reduced pressure to yield 70 mg of 2.4 (68 µmol, 22.7 % yield). 1

H-NMR

(CDCl3):

δ

1.38

(9H,

s,

C(CH3)3),

δ

2.27-2.58

(8H,

m,

NBocCH2CH2NCH2COOH +2x SCH2CH2N), δ 2.97 (6H, m, CH2NCH2COO + CH2NBoc), δ 7.1-7.6 (36H, m, ArH + 2’-, 3’-, 5’-, 6’-, 4-, 5-H), δ 7.95 (1H, d, 6H), δ 8.05 (1H, d, 3-H) +

Mass: [M+H] : 1031 (calculated: 1031) Mp: 69-70.5 °C

2.2.2. Synthesis of BTA-BCL conjugates as precursors for negatively charged 99mTc-BTA derivatives

2.2.2.1. N-BOC-hydrazinonicotinyl-2-[2’-(3-aminopropoxy)-4’-aminophenyl]1,3-benzothiazole

N-BOC-3-amino-1-propanol (2.5)

A solution of 3-amino-1-propanol (7.7 ml, 100 mmol) in 150 ml of dichloromethane was cooled down to 0 °C and triethylamine (TEA) (15.3 ml, 110 mmol) was added dropwise to the mixture, followed by a solution of di-t-

54

Chapter 2

butyl dicarbonate (21.8 g, 100 mmol) in 50 ml of dichloromethane. The mixture was allowed to warm to room temperature and was stirred at room temperature overnight. The reaction mixture was then washed with water (3 x 100 ml) and the organic layer was dried on magnesium sulphate, filtered and evaporated to dryness under reduced pressure to yield 13.1 g (75 mmol, 75 %) of 2.5 as a colourless oil. 1

H-NMR (CDCl3): δ 1.45 (9H, s, tBu); δ 1.67 (2H, m, CH2-CH2-CH2); δ 3.24-

3.33 (3H, m, CH2-NH + OH); δ 3.66 (2H, m, CH2-OH); δ 4.87 (1H, s, NH-CO) -

Mass: [M-H] 174 (calculated: 174)

N-BOC-3-amino-1-propyl-p-tosylate (2.6)

To a solution of 2.5 (12.3 g, 70 mmol) in 40 ml of dichloromethane at 0 °C, TEA (10.8 ml, 77 mmol) was added dropwise followed by a solution of ptoluene sulphonylchloride (13.35 g, 70 mmol) in 25 ml of dichloromethane. The mixture was allowed to warm to room temperature and stirring was continued overnight. The organic layer was washed with water (3 x 100 ml), dried on magnesium sulphate, filtered and evaporated to dryness under reduced pressure to yield a colourless oil. This was purified by column chromatography with hexane/ethyl acetate (80:20) as the eluent to obtain 2.6 as a slightly yellow oil which turned into a white solid after drying under reduced pressure (7.5 g, 23 mmol, 32 %). 1

H-NMR (CDCl3): δ 1.42 (9H, s, tBu); δ 1.85 (2H, m, CH2-CH2-CH2); δ 2.46

(3H, s, CH3); δ 3.15 (2H, m, CH2-NH); δ 4.09 (2H, m, CH2-OS); δ 4.63 (1H, s, NH-CO); δ 7.36 (2H, d, 2-H 6-H); δ 7.78 (2H, d, 3-H 5-H) +

Mass: [M+H] 330 (calculated: 330) Mp: 62.0-63.2 °C

6-Hydrazinonicotinic acid (2.7)

A solution of 6-chloronicotinic acid (8 g, 50.8 mmol) in 80 % hydrazine hydrate (35 ml, 930 mmol) was stirred for 4 h at 100 °C. The homogeneous mixture was cooled and the solvent was removed by reduced pressure (with an oil

Development of thioflavin-T derivatives for SPECT studies

55

pump) at 20 Pa. The residue was taken up in water and the solution was acidified with concentrated hydrochloric acid to pH 5.5 after which a yellowish precipitate formed. The precipitate was filtered off and washed with ethanol (3x) and ether (2x) followed by drying under reduced pressure to yield 6.5 g (42.4 mmol, 83.6 %) of 2.7 as a yellowish powder. 1

H-NMR (DMSO): δ 5.10 (1H, br s, NH); δ 6.83 (1H, d, 5-H); δ 8.1 (1H, dd, 6-δ

8.66 (1H, d, 2-H) +

Mass: [M+H] 154 (calculated: 154) Mp: 285.7-286.4 °C

N-BOC-6-hydrazinonicotinic acid (2.8)

To a solution of 2.7 (4.6 g, 30 mmol) and TEA (5 ml, 36 mmol) in 100 ml of N,N-dimethylformamide (DMF) di-t-butyl dicarbonate (6.55 g, 30 mmol) was added and the mixture was stirred at room temperature overnight. The solvent was evaporated under reduced pressure at 20 Pa. Purification was done by dissolving the residue in ethyl acetate and pouring the solution over a short silica gel column. A polar impurity that was present (as seen on TLC) remained on the silica gel while 2.8 was eluted with ethyl acetate. The filtrate was evaporated under reduced pressure and the residue was dried under reduced pressure to yield 5.3 g of 2.8 (21 mmol, 70 %). 1

H-NMR (DMSO): δ 1.43 (9H, s, tBu); δ 6.55 (1H, d, 5-H); δ 7.99 (1H, dd, 4-

H); δ 8.60 (1H, d, 2-H); δ 8.92 (1H, s, Φ-NH); δ 9.01 (1H, s, NH-CO) -

Mass: [M-H] 252 (calculated: 252) Mp: 237.8-239 °C

Succinimidyl N-BOC-6-hydrazinopyridine-3-carboxylate (2.9)

A solution of dicyclohexylcarbodiimide (3.1 g, 15 mmol) in 15 ml DMF was added to a solution of 2.8 (3.8 g, 15 mmol) and N-hydroxysuccinimide (1.73 g, 15 mmol) in 50 ml of DMF. The mixture was stirred at room temperature for 20 h. The precipitate formed was then filtered off and DMF was removed under reduced pressure at 20 Pa. The residual yellow oil was dissolved in ethyl

56

Chapter 2

acetate and poured over silica gel as described for 2.8. A second purification was done using column chromatography with a mixture of CH2Cl2/MeOH (98:2) as the eluent and yielded 3.7 g of 2.9 (11 mmol, 73 % yield). 1

H-NMR (DMSO): δ 1.44 (9H, s, tBu); δ 2.88 (4H, s, OC-CH2-CH2-CO); δ 6.67

(1H, d, 5-H); δ 8.12 (1H, dd, 4-H); δ 8.75 (1H, d, 2-H); δ 9.20 (1H, s, Φ-NH); δ 9.45 (1H, s, NH-CO) -

Mass: [M-H] 349 (calculated: 349) Mp: 170.4-172 °C

2-(2’-Hydroxy-4’-aminophenyl)-1,3-benzothiazole (2.10)

A dispersion of 4-aminosalicylic acid (7.66 g, 50 mmol) and 2-aminothiophenol (6.4 g, 50 mmol) in 85 g of PPA was heated at 110 °C under mechanical stirring for 1 h. The mixture was allowed to cool down to room temperature and neutralized by adding a 10 % (m/v) Na2CO3 solution. A precipitate was formed which was filtered off and dried in a vacuum oven at 20 °C to yield 4 g (17 mmol, 33 %) of 2.10 as a slightly yellow solid. 1

H-NMR (DMSO): δ 5.97 (2H, s, NH2); δ 6.18 (1H, d, 3’-H); δ 6.26 (1H, dd, 5’-

H); δ 7.34 (1H, t, 5-H); δ 7. 45 (1H, t, 6-H); δ 7.63 (1H, d, 6’-H); δ 7.89 (1H, dd, 7-H); δ 8.03 (1H, dd, 4-H); δ 11.76 (1H; s, OH) -

Mass: [M-H] 241 (calculated: 241) Mp: 186.9-187 °C

2-[2’-(N-BOC-3-aminopropoxy)-4’-aminophenyl]-1,3-benzothiazole (2.11)

2.10 (16.5 mmol, 4 g) was dissolved in 50 ml of acetonitrile/methanol (9:1; v/v) by adding 50 mmol freshly prepared sodium methoxide (1.1 g of sodium dissolved in 12 ml of methanol). To this solution, N-BOC-3-amino-1-propyl-ptosylate (2.6) (11.86 g, 36 mmol) dissolved in 50 ml of acetonitrile/methanol (9:1; v/v) was added dropwise and the mixture was stirred overnight at 50 °C. After cooling down to room temperature, 150 ml of water was added and the mixture was extracted three times with dichloromethane (150 ml). The organic layer was washed with brine (150 ml), dried over MgSO4, filtered and

Development of thioflavin-T derivatives for SPECT studies

57

evaporated to dryness under reduced pressure. The residue was purified with column chromatography using a gradient mixture of CH2Cl2 and methanol (0 → 2 %) as the eluent to yield 3.3 g of 2.11 (8.3 mmol, 50 % yield). 1

H-NMR (CD3OD): δ 1.38 (9H, s, tBu); δ 2.17 (2H, m, CH2-CH2-CH2); δ 3.40

(2H, m, CH2-NH); δ 4.20 (2H, m, O-CH2); δ 5.92 (2H, s, NH2); δ 6.33 (1H, d, 5’-H); δ 6.35 (1H, d, 3’-H); δ 7.00 (1H, s, NH-CO); δ 7.30 (1H, t, 5-H); δ 7.44 (1H, t, 6-H); δ 7.88 (1H, d, 7-H); δ 7.98 (1H, d, 4-H); δ 8.13 (1H, d, 6’-H) +

Mass: [M+H] 400 (calculated: 400) Mp: 170.2-171.6 °C

2-[2’-(N-BOC-6-hydrazinonicotinamido-3-propoxy)-4’-amino]phenyl-1,3benzothiazole (2.12)

A solution of 2.11 (100 mg, 0.25 mmol) in a mixture of 19.5 ml trifluoroacetic acid (TFA) and 0.5 ml of triisopropylsilane was stirred at room temperature for 15 minutes. Trifluoroacetic acid was then removed under reduced pressure and the residue was co-evaporated four times with hexane. The residue was taken up in dichloromethane (50 ml). The solution was extracted twice with 50 ml of a saturated sodium carbonate solution, washed with brine, dried on MgSO4, filtered and evaporated under reduced pressure. The residue (100 mg, 0.334 mmol) was dissolved in 20 ml of dioxane and DIEA (60 µl, 100 mmol) and succinimidyl-6-BOC-hydrazinopyridine-3-carboxylate (2.9) (235 mg, 0.67 mmol) were added. The mixture was stirred overnight at room temperature. The organic solvent was evaporated and the residue was purified by column chromatography with 5 % methanol in CH2Cl2 as the eluent to yield 80 mg of a green-yellow solid (0.15 mmol, 44.9 %). 1

H-NMR (CD3OD): δ 1.47 (9H, s, tBu); δ 2.28 (2H, m, CH2-CH2-CH2); δ 3.71

(2H, m, CH2-NH); δ 4.22 (2H, m, O-CH2); δ 6.37 (1H, d, 5-HPh); δ 6.40 (1H, d, 3-HPh); δ 6.65 (1H, d, 5-HNIC); δ 6.92 (1H, s, NH-CO); δ 7.28 (1H, t, 5-HBe); δ 7.42 (1H, t, 6-HBe); δ 7.87 (2H, d, 4-HBe 7-HBe); δ 7.95 (1H, dd, 4-HNic); δ 8.09 (1H, dd, 6-HPh); δ 8.55 (1H, d, 2-HNIC) +

Mass: [M+H] 535 (calculated: 535) Mp: 147.4-149.5 °C

58

Chapter 2

2.2.2.2. Synthesis of 2-[4’-(S-benzyl-mercaptoacetyl-L-asp(tBu)glycinamido)phenyl]-1,3-benzothiazole (2.13)

To a solution of S-benzyl mercaptoacetyl-L-asp(tBu)-gly (410 mg, 1 mmol) in 10 ml of acetonitrile 2.3 (226 mg, 1 mmol) was added together with HOBT (141 mg, 1 mmol) and EDCI.HCl (198 mg, 1 mmol). The reaction mixture was stirred at room temperature overnight and the precipitate formed was filtered off and washed with acetonitrile. The collected acetonitrile fractions were evaporated. The residual yellow oil was dissolved in dichloromethane and the solution was washed with a saturated sodium bicarbonate solution and with brine, dried over magnesium sulphate, filtered and evaporated to dryness. The residue was purified with column chromatography with dichloromethane as eluent and increasing concentrations of methanol (0 – 5 %). The collected fractions of the product were evaporated to dryness and the residue was redissolved in ether. The precipitate formed was filtered off and dried under reduced pressure to yield 80 mg of 2.13 (0.13 mmol, 13 %). 1

H-NMR (CDCl3): δ 1.46 (9H, s, tBu); δ 2.69 and δ 2.88 (2H, 2x dd, CH2-

COOtBu); δ 3.20 (2H, s, CO-CH2-SBz); δ 3.76 (2H, s, S-CH2-Bz); δ 4.01 and δ 4.18 (2H, 2 x dd, CO-CH2-NH); δ 4.65 (1H, m, CO-CH); δ 7.04 (1H, t, CH2-NHCO); δ 7.28 (5-H, s, C6H5-CH2); δ 7.36 (1H, t, 5-H); δ 7.48 (1H, t, 6-H); δ 7.57 (1H, d, CHNHCO); δ 7.80 (2H, d, 2’-H 6’-H); δ 7.87 (1H, d, 7-H); δ 8.01 (2H, d, 3’-H 5’-H); δ 8.04 (1H, d, 4-H); δ 8.72 (1H, s, Φ-NH-CO) +

Mass: [M+H] 619 (calculated: 619) Mp: 141.1-143.6 °C

2.2.2.3. Synthesis of 2-[4’-(N,N-diacetic acid)-N-acetamidophenyl]-1,3benzothiazole (2.14)

A solution of nitrilotriacetic acid (129 mg, 0.67 mmol) in 1.75 ml pyridine was heated at 50 °C during 10 minutes. Acetic acid anhydride (0.12 ml, 1.2 mmol) was added and the mixture was heated further at 100 °C for 30 minutes. The mixture was allowed to cool down to 50 °C and 2.3 (226 mg, 1 mmol) was added. The mixture was heated again at 100 °C for 1 h and, after cooling

Development of thioflavin-T derivatives for SPECT studies

59

down, pyridine was evaporated under reduced pressure at 20 Pa. The pH was adjusted to 9.5 by the addition of 2 M ammonium hydroxide. The mixture was extracted three times with ether and the organic layer was discarded. Upon acidification of the water layer to pH 2 with 6 M HCl, a precipitate was formed which was filtered off and crystallized from ethanol/water (50:50, v/v) to yield 50 mg of 2.14 (0.13 mmol, 19.4 %). 1

H-NMR (DMSO): δ 3.51 and 3.55 (6H, 3 x s, 3 x CH2); δ 7.42 (1H, t, 5-H); δ

7.53 (1H, t, 6-H); δ 7.69 (1H, s, NHCO); δ 7.86 (2H, d, 3’-H 5’-H); δ 8.02 (2H, d, 2’-H 6’-H); δ 8.10 (2H, d, 4-H 7-H); δ 10.92 (1H, s, COOH) -

Mass: [M-H] 398 (calculated: 398) Mp: 183.9-185 °C

2.2.3. Deprotection and labelling with 99mTc

2.2.3.1. Deprotection and labelling of 2.4

A stock solution of 2.4 (1 mg/ml in acetonitrile) was prepared together with a labelling buffer solution consisting of a mixture of 5 ml 0.5 M phosphate buffer pH 7, 2.5 ml 0.1 M ethylenediaminetetraacetic acid disodium salt (Na2EDTA) and 2.5 ml NaKtartrate solution (40 mg/ml). A mixture of 200 µl of the stock solution and 50 µl of 0.5 M HCl in a labelling vial was heated in a boiling water bath for 15 min. After cooling down, 15 µl SnCl2.2H2O solution (20 mg/5 ml 0.05 M HCl), 200 µl of labelling buffer solution and 1 ml generator eluate containing 400 to 700 MBq of

99m

TcO4- were added. The mixture was heated

again in a boiling water bath for 15 min. After cooling down to room temperature, RP-HPLC analysis was performed.

2.2.3.2. Two-step procedure for deprotection and labelling of 2.12

Deprotection: 6 mg of 2.12 was dissolved in 2 ml of dioxane and 2 ml of HCl/dioxane was added (prepared by bubbling HCl gas at a moderate rate through dioxane for about 10 min). The mixture was stirred at room temperature for about 4 hours after which the precipitate was filtered off and

60

Chapter 2

washed several times with ether. The precipitate (the HCl salt of deprotected 2.12) was used for labelling without further purification. Labelling with 99mTc:

A (with tricine as the co-ligand): to a labelling vial

were successively added 100 µl of a 1 mg/ml solution of deprotected 2.12 (HCl salt) in water/EtOH (50:50, v/v), 200 µl of 0.5 M phosphate buffer pH 7, 15 µl of a 100 mg/ml solution of tricine in water, 12.5 µl of a 4 mg/ml solution of SnCl2.2H2O in 0.05 M HCl and 740 MBq (20 mCi) 99mTcO4- in 1 ml of saline. The mixture was incubated at room temperature for 15 min. B (with EDDA as the co-ligand): performed in the same way as A, but 5 mg of EDDA was used instead of 1.5 mg tricine. C (with tricine/nicotinic acid as the co-ligand): also performed in the same manner as A, but 15 mg tricine and 2 mg nicotinic acid was used and the mixture was heated in a boiling water bath for 15 min.

2.2.3.3. Deprotection and labelling of 2.13

In a labelling vial were successively mixed 100 µl of a 1 mg/ml solution of 2.13 in acetonitrile, 500 µl 0.5 M phosphate buffer pH 9, 250 µl of a 40 mg/ml solution of NaKtartrate in water, 25 µl of a 4 mg/ml solution of SnCl2.2H2O in 0.05 M HCl and 740 MBq 99mTcO4- in 1 ml saline. The mixture was heated in a boiling water bath for 15 min and RP-HPLC was performed after cooling down to room temperature. 2.2.3.4. Labelling of 2.14 with 99mTc [99mTc(I)(CO)3(H2O)3]+ was prepared either by using an Isolink™ kit (Mallinckrodt Medical B.V., Pettten, The Netherlands) or by the CO bubbling method.

When using an Isolink™ kit, 740 MBq (20 mCi) of

99m

TcO4- in a

maximum of 1 ml saline was added to a labelling kit, which was heated in a boiling water bath for 20 minutes. The pH of the reaction mixture was adjusted to 7 with 1 M HCl and 250 µl of this neutralized solution was added to 100 µl

Development of thioflavin-T derivatives for SPECT studies

61

of a 1 mg/ml solution of 2.14 in acetonitrile/MeOH/H2O (3:3:4, v/v)). The mixture was heated at 70 °C for 20 minutes and RP-HPLC analysis was performed after cooling down to room temperature. When [99mTc(I)(CO)3(H2O)3]+ was prepared by the CO bubbling method, a three step procedure was necessary. First CO gas was bubbled through a mixture containing 4.5 mg Na2CO3, 20 mg to 40 mg NaBH4 and 20 mg NaKtartrate. Then 740 MBq of

99m

TcO4- in a maximum of 1 ml saline was

added and the mixture was heated at 75 °C for 20 min. Finally, the pH of the tricarbonyl mixture was adjusted to pH 7 by the addition of approximately 250 µl of a 1 M HCl solution and 250 µl of this mixture was added to 100 µl of a 1 mg/ml solution of 2.14 in acetonitrile/MeOH/H2O (3:3:4, v/v). The mixture was heated at 70 °C for 20 min and RP-HPLC analysis was performed after cooling down.

2.2.4. Analysis of 99mTc-labelled reaction mixtures with RP-HPLC

RP-HPLC analysis of the labelled compounds was carried out using an XTterra™ RP C18 column (5 µm, 250 mm x 4.6 mm). A part of the reaction mixture (25 to 250 µl) was applied to the column that was eluted with gradient mixtures of acetonitrile and 0.05 M ammonium acetate at a flow rate of 1 ml/min. RP-HPLC analysis of 99mTc-labelled 2.4, 2.13 and 2.14 was performed with gradient system 1 and RP-HPLC analysis of system 2 (see Table 4).

99m

Tc-2.12 with gradient

62

Chapter 2

Table 4. Linear gradient elution systems for RP-HPLC analysis with A = acetonitrile and B = 0.05 M ammonium acetate % (v/v)

System 1 Time (min)

A

B

0

0

100

20

90

10

25

90

10

30

0

100

% (v/v)

System 2 Time (min)

A

B

0

0

100

15

90

10

20

90

10

25

0

100

2.2.5. Analysis of 99mTc-labelled compounds with radio-LC-MS

Due to the very low amounts of technetium-99m in a radioactive preparation (typically in the nano- to picomolar range) it is difficult to obtain good mass spectra of technetium-99m labelled compounds in no-carrier added form (nocarrier added means that the radionuclide has not been doped on purpose with one of its stable isotopes, but these can be present). To increase the amount of technetium in the radiolabelled preparation, technetium-99 is added and thus the carrier-added form is used. Technetium-99 emits β--rays and has a half-life of 2.14 x 105 year, so care should be taken when

99

Tc is used to

prevent contamination. Gradient system 1 (see 2.2.4) was used for radio-LCMS analysis of 99mTc-2.4, 99mTc-2.13 and 99mTc-2.14, while an isocratic elution system consisting of 0.05 M ammonium acetate/acetonitrile (75:25, v/v) was used for radio-LC-MS analysis of

99m

Tc-tricine-2.12.

Development of thioflavin-T derivatives for SPECT studies

63

2.2.5.1. Labelling of 2.4, 2.12 and 2.13 for LC-MS analysis

Deprotection and labelling of 2.4, 2.12 and 2.13 was performed as described in 2.2.3 but the solution of 99mTcO4- was spiked with 0.1 ml of a solution of 150 µg NH499TcO4 in 10 ml of water (1.5 µg Tc). In addition, 30 µl, 25 µl and 50 µl of the 4 mg/ml solution of SnCl2.2H2O in 0.05 M HCl was used for 2.4, 2.12 and 2.13, respectively. Mass determination was only carried out with tricine as co-ligand for 2.12.

2.2.5.2. Labelling of 2.14 for LC-MS analysis

To perform a carrier added labelling, the amount of reducing agent in the Isolink™ kit is too low and this cannot be adjusted. To obtain mass data of a 99m

Tc-tricarbonyl labelled compound, it is therefore necessary to perform the

preparation of the

99m

Tc(CO)3(OH2)3 precursor via the CO bubbling method.

The amount of NaBH4 has to be increased because of the high amount of technetium-(99m+99) in the reaction mixture. So, together with the solution of 400 to 700 MBq

99m

TcO4- in 0.9 ml saline, 0.1 ml of a solution of 150 µg

NH499TcO4 in 10 ml of water was added (1.5 µg). The rest of the procedure was followed as described in 2.2.3.4.

2.2.6. Partition coefficient determination

The lipophilicity of the RP-HPLC isolated

99m

Tc-complexes was determined

using a modification of the method described by Yamauchi and co-workers180. To a test tube containing 2 ml of 1-octanol and 2 ml of 0.025 M phosphate buffer pH 7.4, 25 µl of the RP-HPLC isolated

99m

Tc-complex was added. The

test tube was vortexed at room temperature for 3 min and then centrifuged for 10 min at 2700 g. A 50-µl aliquot was taken of the octanol phase and a 500-µl aliquot of the buffer phase, avoiding cross-contamination between the two phases. The activity in each aliquot was counted using a gamma counter and the partition coefficient P was calculated using the following equation:

64

P=

Chapter 2

cpm / ml octanol cpm / ml buffer

with cpm = counts per minute

Experiments were performed in triplicate.

2.3. Results and discussion Thioflavin-T is a rather polar, hydrophilic compound due to the presence of a quaternary amine. Research over the past years has revealed that the quaternary amine is not essential for binding to amyloid β, since uncharged TT derivatives display an even higher affinity for amyloid β plaques than TT itself136,144. Thioflavin-T has affinity for the β-pleated sheet structure of the amyloid plaques and therefore radiolabelled derivatives may be very useful for in vivo detection of amyloid plaques in the human body in a non-invasive way. Such derivatives should be neutral and lipophilic to allow diagnosis of AD, since they have to penetrate the BBB. More polar derivatives are required as tracer agents for diagnosis of PA, since such agents should be cleared fairly rapidly from the abdominal organs and should preferably be excreted with the urine. In this respect, TT is a good starting molecule from which neutral as well as charged derivatives can be synthesized and tested.

2.3.1. Synthesis

2.3.1.1. Synthesis of 2.4, 2.12, 2.13 and 2.14

Synthesis of 2.4

Synthesis of the BAT type BCL S,S'-bis-triphenylmethyl-N-(BOC)-1,2ethylenedicysteamine was done according to published data181. The thiols were protected with trityl groups to prevent oxidation (e.g. disulfide formation) and one amine was protected with a t-butoxycarbonyl group to allow introduction of a single ethyl acetate substituent on one of the amines. The reaction with ethyl bromoacetate was done in dichloromethane with DIEA as base and provided 2.1 in 37.5 % yield. In the next step, the ethyl ester was

Development of thioflavin-T derivatives for SPECT studies

65

removed with NaOH in water/EtOH under reflux, yielding the free acid 2.2 (50 % yield) (Figure 10).

N S

NH3 (liq)

HN

NH

S

S

BOC

tritol TFA

HN

N

S

S

S

Tr

Tr

Tr

HN

NH

S Tr

(BOC)2O

Na

CH2Cl2 DIEA BrCH2COOEt

O

O

BOC HO

Tr = C(C6H5)3 TFA = trifluoroacetic acid DIEA = N,N-diisopropylethylamine

N

N

S

S

Tr

Tr

RT

H2O/EtOH

BOC N

EtO

NaOH 80 °C

2.1

2.2

N

S

S

Tr

Tr

Figure 10. Synthesis of the bifunctional chelating ligand of the BAT type

The

2-(4’-aminophenyl)-1,3-benzothiazole

2.3

was

according to a method described by Shi and co-workers

synthesized 138

. Thus, 2-

aminothiophenol and p-aminobenzoic acid were coupled in PPA at 180 °C during 4 h and this provided 2.3 in 64 % yield (Figure 11) after crystallisation.

NH2

+ SH

HOOC

NH2

PPA 180 °C, 4 h

N NH2 S 2.3

Figure 11. Synthesis of 2-(4’-aminophenyl)-1,3-benzothiazole

Coupling of the aromatic amine of the benzothiazole with the carboxylic acid group of the BAT ligand was performed with EDCI.HCl and HOBT, classical coupling reagents used in peptide chemistry. Although the reagents were added in excess and a long reaction time was used, only a moderate yield of 23 % was obtained (Figure 12). No upscaling of the reaction was done because only low amounts are required for labelling with technetium-99m.

66

Chapter 2

BOC

HOOC N

N NH2

+

S 2.3

S

S

Tr

Tr

2.2 EDCI HOBT CH2Cl2 Room Temperature

EDCI = 1-(3-dimethylaminopropyl)3-ethyl carbodiimide hydrochloride HOBT = hydroxybenzotriazole

O

O

N

N

NH S

2.4

N

N

S

S

Tr

Tr

C O

CH3 CH3 CH3

Figure 12. Coupling of the benzothiazole with the BAT ligand

Synthesis

of

2-[2’-(N-BOC-6-hydrazinonicotinamido-3-propoxy)-4’-

amino]phenyl-1,3-benzothiazole (2.12)

Synthesis of the spacer N-BOC-amino-1-propyl-p-tosylate was done in two steps. First, the amine of 3-amino-1-propanol was protected with a BOC group by

reaction

of

di-t-butyldicarbonate

and

of

3-amino-1-propanol

in

dichloromethane in equimolar amounts. This provided 2.5 in a good yield (75 %). Then, the hydroxyl group was reacted with p-toluenesulphonyl chloride, which gave N-BOC-amino-1-propyl-p-tosylate (2.6, 32 %) (Figure 13).

Development of thioflavin-T derivatives for SPECT studies

CH3

H2N

(CH2) 3

+

OH

H3C

O

O

67

O

CH3

O

O

CH3

CH3

CH3

TEA CH2Cl2

CH3 H3C

O

O

HN

(CH2) 3

OH

2.5

CH3

O TEA CH2Cl2

Cl S

CH3

O CH3 H3C

O

O

O HN

(CH2) 3

CH3

O

S

CH3

O

2.6

Figure 13. Synthesis of the HYNIC-benzothiazole-spacer

The

HYNIC

ligand

(N-BOC-hydrazinonicotinic

acid

N-hydroxy

succinimide ester) was synthesized according to published data176 (Figure 14).

O

85% NH2NH2

O

HO

NH

N

HO

NHS = N-hydroxysuccinimide DCC = N,N'-dicyclohexylcarbodiimide O

NHBOC NH

HO

N 2.8

(tBuOCO)2O

NH2

Cl

TEA, DMF

N 2.7 O N O

NHS DCC, DMF

NHBOC NH

O

O

N

2.9

Figure 14. Synthesis of the succinimidyl activated HYNIC ligand

2-(2’-Hydroxy-4’-aminophenyl)-1,3-benzothiazole

(2.10)

was

synthesized following the same route as 2-(4’-aminophenyl)-1,3-benzothiazole (2.3)138. The temperature was kept at 110 °C but this gave low yields (33 %) due to the high viscosity of polyphosphoric acid at this temperature. The yield of the synthesis can probably be increased by increasing the temperature

68

Chapter 2

during the reaction. Coupling of the benzothiazole to the N-BOC-3-amino-1propyltosylate spacer was done overnight in an acetonitrile/MeOH (9:1, v/v) mixture with sodium methoxide as base at 50 °C. This provided 2.11 in 50 % yield. To allow coupling of the HYNIC ligand to 2.11, the BOC protective group on the amine of the spacer was removed. Initially, this was attempted with a mixture of TFA, triisopropylsilane and a few drops of water, but analysis of the reaction product by RP-HPLC revealed that compound 2.11 was not stable in these circumstances. Removal of this N-BOC in the absence of water was more successful and resulted in a complete deprotection of the amine group of the spacer. This deprotected amine was coupled to the HYNIC ligand in the last step. Coupling occurred in dioxane with DIEA as base and 2 equivalents of the HYNIC ligand. The reaction mixture was kept overnight at room temperature and yielded the desired product 2.12 (45 %) (Figure 15).

N NH2

CH3CN/MeOH (9:1) NaOMe

N NH2 S

2.6

S

O

HO

O NH

2.10

O

2.11

CF3COOH TRIS, RT

TRIS = triisopropylsilane RT = room temperature

N

N

NH2

NH2 S O

NHBOC

O NH NH

N

2.12

2.9

S O

Dioxane DIEA RT overnight

NH2

Figure 15. Synthesis of 2-[2’-(N-BOC-6-hydrazinonicotinamido-3-propoxy)-4’amino]phenyl-1,3-benzothiazole (2.12)

Although there is also a free aromatic amine in 2.11, coupling of the HYNIC ligand with this amine was not observed. This is probably due to the much lower nucleophilic character of this aromatic amine in alkaline conditions.

Development of thioflavin-T derivatives for SPECT studies

69

Synthesis of 2-[4’-(S-benzyl-mercaptoacetyl-L-asp(tBu)-glycinamido)phenyl]1,3-benzothiazole (2.13) The BCL of this compound was synthesized according to published data175 and was coupled to benzothiazole 2.3 in the same manner as described in 2.3.1.1 with EDCI and HOBT to yield 2.13 (Figure 16).

O O

O O

NH

N

HN H2N

+

S

HO

S

2.3

O

Acetonitrile HOBT EDCI.HCl room temperature O EDCI = 1-(3-dimethylaminopropyl)3-ethyl carbodiimide hydrochloride HOBT = hydroxybenzotriazole

O NH O

N

NH

O HN

O

S

S 2.13

Figure 16. Synthesis of 2.13

Synthesis of 2-[4’-(N,N-diacetic acid)-N-acetamidophenyl]-1,3-benzothiazole (2.14)

The reaction was carried out starting from nitrilotriacetic acid anhydride that was formed in situ by mixing acetic acid anhydride and nitrilotriacetic acid in pyridine. After formation of nitrilotriacetic acid anhydride, the benzothiazole 2.3 was added and the mixture was heated at 100 °C for 1 h and yielded compound 2.14 after purification (Figure 17).

70

Chapter 2 O HOOC N N

O O

NH2

NH Pyridine 1h at 100 °C

S

2.3

COOH

O

N

CH2 N

S

COOH

2.14

Figure 17. Synthesis of 2.14

As TT does not contain sufficient metal binding donor atoms in its structure

to

bind

Tc,

we

have

attached

several

BCL’s

to

2-

phenylbenzothiazoles to obtain neutral (99mTc-2.4) and charged (99mTc-tricine2.12,

99m

Tc-2.13,

99m

Tc-2.14)

99m

Tc-labelled TT derivatives. In the case of 2.4

and 2.13 the BCL’s contain two thiol groups, which have to be protected during synthesis to avoid oxidation. Protection of these thiol groups is often carried out using trityl (triphenylmethyl) or benzyl groups. Both are effective protection groups, but deprotection is easier for trityl groups than for benzyl groups. The latter requires rather harsh conditions (sodium in liquid ammonia) to be removed, whereas the trityl group was found to be sufficiently removed by heating in the presence of 0.5 M HCl. In the case of 2.12 and 2.14, other BCL’s were used instead of tetraligands. The hydrazine in 2.12 was protected with a BOC protection group, while 2.14 did not carry any protection groups.

Three of the BTA-BCL conjugates described in this chapter were synthesized starting from the same benzothiazole (2.4, 2.13 and 2.14) by formation of an amide between the free aniline group of the BTA and a carboxyl group of the BCL. In literature, it has been reported that in rats 99mTcBCL-complexes containing amides, pass the BBB less easily than the corresponding complexes in which all amides were substituted by amines182. Therefore, a bis-aminoethanethiol BCL was chosen for 2.4, the technetium complex of which was intended to visualize amyloid β in the brain of AD patients. The other technetium labelled compounds were designed as potential tracer agents for diagnosis of PA. In the case of 2.12, we chose to couple the HYNIC molecule to the 2’ position of the BTA with the aid of a propoxy spacer.

99m

Tc-complexes labelled via a HYNIC BCL have been

Development of thioflavin-T derivatives for SPECT studies

71

reported to be relatively hydrophilic, especially in the presence of a polar coligand, because the hydrazine group as such is not sufficient to efficiently bind technetium. In this study, tricine, EDDA and a mixture of tricine and nicotinic acid have been used as different co-ligands for the labelling of 2.12 with 99m

Tc.

2.3.2. Deprotection, labelling with 99mTc and RP-HPLC analysis

2.3.2.1. One-pot deprotection and labelling of 2.4

In precursor 2.4, the two thiol groups are protected with a trityl group and the secondary amine is protected with a BOC group. Removal of thiol protective trityl groups is normally done with trifluoroacetic acid and triisopropylsilane in an inert solvent. This procedure also removes an N-BOC protective group and allows performing deprotection and labelling with

99m

Tc as a one-pot reaction.

Based on previous work183, however, the method consisting of heating in the presence of hydrochloric acid was preferred. This provides a low amount of the precursor being completely deprotected, while the major part is converted to the mono-trityl mono-BOC form or ditrityl form as shown by mass spectrometry (MS) analysis. Such partial deprotection appears to be sufficient to allow successful exchange labelling of the BAT-BCL with

99m

Tc upon a

subsequent labelling step by heating in a boiling water bath in the presence of stannous ions and tartrate as the weak chelating agent (the formation of an intermediate weak Tc(V)O-tartrate complex prevents the formation of colloidal 99m

Tc). This can be explained by the fact that only nanomolar amounts of Tc

are present in the activities used (400-800 MBq) in the labelling reaction mixture. Apparently, the complex forming force pushes the remaining protection groups out of the molecule, a process that is further aided by continuing heating which produces more deprotected 2.4. Yield of the labelling reaction was 56 % and RP-HPLC analysis showed good separation of the different products present in the labelling reaction mixture (Figure 18).

72

Chapter 2

TOM : 25 0 2 20 04 .R 03 - P ul s e 1

CPS 1. 2 e5

99mTc-BAT-CH 8. 0 e4

99mTc-2.4

2-COOH

0

O

N NH S

N O N Tc

4. 0 e4

99m

Tc-2.4

0. 0 e0 0' 0 0

5' 0 0

10 ' 00

15 ' 00

20 ' 00

25 ' 00

S

30 ' 00

S

m

Figure 18. RP-HPLC chromatogram of the reaction mixture after labelling of 2.4 with

99m

Tc.

99m

Tc-2.4 elutes at 23 min and

99m

Tc-BAT-CH2-COOH at

13’23”. Only the radiometric signal is shown; cps = counts per second, m = minutes. Gradient mixture 1 was used.

Identification of the peak at 13’23’’ was done by comparing its retention time (Rt) with the Rt of 99m

99m

Tc-labelled-2.2. The rather high percentage of

Tc-BAT-CH2-COOH as radiochemical impurity is due to the presence of a

low amount of BAT-CH2-COOH present as an impurity in 2.4, and the fact that this small molecule probably labels more easily than 2.4. The yield of this labelling reaction was, however, sufficient for further biological evaluation.

2.3.2.2. Two-step procedure for deprotection and labelling of 2.12

Deprotection of the N-BOC protected HYNIC ligand was performed with an acidic dioxane solution following a described procedure176, which yielded the HCl salt of deprotected 2.12. This was used for labelling with different coligands. Labelling of HYNIC derivatized compounds with

99m

Tc, requires the

presence of a co-ligand with metal binding donor atoms, as the HYNIC moiety as such does not provide a sufficient number of donor atoms for the formation of a stable complex with

99m

Tc. With tricine as the co-ligand, this yielded 70 %

of 99mTc-tricine-2.12 (Figure 19 A). Gradient system 2 was used.

Development of thioflavin-T derivatives for SPECT studies

73

TOV67 : B IOD I S TR . R0 2- Pu ls e 1

CPS

30 00 0

99mTc-tricine-2.12

A

20 00 0

10 00 0

0 0' 0 0

5' 0 0

10 ' 00

15 ' 00

20 ' 00

25 ' 00

m

TOV67 : TE S TE N .R 0 3- P u ls e 1

CPS 60 00 0

99mTc-tricine/nicotinic

B

acid-2.12

40 00 0

20 00 0

0 0' 0 0

5' 0 0

10 ' 00

15 ' 00

20 ' 00

25 ' 00

m

TOV67 : TE S TE N .R 0 4- P u ls e 1

CPS

C

16 00 0

99mTc-EDDA-2.12 12 00 0

80 00

40 00

0 0' 0 0

5' 0 0

10 ' 00

15 ' 00

20 ' 00

25 ' 00

m

Figure 19. RP-HPLC chromatograms of the reaction products after labelling of 2.12 with

99m

Tc in the presence of tricine (A), tricine/nicotinic acid (B) or

EDDA (C) as co-ligand.

Yield of the labelling reaction in the presence of a mixture of tricine and nicotinic acid was 77 % (Figure 19 B) and in the presence of EDDA it was 52 % (Figure 19 C). From these RP-HPLC chromatograms it is also clear that 99m

Tc-EDDA-2.12 (Rt approximately 15 min) is more lipophilic than

99m

Tc-

tricine/nicotinic acid-2.12 (Rt approximately 13 min). As the goal was to obtain a more hydrophilic complex, we did not focus further on the technetium complex using EDDA as co-ligand.

74

Chapter 2 With tricine as a co-ligand, the resulting Tc-complex

99m

Tc-tricine-2.12

had a Rt on RP-HPLC, which is about the same Rt as observed for the technetium complex obtained when a combination of tricine and nicotinic acid was used as the co-ligand. Although the labelling yield using tricine/nicotinic acid as co-ligand was somewhat higher than with tricine alone (77 % versus 70 %), the labelling procedure is more complex and requires a boiling step. Therefore, the latter technetium complex was also not used for further evaluation.

2.3.2.3. One-step deprotection and labelling of 2.13 Deprotection of 2.13 in a mixture of dichloromethane/TFA (5:1, v/v) at room temperature, as described in literature175, only provided low yields of the desired Tc complex. However, it has been shown that S-protected mercaptotriamides can efficiently be labelled with

99m

Tc via an exchange

labelling reaction, without prior removal of the S-protective group. Apparently, the metal complexation driving force, together with continuing heating, induces the deprotection of the thiol during the complexation reaction. Labelling of S-benzyl protected 2.13 with technetium-99m was indeed performed efficiently by exchange labelling at pH 9 (in the same way as for 2.4, but at a different pH value). Labelling yields were the highest when deprotection and labelling were performed at pH 9 can be seen from Figure 20.

At pH 11 or 12 the labelling yield dropped again to 50 % or less. Therefore, labelling was performed at pH 9 and gradient system 1 was used for RP-HPLC purification.

Development of thioflavin-T derivatives for SPECT studies

75

pH 7 (7 % yield)

TOM : JC3 1 1 1 2 0 . R 0 3 - P ul s e 1

CPS

20 00 0

99mTc-2.13 10 00 0

0 0' 0 0

5' 0 0

10 ' 00

15 ' 00

20 ' 00

25 ' 00

30 ' 00

m

TOM : JC3 1 1 1 2 1 . R 0 1 - P ul s e 1

pH 8 (48 % yield)

CPS 16 00 0

99mTc-2.13 12 00 0

80 00

40 00

0 0' 0 0

5' 0 0

10 ' 00

15 ' 00

20 ' 00

25 ' 00

30 ' 00

m

25 ' 00

30 ' 00

m

25 ' 00

30 ' 00

m

pH 9 (80 % yield)

TOM : JC3 1 1 1 2 1 . R 0 5 - P ul s e 1

CPS 12 00 0

99mTc-2.13 80 00

40 00

0 0' 0 0

5' 0 0

10 ' 00

15 ' 00

20 ' 00

pH 10 (77 % yield)

TOM : JC3 1 1 1 2 1 . R 0 3 - P ul s e 1

CPS

99mTc-2.13

20 00 0

10 00 0

0 0' 0 0

5' 0 0

10 ' 00

15 ' 00

20 ' 00

Figure 20. HPLC chromatograms of the reaction mixtures after labelling of 2.13 with technetium-99m at different pH values 2.3.2.4. Labelling of 2.14 with a 99mTc-tricarbonyl (99mTc(CO)3) core Labelling of a ligand with a

99m

Tc(CO)3 core can be done in two ways: starting

from an Isolink™ labelling kit or using the CO-bubbling method. Both methods yield [99mTc(CO)3 (OH2)3]+, which is reacted with a ligand that has in its structure preferentially three suitable metal binding donor atoms (such as an aromatic nitrogen or amine, a carboxylate oxygen) separated by 2 or 3 atoms. Such donor atoms can easily replace the three loosely bound water molecules

76

Chapter 2

of the [99mTc(CO)3 (OH2)3]+ precursor, resulting in the formation of two 5- or 6rings. An example of a suitable ligand for formation of a Tc-tricarbonyl complex is IDA and its derivatives. The amine nitrogen atom and two carboxylate oxygen atoms function as donor atoms and their binding results in a negatively charged Tc-tricarbonyl-IDA complex.

In the CO-bubbling method CO gas is bubbled through a mixture of Na2CO3, NaBH4 and NaKtartrate. Sodium carbonate is used to obtain an alkaline pH, NaBH4 is a reducing agent, and NaKtartrate is added to stabilize the intermediate complexes formed during the reaction. When pertechnetate is added in a second step and the mixture is heated at 75 °C for 20 minutes, the [99mTc(CO)3(OH2)3]+ precursor is formed. The alkaline pH is necessary to stabilize the reaction mixture over a longer time-period. In a third step, the tricarbonyl complex is added to a solution of the ligand molecule and this mixture is again heated at 70 °C for 20 minutes. Depending on the ligand used it may be necessary to adjust the pH of the [99mTc(CO)3(OH2)3]+ precursor before adding it to the ligand. The more recently introduced Isolink™ kit contains sodium boranocarbonate as the CO source and the [99mTc(CO)3(OH2)3]+

precursor

can

be

prepared

by

simply

adding

pertechnetate to the kit and heating it for 20 minutes in a boiling water bath. A good example of a ligand for this [99mTc(CO)3]+ core is nitrilotriacetic acid, which has a functional group that can be bound to the bioactive compound so that two acetic acid groups remain to bind the [99mTc(CO)3]+ core together with the central amine. This also yields a charged complex that can be used for visualization of PA. Labelling of 2.14, starting from the Isolink™ kit or using the CObubbling method, yielded a Tc-complex with identical retention time on RPHPLC. Because of the alkaline pH of the [99mTc(CO)3(OH2)3]+ solution, it was necessary to adjust its pH to 7 with 1 M HCl before the BTA-precursor was added to the radioactive solution. At pH values above 8, the yield of the labelling reaction was very low. It appeared necessary to use a neutralized [99mTc(CO)3(OH2)3]+ precursor solution because of degradation of the 2.14

Development of thioflavin-T derivatives for SPECT studies

77

ligand at higher pH values. Yield of the labelling reaction at pH 7 was 64 % (Figure 21).

TOM : 1 6 0 1 2 0 0 4 . R 0 3 - P ul s e 1

CPS 1.6 e5

99mTc(CO)

1.2 e5

3-2.14

8.0 e4

4.0 e4

0.0 e0 0' 0 0

5' 0 0

10 ' 00

15 ' 00

20 ' 00

25 ' 00

30 ' 00

m

Figure 21. RP-HPLC chromatogram of the labelling reaction mixture after heating of 2.14 with a [99mTc(CO)3(OH2)3]+ solution at pH 7

2.3.3. Radio-LC-MS analysis of the labelled compounds

The use of liquid chromatography with UV- and radiometric detection in combination with mass spectrometry, allows to determine which mass corresponds to a radioactive product found in the radiometric channel. RadioLC-MS is known for its high sensitivity (detection up to nanomoles), especially a time-of-flight mass spectrometer. However, it remains a challenge to reliably determine the molecular ion-mass of a no-carrier-added radiopharmaceutical, since the mass amount of a radiolabelled product in a classical

99m

Tc-labelled

preparation is too low to allow a reliable detection. However, preliminary experiments on well established routinely used

99m

Tc radiopharmaceuticals184

allowed us to conclude that this technique provides data supporting the hypothesized structure of an experimental compound. As a rule however, it appeared preferable or even necessary to add small amounts of carrier

99

Tc

(about 1.5 µg in the form of ammonium pertechnetate) to be able to determine the mass of the selected peaks. So, radio-LC-MS appeared to be a very efficient tool to obtain supporting data on the identity of experimental radiolabelled compounds.

78

Chapter 2

2.3.3.1. Radio-LC-MS analysis of 99mTc-2.4

Figure 22 B depicts the radiometric signal of the radio-LC-MS analysis of 99m

Tc-2.4, which shows different radioactive peaks. Figure 23 A shows the

background subtracted mass spectrum of the peak with Rt 18.78 min (summed mass spectra obtained prior and after the peak at 18.78 min subtracted from the summed mass spectrum over the peak at 18.78 min), showing one single molecular ion mass of 559.2326 Da, which corresponds to the theoretical mass of

99m

Tc-2.4 (Figure 23 B; 559.0112 Da). Figure 22 A

shows the single ion mass over the mass range 559.172-559.322 Da. As the peak on this single ion mass chromatogram has both an identical retention time and shape as the peak observed in the radiometric channel, the Tc complex eluting at 18.78 min is probably 18.78 min corresponds to

99m

Tc-2.4, while

99m 99m

Tc-2.4. As such, the peak at

Tc-tartrate elutes with the void

volume at 1.64 min. Peaks at 2.95 min and 6.82 min correspond to pertechnetate and the 99mTc-BAT-CH2-COOH, respectively. J C 3 0 2 0 to ta a l p re p a ra a t S m (S G , 5 x 5 ) 1 8 .7 4 100

1: TO F M S ES+ 5 5 9 .1 7 2 _ 5 5 9 .3 2 2 204

A

% 1 3 .1 9

0 J C 3 0 2 0 to ta a l p re p a ra a t 1 .6 4 100

R A NaI An2 8 .5 4 e 5

B

6 .8 2

%

2 .9 5

1 8 .7 8

99mTc-2.4

8 .8 1 9 .9 9

0

1 0 .0 0

2 0 .0 0

T im e 3 0 .0 0

Figure 22. A = single ion mass chromatogram (ES+, 559.172-559.322 Da) and B = radiometric chromatogram of 99mTc-2.4

These data indicate that the peak eluting at 18.78 min in the radiometric channel (Figure 22 B) is indeed 99mTc-2.4.

Development of thioflavin-T derivatives for SPECT studies

79

559.2326

100

A

%

0 112.6081 187.0086

100

280.0093

559.0112

B

%

0 100

187.3428

280.5106

200

300

560.0140

400

500

600

700

800

900

m/z 1000

Figure 23. A = background subtracted mass spectrum of the chromatogram peak at 18.78 min (Figure 22 A) and B = theoretical molecular ion mass of 99m

Tc-2.4

For LC-MS analysis gradient system 1 was used as described in 2.2.4, but the retention times are shorter in comparison with the previous RP-HPLC analysis described in 2.3.2 because of the smaller size of the column.

The molecular ion mass found for

99m

Tc-2.4 (559.2 Da) is in agreement

with a Tc(V)O-complex of a BAT-BCL of 2.4. Upon complexation of the [Tc(V)O]3+ core, both the thiols and one amine of the BAT group lose a proton. Three positive charges of the [Tc(V)O]3+ core are balanced by the loss of these three protons and a neutral complex is thus formed.

In case of a BAT tetraligand, the incorporation of TcO in the mono-Nsubstituted N2S2 ligand can potentially yield two pairs of diastereomers. The complex has an asymmetric amine nitrogen (four bonds of which one is a coordinate bond) and the TcO core can then be oriented syn or anti with respect to the side chain on this chiral amine nitrogen atom. In general, these complexes form the syn diastereomers185. Using conventional C18 RP-HPLC, enantiomers cannot be separated, which declares why we only see one peak on

RP-HPLC.

The

radio-LC-MS

chromatogram

suggests

that

both

diastereomers are nevertheless present, albeit that one diastereomer (Figure 22 A, peak at 13.19 min) is only present in a very small amount.

80

Chapter 2 O

O

R NH

Tc

O

N

R NH

S N S

N S

N

Tc

S

O

Syn

Anti

Figure 24. Syn and anti forms of the 99mTc-labelled BAT ligand

2.3.3.2. Radio-LC-MS analysis of 99mTc-2.12

Figure 25 B shows the radiometric signal of the radio-LC-MS analysis of the reaction mixture of the labelling of 2.12 with

99m

Tc, with tricine as the co-

ligand. The peak at 10.63 min is supposed to be the intended radiolabelled compound

99m

Tc-tricine-2.12. In Figure 26, the background subtracted mass

spectrum of the peak eluting at 10.63 min is depicted and the mass obtained in this spectrum was 749.1918 Da. Figure 25 A depicts the single ion mass over the mass range 749.084-749.312 Da. Although the structures proposed in literature for

99m

Tc-HYNIC complexes contain two tricine molecules, our

radio-LC-MS results suggest a Tc-complex with deprotected 2.12, which includes one molecule of tricine and two molecules of water. TOV 67 8811 gemerkt, isocratisch Sm (Mn, 5x4) 10.62 100

1: TOF MS ES+ 749.084_749.312 53.9

A

%

0 TOV 67 8811 gemerkt, isocratisch 100

RA NaI An1 9.28e5

10.63

B

99mTc-tricine-2.12

%

0 5.00

10.00

15.00

20.00

Time 25.00

Figure 25. A = single ion mass chromatogram (ES+, 749.084-749.312 Da) and B = radiometric chromatogram obtained by LC-MS analysis of 2.12

99m

Tc-tricine-

Development of thioflavin-T derivatives for SPECT studies TO V 67 8811 gem erkt, isocratisch 649 (10.821) C m (624:651-651:652) 749.1918 100

81 1: TO F M S E S + 124

749.1406

750.2062

%

0

m /z 742

744

746

748

750

752

754

Figure 26. Background subtracted mass spectrum of the chromatogram peak eluting at 10.63 min (Figure 25 B)

2.3.3.3. Radio-LC-MS analysis of 99mTc-2.13

Figure 27 B shows the radiometric signal of the radio-LC-MS analysis of the reaction mixture of the labelling of 2.13 with supposed to be

99m

99m

Tc. The peak at 11.38 min is

Tc-2.13, while the peak at 0.87 min represents the void

volume of the column. In Figure 28 B, the background subtracted mass spectrum shows a mass of 583.0304 Da for this compound. Figure 28 A depicts the theoretical mass for the supposed structure of

99m

Tc-2.13 as

582.9573 Da. In Figure 27 A, the single ion mass is depicted over the mass range 583.406-583.474 Da. These data strongly support the assumption that the peak eluting at 11.38 min in the radiometric channel (Figure 27 B), corresponds to a structure as shown in Figure 31 (see further).

The mass found for

99m

Tc-2.13 (583 Da) is in accordance with a

negatively charged complex in which a [Tc(V)O]3+ core is bound to the deprotonated thiol sulphur atom and three deprotonated amide nitrogen atoms of the mercaptotriamide tetraligand. A similar complex is formed in the case of the fully characterized and commonly used 99mTc-MAG3 complex.

82

Chapter 2

1 1 .4 0

100

%

A

1 .2 2

0 100

0 .8 7

B 99m

%

Tc-2.13

1 1 .3 8

0 0 .0 0

5 .0 0

1 0 .0 0

1 5 .0 0

2 0 .0 0

T im e 3 0 .0 0

2 5 .0 0

Figure 27. A = single ion mass chromatogram (ES-, 583.406-583.474 Da) and B = radiometric chromatogram obtained by LC-MS analysis of 99mTc-2.13

582.9573

100

A

% 0 583.0304

100

582.9922

% 0 570

B

m/z 575

580

585

590

Figure 28 A = theoretical molecular ion mass of

595

99m

Tc-2.13 and B =

background subtracted mass spectrum of the peak at 11.38 min (Figure 27 B)

2.3.3.4. Radio-LC-MS analysis of 99mTc-2.14

Figure 29 B shows the radiometric signal of the radio-LC-MS analysis of the reaction mixture after labelling of 2.14 with 99mTc(CO)3. The peak at 12.01 min is supposed to be the intended radiolabelled compound

99m

Tc-2.14. In Figure

30 A, the background subtracted mass spectrum shows a mass of 579.6915 Da for this compound. Figure 30 B depicts the theoretical mass for the supposed structure of

99m

Tc-2.14 as 579.9642 Da. Figure 29 A depicts the

single ion mass chromatogram over the mass range 579.523-580.051 Da. These data strongly support the assumption that the peak eluting at 12.01 min

Development of thioflavin-T derivatives for SPECT studies

83

in the radiometric channel (Figure 29 B) corresponds to a structure as shown in Figure 31 (see further). In such Tc(CO)3 complex, Tc has a valence of +1 whereas two protons of the IDA moiety are lost, resulting in an overall anionic complex (charge -1).

A

12.01

100

%

-1 100

B 99mTc-2.14

%

0 5.00

10.00

15.00

20.00

Time 30.00

25.00

Figure 29. A = single ion mass chromatogram (ES-, 579.523-580.051 Da) and B = radiometric chromatogram obtained by LC-MS analysis of 99mTc-2.14

5 7 9 .6 9 1 5

100

A

5 7 9 .7 8 1 7 5 7 9 .6 6 8 9

% 5 8 0 .7 1 8 0

0 5 7 9 .9 6 4 2

100

B %

0

m /z 571

572

573

574

575

576

577

578

579

580

581

582

583

584

585

586

587

Figure 30. A = background subtracted mass spectrum of the peak at 12.01 min (Figure 29 B) and B = theoretical molecular ion mass of 99mTc-2.14

In Figure 31 the structures of

99m

Tc-2.4,

99m

Tc-2.13 and

99m

Tc-2.14 are

depicted as proposed on the basis of the radio-LC-MS data and findings from literature. Since the structure of 99mTc-tricine-2.12 remains highly hypothetical, this hypothesized structure is not depicted in Figure 31. The definitive

84

Chapter 2

structure determination of this complex requires the synthesis of the corresponding rhenium complex and its crystal x-ray diffraction analysis.

0 O

N NH

N O N

S

Tc 99m

Tc-2.4

S

S

-1

O O N

OH O

N

O

Tc O N

N

S

S 99m

Tc-2.13 -1

O O

N NH

O CH2 N

Tc

S

O 99m

Tc-2.14

CO CO CO

O

Figure 31. Proposed structures of 99mTc-2.4, 99mTc-2.13 and 99mTc-2.14

2.3.4. Partition coefficients

Depending on the functional groups present in the technetium binding ligand and the oxidation state of technetium after reduction (with SnCl2 or NaBH4 in case of the tricarbonyl labelling), a technetium complex is positively charged, negatively charged or neutral.

99m

Tc-2.4, designed as a potential tracer agent

for diagnosis of AD, should be neutral to be able to cross the BBB. Its partition coefficient was found to be 83.28 ± 2.35 (log P = 1.92 ± 0.012) and is in a good range for passive diffusion of neutral molecules over the BBB186. As its molecular mass does not exceed 650 Da186, 99mTc-2.4 may cross the BBB.

Development of thioflavin-T derivatives for SPECT studies For the three other

99m

85

Tc-labelled compounds, penetration through the

BBB is unnecessary. Therefore, they should preferentially be hydrophilic as they were designed for detection of PA, for which renal clearance is the important requirement and their hydrophilic character should be reflected in their partition coefficient. For

99m

Tc-tricine-2.12, the P value was 6.50 ± 0.16

(log P = 0.81 ± 0.011), which is still rather apolar. For 99mTc-2.13, the P value was 0.32 ± 0.013 (log P = - 0.49 ± 0.017), which is rather hydrophilic.

99m

Tc-

2.14 has a P value of 12.31 ± 0.46 (log P = 1.1 ± 0.016). Only for 99mTc-2.13, a value was observed which is in accordance with a relatively hydrophilic compound, whereas the log P values of

99m

Tc-2.12 and

99m

Tc-2.14 indicate

that these two radiolabelled compounds are still rather hydrophobic.

2.4. Conclusion Four derivatives of TT with a technetium binding moiety have been successfully synthesized. For AD, a BAT tetraligand was chosen that would yield an uncharged technetium labelled derivative that potentially could be used to visualize amyloid β plaques in the brain. For PA, three derivatives of TT were synthesized which yield charged technetium labelled complexes. All ligands were synthesized in sufficient yields and their identity and structure was confirmed. Deprotection and labelling with technetium of the studied compounds was done using either direct labelling, exchange labelling or Tc(CO)3 labelling, providing good labelling yields for all four compounds as was seen on RP-HPLC. Radio-LC-MS confirmed the hypothesized structure of all compounds except for 2.12 for which the observed molecular ion mass was not in agreement with earlier literature reports with regard to the structure of Tc-HYNIC complexes. 99m

Tc-2.4 had a log P value in the good range for passive diffusion over

the BBB. Log P of the other compounds showed that introduction of a ligand that provides charged technetium complexes, not necessarily results in a hydrophilic tracer in toto, as was indicated by the relatively high log P value of 99m

Tc-tricine-2.12

and

99m

Tc-2.14.

Although

the

structure

of

the

86

Chapter 2

phenylbenzothiazole is largely maintained in these compounds, the question has to be answered if these compounds still have affinity for amyloid plaques and if they show the desired in vivo pharmacokinetics.

87

CHAPTER 3

3. DEVELOPMENT OF THIOFLAVIN-T DERIVATIVES LABELLED WITH A POSITRON EMITTER

3.1. Positron emission tomography (PET) PET is, just like SPECT, an in vivo imaging technique which provides crosssectional images of the distribution of a radiolabelled tracer in the body. Due to the specific characteristics of positron emitters and PET cameras, PET also allows quantitative assessment of biochemical and physiological processes in a non-invasive manner. The γ-rays generated by the decay of positron emitters nevertheless have a different origin as compared to the single photons of SPECT tracers. One of the possibilities of neutron-poor radioisotopes to become more stable, is the conversion of a proton into a neutron with the concomitant emission of a positron (the positively charged equivalent of an electron) and a neutrino. The positron is slowed down by electrostatic interaction with surrounding matter and combines with an electron to form a positronium. This positronium has a very short half-life and annihilates almost immediately by the conversion of the combined mass of the positron and electron to energy. This energy is electromagnetic radiation which is emitted in the form of two gamma rays with an energy of 511 keV each (the energy equivalent of the mass of an electron), emitted at an angle of 180° opposite to each other. These γ-rays can be detected simultaneously by a PET camera mostly existing of 16 rings with 256 detectors in each ring. The detector material is lutetium oxyorthosilicate (LSO) and these detector crystals are connected to each other with fast coincidence circuits that enable to verify whether two detected γ-rays indeed originate from the annihilation of the same

88

Chapter 3

positronium. The time interval between detection of the two opposite γ-rays should then be approximately 10-20 nanoseconds. The two detectors that simultaneously detected the γ-rays, can be connected by a virtual line that contains the point where the annihilation took place. Due to the large number of detectors, modern PET cameras can detect millions of annihilations at the same time. By combining all these annihilation data with mathematic techniques, an image can be reconstructed that reflects the distribution of the radiolabelled tracer agent in the field depicted on the image. The spatial resolution (the minimal physical distance between two point sources necessary to observe them as two separate objects) of a clinical PET camera is higher than that of a clinical SPECT camera (4-5 mm versus 8-10 mm, respectively). The minimal structural changes made by the incorporation of the PET radionuclides in a BAM are another advantage of PET. These changes are often radioisotopes of atoms such as carbon, nitrogen or oxygen which are present in biologically important molecules and their incorporation yields radiolabelled tracers that are chemically identical to the BAM and as such show the same physiological behaviour. The widely used PET radionuclide fluorine-18 is an exception to this rule, but its incorporation in biologically active compounds most often results in tracers that have only minor structural and biological changes (18F-FDG is a good example). Because of the short half-life of

15

O and

13

N (122 sec and 9.96 min,

respectively), these radionuclides are less attractive for incorporation in a BAM than 11C and 18F (half-life of 20.4 min and 109.8 min, respectively). From a practical point of view, fluorine-18 is the most appropriate of these radionuclides, but 11C-labelled tracer agents are also frequently used. Another advantage of PET is its higher sensitivity than SPECT. This is a consequence of the simultaneous detection of the two annihilation γ-rays on two crystals, which obviates the need of a collimator in front of the detector, as is the case in a SPECT camera. Furthermore, PET allows attenuation-correction (correction for the absorption of photons within the patient’s body), which enables to relate measured count rates to absolute tracer concentrations. All the projections are acquired simultaneously without moving the detectors (in the ring system), which enables rapid sequential imaging followed by an

Development of thioflavin-T derivatives labelled with a positron emitter

89

assessment of the change in tracer concentration over time. This results in a more accurate quantisation of dynamic processes.

The production of positron emitting radionuclides requires a cyclotron to create a nucleus with an artificial surplus of protons by adding a proton to a stable nucleus. Since the nucleus and a proton both are positively charged, the coulomb repulsion between them has to be overcome to allow the fusion of both particles. This is achieved by accelerating the protons to a high kinetic energy in a cyclotron. The accelerated protons will then collide with the target nuclei yielding unstable nuclei due to their proton surplus. Which radionuclide is henceforth produced, depends on the target nucleus and the nuclear reaction, which in turn is dependent on the energy of the incident proton. In the case of carbon-11 production, natural yielding carbon-11 through an

14 7

14 7

N2 gas is bombarded with protons,

N(p,α) 116 C reaction. In the case of fluorine-18,

the starting material is H218O from which 18 8

18

F is produced through an

O(p,n) 189 F reaction. Due to their short half-lives, these radionuclides must be

incorporated in the desired radiolabelled compound as quickly as possible and in an efficient and relatively easy way. Since carbon-11 and fluorine-18 have the longest half-live among the most common PET radionuclides, they are also the most interesting radionuclides for the development of new PET tracer agents.

Although PET has clear advantages over SPECT, there are also some disadvantages. First of all, the technique is much more expensive. Because of the short half-life of the radionuclides, one needs an expensive cyclotron in the nuclear medicine department. PET cameras are also more expensive than SPECT cameras and the emitted γ-rays of 511 keV demand more lead shielding than the 140 keV γ-rays of technetium-99m, resulting in a higher cost for shielding PET workspaces. Secondly, the production of these tracer agents mostly demands purification and/or analysis with RP-HPLC to isolate the desired tracer agent and/or confirm its purity, which makes the production more labour intensive. In the third place, in case of preparations for clinical use, it is also necessary to formulate the radiolabelled compound in a sterile

90

Chapter 3

form, in a solution which is isotonic and pyrogen-free and as such ready for injection in a patient. And last but not least, due to the fast decay of these PET radionuclides, the shelf-life of PET radiopharmaceuticals is limited in comparison to that of technetium-99m labelled tracer agents (half-life of 6 hours). With regard to the almost 2 h half-life of the radionuclide, only fluorine18 labelled agents can be used, allowing a number of doses from the same production batch to be provided. Carbon-11 radiopharmaceuticals, however, require one production per injection.

Because of the interesting characteristics of tracer agents labelled with a PET radionuclide and especially because of the fact that only minor or no structural changes are necessary for the labelling of biologically active compounds, a number of uncharged derivatives of TT labelled with carbon-11 or fluorine-18 were developed. Because of the relatively favourable characteristics of the benzoxazole IBOX (see Figure 7) the question arose if a benzimidazole would also be useful for detection of amyloid plaques. Since no benzimidazoles have been described in literature to date, not only benzothiazoles but also one benzimidazole was included in this series for evaluation.

3.2. Materials and methods For general remarks, see paragraph 2.2.

RP-HPLC system consisted of a Merck-Hitachi ternary gradient pump (model L-6200A intelligent pump, Merck) and an XTterra™ RP C18 column (5 µm, 250 mm x 4.6 mm) (Waters). UV absorbance was measured with a dual wavelength

absorbance

detector

(Waters

2487,

Waters)

(254

nm).

Radioactivity was measured with a 3-inch NaI(Tl) scintillation detector coupled to a single channel analyzer (ACEMate™ amplifier and bias supply, EG & G ORTEC). The mobile phase consisted of a mixture of acetonitrile and 0.05 M ammonium acetate (unless otherwise stated) and gradient as well as isocratic elution were used.

Development of thioflavin-T derivatives labelled with a positron emitter

91

The MPLC-system consisted of a pump (Büchi, model B-680) and a column (internal diameter 2.5 cm, length 45 cm) filled with octadecylsilyl silica gel (LiChroprep®, 15-25 µm, mean pore diameter 100 Å, Merck, Darmstadt, Germany). The column eluate was monitored for UV absorbance at 254 nm (Econo UV monitor model EM-1, Biorad, Hercules, CA). [18F]Fluoride was produced by irradiation of 97 % enriched [18O]water (Rotem HYOX18, Rotem Industries, Beer Sheva, Israel) with 10-MeV protons generated by a Cyclone 10/5 cyclotron (Ion Beam Applications, Louvain-LaNeuve, Belgium). Carbon-11 was produced by irradiation of N2-gas (with 5 % O2) with 10-MeV protons. Partition coefficients were determined as described in 2.2.6.

3.2.1. Synthesis of neutral thioflavin-T derivatives for diagnosis of AD

3.2.1.1. Synthesis of 2-(4’-nitrophenyl)-1,3-benzothiazole (3.1)

A dispersion of p-nitrobenzoic acid (4.18 g, 25 mmol) and 2-aminothiophenol (3.13 g, 25 mmol) in 50 g of PPA was stirred at 150 °C for 1 h. After cooling down to 60 °C, 300 ml of water was added and the pH of the mixture was adjusted to pH 4.5 with 2 M NaOH. A green precipitate formed, which was then filtered off, dried under vacuum and crystallized from methanol. This yielded 2.15 g of a green solid (8.4 mmol, 33.6 % yield). 1

H-NMR (CDCl3): δ 7.47 (1H, dd, 6-H); δ 7.57 (1H, dd, 5-H); δ 7.97 (1H, d, 4-

H); δ 8.14 (1H, d, 7-H); δ 8.27 (2H, d, 2’-H 6’-H); δ 8.36 (2H, d, 3’-H 5’-H) Mp: 226.1-227.4 °C

3.2.1.2. Synthesis of 2-(4’-fluorophenyl)-1,3-benzothiazole (3.2)

To a solution of 4-fluorobenzoyl chloride (1.5 g, 9.5 mmol) and 2aminothiophenol (1.18 g, 9.5 mmol) in 75 ml of toluene, 2 mg of ptoluenesulphonic acid was added. The mixture was refluxed in a flask equipped with a Dean-Starck trap for 2 h, cooled down to room temperature

92

Chapter 3

and concentrated under reduced pressure. The residual white solid was crystallized from ethanol/water to yield 1.2 g of a white solid (5.2 mmol, 55 % yield). 1

H-NMR (CDCl3): δ 7.16 (2H, dd, 3’-H 5’-H); δ 7.38 (1H, t, 5-H); δ 7.50 (1H, t,

6-H); δ 7.90 (1H, d, 7-H); δ 8.06 (2H, d, 2’-H 6’-H); δ 8.10 (1H, d, 4-H) Mp: 97.2-98.5 °C

3.2.1.3. Synthesis of 2-(4’-aminophenyl)-7-hydroxy-1,3-benzothiazole

N-(2-bromo-3-methoxyphenyl)-4-nitrobenzamide (3.3)

A solution of 2-bromo-3-aminoanisole (5.05 g, 25 mmol) and p-nitrobenzoyl chloride (4.64 g, 25 mmol) in 100 ml of pyridine was refluxed for 3 h and then allowed to cool down to room temperature. Upon addition of water (250 ml), a white precipitate formed, which was filtered off, washed with water and dried under reduced pressure. Yield of synthesis: 5 g of 3.3 (14.3 mmol, 57.2 %). 1

H-NMR (DMSO): δ 3.89 (3H, s, CH3); δ 7.07 (1H, d, 6-H); δ 7.17 (1H, d, 4-H);

δ 7.41 (1H, t, 5-H); δ 8.21 (2H, d, 2’-H 6’-H); δ 8.39 (2H, d, 3’-H 5’-H); δ 10.38 (1H, s, NH-CO) +

Mass: [M+H] 350 (calculated: 350) Mp: 242-243.6 °C

N-(2-bromo-3-methoxyphenyl)-4-nitrobenzothioamide (3.4)

A solution of 3.3 (4.6 g, 13 mmol) in 200 ml of dioxane, was heated to 90 °C and Lawesson’s reagent (3.15 g, 7.8 mmol) was added in fractions. The mixture was refluxed for 4 h. After cooling down to room temperature the mixture was poured into water and an orange precipitate formed, which was filtered off and dried under reduced pressure. The product was used without further purification. Yield of synthesis: 4.6 g of 3.4 (12.4 mmol, 95.4 %). 1

H-NMR (DMSO): δ 3.89 (3H, s, CH3); δ 7.07 (1H, d, 6-H); δ 7.13 (1H, d, 4-H);

δ 7.40 (1H, t, 5-H); δ 8.21 (2H, d, 2’-H 6’-H); δ 8.38 (2H, d, 3’-H 5’-H); δ 10.37 (1H, s, NH-CS)

Development of thioflavin-T derivatives labelled with a positron emitter

93

+

Mass: [M+H] 368 (calculated: 368) Mp: 174.5-176 °C

2-(4’-Nitrophenyl)-7-methoxy-1,3-benzothiazole (3.5)

To a solution of 3.4 (4 g, 11 mmol) in 50 ml N-methylpyrrolidone (NMP), sodium hydride (290 mg, 12.1 mmol) was added in fractions while stirring at room temperature. The mixture was heated at 140 °C for 2 h, allowed to cool down to room temperature and then poured into water. A brownish precipitate formed, which was filtered off and dried under reduced pressure. The product was used without further purification. Yield of synthesis: 2 g of 3.5 (7 mmol, 63.6 %). 1

H-NMR (DMSO): δ 3.89 (3H, s, CH3); δ 7.15 (1H, d, 6-H); δ 7.58 (1H, t, 5-H);

δ 7.77 (1H, d, 4-H); δ 8.21 (2H, d, 2’-H 6’-H); δ 8.39 (2H, d, 3’-H 5’-H) Mp: 199.9-201.3 °C

2-(4’-Aminophenyl)-7-methoxy-1,3-benzothiazole (3.6)

To a suspension of 3.5 (1.8 g, 6.3 mmol) in 100 ml ethanol, stannous chloride dihydrate (7.1 g, 31.5 mmol) was added and the mixture was refluxed under nitrogen for 5 h. The mixture was allowed to cool down to room temperature and ethanol was removed by evaporation under reduced pressure. The residue was redissolved in ethyl acetate and the mixture was washed with water (2x 200 ml), 2 M NaOH (2x 150 ml), again with water (150 ml) and then with brine. The organic layer was dried on magnesium sulphate, filtered and evaporated under reduced pressure to yield 1 g of a yellow solid (62 % yield). 1

H-NMR (CDCl3): δ 3.90 (3H, s, CH3); δ 6.67 (2H, d, 3’-H 5’-H); δ 6.77 (1H, d,

6-H); δ 7.38 (1H, t, 5-H); δ 7.63 (1H, d, 4H); δ 7.90 (2H, d, 2’H 6’H) +

Mass: [M+H] 257 (calculated: 257) Mp: 151.1-154 °C

94

Chapter 3

2-(4’-Aminophenyl)-7-hydroxy-1,3-benzothiazole (3.7)

A solution of 3.6 (513 mg, 2 mmol) in 50 ml of dry dichloromethane was cooled down to -70 °C and 12 ml of a 1 M BBr3 solution (12 mmol) in dichloromethane was added dropwise under nitrogen over 30 minutes. The mixture was kept for an additional hour at -70 °C and was then allowed to warm up to room temperature and stirred further at room temperature overnight. The mixture was cooled down to -70 °C and 50 ml of methanol was added dropwise, followed by 50 ml of 2 M NaOH. The water layer was separated, neutralized with 6 M HCl and extracted with CH2Cl2/MeOH (4:1, v/v) (3x 100 ml). The organic layers were collected, washed with water, dried on magnesium sulphate, filtered and evaporated to dryness under reduced pressure. The product was purified on MPLC to yield 180 mg pure 3.7 (0.75 mmol, 37.5 %). 1

H-NMR (DMSO): δ 6.67 (2H, d, 3’-H 5’-H); δ 6.77 (1H, d, 6-H); δ 7.26 (1H, t,

5-H); δ 7.38 (1H, d, 4-H); δ 7.75 (2H, d, 2’-H 6’-H) +

Mass: [M+H] 243 (calculated: 243) Mp: 253.2-255 °C

3.2.1.4. Synthesis of 2-(4’-methylaminophenyl)-7-hydroxy-1,3-benzothiazole (3.8)

A solution of 3.7 (30 mg, 0.12 mmol) and CH3I (80 µl, 1.2 mmol) in 5 ml of acetonitrile was heated at 80 °C for 36 h. The organic solvent was removed under reduced pressure and the residue was dissolved in acetonitrile/0.05 M ammonium

acetate

(50:50,

v/v)

and

purified

with

MPLC

(eluent:

acetonitrile/0.05 M ammonium acetate (50:50, v/v) with a flow rate of 12 ml/min). This yielded 5.7 mg of 3.8 (0.022 mmol, 18.5 % yield). Mass: [M+H]+ 257 (calculated: 257)

Development of thioflavin-T derivatives labelled with a positron emitter

95

3.2.1.5. Synthesis of 2-(4’-aminophenyl)-5-hydroxy-1,3-benzothiazole

N-(2-chloro-5-methoxyphenyl)-4-nitrobenzamide (3.9)

A solution of 6-chloro-m-anisidine.HCl (10 g, 51.5 mmol) in 50 ml of pyridine was cooled down to 0 °C and p-nitrobenzoyl chloride (9.557 g, 51.5 mmol) was added dropwise. The mixture was refluxed for 1 h and poured into 100 ml of water. The formed precipitate was filtered off, washed with water and dried under reduced pressure to yield 12.28 g (40 mmol, 77.7 % yield) of 3.9. 1

H-NMR (DMSO): δ 3.85 (3H, s, CH3); δ 6.70 (1H, dd, 4-H); δ 7.31 (1H, d, 3-

H); δ 8.07 (2H, d, 2’-H 6’-H); δ 8.21 (1H, d, 6-H); δ 8.38 (2H, d, 3’-H 5’-H); δ 8.45 (1H, s, NH-CO) +

Mass: [M+H] : 307 (calculated: 307)

N-(2-chloro-5-methoxyphenyl)-4-nitrobenzothioamide (3.10)

A solution of 3.9 (12.28 g, 40 mmol) and Lawesson’s reagent (11.32 g, 24 mmol) in 200 ml of dioxane, was refluxed for 2.5 h, cooled down and then poured into 200 ml of water whilst stirring. The formed precipitate was filtered off, washed with water and dried under reduced pressure. A bright, orange solid was obtained (12 g, 37.3 mmol, 93.3 %). 1

H-NMR (DMSO): δ 3.82 (3H, s, CH3); δ 6.83 (1H, dd, 4-H); δ 7.38 (1H, d, 3-

H); δ 7.97 (2H, d, 2’-H 6’-H); δ 8.28 (2H, d, 3’-H 5’-H); δ 8.51 (1H, s, 6H); δ 9.4 (1H, s, NH-CS) +

Mass: [M+H] 323 (calculated: 323) Mp: 136.2-138.2 °C

2-(4’-Nitrophenyl)-5-methoxy-1,3-benzothiazole (3.11)

Sodium (362 mg, 15.75 mmol) was dissolved in 20 ml of methanol, the solvent was evaporated to dryness under reduced pressure and the residue was added to a solution of 3.10 (3.38 g, 10.5 mmol) in 50 ml of NMP. The mixture was stirred overnight at 110 °C. The mixture was poured into 200 ml of cold

96

Chapter 3

water and a precipitate formed. This precipitate was filtered off, dried under reduced

pressure

and

further

purified

with

flash

chromatography

(hexane/ethyl acetate 9:1-9:2). This yielded 150 mg of 3.11 (0.52 mmol, 3.3 %). 1

H-NMR (CDCl3): δ 3.91 (3H, s, CH3); δ 7.11 (1H, dd, 6-H); δ 7.58 (1H, d, 4-

H); δ 7.78 (1H, d, 7-H); δ 8.21 (2H, d, 2’-H 6’-H); δ 8.33 (2H, d, 3’-H 5’-H)

2-(4’-Aminophenyl)-5-methoxy-1,3-benzothiazole (3.12)

A dispersion of 3.11 (0.4 mmol, 115 mg) and stannous chloride dihydrate (225 mg, 1.2 mmol) in 3 ml of ethanol was refluxed under nitrogen for 4 h. The mixture was cooled down to room temperature and the organic solvent was removed under reduced pressure. The solution of the residue in ethyl acetate was washed 3 times with 3 M NaOH and once with brine. The organic layer was dried on magnesium sulphate, filtered off and evaporated to dryness under reduced pressure, yielding 79 mg of 3.12 (yield: 77.1 %). +

Mass: [M+H] : 257 (calculated: 257)

2-(4’-Aminophenyl)-5 hydroxy-1,3-benzothiazole (3.13)

A solution of 3.12 (77 mg, 0.3 mmol) in 5 ml of dry dichloromethane under nitrogen was cooled down to -70 °C. A 1 M solution of BBr3 in dichloromethane (1.5 ml, 1.5 mmol) was added dropwise over a 30 min period and the mixture was kept at -70 °C for an additional hour. The mixture was allowed to warm up to room temperature and was stirred overnight. Following that, the mixture was again cooled down to -70 °C and 2 ml of methanol were added dropwise, followed by 5 ml of 2 M NaOH and 10 ml of dichloromethane. The aqueous layer was separated, the dichloromethane layer was washed with 5 ml of 2 M NaOH and the collected aqueous layers were neutralized with 6 M HCl. Upon neutralization, a precipitate formed, which was filtered off and dried under reduced pressure, yielding 55 mg of 3.13 (76 %).

Development of thioflavin-T derivatives labelled with a positron emitter 1

97

H-NMR (DMSO): δ 5.85 (2H, s, NH2); δ 6.66 (2H, d, 3’-H 5’-H); δ 6.85 (1H,

dd, 6-H); δ 7.25 (1H, d, 4-H); δ 7.72 (1H, d, 7-H); δ 7.75 (2H, d, 2’-H 6’-H); δ 9.67 (1H, s, OH) +

Mass: [M+H] 243 (calculated: 243) Mp: 250.5-252.4 °C

3.2.1.6. Synthesis of 2-(4’-methylaminophenyl)-5-hydroxy-1,3-benzothiazole

Synthesis of 4-methoxy-2-amino-thiophenol (3.14)

To a mixture of ethylene glycol (23 ml) and 10 M NaOH (23 ml), 5-methoxy-2methyl-benzothiazole (5 g, 28 mmol) was added and the mixture was refluxed for 2 h under nitrogen. After cooling down, the mixture was neutralized with concentrated HCl and extracted with ethyl acetate (3 x 50 ml). The organic layer was dried on MgSO4, filtered and evaporated to dryness under reduced pressure. The obtained product 3.14 was used without further purification.

Synthesis of 2-(4’-methylaminophenyl)-5-methoxy-1,3-benzothiazole (3.15)

A dispersion of 3.14 (2.4 g, 28 mmol) and 4-(methylamino)benzoic acid (2 g, 28 mmol) in PPA (6 g) was stirred at 180 °C for 3 h. After cooling down, 100 ml of a 10 % (m/v) Na2CO3 solution was added and the water layer was extracted with ethyl acetate (2 x 100 ml). The combined organic fractions were washed with a saturated Na2CO3 solution and brine, dried on magnesium sulphate, filtered and evaporated to dryness under reduced pressure. Purification was done with column chromatography (hexane/ethyl acetate; 8:2, v/v). The fractions containing the pure product were used for further synthesis.

Synthesis of 2-(4’-methylaminophenyl)-5-hydroxy-1,3-benzothiazole (3.16)

A solution of 3.15 (57 mg, 0.3 mmol) in 5 ml of dry dichloromethane was cooled down to -70 °C under nitrogen and a 1 M solution of BBr3 in dichloromethane (1.3 ml, 1.3 mmol) was added dropwise over a 30 min

98

Chapter 3

period. The mixture was kept at -70 °C for 1 h and then stirred overnight at room temperature. The mixture was again cooled down to -70 °C and 2 ml of methanol was added dropwise followed by 5 ml of a 2 M NaOH solution and 10 ml of dichloromethane. The NaOH layer was separated and the dichloromethane layer was washed with 5 ml of a 2 M NaOH solution. The collected NaOH layers were neutralized with a 6 M HCl solution. A precipitate formed, which was filtered off and dried under reduced pressure. This yielded 45 mg of 3.16 (0.18 mmol, 59 %). 1

H-NMR (DMSO): δ 2.75 (3H, d, CH3); δ 6.40 (1H, d, NH); δ 6.63 (2H, d, 2’-H

6’-H); δ 6.84 (1H, d, 6-H); δ 7.25 (1H, s, 4-H); δ 7.73 (1H, d, 7-H); δ 7.77 (2H, d, 3’-H 5’-H); δ 9.64 (1H, s, OH) +

Mass: [M+H] 257 (calculated: 257) Mp: 260-261 °C

3.2.1.7. Synthesis of 2-(4’-aminophenyl)-1,3-benzimidazole (3.17)

A dispersion of p-aminobenzoic acid (5.83 g, 42.5 mmol) and ophenylenediamine (4.6 g, 42.5 mmol) in 65 g of PPA was stirred for 5 h at 180 °C. After cooling down to room temperature, 500 ml of water was added and the precipitate was filtered off and resuspended in a 2 M NaOH solution. The fraction that was not dissolved, was filtered off, dried under reduced pressure

and

purified

with

column

chromatography

(dichloromethane/methanol from 95:5 to 80:20, v/v). This yielded 1.49 g of 3.13 (7.1 mmol, 16.8 % yield). 1

H-NMR (DMSO): δ 5.60 (2H, s, NH2); δ 6.67 (2H, d, 3’-H 5’-H); δ 7.10 (1H, t,

5-H); δ 7.12 (1H, t, 6-H); δ 7.48 (2H, br s, 4-H 7-H); δ 7.85 (2H, d, 2’-H 6’-H). +

Mass: [M+H] : 210 (calculated: 210) Mp: 235.8-236.5 °C

3.2.1.8. Synthesis of 2-(4’-aminophenyl)-N-methyl-1,3-benzimidazole (3.18)

A dispersion of p-aminobenzoic acid (2.74 g, 20 mmol) and N-methyl-1,2phenylenediamine (2.3 ml, 20 mmol) in PPA (10 g) was stirred at 180 °C for 2

Development of thioflavin-T derivatives labelled with a positron emitter

99

h. After cooling down to room temperature, 200 ml of a 10 % (m/v) Na2CO3 solution was added and a precipitate formed, which was filtered off, washed repeatedly with water and crystallized from ethanol/water. This yielded 2.5 g of 3.18 (11.2 mmol, 56 %). 1

H-NMR (CDCl3): δ 3.58 (3H, s, CH3); δ 6.79 (2H, d, 3’-H 5’-H); δ 7.26 (2H, m,

5-H 6-H); δ 7.39 (1H, m, 7-H); δ 7.54 (1H, d, 2’-H 6’-H); δ 7.71 (1H, m, 4-H) +

Mass: [M+H] : 224 (calculated: 224) Mp: 178.7-180 °C

3.2.2. Synthesis of 2-(4’-[18F]fluorophenyl)-1,3-benzothiazole (3.19)

3.2.2.1. Production and isolation of [18F]fluoride After proton irradiation of 500 µl of [18O]water with 10-MeV protons for 50 min in a cyclotron at 20 µA, the content of the target (a solution of [18F]fluoride in [18O]water) was passed over an anion exchange column (Waters Accell™ Plus QMA, Sep-Pak® cartridge (CO 32 − form by rinsing the cartridge with 5 ml of a 0.5 M K2CO3 solution, followed by two rinsing steps with 5 ml water each)). This cartridge captures [18F]fluoride, while the [18O]water is recovered. The anion exchange cartridge was rinsed with 1 ml of water and in the next step [18F]fluoride was eluted from the cartridge with 750 µl of a solution of K2CO3/Kryptofix® 2.2.2. (260 mg Kryptofix® 2.2.2. and 23 mg K2CO3 in a mixture of 6.65 ml acetonitrile and 0.35 ml of water). The eluate was transferred to a reactor vial and heated at 115 °C for 6 min under a stream of nitrogen to evaporate the water and acetonitrile. After addition of 1 ml of acetonitrile, the mixture was again evaporated to dryness under a stream of nitrogen at 115 °C for 5 minutes, leaving a residue of K[18F]F-kryptofix in dry form. 3.2.2.2. Synthesis of 2-(4’-[18F]fluorophenyl)-1,3-benzothiazole 3.19 To the reaction vial containing the K[18F]F-kryptofix complex (3.2.2.1), 3.1 (1.5 mg) in 500 µl anhydrous DMSO was added and the closed vial was heated for

100

Chapter 3

20 min at 150 °C187,188,189. The mixture was allowed to cool down to room temperature and was purified using RP-HPLC on a Xterra™ MS C18 3.5 µm column (2.1 mm x 50 mm) (Waters), eluted with an isocratic mixture of 0.05 M ammonium acetate and tetrahydrofuran/ethanol (75:25, v/v) (50:50, v/v) at a flow rate of 1 ml/min. The radiolabelling yield was 38 % and 3.2 was used to identify the labelled compound 3.19 by comparing their retention times on RPHPLC.

3.2.3. Synthesis

of

2-(4’-[11C]methylaminophenyl)-7-hydroxy-1,3-

benzothiazole (3.20)

3.2.3.1. [11C]Methyltriflate ([11C]CH3OSO2CF3) A volume of 10 ml N2-gas (containing 5 % O2) was irradiated in a cyclotron with 10-MeV protons for 30 min at a beam current of 25 µA. The target gas was transferred to a 5.5 cm lead shielded box in which [11C]CO2 was captured in a loop immersed in liquid nitrogen. The loop was allowed to warm up to room temperature and was flushed with a mixture of N2 and H2 (95:5, v/v) that was directed over a NiO-column (heated at 350 °C) for the reduction of [11C]CO2 to [11C]CH4 and then over a Porapak Q column (Waters) immersed in liquid nitrogen to capture the [11C]CH4 gas. This column was then rinsed with a continuous stream of helium-gas and was allowed to warm up. The [11C]CH4-gas was released from the column and circulated for 5-6 min through a closed loop consisting of a column (diameter 12 mm, length 3 cm) filled with iodine pearls (1-3 mm diameter, Aldrich) heated at 100 °C, followed by a quartz column (diameter 12 mm, length 20 cm) at 620 °C, where [11C]CH4 is transformed into [11C]CH3I. The resulting [11C]CH3I was then captured on a Porapak Q column at room temperature, whereas the unreacted [11C]CH4 was recirculated. After sufficient conversion of [11C]CH4 to [11C]CH3I, the Porapak Q column with the [11C]CH3I was heated with a heat gun to approximately 200 °C and with a stream of He the [11C]CH3I was directed over a column (diameter 5 mm, length 15 cm) packed with silver triflate absorbed on graphite

Development of thioflavin-T derivatives labelled with a positron emitter

101

(heated at 180 °C) yielding [11C]CH3OSO2CF3. A schematic representation of this synthesis route is given in Figure 32.

I2 oven 100 °C

Oven 625 °C

I2 condenser

manometer

ascarite

CH3I/CH3OTf out

pump AgOTf

porapak Q column flow control

180 °C heat gun CV5

flow meter

CV4

waste porapak loop CV3

CV2 C V2

11CH 4

waste lift

N2 liq

in

He in CV1

Figure 32. Schematic diagram of the components of the [11C]CH3OSO2CF3 synthesis module (CVx are direction-guiding valves; ascarite is a drying agent)

3.2.3.2. Synthesis

of

2-(4’-[11C]methylaminophenyl)-7-hydroxy-1,3-

benzothiazole (3.20) With a stream of He, [11C]CH3OSO2CF3 (3.2.3.1) was bubbled through 500 µl of a solution of 3.7 (1 mg/ml in acetone) at room temperature for 30 seconds. The reaction mixture was diluted with 5 ml of water and passed over a SepPak® light C18 cartridge (Waters) (first activated with 5 ml methanol and rinsed two times with 5 ml water). The cartridge was rinsed with 5 ml of water and 3.20 was eluted from the cartridge with 750 µl of methanol, after which the eluate was applied on an Econosphere C18 column (250 mm x 10 mm; 10 µm, Alltech) which was eluted with in isocratic mixture of 0.05 M ammonium

102

Chapter 3

acetate/acetonitrile (50:50) at a flow rate of 3 ml/min. The peak containing 3.20 was identified using radiometric and UV detection (254 nm), and isolated.

3.2.4. Synthesis

of

2-(4’-[11C]methylaminophenyl)-5-hydroxy-1,3-

benzothiazole (3.21)

The same procedure as described in 3.2.3.2 was used, but now 500 µl of a solution of 3.13 (1 mg/ml in acetone) was used for labelling. The obtained reaction mixture of 3.21 was purified as described in 3.2.3.2.

3.2.5. Synthesis of 2-(4’-aminophenyl)-1-N-[11C]methyl-1,3-benzimidazole (3.22)

The same procedure as described in 3.2.3.2 was used, but now 500 µl of a solution of 3.17 (1 mg/ml in acetonitrile) was used for labelling. The obtained reaction mixture of 3.22 was used as such for RP-HPLC purification with an isocratic system of acetonitrile/0.05 M ammonium acetate (30:70, v/v) at a flow rate of 1 ml/min and an Xterra™ MS C18 3.5 µm column (2.1 mm x 50 mm) (Waters).

3.2.6. Analysis of PET tracers with radio-LC-MS

Contrary to the situation with technetium-99m labelled tracer agents, the mass amount of carbon-11 or fluorine-18 labelled tracers is high enough for LC-MS analysis. The labelled compounds 3.20, 3.21 and 3.22 were henceforth used as such for molecular ion mass determination.

Development of thioflavin-T derivatives labelled with a positron emitter

103

3.3. Results and discussion

3.3.1. Synthesis

Whereas synthesis of technetium-99m labelled tracer agents requires the introduction of a rather large technetium binding ligand, the synthesis of tracers for PET-studies can mostly be done without significant structural modifications. It is evident that these different methods of radiolabelling will affect the biological characteristics of the resulting tracer agents in a different way.

3.3.1.1. Synthesis of 2-(4’-nitrophenyl)-1,3-benzothiazole

Synthesis of 2-(4’-nitrophenyl)-1,3-benzothiazole (3.1) was performed using the procedure as described by Shi et al138 with small modifications. The reaction between 4-nitrobenzoic acid and 2-aminothiophenol in the presence of PPA was performed at 150 °C instead of 220 °C, taking into account the experience that lower yields were obtained at the higher temperature. Although the yield of the synthesis was rather low, it was, however, as mentioned before, sufficient for our purposes. The scheme of the synthesis reaction is depicted in Figure 33.

NH2 +

SH

HOOC

NO2

PPA 150 °C, 1 h

N NO2 S

3.1

Figure 33. Synthesis of 2-(4’-nitrophenyl)-1,3-benzothiazole

3.3.1.2. Synthesis of 2-(4’-fluorophenyl)-1,3-benzothiazole

For the synthesis of 2-(4’-fluorophenyl)-1,3-benzothiazole (3.2) the method as described by Klunk and co-workers144 was used. The starting material was a commercially available aromatic acyl chloride, i.e. 4-fluorobenzoyl chloride,

104

Chapter 3

instead of the 4-fluorobenzoic acid. This method was found to be more appropriate than the method with PPA which suffers from the high viscosity of PPA and renders the homogeneity of the reaction mixture more difficult to obtain. This yielded 3.2 in high yield. The scheme of the synthesis of 2-(4’fluorophenyl)-1,3-benzothiazole is depicted in Figure 34.

NH2 +

SH

ClOC

F

toluene p-toluenesulphonic acid Dean-Starck trap, 2 h reflux

N F S

3.2

Figure 34. Synthesis of 2-(4’-fluorophenyl)-1,3-benzothiazole

Since

18

F-labelled 3.2 is not detectable by ESI LC-MS (in the negative

nor in the positive mode), its identity was confirmed using the wellcharacterized non-radioactive authentic product 3.2, through comparison of the retention times on RP-HPLC of the 18F- and 19F-compounds (3.19 and 3.2, respectively).

3.3.1.3. Synthesis of 2-(4’-aminophenyl)-7-hydroxy-1,3-benzothiazole (3.7)

The synthesis of 2-(4’-aminophenyl)-7-hydroxy-1,3-benzothiazole (3.7) (and of 2-(4’-aminophenyl)-5-hydroxy-1,3-benzothiazole (3.13)) caused unexpected problems when the pathway as described by Shi et al138 for similar molecules was used. The reaction of m-anisidine with p-nitrobenzoyl chloride followed by conversion of the amide to a thioamide, proceeded efficiently. However, the subsequent ring closure using potassium ferricyanide, yielded a mixture of two phenylbenzothiazole isomers with the methoxy substituent in respectively the 7-position or 5-position (Figure 35).

Development of thioflavin-T derivatives labelled with a positron emitter H3CO

H3CO

Pyridine

NH2

+

ClOC

NO2

O NH

NO2

Reflux

Dioxane Lawesson's reagent Reflux, 4 h

2-(4'-nitrophenyl)-5-methoxy-1,3-benzothiazole

H3CO

105

N NO2 K3Fe(CN)6, NaOH, EtOH, H2O

S +

H3CO

S NH

NO2

90 °C, 3 h

N NO2 S OCH3 2-(4'-nitrophenyl)-7-methoxy-1,3-benzothiazole

Figure 35. Synthesis of a mixture of 2-(4’-nitrophenyl)-5-methoxy-1,3benzothiazole

and

2-(4’-nitrophenyl)-7-methoxy-1,3-benzothiazole.

The

arrows show the possible ring closure systems.

Separation of these structure isomers proved too difficult to obtain satisfactory results. Attempts to perform the chromatographic separation of the isomers after reduction of the nitro group and after demethylation, also failed. The further exploration of literature revealed that this problem can be avoided using a modified procedure as described by Hutchinson190. When a halogen atom (typically chlorine or bromine) is introduced in the position were the ring closure should occur and NaH or NaOMe is used in combination with NMP as a solvent, then the ring closure only takes places in the desired position. Using this procedure, we were able to obtain 2-(4’-aminophenyl)-5methoxy-1,3-benzothiazole

(3.7)

and

2-(4’-aminophenyl)-7-methoxy-1,3-

benzothiazole (3.13) in pure form. The yields of some of the intermediate steps were not as high as expected, but nevertheless a sufficient amount could be obtained.

For the synthesis of 2-(4’-aminophenyl)-7-hydroxy-1,3-benzothiazole (3.7) the starting compound 2-bromo-3-aminoanisole was obtained according to a reported procedure191. The 2-bromo-3-aminoanisole was reacted with p-

106

Chapter 3

nitrobenzoyl chloride to form an amide (57 % yield), followed by the conversion of the amide to a thioamide with the aid of Lawesson’s reagent. Lawesson’s

reagent

is

2,4-bis(4-methoxyphenyl)-1,3-dithia-2,4-

diphosphetane-2,4-disulphide and is useful as a thiation reagent to substitute the carbonyl oxygen of ketones, amides and esters with sulphur. This procedure yielded 3.4 in high yield (96 %). Next, the benzothiazole ring was closed with NaH in NMP (an inert solvent with a high boiling point), which yielded 3.5 in a 63 % yield. Reduction of the nitro group with stannous chloride in boiling ethanol, resulted in 3.6 in a 62 % yield. The removal of the O-methyl group with boron tribromide in dichloromethane, gave low yields, since the deprotected mixture had to be purified with MPLC. However, a sufficient amount of product could be obtained to perform labelling procedures and characterization using MS and NMR. The MPLC purified product 3.7 was used as such for labelling reactions. The successive reaction steps for the synthesis of 3.7 are depicted in Figure 36.

H2N

H3CO

Br OCH3 + ClCO

NO2

Br O

Pyridine

NO2

NH 3.3

Reflux, 3 h

Dioxane Lawesson's reagent Reflux, 4 h

NMP = N-methylpyrrolidone RT = room temperature

H3CO N NO2 S

Br S

NMP, NaH

NO2

NH

140 °C, 2 h

3.5

3.4

OCH3 Ethanol SnCl2.2H2O Reflux, N2 N

N

BBr 3, CH2Cl2

NH2

NH2 S OCH3

S

RT 3.6

Figure 36. Synthesis of 3.7

OH

3.7

Development of thioflavin-T derivatives labelled with a positron emitter

107

3.3.1.4. Synthesis of 2-(4’-methylaminophenyl)-7-hydroxy-1,3-benzothiazole (3.8)

A small amount of 3.8 was synthesized by methylation of 3.7 with iodomethane (Figure 37). Mass spectrometric analysis of the reaction mixture showed

the

presence

of

the

N-dimethylated

product

next

to

N-

monomethylated 3.8, whereas RP-HPLC analysis further showed that the obtained compound had a different retention time than the methoxy-isomer (3.6, see 3.3.1.3), indicating that methylation occurred at the amino group (comparison of the retention time of 3.8 with 3.6). O-methylation normally requires deprotonation of the phenol and thus more alkaline reaction conditions. Purification was done using MPLC, but the amount of purified 3.8 was too low to allow NMR analysis. Mass spectrometric analysis however confirmed the expected molecular ion mass

N

CH3I

N

Acetonitrile 70 °C

S

NH2 S OH

Figure

3.7

37.

NHCH3 OH

Synthesis

of

3.8

2-(4’-methylaminophenyl)-7-hydroxy-1,3-

benzothiazole (3.8)

3.3.1.5. Synthesis of 2-(4’-aminophenyl)-5-hydroxy-1,3-benzothiazole (3.13)

Synthesis of 2-(4’-aminophenyl)-5-hydroxy-1,3-benzothiazole (3.13, Figure 38) was done in a manner comparable to that of 3.7. However, the commercially available hydrochloric salt of 6-chloro-m-anisidine was used as starting product.

108

Chapter 3

H3CO

H3CO NH2

+

ClOC

NO2

O

Pyridine

NO2

NH

Reflux, 3 h Cl

Cl

3.9 Dioxane Lawesson's reagent Reflux, 4 h

H3CO

H3CO

N NO2

S

NMP, NaOMe

NH

NO2

110 °C, 12 h

S

3.11

Cl

3.10

Ethanol SnCl2.2H2O Reflux, N2

H3CO

BBr 3, CH2Cl2

N

HO

N

NH2

NH2 RT

S

3.12

S

3.13

Figure 38. Synthesis of 2-(4’-aminophenyl)-5-hydroxy-1,3-benzothiazole

This hydrochloric salt was reacted with p-nitrobenzoyl chloride to obtain amide 3.9 in a much higher yield than in the synthesis of 3.3 (78 % as compared to a 57 % yield). The following steps included the transformation into a thioamide with Lawesson’s reagent (3.10), ring closure with NaOMe in NMP (3.11, 5 % yield), reduction of the nitro group with stannous chloride (3.12, 77 % yield) and removal of the O-methyl group with BBr3 (3.13, 76 % yield).

3.3.1.6. Synthesis of 2-(4’-methylaminophenyl)-5-hydroxy-1,3-benzothiazole (3.16)

To avoid a difficult MPLC separation of the mono- and dimethylamino derivative resulting from the methylation reaction, 3.16 was prepared using the method described by Lin202. The benzothiazole ring of the commercially available 5-methoxy-2-methyl-benzothiazole was opened by treating it with 10

Development of thioflavin-T derivatives labelled with a positron emitter

109

M NaOH in order to obtain 2-amino-4-methoxy thiophenol (3.14). After that, 3.14 was reacted with 4-(methylamino)benzoic acid in PPA to yield 3.15 which was purified by column chromatography. The methoxy group was then removed in the classical way with boron tribromide, which yielded 3.16 in a 59 % yield. The final compound 3.16 was purified by extraction, followed by recrystallization. In Figure 39 a schematic representation of the synthesis route is depicted.

H3CO

N CH3 S

H3CO

Ethylene glycol

3.14

10 M NaOH

PPA 180 °C, 3 h HO

N

BBr3, CH2Cl2

H3CO

NH2 SH HOOC

NHCH3

N NHCH3

NHCH3 S

S

RT 3.16

3.15

Figure 39. Synthesis route of 2-(4’-methylaminophenyl)-5-hydroxy-1,3benzothiazole

3.3.1.7. Synthesis of 2-(4’-aminophenyl)-1,3-benzimidazole (3.17)

Synthesis of 2-(4’-aminophenyl)-1,3-benzimidazole was done according to a reported procedure192. Instead of 2-aminothiophenol, o-phenylenediamine was used as the aromatic amine, which was coupled to p-aminobenzoic acid in PPA in a 17 % yield. The reaction scheme of the preparation of 3.17 is shown in Figure 40. The much lower yield than described in literature was probably due to stirring problems during synthesis, resulting in a nonhomogenous reaction mixture.

NH2

+ NH2

HOOC

NH2

PPA 180 °C, 5 h

N NH2 NH

Figure 40. Synthesis of 2-(4’-aminophenyl)-1,3-benzimidazole

3.17

110

Chapter 3

3.3.1.8. Synthesis of 2-(4’-aminophenyl)-1-N-methyl-1,3-benzimidazole (3.18)

This synthesis was performed in the classical way with PPA starting from Nmethyl-1,2-phenylenediamine and gave 3.18 in a reasonable yield (Figure 41).

NH2 +

HOOC

NH2

PPA 180 °C, 2 h

NH

N NH2 N CH3

CH3

3.18

Figure 41. Synthesis of 2-(4’-aminophenyl)-1-N-methyl-1,3-benzimidazole

3.3.2. Synthesis of 2-(4’-[18F]fluorophenyl)-1,3-benzothiazole

3.3.2.1. Production of [18F]fluoride [18F]fluoride was produced in the same way as in the preparation of routinely used PET tracers. Depending on the irradiation time (mostly from 50 min to 1 h) and the ion beam current used, the activity of [18F]fluoride produced varied, but resulted in enough [18F]fluoride to perform one or more labelling reactions. By adsorbing the [18F]fluoride on an anion exchange resin, the residual [18O]water can be recovered. This is done via automated procedures to prevent the manipulator from receiving high radiation doses. The [18F]fluoride is then eluted from the ion exchange column with a solution containing a mixture of potassium carbonate/Kryptofix® 2.2.2. in acetonitrile/water. Kryptofix® 2.2.2. is a so-called crown ether that is used to keep the [18F]fluoride in solution in an organic solvent like acetonitrile. Potassium carbonate is used to provide potassium as a counter ion for [18F]fluoride. The eluate containing the K[18F]fluoride has to be dried thoroughly to remove all water, which would negatively affect subsequent nucleophilic substitution reactions. For the same reason, only anhydrous solvents can be used for these

substitution

reactions

using

the

potassium-[18F]fluoride-kryptofix

complex. After removal of the solvents, the remaining complex of K[18F]fluoride-kryptofix was used as such for labelling of the precursor.

Development of thioflavin-T derivatives labelled with a positron emitter

111

3.3.2.2. Synthesis of 2-(4’-[18F]fluorophenyl)-1,3-benzothiazole (3.19)

To the K[18F]fluoride-kryptofix complex, a solution of 3.1 in DMSO was added. Substitution of a

18

F-atom for the aromatic nitro group was done by heating

this solution of the precursor with the K[18F]kryptofix complex in DMSO for 20 min at 150 °C. Nucleophilic substitution of aromatic nitro groups with [18F]fluoride only occurs if other electron withdrawing substituents are present on the ortho or para position. The presence of the benzothiazole substituent apparently activates the 4’-nitro group in a sufficient way as relatively high yields of the corresponding 2-(4’-[18F]fluorophenyl)-1,3-benzothiazole were obtained.

The reaction mixture was purified using RP-HPLC. Eluent systems consisting of binary mixtures of acetonitrile, methanol or ethanol with buffer were not able to separate 3.19 from its precursor 3.1. A number of other solvent systems were therefore evaluated and the ternary mixture consisting of acetonitrile, tetrahydrofuran and 0.05 M ammonium acetate buffer yielded the best separation of 3.19 and 3.1. RP-HPLC showed a radiolabelling yield of 38 % (not decay corrected) and identification of the radiolabelled 3.19 was done by comparison of the retention times of the authentic analogue 3.2 and the

18

F labelled 3.19 on RP-HPLC (because 3.19 cannot be ionized, see

3.3.1.2). The RP-HPLC chromatogram of the labelling reaction mixture of 3.19 is depicted in Figure 42.

112

Chapter 3 GU Y : D V1 0 11 0 9 . R0 2 - Pu ls e 1

CPS

A

3.19

1.2 e5

8.0 e4

4.0 e4

0.0 e0 0' 0 0

5' 0 0

10 ' 0 0

15 ' 0 0

20 ' 0 0

m

GU Y : D V1 0 11 0 9 . R0 2 - An a l og 1

mV

B

8 00

3.1

4 00

0 0' 0 0

5' 0 0

10 ' 0 0

15 ' 0 0

20 ' 0 0

m

Figure 42. RP-HPLC chromatogram of the labelling reaction mixture of 3.19. A = radioactive channel, B = UV channel (254 nm)

In the UV signal the peak at 14’30’’ is the precursor 3.1, while the peak at 12’40’’ in the radioactive channel is the labelled compound 3.19. The radioactive peaks at the beginning of the run are a combination of [18F]fluoride and one or more (other) polar products. The peaks in the beginning of the run on the UV channel were non-identified impurities. In this RP-HPLC system the labelled compound 3.19 is somewhat more hydrophilic than the precursor, as can be seen in the difference in retention time. The octanol/buffer partition coefficient P for 3.19 was 718.8 ± 75.1 (log P = 2.86 ± 0.045).

3.3.3. Synthesis

of

2-(4’-[11C]methylaminophenyl)-7-hydroxy-1,3-

benzothiazole (3.20)

3.3.3.1. Synthesis of [11C]CH3OSO2CF3 [11C]CH3OSO2CF3 is a 11C-labelled alkylating agent which is a frequently used precursor to prepare a number of carbon-11 labelled radiopharmaceuticals used in clinical studies. The amount of [11C]CH3OSO2CF3 produced varies according to the irradiation time (generally 30 to 40 minutes) and the ion beam

Development of thioflavin-T derivatives labelled with a positron emitter

113

current used to prepare [11C]CO2 in the cyclotron. The production of [11C]CH3OSO2CF3 consists of three steps which are all performed in gas phase. A failure in one of these steps results in capturing the radioactivity in the system and necessitates waiting until the radioactivity has decayed sufficiently. The reduction of [11C]CO2 to [11C]CH4 gas is followed by the iodination of [11C]CH4, yielding [11C]CH3I and finally [11C]CH3I is transformed to [11C]CH3OSO2CF3. The whole system is located in a lead shielded (5.5 cm) cabinet and is remote-controlled to avoid the accumulation of a high radiation dose to the manipulator.

3.3.3.2. Synthesis

of

2-(4’-[11C]methylaminophenyl)-7-hydroxy-1,3-

benzothiazole (3.20) 11

C-methylation of precursor 3.7 was done in a 7 % yield (not decay

corrected) by bubbling [11C]CH3OSO2CF3 through a solution of the precursor. The low yield provided enough 3.20 to perform the biological studies and no attempts were made to increase the yield. Pre-purification of the reaction mixture was done with the aid of a C-18 Sep-Pak® cartridge (which retained 3.20, whereas more hydrophilic compounds (like [11C]methyltriflate) were eluted from the cartridge). After rinsing of the cartridge with water, the labelled compound 3.20 was eluted from the cartridge with methanol. The eluate containing 3.20 was further purified by RP-HPLC on a semi-preparative C-18 column, eluted with an isocratic system with a high percentage of acetonitrile in the mobile phase, as to assure a rapid elution of the labelled compounds in view of the short half-life of carbon-11. The use of a semi-preparative column was necessary to inject the entire purified reaction mixture at once. The retention time of 3.20 was about 10 min. The RP-HPLC chromatogram of the purified 3.20 is shown in Figure 43. The peak of precursor 3.7 in the UVchannel is split because of an overloading effect. When peaks containing the desired

11

C labelled benzothiazole were isolated from the RP-HPLC run, care

was taken to start the collection of the peak only after the UV-signal had returned to baseline, in order to prevent contamination of the carbon-11 labelled benzothiazole by the non-radioactive precursor. As could be expected, the C-11 methylated 3.20 has a longer retention time than 3.7. The

114

Chapter 3

octanol/water partition coefficient of 3.20 is 280.70 ± 63.60 (log P = 2.45 ± 0.096). Identification of the radiolabelled 3.20 was done by comparison of the retention times of the authentic analogue 3.8 and the 11C labelled 3.20 on RPHPLC. Radio-LC-MS showed a molecular ion mass of 255.3162 Da (theoretical mass: 255.0598 Da) in ES- mode (data not shown).

TOM : 0 5 04 2 0 0 4. R 0 4 - P ul s e 1

CPS

3.20

60 00 0

A

40 00 0

20 00 0

0 0' 0 0

5' 0 0

10 ' 00

15 ' 00

20 ' 00

2 5 ' 00

30 ' 00

m

30 ' 00

m

TOM : 0 5 04 2 0 0 4. R 0 4 - Ana lo g 1

mV 6 00

B

3.7

4 00

2 00

0 0' 0 0

5' 0 0

10 ' 00

15 ' 00

20 ' 00

2 5 ' 00

Figure 43. RP-HPLC chromatogram of 3.20, A = radiometric channel, B = UVchannel (254 nm). For conditions, see the experimental part.

3.3.4. Synthesis

of

2-(4’-[11C]methylaminophenyl)-5-hydroxy-1,3-

benzothiazole (3.21)

Starting from 3.13, the same procedure as in 3.3.3 was followed to prepare 3.21 that was obtained with a labelling yield of 5 % (not decay corrected). The radiochromatogram of the tracer agent obtained by HPLC on a semipreparative column, is shown in Figure 44.

Development of thioflavin-T derivatives labelled with a positron emitter

115

TOM : 06 0 4 2 0 0 4 . R 02 - P ul s e 1

CPS

A

3.21 80 00

40 00

0 0' 0 0

5' 0 0

10 ' 00

15 ' 00

20 ' 00

25 ' 00

m

TOM : 06 04 20 04. R 02 - Ana lo g 1

mV

3.13

8 00

B 6 00 4 00 2 00 0 0' 0 0

5' 0 0

10 ' 00

15 ' 00

20 ' 00

25 ' 00

m

Figure 44. RP-HPLC chromatogram of 3.21; A = radiometric channel, B = UVchannel (254 nm). For conditions, see the experimental part.

Log P value of 3.21 is 2.48 ± 0.063 (P = 302.44 ± 43.30). Identification of the radiolabelled 3.21 was done by comparing the retention times of the authentic analogue 3.16 and the

11

C-labelled 3.21 on RP-HPLC.

Radio-LC-MS analysis of 3.21 could not determine the mass of the peak in the radiometric channel due to the low labelling yield coupled to the small injection volume (typically 5-10 µl) on the LC-MS system.

3.3.5. Synthesis of 2-(4’-aminophenyl)-1-N-[11C]methyl-1,3-benzimidazole (3.22)

The same procedure as described in 3.3.3 was followed to label 3.17, but the precursor was dissolved in acetonitrile. The C-11 methylated 3.22 was obtained in a 74 % yield (not decay corrected) and because of this high labelling yield, a Sep-Pak pre-purification of the reaction mixture was not necessary. Furthermore, only a small fraction of the labelling reaction mixture was injected on an analytical column for purification (also the case for 3.19), whereas for the other radiolabelled agents (3.20 and 3.21), the low labelling

116

Chapter 3

yield

necessitated

the

use

of

a

semi-preparative

column.

The

radiochromatogram obtained by RP-HPLC analysis is shown in Figure 45.

TOM : TOV85 C1 1. R 1 1- Pu l se 1

CPS 12 00 0

3.22

A

80 00

40 00

0 0' 0 0

5' 0 0

10 ' 00

15 ' 00

20 ' 00

25 ' 00

30 ' 00

m

TOM : TOV8 5C1 1 . R 11 - An a l og 2

mV

3.17

4 00

B

3 00

2 00

1 00

0 0' 0 0

5' 0 0

1 0 ' 00

1 5 ' 00

2 0 ' 00

2 5 ' 00

30 ' 00

m

Figure 45. RP-HPLC chromatogram of the labelling reaction mixture of 3.22; A = radiometric channel, B = UV-channel (254 nm)l. For conditions, see the experimental part.

N-methylation of the benzimidazole 3.17 can theoretically happen on the aromatic amine nitrogen atom or on a nitrogen atom of the benzimidazole ring. Comparison of the RP-HPLC retention time of the

11

C labelled reaction

product with that of the two possible N-methylated non-radioactive analogues, showed that the

11

C-methyl substituent was introduced on the imidazole

nitrogen atom and not on the aromatic amine. This is probably due to the rather acidic properties of this benzimidazole amine.

The octanol/water partition coefficient P of 3.22 was found to be 95.74

± 6.24 (log P = 1.98 ± 0.029). The benzimidazole 3.22 is also slightly more hydrophilic than the three benzothiazoles, which is reflected in their log P values (about 2 for 3.22 and ~2.5 or higher for 3.19, 3.20 and 3.21).

Development of thioflavin-T derivatives labelled with a positron emitter

117

Identification of the radiolabelled 3.22 was done by comparison of the retention times of the authentic analogue 3.18 and the

11

C labelled 3.22 on

RP-HPLC. Radio-LC-MS showed a molecular ion mass of 224.2102 Da (theoretical mass: 224.1182 Da) in ES+ mode (data not shown).

3.4. Conclusion In this chapter the synthesis and analysis of four phenylbenzothiazoles labelled with carbon-11 or fluorine-18 has been described. Synthesis yields were sufficient enough to allow labelling of the precursors with fluorine-18 or carbon-11. The yield of the radiolabelling of 3.1 was high enough to perform in vivo experiments (see chapter 4) and the half-life of fluorine-18 makes the use of 3.19 promising for in vivo experiments. Radiolabelling of 3.7 and 3.13 was done in low yields and this made it difficult to perform LC-MS analysis and in vivo tests. Nevertheless, because of the use of the corresponding carbon-12 or fluorine-19 products, it was possible to perform in vitro binding affinity studies (see chapter 5). Although the radiolabelling yield was low, we managed to perform in vivo studies (see chapter 4) as well. The cold product 3.2 was also used for in vitro experiments. The carbon-11 labelled benzimidazole 3.22 was labelled in the benzimidazole part of the molecule instead of on the aromatic amine (which had been expected). Due to the higher polarity of this compound and its accordingly lower log P value of ~2, this tracer agent carries the best promise for high uptake in the brain and fast wash-out of the brain and the body. The question remains of course which effect substitution of a thioether by an amine will have on the affinity for amyloid plaques. It can, however, be assumed that the mere introduction of a carbon-11 or fluorine-18 radionuclide in these thioflavin-T derivatives will have little effect on their affinity for amyloid plaques, contrary to the technetium-99m labelled agents where the introduction of a bulky technetium binding ligand is required.

119

CHAPTER 4

4. BIODISTRIBUTION

STUDIES

OF

THE

RADIOLABELLED THIOFLAVIN-T DERIVATIVES

4.1. Introduction The way radiolabelled tracer agents behave in a biological system is of utmost importance for their usefulness and efficacy. It is therefore important to know the tissue distribution of the tracers, together with data concerning their plasma clearance and excretion from the body through the kidneys and the hepatobiliary system. Furthermore, because amyloid plaques are deposited in specific body regions, it is obvious that tracer agents intended for visualization of such plaques, should have the ability to reach the target tissue and to bind there to the amyloid plaques. This means that tracer agents for detection and visualization of AD should be able to cross the BBB and show a high affinity for amyloid plaques. On the other hand, radiopharmaceuticals for visualization of PA in abdominal organs should show a fast clearance from the abdomen to assure a high target to non-target ratio for affected organs in this region.

Tracers for diagnostic studies in humans should be labelled with shortlived radionuclides that emit γ-rays with an energy in the range of 100-200 keV without emission of particular radiation (α or β). Technetium-99m, with its half-life of 6 h and its γ-rays of 141 keV, fully meets these requirements. Moreover, its continuous availability by the use of a

99

Mo/99mTc-generator is

an additional advantage. The positron emitting radionuclides fluorine-18 and carbon-11 are less favourable than technetium-99m in terms of biological effects of the positrons, but because of their short half-life and because of the specific characteristics of positron emitters (i.e. the annihilation gamma radiation), they are of great importance in nuclear imaging. Also the biological

120

Chapter 4

half-life of the radiolabelled tracer is very important as it is one of the factors determining the absorbed radiation dose resulting from the administration of radiopharmaceuticals. It is therefore necessary to study the biological half-life before taking a tracer into the clinic.

The aim of this study was to develop radiolabelled tracer agents for a non-invasive in vivo visualization of AD as well as PA. Tracer agents for diagnosis of AD have to be able to cross the BBB. The BBB is a highly structured barrier that is necessary to protect brain cells from potential nocuous agents. The extracellular fluid in the brain has a specific composition that is maintained within precise limits, which is independent of the composition of the circulating blood. These conditions are necessary to support a suitable environment in which neuronal activity can take place. The so-called continuous capillaries in the brain have a single layer of endothelial cells (from one to several) and these are connected with tight junctions. No discontinuities are found in the capillary wall, which is surrounded by a continuous basal lamina193. The presence of these tight junctions is the main factor in maintaining the impermeability of the BBB because they prevent aqueous paracellular transport of hydrophilic molecules between the cells. Furthermore, these brain capillaries also show no pores. This is also the reason why there are different transport mechanisms to ensure that the brain receives al the necessary compounds to sustain its activity. This can occur through passive diffusion and/or active transport194. Passive diffusion can occur across the endothelium, through the cells themselves or through the tight junctions between the cells. In case of active transport, there is the necessity of carriers that use energy (mostly adenosine triphosphate) to transport certain molecules across this BBB. This is for example the case with glucose, the major energy source of the brain195. On the other hand, passive transport across the endothelium is only possible for compounds with appropriate characteristics. These characteristics relate to lipophilicity, molecular mass, size, charge and the percentage of substance ionized at physiological pH. Of course, the cerebral blood flow and binding of a compound to plasma proteins, as well as its possible affinity for endogenous carriers, also plays a role in the net passage across the BBB.

Biodistribution studies of the radiolabelled thioflavin-T derivatives

121

The solubility of a compound in a lipid bilayer partially depends on its lipophilicity. Lipophilicity is used to describe the affinity for a lipid phase. Water soluble molecules as a rule are lipophobic while compounds that dissolve in organic solvents are more lipophilic and will therefore have more chances of penetrating a lipid bilayer in an in vivo system. The partition coefficient P of a compound reflects its lipophilicity and is determined by partitioning a compound between two immiscible phases, one being aqueous and the other mostly 1-octanol. Dischino et al196 found that carbon-11 labelled compounds with a log P value between 0.9 and 2.5 could freely cross the BBB. When the log P value increases to 2.7, the product is still able to cross the BBB, but with time it concentrates in the white matter as a result of its lipophilicity. Furthermore, a parabolic relationship was found between the log P value of these carbon-11 labelled compounds and the fraction entering the brain. When the log P value increases above 2.7, the uptake in the brain is much lower because of its binding to blood components and macromolecules, a binding that is based on hydrophobic bonding. The ideal log P value for neutral molecules for passive diffusion across the BBB lies around 2.

Another important factor for the passage of a compound over the BBB is its molecular mass. Ideally, this mass should be in the 400 to 700 Da range. The high molecular mass is also the reason why most proteins cannot pass the BBB.

In the case of tracer agents for visualization of PA, the passage across the BBB is of minor importance. The major issue here is the excretion of the labelled compound out of the biological system. The excretion should be fast in order to limit the background noise in the abdomen. Because amyloid plaques in PA are most often found in the liver and the kidneys (together with other locations), it is impossible to obtain images from the abdomen with suitable target-to-background activity ratios if the tracers are retained in the liver and the intestines. For this reason, these tracer agents should preferentially be charged and have rather hydrophilic properties to promote a fast clearance through the kidneys. In addition to the charge, other factors,

122

Chapter 4

such as the log P value and the binding to plasma proteins, will also play a role in the biological behaviour of these agents.

In this chapter the results will be discussed of the biological evaluation of the tracer agents labelled with technetium-99m and the four compounds labelled with a positron emitter in normal mice.

4.2. Materials and methods

4.2.1. Preparation of radiolabelled thioflavin-T derivatives

Deprotection and labelling of 2.4, 2.12, 2.13 and 2.14 was performed as described in 2.2.3. Labelling of 3.1, 3.7, 3.13 and 3.17 was performed as described in 3.2.2, 3.2.3, 3.2.4 and 3.2.5, respectively. The peak of the desired compound was isolated on RP-HPLC and in case of the technetium99m labelled agents, diluted with saline to obtain an appropriate radioactivity concentration for biodistribution studies. In the case of the carbon-11 and fluorine-18 labelled compounds, the mixture containing the isolated peak was concentrated by bubbling nitrogen-gas through it at 50 °C to evaporate the organic solvent (acetonitrile in most cases). This was necessary to reduce the percentage of organic solvent to a sufficiently low level for injection in mice. The percentage of organic solvents was checked using gas-chromatography (less than 0.1 % of acetonitrile in the final solution).

4.2.2. Tissue distribution in normal mice

The radiometric peak containing the desired labelled compound was isolated on RP-HPLC and diluted with saline until a concentration was reached of 150 kBq/ml for technetium-99m labelled compounds. For carbon-11 labelled compounds the isolated and evaporated peak was diluted with saline to a radioactivity concentration of 37 MBq/ml. Fluorine-18 labelled 3.1 was diluted with saline to a concentration of 3.7 MBq/ml. Male NMRI mice were sedated with an intraperitoneal injection of 75 µl Hypnorm® (fentanyl citrate 63 µg/ml +

Biodistribution studies of the radiolabelled thioflavin-T derivatives

123

fluanisone 2 mg/ml). Each of the different compounds (15 kBq/100 µl in case of

99m

Tc-labelled compounds, 3.7 MBq/100 µl in case of carbon-11 labelled

compound and 370 kBq/100 µl in case of fluorine-18 labelled compounds) was injected via a tail vein in a series of eight mice. The mice were sacrificed by decapitation at 2 or 60 min p.i. (n = 4 at each time point). The different organs and other body parts were dissected and transferred in tared vials and the activity of each vial was measured with the automatic gamma counter. The measured activity was expressed in a percentage of the injected dose per organ (% ID), equal to the sum of the number of counts in all organs, corrected for background activity, and in a percentage of the injected dose per gram of organ (% ID/g). To calculate the activity in the blood, blood mass was assumed to be 7 % of body mass.

4.3. Results and discussion The values for uptake in organs and other body parts reported in Table 5 to Table 9 and in Table 11 to Table 14 (see below) are the result of the i.v. injection of the different tracers in healthy mice followed by their distribution over the different organs and their metabolism. Whether the values at 60 min p.i. are the result of the tracer agent as such or of the tracer agent and one or more of its radiolabelled metabolites has not been assessed a priori, since this is only of importance for the compounds with favourable characteristics, i. e. that effectively possess good BBB penetration or good renal excretion and high affinity for amyloid β.

4.3.1. Derivatives labelled with technetium-99m

The results of the biodistribution studies of the technetium-99m labelled phenylbenzothiazoles 2.4, 2.12, 2.13 and 2.14 in mice at 2 and 60 min p.i. are summarized in Table 5 to Table 9.

124

Chapter 4

Table 5. Tissue distribution of

99m

Tc-2.4 after i.v. injection in normal mice (n =

4 at each time point (TP)) at 2 and 60 min p.i. (s.d. = standard deviation) % ID ± s.d. 2 min p.i.

60 min p.i.

Urine

2.5 ± 1.8

6.9 ± 2.1

Kidneys

4.4 ± 0.84

Liver

% ID/g tissue ± s.d. 2 min p.i.

60 min p.i.

2.2 ± 0.15

7.4 ± 1.1

2.9 ± 0.14

32.6 ± 2.6

19.2 ± 4.1

14.8 ± 2.8

7.5 ± 1.0

Intestines

5.1 ± 1.8

24.2 ± 3.3

Spleen + Pancreas

1.3 ± 0.66

0.38 ± 0.10

2.8 ± 0.71

0.92 ± 0.13

Lungs

1.8 ± 0.45

0.45 ± 0.10

7.0 ± 1.5

1.5 ± 0.46

Heart

0.62 ± 0.08

0.21 ± 0.08

3.5 ± 0.35

7.5 ± 12.9

Stomach

1.1 ± 0.19

5.2 ± 1.4

1.4 ± 0.09

8.3 ± 3.4

Cerebrum

0.09 ± 0.03

0.03 ± 0.00

0.30 ± 0.11

0.10 ± 0.02

Cerebellum

0.11 ± 0.11

0.03 ± 0.00

0.86 ± 0.62

0.21 ± 0.04

Blood

18.8 ± 2.8

5.7 ± 0.71

6.7 ± 1.0

1.7 ± 0.22

Table 6. Tissue distribution of

99m

Tc-tricine-2.12 after i.v. injection in normal

mice (n = 4 at each TP) at 2 and 60 min p.i. % ID ± s.d. 2 min p.i.

60 min p.i.

Urine

0.42 ± 0.33

2.0 ± 0.07

Kidneys

3.5 ± 0.28

Liver

% ID/g tissue ± s.d. 2 min p.i.

60 min p.i.

2.6 ± 0.26

5.4 ± 0.42

3.9 ± 0.22

48.0 ± 1.8

32.6 ± 3.6

19.3 ± 2.7

13.3 ± 2.0

Intestines

5.00 ± 0.36

39.4 ± 3.1

Spleen + Pancreas

0.69 ± 0.13

0.34 ± 0.08

1.6 ± 0.18

0.88 ± 0.15

Lungs

1.3 ± 0.30

0.92 ± 0.22

5.1 ± 0.58

3.2 ± 0.49

Heart

0.45 ± 0.09

0.25 ± 0.03

2.5 ± 0.18

1.4 ± 0.22

Stomach

0.60 ± 0.21

0.96 ± 0.48

0.87 ± 0.20

1.6 ± 1.1

Cerebrum

0.12 ± 0.03

0.05 ± 0.01

0.38 ± 0.07

0.17 ± 0.05

Cerebellum

0.07 ± 0.02

0.03 ± 0.01

0.59 ± 0.19

0.23 ± 0.06

Blood

26.1 ± 1.4

14.9 ± 0.89

9.3 ± 0.53

5.6 ± 0.55

Biodistribution studies of the radiolabelled thioflavin-T derivatives 99m

Table 7. Tissue distribution of

Tc-EDDA-2.12 and

125

99m

Tc-tricine/nicotinic

acid-2.12 after i.v. injection in normal mice (n = 4 at each TP) at 2 and 60 min p.i. Time p.i. 99m

Kidneys

Liver

Intestines

Cerebrum

Cerebellum

Blood

Tc-EDDA-2.12 (% ID ± s.d.)

2

4.1 ± 0.66

41.8 ± 2.7

5.0 ± 0.24

0.12 ± 0.03

0.06 ± 0.01

38.2 ± 3.3

60

1.4 ± 0.28

40.3 ± 5.2

37.8 ± 4.3

0.03 ± 0.01

0.02 ± 0.01

7.3 ± 0.58

99m

Tc-EDDA-2.12 (% ID/g tissue ± s.d.)

2

6.4 ± 0.94

19.0 ± 2.0

0.36 ± 0.06

0.61 ± 0.08

13.7 ± 1.4

60

1.9 ± 0.4

18.0 ± 1.2

0.09 ± 0.02

0.21 ± 0.06

2.5 ± 0.27

99m

Tc-tricine/nicotinic acid-2.12 (% ID ± s.d.)

2

2.7 ± 0.31

62.3 ± 3.9

5.9 ± 2.1

0.05 ± 0.01

0.03 ± 0.01

9.4 ± 1.6

60

0.56 ± 0.21

7.8 ± 2.1

82.0 ± 6.7

0.01 ± 0.00

0.01 ± 0.00

0.66 ± 0.08

99m

Tc-tricine/nicotinic acid-2.12 (% ID/g tissue ± s.d.)

2

4.0 ± 0.44

29.8 ± 1.6

0.13 ± 0.03

0.26 ± 0.13

3.4 ± 0.62

60

1.0 ± 0.18

3.9 ± 1.6

0.02 ± 0.01

0.04 ± 0.02

0.23 ± 0.04

Table 8. Tissue distribution of 99mTc-2.13 after i.v. injection in normal mice (n = 4 at each TP) at 2 and 60 min p.i. % ID ± s.d. 2 min p.i.

60 min p.i.

Urine

0.11 ± 0.04

0.78 ± 0.32

Kidneys

7.6 ± 0.13

Liver

% ID/g tissue ± s.d. 2 min p.i.

60 min p.i.

2.3 ± 0.46

11.0 ± 0.54

3.3 ± 0.61

71.7 ± 1.1

20.4 ± 11.5

31.4 ± 3.2

8.1 ± 4.4

Intestines

3.3 ± 1.2

72.7 ± 12.6

Spleen + Pancreas

0.35 ± 0.06

0.07 ± 0.02

1.1 ± 0.22

0.18 ± 0.06

Lungs

0.60 ± 0.31

0.13 ± 0.08

2.1 ± 0.62

0.40 ± 0.27

Heart

0.28 ± 0.09

0.06 ± 0.02

1.8 ± 0.64

0.36 ± 0.12

Stomach

0.38 ± 0.14

0.47 ± 0.39

0.41 ± 0.11

0.59 ± 0.52

Cerebrum

0.09 ± 0.03

0.06 ± 0.02

0.28 ± 0.08

0.20 ± 0.10

Cerebellum

0.07 ± 0.02

0,05 ± 0.02

0.63 ± 0.19

0.50 ± 0.14

Blood

8.62 ± 1.8

0.90 ± 0.41

3.1 ± 0.70

0,32 ± 0.15

126

Chapter 4

Table 9. Tissue distribution of 99mTc-2.14 after i.v. injection in normal mice (n = 4 at each TP) at 2 and 60 min p.i. % ID ± s.d. 2 min p.i.

60 min p.i.

Urine

0.10 ± 0.12

0.34 ± 0.13

Kidneys

8.8 ± 0.95

Liver

% ID/g tissue ± s.d. 2 min p.i.

60 min p.i.

3.3 ± 0.24

13.0 ± 0.19

5.1 ± 0.37

77.0 ± 0.95

43.3 ± 3.0

36.5 ± 1.6

23.4 ± 1.6

Intestines

2.6 ± 1.3

46.3 ± 4.9

Spleen + Pancreas

0.23 ± 0.11

0.11 ± 0.06

0.70 ± 0.32

0.31 ± 0.06

Lungs

0.44 ± 0.07

0.15 ± 0.01

1.67 ± 0.44

0.58 ± 0.06

Heart

0.12 ± 0.02

0.06 ± 0.04

0.74 ± 0.19

0.35 ± 0.22

Stomach

0.19 ± 0.11

3.04 ± 3.8

0.36 ± 0.27

5.0 ± 5.4

Cerebrum

0.02 ± 0.00

0.01 ± 0.00

0.07 ± 0.02

0.03 ± 0.01

Cerebellum

0.02 ± 0.01

0.01 ± 0.00

0.12 ± 0.05

0.08 ± 0.06

Blood

4.9 ± 0.93

1.6 ± 0.14

1.7 ± 0.40

0.61 ± 0.03

For urine and intestines, the uptake is not given in % ID/g because of the fact that the mass of urine and of the contents of the intestines varies as a function of time. Therefore, a value of a radioactivity concentration in excrements is not really meaningful. The tables list the tracer uptake in the most important organs, but not in body parts such as limbs, head and carcass. For this reason, the sum of the activities in the tables does not equal 100 %. On the other hand, the activity in well-perfused organs such as the liver, may be artificially high as blood was not removed from the organs and blood activity was in some cases high, especially at 2 min p.i.

To allow a better comparison of the tracer uptake and kinetics in different organs as a function of some physicochemical characteristics of the respective tracer agents, the relative molecular mass, the log P and the charge of the four complexes are summarized in Table 10.

Biodistribution studies of the radiolabelled thioflavin-T derivatives Table

10.

Physicochemical

characteristics

127

99m

of

Tc-labelled

phenylbenzothiazoles Labelled Compound

Mr

log P

charge

polar group(s)

Tc-2.4

558

1.92

0

no

Tc-tricine-2.12

748

0.81

0?

OH/COOH of tricine

99m

99m

99m

Tc-2.13

582

-0.49

-1

-COOH

99m

Tc-2.14

579

1.1

-1

no

Although the nature of the Tc-binding ligand in these

99m

Tc-labelled

phenylbenzothiazoles is quite different and also the overall charge of the respective complexes varies, their biological behaviour is similar but not identical.

Surprisingly, none of the compounds is excreted efficiently through the renal system. The efficiency of extraction of a compound from the blood by the kidneys is reflected in the kidney value at 2 min p.i. These values vary from 3-4 % ID for

99m

Tc-2.4 and the HYNIC derivatized benzothiazoles

2.12 to 8-9 % ID for the two negatively charged complexes 99m

99m

99m

Tc-

Tc-2.13 and

Tc-2.14. These low values are solely the result of a low extraction

efficiency and not of a rapid transit into the urine; the urinary activity at 60 min p.i. is almost negligible with a maximum of about 7 % ID for low as 0.34 % ID for complexes

99m

99m

99m

Tc-2.4 and as

Tc-2.14. It is remarkable that the negatively charged

Tc-2.13 and

99m

Tc-2.14 show a lower percentage in the renal

system (kidneys + urine) at 60 min p.i. than at 2 min p.i., an observation consistently made in all animals of these studies. This seems to indicate an initial filtration and/or temporary binding of a fraction of these tracer agents in the kidneys followed by a re-uptake in the blood without significant excretion into the urine. The assumption that the presence of one (99mTc-2.14) or two (99mTc-2.13) negative charges on the complex at physiological pH and thus of more polar characteristics, would promote extraction and excretion by the kidneys, was found invalid. It has to be admitted that this strategy to design a tracer agent with rapid clearance of activity from the abdominal organs via the kidneys, completely failed. It is clear that other factors determine renal

128

Chapter 4

handling of a compound (to a higher degree) than charge and polarity. One of these factors might be the binding to plasma proteins. For other types of compounds however, the strategy of adding one or two carboxylate groups to a

99m

Tc-labelled conjugate of a mercaptoacetyldipeptide (SN3 ligand) and a

biologically active peptide, was found successful for significantly increasing the rate and extent of urinary excretion of the tracer agent197.

In line with the poor renal handling of the studied

99m

Tc-labelled

phenylbenzothiazoles, they are all efficiently taken up by the liver and excreted into the intestines. For these four

99m

Tc-compounds (with the

HYNIC-BTA conjugate 2.12 labelled in three ways), Figure 46 compares their uptake in liver and intestines at 2 and 60 min p.i., expressed in percentage of ID, and compares also the total activity in the hepatobiliary system (liver + intestines) at these time points. Liver uptake (in % ID) at 2 min p.i. varies from 33 % (99mTc-2.4) to as high as 77 % (99mTc-2.14) and decreases at 60 min p.i. to values between 8 % (99mTc-tricine/nicotinic acid-2.12) and 43 % (99mTc-2.14). The clearance of the activity from the liver is most rapid and most complete for 99m

99m

Tc-2.13 and

Tc-tricine/nicotinic acid-2.12. In both complexes, a free carboxylate group

is present in the

99m

Tc-binding ligand moiety, but evidence is lacking to prove

that this is the reason for the efficient transfer from the liver to the intestines. Nevertheless, the high activity in the intestines at 60 min p.i. is striking for these two

99m

Tc-complexes (73 % and 82 % of ID, respectively) as compared

to 25-46 % for the other compounds. Interesting to note is the fact that

99m

Tc-

2.4, a neutral and the most lipophilic of these technetium-99m labelled phenylbenzothiazoles (log P = 1.91) shows the lowest uptake in the hepatobiliary system: its total activity in liver + intestines is at each TP clearly lower than that of the other agents. On the other hand, the anionic and most polar complex, i.e.

99m

Tc-2.13 (log P = -0.49), is concentrated most efficiently

and rapidly in the hepatobiliary system (activity in liver + intestines ranging from 75 % at 2 min p.i. to 93 % at 60 min p.i.).

Biodistribution studies of the radiolabelled thioflavin-T derivatives

129

Liver

% ID

99mTc-2.4 99mTc-2.14

99mTc-tricine-2.12 99mTc-EDDA-2.12

99mTc-2.13 99mTc-tri/nic-2.12

100 80 60 40 20 0 2

60 min

Intestines

% ID

99mTc-2.4 99mTc-2.14

99mTc-tricine-2.12 99mTc-EDDA-2.12

99mTc-2.13 99mTc-tri/nic-2.12

100 80 60 40 20 0 2

60 min

Liver + Intestines

% ID

99mTc-2.4 99mTc-2.14

99mTc-tricine-2.12 99mTc-EDDA-2.12

99mTc-2.13 99mTc-tri/nic-2.12

100 80 60 40 20 0 2

60 min

Figure 46. Uptake of

99m

Tc-labelled phenylbenzothiazoles in liver and

intestines of mice at 2 and 60 min p.i. and the total activity in the hepatobiliary system (liver + intestines) at these time points. (tri/nic = tricine/nicotinic acid)

The general assumption that polar compounds are extracted from the blood mainly by the kidneys and lipophilic compounds by the liver, is clearly not valid for the 99mTc-labelled phenylbenzothiazoles of this study.

130

Chapter 4 It is logic that the anionic complexes

99m

Tc-2.13 and

99m

Tc-2.14 do not

show uptake in the brain (cerebrum and cerebellum) any higher than the background activity due to the presence of the tracer agent in the blood. As a rule, charged and polar compounds do not pass the BBB by passive diffusion. A similar argument can be used to explain the absence of brain uptake for the three complexes prepared, starting from the HYNIC-phenylbenzothiazole derivative 2.12. The labelling of 2.12 with

99m

Tc requires the presence of one

or more co-ligands; in this case tricine, EDDA or a mixture of tricine and nicotinic acid have been used. Although the exact structure of the final

99m

Tc-

complex in the case of HYNIC derivatives has not yet been elucidated, the coligand surely takes part in the complex formation. Whether or not the carboxylate group(s) of the co-ligand are bound to Tc or are present in anionic form, is not clear, but the hydrophilic properties of the co-ligand(s) in each case increase the hydrophilicity of the log P value of

99m

99m

Tc-complex, as can be seen in the

Tc-tricine-2.12 (0.81). A HYNIC derivative of an octapeptide

with affinity for somatostatine receptors, labelled with

99m

Tc in the presence of

tricine or EDDA as co-ligand, allowed to clearly improve the rate of renal excretion of the radiolabelled octapeptide as compared to the

111

In-DTPA

(diethylenetriaminepentaacetic acid) labelled analog177.

For

99m

Tc-2.4 however, the negligible brain uptake is not expected.

This tracer agent possesses several characteristics that normally are associated with free diffusion over the BBB: it is neutral, has a molecular ion mass well below 600 and a log P value of 1.92. Also the

99m

Tc-BAT moiety is

known to efficiently pass the BBB173. A potential reason for the absence of brain uptake, is the possible high binding of the compound to plasma proteins. In addition, the presence of an amide function in the molecule might compromise the uptake in cerebrum and cerebellum. In a study of different complexes of 99mTc with N2S2 tetraligands, Kung and co-workers have found a clearly lower brain uptake for amide containing structures173. Not any

99m

Tc

compound with an amide function has been reported so far that shows brain uptake. In this respect, it would therefore be interesting to prepare and evaluate the derivative of converted to an amine.

99m

Tc-2.4 in which the amide group has been

Biodistribution studies of the radiolabelled thioflavin-T derivatives The clearance from the blood is efficient and rapid for 99m

Tc-2.14 and clearly less favourable for

99m

99m

131 Tc-2.13 and

Tc-2.4. The latter observation

could be an additional indication for a higher protein binding. The HYNIC derivative 2.12, labelled with

99m

Tc in the presence of tricine or EDDA, shows

an even poorer plasma clearance. Especially in this case, the liver value measured at 2 min p.i. is artificially high, since it represents the sum of the tracer agent taken up by the liver tissue and the residual activity in the blood present in the liver.

The results of the biodistribution studies of the

99m

Tc-labelled

phenylbenzothiazoles, indicate that none of these tracer agents is suitable for in vivo diagnosis of AD or PA. Even if they would have affinity for amyloid plaques, they would not be able to reach these plaques in the brain, whereas in abdominal organs the target-to-background activity ratio would be unfavourable due to the pronounced uptake in liver and intestines.

4.3.2. Biodistribution studies of phenylbenzothiazoles labelled with carbon-11 or fluorine-18

The results of the biodistribution studies of

18

F-labelled 3.19 and

11

C-labelled

3.20, 3.21 and 3.22 in mice at 2 min and 60 min p.i. are summarized in Table 11 to Table 14.

The organ distribution of these phenylbenzothiazoles is in several aspects clearly different from that of the higher described technetium-99m labelled analogues. The most remarkable characteristic of the

18

F- or

11

C-

labelled agents is their significant brain uptake at 2 min p.i., followed by an efficient wash-out during the first hour. In this respect, no difference is seen between the three benzothiazoles (3.19, 3.20 and 3.21) on the one hand, and the benzimidazole (3.22) on the other hand. When expressed in percentage of ID per gram tissue, the brain uptake of 3.22 is as high as that of 3.21, the phenylbenzothiazole having the highest activity in cerebrum and cerebellum at 2 min p.i.

132

Chapter 4

Table 11. Tissue distribution of 3.19 after i.v. injection in normal mice (n = 4 at each TP) at 2 and 60 min p.i. % ID ± s.d. 2 min p.i.

60 min p.i.

Urine

0.62 ± 0.16

15.5 ± 2.7

Kidneys

11.4 ± 1.6

Liver

% ID/g tissue ± s.d. 2 min p.i.

60 min p.i.

5.9 ± 1.5

17.9 ± 1.7

8.2 ± 1.8

21.8 ± 1.8

11.7 ± 1.5

10.2 ± 1.6

5.3 ± 0.64

Intestines

8.5 ± 0.59

37.8 ± 4.3

Spleen + Pancreas

1.1 ± 0.11

0.59 ± 0.79

3.2 ± 0.17

1.7 ± 2.5

Lungs

3.9 ± 1.0

0.38 ± 0.01

13.8 ± 2.3

1.5 ± 0.32

Heart

0.64 ± 0.11

0.10 ± 0.04

3.5 ± 0.50

0.52 ± 0.18

Stomach

0.95 ± 0.06

0.88 ± 1.5

1.9 ± 0.33

1.8 ± 3.0

Cerebrum

1.0 ± 0.13

0.07 ± 0.01

3.2 ± 0.38

0.21 ± 0.03

Cerebellum

0.33 ± 0.08

0.02 ± 0.01

3.4 ± 0.38

0.21 ± 0.05

Blood

7.8 ± 0.72

1.8 ± 0.26

2.9 ± 0.23

0.64 ± 0.10

Table 12. Tissue distribution of 3.20 after i.v. injection in normal mice (n = 4 at each TP) at 2 and 60 min p.i. % ID ± s.d. 2 min p.i.

60 min p.i.

Urine

0.58 ± 1.1

24.8 ± 2.9

Kidneys

9.0 ± 2.6

Liver

% ID/g tissue ± s.d. 2 min p.i.

60 min p.i.

2.5 ± 0.98

16.9 ± 4.0

4.8 ± 1.8

27.8 ± 2.4

7.1 ± 0.59

14.5 ± 2.0

4.0 ± 0.38

Intestines

9.4 ± 2.2

54.7 ± 2.8

Spleen + Pancreas

1.2 ± 0.24

0.10 ± 0.02

3.8 ± 0.87

0.35 ± 0.05

Lungs

6.3 ± 2.2

0.76 ± 0.09

28.8 ± 9.9

3.1 ± 0.48

Heart

0.70 ± 0.07

0.05 ± 0.02

5.3 ± 1.1

0.36 ± 0.09

Stomach

0.63 ± 0.14

0.11 ± 0.01

0.80 ± 0.14

0.17 ± 0.03

Cerebrum

0.83 ± 0.23

0.05 ± 0.01

2.6 ± 0.76

0.16 ± 0.03

Cerebellum

0.22 ± 0.05

0.01 ± 0.01

2.8 ± 0.49

0.10 ± 0.14

Blood

9.8 ± 0.78

1.2 ± 0.17

4.6 ± 0.40

0.57 ± 0.08

Biodistribution studies of the radiolabelled thioflavin-T derivatives

133

Table 13. Tissue distribution of 3.21 after i.v. injection in normal mice (n = 4 at each TP) at 2 and 60 min p.i. % ID ± s.d. 2 min p.i.

60 min p.i.

Urine

0.35 ± 0.23

14.6 ± 2.2

Kidneys

8.1 ± 1.5

Liver

% ID/g tissue ± s.d. 2 min p.i.

60 min p.i.

0.70 ± 0.70

14.2 ± 1.8

1.2 ± 1.2

25.8 ± 2.4

4.2 ± 1.5

12.1 ± 1.7

2.2 ± 0.91

Intestines

12.9 ± 2.7

73.7 ± 3.3

Spleen + Pancreas

1.5 ± 0.33

0.11 ± 0.05

4.3 ± 0.46

0.31 ± 0.08

Lungs

2.0 ± 0.26

0.13 ± 0.04

7.8 ± 1.1

0.52 ± 0.17

Heart

0.76 ± 0.07

0.02 ± 0.01

5.1 ± 0.67

0.13 ± 0.03

Stomach

1.2 ± 0.24

0.15 ± 0.07

1.4 ± 0.42

0.13 ± 0.04

Cerebrum

1.3 ± 0.25

0.03 ± 0.00

4.3 ± 0.45

0.09 ± 0.02

Cerebellum

0.57 ± 0.08

0.01 ± 0.00

4.4 ± 0.54

0.08 ± 0.07

Blood

6.0 ± 0.70

0.50 ± 0.06

2.5 ± 0.30

0.21 ± 0.03

Table 14. Tissue distribution of 3.22 after i.v. injection in normal mice (n = 4 at each TP) at 2 and 60 min p.i. % ID ± s.d. 2 min p.i.

60 min p.i.

Urine

0.20 ± 0.11

9.0 ± 5.6

Kidneys

5.9 ± 1.0

Liver

% ID/g tissue ± s.d. 2 min p.i.

60 min p.i.

6.0 ± 3.1

10.5 ± 2.3

9.1 ± 4.4

14.7 ± 4.5

10.9 ± 2.2

6.7 ± 2.4

5.0 ± 1.3

Intestines

10.8 ± 1.9

25.7 ± 2.3

Spleen + Pancreas

1.6 ± 0.10

0.39 ± 0.13

4.4 ± 0.44

1.1 ± 0.24

Lungs

1.6 ± 0.29

0.78 ± 0.16

7.1 ± 0.74

3.1 ± 0.34

Heart

1.3 ± 0.11

0.28 ± 0.05

7.9 ± 0.91

1.7 ± 0.9

Stomach

2.0 ± 0.20

5.1 ± 1.1

2.0 ± 0.28

8.6 ± 1.6

Cerebrum

1.1 ± 0.22

0.12 ± 0.02

4.0 ± 0.57

0.38 ± 0.08

Cerebellum

0.49 ± 0.13

0.03 ± 0.00

4.7 ± 0.51

0.53 ± 0.05

Blood

4.8 ± 0.15

11.7 ± 0.75

1.9 ± 0.11

4.5 ± 0.22

The

absence

of

brain

uptake

for

the

99m

Tc-labelled

phenylbenzothiazoles versus a pronounced initial brain activity for

18

F or

11

C

labelled analogues, indicates that the conjugation of the phenylbenzothiazole with a

99m

Tc-complex, apparently has a detrimental effect on the ability to

134

Chapter 4

pass the BBB. The increase of the molecular mass from 220-256 Da for the simple phenylbenzothiazoles to more than 550 for the

99m

Tc-labelled

conjugates (Table 10 and Table 15) is undoubtly one of the factors contributing to the poor passage of the

99m

Tc-labelled agents over the BBB,

but most likely other factors are also involved in this marked difference.

Table

15.

Physicochemical

characteristics

18

of

F-

or

11

C-labelled

phenylbenzothiazoles Labelled Compound

Mr

log P

charge

polar group(s)

3.19

229

2.86

0

no

3.20

256

2.45

0

-OH

3.21

256

2.48

0

-OH

3.22

223

1.98

0

no

The initial brain uptake of these compounds is in the same range as that of the clinically used cerebral perfusion imaging agents

99m

Tc-HMPAO

and 99mTc-ECD174. The activity found in the cerebrum is about 2.2 to 3.8 times higher than compared to the activity in the cerebellum, and ranges from 1.3 % ID (3.21) to 0.83 % ID (3.20). The initial uptake in the brain (cerebrum + cerebellum) in the range of 1 to 1.8 % ID at 2 min p.i., followed by a rapid and almost complete wash-out at 60 min p.i. (Figure 47), indicates that these tracer agents can freely diffuse over the BBB and are not bound to any structure in the brain of healthy animals.

Cerebrum + Cerebellum 3.19

3.20

3.21

3.22

2

% ID

1,5 1 0,5 0 2

60

min

Figure 47. Comparison of the activity in brain (cerebrum + cerebellum) at 2 and 60 min p.i. for the 18F- and 11C-labelled tracer agents

Biodistribution studies of the radiolabelled thioflavin-T derivatives

135

This meets the requirement for a tracer agent to allow visualization of amyloid plaques in AD: it should be able to enter the brain cells, to be retained if amyloid plaques are present or to leave the brain in the absence of such plaques. The slight differences in brain uptake between the four agents is a minor issue, although 3.20 (0.83 % ID in cerebrum at 2 min p.i.) seems to be somewhat less suitable in this respect. For tracer agents intended for visualization of AD plaques, the ability to cross the BBB and the affinity for amyloid are the most important requirements. Their distribution in other organs is, however, of a lesser concern, except for the clearance of the activity from the blood, which should be rapid and efficient in order to increase the target-to-non-target ratio.

For the

18

F- or

11

C-labelled phenylbenzothiazoles, the clearance of the

activity from the blood is relatively efficient: 6-10 % of ID in the blood at 2 min p.i. and 0.5 to 1.8 % of ID at 60 min p.i. For the

11

C-labelled benzimidazole

3.22, the very unusual observation was made that the activity in the blood at 60 min p.i. was 2.4 times higher than at 2 min p.i. This was seen in all animals involved in this study. An explanation for this rather bizarre increase of blood activity over time has not yet been found and further research is therefore necessary.

Compared

to

the

99m

above-described

phenylbenzothiazoles, the derivatives labelled with

18

F or

11

Tc-labelled

C show a clearly

higher uptake in the kidneys and excretion in the urine. The total activity in the renal system at 60 min p.i. varies from 15 % ID for 3.21 and 3.22 to almost 30 % ID for 3.20. No correlation could be found between the lipophilicity of a compound (log P, Table 15) and its degree of renal handling. It is interesting to note as well that the

99m

Tc-labelled agents have a clearly higher hydrophilic

character than the 11C- and 18F-labelled analogues (log P respectively -0.49 to 1.92 versus 1.98 to 2.86, Table 10 and Table 15), whereas the renal handling is more limited for the 99mTc-compounds.

The handling of the

11

C- and

18

F-labelled phenylbenzothiazoles is

represented graphically in Figure 48. Surprisingly, the derivatives bearing a

136

Chapter 4

phenolic group on the benzothiazole moiety (3.20 and 3.21) are extracted most efficiently from the blood into the liver. Their transfer from liver to intestines is also clearly more rapid than for the other phenylbenzothiazole (3.19) and the phenylbenzimidazole (3.22). The total activity in the hepatobiliary system at 60 min p.i. varies from about 36 % of ID for the phenylbenzimidazole 3.22, to 78 % for one of the phenol substituted phenylbenzothiazoles (3.21), which illustrates the pronounced effect of small structural differences on the biological behaviour of this type of compounds. Liver

% ID

3.19

3.20

3.21

3.22

40 35 30 25 20 15 10 5 0 2

60 min

Intestines

% ID

3.19

3.20

3.21

3.22

80 70 60 50 40 30 20 10 0 2

60

min

Liver + Intestines

% ID

3.19

3.20

3.21

3.22

80 70 60 50 40 30 20 10 0 2

60

min

Figure 48. Uptake of the

11

C- and

18

F-labelled phenylbenzothiazoles in liver

and intestines of mice at 2 and 60 min p.i. and the total activity in the hepatobiliary system (liver + intestines) at these time points.

Biodistribution studies of the radiolabelled thioflavin-T derivatives

137

In this respect, it is also noteworthy to point to the markedly different behaviour

of

the

two

structure

isomers,

i.e.

the

OH-derivatized

phenylbenzothiazoles 3.20 and 3.21. The derivative with the phenol group in para-position with respect to the sulphur atom (3.21), shows a 1.8 times higher initial brain uptake (cerebrum + cerebellum), a clearly higher clearance via liver and intestines (78 % of ID versus 62 % at 60 min p.i.), an accordingly lower excretion via the kidneys and urine (15 % of ID versus 25 % at 60 min p.i.) and a faster and more complete clearance from the blood than its structure isomer 3.20. Nevertheless, the log P value of the two structure isomers is almost identical, as is also the case for their Rt on the RP-HPLC system used.

4.4. Conclusion The results of the biodistribution studies in mice show that the simple phenylbenzothiazoles, labelled with

18

F or

11

C, and to a lesser degree the

studied phenylbenzimidazole, have a higher potential as useful diagnostic tracer agents for visualization of amyloid plaques than the derivatives conjugated to a

99m

Tc-complex. The non-conjugated agents are all capable of

diffusing over the BBB, followed by an almost complete wash-out from the brain, a prerequisite for a tracer intended for diagnosis of AD.

On the other hand, none of the tested radiolabelled compounds seems to have optimal biodistribution characteristics that are required for an ideal radiopharmaceutical for diagnosis of PA in the abdominal region. The lowest residual activity in liver + intestines at 60 min p.i. was about 36 % of ID for the phenylbenzimidazole 3.22 and 42 % for

99m

Tc-labelled phenylbenzothiazole

2.4. However, such activity would still interfere with visualization of plaques in this region.

In this discussion, the affinity of the different compounds for amyloid plaques has not yet been taken up, although this is the most important parameter. The evaluation of this characteristic will be discussed in the next

138

Chapter 4

chapter. The real clinical value of a tracer agent, however, is dependent on a combination of both the overall organ distribution characteristics and the affinity for amyloid plaques.

139

CHAPTER 5

5. IN VITRO STUDIES TO ASSESS THE AFFINITY OF THIOFLAVIN-T DERIVATIVES FOR AMYLOID β

5.1. Introduction As mentioned in chapter 1 (1.3.2.2 and 1.4.2.2), different stains can be used to visualize the presence of amyloid plaques. Ammoniacal silver solution (used in the Bielschowsky method), Bodian’s stain (Protargol solution), Congo red, thioflavin-T (TT), thioflavin-S, X-34 and fluorescent antibodies against Aβ, are all stains that can be used to achieve this goal. The greatest disadvantage of these staining methods however, is the need for biopsied tissue and hence their invasive character, as well as the fact that in case of AD these stains can only be used post mortem. The biopsies in case of PA can furthermore lead to false negative results because of the small sample taken and can lead to serious haemorrhage in the patient. A tracer for in vivo visualization of amyloid plaques would provide a useful tool, not only for the diagnosis but also for therapy evaluation. Especially for the latter, the search for a radioactive tracer that enables to visualize the plaques in vivo, is still a scientific priority.

In order to assess the in vitro affinity of a potential tracer agent, different methods can be used. One method is the use of post mortem tissue sections from AD or PA patients which are incubated with a solution of a radiolabelled compound in the presence or absence of an agent (i.e. TT) that competes for the same binding places. After rinsing the non-bound tracer fraction off the section, the specific binding of the tracer can be visualized with autoradiography. In order to ascertain that these products effectively bind to amyloid plaques, adjacent sections have to be stained with antibodies directed against the amyloid plaques. Ideally, the spatial resolution of the

140

Chapter 5

autoradiography images should be comparable to that of the microscopic images obtained after antibody staining.

Another evaluation method consists of performing binding assays using post mortem AD brain homogenates. This implies homogenization of human brain tissue of AD patients and control brain. A solution of the radiolabelled compound is incubated with the brain homogenate in the presence of different concentrations of a reference compound (TT or likewise molecules) in order to obtain data on the specific and non-specific bound fraction of the tracer.

The in vitro affinity of the new tracer agents described in chapter 2 and 3 was assessed, using binding assays with synthetic amyloid β1-40 fibrils. The technique used was a slight modification of the procedure described by Kung and

co-workers153.

This

assay

consists

in

incubation

of

2-(4’-

dimethylaminophenyl)-6-[125I]iodo-1,3-benzothiazole (125I-TZDM)137, a tracer agent with a high affinity (Kd = 6 x 10-11 M) for amyloid β, with synthetic amyloid β fibrils, in the presence of a series of concentrations of the compound to be assessed. The bound and free fraction of

125

I-TZDM are

separated by filtration and quantified by gamma counting. The affinity determination of technetium-99m labelled tracer agents is generally performed indirectly by assay of the corresponding rhenium complex taking into account the fact that technetium has no stable isotopes. Rhenium is a chemical congener of technetium and is frequently used as a substitute for technetium in the crystallographic structure determinations of new complexes. Since the biodistribution

properties

of

the

examined

technetium-99m

labelled

compounds were not ideal, the decision was taken not to include the synthesis of the corresponding rhenium complexes in this study, but to perform the test in a reversed way where the technetium-99m labelled compound is used as the radiotracer instead of

125

I-TZDM, and TT is used as

competing agent. This test provides qualitative rather than quantitative results with regard to the affinity of the Tc-compound. In case of the tracer agents labelled with carbon-11 or fluorine-18, the corresponding authentic carbon-12 and fluorine-19 products were synthesized and used to determine the affinity

In vitro affinity studies of thioflavin-T derivatives for amyloid β

141

for amyloid β fibrils. These authentic compounds were also used to perform affinity tests on post mortem human AD brain homogenates and binding studies on transgenic AD mouse brain sections and human AD brain sections.

5.2. Materials and methods

5.2.1. Preparation of technetium-99m labelled thioflavin-T derivatives

Deprotection and labelling of 2.4, 2.12, 2.13 and 2.14 was done as described in 2.2.3.

5.2.2. Synthesis of the non-radioactive analogues of the tracer agents labelled with carbon-11 or fluorine-18

5.2.2.1. Synthesis of 2-(4’-fluorophenyl)-1,3-benzothiazole (3.2)

This was performed as described in 3.2.1.2.

5.2.2.2. Synthesis of 2-(4’-methylaminophenyl)-7-hydroxy-1,3-benzothiazole (3.8)

This was performed as described in 3.2.1.4.

5.2.2.3. Synthesis of 2-(4’-methylaminophenyl)-5-hydroxy-1,3-benzothiazole (3.16)

This was performed as described in 3.2.1.6.

5.2.2.4. Synthesis of 2-(4’-aminophenyl)-1-N-methyl-1,3-benzimidazole (3.18)

This was performed as described in 3.2.1.8.

142

Chapter 5

5.2.3. Synthesis of a cationic thioflavin-T derivative intended for diagnosis of PA

5.2.3.1. Synthesis

of

2-(4’-dimethylaminophenyl)-6-hydroxy-N-methyl-1,3-

benzothiazole

2-Imino-6-methoxy-N-methyl-benzothiazole (5.1)

A solution of 2-amino-6-methoxybenzothiazole (18 g, 0.1 mol) and methyliodide (6.22 ml, 0.1 mol) in 50 ml of ethanol was refluxed overnight. The precipitate formed was filtered off and dissolved in hot water. The solution was alkalinized to pH 9 with a saturated Na2CO3 solution (upon which a precipitate formed) and extracted three times with dichloromethane. The organic layer was dried on MgSO4, filtered and evaporated under reduced pressure to yield 60 mmol of 5.1 (11.6 g, 60 % yield). The product was used without further purification. 1

H-NMR (DMSO): δ 3.26 (3H, s, NCH3); δ 3.70 (3H, s, OCH3); δ 6.79 (1H, dd,

5-H); δ 6.90 (1H, d, 4-H); δ 7.09 (1H, d, 7-H); δ 7.99 (1H, s, NH) +

Mass: [M+H] : 195 (calculated: 195) Mp: 92.4-93.3 °C

5-Methoxy-2-methylamino-thiophenol (5.2)

A solution of 5.1 (9.7 g, 50 mmol) in 30 ml of a 50% (m/v) solution of KOH was refluxed for 42 h. Heating was stopped and 100 ml of water was added, followed by concentrated acetic acid to adjust the pH to 6. The mixture was extracted three times with ethyl acetate, the organic fractions were combined, dried on MgSO4, filtered and evaporated under reduced pressure. The obtained product 5.2 (4.2 g, 50 %) was used as such without further purification. 1

H-NMR (DMSO): δ 2.70 (3H, d, NHCH3); δ 3.54 (3H, s, OCH3); δ 3.66 (1H, s,

SH); δ 5.11 (1H, m, NH); δ 6.56 (1H, d, 3-H); δ 6.62 (1H, d, 6-H); δ 6.90 (1H, dd, 4-H)

In vitro affinity studies of thioflavin-T derivatives for amyloid β

143

+

Mass: [M+H] 170 (calculated: 170) Mp: 54.4-55.6 °C

2-(4’-Dimethylaminophenyl)-6-methoxy-N-methyl-1,3-benzothiazole (5.3)

A solution of 4-(dimethylamino)benzoyl chloride (1.83 g, 10 mmol) in 10 ml of toluene was heated to 60 °C under nitrogen and 5.2 (1.4 g, 10 mmol) was added. After heating it for 30 minutes at 60 °C, the mixture was refluxed for 1 h. After cooling down, the formed precipitate was filtered off and purified with column chromatography (100 % CH2Cl2 to CH2Cl2/MeOH 80:20) and this yielded 200 mg of product 5.3 (0.67 mmol, 6.7 %). 1

H-NMR (DMSO): δ 3.09 (6H, s, N(CH3)2); δ 3.57 (3H, s, OCH3); δ 4.22 (3H, s,

NCH3); δ 6.95 (2H, d, 3’-H 5’-H); δ 7.68 (1H, d, 5-H); δ 7.81 (2H, d, 2’-H 6’-H);

δ 8.13 (1H, d, 4-H); δ 8.21 (1H, s, 7-H) +

Mass: [M+H] 299 (calculated: 299) Mp: 188-189 °C

2-(4’-Dimethylaminophenyl)-6-hydroxy-N-methyl-1,3-benzothiazole (5.4)

A solution of 5.3 (150 mg, 0.5 mmol) in 5 ml of dry CH2Cl2 was cooled down to -70 °C under nitrogen and a 1 M BBr3 solution in CH2Cl2 (3 ml, 3 mmol) was added dropwise over a 30 min period. After an additional hour at -70 °C, the solution was allowed to warm up to room temperature and stirring was continued overnight. After cooling down again to -70 °C, 3 ml of MeOH was added and the organic layer was evaporated to dryness under reduced pressure.

The

residue

was

purified

with

column

chromatography

(CH2Cl2/MeOH 95:5 to 90:10). This yielded 80 mg of product 5.4 (0.28 mmol, 56 %) 1

H-NMR (DMSO): δ 3.08 (6H, s, N(CH3)2); δ 3.99 (3H, s, NCH3); δ 6.94 (2H, d,

3’-H 5’-H); δ 7.28 (1H, dd, 5-H); δ 7.68 (1H, d, 7-H); δ 7.75 (2H, d, 2’-H 6’-H); δ 8.02 (1H, d, 4-H) +

Mass: [M+H] 285 (calculated: 285) Mp: 205.8-206.4 °C

144

Chapter 5

5.2.4. Determination of affinity of

99m

Tc-labelled compounds for amyloid

β

Aβ1-40 fibrils were prepared as described by Zhuang et al.153 Briefly, the lyophilized Ab1-40 peptide as the HCl salt was obtained from Bachem (Bubendorf, Switzerland). Fibrils were formed by dissolving 1 mg of the peptide in 1.155 ml of a 1 mM Na2EDTA solution in H2O. The solution was sonicated for 5 min, after which 1.155 ml of 1 mM Na2EDTA in 20 mM phosphate buffer pH 7.4 was added. The solution was stirred for 3 days at 30 °C and was then centrifuged at 28,000 g for 15 min at 4 °C (Beckman Avanti 30 centrifuge, Beckman-Coulter, Fullerton, CA, U.S.A.). The supernatant was pipetted off and the pellet was washed twice with 100 µl of a mixture of 1 mM Na2EDTA in 10 mM phosphate buffer pH 7.4. Finally, the pellet was resuspended by adding 2.310 ml of 1 mM Na2EDTA in 10 mM phosphate buffer pH 7.4. The suspension was divided in 30 µl aliquots and stored at –80 °C. The final concentration of the fibrils in each aliquot was 100 µM. The amount of non-fibrillar free peptide was determined by Bio-Rad protein assay (Bio-Rad, München, Germany). For this purpose, a calibration curve was constructed using four test tubes containing non-fibrillar Ab1-40 in a concentration between 10 µM and 10 nM in 5 % formic acid. The amount of protein in the pellet was determined by diluting 100 µl of the pellet suspension with 100 µl formic acid and diluting this mixture with 7.8 ml of a 5 % solution of formic acid. The amount of protein in the supernatant was determined by diluting 200 µl of the supernatant with 1.8 ml 5 % formic acid. To each test tube containing 0.8 ml of the standard or 0.8 ml of the unknowns, 0.2 ml of dye reagent (from the Bio-Rad assay) was added and the reaction mixture was vortexed. To prepare a blank, 0.2 ml dye reagent was added to a test tube containing 0.8 ml of 5 % formic acid and this mixture was vortexed. After 30 min, the absorbance was measured at 595 nm (Shimadzu Recording Spectrophotometer UV-240, Shimadzu, Kyoto, Japan) 30 min after addition of the dye reagent for all samples.

In vitro affinity studies of thioflavin-T derivatives for amyloid β Binding of the

99m

145

Tc-compound to the fibrils was studied using a

procedure similar to the one described by Kung et al198, in which the

99m

Tc-

compound was used as the radioligand and thioflavin-T as the competing agent. Briefly, three series of 12 test tubes were used. Each test tube contained 860 µl of a 10 % EtOH solution in water, 50 µl of the RP-HPLC isolated

99m

Tc-compound (~74 kBq) diluted with 0.025 M phosphate buffer pH

7.4 and 50 µl of the diluted Aβ fibril suspension (2.5 µg/ml). The diluted Aβ fibril suspension was added as the last component. Test tubes 1 to 10 also contained 40 µl of a TT solution in EtOH (concentrations ranging from 1.5 nM to 100 µM), while tube 11 contained 40 µl of a 200 µM TT solution (for determination of non-specific binding) and test tube 12 contained 40 µl of EtOH 10 % (control test tube). The final volume in the test tubes was 1 ml and all test tubes were shortly vortexed before incubation at room temperature for 3 h. After the incubation period, glass fibre filters (Glasfaser Rundfilter, GF 51, 25 mm diameter, Schleicher & Schuell, Dassel, Germany) drenched in a 10 % ethanol solution were put into place in a home-made harvester system, provided by Janssen Pharmaceutica. The harvester was closed and a vacuum pump was attached to the outlet of the harvester in order to remove the fluids from the apparatus. A series of 12 test tubes was poured out over the filters to separate the Aβ fibril-bound and free tracer agent. To each test tube 3 ml of a 10 % ethanol solution was added and the tubes were emptied again over the filters. This rinsing procedure was repeated two more times to ensure a transfer of radioactivity as complete as possible from the test tubes to the filters. Next, each filter hole was rinsed once more with 3 ml of a 10 % ethanol solution, the filters were taken out of the apparatus and their radioactivity counted in an automatic gamma counter. The activity retained on the filter (bound activity) was used for further calculations. Experiments were done in triplicate.

The radioactivity determined on the filter of the control test tube minus the radioactivity measured on the filter corresponding to the highest concentration of the competing ligand, is assumed to correspond to a 100 %

146

Chapter 5

specific binding. The percentage of the specific binding (relative to the control) for each sample is then calculated in the following way:

% specific binding = 100 x

activity on sample filter - activity on non specific binding filter activity on control filter - activity on non specific binding filter

The percentage of the specific binding is then plotted against the logarithm of the concentration of the competitor.

The IC50 value for the competitor is calculated as the concentration of the competitor at which the specific binding of the radiotracer decreases to 50 % of the control.

5.2.5. Determination

of

the

affinity

of

authentic

non-radioactive

compounds for amyloid β

Approximately the same procedure as described in 5.2.4 was used, but in this test

125

I-TZDM was used as the radioligand and was incubated with Aβ fibrils

in the presence of a series of concentrations of the compound to be tested. 125

I-TZDM was kindly provided by Dr. H. F. Kung (University of Pennsylvania,

PA, USA). To each test tube was added 50 µl of a solution of

125

I-TZDM

(specific activity 8.05 x 1016 Bq/mol) in 40 % ethanol (yielding approximately 80 000 to 100 000 counts per minute (cpm)), which corresponds to 20 pmol of 125

I-TZDM and brings the final concentration of

third of the Kd value of

125

125

I-TZDM in the assay to one

I-TZDM (60 pmol). Test tubes 1 to 10 contained 40

µl of a solution of the compounds to be tested (3.2, 3.8, 3.16, 3.18 and 5.3) in concentrations ranging from 1.5 nM to 2 mM. The test was further performed as described in 5.2.4. The percentage of the specific binding was calculated, plotted against the logarithm of the concentration of the competitor and the Kd values of these competitors were calculated using the following equation (Cheng and Prusoff method199):

In vitro affinity studies of thioflavin-T derivatives for amyloid β Kdi =

147

EC50

⎛ [* D] ⎞ ⎟ ⎜1 + ⎜ Kd ⎟ *d ⎠ ⎝

with K d i = equilibrium dissociation constant of the competitor and K d

*d

=

equilibrium dissociation constant of 125I-TZDM (*D).

Since [*D] is one third of K d , the K d i of the five competitors (3.2, 3.8, *d

3.16, 3.18 and 5.4) can be calculated as follows:

Kdi =

EC50 or EC50 = 1.33 K d i 1.33

GraphPad Prism™ version 3.03 (GraphPad Software, Inc.; San Diego, CA, USA) software was used for these calculations.

Due to the long half-life of

125

I (60 days) and the possibility of formation

of volatile radioiodine (upon deiodination of

125

I-TZDM), two necessary

precautions were taken: firstly, the harvester was placed in a plastic dish to prevent contamination of the working area and secondly, the outlet of the vacuum pump was coupled to a flask filled with a 1 M NaOH solution to capture volatile radioiodine.

5.2.6. Determination of the affinity of authentic non-radiolabelled derivatives for post mortem human AD brain homogenates

5.2.6.1. Preparation of brain tissue homogenates

Post mortem brain tissues were obtained from four AD patients and four agematched controls at autopsy, and neuropathological diagnosis was confirmed by current criteria (NIA-Reagan Institute Consensus Group, 1997). The cerebellum and the affected temporal or parietal cortex of AD patients together with the corresponding regions from controls were isolated. Gray matter was carefully separated from white matter from the cortical tissues.

148

Chapter 5

Homogenates were then prepared in phosphate buffered saline (PBS, pH 7.4) at a concentration of approximately 100 mg wet tissue/ml (motor-driven glass homogenizer with setting of 6 for 30 sec). The homogenates were aliquoted into 1 ml portions and stored at –70 °C for 3-6 months without loss of binding signal (Dr. H. F. Kung, personal communication).

5.2.6.2. Binding studies

Binding assays were carried out in 12 mm x 75 mm borosilicate glass tubes. For saturation studies, the reaction mixture contained 50 µl of tissue homogenate (20-50 µg), 50 µl of a [125I]IMPY (see Figure 7) solution in PBS, (0.02-0.04 nM)200 and 40 µl of a solution of the test compound (3.2, 3.8, 3.16 and 3.18; 10-7-10-10 M diluted serially in PBS containing 0.1 % bovine serum albumin) in a final volume of 1 ml. Nonspecific binding was defined in the presence of 600 nM IMPY in the same assay tubes. For the competition binding studies, concentrations of 10-5 –10-10 M of the compounds and 0.06 nM [125I]IMPY were used. The mixtures were incubated at 37 °C for 2 h and the bound and the free radioactivity were separated by vacuum filtration through Whatman GF/B filters using a Brandel M-24R cell harvester followed by 2 x 3 ml washes with PBS at room temperature. Filters containing the bound iodine-125 ligand were counted in a gamma counter (Packard 5000) with 70 % counting efficiency. Under the assay conditions, the specifically bound fraction was less than 15 % of the total radioactivity. Protein determinations were performed with Lowry’s method201 using bovine serum albumin as a standard. The results of saturation and inhibition experiments were subjected to nonlinear regression analysis using EBDA by which Kd and Ki values were calculated.

In vitro affinity studies of thioflavin-T derivatives for amyloid β

149

5.2.7. Determination of in vitro affinity for plaques in mouse and human AD brain sections

5.2.7.1. Paraffin sections

Transgenic mice overexpressing human APP (London mutation) were anaesthetized with Nembutal® (intraperitoneal injection, 2 µl/g) and the brain was flushed transcardially with ice-cold saline. The brain was removed and post-fixed overnight with 4 % paraformaldehyde in PBS. One part of the brain was stored in 0.1 % sodium azide in PBS at 4 °C. The other part of the brain was processed for paraffin embedding. Sagittal sections (8 µm) were cut using a Microm HM 340 E microtome (Microm, Walldorf, Germany). Prior to incubation, paraffin sections were taken through two 5-min washes with xylol, followed by 1-min sequential washes with 100 %, 100 %, 90 %, 70 %, 50 % EtOH and PBS. After rehydration, the sections were incubated for 1 h with one of the cold compounds (3.8 or 3.16) at a concentration of 1 µM. After 1 h, the brain sections were rinsed with tap water and coverslipped with MowiolDABCO. Preparation of Mowiol-DABCO: a mixture of mowiol (Calbiochem, Darmstadt, Germany; 10 g) and 90 ml of PBS pH 7.4 was stirred for 4 to 6 h, DABCO (1,4 diazabicyclo[2.2.2]octane; Sigma-Aldrich; 3.25 g) was added followed by 40 ml of glycerin and the mixture was stirred overnight. Then the mixture was centrifuged at 15 000 rounds per minute for 30 min, aliquoted and stored at -20 °C. The aliquots were thawed before use and can be refrozen after use.

For immunohistochemical staining, the same deparaffinization protocol was used. The sections were boiled for 10 min in a microwave (Miele M 326G) with 0.01 M citric acid pH 6 to enhance signal. After cooling down for 20 min the sections were rinsed with PBS followed by treating them with 1.5 % H2O2 in PBS/MeOH (50:50 v/v) to inhibit endogenous peroxidase. Aspecific binding was blocked by treatment with PBS containing 0.05 % Tween® 20 (PBST) and 10 % fetal calf serum (FCS). The sections were incubated overnight with AβN25 monoclonal antibody, diluted 1/500 in PBST with 10 %

150

Chapter 5

FCS at 4 °C. After rinsing in PBST, the sections were incubated for 1 h with 1/500 diluted goat anti-mouse biotin (Vector Laboratories, Burlingame, CA, U.S.A.) in PBST with 10 % FCS. Then sections were incubated with Avidin Biotin Complex (Vectastain Elite ABC-peroxidase kit, Vector Laboratories) dissolved in PBST with 10 % FCS for 30 min at room temperature. Peroxidase activity was developed with 3,3’-diaminobenzidine, after which the sections were counterstained with haematoxylin.

Paraffin embedded human AD brain was stained following the same procedure as described for mouse brain. AD brain was treated according to the regulations of the Ethical Committee of the hospital.

5.2.7.2. Vibratome sections

Besides paraffin sections (see 5.2.7.1) also vibratome sections were used. For vibratome sections, the paraformaldehyde-fixed brain was cut into sagittal sections of 40 µm thickness with a Microm HM 650 V vibration microtome (Microm). The sections were incubated for 1 h at room temperature with one of the cold compounds (3.8 or 3.16) at a concentration of 1 µM. After incubation, the sections were rinsed for 5 s in tap water and coverslipped using a Mowiol-DABCO solution.

Immunohistochemical staining of vibratome sections was performed using the same protocol as described in section 5.2.7.1 with some modifications. No microwave pretreatment was performed and the solution containing 0.05 % Tween® 20 in PBS buffer was replaced by a 0.1 % Triton X100 solution. The dilution of the AβN25 antibody was 1/2000. After rinsing in PBST, the sections were incubated for 1 h with 1/500 diluted goat anti-mouse horseradish peroxidase (Dako A/S, Denmark) in PBST with 10 % FCS. Peroxidase activity was developed with 3,3’-diaminobenzidine. After washing with PBS, the sections were mounted on silane-coated glasses and after drying they were counterstained with haematoxylin, the sections were

In vitro affinity studies of thioflavin-T derivatives for amyloid β

151

dehydrated by passage through a graded series of alcohol and xylol and coverslipped with DePeX mounting medium.

Fluorescence microscopy was performed using a Leica DMR microscope equipped with a digital Leica DC480 camera and a UV filter set with the following specifications: excitation filter 355-425 nm; dichromatic mirror 455 nm; emission 470 nm longpass filter. The images were collected and processed with Leica IM500 image processing software.

5.3. Results and discussion

5.3.1. Synthesis of 2-(4’-dimethylaminophenyl)-6-hydroxy-N-methyl-1,3benzothiazole (5.4)

Starting from the commercially available 2-amino-6-methoxybenzothiazole, a method described by Lin202 was used to obtain 5.1. Thus, 2-amino-6methoxybenzothiazole was methylated with iodomethane in boiling ethanol. This resulted in the hydroiodide salt of 5.1, which was converted to the amine with saturated Na2CO3 to obtain 5.1 in a 60 % yield. In a next step, the thiazole ring was opened by boiling in a 50 % potassium hydroxide (m/v) solution for 42 hours, which yielded 5.2. The latter was reacted with 4(dimethylamino)benzoyl chloride in toluene, resulting in 5.3. The O-methyl group was then removed with BBr3 to obtain the intended compound 5.4. A schematic representation of the synthesis of 5.4 is given in Figure 49.

Compound 5.3 was synthesized in order to obtain a cationic TT derivative which could be labelled with carbon-11. In order to be sure that the authentic non-radioactive analogue 5.3 has affinity for amyloid plaques, the affinity of 5.3 for amyloid β fibrils was tested first. In a later stadium 5.4 can be labelled with carbon-11 if 5.3 shows affinity for amyloid β fibrils. The only difference with TT is the presence of a 6-methoxy group in 5.3, instead of a 6-methyl group as in TT.

152

Chapter 5

CH3

N NH2 S

H3CO

N

CH3I

NH

EtOH, reflux

H3CO

5.1

S

KOH Reflux

CH3 NH ClOC

N(CH3)2

+

SH

H3CO

Toluene

5.2

N

N

BB3

N(CH3)2 H3CO

S 5.3

N(CH3)2 CH2Cl2

HO

S 5.4

Figure 49. Synthesis of 2-(4’-dimethylaminophenyl)-6-hydroxy-3-N-methyl-1,3benzothiazole

5.3.2. Affinity tests: set-up

A schematic representation of the test set-up for the determination of the affinity of the non-radioactive authentic compounds (3.2, 3.8, 3.16 and 3.18) and 5.3 for synthetic amyloid β fibrils, is given in Figure 50. The authentic compounds function as competitor agents. In the case of the technetium-99m labelled 2.4, 2.12, 2.13 and 2.14, TT functions as the competitor. In all these tests the amyloid β fibril suspension was added in the last step, as to assure that the radiotracer and the competing agent can equally compete for the binding sites on Aβ fibrils.

The affinity tests were performed in triplicate, but in some cases the results of the three tests showed minor inter-test variabilities. The reason for these differences has not yet been elucidated, but possibly has to do with the reproducibility of the vacuum applied on the respective harvester holes during filtration of the samples.

In vitro affinity studies of thioflavin-T derivatives for amyloid β

153

Content of the test tubes

50 µl Amyloid β fibrils

50 µl Amyloid β fibrils 50 µl 125I-TZDM 40 µl competitor

125

50 µl I-TZDM 40 µl EtOH 10 %

860 µl EtOH 10%

860 µl EtOH 10%

increasing concentrations (series of 10)

Control test tube

50 µl Amyloid β fibrils 50 µl 125I-TZDM

40 µl TT 8 µM

860 µl EtOH 10%

Non-specific binding test tube

3 series of 12 test tubes are used per test; each test was performed in triplicate

Incubation for 3 h at room temperature Application of contents to the harvester and rinsing vacuum pump oo oo oo oo

ooo ooo ooo ooo

oooo oooo oooo oooo

o o o o

Cell Harvester Each hole contains a filter Each test tube is rinsed three times with 10% EtOH

Counting of activity on the filters with automatic gamma counter: 1 min/filter Figure 50. Schematic representation of the amyloid β fibrils affinity test set-up 5.3.2.1. Affinity of 99mTc-2.4, 99mTc-2.13 and 99mTc-2.14 for amyloid β

The plot of the percentage specific binding versus the logarithm of the concentration of the TT plot (results not shown), yields a horizontal straight line for 99mTc-2.4, 99mTc-2.13 and 99mTc-2.14, indicating that these compounds do not have affinity for amyloid β fibrils or at least do not share the same binding site as TT.

154

Chapter 5

5.3.2.2. Affinity of 99mTc-tricine-2.12 for amyloid β

In contrast to the other technetium-99m labelled tracers,

99m

Tc-tricine-2.12

showed a simple sigmoidal competition binding profile. This allows concluding that

99m

Tc-tricine-2.12 indeed has affinity for synthetic amyloid β fibrils (EC50

value of 2.7 x 10-6).

% specific binding

200

100 50 0 -11

-10

-9

-8

-7

-6

-5

-4

log of concentration of thioflavin-T (M) Figure 51. Percentage of specific binding of

99m

Tc-tricine-2.12 to synthetic

amyloid β fibrils with TT as inhibitor

To obtain extra information on the affinity of

99m

Tc-tricine-2.12 for

amyloid β fibrils, additional tests have to be performed. The use of post mortem human tissue sections of patients with PA can be very useful to obtain adequate information on the affinity of Since

99m

99m

Tc-tricine-2.12 for amyloid plaques.

Tc-tricine-2.12 is in vivo characterized by a high first pass extraction

by the liver, followed by rapid excretion into the intestines, this tracer can still be useful to detect amyloid plaques in body parts that are positioned outside the splanchnic area. It would therefore be useful to examine the affinity of 99m

Tc-tricine-2.12 in human tissue sections containing amyloid plaques, since

this in vitro technique represents a situation that is more comparable to an in vivo situation. Furthermore, since this compound shows no brain-uptake, it is

In vitro affinity studies of thioflavin-T derivatives for amyloid β

155

necessary to take a closer look at its affinity for amyloid plaques that are not composed of amyloid β, since the latter is only found in the brain of AD patients.

Although the synthesis of rhenium complexes was not yet part of this study, it will be of future interest to synthesize a rhenium complex of

99m

Tc-

tricine-2.12. Such a rhenium complex can be used to unravel the structure of this technetium-99m-HYNIC-tricine-BAT-complex. It will help to understand the biological characteristics of this tracer agent better.

The use of an animal model with PA would be most ideal to test the in vivo properties of this tracer. Although the abdominal region will show high uptake and high background activity, the possible binding of the tracer to plaques present in joints and other non-abdominal body parts, can be visualized with the aid of a micro-SPECT camera or using autoradiography on tissue sections of animals which are sacrificed after injection of the tracer. If amyloid deposits in joints can be visualized via micro-SPECT of the living animal or via autoradiography of tissue sections, this will be very strong proof of the affinity of

99m

Tc-tricine-2.12 for amyloid plaques. In addition to these

tests, it would also be useful to repeat these experiments with

99m

Tc-labelled

2.12 with EDDA or tricine/nicotinic acid as co-ligand in order to determine whether the affinity of

99m

Tc-labelled 2.12 is dependent on or affected by the

nature of the co-ligand.

Although there is not much structural difference in the BTA part of the four technetium-99m labelled tracers, it is striking that only

99m

Tc-tricine-2.12

shows affinity for amyloid β fibrils. This may indicate that the conversion of the amine group on the phenyl ring to an amide, negatively affects the affinity of these phenylbenzothiazoles, since only in 2.12 this amino group is preserved. The presence of one or two methyl groups on this aromatic amine does not cause an on-off phenomenon in case of affinity for amyloid β, which is clear from the affinity profile of TT and numerous derivatives described in literature (see Figure 7). The presence of an amide on the other hand, seems to be

156

Chapter 5

detrimental to the affinity for amyloid. Whether this is only caused by the presence of the amide or by a combination of the amide with a large BCL, is not clear. The affinity of

99m

Tc-tricine-2.12 nevertheless indicates that

derivatization of the phenylbenzothiazole on positions other than the aromatic amine of the phenyl ring, does not necessarily have a negative effect on the affinity for amyloid β. Further studies could include the technetium-99m labelled derivatives described in chapter 2, but with the amide group reduced to an amine group. This will lead to useful information concerning the structure-affinity relationship, as well as the ability to cross the BBB in case of neutral complexes such as 99mTc-2.4.

5.3.2.3. Affinity of authentic non-radioactive products 3.2, 3.8, 3.16, 3.18 and 5.3 for amyloid β

To test the affinity of the non-radioactive analogues of the carbon-11 and fluorine-18 labelled tracer agents, the test as depicted in Figure 50 was performed with

125

I-TZDM as the radioligand.

125

I-TZDM has been reported to

have a high affinity for Aβ (Kd = 6x10-11 M). The graphs of 3.2, 3.8, 3.16 and 3.18 are depicted in Figure 52. Because the concentration of

125

I-TZDM in the

incubation mixture is known, the Ki values could be derived for the authentic products: 443 nM for 3.2, 6.5 µM for 3.8 and 2 µM for 3.16. These values are higher than those observed for TT (265 nM).

No Ki value could be obtained for 3.18, since this compound did not seem to compete with

125

I-TZDM for binding to amyloid β fibrils. The other

three agents have affinities that are lower than that of TT, but still high enough to be useful for good binding to amyloid β fibrils.

The Ki value for TT reported in literature is 116 nM when

125

I-TZDM

was used as the radiotracer137. Although this is in the same nanomolar range, the value found in this study is about two factors higher than the value described in literature. A potential reason for this inconsistency might be the fact that in the study here presented, the separation of bound and unbound

In vitro affinity studies of thioflavin-T derivatives for amyloid β

157

radioactivity was done manually in comparison to the use of an automated

% Specific binding

harvester in the studies described in literature.

100 90 80 70 60 50 40 30 20 10 0 -11

Thioflavin-T 3.2 3.8 3.16 3.18

-10

-9

-8

-7

-6

-5

-4

log concentration competitor (M) Figure 52. Percentage of specific binding of

125

I-TZDM to synthetic amyloid β

fibrils with the authentic products 3.2, 3.8, 3.16 and 3.18 as competitor

As a result, the TP required for separation of the amyloid fibrils after the incubation, is 10 to 15 minutes in our setting, while an automated system only needs 2 or 3 minutes to perform the task. This, together with the fact that an automated system empties and rinses all tubes at the same moment, may be important factors leading to the difference in Ki value. Although hard evidence is lacking that these factors are responsible for the differences between the Ki value for TT found in our study and that reported in literature, this observation led to the assumption that the Ki values found for 3.2, 3.8 and 3.16 are a minor underestimation of the real affinity, meaning that these authentic compounds do have a lower Ki value than the one described above.

Surprisingly the benzimidazole derivative 3.18 showed no affinity for amyloid β fibrils in comparison to

125

I-TZDM. Unlike benzoxazoles,

radiolabelled benzimidazole derivatives have not been described in literature to date as potential tracer agents for visualization of amyloid plaques. It therefore seemed interesting to compare a series of such compounds with

158

Chapter 5

benzothiazoles as potential tracers for diagnosis of AD or PA. Compared to benzothiazoles and benzoxazoles, it is clear that the hydrophilicity of the benzimidazoles is the highest, followed by the benzoxazoles. It was assumed that this characteristic would lead to a better clearance from the body and the brain. However, this was not the case, as could be seen in Table 14. The absence of affinity for amyloid can probably be attributed to the presence of the nitrogen atom in the –imidazole ring, compared to the sulphur atom in the benzothiazoles and the oxygen atom in the benzoxazoles. The nitrogen atom has the possibility of losing a proton, which results in a stabilized structure due to the presence of the phenyl ring next to the –imidazole ring. Although the precise mechanism of binding to amyloid β fibrils has not yet been elucidated, the acidic properties of the nitrogen atom in the –imidazole ring seem to have an influence on the affinity for amyloid plaques. The acidic properties, possibly together with the presence of a methyl group on this nitrogen atom, probably explain the lack of affinity of 3.18 for amyloid β fibrils.

Although the non-radioactive compound 5.3 shows high structural similarity to TT, it showed no affinity for amyloid β fibrils. No explanation was found to date to explain this absence of affinity and therefore labelling of 5.4 with carbon-11 was not performed.

5.3.3. Affinity of the authentic non-radiolabelled derivatives 3.2, 3.8, 3.16 and 3.18 for post mortem human brain homogenates

In order to obtain extra information on the in vitro affinity of these tracers for amyloid β fibrils, determination of their binding to human brain homogenates is a useful test. Because of some practical problems, a cooperation was established with Dr. H. F. Kung in Philadelphia to perform this test. The different Ki values obtained for the authentic non-radioactive compounds are given in Table 16.

In vitro affinity studies of thioflavin-T derivatives for amyloid β Table 16. Ki values of 3.2, 3.8, 3.16 and 3.18 competing with

159 125

I-IMPY

binding to amyloid plaques in human AD brain homogenates Compound

Ki (nM, mean ± SEM)

3.2

9.0 ± 2

3.8

11.2 ± 5

3.16

11.5 ± 3

3.18

1803 ± 300

These values express the amount of non-radioactive compound needed to obtain a 50 % inhibition of the binding of

125

I-IMPY to the amyloid β

plaques present in the post mortem brain homogenates. From these values the same conclusion can be drawn as from the Kd values obtained in 5.3.2.3, namely that the benzothiazoles show good affinity while the benzimidazole shows no affinity for amyloid β plaques.

Non-radioactive 6-OH-BTA-1 (see Figure 7) showed a Ki value of 2.8 ± 0.5 nM in this test, indicating that the new compounds have a similar affinity for amyloid β as 6-OH-BTA-1. Furthermore, it can be concluded that all these compounds bind to the same binding sites in the amyloid β plaques.

5.3.4. In vitro affinity of 3.2, 3.8 and 3.16 for plaques in mouse and human AD brain sections

The potential of binding to amyloid β plaques of the three most promising compounds 3.2, 3.8 and 3.16 was also evaluated in post mortem brain sections of a human AD patient and in brain sections of transgenic AD mice. Both vibratome sections of unparaffinized brain and mictrome sections of paraffinized brain can be used for this purpose. Vibratome sections can be incubated with the non-radioactive compound in free floating conditions, maximizing the contact surface between the solution and the brain section. Microtome sections benefit from the fact that they are thinner, which is more useful for performing co-localization experiments. In such experiments, one

160

Chapter 5

section is incubated with one of the non-radioactive compounds (which can be visualized with fluorescence microscopy as the bound benzothiazole emits fluorescence),

while

a

neighbouring

section

can

be

stained

immunohistochemically with the AβN25 antibody that binds to fibrillar amyloid

β. If the non-radioactive compound binds to fibrillar amyloid β on a particular section, the localization of the fluorescence on this section has to be almost identical to the localization of the antibody on the adjacent section. Such a colocalization experiment is almost impossible to perform with vibratome sections since the diameter of the plaques is about 100 µm and these vibratome sections are 40 µm thick. Treatment of a vibratome section with the non-radioactive compound and the AβN25 antibody at the same time, nevertheless also enables to obtain useful co-localization data. After incubation and rinsing with tap water, the glass slides were coverslipped with a Mowiol-DABCO solution which glues the two glass slides together and reduces the photobleaching effect (prolonging fluorescence conservation).

The results of the incubation of 8 µm microtome sections of transgenic AD mouse brain and post mortem human AD brain with either 3.8 or 3.16 in a 1 µM concentration and of the immunohistochemical staining of neighbouring sections is shown in Figure 53. The region of the mouse brain shown is the subiculum, a structure rich in plaques in AD mice that is located in the hippocampic region. The region shown in the human brain is part of the cortex with unknown specification of the localization.

In vitro affinity studies of thioflavin-T derivatives for amyloid β

161

*

* *

*

3.8

#

*

* *

* AβN25

#

#

# # #

#

# # # #

3.16

# #

Transgenic AD mouse

#

Human AD brain

Figure 53. Fluorescence microscopic images after incubation of transgenic AD mouse brain (left) and human AD brain (right) with 3.8 (top) or 3.16 (bottom). Middle row immunohistochemically stained with AβN25. Bar = 100 µm. From top to bottom are three serial sections. Amyloid β plaques in AD mouse brain highlighted with white arrowheads (top/bottom) or red arrowheads (middle). Amyloid β plaques in human AD brain marked with * for 3.8 (corresponding plaques immunohistochemically stained marked with yellow *) or with # for 3.16 (corresponding plaques immunohistochemically stained marked with red #).

162

Chapter 5 From Figure 53 it can be seen that 3.8 as well as 3.16 stain plaques in

transgenic AD mouse brain as well as in human AD brain. This is in accordance with the findings from the in vitro affinity tests. Comparison of the images of the immunohistochemical staining to the images obtained with fluorescence microscopy shows that the AβN25 antibody stains more Aβ plaques than the two non-radioactive compounds. This is a consequence of the much higher affinity of the AβN25 antibody for the amyloid β plaques in comparison to the affinity of 3.8 and 3.16.

AβN25

AβN25 #

H

*

*

#

u

*

m

#

a n #

* A D

# b

*

*

# *

r

#

a i n

#

* 3.8

3.16

Figure 54. Fluorescence microscopic images after incubation of human AD brain with 3.8 (bottom left) or 3.16 (bottom right). Upper row: images of sections immunohistochemically stained with AβN25. Bar = 100 µm. Three serial paraffin sections (upper section is the same). Co-localization of plaques indicated with * for 3.8 and # for 3.16.

In vitro affinity studies of thioflavin-T derivatives for amyloid β

163

Figure 54 further shows the ability of 3.8 and 3.16 to stain amyloid β plaques. A possible explanation why 3.8 and 3.16 do not stain the same plaques is that these are serial sections starting from bottom left over the immunohistochemically stained section to bottom right. Since these are 8 µm sections, the core of the amyloid β plaques can be located differently. From this figure it also appears that the benzothiazoles primarily stain the core of the amyloid β plaque. To further evaluate this, a vibratome section was incubated first with the AβN25 antibody and then with 3.16. This colocalization is shown in Figure 55.

In these images, the outer rim of the plaque is only stained with the AβN25 antibody while the core of the amyloid β plaque shows intensive fluorescence. This is probably a consequence of the higher amount of βpleated sheets in the core of the plaque and proves the assumption that these benzothiazoles primarily label the core of the amyloid β plaque.

Figure 55. Section of a vibratome section of a transgenic AD mouse brain stained with AβN25 in combination with 3.16. Depicted are two enlarged plaques from an entire section.

Other vibratome sections from transgenic AD mouse brain also show that 3.8 and 3.16 can be used to visualize vascular amyloid as can be seen in Figure 56. In these images the background signal is lower in the left section stained with 3.8 than in the right section stained with 3.16. Both compounds highlight the vascular amyloid present, but the fluorescent signal with 3.16 is

164

Chapter 5

more intensive than with 3.8 and the background signal is also higher for 3.16. Therefore, attempts were undertaken to repeat the test with a 0.1 µM concentration of the compounds (data not shown). As could be expected from Figure 56, this lower concentration still yielded useful fluorescence images with 3.16 with a lower background signal, while the fluorescence was close to the detection limit for 3.8. This is an important finding when looking at the concentrations used in nuclear medicine. When a patient is injected with 370 MBq of a tracer agent, the amount of injected compound is approximately 9 nmol (if the tracer has a specific activity of 4.2 x 1016 Bq/mol). Since this injected amount is diluted in blood and the compound possibly metabolized in the liver, only a small percentage of the injected tracer will reach the brain, pass the BBB and bind to fibrillar Aβ. The concentrations used in these in vitro tests are much higher and represent amounts that can never be used when a radioactive tracer is injected in vivo. For 3.16, the lowest concentration used was 100 nM, which still is much higher than the concentration that can be reached by injection of 9 nmol. The question remains if these compounds will still be effective in vivo after injection because of this difference in concentration.

Transgenic AD mouse brain cortex

Figure 56. Fluorescence microscopic images of vibratome sections of transgenic AD mouse brain cortex stained with a 1 µM solution of 3.8 (left) or 3.16 (right). Arrows indicate vascular amyloid.

Although 3.2 showed good affinity in the test on human brain homogenates, this could not be confirmed in the affinity tests on transgenic AD mouse brain since no fluorescence could be visualized. Different filter sets

In vitro affinity studies of thioflavin-T derivatives for amyloid β

165

were used together with different concentrations of 3.2, but no usable images could be obtained. Further research is necessary to confirm whether this is due to a shift in the emission spectrum of 3.2. Application of 3.2 to post mortem human AD brain might be another possibility to see if there is also absence of fluorescence in human AD brain. Some authors claim that the amyloid β plaques in the brain of transgenic mice are slightly different from human AD amyloid β plaques, characterized by a lower number of binding sites for TT derived tracers in transgenic AD mouse brain (C. A. Mathis, personal communication). However, this is clearly not the case in the above-mentioned experiments. The two compounds stain amyloid plaques as well in human AD brain as in transgenic AD mouse brain.

A possible in vivo test is the i.v. injection of a carbon-11 or fluorine-18 labelled tracer in a transgenic AD mouse, followed by visualization of the distribution of the tracer throughout the body with the aid of a micro-PET camera. In such a study, the test animal is not sacrificed. This technique should enable to visualize retention of the tracer in the brain since amyloid plaques are present in the brain of these transgenic AD mice. A similar ex vivo test has already been performed on transgenic AD mice (provided by Dr. F. Van Leuven), but some practical problems during this test have so far been an obstacle to a successful outcome.

5.4. Conclusion In vitro tests for the determination of the affinity characteristics of tracer agents for amyloid β require the use of the authentic products, since the concentrations of the labelled compounds are far too low for this purpose. With regard to the technetium-99m labelled tracer agents, the authentic products with rhenium were not yet available. However, it would be useful to have the ability to test the affinity of the rhenium complex of for amyloid β fibrils. A modified test for these

99m

99m

Tc-tricine-2.12

Tc-labelled compounds was

166

Chapter 5

used in which TT was used as the competitor. From these data it was clear that in the group of the

99m

Tc-labelled compounds only technetium-99m

labelled tricine-2.12 showed affinity for amyloid β. The non-radioactive analogues of the carbon-11 and fluorine-18 labelled tracer agents were used to perform the affinity test as described in literature. This resulted in Ki values of the authentic products which are equal to the affinity value of their radiolabelled analogues. The results indicate that the benzimidazole, apart from its favourable biological behaviour, has no affinity

for

amyloid β.

The

three

studied

positron

emitter

labelled

benzothiazoles (3.2, 3.8 and 3.16), on the other hand, showed affinity profiles comparable to that of TT and also bind with high affinity to post mortem human AD brain homogenates. Because of these useful characteristics, binding studies were also performed with the authentic non-radioactive analogues of these three positron emitter labelled benzothiazoles on transgenic AD mouse brain sections and human AD brain sections.

It may be concluded from these findings that technetium-99m labelled TT derivatives have to be derivatized at other sites than the aromatic amine in order to maintain their affinity for amyloid β, although reduction of the amide to an amine might solve the problem. The best results were obtained with carbon-11 or fluorine-18 labelled benzothiazoles which have good biological properties combined with a good affinity for amyloid β. The two carbon-11 labelled benzothiazoles also show good binding to AD brain sections and further evaluation of these tracer agents or derivatives thereof has a reasonable chance to result in a radiopharmaceutical with suitable characteristics for diagnosis and evaluation of AD. This will definitely be of high importance for the diagnosis of the disease, but also for the follow-up of the effects of newly developed drugs for the treatment of this disease, which is at present thoroughly being examined.

167

CHAPTER 6

6. GENERAL CONCLUSION

The general aim of this study was the development of radiolabelled thioflavinT derivatives for the in vivo visualization of amyloid deposits in peripheral organs in case of peripheral amyloidosis (PA), and of amyloid β deposits in the brain in case of Alzheimer’s disease (AD).

Different issues must be kept in mind with regard to the studied technetium-99m labelled tracers for PA and AD. A technetium labelled tracer for use in clinical practice should be available in kit form. Such kit should be formulated in a way that only the addition of pertechnetate from a

99

Mo/99mTc

generator is needed, combined with an incubation step (at room temperature or with heating). From the RP-HPLC chromatograms of the technetium-99m labelled tracers described in this work, it is clear that a purification step with RP-HPLC is necessary. This would result in an impractically long preparation time for use in routine clinical practice, but this problem could be avoided by further optimization of the labelling conditions for a selected compound. Furthermore, the biological characteristics of these agents are far from ideal, although they can still be useful to detect PA in other body parts than the abdomen if they show affinity for amyloid plaques. From this point of view, it would be useful to test the affinity of

99m

Tc-EDDA-2.12 and of

99m

Tc-

tricine/nicotinic acid-2.12 for amyloid β fibrils in comparison to TT. In this way it could be elucidated whether derivatisation at other positions than the aromatic amine yields thioflavin-T derivatives with affinity for amyloid plaques. Another possibility to test

99m

Tc-tricine-2.12 would be the use of human tissue

sections containing amyloid plaques (e.g. kidney, liver, skin tissue). If this would confirm the affinity of 99mTc-tricine-2.12 for amyloid, toxicity data are still necessary before a clinical evaluation can be considered. For tracer agents, the potential toxicity of the final

99m

99m

Tc-labelled

Tc-complex is of minor or no

168

Chapter 6

concern, as it is present in nano- to picomolar amounts. However, if a kit preparation is used, the precursor and additional materials (stannous salt, buffer, weak chelating agent…) remain present in (usually) milligram amounts and their safety has to be confirmed. Moreover, the uptake in the abdomen should be reduced. Potential strategies to realise this goal consist in adding other hydrophilic groups to the compound or increasing the charge on the tracer. Research into a technetium-99m labelled tracer remains an important issue because of the ideal characteristics of this radionuclide for imaging and because of its easy and continuous availability. Further research on

99m

Tc-

tricine-2.12 is justified as this tracer agent may have some potential for in vivo visualization of amyloid plaques in PA. Both the in vivo and the in vitro properties of the three other studied technetium-99m labelled tracer agents (99mTc-2.4,

99m

Tc-2.13 and

99m

Tc-2.14) are unfavourable, so further study of

these tracers does not seem to be meaningful.

The three carbon-11 labelled compounds and the fluorine-18 labelled tracer developed in this study, show a more promising picture than the technetium-99m labelled tracers. First of all, they have a relatively good in vivo behaviour with reasonable uptake in cerebrum and cerebellum and fast wash-out of the brain in healthy mice. With regard to the affinity for amyloid of the positron emitter labelled TT derivatives, it became clear that the authentic non-radioactive benzothiazoles show good amyloid β affinity as opposed to the authentic non-radioactive benzimidazole. A few other benzimidazoles were also tested and also lacked affinity for amyloid β (data not described in this thesis). The in vitro affinity of these benzothiazoles for amyloid β fibrils and post mortem human AD brain homogenates together with their good BBB passage, make them possible candidates for in vivo visualization of amyloid β in the brain of AD patients. The affinity tests on transgenic AD mice brain sections and post mortem human AD brain sections nevertheless show that only the two carbon-11 labelled benzothiazoles have a sufficient affinity for the amyloid β plaques present in the brain of these mice and of human AD brain. These two carbon-11 labelled tracer agents therefore are the most promising candidates for further in vivo tests. At this stage of the study, their clinical use

General conclusion

169

is, however, hampered by the low radioactive synthesis yield, but optimization is probably possible. Surprisingly, a structure analogue of TT (including the positive charge useful for visualization of PA) did not show in vitro affinity for amyloid β fibrils. Further tests are necessary to confirm this lack of affinity.

In general, it can be concluded that some of the newly developed tracers show promising characteristics to justify their clinical evaluation. From this point of view, a technetium-99m labelled tracer (ideal physical characteristics and continuous availability) or a fluorine-18 labelled tracer (longer half-life than carbon-11) receives the most interest. It is, however, rather unlikely that a

99m

Tc-labelled agent could be developed with suitable

characteristics for in vivo diagnosis of AD.

Future perspectives include the elaboration of a suitable formulation form for the two most promising tracer agents (3.20 and 3.21) and the search for tracer agents that can interact with other steps in the amyloid cascade hypothesis (e.g. BACE, γ-secretase, soluble amyloid…).

171

CHAPTER 7

7. SUMMARY

Amyloidosis is characterized by the presence of pathological deposits of amyloid masses which can be present in almost every part of the human body. The consequence of such amyloid deposits is a change in organstructure which will lead to a loss of function and ultimately to the death of the patient. All forms of amyloidosis are caused by an abnormal folding of proteins which bear the capacity of misfolding and are therefore named chameleon proteins. This abnormal folding leads to proteins with a high content of βpleated sheets, characterized by their high insolubility and rigidity. The amyloid deposits are present in a wide variety of clinical disorders including rheumatoid arthritis, diabetes, chronic renal dialysis, Alzheimer’s disease (AD) … Their widespread occurrence and the difficulties in diagnosing these diseases make it impossible to obtain prevalence data. Diagnosis is still only possible by taking tissue biopsies in case of systemic amyloidosis and by staining them, e.g. with Congo red which leads to a specific apple green birefringence of the plaques when viewed under polarized light. The same is true for AD where a definite diagnosis of the disease is only possible post mortem when brain autopsy sections can be stained with Congo red or other stains. The techniques necessary for diagnosing amyloidosis suffer from the dangers accompanying tissue biopsying, the small parts biopsied and the necessity of experienced personnel. In the case of AD, relevant data on the patient’s symptoms can only be obtained post mortem. As a result of all these factors amyloid plaques are often not diagnosed or are missed due to the presence of other symptoms. Furthermore, some of these protein deposits become common in an aging population like the amyloid β protein in AD and the amyloid deposits in the islets of Langerhans in type II diabetes mellitus. This compromises the diagnosis even more because the presence of these protein deposits does not necessarily mean that the disease is also present.

172

Summary Although there are a few techniques to diagnose amyloidosis like the

use of

123

I-SAP, these methods are all restricted to specialized laboratories

and they are not without risk (in case of use of a human protein). Furthermore,

123

123

I-SAP the risks associated with the

I-SAP cannot be used to detect AD.

The latter can up-to-now only be diagnosed during life with the aid of psychological tests and clinical examination of the patient, techniques which offer only indirect evidence of the disease. Although the diagnosis of dementia is often easy, the differential diagnosis between AD and other forms of dementia is not. It would therefore be useful to have a radiolabelled tracer which can be used to visualize the amyloid deposits in vivo in a non-invasive way.

The aim of this study was therefore the synthesis, labelling and evaluation of radiolabelled tracer agents which would allow a non-invasive diagnosis of peripheral amyloidosis (PA) and AD during life, preferentially at an early stage. On the basis of experience acquired in our and other research groups, it was decided to develop and evaluate derivatives of thioflavin-T (TT) in view of its affinity for amyloid. Labelling of the tracer should occur with a radionuclide emitting γ-rays which can be detected outside the body with the aid of a SPECT or PET camera. Furthermore, the tracer agents should show affinity for amyloid plaques. Thioflavin-T as such has a benzothiazole structure with a quaternary nitrogen atom in the thiazole part which generates a positively charged molecule. It has been reported however, that uncharged derivatives of TT show an even higher affinity for amyloid than the charged TT.

The second chapter of this study describes the preparation of four new technetium-99m labelled derivatives of TT, consisting of a conjugate of a 2phenylbenzothiazole (PBT) and a complex of

99m

Tc with a bifunctional

chelating ligand (BCL). One thioflavin-T derivative was coupled to a bisaminoethanethiol (BAT) ligand of which the

99m

Tc-complex is known to show

good uptake in the brain. The protection groups present in the BCL were partially removed by heating the precursor in diluted hydrochloric acid and an

Summary

173

exchange labelling with tartrate as the weak technetium-99m chelating agent was then performed. To allow mass spectrometry of these

99m

Tc-labelled

tracer agents carrier technetium-99 was added to the labelling reaction mixture and this enabled identity confirmation of the structure. The three other 99m

Tc-labelled derivatives of TT were designed as potential tracer agents for

detection of PA. In one of these tracers, the PBT was coupled to a mercaptotriamide tetraligand of which the thiol was protected with a benzyl group. A one-step deprotection and labelling procedure was developed using an exchange labelling with tartrate at pH 9. The yield of the labelling reaction was high and LC-MS data for this agent were in accordance with the proposed structure. In the second compound a PBT was coupled to a HYNIC precursor in view of reported hydrophilic properties for

99m

Tc-tracers bearing

this HYNIC precursor. Labelling of this precursor requires the presence of a co-ligand for which tricine, EDDA and a combination of tricine/nicotinic acid were tested. The complex with tricine as co-ligand showed the most favourable properties. LC-MS of this agent suggested a radiolabelled complex consisting of the PBT-HYNIC conjugate, a technetium atom, a tricine molecule and two molecules of water. The third tracer consisted of a PBT coupled to nitrolotriacetic acid and was labelled as a technetium tricarbonyl complex. This resulted in a charged complex in high yield and LC-MS confirmed the identity of the compound. Technetium tricarbonyl complexes are as a rule more lipophilic than technetium(V)oxo complexes.

Chapter 3 describes four derivatives of thioflavin-T labelled with carbon-11 or fluorine-18 and designed for in vivo visualization of AD. The labelled compound was obtained by reaction of

18

18

F-

F-fluoride with a 4’-

nitrophenylbenzothiazole precursor, of which the nitro group in para-position of the thiazole moiety is activated. Identity of the radiolabelled PBT was confirmed by comparison with the retention time on RP-HPLC of the authentic non-radioactive compound, as LC-MS analysis of this non-ionisable was not possible. Moreover, two 11

11

18

F-PBT

C-labelled phenylbenzothiazoles and one

C-phenylbenzimidazole were prepared. The two phenylbenzothiazoles are

structure isomers in which only the position of a hydroxyl-group on the benzothiazole ring differs. Labelling was performed in both cases by

11

C-

174

Summary

methylation of the aromatic amine on the 2-phenylring, but the yield was very low. This made it impossible to obtain relevant LC-MS data of these radiolabelled compounds due to the low amounts present in the labelling reaction mixtures. For this reason, the identity of the labelled compounds was confirmed by comparison with the retention time of the authentic nonradioactive compounds on RP-HPLC. The labelled 2-phenylbenzimidazole was also prepared by

11

C-methylation, but the

11

C-methyl group was in this

case introduced on the imidazole nitrogen atom. The identity of the labelled compound was supported by LC-MS analysis.

Chapter four describes the biological properties of the eight newly developed tracer agents. Biodistribution characteristics were studied in healthy mice. They all showed high liver uptake followed by excretion to the intestines. Although some of the technetium-99m labelled tracer agents are rather hydrophilic, the uptake in kidneys and excretion to urine was higher for the carbon-11 and fluorine-18 labelled tracer agents. The presence of a hydroxyl group on the benzothiazole moiety resulted in a higher renal excretion. For the technetium-99m labelled tracers designed for diagnosis of PA the abdominal uptake was too high to be useful for in vivo visualization of amyloid deposits in the abdomen. They may nevertheless be useful for visualizing deposits in regions located out of the splanchnic area. The reason for this high abdominal uptake is most certainly the presence of a bulky technetium binding ligand which renders the tracer agents more lipophilic and leads to an increase in molecular mass. The latter probably also is partially responsible for the low brain uptake of the

99m

Tc-labelled agent designed for

diagnosis of AD. This can be concluded from comparison with the properties of the positron emitter labelled tracers which have a lower molecular mass, lack a bulky BCL and show a clearly higher brain uptake. Passage over the BBB was indeed favourable for the carbon-11 and fluorine-18 labelled tracers which grants these tracers with the best in vivo characteristics for visualization of amyloid β plaques. Nevertheless, upscaling of the labelling yield of the two carbon-11 labelled benzothiazoles is necessary to allow a potential clinical evaluation of these promising in vivo tracers.

Summary

175

Determination of the affinity for synthetic amyloid β fibrils was not straightforward

for

the

technetium-99m

labelled

tracers.

Since

the

corresponding authentic non-radioactive rhenium complexes were not available, the affinity test was modified to obtain information on the affinity of the technetium-99m labelled compounds for synthetic amyloid β fibrils. Only the labelled compound that was not derivatized on the aromatic amine of the PBT showed affinity for amyloid β fibrils with thioflavin-T as the competitor. In the case of the positron emitting tracers we synthesized the corresponding authentic non-radioactive products and these were used in the affinity tests. The results showed a good affinity of the three phenylbenzothiazoles for synthetic amyloid β fibrils, as well as for post mortem human AD brain homogenates, while the phenylbenzimidazole did not show any affinity in both tests. The two carbon-11 labelled benzothiazoles furthermore showed good binding affinity to transgenic AD mice brain sections and to post mortem human AD brain sections, while this could not be assessed for the fluorine-18 labelled benzothiazole.

The results of this research clearly show that the two benzothiazoles labelled with carbon-11 show the best in vivo and in vitro properties. These tracer agents thus have potential for the visualization of amyloid β in vivo. The development of such compounds was essentially the aim of this study.

177

Samenvatting

SAMENVATTING

Amyloïdose wordt gekenmerkt door de aanwezigheid van pathologische amyloïddeposities die in bijna elk orgaan van het menselijk lichaam kunnen aanwezig zijn. Het gevolg van dergelijke deposities is een wijziging in de orgaanstructuur die leidt tot functieverlies en uiteindelijk eventueel tot het overlijden van de patiënt. De verschillende vormen van amyloïdose worden veroorzaakt door een abnormale opplooiing van eiwitten en de mogelijkheid tot het vormen van verschillende tertiaire structuren heeft hen de naam ‘kameleon eiwitten’ bezorgd. Het abnormale opvouwen geeft aanleiding tot eiwitten met een hoog ‘β-sheet’ gehalte die worden gekenmerkt door hun rigiditeit en geringe oplosbaarheid. Deze amyloïddeposities zijn aanwezig in een waaier van klinische aandoeningen waaronder reumatoïde arthritis, diabetes, chronische nierdialyse, de ziekte van Alzheimer (zvA)… De diagnose van perifere amyloïdose (PA) is nog altijd alleen maar mogelijk aan de hand van weefselbiopsies die worden gekleurd met Congorood, wat aanleiding geeft tot een specifieke appel-groene kleur van de deposities onder gepolariseerd licht. De huidige technieken voor de diagnose van amyloïdose hebben de nadelen en beperking van weefselbiopsies, namelijk de beperkte biopsieregio en de noodzaak van gespecializeerd personeel. Ook bij de zvA is een definitieve diagnose alleen mogelijk post mortem als autopsie hersencoupes kunnen gekleurd worden met Congorood of andere stoffen. Als gevolg hiervan worden amyloïd plaques vaak niet worden gediagnosticeerd of over het hoofd gezien omwille van de aanwezigheid van andere symptomen. Bovendien zijn sommige van deze eiwitdeposities ook onder normale omstandigheden aanwezig in een ouder wordende populatie zoals het amyloïd β eiwit in de zvA en de amyloïddeposities in de eilandjes van Langerhans bij type II diabetes mellitus. Dit maakt het stellen van een diagnose nog ingewikkelder.

178

Samenvatting Er bestaan weliswaar enkele technieken voor de diagnose van

amyloïdose zoals het gebruik van

123

I-SAP, maar de toepassing ervan is

beperkt tot gespecializeerde laboratoria en ze zijn niet helemaal zonder risico (in het geval van

123

I-SAP omwille van het gebruik van humane eiwitten).

Bovendien kan 123I-SAP niet gebruikt worden om de zvA op te sporen. De zvA kan momenteel tijdens het leven alleen worden gediagnosticeerd met behulp van psychologische testen en een klinisch onderzoek van de patiënt. Beide testen leveren alleen een indirect bewijs van de ziekte. Alhoewel de diagnose van dementie vaak eenvoudig te stellen is, is de differentiële diagnose tussen de zvA en andere vormen van dementie dat niet. Het zou daarom nuttig zijn een radioactief gemerkt product te hebben dat kan gebruikt worden om amyloïddeposities in vivo aan te tonen op een niet-invasieve manier via scintigrafische beeldvorming van buiten de patiënt.

Het doel van dit onderzoek was daarom de synthese, merking en evaluatie van radioactief gemerkte verbindingen die een niet-invasieve diagnose van PA en de zvA mogelijk maken in vivo, liefst in een vroeg stadium. De opgedane ervaring in onze en andere onderzoeksgroepen deed ons besluiten derivaten van thioflavine-T (TT) te ontwikkelen omwille van zijn affiniteit voor amyloïd. Merking van deze derivaten moet gebeuren met een radionuclide dat gammastralen uitzendt die kunnen gedetecteerd worden buiten het lichaam met behulp van een SPECT of PET camera. Bovendien moeten deze derivaten affiniteit vertonen voor amyloïd plaques. Thioflavine-T heeft een benzothiazool structuur met een quaternair stikstofatoom in het thiazool gedeelte en is daarom een positief geladen molecule. In de literatuur werd gerapporteerd dat ongeladen derivaten van TT een zelfs hogere affiniteit voor amyloïd hebben dan TT zelf en daarom leken ze het beste uitgangspunt.

Het tweede hoofdstuk van dit onderzoek beschrijft de bereiding van vier nieuwe technetium-99m gemerkte derivaten van TT die zijn opgebouwd uit een conjugaat van een 2-fenylbenzothiazool (FBT) en een complex van 99m

Tc met een bifunctionele chelaterende ligand (BCL). Een eerste TT

derivaat werd gekoppeld aan een bis-amino-ethaanthiol (BAT) ligand waarvan geweten is dat het 99mTc-complex goede opname vertoont in de hersenen. De

Samenvatting

179

schermgroepen op de thiolen en één amine in deze BCL werden gedeeltelijk verwijderd in verdund zoutzuur en een uitwisselingsmerking met technetium99m werd vervolgens uitgevoerd met tartraat als intermediaire chelaterende stof. Technetium-99 werd toegevoegd aan het merkingsmengsel om massaspectrometrie mogelijk te maken en zo de identiteit van het tracerproduct te bevestigen. De drie andere

99m

Tc gemerkte derivaten van TT

werden ontwikkeld als potentiële tracerproducten voor de detectie van PA. Het

FBT

in

een

tweede

tracerproduct

werd

gekoppeld

aan

een

mercaptotriamide tetraligand met een S-benzyl beschermde thiolgroep. Ontscherming en merking gebeurden in één stap via een uitwisselingsmerking met tartraat bij pH 9. Massaspectrometrie gekoppeld aan hoge druk vloeistofchromatografie (HDVC) bevestigde de vooropgestelde structuur van dit product. In een tweede product werd het FBT gekoppeld aan een hydrazinonicotinezuur (HYNIC) precursor omwille van de gerapporteerde hydrofiele eigenschappen van

99m

Tc tracers die een HYNIC gedeelte

bevatten. Voor de merking van deze precursor werden tricine, EDDA of een combinatie van tricine/nicotinezuur als co-ligand uitgetest. Het complex met tricine

als

co-ligand

vertoonde

de

meest

gunstige

eigenschappen.

Massaspectrometrie gekoppeld aan HDVC van dit product suggereerde een radioactief complex bestaande uit het FBT-HYNIC conjugaat, een technetium atoom, een tricine molecule en twee molecules water. Het derde tracerproduct bestond uit een FBT gekoppeld aan nitrilotriazijnzuur en werd gemerkt als een negatief geladen technetium tricarbonyl complex waarvan de identiteit werd bevestigd met behulp van massaspectrometrie gekoppeld aan HDVC

Hoofdstuk 3 beschrijft synthese en merking van vier derivaten van TT gemerkt met koolstof-11 of fluor-18, ontworpen als tracerproducten voor de in vivo visualisatie van de zvA. Het fluor-18 gemerkte product werd bekomen door reactie van

18

F-fluoride met een 4’-nitrofenylbenzothiazool precursor

waarvan de nitrogroep geactiveerd is omwille van zijn para-positie ten opzichte van het thiazool gedeelte. De identiteit van het radioactief gemerkte FBT werd bevestigd door vergelijking van zijn retentietijd op HDVC met de retentietijd van zijn authentiek, niet-radioactief analoog omwille van het feit dat dit product niet ioniseerbaar is bij massaspectrometrie. Verder werden twee

180

Samenvatting

koolstof-11 gemerkte FBT’s en een koolstof-11 gemerkt fenylbenzimidazool gemaakt. De twee FBT’s zijn structuur isomeren waarbij enkel de positie van een hydroxylgroep op de fenylring van het benzothiazoolgedeelte verschilt. Merking werd in beide gevallen uitgevoerd door

11

C-methylering van het

aromatische amine op de 2-fenylring, maar de opbrengst was zeer laag. Dit maakte het onmogelijk relevante data te verkrijgen via massaspectrometrie van deze tracerproducten. Daarom werd de identiteit van deze gemerkte tracerproducten bevestigd door vergelijking met de retentietijden van de authentieke niet-radioactieve analogen op HDVC. Het 2-fenylbenzimidazool werd ook gemerkt via

11

C-methylering maar de

11

C-methylgroep kwam in dit

geval op het imidazool stikstofatoom. De identiteit van het gemerkte tracerproduct werd ondersteund door massaspectrometrie gekoppeld aan HDVC.

In hoofdstuk 4 worden de biologische eigenschappen van de acht nieuw ontwikkelde tracerproducten in gezonde muizen weergegeven. Al deze verbindingen vertonen hoge opname in de lever gevolgd door excretie naar de ingewanden. Alhoewel sommige van deze technetium-99m gemerkte tracerproducten tamelijk hydrofiel zijn, was de opname in de nieren en excretie naar de urine hoger voor de koolstof-11 en fluor-18 gemerkte tracerproducten.

De

aanwezigheid

van

een

hydroxylgroep

op

het

benzothiazool gedeelte resulteerde in een hogere renale excretie. Voor de technetium-99m gemerkte producten ontworpen voor diagnose van PA was de abdominale opname veel te hoog om bruikbaar te zijn voor in vivo visualisatie van amyloïddeposities in het abdomen. Deze verbindingen kunnen eventueel toch nuttig zijn voor visualisatie van deposities in organen verder verwijderd van het abdomen. De reden voor deze hoge abdominale opname is ongetwijfeld de aanwezigheid van een omvangrijke technetium bindende ligand die de moleculaire massa en het lipofiele karakter verhoogt. De hoge moleculaire massa is mee verantwoordelijk voor de lage hersenopname van de

99m

Tc gemerkte tracerproducten ontworpen voor

diagnose van de zvA. Dat kan worden afgeleid uit de vergelijking met de eigenschappen van de positron emitter gemerkte tracers die een lagere moleculaire massa hebben, geen grote BCL dragen en toch een duidelijk

Samenvatting

181

hogere hersenopname vertonen. De passage over de bloed-hersen barrière was

inderdaad

gunstig

voor

de

koolstof-11

en

fluor-18

gemerkte

fenylbenzothiazolen en deze tracerproducten vertoonden dan ook de beste in vivo karakteristieken voor visualisatie van amyloïd β plaques. Niettemin moet de merkingsopbrengst van de twee koolstof-11 gemerkte benzothiazolen nog worden

verhoogd

om

een

potentiële

klinische

evaluatie

van

deze

veelbelovende in vivo tracerproducten toe te laten. De bepaling van de affiniteit voor synthetische amyloïd β fibrillen was niet eenvoudig voor de technetium-99m gemerkte tracerproducten. Omdat de overeenkomstige

authentieke

niet-radioactieve

renium

complexen

niet

beschikbaar zijn, werd de affiniteitstest aangepast om toch gegevens te verkrijgen

over

de

affiniteit

van

deze

technetium-99m

gemerkte

tracerproducten. Alleen het tracerproduct dat niet gederivatiseerd was op het aromatische amine van het FBT vertoonde affiniteit voor amyloïd β fibrillen met TT als competitor. In het geval van de positron emitter tracerproducten werden

de

analoge

authentieke

niet-radioactieve

producten

wel

gesynthetiseerd en deze werden gebruikt in de affiniteitstesten. De resultaten toonden aan dat de drie benzothiazolen een goede affiniteit bezitten zowel voor

synthetische

amyloïd

β

fibrillen

als

voor

post

mortem

hersenhomogenaten van een zvA patiënt, terwijl het benzimidazool geen enkele affiniteit vertoonde in beide testen. De twee koolstof-11 gemerkte benzothiazolen

vertoonden

bovendien

goede

bindingsaffiniteit

aan

hersensecties van transgene zvA muizen en aan post mortem hersencoupes van een patiënt met de zvA. Dit kon niet worden aangetoond voor het fluor-18 gemerkte benzothiazool.

De resultaten van dit onderzoek tonen aan dat de twee benzothiazolen gemerkt met koolstof-11 de beste in vivo en in vitro eigenschappen bezitten. Het is duidelijk dat deze tracerproducten potentiële kandidaten zijn als geschikt tracerproduct voor in vivo visualisatie van amyloïd β wat uiteindelijk het doel was van deze studie.

183

REFERENCES

REFERENCES 1

Glenner, G. G. Amyloid deposits and amyloidosis. The β-fibrilloses (first of two parts). N Engl J Med, 1980; 302: 1283-1292

2

Virchow, R. Ueber eine im Gehirn und Rückenmark des Menschen aufgefundene Substanz mit der chemischen Reaktion der Cellulose. Virchows Arch Pathol Anat Physiol Klin Med, 1854; 6: 135-138, 268-271, 416-426

3

Kyle, R. A.; Bayrd, E. D. Amyloidosis: review of 236 cases. Medicine, 1975; 54: 271-299

4

Glenner, G. G.; Harada, M.; Isersky, C. The purification of amyloid fibril proteins. Prep Biochem, 1972; 2: 39-51

5

Ghiso, J.; Frangione, B. Amyloidosis and Alzheimer’s disease. Adv Drug Deliv Rev, 2002; 54: 1539-1551

6

Merlini, G. M.; Bellotti, V. Mechanisms of disease. Molecular mechanisms of amyloidosis. N Engl J Med, 2003; 349: 583-596

7

Westermark, P.; Benson, M. D.; Buxbaum, J. N.; Cohen, A. S.; Frangione, B.; Ikeda, S.; Masters, C. L.; Merlini, G.; Saraiva, M. J.; Sipe, J. D. Amyloid fibril protein nomenclature 2002. Amyloid, 2002; 9: 197-200

8

Perutz, M.F. Amyloid fibrils: mutations make enzyme polymerize. Nature, 1997; 385: 773775

9

Cohen, A. S. Amyloidosis. Bull Rheum Dis, 1991; 40: 1-12

10

Lachmann, H. J.; Hawkins, P. N. Novel pharmacological strategies in amyloidosis. Nephron Clin Pract, 2003; 94: c85-c88

11

Kisilevsky, R. From arthritis to Alzheimer’s disease: current concepts on the pathogenesis of amyloidosis. Can J Physiol Pharmacol, 1987; 65: 1805-1815

12

WHO-IUIS nomenclature sub-committee. Nomenclature of amyloid and amyloidosis. Bull World Health Organ, 1993; 71: 105-108

13

Dobson, C. M.; Karplus, M. The fundamentals of protein folding: bringing together theory and experiment. Curr Opin Struct Biol, 1999; 9: 92-101

14

Hurle, M. R.; Helms, L. R., Li, L.; Chan, W.; Wetzel, R. A role for destabilizing amino acid replacement in light-chain amyloidosis. Proc Natl Acad Sci USA, 1994; 91: 5446-5450

15

Stevens, F. J. Four structural risk factors identify most fibril-forming kappa light chains. Amyloid, 2000; 7: 200-211

16

Chiti, F.; Calamai, M.; Taddei, N., Stefani, M.; Ramponi, G.; Dobson, C. M. Studies of the aggregation of mutant proteins in vitro provide insights into the genetics of amyloid diseases. Proc Natl Acad Sci USA, 2002; 99: suppl 4: 16419-16426

17

Looi, L. M. The pattern of amyloidosis in a Malaysian patient population. Histopathology, 1991; 18: 133-141

18

Gejyo, F.; Yamada, T.; Odani, S.; Nakagawa, Y.; Arakawa, M.; Kunitomo, T.; Kataoka, H.; Suzuki, M.; Hirasawa, Y.; Shirahama, T.; Cohen, A. S.; Schmid, K. A new form of amyloid protein associated with chronic hemodialysis was identified as β2-microglobulin. Biochem Biophys Res Commun, 1985; 129: 701-706

19

Stone, W. J.; Hakim, R. M. Beta-2-microglobulin amyloidosis in long-term dialysis patients. Am J Nephrol, 1989; 9: 177-183

20

Cornwell, G. G.; Westermark, P. Senile amyloidosis: a protean manifestation of the aging process. J Clin Pathol, 1980: 33: 1146-1152

21

Glenner, G. G. Amyloid deposits and amyloidosis. The β-fibrilloses (second of two parts). N Engl J Med, 1980, 302: 1333-1343

184

References

22

Cohen, A. S. Amyloidosis. N Engl J Med, 1967; 277: 522-530, 574-583, 628-638

23

Bennhold, H. Ein spezifische amyloid-farbung mit Kongorot. Munch Med Wochenschr, 1922; 69: 1537-1538

24

Puchtler, H.; Sweat, F.; Levine, M. On the binding of Congo red by amyloid. J Histochem Cytochem, 1962; 10: 355-364

25

Styren, S. D.; Hamilton, R. L.; Styren, G. C.; Klunk, W. E. X-34, a fluorescent derivative of congo red: a novel histochemical stain for Alzheimer's disease pathology. J Histochem Cytochem, 2000; 28: 1223-1232

26

Sen, S.; Basdemir, G. Diagnosis of renal amyloidosis using Congo red fluorescence. Pathol Int, 2003; 53: 534-538

27

Linke, R. P. Highly sensitive diagnosis of amyloid and various amyloid syndromes using Congo red fluorescence. Virchows Arch, 2000; 436: 439-448

28

Tan, S. Y.; Pepys, M. B. Amyloidosis. Histopathology, 1994; 25: 403-414B.

29

Binette, P.; Binette, M.; Calkins, E. The isolation and identification of the P-component of normal human plasma proteins. Biochem J, 1974; 143: 253-254

30

Haupt, H.; Heimburger, N., Kranz, T.; Baudner, S. Humanserumproteine mit hoher Affinität zu Carboxymethyl-cellulose, III. Hoppe Seylers Z Physiol Chem, 1972; 353: 1841-1849

31

Painter, R. H.; De Escallon, I.; Massey, A.; Pinteric, L.; Stern, S. B. The structure and binding characteristics of serum amyloid protein. Ann N Y Acad Sci, 1982; 389: 199-215

32

Hawkins, P. N.; Lavender, J. P.; Pepys, M. B. Evaluation of systemic amyloidosis by scintigraphy with 123I-labeled serum amyloid P component. N Engl J Med, 1990; 323: 508513

33

Hawkins, P. N.; Myers, M. J.; Lavender, J. P.; Pepys, M. B. Diagnostic radionuclide imaging of amyloid: biological targeting by circulating human serum amyloid P component. Lancet, 1988; i:1413-1418

34

Saïle, R.; Deveaux, M.; Hachulla, E.; Descamps, J.; Duquesnoy, B.; Marchandise, X. Iodine-123-labelled serum amyloid P component scintigraphy in amyloidosis. Eur J Nucl Med, 1993; 20: 130-137

35

Hawkins, P. N.; Richardson, S.; Vigushin, D. M.; David, J.; Kelsey, C. R.; Gray, R. E. S.; Hall, M. A.; Woo, P.; Lavender, J. P.; Pepys, M. B. Serum amyloid P component scintigraphy and turnover studies for diagnosis and quantitative monitoring of AA amyloidosis in juvenile rheumatoid arthritis. Arthritis Rheum, 1993; 36: 842-851

36

Hutchinson, W. L.; Mather, S. J.; Staltieri, M.; Noble, G. E.; Vigushin, D. M.; Tan, S. Y.; Seymour, A.; Pepys, M. B.; Hawkins, P. N. Scintigraphic imaging of amyloid deposits with 99m Tc-labelled serum amyloid P component. In: Kisilevsky, R.; Benson, M. D.; Frangione, B.; Gauldie, J.; Muckle, T. J.; Young, I. D.; eds. Amyloid and amyloidosis, 1993. Pearl River-New York: Parthenon Publishing; 1994: 682-684

37

Meegalla, S. K.; Plössl, K.; Kung, M-P.; Chumpradit, S.; Stevenson, D. A.; Kushner, S. A.; McElgin, W. T.; Mozly, P. D.; Kung, H. F. Synthesis and characterization of technetium99m-labeled tropanes as dopamine transporter-imaging agents. J Med Chem, 1997; 40: 917

38

Urbanyi, Z.; Medzihradszky, D. Rapid method to isolate serum amyloid P component from human plasma. Characterization of the isolated protein. J Chromatog, 1992; 578: 130-133

39

De Beer, F. C.; Pepys, M. B. Isolation of human C-reactive protein and serum amyloid P component. J Immunol Methods, 1982; 50: 17-31

40

Gertz, M. A.; Brown, M. L.; Hauser, M. F.; Kyle, R. A. Utility of technetium Tc 99m pyrophosphate bone scanning in cardiac amyloidosis. Arch Intern Med, 1987; 147: 10391044

References

185

41

Falk, R. H.; Lee, V. W.; Rubinow, A.; Hood, W. B.; Cohen, A. S. Sensitivity of technetium99m-pyrophosphate scintigraphy in diagnosing cardiac amyloidosis. Am J Cardiol, 1983; 51: 826-830

42

Worsley, D. F.; Lentle, B. C. Uptake of technetium-99m MDP in primary amyloidosis with a review of the mechanisms of soft tissue localization of bone seeking radiopharmaceuticals. J Nucl Med, 1993; 34: 1612-1615

43

Schaadt, B. K.; Hendel, H. W.; Gimsing, P.; Jonsson, V.; Pedersen, H.; Hesse, B. aprotinin scintigraphy in amyloidosis. J Nuc Med, 2003; 44: 177-183

44

Aprile, C.; Marinone, G.; Saponaro, R.; Bonino, C.; Merlini, G. Cardiac and pleuropulmonary AL amyloid imaging with technetium-99m labelled aprotinin. Eur J Nuc Med, 1995; 22: 1393-1401

45

Hirschfield, G. M.; Hawkins, P. N. Amyloidosis: new strategies for treatment. Int J Biochem Cell Biol, 2003; 35: 1608-1613

46

Katzman, R. Alzheimer’s disease. N Engl J Med, 1986; 314: 964-973

47

Selkoe, D. J. Alzheimer’s disease: genes, proteins and therapy. Physiol Rev, 2001; 81:741-766

48

Clark, C. M.; Karlawish, J. H. T. Alzheimer disease: current concepts and emerging diagnostic and therapeutic strategies. Ann Intern Med, 2003; 138: 400-411

49

Jorm, A. F.; Jolley, D. The incidence of dementia. A meta-analysis. Neurology, 1998; 51:728-733

50

Forsyth, E.; Ritzline, P. D. An overview of the etiology, diagnosis and treatment of Alzheimer’s disease. Phys Ther, 1998; 78: 1325-1331

51

Prigerson, H. G. Costs to society of family caregiving for patients with end-stage Alzheimer’s disease. N Eng J Med, 2003; 349: 1891-1892

52

Wisniewsky, T.; Ghiso, J.; Frangione, B. Biology of Aβ amyloid in Alzheimer’s disease. Neurobiol Dis, 1997; 4: 313-328

53

Selkoe, D. J.; Schenk, D. Alzheimer’s disease: molecular understanding predicts amyloid based therapeutics. Annu Rev Pharmacol Toxicol, 2003; 43: 545-584

54

Strittmatter, W. J.; Saunders, A. M.; Schmechel, D.; Pericak-Vance, M.; Enghild, J.; Salvesen, G. S.; Roses, A. D. Apolipoprotein E: high avidity binding to β-amyloid and increased frequency of type 4 allele in late-onset familial Alzheimer disease. Proc Natl Acad Sci USA, 1993: 1977-1981

55

Blacker, D.; Wilcox, M. A.; Laird, N. M.; Rodes, L.; Horvath, S. M.; Go, R. C.; Perry, R.; Watson, B. Jr.; Bassett, S. S.; McInnis, M. G.; Albert, M. S.; Hyman, B. T.; Tanzi, R. E. Alpha-2-macroglobulin is genetically associated with Alzheimer disease. Nat Genet, 1998; 19: 357-360

56

Bertram, L.; Blacker, D. ; Mullin, K.; Keeney, D.; Jones, J.; Basu, S.; Yhu, S.; McInnis, M. G.; Go, R. C.; Vekrellis, K.; Selkoe, D. J.; Saunders, A. J.; Tanzi, R. E. Evidence for genetic linkage of Alzheimer’s disease to chromosome 10q. Science, 2000; 290: 23022303

57

Kisilevsky, R.; Fraser, P. E. Aβ amyloidogenesis: unique, or variation on a systemic theme? Crit Rev Biochem Mol Biol, 1997; 32: 361-404

58

Esteban, J. A. Living with the enemy: a physiological role for the Trends Neurosci, 2004; 27: 1-3

59

Kamenetz, F.; Tomita, T.; Hsieh, H.; Seabrook, G.; Borchelt, D.; Iwatsubo, T.; Sisodia, S.; Malinow, R. APP processing and synaptic function. Neuron, 2003; 37: 925-937

60

Hung, A. Y.; Selkoe, D. J. Selective ectodomain phosphorylation and regulated cleavage of β-amyloid precursor protein. EMBO J, 1994; 13: 534-542

99m

Tc-

β−amyloid peptide.

186

References

61

Oltersdorf, T.; Ward, P. J.; Henriksson, T.; Beattie, E. C.; Neve, R.; Lieberburg, I.; Fritz, L. C. The Alzheimer amyloid precursor protein. Identification of a stable intermediate in the biosynthetic/degradation pathway. J Biol Chem, 1990; 265: 4492-4497

62

Walter, J.; Capell, A.; Hung, A. Y.; Langen, H.; Schnolzer, M.; Thinkaran, G.; Sisodia, S. S.; Selkoe, D. J.; Haass, C. Ectodomain phosphorylation of β-amyloid precursor protein at two distinct cellular locations. J Biol Chem, 1997; 1896-1903

63

Weidemann, A.; Konig, G.; Bunke, D.; Fischer, P.; Salbaum, J. M.; Masters, C. L.; Beyreuther, K. Identification, biogenesis and localization of precursors of Alzheimer’s disease A4 amyloid protein. Cell, 1989; 57: 115-126

64

Slunt, H. H.; Thinakaran, G.; Von Koch, C.; Lo, A. C. Y.; Tanzi, R. E.; Sisodia, S. S. Expression of a ubiquitous, cross-reactive homologue of the mouse β-amyloid precursor protein (APP). J Biol Chem, 1994; 269: 2637-2644

65

Wasco, W.; Bupp, K.; Magendantz, M.; Gusell, J.; Tanzi, R. E.; Solomon, F. Identification of a mouse brain cDNA that encodes a protein related to the Alzheimer diseaseassociated amyloid β-protein precursor. Proc Natl Acad Sci USA, 1992; 89: 10758-10762

66

Esch, F. S.; Keim, P. S.; Beattie, E. C.; Blacher, R. W.; Culwell, A. R.; Oltersdorf, T.; McClure, D.; Ward, P. J. Cleavage of amyloid β-peptide during constitutive processing of its precursor. Science, 1990: 248: 1122-1124

67

Sisodia, S. S.; Koo, E. H.; Beyreuther, K.; Unterbeck, A.; Price, D. L. Evidence that βamyloid protein in Alzheimer’s disease is not derived by normal processing. Science, 1990; 248: 492-495

68

Seubert, P.; Oltersdorf, T.; Lee, M. G.; Barbour, R.; Blomqvist; Davis, D. L.; Bryant, K.; Fritz, L. C.; Galasko, D.; Thai, L. J.; Lieberburg, I.; Schenk, D. B. Secretion of β-amyloid precursor protein cleaved at the amino-terminus of the β-amyloid peptide. Nature, 1993; 361: 260-263

69

Busciglio, J.; Gabuzda, D. H.; Matsudaira, P.; Yankner, B. A. Generation of β-amyloid in the secretory pathway in neuronal and non-neuronal cells. Proc Natl Acad Sci USA, 1993; 90: 2092-2096

70

Haass, C.; Schlossmacher, M.; Hung, A. Y.; Vigo-Pelfrey, C.; Mellon, A.; Ostaszwewski, B.; Lieberburg, I.; Koo, E. H.; Schenk, D.; Teplow, D.; Selkoe, D. J. Amyloid β-peptide is produced by cultured cells during normal metabolism. Nature, 1992: 359: 322-325

71

Seubert, P.; Vigo-Pelfrey, C.; Esch, F.; Lee, M.; Dovey, H.; Davis, D.; Sinha, S.; Schlossmacher, M. G.; Whaley, J.; Swindlehurst, C.; McCormack, R.; Wolfert. R.; Selkoe, D. J.; Lieberburg, I.; Schenk, D. Isolation and quantitation of soluble Alzheimer’s β-peptide from biological fluids. Nature, 1992; 359: 325-327

72

Shoji, M.; Golde, T. E.; Ghiso, J.; Cheung, R. T.; Estus, S.; Shaffer, L. M.; Cai, X.; McKay, D. M.; Tintner, R.; Frangione, B.; Younkin, S. G. Production of the Alzheimer amyloid β protein by normal proteolytic processing. Science, 1992; 258:126-129

73

Haass, C.; Hung, A. Y.; Schlossmacher, M. G.; Teplow, D. B.; Selkoe, D. J. β-amyloid peptide and a 3-kDa fragment are derived by distinct cellular mechanisms. J Biol Chem, 1993; 268: 3021-3024

74

Koo, E. H.; Squazzo, S. Evidence that production and release of amyloid-β protein involves the endocytic pathway. J Biol Chem, 1994; 269: 17386-17389

75

Perez, R. G.; Soriano, S.; Hayes, J. D.; Ostaszewski, B. L.; Xia, W.; Selkoe, D. J.; Chen, X.; Stokin, G. B.; Koo, E. H. Mutagenesis identifies new signals for β-amyloid precursor protein endocytosis, turnover and the generation of secreted fragments, including A42. J Biol Chem, 1999; 274: 18851-18856

76

Motter, R.; Vigo-Pelfrey, C.; Kholodenko, D.; Barbour, R.; Johnson-Wood, K.; Galasko, D.; Chang, L.; Miller, B.; Clark, C.; Green, R.; Olson, D.; Southwick, P.; Wolfert, R.; Munroe,

References

187

B.; Lieberburg, I.; Seubert, P.; Schenk, D. Reduction of beta-amyloid peptide 42 in the cerebrospinal fluid of patients with Alzheimer’s disease. Ann Neurol, 1995; 38: 643-648 77

Scheuner, D.; Eckman, C.; Jensen, M.; Song, X.; Citron, M.; Suzuki, N.; Bird, T. D.; Hardy, J.; Hutton, M.; Kukull, W.; Larson, E.; Levy-Lahad, E.; Vittanen, M.; Peskind, E.; Poorkaj, P.; Schellenberg, G.;Tanzi, R.; Wasco, W.; Lannfelt, L.; Selkoe, D. J.; Younkin, S. Secreted amyloid β-protein similar to that in the senile plaques of Alzheimer’s disease is increased in vivo by presenilin 1 and 2 and APP mutations linked to familial Alzheimer’s disease. Nat Med; 1996; 2: 864-870

78

Vassar, R.; Citron, M. Aβ-generating enzymes: recent advances in β- and γ-secretase research. Neuron, 2000; 27: 419-422

79

Haass, C. Take five-BACE and the γ-secretase quartet conduct Alzheimer’s amyloid peptide generation. EMBO J, 2004; 23: 483-488

80

De Strooper, B. Aph-1, Pen-2, and Nicastrin with Presenilin generate an active γ-secretase complex. Neuron, 2003; 38: 9-12

81

De Strooper, B.; Saftig, P.; Craessaerts, K.; Vanderstichele, H.; Guhde, G.; Annaert, W.; Von Figura, K.; Van Leuven, F. Deficiency of presenilin-1 inhibits the normal cleavage of amyloid precursor protein. Nature, 1998; 391: 387-390

82

Yu, G.; Nishimura, M.; Arawaka, S.; Levitan, D.; Zhang, L.; Tandon, A.; Song, Y. Q.; Rogaeva, E.; Chen, F.; Kawarai, T.; Supala, A.; Levesque, L.; Yu, H.; Yang, D-S.; Holmes, E.; Milman, P.; Liang, Y.; Zhang, D. M.; Xu, D. H.; Sato, C.; Rogaev, E.; Smith, M.; Janus, C.; Zhang, Y.; Aebersold, R.; Farrer, L.; Sorbi, S.; Bruni, A.; Fraser, P.; St George-Hyslop, P. Nicastrin modulates presenilin-mediated notch/glp-1 signal transduction and βAPP processing. Nature, 2000; 407: 48-54

83

Goutte, C.; Tsunozaki, M.; Hale, V. A.; Priess, J. R. APH-1 is a multipass membrane protein essential for the Notch signaling pathway in Caenorhabditis elegans embryos. Proc Natl Acad Sci USA, 2002; 99: 775-779

84

De Strooper, B.; Annaert, W.; Cupers, P.; Saftig, P.; Craessaerts, K.; Mumm, J. S.; Schroeter, E. H.; Schrijvers, V.; Wolfe, M. S.; Ray, W. J.; Goate, A.; Kopan, R. A presenilin-1-dependent γ-secretase-like protease mediates release of Notch intracellular domain. Nature, 1999; 398: 518-522

85

Golde, T. E.; Eckman, C. B.; Younkin, S. G. Biochemical detection of Aβ isoforms: implications for pathogenesis, diagnosis, and treatment of Alzheimer’s disease. Biochim Biophys Acta, 2000; 1502: 172-187

86

Joachim, C. L.; Morris, J. H.; Selkoe, D. J. Diffuse senile plaques occur commonly in the cerebellum in Alzheimer’s disease. Am J Pathol, 1989; 135: 309-319

87

Tagliavini, F.; Giaccone, G.; Frangione, B.; Bugiani, O. Preamyloid deposits in the cerebral cortex of patients with Alzheimer’s disease and nondemented individuals. Neurosci Lett, 1988; 93; 191-196

88

Yamaguchi, H.; Hirai, S.; Morimatsu, M.; Shoji; M.; Harigaya, Y. Diffuse type of senile plaques in the brains of Alzheimer-type dementia. Acta neuropathol, 1988; 77: 113-119

89

Lemere, C. A.; Lopera, F.; Kosik, K. S.; Lendon, C. L.; Ossa, J.; Saido, T. C.; Selkoe, D. J. Sequence of deposition of heterogeneous amyloid β-peptides and Apo E in Down syndrome: implications for initial events in amyloid plaque formation. Neurobiol Dis, 1996; 3: 16-32

90

Yankner, B. A.; Mesulam, M-M. β-amyloid and the pathogenesis of Alzheimer’s disease. N Engl J Med, 1991; 325: 1849-1857

91

Gamblin, T. C.; Chen, F.; Zambrano, A.; Abraha, A.; Lagalwar, S.; Guillozet, A. L.; Lu, M.; Fu, Y.; Garcia-Sierra, F.; LaPointe, N.; Miller, R.; Berry, R. W.; Binder, L. I.; Cryns, V. L.

β-

188

References

Caspase cleavage of tau: linking amyloid and neurofibrillary tangles in Alzheimer’s disease. Proc Natl Acad Sci USA, 2003; 100: 10032-10037 92

Rapoport, M.; Dawson, H. N.; Binder, L. I.; Vitek, M. P.; Ferreira, A. Tau is essential to βamyloid-induced neurotoxicity. Proc Natl Acad Sci USA, 2002; 99: 6364-6369

93

Rank, K. B.; Pauley, A. M.; Bhattacharya, K.; Wang, Z.; Evans, D. B.; Fleck, T. J.; Johnston, J. A.; Sharma, S. K. Direct interaction of soluble human recombinant tau protein with Aβ1-42 results in tau aggregation and hyperphosphorylation by tau protein kinase II. FEBS Lett, 2002; 514: 263-268

94

Hardy, J. A.; Higgins, G. A. Alzheimer’s disease: the amyloid cascade hypothesis. Science, 1991; 256: 184-185

95

Friedland, R. P. Alzheimer’s disease: clinical features and differential diagnosis. Neurology, 1993; 43 (Suppl 4): S45-S51

96

McCarten, J. R. Recognizing dementia in the clinic: whom to suspect, whom to test. Geriatrics, 1997; 52 (suppl 2); S17-S21

97

Morris, J. C. Alzheimer’s disease: a review of clinical assessment and management issues. Geriatrics, 1997; 52 (suppl 2): S22-S25

98

McKhann, G.; Drachman, D.; Folstein, M.; Katzman, R.; Price, D.; Stadlan, E. M. Clinical diagnosis of Alzheimer’s disease: report of the NINCDS-ADRDA work group under the auspices of the department of health and human services task force on Alzheimer’s disease. Neurology, 1984; 34: 939-944

99

Tierney, M. C.; Fisher, R. H.; Lewis, A. J.; Zorzitto, M. L.; Snow, W. G.; Reid, D. W.; Nieuwstraten, P. The NINCDS-ADRDA work group criteria for the clinical diagnosis of probable Alzheimer’s disease: a clinicopathologic study of 57 cases. Neurology, 1988; 38: 359-364

100

Folstein, M.; Folstein, S.; McHugh, P. R. Mini-mental state: a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res, 1975; 12: 189-198

101

Blessed, G.; Tomlinson, B. E.; Roth, M. The association between quantitative measures of dementia and of senile change in the cerebral gray matter of elderly subjects. Br J Psychiatry, 1968; 114: 797-811

102

Reisberg, B.; Ferris, S. H. Brief cognitive rating scale (BCRS). Psychopharmacol Bull, 1988; 24: 629-636

103

Mohs, R. C.; Cohen, L. Alzheimer’s disease assessment scale (ADAS). Psychopharmacol Bull, 1988; 24: 627-628

104

Morris, J. C.; Heyman, A.; Mohs, R. C.; Hughes, J. P.; van Belle, G.; Fillenbaum, G.; Mellits, E. D.; Clark, C.; CERAD investigators. The consortium to establish a registry for Alzheimer’s disease (CERAD). Part I; Clinical and neuropsychological assessment of Alzheimer’s disease. Neurology, 1989; 39: 1159-1165

105

Khachaturian, Z. S. Diagnosis of Alzheimer’s disease. Arch Neurol, 1985; 42: 1097-1105

106

Davis, P. C.; Gray, L.; Albert, M.; Wilkinson, W.; Hughes, J.; Heyman, A.; Gado, M.; Kumar, A. J.; Destian, S.; Lee, C.; Duvall, E.; Kido, D.; Nelson, M. J.; Bello, J.; Weathers, S.; Jolesz, F.; Kikinis, R.; Brooks, M. The consortium to establish a registry for Alzheimer’s disease (CERAD). Part III. Reliability of a standardized MRI evaluation of Alzheimer’s disease. Neurology, 1992; 42: 1676-1680

107

Small, G. W. Use of neuroimaging to detect early brain changes in people at genetic risk for Alzheimer’s disease. Adv Drug Deliv Rev, 2002; 54: 1561-1566

108

Rossor, M. N. Neuroimaging in Alzheimer’s disease. Acta Neurol Belg, 1998; 98: 229-230

109

Majocha, R. E.; Reno, J. M.; Friedland, R. P.; VanHaight, C.; Lyle, L. R.; Marotta, C. A. Development of a monoclonal antibody specific for β/A4 amyloid in Alzheimer’s disease

References

189

brain for application to in vivo imaging of amyloid angiopathy. J Nuc Med, 1992; 33: 21842189 110

LeVine, H. Quantitation of β-sheet amyloid fibril structures with thioflavin T. Methods Enzymol; 1999; 309: 274-284

111

Yamamoto, T.; Hirano, A. A comparative study of modified Bielschowsky, Bodian and thioflavin S stains on Alzheimer’s neurofibrillary tangles. Neuropathol Appl Neurobiol, 1986; 12: 3-9

112

Vinters, H. V. Cerebral amyloid angiopathy. A critical review. Stroke, 1987; 18: 311-324

113

Mirra, S. S. The CERAD neuropathology protocol and consensus recommendations for the postmortem diagnosis of Alzheimer’s disease: a commentary. Neurobiol Aging, 1997; 18 (S4): S91-S94

114

Mirra, S. S.; Heyman, A.; McKeel, D.; Sumi, S. M.; Crain, B. J.; Brownlee, L. M.; Vogel, F. S.; Hughes, J. P.; van Belle, G.; Berg, L. et al. The consortium to establish a registry for Alzheimer’s disease (CERAD). Part II. Standardization of the neuropathologic assessment of Alzheimer’s disease. Neurology, 1991; 41: 479-486

115

Cutler, N. R.; Sramek, J. J. Review of the next generation of Alzheimer’s disease therapeutics: challenges for drug development. Prog Neuropsychopharmacol Biol Psychiatry, 2001; 25: 27-57

116

Mohr, E.; Nair, N. P.; Sampson, M.; Murtha, S.; Belanger, G.; Pappas, B.; Mendis, T. Treatment of Alzheimer’s disease with Sabeluzole: functional and structural correlates. Clin Neuropharmacol, 1997; 20: 338-345

117

Schenk, D.; Barbour, R.; Dunn, W.; Gorden, G.; Grajeda, H.; Guido, T.; Hu, K.; Huang, J.; Johnson-Wood, K.; Khan, K.; Kholodenko, D.; Lee, M.; Liao, Z.; Lieberburg, I.; Motter, R.; Mutter, L.; Soriano, F.; Shopp, G.; Vasquez, N.; Vandevert, C.. Walker, S.; Wogulis, M.; Yednock, T.; Games, D.; Seubert, P. Immunization with amyloid-β attenuates Alzheimerdisease-like pathology in the PDAPP mouse. Nature, 1999; 400:173-177

118

Frenkel, D.; Katz, O.; Solomon, B. Immunization against Alzheimer’s β-amyloid plaques via EFRH phage administration. Proc Natl Acad Sci USA, 2000; 97: 11455-11459

119

Schenk, D. B.; Seubert, P.; Lieberburg, I.; Wallace, J. β-peptide immunization. Arch Neurol, 2000; 57: 934-936

120

DeMattos, R. B.; Bales, K. R.; Cummins, D. J.; Paul, S. M.; Holtzman, D. M. Brain to plasma amyloid-β efflux: a measure of brain amyloid burden in a mouse model of Alzheimer’s disease. Science, 2002; 295: 2264-2267

121

Zhang, G.; Leibowitz, M. J.; Sinko, P. J.; Stein, S. Multiple-peptide conjugates for binding β-amyloid plaques of Alzheimer’s disease. Bioconjug Chem, 2003; 14: 86-92

122

Roberds, S. L.; Anderson, J.; Basi, G.; Bienkowski, M. J.; Branstetter, D. G.; Chen, K. S.; Freedman, S.; Frigon, N. L.; Games, D.; Hu, K.; Johnson-Wood, K.; Kappenman, K. E.; Kawabe, T. T.; Kola, I.; Kuehn, R.; Lee, M.; Liu, W.; Motter, R.; Nichols, N. F.; Power, M.; Robertson, D. W.; Schenk, D.; Schoor, M.; Shopp, G. M.; Shuck, M. E.; Sinha, S.; Svensson, K. A.; Tatsuno, G.; Tintrup, H.; Wijsman, J.; Wright, S.; McConlogue, L. BACE knockout mice are healthy despite lacking primary-secretase activity in the brain: implications for Alzheimer’s disease therapeutics. Hum Mol Genet, 2001; 10: 1317-1324

123

Luo, Y.; Bolon, B.; Kahn, S.; Bennett, B. D.; Babu-Kahn, S.; Denis, P.; Fan, W.; Kha, H.; Zhang, J.; Gong, Y.; Martin, L.; Louis, J-C.; Yan, Q.; Richards, W.G.; Citron, M.; Vassar, R. Mice deficient in BACE1, the Alzheimer’s β-secretase, have normal phenotype and abolished b-amyloid generation. Nat Neurosci, 2001; 4: 231-232

124

Dovey, H. F.; John, V.; Anderson, J. P.; Chen, L. Z.; de Saint Andrieu, P.; Fang, L. Y.; Freedman, S. B.; Folmer, B.; Goldbach, E.; Holsztynska, E. J.; Hu, K. L.; Johnson-Wood, K. L.; Kennedy, S. L.; Kholodenko, D.; Knops, J. E.; Latimer, L. H.; Lee, M.; Liao, Z.; Lieberburg, I. M.; Motter, R. N.; Mutter, L. C.; Nietz, J.; Quinn, K. P.; Sacchi, K. L.;

190

References

Seubert, P. A.; Shopp, G. M.; Thorsett, E. D.; Tung, J. S.; Wu, J.; Yang, S.; Yin, C. T.; Schenk, D. B.; May, P. C.; Altstiel, L. D.; Bender, M. H.; Boggs, L. N.; Britton, T. C.; Clemens, J. C.; Czilli, D. L.; Dieckman-McGinty, D. K.; Droste, J. J.; Fuson, K. S.; Gitter, B. D.; Hyslop, P. A. ; Johnstone, E. M.; Li, W-Y.; Little, S. P. ; Mabry, T. E.; Miller, F. D.; Ni, B.; Nissen, J. S.; Porter, W. J.; Potts, B. D. ; Reel, J. K. ; Stephenson, D. ; Su, Y.; Shipley, L. A.; Whitesitt, C. A.; Yin, T.; Audia, J. E. Functional gamma-secretase inhibitors reduce beta-amyloid peptide levels in brain. J Neurochem, 2001; 76: 173-181 125

Weggen, S.; Eriksen, J. L.; Das, P.; Sagi, S. A.; Wang, R.; Pietrzik, C. U.; Findlay, K. A.; Smith, T. E.; Murphy, M. P.; Bulter, T.; Kang, D. E.; Marquez-Sterling, N.; Golde, T. E.; Koo, E. H. A subset of NSAIDs lower amyloidogenic Aβ42 independently of cyclooxygenase activity. Nature, 2001; 414: 212-216

126

in ’t Veld, B. A.; Ruitenberg, A.; Hofman, A.; Launer, L. J.; van Duijn, C. M.; Stijnen, T.; Breteler, M. M. B.; Stricker, B. H. C. Nonsteroidal anti-inflammatory drugs and the risk of Alzheimer's disease. N Engl J Med, 2001; 345: 1515-1521

127

Han, H.; Cho, C-G.; Lansbury, P. T. Technetium complexes for the quantification of brain amyloid. J Am Chem Soc, 1996; 118: 4506-4507

128

Ashburn, T. T.; Han, H.; McGuinness, B. F.; Lansbury, P. T. Amyloid probes based on Congo red distinguish between fibrils comprising different peptides. Chem Biol, 1996; 3: 351-358

129

Klunk, W. E.; Debnath, M. L.; Pettegrew, J. W. Chrysamine G binding to Alzheimer and control brain: autopsy study of a new amyloid probe. Neurobiol Aging, 1995; 16: 691-698

130

Kroll, R. A.; Neuwelt, E. A. Outwitting the blood-brain barrier for therapeutic purposes: osmotic opening and other means. Neurosurgery, 1998; 42: 1083-1100

131

Mathis, C. A.; Mahmood, K.; Debnath, M. L.; Klunk, W. E. Synthesis of a lipophilic, radioiodinated ligand with high affinity to amyloid protein in Alzheimer's disease brain tissue. J Labelled Comp Radiopharm, 1997; 40: 94-95

132

Dezutter, N. A.; Dom, R. J.; de Groot, T. J.; Bormans, G. M.; Verbruggen, A. M. 99mTcMAMA-chrysamine G, a probe for beta-amyloid protein of Alzheimer’s disease. Eur J Nuc Med, 1999; 26: 1392-1399

133

Dezutter, N. A.; de Groot, T. J.; Busson, R. H.; Janssen, G. A.; Verbruggen, A. M. Preparation of 99mTc-N2S2 conjugates of chrysamine G, potential probes for the betaamyloid protein of Alzheimer’s disease. J Labelled Comp Radiopharm, 1999; 42: 309-324

134

Pardridge, W. P. CNS drug design based on principles of blood-brain barrier transport. J Neurochem, 1998; 70: 1781-1792

135

Morton, K. C.; King, C. M.; Baetcke, K. P. Metabolism of benzidine to N-hydroxy-N,N’diacetylbenzidine and subsequent nucleic acid binding and mutagenicity. Cancer Res, 1979; 39: 3107-3113

136

Klunk, W. E.; Wang, Y.; Huang, G-F.; Debnath, M. L.; Holt, D. P.; Mathis, C. A. Uncharged thioflavin-T derivatives bind to amyloid-beta protein with high affinity and readily enter the brain. Life Sci, 2001; 69: 1471-1484

137

Zhuang, Z-P.; Kung, M-P.; Hou, C.; Skovronsky, D. M.; Gur, T. L.; Plössl, K.; Trojanowski, J. Q.; Lee, V. M-Y.; Kung, H. F. Radioiodinated styrylbenzenes and thioflavins as probes for amyloid aggregates. J Med Chem, 2001; 44: 1905-1914

138

Lee, C-W.; Zhuang, Z-P.; Kung, M-P.; Plössl, K.; Skovronsky, D.; Gur, T.; Hou, C.; Trojanowski, J. Q.; Lee, V. M-Y.; Kung, H. F. Isomerization of (Z,Z) to (E,E)1-bromo-2,5bis-(3-hydroxycarbonyl-4-hydroxy)-styrylbenzene in strong base: probes for amyloid plaques in the brain. J Med Chem, 2001; 44: 2270-2275

139

Shi, D-F.; Bradshaw, T. D.; Wrigley, S.; McCall, C. J.; Lelieveld, P.; Fichtner, I; Stevens, M. F. G. Antitumor benzothiazoles. 3. Synthesis of 2-(4-aminophenyl)benzothiazoles and evaluation of their activities against breast cancer cell lines in vitro and in vivo. J Med Chem, 1996; 39: 3375-3384

References

191

140

Kashiyama, E.; Hutchinson, I.; Chua, M-S.; Stinson, S. F.; Philips, L. R.; Kaur, G.; Sausville, E. A.; Bradshaw, T. D.; Westwell, A. D.; Stevens, M. F. G. Antitumor benzothiazoles. 8. Synthesis, metabolic formation, and biological properties of the C- and N-oxidation products of antitumor 2-(4-aminophenyl)-benzothiazoles. J Med Chem, 1999; 42: 4172-4184

141

Chua, M-S.; Shi, D-F.; Wrigley, S.; Bradshaw, T. D.; Hutchinson, I.; Shaw, P. N.; Barrett, D. A.; Stanley, L. A.; Stevens, M. F. G. Antitumor benzothiazoles. 7. Synthesis of 2-(4acylaminophenyl)benzothiazoles and investigations into the role of acetylation in the antitumor activities of the parent amines. J Med Chem; 1999; 42: 381-392

142

Hutchinson, I.; Chua, M-S.; Browne, H. L.; Trapani, V.; Bradshaw, T. D.; Westwell, A. D.; Stevens, M. F. G. Antitumor benzothiazoles. 14. Synthesis and in vitro biological properties of fluorinated 2-(4-aminophenyl)benzothiazoles. J Med Chem, 2001; 44: 14461455

143

Wang, Y.; Klunk, W. E.; Huang, G-F.; Debnath, M. L.; Holt, D. P.; Mathis, C. A. Synthesis and evaluation of 2-(3’-iodo-4’-aminophenyl)-6-hydroxybenzothiazole for in vivo quantitation of amyloid deposits in Alzheimer’s disease. J Mol Neurosci, 2002; 19: 11-16

144

Mathis, C. A.; Backsai, B. J.; Kajdasz, S. T.; McLellan, M. E.; Frosch, M. P.; Hyman, B. T.; Holt, D. P.; Wang, Y.; Huang, G-F.; Debnath, M. L.; Klunk, W.E. A lipophilic thioflavin-T derivative for positron emission tomography (PET) imaging of amyloid in brain. Bioorg Med Chem Lett, 2002; 12: 295-298

145

Klunk, W. E.; Wang, Y.; Huang, G-F.; Debnath, M. L.; Holt, D. P.; Shao, L; Hamilton, R. L.; Ikonomovic, M. D.; DeKosky, S. T.; Mathis, C. A. The binding of 2-(4’methylaminophenyl)benzothiazole to postmortem brain homogenates is dominated by the amyloid component. J Neurosci, 2003; 23: 2086-2092

146

Klunk, W. E.; Engler, H.; Nordberg, A.; Wang, Y.; Blomqvist, G.; Holt, D. P.; Bergström, M.; Savitcheva, I.; Huang, G-F.; Estrada, S.; Ausén, B.; Debnath, M. L.; Barletta, J.; Price, J. C.; Sandell, J.; Lopresti, B. J.; Wall, A.; Koivisto, P.; Antoni, G.; Mathis, C. A.; Langström, B. Imaging brain amyloid in Alzheimer’s disease with Pittsburgh coumpound-B. Ann Neurol, 2004; 55: 306-319

147

Klunk, W. E.; Engler, H.; Nordberg, A.; Backsai, B. J.; Wang, Y.; Price, J. C.; Bergström, M.; Hyman, B. T.; Langström, B.; Mathis, C. A. Imaging the pathology of Alzheimer’s disease: amyloid-imaging with positron emission tomography. Neuroimaging Clin N Am, 2003; 13: 781-789

148

Wang, Y.; Mathis, C. A.; Huang, G-F.; Debnath, M. L.; Holt, D. P.; Shao, L.; Klunk, W. E. Effects of lipophilicity on the affinity and nonspecific binding of iodinated benzothiazole derivatives. J Mol Neurosci, 2003; 20: 255-260

149

Backsai, B. J.; Hickey, G. A.; Skoch, J.; Kajdasz, S. T.; Wang, Y.; Huang, G-F.; Mathis, C. A.; Klunk, W. E.; Hyman, B. T. Four-dimensional multiphoton imaging of brain entry, amyloid binding, and clearance of an amyloid-β ligand in transgenic mice. Proc Natl Acad Sci USA, 2003; 100: 12462-12467

150

Mathis, C. A.; Wang, Y.; Klunk, W. E. Imaging β-amyloid plaques and neurofibrillary tangles in the aging human brain. Curr Pharm Des, 2004; 10: 1469-1492

151

Skovronsky, D. M.; Zhang, B.; Kung, M-P.; Kung, H. F.; Trojanowski, J. Q.; Lee, V. M-Y. In vivo detection of amyloid plaques in a mouse model of Alzheimer’s disease. Proc Natl Acad Sci USA, 2000; 97: 7609-7614

152

Zhuang, Z-P.; Kung, M-P.; Hou, C.; Lee, C-W.; Trojanowski, J. Q.; Lee, V. M-Y.; Kung, H. F. Novel probes for imaging amyloid: iodinated benzothiazole derivatives. J Labelled Comp Radiopharm, 2001; 44 (suppl 1): S29-SS32

153

Zhuang, Z-P.; Kung, M-P.; Wilson, A.; Lee, C-W.; Plössl, K.; Hou, C.; Holtzman, D. M.; Kung, H. F. Structure-activity relationship of imidazo[1,2-a]pyridines as ligands for detecting β-amyloid plaques in the brain. J Med Chem, 2003; 46: 237-243

192

References

154

Kung, M-P.; Hou, C.; Zhuang, Z-P.; Zhang, B.; Skovronsky, D.; Trojanowski, J. Q.; Lee, V. M-Y.; Kung, H. F. IMPY: an improved thioflavin-T derivative for in vivo labeling of βamyloid plaques. Brain Res, 2002; 956: 202-210

155

Zhuang, Z-P.; Kung, M-P.; Hou, C.; Plössl, K.; Skovronsky, D.; Gur, T. L.; Trojanowski, J. Q.; Lee, V. M-Y.; Kung, H. F. Ibox (2-(4’-dimetheylaminophenyl)-6-iodobenzoxazole): a ligand for imaging amyloid plaques in the brain. Nuc Med Biol, 2001; 28: 887-894

156

Ono, M.; Kung, M-P.; Hou, C.; Kung, H. F. Benzofuran derivatives as Aβ-aggregate specific imaging agents for Alzheimer’s disease. Nuc Med Biol, 2002; 29: 633-642

157

Lee, C-W.; Kung, M-P.; Hou, C.; Kung, H. F. Dimetylamino-fluorenes: ligands for detecting β-amyloid plaques in the brain. Nuc Med Biol, 2003; 30: 573-580

158

Ono, M.; Wilson, A.; Nobrega, J.; Westaway, D.; Verhoeff, P.; Zhuang, Z-P.; Kung, M-P.; Kung, H. F. 11C-labelled stilbene derivatives as Aβ-aggregate-specific PET imaging agents for Alzheimer’s disease. Nuc Med Biol, 2003; 30: 565-571

159

Zhuang, Z-P.; Kung, M-P.; Ono, M.; Hou, C.; Kung, H. F. Synthesis of two potential ligands for Tc-99m labeling as diagnosis agents of Alzheimer’s disease. Chinese Journal of Chemistry, 2003; 21: 824-832

160

Kung, H. F.; Kung, M-P.; Zhuang, Z. P.; Hou, C.; Lee, C-W.; Plössl, K.; Zhuang, B.; Skovronsky, D. M.; Lee, V. M-Y.; Trojanowski, J. Q. Iodinated tracers for imaging amyloid plaques in the brain. Mol Imaging Biol, 2003; 5: 418-426

161

Jacobson, A.; Petric, A.; Hogenkamp, D.; Sinur, A.; Barrio, J. R. 1,1-dicyano-2[6(dimethylamino)naphthalen-2-yl]propene (DDNP): a solvent polarity and viscosity sensitive fluorophore for fluorescence microscopy. J Am Chem Soc, 1996; 118: 5572-5579

162

Agdeppa, E. D.; Kepe, V.; Liu, J.; Shoghi-Jadid, K.; Satyamurthy, N.; Small, G. W.; Petric, A.; Vinters, H. V.; Huang, S. C.; Barrio, J. R. In vitro and in vivo binding characteristics of two biological probes for plaques and tangles in Alzheimer’s disease. J Labelled Comp Radiopharm, 2001; 44 (suppl 1): S242-S244

163

Shoghi-Jadid, K.; Small, G. W.; Agdeppa, E. D.; Kepe, V.; Ercoli, L. M.; Siddarth, P.; Read, S.; Satyamurthy, N.; Petric, A.; Huang, S-C.; Barrio, J. R. Localization of neurofibrillary tangles and beta-amyloid plaques in the brains of living patients with Alzheimer disease. Am J Geriatr Psychiatry, 2002; 10: 24-35

164

Small, G. W.; Agdeppa, E. D.; Kepe, V.; Satyamurthy, N.; Huang, S-C.; Barrio, J. R. In vivo brain imaging of tangle burden in humans. J Mol Neurosci, 2002; 19: 323-327

165

Agdeppa, E. D.; Kepe, V.; Liu, J.; Small, G. W.; Huang, S-C.; Petric, A.; Satyamurthy, N.; Barrio, J. R. 2-Dialkylamino-6-acylmalononitrile substituted naphtalenes (DDNP analogs): novel diagnostic and therapeutic tools in Alzheimer’s. Mol Imaging Biol, 2003; 5: 404-417

166

Agdeppa, E. D.; Kepe, V.; Petric, A.; Satyamurthy, N.; Liu, J.; Huang, S-C.; Small, G. W.; Cole, G. M.; Barrio, J. R. In vitro detection of (S)-naproxen and ibuprofen binding to plaques in the Alzheimer’s brain using the positron emission tomography molecular imaging probe 2-(1-{6-[(2-[18F]fluorethyl)(methyl)amino]-2-naphtyl}ethylidene) malononitrile. Neuroscience, 2003; 117: 723-730

167

Bresjanac, M.; Smid, L. M.; Vovko, T. D.; Petric, A.; Barrio, J. R.; Popovic, M. Molecularimaging probe 2-(1-{6-[(2-fluoroethyl)methyl)amino]-2-naphtyl}ethylidene) malononitrile labels prion plaques in vitro. J Neurosci, 2003; 23: 8029-8033

168

Suemoto, T.; Okamura, N.; Shiomitsu, T.; Suzuki, M.; Shimadzu, H.; Akatsu, H.; Yamamoto, T.; Kudo, Y.; Sawada, T. In vivo labeling of amyloid with BF-108. Neurosci Res, 2004; 48: 65-74

169

Okamura, N.; Suemoto, T.; Shimadzu, H.; Suzuki, M.; Shiomitsu, T.; Akatsu, H.; Yamamoto, T.; Staufenbiel, M.; Yanai, K.; Arai, H.; Sasaki, H.; Kudo, Y.; Sawada, T. Styrylbenzoxazole derivatives for in vivo imaging of amyloid plaques in the brain. J Neurosci, 2004; 24: 2535-2541

References

193

170

Shimadzu, H.; Suemoto, T.; Suzuki, M.; Shiomitsu, T.; Okamura, N.; Kudo, Y.; Sawada, T. Novel probes for imaging amyloid-β: F-18 and C-11 labeling of 2-(4aminostyryl)benzoxazole derivatives. J Label Comp Radiopharm, 2004; 47: 181-190

171

Verbeke, K.; Kieffer, D.; Vanderheyden, J-L.; Reutelingsperger, C.; Steinmetz, N.; Green, A.; Verbruggen, A. Optimization of the preparation of Tc-99m-labeled Hynic-derivatized Annexin V for human use. Nuc Med Biol, 2003; 30: 771-778

172

Kung, H. F. New technetium 99m-labeled brain perfusion imaging agents. Semin Nuc Med, 1990; 20: 150-158

173

Kung, H. F.; Guo, Y-Z.; Yu, C-C.; Billings, J.; Subramanyam, V.; Calabrese, J. C. New brain perfusion imaging agents based on 99mTc-Bis(aminoethanethiol) complexes: stereoisomers and biodistribution. J Med Chem, 1989; 32: 433-437

174

Verbruggen, A. M. Lipophilic tracers for the study of regional cerebral blood flow. In: Murray, I. P. L.; Ell, P. J.; eds. Nuclear medicine in clinical diagnosis and treatment, vol 1. Churchill Livingstone, 1994: 465-482

175

Verbeke, K.; Snauwaert, K.; Cleynhens, B. J.; Scheers, W.; Verbruggen, A. M. Influence of the bifunctional chelate on the biological behavior of 99mTc-labelled chemotactic peptide conjugates. Nuc Med Biol, 2000, 27, 769

176

Abrams, M. J.; Juweid, M.; tenKate, C. I.; Schwartz, D. A.; Hauser, M. M.; Gaul, F. E.; Fuccello, A. J.; Rubin, R. H.; Strauss, H. W.; Fischman, A. J. Technetium-99m-human polyclonal IgG radiolabeled via the hydrazine nicotinamide derivative for imaging focal sites of infection in rats. J Nuc Med, 1990; 31: 2022-2028

177

Decristoforo, C.; Mather, S. J.; Cholewinski, W.; Donnemiller, E.; Riccabona, G.; Mancayo, R. 99mTc-EDDA/HYNIC-TOC: a new 99mTc-labelled radiopharmaceutical for imaging somatostatin receptor-positive tumours: first clinical results and intra-patient comparison with 111In-labelled octreotide derivatives. Eur J Nuc Med, 2000; 27: 1318-1325

178

Decristoforo, C.; Mather, S. J. Preparation, 99mTc-labeling, and in vitro characterization of HYNIC and N3S modified RC-160 and [Tyr3]octreotide. Bioconjug Chem, 1999; 10: 431438

179

Wald, J.; Alberto, R.; Ortner, K.; Candreia, L. Aqueous one-pot synthesis of derivatized cyclopentadienyl-tricarbonyl complexes of 99mTc with an in situ CO source: application to a serotonergic receptor ligand. Angew Chem Int Ed Engl, 2001; 40: 3062-3066

180

Yamauchi, H.; Takahashi, J.; Seri, S.; Kawashima, H.; Koike, H.; Kato-Azuma, M. In vitro and in vivo characterization of a new series of 99mTc-complexes with N2S2 ligands. In: Nicolini, M.; Bandoli, G.; Mazzi, U.; eds. Technetium and rhenium in chemistry and nuclear medicine 3. Verona, Italy: Cortina International, 1989: 475-502

181

Bormans, G.; Cleynhens, B.; de Groot, T. J.; Mortelmans, L.; Moretti, J.-L.; Verbruggen, A. Synthesis, radio-LC-MS analysis and biodistribution in mice of 99mTc-NIM-BAT. J Labelled Comp Radiopharm, 2003, 46, 575.

182

Oya, S.; Plössl, K.; Kung, M-P.; Stevenson, D. A.; Kung, H. F. Small and neutral TcVO BAT, bisaminoethanethiol (N2S2) complexes for developing new brain imaging agents. Nuc Med Biol, 1998; 25: 135-140

183

Cleynhens, B. J.; Bormans, G. M.; Vanbilloen, H. P.; Vanderghinste, D. V.; Kieffer, D. M.; de Groot, T. J.; Verbruggen, A. M. Trityl protection of bis-amino bis-thiol (BAT) chelator enables flexible derivatisation and facile labelling with technetium-99m. Tetrahedron Lett, 2003; 44: 2597-2600

184

Verduyckt, T.; Kieffer, D.; Huyghe, D.; Cleynhens, B.; Verbeke, K.; Verbruggen, A.; Bormans, G. Identity confirmation of 99mTc-MAG3, 99mTc-Sestamibi and 99mTc-ECD using radio-LC-MS. J Pharm Biomed Anal, 2003; 32: 669-678

185

Meegalla, S. K.; Plössl, K.; Kung, M-P.; Chumpradit, S.; Stevenson, D. A.; Kushner, S. A.; McElgin, W. T.; Mozley, P. D.; Kung, H. F. Synthesis and characterization of technetium-

194

References

99m-labeled tropanes as dopamine transporter-imaging agents. J Med Chem, 1997; 40; 917 186

Young, R. C.; Mitchell, R. C.; Brown, R. H.; Ganellin, C. R.; Griffiths, R.; Jones, M.; Rana, K. K.; Saunders, D.; Smith, I. R.; Sore, N. E.; Wilks, T. J. Development of a new physicochemical model for brain penetration and its application to the design of centrally acting H2 receptor histamine antagonists. J Med Chem, 1988; 31: 656-671

187

VanBrocklin, H. F.; O’Neil, J. P.; Hom, D. L.; Gibbs, A. R. Synthesis of [18F]fluoroanilines: precursors to [18F]fluoroanilinoquinazolines. J Labelled Comp Radiopharm, 2001; 44 (suppl 1): S880-S882

188

Patt, J. T.; Patt, M. 18F-labelled diazonium cation: a possible precursor for radiolabeling of peptides. J Labelled Comp Radiopharm, 2001; 44 (suppl 1): S154-S156

189

Lasne, M-C. ; Perrio, C.; Rouden, J.; Barré, L.; Roeda, D.; Dolle, F.; Crouzel, C. Chemistry of β+-emitting compounds based on fluorine-18. In: Topics in current chemistry, vol 222. Springer-Verlag Berlin Heidelberg, 2002: 201-258

190

Hutchinson, I.; Stevens, M. F. G.; Westwell, A. D. The regiospecific synthesis of 5- and 7monosubstituted and 5,6-disubstituted 2-arylbenzothiazoles. Tetrahedron Lett, 2000; 41: 425-428

191

Krolski, M. E.; Renaldo, A. F.; Rudisill, D. E.; Stille, J. K. Palladium-catalyzed coupling of 2-bromoanilines with vinylstannanes. A regiocontrolled synthesis of substituted indoles. J Org Chem, 1988, 53: 1170-1176

192

Brana, M. F.; Castellano, J. M.; Yunta, M. J. R. Synthesis of benzimidazo-substituted 3quinolinecarboxylic acids as antibacterial agents. J Heterocycl Chem, 1990; 27: 1177-1180

193

Sage, M. R.; Wilson, A. J. The blood-brain barrier: an important concept in neuroimaging. AJNR Am J Neuroradiol, 1994; 15: 601-622

194

Goldstein, G. W.; Betz, A. L. The blood-brain barrier. Sci Am, 1986; 255:74-83

195

Laterra, J.; Keep, R.; Betz, A. L.; Goldstein, G. W. Blood-brain-cerebrospinal fluid barriers. In: Siegel, G. J.; Agranoff, B. W.; Albers, R. W.; Fisher, S. K.; Uhler, M. D., eds. Basic neurochemistry: molecular, cellular and medical aspects, 6th edition. Lippincott-Raven, 1998: 671-690

196

Dischino, D. D.; Welch, M. J.; Kilbourn, M. R.; Raichle, M. E. Relationship between lipophilicity and brain extraction of C-11-labeled radiopharmaceuticals. J Nuc Med, 1983; 24: 1030-1038

197

Verbeke, K.; Verbeke, A.; Vanbilloen, H.; Verbruggen A. Preparation and preliminary evaluation of Tc-99m-EC-For-MLFK. Nuc Med Biol, 2002; 29: 585-592

198

Klunk, W. E.; Debnath, M. L.; Pettegrew, J. W. Development of small molecule probes for the beta-amyloid protein of Alzheimer’s disease. Neurobiol Aging, 1994; 15: 691-698

199

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

200

Kung, M-P.; Hou, C.; Zhuang, Z-P.; Cross, A. J.; Maier, D. L.; Kung, H.F. Characterization of IMPY as a potential imaging agent for β-amyloid plaques in double transgenic PSAPP mice. Eur J Nucl Med Mol Imaging, 2004; 31: 1136-1145

201

Lowry, O. H.; Rosebrough, N. J.; Farr, A. L.; Randall, R. J. Protein measurement with Folin phenol reagent. J Biol Chem, 1951; 193: 265-275

202

Lin, A. J.; Kasina, S. Synthesis of 3-substituted 7-(3,3-dimetyl-1-triazeno)-10methylphenothiazines as potential antitimor agents. J Heterocycl Chem, 1981; 18: 759-761