UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Radiolabelled carbohydrate conjugates : studies of Alzheimer's disease therapeutics and tumor imaging Bowen, Meryn Louisa 2009

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
[if-you-see-this-DO-NOT-CLICK]
ubc_2009_spring_bowen_meryn_louisa.pdf [ 4.77MB ]
[if-you-see-this-DO-NOT-CLICK]
Metadata
JSON: 1.0061548.json
JSON-LD: 1.0061548+ld.json
RDF/XML (Pretty): 1.0061548.xml
RDF/JSON: 1.0061548+rdf.json
Turtle: 1.0061548+rdf-turtle.txt
N-Triples: 1.0061548+rdf-ntriples.txt
Original Record: 1.0061548 +original-record.json
Full Text
1.0061548.txt
Citation
1.0061548.ris

Full Text

RADIOLABELLED CARBOHYDRATECONJUGATES: STUDIES OF ALZHEIMER’SDISEASE THERAPEUTICS ANDTUMOR IMAGINGbyMERYN LOUISA BOWENB.Sc.(Hons.), The University of Canterbury (New Zealand),2002A THESIS SUBMITTED IN PARTIAL FULFILLMENTOFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIES(Chemistry)THE UNIVERSITY OF BRITISH COLUMBA(Vancouver)March 2009© Meryn Louisa Bowen, 2009ABSTRACTThis thesis is split into two distinctparts, with the common theme beingthe radiolabeling ofcarbohydrate-conjugates. Chapter2 discusses radioiodinating 3 -hydroxy-4-pyridinones,ofinterest in treating Alzheimer’sdisease. Chapters 3 - 5 describeglucosamine conjugates of99mTcinvestigated as potential carbohydrate-basedSPECT imaging agents.Alzheimer’s disease (AD) sufferersdevelop characteristic beta-amyloid plaquesin their brains,made up of beta-amyloid protein withhigh concentrations of zinc and copper. Theredox activecopper ion can form reactive oxygen species(ROS) which damage surrounding tissueand leadto cell death. 3-Hydroxy-4-pyridinonesare of interest in the treatment of AD becausethey areantioxidants and metal chelators,targeting both the plaques and ROS. Functionalisationof thesepyridinones with a glucose moietymasks the chelating portion of the molecule,and mayfacilitate blood brain barrier (BBB)crossing via the GLUT glucose transporters.To determinethis BBB permeability, two compoundswere labelled with1251and then assessed using a ratbrain perfusion procedure. They wereobserved to cross the BBB, a crucialfinding in thedecision to pursue this line of research forAD therapy.A99mTc..basedSPECT tumor imaging agent would increaseworldwide access to the importantdiagnostic tools of nuclear medicine. Chapters3 — 5 discuss the synthesis, characterization andassay results of several monoanionic glucosamine-appendedtridentate ligands and theircomplexes with the99mTcand Re tricarbonyl cores. The length of the linkerbetween theglucosamine and the metal binding portionof the molecule range between two and elevencarbons. The binding moiety wasalso varied to give a library of molecules with differentbinding groups and linker lengths;useful for structure-activity relationship determination inarange of assays. The interaction of the compoundswith hexokinase was assessed, and none ofthe compounds were foundto be substrates for hexokinase. Transport of the compounds intocells by the GLUT transporters was also assayed,and found to be insignificant for allcompounds tested. Valuable information on the tolerancesof these proteins was discovered andChapter 6 includes ideas for improvements in futurecompounds.11TABLE OF CONTENTSAbstractiiTable of ContentsiiiList of TablesviiList of FiguresviiiList of SchemesxiAbbreviationsxiiAcknowledgmentsxviiiCo-Authorship StatementxixCHAPTER 1 Introduction11.1 Medicinal Inorganic Chemistry 11.2 Nuclear Medicine 21.2.1 Types of Radioactivity31.2.2 Therapy in Nuclear Medicine51.2.3 Imaging in Nuclear Medicine71.3 Carbohydrates — Some Relevant In Vivo Considerations101.4 Technetium and its Use in Nuclear Medicine141.4.1 Oxidation States of Technetium151.4.2 Technetium (I) Bioconjugates181.5 Alzheimer’s Disease 211.5.1 The Amyloid Cascade Hypothesis 211.5.2 Current Treatments for Alzheimer’ s Disease 251.5.3 Metal Chelators to Treat Alzheimer’s Disease 281.6 Thesis Overview321.7 References34CHAPTER 2 Radioiodination of Glycosylated Pyridinones as Potential Therapeuticsfor Alzheimer’ s Disease 422.1 Introduction 422.1.1 Metals in Alzheimer’ s Disease 422.1.2 The Multifunctional Approach to Treating111Alzheimer’s Disease.432.1.3 Our Approach to Multi-Target-DirectedLigandsfor Alzheimer’s Disease462.2 Experimental482.2.1 Instruments and Materials482.2.2 Chemical Synthesis492.2.3 GLUT-I Cell Uptake Assay522.3 Results and Discussion532.3.1 Synthesis532.3.2 In Vivo Studies562.3.3 In Vitro Studies582.4 Conclusion602.5 References61CHAPTER 3 Monoanionic Glucosamine-basedLigands for the Formation ofNeutral Complexes with the[M(CO)3]Core (M Re,99mTc)643.1 Introduction643.1.1 Designing a Carbohydrate-based ImagingAgent 643.1.2 Technetium Glycoconjugates653.1.3[mTc(CO)3]+Glycoconjugates693.1.3.1 Bidentate Ligand Systems693.1.3.2 Tridentate Ligand Systems733.2 Experimental793.1.2 Instruments and Materials793.2.2 Synthesis803.2.399mTcComplex Formation883.2.4 Cysteine and Histidine Challenges883.2.5 GLUT-i Cell Uptake Studies893.2.6 Hexokinase PhosphorylationStudies 903.3 Results and Discussion903.3.1 Synthesis, Characterization and Complex Stability903.3.2 Cell Uptake Studies1023.3.3 Hexokinase Phosphorylation Studies104iv3.4 Conclusion.1043.5 References106CHAPTER 4 99’Tc-Labe1lingand In Vitro Assays of Long ChainGlucosamine-basedTridentate Ligands1094.1 Introduction1094.1 .1 Long Chain Bioconjugatesof[99mTC(CO)3]+1094.1.2 Long Chain CarbohydrateConjugates of{99mTC(CO)3]+1104.2 Experimental1124.2.1 Instruments and Materials1124.2.299mTecetiComplex Formation 1124.2.3 Cysteine and Histidine Challenges1134.2.4 GLUT-i Cell Uptake Assay1134.2.5 Hexokinase Inhibition Assay1134.2.6 Hexokinase Substrate Assay1134.2.7 MTT Cytotoxicity Assay1144.3 Results and Discussion1154.3.199mTCLabelling and Cysteine/Histidine Challenges1154.3.2 Cell Uptake Assay1174.3.3 Hexokinase Inhibition Studies1194.3.4 Hexokinase Substrate Studies1234.3.5 MTT Assay1234.4 Conclusion1254.5 References127CHAPTER 5 Long Chain GlucosamineCyclopentadienyl Ligands for the [M(CO)3]Core(M = Re,99mTC)1295.1 Introduction1295.1.1 Synthesis of [CpM(CO)3]Compounds1305.1.2 Functionalization and BiologicalStudies of[CpM(CO)3]Complexes1375.2 Experimental1405.2.1 Instruments and Materials140V5.2.2Synthesis.1415.2.399mTechnetiumLabeling via SLT1495.2.4 Hexokinase SubstrateAssay1495.2.5 GLUT-i Cell Uptake Assay1495.3 Results and Discussion1495.3.1 Synthesis and Characterization1495.3.2 Hexokinase SubstrateAssay1585.3.3 GLUT-i Cell Uptake Studies1595.4 Conclusion1605.5 References161CHAPTER 6 Conclusions andFuture Work1616.1 Multifunctional MetalChelators as Potential TherapeuticsforAlzheimer’s Disease1616.1 .1 3-Hydroxy-4-pyridinones1616.1.2 NN 0, 0-TetradentateAminophenols1646.2 Carbohydrate Conjugatesof99mTCfor Use in Molecular Imaging1656.3 References170APPENDIX 1 Procedurefor rat brain perfusion experiment172viLIST OF TABLESTable 1.1 Three majortypes of radiation used in nuclearmedicine4Table 1.2 Properties and comparisonof PET vs. SPECT8Table 3.1 Synthesis and stabilityof99mTccomplexes in Chapter 3101Table 4.1 Synthesis and stabilityof99mTCcomplexes in Chapter 4116Table 5.1 Synthesis of99mTccomplexes in Chapter 5154viiLIST OF FIGURESFigure 1.1 Facets of medicinalinorganic chemistry2Figure 1.2 Examplesof diagnostic and therapeutic radiopharmaceuticals3Figure 1.3 Generic bioconjugatesschematic10Figure 1.4 Compounds recognisedby GLUT-i and/or hexokinase12Figure 1.5 MAG3syn and antiisomers16Figure 1.6 The HYNIC ligandsystem17Figure 1.7 The technetium tricarbonylcore18Figure 1.8 Small biologicallyrelevant molecules labelled with {M(CO)3J20Figure 1.9 AD drugs currently onthe market in Canada26Figure 1.10 Compounds tested in vivofor AD30Figure 2.1 Multi target directed ligandsin AD research44Figure 2.2 3-Hydroxy-4-pyridinones;their prodrugs and metal complexes47Figure 2.3 Compounds synthesizedin this chapter47Figure 2.4 A radio-TLC traceof pyrA56Figure 2.5 Graph of BBB permeabilityvs. octanol/water coefficient57Figure 3.1 Comparison of D-glucoseand FDG64Figure 3.2 Two glucosamine-conjugatedligands for the TcO(V) core66Figure 3.3 Three glucosamine-based ligandsfor the TcO(V) core 68Figure 3.4 A bidentate glucosamine-basedligand for the [Tc(CO)3]core70Figure 3.5 Pyridinone-carbohydrate ligandsfor the [Tc(CO)3]core 71Figure 3.6 Diamino carbohydrate based ligandsfor the [Tc(CO)3]core72Figure 3.7 Bipyridyl-thioglucose conjugatesfor the [Tc(CO)3)core 73Figure 3.8 Two tridentate glucose conjugates forthe [Tc(CO)3]core 74viiiFigure 3.9 C-3 functionalized glucoseconjugates for the [Tc(CO)3Jcore74Figure 3.10 Five carbohydrate-conjugatedligands for the [Tc(CO)3]core 75Figure 3.11 Four carbohydrate -baseddipicolylamine ligands76Figure 3.12 Fluorescent glucosamine conjugatesfor the [Tc(CO)3]core77Figure 3.13 Compounds made and studiedin Chapter 3 90Figure 3.14 JR spectra of 7 and ilL295Figure 3.15 JR spectra of ilL’ and ReL196Figure 3.16 ‘H NMR spectra of 7, HL2 and ReL298Figure 3.17 ‘3C NMR spectra of 7, HL2and ReL2 100Figure 3.18 Radiation trace of99mTcLl102Figure 3.19 Graph of GLUT-i cell uptakeassay for compounds in Chapter 3 103Figure 4.1 Thymidine and amino acid analoguesfor the [Tc(CO)3]core 110Figure 4.2 Two long chain glucose-conjugatesfor the [Tc(CO)3]core 110Figure 4.3 Three compounds studied in Chapter 4115Figure 4.4 Radiation trace ofCll99mTc117Figure 4.5 Graph of GLUT-i cell uptakeassay for compounds in Chapter 4 118Figure 4.6 Plots manipulating data to obtainK values 121Figure 4.7 Compounds used in the MTT assayin Chapter 4 124Figure 4.8 Size comparison of glucosamine andcompounds in Chapter 4 126Figure 5.1 ACp99fhTcconjugate with biological activity127Figure 5.2 Some bioconjugates of the [CpM(CO)3]core 136Figure 5.3 A glucosamine-Cp conjugated ligandfor the [Tc(CO)3]core 138Figure 5.4 ‘H NMR spectra of C6Fc, C6(Ac)4Reand C6Re 151Figure 5.5 ‘3C NMR spectra of C8(Ac)4ReandC8Re 152Figure 5.6 JR spectra of C8Re and C8(Ac)4Re153ixFigure 5.7 Radio-HLPC traceofC899mTc(CO)3156Figure 5.8 Graph of GLUT-i celluptake assay for compounds in Chapter 5157Figure 6.1 Prototype glycosylatedand non-glycosylated pyridinones162Figure 6.2 An aminophenol for radioiodination165Figure 6.3 Carbohydrate-conjugates interactingwith GLUT-i and /or HK 166Figure 6.4 Possible bifunctional ligandsto bind to the [Tc(CO)3]core 169xLIST OF SCHEMESScheme 1.1 Phosphorylation of FDG11Scheme 2.1 Reaction scheme for compoundssynthesized in Chapter 254Scheme 3.1 “Click to chelate” — synthesisof carbohydrate-appended ligand for the[Tc(CO)3]core using click chemistry78Scheme 3.2 Reaction scheme for synthesisof HL1 and HL2 93Scheme 4.1 Phosphorylation and subsequentdehydration of glucose 120Scheme 5.1 Double ligand transfer (DLT)reaction 129Scheme 5.2 Proposed mechanismof ring transfer in DLT reaction130Scheme 5.3 Single ligand transfer (SLT) reaction131Scheme 5.4 Four alternative methodsof forming [CpM(CO)3]complexes 133Scheme 5.5 Synthesis of non-radioactive compoundsmade in Chapter 5 148Scheme 5.6 Labelling scheme forC699mTC155xiABBREVIATIONSapproximate2D two dimensional3D three dimensionala. alpha or alpha particleA angstrom, 1 x10b0metre13betabeta particlepositrongamma ray6 chemical shift in parts permillion (ppm)extinction coefficient (L.mol’.cm’)(UV-visible)wavelengthmicro(10-6)v stretching frequency (IR)Af3 f3-amyloidAc acetylACN acetonitrileAcOH acetic acidAD Alzheimer’s diseaseADEPT antibody directed enzyme prodrug therapyADP adenosine diphosphateALS amytrophic lateral sclerosisxiiAPOE apolipoproteinEAPP amyloid precursorproteinAPT attached protontest (NMR)atm atmosphereATP adenosine triphosphateBBB blood brain barrierBn benzylBNCT boron neutroncapture therapybr broad (JRsignal)Bq Becquereldegrees celsiusCaic. calculatedCi CurieCIS coordination inducedshiftcm1 wavenumber(s), reciprocal centimetreCNS central nervous systemCOSY correlation spectroscopy (NMR)Cp cyclopentadienylCQ clioquinolCSF cerebrospinal fluidCT computed tomographyd day(s) OR doublet (NMR)DA Diels-Alder reactionDCC dicyclohexylcarbodiimideDCM dichloromethanexliiDFO desferrioxamineDFT density functional theoryDLT double ligand transferDMAP 4-dimethylaminopyridineDMF dimethylformamideDMSO dimethylsulfoxideDTPA diethylenetriaminepentaacetic acidEA elemental analysisEC electron captureEDC 1 -ethyl-3 -(3 -dimethylaminopropyl)carbodiimidehydrochlorideElMS electron impact massspectrometryESIMS electrospray ionization massspectrometryEtOH ethanolEtOAc ethyl acetateeq equivalents (of a reagent)fac facial (ligand arrangement inOhcomplexes)FAD familial Alzheimer’s diseaseFc ferroceneFDA Food and Drug Administration (USA)FDG[18F]2-deoxy-2-flouro-D-glucoseg gramG6P glucose-6-phosphateG6PDH glucose-6-phosphate dehydrogenaseGLUT glucose transporterhr hour(s)xivHK hexokinaseHMBC heteronuclear multiplebond coherence (NMR)HMQC heteronuclear multiplequantum coherence (NMR)HPLC high performance liquidchromatographyHR-MS high resolutionmass spectrometryHSAB hard soft acidbaseHz hertz(s’)IC50 compound concentrationwhich kills 50 % of the cells present%ID/g the percentage of thetotal injected dose (of radioactivity into ananimal) that ends up in one gram ofa particular tissue typeJR infraredJ coupling constant(NMR)k kilo (l0)K KelvinKa acidity constantK1 inhibition constantL litre OR ligandm metre OR milli- (l0)OR medium (IR) OR multiplet (NMR)M molar (moles/litre) ORmetal OR mega(106)MAO3 mercaptoacetyltriglycineMeOH methanolmm minute(s)mol mole(s)MRI magnetic resonance imagingMS mass spectrometryxvMTT 3 (4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromidem/z mass per charge (MS)n nano- (1 0) OR numberof samplesNADH nicotinamide adenine dinucleotide,reduced formNFT neurofibrillary tanglesNMR nuclear magnetic resonanceNSAID non-steroidal anti-inflammatorydrugORTEP Oak Ridge Thermal EllipsoidProgramPBS phosphate bufferedsalinePd/C palladium on carbon(10 % by weight)PET positron emission tomographyPEG polyethyleneglycolpH -log[H3O]ppm parts per millionpsi pounds per square inchq quartet (NMR)RBF round bottom flaskRf retention factor (TLC)ROS reactive oxygen speciesRPM revolutions per minutert retention time (HPLC)s strong (IR) OR singlet (NMR)SD standard deviationsec second(s)SLT single ligand transferxviSPECT single photon emission computedtomographyt triplet (NMR)T temperaturetJ/2 half-lifeTEOA triethanolamineTFA trifluoroacetic acidTg transgenicTLC thin layer chromatographyTMtrademarkTMEDA tetramethylethylenediamineTOCSY total correlation spectroscopyUBC University of British ColumbiaUV ultravioletw weak (IR)xviiACKNOWLEDGMENTSThanks to Dr. Chris Orvigfor taking me into hisgroup, always being positiveand encouraging,and allowing me the freedomto carry my research in the directionI wanted. Dr. Mike Adam hasmade an excellent secondboss, he has always been cheerful,happy to talk over ideas, andveryhelpful. Thanks to all the Orviggroup members who havemade for a great workplace overtheyears: particularly Tim, Neil,Cara, Eszter, Dr. Chen, Adrienne,and Michael who either workedwith me, taught me variousaspects of chemistry or havebeen the source of useful discussions.The PET group at TRIUMF havebeen very accommodating, andI would especially like to thankWade English and Ken Buckleyfor continuously havingto fix the HPLC, and Dr. JianmingLufor the use of his lab andisotopes. MDS Nordion generouslysupplied123jand the UBCHospital Department of NuclearMedicine the99mTCThanks to our collaboratorsat Texas Tech University who carriedout the in vivo Alzheimer’swork: Fancy Thomas withDr. David Allen and Dr. Paul Lockman.I am very grateful to Dr. UrsHäfeli and Ripen Misriin the Faculty of Pharmaceutical Sciencesat UBC, who have been verygenerous with their time and expertisein setting up the cell uptake studies.UBC NMR, MS and BiologicalServices staff as wellas members of the mechanicalandelectrical shops have been nothingbut helpful in my interactions withthem.Finally I would like to thankmy family — my parents encouragedme into chemistry when I maynot have chosen that path myself,as they could see that was wheremy interests lay. And theyhave been nothing but supportivesince. My brothers who have never understoodwhy anyonewould do chemistry whenit obviously meant getting ajob mixingpaints in a factory, have mademe think, and always made me laugh.And Michael, who still has noidea about chemistry butnever ceases to ask questions anyway.He is always encouraging and challenging,and able tomake me laugh, with chemistry aswith all aspects of life. It certainlywouldn’t have been asenjoyable a ride without him.xviiiCO-AUTHORSHIP STATEMENTEverything in this thesis wasdesigned, researched, performed, analysedand written by myselfexcept the following:Chapter 2- The non-radioactive form of pyrB wasmade by Dr. Michael Merkel. Dr. Merkel andIdeveloped the radiolabeling forpyrB together, based on work I had alreadydone with pyrA.- The rat brain perfusion experiments werecarried out by collaborators: Dr. David Allen,Dr.Paul Lockman and Fancy Thomasat the Texas Tech University Health SciencesCentre,Armadillo, TX.Chapter 3- Compounds HL3,HL4 werefirst made by Chuck Ewart and Dr. NeilLim, respectively.Compound L5 was made by Dr. Tim Storr.Chapter 4- The C6, C8 and Cli ligands andtheir rhenium complexes were first madeby Dr. Zhen-FengChen under my guidance. I designed andcarried out some research into these compounds,andDr. Chen did further research and was responsiblefor performing the synthesis.- Adrienne Roos was a summer studentunder my supervision who performed theMTT andhexokinase phosphorylation assays.I was responsible for the design and researchof theseassays. Adrienne performed the experimentsand manipulated the data with guidance and someassistance from me.Chapters 2- 5- The GLUT-i cell uptake assay was designedby myself with assistance from Dr. Urs Häfeli,Faculty of Pharmaceutical Sciences, UBC.Cell maintenance and counting was doneby RipenMisri. The experiments were all carried outby myself.xixCHAPTER 1Introduction*1.1 Medicinal Inorganic ChemistryMedicinal inorganic chemistryis the application of inorganicchemistry to the therapy ordiagnosis of disease. As shownin Figure 1.1, this can involveeither addition or removal of aninorganic substance to orfrom a biological system in a controlled maimer.The purpose of theseactions is either diagnosticor therapeutic, to determine the natureand severity of a problem, orto treat a determined disease state.As seen in Figure 1.1, therapycan take the form of removing excess metalsfrom biologicalsystems by administration of an organicmetal chelator. Desferrioxamineis used to treat patientswith an excess of iron in their bodies,by binding tightly to Fe3 and solubilizingit so it can beexcreted from the body.3Before a species like this is introducedto a living system, theinteraction of the chelator withthe metal ion(s) of interest willhave been thoroughlyinvestigated. Therapy inmedicinal inorganic chemistry may alternativelyinvolve administrationof a metal complex where theproperties of the metal imparta therapeutic effect to the drug.Cisplatin kills cancer cells because theweakly bound chloride ligandson the platinum arereplaced by guanine N-7 donorsin DNA. This causes cross linking ofDNA strands, affectingthe cells’ ability to replicate and eventuallykilling the cells.’ In terms of diagnostics,variousproperties of metals can be exploitedto impart functionalities not available from purely organiccompounds. An excellent exampleof this is the gadolinium based MRI contrast agents,wherethe large paramagnetic field of the metalions (7 unpaired electrons on each Gd3)leads to animproved signal from nearby water molecules.2The examples given here illustrate some of theadvantages that can be garnered by using ametal ion to impart certain properties thatmay not beavailable in a purely organic molecule.However, adding a metal to a potentialdrug alsoincreases the need for caution, asthe complex needs to be shown to stay intact underbiologicalconditions. For example, in the caseof the Gd3 contrast agents, the metal ion itself is toxicandmust stay bound to its chelate in vivo toprevent detrimental side effects. A largebattery of*A portion of this chapter has been published: Bowen,M. L. and Orvig, C.9smTechnetiumcarbohydrate conjugatesas potential agents in molecular imaging.C’hem. Commun. 2008, 5077 — 5091.References begin on Pg 34 1toxicological and stabilitystudies are required for any compoundthat may be used in humans,regardless of the type of compoundbeing investigated. The wide range ofproperties availablewhen utilising most of the Periodic Table,as inorganic chemists do, have potential benefits thatmake these many elements worth exploring.MEDICINAL INORGANICCHEMISTRYAddition of metalsRemoval of metals fromto biological systemsbiological systemsLi__ ____DIAGNOSIS THERAPYTHERAPYe.g. e.g.e.g. desferrioxamine in J3-Magnevist — cisplatinfor the thalassemia (to relieve ironGd-DTPA used as an MRJ treatmentof overload)3contrast agent2 cancer1HNH,NlN/\\CIpCI5HHOOC”\ç’COOHH3N’ NH3OHH/JHOOCCOOH COOHLi___IN THIS THESISIN THIS THESISTc-carbohydrate conjugateAlzheimer’s project Ch 2project Ch3 -5Figure 1.1 Medicinal inorganic chemistry isthe application of inorganic chemistry to diagnosingand treating disease states.1.2 Nuclear MedicineNuclear medicine is the use of radioactive isotopesincorporated into diagnostic ortherapeutic agents. The compounds used in nuclearmedicine are often referred to asradiopharmaceuticals, pharmaceuticals containingradioactivity. Therapeutic effects areachieved by damaging and killing diseased tissue,normally cancer cells, with the radiationReferences begin on Pg 34 2given off by the radioisotope.Diagnosis is achieved by usingspecial cameras to image theradioactivity given off by decayevents.OHO\HO-—-- OHabCD 20receptorCD 20MAbNH2 Il-IdCeO\Figure 1.2 Some examples of therapeutic(a, b and c) and diagnostic (c, d and e) compoundsused in nuclear medicine:a)‘88Re-P2045 is undergoing clinical trials as a treatmentof non-small cell lung carcinoma4 b) Zevalin— an antibody that targets non-Hodkin’s lymphomacells, and kills them with the 3 radiationgiven off by the 90Y c) MIBG —metaiodobenzylguanidine used for imaging(123J)and therapy(1311)of adrenal cancers6 d)Cardiolite— 99mTCsestamibi, used for imaging heart function withSPECT7’8e) FDG, 2-’8F-fluoro-2-deoxyglucose, used to imageareas of increased (e.g. cancer)9and decreased (e.g.dementia or heart ventricle dysfunction)’° glucosemetabolizm using PET.1.2.1 Types of RadioactivityNuclear decay can be divided into three categories;a., f3 andy.Table 1.1 shows someproperties of these types of radiation as wellas some commonly used isotopes of each kind.Radioactive particles in nuclear medicine comefrom nuclear decay, but it is important tonote that they are simply high energy particles; forexample, a. radiation is a He2 nucleuswith a certain amount of kinetic energy. Radiationinteracts with, and sometimes damagesReferences begin on Pg 343different materials by transferringsome of its energy. Manyforms of radioactivity containlarge amounts of kineticenergy gained from thenuclear decay event, and dependingon themass and the speed of theparticle will distribute that energyto surrounding matter over acertain time and area.The smaller the space over whichthe energy is transfered, themoredamaging the interaction willbe to the receiving tissue.Table 1.1 Some importantproperties and uses in nuclearmedicine of the three majortypesof radiation.Radiation Particle orRange in Range inShieldinglonising? UseType TypeAir Tissue112 protons anda sheet of2 neutrons 1- 10 cm 2 — 10 cells stronglytherapy(He2)paperperspextherapy,electron (f3) or10 — 500or 1 - 10 m mediumimagingpositron (f3)cellsaluminum(PET)lead or100 - 1000 wholeimagingy photon thicklowm bodies (SPECT)concreteAlpha particles are relativelyheavy, travel quite slowly and only havea small amount ofkinetic energy to lose before comingto rest. They are not hazardous unlessingested, as theouter layers of skin are capableof stopping their progress. Beta particles havemasses about1/2000 the size of alpha particles, sofor a set energy input will have more velocitythan analpha, and are able to travel moderatedistances through air. In tissuethey will only travelshort distances, and so are moreof a hazard if internalised than if exposureis from anexternal source. Gamma rays area form of electromagnetic radiation, with nomass and highenergy. They travel very quickly (at the speedof light) and with a short wavelength, so willtravel a long way if not attenuated bya dense material such as concrete or lead. These poseas much risk of tissue damage whetherthe exposure is internal or external to the body,asthey can easily penetrate living tissue (itis not dense enough to attenuate them). ThisalsoReferences begin on Pg 34 4means that the amount of energy transferto the tissue is small compared to other radiationtypes, so long exposure is requiredfor enough energy to accumulate ina given area fordamage to be observed.These properties are what make each typeof radiation useful for a certain task. As showninTable 1.1, alpha and betaradiation do not travel very far in tissue,so energy from theradiation they emit will allbe deposited in nearby tissue. This makesthem useful fordamaging tissue in a certain area ifthey can be made to accumulate in thatarea. Gammarays on the other hand are not particularlydamaging and travel a long way throughtissuewithout their path being significantlyperturbed, making them ideal foruse in imaging.Again the isotope needs to accumulatein the tissue to be imaged and then,using camerassensitive to gamma rays, theposition of the radiation source can be determined.This type ofdecay can be imaged using Single Photon EmissionComputed Tomography — SPECT.Positrons are the antiparticles to electronsand are denoted . These have similar propertiesto f3 radiation, but behave quite differently. Becausethey are antiparticles of electrons,whenever they come into contact withan electron the two annihilate with each other andaretransformed into energy. Conservationof both mass and energy means that this interactionproduces two gamma rays at 180°to each other with energies of 511 keV. Thesegammarays can then be imaged in a similarmanner to the gamma rays produced by simplegammadecay, although in this case the two parallel raystravelling at the same speed in oppositedirections give additional spatial information.This type of radiation can be imaged using atechnique known as Positron Emission Tomography— PET.1.2.2 Therapy in Nuclear MedicineTherapy in nuclear medicineis generally aimed towards killing cells rather than healingthem. Nuclear medicine therapeutics are thereforemainly used in cancer, although there areother applications too, such as overactive thyroid(hyperthyroidosis), and bone painpalliation. Hyperthyroidosis is treated withan oral dose of ‘3’i, where the f3 emissiondamages the surrounding tissue to reduce the activityof the thyroid.’2 As131jalso has some‘ emission, it can also be used for imaging of the thyroid.89Sr strontium chloride —Metastron - is administered intraveneously for the reductionof bone pain in bone cancerReferences begin on Pg 34 5patients.’3 It works becauseSr2,behaving similarly to Ca2,accumulatesin bone, and the f3emission in the bone reducespain, though the exact mechanism of this pain reductionis notwell understood. 89Sr has ahalf life of 50 days, so pain is alleviated for monthsat a timewith a single injection.Zevalin (Figure 1 .2b) was approvedby the Food and Drug Administrationin the USA in2002 as an anti-cancer agent. It containsan antibody that targets non-Hodgkin’slymphomacells by binding to the CD2O antigenexpressed on their surface.’4 90Yis bound to theantibody, and the 3 radiation givenoff by its decay kills nearby cells. An advantageofantibody therapy is that antibodiesbind specifically to their antigens,allowing for morespecific delivery of the radioisotopeto the desired site than most smallmolecules canattain.’4 More specific targetting leadsto reduced damage to healthy tissue, and thereforetoless side effects. An exampleof a small molecule radiotherapeutic is MIBG,metaiodobenzylguanidine, which isused for both therapy and imaging(131Jand123jrespectively).6MIBG is specific forcells that produce adrenalin, and as such is used mainlyfor cancer of the adrenal gland and certain kindsof brain cancer. It is more commonly usedfor imaging.’5There are currently no commerciallyavailable alpha emitter therapeutics.’6There arehowever several that have been, or are currently,in advanced stage clinical trjals.’6 Progressin this area has been quite slow, likely for several reasons.Alpha emitters are all heavyisotopes, often of toxic elements, and mostdo not have a large depth of research into theircoordination chemistry. This means therehas to be a substantial amount of work done on thecoordination chemistry of these species to ensurecontainment of the radioisotope once themolecule is introduced to a biological system.Alpha emitters do have the advantage ofbeing much easier to shield and transport than othertypes of radiation as their emissions arenot able to penetrate even standard laboratorycontainers. They generally have quite longhalf lives, meaning they can be transported considerabledistances without a significant lossof activity.The alpha emitters that have undergone the most research include21‘At, 213Bi and 223Ra.Aipharadin(223RaC12,t,12 = 11 days) is currently in Phase IIIclinical trials for the treatmentof skeletal metastases in prostate cancer patients. It successfully completedphase II clinicalReferences begin on Pg 34 6trials, where it was foundto improve disease statemarkers and extend patient survival timecompared to placebo treated controls.’7223Ra2 is thought to target bonebecause of itssimilarity to Ca2,and becauseof this mechanism of action the authorspropose it will beuseful for treating bone cancerand metastases from other cancer types as well.As 223Ra isadministered as an ioniccompound, strength of chelation is nota concern, as is often thecase with other alpha emitters.213Bi has been conjugated to antibodiesand investigated as a treatment forprostate cancer. Itis chelated via a DTPA moiety attachedto the J591 antibody which is targetedto the prostatespecific membrane antigen expressedin high concentrations on the surfacesof prostatecancers and their metastases.18 Ithas been investigated for treating lungcancer, also byconjugation to an antibody,’9but no commercial product has yet come out of thesestudies.21is being investigated by several groups.2°Astatine is a large element, and asa halogenrequires different types of bindingto the other large elements used in radiochemistry,whichare normally metals. The Wilbur groupat the University of Washington workswith astatinebound to carboranes, and they find the resultingcompounds to be very stable, and relativelyeasy to conjugate to a wide range of antibodiesand other targeting molecules.21’22Otherapproaches to astatine conjugation include bindingit via a single covalent bond to an arylgroup, which has been found tobe of adequate in vivo stability depending on the substitutionpattern of the aryl ring.23’241.2.3 Imaging in Nuclear MedicineImaging in nuclear medicine is achieved using eithera ‘ (gamma) emitting nucleus whichcan be imaged with single photon emission computed tomography(SPECT), or a(positron) emitting isotope, via positron emissiontomography (PET). In SPECT a singleray is emitted by each decay event, and three dimensional(3D) data can be recorded oncameras and reconstructed to give a meaningful imageof the area of interest. In PET anemitted positron travels until it meetsan electron (its anti-particle), at which time the twoannihilate, resulting in two opposingy rays which can be recorded on 3D cameras and usedto reconstruct an image. The most commonly used PET diagnosticin nuclear medicine, is2-’8F-fluoro-2-deoxyglucose (FDG) (Figure1 .2d). In this compound, the 2- position ofReferences begin on Pg 34 7glucose contains an ‘8Fatom instead of the native hydroxylgroup.25 ‘8F is a emittingisotope, t112 = 110 mm. BecauseFDG is very similar to glucose, itis used as a measure ofglucose metabolizm.25 The mostcommonly used SPECT compound,in nuclear medicine isCardiolite—99mTcSestamibi(shown in Figure 1.2d). This worksas a blood perfusion agentfor imaging heart function asits charge (+1) and lipophilicity(which comes from thebranched organic chains on each ofthe six ligands) lead to a favourableaccumulation in thistissue. The important propertiesof the two imaging modalities, PET and SPECT,arerecorded in Table 1.2.Table 1.2 Properties and comparisonof PET vs. SPECT for imaging in nuclearmedicine.PETSPECTSingle Photon Emission ComputedPositron Emission TomographyTomographyPositron (f3 particle) emittersGamma (y photon) emittersShort lived isotopesRange of half livese.g. ‘1C, ‘3N, ‘SO, 18F e.g.99mTCii1jl23jAnnihilates with e —* 2 ‘ rays 180°apartImaged with a 3D cameraScanners less widely availableScanners available at all major hospitalsPart of this thesis is concerned with the designof new SPECT based imaging agents. Thereasons for this interest can be seen in comparingthe two imaging modalities in Table 1.2.SPECT isotopes are generally longer lived than PETisotopes, and many do not need to beproduced in a cyclotron. One major effect of thisis that SPECT isotopes are generallycheaper than PET isotopes.26 SPECT isotopesare generally more widely available too,because they can be produced in morelocations and transported further without excessdecay. In addition, there are many moreSPECT than PET cameras available - as ofDecember 2006 there were 616 SPECT scanners (including3 SPECT/CT instruments) and27 PET scannera (including PET/CT instruments)in use in Canada.27 It can be seen thatthese properties combined make SPECT muchmore widely available than PET in terms ofboth geography and cost.References begin on Pg 348For these PET or SPECTimages to be useful for diagnosis,28 the radionuclidemustaccumulate in tissue of a certain type.This is often achieved by adding a radioactiveisotopeto a compound known to accumulatein the target tissue. These targeting compoundsareoften biomolecules, like peptides,28antibodies,14 lipids,29 carbohydrates,25nucleicacids3°orsmall molecules such as biotin,3’as naturally occuring substrates havehigh affinities fortheir receptors. In this approach the radionuclidemust be added in a way that minimizesitseffect on the biodistribution of the parent biomolecule.In this regard, PET has the advantagethat its nuclei such as ‘8F and “C canbe covalentlyincorporated into a molecule.They can often be used to replace a non-radioactive,naturallyoccuring isotope such that no chemical change isseen in forming the radioactive species.Asmost of the commonly used SPECTisotopes are metallic (e.g.99mTc,“In, 67Ga), and do notoccur naturally in vivo, an important partof the design of SPECT radiopharmaceuticals isthechelation of the radionuclide. Thechelation needs to be strong such that the radioisotopestays bound to the rest of the molecule.The free radioisotope will probably not accumulatein the tissue of interest as well as the complexdoes, so its biodistribution will becomeunpredictable, leading to increased backgroundnoise which decreases image quality andtherefore diagnostic capability, as wellas raising the dose to the patient as the unboundisotope may not clear well from the body. Mostbiomolecules or small molecules that targetcertain tissue types do not bind strongly to metals,so a chelating portion needs to be addedto the target molecule. The compounds madeto address these issues are a class ofbioconjugates. These are illustrated in Figure 1.3,where it can be seen that there are severalfunctionalities designed into molecules of thistype. The metal binding portion must attachstrongly to the radionuclide. The biomoleculewill hopefully be responsible for thebiological activity and distribution of the resultingcomplex. The linker needs to be robust,so the complex does not fall apart, and may alsobe used to impart other properties to thecomplex such as sufficient separation of thetwo functional parts of the molecule oradjustment of the overall lipophilicity.In the carbohydrate-based imaging that is the focus of this thesis,these bioconjugates areformed by conjugation of a radioactive atom viaa metal binding moiety and a linker to asugar, in the hopes that the biodistribution of that sugar is essentially maintained.In thiswork, the tether, the linker and the metal binding portionshave all been varied and theReferences begin on Pg 34 9effects of these on the activityof the compounds with the relevantenzymes investigated.Given the focus of this thesis oncarbohydrate derivatives, the compounds examinedherewill have the ultimate aim of beingused to image areas of unusualsugar metabolism; eitherraised, as in cancer, or diminished,as in stroke, heart attack, or Alzheimer’sdisease. TheSPECT analogues we are attemptingto produce have the potential to make a veryusefuldiagnostic technique available toa much larger number of patients, due to thewideravailability of SPECT compared with PET.Bio-molecule E)Zlr— —si________L. J_______Tether_______Conjugation“M”/Linker cz)MMetal bindingrportionFigure 1.3 A generic model of a bioconjugateused to bind to a radioactive metal such astechnetium, resulting in a bioconjugated radiotracer.1.3 Carbohydrates — Some Relevant In Vivo ConsiderationsCarbohydrates are the main source of fuel for the body.As well as providing energy they areintimately involved in many vital processes, suchas cell signaling, molecular recognition,surface adhesion, providing structure, the inflammation response,and in the fertilization anddevelopment of the fetus.32 Because carbohydrates aresuch important biological molecules,life forms have many enzymatic systems inplace to monitor their levels as well as totransport, chemically modify, and metabolize them.The major use of glucose is in metabolizm, through the complex,tightly regulated processesof glycolysis, the citric acid cycle and oxidative phosphorylation.Because of this, life hasdeveloped very efficient and particular processes for thetransport of glucose into and out ofcells, and for the processing and breakdown of the glucose once insidecells. Glycolysis isone way cells make ATP — the biological energy currency. Cellsalso use oxidativephosphorylation for this purpose; this is a muchmore efficient process, producing around 32References begin on Pg 34 10ATP per glucose, comparedto two ATP per glucose for glycolysis.33One of the mostpronounced and well studieddifferences between cancerousand normal tissue is the increasein glycolysis and decrease inoxidative phosphorylation in cancerousvs. normal cells. Thisresults in inefficiency in the metabolizmof cancer cells, so their glucose intake requirementsare much higher than those ofnormal cells to achieve the same energyproduction level, afundamental difference knownas the Warburg effect.34 The consequentenzymaticenhancements of tumor cellsover many other cell types can be exploitedby using glucose asa targeting vector for diagnosticor therapeutic compounds.A key carbohydrate processingenzyme that is notably overexpressed in cancercells ishexokinase (HK).35 As HK isthe first enzyme in the glycolysis pathway, changes in itsactivity have a large affect on overall energyproduction of a cell. HK has many isoforms,but most found in humans are membranebound and are about 100 kDa in size. Each enzymeis comprised of two similar domains, withthe active site residing in a cleft between thetwo.36 When a substrate, such as glucose,is bound in the active site, the two portions of theenzyme rotate together by 12° to narrowthe cleft and allow for the transfer of a phosphategroup from an ATP (also present at the active site)to the hydroxyl group at the C-6 positionof the sugar.36 The products of this reaction(as shown in Scheme 1.1) are ADP and glucose-6-phosphate, which is negativelycharged at physiological pH, and therefore notable topermeate the cell membrane./0/POHHKHV\L0\HO0\HO--.—--OHOHF F+ATP+ADPScheme 1.1 Phosphorylation of FDG by hexokinase(HK) leads to a negatively chargedproduct which is trapped in cells, in the process ATP is convertedto ADP.This means that for a given amount of glucose transportedinto a cell, the more hexokinaseactivity the cell exhibits, the higher its concentrationof trapped carbohydrate metaboliteswill be. In cancer cells, HK activity in often highdue to either overexpression of the protein,or incorporation of a larger proportion of HK intoa membrane, which is known to increaseits activity.37 Compounds that are known inhibitorsof hexokinase are being investigated asReferences begin on Pg 34 11potential chemotherapies for varioustypes of cancer.38 3-Bromopyruvate (Figure 1 .4a)is aninhibitor of HK that has been shownto deplete ATP in cancer cells.38 N-appendedglucosamine derivatives were testedby Bertoni and Weintraub,39 who found severalcompounds that exhibited competitiveinhibition of hexokinase. N-Benzoylglucosamine(Figure 1 .4b) exhibited the highestinhibition of human brain hexokinase of the compoundstested, with K values of 8.6 - 22 nMdepending on the isoform of hexokinase used.39Thefluorescent dye 2-NBDG (Figure 1 Ac), was foundto be phosphorylated by hexokinase in Kcoil cells.40BrfO- bOHNO2Figure 1.4 Compounds found to interact withkey carbohydrate metabolizing proteins in vivo.Competitive inhibitors of hexokinase:a) 3-bromopyruvate38b) N-benzoylglucosamine39c)2-NBDG.4’ Functionalized glucose and glucosamineanalogues that exhibit GLUT glucosetransporter activity: c) 2-NBDG41d) a near-JR dye appended to glucosamine.43Carbohydrates are very polar molecules,and as such have very poor membrane (lipid)permeability. The GLUTs are a class oftransmembrane proteins used to facilitate transportof glucose and other sugars, inan energy independent manner, across key membranes. Theyare ubiquitous in mammalian cells, although cellsexpress varying amounts of differentisoforms, depending on their energyrequirements. In humans there are thirteen knownisoforms of the GLUT family; they range from 45- 60 kDa in size, and all contain twelvemembrane spanning domains with high sequencehomology.44 GLUT transporters are foundin high concentrations in the membranesof cells with high energy requirements, and at theblood brain barrier (BBB).The BBB is a series of tight endothelial junctions betweenthe cells that line the interiorwalls of the blood vessels in the brain. The purposeof these tight junctions is to stop theReferences begin on Pg 34 12dinflux, to the very importantand sensitive organ that is the brain,of any unwanted andpotentially harmful components thatare circulating in the blood stream.As glucose is theprimary source of energy forthe brain, which is responsible forthe consumption of 25 % ofthe body’s glucose despitemaking up only 2 % of its mass,it is crucial that there bemechanisms in place to ensurean adequate supply of carbohydratesto the brain. Thus thereis a high concentration of GLUTs,particularly GLUT-i, found at theBBB. This is utilisedin our approach to Alzheimer’sdisease therapeutics, as addressed inChapter 2 of this thesis.As tumors grow rapidly andhave altered metabolizm, they have correspondinglyhigh energyrequirements.45’46This leads to a higher demand forcarbohydrates, which means that theiruptake into cells must be increased,necessitating an increased expressionof GLUTscompared to normal cells. Thisfact was used in the design of the99mTcbased carbohydrateconjugates explored as potential imagingagents in Chapters 3 — 5.Although many of the GLUT proteinsare known to exhibit high substratespecificity (e.g.GLUT-i transports D-glucose,but not its enantiomer L-glucose), thetransport of non-naturalsubstrates across otherwise impermeablemembranes has been observed. Appendingacarbohydrate moiety to a compound toenhance either BBB or cell permeabilityhas proveduseful with a number of different molecules.42However, C-2-functionalised glucosamineanalogues have perhaps shown themost promising GLUT substrate scope. Zheng andcoworkers appended a large porphyrinthat acts as a near-infrared dye to the nitrogenofglucosamine (Figure 1 .4e) and foundthat the compound was transported into cells.43 2-NBDG (Figure 1 .4c) has a fluorescentprobe attached to the nitrogen of glucosamine, andwas found to be transported into human erythrocytes.This cellular uptake was competitivelyinhibited by D-glucose, meaning the transport couldbe attributed to GLUT-i rather thanpassive diffusion, as this would notbe affected by the presence of glucose.4’As mentionedabove 2-NBDG is also phosphorylatedby hexokinase.An example of a molecule that has biologicalactivity with both the GLUT transporters andhexokinase is FDG (seen above in Figurei.2e and Scheme 1.1). FDG is taken up into cellsby the GLUT transporters due to its similarity toglucose.47 Once inside a cell, glucose isphosphorylated at the 6 position by HK,then dehydrated across the C-i — C-2 bond byglucose-6-phosphatedehydrogenase (G6PDH). FDGis phosphorylated at C-6 by HK, but theReferences begin on Pg 34 13resulting FDG-6-phosphate isnot active under G6PDH because ithas a fluorine in the 2-position, not the requisite hydroxyl(see Scheme 1.1). FDG-6-phosphate is not a majorsubstrate for any other enzymes;and because it is negatively charged and cannotdiffusethrough membranes, it simply accumulatesin the cells with the greatest GLUTandhexokinase activities; normally cancercells.48 These properties explain whyFDG is such agood imaging agent for glucose metabolizm,and these are the properties we aim to mimicindesigning new SPECT carbohydrateanalogues.1.4 Technetium and its Use in NuclearMedicine99mTcis the most commonly used isotope in nuclear medicine,accounting for about 90 % ofall diagnostic nuclear medicine scans worldwide.49This is due to the near ideal physicalproperties of the99mTCisotope; it has a six hour half life andemits ‘ rays with an energy of141 keV. This means there is sufficienttime to chemically manipulate the isotope beforeinjection into a patient, and allow for its accumulationin target tissue while still having asignificant amount of the original activity left to image.This emission energy is fairly low;meaning the radiation dose to the patient is minimized,comparable to that from aconventional medical X-ray.The other appealing practicality of99mTcis that it is producedin a generator from 99Mo, t112 66 hr. The technetiumis eluted as required in the form of[99mTc04ftaking advantage of the charge difference between[Mo04]2 and [TcO4fto elute99mTcselectively. This means that technetium is inexpensiveand easily transportable,making a useful SPECT isotope available whereverthere is a SPECT scanner.There is also interest in the third row congener of technetium, rhenium,which has twopotentially useful radioactive isotopes — ‘86Re and‘88Re. ‘86Re, ti,’2 = 3.7 days, decays viaemission of both f3(Emax = 1.07 MeV, maximum distance in tissue = 5 mm) and ‘y (9 %, 137keV) radiation. ‘88Re, ti,2 = 17 hr, also emits both f3(Emax= 2.12 MeV, maximum distance intissue = 11 mm) and (15 %, 155 keV) radiation.13emission is capable of killing cells, andif the isotope can be targeted to a tumor then the short tissuerange of the radiation allows forselective cell death, resulting in reduced sideeffects compared to less specific radio- orchemotherapies. The ‘y emission of these compoundsmay allow for concomitant imaging, toReferences begin on Pg 34 14help tailor dosage, and to monitoraccumulation of the compound in the areas of concern.Technetium, whose name comes fromtechnetos, greek for artificial, does not have any non-radioactive isotopes. Thus itis standard practice when working with technetiumto use non-radioactive (or cold) rhenium, to performlarger scale chemistry and characterization. Inthepast, a J3 emitter, 99Tc, t112 = 2.13x105years, was used, but as radioactive licensing becomesstricter and our understanding of the similaritiesbetween technetium and rhenium chemistryincreases, this is becoming less common. Rheniumexhibits very similar chemistry totechnetium. Rhenium is slightly larger,and because of this is a little softer,so may haveslightly different affinities for differentligating atoms. Rhenium has a slightly lowerreduction potential than technetium, meaningthat it is harder to reduce from the +7 oxidationstate in which these metals are commonly suppliedas the {M04fanion. Given that theselimitations are fairly well understood,and that all other aspects of the chemistry of these twoelements are comparable, similarity can be assumed.For those not very well acquaintedwith radiochemistry, a note on the characterizationof radioactive compounds: due to theminute concentration ranges being dealt with, aswell as the hazards involved with the useofradioactivity, radioactive compounds cannot becharacterized via the same methods as arenon-radioactive “cold” compounds. HPLC retentiontime comparison with the analogous,thoroughly characterized rhenium compoundis used as the primary method of identificationof technetium complexes. If coinjection of the rheniumcomplex and the technetium reactionmixture give peaks at the same retention time,this is considered proof that the equivalenttechnetium complex has been formed.The retention times of starting materials are alsoknown, and these must be different to thoseof the complexes for the results to bemeaningful.The radiochemistry of rhenium is less developedthan that of technetium, so many of thecompounds made in this area of research are madefor cold rhenium and for technetium butnot for radioactive rhenium.1.4.1 Oxidation States of TechnetiumAlthough all oxidation states from -1 to +7 are knownfor technetium, the vast majority ofthe chemistry to date has focused on two of them, +5 and+1.50There has been a lot of work,References begin on Pg 34 15over a long period of time,on the use of99mTc(V>.basedimaging agents.51’52A commonlystudied species in this oxidationstate is the technetium oxospecies,[99mTcO]3+.Thecoordination chemistry of this systemhas been well studied, and it is knownto form stabledistorted square pyramidalcompounds with tetradentate ligands.Although a wide range ofdonor atoms have been investigatedwith this core, much of the workdirected towardsradiopharmaceutical application hasinvolvedNS4 coordination spheres, first reported byDavison, Orvig and coworkers.53A derivative of this complex typethat has foundsignificant use is the N3S mercaptoacetyltriglycineor MAG3 core, which is often appendedwith various groups, as shown inFigure 1.5. There are many 99Tc(V)compounds inclinical use today,5’ most of whichowe their favourable biodistributionand tissueaccumulation to their overallsize, charge and lipophilicity, rather that the presenceof adirecting group such as a biomolecule.The major problems associated with these speciesarethe lack of control over the isomersformed54 (the syn and anti isomers are illustratedinFigure 1.5), and difficulties in characterizingthe protonation states of the complexesatphysiologically relevant pH.55 Theresulting diastereoisomers display different physiologicalproperties.00S—TCj-000aFigure 1.5 MAG3 isomers — syn and anti (namedfor the relative orientation of the TcO bondand the pendant carboxylic acid).As an example,99mTcDepreotideis a cyclic oligopeptide that acts as a somatostatin receptor(overexpressed by certain types of cancer cells) ligandand has been FDA approved for use inimaging lung cancer since 1999. It was oniyin 2007 that the syn and anti diastereomers ofthis compound were separated, thoroughly chemically identified,and their individualreceptor affinities and biodistributions examined.56’57The syn diastereomer makes up about90 % of the complex when it is synthesized viathe kit preparation used in making theradiopharmaceutical. This isomer has an 1C50 of0.15 nM, and a tumor uptake in mice of6.58 % ID/g (the percentage of total injected dose that ends upin one gram of a particularReferences begin on Pg 34 16tissue) compared to the loweraffinity anti isomer which exhibits an1C50 of 0.89 nM and atumor uptake of 3.38 % ID/g.Given that both isomers have favourableimagingcharacteristics, the small percentageof the lower affinity anti isomer doesnot proveproblematic, however a gap ofeight years between the introduction of99mTc..Depreotideontothe market and the full characterizationand understanding of its componentsis quitesurprising.The other technetium core that has historicallybeen the subject of much research alsoseestechnetium in the (proposed) +5oxidation state,58 here combinedwith ahydrazinonicotinamide (HYNIC)ligand system (see Figure 1.6). These compoundsaresynthesized from the[99mTcO]3+core by reaction with a functionalised hydrazinewhichreplaces the oxo group. As the HYNICligand does not fill the coordinationsphere, theretend to be multiple products formedas other ligand types and denticities bind to the metal,and although stable (no free pertechnetateis detected in vivo), the lack of thorough chemicalcharacterisation of these compoundsis problematic. The choice of co-ligand is crucialto thestability and biodistribution of these complexes,and work in this area is ongoing.586°OHQL1LfIL,Figure 1.6 A generic HYNIC ligand system — thepyridyl nitrogen may bind to the metal inone of the L sites. The acid can be used for conjugationto a biomolecule such as a protein.A simple and elegant synthesis of the very useful[99mTc(CO)3(H20)]+core (Figure 1.7) byAlberto et al.6’ has sparked great interest inresearching new SPECT radiotracers of thisradionuclide. The core is attractive for several reasons;it is small, kinetically inert, stable tooxidation, and is amenable to chelationby several types of ligating atoms. Given thisexciting combination of properties, MallinckrodtInc. (now called Covidien) has developed akit preparation of this core by boranocarbonate reductionfrom the pertechnetate anion:IsolinkTM.This has prompted an explosion of research interest,and has lead to the discoveryReferences begin on Pg 34 17of many new and potentiallyinteresting coordination compounds.Using this tricarbonylcore, the formation ofwell defined and thoroughly characterizedrhenium and technetiumcomplexes is now possible.COOH2ITcOHH20Figure 1.7 The technetium(I) tricarbonylcore pioneered by Alberto and coworkers;6’theaqualigands can be readily replaced, creatinga useful platform for radiopharmaceuticaldevelopment.1.4.2 Technetium (I) BioconjugatesJaouen and coworkers were pioneers inthe field of protein labeling with the organometallictricarbonyl core. In 1993 they reportedthe labeling of proteins with [CpRe(CO)3](whereCpis cyclopentadiene).62 An N-hydroxysuccinimideester bound to the Cp was reacted with freeamines on the protein of interest, andthe monoclonal antibody JOSS2-2 was foundto retainsatisfactory receptor recognition uponbeing labeled on 15 % of its available sites inthisfashion.62 After the development of the aqueousprecursor for the tricarbonyl core,6’a lotmore research began to focus on these M(I) labelledbioconjugates. The first synthesis of anorganometallic bioconjugate starting fromthe[99mTc(H20)3(CO)1+core was by Alberto,Schibli and Schubiger,63 who successfully labeleda 5-HT1A serotonergic receptor ligand(Figure 1.8 a), and found it to retain its receptor affinityonce labeled. Since then, the majorclasses of biomolecules (nucleic acids, lipids, peptidesand carbohydrates) plus examples ofother small molecule receptor ligands haveall been labelled with the tricarbonyl core.Two examples of small molecule labellingwith99mTcare shown in Figure 1.8. 1-(2-Methoxyphenyl)-piperazine has beenbound to the tricarbonyl core via a bidentate chelate(Figure 1 .8a), as well as by way of a cyclopentadieneyl(Cp) moiety (see Chapter 5 for moreinformation). The rhenium complex of thisCp ligand has a very high affinity (1C50 = 6 nM)for the 5-HT,A serotonergic receptor.64 Anotherexample of small molecule bioconjugateformation came from Zubieta andcoworkers;65 biotin was linked via a five carbon alkylchain to a fluorescent tridentate bindingsite (Figure 1.8b). This ligand forms stablecomplexes with both the rhenium and technetium tricarbonylcores, and these complexesretain very high binding affinity for the biotin receptor avidin.65References begin on Pg 34 18Single nucleotide bases have beenbound to the tricarbonyl cores. Guanine binds inamonodentate fashion through its N-7 atom,and two guanine molecules may be bound to eachmetal (Figure 1 .8c).66 In the solidstate these two guanines are found in a headto tailarrangement. An in depth NMR studyof the solution isomers and conformations resultingfrom the binding of nucleoside monophosphates(NMPs) to the rhenium tricarbonyl coreshowed that many binding modes werepossible with NMPs — with monodentate bindingbeing the most common.67 There isparticular interest in these species as it has beenproposed that [Re(CO)3(H20)f’mayexhibit anticancer properties in cell studies, possiblyby crosslinking DNA bases, in a way mechanisticallysimilar to cisplatin.67 Recently,chelating moieties have been appendedto a structural mimic of DNA, in the hope of usingthis as an antisense oligodeoxynucleotidethat is recognized by mRNA and incorporated intoareas with upregulated gene expressionof a targeted gene. No biological data is yetavailable, but this technology would be an importantdevelopment in cancer diagnostics, aschanges in patterns of gene expressionmay point to disease states before symptoms appear.68Fatty acid type labeling has been achievedusing C18 alkyl chains as Lipiodol surrogates —Lipiodol is a mixture of iodinated (with131jfor therapy in liver cancer) fatty acid esters frompoppy seed oil found to accumulate in andbe retained by the liver. Coordination via bothbidentate and tridentate ligand sets with pendantC18 chains to the Re,99mTcand ‘86Retricarbonyl cores was reported by Albertoand coworkers.29 The resulting complexes (Figure1 .8d) were stable for 24 hr in Lipiodol and 48 hr in an ethanol/water mixture,and are beinginvestigated as potential diagnostic(99mTc)and radiotherapeutic (‘86”88Re) pairs for thetreatment of liver cancer.Most work on bioconjugates of technetium has focused on the functionalisationof peptides.Santos and coworkers have used pyrazole-based tripodalligands to attach a fragment of thepeptide bombesin to the tricarbonyl core, and found thatthe resulting complexes were stableand retained affinity for cells that contained the receptor forbombesin.28 Zubieta, Valliantand coworkers developed the single amino acid chelate(SAAC) technology, whichincorporates a non-natural amino acid containing a tridentatemetal binding group in itssidechain, which can then be added into a peptide chain.7’Anelegant extension of this workinvolved the incorporation of the non-natural, chelating aminoacid, either alone or with therhenium tricarbonyl core bound, into the automated synthesis ofa peptide, with the resultingReferences begin on Pg 34 19compounds having similar affinityfor the receptor as the parent (Figure 1 .8e).69 Thesamegroup has reported a methodfor the direct labeling of proteins by reaction withmaleimidelinked to one end oftheir bifunctional chelate. The maleimide reactsselectively withsulfhydryl groups to producea stable thioether linkage between themetal chelate and theside chain of cysteine residues.72 RecentlyAlberto and coworkers reported the synthesis ofasingle amino acid labeled with{99mTc(Co)31+(Figure 1 .8f).7° Its rhenium analog is takenupinto cells via the LAT-1 aminoacid transporter. This is an important discovery, asit showsthat adding significant bulk (a linker,binding group and a metal ion) to a relatively smallamino acid, does not always meansacrificing the enzymatic recognition and activityof theparent molecule.Figure 1.8 Some examples of small biologically active moleculeslabeled with the tricarbonylcore: a) a5-HT1A serotonergic receptor ligand with a bidentate chelate, M= 99mTC63b) afluorescent biotin-conjugated compound, M =Re,99mTC65c) the binding of guanine to thetricarbonyl core66 d) a long Lipiodol-like alkyl chain labeled withthe tricarbonyl core,29 M Re,99mTC l8bRee) a peptide incorporating a single amino acid chelate69f) a non-natural amino acidthat retains LAT- 1 amino acid transporter activity.°0aCHNNT NHH2NNNNNNHebHNNHdCoOC ‘COf0OHOHNH2OC’ ‘COCOReferences begin on Pg 34 20Carbohydrates have also beenconjugated to the technetium tricarbonyl core, but asthis areais the topic of a large part of thisthesis, it is covered in more detail in the introductionsections of the relevant chapters.1.5 Alzheimer’s DiseaseAlzheimer’s Disease (AD) is a debilitating conditionthat is the fourth leading cause of death inthe Western world.73 Although AD wasfirst discovered by Alois Alzheimer in 1906, over acentury later we still rely on the same primitive postmortem analysis to definitively diagnoseAD. This is done by looking for two histologicalfeatures in a brain biopsy; neurofibrillarytangles and f3-amyloid plaques. The tangles aremade of hyperphosphorylated ‘v-protein, whilethe plaques consist mainly of the 3-amyloid protein.AD affected 26.6 million people worldwidein 2006, with incidence predicted to triple by2050. The Canadian health care systemis estimated to have spent $1.7 billion in 2001 on thetreatment and care of people with AD,76 and that amount will continueto increase as incidencegrows.77 The symptoms of AD are a loss of memoryand cognitive function, which lead to agradual loss of overall function, and an inabilityto live an independent life. There is no knowncure, so this decline in function and abilities, after a variable lengthof time, culminates in death.As the populations in developed countries continue toage, AD will place an increasing burdenon society, both socially and financially.1.5.1 The Amyloid Cascade HypothesisWhile 95 % of AD cases are late-onset, where age is theonly firmly verified risk factor, there isa small percentage of the affected population that develops thedisease at a young age (below60). These people are said to have early onset AD. Genetic testing has shown mutationin fourkey genes to be a risk factor for developing AD: amyloid precursorprotein (APP), presenilin 1(PSi), and presenilin 2 (PS2) mutants have been linked with earlyonset AD. A certain allele(E4) of apolipoprotein, ApoE4, has been shown to be a risk factor forlate onset AD, with thosehaving two copies of E4 having increased risk over those with onecopy. Amyloid PrecursorProtein (APP) is a membrane bound protein that is cleaved to produce smallerA13fragments.Depending on the enzymes that control the cleavage, A3140 orA13142will be produced.References begin on Pg 34 21Although both these fragmentsare found ubiquitously and are normalin undiseased individuals,they are found in high concentrationsin the brains of people with AD, and theA13i42that isnormally the minor componentis seen in higher relative concentrations. Althoughthe exact roleof PSi and 2 are not known,they are known to be involved inthe processing of APP, so may beinvolved with an overproductionof Af3 and/or the alteration in theAf3io/A13i42 ratio seen mAD. ApoE is a protein found inmany organs of the body and with severalfunctions involved inthe binding and transporting oflipids. There are two main waysin which it has been implicatedin AD; one is in the proteolytic degradationofA13,leading to a decrease in brain concentrationofsoluble Af3,78 and the other isin the binding and cross linkingof theA13fragments.79 Work inthis area is ongoing, and thereis still contention as to how these functionsfit together to increasethe risk of AD.These genetic studies have helped to elucidatethe key proteins in the cause and progression ofAD, and aid our current understanding ofthe disease. As these mutations accountfor only avery small proportion of peoplewho develop AD, work into these geneticcausal and predictivefactors is continuing.80 Although apoE4was identified as a risk factor for old age ADin1993,81and the usefulness of such genetic markersis widely recognized, it remains the only reliablegenetic determinant of late onset AD.82The amyloid cascade hypothesis was first postulatedby Hardy and Higgins in1992.83Itproposed that the deposition of 3-amyloidwas the cause and first step in the progression of AD,and that all other histologies and symptomsare downstream effects. In the years following thisproposal, a great deal of researchhas verified and expanded upon this idea, and itis nowaccepted by many working on AD thatthe errant behavior of the amyloid protein is a key causalfactor in the developmentof AD. The key evolution to this theory is that it is not simply thepresence of the characteristic amyloid deposits themselvesthat are the toxic, problematicspecies, but it is now thought to be solubleforms of amyloid in close proximity with metal ionsthat are the key toxic species.8413-Amyloid (amyloid, or Af3) isa protein found ubiquitously in human cells, butits naturalfunction is not yet well understood. It is formed bythe precise cutting of a amyloid precursorprotein (APP), and there are several enzymes involvedin this cleavage; c, f3, and ‘y-secretases. 3-secretase cleaves to form the N-terminalend of the Af3, while the position the y-secretase cutsReferences begin on Pg 34 22determines the exact positionof the C-terminus, and therefore the lengthof the resulting peptide.Normally y-secretase cleavesto give a 40 amino acid peptide calledAf3140,but in AD, forreasons not fully understood, APPis more frequently cleaved between residue 42 and 43to givea larger proportion of Af3142. Residues41 and 42 are alanine and isoleucine respectively,bothquite hydrophobic amino acids, whichgivesA13142a much higher propensity for aggregatingwith itself, meaning it is more proneto forming oligomers and plaques inthe aqueousenvironment of our bodies.3-Amyloid contains some very highaffinity copper binding sites,85 reminiscent of thosefoundon prion proteins that are known to bindcopper in vivo. It is thought that the natural role off3-amyloid may somehow involvethe transport and/or storage of copper. There has beenmuchspeculation as to why a protein that naturallyoccurs throughout our bodies can lead tosuchtoxicity in this disease state. Onewell documented answer is that it is the positionof theamyloid within the cell that is problematic.Under normal circumstances the vast majority ofamyloid is membrane-associated. Whenbinding metals, amyloid forms a hexamer, which isembedded in a membrane, and is not redox active.The propensity of the amyloid to be insertedinto the membrane is affected by several factors suchas the pH, lipid composition and thepresence of metal ions, and as such is very delicately balanced.86If the amyloid becomesoxidized, it is proposed that it becomes redoxactive, and less strongly associated with themembrane. This initial oxidation is likely causedby the increased concentrations of freelyreleasable redox active copper in the aging brain.As the brain becomes aged and worn, themany enzymes required for the tightly regulated homeostasis of potentiallyreactive metal ionsmay not do their job so well, and the concentrations offree metal ions may increase.There are high levels of iron, copper and zinc found in theneuropil of the brains of ADsufferers.87 The levels of these metals found are three to five times thoseoccurring in the brainsof non-AD sufferers of the same age.87 There is evidence that the normal enzyme-assistedhomeostasis of these metal ions is drastically altered in AD brains.88 Themechanisms thatcontrol the homeostasis of these ions become worn and less efficient with age,which may helpexplain why age is the main causal factor in developing AD. Also,as the blood brain barrierbecomes worn with age it may not do such a good job of regulating thebrain’s uptake andregulation, so this may lead to altered metal concentrations.References begin on Pg 34 23Free copper may be takenup by the high affinity binding sites onamyloid, but if the systembecomes overwhelmed with copper,it is possible that other, redox active siteson the Af3 becomeoccupied, or that the copper is not bound,so will react when it comes into contactwith a suitableoxidation partner. Cu2 canbe reduced to Cu quite readily in thepresence of an easily oxidizedsubstrate such as a methionineor tyrosine residue.89 Af3 oxidized at Met35,a very commonsiteof oxidation, has been shown to emergefrom within the membrane, and atthis point thesolubilised peptide, bound to, or in closeproximity to free metal ions, becomescapable of redoxactivity to produce peroxide. Once peroxideis produced in the vicinity of the metalions copperand iron, Fenton type chemistry canlead to the production of reactive oxygenspecies (ROS)such as OW, a highly reactive radical. The reactivityof ROS means that they are unselective intheir activity, and will react with anddamage whatever molecules they firstcome into contactwith. In vivo they are commonly in closeproximity and therefore react with peptides, lipids,orDNA. Damage to such importantmolecules leads to a loss of function, may stimulatetheinflammatory response with all the complicationsthat invokes, and particularly in the caseofDNA damage can lead to cell death.In 1999 it was proposed that soluble formsof amyloid were responsible for the problemsassociated with AD.9° Although amyloid plaque loadingcould be used to categorize patients asAD and non-AD, it did not vary reliablywith disease severity, and was in fact found tobeinversely correlated to oxidative damage.9’On the other hand, soluble amyloid was shown tovary in line with stage and severity of disease.If the species that produce ROS are soluble, theyare not confined to one specific areawithin the brain. They can diffuse into and aroundtheinterstitial spaces where they continue to catalyticallyproduce ROS and damage cells and undercertain circumstances they can also run into highconcentrations of zinc ions. Zinc is releasedfrom synapses upon activation by electrical pulse.This is the method via which signals aretransmitted and information is passed throughthe nervous system. Under normal conditions, assoon as the nerve impulse has passed, the synapsereabsorbs the zinc. But if amyloid oligomersare nearby, they will compete for thezinc by binding it in one of their many possible metalbinding sites. It has been shown thatsome of the binding sites on Af3 have a high enoughaffinity for Zn2 that they can compete for the zincon the time scale of reabsorption by thesynapses.92 The metal ions can catalyse the cross-linking ofnearby peptides, so these zincbound species can get quite large, and become insolublefibrils with all kinds of oxidative crosslinking taking place to cement the aggregate together.The binding of zinc to the oligomersReferences begin on Pg 34 24quenches some of the redoxactivity, possibly through the replacementof the copper by theredox silent zinc.It has been proposed thatthe amyloid plaques are actually protective,and that their formationleads to the trapping and redox-silencingof potentially dangerous metal ions.91 Despite theideathat it is likely the soluble oligomersof amyloid that cause the most damage,the presence ofplaques is also detrimental.Plaques are still found to be able to produceperoxide and ROS, andcause secondary effects such as the inflammatoryresponse and microglial activation.93It is alsopossible that the plaques act as reservoirswhich are in equilibrium with the soluble amyloid,andrelease oligomers over timeto keep the levels of soluble speciesat a constant concentration.941.5.2 Current Treatments for Alzheimer’sDiseaseThere is no cure for Alzheimer’s disease.Symptoms develop due to extensive cell damageanddeath, and these effects are not reversible. Currenttreatments are administered with the hope ofslowing the progression of thedisease. There is significant research going oninto bothtreatments and early diagnostic tools forAD.It is possible to predict and diagnose thedevelopment of AD with currently existing methods.For example, FDG (shown in Scheme 1.1)has been used in many clinical studies, and providesreliable differentiation between AD patientsand age matched controls.95 Several studies havefound that glucose metabolizm is lowered in certain brainregions in patients who will go on todevelop dementia compared to those who will not.96However, it is not yet accurate enough tobe used in routine screening, and does not do a reliablejob of predicting AD compared to othertypes of dementia. Practically this type of diagnosisis not viable, as the whole population wouldneed to be scanned before any symptoms were visible,but it is important to help understand thedevelopment of the disease.There are four compounds currently approved forthe treatment of AD in Canada; these areshown in Figure 1.9. Galantamine, rivastigimine anddonepezil (Figure 1 .9a-c) are approved foruse in mild to moderate disease states, and all workas acetylcholinesterase inhibitors. The mainfunction of these compounds in treating ADis in inhibiting the action of the protein that breaksdown acetyicholine — a neurotransmitter that has reducedlevels in AD.97 By preventing thisReferences begin on Pg 34 25breakdown, the brain hasmore of this important neurotransmitter.This is now agreed to be avery downstream affect inAD; given the efficacy of these treatmentsit has been proposed thatthey may act on some other,upstream, targets as well,and that this may lead to the observedtherapeutic effect.bOMe0dNH2MeOFigure 1.9 Drugs currentlyon the market in Canada for the treatmentof AD: a) galantamine b)rivastigimine c) donepezild) memantine.Memantine (Figure 1 .9d) is used inmoderate to advanced disease states, and worksas anNMDAR antagonist.98 NMDA isN-methyl-D-aspartic acid, and its receptor NMDARis an ionchannel (especially important for Ca2transmission) that is open when bound to its naturalsubstrate glutamate. In AD, glutamate,an excitatory neurotransmitter,is often overproduced,and this can lead to overstimulationof the NMDA receptor and cell death. Memantine bindstothe NMDA receptor in the brain and inhibitsthe natural substrate glutamate from stimulating thereceptor. Memantine is of low enough affinitythat it can be displaced by high concentrations ofglutamate, allowing for normal neurotransmissionto take place.Antioxidants are currently used in thetreatment of AD. These react with free radicals inthebrain to prevent them reacting with other biologicallyimportant molecules, and thus to preventthe furthering of oxidative damage. As the oxidativedamage is only one facet of a complicateddisease, this treatment appears to be somewhat symptomatic,and does not fully prevent theadvancement of AD, only slowsit down.99 A clinical trial of a-tocopherol (vitamin E), a knownantioxidant, administered to AD patients showed thatthis strategy has merit.’00 The patientswho were given vitamin E lived on average 230 days longer,and had delayed admission to aninstitution compared to the non-treated control group.’°° However,there was no significantimprovement in the cognitive test scoresof the treated vs. the non-treated Thissuggests that while antioxidants have a role to play in reducingthe downstream effects of AD,alone they are not powerful enough to stop or reversethe disease’s progress.aReferences begin on Pg 34 26Non-steroidal anti-inflammatorydrugs (NSAIDs) have been postulatedto help in AD for quitesome time.’0’ They are expectedto help relieve the symptoms of AD by reducing inflammationthat occurs as part of thenatural inflammatory response. Alzheimer’spatients exhibit asignificant inflammatory response as aresult of the biochemical trauma thatoccurs in theirbrains. A study of transgenicmice who were given large daily doses of ibuprofenfor six monthsfrom the time their AD symptoms first starteddeveloping showed a significant reductionin thenumber and size ofA13plaques compared to non-treated controls.’°2There have been conflictingreports on the effectiveness of such treatmentsin humans. One of the most comprehensivestudies was published recently, where seven differentdoses of 21 NSAIDs were given toover49,000 patients over a five year time period.’03 Fourof the NSAIDs were also known to haveanti-Af3,42 effects, and decrease serumlevels of the peptide as seen in transgenicmice. Overallthere were some protective effectsseen, and these were more noticeable in thecases of knownanti-Af3 NSAIDs, and the longer the subjectwas given the drug.Beta (J3) and gamma (y) secretases are responsiblefor cleaving the ends of Af3 from APP.’°4Inhibition of these enzymes as amethod to control the amount of amyloid producedhave beenthe subject substantial research over thepast decade or so, including at big pharmaceuticalcompanies. Gamma secretase cleavesthe C-terminus of A, and can do so in several positionsto form a peptide between 39 - 42 residueslong.’04 Gamma secretase is a complex structuremade up of several proteins embedded inthe membrane,’°5and known to contain PSi and 2,genetic mutations of which are linkedwith early onset AD (see above). Beta secretase orBACE- 1 (beta site of APP cleaving enzyme1) is responsible for the first step in the formation ofA13- the cleavage of the N terminus.104 This area of theAPP can alternatively be cleaved byalpha secretase, and when this occurs a different fragmentis produced that does not go on toproduce any of the effects of AD. BACEhas not been genetically linked to AD, but it is foundin high levels in the brains of people with late onsetAD.Inhibition of either of these secretases reduces the amountofA13in the brain, so inhibitors ofboth have been under investigation as AD therapeutics.A gamma secretase inhibitor was testedin a trial of 51 AD patients by Eli Lilly for12 weeks.’°6 They found a significant reduction intheA13140in the plasma (‘— 60 %) of treated patients, and could not measuretheAI3i2.Therewas no significant change in the levels of A1o orA13,42in the CSF (cerebrospinal fluid) oftreated patients, as was expected. This may be due to the lag timeof A3 levels equilibratingReferences begin on Pg 34 27between the serum and the CSF.There were no changes in the behaviouraltesting of the treatedgroup, but the authors state thatgiven the relatively short timeframe of the study and the slowrate of disease progression, thismay be affected if treatment was sustainedover a longer timeperiod.’0613-secretase (BACE) inhibitors are a little behind the y-secretases in development,but a lot ofresearch is now being done inthis area. Many research groups and companieshave come upwith potent inhibitors of -secretase,’°7including some that show reductionof Aj3 in animalmodels.’°8 But it is yet to be seenhow these perform in large scale humantrials.Another strategy for the potential treatmentof AD has been the development ofan anti-amyloidvaccine.109 The way a vaccine wouldwork is by the immunisedbody raising antigens to thepeptide, allowing it to developincreased clearance mechanisms forA13.Problems occurred inclinical trials when a smallbut significant percentage of the population developedmeningoencephalitis due to reactionof the This problem is being circumventedbyusing shorter chainA13analogs, which are showing promise in mouse andmonkey models.”A slightly less orthodox suggestionis that the restriction of calories may help to treatAD.”2This is proposed to work becausea significant lack of calories may trigger a mildstress responsewhich is known to increase the productionof proteins that promote the growth and survivalofneurons.As AD is such a complex conditionthat is still not thoroughly understood, there are several otheravenues of possible treatment being investigated.Those outlined here are either currently useddrugs or those with a large amount of researchinterest that have shown promise as potentialtherapeutics.1.5.3 Metal Chelators to Treat Alzheimer’s DiseaseDesferrioxamine (DFO) (Figure 1.1, righthand box), a well known metal chelator, with aparticularly high affinity for trivalent metal ions,was investigated as a treatment for AD.’13 Atthe time it was thought that aluminum was involvedin AD, so the hypothesis behind these testswas that the DFO would chelate the Al3 ions andthereby slow the progression of dementia.References begin on Pg 34 28After two years of twicedaily intramuscular DFO injections, it wasfound that the DFO had hada significant effect, with thenon-treated group intellectually deterioratingtwice as fast as thetreated group. There wereno improvements, just a slowing of decline observedin the treatedpatients. The authors note thatthe DFO would chelate Fe3 even more readily than Al3,thoughwhen iron levels were lowa threefold increase in aluminum levelsin the urine was induced bythe DFO. However, it is mentionedin this initial paper, and research since this publicationhasadded much weight to the argument, thatit may be the chelation of the iron that was responsiblefor a significant amount or eventhe majority of the observed effect. It is noted in the paperthatthe formation of radicals is promoted byfree iron, so the removal of this would havethe capacityto reduce the oxidative stress of the patient,slowing oxidative damage and therefore neurologicaldecline.There was a large time lapse betweenthe DFO studies and the next investigations ofmetalchelators as a treatment option for AD. Clioquinol(CQ) (Figure 1.1 Oa) is a bidentate chelator,with an N 0 binding sphere, known tobind to a range of metals.114’115It was used as an off-the-shelf, proof of concept therapeutic, as it had FDAapproval as an anti-infective,”6meaning itcould relatively easily be used in humantrials. In vitro studies had established that CQ coulddissolve Af3 aggregates and inhibit the toxicity of Af3to neuronal cells.”7 Preclinical trials werecarried out in Tg2576 mice that weregenetically engineered to overexpress AJ3,42 and thusdevelop the histologies and symptoms ofAD.”7 These mice were dosed with CQ at 30mg/kg/day for 9 weeks and at the end of this timethey showed improved cognitive abilities, asmeasured in a water maze, compared with their controltreated transgenic peers.”7 The treatedmice were also found to have a 49 % reduction inAf3 deposition over this time period. By theend of the nine weeks, the treated animals weremore healthy than the non-treated controls interms of weight and various toxicology measures.117 Interestinglytriethylaminetetraamine(TETA) (Figure 1.1 Ob) was used as a control for generalmetal chelation throughout the bodybeing responsible for the observed effects, and it was found notto have a significant effect onthese mice. This is proposed to be because unlikeCQ, TETA does not cross the BBB, so cannotaccess the brain and therefore cannot chelate the metals in the crucial places.”7This findingshows the lack of utility in the approach to systemic metal chelationas a treatment for AD — thepotential therapeutic must have brain access tobe effective, at least in the area of metalchelation.References begin on Pg 34 29/—COO(CH2)0C18H37Cab//\H2N N N NH2H HOH( 2)2 18 37Figure 1.10 a) Clioquinol (CQ) andb) triethylenetetraamine (TETA)”7c) DP-109,”8chelatorsused in mouse model studies of AD.These results from the CQ mouse study were promising,so a phase II clinical trial was carriedout in humans (phase I was not required asCQ had prior FDA approval for human use, and themajor side effects are known).116Here the results were mildly encouraging. Thirty sixpatientswith varying stages of AD were given oral doses ofCQ twice daily for 36 weeks with all dosesincreasing over time.116 Benefit wasmost obvious in the patients with more advanced AD,but itwas not found to be statistically significant over thosegiven placebo after 36 weeks as measuredby cognitive testing results. After fourweeks there was a small improvement seen in the severecases, and the authors propose this is due to theability of CQ to neutralize the pooi of solubleA.”6 Plasma levels ofA13,42were measured and were found to decrease in the treatedpatients,and to maintain a lower level over the course of thetrials.”6 This finding was more pronouncedin the more severe cases of AD. Plasma metal levelswere also monitored, and the amount ofzinc in the plasma increased upon treatment withCQ, whereas the copper plasma levels were notaffected.”6Zinc levels prior to beginning the study werelower than age matched controls, so anincrease in zinc may indicate some kindof restorative effect. A study of two patients with earlyonset AD showed similar results.”9 The natureof this study means that no hard conclusions canbe drawn, but one patient was observed to be stableover the fourteen months of treatment, whilethe other showed a slight improvement followedby stabilization over the nine months of CQadministration.The authors propose CQ as a prototypical metal-proteinattenuating compound (MPAC).’6 CQis a moderate affinity chelator, but is not strong enoughto, for example, remove copper from thehighest affinity binding site on Af3 — with approx 8 aM affinity.85CQ could effectively treat ADdespite having a lower affinity for key metal ions than some of theprotein sites with which itwas competing. This led the authors to suggest that the metal didnot have to be fully removedReferences begin on Pg 34 30from the M — Af3 interaction,but rather that interaction just neededto be perturbed so that thebrain’s homeostatic mechanismscould have a chance to catch up andreinstate its normal state ofThus the attenuation of the diseasestate was all that was required, and the brainwouldhelp itself once some of theplaques and oligomers were brokenup, and the rate of ROSproduction was slowed.A compound very similar to CQ, PBT2whose structure has never been disclosed,was designedto overcome some of the weaknesses inCQ.’2’ It is also an 8-hydroxyquinoline,but without aniodine attached, and is easierto synthesize and has improved solubility and increasedBBBpermeability as compared toCQ.’22 In vitro PBT2 displayed the same traits asCQ in its abilityto inhibit the formation of crosslinkedA13oligomers, to redissolve Zn-aggregatedspecies and toprevent the redox activity that leads tothe formation ofH20.12’ In AD modelsin mice it wasable to improve the learning and memoryin a matter of hours after administration.’22 Thisisproposed to be due to the rapid clearanceand redistribution of solubleA13in the interstitial fluid,which the authors claim is maybewhat causes the symptoms of AD. Thecompounds weretested in two different strains of mice andthis theory is the only one consistent with the resultsseen in the two varieties of mice.122 The authorssuggest that this interstitial A3 affects theneurons ability to transmit signals, and thatthis may be an underappreciated facet of the humandisease as well. PBT2 is proposedto chaperone Cu and Zn ions away from Af3 andinto cellswhere the normal homeostatic mechanismscan take over. As discussed above, the complicatedrelationship between metal ion levelsandA13can be easily perturbed, and this helping hand maybe all that is needed to let the body heal itselfby utilizing its normal feedback cycles of proteinexpression and regulation. PBT2 treatmentof transgenic mice did not affect the levels of metalions seen in either the brain or the body, suggestingto the authors that it acts as an ionophore,chaperoning metal ions for a time only.PBT2 has recently finished a phase ha clinical trialwith good results.’23 It was administered to49 AD patients in two different doses in a once dailyoral dose for three months. There were nosignificant adverse effects seen. The higher doseof PBT2 elicited a significant reduction in theconcentration of Af3,42 in the CSF, but not in theplasma. This shows that PBT2 is somehowaffecting central Af3 clearance in vivo. Serummetal concentrations were not affected, againsupporting the fact that this compound helps redistribute metalsrather than strongly bindingthem. Most notably, there was a statistically significantimprovement in executive function seenReferences begin on Pg 34 31in the patients receivingthe higher dose of PBT2. Although theauthors note that these findingsneed to be verified in a largersample size over a longer period oftime, these results are veryexciting for the future of theMPAC approach to the treatment of AD.Other chelators have sincebeen investigated in what others believecould be a usefulmethodology for treating this debilitatingdisease. DP-l09 is a lipophilic chelator,”8thestructure of which is given in Figure1.1 Oc. It was found to reduce the numberof amyloidplaques while increasing levelsof dissolved Af3 when administered daily for threemonths inTg2576 mice.118 The authors claim DP-109shows some selectivity for copper andzinc overother metal ions and that it likely crossesthe BBB due to its lipophilicity and efficacy.”8Our approach is to incorporatethese metal binding characteristics that showsuch promisingresults with more biologically relevantfeatures (vide infra). This involves the additionofantioxidant capability into the moleculesto passivate ROS, as well as glycosylationto improvewater solubility, mask the chelating partof the molecule and target the resulting prodrugstocross the blood brain barrier.1.6 Thesis OverviewThis thesis focuses on the use of carbohydrateconjugation as a method for controlling thebiodistribution of compounds for use inmedicinal inorganic chemistry. The thesis covers twodistinct projects. This work fits into the categories ofmedicinal inorganic chemistry as outlinedin Figure 1.1.The first focuses on Alzheimer’s disease therapeuticsdeveloped in the Orvig group. The role ofthis work in that project was to use radiochemistry asa tool to determine the in vivo distributionof the compounds. As Alzheimer’s disease isa disease of the brain, the radiolabelledcompounds that were synthesized in this work were testedto see if they crossed the blood brainbarrier into the brain. This was a crucial step in the progress ofthis project, and as the outcomewas positive for the compounds crossing into the brain,work on this project has progressed withincreased hope of finding a multifunctional wayto treat Alzheimer’s disease. This work isdiscussed in Chapter 2.References begin on Pg 34 32The second part of this thesis focuseson the development of carbohydrate conjugates formolecular imaging with99mTc.This work includes a study of different tridentate, monoanionic,glucosamine based ligands and their binding tothe [M(CO)3]core (M = Re,mTc).It alsoexamines the effects of varying the lengthof the linker between the carbohydrate and the metalbinding portion of the molecule. The variation inlinker length is applied to both the tridentateligand sets and Cp (cyclopentadienyl) based compounds.These compounds are synthesized(both cold and radioactive analogs), characterised,and tested for stability under biologicallyrelevant conditions. They are then subjected to assayswith the key enzymes in carbohydratebiodistribution; GLUT-i and hexokinase. This workis outlined in Chapters 3, 4 and 5.References begin on Pg 34 331.7 References1. Alderden, R. A.; Hall, M. D.; Hambley, T.W., I Chem. Ed. 2006, 83, 728 - 734.2. Chan, K. W.-Y.; Wong, W.-T.,Coord. Chem. Rev. 2007, 251, 2428 -2451.3. Singh, S.; Khodr, H.; Taylor, M. I.; Hider,R. C., Therapeutic Iron Chelators and TheirPotential Side-effects. In Free Radicals andOxidative Stress: Environment, Drugs andFood Additives, Rice-Evans, C.; Halliwell, B.; Lunt,G. G., Eds. Portland Press: London,1995;pp127-137.4. Cyr, J. E.; Pearson, D. E.; Wilson,D. W.; Nelson, C. E.; Guaraldi, M.; Azure, M. T.;Lister-James, J.; Dinkelborg, L. M.; Dean, R. T.,I Med. Chem. 2007, 50, 1354 - 1364.5. Sharkey, R. M.; Goldenberg, D. M., I Nuci. Med. 2005, 46,Si 15 - S 127.6. Rose, B.; Matthay, K. K.; Price, D.; Huberty,J.; Klencke, B.; Norton, J. A.; Fitzgerald, P.A., Cancer 2003, 98, 239 - 248.7. Abrams, M. J.; Davison, A.; Jones, A. G.;Costello, C. E.; Pang, H., Inorg. Chem. 1983,22, 2798 - 2800.8. Boucher, C. A.; Wackers, F. J.; Zaret, B. L.; Mena, I.G., Am. I Cardiol. 1992, 69, 22 -27.9. Delbeke, D., I Nuci. Med 1999, 40, 591 - 603.10. Foster, N. L., Exp. Neurol. 2003, 184, 2 - 8.ii. Grossband, M. L., Monoclonal Antibody-Based Therapy ofCancer. Marcel Dekker:NewYork, 1998.12. Gittoes, N. J.; Franklyn, J. A., Drugs 1998, 55, 543 - 553.13. Lee, C. K.; Aeppli, D. M.; Unger, J.; Boudreau, R. J.; Levitt,S. H., Am. I Clin. Oncol.1996, 19, 102 - 107.14. Davies, A. J., Oncogene 2007,26,3614-3628.15. Rufini, V.; Fisher, G. A.; Shulkin, B. L.; Sission, J. C.; Shapiro, B.,I Nuci. Med. 1996,37, 1464 - 1468.16. Sgouros, G., Adv. Drug. Deliver. Rev. 2008, 60, 1402 - 1406.17. Nilsson, S.; Franzen, L.; Parker, C.; Tyrrell, C.; Blom, R.; Tennvall, J.; Lennernas,B.;Petersson, U.; Johannessen, D. C.; Sokal, M.; Pigott, K.; Yachnin,J.; Garkavij, M.;Strang, P.; Harmenberg, J.; Boistad, B.; Bruland, 0.S., Lancet Oncol. 2007, 8, 587 - 594.3418. Li, Y.; Tian, Z.; Rizvi, S. M.A.; Bander, N. H.; Allen, B. J., Prostate CancerP. D. 2002,5,36-46.19. Allen, T. M., Nat. Rev. Cancer2002, 2, 750 - 763.20. Zalutsky, M. R.; Vaidyanathan,G., Curr. Pharm. Design 2000, 6, 1433 - 1455.21. Wilbur, D. S.; Hamlin, D. K.;Chyan, M.-K.; Brechbiel, M. W., Bioconjugate Chem.2008, 19, 158 - 170.22. Steffen, A. C.; Almqvist, Y.; Chyan,M.-K.; Lundqvist, H.; Tolmachev, V.; Wilbur, D.S.; Carisson, J., Oncol. Rep. 2007, 17, 1141 - 1147.23. Talanov, V. S.; Yordanov, A. T.; Garmestani,K.; Milenic, D. E.; Arora, H. C.; Plascjak,P. S.; Eckelman, W. C.; Waldmann, T.A.; Brechbiel, M. W., Nuci. Med. Biol. 2004, 31,1061 - 1071.24. Robinson, M. K.; Shaller, C.; Garmestani,K.; Plascjak, P. S.; Hodge, K. M.; Yuan,Q.A.; Marks, J. D.; Waldmann, T. A.; Brechbiel, M.W.; Adams, G. P., Clin. Cancer Res.2008, 14, 875 - 882.25. Adam, M. J., I Labelled Compd. Radiopharm.2002, 45, 167 - 180.26. Knesaurek, K.; Machac, J., BMC. Nuci.Med. 2006, 6, 5.27. Canadian, I. f. H. I., Medical Imaging Technologiesin Canada, 2006 - Supply,Utilization and Sources ofOperating Funds 2006.28. Alves, S.; Paulo, A.; Correia, J. D. G.; Gano, L.;Smith, C. J.; Hoffman, T. J.; Santos, I.,Bioconjugate Chem. 2005, 16, 438 - 449.29. Saw, M. M.; Kurz, P.; Agorastos, N.; Hor, T. S. A.; Sundram,F. X.; Yan, Y. K.; Alberto,R., Inorg. Chim. Acta 2006, 359, 4087 - 4094.30. Lawrentschuk, N.; Poon, A. M. T.; Scott, A. M., Cliii. Nucl. Med 2006,31, 788 - 789.31. Lazzeri, E.; Manca, M.; Molea, N.; Marchetti, S.; Consoli,V.; Bodei, L.; Bianchi, R.;Chinol, M.; Paganelli, G.; Mariani, G., Eur. I Nuci.Med. 1999, 26, 606 - 614.32. Stick, R. V., Carbohydrates: The Sweet Moleculesof Lfe. Academic Press: London,2001.33. Garrett, R. H.; Grisham, C. M., Biochemistry. 2 ed.; SaundersCollege Publishing:Orlando, 1999;p705.34. Warburg, 0., Science 1956, 123, 309 - 314.35. Brown, R. S.; Goodman, T. M.; Zasadny, K. R.; Greenson, J. K.; Wahl, R.L., Nucl. Med.Biol. 2002, 29, 443 - 453.3536. Bennett, W. S. J.; Steitz,T. A., Proc. Nat!. Acad. Sci. 1978, 75, 4848 - 4852.37. Rempel, A.; Mathupala,S. P.; Pedersen, P. L., Glucose Catabolism in CancerCells: Roleand Regulation of Hexokinase Overexpression.In Cell Growth and Oncogenesis,Bannasch, P.; Kanduc, D.; Papa, S.; Tager, J. M.,Eds. Birkhauser: Basel, 1998;pp3 -14.38. Pelicano, H.; Martin, D. S.; Xu, R.-H.; Huang,P., Oncogene 2006, 25, 4633 - 4646.39. Bertoni, J. M.; Weintraub, S. T., I Neurochem.1984, 42, 513 - 518.40. Yoshioka, K.; Saito, M.; Oh, K.-B.; Nemoto,Y.; Matsuoka, H.; Natsume, M.; Abe, H.,Biosci. Biotech. Biochem. 1996, 60, 1899 - 1901.41. Speizer, L.; Richard, H.; Howard, K., Biochim.Biophys. Acta 1985, 815, 75 - 84.42. Battaglia, G.; La Russa, M.; Bruno, V.;Arenare, L.; Ippolito, R.; Copani, A.; Bonina, F.;Nicoletti, F., Brain Res. 2000, 860, 149 - 156.43. Zhang, M.; Zhang, Z.; Blessington, D.;Li, H.; Busch, T. M.; Madrak, V.; Miles, J.;Chance, B.; Glickson, J. D.; Zheng, G., BioconjugateChem. 2003, 14, 709 - 714.44. Joost, H.-G.; Thorens, B., Mo!. Membr. Biol.2001, 18, 247 - 256.45. Medina, R. A.; Owen, G. I., Biol. Res. 2002,35, 9 - 26.46. Rudlowski, C.; Moser, M.; Becker, A. J.; Rath,W.; Buttner, R.; Schroder, W.;Schurmann, A., Oncology 2004, 66, 404 - 410.47. Maschauer, S.; Prante, 0.; Hoffmann, M.; Deichen,J. T.; Kuwert, T., I Nuc!. Med 2004,45,455-460.48. Smith, T. A. D., Nuc. Med. Biol. 2001, 28, 1 - 4.49. Dilworth, J. R.; Parrott, S. J., Chem. Soc. Rev. 1998, 27, 43 - 55.50. Banerjee, S. R.; Maresca, K. P.; Francesconi, L. C.; Valliant, J. F.; Babich,J. W.;Zubieta, J., Nuc. Med. Biol. 2005, 32, 1 - 20.51. Mahmood, A.; Jones, A. G., Technetium Radiopharmaceuticals.In Handbook ofRadiopharmaceutica!s, 1 ed.; Welch, M. J.; Redvanly,C. S., Eds. John Wiley & SonsLtd.: Chichester, England, 2003;pp323 - 362.52. Liu, S., Chem. Soc. Rev. 2004, 33, 445 - 461.53. Davison, A.; Jones, A. G.; Orvig, C.; Sohn, M., Inorg. Chem. 1981, 20, 1629- 1632.54. O’Neil, J. P.; Wilson, S. R.; Katzenellenbogen, J. A., Inorg. Chem. 1994, 33, 319 - 323.3655. Marzilli, L. G.; Banaszczyk,M. G.; Hansen, L.; Kukienyik, Z.; Ci, R.; Taylor,A. J.,Inorg. Chem. 1994, 33, 4850 - 4860.56. Cantorias, M. V.; Howell, R.C.; Todaro, L.; Cyr, J. E.; Berndorff, D.; Rogers, R.D.;Francesconi, L. C., Inorg. Chem. 2007, 46, 7326- 7340.57. Cyr, J. E.; Pearson, D. E.; Nelson,C. A.; Lyons, B. A.; Zheng, Y.; Bartis, J.; He, J.;Cantorias, M. V.; Howell, R. C.; Francesconi, L.C., I MecL Chem. 2007, 50, 4295 -4303.58. King, R. C.; Surfraz, M., B.-U.; Biagini,S. C. G.; Blower, P. J.; Mather, S. J., DaltonTrans. 2007, 4998 - 5007.59. Babich, J. W.; Coco, W. G.; Barrow, S.; Fischman,A. J.; Femia, F. J.; Zubieta, J., Inorg.Chim. Acta 2000, 309, 123 - 136.60. Kovacs, M. S.; Hem, P.; Sattarzadeh,S.; Patrick, B. 0.; Orvig, C., I Chem. Soc., DaltonTrans. 2001, 3015 - 3024.61. Alberto, R.; Schibli, R.; Elgi, A.; Schubiger, P.A., I Am. Chem. Soc. 1998, 120, 7987 -7988.62. Salmain, M.; Gunn, M.; Gorfti, A.; Top,S.; Jaouen, G., Bioconjugate Chem. 1993, 4, 425- 433.63. Alberto, R.; Schibli, R.; Schubiger, P. A.; Abram, U.; Pietzsch, H.-J.; Johannsen,B., IAm. Chem. Soc. 1999, 121, 6076 - 6077.64. Bernard, J.; Ortner, K.; Spingler, B.; Pietzsch, H.-J.; Alberto,R., Inorg. Chem. 2003, 42,1014- 1022.65. James, S.; Maresca, K. P.; Babich,J. W.; Valliant, J. F.; Doering, L.; Zubieta, J.,Bioconjugate Chem. 2006, 17, 590 - 596.66. Zobi, F.; Springier, B.; Fox, T.; Alberto, R., Inorg. Chem. 2003, 42, 2818 - 2820.67. Adams, K. M.; Marzilli, L. G., Jnorg. Chem. 2007, 46, 4926 - 4936.68. Xavier, C.; Pak, J. K.; Santos, I.; Alberto, R., I Organomet. Chem. 2007,692, 1332 -1339.69. Stephenson, K. A.; Banerjee, S. R.; Besanger, T.; Sogbein,0. 0.; Levadala, M. K.;McFarlane, N.; Lemon, J. A.; Boreham, D. R.; Maresca, K. P.; Brennan,J. D.; Babich, J.W.; Zubieta, J.; Vailiant, 3. F., I Am. Chem. Soc. 2004, 126, 8598 -8599.70. Liu, Y.; Pak, 3. K.; Schmutz, P.; Bauwens, M.; Mertens, J.; Knight,H.; Alberto, R., IAm. Chem. Soc. 2006, 128, 15996 - 15997.3771. Banerjee, S. R.; Levadala,M. K.; Lazarova, N.; Wei, L.; Valliant,J. F.; Stephenson, K.A.; Babich, J. W.; Maresca,K. P.; Zubieta, J., Inorg. Chem. 2002, 41, 6417 - 6425.72. Banerjee, S. R.; Schaffer,P.; Babich, J. W.; Valliant, J. F.; Zubieta,J., Dalton Trans.2005, 3886 - 3897.73. Cavalli, A.; Bolognesi, M. L.;Minarini, A.; Rosini, M.; Tumiatti, V.; Recanatini,M.;Meichiorre, C., I Med. Chem. 2008, 51,347 - 372.74. Alzheimer, A., Allg. Zschr. fPsychiatr.Psychisch-Gerichtl. Mediz. 1907, 64, 146-148.75. Brookmeyer, R.; Johnson, E.; Ziegler-Graham,K.; Arrighi, H. M., Alz. Dement. 2007, 3,186 - 191.76. Tator, C.; Bray,G.; Morin, D., Can. I Neurol. Sd. 2007, 34, 268 - 269.77. Ostbye, T.; Crosse, E., Can. Med. Assoc.1 1994, 151, 1457 - 1464.78. Jiang,Q.;Lee, C. Y. D.; S., M.; Wilkinson, B.;Cramer, P.; Zelcer, N.; Mann, K.; Lamb,B.; Wilison, T. M.; Collins, J. L.; Richardson, J. C.;Smith, J. D.; Comery, T. A.; Riddell,D.; Holtzman, D. M.; Tontonoz, P.; Landreth, G. E., Neuron 2008,58, 681 - 693.79. Gunzburg, M. 3.; Perugini, M. A.; Howlett,G. J., I Biol. Chem. 2007, 282, 35831 -35841.80. Bertram, L.; Hiltunen, M.; Parkinson,M.; Ingelsson, M.; Lange, C.; Ramasamy, K.;Mullin, K.; Menon, R.; Sampson, A. J.; Hsiao,M. Y.; Elliott, K. J.; Velicelebi, G.;Moscarillo, T. 3.; Hyman, B. T.; Wagner,S. L.; Becker, K. D.; Blacker, D.; Tanzi, R., NEngi. I Med 2005, 352, 884 - 894.81. Strittmatter, W.; Saunders, A. J.; Schmechel, D.; Pericak,V.; Enghild, 3.; Salvesen, G. S.;Roses, A. D., Proc. Nati. Acad. Sci. 1993,90, 1977 - 1981.82. Lambert, J.-C.; Amouyer, P., Psychoneuroendocrinol. 2007,32, S62 - S70.83. Hardy, J. A.; Higgins, G. A., Science1992, 256, 184-5.84. Gong, Y. S.; Chang, L.; Viola, K. L.; Lacor, P. N.; Lambert,M. P.; Finch, C. E.; Krafft,G. A.; Klein, W. L., Proc. Natl. Acad. Sci. 2003, 100, 10417 - 10422.85. Atwood, C. S.; Scarpa, R. C.; Huang, X.; Moir, R. D.; Jones,W. D.; Fairlie, D. P.; Tanzi,R. E.; Bush, A. I.,I Neurochem. 2000, 75, 1219- 1233.86. Curtain, C. C.; Ali, F. E.; Smith, D. G.; Bush, A. I.; Masters,C. L.; Bamham, K. J., IBiol. Chem. 2003, 278, 2977-2982.87. Lovell, M. A.; Robertson, 3. D.; Teesdale, W. J.; Campbell, J. L.; Markesbery,W. R., INeurol. Sci. 1998, 158, 47-52.3888. Atwood, C. S.; Huang,X.; Moir, R. D.; Tanzi, R. E.; Bush, A. I., Met. IonsBiol. S’yst.1999, 36, 309 - 364.89. Atwood, C. S.; Perry, G.; Zeng,H.; Kato, Y.; Jones, W. D.; Ling, K.Q.;Huang, X. D.;Moir, R. D.; Wang, D. D.; Sayre,L. M.; Smith, M. A.; Chen, S. G.; Bush, A. I.,Biochemistry 2004, 43, 560 - 568.90. McLean, C. A.; Cherny, R. A.;Fraser, F. W.; Fuller, S. J.; Smith, M. J.; Beyreuther,K.;Bush, A. I.; Masters, C. L., Ann. Neurol. 1999,46, 860-866.91. Cuajungco, M. P.; Goldstein, L. E.; Nunomura,A.; Smith, M. A.; Lim, J. T.; Atwood, C.S.; Huang, X.; Farrag, Y. W.; Perry,G.; Bush, A. I., I Biol. Chem. 2000, 275, 19439 -19442.92. Noy, D.; Solomonov, I.; Sinkevich,0.; Arad, T.; Kjaer, K.; Sagi, I., I Am. Chem. Soc.2008, 130, 1376 - 1383.93. Giulian, D.; Haverkamp, L. J.; Yu, J. H.; Karshin,W.; Tom, D.; Li, J. H.; Kirkpatrick, J.;Kuo, Y.-M.; Roher, A. E., I Neurosci. 1996, 16,6021 - 6037.94. Barnham, K. J.; Masters, C. L.; Bush, A. I.,Nat. Rev. Drug Discov. 2004, 3, 205-2 14.95. Herholz, K.; Salmon, E.; Perani, D.; Baron,J.-C.; Holthoff, V.; Frolich, L.; Schonknecht,P.; Ito, K.; Mielke, R.; Kalbe, E.; Zundorf, G.; Delbeuck, X.;Pelati, 0.; Anchisi, D.;Fazio, F.; Kerrouche, N.; Desgranges, B.; Eustache, F.; Beuthien-Baumann,B.; Menzel,C.; Schroder, J.; Kato, T.; Arahata, Y.; Henze, M.; Heiss, W.-D., Neurolmage 2002, 17,302-316.96. de Leon, M. J.; Mosconi, L.; Blennow, K.; DeSanti, S.; Zinkowski,R.; Mehta, P. D.;Practico, D.; Tsui, W.; Saint Louis, L. A.; Sobanska, L.; Brys,M.; Li, Y.; Rich, K.;Rinne, J.; Rusinek, H.,Ann. NY Acad. Sd. 2007, 1097, 114- 145.97. Kasa, P.; Rakonczay, Z.; Gulya, K., Frog. Neurobiol. 1997, 52, 511- 535.98. Robinson, D. M.; Keating, G. M., Drugs 2006, 66, 1515 - 1534.99. Gilgun-Sherki, Y.; Melamed, E.; Offen, D., I Mo!. Neurosci. 2003, 21, 1 - 11.100. Sano, M.; Ernesto, C.; Thomas, R. G.; Klauber, M. R.; Schafer, K.;Grundman, M.;Woodbury, P.; Growdon, J.; Cotman, D. W.; Pfeiffer, E.; Schneider,L. S.; Thal, L. J., NEngi. I Med. 1997, 336, 1216-1222.101. Breitner,J. C. S.,Annu. Rev. Med 1996, 47, 401 -411.102. Lim, G. P.; Yang, F.; Chu, T.; Chen, P.; Beech, W.; Teter, B.; Tran, T.; Ubeda, 0.;Ashe,K. H.; Frautschy, S. A.; Cole, G. M., I Neurosci. 2000, 20, 5709- 5714.103. Viad, S. C.; Miller, D. R.; Kowall, N. W.; Felson, D. T., Neurology 2008,70, 1672 -1677.39104. Selkoe, D. J., Nature 1999,399, A23 - A31.105. Ogura, T.; Mio, K.; Hiyashi, I.;Miyashita, H.; Fukuda, R.; Kopan, R.; Kodama,T.;Hamakubo, T.; Iwatsubo, T.; Tomita, T.; Sato,C., Biochem. Biophys. Res. Commun.2006, 343, 525 - 534.106. Fleisher, A. S.; Raman, R.; Siemers, E.R.; Becerra, L.; Clark, C. M.; Dean, R. A.;Farlow, M. R.; Galvin, J. E.; Peskind,E. R.; Quinn, J. F.; Sherzai, A.; Sowell, B. B.;Aisen, P. S.; Thai, L. J.,Arch. Neurol. 2008, 65, 1031- 1038.107. Baxter, E. W.; Conway, K. A.; Kennis,L.; Bischoff F.; Mercken, M. H.; De Winter, H.L.; Reynolds, C. H.; Tounge, B. A.; Luo, C.; Scott,M. K.; Huang, Y.; Braeken, M.;Pieters, S. M. A.; Berthelot, D. J. C.; NMasure,S.; Bruinzeel, W. D.; Jordan, A. D.;Parker, M. H.; Boyd, R. E.; Qu, J.; Alexander, R. S.;Brenneman, D. E.; Reitz, A. B., IMed. Chem. 2007, 50, 4261 - 4264.108. Hussain, I.; Hawkins, J.; Harrison, D.; Hille, C.;Wayne, G.; Cutler, L.; Buck, T.; Waiter,D.; Demont, E.; Howes, C.; Naylor,A.; Jeffrey, P.; Gonzalez, M. I.; Dingwall, C.;Michel, A.; Redshaw, S.; Davis, J. B., I Neurochem. 2007,100, 802 - 809.109. Imbimbo, B. P., Drug Dev. Res. 2002, 56,150-162.110. Orgogozo, J.-M.; Gilman, S.; Dartigues, J.-F.; Laurent,B.; Puel, M.; Kirby, L. C.;Jouanny, P.; Dubois, B.; Eisner, L.; Flitman, S.; Michel, B. F.; Boada,M.; Frank, A.;Hock, C., Neurology 2003, 61, 46 - 54.111. Maier, M.; Seabrook, T. J.; Lazo, N. D.; Jiang,L.; Das, P.; Janus, C.; Lemere, C. A,, INeurosci. 2006, 26, 4717 - 4728.112. Mattson, M. P., Neurology 2003, 60, 690 - 695.113. McLachlan, D. R. C.; Dalton, A. J.; Kruck, T.P.; Bell, M. Y.; Smith, W. L.; Kalow, W.;Andrews, D. F., Lancet 1991, 337, 1304 - 1308.114. Tamura, Z.; Yoshioka, M.; Imanari, T.; Fukaya, J.; Kusaka, J.; Samejima,K., Clin. Chim.Acta 1973, 47, 13 - 20.115. Arbiser, J. L.; Kraeft, S. K.; van Leeuwen, R.; Hurwitz,S. J.; Selig, M.; Dickersin, G. R.;Flint, A.; Byers, H. R.; Chen, L. B., Mol. Med. 1998, 4, 665-670.116. Ritchie, C. W.; Bush, A. I.; Mackinnon, A.; Macfarlane, S.; Mastwyk,M.; MacGregor,L.; Kiers, L.; Chemy, R.; Li,Q.X.; Tammer, A.; Carrington, D.; Mavros, C.; Volitakis,I.; Xilinas, M.; Ames, D.; Davis, S.; Beyreuther, K.; Tanzi, R. E.; Masters,C. L., Arch.Neurol. 2003, 60, 1685-1691.117. Chemy, R. A.; Atwood, C. S.; Xilinas, M. E.; Gray, D. N.;Jones, W. D.; McLean, C. A.;Bamham, K. J.; Volitakis, I.; Fraser, F. W.; Kim, Y. S.; Huang, X.D.; Goldstein, L. E.;40Moir, R. D.; Lim, J. T.; Beyreuther,K.; Zheng, H.; Tanzi, R. E.; Masters, C. L.; Bush,A.I., Neuron 2001, 30, 665-676.118. Lee, J.-Y.; Friedman, J. E.;Angel, I.; Kozak, A.; Koh, J. Y., Neurobiol. Aging 2004,25,1315-1321.119. Ibach, B.; Haen, E.; Marienhagen,J.; Hajak, G., Pharmacopsychiatry 2005, 28, 178 -179.120. Bush, A. I., Trends Neurosci. 2003,26, 207-2 14.121. Cherny, R. A.; Barnham, K.J.; Bush, A. I.; Cappai, R.; Gautier, E. C. L.; Masters,C. L.;Carrington, D.; Kocak, G.; Volitakis,I.; Kok, G. B., Alz. Dement. 2006, 2, S646.122. Adlard, P. A.; Chemy, R.A.; Finkelstein, D. I.; Gautier, E.; Robb, E.; Cortes,M.;Volitakis, I.; Liu, X.; Smith, J. P.; Perez, K.; Laughton,K.; Li, Q.-X.; Charman, S. A.;Nicolazzo, J. A.; Wilkins, S.; Deleva,K.; Lynch, T.; Kok, G.; Ritchie, C. W.; Tanzi,R.E.; Cappai, R.; Masters, C. L.; Bamham,K. J.; Bush, A. I., Neuron 2008, 59, 43 - 55.123. Lannfelt, L.; Blennow, K.; Zetterberg,H.; Batsman, S.; Ames, D.; Harrison, J.; Masters,C.; Targum, S.; Bush, A.; Murdoch, R.; Wilson,J. E.; Ritchie, C. W., Lancet Neurol.2008, 7, 779 - 786.- 41CHAPTER 2Radioiodination of Glycosylated Pyridinones as Potential Therapeuticsfor Alzheimer’sDisease*2.1 Introduction2.1.1 Metals in Alzheimer’s DiseaseAf3 (f3-amyloid) plaques found in Alzheimer’s brains contain elevated levels of the metal ionscopper, zinc and iron; up to 400 jiM for copper, 1 mM for zinc, and 1 mM for iron.1’2hasbeen found that iron does not actually bind to the amyloid, or induce fibril formation, but isassociated with the plaques in vivo.3 This is not the case for Cu and Zn, however, and the ions ofthese metals are thought to play key roles in binding to the Af3 protein and in causing some of thesubsequent aggregation, redox activity and downstream problems. There are many metalbinding sites provided by the residues present in the Aj3 proteins, particularly the histidine andtyrosine side chains. It has been postulated that the natural role of endogenousA13is as a copperbinding and regulating protein. This is a very complicated cycle, the full details of which havenot yet been conclusively determined, but it is proposed to involve free copper triggering theproduction of more APP and Af3.4 In healthy brains this system works because A binds freecopper, but in Alzheimer’s disease (AD), both the copper and theA13are required for theobserved toxicity, so one increasing the other only serves to increase progression of the disease.Copper is found in vivo in both the monocationic and dicationic states. These two states are bothreadily accessible under biological conditions, making copper ions potential producers of freeradicals. For this reason the body has many control mechanisms for copper, and in a normallyfunctioning organism nearly all the copper present is protein bound so it is either being utilizedor stabilized in one oxidation state. As these control mechanisms become worn with age andwork less efficiently, small amounts of copper ions are left free in vivo, where they will react to‘Parts of this chapter have been published: Schugar, H. J.; Green, D. E.; Bowen, M. L.; Scott, L. E.; Storr, T.;Böhmerle, K.; Thomas, F.; Allen, D. D.; Lockman, P. R.; Merkel, M.; Thompson, K. H.; Orvig, C. CombatingAlzheimer’s Disease With Multifunctional Molecules Designed for Metal Passivation. Angew. Chem. mt. EL2007, 46, 1716- 1718.References begin on Pg 61 42give or take electrons from whatever source they find themselves near. This is proposed to be amechanism at play early on in the aberrant metal binding and oxidative damage seen in AD.5Oxidative damage is caused by radicals, very reactive species that react non-selectively withwhatever is in their proximity. In this way essential molecules such as proteins, cell membranesand DNA are damaged, which leads to the inflammatory response being triggered, andeventually to cell death. Reactive oxygen species (ROS) do not form without the presence ofmetal ions. Amyloid is found not to be toxic to cells without the associated copper or iron.6 It isnot yet known if amyloid oligomerises in vivo in the absence of metal ions. Thus an interventionthat has received a considerable amount of attention recently is the use of metal chelators.Proponents of the MPAC (metal protein attenuating compound) approach to treating AD havepreviously used an FDA-approved agent clioquinol (CQ) as a proof of principle drug.7 CQ, anda similar compound, known as PBT2, are 8-hydroxyquinolines, and bind in a bidentate fashionwith high affinity to a range of metal ions.8 These compounds are able to slow the progressionof dementia in Tg2576, AD type transgenic mice.7’9 They are also able to affect the levels of Af3found in the plasma of human subjects with AD,’°’ “though it is not yet known if they will beeffective at alleviating symptoms in people over the long term. As compounds of these typesonly treat one specific point in the complex cascade that leads to the effects of AD, currentresearch often tries to build on the basic promise seen in approaches like this by adding multiplefunctionalities into one molecule to better treat the many factors occurring simultaneously inAD.2.1.2 The Multifunctional Approach to Treating Alzheimer’s DiseaseAs AD is such a multi-faceted and overall poorly understood disease, it is widely acknowledgedthat a combination of two or more types of therapy is much more likely to form a successfultreatment than any one could alone, as has been found to be the case with other complex diseaseslike cancer and HIV.’2’13The investigation of multifunctional compounds to treat AD has beenreviewed recently,’4in a dense 26 page article, indicating the recent enthusiasm for therapies ofthis nature. Some examples of proposed multifunctional compounds are shown in Figure 2.1.The benefits of combining different functionalities into one molecule include alleviating the needReferences begin on Pg 61 43for multiple pills and the increased complications that come from drugs interacting with eachother in vivo.14 Cavalli et al. encourage the use of the term multi-target-directed ligands(MTDLs) for compounds that are designed to interact with multiple targets, and they stress theimportance of this approach in complex multi-faceted diseases like AD.’4It is interesting to notethat the number of patents for multiple compound medications has overtaken those for singlemolecular entities for the treatment of AD.’4baH0oH HCOOHCOOH COOHfFigure 2.1 Multi-target-directed-ligands under research as AD therapeutics: a) an AChEI thatalso stops Af3 aggregation b) an AChEI with antioxidant potential c) Memoquin has fourdifferent anti-AD functionalities d) and e) combine metal chelating abilities with antioxidantcapacity f) a metal chelator that also targets Af3 aggregates.As three of the four AD therapeutics currently on the market in Canada are acetylcholinesteraseinhibitors (AChEI) (see Chapter 1), most of the work towards multifunctional compoundsincorporates this functionality. Two examples of molecules of this kind are shown in Figure2.1 a and b. Figure 2.1 a shows a molecule that was designed to be an AChEI that can also inhibittBuReferences begin on Pg 61 44the self-aggregation ofA13.’5Although it was found to be successful in these two functions withan 1C50 of 0.18 tM against AChE and the ability to prevent 60 % of the Af3 self aggregation at aconcentration of 100 jiM, the charge on the compound will likely limit its BBB permeability andtherefore its therapeutic potential. It is still a contentious issue as to whether the self aggregationof Af3 is problematic in the absence of metal ions; if this is determined not to be the case it wouldlimit the use of this kind of a therapeutic agent. Figure 2.lb shows a compound that combinesAChEI activity with antioxidant potential.’6 The compound has a very high affinity for AChEwith an 1C50 of 8 pM, and also shows good antioxidant capacity with a TEAC (TroloxEquivalent Antioxidant Capacity) value of 2.5, where Trolox is a water soluble vitamin Eanalogue normalized to a TEAC value of one. The authors predict it will cross the BBB,although no experimental evidence of this was presented.Figure 2. ic shows Memoquin, a molecule with four different AD functionalities built in.17 It is anM inhibitor of AChE, shows antioxidant properties, and can inhibit the aggregation ofA13,either when left to self-aggregate or when induced to do so by AChE. Finally, Memoquin canact as a BACE- 1(13secretase — see Chapter 1.5) inhibitor to reduce the amount of Af3 produced.These properties were tested in vitro, and then in vivo in mouse AD models, with verypromising, and multi-faceted results. Memoquin prevented neurodegeneration, and reversed orprevented (depending on the advancement of the disease state) choline deficit, thyperphosphorylation and behavioral / memory problems. This is a very promising strategy, andshows the potential benefits of the MTDL approach to AD.Other groups, including our own, have been interested in the potential of combining metalchelating ability with other relevant functionalities. Figure 2.1 d shows a compound proposed tobe useful for AD as it combines metal chelation and antioxidant capabilities.’8Interestingly, thiscompound was proposed by theoretical chemists, though it has not been practically tested as ofthis writing. The reasons they chose this type of compound is because of its high affinity forcopper, and ability to still act as an antioxidant (electron acceptor) once bound to a copper ion.They also note that these compounds should be easy to modify to add further functionality suchas A13 aggregation inhibitors, and that they are expected to be BBB permeable and non-toxic,according to their calculations.References begin on Pg 61 45There has long been interest in pyridinones as potential therapeutics fora range of conditions.19-21Bebbington et al. made a series of N-substituted pyridinones and tested them as antioxidantswith an interest in AD among other applications.22 The most promising candidate of this study isshown in Figure 2. le.22 This compound showed synergistic properties in that the antioxidantcapacity observed was greater than that for the administration of the two parts of the compoundseparately: the pyridinone and the radical scavenger substituted phenol. This is veryencouraging for the line of compounds our group has been functionalizing.XH1 (Figure 2.lf) contains both a metal chelating and an amyloid binding portion in onemolecule.23 It was found to decrease Af3 aggregation in vitro, and to reduce APP expression inhuman neuronal cells. XH1 was found to reduceA13140concentrations and plaque formation inAD expressing transgenic mice after four weeks of oral treatment.23 From these results it wasproposed that the compound crosses the BBB, though no experiments were done to determinethis.2.1.3 Our Approach to Multi-Target-Directed-Ligands for Alzheimer’s DiseaseIn the Orvig group, we are interested in designing MTDLs based on the bidentate chelatingpyridinones shown in Figure 2.2. We have chosen to funstionalize them in several ways toimprove their usefulness as AD therapeutics. The pyridinone nitrogen can be substitutedrelatively easily, and the two functionalities we have successfully employed in this position areantioxidant24and Af3 plaque targeting moieties.25 We have also chosen to form the glycosylatedprodrugs of the metal chelators, which we propose will confer several benefits.24 The glucosemoiety increases the water solubility of the compounds. While the glucose is in place, thecompounds will not chelate metal ions. This is important because the 3-hydroxy-4-pyridinonesare fairly high affinity chelators, especially for iron, so by adding a glucose molecule to maskthis chelating functionality, we hope to prevent systemic chelation of essential elements in thebody. We have shown that the glycosyl bond can be cleaved enzymatically to reveal the activechelator,24 and we propose this may occur in vivo. Finally, the glucose moiety may increaseBBB permeability of our compounds, an advantageous property perhaps available via facilitatedtransport into the brain on the GLUT-i glucose transporters. These transporters are found inReferences begin on Pg 61 46high concentration at the BBB, and as such may increasethe amount of our compounds that getinto the brain as compared to entering by passive diffusionin the non-glycosylated form.deglycosylationHoXchelationR = H, Me, Et, phenyl, 4-hydroxyphenyl etc.Figure 2.2 3-Hydroxy-4-pyridinones synthesized in the Orvig group.24A representative rangeof R groups are shown, but many more have been made.It can be seen that once deglycosylatedthe free pyridinone is able to act as a bidentate ligand.The aim of the work in this thesis chapter was to createa radiolabelled pyridinone for use inassessing the blood brain barrier permeability of these glycosylated prodrugsin vivo. The twoiodinated compounds prepared in this work are shown in Figure 2.3.OHbaHOOH-N7pyrA (5)Figure 2.3 The two iodinated compounds in this chapter: a) pyrA (5) , madeby myself b) pyrB,first made by myself and Dr. Michael Merkel; the non-radioactive I123jand125janalogues wereprepared for each.pyrBReferences begin on Pg 61 472.2 Experimental2.2.1 Instruments and MaterialsAll solvents and reagents were used as received. Reagents were purchased from Sigma-Aldrichunless otherwise stated. Maltol (3-hydroxy-2-methyl-4-pyrone) was purchased from CultorFood Sciences, New York. Solvents were HPLC grade, and were purchased from FisherScientific. Toluene was dried over activated 4A molecular sieves for at least 48 hr prior to use.Na[’2311 was generously supplied by MDS Nordion. Na[’251] was a gift from Dr. Jianming LuofTRIUMF, Vancouver.The tributylstannyl derivative of compound pyrB shown in Figure 2.3b (2-methyl-3-(2,3,4,6-tetra-O-acetyl-f3-D-glucopyranosyloxy)- 1- [4-(4-tributylstannylphenyl)-2-thiazolyl]-4pyridinone)was made by Dr. Michael Merkel. The radiolabelling of this compound was carried out bymyself and Dr. Merkel in a manner similar to that used for the radioiodination of compoundpyrA.The analytical TLC plates, which were aluminum backed ultra pure silica gel 60, 250 Jim, andthe flash column silica gel (standard grade, 60 A, 32-63 mm) used were provided by Silicycle.Sep-paks were provided by Supelco. ‘H and ‘3C NMR, ‘3C NMR APT, 2D ‘H-’H COSY and‘H-’3C HMQC spectra were recorded at ambient temperature on Bruker AV300, AV400 orDRX400 instruments. The NMR spectra are expressed on the ö scale and were referenced to theresidual solvent peaks. Infrared spectra were recorded on a Nicolet 6700 FT-JRspectrophotometer in transmission mode between 400 and 4000 cm1 at a resolution of ± 0.09cm1. ESI mass spectra were recorded on a Micromass LCT instrument. High resolution massspectra (Micromass LCT TOFMS) and elemental analysis (Carlo Erba EA 1108 ElementalAnalyzer) were provided by the Analytical Services Facility, Department of Chemistry,University of British Columbia. Radioactive TLC analysis was done using a Bioscan System200 Imaging Scanner fitted with a Bioscan 1000 Autochanger.References begin on Pg 6] 48In vivo rat brain perfusion experiments were carriedout by Dr. David Allen, Dr. Paul Lockmanand Fancy Thomas at Texas Tech University HealthSciences Centre, using the proceduredetailed in Appendix 1.2.2.2 Chemical Synthesis1-(4-Bromophenyl)-3-hydroxy-2-methyl-4-pyridinone (1)9Maltol (2-methyl-3-hydroxy-4-pyrone) (10.75g, 85.25 mmol) and 4-H342bromoaniline (29.21 g, 170.0 mmol) were placed in a round bottom flask with 806mL methanol and 200 mL dilute hydrochloric acid (5 mL 12 M HC1 in water).8The resulting mixture was heated, dissolving to give an orange solution, and1Or refluxed for 48 hr. The contents of the flask were filtered to give a dark pinkprecipitate which was recrystallised from hot methanol to give a pale pink product(12.56 g, 53 % yield). ‘H NMR (300 MHz, DM50—cl6,6): 7.75 (d, 3J8,9 = 8.1 Hz, 2H; H8), 7.54(d, 3J45 = 7.1 Hz, 1H; H4), 7.43 (d, 3J98 = 8.1 Hz, 2H; H9), 6.20 (d, 3J54 = 7.1 Hz, 1H; H5), 1.95(s, 3H; H6). HR-MS (ES+ of MNaj: mlz calcd for C,2H,079BrNONa : 301.9793, found:301.9784. Anal. Calcd. (found) forC12H1OBrNO2:C, 51.45 (51.82); H, 3.57 (3.81); N, 5.00(5.29).1-(4-Bromophenyl)- 2-methyl -3-(2,3,4,6-tetra-O-acetyI--D-g1ucopyranosyIoxy)-4-pyridinone (2)OAc1 (6.25 g, 22.3 mmol), acetobromo-cL-D-glucose (3.35g, 8.15AcO°\1 mmol) and tetrabutylammonium bromide (2.66 g, 8.24 mmol)2were suspended in 35 mL CH2C1 and heated to 35 °C. NaOH (35mL, 1 M) was added and the resulting heterogeneous mixture1415L1stirred vigorously for 3 hr, during which time it turned dark red.1 6The reaction was cooled to room temperature, and EtOAc (50 mL)added. The organic layer was separated and washed with 3 x 50mL 1M NaOH, 50 mL water then 50 mL saturated brine. The organic layer was then dried overMgSO4,the drying agent filtered off, and the resulting liquid evacuated on a rotary evaporator.Column chromatography on silica gel using EtOAc:MeOH (9:1) gave a white solid (1.12g, 23% yield). ‘H NMR (400 MHz, MeOH-d4,6): 7.76 (d,3Jj,j4= 8.8 Hz, 2H, H15), 7.67 (d, 3J11,jo =References begin on Pg 61 497.5 Hz, 1H, Hi 1), 7.37 (d,3J1415 8.5 Hz, 2H, H14), 6.49 (d, 3J10,jj = 7.5 Hz, 1H, H10), 5.57 (d,= 7.9 Hz, 1H, Hi), 5.34 (dd, 3J43 = 9.6 Hz, 3J45 = 9.6 Hz, 1H, H4), 5.10 (m, 2H, H2, H3),4.26 (dd,3J6a5= 4.6 Hz,3J6ab12.3 Hz, 1H, H6a) 4.14 (dd,3J6b,5= 2.4 Hz,3J6b,a= 12.5 Hz,1H, H6b), 3.81 (ddd, 3J5,4 = 9.6 Hz,3J56a4.5 Hz,3J5,6b= 2.4 Hz, 1H, H5), 2.12 (s, 3H, H12),2.10 (s, 3H, COCH3),2.01 (s, 3H, COCH3),2.00 (s, 3H, COCH3),1.98 (s, 3H, COCH3).HR-MS(ES+ of MNa): mlz calcd forC26H2879BrNO11Na : 632.0743, found: 632.0736. Anal. Calcd.(found) forC26H28BrNO11:C, 51.16 (51.49); H, 4.62 (4.75); N, 2.29 (2.66).2-Methy1-3-(2,3,4,6-tetra-O-acety1--D-gIucopyranosy1oxy)-1-(4-tributy1stanny1pheny1) -4-pyridinone (3)OAc2 (0.52 g, 0.84 mmol) andAcO’\10tetrakis(triphenylphosphine)palladium(0) (0.10 g, 0.08 8 mmol)2OAcwere weighed into a two necked RBF under an inert atmosphere.127Dry toluene (15 mL) and hexabutylditin (2. i mL, 4.2 mmol) were14added via syringe, and the resulting solution brought to reflux1516(under Ar. After 16 hr at reflux the solution was cooled to room(Sn .3temperature and filtered through Celite with thorough washing withCH2C1.The resulting solution was evaporated to dryness on a rotary evaporator, dissolved in10 mL acetonitrile and washed three times with 30 mL hexanes to remove any remainingbutyltin species. The acetonitrile layer was dried and then purified by silica gel columnchromatography using 2 % MeOH in EtOAc as eluent. A colourless oil was produced thatbecame a clear white glassy solid after vacuum drying (0.16 g, 38 % yield). ‘H NIvIR (400 MHz,DMSO—d6,6): 7.64 (d,3Jii,o= 7.5 Hz, 1H, H11), 7.62 (d, 3J15,14 = 8.1 Hz, 2H, H15), 7.38 (d,= 8.2 Hz, 2H, H14), 6.22 (d,3J10= 7.6 Hz, iH, H10), 5.42 (d, 3J1,2 = 7.9 Hz, 1H, Hi),5.37 (dd, 3J4,= 9.7 Hz, 3J4,5 = 9.6 Hz, 1H, H4), 4.96 (dd, 3J2,, 7.9 Hz, 3J2,= 9.9 Hz, 1H, 112),4.93 (dd, 3J,2 = 9.8 Hz, 3J,4 = 9.6 Hz, 1H, H3), 4.13 (dd,3J6a,s= 5.2 Hz,3J6a,i= 12.4 Hz, iH,H6a), 4.02 (dd,3J6b,5= 2.3 Hz,3J6b,óa= 12.2 Hz, 1H, H6b), 3.96 (ddd, 3J5,4 = 9.7 Hz,3J5,6a= 4.9Hz,3J5,6b= 2.4 Hz, iH, H5), 2.04 (s, 3H, H12), 1.99 (s, 3H, COCH3), 1.97 (s, 3H, COCH3)1.93 (s, 6H, COCH3),1.52 (m, 6H, SnCH2CH3),1.29 (m, 6H, SnCH2CH3),1.09(m, 6H, SnCH2CH),0.85 (m, 9H, SnC3H6CH).HR-MS (ES+ of MNa): mlz calcd forC38H55N011’‘8SnNa: 842.2689, found: 842.2680.References begin on Pg 61 501-(4-Iodopheny1)-2-methy1-3-(2,3,4,6-tetra-O-acety1--D-gIucopyranosyIoxy)-4-pyridinone(4)QAc3 (0.060 g, 0.073 mmol) was dissolved in 5 mL of MeOH.AcO\jChloramine-T (8.3 mg, 0.037 mmol) was added, followed by NaT2OAcJ[j10(5.6 mg, 0.037 mmol). The reaction mixture was stirred for 30_11121,Lmins. It was then quenched with an aqueous solution of NaHSO3[‘(0.1 M) until the yellow coloured reaction mixture became1516Jcolourless (3.1 mg, 64 % yield). ‘H NMR (300 MHz, MeOH-d4,):7.95 (d, 3J11,jo = 7.4 Hz, 1H, H1i), 7.68 (d, 3J15,14 = 8.5 Hz, 2H,H15), 7.34 (d, 3J14j= 8.4 Hz, 2H, H14), 6.52 (d, 3J1011 = 7.4 Hz, 1H, H10), 5.56 (d, 3J12 = 7.8Hz, 1H, Hi), 5.33 (dd, 3J43 9.5 Hz, 3.J4,5 = 9.6 Hz, 1H, H4), 5.09 (m, 2H, H2,3), 4.27 (m, 1H,H6a), 4.14 (m, 1H, H6b), 3.80 (m, 1H, H5), 2.15 (s, 3H, H12) 2.12, 2.10, 2.01, 1.99 (s, 3H,COCH3). ESI-MS (ES+ of MH): m!z calcd forC26H291N011 : 658.0, found: 658.0.3-(-D-G1ucopyranosyIoxy)-1-(4-iodopheny1)-2-methyI-4-pyridinone (5, pyrA)OH 4 (3.1 mg, 0.047 mmol) was dissolved in MeOH, and an excess ofHO’180910NaOMe (5.4 mg, 0.1 mmol) added. The reaction mixture was3Istirred for 10 mm, and then quenched with Amberlite. The solution12N14was filtered and evacuated on a rotary evaporator to give crude6product. This was purified by semipreparative HPLC (1 mg, 43 %yield). ‘H NMR (400 MHz, MeOH-d4,)7.95 (d, 3J11,jo 7.4 Hz,1H, H11), 7.76 (d,3J15,14=8.5 Hz, 2H, HiS), 7.34 (d, 3J14,15 = 8.5 Hz, 2H, H14), 6.55 (d, 3J10,jj =7.4 Hz, 1H, H10), 4.73 (d, 3J1,2 = 7.4 Hz, 1H, Hi), 4.84 (m, 1H, H4), 3.65 (m, 2H, H2,3), 3.43(m, 3H, H5,6), 2.26 (s, 3H, H]2). ESI-MS (ES+ of MH): m!z calcd forC18H2,1N07: 490.0,found: 489.9.3-(-D-G1ucopyranosy1oxy)-1-(4-[125/12311iodophenyl)-2-methyl-4-pyridinone(1251-pyrA or1231-pyrA)To a sample of Na{’251](0.5 mCi) or Na[’231](‘—S2 mCi) was added 2.5 iL of 0.05 N H3P04,and the resulting neutralised solution then added to an ethanolic solution of the tributyistannylprecursor 3 (1.3 mg) in 200 iL of ethanol. A freshly prepared aqueous solution of Chloramine-T(1 mg in 100 iL) was added to the reaction mixture that was then stirred at RT. After 20 mm theReferences begin on Pg 61 51OHreaction was quenched by the addition of sodium chloride solution(100 tL; 20 % conc) and the radioiodination product (4) was2OH810extracted with ethyl acetate (x 2). The solvent was removed from121the combined organic layers under a stream of air and the residue14dissolved in 200 1iL methanol for the subsequent deprotection ofthe glucose moiety. Sodium methoxide (1 mg in 50 jiL ofmethanol/water (1:1)) was added and the reaction mixture stirredfor 15 mm. After addition of Amberlite resin (H-form) and stirring for another 7 mm at RT, thesolution was diluted with 5 mL of water and passed through a Sep-Pak cartridge (6 mL size, C-18 silica). The cartridge was washed with water (5 mL) and finally the product eluted with2.5 mL of methanol. According to TLC analysis the radiolabelled product was obtained in aradiochemical purity of 97 %. Its identity was confirmed by comparing the R.values (0.25; onsilica using ethyl acetate/methanol 4:1) of the labelled and cold analogues of 5.2.2.3 GLUT-i Cell Uptake AssayThese experiments were conducted by me using LCC6-HER2 cells — a human breast cancer cellline chosen for its overexpression of the glucose transporter GLUT-i26LCC6-HER2 cells wereplated with Dulbecco’s Modified Eagle Medium (D-MEM) (lx), liquid (high glucose)supplemented with 10 % Fetal Bovine Serum and 1 % PenicillinlStreptomycin. The cells wereallowed to adhere at 37 °C in a humidified atmosphere containing 5 % CO2.The plating wascarried out in 75 cm2 tissue culture flasks with 0.2 tm vented caps. The cultures were maintainedin a humidified 5 % CO2 atmosphere, with medium changes every alternate day. Subculturingwas carried out every 3-4 days using Trypsin-EDTA lX (0.25 % Trypsin with Na4EDTA),incubated for about 5 mm at 37 °C for cell detachment. A hemacytometer was used for countingthe cells to monitor cell proliferation.On the experiment day, lxl06 cells/mL concentration of cells was prepared in 1 % PBS (pH 7.4)and aliquoted into 1.5 mL Eppendorf vials to get a final volume of 0.5 mL in each vial (5x105cells). The compound to be tested was then added to the cells and the vial gently inverted. The123jcomplex (15 tCi) and‘8F-FDG (30 jiCi) were used in 500 jtL PBS solution. Eachcompound was added to the cells in glucose-free conditions, and in a final concentration of 5References begin on Pg 61 52mM D-glucose. The resulting suspensions were incubatedat 37 °C for 30 mm with shaking at400 RPM (revolutions per minute). Following incubationthe vials were centrifuged at 1100RPM for 5 mm then 900 jiL of the supernatant was removed. Cold PBS solution(900 jiL) wasadded, and gentle mixing of the cells into the solution was achieved by gentle uptake and releaseof the solution from the pipette tip three times. This centrifuging and washing procedure wasrepeated four times in total to give the original supernatant and three washing supematants.Finally the cells were vortexed in cold PBS (1 mL) to remove them from the vial wall andtransferred to a tube for gamma counting.The activity in each cell sample was divided by the sum of all the supernatants forthat sample togive a percentage of the original amount of activity that ended up associated with the cells. Allexperiments were carried out in quadruplicate. FDG was used as a positive control, and wasrepeated each day that these experiments were carried out to ensure experimental integrity.2.3 Results and Discussion2.3.1 SynthesisThe synthetic route used to make the radioiodinated, glycosylated pyridinone pyrA isshown inScheme 2.1. Synthesis of the para-bromophenyl pyridinone from maltol via an acid catalysedamine insertion and condensation proceeded in adequate yield. This is comparable to thatobserved for the insertion reaction of other aromatic amines to maltol.27Many methods for the glycosylation of the hydroxyl group were investigated. Initially theglucose was installed as the tetrabenzyl protected species. This was made using the a-bromotetrabenzylated sugar starting material with BF3.OEt2 as a Lewis acid catalyst for theSN2substitution; however, this benzylated substrate was not suitable for use in radiosynthesis, as theremoval of the benzyl groups proved problematic. As radioactivity decays over time, tomaximize the useful yield of a reaction sequence, reactions must proceed quickly. Removal ofthe four benzyl groups via hydrogenation does not proceed quickly under ambient pressures ofhydrogen, and the use of high pressure equipment was not feasible in this case. Removal of thebenzyl groups of the bromopyridinone with HBr resulted in loss of the bromine from the phenylReferences begin on Pg 6] 53ring. As the C-I bond is weaker than the C-Br bond, it was assumed that the iodine would alsobe lost under these reaction conditions.OAcAcV\AcO--\—.a-- O- -OAc2BrAcetyl protecting groups were used, as these can be quickly and reliably removed underconditions not expected to affect the C-I bond. Several methods were investigated for theaddition of the tetraacetyl protected glucose. The Mitsunobu reaction and Koenigs-Knorrreaction were used (starting from the sugar with a free hydroxyl at C-i), as well as Lewis acidcatalysis with BF3. OEt2 and Bronsted base catalysis with hydroxide (the latter two starting withthe cL-bromo sugar). Despite finding some success with each of these reactions, it was a phasetransfer reaction that became the synthetic method of choice. This involves a biphasic reactionmixture consisting of methylene chloride and an aqueous solution of 1 M NaOH. The twopartners to be coupled were added to this mixture along with tetrabutylammonium bromidewhich behaves as a phase transfer catalyst; the reaction mixture was heated at 30 °C and stirredvigorously for three hours. After workup and column purification, only modest yields areb0HOiJaOAcAcOoJ3ySnBuHOkdCeOHOHpyrA, 54Scheme 2.1 Reaction scheme for synthesis described in this chapter: a) HC1,H20/MeOH, reflux,48 hr b) acetobromo-cL-D-glucose, NBu4Br, CH2C1/1 M NaOH, 35 °C, 3 hr c) Sn2Bu6,Pd(PPh3)4,toluene, reflux, 16 hr d) NaT(123/125/127J)Chloramine-T, MeOH, 30 mm e) NaOMe,MeOH, 1 hr.References begin on Pg 61 54achieved, but this reaction is favourable in itsshort reaction time, and ease of productpurification.A Stille reaction was then performed to obtain a relatively weak C-Sn bond that is known to bereadily cleaved by an I electrophile.28 The chemistry of this typeof palladium cross-couplingreaction is fairly well understood,29 and standard reaction conditions were used to obtainadequate yields of the tributyltin substituted pyridinone. As an excess of tin reagent was used,an acetonitrile:hexanes extraction was very helpful in removing some of the excess butyltinspecies prior to column chromatography, where they tended to streak. Substitution of the tinwith iodine proceeded smoothly, regardless of iodine isotope used. This reaction is performedusing Nal and an oxidant which oxidises the F to I, which is believed to be the active species.In this work the oxidant that gave the best results was Chloramine-T. Removal of the acetylprotecting groups proceeded in a straightforward manner using NaOMe in MeOH in anessentially quantitative yield.When performing radioiodination the same chemical reagents were utilized, the only differencebeing the isotope of iodine used:123jfor preliminary work and1251for the animal studies.1231has a half life of 13 hr, decaying quickly enough that it can be conveniently left in shieldedstorage in the radiochemistry lab. It is a very useful tool for optimization of chemistry,purification, analysis and storage conditions. Once these systems had been optimized, thechemistry was transferred to the longer lived125jisotope, ti,2 60 days. As expected, thechemistry proceeded in an analogous manner. The final125jlabelled product was shipped to ourcollaborators in Texas for the in vivo studies. A sample was retained and kept under the sameconditions (in ethanolic solution at room temperature) and tested to ensure stability over the timeframe required for animal studies. The sample did not degrade appreciably over the several daysduring which it was tested. A representative TLC trace is shown in Figure 2.4.References begin on Pg 61 555500:5000-i/4500:/\4000’3500-/30002500-)\20001500—/1000./1r•rj’0.0 ‘10.0 20.0 30.040.0 50.0 60.0 70.0 80.0 90.0mmFigure 2.4 Radio-TLC trace of final iodinated pyridinone pyrA, Rf = 0.24 — 97.8% pure. Theline at 10 mm represents the baseline, and that at 70 mm shows the solvent front. Units on thelefthand side are counts of activity.2.3.2 In Vivo StudiesIn vivo rat brain perfusion experiments were carried out by Dr. David Allen, Dr. Paul Lockmanand Fancy Thomas at Texas Tech University Health Sciences Centre. The key finding from theperspective of the advancement of our Alzheimer’ s project is that these glycosylated prodrugs doindeed cross the blood brain barrier.This type of brain perfusion experiment is very fast, with only 60 sec available for BBBpenetration. The reason that such a short timeframe is used is to render any outflow from thebrain negligible. It can then be assumed that the entire amount of compound taken into the brainin that 60 sec is still in there at the end of the measured period, greatly simplifying theinterpretation of the results. Choline was used as a control in this experiment, as it is known tocross the BBB.3°-aReferences begin on Pg 61 56IC’)-jCl)000-JOctanol/water Coefficient (CIogD)Figure 2.5 Representation of BBB permeability of pyrA and pyrB shown with variouscompounds of known permeability for context. ClogD is the calculated octanol/water partitioncoefficient, a measure of lipophilicity, and BBB PS is a measure of the brain permeability ofeach compound.The results of this experiment are shown graphically in Figure 2.5, where they are contextualizedwith some compounds of known BBB permeability. The diagonal line running through thegraph corresponds to the expected uptake from passive diffusion. Sucrose crosses the BBB veryslowly, and is commonly used as a negative control in experiments of this type, where itspresence inside the brain would show either a compromised BBB or poor experimentaltechnique. Caffeine and nicotine are very commonly used drugs, with their rapid effectsstemming from the ease with which they enter the brain. Nicotine is a very small molecule (MW162 g.mol’) which crosses the BBB via passive diffusion and is therefore situated on thediagonal trend line in the graph above. Caffeine (MW 194 g.mor’), while also small, is aDiazepam—ICaffeinee pyrAApyrBASucrose-2 0 2 4 6References begin on Pg 61 57structural analogue of the purine base adenosine, andas such cross the BBB via both passivediffusion and carrier mediated transport.3’This is whycaffeine sits above the trend line on thegraph above — it is taken up into the brain faster thanif it were transported by passive diffusionalone.The BBB uptake of pyrA was found to be 6.4x1 ü ± 4.2x1 mL/sec/g andthat of pyrB wasfound to be 8 .9x 1 0 ± 4. lxi mL/sec/g. In Figure 2.5 both pyrA and pyrB arereasonably faraway from the passive diffusion trend line. As they are below this line, thebrain uptake rate thatthey experience is lower than that expected from passivediffusion. The timescale of the BBBperfusion experiment used in this study is very short (60 sec), meaning efflux is highly unlikely.This indicates that the compounds either undergo passive diffusion very slowly,or that there isanother mechanism at play. However the compounds are getting into the brain,they are doing itrelatively slowly.Initially it was thought that the mechanism of the two compounds’ entry into thebrain was bytransport on the glucose transporter GLUT-i. This is one of a family of GLUTs that facilitatetransport of sugars through various membranes, as detailed in Chapter 1. TheGLUT-i isoformis known to be expressed in high concentrations at the blood brain barrier and in erythrocytes.Because of this, and the fact that that there are sugars present in these molecules, it is possiblethat the compounds entered the brain via the GLUT-i transporters.It is unlikely that thesecompounds, with their deprotected sugars, would be lipophilic enough to cross into the brain viapassive diffusion. As the compounds tested were quite different to glucose, it is plausible thatany GLUT-mediated transport would occur at a considerably slower rate than thatof nativeglucose.2.3.3 In Vitro StudiesIn vitro cell studies were done to help verify the in vivo results, and alsoto deduce a little moreabout the mechanism of the brain uptake observed in the rats.The in vitro cell uptake studies were performed in LCC6-HER2 cells — a human breast cancercell line chosen for its overexpression of the glucose transporter GLUT-i26The cellular uptakeReferences begin on Pg 61 58of an123jlabelled compound - pyrB - was examined under both glucose free and high glucose(10 mM) conditions, and was found to be 0.10 ± 0.02 % and 0.12 ± 0.02 % of the total radiationrespectively.Any changes in amount of radioactive species seen in the cells under these two conditions wereattributed to the presence of glucose, as all other conditions were kept constant. If the testcompound was utilizing the same transport mechanism as the glucose, the large excess ofglucose molecules present would occupy and be transported by GLUT-i such that lesstransporters would be free to move the test compound. The presence of the glucose does notmake a significant difference to the amount of test compound getting into the cells, implying thatthe compound tested does not enter the cells via the GLUT-i transporters.A limitation of this assay is that it does not distinguish between the radioactive compound beinginside the cell or just being embedded in the membrane. However, the fact that these compoundsget into the brain leads us to believe that they also getting into the cells, as the BBB is generallyless permeable than are cell membranes. The mechanism for this brain entry is still not known.Possibilities are passive diffusion, though is seems unlikely as outlined above, or entry throughanother transporter such as the Nat’K pump, as has been observed with other compounds.32The amount of activity associated with the cells after a 30 mm incubation is very small. This isin keeping with the findings of the rat brain perfusion studies which found that the samecompound was able to cross the blood brain barrier, though only in small amounts in the giventime frame. Although the uptake observed in both these studies is not particularly fast, it islikely sufficient for the proposed purpose of compounds of this type as AD therapeutics. Ifglycosylated pyridinones of this type were found to be useful in the treatment of AD, this rate ofBBB permeation would likely be adequate to elicit a therapeutic effect, as once the compoundswere administered, they would have much longer than 60 sec in which to cross the BBB.This raises the possibility of administering the non-glycosylated version of the compounds, asthese may permeate key membranes more quickly. As mentioned above, there were severalproposed benefits of appending the glucose, and the water solubility and protection of thechelating moiety are both important features that may be compromised if the glucose pro-drugReferences begin on Pg 61 59form of the molecule is not used. From the studies performed to date we cannot be sure that theglycosidic bond remained intact during either of these assays. A sample of iodinated compoundwas retained and tested for degradation after 3 days, and it was found to be intact. However,once inside an animal or interacting with cells, we cannot be sure that the glucose remainedbound to the pyridinone. More work needs to be done in this area to ensure the integrity of thisbond under more biologically relevant conditions, and to assess the BBB permeability of theglycosylated vs. non-glycosylated compounds.2.4 ConclusionThis was a critical set of experiments in order to determine the potential utility of this approachin treating AD. In this work two glucose-protected 3-hydroxy-4-pyridinones were successfullyradioiodinated. The radiolabelled compounds were assessed using an in vivo brain perfusionstudy in rats, and were found to cross the BBB. If these glycosylated pyridinones were not BBBpermeable there would be little point in continuing this line of investigation and the currentcompounds would require significant modification to improve their in vivo properties. Althoughthe brain uptakes observed were relatively low compared to those seen in native brain substrates,this approach does not necessarily require very fast brain uptake. As these glycosylatedpyridinones have been found to be non-toxic to human cells,24 it is possible that a larger dosecould be administered and slow BBB permeation could lead to a constant supply of deprotectedprodrug in the brain. It is not known exactly how these many factors will interplay, but moreand more information is being gathered as work on this project is continuing in the Orvig group.The aim of this particular study was met with a successful determination of BBB permeation atwhat we deemed to be a great enough rate to continue this work.References begin on Pg 61 602.5 References1. Lovell, M. A.; Robertson, J. D.; Teesdale, W. J.; Campbell, J. L.; Markesbery, W. R., INeurol. Sci. 1998, 158, 47-52.2. Smith, M. A.; Harris, P. L. R.; Sayre, L. M.; Perry, G., Proc. Nat!. Acad Sci. U S. A.1997, 94, 9866-9868.3. Dong, J.; Atwood, C. S.; Anderson, V. E.; Siedlak, S. L.; Smith, M. A.; Perry, G.; Carey,P. R., Biochemistry 2003, 42, 2768-2773.4. Armendariz, A. D.; Gonzalez, M.; Loguinov, A. V.; Vulpe, C. D., Physiol. Genomics2004, 20, 45 - 54.5. Bush, A. I., Trends Neurosci. 2003, 26, 207-2 14.6. Rottkamp, C. A.; Raina, A. K.; Zhu, X.; Gaier, E.; Bush, A. I.; Atwood, C. S.; Chevion,M.; Perry, G.; Smith, M. A., Free Radic. Biol. Med. 2001, 30, 447-450.7. Cherny, R. A.; Atwood, C. S.; Xilinas, M. E.; Gray, D. N.; Jones, W. D.; McLean, C. A.;Barnham, K. J.; Volitakis, I.; Fraser, F. W.; Kim, Y. S.; Huang, X. D.; Goldstein, L. E.;Moir, R. D.; Lim, J. T.; Beyreuther, K.; Zheng, H.; Tanzi, R. E.; Masters, C. L.; Bush, A.I., Neuron 2001, 30, 665-676.8. Yurdakul, S.; Arici, K., I Mol. Struct. 2004, 691, 45 - 49.9. Adlard, P. A.; Cherny, R. A.; Finkelstein, D. I.; Gautier, E.; Robb, E.; Cortes, M.;Volitakis, I.; Liu, X.; Smith, J. P.; Perez, K.; Laughton, K.; Li, Q.-X.; Charman, S. A.;Nicolazzo, J. A.; Wilkins, S.; Deleva, K.; Lynch, T.; Kok, G.; Ritchie, C. W.; Tanzi, R.E.; Cappai, R.; Masters, C. L.; Barnham, K. J.; Bush, A. I., Neuron 2008, 59, 43 - 55.10. Ritchie, C. W.; Bush, A. I.; Mackinnon, A.; Macfarlane, S.; Mastwyk, M.; MacGregor,L.; Kiers, L.; Cherny, R.; Li,Q.X.; Tammer, A.; Carrington, D.; Mavros, C.; Volitakis,I.; Xilinas, M.; Ames, D.; Davis, S.; Beyreuther, K.; Tanzi, R. E.; Masters, C. L., Arch.Neurol. 2003, 60, 1685-1691.11. Lannfelt, L.; Blennow, K.; Zetterberg, H.; Batsman, S.; Ames, D.; Harrison, J.; Masters,C.; Targum, S.; Bush, A.; Murdoch, R.; Wilson, J. E.; Ritchie, C. W., Lancet Neurol.2008, 7, 779 - 786.12. McKelvey, E. M.; Gottlieb, J. A.; Wilson, H. E.; Haut, A.; Talley, R. W.; Stephens, R.;Lane, M.; Gamble, J. F.; Jones, S. E.; Grozea, P. N.; Gutterman, J.; Coitman, C.; Moon,T. E., Cancer 1976, 38, 1484 - 1493.13. Egger, M.; Hirschel, B.; Francioli, P.; Sudre, P.; Wirz, M.; Flepp, M.; Rickenbach, M.;Malinvemi, R.; Vernazza, P.; Battegay, M., Brit. Med. 1 1997, 315, 1194 - 1199.6114. Cavalli, A.; Bolognesi, M. L.; Minarini, A.;Rosini, M.; Tumiatti, V.; Recanatini, M.;Meichiorre, C., I Med. Chem. 2008, 51,347 - 372.15. Kapkova, P.; Aiptuzun, V.; Frey,P.; Erciyas, E.; Holzgrabe, U., Bioorg. Med. Chem.2006, 14, 472 - 478.16. Rodriguez-Franco, M. I.; Fernandez-Bachiller,M. I.; Perez, C.; Hernandez-Ledesma, B.;Bartolome, B., I Med. Chem. 2006, 49,459 - 462.17. Cavalli, A.; Bolognesi, M. L.; Capsoni,S.; Andrisano, V.; Bartolini, M.; Margotti, E.;Cattaneo, A.; Recanatini, M.; Meichiorre, C., Angew.Chem. mt. Ed. 2007, 46, 3689 -3692.18. Ji, H. F.; Zhang, H. Y., Bioorg. Med. Chem.Lett. 2005, 15, 2 1-24.19. Hider, R. C.; Liu, Z. D., I Pharm. Pharmacol.1997, 49, 59-64.20. Hider, R. C., Toxicol. Lett. 1995, 82-3,96 1-967.21. Dobbin, P. S.; Hider, R.C.; Hall, A. D.; Taylor, P. D.; Sarpong, P.; Porter, J. B.; Xiao, G.Y.; Vanderheim, D., I Med. Chem. 1993, 36, 2448-2458.22. Bebbington, D.; Monck, N. J. T.;Gaur, S.; Palmer, A. M.; Benwell, K.; Harvey, V.;Malcolm, C. S.; Porter, R. H. P., 1 Med. Chem. 2000,43, 2779-2782.23. Dedeoglu, A.; Cormier, K.; Payton,S. M.; Tseitlin, K. A.; Kremsky, J. N.; Lai, L.; Li, X.;Moir, R. D.; Tanzi, R. E.; Bush, A.I.; Kowall, N. W.; Rogers, J. T.; Huang, X., Exp.Gerontol. 2004, 39, 1641 - 1649.24. Schugar, H.; Green, D. E.; Bowen, M.L.; Scott, L. E.; Storr, T.; Bohmerle, K.; Thomas,F.; Allen, D. D.; Lockman, P. R.; Merkel, M.; Thompson, K.; Orvig,C., Angew. Chem.mt. Ed. 2007, 46, 1716- 1718.25. Scott, L. E.; Page, B. P.; Merkel, M.; Orvig,C., unpublished results.26. Dragowska, W. H.; Ruth, T. J.; Adam,M. J.; Kozlowski, P.; Skov, K.; Bally, M. B.;Yapp, D. T. T. In Studies of Tumor Microenvironment andMetabolic Activity in HER2/neu Overexpressing Breast Cancer Xenograftsby MicroPET and MRL AmericanAssociation for Cancer Research, 2005; 2005;p900.27. Zhang, Z.; Rettig, S. J.; Orvig, C.,Can. I Chem. 1992, 70, 763 - 770.28. Lindegren, S.; Skarnemark, G.;Jacobsson, L.; Karlsson, B., Nuci. Med Biol. 1998, 25,659 - 665.29. Espinet, P.; Echavarren, A. M., Angew. Chem.mt. Ed 2004, 43, 4704 - 4734.30. Allen, D. D.; Smith,Q.R., Journal ofNeurochemistry 2001, 76, 1032 - 1041.6231. Di, L.; Kerns, E. H.; Fan, K.; McConnell, 0. J.; Carter,G. T., Eur. I Med. Chem. 2003,38,223-232.32. Kishimoto, S.; Kawazoe, Y.; Ikeno, M.; Saitoh, M.; Nakano, Y.; Nishi,Y.; Fukushima,S.; Takeuchi, Y., Cancer Chemoth. Pharm. 2006, 57, 84 - 90.63CHAPTER 3Monoanionic Glucosamine-based Ligands for theFormation of Neutral Complexes withthe [M(CO)3]Core (M = Re,99mTC)*3.1 Introduction3.1.1 Designing a Carbohydrate-based Imaging AgentThe ability to image the location of carbohydrate metabolism is a very usefuldiagnostic tool.It can be used to derive information on cell function, with increased carbohydrateprocessingactivity found in areas such as cancerous tissue,1 and decreased activity indicatingareas witha loss of metabolic function, such as atrophied brain areas in Alzheimer’sdisease.2Although there is currently a very useful carbohydrate based imagingagent available in theform of FDG (2-’8F-fluoro-2-deoxyglucose, Figure 3.lb), its availability is somewhatlimited. The mechanism of action of FDG is discussed in more detailin Chapters 1.2 and1.3. In summary, the structural similarity between glucose and FDG (illustrated in Figure3.1) means that FDG is transported and to some extent metabolised as if it were glucose.The difference between the two means that the metabolism of FDG can only proceedso far,and as such FDG does not get broken down, but instead gets trapped in cells, particularlythose which overexpress certain proteins; GLUT and hexokinase.aOHbHO\” HOqOHHO—-’-- OHOH ‘8FFigure 3.1 a) D-glucose b)2-’F-fluoro-2-deoxyglucose (FDG). The substitution of ‘8F for -OH at C-2 is the only difference between the two molecules.Access to the ‘8F isotope used in FDG is not widespread as it has a short half life (t112 — 110*A portion of this chapter has been published: Bowen, M. L. and Orvig, C.99mTechnetiumcarbohydrate conjugatesas potential agents in molecular imaging. Chem. Commun, 2008, 5077—5091. A version of this chapter will besubmitted for publication: Bowen, M. L., Lim, N. A. C., Ewart, C. B., Adam, M. J. and Orvig, C. GlucosamineConjugates Bearing JVJ 0-donors: Potential Imaging Agents Utilizing [M(CO)3]Core (M = Re, Tc).References begin on Pg 106 64mm) and is produced in a cyclotron.3’ This meansthat the patient must be geographicallyclose to a cyclotron in order for it to be feasible touse such a short lived isotope without toomuch loss of activity. Countries withno cyclotron, such as New Zealand, do not haveregular access to this very versatile and useful diagnosticagent. This is one of the mainreasons behind the interest in the nuclear medicine communityin discovering a SPECTcarbohydrate-based imaging agent.In designing a carbohydrate based imaging agent, itis vital to retain a reasonable level ofactivity with both GLUTs (glucose transporters) and HK(hexokinase) compared to theparent carbohydrate. To produce useful images,an agent must accumulate in target tissue ata faster rate than it does in background tissue.For this to happen the radiolabelledcompound needs to access cells (GLUT), andbe trapped there (HK), so cells that haveoveractive GLUT and HK, as do most cancer cells,are the cells that are targeted by thisapproach. This was discussed in detail in Chapter 1.3.3.1.2 Technetium GlycoconjugatesThere have been many99mTcglycoconjugates made over the years utilizing several differentapproaches to metal binding. Due to the high dilution involved in99mTcchemistry, sugarswere often modified by adding a binding moiety with a higher affinity for the metal ion thanthe native hydroxyl groups exhibit.5 As with much early work on sugar coordinationchemistry, the compounds were often not subjected to rigorous chemical characterization, sothe exact structures were not known.5’6In vivo studies of99mTclabelled 5-thio-D-glucoseshowed that the compound did not behave like a glucose analogue, suggesting it containedtoo many modifications compared to glucose for it to be recognized and/or used by the nativetransport and metabolizing proteins.7 Recent studies of99mTClabelled 1-thio-f3-D-glucose8result in very different HPLC traces depending on the pH of the solution and theconcentration of the ligand. The complex formed is not particularly stable, given that atneutral pH the99mTccomplex is no longer intact after 2 hr.8 From a coordination chemistryperspective, this also suggests that there may be more than one chemical species presentunder physiological conditions, which greatly complicates interpretation of anybiodistribution data that may be obtained. The authors note they cannot form the rheniumanalogues, which is unusual given that technetium and rhenium normally exhibit very similarReferences begin on Pg 106 65coordination chemistry.8 Cellular uptakeis not affected by the presence of glucose,suggesting that the compound is not transportedby the GLUT transporters.8Yang and coworkers have made and studieda species with99mTC(V)bound toethylenedicysteine-deoxyglucose (ECDG) (Figure3 .2a). This is a tetradentate ligand withan N2S binding sphere and two appended glucosamines,bound to the chelate via amidebonds at their C-2 nitrogens. Radiolabelling proceedsfrom{99mTcO4fin the presence ofSnC12 in 94 % radiochemical yield.9 The ECDGligand was subjected to a hexokinase assay,and the authors claim that ECDG couldbe phosphorylated by hexokinase, though theexperimental results presented in this paper do notseem to provide enough information toverify this claim.9 An in vitro cell uptake study of99mTcECDGwas performed in humanlung cancer cells, and it was found that cellular uptake wassimilar to that observed for FDG(0.5 vs. 0.6 % of total activity used).9 It was also found that uptakeof99mTc..ECDGdecreased in the presence of D-glucose, but not in the presence of L-glucose,suggesting thatthe uptake observed is due to the same transport mechanismsas those used for D-glucose i.e.GLUT transporters.9OHaH°’\OHH\OHb:oCOOHSH HS’ HOOC—’ ‘—COOHFigure 3.2 a) ECDG9b) DTPA-DG;’° protonation states when complexed to metal ions arenot discussed in the original papers.In vivo it was found that rats which had been pretreated with insulin had increased tumoruptake of99mTcECDG,again suggesting the cellular uptake is due to a glucose relatedmechanism.” The authors suggest the complex is stable as there is no accumulation ofactivity in the thyroid in technetium were to become uncomplexed, underphysiological conditions it would be oxidised to pertechnetate, which is known toaccumulate in the thyroid.’2 Biodistribution results show good tumor-to-muscle ratios 4 hrpost injection, suggesting some potential utility as an imaging agent. The amount of activityin the blood is twice as much as in the tumor at all time points, and this ratio does notReferences begin on Pg 106 66increase over time, as it does for FDG.” It istherefore possible that ECDG is notphosphorylated by hexokinase as the authors suggest,and is therefore not trapped in cells,but reaches an equilibrium concentration due toconstant influx and efflux rates.Nonetheless, images obtained from tumor-bearing rats with99mTcECDGclearly show thetumor, while those from the nonglycosylated99mTcECcontrol complex only show activityin the excretory organs.” This demonstrates theutility of the sugar conjugation approach inmolecular imaging.99mTc.ECDGwas also used to monitor treatment efficiency, by imagingtumor-bearing rats before and after treatment with the known anticancer agentsPaclitaxeland Cisplatin.” Differences were seen in the imagesobtained before and after treatment,and the authors suggest99mTCECDGmay be useful in assessing the therapeutic response ofcancer treatments.1’Another study on the use of carbohydrate-containing ligands for the[99mTc013+core utilizesglucosamine bound to a diethylenetriaminepentaaceticacid (DTPA) chelate via an amidebond between the C-2 nitrogen of the sugar and the acetate moiety on the centralnitrogen ofthe DTPA, giving diethylenetriaminepentaacetic acid deoxyglucose (DTPA-DG) (Figure3.2b).1° DTPA is known to be a strong metal chelator, and here the labellingwith99mTcproceeded in excellent radiochemical purity, although there is little characterization dataprovided for either the ligand or the technetium complex)°The ligand, as drawn in theoriginal paper, has seven obvious chelating atoms, and no mention is madeas to which ofthese atoms are believed to be bound to the technetium. The complexwas found to be stablein solution for six hours, and is proposed to stay intact in vivo due to lack of activityin thethyroid,’° a known site of accumulation for free pertechnetate. In vitro cellular uptake of99mTcDTpADGwas found to be about the same as for99mTcECDG- 0.5 % ofadministered activity compared to 0.6 % for FDG. Similar to the findingsfor99mTcECDG, the tumor-to-muscle ratios for the test compound were significantly higher than forFDG, while the tumor-to-blood ratios were lower. Both the tumor-to-blood and tumor-to-muscle ratios for the test compound increase over time to the two hour time point and thenremain constant out to eight hours.’° Images with99mTc.DTpADGof tumor bearing ratsenable visualization of the tumor, while control images with99mTcDTPAshow only theliver and kidneys.’° In a subsequent study, selectivity of99mTc..DTpA..DGfor cancerousrather than inflamed tissue was also demonstrated.’3The ‘88Re analogue of this compoundwas prepared and --9 MBq was injected into the tail vein of tumor-bearing mice.14 PlanarReferences begin on Pg 106 67scintigraphy imaging clearly shows the tumortissue.’4 Twenty one days after this treatmentthe tumors of the treated animals had reducedin size significantly (around 30 %) withrespect to those of a control group.’4 This findingillustrates the potential of using99mTCand‘88Re analogues of the same ligand system for imaging and therapy, respectively.Liu and coworkers examined three deoxyglucose conjugatesas ligands for the[99mTCO]3+core (Figure3•3)15The ligands were made by conjugation via an amide bondto the nitrogenat the C-2 position of glucosamine.’5 One of the ligands (Figure 3.3a)appears to be NSbidentate (though there is no discussion rulingout the involvement of the proximal sugarhydroxyl groups in binding to the metal, or the formation of an ML2species) and the twoothers are tetradentate, one N3S (Figure 3.3b),and the other N2S (Figure 3.3c).Radiolabelling via an exchange reaction with99mTc..glucoheptonateprovided the desiredcompounds in very good radiochemical yields.’5 Biodistribution studiesin tumor-bearingmice show these compounds to all exhibit tumor-to-muscle ratios of2 - 3 four hours postinjection, and tumor-to-blood ratios of 0.5 - L’5 The compound based ona derivatizedMAG3 shows the most promise for further studies (Figure 3.3b),’5thoughthe tumor-to-bloodratio would ideally need to be increased to give improved image quality.OHOHaHo\OHbHOL/NHCHOOH0NH0HONHOHJr__- N NH)NHHS0SH HSFigure 3.3 Three ligands used for binding to[99mTCO]3+lsIt seems that despite significant work in this area over several decades, there have not beenany major breakthroughs made towards a useful glycosylated[99mTcOl3+..basedcompoundfor molecular imaging. As discussed in Chapter 1.4.1, with any ligand that is not C2symmetric, it is not possible to synthesize a single diastereomeric complex of this core, andthe resulting isomers may exhibit different biological properties. The lack of thoroughmolecular characterization of these compounds is also problematic, in terms ofunderstanding results and improving on design. A lack of stability studies may also castReferences begin on Pg 106 68doubt on the exact nature of the99mTcspecies that is being imaged in a given in vivoexperiment.3.1.3[99mTc(CO)31+GlycoconjugatesThe advent of the [M(CO)3Jcore has revolutionalizedthe field of99mTCchemistry.’6 Thiscore brings together many properties of importance in radiopharmaceuticalchemistry toprovide a very attractive base upon which to build SPECT tracers.[M(CO)3]is small,kinetically inert, of moderate lipophilicity, and easyto prepare, especially when made via aMallinkcrodt (now Covidien)IsolinkTMkit. These properties, combined with the favourablecharacteristics of99mTcas a SPECT tracer (ti,2 = 6 hr,y = 140 keV, generator produced),give a very attractive starting point for radiopharmaceuticaldesign.As this is the core used in the work described in this thesis,the glycoconjugates previouslymade on this core are very relevant, so will be coveredin some detail here.3.1.3.1 Bidentate Ligand SystemsGiven the promising results seen in the enzymatic activity retention of N-functionalizedglucosamines, the Orvig group’s first work in this field involved aglucosamine conjugatedligand.’7 This was designed to bind in a bidentate fashion via a phenolateoxygen and asecondary amine: the nitrogen of the glucosamine (Figure3•4)•17‘H NMR spectroscopicstudies of the rhenium complex of this ligand indicated binding of the C-3 hydroxyl group ofthe sugar in addition to the phenolate and amino donors, giving a tridentate faciallycoordinated ligand in solution.’7 As this ligand was not designed to have threeatomsbinding to the metal, the donor atoms were not optimised, and the NO2 bindingsphere didnot show the desired stability when incubated with high concentrationsof cysteine andhistidine ligands for 24 hr.’7 This finding confirms the discussion in Chapter 1: thatcarbohydrates themselves do not make very good ligands.In this case the C-3 hydroxyl isheld very close to the vacant coordination site on the metal, so this proximalgeometry mayencourage binding of a rather weak donor atom that may not bind under other conditions.This type of ligand system was not investigated further, as it was thought that if thecarbohydrate is bound to a metal, it would not be recognized and metabolized in vivo as theReferences begin on Pg 106 69parent carbohydrate would, andthus the complex would likely not be usefulas acarbohydrate based imaging agent.Figure 3.4 A ligand designed tobe bidentate shows coordination of the C-3 hydroxylgroupof the carbohydrate in solution, M = Re,99mTc,‘86Re.’7The coordination of 3-hydroxy-4-pyridinoneligands with pendant carbohydrates to the[M(CO)3]core was then examined, as it was thoughtthat a larger gap between the metalbinding atoms and the carbohydrate would likelyforce the latter to remain pendant.’8 Theseligands consist of two oxygen atoms ortho to oneanother on a six membered, aromatic ring(Figure 3.5). The hydroxyl group is deprotonatedupon metal binding, giving a monoanionicligand capable of neutralizing the positive chargeon the tricarbonyl core to yield a neutraloverall complex with a stable five-membered chelate ring.There was reason to believe thatthis kind of ligand system could be useful for the[M(CO)3]core as the pyridinone basicityis comparable to that of the aromatic amines that are knownto be excellent donors for thesemetal centers. Five ligands were examined in thisstudy, and they were composed of threedifferent carbohydrate attachment methods: via an etherlinkage to C-i of glucose, an amidelinkage at the C-6 of glucose, and to the nitrogen at C-2 of glucosamine, while alsovaryingthe distance between the carbohydrate and the chelating oxygenatoms.’8 These ligands werefound to bind in a bidentate fashion, with the remaining coordination siteson the metal beingoccupied by the three carbonyl ligands, and a (predicted) water moleculein aqueoussolution.’8 The labelling with both99mTcand ‘86Re proceeded in excellent yields, and thestability in excess cysteine and histidine was essentially quantitative aftera 4 hr incubation,with some degradation seen in the histidine experiment after 2418The interaction ofthese compounds with the glucose metabolizing enzyme hexokinase (HK)was examined, butthese compounds were found not to inhibit, orbe phosphorylated by, HK under theconditions tested.’8 This means that the change in environmentof the sugar disrupts theinteractions between the carbohydrate and the enzyme to such an extentthat the compoundeither is not recognized or cannot be phosphorylated.References begin on Pg 106 70OH0 HO\OHOHORHNNOHOMFigure 3.5 Five glucose appended ligands investigatedas bidentate chelates for the[M(CO)3]core where M = Re,99mTc‘86Re.’8In collaboration with the Yano/Mikata groupat Nara Women’s University in Japan, theOrvig group has complexed the Re and99mTctricarbonyl cores to a range of small, welldefined, N2 bidentate, carbohydrate bearing ligands (Figure 3.6).’The sugars were alllinked to a chelating 1,3 -diaminopropyl group via anether linkage at cl20The ligandswere bound to the metal tricarbonyl core via two primary amine donors, andthe remainingcoordination sites on the metal ion were occupied by three carbonylsand one bromide ligandfor the rhenium complexes, or a proposed aqua ligand under the more dilutetracer conditionsused for technetium.’9 The desired pendant nature ofthe various carbohydrates wasconfirmed by a lack of coordination induced shift in the ‘H NMR spectroscopy solutionstudies, and by X-ray crystallographic analysis of two ofthe complexes in the solid state.19The99mTccomplexes of these ligands were formed very readily, and their stabilitywhenincubated with high concentrations of the biologically relevant, potential ligands cysteineand histidine for 24 hours were >90 % for cysteine and >68 % for histidine (whichhas beenshown to have a high affinity for the [M(CO)3jcore21).’9 It is interestingto note that therelative stability of the complexes is in line with the steric size of the ligands,suggesting thatthe larger ligands may not leave enough space for an adventitious amino acid to reach thevacant coordination site on the metal.R = H, AcReferences begin on Pg 106 71aOHbHOO\NHO—<OHOHCOHd-----°HO—\--.- OHo\OH OH N..HINH2 OHHOHNHNH2Figure 3.6 Diamino carbohydrate-based ligands:2°a) the f3 anomer was synthesized forglucose (shown) as well as xylose and galactoseb) the c anomers were synthesized formannose (shown) and galactose, c) and d) ligands whichshowed higher stability in vitroperhaps due to their increased steric bulk.Gottschaldt, Yano and coworkers have developeda set of ligands where the carbohydrate islinked to a metal chelate by a thioether moiety (Figure37)22This is proposed to helpcircumvent the enzymatic cleavage that can occurwith O-glycosidic bonds in vivo. Thethree ligands examined in this study consist of N2, 2,2’-bipyridyl, bidentatechelates,appended with two identical sugars at the 4 and 4’positions of the bipyridine.22 Glucose,galactose and mannose were attached to the chelate via thioether linkages at theC-ipositions. These ligands were bound to the rhenium tricarbonylcore in good yields,resulting in complexes where the metal is bound to thethree carbonyl ligands, the twonitrogen atoms of the bipyridyl, and one anionic chlorideto fill out the binding sphere andprovide a neutral complex.22 A solid state X-ray crystal structureof the acetyl protectedglucose derivative was obtained and it verifies the above coordination sphereas well asconfirming that the carbohydrates do indeed remain pendant, and outsidethe binding sphereof the metal, at least while acetyl protected.22 The analogous99mTCcomplexes were alsoformed in good yields (>95 % radiochemical purity), and were foundto exhibit satisfactorystability when incubated with excess histidine for 4.5 hr.22 After incubating for 24 hrsignificant degradation was observed, though the product of this process is not the[99mTc(His)(CO)3]species that may be expected (as verified by HPLC comparison with thepresynthesized histidine complex).22 The authors speculate that the observed products are aresult of the replacement of the one labile position (occupied by a water molecule or ahalide) with a histidine, while the rest of the coordination sphere remains intact.22References begin on Pg 106 72OHx+1LOH IM/CO1fCOFigure 3.7 An example of a bipyridine complex madeby Gottschaldt et al.22 M = Re, L = Cl,x=OandM=99mTc,L=H20,x=1.Other workers have also found the stability of bidentate complexesto be lower than requiredfor successful in vivo application.23’24 This is thoughtto be due to the binding of some othercompeting endogenous ligand (such as a proteinin vivo), to the vacant coordination site onthe metal, and eventual replacement of the originalligand over time. Consequently, work onthe tricarbonyl cores now mainly focuseson tridentate ligands. Initial work on ligandbinding experiments,23 and more recent verificationin the form of DFT (density functionaltheory) calculations,25’26has shown that the higher the nitrogen content of the bindingsphere, the more stable the resulting complexes willbe. This means that most work in thisarea now incorporates at least one nitrogen donoratom into a tridentate core. Work on basiccoordination chemistry to optimize the metal binding portion of bioconjugatesis ongoing inour labs and others.27’283.1.3.2 Tridentate Ligand SystemsThe first organometallic carbohydrate containing complexes ofgroup 7 metals were reportedin 2001 (Figure3.8).29Two ligands were synthesized, one containing glucose and theother2-deoxyglucose, both connected via a linker at C-i to anNO2 tridentate binding sphere madeup of a tertiary amine and two carboxylates. These ligands werecomplexed to both therhenium and technetium tricarbonyl cores, and were foundto quickly form stable complexes,as evidenced by the small amount of decomposition seen after incubationof the complexes inserum for 24 hr.29 Detailed NMR spectral studies support the coordinationof the ligand viathe three intended donor atoms with the carbohydrate remaining pendant.29 This wasfurtherverified by failed attempts to coordinate the sugars alone, where the same reactionconditionsdid not produce any stable compounds.29References begin on Pg 106 73OHHOCOOHOHN“—CO OHOH1—COOHN“—COOHFigure 3.8 The two ligands used in the firstorganometallic carbohydrate containingcomplexes of group 7 metals.29In 2005, Schibli and coworkers reported the synthesisof three tridentate ligands where themetal binding portion of the molecule was joinedto the C-3 position of a glucose (Figure39)30In each case there was a C2 linker between the oxygenat the C-3 position and themetal binding sphere.3° The donor atoms were made upof an alkyl amine in combinationwith aromatic nitrogens and/or carboxylic acidsto give either an N20 or an NO2 bindingsphere.3° Interestingly, coordination to the rhenium tricarbonylcore proceeded extremelywell for the NO2,dianionic ligand (94 %), whereas onlymoderate yields were obtained withthe two monoanionic ligands (54 — 57%)31The opposite was observed on the tracer level,where reaction with99mTcgave over 90 % for the monoanionic N20 ligands and 78 % for theNO2 ligand.3’Figure 3.9 Glucose-based C-3 functionalized tridentate ligands for the technetiumtricarbonyl core.30A feasibility study of key in vitro tests of these C-i and C-3 linked compounds (Figures 3.8left and 3.9) with some previously synthesized C-2 and C-6 analogues32 (see Figure 3.10)OHReferences begin on Pg 106 74was reported in 2005.’ Labellingwith99mTcproceeded in good yields regardless of linkerlength or position of attachmentto the glucose.3’ Complex stability was examined byincubation at 37 °C with excess cysteine or histidine.3’A small amount of degradation (5—10% after 24 hr) was observed with the aminodiacetateligand sets, whereas the N20 and N3binding groups showed no detectable exchangeunder the conditions tested.3’abNHNOHcOOHCOHHO\COOH,COOH N/COOH/ BCOOH \ /‘—COOH‘—COOHFigure 3.10 Some a) C-3 b) C-6 c) C-2 functionalizedglucose-based tridentate ligandsexamined for binding to the[99mTc(Co)31+core and subjected to in vitro assays.30’3These nine ligands and their metal tricarbonyl complexes were subjectedto a series of invitro assays to examine their interactions with key enzymes in the uptakeand metabolism ofglucose31 — vital interactions to maintain if these99mTccompounds are to be useful asSPECT imaging agents. The ability of the compounds to inhibit thephosphorylation ofglucose by hexokinase was examined, and two of the rhenium complexes were foundto bemillimolar inhibitors of this process (K1 0.25— 5.8 mM).3’ Interestingly, the correspondingfree ligands did not show any inhibitory character.3’As was predicted bymolecular dockingstudies where the size and shape of the active site cleft ofhexokinase was modeled, the twospecies that showed HK activity were those with long alkyl chains linking thesugar andmetal binding portions of the molecule (the latter two ligands shown in Figure 3.1 Oc).3’ It isthought that this type of linker is thin enough to fit between the two domainsof hexokinasewhile the bulky metal chelating portion is far enough removed that it can remainoutside thiscleft. Due to these promising results, the compounds were then examined to see if they weresubstrates for hexokinase, but none of them were - i.e. they were not themselvesReferences begin on Pg 106 75phosphorylated.3’Finally, the cellular uptake ofthe99mTccomplexes was examined in HT29cells which are known to overexpress the GLUT-iglucose transporter.31 Small amounts ofsome complexes were found to get into the cells;3’however, these amounts were low, andthey were neither dose dependent nor affectedby the addition of cytochalasin B, a knownGLUT-i inhibitor.33 These observations, combinedwith the fact that the highest uptake wasobserved for the most lipophilic compound (usingthe ligand shown in Figure 3.1 Oa), led tothe conclusion that the uptake observed was from unspecific passivediffusion rather thanGLUT facilitated transport.31Meanwhile, the Orvig group had started working on a set of carbohydrate-pendantligandswhere the metal binding portion was a dipicolylamineentity.34 This provides an N3 bindingsphere consisting of one tertiary amine and two aromatic amines. Fourcompounds wereinvestigated: a glucose (Figure 3.1 la), xylose (Figure 3.1 ib) and mannose(Figure 3.1 ic)sugar linked via an ethylene spacer to the metal binding moiety at C-i,35 anda glucosaminelinked via a glycine to the nitrogen at the C-2 position (Figure 3.lld).36bOH OHCHOH00dHO\flFigure 3.11 Dipicol6ylamine tridentate ligands for binding to the rhenium and technetiumtricarbonyl cores:343 carbohydrate = a) glucose b) xylose c) mannosed) glucosamine.The expected pendant nature of the carbohydrate was confirmed in each caseby ‘H NMRspectroscopy of the rhenium complexes in solution, and crystallographic analysis of theglucose rhenium complex in the solid state.35 These ligands all reacted quantitatively withthe{99mTc(CO)3]+core, and the resulting complexes were at least 94 % stable in 100-foldReferences begin on Pg 106 76excess of cysteine or histidine afterincubation for 24 hr.35 Biodistribution of these andsimilar complexes in tumor-bearing micesuggest that they are not taken up into cells and/ornot phosphorylated by hexokinase.37This is deduced by the fact that the washout rates of thetumors parallel those of blood; they decrease at similarrates over time. This suggests thatthe observed increase in activity of the tumors overthe background tissue is perhaps due toincreased vasculature that is known tooccur in tumors.37Recently, Zubieta and coworkers have reportedcarbohydrate conjugates that utilizefluorescent quinolinoyl moieties as metal bindingagents for the tricarbonyl cores (Figure3.12).38These ligands consist of glucosamine (with either freeor acetyl protected hydroxylgroups) conjugated to C-2 via a linker (of varying length) leadingto an N3 tridentate bindingcore made up of one tertiary amine and two aromatic amines that arepart of the fluorescentquinolinoyl groups.38 The idea behind this ligandset is that it could be bound to[99mTc(CO)3]+to form a complex suitable for SPECT imaging, and also bound to [Re(CO)3]to give a complex suitable for fluorescence imaging— useful for in vitro testing of cellularuptake and visualizing areas of localization within cells.38 Although99mTcbinding was notreported,38 the binding to rhenium proceeded ingood yield, so it is likely the same will beobserved on the tracer level.HO\OHRO\ORNNH 4/NR=HAc H IFigure 3.12 Fluorescent glucosamine conjugates for the tricarbonylcore.38Schibli and coworkers recently developed an elegant method for forming tridentatechelatesfor the [M(CO)3] core using “click chemistry”.39 This approach was applied to a galactoseanalogue as well as to examples of the other major classes of biologically relevant moleculesReferences begin on Pg 106 77as proof of principle.39 1 -Azido- 1 -deoxy-3-D-galactopyranosewas reacted with L-propargylglycine in the presence of Cu(OAc)2and sodiumascorbate at 100 °C for 30 mm to give aquantitative yield of the expected triazole (Scheme3.1)39This reaction proceeds withoutthe need for protecting groups, and forms anN20 tridentate binding sphere comprised of anitrogen from the newly formed triazole ring, anda primary amine and carboxylic acid thatwere part of the original glycine starting material.39This compound was found to bind toboth the rhenium and technetium tricarbonyl coresin excellent yields.39 Importantly, neitherthe glycine nor the carbohydrate precursors producedany kind of stable complex with99mTealone.39 This allowed for the development ofa one pot synthesis wherein the reagentsnecessary for the ligand formation via click chemistrywere heated together for 30 mm; the[99mTc(CO)3(H20)]+precursor was then added into the reaction mixture and, following anadditional 30 mm of heating, the desired complexwas determined to have been produced invery high radiochemical yield.39 The bombesin analog — the bombesin oligopeptidereplacesthe galactose but the metal chelating portion of the molecule remainsunchanged wasassessed for in vivo and in vitro stability, and both of these were foundto be very high.39This is an exciting new development with a myriad of possibilities forfuture developmentsin radiopharmaceuticals, where fast and efficient synthesis is crucial.OH1. Cu(OAc), sodium ascorbateHON3OH100 °C, 30 mmScheme 3.1 “Click to chelate” — a fast and efficient synthesis of a galactose-appendedtridentate ligand suitable for binding to the Re and99mTCtricarbonyl cores.39Although there has been significant work performed with the aim of makinga99mTccarbohydrate-based imaging agent, to date none of these efforts have been successful. Theabove work has pointed out the neccessity of a tridentate binding group and nitrogenousdonors for complex stability. From previous work in the Orvig group it is known thatmonocationic compounds of this type give biodistributions characteristic of charged ratherthan sugar-type species,37 so this work focuses on monoanionic ligands to provide neutralmetal complexes. Other studies in the Orvig group have shown the utility of using aphenolate as an anionic donor group for the tricarbonyl core, as this results in stableReferences begin on Pg 106 78complexes, while increasingthe lipophilicty compared to the commonly usedacid group.27The work in this chapter aimsto combine these factors by examining variouscombinationsof binding groups that result in monoanionic,tridentate ligands. The resulting rheniumandtechnetium-99m tricarbonyl complexesare tested in appropriate assaysto assess theirpotential as molecular imaging agents.3.2 Experimental3.2.1 Instruments and MaterialsAll solvents and chemicals were reagent gradeand used as received unless specifiedotherwise. Reagents were purchased from Acrosunless otherwise stated. Salicylaldehydewas purchased from Alfa Aesar. Sodiumborohydride was purchased from Fisher Scientific.Pd(OH)2 (20 % on carbon), 1 -methyl-2-imidazolecarboxaldehydeand 2-pyridinecarboxaldehyde were purchasedfrom Aldrich. Glucose (hexokinase) assay kits,glucose standard solution and hexokinase wereobtained from Aldrich. Re(CO)sBr iscommercially available (STREM). [Re(CO)3(H20)]Br,4°1VN-dibenzylglycine,4’and 1,3,4,6-tetra-O-acetyl-3-D-glucosamine.HCl42were preparedas previously described. Solvents wereHPLC grade, and were purchased from Fisher Scientific.Ethanol was dried over activated4A molecular sieves for at least 48 hoursprior to use. Hydrogen and argon were purchasedfrom Praxair.IsolinkTMkits were provided by Mallinckrodt Inc.(now Covidien).Na99mTcO4was provided by the Nuclear Medicine Departmentat the University of BritishColumbia Hospital. HL3 and ReL3were first made by Mr. Chuck Ewart,43 HL4 and ReL4were made by Dr. Neil Lim.44 Compound numbersrefer to Figure 3.13 on page 89.The analytical TLC plates, which were aluminumbacked ultra pure silica gel 60, 250 rim,and the flash column silica gel (standard grade, 60A, 32-63 mm) used were provided bySilicycle. 1H and ‘3C NMR, ‘3C NMR APT,2D ‘H-’H COSY and ‘H-’3C HMQC spectrawere recorded on Bruker AV400 or DRX400 instrumentsat ambient temperature. The NMRspectra are expressed on the scaleand were referenced to the residual peaks of thedeuterated solvent. Infrared spectra were recordedon a Nicolet 6700 FT-JR (Fouriertransform infrared) spectrophotometer in transmissionmode between 400 and 4000 cm’ at aresolution of ± 0.09 cm1. ESI mass spectra were recordedon a Micromass LCT instrument.References begin on Pg 106 79High resolution massspectra (Micromass LCT TOF-MS) andelemental analysis (Carlo ErbaEA 1108 Elemental Analyzer)were provided by the Analytical ServicesFacility, Departmentof Chemistry, Universityof British Columbia. HPLC analysisof non-radioactive compoundswas done on a PhenomenexSynergi 4 im Hydro-RP 80A column(250 x 4.6 mm) in aWaters WE 600 HPLC systemequipped with a 2478 dual wavelengthabsorbance UVdetector run using the Empowersoftware package. HPLC analyses ofradiolabelledcomplexes were performedon a Knauer Wellchrom K-1001 HPLCequipped with a K-2501absorption detector and a Capintecradiometric well counter. A PhenomenexHydro-Synergi4 C18 RP analytical columnwith dimensions of 250 x 4.6 mm wasused.3.2.2 Synthesis2-((N, N-Dibenzylamino)acetamido)-2-deoxy-D-glucopyranose(2)The synthesis of this compound wasimproved from thatOAtreported previously.43 N N-dibenzylglycine(10.7 g 0.0421,_,___50Ac042’\iAcO— OAc12mol) was dissolved in hot dimethylformamide (150mL), andthe resulting solution cooledon ice. 1-Ethyl-3-(3-dimethylaminopropyl) carbodumide hydrochloride(EDC)(8.07 g, 0.0421 mol) and 4-dimethylaminopyridine(DMAP) (0.514g, 4.21 mmol) wereadded; the solution was purged with argon,and stirred for one hour. Meanwhile, 1,3,4,6-tetra-O-acetyl-3-D-glucosamine.HCl(1) (18.0 g, 0.0468 mol) was stirred vigorously with 1M Na2CO3solution (250 mL) and dichloromethane(200 mL) for 30 mm. The organic layerwas separated and the aqueous layer washed withtwo further portions of dichloromethane(100 mL). The organic extracts were combined,dried over MgSO4,filtered and reducedon arotary evaporator to give the free aminewhich was added to the dimethylformamide solutionand the reaction mixture was stirred underinert atmosphere at room temperature for 16 hr.The dimethylformamide was removed ona rotary evaporator to give a very thick off-whiteoil which was taken up in dichioromethaneand washed sequentially with saturated Na2CO3solution, 1 M HC1 solution, water, andthen brine. It was then dried over Mg504,filteredand reduced on a rotary evaporator toan off white solid. This solid was dissolved inaminimum amount of warm dichloromethane,and precipitated by addition of ethyl acetate(200 mL). Filtration and washing with cold ethylacetate gave a pure white solid (14.3g, 52% yield). ‘H NMR (DMSO-d6,400 MHz,): 7.53 (m, 4H, H]2), 7.43 (m, 6H, H13,14), 5.85References begin on Pg 10680(s, 1H, Hi), 5.33 (s, 1H, H3), 4.90(dd,3J4310.0 Hz, 3J4,5 = 9.6 Hz, 111, H4), 4.19(dd,3J6a,5= 4.8 Hz,3J6a,6b= 7.8 Hz, 1H, H6a), 4.08 (m, 1H, H5),4.00 (m, 4H, H2,6b,9), 3.54 (s,4H, H10), 2.00, 1.97, 1.97,1.86 (s, 3H, OCOCH3).2-(2-Aminoacetamido)-2-deoxy-1,3,4,6-tetra-O-acety1--D-g1ucopyranoseacetate salt (3)OAcThis was made by a similar,or improved method to that reportedAc0°previously.43 Larger scale reactionswere done as outlined here, andAc0— N\1OAc7NHsmaller scale reactions couldbe done in much shorter reaction times—NH2HOAc (2 hr) under 200 psi of H2 in a Parr hydrogenation bomb. Briefly,2-((1V N-dibenzylamino)acetamido)-2-deoxy-D-glucopyranose(2)(2.36 g, 4.04 mmol) was dissolved inglacial acetic acid (30 mL); Pd(OH)2on carbon (20 %)(0.500 g, 0.7 14 mmol) was added, andthe flask evacuated and filled with H2.The reactionwas stirred at room temperature undera positive H2 pressure from a balloon for12 — 20 hr.The reaction was monitored by TLC,and once the absence of both mono and dibenzylatedmaterials was confirmed, the mixture was filteredthrough a celite plug, evacuated onarotary evaporator to give a pale orangeoil and analysed using ‘H NMR spectroscopy.The‘H NMR spectrum was used in each caseto quantify the amount of acetic acid associatedwith the product both for calculatingyield and ensuring addition of sufficientbase toneutralise all the acid in subsequentreactions. The crude reaction mixturewas used as is inthe next step (3.45 mmol, 85 % yieldby ‘H NMR). ‘H NMR (CDC13,400 MHz,ö): 6.00 (d,= 9.4 Hz, 1H, H7), 5.67 (d, 3J12 8.8Hz, 1H, Hi), 5.15 (dd, 3J,2 10.4 Hz, 3J,4 9.6Hz, 1H, H3), 5.07 (d, 3J43= 9.6 Hz, 3J45 = 9.6 Hz, 1H, H4), 4.22 (m, 2H, H2,6a), 4.08(dd,3J6b,5= 2.1 Hz,3J6a,b= 12.5 Hz, 1H, H6b), 3.81 (dd,3J5,6a= 4.6 Hz,3J56b= 2.2 Hz, 3J5,4 =9.8 Hz, 1H, H5), 2.06 (s, 2H, H9), 2.03, 2.00,1.99, 1.88 (s, 3H, OCOCH3).2-((N-(Pyridin-2-ylmethyl)amino)acetamido)-2-deoxy-1,3,4,6-tetra-O-acety1--D-glucopyranose (4)QAc Thesynthesis of this compoundwas improved from thatAcO40described previously.43 2-(2-Aminoacetamido)-2-deoxy-AcO OAc7NH12131,3,4,6-tetra-O-acetyl-3-D-glucopyranose acetatesalt (3) (5.04mmol) was dissolved in dry ethanol(30 mL), and anappropriate amount of sodium carbonate (to bein excess of theacetic acid present in the starting material) was added(3.73 g, 40.0 mmol). After stirring forReferences begin on Pg 106 81ten minutes to ensure all the acid wasquenched, 2-pyridine-carboxaldehyde (478 1iL, 5.00mmol) was added to the reaction mixture.A very fast colour change to purple indicated anearly instantaneous reaction, and withinten minutes the reaction was complete as verifiedby TLC. At this time, sodium borohydride (0.56g, 15.0 mmol) was added and the reactionmixture stirred for ten minutes longer,as the purple colour faded. The reaction mixturewasquenched with water, which also servedto dissolve the excess sodium carbonate. Theresulting mixture was extracted three times with dichloromethane;the organic fractions werecombined and washed once with water, once withbrine, then dried over magnesium sulfatebefore being filtered and reduced to a yellow oilon the rotary evaporator. This oil waspurified using column chromatography (2% methanol in dichioromethane) and the solventsremoved in vacuo to give a white solid(1.31 g, 52 % yield). ‘H NMR (DMSO-d6,400 MHz,8): 8.50 (d, 3J7,2 = 4.3 Hz, 1H, H7), 8.09 (d, 3J15,14= 9.8 Hz, 1H, H15), 7.75 (ddd, 3Ji,12 =7.5 Hz, 3J13j= 7.6 Hz, 4J13j= 1.8 Hz, 111, H13), 7.38 (d, 3J12,13= 7.5 Hz, 1H, H12), 7.25(m, 1H, H14), 5.88 (d, 3J12 = 8.8 Hz, 1H,Hi), 5.36 (dd, 3J32 = 10.4 Hz, 3J,4 = 9.6 Hz, 1H,H3), 4.90 (dd, 3J4,= 9.7 Hz, 3J4,5 = 9.8 Hz, 111, H4),4.19 (dd,.J6a,5= 4.8 Hz,J6a,b= 12.9Hz, 1H, H6a), 4.01 (m, 3H, H2,5,6b),3.67 (s, 2H, HiO), 3.11 (s, 2H, H9), 2.03, 2.01, 1.98,1.92 (s, 3H, OCOCH3).2-((N-(Pyridin-2-ylmethyl)-N-(methoxyethanoic acid)amino)acetamido)-2-deoxy-1,3,4,6-tetra-O-acety1--D-g1ucopyranose (5)OAc2-((N-(Pyridin-2-ylmethyl)amino)acetamido)-2-deoxy-1,3,4,6-AcO°tetra-O-acetyl--D-g1ucopyranose (4) (0.950g, 1.92 mmol) wasAcO- N QAc7NH1213 dissolved indry dichioromethane (15 mL). Na2CO3 (0.406 g,3.84 mmol) and methyl bromoacetate (0.241 mL, 2.88 mmol)17were added. The flask was evacuated and filled with argon, and“18the reaction mixture stirred at room temperature for 48 hr. Waterwas added, the two layers were separated, and the organic layerwas washed again with water,and once with brine, before being dried over MgSO4.The drying agent was filtered off; thefiltrate was evaporated on a rotary evaporator, andthen purified by column chromatography onsilica gel with 5 % methanol in dichioromethane as eluent. Thesolvents were removed in vacuoto give an off white solid (0.790g, 72 % yield). ‘H NMR (CDC13,400 MHz, 6): 8.98 (d, 3J7,2 =9.4 Hz, 1H, H7), 8.60 (d, 3J15j= 4.5 Hz, 1H, H15), 7.67 (ddd, 3J13,12 = 7.7 Hz, 3J1314 = 7.6 Hz,= 1.7 Hz, 1H, H13), 7.23 (m, 2H, Hi2,i4), 5.84 (d,.J]2= 8.8 Hz, 1H, Hi), 5.31 (dd, 3J,2References begin on Pg 106 82= 10.2 Hz, 3J34 = 9.5 Hz, 1H, H3),5.13 (dd, 3J4,= 9.6 Hz, 3J45 9.8 Hz, 1H, H4), 4.30 (m, 2H,H2,6a), 4.13 (dd,3J6b,5= 2.0 Hz,3J6ab12.4 Hz, 1H, H6b), 3.88 (m, 3H, H5,iO), 3.71 (s, 3H,H18), 3.40 (d,3Jl6a,b= 3.1 Hz, 2H, H16), 3.31 (s, 2H, H9), 2.09,2.02, 1.99, 1.94 (s, 3H,OCOCH3).‘3C NMR (CDC13,100MHz, ö): 171.90, 171.87, 170.89, 170.55, 169.59, 169.34(OCOCH3, C8,17), 157.77 (Cii),149.87 (C15), 137.08 (C13), 123.34 (Ci2), 123.00(C14),92.62 (Ci), 73.12 (C3), 72.93 (C5),68.43 (C4), 61.91 (C6), 59.60 (dO), 58.64 (do),54.76(C9), 52.76 (C2), 51.99 (C18), 20.96, 20.93, 20.80,20.78 (OCOCH3).IRVmax (cm’): 3321 (w),2958 (w), 1744 (s), 1663 (m), 1508(m), 1390 (m), 1218 (s), 1038 (m). HR-MS (ES+ of MNa):mlz calcd forC25H33N3O12Na: 590.1962, found:590.1948.2-((N-(Pyridin-2-ylmethyl)-N-(ethanoic acid)amino)acetamido)-2-deoxy-D-glucopyranose (HL1)OH2-((N-(Pyridin-2-ylmethyl)-N-(methoxyethanoicacid) amino)HO\1OHacetamido)- 1,3 ,4,6-tetraacetyl-2-deoxy-D-glucopyranose(5) (407’JH1213mg, 0.071 mmol) was suspended in 1M NaOH (2 mL) andstirred vigorously. After about 5 mm the solution becameO=17homogenous and slightly yellow. The reactionwas completebHaccording to ESI-MS after about 60 mm. Amberlite CG-50 ionexchange resin was added and the resulting slurrystirred vigorously for 15 mm, before theresin was filtered off. The aqueous filtrate was reduced toa yellow solid on the rotaryevaporator before being taken up in a minimumamount of methanol and filtered to removemost of the NaOH. The methanolic solution was then purifiedby reverse phase HPLC, andthe solvents removed in vacuo to give a pale orange oil (15mg, 55 % yield). ‘H NMR(MeOH-d4,400 MHz, ö): 8.79 (s, 1H, HiS), 8.41(dd, 3J13,12 = 7.4 Hz,3J13,14 = 7.0 Hz, 1H,Hi3), 7.93 (m, 1H, H12), 7.86 (m, 1H, Hi4), 5.10(d, 3J1,2 = 3.1 Hz, 0.75H, Hia), 4.66 (d,8.2 Hz, 0.25H, H]j3), 4.41 (d,3JlOaiOb= 4.7 Hz, 2H, H]O), 4.30, 4.02, 3.66 — 3.90, 3.50— 3.64, 3.34 — 3.42 (m, C2-6, 9,16a and /3). ‘3C NMR (MeOH-d4,100 MHz, 6): 173.79,172.72 (Ci7), 162.63, 162.29 (C8), 156.53 (Cii), 145.78,145.66 (Ci3), 143.35 (Ci5),126.71 (C12), 126.22 (C14), 96.40, 94.27, 92.10 (Ci), 77.58,75.38, 73.16, 72.76, 72.26,71.95, 71.71, 68.21 (C3,4,5), 62.28, 61.88, 58.31 (C9,iO,i6), 58.23,55.48 (C2), 57.31, 57.17(C6). JR Vmax (cm’): 3272 (s, br), 2918 (m), 1667 (s, br), 1538 (m),1417 (w), 1188 (s), 1031(s). HR-MS (ES+ of MNa): m/z calcd forC16H23N3O8Na: 408.1383,found: 408.1388.References begin on Pg 106 83(2-((N-(Pyridin-2-ylmethyl)-N-(ethanoic acid)amino)acetamido)-2-deoxy-D-glucopyranosyl)tricarbonylrhenium(I) (ReL1)OH 2-((N-(Pyridin-2-ylmethyl)-N-(ethanoic acid)amino)acetamido)2-deoxy-D-glucopyranose (HL1) (0.044g, 0.11 mmol) and75’4H1213[Re(H20)3(CO)]Br (0.046g, 0.11 mmol) were refluxed inOEc8 10/14methanol (5 mL) for 2 hr. After cooling to roomtemperature,15Qk 17Re(CO)3the complex was purified by semi-preparative HPLC,and the0solvents removed on a rotary evaporator to givea clear oil(0.049 g, 66 % yield). ‘H NMR (MeOH-d4,400MHz, 6): 8.81 (d, 3J15,14 = 5.4 Hz, 1H, H15),8.09 (dd, 3J13,14 = 7.8 Hz,3J13,12 7.8 Hz, 111, H13),7.73 (d, 3J12,13 = 7.6 Hz, 1H, H]2), 7.55(dd, 3J1413 7.40 Hz,3J14,15 = 5.8 Hz, 1H, H14),5.16—5.33,4.34—4.70,4.06—4.12,3.65 —3.91, 3.25 — 3.53 (m, C1-6,8-1O,16a and 8).‘3C NMR (MeOH-d4,100 MHz, 6): 197.8,197.0, 196.4 (Re-CO), 183.6, 180.5 (C8), 168.4(C17), 159.8, 159.6 (Cli), 152.2 (CiS),140.4 (C13), 125.8 (C12), 123.9 (C14), 95.7,93.7, 91.3 (Ci), 76.9, 74.8, 72.1, 71.9, 71.3,70.9 (C3,4,5), 69.4, 69.2, 69.1, 69.0, 68.8, 68.7 (C9,]O,i6) 61.5,60.9 (C6), 57.1, 54.4 (C2).JR Vmax (cm’): 3292 (w, br), 2360 (m), 2341(m), 2027 (m), 2015 (s), 1910 (s), 1863 (s),1635 (m, br), 774 (m). HR-MS (ES+ of MNa):m/z calcd forC19H22N3O11187ReNa678.0710, found: 678.0704.2-((N-(1-Methylimidazol-2-ylmethyl)amino)acetamido)-2-deoxy-1,3,4,6-tetra-O-acetyl--D-g1ucopyranose (6)OAc2-(2-Aminoacetamido)-2-deoxy- 1,3 ,4,6-tetra-O-acetyl-3-D-6Lo glucopyranose (3) (6.52 mmol) was dissolved in dry ethanolAc04 2’.AcO— \_—OAC7NH13(20 mL), and an appropriate amount of sodium carbonate wasO8 j/1. .—14added to neutrahse the associated acetic acid, as determinedby9HN‘H NMR spectroscopy of starting material (4.60g, 43.4 mmol).After stirring for ten minutes to ensure all the acid wasquenched, 1 -methylimidazole-2-carboxaldehyde (0.770g, 6.99 mmol) was added to the reaction. A nearly immediate colourchange to green was observed, and within ten minutes the reactionwas complete, as testedby TLC. Sodium borohydride (0.780g, 21.1 mmol) was added and the reaction mixturestirred for ten minutes longer, as the green colour faded. The reaction mixture was quenchedwith water (30 mL), which also dissolved the excess sodium carbonate.The solution wasextracted three times with dichloromethane (3 x30 mL), the organic fractions wereReferences begin on Pg 106 84combined and washed once with water(30 mE), once with brine (30 mE), then dried overmagnesium sulfate before being filtered and reducedto an oil on the rotary evaporator. Thisoil was purified using column chromatography (5 %methanol in dichloromethane) and thesolvents evaporated in vacuo to give an off white solid(0.81 g, 25 % yield). ‘H NMR(MeOH-d4,400 MHz, ): 7.67 (d, 3J72 = 9.0Hz, 1H, H7),6.95 (s, 2H, HiS), 6.84 (s, 2H,H14), 5.80 (d, 3J12 = 8.7 Hz, 1H, H]), 5.24 (dd, 3J32= 9.4 Hz, 3J34 10.2 Hz, JH, H3), 5.13(dd, 3J4,= 9.4 Hz, 3J4,5 = 9.7 Hz, lH, H4), 4.27 (m, 2H, H2,6a), 4.13 (dd,3J6b,5= 2.2 Hz,J6b,a= 12.4 Hz, JH, H6b), 3.83 (ddd, 3J,4 = 9.8 Hz,3J5,6a= 2.2 Hz,3J5,6b= 4.7 Hz, 1H,H5), 3.77 (d,3J10a,JOb= 6.4 Hz, 2H, HiO), 3.63 (s, 3H, H13), 3.29 (d,3J9a9b= 4.6 Hz, 2H,H9). ‘3C NMR (CDCI3,100 MHz, ö): 171.76 (C8), 170.66, 169.32 (OCOCH3),144.21 (C12),127.29 (C14), 121.31 (C13), 92.46 (Ci), 72.81 (C3),72.70 (Cs), 67.89 (C4), 61.66 (Ca),52.73 (C2), 52.12 (C9), 44.98 (Cii), 32.46 (Ci5), 20.90,20.69, 20.56 (OCOCH3). JRVmax(cm’): 3324 (w), 2953 (w), 2360 (m), 2341 (m), 1743(s), 1673 (m), 1519 (m), 1367 (m),1214 (s), 1034 (s), 728 (m). HR-MS (ES+ of MNa):m/z calcd forC2,H30N4O10Na521.1860, found: 521.1863.2-((N-(1 -Methylimidazol-2-ylmethyl)-N-(2-hydroxybenzyl)amino)acetamido)-2-deoxy-1,3,4,6-tetra-O-acety1--D-g1ucopyranose (7)QAc2-((N-( 1 -Methylimidazol-2-ylmethyl)amino)acetamido)-2-AcO’deoxy-D-glucopyranose (6) (0.800 g, 1.62 mmol) wasAcO—_ N\1QAc127NH dissolved in 1,2-dichioroethane (15 mL) and sodium carbonate—N414(0.343 g, 3.24 mmol) was added. The reaction flask wasN16\/OHpurged for ten minutes with argon and salicylaldehyde (33717j.iL, 3.21 mmol) was added, and the reaction mixture stirred at1819room temperature for 30 mm. Sodium triacetoxyborohydride(1.02 g, 4.83 mmol) was added and the reaction mixture stirred for 24 hr. Upon completionof the reaction, as determined by loss of starting materials by TLC, the solvent was removedby rotary evaporator and the off white solid partitioned between dichioromethane and water(20 mL each). The layers were separated and the aqueouslayer washed twice withdichloromethane (20 mL), the two fractions of which were then combined. These were driedwith brine (30 mL) then MgSO4,filtered and reduced in volume on the rotary evaporator.The yellow solution was purified by column chromatography on silica gel using 4 %methanol in dichioromethane as an eluent, and the solvent was removed in vacuo to give anReferences begin on Pg i06 85off white solid (0.626g, 64 % yield). ‘H NMR (CDC13,400 MHz, ö): 7.97 (d, 3J7,2 = 9.0 Hz,1H, H7), 7.14 (ddd, 3J1918 = 7.2 Hz, 3J1920 =8.2 Hz, 4J19,17 1.5 Hz, 1H, H19), 7.07 (d,= 1.1 Hz, 1H, H]4), 6.96 (dd, 3J1718 7.4 Hz,4J17j= 1.3 Hz, 1H, H]7), 6.87 (s, 1H,H]3), 6.85 (s, 1H, H20), 6.71 (dd, 3J18,17 = 7.4Hz, 3J18,19 = 7.3 Hz, 1H, H18), 5.72 (d, 3J1,28.4 Hz, 1H, Hi), 5.10 (dd, 3J32 = 9.6 Hz, 3J34= 9.9 Hz, 1H, H3), 5.00 (dd, 3J4,= 9.8 Hz,9.4 Hz, 1H, H4), 4.23 (dd,3J6a512.4 Hz,2J6a,/,= 4.5 Hz, 1H, H6a), 4.02 (dd,3J6b,5= 2.1 Hz,2J6a6b= 12.6 Hz, 1H, H6b), 3.96 (dd, 3J23 9.4Hz, 3J27 = 9.6 Hz, 1H, H2), 3.84(d,2JIOa,IOb= 13.1 Hz, 1H, HiO), 3.78 (ddd,3J56a= 4.5 Hz,3J5,6b= 2.0 Hz, 3J5,4 = 9.9 Hz, 1H,H5), 3.67 (dd,2Jl5a,Ib= 15.6 Hz,2JJ5b,15a= 14.1 Hz, 2H, H15), 3.62 (d,2JJQb,JOa13.1 Hz,1H, HiO), 3.52 (s, 3H, Hi2), 3.18 (dd,2J9a9b= 16.7 Hz,2J9b,a= 10.6 Hz, 211, H9), 2.03,1.94, 1.85, 1.76 (s, 12H, OCOCH3).‘3C NMR (CDC13,100 MHz, ö): 171.71 (C8), 170.85,170.26, 169.55, 169.04 (OCOCH3),157.42 (C21), 145.28 (Cii), 131.14 (Ci7), 129.88(C]9), 125.93 (Ci4), 122.35 (Ci6), 121.82 (C20),119.06 (C18), 117.85 (Ci3), 92.04 (Ci),72.84 (C3), 72.56 (C5), 68.27 (C4), 61.85 (Co), 56.57(C9), 56.49 (dO), 53.24 (C2), 50.27(Cu), 32.70 (C12), 20.95, 20.87, 20.76, 20.68 (OCOCH3).IR Vmax (cm’): 3274 (w), 1744(s), 1674 (m), 1388 (m), 1211 (s), 1072 (m), 1035(s), 760 (s). HR-MS (ES+ of MNa): m!zcalcd forC28H36N4O11Na: 627.2278, found: 627.2288.2-((N-(1-Methylimidazol-2-ylmethyl)-N-(2-hydroxybenzyl)amino)acetamido)-2-deoxy-D-glucopyranose (HL2)OH 2-((N-( 1 -Methylimidazol-2-ylmethyl)-N-(2-hydroxybenzyl)HOH12amino) acetamido)-2-deoxy- 1,3 ,4,6-tetraacetyl-D-glucopyranose(7) (0.225 g, 0.372 mmol) was dissolved in methanol (10 mL).An excess of NaOMe (0.110g, 20.4 mmol) was added and the16/OHreaction mixture stirred at room temperature fortwo hours.17?2120Amberlite CG-50 ion exchange resin was added, and after ten19mm of rapid stirring, was removed by filtration. The filtrate wasreduced in volume on a rotary evaporator, and the resulting yellow-orangeoil purified bysemi-prep HPLC. The solvents were removed in vacuo to give a paleyellow oil (0.040 g, 25% yield). ‘H NMR (MeOD, 400 MHz, ö): 7.28 — 7.32 (m, 2H, H20,i4),7.18 —7.20 (m, 111,H18), 7.09 - 7.14 (m, 1H, Hi7), 6.76— 6.82 (m, 2H, Hi9,i3),5.09 (d, 3J1,2 = 3.4 Hz, 0.63H,H]a), 4.67 (d, 3J12 = 8.3 Hz, 0.37H, Hi/i), 4.00 —4.10 (m, 2H, HiO),3.79 — 3.95 (m, 4H, aand/I sugar protons) 3.74 (s, 311, Hi2), 3.72 (s, 211, H15), 3.44 (s, 2H, H9)3.35—3.43 (m,References begin on Pg 106 862H, a and/3 sugar protons). ‘3C NMR (MeOD, 100MHz, ): 173.64, 173.24 (C8), 157.74,157.59 (Cl]), 146.42 (C21), 133.31, 133.15(C18), 131.20, 131.12 (C]7), 125.48 (C20),123.97 (C]6), 121.25, 121.09 (C19), 119.65 (C]4),116.92 (C13), 97.23, 92.98 (Cl), 78.49,76.31, 73.54, 73.01, 72.93, 72.65 (C3,4,5), 63.20,63.12 (C6), 59.96, 59.89 (C9), 58.88,56.00 (C2), 57.61, 57.55 (C15), 49.50 (obscured bysolvent peak dO), 35.26, 35.18 (C12).JR Vmax (cm’): 3293 (s, br), 2942 (m, br), 1671(s), 1609 (w), 1531 (m), 1459 (m), 1210 (s),1133 (s), 757 (m). HR-MS (ES+ of MNa) m/zcalcd forC20H28N4O7Na: 459.1856, found459.1848.(2-((N-(1-Methylimidazol-2-ylmethyl)-N-(2-hydroxybenzyl)amino)acetamido)-2-deoxy-D-glucopyranosyl)tricarbonylrhenium(I) (RCL2)2-((N-( 1 -Methylimidazol-2-ylmethyl)-N-(2-hydroxybenzyl)amino)acetamido)-2-deoxy-D-glucopyranose (HL2) (0.040g,0.92 mmol) was dissolved in methanol (5 mL),and[Re(H20)3(CO)]Br (0.036g, 0.097 mmol) added. Thereaction mixture was refluxed for 8 hr then the volumeOHHOH127NHN.13O—8—iiKH-14/-. N.16LQRe(CO)317V\211820reduced on a rotary evaporoator before purificationby semi-19preparative HPLC, followed by removal of the solvents invacuo to give a pale yellow oil (0.028g, 43 % yield). ‘H NMR (MeOD, 400 MHz, ö): 7.53(d, 3J20,19 = 7.4 Hz, 1H, H20), 7.35 (dd, 3J18 p = 7.1 Hz,3J18,19 7.1 Hz, 1H, H18), 7.20 (d,= 5.8 Hz, 1H, H17), 7.14 (s, 1H, H14), 6.98 (m, 2H, H13,19), 5.10 (d, 3J,,2= 3.3 Hz,0.45H, Hia), 4.40 (s, 0.55H, Hl/3), 4.66 — 4.95,4.08 — 4.14, 3.64 — 3.81, 3.01 — 3.55, (m,1311, H2-6, 9,lO,15a and/3 anomers),3.64 (s, 311, H]2). ‘3C NMR (MeOD, 100 MHz, ö):197.94, 197.44, 197.35, 196.55, 196.35 (GO), 181.29, 181.23(C8), 158.24, 158.20 (Cli),152.17, 152.04 (C16), 135.03, 135.00 (C20), 133.00(C18), 129.61 (C14), 125.88, 125.80(C17), 121.49 (C]9), 120.15, 120.01 (C21), 117.58 (C13),96.19, 92.53, 91.99 (Cl), 78.46,75.69, 75.18, 73.51, 73.44, 72.36, 72.27, 72.13, 72.05 (C3,4,5),68.18, 67.92 (C15), 65.29,64.81, 64.43 (dO), 62.92, 62.74 (C9), 59.14, 58.95(C6), 57.47, 57.42 (C2), 35.40, 35.28,35.21 (C12). JRVmax(cm’): 3232 (m, br), 2360 (w), 2028 (s), 1884 (s), 1672 (m), 1620(s),1198 (s), 1134 (s), 757 (m), 527 (s). HR-MS (ES+ of MH) m/z calcd forC23H28N4010’87Re:707.1363, found 707.1365.References begin on Pg 106 873.2.399mTCComplex FormationNa99mTcO4was obtained in saline solution from theUBC Hospital Department of NuclearMedicine. It was added to an IsolinkTMkit (Mallinckrodt) and the volume made up to1 mLwith 0.9 % saline solution. The resulting solutionwas heated at approx. 90 °C for 30 mm.After cooling, the solution was neutralized withabout 0.1 mL of 1 M HC1 to pH 7 — 10depending on the particular compound. Meanwhile,1 0 M solutions of the ligands wereprepared, and 0.5 mL of each was transferredto a reaction vial which was fitted with aseptum and purged with N2. Ligands containinga phenol as a chelating moiety were madeup in ethanol and three equivalents of NaOEt wereadded to deprotonate the phenolate (0.15mL of a 1 0 M solution in ethanol) prior to additionof[99mTc(CO)3(H20)]+.Ligandscontaining an acid chelating moiety were madeup in aqueous PBS buffer (110 mM, pH 7.4,0.9 % saline) and were used without the additionof base. Between 0.1 and 1 mL of thetechnetiumIsolinkTMreaction was added to each vial, and these were then heated between70— 80 °C for 30 mm. The solutions were cooledthen injected into the HPLC (with thecorresponding cold standard rhenium complex)for analysis.3.2.4 Cysteine and Histidine ChallengesThe99mTccomplexes were synthesized as outlined above. Cysteine andhistidine solutionswere freshly made at i0 M in aqueous PBS buffer (110 mM,pH 7.4, 0.9 % saline).Cysteine or histidine solution (0.9 mL) was placed in a vial, and 0.1mL of the technetiumcomplex (in the reaction solution described above) was added to givea final volume of 1 mLwith the concentration of competing amino acid one hundred timesgreater than that of theligand used for making the technetium complex. The solutionwas incubated at 37 °C andaliquots removed at 1 hr, 4 hr and 24 hr for HPLC analysis. Each timepoint wasdone intriplicate from three separate incubation vials. The percentageof the initial complexremaining was examined, and the mean and standard deviation of three trials foreach ligandwith each of cysteine and histidine was calculated.References begin on Pg 106 883.2.5 GLUT-i Cell Uptake StudiesThese experiments were conducted byme in the lab of Dr. Urs Häfeli in theUBC Faculty ofPharmaceutical Sciences. They wereperformed using LCC6-HER2 cells — a human breastcancer cell line chosen for its overexpressionof the glucose transporter GLUT- LCC6-HER2cells were plated with Dulbecco’s ModifiedEagle Medium (D-MEM) (lx), liquid (high glucose)supplemented with 10 % Fetal Bovine Serum and1 % PenicillinlStreptomycin. The cellswereallowed to adhere at 37 °C in a humidifiedatmosphere containing 5 % CO2.The plating wascarried out in 75 cm2 tissue culture flasks with 0.2jim vented caps. The cultures were maintainedin a humidified 5 % CO2 atmosphere, with mediumchanges every alternate day. Subculturingwas carried out every 3-4 days using Trypsin-EDTA(0.25 % Trypsin with Na4EDTA) (1X)incubated for about 5 mm at 37 °C for cell detachment.A hemacytometer was used for countingthe cells to monitor cell proliferation.On the experiment day, suspensions of cells with concentration lxi06cells/mL were prepared in1 % PBS (pH 7.4). The suspension was aliquotedinto 1.5 mL Eppendorf vials to get a finalvolume of 0.5 mL in each vial (5xi05 cells). Thecompound to be tested was then added to thecells and the vial gently inverted. Each99mTCcomplex (15 jiCi) or‘8F-FDG (30 jiCi)was usedin 500 jiL PBS solution. Each compound was addedto the cells in glucose-free conditions, andin a final concentration of 5 mM D-glucose.The resulting suspensions were incubated at 37 °Cfor 30 mm with shaking at 400 RPM. Following incubation, thevials were centrifuged at 1100RPM for 5 mm then 900 j.tL ofthe supernatant was removed. Cold PBS solution (900 jiL) wasadded, and gentle mixing of the cells into the solution was achievedby gentle uptake and releaseof the solution from the pipette tip three times. This centrifugingand washing procedure wasrepeated four times in total to give the original supernatant andthree washing supernatants.Finally the cells were vortexed in cold PBS (1 mL)to remove them from the vial wall andtransferred to a tube for gamma counting. The activity in each cell samplewas divided by thesum of all the supernatants for that sample to give a percentageof the original amount of activitythat ended up associated with the cells. All experiments were carriedout in quadruplicate. FDGwas used as a positive control, and was repeated eachday that these experiments were carriedout to ensure experimental integrity.References begin on Pg 106 893.2.6 Hexokinase PhosphorylationStudiesA 5 mM aqueous solution of test compound (100j.tL) was added to a reaction vial such that afinal concentration of 0.5 mM test compound, 10 U/mLhexokinase, 1 mM ATP, 4 mM MgC12ina 30 mM TEOA (triethanolamine,pH = 6) buffer was achieved. The resulting solutionwasmixed gently by inverting each vialthree times, and then incubated for 16 hr at 37°C. Eachsolution (20 j.tL) was analysed by HPLCto determine whether any of the ATP had beenconverted to ADP, as would accompany thephosphorylation of a substrate. UsingaPhenomenex Synergi 4 jim Hydro-RP80 A column (250 x 4.6 mm) and eluting with 30 mMKJ-12P04,the retention time of ATP is8.5 mm, whereas that of ADP is 12.5 mm.3.3 Results and Discussion3.3.1 Synthesis, Characterisation and Complex Stability1OHHL2OHHO’\HOH\O:bHHL3HLHO\LHOOJFigure 3.13 Compounds made and studied in this chapter. HL3and HL4 were first made byMr. Chuck Ewart,43 and Dr. Neil Lim,44 respectively,and both syntheses were subsequentlyimproved and finalised by myself. L5 was supplied by Dr. TimStorr.36References begin on Pg 106 90The acetylated glucosamine startingmaterial 1 was made in four steps from commerciallyavailable glucosamine, and could be obtainedon a relatively large scale in moderate to goodyields.42 Dibenzylglycine was preparedfrom commercially available glycine in one step byanSN2 reaction with benzylbromide in the presence of base, followed by precipitation.4’Theamide coupling of these two protected compoundsto give 2 was initially performed byChuck Ewart,43 and subsequently investigatedand optimized by myself (Scheme 3.2). Anumber of different coupling agents were tried andnone were found to work better thanEDC, so the main improvement made inthis synthesis was the development of aprecipitation protocol which provided very cleanmaterial without the need for columnchromatography. The removal of the benzyl protectinggroups from the primary amine43 wasimproved in this work with the use of a Parr hydrogenationbomb. When working with up to- 0.5 g of starting material, 200 psi of H2 pressure(which can easily be obtained in thisapparatus) gave reaction times on the order of 1.5— 2 hr, as opposed to 12 — 20 hr if a H2balloon was employed. Due to the volume constraintsof the hydrogenation bomb, thisimproved deprotection was not always employed,but it did prove very useful and efficientfor small to medium scale reactions.The deprotected amine 3 was functionalized by additionof the metal binding groups in astepwise manner (Scheme 3.2). The pyridylarm had been utilized in compounds HL343 andHL4 (Figure 3.13). the major improvement to theirmethod for addition of the first bindingarm (via reductive amination to give 4),was the use of NaBH4 in place of NaBH(OAc)3.The overall reaction time for the formation of the secondary amine was reducedfrom up to 2days to 20 mm. NaBH(OAc)3is a milder reducing agent than NaBH4;it does not reducealdehydes and as such works in situ so the two step reductive aminationcan be carried out inone pot. Reduction with NaBH(OAc)3lead to the formation of the tertiaryamine, and wasinconveniently slow, so the use of NaBH4was investigated. Thechange of solvent from 1,2-dichloroethane to ethanol, as necessitatedby the solubilities of the particular reducingagents, resulted in the imine formation proceeding very rapidly,as evidenced by the nearlyinstantaneous colour change upon addition of aldehyde. TLC after5 mm was used toconfirm the absence of any starting materials, and excess NaBH4was added. As NaBH4reduces aldehydes as well as imines, it was thought that the reductionof any excess aldehydewould prevent the formation of the tertiary amine, and this was found tobe the case. Asadequate yields were obtained via this method, it was not necessary to isolate the iminepriorReferences begin on Pg 106 91to reduction. The whole reductive aminationcould be performed in one pot in a very shortperiod of time. This methodolgy wasapplied to the synthesis of compound6, the methylimidazole analogue, and was found to worksuccessfully in this case too.The second chelating arm was installedby either a second reductive animation ofsalicylaldehyde to give compound 7,or an alkylation to add a protected acetate armtocompound 5 (Scheme 3.2). As the reductive aminationin this case was being used to make atertiary amine from a secondary amine,there was no danger of overalkylation; NaBH(OAc)3was the reducing agent of choice as it could reducethe imminium ion formed in situ. Thealkylation of the secondary amine viaSN2 attack on methylbromoacetate proceeded asexpected to give compound 5.The methyl protecting group on the acetate in compound5 was chosen because it could beremoved by NaOH, which would allow for concurrentdeprotection of this group and the fouracetyl groups on the sugar hydroxyls. This wasfound to be the case, and HL1 was preparedfrom 5 by vigorous stirring in 1M NaOH. TLC and MS analysis showed that this reactionproceeded in very high yield. As theproduct of this reaction is very polar, reverse-phaseHPLC (RP-HPLC) was needed to purifythe resulting oil, and this dramatically lowered theoverall isolated yield of the reaction.To produce HL2 from 7 required onlythe deprotection of the glucosamine acetyl groups,which was achieved in good yield using NaOMein MeOH. Although HL2 is not quite ashydrophillic as HL’, it still required purificationby RP-HPLC, which again resulted in theisolated yield being much lower than thecrude yield.References begin on Pg 106 92OAcAcO\AcO----OAcNH2IglycineBnEDC,DMAP,DMFOAc0AcOAc0—.\OAcNH2Ozi/Pd(OH)2OAcH2AcO\AcO-—-._- OAc“NH2\\2-methylimidazolecarboxaldehyde,Na2CO3,EtOH, NaBH4OAc0AcO\oAc\MethyibromoacetateNa2CO3,DCM2-pyridinecarboxaldehyde,Na2CO3,EtOH, NaBH4///OAcAco\oM5_/O—0/H0\oHO,jHL1OH61 M NaOHNaOMe, MeOH7HO\M(CO)3 /OHM= 99mTCNaQEt,HL275°C, 40 mmM = Re, reflux, 8 hr OH\NHNM(CO)r0ML2M (CO)3M= 99mTC75 °C, 30 mmYoF-I M=Re,refiux,2hrHO\OHML1M(CO)3Scheme 3.2 The syntheses of HL1,HL2 and theirmetal complexes.References begin on Pg 106 93Complexation of either HL1 or HL2 to therhenium tricarbonyl core proceeded in astraightforward manner by refluxing in methanol.The ligand containing the acidfunctionality bound to the metal core more quicklythan the phenolate ligand HL2,presumably due to the differing protonation states of the two ligands. Thesecomplexes wereagain purified by RP-HPLC, as the free glucose makesit very difficult to purify them onsilica, by precipitation or by crystallization.The ligands HL’ and HL2,their rhenium complexes,and most chemical intermediates in thesyntheses were characterized by HR-MS, IR, ‘H and‘3C NMR spectroscopy. A variety of 2-D NMR techniques such as COSY, HMQC, HMBCand TOCSY were used, as appropriate,for full assignment of ‘H and ‘3C NMR signals.HR-MS gave the expected peaks in each case for either M+H andJor M+Na,verifying thepresence of the target compound. The rhenium complexes exhibited thecorrect isotopepattern (‘85Re = 37 % and ‘87Re = 63 % natural abundance)to verify the presence of onerhenium. In a typical example, the HR-MS of compoundHL’ shows peaks at 676.0610 (—55% intensity) and 678.0704 (100% intensity) corresponding to C,gH22N3O,1185ReNa andC,9H22N3011‘87ReNa, respectively.Infrared (IR) spectroscopy was used to verify the presence of the targetcompound byidentification of key peaks corresponding to certain functional groupsin that molecule. Thedeprotection of the acetyl groups on the sugars to reveal the proligands canbe seen in the loss ofthe acetyl peak in the JR. The tetraacetyled proligand (5) had a strong peak at 1745cm1 (Figure3.14), characteristic of ester groups which disappearedin the spectrum of HL1,while strongbroad peaks at 3272 cm’ and 2918cm’ grew in, characteristic of the free hydroxyl groupsrevealed by the deacetylation (Figure 3.14). Deprotection of compound7 to give HL2 causedthe disappearance of a strong ester band at 1744 cm1 accompanied by the appearanceof bandscorresponding to the —OH stretches of the acid and the hydroxyl groupsat 3293 cm1 and 2942cm1.References begin on Pg 10610090 -80 -70%T60504030 -4000Figure 3.14 JR spectrophotometric traces of compound 5 (in black, starting at 98%transmittance in the top left corner) and its deacetylated analogueHL1 (in gray, starting at 100 %transmittance in the top left corner).JR spectrophotometry can also be very diagnostic of metal binding,so the spectra of theproligands and their metal complexes were compared. Figure 3.15 shows the overlaid spectra ofHL2 and its rhenium tricarbonyl complex ReL2. The decrease in relative intensity of the broadpeaks at high wavenumber (around 3000 cm’) is due to the coordination of the acid functionalgroup to rhenium, and the small peaks at similar wavenumber in the spectrum of ReL2 are due tothe free —OH groups of the deprotected carbohydrate. The rhenium complex shows several morestrong peaks than the corresponding ligand ilL2. The peaks at 2028 and 1884 cm’ are due to the— + -1ReC=O bonds. In [Re(H20)3(CO)]these peaks appear around 2000 and 1868 cmrespectively, and a shift to higher wavenumber as observed here is characteristic of metalbinding in complexes of this type.27 The carbonyl peak appears at 1671 cm1 in the proligandNi2 ahd at 1672 cm1 in the spectrum of the rhenium complex ReL2. As this peak correspondsto the amide at C-2, it is not close enough to the metal coordinating portion of the molecule to besignificantly affected by binding to the metal. This supports the metal binding of the desiredcoordination sphere and the pendant nature of the carbohydrate.3500 3000 2500 2000 1500 1000 500Wavenumbers (cm1)References begin on Pg 106 951009080%T7060504000Figure 3.15 Overlaid IR spectra of HL2 (in black, beginningat 100 % transmittance at topleft) and ReL2 (in gray, beginning at 97 % transmittanceat top left).Similar observations can be made when comparingproligand HL1 with its rhenium complex,ReL1. Proligand peaks in the 2800 - 3300 cm1 range decreased in relativeintensity uponmetal coordination as the acid functionaility became boundto the metal centre. Peakscorresponding to the ReCO bonds appeared at 2020 and 1910 cm’.The carbonyl peakshifted upon metal coordination; appearing at 1667 cm’ inthe proligand and 1635 cm1 inthe complex. This is because this peak corresponds to the acid moiety, whichis very close tothe metal, so would be expected to be affected by the coordination.This large shift supportsthe binding of the acid group to the metal.‘H and ‘3C NMR spectroscopy were used to further probe the identity of each compoundmade, as well as the coordination modes of the final complexes ReL’ and ReL2. The spectraof the chemical intermediates and the protected proligands 5 and 7 were unremarkable. Allwere rigorously assigned to ensure the identity of each compound. Deprotectionof 5 and 73500 3000 2500 2000 15001000 500Wavenumbers (cm1)References begin on Pg 106 96to give proligands HL’ and HL2,respectively,led to more complicated NMR spectra. Thisis typical of sugar compounds where allthe free hydroxyl groups lead the hydrogens on thepyranose ring all to be in similar chemical environments,especially those at C-3-C-5. Theseparts of both the ‘H and ‘3C NMR spectra are difficultto fully assign and interpret, as can beseen in Figure 3.16, where 3.16a is the ‘H NMRspectrum of the protected 7, and and 3.16bis that of HL2. Comparing the two, it is easy to identifythe large upfield shift and narrowingtogether of the sugar peaks. However, the aromaticpeaks remain very diagnostic, and theanomeric protons at the C-i position (5.1 and 4.7ppm) confirm the presence of a non-protected glucopyranose ring. The same trends areobserved when comparing the111NMRspectra of protected 5 and HL1,where the aromaticregion remains essentially unchangedand the sugar peaks all undergo large shifts andnarrowing with the Cl anomeric peaks beingthe only ones still clearly defined.Figure 3.16c shows the ‘H NMR spectrum of ReL2.Comparing this with Figure 3.16.b(HL2) there are a number of changes that pointtoward the desired coordination mode.Firstly, the largest shifts seen are for the aromaticprotons, which have collectively spreadout and shifted downfield. These large shifts are indicative of metalbinding occurring closeto these hydrogen atoms. The peaks between 3 - 5ppm have also shifted somewhat, thoughdue to the overlapping it is difficult to tell exactly which peak has moved where.The onlysugar peak that can be clearly seen is that of C-lct proton whichappears at 5.1 ppm in bothcompounds. That it does not shift upon coordination means thatits chemical environmenthas not changed significantly, and therefore that it is not close to thesite of metal binding.The same trends are observed upon comparison of the ‘H NMR spectraof HL1 and ReL1.The aromatic peaks spread and move downfield while the region between3 - 5 ppm does notundergo such large changes, though again this is difficultto interpret in detail.Again, the Clct peak is clearly seen andis found not to move significantly upon metalcoordination.References begin on Pg 106 9712 OCOCH3aOAc*AcO’215,10,1513j5N19,141T26b1017’\\2119418197183OHb *122014regon7NH \*19,1315OHN17c”218,1711113OH19JLJL LiP..C6L5* *HOH127NH_JN131413/- N. , sugar16LQRe(CO)319regionl7(Y2120Ila.__U10.5 10 9.5 9 8.5 8 7.5 7 6.5 6 5.5 5 4.5 4 3.53 2.5 2 1.5ppmFigure 3.16 ‘H NMR spectra of a) 7 b) HL2 c) ReL2,*shows the residual solvent peaks.‘3C NMR spectroscopy reinforces the conclusions drawn from analysis of the ‘H NMR data.Addition of the binding arms to build up HL1 and HL2in a stepwise manner showedproduction of the correct number of new peaks in the expectedrange. Deprotection of 5and 7 to reveal the proligands HL1 and HL2 led to several changes in the ‘3C NMR spectra.These are illustrated by representative traces in Figure 3.17a and b, the ‘3C APT spectra of 7References begin on Pg 106 98and HL2 respectively. These spectrawere recorded in different solvents for solubilityreasons, so the exact chemical chifts arenot comparable, although the trends still exist.Onemajor change upon deprotection is thedisappearance of the acetyl peaks inthe regionsaround 170 ppm and 20 ppm, anda downfield shift of the peaks corresponding to C-i -C-6ofthe carbohydrate. Due to the productionof two anomers, the spectrum becomes much morecomplicated, with two peaks visible formany carbon atoms. As the anomers differ intheirstereochemistry only at the C-i position, the closeran atom is to this centre the more likely itis to be split, and the larger the gap between thetwo peaks. For example, the two C-2 peaksin HL’ are 2.7 ppm apart, as they are adjacent tothe epimeric centers, whereas the aromaticcarbons C- 12-C-i 5 in the same moleculeeach show up as only one distinguishable peak,asthey are fairly well removed from the amomericcentre. These features are all wellillustrated in the examples shown in Figure 3.17a andb.Coordination of the proligands to [Re(CO)3jresultsin several coordination induced shifts(CIS). A major feature of the spectra of themetal complexes is the appearance of threepeaks due to the carbon atoms on the CO ligandsaround 194 ppm (see Figure 3.17c), anormal chemical shift for compounds ofthis type.27 Similar to what is seen in the ‘H NMRspectra of these complexes, the aromatic signals move significantlycompared to the signalscoming from the rest of the molecule. Somearomatic signals shift very little, and others upto about 10 ppm either upfield or downfield,depending on that carbon’s position withrespect to the ligating atom. These shiftscan be observed by direct comparison of Figures3.i7b and c. It can also be observed that the sugar peaksdo not shift significantly from thoseof the free ligand.Taken together this characterization information is all consistent withthe structures putforward and expected to form under the given reaction conditions.References begin on Pg 106 99OAc0Ao04\ 2\7NHa8,OAc16 * 915N16L/OH216,191313,54212OHOAcbi5 0HO4\\OH 1261516*(15N8 16\,OH11l72121_________fr&Lrf9-wwi !T”20210412C181192OH131’L5o1217 7NH\21 *811161615co 17/\21IPIuI I’;:17141201,191’21218133’,4’,5j200 190 180 170 160 150 140 130 120 110 10090 80 70 60 50 40 30 20 10 0ppmFigure 3.17 ‘3C APT spectra of a) 7 b) HL2 c) ReL2,*marks the residual solvent peaks.References begin on Pg 106 100As the rhenium complexes formedwell, it was expected that formation of the99mTccomplexes would also proceed smoothly. Thiswas found to be the case, and highradiochemical yields were achieved(Table 3.1). As was observed with rhenium complexformation, the acid-containing ligandHL1 gave faster coordination and higher yields. It alsoproceeded under milder conditions — an aqueoussolution buffered to pH 7.4 gave excellentyields. At this pH all the acid groupsare deprotonated and therefore ready to bind to themetal ion. The phenolate-based proligand HL2was deprotonated with NaOEt prior toaddition of the99mTCprecursor. This enabled good labelling yields in30 — 40 mm at 75 °C.Table 3.1 Summary of labelling and stability resultsfor the two novel99mTccomplexesmade in this chapter.Stability in Stability inRT (Re RT(99mTCRadiochemicalLigandcysteine — histidine —compound)* compound)*yield24hr 24hrHL1 12.9mm 13.2mm 94% 94±2% 85±3%HL2 19.6mm 19.4mm 96% 100±4% 94±3%*HPLC conditions — 100 % H20 (with 0.1 % TFA) linear gradient to100 % ACN at 30 mm.The99mTccomplexes were analyzed using RP-HPLC. As discussed in Chapter1.4, this isstandard practice when performing radiochemistryas the radioactivity and very smallamounts involved preclude the use of conventional chemistrycharacterization techniques.The thoroughly characterized rhenium complex was coinjectedwith the99mTCcomplex andagreement of their retention times taken to mean that the technetiumanalogue is of the samechemical structure as the rhenium complex. There is normallya slight difference betweenthe two retention times because 1) the UV (rhenium) and radiation(99mTc)detectors are inseries, and 2) technetium and rhenium are not exactly the same,so their small differencesmay be enough to affect their relative retention times. Percentage radiochemicalyield iscalculated by integration of all the peaks that appear in the radiation trace. An exampleof aradioactive HPLC trace of99mTCL1is shown in Figure 3.18.All99mTccomplexes were tested for stability in excess amounts of cysteineand histidine forReferences begin on Pg 106 10124 hr at 37 °C. The results are shown in Table 3.1.Cysteine and histidine are strong ligandsfor the tricarbonyl core, and are found ubiquitouslyin vivo. This method of measuringcomplex stability is routinely used in radiochemistryas an indicator of in vivo applicabilityof the given chelating system. The compounds shownhere both exhibit satisfactory stabilityto allow for their further investigation as molecular imaging agents.o i 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25Time (mm)Figure 3.18 Radiation trace from HPLC of99mTCLl.Retention time is around 13 mm, withthe small peak at 16 mm due to unreacted starting material. The x-axis is arbitrary unitscorresponding to amount of radiation detected.3.3.2 Cell Uptake StudiesCell uptake of several99mTccomplexes was examined in LCC6-HER2 cells — a human breastcancer cell line with high concentrations of GLUT-i glucose transporters on its surface.45The uptake was studied in the presence and absence of glucose to help elucidate themechanism of any cell uptake observed — if uptake was foundto be inhibited by the presenceof high concentrations of glucose it would mean that the test compound and glucose weresharing the same cell uptake pathways. Figure 3.19 shows the results of this assay. It isclear from this graph that the cell uptake of each of the99mTccompounds examined isindependent of glucose. This means that the uptake seen is not due to the compounds beingtransported along the GLUT-i transporters. It can also be seen that the actual percentageuptake for each of the compounds is quite low. For comparison, FDG, the positive control,showed uptake of around 3.4 % of the total activity when in a glucose free environment, andonly 0.1 % when 10 mM glucose was present (results not shown for graphical clarity). ThisReferences begin on Pg 106 102illustrates both the amount of uptake that canbe achieved by utilizing the GLUT-itransporters, and the strong effect that 10 mMglucose will exhibit on a compoundtransported by such a mechanism.0.12% -0.08%0.04%0.00%Figure 3.19 Results of cell uptake studies. Each compounds’ uptakewas examined in aglucose free environment and in 10 mM glucose solution. Average% of total activity in thecells after a 30 mm incubation for four independent experiments are shown- error bars arestandard deviations of the four.Five technetium compounds (Figure 3.13) were tested in thisassay. There were four neutralcompounds99mTcL4,and the cationic compound99mTcL5.Given that less than 0.1 % ofeach of these compounds was taken up into cells after 30 mm, it can be concludedthat noneof these compounds have any significant cellular uptake via GLUT-l or any othermechanism. The overall trend is that lipophilic compounds have higheruptake, thoughagain given the small percentages involved this is not a particularly robustconclusion. Thecationic compound did exhibit an uptake lower than some ofthe neutral complexes, but hadhigher uptake than the neutral99mTCL1.In a practical sense the exact mechanism ofinteraction is not of great importance, as the size of the uptakes observed mean thatcompounds of this type will not be of use as molecular imaging agents,as they will not bej: glucose free10mM glucoseReferences begin on Pg 106 103able to enter cells in useful amounts.3.3.3 Hexokinase Phosphorylation StudiesAs discussed in Chapter 1.3, hexokinase phosphorylatesglucose in the C-6 position, and isalso known to phosphorylate certain glucoseanalogues. It was of interest to see if thesecompounds would be phosphorylated by hexokinase.This is an important property to beassessed when looking for a glucose based imaging agent,as without phosphorylation tomake the compound negatively charged, it can diffuse outof a cell just as quickly as itdiffused in. This would mean there would be no significant enhancement of signal incellsthat overexpress hexokinase, regardless of GLUTexpression and cell uptake.Unfortunately none of the compounds tested were foundto be phosphorylated by hexokinase.This may be because the metal-binding portions of themolecules are too large and bulky toallow the cleft in hexokinase to narrow as is required forphosphorylation to occur. Trying tocircumvent this likely cause of the activity loss ledto the idea of making long-chainglucosamine conjugates. The idea behind these is that the metal-binding portionof themolecule is held at a large enough distance from the active site that the closing of the cleftcan occur without being blocked by the metal ion and chelate.This is the reasoning behindthe long-chain compounds discussed in Chapters 4 and 5 of this thesis.3.4 ConclusionThe phenolate ligand system is one that has been developed by theOrvig group relativelyrecently,’7’27and to our knowledge we are the only group to utilize this functionality as aligand for the tricarbonyl core. This group was investigated as an alternative anionic bindinggroup to provide a neutral complex when bound to the monocationic tricarbonylcore. Thefunctional group normally used for this purpose in[99mTc(CO)3]+chemistry is the carboxylicacid. We were interested in exploring different possibilities for binding groups andwideningthe cordination chemistry knowledge of this important radiochemistry core. It has beendetermined27that the phenolate group is indeed a suitable donor for the tricarbonyl core,as itbinds to both the technetium and rhenium centres within a radiochemistry-appropriatetimeframe and gives complexes stable to cysteine and histidine challenges. As the pKa ofReferences begin on Pg 106 104the phenol is around 10, the reaction mixtureneeds to be made basic to enable sufficientligand deprotonation for complex formation in a timely manner. In oursystems thisproceeded well and with no sign of degradationproducts, but these conditions may limit itsgeneral applicability. This potential drawback maybe outweighed, however, by theincreased lipophilicity of a phenol compared toan acid group, which in certain cases may beable to improve the biodistribution and solubilitiesof the resulting compound. As thesebiological properties can be very sensitive to small changes,this is a very useful addition tothe expanding knowledge on the coordination chemistryof technetium.The practical application of the compounds made in this chapter, whetherbased onphenolates or other ligand types, is not as we had initially hoped. To be a contender as amolecular imaging agent, a compound has to be taken upby and trapped in cells thatoverexpress the enzymes of interest. In the case of carbohydrate based imaging agents thoseenzymes are hexokinase and the GLUT family. The in vitro assays performedon thecompounds in this chapter show no substrate activity either in terms of cell uptake viaGLUT-i or phosphorylation by hexokinase. The reason for these results is that thesecompounds are too different from glucose for the very specific enzymes to recognize andprocess the compounds as they would their native substrates. Figure 3.1 shows glucose andFDG, where the similarities between the two are very apparent. When trying to utilize ametallic radionuclide, it is not possible to make such small changes, as a metal binding groupmust always be added. But in this case at least, that perturbation is too large to allow for afunctional imaging agent to be formed.References begin on Pg 106 1053.5 References1. Kubota, K., Ann. Nuci. Med. 2001, 15, 471-486.2. Herholz, K.; Salmon, E.; Perani, D.; Baron,J.-C.; Hoithoff, V.; Frolich, L.;Schonknecht, P.; Ito, K.; Mielke, R.; Kalbe, E.; Zundorf,G.; Delbeuck, X.; Pelati, 0.;Anchisi, D.; Fazio, F.; Kerrouche, N.; Desgranges,B.; Eustache, F.; BeuthienBaumann, B.; Menzel, C.; Schroder, J.; Kato, T.; Arahata, Y.;Henze, M.; Heiss, W.D., Neurolmage 2002, 17, 302-316.3. Adam, M. J., I Label. Compd. Radiopharm.2002, 45, 167 - 180.4. Adam, M. J.; Wilbur, D. S., Chem. Soc. Rev.2005, 34, 153 - 163.5. Risch, V. R.; Honda, T.; Heindel, N. D.; Emrich,J. L.; Brady, L. W., Radiology 1977,124, 837 - 838.6. Caner, B. E.; Ercan, M. T.; Bekdik, C. F.; Varoglu,E.; Muezzinoglu, S.; Duman, Y.;Erbengi, G. F., NuklearMedizin 1991, 30, 132 - 136.7. Ozker, K.; Collier, B. D.; Lindner, D. J.; Kabasakal,L.; Liu, Y.; Krasnow, A. Z.;Heliman, R. S.; Edwards, S. D.; Bourque,C. R.; Crane, P. D., Nuci. Med. Commun.1999, 20, 1055 - 1058.8. Oh, S. J.; Ryu, J.-S.; Yoon, E.-J.; Bae, M.S.; Choi, S. J.; Park, K. B.; Moon, D. H.,Appi. Radiat. Isot. 2006, 64, 207 - 215.9. Yang, D. J.; Kim, C.-G.; Schechter, N. R.; Azhdarinia, A.;Yu, D.-F.; Oh, C.-S.;Bryant, J. L.; Won, J.-J.; Kim, E. E.; Podoloff, D. A., Radiology2003, 226, 465 - 473.10. Chen, Y.; Huang, Z. W.; He, L.; Zheng,S. L.; Li, J. L.; Qin, D. L., Appi. Radiat. Isot.2006, 64, 342 - 347.11. Yang, D. J.; Yukihiro, M.; Yu, D. F.; Ito,M.; Oh, C.-S.; Kohanim, S.; Azhdarinia, A.;Kim, C.-G.; Bryant, J. L.; Kim, E. E.; Podoloff, D. A., Cancer Biother. Radio.2004,19, 443 - 456.12. Zuckier, L. S.; Dohan, 0.; Li, Y.; Chang, C. J.; Carrasco, N.; Dadachova,E., I Nuci.Med. 2004, 45, 500 - 507.13. Chen, Y.; Xiong,Q.;Yang, X.; Huang, Z.; Zhao, Y.; He, L., Cancer Biother. Radio.2007, 22, 403 - 405.14. Chen, Y.; Xiong,Q.;Yang, X.; Huang, Z.; He, L., Cancer Biother. Radio. 2007, 22,400 - 402.15. Chen, X.; Li, L.; Liu, F.; Liu, B., Bioorg. Med. Chem. Lett. 2006, 16,5503 - 5506.16. Alberto, R.; Schibli, R.; Elgi, A.; Schubiger, P. A., I Am. Chem.Soc. 1998, 120,7987 - 7988.17. Bayly, S. R.; Fisher, C. L.; Storr, T.; Adam,M. J.; Orvig, C., Bioconjugate Chem.2004, 15, 923 - 926.18. Ferreira, C. L.; Bayly, S. R.; Green, D. E.; Storr,T.; Barta, C. A.; Steele, J.; Adam,106M. J.; Orvig, C., Bioconjugate Chem. 2006, 17, 1321 - 1329.19. Storr, T.; Obata, M.; Fisher, C. L.; Bayly,S. R.; Green, D. E.; Brudzinska, I.; Mikata,Y.; Patrick, B. 0.; Adam, M. J.; Yano, S.; Orvig, C., Chem. - Eur.J 2005, 11, 195 -203.20. Mikata, Y.; Shinohara, Y.; Yoneda, K.; Nakamura, Y.; Esaki, K.; Tanahashi,M.;Brudzinska, I.; Hirohara, S.; Yokotama, M.; Mogami, K.; Tanase, T.; Kitayama,T.;Takashiba, K.; Nabeshima, K.; Takagi, R.; Takatani, M.; Okamoto, T.; Kinoshita,I.;Doe, M.; Hamazawa, A.; Morita, M.; Nishida, F.; Sakakibara, T.; Orvig,C.; Yano, S.,I Org. Chem. 2001, 66, 3783 - 3789.21. Pak, J. K.; Benny, P.; Spingler, B.; Ortner, K.; Alberto, R., Chem. - Eur.1 2003, 9,2053-2061.22. Gottschaldt, M.; Koth, D.; Muller, D.; Klettte, I.; Rau, S.; Goris, H.; Baum, R. P.;Yano, S., Chem. - Eur. 1 2007, 13, 10273 - 10280.23. Schibli, R.; La Bella, R.; Alberto, R.; Garcia-Garayoa, E.; Ortner, K.; Abram,U.;Schubiger, P. A., Bioconjugate Chem. 2000,11, 345 - 351.24. Alberto, R.; Pak, J. K.; van Staveren, D.; Mundwiler, S.; Benny,P., Biopolymers2004, 76, 324 - 333.25. Lipowska, M.; Ci, R.; Tamasi, G.; Xu, X.; Taylor, A. T.; Marzilli, L.G., Inorg.Chem. 2004, 43, 7774 - 7783.26. Safi, B.; Mertens, J.; De Proft, F.; Geerlings, P., 1 Phys. Chem.A 2006, 110, 9240 -9246.27. Lim, N. C.; Ewart, C. B.; Bowen, M. L.; Ferreira, C. L.; Barta, C. A.; Adam, M. J.;Orvig, C., Inorg. Chem. 2008, 47, 1337 - 1345.28. Chiotellis, A.; Tsoukalas, C.; Pelecanou, M.; Raptopoulou, C.; Terzis, A.;Papadopoulos, M.; Papadopoulou-Daifoti, Z.; Pirmettis, I., Inorg. Chem. 2008, 47,2601 - 2607.29. Petrig, J.; Schibli, R.; Dumas, C.; Alberto, R.; Schubiger, P. A., Chem. - Eur. 1 2001,7, 1868 - 1873.30. Dumas, C.; Petrig, I.; Frei, L.; Spingler, B.; Schibli, R., Bioconjugate Chem. 2005,16, 421 - 428.31. Schibli, R.; Dumas, C.; Petrig, J.; Spadola, L.; Scapozza, L.; Garcia-Garayoa, E.;Schubiger, P. A., Bioconjugate Chem. 2005, 16, 105 - 112.32. Dumas, C.; Schibli, R.; Schubiger, P. A., I Org. Chem. 2003, 68, 512 - 518.33. Yang,Q.;Lundahi, P., Biochemistry 1995,34,7289-7294.34. Mikata, Y.; Sugai, Y.; Yano, S., Inorg. Chem. 2004, 43, 4778 - 4780.35. Storr, T.; Sugai, Y.; Barta, C. A.; Mikata, Y.; Adam, M. J.; Yano, S.; Orvig, C.,Inorg. Chem. 2005, 44, 2698 - 2705.36. Storr, T.; Fisher, C. L.; Mikata, Y.; Yano, S.; Adam, M. J.; Orvig, C., Dalton Trans.2005, 654 - 655.10737. Ferreira, C. L.; Marques, F.; Okamoto, M.R. Y.; Otake, A. H.; Sugai, Y.; Mikata, Y.;Storr, T.; Bowen, M. L.; Yano, S.; Adam, M. J.; Chammas, R.;Orvig, C., submittedfor publication.38. Banerjee, S. R.; Babich, J. W.; Zubieta, J.,Inorg. Chim. Acta 2006, 359, 1603 - 1612.39. Mindt, T. L.; Struthers, H.; Brans, L.; Anguelov,T.; Schweinsberg, C.; Maes, V.;Tourwe, D.; Schibli, R., I Am.Chem. Soc. 2006, 128, 15096 - 15097.40. Banerjee, S. R.; Levadala, M. K.; Lazarova,N.; Wei, L.; Valliant, J. F.; Stephenson,K. A.; Babich, J. W.; Maresca, K. P.; Zubieta, J., Inorg. Chem. 2002,41, 6417 - 6425.41. Breitenmoser, R. A.; Heimgartner, H., Helv.Chim. Acta 2001, 84, 786 -796.42. Silva, D. J.; Wang, H.; Allanson, N. M.; Jam,R. K.; Sofia, M. J., I Org. Chem. 1999,64, 5926 - 5929.43. Ewart, C. B. M.Sc. Thesis. University of British Columbia, Vancouver,2006.44. Lim, N. C.; Adam, M. J.; Orvig,C., unpublished results.45. Dragowska, W. H.; Ruth, T. J.; Adam, M. J.; Kozlowski, P.;Skov, K.; Bally, M. B.;Yapp, D. T. T. In Studies of Tumor Microenvironment and Metabolic ActivityinHER-2/neu Overexpressing Breast Cancer Xenografts by MicroPETand MRLAmerican Association for Cancer Research, 2005,p900.108CHAPTER 499mTc..Labellingand In Vitro Assays of Long Chain Glucosamine-basedTridentateLigands*4.1 Introduction4.1.1 Long Chain Bioconjugates of[99mTc(CO)3]+The effect of linker length on phosphorylation ofa substrate has been observed in non-sugarsystems.’ Thymidine kinases are proteins responsible for the phosphorylationof thymidine(Figure 4.1 a), forming a nucleotide from the nucleosideby addition of a phosphate in the 5’position.’ As these kinases are overexpressed in sometypes of cancer, they are of interest inmolecular imaging in oncology. A series of compounds have been synthesizedwith a carbonchain between the tridentate technetium binding sphere and the thymidinemoiety being varied inlength between two and ten carbons.1 Rates of phosphorylation of thesecompounds bythymidine kinase were tested relative to the native substrate thymidine, and rates werefound toincrease with increasing chain length.’ The compound with a C,0 linker shownin Figure 4.1 hasa rate of phosphorylation of 71.5 (± 4.7) % that of thymidine, the native substrate. This is a verysignificant percentage given the large change in molecular structure between the twocompounds, as illustrated in Figure 4.1 a and b. The thymidine conjugates synthesized were alsoexamined for their cell internalization properties.’ The authors were curious as to whether therewould be any nucleoside transporter mediated cell uptake. This was found not tobe the case;there was a strong correlation between lipophilicity and cell uptake, indicating passive diffusionas the likely mechanism of transportation.’An amino acid-conjugate (Figure 4.1 d) was synthesized and examined for cell uptake via theLAT-1 amino acid transporter.2 LAT-1 transports lipophilic amino acids and their analogues,and is found in high concentrations at the blood brain barrier (BBB).3 Brain delivery of drugsand imaging agents is a large problem in medicine as the BBB excludes most foreign materials.4Uptake studies were performed on a range of compounds, and though most analogues showed no*A version of this chapter will be submitted for publication: Bowen, M. L., Chen, Z.-F., Roos, A., Adam, M. J. andOrvig, C. Synthesis and Characterization of Organometallic Rhenium and Technetium Glucose Complexes againstHexokinase.References begin on Pg 127 109or poor activity, the analogue in Figure 4. id showed K = 308pM.2 This is the first time that ametal-labelled compound has shown transport activitywith the LAT- 1 transporter? Althoughthere is some substrate scope accepted by this enzyme, the activity quickly drops if certain smallmodifications are made.2a o cOHdNH N—NH2H\Jcc”HgFigure 4.1 a) Thymidine b) a long chain (C10 linker) thymidine conjugate’ c) a generic aminoacid and d) long chain (C5 linker) amino acid derivative;2these bioconjugates retain significantbiological recognition of the parent compound in relevant assays, M = Re,99mTc.4.1.2 Long Chain Carbohydrate Conjugates of(99mTc(CO)31+There has only been one previous study, to our knowledge, investigating the potential merits oflong chain carbohydrate conjugates as molecular imaging agents.5 This included aninvestigation of phosphorylation by hexokinase and transport by GLUT-i of the compoundssynthesized,5and was discussed briefly in Chapter 1. The two long chain compounds in thatwork that are relevant to this Chapter will be examined in more detail here (Figure 4.2).HV0 COOH/0\(SN,/—COOH—COOHCOOHFigure 4.2 Two long chain glucose-based ligands for binding to the tricarbonyl core.5The cellular uptake of the99mTClabelled compounds (Figure 4.2) was found to parallel theirlipophilicity.5 The uptakes were also found to be independent of the concentration of both theReferences begin on Pg 127 110test compound and of cytochalasin B, a known inhibitorof GLUT-i This information points tothe cellular uptake being due to unspecific passive diffusion rather thana GLUT-mediateduptake mechanism. The authors also modeled the interactionof these compounds with the activesite of hexokinase.5 Their rationale for synthesizing the long chain analogueswas that thiswould allow for the carbohydrate to be in the active site of hexokinase, while the bulky metaltricarbonyl core with its tridentate chelate would be held far away from the activesite. Becausethe active site of hexokinase is in a cleft which has to narrow in order for phosphorylation tooccur,6having a bulky obstacle in the way may block the cleft from closing, andtherefore thephosphate from transferring. In silico it appeared that the length of linker requiredto preventthis blocking was seven methylene units.5 A hexokinase inhibition assayshowed that twocompounds, those with long alkyl chains (Figure 4.2), were strong inhibitorsof hexokinase - K0.25 — 5.8 mM.5 The shorter chain compounds did not exhibit any inhibitorycharacter.5 Itwas found that even the long chain compounds, which have nine atoms in their linker chains,were not themselves phosphorylated by hexokinase.5The Orvig group has always been interested in glucosamine derivatives of these types ofcompounds due to the higher tolerances in structural variation of glucosamine analoguescompared to glucose analogues exhibited by hexokinase and GLUTs.7’8The extracellularbinding of substrate to GLUT-i is known to depend on interactions with the hydroxyl groups atC-i, C-2 and C-3.9 When the hydroxyl group at C-2 of glucose is substituted, it loses its abilityto hydrogen bond as the 0 no longer has a H attached. In contrast, when the N at the C-2 ofglucosamine is monosubstituted, there is still one hydrogen attached to the nitrogen, so it is stillable to hydrogen bond. For example, 2-NBDG (Figure 1 .4c), with a dye appended directly to thesecondary nitrogen at the C-2 of glucosamine is found to be phosphorylated by hexokinase.’°The same compound exhibits cell uptake properties characteristic of transport via GLUTs - theuptake is inhibited by both glucose and cytochalasin B.’°The work in this chapter exploits these properties of functionalized glucosamines and explorestheir abilities to accommodate long chain99mTclabelled moieties and still retain certain enzymeactivities. This chapter examines the effect of chain length on the in vitro properties of rheniumand technetium tricarbonyl complexes, while keeping the metal-binding and carbohydrateportions of the molecules constant.References begin on Pg 127 1114.2 Experimental4.2.1 Instruments and MaterialsAll solvents and chemicals were reagent grade and wereused as received unless statedotherwise. Solvents were HPLC grade, and were purchasedfrom Fisher Scientific.IsolinkTMkits were provided by Mallinckrodt Inc. (now Covidien).Na99mTcO4was provided by theNuclear Medicine Department at the University of British Columbia Hospital. The C6,C8, andCli ligands and their Re complexes (Figure 4.3) were first madeby Dr. Zhen-Feng Chen undermy guidance.” The MTT cell toxicity assay and hexokinase inhibitionstudies of the C6 and C8compounds were performed by an undergraduate summer student,Adrienne Roos, under mysupervision.The analytical TLC plates, which were aluminum backed ultrapure silica gel 60, 250 jim, andthe flash column silica gel (standard grade, 60 A, 32-63 mm) used were provided by Silicycle.HPLC analyzis of non-radioactive compounds was done on a Phenomenex Synergi4 jim HydroRP 80 A column (250 x 4.6 mm) in a Waters WE 600 HPLC system equipped with a 2478 dualwavelength absorbance UV detector run controlled by Empower softwarepackage. HPLCanalyzes of radiolabelled complexes were performed on a Knauer Wellchrom K-i001 HPLCequipped with a K-2501 absorption detector and a Capintec radiometric well counter using aPhenomenex Synergi 4 jim Hydro-RP 80 A column (250 x 4.6 mm).4.2.299mTechnetiumComplex FormationNa99mTcO4(200 MBq) was obtained as a saline solution from the University of British ColumbiaHospital, Department of Nuclear Medicine. It was added to an IsolinkTM kit (Mallinckrodt) andthe volume made up to 1 mL with 0.9 % saline solution. The resulting solution was heated at90 °C for 30 mm. After cooling, the solution was neutralized with 1 M HC1 (0.12 mL) to pH 7as tested with pH paper. Meanwhile, 1 0 M ethanolic solutions of the ligands were prepared,and 0.5 mL of each was transferred to a reaction vial with three equivalents of NaOEt (0.15 mLof a i0 M solution in ethanol). The vial was purged with nitrogen for 10 mm to provide aninert atmosphere and to ensure time for deprotonation of the phenolates. Between 0.1 and 1 mLof[99mTC(CO)3(H20)]+was added to each vial, and the vial heated at 80 °C for 40 mm. AfterReferences begin on Pg 127 112cooling, the solutions were injected into the HPLC (withthe analogous cold standard rheniumcomplex) for analyzis.4.2.3 Cysteine and Histidine ChallengesThese experiments were performed as detailed in Chapter 3.2.4.4.2.4 GLUT-i Cell Uptake AssayThis was performed as detailed in Chapter 2.2.4.4.2.5 Hexokinase Inhibition AssayHexokinase kits made up to contain 1 U/mL hexokinase, 1 U/mL glucose-6-phosphatedehydrogenase, 1 mM ATP and 1.5 mM NAD were used in this assay. An aliquot (500 1iL)of this solution was heated to 25 °C for 7 — 8 mm and then transferred to a 1 cm pathlengthcuvette for each trial. Rates were determined by adding one of four amounts of glucose: 10,25, 50 or 75 jiL of a standard solution (1.0 mg/mL glucose) to the cuvette, and the volumemade up to 1250 iL with distilled water. The cuvette was gently inverted three times beforebeing placed in a UV-vis spectrometer fitted with a water jacket equilibrated to 25 °C. Theabsorbance at 340 nm was recorded every 30 sec for at least five mm. For each testcompound five different concentrations were examined with each of the four glucoseconcentrations used above. The inhibitors were tested by adding either 25, 75, 100, 125 or250 1iL of a 5 mM stock solution of the compound to the assay mixture described above(final concentration of 0.1 — 0.5 mM) prior to addition of glucose. The same volume ofhexokinase kit was used each time, with the total volume in the cuvette always being madeup to 1250 tL with distilled water. As these factors were kept constant, changes in the rateof phosphorylation of glucose in the presence of a test compound were attributed toinhibition of the process by that test compound.4.2.6 Hexokinase Substrate AssayThis was performed as detailed in Chapter 3.2.5.References begin on Pg 127 1134.2.7 MTT Cytotoxicity Assay3 -(4,5-Dimethylthiazol-2-yl)-2,5 -diphenyltetrazolium bromide(MTT) is a dye commonly usedin a colorimetric assay for studying the toxicityof compounds in a given cell line.’2 Humanhepatocytes were used in this study, and the cells were cultured and maintained by Jessie Chenin the UBC Biological Services facility. Onday 1 of the study, the cells were trypsinized toremove them from the flask wall to which they had adhered,then rinsed and counted on ahemocytometer. They were then diluted with media to a finalconcentration of lxi cells/mL,and 100 pL of this solution was added to each well of a 96 well plate. The plates were incubatedat 37 °C in a 5 % CO2 atmosphere overnight. On day 2, test solutions were made upbydissolving varying amounts of the test compounds in media and 10 % fetal bovine serum. Testcompound (100 jiL of various concentration) was added to each well to bring the finalwellvolume to 200 jiL. Control wells were made up to 200 1iL by addition of media containing 10 %fetal bovine serum. The plates were incubated at 37 °C in 5 % CO2 for 72hr. On day 5, a 2.5mg/mL solution of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) in PBSwas prepared and added to each well (50 1iL), and the plates were then incubated for a further 3hr at 37 °C. The solution was removed, leaving the cells adhered to the sides of the wells, anddimethylsulfoxide (150 jiL) was added to each well. The plates were gently shaken to solubilizethe purple formazan crystals that had formed. The formazan was quantified using a plate readerto measure absorbance at 570 nm. At least four concentrations of each compound were tested,and all were done in quadruplicate.Data were analyzed by setting the average absorbance of the control wells of each plate to 100 %cell viability. The average absorbance of each concentration of each compound was divided bythe average absorbance of the control wells on that plate, and reported as an average percentagecell viability. Cisplatin was used as a positive control for cytotoxicity, and was found to have an1C50 value of 2.6 ± 0.3 jiM, which is consistent with the literature.’3References begin on Pg 127 1144.3 Results and Discussion4.3.199mTCLabelling and Cysteine/Histidine ChallengesOHaHO\OHOHbCOHFigure 4.3 Long chain glucosamine conjugates studied in various assays in thisChapter: a) C6,b) C8 and c) Cli, where the number refers to the number of carbons between Nof theglucosamine and the N of the metal binding sphere; M = Re,99mTc.Although the original synthesis of the ligands and rhenium complexes (Figure 4.3 ligands, M =Re) were developed by myself and Dr. Z.-F. Chen, they were repeated several times andin somecases optimized by myself.” Once the sugars were deprotected to form the final proligands, thecompounds decomposed over the course of days to weeks, and were repurified or synthesizedjust prior to use.As these three ligands all possessed the same tridentate binding sphere, it was expected that theirlabelling would proceed in a similar manner. This was found to be the case, and the labellingyields were largely independent of chain length (Table 4.1).References begin on Pg 127 115Table 4.1 Results of99mTclabelling and stability studies of three long chain glucosamine-basedligands.Stability in Stability inRT (Re RT(99mTCRadiochemicalCompound cysteine— histidine —compound)* compound)*yield24hr 24hrC6 18.8mm 17.9mm 87% 99±2% 95±4%C8 20.8mm 21.7mm 93% 100±3% 99±3%Cl]. 22.0 mm 22.2 mm 93 % 100 ± 3 %93 ± 3 %*HPLC conditions — 100 % H20 (with 0.1 % TFA) linear gradient to 100% ACN at 30 mm.As these ligands utilize a phenolate donor in the coordination sphere, a basewas required for thelabelling reactions to proceed quickly. The rhenium complexes could be formedwithout base byrefluxing each ligand with the rhenium tricarbonyl precursor in methanol overnight.This is notpractical when dealing with technetium and its six hour half life, so when radiolabelling,a basewas used. Although labelling in aqueous media is desirable, there was concern over the labilityof the amide bond when heated in the presence of hydroxide ions. Theuse of saline solutionwith addition of sodium hydroxide to pH 10 was explored, but was found to be unreliable. Themethod outlined in the Experimental section above had worked for similar compounds,’4so wasapplied here, with good results. The proligand was dissolved in ethanol and three equivalents ofNaOEt were added. This mixture was purged with nitrogen for 10 - 15 mm, ensuring plenty oftime for both ligand deprotonation and oxygen removal before addition of[99mTc(CO)3(H20)]+.The reaction mixture was heated at 80 — 90 °C for about 40 — 50 mm, and gave consistently goodradiochemical yields regardless of slight variations in time or temperature.As expected, the retention times (RTs) increased with the length of the linker between theglucosamine and the metal binding sphere (Table 4.1). The C18 colunm used was non-polar andthe eluent was 100 % H20 (with 0.5 % TFA) to 100 % ACN over 30 mm, from very polarto lesspolar. Thus polar compounds would be eluted earlier, as was observed. There are smalldiscrepancies between the retention times of analogous technetium and rhenium complexes. Asdiscussed in Chapter 3.3, the detectors for the two compounds (radiation and UV-vis detectorsrespectively) are connected in series, which means that their retention times would be expectedReferences begin on Pg 127 116to be different. Also the two compounds are not identical,as they contain different metal ionsand are present in very different concentrations, so itis normal that their retention times willdiffer slightly. A representative radiation trace from the HPLCis shown in Figure 4.4.__________-______0 5 10 15 2025 30Time (mm)Figure 4.4 A representative radio-HPLC trace, forCll99mTc:the peak corresponding to thisspecies appears at 26.6 mm and represents a 91 % radiochemical yield. The small peak at 10- 11mm corresponds to unreacted starting material. The y-axis represents amount of radioactivitydetected at a given time (arbitrary units).These compounds share the same tridentate binding moiety and theirstabilities when challengedwith high concentrations of cysteine and histidine should be very similar. This was indeed true,with all three ligands binding tightly to99mTcfor 24 hr in the presence of large excesses ofcompeting amino acids (Table 4.1). The cysteine and histidine challenges represent a simplemodel of an in vivo system where the synthesized99mTccomplex would be present in very smallconcentrations compared to a wide range of other potential donor molecules. It is very importantto know that the ligand is not easily replaced by an endogenous competing ligand, as this wouldalter the biodistribution of the radiation. The complexes were essentially unchanged after 24 br,a very good indication that they would be stable enough for in vivo applications.4.3.2 Cell Uptake AssayThe compounds tested here showed modest cellular uptakes, which were not affected by thepresence of glucose, suggesting that the cellular uptake observed is not due to the samemechanisms that are used to transport glucose. As can be seen in Figure 4.5, the uptake of theReferences begin on Pg 127 117compounds parallel their lipophilicity, with the Cli compoundexhibiting the highest uptake andthe C6 compound the lowest. The C2 compound withthe analogous binding sphere(99mTCL3discussed in Chapter 3) is shown here for context, and verifiestrends in lipophilicity, as it has alower cellular uptake than the longer chain compounds. Figure 4.5 alsoshows that addition of alarge excess of glucose does not significantly affect the cellular uptake of any of thetestcompounds. These data together suggest that the observed uptake is due to passive diffusionthrough the lipid bilayer, where more lipophilic compounds candiffuse more easily.0.4%0.3%0.2%__ __IC2-Tc(99m)C6-Tc(99m) C8-Tc(99rn) Cl l-Tc(99m)Figure 4.5 Percentage of activity taken up into LCC6-HER2 cells during a 30 mm incubation.For each compound the results shown are the average of four separate experiments with errorbars indicating the standard deviation. Each compound was tested in glucose free media (dottedbars) and in 10 mM glucose (striped bars).A limitation of this assay is that it is not able to distinguish between cell uptake and membraneassociation. Further studies could be performed to distinguish between these two possibilities,but given the low level of activity associated with the cells, this is not really worthwhile, as thelevel at which they are being taken up (if at all) is too low to be useful in the context of aReferences begin on Pg 127 118molecular imaging agent. The mechanismby which they are being taken up (if at all) is not thesame as that used for glucose, meaning the uptakewill not increase in cells overexpressingglucose transporters. This is a necessary requirementfor a cancer-targeting molecular imagingagent so as to get selective enhancement of uptake inthe area to be imaged.These same uptake tests were performed by Schibliand coworkers on the long-chain glucosebased compounds (Figure 4.2) discussed in Chapter 4.1.Their experiments were performed inHT29 cells, a human colon cancer cell line also knownto overexpress GLUT- i. As theseexperiments were performed in different cells in a differentplace, there may be some variationbetween the data sets. The uptake percentage values obtainedfor all the long chain compoundsare quite similar, with the two glucose analogues foundto have cell uptakes of 0.1 — 0.2 % of theactivity to which they were exposed.5 These uptakevalues did not vary significantly with thepresence of glucose, the concentration of test compound, or the presence ofa GLUT-i inhibitor,meaning that the uptake observed is not due to GLUT-i mediated transport.It seems unlikelythat there is a significant difference betweenGLUT-i interactions of the glucose andglucosamine substituted bioconjugates of this type.4.3.3 Hexokinase Inhibition StudiesHexokinase is the first enzyme in the oxidative phosphorylation chain, andis responsible fortransferring a phosphate group from ATP to the C-6 position of a hexose.The details of thiswere discussed in Chapter 1.3, and the reaction schemeis shown in Scheme 4.1. Aconvenient way to quantitate this process is to couple this phosphorylation tothe next step inthe oxidative phosphorylation chain where glucose-6-phosphate dehydrogenasedehydratesglucose-6-phosphate and, in the process, converts NADinto NADH. NADH absorbsstrongly at 340 nm, and this can be used to monitor the rate of thetwo reactions. Althoughthe increase in absorbance observed at this wavelength is dueto the production of NADH,the phosphorylation catalysed by hexokinase is the rate limitingstep in this scheme and theproduction of NADH effectively reports the hexokinase reaction rate.This assay examined the test compounds to see if they were able to perturbthe hexokinasecatalysed phosphorylation of glucose. Absorbance at 340 nm was monitoredat 30 secintervals to determine the reaction rate at a range of substrate (glucose)concentrations. ThisReferences begin on Pg 127 119was done in the absence and presence of variousconcentrations of test compound. Theinitial rates were taken from the linear portionof the A340 vs. time graphs (normally the first210 — 240 sec), and these rates were plotted vs.glucose (substrate) concentration. This wasdone for all combinations of glucose and test compound concentrations.A Lineweaver-Burkplot’5 (1/[glucose] vs. 1/V1) was generated for each test compound andleast squares fittingused to determine the trend line for each concentrationof test compound (Figure 4.6a). Theposition at which the lines for each concentration of test compoundcross gives informationon the mechanism of inhibition. Lines intersecting at the y-axis indicatecompetitiveinhibition, those at the x-axis indicate non-competitive inhibition,and lines that meetbetween the two show mixed inhibition.’6 To determine the inhibition constant,K1, for eachtest compound a plot of [inhibitor] vs. [glucose]/V1was prepared (Figure4.6b), and a leastsquares line fitted to this data. The intercept of this line with the x axisis equal to the —K1value of the test compound.Q\OoOHHK oo G6PDHHO\ATPNAD NADHHOScheme 4.1 Reaction scheme for phosphorylation of glucose byhexokinase (HK) andsubsequent dehydration of the product by glucose-6-phosphate dehydrogenase(G6PDH).The compounds tested in this work had Lineweaver Burk line intersectionpoints close to they axis, as can be seen for C6L in Figure 4.6a. This indicates either competitive inhibition ormixed inhibition with a large amount of competitive character is takingplace. This meansthat the inhibitors must be able to bind to the active site of hexokinase. Thisis an excellentfirst step for these compounds, as if they are to be phosphorylated by hexokinase, they mustbe able to fit into the active site.References begin on Pg 127 120aU)a)U)0()20001500U)125uL C6L1000 -i75uL C6L25uIL C6LOuLC6L-0.03 -0.01b0.01 0.02-500 -1/[glu] (iL1)0.03600005000040000.30000-300 -200 -100-10000100 200 300[C6L] (jiM)Figure 4.6 Plots showing data manipulation to determine K1 values: a) Lineweaver-Burk plot forC6L, b) K plot for C6L.References begin on Pg 127 121The Cli compounds were not tested in this assay dueto solubility constraints. The C6 and C8ligands and complexes that were tested all showed interactionwith hexokinase: C6L K1 = 220 ±30 jiM, C6Re K, = 500 ± 60 pM, C8L K, 210±40jiM, C8Re K, = 70±20 M. The two freeligands C6L and C8L have the same K, values aseach other, C8Re has a stronger interactionand C6Re a weaker interaction with hexokinase. We do not fully understandwhy this is thecase. One possible reason C8Re exhibits a stronger interaction with hexokinasethan C8L is thatby coordinating the metal ion the chelating portion of the ligand becomes moreordered. Thechelating arms no longer have the ability to rotate freely as they do in theligand, so the metalcomplex may end up taking up less space. The reverse is seen for the C6 compoundswhere theligand displays a stronger interaction with the enzyme thandoes the metal complex. Perhaps inthis case the extra bulk added by the metal ion makes the compound too bigto fit into the cleft.In the C6 compounds the metal binding moiety is closer to the active site thanin the equivalentC8 compound, so making the complex bulkier would have a greater effect. Itmay also be thatone of the compounds has a functional group in place to specifically interact witha residue onhexokinase, either in a favourable or an unfavourable manner.This is a positive outcome for these compounds. Despite some minor differences between thetest compounds, they all displayed tM inhibition constants with hexokinase. For comparison,FDG has an inhibition constant for hexokinase of 1060 jiM.’7 The K values suggest that thesecompounds interact strongly enough with hexokinase to make it possible that they will bephosphorylated themselves.Schibli and coworkers investigated the hexokinase inhibition properties of their glucose-basedlong chain conjugates and found results similar to what were found in this study.5 They tested arange of compounds in their assay (see Figures 1.8, 1.9 and 1.10), the majority of which hadshort chains. The two long chain analogues (Figure 4.2) were the only ones that exhibited anyappreciable inhibition of hexokinase. The PEGylated compound had K = 250 jiM and the C8compound K, = 5800 jiM.5 The addition of two oxygen atoms in the alkyl linker chain changesthe affinity for hexokinase by more than an order of magnitude.The affinities of the long chain glucose and glucosamine analogues for hexokinase are verysimilar. Although the protocols for the two sets of experiments were the same, potentialdifferences in equipment and experimental technique make it difficult to know how comparableReferences begin on Pg 127 122the data are. The glucosamine compounds seem to interactmore strongly with the active site ofhexokinase. Comparing the K values of the non-PEGylated glucose rhenium complex with thatof C8Re, the glucosamine analogue has an interaction with hexokinase that is nearly two ordersof magnitude stronger than the glucose compound (K1 = 70 vs. 5800 1iM). As the PEGylatedglucose compound has a much stronger interaction than its all carbon counterpart, this suggeststhat PEGylated glucosamine compounds could be worth investigating further.4.3.4 Hexokinase Substrate StudiesFour compounds were tested in this assay: C6L, C8L and their respective [Re(CO)3jcomplexes.Unfortunately none of them showed any significant phosphorylation by hexokinase. The assaysolutions were examined using ESI-MS to search for any trace of phosphorylated species notdetected by the HPLC assay. Although the concentrations involved were extremely low, therhenium isotope pattern is a good handle to identify the metal complexes. Unfortunately noevidence of any phosphorylated products were seen. In combination with the HPLC resultswhich show no significant ADP production above the hydrolysis levels of the control samples,these results suggest that no appreciable phosphorylation is occurring.As discussed in Chapters 4.1 and 1.3, when hexokinase phosphorylates a substrate the twodomains of the enzyme rotate together around the active site, narrowing the cleft in the middle.For a compound to be a competitive inhibitor of glucose phosphorylation it only has to bind tothe open conformation of hexokinase; the cleft does not have to close. For a compound to be asubstrate of hexokinase, the cleft needs to be able to close around the molecule in the active sitein order for the phosphate group to be transferred. This could explain why the compoundsexamined in this Chapter can inhibit hexokinase in a manner that suggests they interact with itsactive site, but cannot themselves be phosphorylated.4.3.5 MTT Assay3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MIT) is a yellow compound thatis metabolized by mitochondrial reductase in living cells to purple formazan (Figure 4.7a). Thepurple colour is quantified by measuring absorbance at 570 nm (A570). MIT metabolism is usedto indicate the number of viable cells, as the more living cells there are, the more dye isReferences begin on Pg 127 123metabolized and the darker the purple colour produced.A570 is compared between control wellswith no test compound and wells with varyingconcentrations of test compound. The controlwells are taken to have 100 % cell viability, andthe relative percentage of viable cells in testwells on the same plate are calculated with respectto this.a\‘ \\jlive cellsBr/MU (yellow)Formazan (purple)OHbHNCiC d7’“NH—- eOHH0MCC COCo OC_ COCOFigure 4.7 a) MTT dye, and its purple metabolic product formazan; b — e) compounds tested fortoxicity in an MTT assay but not shown in Figure 4.3 above: b) cisplatinc - e) possiblefragments of the compounds in Figure 4.3.Due to solubility problems, the Cli compound was not tested. The C6 andC8 ligands andrhenium complexes were tested, along with selected ligand fragments (Figure4.7c-e). Toxicityis normally less important in radiopharmaceutical compounds than in other drugs. Theconcentrations administered are very low and diagnostics of this type need only be administeredonce or a few times, not chronically like most therapeutics. Understanding toxicity addsto ourknowledge of how these compounds may behave in vivo, and as such, the more information thatcan be gathered the better.None of the ligands or rhenium complexes tested were found to exhibit any toxicity up to themaximum concentrations tested. C6L, C8L and C6Re were tested to a concentration of 1 mM,and (due to solubility constraints) C8Re was tested up to 100 M. The tridentate binding unitalone (Figure 4.7c) showed some toxicity, with an IC50 value of 158 mM. Although this bindingReferences begin on Pg 127 124unit does show significant toxicity, it is two ordersof magnitude less toxic than cisplatin (used aspositive control, 1C50 = 2.6 mM). Two analoguesof this compound (Figure 4.7d and e) weretested and found not to be toxic. This tridentateligand bound to [Re(CO)3](Figure 4.8d) wasfound to be non-toxic to 200 1iM, its solubility limit in aqueoussolution. The same tridentatecore but with ethanoic acid appended to the aliphaticnitrogen (Figure 4.7e) was found to be non-toxic to 1 mM. It is highly unlikely that the bindinggroup that exhibited the measureabletoxicity (Figure 4.7c) would ever form in vivo. The compoundwould be administered bound to[99mTc(CO)3]+meaning formation of this group would require the decomplexation of the metaltricarbonyl core and the breaking of a C-N bond. Cysteine/histidine challenge experiments showthat decomplexation of this tridentate ligand happens at a negligible rate, while the spontaneousbreakage of such a C-N bond would be unprecedented. If degradation were to occur, it wouldlikely be at the amide bond resulting in a pendant acid group, where this type of compound wasfound not to be toxic. The typical amount of a99mTccomplex injected into a patient is 800 MBq,45 pmol.’8 As this is distributed throughout the wholebody, the concentration found in anyone place, even sites of preferential accumulation, would be so low that chemical toxicity is notof great concern.4.4 ConclusionThis chapter details the formation and in vitro assays of long chain glucosamine-based tridentateligands. Unfortunately these compounds are not transported or phosphorylated as carbohydrateanalogues. In synthesising these metal chelates, the sugar has been modified to such an extentthat it is no longer enzymatically recognized as such. It is not known how much change is toomuch, but work like that shown here aims to determine, and take advantage of, these acceptablelimits. These compounds are much bigger than a monosaccharide, as shown in Figure 4.8.Comparing the molecular weights, glucosamine is 179 g.mol’,C699mTc(the lightest metalconjugate made in this work) is 683 g.moV’, and CuRe (the heaviest metal complex made inthis work) is 829 g.mo[’. Glucosamine consists of one six membered ring; these metalcomplexes comprise three six membered rings with a heavy metal atom and a long carbon linker.As the linker length is required for the hexokinase activity, it may be advantageous to decreasethe size of the metal chelating portion.References begin on Pg 127 125OHa,HO-OHHO-NH2b)OHi’/TCcoCoc)Figure 4.8 Size comparison of a) glucosamine, MW 179 g.mol’ b)C699mTc,MW 683 g.mor’and c) CuRe, MW 829 g.moF’.Strong metal binding can be achieved with smaller chelating arms provideda tridentate bindingsphere is maintained.’9The thymidine kinase and amino acid long chain conjugates (Figure 4.1)discussed in the Introduction to this Chapter both have small tridentate bindingspheres andretain at least some of the biological recognition of the parent compounds.”2The stability ofthese compounds was found to be adequate, with the thymidine kinase mimicsremaining at least95 % intact after incubation for 24 hr in PBS buffer,’ and the amino acid analogue being stablefor 24 hr in serum.2Both these bioconjugates (in Figure 4.1) have a carbon chain of a significant size with respect tothe overall size of the molecule as a spacer between the recognition part of the molecule and themetal chelate. The distance between the two portions of the molecule is proposed to allow thebiologically active part to be recognized as a substrate, while the labelled part is far enough awaynot to interfere with this process. Although the recognition and catalytic mechanisms of eachenzyme are different, these findings reinforce the merit in our approach to the long chainglucosamine conjugates.References begin on Pg 127 1264.5 References1. Desbouis, D.; Struthers, H.; Spiwok, V.; Kuster,T.; Schibli, R., I Med. Chem. 2008, 51,6689 - 6698.2. Liu, Y.; Pak, J. K.; Schmutz, P.; Bauwens,M.; Mertens, J.; Knight, H.; Alberto, R., IAm. Chem. Soc. 2006, 128, 15996 - 15997.3. Killian, D. M.; Chikhale, P. J., Neurosci. Lett.2001, 306, 1 - 4.4. Pardridge, W. M.; Oldendorf, W. H.; Cancilla,P.; Frank, H. J. L., Ann. Intern. Med 1986,105, 82-95.5. Schibli, R.; Dumas, C.; Petrig, J.; Spadola,L.; Scapozza, L.; Garcia-Garayoa, E.;Schubiger, P. A., Bioconjugate Chem. 2005, 16, 105 - 112.6. Steitz, T. A.; Anderson, W. F.; Fletterick, R. J.; Anderson,C. M., I Biol. Chem. 1977,252, 4494 - 4500.7. Speizer, L.; Haugland, R.; Kutchai, H., Biochim. Biophys. Acta - Biomembranes1985,815, 75-84.8. Yoshioka, K.; Saito, M.; Oh, K.-B.; Nemoto, Y.; Matsuoka, H.; Natsume, M.;Abe, H.,Biosci. Biotech. Biochem. 1996, 60, 1899 - 1901.9. Kahienbe, A.; Dolansky, D., Can. I Biochem. 1972, 50, 638 - 643.10. Speizer, L.; Richard, H.; Howard, K., Biochim. Biophys. Acta1985, 815, 75 - 84.11. Bowen, M. L.; Chen, Z.-F.; Roos, A.; Adam, M. J.; Orvig, C., manuscript in preparation.12. Mosmann, T., I Immunol. Methods 1983, 65, 55-63.13. Wu, A.; Kennedy, D. C.; Patrick, B. 0.; James, B. R., Inorg. Chem. 2003, 42,7579-7586.14. Lim, N. C.; Ewart, C. B.; Bowen, M. L.; Ferreira,C. L.; Barta, C. A.; Adam, M. J.;Orvig, C., Inorg. Chem. 2008, 47, 1337 - 1345.15. Lineweaver, H.; Burk, D., I Am. Chem. Soc. 1934, 56, 658 - 666.16. Suzuki, Y.; Nakabayashi, Y.; Nakata, K.; Reed, J. C.; Takahashi, R., I Rio!. Chem. 2001,276, 27058 - 27063.17. Ferreira, C. L.; Ewart, C. B.; Bayly, S. R.; Patrick, B. 0.; Steele, J.; Adam, M. J.; Orvig,C., Inorg. Chem. 2006, 45, 6979 - 6987.18. Sampson, C. B., Text Book ofRadiopharmacy: Theory and Practice. Taylor and Francis:Oxford, 1994;p17.12719. Schibli, R.; La Bella, R.; Alberto, R.; Garcia-Garayoa,E.; Ortner, K.; Abram, U.;Schubiger, P. A., Bioconjugate Chem. 2000, 1], 345 - 351.128CHAPTER 5Long Chain Glucosamine Cyclopentadienyl Ligands for the[M(CO)3]Core (M = Re,99mTC)*5.1 IntroductionCyclopentadienyl (Cp) ligands have been shown to bind well to the [M(CO)3](M= 99mTC,Re)core via a95coordination mode to give 18-electron organometallic complexes.’ Thesecomplexes are referred to as “piano stool” because of their shape, a flat Cp ring on top with threeCO “legs” beneath. Cp ligands are monoanionic so the complexes they form with themonocationic metal tricarbonyl cores are neutral. These ligands are also attractive for their smallsize and lipophilicity, as well as the stability of the resulting complexes. The stability comesfrom the fact that these inert Tc(I) compounds have a low spin d6 electronic configuration. Thesmall size and lipophilicity can be exploited in the ability of complexes to fit into active sites ofenzymes or for enhancing passive diffusion across membranes such as the blood brain barrier(BBB). For example, a long-chain fatty acid conjugate (Figure 5.1) has been labelled withCp99mTc(CO)3,2and was found to undergo beta-oxidation in the heart. Beta-oxidation is theprocess by which fatty acids are metabolized. They are broken down two carbon atoms at a timeto produce acetyl-CoA, NADH and FADH2.3 The enzymes involved in beta-oxidation were ableto utilize this bioconjugate as a substrate because the CpTc portion of the molecule was smallenough not to significantly detract from the biological recognition and activity of the fatty acidportion.2OH99mTCoC__CoCOFigure 5.1 ACp99mTcconjugate that is metabolized by beta-oxidation in the heart.2*A version of this chapter will be submitted for publication: Bowen, M. L., Adam, M. J., Orvig, C. Investigation ofLong Chain Glucosamine-[CpM(CO)3]Complexes (M = Re,99mTc)for Use in Molecular Imaging.References begin on Pg 161 1295.1.1 Synthesis ofLCpM(CO)31CompoundsCp is a widely used ligand known to bind to a rangeof metal ions with several differenthapticities (normally 1, 3 or5)4Its coordination to rhenium tricarbonyl has been knownsince the 1970’s,5but very harsh reaction conditionswere needed to make compounds of thistype until relatively recently.6 The introduction of a simple synthesisof[99mTc(CO)3(H20)]+by Alberto and coworkers7has renewed interest in molecules of thistype, and Cp complexes of this core are now an activefield of investigation due to thefavourable physical properties outlined above. Studies involve the functionalizationof{CpM(CO)31complexes in a bid to affect their in vivo behaviour. There are two waystofunctionalize a Cp-metal complex: either by functionalizing the Cp and then binding it to themetal, or by forming the metal-Cp complex and then adding functionality to that core.Within this framework there are many different methodologies that have been developed,8and the key ones are outlined below with a focus on the development of reaction types thatare used in the syntheses carried out in this chapter.The practical synthesis ofCp99mTccomplexes was pioneered in 1992 by Wenzel.9’ Adouble ligand transfer (DLT) reaction was used (see Scheme 5.la) to form99mTccomplexesfrom pertechnetate, a derivatized ferrocene and Mn(CO)5Br(as a carbonyl donor) at 150 °Cin 30 — 90 % yield. The DLT is named as such because both the ligandtypes on the Re(I) inthe reaction product were attached to different metal ions in the reactant species — both weretransfered during the DLT reaction. Others have since remade these compounds forexamination of their biodistribution,” and found that after heating at 150 °C for one hourand purifying the reaction mixtures by preparative TLC, only 20 — 25 % yield is reliablyrecoverable. Analysis of an aqueous solution 24 hr after synthesis, and the biodistribution ofthe compounds both strongly suggest the compounds are stable on the timescale relevant toimaging, both in aqueous solution and in vivo.References begin on Pg 161 130aMn(CO)5Br99rnTCO4Ire 99mTC,oc-_ 4Cc0MeOHorTHF140 °C, 1 hrR = carbonyl groupR’ = carbonyl group or HCr(CO)6bP CrCI30--_-___)/ 0—FeRe,MeOH ocy160 °C, 1 hr0Scheme 5.1 Double Ligand Transfer (DLT) reaction: a) as pioneeredby Wenzel9and b) asimproved upon by Spradau and Katzenellenbogen.’2This DLT was improved upon and extended to rhenium by Spradau and Katzenellenbogen(Scheme 5.lb).’2 A major flaw in Wenzel’s system that Katzenellenbogen addressed wasthat the Cp substituted Mn compound was also formed in the reaction, and was virtuallyimpossible to separate from the desired99mTcproduct, thus reducing the specific activity ofthe99mTccompound. The modifications made by Katzenellenbogen included the use ofCr(CO)o as a CO source, and extensive optimization of solvents, temperatures, Cpsubstitution patterns and other additives.’2Although this system works well for the synthesisof simple Re precursors like the one shown in Scheme 5.la, which can then be coupled toother molecules of interest (as we have done in this work), the harsh experimental conditionspreclude its widespread application. The reaction only produces useful yields whenperformed in methanol, meaning it cannot currently be used for radiochemistry whereaqueous conditions are the norm. Another requirement that limits its usefulness is the needfor the Cp ring that is transfered in the reaction to have an electron withdrawing groupattached. The importance of this electron withdrawing group can be understood byexamination of the proposed reaction mechanism (Scheme 5.2). The ring that is transfered inthe course of the reaction has to bridge the two metals involved, and transfer from the iron offerrocene to the rhenium carbonyl centre. The electron withdrawing group is proposed tostabilize intermediate A (in Scheme 5.2) where the Cp is bridging the two metal centres.12References begin on Pg 16] 131Scheme 5.2 Proposed mechanism of ring transfer in the DLT reaction.12The first stepinvolves reduction of the metal centre and transfer of the CO ligands. This isfollowed by amultistep transmetallation sequence that tranfers the substituted Cp ringfrom the ferroceneto the other metal centre,12 M= 99mTC,Re.An alternative to the DLT is the single ligand transfer reaction (SLT), where the carbonylligands are already attached to the metal ion of interest, and only the Cp ligand transfersduring the reaction. When an SLT reaction between [Re(CO)6]and acetylferrocene wasperformed in water alone no desired product was observed, but in a 1:1 mixtureof DMSOand water, 60 % yield was obtained (Scheme 5.3a).’3 The DLT does not have this range ofsolvent availability due to the CO transfer step requiring donation from thebulk solvent,meaning it is limited to MeOH, and is therefore not readily transferableto the synthesis ofradiopharmaceuticals.’2The SLT with [Re(CO)6Jis proposed to occur viatwo steps; thesubstitution of three carbonyl ligands with solvent molecules, followedby Cp exchange, asfor the DLT mechanism shown above (Scheme 5.2). The observations inthis work supportthe mechanism of ring exchange proposed above in the DLT reaction(see Scheme 5.2),where r tori3ring slippage of the Cp gives it the ability to coordinate tothe tricarbonylmetal centre and subsequent transmetallation results in the product.CrCI3C r(CO)6MeOH160°CocoHHCH3CHcFI3M04ocMco+- MeOH-4———CH3( CoCp—Fe /\7—M—CO-MeOHCO4CH3Co0COAReferences begin on Pg 161 132qaFe[Re(CO)6][BF4]DMSO160 °C, 3 hr0Re,oc_71 %bFe[Re(CO)6][BF4]DMSO160°C ,--‘30 rn4hoc’.J ‘DMSO/H20[99mTC(CO)(HO)]+oc965 %Scheme 5.3 Single Ligand Tranfer (SLT) reactions: a) using [Re(CO)6]starting materials’3b) illustrating the difference in reaction conditions and yieldsbetween rhenium andtechnetium’4c) lower yields result when using alternative rhenium startingmaterials.’5As this SLT reaction was proposed to go via a [M(CO)3(solvent)]intermediate,Jaouen,Alberto and coworkers attempted to extendit to the[99mTc(CO)(H20)]+core.14 This wouldenable transfer of a substituted Cp ring from the iron centreof ferrocene directly to the[M(CO)3]core (Scheme 5.3b).’4 For the99mTCreaction, the authors successfully used[99mTc(CO)(H20)]+as a precursor, and the reaction proceeded in the presence of water ingood yield.’4 DMSO was needed to solubilize the ferrocene precursorso the reaction wasperformed in 1:1 DMSO:H20.’4 For the non-radioactive chemistry,the [Re(CO)6][BF4]precursor was used and the reaction was carried out in DMSO due to the high temperatureC> 80 %Re(CO)5BrMeOH,160°C,lhrAB[NEt4]2[Re(CO)3BrMeOH, 160 °C, 1 hr0iL--Re ,oc_A =40 %B = 16%References begin on Pg 16] 133requirements of the reaction.14 No discussion is made ofstarting with the [Re(CO)3(H2O)fprecursor.14The Orvig group’5 has also attempted to extendthe scope of the SLT reaction to allow theuse of commercially available ([Re(CO)5Br]) or easily accessible ([NEt4]2[Re(CO)3Br])’6starting materials. Although some success was achieved with thesesubstrates in the SLTreaction (Scheme 5.3c), the yields produced were lowerthan those attained using [Re(CO)6]as a starting material.’5Alberto and coworkers pioneered the use of substituentsto lower the pKa of the Cp ring.’7This allowed for a significant amount of the Cp to be deprotonated at pH7.4, so the labellingreaction could take place under ideal radiochemical conditions.’7This was achieved viaaddition of an acetyl, an electron withdrawing group, to the Cp ring, which increased theamount of conjugation in the system and lowered the pKa to 8.6.’ This meant5 % of theCp rings were deprotonated at pH7417The substituted cyclopentadienyls were synthesizedby reacting NaCp with the appropriately substituted ester (with loss of the alcohol portion ofthe ester) in yields of 50 - 60 % (Scheme 5.4a).’7 Once substituted,these Cp derivativeswere found to be water soluble and relatively stable.’7 To form the metalcomplexes, thetricarbonyl precursor and the substituted NaCp were stirred inbuffer (pH 7.4) at 85 °C for 2hr.’7 The yields for the formation of the rhenium complexes were around20 %, which theauthors propose to be low due to competing metal cluster formation.’7The analogous99mTCreaction proceeds in 15 mm at 90 °C to give a quantitative yield of the desired product(Scheme 5.4a).’7 These compounds were all found to bestable in solution for 24 hr at 37The three component procedure (Scheme 5.4b) was discoveredby Minutolo andKatzenellenbogen, who found they could add three components together in onepot to obtaintheir desired products in good yields.’8 The Cp starting material for thisreaction is a diazasubstituted Cp, which upon reaction with a nucleophile in thepresence of the metaltricarbonyl core produces the nucleophile-substituted Cp bound to the metal.’8 This reactionrequires the use of relatively mild conditions: 85 °C for 45 mm.18 The yields vary dependingon the nucleophile, ranging from 53 - 72 % isolated yield for thesubstituted rheniumcomplexes.’8The drawbacks to this approach are the reactivity of the diazastarting materialReferences begin on Pg 16] 134and the requirement of exchanging the Br counterionof the rhenium starting metal toprevent bromination of the Cp ring (unless that is the desiredproduct).8[Re(CO)3(H2O)]0 pH 7.4, 2 hr, 85 °CaNa-ORfrR25%Na[99mTC(CO)(HO)]+ QRpH 7.4, 15 mm, 90 C ImTci-”- ccCH3CN Nub [Re(CH3CN)(CO)][OTf]++ Nu80 °C, 45 mmYield: Nu = 1 - 53 %Nu=CH3COO K 62%cphoto’sis,Rn-BuLiRe(CO)5Broxidation[LTHF, 78 °CLiocj30[Re(CO)3(H2O)JH20, 4 hr, 160 °cocj°25 %d_—- COCHPhPhH2COC0[99mTC(CO)(HO)]+MeOH/H20,30 - 120 mm, 95 °C oTSco>95 %Scheme 5.4 Four alternative methods for forming [CpM(CO)3]complexes: a) deprotonationof a substituted Cp b) the three component system c) a decomplexation!recomplexationmethod and d) via a retro Diels-Alder reaction.References begin on Pg 161 135Jaouen and coworkers have utilizeda method of decomplexation and recomplexation tomake [CpRe(CO)3Jcompounds (Scheme5.4c).’9 They began with their functionalized Cpligand bound to a metal ion (normally Mn) to stabilizeit and allow for long-term storage.’9The ligand is decomplexed in situ using lightor oxidation, and following the addition of baseand another metal ion, recomplexation occurs.19’20This was applied to the synthesis of Recomplexes to give 45 % yield.2° As wellas the photolysis or oxidation inducer beingnecessary, this reaction also requires the use ofn-BuLi at -780C20which may preclude itswidespread applicability. To our knowledge,details of the application of this method toradiochemistry with99mTchave not been discussed.Recently Alberto and coworkers have discovereda synthesis of [CpM(CO)3](M = Re,99mTc)using a retro Diels-Alder (DA) reaction catalyzedby the metal ion which becomessubstituted (Scheme 5.4d).2’ The synthesis of the[99mTc(CO)3(H20)]+precursor can beperformed in situ from{TcO4rand areducing agent rather than requiring a two step reactionto make the precursor prior to addition of the proligand.2’The authors have found that forthe reaction to proceed, both the former diene anddienophile must be substituted with aweak metal binder like a carboxylate.2’They observeda complete lack of free retro DAproduct (not metal coordinated).2’These findings led theauthors to conclude the mechanismof reaction first involves weak coordination of the dimerized ligand to the metal,followed bya metal-mediated retro DA reaction andii5-Cpcoordination.2’The starting materials for thisreaction can be made by either conjugating biomolecules tothe acids on the DA dimerizedproduct of (HCp-COOH)2(Thiele’s acid), or by reacting NaCp with an esterof the desiredtargeting molecule and then dimerizing the resulting Cp.2’ By reactingthe water sensitiveNaCp to make a dimeric complex stable to both heat and water, the complexation reactioncan take place in aqueous solution, as required for use in99mTclabelling reactions (Scheme5.4d).2’ The full scope of this reaction is yet to be explored, but work is continuing onlabelling biomolecules and making polymer supported starting material to provide complexeswith very high specific activities.21There are some significant differences in the reactions of Cp with the tricarbonyl cores ofrhenium compared to technetium (Scheme 5.3b and 5.4a,d). The difference between thesetwo metals is much more noticeable in this area of chemistry than in the simple tridentatecoordination chemistry seen in Chapters 3 and 4 of this thesis, where the two metals behavedReferences begin on Pg 161 136very similarly. The reasons for the differencesin the Cp chemistry of the two metals seem tobe a manifestation of their slightly different physicalproperties being exaggerated by theharsh reaction conditions required. The smalldifferences in size and reactivity of rheniumand technetium combined with concentration differencesin the tracer vs. macroscopic scalesand solubility constraints of some starting materialslead to analogous compounds of the twocongeners behaving quite differently. Detailsof this phenomenon, as applied to Cpchemistry, have not been thoroughly investigated toour knowledge.5.1.2 Functionalization and Biological Studies of [CpM(CO)3]ComplexesJaouen and coworkers were pioneers in the field ofprotein labelling with the organometallic[CpM(CO)3]core, reporting on the first such rhenium complexin1993.22A Cp wassubstituted with an N-hydroxysuccinimide ester and was then bound tothe metal centre,before reaction of the activated ester with free amineson the lysine sidechains of theprotein.22 Monoclonal antibody JOSS2-2 was found to retain satisfactory receptorrecognition upon being labelled at 15 % of its available sitesin this fashion.22 The authorsdid not report any radioactive complexes, and the reaction times (15 hr)22do not lendthemselves to radiochemistry. This was an important first step, and a keyfinding was thestability of the resulting compounds — no degradation was observed after incubation at37 °Cfor 24 hr.22The same researchers have applied the decomplexation/recomplexation method (Figure 5.5c)to the synthesis of rhenium-Cp tamoxifen derivatives (Figure 5.2a).’9’20They have gone onto make several derivatives and tested the rhenium compounds for their binding affinities tothe estrogen receptors overexpressed in certain types of breast cancers.23 The compoundshown in Figure 5.2a showed the highest affinity of the compounds tested, with a relativebinding affinity (compared to that of tamoxifen) of 31 ± 3%23Using computer modeling ofthe ligand binding sites of the receptors the researchers observed that the slightly larger sizeof the [CpRe(CO)3jcore compared to the phenyl group of tamoxifen is the cause of thedecreased binding affinity.23 The authors claim the need to find improved labelling reactionsbefore being able to apply this work towards radioactive imaging agents.231 -(2-Methoxyphenyl)-piperazine has been linked to a Cp ring, and the resulting ligand boundReferences begin on Pg 16] 137to the tricarbonyl core (Figure 5.2b) via a deprotonationreaction (Scheme 5.3a).24 A seriesof chain lengths were examined and were found to have a largeeffect on the recognition ofthe compounds.24 The [CpRe(CO)3]complex with a C5 linker (Figure 5.2b) has a very highaffinity in vitro (1C50 = 6 nM) for the5-HT1A serotonergic receptor.24 This is of interest inimaging brain function as it has been linked to mental conditionssuch as anxiety anddepression.25 An analogous complex, but with the metal bound via a bidentate17N-chelaterather than a Cp was also investigated, and was found to have an 1C50 value of 5 ± 2 nM forthe 5-HTIA receptor in vitro.26 Affinities of the two compounds are essentially the same,soin this case the use of Cp does not seem to confer any real advantages to the metal complex,though in vivo testing would be required to compare properties like BBB permeability.aOHb1/ oHN—/—/I /CCI ‘CO/—_Z‘I 10cc0C’’C0CCNDCLDC—Figure 5.2 Some bioconjugates of the [CpM(CO)3]core: a) a tamoxifen derivative made viaa decomplexation/recomplexation reaction20 b) a 1 -(2-methoxyphenyl)-piperazine conjugate— a 5-HTIA receptor ligand24 c) a D1 dopamine receptor ligand analogue27 d) an estradiolanalogue made via an amide coupling to the rhenium precursor28 e) an alternatively linkedethynylestradiol analogue made via a SLT reaction.14Known dopamine D1 receptor ligands have been labelled with the {CpRe(CO)3]core (Figure5.2c).27 The addition of the —COC5H4Re(CO)3group to the native ligand reduced thereceptor affinity by half, but increased its selectivity for D1 over other similar receptor types,showing promise as a target for99mTcbased imaging.27 The authors claim this is a model forthe99mTccomplex, but this extension has not, to our knowledge, been published. As therhenium compound is made via an indirect and somewhat circuitous route, this synthesis iseReferences begin on Pg 161 138not readily applicable to radiolabelling, and the99mTcreaction may not have been developed.Steroid derivatives have been of interest inthe field of molecular imaging for quite sometime, and the Cp analogues of various steroids have beenmade.’ Estradiol is of particularinterest in imaging estrogen receptors overexpressed in many types of breastcancer.14 Aseries of estradiol derivatives bound to variouschemical forms of rhenium have beenstudied.28 The Cp-Re analogues of interest here were synthesizedvia an indirect route,where the rhenium-Cp complex was made and thencoupled to the functionalized steroid togive the bioconjugate.28 The position of substitution on the estradiol andthe type and lengthof the linker to the metal ion are both very important.28A 35 % binding affinity relative toestradiol (normalized to 100 %) could be achieved with a7a substituion and a medium sizedlinker (Figure 5.2d).28’29 The analogous ligand with a thioether S,S-bidentatebinding moietywas also synthesized, and was found to have inferior biological properties compared with itsCp counterpart.29 The bidentate complex showed a relative binding affinity of 12 %compared to the analogous Cp compound (with no 0 in the linker comparedto the oneshown in Figure 5 .2d) with an affinity of 29%29The authors do not speculate as to whysuch large differences are observed between the compounds. Others have suggested severalfactors that could contribute to this difference;’ coordination of the bidentate thiol givesdiastereomeric complexes which will have different biological properties to each other, andthe coordination of Bf to fill out the octahedral binding sphere increases the bulkat themetal centre.1 Preliminary biodistributions of these very non-polar compounds showed highnon-specific binding, proposed to be due to their high lipophilicities.28 This study illustratesthe large changes in in vitro properties that can be brought about by relatively smalladjustments such as the substitution of a Cp ring for a bidentate chelate.A different set of estradiol derivatives was made via a SLT reaction from the appropriatelysubstituted ferrocene and either[99mTC(CO)3(H2O)3jor [Re(CO)o] [BF4j (Figure 5 .2e).’The rhenium reaction occurred in moderate yield in DMSO at 130 °C after 1 hr, whereas the99mTcreaction took place at lower temperatures — a nearly quantitative yield was achievedafter 3.5 hr at 95 °C in 1:1 H20: DMSO.’4 The estradiol-substituted Cp ring transfered inbetter yield than the acetylferrocene that was used in the test rection,’4suggesting steric bulkis well tolerated in this reaction. Also of note is that the SLT to the rhenium requiredprotecting groups on the hydroxyl functionality of the steroid, whereas for the technetiumreaction an excellent yield was obtained without the need for such protection.’4 TheseReferences begin on Pg 16] 139reaction conditions were adapted for the99mTcglucosamineconjugates in this chapter.Two glucosamine appended Cp ligands (Figure5.3) and their Re and99mTctricarbonylcomplexes were made previously in theOrvig group.’5 Their ability to inhibit thephosphorylation of glucose by hexokinase wasexamined, and they were found to becompetitive inhibitors of this process, with the mostefficacious (Figure 5.3, R H, M = Re)having K, = 330 ± 70 tiM. This suggests that thesmall size of the Cp ligand may indeed bebeneficial in helping to retain some activity of the carbohydrate in vivo.The compoundswere assayed to determine whether they couldthemselves be phosphorylated by hexokinase,but unfortunately they were found not to be substrates.R0\ROoc__‘90Figure 5.3 Cyclopentadienyl ligands functionalized with glucosamine ligands (RH, Ac), andbound to the tricarbonyl core (M = Re,99mTC)15The current project builds on the promising results of this initial study and further probes theusefulness of compounds with a long alkyl linker between the carbohydrate and the metalchelate. As detailed in Chapters 1.3 and 4.1, the distance between the relatively bulky metalchelate and the carbohydrate (the part expected to interact with key enzymes) is crucial toretaining hexokinase’s recognition of the modified compounds. Thus the purpose of thiswork was to combine the long chain linkers used in Chapter 4 with a significantly smallermetal binding sphere (Cp), with the aim of retaining key biological activities.5.2 Experimental5.2.1 Instruments and MaterialsAll solvents and chemicals were reagent grade and were used as received unless statedotherwise. Reagents were purchased from Acros unless otherwise stated. Solvents were HPLCReferences begin on Pg 161 140grade, and were purchased from Fisher Scientific.Dichioromethane was dried in a solventpurification tower, and was used as dispensed.Hydrogen and argon were purchased fromPraxair.IsolinkTMkits were provided by Mallinckrodt Inc. (now Covidien).Na99mTcO4wasprovided by the Nuclear Medicine Department atthe University of British Columbia Hospital.Tricarbonyl(cyclopentadienylcarboxylic acid)rhenium wasmade in three steps from 1,1’ -ferrocenedicarboxylic acid, according to literature procedures.’2’308-((NN-Dibenzylamino)octylamido)-1,3 ,4,6-tetraacetyl-2-deoxy-D-glucopyranoseand 6-((1V Ndibenzylamino)hexylamido)- 1,3 ,4,6-tetraacetyl-2-deoxy-D-glucopyranosewere first synthesizedby Dr. Z.-F. Chen under my supervision.3’The analytical TLC plates, which were aluminum backed ultra pure silicagel 60, 250 rim, andthe flash column silica gel (standard grade, 60 A, 32 - 63 mm) used wereprovided by Silicycle.‘H and ‘3C APT, 2D COSY, HMQC and HMBC spectra were recorded ona Bruker DRX400instrument at ambient temperature. The NMR spectra are expressed onthe ö scale and werereferenced to the residual peaks of the deuterated solvent. Infrared spectrawere recorded on aNicolet 6700 FT-JR spectrophotometer in transmission mode between 400 and 4000cm’ at aresolution of ± 0.09 cm1. ESI mass spectra were recorded on a MicromassLCT instrument.High resolution mass spectra (Micromass LCT TOF-MS) were providedby the AnalyticalServices Facility, Department of Chemistry, University of British Columbia. HPLC analysis ofcold compounds was done on a Phenomenex Synergi 4 im Hydro-RP 80 A column (250 x 4.6mm) in a Waters WE 600 HPLC system equipped with a 2478 dual wavelength absorbance UVdetector run using the Empower software package. HPLC analyses of radiolabelled complexeswere performed on a Knauer Wellchrom K-l00l HPLC equipped witha K-2501 absorptiondetector and a Capintec radiometric well counter. A Phenomenex Synergi 4 im Hydro-RP 80 Acolumn (250 x 4.6 mm) was used.5.2.2 Synthesis6-Amino-hexylamido-1 ,3,4,6-tetra-O-acety1-2-deoxy--D-gIucopyranose (1)6-((N,N-Dibenzylamino)hexylamido)- 1,3 ,4,6-tetra-O-acetyl-2-deoxy-D-3-glucopyranose (0.50g,0.78 mmol) was dissolved in glacial acetic acid (5 mL), and Pd(OH)2(0.30g, 0.43 mmol) wasadded. The black suspension was stirred vigorously in a Parr hydrogenation bomb. The systemReferences begin on Pg 161 141OAcwas filled with H2 and purged three times, beforebeing filled to 200 psiwith H2 and stirred at room temperature for fourdays. TLC and MS wereAcO—S\1OAc,NHused to confirm the absence of any benzylatedmaterials, and then theo 8N mixturewas filtered through celite and evacuated on a rotary evaporator.‘H NMR spectroscopy showed that the material was pure enoughto beused in the next step, and was also used to quantifrthe amount of aceticAcOH.H2N’acid associated with the product to ensure enoughbase was added toneutralize this. A typical batch was a pale orange oil withtwo acetic acid molecules per productmolecule (0.40 g, 89 % yield by ‘H NMR spectroscopy,used without further purification).8-Amino-octylamido-1,3,4,6-tetra-O-acety1-2-deoxy-D--g1ucopyranose(2)OAc8-((A N-Dibenzylamino)octylamido)- 1,3 ,4,6-tetra-O-acetyl-2-deoxy-D-J3-AcOglucopyranose (0.29g, 0.43 mmol) was hydrogenated over Pd(OH)2(0.30AcO—\i_OAcg, 0.43 mmol) under 200 psi of H2 in a manner similar to that used above.Q 8N 9 After five days the reaction mixture wasfiltered through celite, and the10iivolume reduced on a rotary evaporator and used without further13purification. ‘H NMR spectroscopy showed that the material waspureenough to be used in the next step, and quantified the amount of acetic acidAcOH.H2F’ present (normally around 2 equivalents per product molecule) toensureenough base was added to neutralize this (0.37 mmol,85 % crude yield by 1H NMRspectroscopy).Tricarbonyl{6-amino-hexylamido-N-(1 ,3,4,6-tetra-O-acety1-2-amino-2-deoxy--D-glucopyranose)cyclopentadienyl carboxamide} rhenium (I) (C6(Ac)4Re)OAc Tricarbonyl(cyclopentadienyl carboxylicacid)rhenium (0.058 g, 0.1546[AcO mmol)and 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochlorideAcO— \_OAc7NH(EDC) (0.044 g, 0.23 mmol) were dissolved in dry dichloromethane (5mL). The flask was purged with argon for ten minutes and the resulting11solution stirred at room temperature for three hours, 6-Amino-hexylamino-13 1,3 ,4,6-tetra-O-acetyl-2-deoxy-D--glucopyranose (1) (0.070 g, 0.090mmol) in dry dichloromethane (3 mL) was added to the reaction mixture,cz followed by dusopropylethylamine (0.18 mL, 1.0 mmol). The resultingoc solution was stirred under inert atmosphere at room temperature for 24 hr.References begin on Pg 161 142The reaction mixture was washed twice with water(10 mL), once with brine (10 mL), and driedover MgSO4.The drying agent was filteredout and the solution volume reduced on the rotaryevaporator before purification by column chromatographyon silica gel with 2 % methanol indichioromethane as eluent. The solvent was removed in vacuoto give a colourless oil (17 mg,14 % yield). ‘H NMR (DMSO-d6,400 MHz, ö):8.17 (dd,3J1413a= 5.6 Hz,3J14131,= 5.6 Hz, 1H,H14), 7.92 (d, 3J7,2 = 9.2 Hz, 1H, H7), 6.25 (dd,3Jl7J8a= 2.4 Hz, 3J17],g = 2.0 Hz, 2H, H17),5.73 (d, 3J12 = 7.2 Hz, 1H, H]), 5.70 (dd,3JI8I7a2.4 Hz,3J]8,17b= 1.6 Hz, 2H, H]8), 5.18 (dd,= 10.0 Hz, 3J34 10.0 Hz, 1H, H3), 4.88 (dd,3J4, = 10.0 Hz, 3J4,5 10.0 Hz, 1H, H4), 4.19(dd,3J6a,6b12.8 Hz,3J6a,S4.8 Hz, 1 H, H6a), 3.96 (m, 3H, H2, 5, 6b), 3.09 (m, 2H, H]3), 2.03,2.00, 1.97, 1.94 (s, 3H, OCOCH3,and 2H, H9 overlapped), 1.42,(m, 4H, H]O,]2), 1.23 (m, 2H,Hi]). ‘3C NMR (DMSO-d6,400 MHz, ): 194.06 (Re-CO), 172.32(C8), 169.96, 169.45,169.20, 168.75 (OCOCH3), 161.11 (C]5), 96.15 (C6), 91.71 (C]), 87.05 (C17), 86.14(C]8),72.13 (C3), 71.50 (C5), 68.10 (C4), 61.46 (C6), 51.77 (C2), 38.51 (C13),35.39 (C9), 28.66(C]2), 25.69 (Ci]), 24.95 (dO), 20.43, 20.35, 20.26 (OCOCH3).JRVm (cm’): 3305 (w, br),2936 (w), 2359 (w), 2023 (s), 1916 (s), 1747 (s), 1643 (m), 1544 (m), 1388 (m), 1213(s), 1034(s), 737 (m), 596 (m), 511(m). HR-MS (ES+ of MNa): mlz calcd forC29H35N2O14187ReNa:845.1544, found: 845.1567.Tricarbonyl {8-amino-octylamido-N-(1,3,4,6-tetra-O-acety1-2-amino-2-deoxy--D-glucopyranose)cyclopentadienyl carboxamide) rhenium (I) (C8(Ac)4Re)QAcTricarbonyl(cyclopentadienyl carboxylic acid)rhenium (0.024g, 0.06346LoAc1QAcmmol) and 1 -ethyl-3 -(3 -dimethylaminopropyl)carbodiimide hydrochlorideNH8(EDC) (0.015 g, 0.076 mmol) were dissolved in dry dichioromethane (5mL). The flask was purged with argon and the resulting solution stirred at)11room temperature for four hours. 8-Amino-octylamido-1,3,4,6-tetra-O-13acetyl-2-deoxy-D-f3-glucopyranose (2) (0.070g, 0.090 mmol) and\1415 diisopropylethylamine (0.18 mL, 1.0 mmol) were stirred in dryHN1 6dichloromethane (2 mL) for 10 mm, and then added to the activated acid17Osolution. The resulting solution was stirred under inert atmosphere at room20oco temperature overnight. Once TLC had confirmed the consumption ofstarting material, the solution was washed twice with water (7 mL), once with brine (10 mL),and dried over MgSO4.The drying agent was filtered out and the solution reduced on the rotaryevaporator before purification by column chromatography on silica gel with 1.5 % methanol inReferences begin on Pg 16] 143dichioromethane as eluent. The solvent was removedin vacuo to give the product as acolourless oil (2 mg, 5 % yield). ‘H NMR (MeOH-d4,400 MHz, ö): 6.16 (dd,3J19,20a= 2.3 Hz,3J19,20b2.0 Hz, 2H, H19), 5.78 (d, 3J1,2 = 8.6 Hz, 1H,Hi), 5.57 (dd,3J20,19a= 2.4 Hz,3J2o,19b2.4 Hz, 2H, H20), 5.28 (dd, 3J,2 = 10.6 Hz, 3J34 9.4 Hz, 1H, H3), 5.02(dd, 3J4, = 9.4 Hz, 3J4,5= 10.2 Hz, 1H, H4), 4.28 (dd,3J6a6b12.5 Hz,J6a54.7 Hz, 1H, H6a), 4.10 (dd,J6b,a= 12.5Hz,3J6b5= 2.4 Hz, 1H, H6b), 4.06 (d, 3J21 = 9.0 Hz, 1H, H2), 3.91(ddd,3J5,6a= 4.7 Hz,3J56b2.3 Hz, 3J54 = 10.2 Hz, 1H, H5), 3.26 (dd,3Jj5J4a= 6.6 Hz,3J1514b= 6.6 Hz, 2H, HiS), 2.13(dd,3J9Sa= 7.3 Hz,3J9,8b= 7.3 Hz, 2H, H9), 2.07, 2.05, 2.01, 1.98 (s, 3H, OCOCH3),1.55,(m,411, HiO,i4), 1.29 (m, 611, Hii,12,]3). ‘3C NMR (MeOH-d4,100 MHz,): 194.34 (Re-CO),176.65 (C8), 172.45, 171.82, 171.37, 170.74 (OCOCH3),164.86 (Ci7, 96.23 (Ci8),93.51 (Ci),87.94, 87.91 (C19), 86.62, 86.59 (C20), 73.97 (C5), 73.91 (C3), 69.94 (C4), 63.10 (C6), 54.19(C2), 40.64 (C15), 37.26 (C9), 30.51 (Cii), 30.13 (Ci3), 30.12 (C14), 27.85 (C]2), 26.90 (CiO),20.82, 20.79, 20.70 (OCOCH3).IRVmax (cm1): 3308 (w, br), 2931 (m), 2360 (s), 2341 (s), 2023(s), 1920 (s), 1750 (s), 1646 (m), 1545 (m), 1218 (s), 1038 (m). HR-MS (ES+ of IVfNa): mlzcalcd forC31H39N20,4187ReNa: 873.1857, found: 873.1875.6-Amino-hexy1amino-N-(2-deoxy--D-g1ucopyranose) (3)OH6-((iV N-Dibenzylamino)hexylamido)- 1,3 ,4,6-tetra-O-acetyl-2-deoxy-D-f3-H2\1OHglucopyranose (0.717 g, 1.12 mmol) was dissolved in methanol (15 mL).NH NaOMe (0.907g, 16.8 mmol) was added, and the reaction mixture stirred atOK8room temperature for 2 hr. Amberlite resin was added and theheterogeneous mixture stirred for 15 mm before being filtered. The crudereaction mixture was purified by semi-preparative HPLC to give a paleAcOH.H2Nbrown oil (0.18 g, 32 % yield). The material obtained from the HPLC wasdissolved in glacial acetic acid (5 mL) and added to a hydrogenation bomb with Pd(OH)2(0.18g, 0.26 mmol). The bomb was sealed and purged with H2 three times before being filled with H2to a pressure of 200 psi and the reaction mixture in the bomb stirred at room temperature for fivedays. The mixture was filtered through celite, washed with glacial acetic acid (10 mL) andevacuated on a rotary evaporator before being purified by semi-preparative HPLC. MS and 1HNMR were used to confirm the nature of the oily product that resulted. The amount wasestimated based on a comparison of the integration of the anomeric hydrogens with the Cl3hydrogens of acetic acid.References begin on Pg i6i 1448-Amino-octy1amino-N-(2-deoxy-D--g1ucopyranose) (4)OH8-(N N-Dibenzylamino)octylamido)- 1,3 ,4,6-tetra-O-acetyl-2-deoxy-D-13-glucopyranose (0.256 g, 0.383 mmol) was dissolved in methanol (10 mL).3 7/ NaOMe (0.207 g, 3.83 mmol) was added, and the reaction mixture stirred atO 8room temperature for 4 hr. The reaction was quenchedby stirring withAmberlite ion exchange resin for 10 mm, filteredand purified by semipreparative HPLC (0.090 g, 47 % yield). This product was transfered to a\14hydrogenation bomb with glacial acetic acid (5 mL) andPd(OH)2(0.18 g,AcOH.H2N 0.26 mmol). The system was sealed and purged three times with hydrogenbefore being filled to a pressure of 200 psi and then stirredat room temperature for five days.The mixture was filtered through celite, washed with glacial acetic acid(10 mL) and evacuatedon a rotary evaporator. MS and ‘H NMR spectroscopy were used to confirmthe nature of theoily product that resulted. The amount was estimated basedon a comparison of the integrationof the anomeric hydrogens with the hydrogens of acetic acid.Tricarbonyl{6-amino-hexylamido-N-(2-amino-2-deoxy-D-glucopyranose)cyclopentadienylcarboxamide} rhenium (I) (C6Re)OHTricarbonyl(cyclopentadienyl carboxylic acid)rhenium (0.074g, 0.20 mmol)‘and N,N’-dicyclohexylcarbodiimide (DCC) (0.041g, 0.20 mmol) wereH\\1OHdissolved in dry dichloromethane (10 mL). Activated 4 A molecular sieveswere added, the flask was purged with argon and then stirred at room(\1temperature for four hours. 6-Amino-hexy1amino-N-(2-deoxy-3-D-glucopyranose) (3) (-0.009 g, 0.030 mmol, used as a crude oil fromHN\14debenzylation reaction above) and diisopropylethylamine (1.0 mL, 5.715EzQnimol) were stirred in dimethylformamide (6 mL) for 10 mm, before1addition to the activated rhenium precursor solution. The resulting solutionwas stirred under inert atmosphere at room temperature for 40 hr. The reaction mixture wasfiltered, washed with dimethylformamide (5 mL), reduced on a rotary evaporator, taken up indimethylsulfoxide (7 mL) and the resulting precipitate filtered off. The solution was purified bysemi-preparative HPLC, and the solvent removed in vacuo to give a pale brown oil (0.0080g,41% yield). ‘H NMR (MeOH-d4,400 MHz, 6): 6.16 (dd,3J17,18a= 2.3 Hz, 3J17,18,, = 2.2 Hz, 2H,H]7), 5.57 (dd,3J18,170 = 2.4 Hz,3J18,17b= 2.2 Hz, 2H, H18), 5.10 (d, 3J12 = 3.6 Hz, 0.8 H, H1/3),4.59 (d, 3J1,2 = 8.4 Hz, 0.2 H, Hia), 3.82, 3.70, 3.31, 3.17 (m, 8H, H2-6, a and /3,H13), 2.27 (t,References begin on Pg 161 145‘19,IOa7.6 Hz,3J9,lOb= 7.6 Hz, 2H, H9), 1.57, 1.35 (m, 8H, H10-12). ‘3C NMR (MeOH-d4,100 MHz, 6): 194.36 (Re-CO), 175.24 (C8), 166.38 (C15), 97.22, 92.76 (Cl), 93.92 (C16),87.89 (C17), 86.63 (C]8), 78.20, 78.05, 73.25, 72.75, 70.84 (C3,4,5), 63.01 (C6), 58.86, 55.99(C2), 40.63, 40.60 (C13), 37.03 (C9), 32.11, 30.24, 27.62, 27.18, 26.73 (Cl 0,11,12). JRVm(cm1): 3305 (m, br), 2931 (w), 2360 (m), 2341 (m), 2024 (s), 1921 (s), 1675 (m), 1639 (m),1553 (m), 1203 (m), 1136 (m). HR-MS (ES+ of MNa): mlz calcd forC21H27N20,0185ReNa675.1093, found: 675.1086.Tricarbonyl{8-amino-octylamido-N-(2-amino-2-deoxy-D-glucopyranose)cyclopentadienylcarboxamide} rhenium (I) (C8Re)OHTricarbonyl(cyclopentadienyl carboxylic acid)rhenium (0.085g, 0.22 mmol)OHand dicyclohexylcarbodiimide (DCC) (0.052 g, 0.25 mmol) were dissolvediNHin dry dichloromethane (5 mL). The flask was purged with argon and theO 89 resulting solution stirred at room temperature for 2.5 hr. 8-Amino-octylamino-N-(2-deoxy-D--glucopyranose) (4) (0.25g, 0.23 mmol, asi3estimated from crude ‘H NMR studies of debenzylation reaction shownabove) and diisopropylethylamine (0.74 mL, 4.5 mmol) were stirred in1 15dimethylformamide (5 mL) for 10 mm, before addition to the activated17,-Orhenium precursor. The resulting reaction mixture was stirred under inert19 --—-20atmosphere at room temperature for 40 hr, then the volume reduced on a—rotary evaporator and the compound purified by semi-preparative HPLC.The solvent was removed in vacuo to give a pale brown oil (0.076 g, 50 % yield). ‘H NMR(MeOH-d4,400 MHz, 6): 6.16 (s, 2H, H19), 5.57 (s, 2H, H20), 5.10 (d, 3J1,2 3.2 Hz, 0.75 H,H1/3), 4.59 (d, 3J12 = 8.4 Hz, 0.25 H, Hia), 3.73, 3.24 (m, 8H, H2-6, a and/3, H15), 2.25 (dd,7.4 Hz,3J9,IOb= 7.4 Hz, 2H, H9), 1.84 (m, 4H, H10,14), 1.39 (m, 6H, H11,]2,13). ‘3CNMR (MeOH-d4,100 MHz, 6): 193.80 (Re-CO), 176.33 (C8), 164.33 (C17), 96.78, 92.20 (Cl),95.63 (C18), 87.39 (C19), 86.07 (C20), 77.60, 75.64, 72.68, 72.19, 72.16, 71.88 (C3,4,5), 62.43(C6), 58.27, 55.41 (C2), 40.28, 40.15 (C]5), 36.96, 36.60 (C9), 29.91, 29.71, 29.65, 29.62,29.38, 27.35, 26.48 (C10,]],l2,13,14). IRVmax (cmj: 3037 (w, br), 2934 (w), 2360 (s), 2341 (s),2024 (s), 1921 (s), 1672 (m), 1636 (m), 1557 (w), 1201 (w), 1134 (w). HR-MS (ES+ ofMNa):mlz calcd forC23H31N20,0187ReNa: 705.1434, found: 705.1420.References begin on Pg 161 1462-N-(6-Amino-hexylamido-1,3,4,6-tetra-O-acetyl-2-amino-2-deoxy-D-glucopyranose)-carboxamide -1-ferrocene (C6Fc)QAcFerrocenecarboxylic acid (0.359g, 1.56 mmol) and 1-ethyl-3-(3-AcOdimethylaminopropyl)carbodiimide hydrochloride (EDC) (0.329g, 1.72AcO—-- OAc3 7mmol) were dissolved in dry dichioromethane (10 mL). The flask wasOE 8g purged with argon and the resulting mixture stirred at room temperaturefor1osix hours. 6-Amino-hexylamido-1 ,3,4,6-tetra-O-acetyl-2-deoxy-D-f3-12. i13glucopyranose (1) (amount of debenzylated product quantified byHHN14NMR, 1.40 mmol) and diisopropylethylamine (2.8 mL, 16 mmol) werestirred in dry dichioromethane (3 mL) for 10 mm, before additionvial8’j’ .19syringe to the activated ferrocene. The resulting dark red-brown solutionwas stirred under inert atmosphere at room temperature for40 hr. Thesolution was washed twice with water (2 x 10 mL), once with brine (15 mL), and dried overMgSO4. The drying agent was filtered out and the solution reduced on the rotary evaporatorbefore purification by column chromatography on silica gel with 1 % methanol indichloromethane followed by 2.5 % methanol in dichloromethane as eluent.The solvent wasremoved in vacuo to give an orange oil (0.124g, 12 % yield). ‘H NMR (CDC13,400 MHz, 6):6.05 (s, 1H, H14), 5.93 (d, 3J7,2 = 9.0 Hz, 1H, H7), 5.75 (d, 3J12 9.0 Hz, 1H, Hi), 5.21 (dd, 3J32= 9.4 Hz, 3J,4 9.4 Hz, 1H, H3), 5.14 (dd, 3J43 = 9.4 Hz, 3J45 = 9.0 Hz, 1H, H4), 4.70 (dd,J17,18a= 2.0 Hz,3J17,JSb= 2.0 Hz, 2H, H]7), 4.35 (dd,3J/817a= 2.0 Hz,3J18,]7b2.0 Hz, 2H,H18), 4.28 (m, 2H, H2,6a), 4.22 (s, 5H, Cp-R), 4.13 (dd,3J6b,6a= 1.9 Hz,3J6b,5= 12.4 Hz, 1H,H6b), 3.80 (ddd, 3J5,4 9.8 Hz,3J5,6a= 4.8 Hz,3J5,6b2.3 Hz, 1H, H5), 3.37 (dd, 3J13,12 = 6.6Hz, 3J1312’ = 6.2 Hz, 2H, H13), 2.17 (dd, 3J910 = 7.4 Hz, 3J910’ = 7.4 Hz, 2H, H9), 2.10, 2.10,2.04, 2.03 (s, 3H, OCOCH3),1.61 (m, 4H, HiO,12), 1.38 (m, 2H, Hi]). ‘3C NMR (CDC13,100MHz, 6): 173.00 (C8), 170.93, 170.91 (OCOCH3),169.38 169.20 (C]5), 99.91 (Cl), 92.51 (Cl),76.63 (C16), 72.82 (C18), 72.76 (C]7), 70.32 (C4), 69.95 (C]9), 68.04 (C3), 67.85 (C5), 61.60(C6), 52.82 (C2), 38.92 (C13), 36.10 (C9), 29.16 (Cii), 25.91 (Ci2), 24.56 (dO), 20.84, 20.65,20.51 (OCOCH3).JR Vm (cm’): 3295 (w), 2934 (w), 1740 (s), 1633 (s), 1538 (s), 1366 (m),1222 (m), 1035 (s). HR-MS (ES+ of MNa): mlz calcd forC31H4056FeN2O11Na : 695.1879,found: 695.1895.References begin on Pg 161 1472-N-(8-Amino-octylamido..1,3,4,6-tetra-O-acetyl-2-amino-2-deoxy-D-glucopyranose)-carboxamide -1 -ferrocene (C8Fc)OAcFerrocenecarboxylic acid (0.050g, 0.22 mmol) and 1-ethyl-3-(3-AcO6dimethylaminopropyl)carbodiimide hydrochloride (EDC) (0.046g, 0.24AcO—. ‘J-OAc7,NHmmol) were dissolved in dry dichloromethane(8 mL). The flask wasO 8purged with argon and the resulting mixture stirred atroom temperaturefor four hours. 8-Amino-octylamido- 1,3 ,4,6-tetra-O-acetyl-2-deoxy-D--glucopyranose (2) (0.116g, 0.240 mmol) and diisopropylethylamineç14(0.42 mE, 2.4 mmol) were stirred in dry dichioromethane(5 mL) for 10/15HN1 6 mm, before addition to the activated acid mixture. The resulting solution1 7/019_J8 was stirred under inert atmosphereat room temperature overnight. Once20TLC had confirmed the consumption of starting material the solutionwaswashed twice with water (20 mL), once with brine (20 mE),and driedover MgSO4. The drying agent was filtered out and thesolution reduced on the rotaryevaporator before purification by column chromatography on silica gel with1.5 % methanol indichloromethane as eluent. The solvent was removed in vacuo to give an orangeoil (0.022 g, 15% yield). ‘H NMR (CDC13,400 MHz, ö): 5.94 (d, 3J7,2 = 9.4 Hz, 1H,H7), 5.84 (dd,3J16,15a= 8.6Hz,3JJ6,15b= 8.6 Hz, 1H, H]6), 5.71 (d, 3J12 = 8.6 Hz, 1H, H]), 5.19 (dd, 3J32 9.4 Hz,3J,4 =9.4 Hz, 1H, H3), 5.12 (dd, 3J4, 9.4 Hz, 3J45 = 9.4 Hz, 1H, H4), 4.68 (d, 3J19,20 = 2.0 Hz, 2H,H19), 4.34 (d, 3J20j= 2.0 Hz, 2H, H20), 4.26 (m, 2H, H2,6a), 4.20 (s, 5H, Cp-H), 4.11 (dd,J6ba= 2.4 Hz,3J6b512.5 Hz, 1H, H6b), 3.77 (ddd, 3J5,4 = 9.8 Hz,3J5,6a= 4.7 Hz,3J5,6b= 2.3Hz, 1H, H5), 3.36 (dd, 3J1514 = 6.4 Hz, 3J15,14’= 7.0 Hz, 2H, H15), 2.13 overlapped (d, 3J9,jo =7.4 Hz, 7.4 Hz, 2H, H9), 2.10, 2.09, 2.03, 2.02 (s, 3H, OCOCH3),1.57 (m, 4H, H]O,14), 1.32(m, 6H, H]],]2,]3). ‘3C NMR (CDC13,100 MHz, ö): 173.39(C8), 171.18, 170.88, 170.51(OCOCH3),169.66 169.49 (C]7), 92.84 (Cl), 76.65 (C]8), 73.05 (C3), 72.81 (C5), 70.56 (C20),69.94 (C21), 68.29 (C19), 68.13 (C4), 61.90 (C6), 52.97 (C2), 39.50 (C15), 36.60 (C9), 29.96(C]4), 28.76 (C]2), 28.61 (C13), 26.58 (Cli), 25.46 (dO), 21.12, 20.93, 20.89 (OCOCH3).JRVm (cm’): 3292 (w), 2939 (w), 2360 (s), 2341 (s), 1750 (s), 1630 (m), 1541 (m), 1218 (s), 1038(m). HR-MS (ES+ of MNa): m!z calcd forC33H4456FeN2O11Na:723.2192, found: 723.2187.References begin on Pg 16] 1485.2.399mTechnetiumLabelling via SLTThe appropriate ferrocene precursor was dissolvedin dimethylsulfoxide to a final concentrationof 1 mg’hiL. This precursor solution (300 jiL) was transferedto a 3 mL reaction vial which wassealed with a rubber septum and purged with nitrogenfor 10 mm. An aqueous solution of[99’Tc(CO)3(H2O)j(made using an IsolinkTM kit as detailed in Chapter3.2.3) adjusted to pH 7(300 jiL) was added through the septum, and thereaction mixture stirred at 80 °C for 120 mm.Upon cooling, a sample was injected into the radioHPLCfor analysis. For cell uptake studiesthe reaction mixture was purified by HPLC; the appropriatepeaks were collected and used in thestudies.5.2.4 Hexokinase Substrate AssayThis was performed as detailed in Chapter 3.2.5.5.2.5 GLUT-i Cell Uptake AssayThis was performed as detailed in Chapter 2.2.4.5.3 Results and Discussion5.3.1 Synthesis and CharacterisationThe single ligand transfer (SLT) refers to the transfer of a Cp ligand to the [M(CO)3]corefrom the iron of ferrocene. In the double ligand transfer (DLT) reaction, the metal of interest(in this case rhenium) begins in an oxidised state and over the course of the reaction isreduced, and accepts three carbonyl ligands and a Cp ligand from another metal speciespresent in the reaction mixture. These reactions (detailed in Chapter 5.1.1) have both beenshown to be selective for the transfer of Cp rings bearing electron-withdrawingsubstituents.12These reactions are both employed in the syntheses in this Chapter. The SLT was utilized inthe99mTCchemistry;[99mTc(CO)3(H20)1is reacted with a substituted ferrocene whichReferences begin on Pg 16] 149transfers its functionalized ring to the technetiumcentre during the reaction. This reactionwas not applied to the rhenium compounds becauseit has been found not to work for similarcarbohydrate-based ligands.’5 The reaction conditionsare likely too harsh, and thecombination of high temperatures and solventrequirements produced only degradationproducts.’5 Instead, the DLT was used to synthesize the rheniumcomplexes via an indirectroute of making [CpM(CO)3]followedby coupling to the biomolecule of interest. Thesynthesis of rhenium compounds of this type viaan indirect route where the Cp is bound tothe Re and then functionalized further is not unique, and iscommonly utilized whenlabelling proteins and other chemically sensitive biomolecules.22’27, 32OAcAc0AcO———._- OAcNHOH(-NHF?0OAcn=5 C6FcAcO_-OAcn=7 C8FcNHoOAc------NH2HOAcAcO\\Re0AcO—-—-_-OAcn = 5 1 aN Hn = 7 2S n)oocn 5 C6(Ac)4Ren = 7 C8(Ac)ReOH0HO\—- OHOHNH- 0 CO A Is.HOHO—‘OH__________NH‘—)—-NH2HOAcReOCCOn=5 3n=7 4n=5 C6Ren=7 C8ReScheme 5.5 Synthetic route to cold compounds used in this chapter: a) dry DCM, EDC, DIPEAb) dry DCM / DMSO, DCC, DIPEA.References begin on Pg 161 150As discussed in Chapter 5.1, there are several ways in whichCp compounds of this type can besynthesized. The aim of this work was to make long chaincarbohydrate Cp complexes, andexamine their in vitro properties relevant to molecularimaging. To this end, reactions that werethought to be most likely to work in systems of thiskind were employed. If the resultingcomplexes were found to have promising properties, synthetic routescould be optimized byinvestigation of the wide variety of methods available.The rhenium compounds were made via a two step process.The DLT was used to synthesizethe rhenium starting material: [HO2CC5H4Re(CO)3],’followedby an amide coupling to theappropriate sugar-linker compound to give the functionalized metal complex (Scheme 5.5).Once the protected sugar was bound to the [CpRe(CO)3]core, the removal of the acetylprotecting groups was attempted using NaOMe in methanol. This was notsuccessful, and alarge number of compounds were evidenced in both the MS and the TLC, including somewithout the rhenium isotope pattern. The use of NaOMe induced decomposition,as alsooccurred in the short-chain analogues of these Cp compounds.33 Because of this the free-sugar complexes were made by deprotection of the sugar prior to coupling of the sugar-linkercompound and the [CpRe(CO)3]core (Scheme 5.5). Sugar-substituted ferrocenes wereprepared as substrates for the SLT reaction to access the[Cp99mTc(CO)3]complexes. Thesewere made utilizing an amide coupling between the terminal N of the appropriate sugar-linker compound and the acid group of ferrocenecarboxylic acid (Scheme 5.5).The ligands, their rhenium complexes, and ferrocene analogues used in the99mTclabellingwere all characterised by HR-MS, IR, ‘H and ‘3C APT NMR spectroscopy. A variety of 2-DNMR spectroscopic techniques such as COSY, HMQC, HMBC and TOCSY were used asappropriate for full assignment of ‘H and ‘3C NMR signals.Deprotection of the long chain sugar starting materials was assessed using crude ‘H NMRspectroscopy. For the debenzylation reactions to make 1 and 2, ‘H NMR spectra were used toconfirm the absence of any aromatic resonances. As these reactions were performed in aceticacid, integrals of the acetic acid CH3 peak compared to one of the product peaks gave the yieldof product as well as the stoichiometry of acetic acid so sufficient base could be added in thenext step to neutralize it completely. Compounds 3 and 4 were deacetylated prior to coupling tothe metal centre, and ‘H NMR spectroscopy was used to ensure loss of both the acetyl and theReferences begin on Pg 161 151aromatic peaks. Once the acetyl groups were removed,the free hydroxyl groups complicate the‘H NMR spectra to the point where they are not readilyassignable, and their main use is toconfirm the lack of acetyl peaks. These free sugar compounds were then debenzylated, and the‘H NMR spectra became even more difficult to interpret. The spectra were used to confirm thelack of aromatic peaks as well as to quantify the acetic acid presentby comparing the integrals ofthe C-i anomeric peaks (the only quantifiable ones in the spectrum), with the acetic acid peak,asdetailed above.‘H and ‘3C APT NMR spectroscopy were performed to confirm the identity of all final products.Figure 5.4 shows the ‘H NMR spectra of the three non-radioactive C6 metal complexes; thetrends are exactly the same in the C8 compounds. The ferrocene complex C6Fc and theprotected rhenium complex C6(Ac)4Re (Figure 5.4a and b) both display clean spectrawith thesugar peaks well defined and separated. The protons in the aliphatic linker chains are not welldistinguished, but the chemical shifts and integrals match those of the expectedproducts.Comparing the ‘H NMR spectra corresponding to these two compounds (Figure 5.4a and b), thelarge chemical shift changes affected by the different metal ions can be observed. The sugarpeaks in the rhenium complex are spread further apart than those of the ferrocene. This is a verylarge effect, especially given the substantial distance between the atoms of the carbohydrate andthe metal ions. The peaks corresponding to the Cp protons are shifted apart and substantiallydownfield in the rhenium compound compared to the analogous ferrocene. The fact that theseprotons exhibit the largest differences between the two molecules helps to confirm that the Cprings are indeed the part of the ligand system bound to the metal ion.The ‘H NMR spectra of the protected and deprotected rhenium complexes C6(Ac)4Re and C6Recan be compared by looking at Figure 5.4b and c. Due to solubility constraints these two spectrawere run in different solvents, meaning the chemical shifts are not directly comparable. Thegeneral trend can be observed, and the sugar protons seen to move upfield, narrow together andbecome very messy and difficult to interpret with deprotection of the acetyl groups. This isnormal for deprotected sugars, and can also be observed in the ‘H NMR spectra shown inChapters 2 and 3 of this thesis. These ‘H NMR spectra are not easy to interpret and therefore notas diagnostic as those of protected sugars, so ‘3C and 2-D NMR spectroscopies are particularlyuseful in characterizing compounds containing free sugars.References begin on Pg 161 1521.5 1 0.5ppmFigure 5.4 ‘H NMR spectra of: a) C6Fc b) C6(Ac)4Rec) C6Re,*indicates solvent peak.Figure 5.5 shows the ‘3C APT spectra of the two C8 rhenium complexes made in this chapter.Figure 5.5a corresponds to the tetraacetylated compound C8(Ac)4Re, and contains four peaks ateach of 180 ppm and 20 ppm corresponding to the acetyl carbons which are absent in thespectrum of the deprotected analogue C8Re (Figure 5.5b). The other main difference betweenthe two spectra is the number of peaks present. Upon deprotection of the sugar an anomericcentre is produced at C-i, meaning two diast ereomeric products are formed with differentchemical shifts. As the two diastereomers are the same at all centres bar one, the signals of someatoms (particularly those further away from the anomeric centre) are the same for the twostructures, meaning not all peaks are observed to double. For example, the CO peaks at —193ppm do not double as they are distant from C-I, but the proximal C-2 goes from being a singlepeak to two peaks separated by 2.9 ppm upon deprotection of the acetyl groups.a1817719!4b1012‘3NIl 4C13182,6ab13118 7.5 7 6.5 6 5.5 5 4.5 4 3.5 3 2.5 2References begin on Pg 161 153OH46oHO’1OH\1415HN 1617’O198CO 20OCOAcAcO1OAc3/ Iic1213K14/152,10HN16COFigure 5.5 ‘3C APT spectra of the two C8 rhenium complexes made in this chapter: a)C8(Ac)4Re with acetylated sugars b) C8Re the deprotected sugar analogue,*indicates solventpeak.aCo68OAc171511,13149bI I I.18192043,517OAc10,11,12,13,141561891’22200 190 180 170 160 150 140 130 120 11019,20100 90 80 70 60 50 40 30 20 10 0References begin on Pg 16] 154The peak at - 193 ppm in the spectra of the rheniumcomplexes correspond to the three COligands, and there is only one of these peaks observed foreach Cp compound. This is in contrastto the tridentate ligands discussed in Chapter 3 where the asymmetric coordination environmentof the metal ion led to two or three CO peaks being present. The Cp ligandbinds to the metal inar15fashion in these compounds, and is free to rotate. This means only one peakis produced bythe three CO ligands, as they all experience the same chemical environmenton the NMRtimescale.The JR spectra of compounds can be used to identify the presence, absence or changes in theenvironment of major functional groups. As expected, the analogous protected and deprotectedrhenium compounds showed very similar JR traces (Figure 5.6 shows representativeC8 spectra).The key difference between the two is the presence of an ester carbonyl stretch at 1750 cm1 inthe protected compounds (e.g. C8(Ac)4Re) that is not seen in the deprotected analogues(e.g.C8Re) which do not contain acetyl (ester) groups. The peak at 1218 cm1 is assignedto the C-Ostretch of the esters, and this peak also disappears following deprotection. Concurrentenhancement of the —OH peaks at 3300 cm1 confirms the presence of free hydroxyl groups.1 Re100tL.490 - C8(Ac)4Re80 -%T70 -60 -50+4000 3500 3000 2500 2000Wavenumbers (cm1)Figure 5.6 Overlaid JR spectra of C8Re (gray line,C8(Ac)4Re (black line, top left 95 % transmittance).top left, 100 % transmittance) and1500 1000 500References begin on Pg 161 155The JR spectra of the ferrocene and (Ac)4Re analogues are also very similar. The majordifference is the presence of the strong peaksaround 2022 and 1918 cm1 in the rheniumcompounds, corresponding to the CO stretches of thecarbonyl ligands. As the remainder ofeach of these two molecules is essentially the same and all other peaks areas expected, this is inline with the proposed structures. For example the peaks correspondingto the amide carbonylsappear at 1747 cm in C6(Ac)4Re and 1740 cm1 in C6Fc, and the ester peaks at 1643 cm1 inC6(Ac)4Reand 1633 cm1 in C6Fc.HR-MS gave the expected peaks for M+H and/or M+Na, with the rhenium complexesexhibiting their diagnostic isotope pattern (‘85Re = 37 % and‘87Re = 63 % natural abundance).A diagnostic example is the HR-MS of C6Re with peaks at 675.1086(--45 % intensity) and677.1046 (100 % intensity) corresponding toC18H26N4O12185Re andC18H26N4O12187Re,respectively.The characterization data discussed above is all consistent with the proposed structures of thecompounds synthesized.Table 5.1 Retention times (RT) and yields of99mTclabelling and HPLC analysis of two longchain glucosamine-based Cp ligands.RT*RT*RT*RadiochemicalLigandC11(Ac)4Re CAReCn99mTcyieldC6 23.2mm 17.0mm 16.8mm 25%C8 29.7 mm 20.2 mm 20.3 mm 40 %*HPLC conditions— 100 % H20 (with 0.1 % TFA) linear gradient to 100 % ACN at 30 mm.The labelling reaction to make the[Cp99mTc(CO)31complexes required slightly longer reactiontimes than those needed for the tridentate ligands discussed in Chapters 3 and 4; however, giventhe half life of99mTc(6 hr), 2 hr is an acceptable reaction time. The aqueous solution of[99mTc(CO)3(H2O)]+was neutralized to pH 7 before being added to the ferrocene compound tobegin the SLT reaction. It was, therefore, somewhat surprising to observe that the acetylReferences begin on Pg 161 156protecting groups on the glucosamine were removedduring the reaction (Scheme 5.6). This wasevidenced by the HPLC retention time of a major componentof the reaction mixture coincidingwith that of the deprotected rhenium standard. There wereno compounds with higher retentiontimes to coincide with those of the protected rhenium compound (see Table5.1). Furtherevidence of the deacetylation was seen in that reactionwith NaOMe in MeOH did not change theHPLC trace of either the C6 or the C8 system, as it wouldif there were acetyl groups in themolecule. Upon further reflection, this in situ deprotectionis probably to be expected, as thereaction involves elevated temperatures for two hours,and although the pH is neutral as testedby pH paper, there will be some OW ions in the reaction mixture, especially giventhat theIsolinkTMkit used to make the99mTcprecursor is very basic.OAcOHAcO°’\OAcHOR\OH[99mTC(CO)(HO)]+/ /HN DMSO:H201:180 °C, 2 hr99m]JCOc_Scheme 5.6 Labelling scheme forC699mTc.The same is used for the C8 analogue.The radiation traces of these reaction mixtures (Figure 5.7) show unreacted starting material,product and small amounts of other compounds. As there is still99mTcstarting material presentafter 2 hr, it seems that it would be possible to increase the yield by simply reacting at highertemperature or for a longer time. This is not the case though, and longer reaction times lead tothe appearance of more peaks and decreased yields of the desired product. It seems that by thistime point the ferrocene precursor is no longer able to transfer a ring. One possible explanationfor this is that the protecting groups are all removed by the 2 hr time point in the reaction. Thiswould rationalize the observations that either increasing the reaction temperature or time doesnot increase the yield as it may be that only protected ligands will transfer. There wouldtherefore be two competing reactions occurring — the glucosamine-bearing ring could eitherReferences begin on Pg 16] 157transfer or be deprotected, and the relative timescales under these conditions may be such thatthe maximum yield is obtained within 2 hr. Other decompositionroutes are also possible. It isworth noting that no reaction is observed after 3hr at60 °C by radioHPLC. This does not ruleout the possibility that the ferrocene precursor is decomposing at this temperature, but informsthat the speciation of the99mTcdoes not change under these conditions.Others have also found the need for the protection of hydroxyl groups in reactions of this type.’4When making rhenium estradiol derivatives via a SLT, it wasfound that the deprotectedcompound yielded only 12 % while protection of the two hydroxyl groups increased the yieldto41 % under the same reaction conditions.’4.1’II rJI:‘0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25Time (mm)Figure 5.7 Radiation HPLC trace ofC899mTc(CO)3after 2 hr reaction time. The peak at 20 mmcorresponds to the deprotected product, 18 mm is tricarbonyl starting material and 17 mm is anunidentified side product of the labelling reaction.5.3.2 Hexokinase Substrate AssayThe rhenium complexes C6Re and C8Re were tested as substrates for the enzyme hexokinase.They were both found not to be substrates as they did not undergo a significant amount ofphosphorylation under the conditions tested. As with the long-chain tridentate compounds inChapter 4, the exact reasons for this have not been elucidated. It may be that the presence of thechain itself perturbs the carbohydrate too much for it to be recognized, or that the chain needs tobe a little longer to place the metal binding sphere far enough away from the active site to allowhexokinase to fully close and perform the phosphorylation.References begin on Pg 16] 1585.3.3 GLUT-i Cell Uptake StudiesThe two99mTccompounds C699’Tc andC899mTCwere tested for their uptake in cells with highconcentrations of GLUT-i transporters on their surface (Figure5.8). Both compounds showednon-negligible uptake; however, there was no significant decreasein their uptake in the presenceof glucose, indicating that the compounds were not transported intothe cells as carbohydrates.Comparing the magnitude of these cellular uptakes tothose of the long-chain tridentate ligandsdiscussed in Chapter 4 reveals no significant variation inuptake with binding sphere (Figure5.8). The cell uptake is not significantly dependent on the size or lipophilicity of the metalbinding portion in these compounds. It may be that the additionof a long aikyl linker attached tothe glucosamine is enough of a perturbation that it prevents recognition and transport of thecompounds as carbohydrates.0.20% -0.15%0.10%0.05%0.00%Figure 5.8 Comparison of cell uptake ofC699mTcandC899mTc(black) with the equivalentchain length tridentate99mTccomplexes (blue) in the absence (spotted) and presence (striped) ofglucose.::::::::IPBSPBS+lOmMglucoseC6Cp C8Cp C6tri C8triReferences begin on Pg 16] 1595.4 ConclusionSynthesis of two long chain glucosamine basedcyclopentadienyl complexes of both thetechnetium and rhenium tricarbonyl coreswas successfully completed. The resultingcompounds were not recognized or metabolizedby either GLUT-i or hexokinase, so willunfortunately not be useful as carbohydrate basedmolecular imaging agents. There are manymethods known for the preparation of these typesof substituted Cp complexes and furthersynthetic optimization could be carried out if morepromising analogues were discovered.References begin on Pg 161 1605.5 References1. Schibli, R.; Schubiger, P. A., Eur. J Nuci.Med. 2002, 29, 1529 - 1542.2. Uehara, T.; Uemara, T.; Hirabayashi, S.; Adachi,S.; Odaka, K.; Akizawa, H.; Magata,Y.; Irie, T.; Arano, Y., J Med. Chem. 2007,50, 543 - 549.3. Schulz, H., Biochim. Biophys. ActaLipid LipidMet. 1991,1081, 109- 120.4. Cotton, F. A.; Wilkinson, G., Advanced InorganicChemistry. John Wiley & Sons: NewYork, l988;p 1171-1177.5. Khotsyanova, T. L.; Kuznetsov, S. I.; Bryukhova,E. V.; Makarov, Y. V., I Organomet.Chem. 1975, 88, 351 -356.6. Crocker, L. S.; Gould, G. L.; Heinekey, M., I Organomet.Chem. 1988, 342, 243 - 244.7. Alberto, R.; Schibli, R.; Elgi, A.; Schubiger, P. A.,I Am. Chem. Soc. 1998, 120, 7987 -7988.8. Le Bideau, F.; Salmain, M.; Top, S.; Jaouen, G., Chem.- Eur. 1 2001, 7, 2289 - 2294.9. Wenzel, M., I Label. Compd Radiopharm. 1992, 3], 641 - 649.10. Wenzel, M.; Saidi, M., I Label. Compd. Radiopharm.1993, 33, 77 - 80.11. Saidi, M.; Kothari, K.; Pillai, M. R. A.; Hassan, A.; Sarma, H. D.; Chaudhari, P.R.;Unnikrashnan, T. P.; Korde, A.; Azzouz, Z., I Label. CompcL Radiopharm. 2001, 44,603 -618.12. Spradau, T. W.; Katzenellenbogen, J. A., Organometallics 1998, 17, 2009 - 2017.13. Top, S.; Masi, S.; Jaouen, G., Eur. I Inorg. Chem. 2002, 1848 - 1853.14. Masi, S.; Top, S.; Boubekeur, L.; Jaouen, G.; Mundwiler, S.; Spingler, B.; Alberto,R.,Eur. I Inorg. Chem. 2004, 2013 - 2017.15. Ferreira, C. L.; Ewart, C. B.; Bayly, S. R.; Patrick, B. 0.; Steele, J.; Adam, M. J.; Orvig,C., Inorg. Chem. 2006, 45, 6979 - 6987.16. Alberto, R.; Egli, A.; Abram, U.; Hegetschweiler, K.; Gramlich, V.; Schubiger, P. A.,IChem. Soc., Dalton Trans. 1994, 2815 - 2820.17. Wald, J.; Alberto, R.; Ortner, K.; Candreia, L., Angew. Chem. mt. Ed. 2001, 40, 3062 -3066.18. Minutolo, F.; Katzenellenbogen, J. A., I Am. Chem. Soc. 1998, 120, 4514 - 4515.16119. Top, S.; Kaloun, E. B.; Jaouen,G., I Am. Chem. Soc. 2000, 122, 736 - 737.20. Top, S.; Kaloun, E. B.; Toppi,S.; Herrbach, A.; McGiinchey, M. J.; Jaouen, G.,Organometallics 2001, 20, 4554 - 4561.21. Liu, Y.; Springier, B.; Schmutz,P.; Alberto, R., I Am. Chem. Soc. 2008, 130, 1554 -1555.22. Salmain, M.; Gunn, M.; Gorfti, A.; Top, S.; Jaouen,G., Bioconjugate Chem. 1993, 4, 425-433.23. Top, S.; Vessieres, A.; Pigeon, P.; Rager, M.-N.;Huche, M.; Salomon, E.; Cabestaing,C.; Vaissermann, J.; Jaouen, G., ChemBioChem 2004, 5, 1104 - 1113.24. Bernard, J.; Ortner, K.; Spingler, B.; Pietzsch, H.-J.; Alberto, R., Inorg.Chem. 2003, 42,1014- 1022.25. Graeff, F. G.; Guimaraes, F. S.; De Andrade,T. G.; Deakin, J. F., Pharmacol. Biochem.Be. 1996, 54, 129- 141.26. Alberto, R.; Schibli, R.; Schubiger,P. A.; Abram, U.; Pietzsch, H.-J.; Johannsen, B., IAm. Chem. Soc. 1999, 121, 6076 - 6077.27. Tamagnan, G.; Baldwin, R. M.; Kula,N. S.; Baldessarini, R. J.; Innis, R. B., Bioorg.Med. Chem. Lett. 2000,10, 1113- 1115.28. Skaddan, M. B.; Wust, F.; Katzenellenbogen, J. A., I Org. Chem. 1999, 64, 8108- 8121.29. Wust, F.; Skaddan, M. B.; Leibnitz, P.; Spies, H.; Katzenellenbogen, J. A.; Johannsen,B., Bioorg. Med. Chem. 1999, 7, 1827 - 1835.30. Spradau, T. W.; Katzenelienbogen, J. A., Bioconjugate Chem. 1998,9, 765 - 772.31. Bowen, M. L.; Chen, Z.-F.; Roos, A.; Adam, M. J.; Orvig,C., manuscript in preparation.32. Spradau, T. W.; Edwards, W. B.; Anderson, C. J.; Welch, M. J.; Katzenellenbogen, J. A.,Nuci. Med. Biol. 1999, 26, 1 - 7.33. Ferreira, C. L.; Orvig, C., unpublished results.162CHAPTER 6Conclusions and Future Work6.1 Multifunctional Metal Chelators as Potential Therapeutics for Alzheimer’sDisease6.1.1 3-Hydroxy-4-pyridinonesChapter 2 of this thesis focused on the radioiodinelabelling and subsequent study of twoprototypical 3-hydroxy-4-pyridinones.’ These two compounds were chosenas they represent therange of compounds synthesized as part of the Orvig group’s Alzheimer’sdisease (AD) researchprogramme. Iodine, specifically125jwas chosen as the radioactive isotope for this work becauseit was suited to our collaborators’ experimental apparatus, was readily availablethroughTRIUMF, and has well established labelling chemistry.2’3lododestannylation is a commonlyused method for regioselective addition of a radioactive iodine. This methodologywas takenadvantage of in the labelling reactions in this thesis, and as it can only be applied to substrateswith aromatic rings, this restricted the choice of which compoundsto label. The substraterestriction comes from the fact that the Sn-alkyl bond is stronger than the Sn-aryl bond. AsSn(alkyl)3species are the leaving groups substituted by an iodine, the substrate to be labelledmust be aromatic to ensure that that is the bond that breaks and the iodine will endup bound tothe aryl ring, not an alkyl group. It may be that other, non-aromatic analogues will need to beradiolabelled and investigated as this project progresses, and in thiscase different labellingmethodologies will need to be employed.The in vivo experiments performed on the iodinated compounds showed that they are able topermeate the blood brain barrier (BBB) to enter into the brain.’ The concentrations in whichthey access the brain are fairly modest, but may be sufficient to show some effects in vivo.These labelling experiments could be extended to elucidate more information on the mechanismof brain uptake as well as biodistribution and glycosidic cleavage of the compounds in vivo.Further in vivo experiments with the same radiolabelled compounds could provide moremechanistic information. There are two experiments that would potentially give useful insight:comparing the brain uptake of glycosylated and non-glycosylated analogues (e.g. those in FigureReferences begin on Pg 173 1636.1) of a given pyridinone, or by performing uptakestudies in the presence of compounds knownto inhibit certain BBB transport proteins. Comparingthe brain uptake of glycosylated and nonglycosylated analogues would give information on whether the glucose moietywas increasing ordecreasing the BBB permeation. The non-glycosylated molecules are smaller,and less polar,and therefore would be expected to undergo passive diffusion across the BBBmore rapidly.4 Ifthe GLUT transporters were being utilized by the glycosylated compoundsthen removal of theglucose group would decrease the amount of compound seen in the brain.This experimentwould not give a definitive mechanism of uptake, but would add more valuableinformation toour overall knowledge. Some work toward this goal was carried outby synthesizing the nonglycosylated analogue of pyrA (Figure 6.1). This was not completed in timefor the scheduledin vivo experiments, and given the relatively limited mechanistic informationit would haveprovided it was decided not to carry out another roundof experiments for this alone. Thesynthesis of this compound is fairly analogous to that performedin Chapter 2; an acetylprotecting groups is used, and cleaved following iodination, though there isno glycosylationstep.OH00HO. HO.HO-- 0.OHFigure 6.1 Non-glycosylated (left) and glycosylated (right) 3 -hydroxy-4-pyridinoneprototypes.The addition of known inhibitors of BBB transport proteins to the perfusion experiments wouldgive useful information on how our compounds access the brain. This would involve repeatingthe rat brain perfusion experiments, as before and in the presence of large amounts ofcompounds such as D-glucose and cytochalasin B, both of which are known to decrease theamount of a compound taken up via the GLUT transporters.5 A decrease in the amountof testcompound accessing the brain in the presence of such inhibitors would provide evidence for theiruptake being via the GLUT route. If no such difference was seen between trials it wouldstrongly suggest that the compounds were not taken up by the same route as glucose. There areReferences begin on Pg 173 164many different channels and proteins used for transportthrough cell membranes, and althoughhaving a glucose appended increases a compound’schance of being transported by a GLUT, itdoes not preclude its acceptance by other typesof transport proteins. To test for receptormediated uptake in general, an experiment could bedone using increasing concentrations of testcompound. If receptors were being utilized for transport,then above a certain concentration thereceptors would become saturated, and uptake wouldno longer increase. If uptake was viapassive diffusion then it would increase in a linear mannerwith increasing concentration of testcompound. If this experiment showed our compoundsto be transported via some kind ofreceptor, a battery of inhibitors of various channelscould be tested. For example, theNa/KATPase channel has been found to bethe point of access for compounds such ascisplatin;6the known Na7KATPase inhibitor ouabain7could be added to the assay and anyaffect this had on uptake would suggestan interaction of the test compound with Na/KATPase.Another very useful, but much more elaborate, in vivoexperiment would be the monitoring ofglycosyl bond cleavage and biodistribution by a doublelabelling procedure.8 The two differentlabels would be chosen to emit radiation with significantly different energies, andeach detectorset to monitor just one of these energies. In this way the location of the twodifferent atomscould be separately tracked. For the glycosylated pyridinones it would be mostinteresting toplace one radioactive atom on the pyridinone portion of the molecule andthe other on thecarbohydrate. To minimize chemical changes to the carbohydrate a radioactiveisotope ofcarbon or hydrogen would be preferable. 3H (18.6 keV, ti,2 12.3 yr)and 14C (156 keV, ti,2 =5730 yr) could be used in combination with1231(159 keV) and1251(35 keV) respectively.Stability studies in plasma or a similar medium prior to in vivo experiments wouldbe used todetermine the relative stabilities of the different bonds in the molecule. It is expectedthat theglycosidic bond would be the weakest, and if the isotopic distributions werefound to differ fromeach other in vivo, this would indicate cleavage of the glycosidic bond. Thein vivoexperimentation would be quite complex and the animals would needto be monitored over asignificant length of time. Given the effort required for these experiments, it maybe a better useof resources to use non-labelled compounds and have collaborators perform a generalpharmacokinetic/pharmacodynamic study of the compounds. This could be donein normal miceto ensure the glycosidic bond cleaved and examine what portion of the administered compoundsended up in the brain. After these initial experiments it would be extremely interestingto seeReferences begin on Pg 173 165how these compounds fared in a mouse model of AD,and to observe if there was any therapeuticeffect, as we would hope.Another complicating factor in this work is that the BBB is often compromisedin AD patients.9This is thought to result in a decrease in concentrationor activity of the GLUT transporters ofthe BBB10, 11It has been suggested that a reason for this is the oxidativedamage that occursbecause of the Af3—metal ion complexes.’2 Thisshows another potential advantage of ourmultifunctional approach to AD, but also complicatespredictions of efficacy of compounds indifferent patients. Because of these complications it wouldbe very useful to examine the BBBpermeability of these glycosylated pyridinones in normal vs.AD human brain. It may be that adamaged BBB is more permeable to such compounds as the tight endothelialjunctions that makeup this barrier may be weakened such that brain access is increased. Experimentswould need tobe done to determine this.Before such detailed studies would be carried out, basic efficacy must be provenby determiningwhether these compounds elicit any response in an animal model of AD. Any future work,be itbiological testing or compound development and synthesis, will need to be based on the outcomeof such studies.6.1.2 N,N,O,O-Tetradentate AminophenolsAnother major class of compounds that have been studied in the Orvig group as potential ADtreatments are the tetradentate N20 aminophenols.13’ An example of a moleculeof this classis shown in Figure 6.2. Like the pyridinones that were the focusof Chapter 2 of this thesis, someof these compounds were glycosyl protected to form prodrugs. Some work has been put into theradiolabelling of compounds of this type, such that they too could be testedfor BBBpermeability. Labelling was attempted, in an effort to synthesize the compound shown in Figure6.2, using a simple electrophilic aromatic substitution reaction.3 This particular substrate waschosen for its large degree of substitution on the aromatic rings, restricting the number ofpossible iodination products. The different substituents exert conflicting directing effects suchthat there is no clear preference of position for the iodination to occur. In addition, there are tworings present, so the possibility of one or both rings being iodinated adds further complications.As the amount of activity in the rat brain is what is being quantified, it is vital that each moleculeReferences begin on Pg 173 166contains the same amount of activity, meaning thesecompounds would need to be separatedbefore testing. It would be interestingto see the results of BBB permeation tests on compoundsof this type, and if positive, to pursuethe other in vivo experiments listed above in Chapter 6.1.1with some prototypical aminophenols.Figure 6.2 The aminophenol that has undergonesome radioiodination attempts. The position ofexpected iodination is shown.6.2 Carbohydrate Conjugates of99mTcfor Use in Molecular ImagingThe ultimate goal of the work in Chapters 3 - 5 of this thesiswas to make aglucosamine99mTcbioconjugate for use in molecular imaging. As discussed in Chapter 1 and theIntroductions tosubsequent Chapters, a successful carbohydrate-based imaging agentmust be able to both entercells and be phosphorylated in a manner similar to glucose. The end use of animaging agent ofthis type would be as a marker of glucose metabolism, and as such the compoundneeds tobehave similarly to glucose. As glucose metabolism is altered in heart disease,brain disorders,and cancer, this is where a compound of this type wouldbe most useful. As detailed in Chapter1.3, many cancer cells have increased energy requirementsdue to their rapid growth andinefficient metabolic pathways. For an imaging agent to be selectivefor cancerous tissue it mustenter into and get trapped in those cells faster than in healthy tissue. Theincreased glucoseneeds of many cancer cells have led to them expressing higherconcentrations of key glucoseprocessing enzymes than normal cells. This is a key difference that we, andothers, hope to takeadvantage of with carbohydrate-based radiopharmaceuticals. The two key enzymesto betargeted are the GLUTs and hexokinase. GLUTs are the major classof glucose transportenzymes, which, under ideal conditions, may be utilizedto transport bioconjugates insidecells.’’7 Hexokinase is the enzyme that catalyses the first step in glycolysisby addition of aphosphate group to the C-6 position of a carbohydrate. The resulting moleculeis negativelycharged, and as it is unable to diffuse back through the cell membrane, it is retained within theReferences begin on Pg 173 167cell. Thus the action of these twoproteins working together results in molecules that aremetabolized as carbohydrates entering(GLUT) and being trapped (hexokinase) selectively incells that express high concentrationsof both proteins.The way in which the Orvig group has approachedthe problem of retaining GLUT transport isby choosing glucosamine as our sugar analog. Several functionalisedglucosamine analogueshave shown cell uptake by GLUT transporters. 2-NBDG(Figure 6.3a) is a compound with arelatively bulky group attached to the N of glucosaminethat undergoes transport by GLUT-iinto human erythrocytes, as well as hexokinasephosphorylation.’ Two other bulkyglucosamine derivatives that are transported by the GLUT family areshown in Figure 1.4 inChapter1.3.16,17We propose that the reason substituted glucosaminescan be tolerated wheresubstituted glucoses cannot is because of theability of the N-H to participate in hydrogenbonding. An accepted mechanism of transport forcompounds through the GLUT channels is bya rolling mechanism where each group around the carbohydrate must beable to form a hydrogenbond at various stages of the movement.’8 It is possiblethat investigation of other sugars forthese bioconjugates would be advantageous, but given the evidencepresented above it seemsworth pursuing some of the other suggestions in this chapter before movingon to variations insugar.aOHbOHcOHdOHH°’\HO\°\ HO\SN02/(COOH COOH‘—C00H ‘—COOHFigure 6.3 Carbohydrate based bioconjugates of interest for their interactions withkey enzymesand/or investigation as molecular imaging agents, a) 2-NBDG15 b) N-benzoyl glucosamine’9c) along alkyl chain glucose analogue investigated for use in molecular imaging5d)a glucose-basedcompound with a long polyethyleneglycol (PEG) chain investigated foruse in molecularimaging.5References begin on Pg 173 168The Orvig group’s approach to synthesizing compoundswith hexokinase activity has been theuse of long, alkyl linkers between the carbohydrate and the metal bindingmoiety. This approachis justified by the fact that the active site of hexokinase isin a narrow cleft that must close inorder for phosphorylation to occur. If the bulky metal bindingsubstituent is far enough from thecarbohydrate in the active site it may be possible for this processto occur without the chelatingarms getting in the way and blocking the subunits from closing together.The minimum distancerequired to allow for this to occur was calculatedto be seven methylene units long.5 This thesishas explored the in vitro properties of some compounds with linkers up to seven methylene unitslong, and these were all found not to be phosphorylated by hexokinase.It may be that the bulkof the metal binding groups used required the linker to be a little longerto move them slightlyfurther away and allow the cleft to close completely. To that end it would be very interesting tomake the C9 andior Cl 0 analogue of the compounds outlined in this thesis. The C9 compoundswould likely have sufficient water solubility to allow for the appropriate assays to be performed,and certain analogues of the ClO chain compound may too.Before making these new analogues, it may be worth investigating whether the Cli compoundmade in Chapter 4 is phosphorylated. This compound was not soluble enough to be tested as theshorter chain compounds were, but the solubility issues could be circumvented by examining thetechnetium complex instead of the rhenium. From the labelling and stability studies in Chapter4, we know that the Cli ligand is soluble enough to form the99mTCcompound. Radio-HPLCcould be utilized to determine the phosphorylation of the99mTCcomplex directly. This wouldrequire the synthesis of the phosphorylated compounds as standards i.e. chemical synthesis ofCu-Re-phosphate. This assay could be very informative as to the capacity of hexokinase tohandle modified substrates, and has the added advantage of involving the exact species thatwould be the potential imaging agent, the technetium rather than the rhenium complex.It may be possible to alter the longer chain compounds to increase their water solubility. Oneway in which this could be done is by converting the methylene chain to a polyethyleneglycol(PEG) chain. For example, in a paper by Schibli et. al, two long chain glucose compounds thatwere synthesized were found to exhibit an appreciable inhibition of hexokinase (Figure 6.3).The PEGylated compound (Figure 6.3d) had K = 250 jiM and the C8 compound (Figure 6.3c) K= 5800 jiM.3 This is a very significant difference; addition of two oxygen atoms down the alkylchain improved the affinity twenty fold. This would be a very interesting adjustment toReferences begin on Pg 173 169investigate with our glucosamine compounds, as itcould potentially have two benefits, increasedwater solubility and increased affinity for hexokinase(for example, Figure 6.4a).N- Benzoyl glucosamine (Figure 6.3b) is a very highaffinity competitive inhibitor of hexokinasewith a K1 = 22 .tM.’9 By comparison, N-acetyl glucosaminehas K = 98 1iM.’9 The compoundsdiscussed in this work are similar to N-acetyl glucosaminein that they have a carbonyl groupattached to the C-2 nitrogen, with an alkyl carbon inthe alpha position. Although the alkyl“chain” of N-acetyl glucosamine ends after onecarbon, whereas in our compounds it continuesfor quite some time before coming to a very bulky groupat the end, in terms of what interactswith the actual active site, these compounds are verysimilar. The fact that adding an aromaticgroup to the carbon alpha to the carbonyl to giveN-benzoyl glucosamine increases the affinityfor the active site means this could be an avenue worth investigating with ourcompounds (forexample, Figure 6.4c). One must be aware that thesesuggestions are only expected to have aneffect on the compounds’ interaction with hexokinase as a competitiveinhibitor, not as asubstrate. However, the two are not unrelated, and improvingthe way in which something fitsinto the active site can only be advantageous in the stepwise progression towardsdiscovering anew substrate molecule.The other part of the molecule that could be varied is the metal binding portion. Work towardsthis end was shown in Chapter 3, and from the in vitro studies reported therein itwould seemthat variations in the metal binding groups had no effect on the enzymaticrecognition andprocessing of the molecules. For GLUT this is probably true,but for hexokinase the fact thatthese were short chain molecules means they are limited by factors otherthan their bindinggroups (their short chain length). Varying the binding groups may havean affect on hexokinaseaffinity, as well as on the overall solubility and biodistribution of these compounds.One way in which the coordination chemistry of the tricarbonyl core couldbe expanded (thatmay be of use in building libraries of this type) is by the investigation of thiol binding groups.Thiols have been underutilized as ligands in this area of chemistry, especially as they havebeencalculated to be good donors for the tricarbonyl core.20 There have been a few examples ofthiols in this area of the literature,2’but most work with sulfur uses thioethers, which have beenfound not to bind as well as thiols.22’23Work in the Orvig lab has led to successful applicationof the phenolate group as a donor for the tricarbonyl core.24 Although this group binds well, theReferences begin on Pg 173 170synthesis of the99mTccomplex requires use of a base, which may preclude itsapplication incertain situations. A plausible alternative is the use ofa thiophenol group (Figure 6.4b). The pHof the thiol proton is around 7, meaning it willbe deprotonated at physiological pH so thelabelling could likely proceed without the needfor base. Work towards this goal has begun.aH°’OHbH0\°\OH//0‘NRH‘SHNJFigure 6.4 Compounds that could be explored as possible biflinctional ligandsfor the99mTccore, a) a PEG chain analogue, R = any metal binding groups, a range shouldbe investigated b)an analogue incorporating an aromatic ring into the linker c) a thioohenol ligand.The key considerations when designing a binding group for the tricarbonyl coreare affinity forthe metal centre, size of the binding sphere, and overall charge and lipophilicityof the complex.In potential carbohydrate-based imaging agents, a neutral charge is desirable, which means that amonoanionic ligand set is needed. The ligand needs to have good affinity for the metal centre toensure rapid complexation and that the complexes are stable once formed. This isbest achievedby having at least one, and preferably multiple, nitrogen atoms as part of a tridentate bindingsphere.20’25Aromatic nitrogen atoms in particular are favoured, thoughtheir presencenecessitates the presence of hydrophobic rings. Replacing one of these aromatic rings with aprimary amino group would have an affect on the lipophilicity of the compoundas well as thestrength with which the ligand binds the metal ion. Lipophilicity and size are very importantproperties in determining the biodistribution of a compound, and require fine tuning for eachmolecule. These factors are all interrelated and need to be balanced with each other by synthesisand testing of a large range of compounds in a search for observable structure-activityrelationships. The best chelate for each system will have to be investigated on a case by caseReferences begin on Pg 173 171basis, as there is always a balance to bestruck between the relative advantages of small size,strong metal binding, and lipophilicity.The design of a99mTc..carbohydrateconjugate for use in nuclear medicineis a difficult task. Webelieve it is a worthwhile goal becauseof the great benefits it would provide society ifsuccessful.References begin on Pg 173 1726.3 References1. Schugar, H.; Green, D. E.; Bowen, M. L.; Scott,L. E.; Storr, T.; Bobmerle, K.; Thomas,F.; Allen, D. D.; Lockman, P. R.; Merkel, M.; Thompson, K.; Orvig,C., Angew. Chem.mt. Ed. 2007, 46, 1716 - 1718.2. Adam, M. J.; Wilbur, D. S., Chem. Soc. Rev. 2005,34, 153 - 163.3. Seevers, R. H.; Counsell, R. E., Chem. Rev. 1982, 82,575 - 590.4. Levin, V. A., I Med. Chem. 1980, 23, 682-684.5. Schibli, R.; Dumas, C.; Petrig, J.; Spadola, L.; Scapozza, L.; Garcia-Garayoa, E.;Schubiger, P. A., Bioconjugate Chem. 2005, 16, 105 - 112.6. Kishimoto, S.; Kawazoe, Y.; Ikeno, M.; Saitoh, M.; Nakano, Y.; Nishi, Y.; Fukushima,S.; Takeuchi, Y., Cancer Chemoth. Pharm. 2006, 57, 84 - 90.7. McGowan, M. H.; Russell, P.; Carper, D. A.; Lichtstein, D.,I Pharmacol. Exp. Ther.1999, 289, 1559 - 1563.8. Salin, H.; Maitrejean, S.; Mallet, J.; Dumas, S., I Histochem. Cytochem. 2000, 48, 1587 -1592.9. Kalaria, R. N.,Ann. NY. Acad. Sci. 1999, 893, 113- 125.10. Mooradian, A. D.; Chung, H. C.; Shah, G. N., Neurobiol. Aging 1997, 18, 469 - 474.11. Simpson, I. A.; Chundu, K. R.; Davies-Hill, T.; Honer, W.; Davies, P. D., Ann. Neurol.1994,35,546-551.12. Mark, R. J.; Pang, Z.; Geddes, J. W.; Uchida, K.; Mattson, M. P.,1 Neurosci. 1997, 17,1046 - 1054.13. Storr, T.; Merkel, M.; Song-Zhao, G. X.; Scott, L. E.; Green, D. E.; Bowen, M. L.;Thompson, K. H.; Patrick, B. 0.; Schugar, H. J.; Orvig, C., I Am. Chem. Soc. 2007, 129,7453 - 7463.14. Storr, T.; Scott, L. E.; Bowen, M. L.; Green, D. E.; Thompson, K.; Schugar, H.; Orvig,C., Submitted for publication.15. Speizer, L.; Haugland, R.; Kutchai, H., Biochim. Biophys. Acta - Biomembranes 1985,815, 75-84.16. Battaglia, G.; La Russa, M.; Bruno, V.; Arenare, L.; Ippolito, R.; Copani, A.; Bonina, F.;Nicoletti, F., Brain Res. 2000, 860, 149-156.17317. Zhang, M.; Zhang, Z.; Blessington, D.; Li, H.; Busch,T. M.; Madrak, V.; Miles, J.;Chance, B.; Glickson, J. D.; Zheng, G., Bioconjugate Chem. 2003, 14,709 - 714.18. Kahlenbe, A.; Dolansky, D., Can. J Biochem. 1972,50, 638 - 643.19. Bertoni, J. M.; Weintraub, S. T., J Neurochem. 1984, 42, 513 - 518.20. Safi, B.; Mertens, J.; De Proft, F.; Geerlings, P.,J Phys. Chem. A 2006, 110, 9240 -9246.21. Lazarova, N.; Babich, J. W.; Valliant, J. F.; Schaffer,P.; James, S.; Zubieta, J., Inorg.Chem. 2005, 44, 6763 - 6770.22. Gorshkov, N. I.; Lumpov, A. A.; Miroslavov, A. E.; Suglobov, D. N., Radiochemistry2005, 47, 45 - 49.23. Wust, F.; Skaddan, M. B.; Leibnitz, P.; Spies, H.; Katzenellenbogen,J. A.; Johannsen,B., Bioorg. Med. Chem. 1999, 7, 1827 - 1835.24. Lim, N. C.; Ewart, C. B.; Bowen, M. L.; Ferreira,C. L.; Barta, C. A.; Adam, M. J.;Orvig, C., Inorg. Chem. 2008, 47, 1337 - 1345.25. Schibli, R.; La Bella, R.; Alberto, R.; Garcia-Garayoa, E.; Ortner, K.; Abram,U.;Schubiger, P. A., Bioconjugate Chem. 2000, 11, 45 -351.174APPENDIX 1Rat Perfusion Experiments - as discussedin Chapter21,2These experiments took place in the laboratories ofDr. David Allen, Dr. Paul Lockman andFancy Thomas at Texas Tech University Health SciencesCentre, Amarillo, TX. Male Fischer-344 rats (220-330 g; Charles RiverLabs, Kingston, N.Y.) were anesthetized with sodiumpentobarbital (50 mg/kg). A PE-60 catheter filledwith heparinized saline (100 units/mL) wasplaced into the left common carotid arteryafter ligation of the left external carotid, occipital andcommon carotid arteries (common carotid arteryligation is accomplished caudal to the catheterimplantation site). The pterygopalatine artery wasleft open during the experiments. Rat rectaltemperature was monitored and maintained at 37C by a heating pad connected to a feedbackdevice (YSI Indicating Controller, Yellow Springs, Ohio). Thecatheter to the left commoncarotid artery was then connected to a syringe containinga physiologic perfusion fluid with 1.0jiCi/mL3H-choline, and 1251-pyrA or 1251-pyrB as described above.Perfusion fluid was filtered,warmed to 37 °C and gassed with 95 % air and5 % CO2. The perfusion fluid was infused intothe left carotid artery with an infusion pump for 60 s at 10 mL/min (HarvardApparatus, SouthNatick, MA). This perfusion rate maintained a carotid arterypressure of 120 mm Hg. Rats werethen decapitated and cerebral samples obtained; the brains were removedfrom the skull, and theperfused cerebral hemisphere dissected on ice after removalof the arachnoid membrane andmeningeal vessels. Brain regions were placed in scintillation vialsand weighed. In addition, a50 tL aliquot of the perfusion fluid was transferred to a scintillation vialand weighed. The brainand perfusion fluid samples were digested overnight at 50 °C in 1 mL of 1M piperidine.FisherChemical scintillation cocktail (10 mL) (Beckman, Fullerton, CA) wasthen added to each vialand the tracer contents assessed by dual-label liquid scintillation counting. Duallabeledscintillation counting of brain and perfusate samples was accomplished withcorrection forquench, background and efficiency. Brain vascular space was estimated fromseparateexperiments and used to correct for radiotracer remaining within brain capillaries;apparentpermeability surface-area coefficients (PA; BBB “permeability”)were determined using theCrone-Renkin equation (Equation A. 1):PA = -F in (1-K1/F) Equation A.1175PA = permeability in mL/sec/g; F = flow rate throughall capillaries of material being tested,normalised to radiotracer concentration in plasma; K1= rate constant for the influx from blood-to-brain.3References1. Smith,Q.R.; Allen, D. D., Meth. Mol. Med 2003, 89, 209 - 218.2. Schugar, H.; Green, D. E.; Bowen, M. L.; Scott, L. E.; Storr, T.; Bolimerle, K.; Thomas,F.; Allen, D. D.; Lockman, P.R.; Merkel, M.; Thompson, K.; Orvig, C.,Angew. Chem.mt. Ed. 2007, 46, 1716- 1718.3. Pardridge, W. M., Introduction to the Blood-Brain Barrier. Cambridge University Press:Cambridge, 1998; Pg 130.176

Cite

Citation Scheme:

    

Usage Statistics

Country Views Downloads
United States 43 1
Norway 4 0
India 4 1
China 4 6
Canada 2 0
Vietnam 1 0
Republic of Korea 1 0
Mexico 1 0
United Kingdom 1 0
Pakistan 1 0
City Views Downloads
Washington 30 0
Unknown 7 4
Bangalore 4 1
Los Angeles 4 0
Ashburn 3 0
New York 3 1
Guangzhou 2 0
Beijing 2 0
Mountain View 1 0
Calgary 1 0
Hanoi 1 0
Kansas City 1 0
Montreal 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}
Download Stats

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0061548/manifest

Comment

Related Items