Acid Comparison Essay

2008

A PROTEIN-APPENDED ROTAXANE
BY

HAYTHIM HASSANEIN
SUPERVISOR PALL THORDARSON

2

Contents

Certificate of Orignality .............................................................................................................. i Abstract ...................................................................................................................................... ii Acknowledgements ................................................................................................................... iii Table of Contents ....................................................................................................................... 2 1. Introduction ............................................................................................................................
5 2. Natural Protein-Appended Rotaxanes.................................................................................... 6 3. Bioconjugate Chemistry and Rotaxanes ................................................................................ 8 4. Earliest Rotaxane Synthesis ................................................................................................... 9 5. Rotaxanes & Nanodevices ................................................................................................... 10 5.1 Molecular Motors................................................................................................... 10 5.2 Molecular Shuttles ................................................................................................. 12 5.3 Molecular Muscles ................................................................................................. 14 5.4 Molecular Electronics ............................................................................................ 16 6. Approach and Synthesis ....................................................................................................... 17 6.1 Pre-assembly in Non-Aqueous Solvents ............................................................... 17 6.2 Axial Component Motifs ...................................................................................... 17 6.3 Macrocycle Component Motifs ............................................................................ 18 6.4 Stopper Component Motifs ................................................................................... 20 6.5 Approaches ........................................................................................................... 20 6.6 Characterisaiton .................................................................................................... 26 7. Applications ......................................................................................................................... 27 7.1 Drug Delivery ........................................................................................................ 27 7.2 Biosensor Applications .......................................................................................... 28 8. Experimental ........................................................................................................................ 30 8.1 Azo-based Reactions ........................................................................................ 30 8.1.1 8.1.2 Synthesis of 4,4-Azodiphenol (1)........................................................ 30 Synthesis of 4-Bromobutoxy Nitrobenzene (2) .................................. 31

3

8.1.3 8.1.4

Synthesis of trans-Bis-(E)-((Bromobutoxy)phenyl)diazene) (3) ........ 31 Synthesis of trans-4,4'-Bis((6-hydroxyhexoxy)azobenzene) (4) ......... 32

8.1.5 Attempted Synthesis of trans-4,4‟-Bis((6maleimidohexoxy)azobenzne) (5) ................................................................... 32 8.2 Viologen-Based Reactions ............................................................................... 33 8.2.1 Synthesis of 4,4‟-Bipyridinium-N,N-di(propylammonia) hexaflurophosphate (6) .................................................................................... 33 8.2.2 33 Synthesis of 4,4‟-Bipyridinium-N,N-di-(hydroxypropyl) dibromide (7)

8.2.3 Attempted synthesis of 4,4‟-Bipyridinium-N,N-di(maleimidopropyl) hexaflurophosphate (8) .................................................................................... 34 8.2.4 Attempted synthesis of 4,4‟-Bipyridinium-N,N-Bis(3-(6-maleimido hexanamido)propyl hexaflurophosphate) (9) ................................................... 34 8.2.5 Attempted synthesis of 4,4‟-Bipyridinium-N,N-di(propylazide dibromide) (10) ................................................................................................ 35 8.3 Polyethylene glycol (PEG)-Based Reactions................................................... 35 8.3.1 Synthesis of 4,7,10,trioxa-1, 13-bismaleimidotridecane (maleimide PEG) (11) ......................................................................................................... 35 8.4 Azo Pseudorotaxane Complexes...................................................................... 36 8.4.1 Complex A - β Cyclodextrin and trans-4,4'-bis(6hydroxyhexoxy)azobenzene pseudorotaxane .................................................. 36 8.4.2 Complex B - α-Cyclodextrin and trans-4,4'-bis(6hydroxyhexoxy)azobenzene pseudorotaxane .................................................. 37 8.4.3 Complex C - Cucurbit[7]uril and trans-4,4'-Bis(6hydroxyhexoxy)azobenzene pseudorotaxane .................................................. 37 8.5 Viologen Pseudorotaxane Complexes ............................................................. 38 8.5.1 Complex D - Cucurbit[7]uril and 4,4-Bipyridinium-N,N-di(propylammonia hexaflurophosphate complex............................................... 38 8.5.2 Complex E - Cucurbit[7]uril and 4,4‟-Bipyridinium-N,N-di(hydroxypropyl) dibromide complex .............................................................. 38 8.6 Polyethylzene glycol (PEG) Pseudorotaxane Formations ............................... 39 8.6.1 Complex F - α-Cyclodextrin and of 4,7,10-trioxa-1,13tridecanediamine .............................................................................................. 39 8.6.2 Complex G - β-Cyclodextrin and of 4,7,10-trioxa-1,13tridecanediamine .............................................................................................. 40 8.7 Maleimide-capped Rotaxane Formations ........................................................ 40

4

8.7.1 Formation H - α-Cyclodextrin and 4,7,10,trioxa-1, 13bismaleimidotridecane ..................................................................................... 40 8.7.2 Formation I - β-Cyclodextrin and 4,7,10,trioxa-1, 13bismaleimidotridecane complex ...................................................................... 41 8.7.2 Formation J - β-Cyclodextrin and 4,7,10,trioxa-1, 13bismaleimidotridecane complex ...................................................................... 41 8.8 Bioconjugate Rotaxane Formation .................................................................. 42 8.8.1 Cytochrome c biorotaxane with α & β-cyclodextrin Maleimide-PEG 42 8.8.2 Bovine Serum Albumin biorotaxane with α & β-cyclodextrin Maleimide-PEG ............................................................................................... 44 9. Results and Discussion ........................................................................................................ 46 9.1 Synthesis .......................................................................................................... 46 9.1.1 Azo-based Synthesis .............................................................................. 46 9.1.2 9.1.3 9.2 Viologen-based Synthesis ................................................................... 48 PEG-based Synthesis........................................................................... 50

Pseudorotaxane Complexes ............................................................................ 52 9.2.1 Synthesis of Azo Pseudorotaxanes ........................................................ 52 9.2.2 9.2.3 Synthesis of Viologen Pseudorotaxanes ............................................. 56 Synthesis of PEG Pseudorotaxanes ..................................................... 59

9.3

Rotaxane Bioconjugation ................................................................................. 60 9.3.1 9.3.2 Gel Electrophoresis ............................................................................. 60 MALDI Mass Spectrometry ................................................................ 63

Conclusion ............................................................................................................................... 65 References ................................................................................................................................ 66

5

1. INTRODUCTION
Bionanotechnology is an emerging field of science and technology that aims to control and assemble „Molecular Machines‟4. One type of molecular machine is the Rotaxane which comprises of macrocycle „ring‟ structure threaded onto an „axle‟ component then capped with suitable „stoppers‟; a molecular dumbbell-shaped structure5. An important feature of this system is its switching properties that give rise from the non covalent attachment of the „ring‟ allowing it to slide back and forth on the „axle‟ (reference). Formation of these structures could be categorised by three methods (Figure 1.1): 1. Capping – a Pseudorotaxane consisting of the axle and macrocycle is pre-synthesised stoppers added through chemical reaction 2. Clipping – a formation of a half completed macrocycle is added to an axle-stopper unit and then completed through a chemical reaction 3. Slippage – involves macrocycle component that can just thread into the completed axle-stopper unit through the addition of heat One of the most challenging prospects for many scientists participating in Rotaxane synthesis is the ability to create a water-soluble Rotaxane6. A proposal to this dilemma is to incorporate proteins/enzymes at the ends of these stoppers or even on the ring component. Past research in bioconjugate chemistry7 and click chemistry8 has been conducted with polypeptides axles and porphyrin “rings” to form porphyrin-rotaxanes 9. A similar structure will be conducted for synthesis these protein/enzyme bio-Rotaxanes. The rotaxanes based nanodevices have been developed in recent years, which includes functions as „molecular motors‟10, „molecular shuttles‟11, „molecular muscles‟12 and „molecular electronics‟13.

Figure 1.1a) Structure of a rotaxane b) Formations of rotaxanes

6

2. NATURAL PROTEIN-APPENDED ROTAXANES
Embedded in the inner membrane is the enzyme ATP synthase, which is responsible for most ATP hydrolysis in the mitochondria. The structure of ATP synthase, in particular the rotating rod inside a static wheel, mimics the proposed artificial models of a protein-appended rotaxanes2,14-19. This biological “power plant” consists of a Fo domain (composed of 12 identical aspartates protein chains) fixed within the membrane and the adjoining F1 domain (reminiscent of a macrocycle), which extends above the membrane3,15. The F1 contains three α and β subunits where the site of ATP synthesis occurs. These surround a central “axle" structure containing γ, δ and ε subunits. Within its structure, the ATP synthase represents a prime example of a natural rotaxane formed by the F1-axle-F0 assembly.

Intermembrane space Inner Membrane

Figure 2.1 - Embedded within the inner membrane of mitochondria1 are the nano “power plants’ ATP sysnthase3

γ, δ, ε Matrix

α

β

β

The Stator composed of b and d subunits extends from the membrane to the top of the F1 domain effectively anchoring the α and β subunits so they do not rotate with the axle. Interactions with one rotation of axle induce three conformational changes within the α and β subunits which effectively provide the free energy for the ATP synthesis or ATP decomposition. The mechanism of the ATP synthesis involves association of ADP to the binding site of the α- and β-subunit interface in a favoured orientation and stabilised with hydrogen bonding with surrounding amino acids and water molecules and salt bridges with a magnesium divalent cations. Once ADP and inorganic phosphate Pi are positioned within the

7

subunits, a planar transition state of inorganic phosphate (which is initially tetrahedral) is formed resulting in a covalent attachment to the ADP molecule, yielding an ATP molecule. The energy provided for F1 domain axle rotation responsible for ATP synthesis, originates from embedded Fo domain of ATP synthase. An analogy of Fo domain is with a hydroelectric dam, where a height gradient in water (potential energy) provides the necessary kinetic energy to drive a turbine in the power station. Similarly, the regulated flow of protons through an ATP synthase channel „a‟ subunit rotates an array of identical „c‟ subunits just like a turbine in a hydroelectric dam. In order to hydrolyse ATP, ATP synthase must release energy from the flux of the protons through an ATP synthase channel. A high concentration of H+ ions on one side of the membrane causes the H+ ions to travel through the ATP synthase channel to the other side of the membrane where the H+ ion concentration is lower. Energy for ATP synthesis only occurs in presence of this electrochemical proton concentration gradient. In absence of this
Figure 2.3 – ATP synthase can be seen in analogy with a hydroelectric dam, both requiring a gradient to produce energy2

gradient, the mitochondria system will be in equilibrium and ATP synthesis will cease, resulting in the death of the cell. Active electron transport systems drive proton pumps and, hence the translocation of protons within the mitochondria. A number of different coenzymes are evolved in the electron transport systems electrons, oxidised from of these coenzymes, are donated and propagate through the inner membrane of the mitochondria activating these proton pumps and translocating protons thus maintaining the proton gradient. ATP synthase is an electrochemically-driven protein-appended rotaxane, which makes it a very complex system. Light-driven protein-appended rotaxane would be desired less complex proposed system involves and hence the motivation of this project20.

8

3. BIOCONJUGATE CHEMISTRY AND ROTAXANES
One emerging field in Bionanotechnology is the synthesis of novel water-soluble bioconjugate rotaxanes. Herein, bioconjugation refers to the coupling of small organic compounds to larger biomolecules such as proteins & enzymes. Bioconjugation of such systems results in evolved architectures and properties and enable them to be efficiently synthesised and manipulated 6. The enzyme Cytochrome c is one such enzyme which has been demonstrated to exhibit this type of behaviour 21.

Investigations toward Cytochrome c bioconjugate reactions include Lys, Cys and Tyrmodifying reactions. The Lys-modifying reactions involve coupling a carboxylic acid or an aldehyde with the lysine (Lys) residue of the Cytochrome c, resulting in either amide or secondary amine substitution respectively. These Cys-modifying reactions involve coupling either a malemide, -halocarbonyl or orthopyrifyl disulfide with the residue cysteine (Cys) residue of the Cytochrome c resulting in either thioether or disulfide formations. Finally, Tyrmodifying reactions mainly involves coupling such compounds as diazonoium salts and a number alkylation & acylation reactions with the tyrosine (Tyr) residue of the Cytochrome c which could result in substitutions. diazo-compound and ether

Click Chemistry provides an interesting alternative for the synthesis of these water-soluble bioconjugate rotaxanes. Recent heterofunctional spacers that bear functionalities, such as azides and alkynes, have been used to couple a secondary target with a high degree of
Figure 3.1– Cytochrome

selectivity22.

Previous studies have shown that the formation of water-soluble in peptide-based rotaxane systems is quite feasible7. In this system glycine residues are coupled onto the axle component of the rotaxane system. The macrocycle of this system was free to interact with the glyecine units, when environmental conditions were changed. The combination of these results and the power of bioconjugate chemistry indicates that the formation of bioconjugate rotaxanes is a feasible target.

9

4. EARLIEST ROTAXANE SYNTHESIS
One of the earliest reported literatures, depicting rotaxane synthesis and analysis, was conducted in 196723. This synthesis relied on the probability that a basic Rotaxane structure could form through the aid of a resin adduct attached to the macrocycle „ring‟ component (2hydroxycyclotriacontanone), which would later be threaded through an ether chain and hydrolysed to yield the threaded complex seen in Figure 4.1. Synthesis of the ether chain involved a reaction of decane-1,10-diol and triphenylmethyl chloride in a mixture of pyridine, toluene and dimethylformamide. However, to obtain a substantial yield of 6% of the Rotaxane complex, the authors required to undertake 70 treatments of the reaction in a column with the resin bound macrocycle and the removal of the excess reagents through reflux with sodium bicarbonate. The final Rotaxane complex was characterised as oil, supported with infrared spectra and thin film chromatography (TLC). The macrocycle interaction in the Rotaxane only involves weak van der Waals forces, which accounts for the low yield obtained in the experiment. Furthermore, presence of the large hydrocarbon chain and trityl phenyl groups in the ether „axle‟ accounts for the Rotaxane‟s insolubility. The literature demonstrates difficult and inefficient Rotaxane synthesis that results in low yield and insoluble product, however, efficient modern Rotaxane synthesis utilise a number of different interactions to template their synthesis including hydrogen bonding, metal coordination, hydrophobic forces, and columbic interactions that overcome these predicaments. With introduction of these interactions, modern Rotaxanes now have various functions and variables that allow them to be used in many different applications.

Figure 4.1- Earliest Rotaxane synthesis

10

5. ROTAXANES & NANODEVICES
5.1 MOLECULAR MOTORS
One literature has examined pseudorotaxanes (Rotaxanes without „Stopper‟ components) as possible “piston-cylinder” molecular motors which undergo assembly/disassembly processes through the use of photochemical catalysts10. Three main components were considered in the design pseudorotaxane photochemical dethreading system (Figure 5.2a): A pseudorotaxane composed of the macrocycle „ring‟ and cylindrical „axle‟ units. An external electron-transfer photosensitiser A sacrificial electron-donor agent (reductant) The mechanism involved in the pseudorotaxane disassembly involves the excitation of the metal cation in the external photosensitiser, through oxidation by an ultraviolet radiation source, which leads to an electron-transfer with the macrocycle component. This results in the decrease of the non-covalent-bonding interactions responsible for the pseudorotaxane assembly. However, the system can only become efficient with the presence of a sacrificial electron-donor (reducing agent) that reduces the external photosensitiser back to its original state so it becomes readily available to oxidise again and, hence, prevents any possibility that a back electron-transfer reaction (between the macrocycle & external photosensitiser) to occur. Exposing the irradiated solution to O2 results in the reoxidiation of the photosensitiser and the reformation of the pseudorotaxane (Figure 5.2b). Design enhancement is accomplished by incorporating the external photosensitiser on the actual macrocycle component, avoiding effects that any impurities may have on the system (Figure 7(b)). The authors used cyclophane L14+ as the macrocycle component in the pseudorotaxane system which contains a 2,2‟-bipyridine coordinating ligand with two 4,4‟bipryidinium electron-acceptors (Figure 8). The external photosensitisers used in the literature are [Re(CO)3Cl] and [Ru(bpy)2]2+ and they were incorporated with coordination upon the cyclophane L14+ resulting in the macrocycle- photosensitiser complex [Re(CO)3L1Cl]4+ and [Ru(bpy)2L1]6+ (Figure 5.3).

11

The authors used 1,5-bis(2-(2-hydroxyethoxy)-ethoxy)naphthalene (1/5BHEEN) as the cylindrical „axle‟ component of the pseudorotaxane system. 365 nm light was used as the photocatalyst in the experimental procedures. Characterisation of the disassambely process transpired through Fluorescence absorption spectra. Upon assembly of the pseudorotaxane, no peaks were visible between 500700 nm. However, of upon

disassembly

pseudorotaxane,

large peaks become visible in this range (Figure 5.1).
Figure 5.1 - Absorption spectra of the photosensitiser Re(CO3)L1Cl a -> Solution before irradiation b -> 7min irradiated c -> 15min irradiated d -> Oxidation with air

Figure 5.2 Molecular Motor Model (a) Model representing the basic systematic components required for dethreading process to occur. Red= Sacrificial reductant P = photosensitiser (b) Incorporation of the photosensitisers creates a more efficient system

12

Figure 5.3 - Photosensitisers Photosensitisers used in the molecular motor system. [Re(CO3)L1Cl) compound was chosen over the others since in produced a highly resolved Charge-Transfer absorption band

5.2 MOLECULAR SHUTTLES
The term „molecular shuttle‟ refers to a molecular machine that is capable of transporting molecules or ions from one location to another. Extensive studies on Rotaxanes as molecular shuttles have been conducted and reported in many literatures. One model, similar to the Rotaxane structure used in studies of molecular motors24 confines macrocycle shuttling between two stations11. Benzidine and biphenol substituents were used as the units for the two stations in the axial component of the Rotaxane. Cyclophane, comprising of two bipyridinium units bridged by two p-xylyl spacers, was chosen as the macrocycle for the system. The stopper components of the Rotaxane, tri-isopropylsilyl group, prevent

dissociation of the complex. The Rotaxanes were presynthesised in a previous experiment and made in a solution of CD3CN. TwoDimensional
Figure 5.4 – Molecular Shuttling Model
1

H NMR (NOESY), ultraviolet/visible

spectroscopy and cyclic voltammetry were used to characterise the dynamics of the macrocycle shuttling

along the axial component (Figure 5.4).

13

Integration of the NOESY spectrum depicted that the Rotaxane was kinetically favourable with the macrocycle bound to the benzidine station (84% of the Rotaxanes in solution were in this state), rather than the macrocycle bound to the biphenol station (16% of the Rotaxanes in solution were in this state) (Figure 5.4).
Macrocycle shuttling process occurs either through electrochemical oxidation of the benzidine residue or protonation of the basic nitrogens on this station. Both of these methods rely on the repulsive electrostatic interactions between the positively charged axial component and the post-switching cationic nature of the benzidine station (Figure 5.4). Deuterated trifluoroacetic acid (d-TFA) was added in the solution which resulted in the protonation of the benzidine residue and shuttling of macrocycle in to the biphenol station of the Rotaxane. This was supported by NOESY 1H NMR that showed a shift in the peaks in retrospect to Rotaxane solution containing CD3CN (Figure 5.5). Addition of deuterated pyridine (d5-pyridine) neutralised the d-TFA and hence oxidiation of the benzidine residue occurred and the macrocycle had shuttled back to bind to the residue in its favourable conclusions obtained from the 1H NMR spectroscopic studies are supported by room-temperature ultraviolet/visible spectroscopy state (Figure
5.5).

Figure 5.5- Molecular Shuttle NMR data

14

5.3 MOLECULAR MUSCLES
A1 A
A A

2

2 3
A A

3 4
A A

4

A A

Synthetic Molecular Muscles are one the most interesting applications involved in Rotaxane analysis and models have been designed that duplicate real muscular systems, such as the sacromere in the human muscle (containing myosin and actin filaments)12. They are defined as molecular motors which have the ability to expand and contract. In this model, the principle of the Rotaxane

B A

contraction/expansion process involves motion of two threaded identical filament conjugates along one another without disassociation. The filament constitutes four main components (Figure 5.6):

Figure 5.6- Molecular Muscle Model

1. Bidentate embedded macrocycle that is covalently attached to the filament. 2. Bidentate chelate 3. Tridentate coordinating fragment 4. Bulky stoppers whose function is to prevent dethreading of the filaments.

The mechanism for filament motion is triggered by electrochemical reaction. Copper complexes have been previously shown to efficiently trigger molecular motion through CuI/CuII electrochemical reactions. In the rotaxane‟s contracted state, the system contains copper(I) as the assembling and templating metal. However, for Rotaxane expansion to occur, the solution must be subjected to a metal-exchange process induced through a chemical reaction. Synthesis of the rotaxane molecular muscle (32+) structure involved the reaction of a copper(II) precursor containing phenolic functions with the equivalent bromide stopperbearing disymmetrical 2,2‟,6‟,2‟‟-terpyridine (Figure 5.7). The copper complex can only form using the bidentate chelates in the ring and filament, hence, the molecular muscle is in its contracted form. Treatment of the rotaxane molecular muscle in excess potassium cyanide (KCN) removed the copper ions and hence creating a free ligand (4), (Figure 5.7). Addition of the zinc nitrate induced a metal-exhange process (CuI/ZnII), resulting in a ZnII transfer into the

15

free rotaxane ligand and forming a complex with the only bidentate chelate on the ring component and the tridentate coordinating fragment on the filament. This new formation induces expansion of the molecular muscle. Two-dimensional NMR (ROESY) as well as high-resolution mass spectrometery (FAB) were used in the characterisation of the molecular muscle. This included identification that the expansion had occurred by characterising peak shifting as the new metal complexes are formed.
Figure 5.7 - Scheme of the expansion/contraction for molecular muscles

B

16

5.4 MOLECULAR ELECTRONICS
Chemically assembled electronic nanocomputers

(CAENs) could be potential alternatives to the maturing complementary metal-oxide semiconductor (CMOS) based integrated circuits13.
The focus of these chemically synthesised systems centres the selfFigure 5.8 – Structure of the CAEN Teramc system with the rotaxane monolayer embedded between two electrodes

assembly of nano-scale components. Literature has been reported on the use of rotaxanes in these CAEN

models called the Teramac25. The Teramac system consists of a rotaxane monolayer embedded between two perpendicularly oriented electrodes. Synthesis of this system was via step-wise deposition on a silicon substrate. The rotaxane monolayer was deposited through the Langmuir-Blodgett apparatus. The rotaxane structure is composed of two bipyridinium units with bis-paraphenylene-34-crown-10 macrocycle rings. Reduction and oxidation processes controlled the switching properties of the Termac system and characterisation of any signal produced by the rotaxane monolayer was via normalised density of state plots (NDOS). Tunnelling and biasing was observed in the rotaxane junction after the system was exposed to oxidation processes. AND and OR gate simulations were commenced on these rotaxane monolayer systems which demonstrate application in futuristic computational technologies. One major disadvantage of this system was whether the rotaxane monolayer could survive the deposition of the top electrode.
The reason for this was due to the highly hydrophobic heads pointing away from the surface which created a buffer layer between the electrodes and the centrally located crown ether rings bound to the bipyridinium donor site. However, literature reported acceptable
Figure 5.9 – Rotaxane Structure used in the Termac system

translation of redox signals for sold-state device properties.

17

6. APPROACH AND SYNTHESIS
One of the greatest obstacles in Rotaxane development includes synthesis of novel water soluble rotaxane species. Water insolubility is usually due to the rotaxane‟s long hydrocarbon axial components and its bulky stoppers groups. The benefit of high yield water soluble rotaxanes synthesis will result in a new line of research in bionanotechnology. Many rotaxane models and designs have been devised to counteract insolubility issues within rotaxane structures, which include using protecting groups and ion selectivity26

6.1 PRE-ASSEMBLY IN NON-AQUEOUS SOLVENTS
One method in water-soluble rotaxane synthesis is to use protecting groups that reduce reactivity of a certain functional hydrophilic structures and can be later removed after rotaxane finalisation27. One system uses benzoyl protecting groups (2-aminoethyl 2,3,4,6tertra-O-benzoyl-β-D glucopyranoside) during rotaxane synthesis, then removed using NaOMe/MeOH and replaced with hydroxyl groups forming a carbohydrate stopper. This allows hydrogen bonding interactions to occur between water molecules and these carbohydrate stoppers resulting in a water soluble rotaxane27.

Figure 15 – Slippage formation of bipyridinium rotaxane and removal of the protective groups

6.2 AXIAL COMPONENT MOTIFS
Three axial component motifs have been desired for synthesising water-soluble rotaxanes. The first class of axial motif is the azo functional group. Azo compounds are a major commercial synthetic colorant and have been desired in water-soluble rotaxanes20,28,29. These azo compounds can be produced efficiently with high yields and are also easier to characterise using nuclear magnetic resonance spectroscopy30-34. Conformational changes can be induced through photoexictation between the cis (Z) and trans (E) isomers35, hence, a desired motif for protein-appended rotaxanes (Figure 6.2).
Figure 6.2 trans-cis Azo isomerisation

18

The second class of axial motifs is the viologen functioncal group. There have been reported syntheses of viologen based rotaxanes motifs for axial components6,7,9,36-38. Viologen charged properties posses desired applications for protein-appended rotaxane. Redox processes can induce transformations of the viologen to its radical or neutral forms (Figure 6.3). Furthermore, synthesis of the compounds is more efficient then its counterparts, due to the tendency of the compound to form ionic bonds and precipitate as salts.

Figure 6.3 - Viologen transformations

The third class of axial motif is the polyethylene glycol (PEG) functional group. Reported bioconjugate protein-appended rotaxanes have been reported with this class of functional groups39,40. Their polar structures are favoured for macrocycle threading and watersoluble rotaxane formation. Furthermore, their polymer properties and high stability allow for a range of different synthesis approaches.

Figure 6.4 – PEG motif

6.3 MACROCYCLE COMPONENT MOTIFS
Two macrocycle components motifs have been desired for synthesising water-soluble rotaxanes. Rotaxane formation has been investigated and synthesised with cyclodextrins28,41. Cyclodextrins are composed of cyclic oligosaccharides and are considerably water soluble. They have found a wide range of applications in food, pharmaceutical and chemical industries as well as agriculture and environmental engineering. The three main types of cyclodextrin components are α- (6 oligosaccharides), β- (7 oligosaccharides) and γcyclodextrins (8 oligosaccharides) (Figure 6.5). .

19

Figure 6.5 - Structures of cyclodextrins (CD)

Recent studies also reveal a new multimeric macrocycle compound known as Curcbituril, which has a more rigid structure than cyclodextrin and is capbable of forming stable complexes with molecules and ions, thus attractive as a building block for supramolecular architectures (Figure 6.6)42,43.

Figure 6.6 – Structures of cucurbitrul

20

6.4 STOPPER COMPONENT MOTIFS
The stopper motif will be focused on the functional group maleimide. Maleimide-capped rotaxane will be used be form a bioconjugate with the target protein7-9,22. The maleimide double bond reacts favourably in a Michael reaction with the thiol group found on a cysteine amino acid to form a stable carbon-sulfur thioether. Maleimide and its derivatives are prepared from maleic anhydride by treatment with amines followed by dehydration44-46. Recent studies also suggest Mitsunobu methods for maleimide substitution with alcohols studies47. Bismaleimides are a class of compounds with two maleimide groups connected through a molecular unit (Figure 6.7).

Figure 6.7 - Maleimide and bioconjugate methods

6.5 APPROACHES
Addition of the macrocycle will be either through capping or slippage methods depending on which method is the most efficient for this proposed system. Two proposed approaches have been devised and investigated for maleimide substitution within the azo-rotaxane system. The first approach (Figure 6.8, Approach 1) involves click chemistry of alkyne substituted maleimide with an azide substituted azo compound. The second approach (Figure 6.9, Approach 2) involves direct substitution of the maleimide on the azo compound through a mitsunobu reaction. A common reaction in azo formations is the reduction of the p-nitrophenol precursors20. A bimolecular nucleophilic SN2 substitution reaction with 1,4-bromobutane and the resulting azo compound (yellow) will be

21

conducted to extend the axle component in preparation of for malamide and macrocycle additions. A 4,4‟-bipyridine derivative will act as the precursor for the viologen-rotaxane approach (Figure 6.10, Approach 3). The syntheses of this precursor will involve a SN2 reaction with a halogenated linker. Direct maleimide substitution through a reaction with maleic anyhdride and the dehydration of the maleinlic acid intermediate will be the motivation for this approach. The polyethylene glycol approach (Figure 6.11, Approach 4) will involve similar conditions as the viologen-rotaxane system.

22

Figure 6.8 - Approach 1 (Azo-rotaxane system) – Maleimide substitution through click chemistry

23

Figure 6.9 - Approach 2 (Azo-rotaxane system) – Direct maleimide substitution through mitsunobu reaction

24

Figure 6.10 - Approach 3 (Viologen-rotaxane system) – Maleimide substitution through dehydration

25

Figure 6.11 - Approach 4 (PEG-rotaxane system) – Maleimide substitution through dehydration

26

6.6 CHARACTERISAITON
Evidence of pseudorotaxane and rotaxane formation is usually obtained by 1H NMR. Recent evidence suggest that complexation of the axial-macrocycle components result in a broadening of the proton peaks with the 1H NMR spectrum37. This may be due to fast protonproton exchanges between these components48. Recent literature also suggests twodimensional
1

H NMR, especially ROESY

(rotational nuclear
49

Overhauser effect

spectroscopy), as a possible characterisation method . In this method, two perpendicular magnetic pulses (F1, F2) are used to correlate proton-proton coupling which arise from macrocycle-axial component interactions. The rotational Overhauser effect (rOe) involves the transfer of spin polarization from one spin population to another via cross-relaxation and results are negative cross peaks. Reported literature depicts ROESY cross peaks between the aromatic protons of a synthesised azo compound when complexed with α & β cyclodextrins. Bioconjugate rotaxanes formation will be analysed by SDS gel electrophoresis. Evidence of bioconjugate rotaxanes will be supported and characrterised with MS MALDI (matrix-assisted laser desorption/ionization).

α-cyclodextrin

β-cyclodextrin

Figure 6.12 – ROESY of the reported azo-compound which depict macrocycle-azo interactions49

27

7. APPLICATIONS
7.1 DRUG DELIVERY
Rotaxanes have attracted a large attention in drug delivery applications due to their intriguing topologies39. Rotaxane systems have been developed that improve therapeutic efficiency of a range of different proteins50. This includes favourable characteristics such as increased circulating half-life, enhanced proteolytic resistance, reduced antigenicity and immunogenicity, reduced aggregation and improved bioadaptability and bioavailability51. One system, as mentioned on page 18, involves polyethylene glycol (PEG) and the macrocycle cyclodextrin (CyDs) compounds. The bioconjugation product between PEG and proteins are known as pegylated proteins and many proteins that have been pegylated include insulin, α-chymotrypsin, β-lactoglobulin and lipase. The Rotaxane formation of the CyDs and PEG involve multiple cyclic molecules spontaneously threading onto the polymer chain. This type of rotaxane complexation is known as polypseudorotaxane. The three types of CyDs that have been investigated for the PEG interactions are α-CyD (6 glucopyranose units), β-CyD (7 glucopyranose units) and γCyD (8 glucopyranose units). Pegylated Insulin polypseudorotaxane with these CyDs has been reported to engage in sustained release systems for drug delivery through dethreading and threading simulations. Synthesis of the pegylated Insulin polypseudorotaxane first involved formation of the pegylated Insulin by incubating insulin with α-succinimidyl-oxysuccinyl-ω-methoxypoloxyethylene in 3:2 DMF/water solution at room temperature. The pH was then adjusted to 2 by hydrochloric acid and the resulting solution was dialyzed using a membrane filter. Purification was performed with high performance liquid chromatography (HPLC) and characterised using mass spectrometry. The pegylated Insulin was then added to solutions of the CyDs which resulted in the precipitates of the polypseudorotaxane. Precipitates were only observed for α-CyD and γ-CyD; β-CyD polypseudorotaxane did not precipitate. The threading and dethreading of the polypseudorotaxane account for the release mechanisms for the pegylated Insulin. Analysis of the pegylated insulin release rate in a polypseudorotaxane and phosphate buffer was measured in intervals over a period of 12 hours. At its appropriate intervals, the samples were treated and characterised with HPLC. Release rates of the pegylated drug can be controlled by regulating the threading and dethreading rates by adjustment of the administration conditions such as injection medium

28

and concentrations of CyDs in the medium. Figure 7.2 depicts the release rate of the pegylated Insulin with α-CyD and γ-CyD.

Figure 7.1 - Pegylated insulin Puesdorotaxane

Figure 7.2 – Release rate of insulin

7.2 BIOSENSOR APPLICATIONS
Potential biosensor applications for rotaxane structure have been depicted and reported in many literatures52. Rotaxane structures immobilised on surface electrodes can be electrochemically induce mechanical force shuttling of the macrocycle. With the additional recognition protein as the stopper unit, surface-confined rotaxane structures undergo
Figure 7.3 - A rotaxane-based glucose oxidase biosensor

reduction

and

oxidation generate

processes amplify

that

can

alternatively

and

signal.

Furthermore, the macrocycle component can act as a mediator compounds which possess properties for biosensor applications. Analysis has been reported on surface-confined rotaxane structures which consist of threaded sites that are linked to a gold surface at one end and an adamantate stopper complexed with cyclophane and a -donor diiminobenzene unit imbedded in the axial component of the rotaxane53. Redox cyclophane shuttling was demonstrated under two different states (Figure 7.3): 1. Cyclophane acceptor units bind with the diiminobenzene donor unit and become stabilised through - interactions.

29

2. Cyclophane becomes reduced which results in the reduction of the - interactions with the diiminobenzene donor group and electrostatic attraction of the reduced cyclophane to the electrode surface. The electrode surface properties differed in the two states and included hydrophilic and hydrophobic properties, temperature-dependent and temperature-independent

interactions. Control of the kinetics of cyclophane shuttling was dependent on the viscosity (composition) of the medium. Characterisation of the redox potential was derived from cyclic voltammetry.

Figure 7.4 - Redox properties of rotaxane-based biosensor

30

8. EXPERIMENTAL
1

H NMR and

13

C NMR were conducted on the Bruker DPX 300 NMR spectrometer with

Sample Changer at the University of New South Wales Analytical Centre NMR facility (residual peaks: CDCl3= 7.26 ppm, D2O = 4.79 ppm, DMSO = 2.50 ppm, d-acetonitirile = 1.94 ppm, d-methanol = 3.31 ppm, d-Acetone = 2.05 ppm). Two dimensional COSY, ROESY, HQSC NMR spectra were conducted on the Bruker Avance III 400 NMR spectrometer at the University of New South Wales Analytical Center NMR facility. Gel electrophoresis was performed using Invitrogen Novex® NuPage® 12% Bis-Tris 1 mm, 10-well gels, See Blue® Plus2 molecular weight marker, NuPage® LDS Sample Buffer (4x), NuPage® sample reducing agent (10x), NuPage® MES SDS running buffer, and SimplyBlue™ Safestain. Samples were heated at 70 oC for 10 minutes prior, to loading at 1.0 µg per well, and the gels run using Zoom Dual Power supply (model ZP10002, Invitrogen) at constant 200 V. Protein and bioconjugate samples for MALDI were prepared by (1:1) with either a saturated solution of sinapic acid and α-cyano-4-hydroxycinnamic acid (10:1 w/w ratio in acetonitrile/water/trifluroacetic acid (70:30:0.03, v/v/v)).

8.1

AZO-BASED REACTIONS

8.1.1 Synthesis of 4,4-Azodiphenol (1)

A mixture of p-nitrophenol (10 g, 26 mmol), and potassium hydroxide (50 g, 91 mmol) was finely powdered resulting in a yellow mixture containing the potassium nitrophenolate intermediate. The mixture was heated to 120 °C for 60 min resulting in a yellow-orange paste conversion. After the elapsed time, the heat was raised to 200 °C for 2 h resulting in the orange-dark brown conversion with vigorous emission of gas (dominantly ammonia). After gas emission completion, water (150 mL) was poured into viscous dark-brown liquid, allowing all products to dissolve and concentrated aqueous hydrochloric acid (100 mL, 8 M)

31

was slowly added, resulting in a dark brown to beige colour change along with brown precipitate. The brown precipitate was filtered and dried in vacuo to obtain a crude solid. The crude solid was then extracted with diethyl ether (3 x 50 mL), and the combined organic extracts were dried over anhydrous magnesium sulphate and purified via silica chromatography to yield 4,4-azodiphenol as a yellow solid (1.95 g, 25%). 1H NMR (300 MHz, d6-DMSO), δ 10.09 (s, 2H), 7.71 (d, J = 8.6 Hz, 4H), 6.90 (d, , J = 8.6 Hz, 4H). requires 213.1). These results are in agreement with those in the reported literature33. 8.1.2 Synthesis of 4-Bromobutoxy Nitrobenzene (2)
13

C

NMR (75 MHz, DMSO-d6), δ 160.4, 145.6, 124.5, 116.2. MS (ESI): m/z 213.1 ([M-H]-

In a typical experiment, 1,4-dibromobutane (5.87 g, 27 mmol) was dissolved in acetone (20 mL). Potassium nitrophenolate (1.29 g, 8.7 mmol) and anhydrous potassium carbonate (3.77 g, 27 mmol) was added and the mixture was refluxed at 100 oC for 60 min until vigorous bubbling and occasional expansion of the mixture was observed. The temperature was then reduced to 75 oC and the mixture was left to reflux overnight. Water (100 mL) was added to dissolve all products and the solution was evaporated to yield a brown oil. Sonication through hexane washes (4 x 20 mL) aggregated white precipitate from the brown oil, which was filtered and dried in vacuo yielding 4-bromobutoxy nitrobenzene as a white solid (0.317 g, 11%). 1HNMR (300 MHz, CDCl3), δ 8.20 (d, J = 7.2 Hz, 2H), 6.95 (d, J = 7.5 Hz, 2H), 4.10 (t, J = 5.8 Hz, 2H), 3.49 (t, J = 6.2 Hz, 2H), 2.04 (m, 2H), 1.25 (m, 2H). These results are in agreement with those in the reported literature54. 8.1.3 Synthesis of trans-Bis-(E)-((Bromobutoxy)phenyl)diazene) (3)

A mixture of 4,4-azodiphenol (1) (0.5 g, 2.3 mmol), anhydrous potassium carbonate (1.291 g, 9.3 mmol) and dry acetonitrile (20 mL) were added to a three-necked round bottom flask and refluxed at 75 oC for 10 min with an equipped pressure-equalizing funnel. A solution of 1,4dibromobutane (2.02 g, 9.3 mmol) and acetonitrile (5 mL) was added dropwise, along with of

32

addition of a catalytic amount of potassium iodide (≈ 30 mg) and stirred for 24 h at 75 oC. The red solution was cooled to ambient temperature and poured into water (200 mL) resulting in formation of fine brown precipitate. Crude solid was collected, washed with hexane and extracted into dichloromethane and purified via silica column chromatography with dichloromethane (3 x 50 mL) as the eluant. The purified fractions showed no indication of product by 1H NMR analysis20.

8.1.4 Synthesis of trans-4,4'-Bis((6-hydroxyhexoxy)azobenzene) (4)

In a typical experiment 4,4-azodiphenol (1) (0.24 g, 1.09 mmol) was dissolved in dry acetonitrile (20 mL) and anhydrous potassium carbonate (0.604 g, 4.37 mmol) was added resulting in a yellow-orange mixture. The mixture was refluxed at 80 oC for 45 min and 6bromo-1-hexanol (0.79 g, 4.37 mmol) was added drop wise over 5 min and refluxed overnight. Yellow solid was observed to precipitate after the overnight reflux and the mixture was poured into water (500 mL) to dissolve any excess potassium carbonate. The precipitate was filtered, washed with hexane (2 x 20 mL) and dried in vacuo yielding trans-4,4‟-bis(6hydroxyhexoxy)azobenzene as a yellow solid (0.344 g, 76%). 1H NMR (300 MHz, CDCl3), δ 7.86 (d, J = 8.7 Hz, 4H), 6.98 (d, J = 8.7 Hz, 4H), 4.04 (t, J = 6.6 Hz, 4H), 3.68 (m, 4H), 1.76(m, 4H), 1.54 (m, 4H), 1.22 (m, 8H). MS (ESI): m/z 415.3 ([M+H]+ requires 415.2). These results are in agreement with those in the reported literature34.

8.1.5 Attempted Synthesis of trans-4,4’-Bis((6-maleimidohexoxy)azobenzne) (5)

In a typical experiment, triphenylphosphine (63.3 g, 0.241 mmol) was dissolved with dry tetrahydrofuran (50 mL) and the clear solution was cooled to -78 o C. Diethyl

azodicarboxylate/toluene (2:3 (v/v), 42 mg, 0.242 mmol) was added dropwise to the frozen solution over 3 min. The yellow compound, trans-4,4‟-bis(6-hydroxyhexoxy)azobenzene (4)

33

(50 mg, 0.121 mmol), was dissolved in minimal dry tetrahydrofuran and added to the solution. A catalytic amount of neophentyl alcohol (1 drop) was also added to solution. Maleimide (23.4 mg, 0.242 mmol) was dissolved in minimal dry tetrahydrofuran and added and the solution was stirred in ambient temperature over a period of 48 hr. Thin Layer Chromatography showed partial consumption of maleimide, and the solution was purified via silica column chromatography (dichloromethane/methanol (98:2, (v/v)). The purified fractions showed no indication of product by 1H NMR analysis.

8.2

VIOLOGEN-BASED REACTIONS

8.2.1 Synthesis of 4,4’-Bipyridinium-N,N-di(propylammonia) hexaflurophosphate (6)

A mixture of 4,4‟-bipyridine (2.91 g, 18.6 mmol), and 3-bromopropylamine hydrobromide (10.18 g, 46.5 mmol) were dissolved in acetonitrile (70 mL) and refluxed overnight at 100 oC. The resultant pale yellow precipitate was resuspended in diethyl ether (200 mL), filtered and washed further with diethyl ether (50 mL) then dried in vacuo to yield the dibromide salt. The solid was dissolved in minimal water and ammonium hexaflurophosphate (NH4PF6) was dissolved in small amounts. The resultant white precipitate was filtered, washed with water (20 mL) and dried in vacuo to yield 4,4‟-bipyridinium-N,N-di-(propylammonia) hexaflurophosphate as a white solid (5.64 g, 72 %). 1H NMR (300 MHz, d6-DMSO) δ 9.32 (d, J = 6.8 Hz, 4H), 8.77 (d, J = 6.8 Hz, 4H), 7.68 (s, 6H), 4.72, (t, J = 7.4 Hz, 4H), 2.88 (t, J = 7.4 Hz, 4H), 2.23 (m, 4H). MS (ESI): m/z 417.2 ([M-PF6]+ requires 417.2). These results are in agreement with those in the reported literature38.

8.2.2 Synthesis of 4,4’-Bipyridinium-N,N-di-(hydroxypropyl) dibromide (7)

A mixture of 4,4‟-bipyridine (620 mg, 4 mmol) and 3-bromo-1-propanol (8.3 g, 60 mmol)

34

were dissolved in acetonitrile (50 mL) and refluxed overnight at 80 oC. The mixture was poured into acetone (300 mL) and the precipitate was filtered and washed several times with acetone. The precipitate was further washed with warm ethanol (2 x 20 mL), then again with acetone and dried in vacuo to yield 4,4‟-bipyridinium-N,N-di-(hydroxypropyl) dibromide as a yellow solid (0.793 g, 46%). 1H NMR (300 MHz, D2O), δ 9.15 (d, J = 6.8 Hz, 4H), 8.56 (d, J = 6.8 Hz, 4H), 4.85 (t, J = 7.1 Hz, 4H), 3.71 (t, J = 5.8 Hz, 4H), 2.32 (m, 4H). MS (ESI): m/z 355.2 ([M-Br]+ requires 355.1). These results are in agreement with those in the reported literature37.

8.2.3 Attempted synthesis of 4,4’-Bipyridinium-N,N-di(maleimidopropyl) hexaflurophosphate (8)

A solution of 4,4‟-bipyridinium-N,N-di-(propylammonia) hexaflurophosphate (6) (1 g, 2 mmol), maleic anhydride (3.507 g, 35.8 mmol) in glacial acetic acid (100 mL) was heated at 100 oC for 48 hr. The resultant precipitate was filtered, washed with water (20 mL) and dried in vacuo. The remaining mixture was added to water (125 mL), extracted into dichloromethane (3 x 50 mL), slowly washed with saturated sodium hydrogen carbonate (100 mL), dried over anhydrous sodium sulphate, and the solvent removed in vacuo yielding a white solid. 1H NMR (300 MHz, D2O) analysis showed no indication of product.

8.2.4 Attempted synthesis of 4,4’-Bipyridinium-N,N-Bis(3-(6-maleimido hexanamido)propyl hexaflurophosphate) (9)

A solution of 4,4‟-bipyridinium-N,N-di-(propylammonia) hexaflurophosphate (6) (15 mg, 3.65x10-5 mol) and dry tetrahydrofuran (10 mL) was allowed to stir at ambient temperature to

35

allow the precursor to dissolve. 6-Maleimidohexanoic acid N-hydroxysuccinimide ester (EMCS) (22.5 mg, 7.29x10-5 mol) was dissolved in minimal dry tetrahydrofuran and added to the solution. N,N-diisopropylethylamine (DIPEA) (100 µL) was added and the solution was stirred overnight at ambient temperature. The resultant precipitate was filtered, dried in vacuo. 1H NMR (300 MHz, D2O) analysis showed no indication of product. The remaining THF solution was evaporated and residue placed in vacuo to yield a white solid. 1H NMR (300 MHz, d-acetonitrile) analysis of white solid showed no indication of product.

8.2.5 Attempted synthesis of 4,4’-Bipyridinium-N,N-di(propylazide dibromide) (10)

A mixture of 4,4-bipyridinium-N,N-di-(hydroxypropyl) dibromide (7) (0.4 g, 0.912 mmol) and diphenyl phosphoryl azide (DPPA) (1.01 g, 3.69 mmol) was dissolved in N,Ndimethylformamide (18 mL). The 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) (0.561 g, 3.689 mmol) was then added to the mixture resulting in a yellow to brown colour change which was left to stir at ambient temperature overnight. 5 mL of the brown solution was diluted with water (10 mL) and extracted into dichloromethane (3 x 10 mL) and evaporated in vacuo. Dichloromethane failed to extract any product from original solution. Ammonium hexaflurophosphate was added in small amounts to remaining diluted solution but failed to yield any precipitate. Recrystalistion in acetone failed to yield any precipitate. The remaining 10 mL of the brown solution was evaporated to remove the solvent resulting in a thick black residue. 1H NMR (300MHz, d6-DMSO) showed no indication of product.

8.3

POLYETHYLENE GLYCOL (PEG)-BASED REACTIONS

8.3.1 Synthesis of 4,7,10,trioxa-1, 13-bismaleimidotridecane (maleimide PEG) (11)

36

A clear solution of 4,7,10-trioxa-1,13-tridecanediamine (4 g, 18 mmol), maleic anhydride (7.12 g, 35.8 mmol) in glacial acetic acid (50 mL) was heated at 100 oC for 48 hr. The brown solution was poured into water (300 mL), extracted into dichloromethane (3 x 50 mL), washed with saturated sodium hydrogen carbonate, dried with anhydrous sodium sulphate and solvent removed in vacuo yielding 4,7,10,trioxa-1,13-bismaleimidotridecane as a brown oil (4.59 g, 67%). 1H NMR (300 MHz, d-acetone). δ 6.84 (s, 4H), 3.52 (m, 8H), 1.78 (m, 12H). MS (ESI): m/z 403.1 ([M+Na]+ requires 403.1).

8.4

AZO PSEUDOROTAXANE COMPLEXES

8.4.1 Complex A - β Cyclodextrin and trans-4,4'-bis(6hydroxyhexoxy)azobenzene pseudorotaxane

In a typical experiment, β-cyclodextrin (5 mg, 4.4x10-6 mol) and trans-4,4'-bis(6hydroxyhexoxy)azobenzene (4) (1.83 mg, 4.4x10-6 mol) mixture (1:1 mol/mol) was prepared in deuterated dimethyl sulfoxide (1 mL). The mixture was stirred for 12 h at 40 oC and further stirred for 12hr at 60 oC. The resulting mixture was analysed by two-dimensional 1H NMR COSY, ROESY, HSQC on the 400 Hz NMR spectrometer.

37

8.4.2 Complex B - α-Cyclodextrin and trans-4,4'-bis(6-hydroxyhexoxy)azobenzene pseudorotaxane

In a typical experiment, α-cyclodextrin (20 mg, 2.1x10-5 mol) and trans-4,4'-bis(6hydroxyhexoxy)azobenzene (4) (8.5 mg, 2.1x10-6 mol) mixture (10:1 mol/mol) was prepared in deuterated water (1 mL). The mixture was stirred 12 hr at 40 oC, and further stirred for 12 hr at 60 oC. The resulting mixture was analysed by two-dimensional 1H NMR, ROESY on a 400 Hz NMR spectrometer. 8.4.3 Complex C - Cucurbit[7]uril and trans-4,4'-Bis(6hydroxyhexoxy)azobenzene pseudorotaxane

In a typical experiment, cucurbit[7]uril (5 mg, 4.3x10-6 mol) and trans-4,4'-bis(6hydroxyhexoxy)azobenzene (4) (1.78 mg, 4.3x10-6 mol) mixture (10:1 mol/mol) was prepared in deuterated dimethyl sulfoxide (1 mL). The mixture was stirred for 12 hr at 40 oC,

38

and further stirred for 12 hr at 60 oC. The resulting mixture was analysed by two-dimensional
1

H NMR, ROESY on a 400 Hz NMR spectrometer.

8.5

VIOLOGEN PSEUDOROTAXANE COMPLEXES

8.5.1 Complex D - Cucurbit[7]uril and 4,4-Bipyridinium-N,N-di-(propylammonia hexaflurophosphate complex

In a typical experiment, cucurbit[7]uril (5 mg, 4.3x10-6 mol) and 4,4‟-Bipyridinium-N,N-di-( propylammonia hexaflurophosphate) (6) (2 mg, 4.7x10-6 mol) mixture (1:1 mol/mol) was prepared in a solution 0.2 M sodium chloride/deuterated water (1 mL). The mixture was stirred for 12 hr at ambient temperature. The resulting mixture was analysed by twodimensional 1H NMR, ROESY on a 400 Hz NMR spectrometer. 8.5.2 Complex E - Cucurbit[7]uril and 4,4’-Bipyridinium-N,N-di-(hydroxypropyl) dibromide complex

In a typical experiment, cucurbit[7]uril (5.4 mg, 4.6x10-6 mol) and 4,4‟-bipyridinium-N,N-di( hydroxypropyl) dibromide (7) (2 mg, 4.7x10-6 mol) mixture (1:1 mol/mol) was prepared in

39

deuterated water (1 mL). The mixture was stirred for 12 hr at 60 oC. 1H NMR of the mixture was conducted on a 300 Hz NMR spectrometer.

8.6 POLYETHYLZENE GLYCOL (PEG) PSEUDOROTAXANE FORMATIONS
8.6.1 Complex F - α-Cyclodextrin and of 4,7,10-trioxa-1,13-tridecanediamine

In a typical experiment, α-cyclodextrin (44 mg, 4.5x10-5 mol) and 4,7,10-trioxa-1,13tridecanediamine (16 mg, 7.3x10-5mol) mixture (1:2 mol/mol) was prepared in deuterated water (1 mL). The mixture was stirred for 12 hr at 60 oC and cooled to 4 oC for 1hr. 1H NMR of the mixture was conducted on a 400 Hz NMR spectrometer.

40

8.6.2 Complex G - β-Cyclodextrin and of 4,7,10-trioxa-1,13-tridecanediamine

In a typical experiment, β-cyclodextrin (52 mg, 4.5x10-5 mol) and 4,7,10-Trioxa-1,13tridecanediamine (26 mg, 1.2x10-4mol) mixture (1:2.5 mol/mol) was prepared in deuterated water (1 mL). The mixture was stirred for 12 hr at 60 oC, cooled to 4 oC for 1 hr and left in ambient temperature for 48 hr. 1H NMR of the mixture was conducted on a 400 Hz NMR spectrometer.

8.7

MALEIMIDE-CAPPED ROTAXANE FORMATIONS

8.7.1 Formation H - α-Cyclodextrin and 4,7,10,trioxa-1, 13bismaleimidotridecane

41

In a typical experiment, α-cyclodextrin (32.8 mg, 3.4x10-5 mol) and 4,7,10-trioxa-1,13bismaleimidiotridecane (16.5 mg, 4.3x10-5 mol) mixture ( 1:1 mol/mol) was prepared in deuterated water (1 mL). The mixture was for 12 hr at 60 oC and cooled to 4 oC for 1hr. 1H NMR of the mixture was conducted on a 400 Hz NMR spectrometer. 8.7.2 Formation I - β-Cyclodextrin and 4,7,10,trioxa-1, 13-bismaleimidotridecane complex

In a typical experiment, β-cyclodextrin (40 mg, 3.5x10-5 mol) and 4,7,10-Trioxa-1,13bismaleimidiotridecane (2 mg, 1.3x10-4mol) mixture (1:4 mol/mol) was prepared o o

in

deuterated water (1 mL). The mixture was stirred for 12 hr at 60 C and cooled to 4 C for 1hr. 1H NMR of the mixture was conducted on a 400 Hz NMR spectrometer. 8.7.2 Formation J - β-Cyclodextrin and 4,7,10,trioxa-1, 13-bismaleimidotridecane complex

42

In a typical experiment, cucurbit[7]uril (61 mg, 5.2x10--6 mol) and 4,7,10-Trioxa-1,13bismaleimidiotridecane (2 mg, 5.3x10--5mol) mixture (1:2) mol/mol) was prepared in deuterated water (1 mL). The mixture was stirred for 12 hr at 60 oC and cooled to 4 oC for 1h. The precipitate was collected, washed with water and dried in vacuo. 1H NMR of the mixture was conducted on a 400 Hz NMR spectrometer.

8.8

BIOCONJUGATE ROTAXANE FORMATION

8.8.1 Cytochrome c biorotaxane with α & β-cyclodextrin Maleimide-PEG

43

In a typical experiment, complexes H, I, or J containing, α-cyclodextrin, βcyclodextrin or cucurbit[7]uril and (11) (0.013 M, 500 µL) in deuterated water were diluted (1:8) with Phosphate Buffer (0.1 M, 3.5 mL, pH 7). Cytochrome c (6 mg, 4.9 x 10-7) was then added to 50 µL of these macyrocyles and maleimido PEG (11) mixtures (1.62 mM, 8.1 x 10-8 mol). The solutions were allowed to sit 12 h at ambient temperature and the samples then stored at -20 o C. The resulting reaction mixtures were then analysed by SDS gel

electrophoresis in both the non-reduced and reduced form showing a clear difference for the latter between the reference (cyt c) and the reaction mixtures. Mass spectrometry analysis was then carried using MALDI-TOF with the following main product identified by MS-MALDI: i) Reaction of Complex H (α-cyclodextrin+(11)) + cyt c: m/z 13071 ([cyt-11 monostoppered dumbbell]+ requires 13090), 25972 ([cyt-11-cyt bis-stoppered dumbell]+ requires 25800). ii) Reaction of Complex I (β-cyclodextrin+(11)) + cyt c: m/z 13069 ([cyt-11 monostoppered dumbbell]+ requires 13090), 25975 ([cyt-11-cyt bis-stoppered dumbbell]+ requires 25800). iii) Reaction of Complex J (cucurbit[7]uril+(11)) + cyt c: m/z 13069 ([cyt-11 monostoppered dumbbell]+ requires 13090), 14277 ([cyt-11 mono-stoppered dumbbell + cucurbit[7]uril pseduorotaxane+Na]+ requires 14276), (25817 ([cyt-11-cyt bis-stoppered dumbbell]+ requires 25800), 26989 (([cyt-11-cyt:cucurbit[7]uril rotaxane+Na]+ requires 26989.

44

8.8.2 Bovine Serum Albumin biorotaxane with α & β-cyclodextrin Maleimide-PEG

In a typical experiment, complexes H, I, or J, containing α-cyclodextrin, β-cyclodextrin or cucurbit[7]uril and 11 mixtures (0.013 M, 500 µL) in deuterated water were diluted (1:8) with

45

Phosphate Buffer (0.1M, 3.5 mL, pH 7). Bovine Serum Albumin (10 mg, 1.62 x 10-7) was then added to 50 µL of the macyrocyles and 11 mixtures (1.62 mM, 8.1 x 10-8 mol). The solutions were allowed to sit for 12 h at ambient temperature and samples then stored at -20 o C. The resulting reaction mixtures were then analysed by SDS gel electrophoresis in both the

non-reduced and reduced form which did not indicate any difference between the reaction mixtures and the reference (BSA). Mass spectrometry analysis was then carried using MALDI-TOF but no indication of products other than the cyt-BSA-11 mono-stoppered dumbbell were obtained.

46

9. RESULTS AND DISCUSSION
9.1 SYNTHESIS

9.1.1 Azo-based Synthesis The mechanism of the synthesis of 4,4-Azodiphenol (1) involves the decomposition and disproportionation of (12) intermediate. The immediate formation of the intermediate potassium nitrophenolate (12) occurs upon addition of the potassium hydroxide to p-nitrophenol. This accounts for the greyyellow colour change in mixture and supported through
1

the p-nitrophenol

H NMR. To initiate the

decomposition and disproportionation, the intermediate must undergo two heat transition stages. The first stage utilizes the intermediate in liquid phase at 120 °C resulting in a yelloworange paste conversion, and the second stage results in the decomposition and disproportionation of the intermediate at 200 °C along with a visible orange-dark brown conversion and vigorous emission of gas (dominantly ammonia). Decomposition and disproportionation only occurs until this temperature threshold of 200 °C is achieved.

In preliminary experiments, decomposition and disproportionation did not occur because initial studies indicated that formation of the 4,4-azodiphenol can occur at 180 °C20. The result was the yield of the yellow solid intermediate, and without supported evidence from any characterisation data, the intermediate potassium nitrophenolate was mistakenly assumed to have been the 4,4-azodiphenol compound. The attempted synthesis of trans-4,4‟bis(6-hydroxyhexoxy)azobenzene (3) only resulted in a SN2 reaction of the intermediate potassium nitrophenolate with 1,4-dibromobutane forming the 4-Bromobutoxy nitrobenzene (12) compound in low yields. Purification of the 4,4-azodiphenol compound was most efficient with silca chomotagaphy column with diethyl ether as the eluant. Synthesis of the trans-4,4‟-bis(6-hydroxyhexoxy)azobenzene (4) was successful with the purified 4,4-azophenol. However, 1H NMR results suggested evidence of the cis (Z) isomer within the yield (30%). Silica column chromatography could not remove the isomer which initially thought to be a by-product of the reaction. A thermal treatment of the product was conducted with acetonitrile by shielding it from light and refluxing at 90 oC. The result was an isomer reduction within the yield to (4%) (Figure 9.1). The azo compound was also insoluble in water, which presented an obstacle for further bioconjugation synthesis.

47

Figure 9.1 – Cis Isomer Reduction

48

Attempted synthesis of trans-4,4‟-bis((6-maleimidohexoxy)azobenzene) (5) was based on a Mitsunobu reaction as the method of maleimide substitution due to the its high dissociation constant. However, after numerous attempts with Ph3P/DEAD as the catalyst, substantial purification on these solutions could not identify any product by 1H NMR, which instead shows the presence of the unreacted maleimide and azo compounds. This suggests that the conditions set for the reaction were insufficient to promote dehydration for formation of the -4,4‟-bis((6-maleimidohexoxy)azobenzene) (5).

Scheme 9.1

9.1.2 Viologen-based Synthesis Synthesis of 4,4‟-bipyridinium-N,N-di-(propylammonia) hexaflurophosphate (6) was successful and in high yields (72%). The attempted synthesis of 4,4‟-bipyridinium-N,Ndi(maleimidopropyl) hexaflurophosphate (8) only resulted in open-form maleimeide substitution (carboxylic acid) with the 4,4‟-bipyridinium-N,N-di-(propylammonium

hexaflurophosphate). A different approach in maleimide substitution was investigated with 6maleimidohexanoic acid N-hydroxysuccinimide ester (EMCS) with the aim to synthesis 4,4‟bipyridinium-N,N-bis(3-(6-maleimido hexanamidopropyl) hexaflurophosphate) (9). Similarly, only open-form maleimeide substitution (carboxylic acid) was evident in 1H NMR characterisation data. This suggests that the conditions set for the reaction were insufficient to promote dehydration, similar to that of the Mitsunobu reaction in the attempted synthesis of trans-4,4‟-bis((6-maleimidohexoxy)azobenzne) (5).

49

Scheme 9.2

A different scheme was proposed for maleimide substitution which involves synthesising 4,4‟-bipyridinium-N,N-di-(hydroxypropyl) dibromide (7) and further reacting to form the azide derivative 4,4‟-bipyridinium-N,N-di(propylazide) dibromide that can be used as the basis for click chemistry8,22. Synthesis of 4,4‟-bipyridinium-N,N-di-(hydroxypropyl) dibromide (7) gave a yield of 46% in acetonitrile, however, 1H NMR still showed evidence of precursors along with mono-substituted by-products, even after attempted purification via recystallisation. Despite this, the sample was reacted with DPPA to attempt to synthesise the 4,4‟-bipyridinium-N,N-di(propylazide) dibromide. The result was a thick black residue which did not contain the product identifiable by 1H NMR.

50

DBU

Scheme 9.3

The main explanation for these unsuccessful reactions may be due to the ease of the viologen to be reduced to its radical monocation (Scheme 9.4). Evidence may also suggest that cleavage of the bipyridinium bond may occur in the presence of a strong base such as DBU.

Scheme 9.4 – Viologen transformations

9.1.3 PEG-based Synthesis One of the more successful syntheses within this project was based on the polyethylene glycol (PEG) motif. Synthesis of trans-4,4‟-bis((6-maleimidohexoxy)azobenzne) was successful with a yield of 67%. Conditions of the reaction were similar to the reaction undertaken in the attempted synthesis of 4,4‟-bipyridinium-N,N-di(maleimidopropyl hexaflurophosphate) (8). Dehydration of the intermediate occurred and the final product was

51

a brown oil. The product was fully soluble in water, which is a major advantage for bioconjugation synthesis (Scheme 9.5).

Scheme 9.5

52

9.2

PSEUDOROTAXANE COMPLEXES

The purpose of the pseudorotaxane complexes were to simulate whether the macrocycle could bind to the motif binding site on the axial component in absence of the maleimide stopper. The approach would depict if any macrocycle–motif binding site electrostatic interactions would occur. The interactions could be detected using 1H NMR, ROESY and HSQC. In typical experiments, 1:10 axial-macrocycle ratios were used. This increased the probability that associative interactions between the components would occur. Table 9.1 below gives a summary of the pseudorotaxane formation.
Table 9.1 - Pseudorotaxane formations

Macrocycle Axial Component Azo (4) Viologen 1 (6) Viologen 2 (7) PEG (11)

α-Cyclodextrin

β-Cyclodextrin

Cucurbit[7]uril

No No

Inconclusive No

Yes Yes Yes Yes

Results show that neither α- or β-cyclodextrins could form pseudorotaxane complexes with any of the motif axial components. However, the macrocycle cucurbit[7]uril did form pseudorotaxane complexes with all of the motif axial components. This indicates that there are solubility issues that arise with the α- & β-cyclodextrins in water. The α- & βcyclodextrins macrocycles are, therefore, found to be inefficient for rotaxane formation under these condtions. 9.2.1 Synthesis of Azo Pseudorotaxanes One of the major disadvantages of the azo axial component trans-4,4‟-bis(6hydroxyhexyloxy)azobenzene (4) is its insolubility in water. To overcome this issue, tests were also undertaken in deuterated dimethyl sulfoxide. All three macroycles were examined for pseudorotaxane complexation with the azo compound and interactions were characterised by 1H NMR and two-dimensional ROESY. Figure 9.2 depicts the ROESY spectrum of the proton-proton coupling between the trans-4,4‟-bis(6-hydroxyhexoxy)azobenzene (4) (highlighted yellow) and β-cyclodextrin (highlighted red). The proposed region of interaction between the aromatic protons of the azo compound and the protons of the β-cyclodextrin do not depict any ROESY peaks. Although

53

there is some evidence that the spacer of the azo component show cross peaks with the β cyclodextrin peaks, insufficient data can prove that formation of pseudorotaxanes formed.

Figure 9.2– Azo -β cyclodextrin psusedorotaxane

Figure 9.3 depicts the ROESY spectrum of the proton-proton coupling between the trans4,4‟-bis(6-hydroxyhexoxy)azobenzene (4) (highlighted yellow) and α-cyclodextrin

(highlighted blue). The proposed region of interaction between the aromatic protons of the azo compound and the protons of the α-cyclodextrin do not depict any ROESY peaks. No other interactions between the two components could be identified.

54

Figure 9.3 - Azo -α cyclodextrin psusedorotaxane

Figure 9.4 depicts the ROESY spectrum of the proton-proton coupling between the trans4,4‟-bis(6-hydroxyhexoxy)azobenzene (4) (highlighted yellow) and cucurbit[7]uril

(highlighted purple). The proposed region of interaction between the aromatic protons of the azo compound and the protons of the cucurbit[7]uril contains ROESY peaks indicated by the dashed area of the spectrum. This confirms that proton-proton coupling occurs between the aromatic protons of the azo compound and the protons of the cucurbit[7]uril, hence conformation of a pseudorotaxane complex.

55

ROESY 9.4 - Azo – cucurbit[7]uril pseudorotaxane

56

9.2.2 Synthesis of Viologen Pseudorotaxanes A

B

Pseudorotaxane

Figure 9.5 – Complex D: 4,4’-bipyridinium-N,N-di-(propylammonia) hexaflurophosphate: cucurbit[7]uril pseudorotaxane

Results from the 1H NMR (Figure 9.5) depict proton peak broadening of the, 4,4‟bipyridinium-N,N-di-(propylammonia) hexaflurophosphate (6), protons upon addition of

cucurbit[7]uril macrocycle to form Complex D. Furthermore, formation of a new proton peak (δ 7.2 ppm) concludes formation of a pseudorotaxane. Heat treatment of the solution resulted in the increase of the proton peak. The result of the broadening may be due to the fast proton exchanges between the cucurbit[7]uril and the viologen. Similar proton broadening could be seen with Complex E from, 4,4‟-bipyridiniumN,N-di-(hydroxypropyl) dibromide (7) and cucurbit[7]uril, which also indicated

pseudorotaxane formation (Figure 9.6).

57

Pseudorotaxane

Figure 9.6 – Complex E: 4,4’-bipyridinium-N,N-di-( hydroxypropyl) dibromide cucurbit[7]uril pseudorotaxane

A two-dimensional ROESY (Figure 9.7) was conducted on the Complex Dcucurbit[7]uril mixture. No cross peaks between observed between the aromatic viologen protons and the cucurbit[7]uril protons. Although sufficient evidence from the 1H NMR peak broadening indicates that pseudorotaxane formation occurs in the viologen systems, formation in the viologen mixture was low, indicated by the formation of a new small proton peak. This means that a large amount of uncoupled components were existent in the mixtures and any pseudorotaxane interactions would be too weak for detection in a normal ROESY experiment.

58

Figure 9.7 – ROESY of the Complex D: 4,4’-bipyridinium-N,N-di-( hydroxypropyl) dibromide: cucurbit[7]uril pseudorotaxane

59

9.2.3 Synthesis of PEG Pseudorotaxanes Results of the 1H NMR α- and β-cyclodextrin – PEG precursor mixture depict no formation of Pseudorotaxanes. Due to time constraints, no analysis on the cucurbit[7]uril – PEG precursor pseudorotaxane formation was commenced. Results of the 1H NMR (Figure 9.8) α and β cyclodextrin – bismaleimido PEG mixture aldo depict no formation of pseudorotaxane. However, successful synthesis of pseudorotaxane the cucurbit[7]uril – bismaleimido PEG was depicted in the 1H NMR with the formation of a large peak. This suggests fast proton-proton exchanges between the components, resulting in the peak formation at δ 4.73 ppm. Evidence of uncoupled cucurbit[7]uril and bismaleimido PEG (11) is visible in the 1H NMR. However, the cucurbit[7]uril peaks has broaden and almost disappeared, suggesting at least 93% successful pseudorotaxane formation.

pseudorotaxane

Figure 9.8 - 1H NMR of cucurbit[7]uril and bismaleimido PEG (11)

60

9.3

ROTAXANE BIOCONJUGATION

11

Figure 9.9 – Scheme for Bioconjugation

The synthesised macrocycle-bismaleimido-PEG (11) mixtures (complexes H, I and J) were used as precursors for bioconjugation reactions. Previous results (9.2.3) indicated that pseudorotaxane formation occurred only the cucurbit[7]uril mixture, but no formation occurred in the α- and β-cyclodextrin mixtures. Hence it is expected that only proteinappended rotaxane will only form with cucurbit[7]uril mixtures. Macroycle-bismaleimidoPEG (11) interactions are still assumed to be present during transfer into the PBS buffer, since the environment exhibits the same conditions. Bioconjugation will be a result a Michael reaction with the thiol group found on a cysteine amino acid, from either Cytochrome c or Bovine Serum Albumin (BSA), to form a stable carbon-sulfur thioether. Cytochrome c can exists as a dimer via a disulfide linkage and Bovine Serum Albumin exist also as a dimer and tetramer in solution7,14,21. 9.3.1 Gel Electrophoresis Gel SDS electrophoresis of all samples were conducted on both the non reduced and reduced samples from bioconjugation reactions. The purpose of the reduced sample is to ensure that all cytochrome c and BSA proteins are in monomer structure by cleavage of disulfide

61

linkages. The non-reduced samples depict both the monomer and dimer for Cytochrome c and the monomer, dimer and tetramer for BSA. Evidence of bismaleimide-PEG (11) bioconjugated cytochrome c dumbbell and/or rotaxane formation is obtained from the reduced samples which depicts dimer bands (≈ 28 kDa) (Figure 9.10), whereas no dimer is observed for the unreacted Cytochrome c. Furthermore the intensity of the dimer bands compared to the monomer in the reduced samples is considerably less (25:75) ratio, indicating about 20-30% yield of the cyt-11-cyt dumbbell and/or the desired rotaxane. However, evidence of a biorotaxane system cannot be obtained from this method as the molecular weight difference is too small (