Vrije Universiteit Brussel Faculteit Wetenschappen

Vakgroep Bio-Ingenieurswetenschappen Onderzoeksgroep Ultrastructuur

Academiejaar 2007-2008

Therapeutic Nanoreactors: Combining Chemistry and Biology in a novel Triblock

Coploymer Drug Delivery System

Proefschrift voorgedragen tot het bekomen van de graad van doctor in de toegepaste biologische wetenschappen

Thesis submitted in fulfillment of the requirements for the degree of doctor

(PhD) in applied biological sciences

Ir. An Ranquin

Promotoren: Prof. Dr. ir. Jan Steyaert Dr. Patrick Van Gelder Dit werk kwam tot stand in het kader van een specialisatiebeurs van het IWT-Vlaanderen (Instituut voor de aanmoediging van Innovatie door Wetenschap en Technologie in Vlaanderen) en een BOF mandaat van de Vrije Universiteit Brussel

Published by the research Group of Ultrastructure. ULTR, Vrije Universiteit Brussel, Pleinlaan 2, 1050 Brussels, Belgium Departement of Molecular and Cellular Interactions, VIB, Belgium. Appart from any fair dealing for the purpose of research or private study or criticism or review, this publication may not be reproduced, stored in a retrieval system, or transmitted in any form or any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, without the prior permission in writing of the publisher. Ranquin – Therapeutic Nanoreactors: Combining Chemistry and Biology in a Novel Triblock Copolymer Drug Delivery System – 2008. PhD Thesis Vrije Universiteit Brussel, Brussels, Belgium Keywords: ADEPT, nanoreactors, triblock copolymers, enzyme-prodrug therapy, drug delivery. ISBN: 9789081294829

I

Acknowledgements

This research was funded by the Instituut voor de aanmoediging van Innovatie door

wetenschap en technologie in Vlaanderen (IWT) and the Vrije Universiteit Brussel

(VUB) in collaboration with Prof. W. Meier from the University of Basel (UB) who

kindely provided us with PMOXA-PDMS-PMOXA polymers.

The past six years have been filled with periods of joy and sorrow, happiness and

despair so it goes without saying that I have a ‘few’ people to thank for support

during these periods. First I would like to thank my promoter Prof. Dr. J. Steyaert

for giving me the opportunity to conduct this research at the Ultrastructure lab. A

special thanks goes to my co-promotor Dr. P. Van Gelder who stood by me during

the last six years. Thanks a lot Patrick for the meaningful discussions, great ideas,

guidance and for revising this manuscript. I would also like to thank the members or

former members of the membrane group: Kris, Kim and Gerard and Anneke and

Wim V. for introducing me into the world of enzymology. A special thanks goes to

Caroline who I could always depend on for experiments, meaningful discussions and

her organization skills which makes Caroline a great colleague and a good friend. I

would also like to thank Ronny for the help with AFM experiments, Els for the

nanobody constructs and info as well as her knowledge on yeast expression and

Lieven in times of computer crisis. And off course I have to thank all of my

colleagues at the ULTR lab for the amazing atmosphere: Elke, Karo, Adinda, Celine,

Sarah, Khadija, Inge, Maia DK, Nathalie, Mike, Klaas, Wim C., Kathy, Nadine,

Marieke, Bruno,…

There are also several people from other labs that deserve my gratitude. The people

from CMIM: Nick, Lea, Jo and Benoit thanks a lot for the good advice. The people

from SWITCH: Maia DB and Hannah, thanks for supplying me week after week with

fresh neuroblastoma cell cultures. Also a special thanks to the people from Basel:

Alexandra, Samantha, Caroline, Per, Fabian, Olivier, Sandrine and off course

Wolfgang for your opinions and knowledge and for entertaining me on my trips to

Basel.

II

There are also some people from outside work that I would like to thank. First and

foremost my partner Jeroen who sometimes had to put up with my bad moods

when things didn’t go well and for always believing in me and supporting me. Also

thanks a lot baby for making yummy dinner each night for the last two months, I

really appreciate it. Special thanks to my baby boy Stanneke who was always there

(for the last 9,5 months anyway) to put a smile on my face, even after only 5 hours

of sleep. Also a thanks to my parents for supporting me and believing in me. And

thanks mum for the beautiful cover.

I can only hope that I didn’t forget anyone at the end of this long list. If I did, I

sincerely apologise.

III

Table of contents Chapter 1: Introduction 1

1.1 Chemotherapy: current status 3 1.2 Enzyme-prodrug therapy 6

1.2.1 ADEPT 6 1.2.2 GDEPT 11

1.3 Liposomes as drug delivery vehicles in cancer therapy 16 1.3.1 Long-circulating liposomes 17

1.3.2 Targeting liposomes to tumor tissue 18 1.3.3 Stimuli-sensitive liposomes 21 1.3.4 Conclusion 27

1.4 Polymer vesicles 28 1.5 References 30

Chapter 2: State of the art: a nanoreactor coming to life 39

2.1 Aim of the work: a novel directed enzyme-prodrug strategy 41

2.2 PMOXA-PDMS-PMOXA triblock copolymers 42 2.3 Bacterial outer membrane proteins 44

2.3.1 Aspecific porines OmpF and PhoE 45 2.3.2 Specific porin Tsx 47

2.4 Nucleoside hydrolase of Trypanosoma vivax 49 2.5 Purine nucleoside analogs as prodrugs 54

2.5.1 Deoxyadenosine analogs 54 2.5.2 6-Thioguanosine 57

2.6 Targeting of nanoreactors 58 2.6.1 Passive targeting 58 2.6.2 Active targeting with single domain

antibodies 59 2.7 References 65

Chapter 3: Encapsulation of therapeutic nucleoside hydrolase in functionalised nanocapsuls 71

3.1 Introduction 73 3.2 Materials and methods 75

3.2.1 Purification of T. vivax nucleoside hydrolase 75 3.2.2 Purification of E. coli porins OmpF and PhoE 76 3.2.3 Electrophysiology 76

IV

3.2.4 Preparation of proteoliposomes and swelling assays 77

3.2.5 Preparation of enzyme encapsulating proteoliposomes 78

3.2.6 Fluorescence measurements 78 3.3 Results 78

3.3.1 Purification and functional characterisation of TvNH and porins 78

3.3.2 Solute transport through OmpF and PhoE 80 3.3.3 Activity of nucleoside hydrolase

encapsulated in proteoliposomes 81 3.3.4 Interactions between TvNH and liposomes 83

3.4 Discussion 86 3.5 References 88

Chapter 4: Comparison and characterisation of PMOXA-PDMS-PMOXA vesicles 91

4.1 Introduction 93 4.2 Materials and methods 94

4.2.1 Purification of T. vivax nucleoside hydrolase 94 4.2.2 Cloning, expression and purification of Tsx 94 4.2.3 Preparation of nanoreactors 96 4.2.4 DLS measurements 97 4.2.5 Reducing sugar assay 97 4.2.6 Atomic Force Microscopy (AFM) 97 4.2.7 Transmission Electron Microscopy (TEM) 98 4.2.8 Fluorescent labelling of TvNH 98 4.2.9 In vitro nanoreactor-macrophage

interactions 99 4.3 Results 100

4.3.1 Polymer batch variability 100 4.3.2 AFM and TEM images of nanoreactors 104 4.3.3 Fluorescently labelled Tsx-TvNH

nanoreactors 106 4.3.4 In vitro interaction of fluorescent

nanoreactors with macrophages 107 4.4 Discussion 109 4.5 References 111

Chapter 5: Therapeutic nanoreactors: combining chemistry and biology in a novel triblock copolymer drug delivery system 113

5.1 Introduction 115 5.2 Materials and methods 119

V

5.2.1 Purification of TvNH 119 5.2.2 Purification of E. coli porins OmpF and Tsx 119 5.2.3 Preparation of nanoreactors 120 5.2.4 Trypsin digestion 120

5.3 Results 121 5.3.1 Kinetic parameters of TvNH 121 5.3.2 Preparation of nanoreactors 122 5.3.3 Encapsulation efficiency of TvNH 123 5.3.4 Activity of nanoreactors 124

5.4 Discussion 127 5.5 References 128

Chapter 6: Nanoreactor mediated prodrug activation and killing of neuroblastoma cells 131

6.1 Introduction 133 6.2 Materials and methods 135

6.2.1 Nucleoside analogs and nucleobase analogs 135 6.2.2 Cytotoxicity assay 136 6.2.3 EC50 determination 136 6.2.4 Swelling assay with Tsx proteoliposomes 137 6.2.5 Preparation of TvNH nanoreactors 137 6.2.6 Enzymatic activity assay 138 6.2.7 Limulus Amebocyte Lysate (LAL) assay 138

6.3 Results 139 6.3.1 Cytotoxicity screening of prodrugs and drugs 139 6.3.2 EC50 determination of 6-thioguanosine and

6-thioguanine 141 6.3.3 6-Thioguanosine activation by human purine

nucleoside phosphorylase (hPNP) 143 6.3.4 Transport of 6-thioguanosine and 6-

thioguanine by Tsx 144 6.3.5 Kinetics of Tsx-TvNH nanoreactors 145 6.3.6 Cytotoxicity of 6-thioguanosine activated by

TvNH encapsulating nanoreactors 146 6.4 Discussion 148 6.5 References 153

Chapter 7: Production of cAbLys-3 mutants for selective coupling to nanoreactors 155

7.1 Introduction 157 7.2 Materials and methods 159

7.2.1 Site directed mutagenesis of cAbLys-3 159

VI

7.2.2 Bacterial expression and purification of cAbLys-3 mutants 160

7.2.3 Expression of cAbLys-3 S17C in Pichia pastoris 161

7.2.4 Enzyme-Linked Immuno Sorbent Assay (ELISA) 163

7.3 Results 164 7.3.1 Production of cAbLys-3 mutants in E. coli 164 7.3.2 Functionality of cAbLys-3 mutants 167 7.3.3 Production of cAbLys-3 S17C in P. pastoris 167

7.4 Discussion 168 7.5 References 170 7.6 Appendix 171

Chapter 8: General discussion 175 Summary 183 Samenvatting 187 Publications 191

Chapter 1 Introduction

1

Chapter 1:

Introduction

Chapter 1

2

Introduction

3

1.1 Chemotherapy: current status

The first drug used for cancer chemotherapy, Mustard gas, was not originally

intended for that purpose, but used as a chemical warfare agent. Scientist

discovered that it had a toxic effect on rapidly growing tumor cells. That

experience led researchers to look for other substances that might have similar

effects. As a result, many other drugs have been developed to treat cancer, and

drug development since then has exploded into a multi-billion dollar industry.

Most chemotherapeutic drugs work by impairing mitosis (cell division),

therefore specifically targeting fast-dividing cells and inducing apoptosis.

Unfortunately this means that non-malignant fast dividing cells such as those

responsible for hair growth and for replacement of the intestinal epithelium

(lining) are also often affected. This causes severe side effects to patients

receiving chemotherapy. Since most chemotherapeutics affect rapidly growing

tumors, chemotherapy is mostly used to treat tumors with high growth rates

such as acute myelogenous leukaemia and the aggressive lymphomas, including

Hodgkin's disease. Young tumors are treated more effectively because

mechanisms regulating cell growth are usually still preserved. With succeeding

generations of tumor cells, differentiation is typically lost, growth becomes less

regulated, and tumors become less responsive to most chemotherapeutic agents.

Near the centre of some solid tumors, cell division has effectively ceased, making

them insensitive to chemotherapy. Therefore, solid tumors are usually treated by

radiation therapy and surgery due to the fact that the chemotherapeutic agent

often does not reach the core of the tumor. Another problem occurs when cancer

cells become more resistant to chemotherapy treatments as a result of reduced

drug accumulation in the tumor cells. In 1976, Juliano and Ling discovered the

first surface glycoprotein responsible for altering drug permeation in drug

resistant tumor cells, p-glycoprotein (Juliano and Ling, 1976). It is responsible

for the active efflux of drugs from cancer cells. Since then, other multidrug

resistance proteins have been identified, including the multidrug resistance-

associated protein (MRP1, ABCC1) (Abe et al., 1995), identified in small cell lung

carcinoma and the breast cancer resistance protein (mitoxantrone resistance

Chapter 1

4

protein, ABCG2) (Doyle et al., 1998). Medications to inhibit the function of these

multidrug resistance proteins can enhance the efficacy of chemotherapy.

The majority of chemotherapeutic drugs can be divided into: alkylating agents

(cisplatin, carboplatin, oxaliplatin,…) (Alderden et al., 2006; Canetta et al., 1985;

Graham et al., 2004), antimetabolites (fludarabine, 2-chlorodeoxyadenosine, 2'-

deoxycoformycin...) (Cheson, 1992), anthracyclines (doxorubicin,

mitoxantrone,...) (Lown, 1993), alkaloids (camptothecin, paclitaxel,...) (Wall and

Wani, 1995), monoclonal antibodies (trastuzumab, cetuximab, rituximab,...)

(Albanell et al., 2003) and other antitumor agents. All of these drugs affect cell

division or DNA synthesis and function in some way. Some newer agents don't

directly interfere with DNA. These include the new tyrosine kinase inhibitor

imatinib mesylate (Gleevec® or Glivec®) (Droogendijk et al., 2006), which

directly targets a molecular abnormality in certain types of cancer (chronic

myelogenous leukaemia, gastrointestinal stromal tumors).

In addition, some drugs may be used which modulate tumor cell behaviour

without directly attacking those cells. Hormone treatments fall into this category

of adjuvant therapies.

One of the major difficulties in chemotherapy is the dosage: if the dose is too

low, it will be ineffective against the tumor, while at excessive doses the toxicity

(side-effects) will be intolerable to the patient. This has led to the formation of

detailed "dosing schemes" in most hospitals, which give guidance on the correct

dose and adjustment in case of toxicity. Harmful and lethal toxicity from

chemotherapy limits the dosage of drugs that can be given. Some tumors can be

destroyed by sufficiently high doses of chemotherapeutic agents. Unfortunately,

these high doses cannot be given because they would be fatal to the patient.

Several efforts have been undertaken to increase the tolerated dose of

chemotherapeutics. These include Haematopoietic stem cell transplant

approaches, isolated infusion approaches, targeted delivery mechanisms and

prodrug therapies. Since chemotherapeutic toxicity mainly affects haematopoietic

cells, cell transplants are used in combination with chemotherapeutics to

decrease damage to haematopoietic tissue. However, years of research has

yielded little proof of efficacy. Therefore haematological malignancies such as

myeloma, lymphoma, and leukaemia remain the main indications for stem cell

transplants.

Introduction

5

Isolated limb perfusion (often used in melanoma), or isolated infusion of

chemotherapy into the liver or the lung have been used to treat some tumors.

This way, a very high dose of chemotherapy can be delivered to tumor sites

without causing overwhelming systemic damage. Unfortunately, while these

approaches can be useful against solitary or limited metastases, they are - by

definition - not systemic and therefore do not treat distributed metastases or

micro metastases. Another way of increasing the dosage of therapeutics is the

use of specifically targeted delivery vehicles to increase effective levels of

cytotoxins for tumor cells while reducing effective levels for other cells. This

should result in an increased tumor kill and/or reduced toxicity. Specifically

targeted delivery vehicles have a differentially higher affinity for tumor cells by

interacting with tumor specific or tumor associated antigens. In addition to their

targeting component, they also carry a payload which is the chemotherapeutic

agent. Specifically targeted delivery vehicles vary in their stability, selectivity and

choice of target, but in essence they all aim to increase the maximum effective

dose that can be delivered to the tumor cells. Reduced systemic toxicity means

that they can also be used in sicker patients, and that they can carry new

chemotherapeutic agents that would have been far too toxic to deliver via

traditional systemic approaches.

Finally another way to increase dosage is the use of prodrugs. Prodrugs are

non toxic precursors that are designed to be transformed after administration to

form a pharmacologically active species. Such prodrugs are divided into two

categories: prodrugs designed to increase bioavailability and prodrugs designed

to deliver anticancer agents locally at the site of the tumor to increase specificity

and decrease systemic toxicity. The first category of prodrugs is used in case of

poorly soluble anti cancer agents. They are usually metabolically transformed to

their active compound. The latter category however is specifically activated at

the tumor site. Activation of the prodrugs can be achieved by the tumor

environment, enzymes specifically up-regulated in tumor tissue, enzymes

excreted by tumor cells or exogenous enzymes directed to tumor tissue.

Chapter 1

6

1.2 Enzyme-prodrug therapy

Enzyme-prodrug therapies were developed to deliver exogenous enzymes to

tumor tissue to selectively convert a relatively non-toxic prodrug to an active

drug. This leads to a higher local drug concentration in the tumor, improving the

anti tumor effect. Additionally, this lowers the systemic drug concentration

hereby reducing unwanted side effects that accompany conventional cancer

chemotherapy. The enzyme can either be delivered by an antibody-enzyme

fusion protein (antibody-directed enzyme-prodrug therapy, ADEPT) or by a

vector carrying the gene encoding for the exogenous enzyme (gene-directed

enzyme-prodrug therapy, GDEPT). When a viral vector is used for gene delivery,

the latter is also referred to as viral-directed enzyme-prodrug therapy (VDEPT)

(Figure 1).

Figure 1: Schematic overview of enzyme-prodrug therapy

1.2.1. ADEPT

Antibody directed enzyme-prodrug therapy is a two step approach where

selectivity for the target is achieved by an antibody in an antibody-enzyme fusion

complex. The antibody binds to antigens that are preferentially expressed on the

surface of tumor cells, or in the tumor interstitium. In the first step, the

active enzyme

prodrug

active drugantibody

antigen

active drug

prodrug

prodrugactiveenzyme

enzyme cDNA

enzyme mRNAtranscription

enzymetranslation

posttranslationalmodification

??

enzyme cDNA

viral transduction

physical transduction

VDEPT

GDEPT

ADEPT

CYTOPLASM

active enzyme

prodrug

active drug

prodrug

active drugantibody

antigen

active drug

prodrug

prodrugactiveenzyme

enzyme cDNA

enzyme mRNAtranscription

enzyme cDNA

enzyme mRNAtranscription

enzyme

posttranslationalmodification

??

enzyme cDNA

viral transduction

physical transduction

VDEPT

GDEPT

ADEPT

CYTOPLASM

active enzyme

prodrug

active drug

prodrug

active drugantibody

antigen

active drug

prodrug

prodrugactiveenzyme

enzyme cDNA

enzyme mRNAtranscription

enzyme cDNA

enzyme mRNAtranscription

enzymetranslation

posttranslationalmodification

??

enzyme cDNA

viral transduction

physical transduction

VDEPT

GDEPT

ADEPT

CYTOPLASM

active enzyme

prodrug

active drug

prodrug

active drugantibody

antigen

active drug

prodrug

prodrugactiveenzyme

enzyme cDNA

enzyme mRNAtranscription

enzyme cDNA

enzyme mRNAtranscription

enzyme

posttranslationalmodification

??

enzyme cDNA

viral transduction

physical transduction

VDEPT

GDEPT

ADEPT

CYTOPLASM

Introduction

7

antibody-enzyme conjugate is administered and accumulates in tumor tissue.

After clearance of non bound antibody-enzyme conjugates, a non toxic prodrug is

injected in the second step. This non-toxic prodrug can then be converted to a

cytotoxic drug by the antibody-enzyme complex which is located in the tumor.

There are two important features of this system. First, one molecule of antibody-

enzyme conjugate is able to catalyse the conversion of many molecules of

prodrug which enables higher drug concentrations at the tumor compared to

direct injection of the drug. Secondly, tumor cells which do not express the

antigen that is targeted by the antibody-enzyme complex are killed due to the

bystander effect.

There are some specific requirements for enzymes that are used in ADEPT. It

is important that they exert catalytic properties that are different from any

endogenous enzyme to prevent activation of prodrugs at other sites in the body.

They should also catalyse a scission reaction and be active and stable under

physiological conditions. Finally, they should affect high catalytic turnover.

The enzymes used in ADEPT can be characterised in three categories: (i)

enzymes of non-mammalian origin and with no mammalian homologue. These

enzymes have great potential to be used in ADEPT since activation of prodrugs

by endogenous enzymes in the blood and healthy tissue is avoided. Enzymes of

bacterial origin in particular are of interest since they are readily available on a

large scale due to their lack of posttranslational modifications. Their main

disadvantage is their potential to elicit an immune response. This is very

problematic since circulating host anti-conjugate antibodies may interfere with

successive treatment. Enzymes in this category include carboxypeptidase G2

(CPG2), cytosine deaminase (CD), β-lactamase (β-L), penicillin G amidase (PGA),

penicillin V amidase (PVA), etc. (ii) Enzymes of non-mammalian origin with a

mammalian homologue. It is very important that the mammalian homologue of

the enzyme used is only present at low concentrations in the blood or healthy

tissue. Furthermore the mammalian homologue should have different catalytic

properties than the enzyme used to activate the prodrug. This includes different

turnover rates, different optimal conditions such as pH or different structural

requirements for substrates. Enzymes in this category include E. coli β-

glucuronidase (β-G) which has an optimal pH of 6.8 compared to 5.3 for the

mammalian homologue and a higher turnover rate and E. coli nitroreductase that

Chapter 1

8

exhibits different structural requirements for substrates than the human

homologue DT-diaphorase. (iii) Enzymes of mammalian origin. Their main

advantage is the reduction to elicit a host immune response. Unfortunately use in

ADEPT is limited due to the fact that their presence in humans is likely to elicit

prodrug activation in the blood and healthy tissue. Examples include alkaline

phosphatase (AP), α–galactosidase (α-G) and carboxypeptidase A (CPA). In case

of CPA, the risk of conversion of the prodrug by the human enzyme was

circumvented by site specific mutation of the enzyme. This mutant T268G CPA is

able to convert bulky analogues such as 3-cyclopentyltyrosine methotrexate

whereas the wild-type enzyme is not (Smith et al., 1997). Table 1 gives an

overview of selected examples of ADEPT in enzyme-prodrug cancer therapy.

The antibodies (Abs) that bind to tumor-associated antigens are a key

component in ADEPT since they ensure the specific activation of prodrugs in

tumor tissue. There are some specific requirements for Abs used in ADEPT. For

instance, they should bind to tumor cells with high affinity but exert minimal

binding to normal tissue. Furthermore, covalent binding to the enzyme must not

impair the ability of the Abs to bind the associated antigen, nor should it alter the

enzyme activity. Ideally, the non-bound Ab-enzyme conjugate should be rapidly

cleared from the blood. This can be achieved by using a secondary antibody that

binds the Ab-enzyme conjugate. This was demonstrated by Sharma et al

(Sharma et al., 1994). Their study was performed with a monoclonal anti-

carcinoembryonic antigen antibody fragment A5B7-F(ab’)2 conjugated to the

bacterial enzyme, carboxypeptidase G2 (CPG2), in LS174T xenografted mice. A

monoclonal antibody (SB43), directed at CPG2, was used, which inactivates

CPG2 in vitro and in vivo. SB43 was galactosylated so that it had sufficient time

to form a complex with plasma CPG2, resulting in the inactivation and clearance

of the complex from plasma via the carbohydrate-specific receptors in the liver.

Injection of SB43gal 19 hours after administration of the conjugate significantly

reduced the amount of conjugate present in the blood without affecting

conjugate levels in the tumor. They also used a different approach in which the

conjugate was galactosylated so that it is rapidly cleared from the blood by the

asialoglycoprotein receptors in the liver. Localization of the conjugate was

achieved by blocking this receptor for about 8 hours with a single injection of an

inhibitor that binds competitively to the receptor. This allowed tumor localization

Introduction

9

of the conjugate followed by a rapid clearance of the galactosylated conjugate

from the blood as the inhibitor was consumed. The conjugate had a tumor to

blood ratio of 45:1 after 24 hours, which increased to 100:1 at 72 hours after the

conjugate injection.

Table 1: Selected examples of ADEPT in enzyme-prodrug therapy

Enzyme Antibody Prodrug Ref

β-glucosidase HMFG1 Amygdalin (Syrigos et al., 1998)

human β-glucosidase

humanised CEA specific binding region

Prodrugs of anthracyclines

(Florent et al., 1998)

human β-glucosidase

anti-CD20 1H4 Doxorubicin (Haisma et al., 1998)

human β-glucosidase

anti-CEA BW431 Doxorubicin (Bosslet et al., 1992)

carboxypeptidase G2

anti-CEA A5B7 CMDA (Stribbling et al., 1997)

carboxypeptidase G2

W14

p-N-bis (2-Chloroethyl)benzoyl glutamic acid

(Bagshawe et al., 1988; Springer et al., 1991)

alkaline phosphatase

anti-tumor associated carbohydrate L6

Etoposide phosphate (Senter et al., 1989)

Penicillin amidase anti-tumor associated carbohydrate L6

Doxorubicin

(Kerr et al., 1990; Vrudhula et al., 1993)

β-lactamase anti-CEA CEM2314 Desacetylvinblastine hydrazide

(Meyer et al., 1993)

β-lactamase cab-CEA5 nanobody 7-(4-carboxybutanamido) cephalosporin

(Cortez-Retamozo et al., 2004)

Another important factor in ADEPT is the ability of the Ab-enzyme conjugate to

penetrate the tumor. It is well known that tumor vasculature is leaky which

enables macromolecules to extravasate to the tumor interstitium. Furthermore,

the lymphatic drainage system is impaired so that macromolecules are retained

in the interstitium for a prolonged time. This is called the enhanced permeability

and retention effect (EPR) (Figure 4) (Jain, 1987). Due to this EPR, it is relatively

Chapter 1

10

easy for Ab-enzyme conjugates to enter the tumor interstitium. However, further

penetration of the tumor is much more difficult for macromolecules. This leads to

an inadequate distribution of Ab-enzyme conjugates in the tumor tissue (Jain,

1990). To overcome this limitation, Ab fragments, which include F(ab’)2, F(ab’),

scFv and nanobodies, rather than intact antibodies have been used (Figure 2).

These Ab fragments show an increased interstitial rate of transport and

additionally a more rapid clearance from the blood. Using shorter Ab fragments

also decreases the chance of eliciting a host immune response towards the Ab-

enzyme conjugate. A further decreased immunogenicity can be achieved by

combining the regions of the murine antibody that are responsible for the

antigen recognition with human antibody fragments. Such antibodies can be

“chimerized” (murine variable region and human constant region; antibodies

called –iximab) or “humanized” (additional replacement of the murine framework

regions within the complementarity determining regions by human residues,

antibodies called –umab) (Vaughan et al., 1998). Initial comparisons between

murine and chimeric monoclonal antibodies were very promising. For example

use of the chimeric MAb 17-1A in the treatment of colorectal adenocarcinoma

showed improved pharmacokinetics together with a significant reduction in

immunogenicity (LoBuglio et al., 1989). However, not all chimeric MAb’s are

successful in reducing immunogenicity. Nevertheless, many of them as well as

humanised MAb’s have progressed through clinical trials (Reichert et al., 2005).

Figure 2: schematic representation of different antibody fragments used in ADEPT

VL

CH3

CH1

CH2

VH

CL

CH1CL VH

VL

CH1VH

CL

VL

F(ab’) 50 kDa F(ab’)2 100kDa

Monoclonal antibody

CH3

VHH

VHH

CH2

Single chain camel antibody

Nanobody 15 kDa

VLVH

scFv 25 kDa

VL

CH3

CH1

CH2

VH

CL

CH3

CH1

CH2

VH

CL

CH1CL VH

VL

CH1CL VH

VL

CH1VH

CL

VL

F(ab’) 50 kDa F(ab’)2 100kDa

Monoclonal antibody

CH3

VHH

VHH

CH2

Single chain camel antibody

Nanobody 15 kDa

VLVH

scFv 25 kDa

Introduction

11

Since many ADEPT strategies make use of a bacterial enzyme, there is a high

risk that a host immune response towards this bacterial enzyme is elicited. A way

of overcoming this is to use so called abzymes. Abzymes are antibodies raised

against the transition intermediate of an enzyme substrate. Given the strong

binding affinity towards this transition state, they are able to catalyse the

conversion from substrate to product. Since it is an antibody, it can be

chimerized or humanised to minimize immunogenicity. In antibody-directed

abzyme prodrug therapy (ADAPT) a bispecific antibody is used were one arm is

the catalytic antibody and the other arm is the targeting antibody (Kakinuma et

al., 2002).

In conclusion, although ADEPT offers much advantages compared to

conventional therapy, there are many clinical limitations. In poorly vascularised

tumors, delivery of large conjugates is restricted and it is impossible to deliver

conjugates to all tumor cells. Because the enzyme level is low, it is difficult to

generate adequate levels of active drug to reach lethal doses. Furthermore, the

binding of conjugates is limited by antigen heterogeneity. The biggest problem

with ADEPT however, remains the immunogenicity of the antibody-enzyme

conjugate. Although there are many ways of overcoming this immunogenicity,

development is very costly and difficult.

1.2.2. GDEPT

GDEPT, or suicide gene therapy, is also a two step approach. In the first step

the gene encoding a prodrug activating enzyme is directed to tumor tissue. In a

second step the prodrug is administered and subsequently activated by the

prodrug activating enzymes present in tumor cells. Success or failure of GDEPT

strategies not only rely on choosing the right enzyme-prodrug combination but

also on the efficient gene transfection and sustained expression of the genes in

tumor tissue. Therefore the design of appropriate delivery vectors remains the

biggest challenge in suicide gene therapy. In the past two decades, numerous

viral and non-viral methods for transduction were developed.

Since viral vectors naturally evolved to efficiently transfect host cells and

posses the appropriate molecular mechanisms for gene delivery, they are widely

used for suicide gene therapy. In these vectors, the genes necessary for the

replication phase of the virus are replaced by the gene encoding the prodrug

Chapter 1

12

activating enzyme. In this manner, non replicative viruses are made that can

infect target cells and introduce genes either by integrating them into the

genomic DNA of the target cell or by residing on a plasmid. Since viruses have

evolved as parasites, they elicit an immune response that reduces their clinical

use. Despite this, several viral vectors have been engineered for suicide gene

therapy. Interest has centred on four types: adenoviruses, retroviruses

(including lentiviruses), adeno-associated viruses and herpes simplex virus type

1. One major challenge of viral vectors is the efficient targeting of the vectors to

the cells of interest: to achieve successful gene therapy, the appropriate genes

must be delivered to and expressed in target cells, without harming non-target

cells.

One way to obtain tumor specific expression of the prodrug activating enzyme

is by transcriptional targeting. In this approach tumor-specific enhancer-

promoters are used thus allowing transcription of the gene in tumor cells only.

Many tumor-specific enhancer-promoter sequences were discovered and used to

target adenoviruses to various tissues, for a summery see Table 2. These

adenoviral vectors carrying suicide genes controlled by tumor specific regulatory

elements demonstrated both targeting and efficacy. However, transcriptional

targeting does not prevent transfection of healthy tissue and toxic effects related

to this dislocation of viral vectors.

Another way to achieve tumor specific expression of the suicide gene is by

direct targeting of the viral vectors to the surface of tumor cells by using the

native tropism (host range) of the viral vector. However, this native tropism

often does not meet the therapeutic need. Native tropism may not be able to

specifically transfect tumor tissue and therefore needs to be diminished to avoid

toxic side effects. Therefore many different mechanisms were developed to

approve the targeting of viral vectors to specific tissues, including pseudotyping,

adaptor systems and genetic approaches. For a complete overview see Table 3.

Introduction

13

Table 2: selected examples of transcriptional targeting to various tumors

Enhancer-promotor Target tissue Ref

hAFP (human α-fetoprotein) Hepatocellular carcinoma (Kaneko et al., 1995; Li et al., 2001)

hALA (human α-lactalbumin) Mammary cells (breast cancer)

(Anderson et al., 2000)

hCEA (human carcinoembryonicantigen)

Colorectal carcinoma (Zhang et al., 2003)

Cox-2 (cyclooxygenase-2) Gastrointestinal cancer (Yamamoto et al., 2001)

GRP (gastrin releasing peptide) Lung cancer (Morimoto et al., 2001)

L-plastin Ovarian and bladder cancer (Peng et al., 2001)

rPB (rat probasin) Prostate cancer (Andriani et al., 2001; Lu and Steiner, 2000)

PSA (human prostate-specific antigen)

Prostate cancer (Gotoh et al., 1998)

SLPI (secretory leukoprotease inhibitor)

Cervical and ovarian cancer (Barker et al., 2003)

Survivin Suvivin-positive tumor cells (Zhu et al., 2004)

hTERT (human telomerase reverse transcriptase)

Telomerase-positive tumors (Bilsland et al., 2003; Gu et al., 2000; Huang et al., 2004)

Tg (rat thyroglobulin) Thyroid carcinoma (Shimura et al., 2001; Zhang et al., 2001)

Tyrosinase Melanoma (Nettelbeck et al., 2002; Siders et al., 1996)

BLG (ovine β-lactoglobulin) Mammary cells (breast cancer)

(Anderson et al., 2000)

ErbB2 ErbB2-positive tumor cells (Vassaux et al., 1999)

Chapter 1

14

Table 3: Tropism changing strategies for viral targeting

Approach Principle Ref

Adaptor systems

Receptor-ligand A tumor specific ligand is fused with the viral

receptor

(Pereboev et al., 2004; Snitkovsky and Young, 2002)

Bispecific antibodies

Two antibodies that bind the viral vector and the target cell are coupled

(Bartlett et al., 1999; Wurdinger et

al., 2005)

Chemical linkage The targeting molecule is chemically coupled

to the viral particle (Eto et al., 2005)

Avidin-Biotin Coupling biotin to the vector and avidin to

the targeting molecule (Arnold et al., 2006)

Antibody A tumor specific antibody that binds to a

genetically incorporated Ig-binding domain of the vector

(Tai et al., 2003)

Genetic systems

Serotype switching Using a different serotype of the same family (Wu et al., 2006)

Pseudotyping Using a viral attachment protein from a

different strain or family (Cronin et al., 2005)

Targeting motifs Small targeting peptides that are inserted into the capsid or viral attachment protein

(Gollan and Green, 2002; Stachler and

Bartlett, 2006)

Single-chain antibody

A single chain antibody is inserted in the viral attachment protein

(Chowdhury et al., 2004; Hedley et al., 2006; Yang et al.,

1998)

Mosaic viral attachment proteins

Two viral attachment proteins with different properties are combined

(Pereboeva et al., 2004)

In 2007, 70 % of gene therapy clinical trials used viral vectors for gene

delivery. However, there are many drawbacks in using viral vectors for gene

delivery such as their immunogenicity, cytotoxicity as well as the risk of

insertional mutagenesis. Therefore non-viral vectors have important safety

advantages over viral vectors.

Introduction

15

Conventional non-viral vectors include diverse liposomal formulations (DNA

lipoplexes) (Karmali and Chaudhuri, 2007; Liu et al., 2003), cationic peptides

(Fabre and Collins, 2006) and polymers (DNA polyplexes) such as

polyetylenimine (Boussif et al., 1995; Lungwitz et al., 2005). They interact with

DNA to facilitate cell entry by binding or enveloping DNA through a charge

interaction. These vectors however have poor in vivo transfection efficiency and

only confer transient gene expression. This is partially due to the ability of the

non-viral vector–DNA complex to interact with blood plasma proteins,

undesirable cells and the extracellular matrix. Once inside the target cell,

additional requirements for transfection include the need for the DNA to escape

the liposome or endosome, resistance to cytoplasmic degradative enzymes such

as nucleases, and passage through the double-membrane structure of the

nuclear envelope. Finally, but importantly, plasmid DNA that is delivered to the

nucleus is generally not replicated and is lost when the nuclear envelope is

degraded during mitosis.

To overcome these problems, novel non-viral vectors with increased in vivo

stability have been developed, which have a reduced affinity for extracellular

proteins and cell surfaces, enabling them to reach target cells (Kichler, 2004;

Knorr et al., 2007). Furthermore, the inclusion of ligands for receptor-mediated

endocytosis (Kircheis et al., 2001; Kursa et al., 2003), endosomal disruption

sequences (Cho et al., 2003; Funhoff et al., 2005) and nuclear-import signals

(Bremner et al., 2004; Zanta et al., 1999) have improved the passage of non-

viral vectors into the cell, and into its nucleus. In combination, these modular

non-viral vectors mimic the ability of viruses to overcome the cellular barriers to

DNA delivery through mechanisms that are analogous to those of viral vectors

(Uherek et al., 1998).

The enzyme-prodrug combinations used in GDEPT are similar to those used in

ADEPT. The most frequently used and best characterised prodrug activating

enzymes include thymidine kinase from herpes simplex virus type 1, cytosine

deaminase, cytochrome P450 reductase, nitroreductase, carboxypeptidase G2

and purine nucleoside phosphorylase. For an overview see Table 4.

Taken together, suicide gene therapy shows great potential to improve

chemotherapy by decreasing unwanted cytotoxic side effects. However, both

viral and non-viral approaches suffer from several drawbacks such as inefficient

Chapter 1

16

gene transfection and prolonged gene expression, pathogenicity and

immunogenicity.

Table 4: Selected examples of suicide gene therapy systems

Enzyme Prodrug Vector Ref

herpes simplex virus 1 Thymidine kinase

ganciclovir Adenovirus Ad-CMV-

UTk, Polyethylenimine

(Mathis et al., 2006) (Iwai et

al., 2002)

yeast Cytosine deaminase

5-fluorocytosine Adenovirus Ad-CMV-CD (Zeng et al., 2007)

Cytochrome P 2B1 Cyclophosphamide Adenovirus Ad-CYP2B1 (Huch et al., 2006)

E. coli Nitroreductase

Nitrocompound CB1954

Adenovirus Ad-hTR-NTR (Bilsland et al., 2003; Plumb et

al., 2001) E. coli Purine nucleoside phosphorylase

6-methylpurine 2’-deoxyriboside

Plasmid phTERT-ePNP, Adenovirus Ad2-Tyr2-

PNP

(Zhou et al., 2007) (McCart et al., 2002)

Carboxypeptidase G2

Mustard compound ZD2767P

Adenovirus AdV-hTERT-CPG2

(Schepelmann et al., 2007)

Human deoxycytidine kinase

Cytosine arabinoside Adenovirus Ad-hdCk, Retrovirus Rv-hdCK

(Manome et al., 1996)

D. melanogaster Deoxyribonucleoside kinase

Cytosine arabinoside, 5-fluorocytosine

lipoplexes (Kamiya et al., 2006)

β-glucuronidase glucuronide prodrug of

doxorubicin (DOX-GA3)

poly(2-(dimethylamino)ethyl

methacrylate) polyplexes

(Fonseca et al., 1999)

1.3 Liposomes as drug delivery vehicles in cancer therapy

Liposomes were suggested as drug carriers in cancer chemotherapy by

Gregoriadis et al. in 1974 (Gregoriadis et al., 1974). Since then, the interest in

liposomes has increased and liposome systems are now being extensively

studied as drug carriers. Moreover, liposomes are the most advanced of the

particulate drug carriers and are now considered to be a mainstream drug

delivery technology with breakthrough developments resulting in the approval of

Introduction

17

several liposomal drugs, such as Doxyl® (Ortho Biotech), DaunoXome® (Gilead

Sciences, Inc.), caelyx® (Schering-Plough, Inc.), ...

In general, hydrophilic drugs can be entrapped in the aqueous interior and

hydrophobic drugs can be incorporated in the bilayer (Figure 3). Amphiphilic

drugs that are weak bases or weak acids can also be loaded into the liposome

interior using remote loading methods. Three basic requirements need to be

fulfilled if liposomes are to be successful in delivering drugs specifically to

cancerous tissue: (i) prolonged blood circulation, (ii) sufficient tumor

accumulation, (iii) controlled drug release and uptake by tumor cells with a

release profile matching the pharmacodynamics of the drug.

Initially, the research in liposome drug delivery systems suffered from the

very fast blood clearance, due to adsorption of plasma proteins (opsonins) to the

‘naked’ phospholipid membrane and complement activation. Subsequently

triggering recognition and uptake of the liposomes predominantly by Kuppfer

cells of the reticuloendothelial system (RES) (Ishida et al., 2002). Liposomes

typically have a half-life of approximately 0,6 hours resulting in complete

removal from the bloodstream within several hours (Blume and Cevc, 1993). A

major advance in the field of liposomes came with the development of long-

circulating liposomes or Stealth® liposomes.

1.3.1. Long-circulating liposomes

Different methods have been suggested to achieve long circulation of

liposomes in vivo, including coating the liposome surface with inert,

biocompatible polymers, such as poly(ethylene glycol) (PEG). These PEG chains

form a protective layer over the liposome surface and prevent liposome

interactions with opsonins and subsequent clearance of liposomes (Klibanov et

al., 1990) (Figure 3). An important feature of protective polymers is their

flexibility, which allows a relatively small number of surface-grafted polymer

molecules to create an impermeable layer over the liposome surface (Torchilin et

al., 1994). Long-circulating liposomes demonstrate dose-independent, non-

saturable, log-linear kinetics and increased bioavailability (Allen and Hansen,

1991). Although, PEG remains the standard for the steric protection of

liposomes, other polymers also posses stealth-like properties such as methyl and

ethyl polyoxazolines (Woodle et al., 1994; Zalipsky et al., 1996), poly-N-

Chapter 1

18

vinylpyrrolidones (Torchilin et al., 2001), and polyvinyl alcohols (Takeuchi et al.,

2001). Long-circulating liposomes, or stealth liposomes, are now being

investigated in detail and are widely used in biomedical in vitro and in vivo

studies and have found their way into clinical practice (Gabizon, 2001).

More recently, research has focussed on attaching PEG in a removable fashion

to facilitate liposome capture by cells. For instance, in suicide gene therapy

uptake of the liposomal vector is required for therapeutic activity. After PEG-

liposomes accumulate at the target site, through the enhanced permeability and

retention (EPR) effect (Jain, 1987), the PEG coating is detached locally by

proteolysis (Hatakeyama et al., 2007) or mild thiolysis (Zalipsky et al., 1999) at

the tumorsite allowing facilitated endocytosis of the lipoplexes

Figure 3: Schematic representation of a long circulating liposome

1.3.2. Targeting liposomes to tumor tissue

Both conventional and stealth liposomes are able to accumulate in tumor

tissue due to the enhanced permeation and retention effect (EPR). Leaky tumor

vasculature in combination with an impaired lymphatic drainage, allows

macromolecules ranging from 10 to 500nm to infiltrate solid tumor tissue (Figure

4) (Jain, 1987; Matsumura and Maeda, 1986). This passive targeting results in

Introduction

19

several-fold increased drug concentrations in solid tumors relative to those

obtained with non-macromolecular-complexed free drugs (Northfelt et al., 1996).

Figure 4: Schematic overview of the enhanced permeation and retention effect (EPR). Vascular endothelial cells are depicted in light blue, macromolecules in dark blue.

In attempts to further increase the bioavailability of liposomal drugs at the

target site, recent efforts in the liposome field have focused on the active

targeting of liposomes to specific tissues by coupling ligands to their surface.

Targeting moieties can essentially be any molecule that selectively recognizes

and binds to target antigens or receptors over-expressed or selectively expressed

on cancer cells and include antibody molecules, or fragments thereof, peptides,

carbohydrates, glycoproteins or receptor ligands. To date antibodies or antibody

fragments that bind tumor associated antigens, folate -that binds to the folate

receptor induced on the surface of actively growing cancer cells- or transferrin –

that binds to the integrin receptor expressed on the surface of endothelial cells of

the neo-vasculature of growing tumors - are the most extensively researched

ligands for targeting liposomes to tumor tissue.

Liposomes can be coated with antibodies or antibody fragments either by

directly attaching the antibody to the liposome phospholipid head group or to the

distal end of the PEG polymer. The latter approach has proven most successful,

due to better accessibility of the antibody towards its target (Maruyama et al.,

1999; Sapra and Allen, 2003). However, the attachment of antibodies directly to

the liposome surface is also a viable method (Maruyama et al., 1995) since

shielding of the bound antibody by the PEG chains decreases the antibody-

Chapter 1

20

mediated clearance and immunogenicity of the liposomes. The coating of

liposomes with antibodies directed against tumor-associated targets consists of a

fine balance between coating with a sufficient amount of antibodies to achieve

target binding and tumor retention on one side and enhanced RES clearance with

an increased number of antibodies per liposome on the other (Maruyama et al.,

1999). An optimal coating ratio of 10–30 antibody molecules per liposome was

shown to give the most efficient delivery of drugs to tumors with limited increase

in RES uptake (Maruyama et al., 1995; Sapra and Allen, 2003).

Immunoliposomes can be targeted to surface molecules expressed either in the

vascular system or in the extravascular system on tumor cell membranes. The

most readily accessible target sites for immunoliposomes are the vascular

endothelial surface of growing tumors and circulating cells related to the immune

system.

1.3.2.1. Coupling mechanisms

In general, covalently linking ligands either to the surface of liposomes or the

distal end of PEG chains is based on three main reaction methods, which are

quite efficient and selective: reaction between activated carboxyl groups and

amino groups, which yields an amide bond; reaction between pyridyldithiols and

thiols, which yields disulphide bonds; and reaction between maleimide

derivatives and thiols, which yields thioether bonds.

Coupling of ligands directly to the surface of liposomes can be achieved via

linker lipids such as N-glutaryl-phosphatidylethanolamine (NGPE), N-(4'-(4"-

maleimidophenyl)butyroyl)dioleoylphosphatidylethanolamine (MPB-DOPE) or N-

(3'-(pyridyldithio)propionoyldioleoylphosphatidylethanolamine (PDP-DOPE) that

are incorporated in the liposome membrane. NGPE is activated by 1-ethyl-3(3-

dimethylaminopropyl)carbodiimide (EDC) and N-hydroxy-sulphosuccinimide

(sulpho-NHS) to interact with an amino group of the ligand to yield an amide

bond. MPB-DOPE and PDP-DOPE can be linked to thiol-containing ligands forming

a thioether bond or disulphide bond, respectively. In addition PDP-DOPE can be

linked to ligands that are activated by succinimidyl-4-(p-

maleimidophenyl)butyrate (SMPB) by means of a thioether bond.

For coupling ligands to the distal end of the PEG chain, several commercially

available PEG derivatives can be used such as maleimide-PEG (Mal-PEG), PDP-

Introduction

21

PEG, hydrazide-PEG (Hz-PEG) and p-nitrophenylcarbonyl-PEG (pNP-PEG). Mal-

PEG and PDP-PEG coupling is identical to MPB-DOPE and PDP-DOPE. Hz-PEG

reacts with oxidized carbohydrates on the ligand to form a hydrazon bond

whereas pNP-PEG binds aminogroups via a carbamate bond.

Non covalent linking of ligands to the liposome surface or distal end of the PEG

chain is also possible via the biotin-avidin coupling method. Avidin that has four

binding sites for biotin, serves as a crosslinker between biotinylated ligand and

biotinylated lipids or biotinylated PEG chains.

1.3.3. Stimuli-sensitive liposomes

Triggered release of liposome encapsulated drugs at the site of the tumor can

enhance tumor specific drug delivery. In conventional liposomes, the release of

encapsulated drug occurs gradually due to leaking of the substance or

degradation of the liposome. To date many stimuli-sensitive liposomes are being

developed that release their substance in one single burst as a result of

destabilization of the liposome membrane caused by a change in the

environment.

1.3.3.1. pH-sensitive liposomes

It is well known that the extracellular environment of solid tumors is acidic

with a pH ranging from 6.5 to 7.2 compared to the pH of the blood and normal

tissue (7.5) (Stubbs et al., 2000; Tannock and Rotin, 1989). However, the most

acidic regions in solid tumors are located far from the tumor vasculature thus

making it difficult for liposomes to reach these acidic regions (Helmlinger et al.,

1997). In addition, it is technically very difficult to design pH responsive

liposomes that are stable in the blood (pH 7.5) and instable at pH 6.5 due to the

small change in acidity. Therefore using the acidic tumor microenvironment for

triggered release has not been very successful. A more viable strategy is the

exploitation of the very acidic environment in endosomes and lysosomes where

the pH is lower than 5.0. When it is desired to deliver the drugs to the

cytoplasm, liposomes can be non-specifically or specifically (receptor mediated)

internalized to endosomes by endocytosis and subsequently delivered to

lysosomes. In the lysosome, both liposomes and encapsulated drug can be

Chapter 1

22

degraded by metabolic enzymes. Therefore it is important that liposomes can

escape the endosome after internalization. Intensive research has focussed on

acid triggered drug delivery using fusogenic liposomes. After cell internalization,

the drop in pH in the endosome triggers a phase transition of the lipid bilayer,

resulting in the fusion of the liposome membrane with the endosomal membrane

(Hafez and Cullis, 2001) (Figure 5).

Figure 5: Destabilization of the endosomal membrane: After being endocytosed by the cell and taken inside the endosome, the liposome containing stimuli (pH)-sensitive components, such as lipids (a) in the membrane can undergo pH-dependent membrane destabilization and initiate the destabilization of the endosomal membrane allowing drug (b) efflux into the cell cytoplasm.

A main strategy has been to stabilize the non-bilayer forming lipid

diacylphosphatidylethanolamine (DOPE) with micellar forming lipids such as PEG-

coated lipids. Mixtures of these lipids can form stable liposomes. After acid

catalyzed cleavage of the PEG chain from the PEG-lipids in the liposomes, the

liposomes become fusogenic and fuse with the endosome membrane leading to

drug release into the cytoplasm of the cell (Gerasimov et al., 1999; Guo and

Szoka, 2001; Ishida et al., 2001). Mildly acidic amphiphiles such as

diacylsuccinylglycerols, cholesterol hemisuccinate and oleic acids can also be

used to destabilize the liposome membrane and promote fusion by protonation at

acidic pH (Chu et al., 1990; Collins et al., 1990; Duzgunes et al., 1985)

1.3.3.2. Thermosensitive liposomes

Since it is possible to induce local hyperthermia (LHT) by external heating via

high frequency waves or internal heating, research has also focussed on

producing heat sensitive liposomes. The idea originated from the understanding

Nature Reviews Drug Discovery 4, 145-160 (2005)

Introduction

23

that liposomes become highly leaky to water soluble contents near the gel-to-

liquid crystalline phase transition temperature of its membrane (Blok et al.,

1975). Yatvin and co-workers introduced this concept in 1978. They used

dipalmitoylphosphatidylcholine (DPPC) liposomes with a phase transition

temperature of 41°C and added small amounts of distearoylphosphatidylcholine

(DSPC) as a co-lipid to adjust its transition temperature to 42°C. They showed

that these liposomes delivered more than four times as much methotrexate to

murine tumors heated to 42°C as to unheated control tumors (Weinstein et al.,

1979; Yatvin et al., 1978). Several other formulations based on this composition

have been designed including sterically stabilized liposomes (Maruyama et al.,

1993; Merlin, 1991). For instance Gaber et al. designed long circulating

thermosensitive liposomes made of a mixture of DPPC/hydrogenated soy

phosphatidylcholine (HSPC)/cholesterol (CHOL)/DSPC-PEG lipids that released

more than 60% of their contents when heated at 42 °C for 0.5 h in vitro (Gaber

et al., 1995). More recently, thermosensitive polymers have been used to

produce heat-sensitive liposomes. These polymers become water insoluble above

critical temperature called the lower critical solution temperature (LCST). At the

molecular level this means that chains of these polymers undergo a coil-to-

globule transition above the LCST (Figure 6). When such thermosensitive

polymers are fixed on liposome membranes, the temperature-dependent change

of their solubility can destabilize liposomes resulting in the release of their

content (Kono, 2001). Among thermosensitive polymers, poly(N-

isopropylacrylamide) (poly(NIPAM)) is most frequently used. This polymer is

known to exhibit a drastic change in water-solubility in a very narrow

temperature region near 32°C (Schild, 1992). In addition, its LCST can be

adjusted to a desired temperature by copolymerization with co-monomers with

varying hydrophilicity or hydrophobicity. In general, incorporation of hydrophobic

co-monomers decreases the LCST whereas hydrophilic co-monomers lead to an

increase in LCST (Feil et al., 1993). Han et al. were able to produce

thermosensitive liposomes by modification of the surface of

DPPC/HSPC/CHOL/DSPC-PEG liposomes with polyNIPAM/acrylamide copolymers.

By varying the acrylamide content they obtained copolymers with LCST’s ranging

from 33°C to 47°C. These copolymer modified liposomes showed increased burst

release of encapsulated doxorubicin around the LCST of the

polyNIPAM/acrylamide copolymer (Han et al., 2006).

Chapter 1

24

The use of thermosensitive liposomes has a significant advantage over other

triggering concepts because hyperthermia increases the tumor blood flow and

microvascular permeability (Kong and Dewhirst, 1999; Song et al., 1984). This

results in higher liposome tumor accumulation. Unfortunately, thermosensitive

liposomes can not be used to treat metastatic tumors since the location of the

tumor must be known and the tumor site must be accessible to local

hyperthermia.

Figure 6: Effect of temperature on the solubility of thermosensitive polymers grafted on the liposomal surface

1.3.3.3. photosensitive liposomes

Liposomes can be made photosensitive by using lipids that undergo various

transformations such as isomerisation, fragmentation or polymerization upon

photoexcitation, hereby destabilizing the liposomal membrane.

Firstly, photoisomerizable lipids were exploited by Bisby et al. to allow UV-

triggered release from liposomes (Bisby et al., 2000a). For this purpose they

used the lipid azobenzene-glycero-phosphocholine. Trans-cis isomerisation upon

UV irradiation resulted in destabilization of the liposome membrane and fast

release of encapsulated doxorubicin (Figure 7). However, the use of UV-light is

not very suitable for biological applications due to the potential damage to

healthy tissue. Bisby and coworkers also discovered that incorporation of

cholesterol (up to 25 mol%) in the liposomal membrane induces fast release of

encapsulated drug upon illumination at 470 nm (Bisby et al., 2000b). These

cholesterol containing liposomes therefore have greater therapeutic potential.

Adv Drug Deliv Rev 53 (2001) 307– 319

Introduction

25

Figure 7: Trans-cis isomerization upon UV irradiation of Bis-Azo PC

Secondly, sensitized photo-oxidation has proven to be another viable method

for light triggered release. Thompson et al. used plasmenylcholine that is cleaved

into single chain surfactants via sensitized photo-oxidation of the plasmalogen

vinyl ether linkage (Thompson et al., 1996) (Figure 8). This results in

destabilization of the liposome membrane and an increased permeability.

Zinc phthalocyanine, tin octabutoxy phthalocyanine and bacteriochlorophyll-a

were investigated as sensitizing agents. Bacteriochlorophyll-a produced the

fastest release and absorbs at a wavelength of 820 nm which allows tissue

penetration of ≥ 0,8 cm.

Biochem Biophys Res Commun 276(1), 169–173 (2000)

Chapter 1

26

Figure 8: Singlet oxygen-mediated photo-oxidation of plasmalogen vinyl ether linkage.

Finally, drug release from liposomes by photopolymerization was addressed by

Bondurant et al. who reported a PEG-liposome formulation containing 1,2-bis[10-

(2′,4′-hexadienoyloxy)-decanoyl]-sn-glycero-3-phosphocholine (bis-SorbPC), a

photosensitive lipid that forms a cross-linked lipid network upon exposure to UV-

light (Bondurant et al., 2001). Leaking of encapsulated drugs occurs during the

polymerization process due to the formation of defects in the bilayer (Spratt et

al., 2003) (Figure 9). Again, the use of UV-light is not very suitable for biological

applications. The incorporation of a cyanine dye into the PEG-liposomes made

them also sensitive to visible light, thus increasing the therapeutic potential

(Mueller et al., 2000).

Biochim Biophys Acta 1279 25-34(1996)

Introduction

27

Figure 9: Photopolymerization triggered release via Bis-SorPC. The photopolymerization-induced reduction in the surface area of the polymerizable domains during UV irradiation is shown on the right.

1.3.4. Conclusion

Since the 1970s, when it was first suggested that liposomes could be used as

drug carriers in the treatment of cancer, a significant amount of research has

been performed to optimize and utilize the liposomal carriers successfully in the

treatment of various diseases. Cancer has especially been a disease where

considerable efforts have been made to use liposomal drug delivery systems to

gain increased efficacy and limited toxicity of various chemotherapeutics. A few

drug carriers have appeared on the market, however, the clinical success with

respect to efficacy in cancer therapy, when compared with the free drug, has

been limited even though improved toxicity profiles are found and many

promising preclinical experiments have been reported. Although targeting

liposomal formulations to tumor tissue increases the accumulation in tumor

tissue, this thus not necessarily result in improvement of therapeutic efficacy

(Andresen et al., 2005). This is mainly caused by the obstruction of liposome

extravasation after the first liposomes have bound to target cells lining the blood

vessels. Another obstacle relates to the fact that immunoliposomes in general

Biochim Biophys Acta 1611, 35–43 (2003)

Chapter 1

28

show enhanced liposome clearance (Andresen et al., 2005). Furthermore, it is

important to notice that doxorubicin is often the drug of choice when developing

active release strategies. This drug is able to diffuse over an intact liposome

membrane, caused by the destruction of the pH gradient used to load the

liposomes with doxorubicin. Therefore it is not the best drug to prove an active

release concept.

It is clear that the use of liposomal formulations in cancer therapy holds great

promise but many hurdles still have to be overcome before resulting in clinical

success.

1.4 Polymer vesicles

Nowadays, many different delivery vehicles are being investigated but the best

known examples are lipid vesicles or liposomes that are made of closed lipid

bilayers. Although liposomes were originally used to study biological membranes,

they were introduced in the 1970’s as drug delivery vehicles. Owing to their low

molecular weight (MW < 1 kDa), aggregation of lipids results in molecularly thin

membranes that posses a dynamical, physical softness. As a consequence, many

lipid vesicle properties such as encapsulant retention, membrane stability, and

degradation are not particularly well controlled. In order to obtain more robust

membranes with controllable properties, extensive efforts were made within the

last decade to design polymeric vesicles. This has led to a wide range of

container systems made of diblock copolymers, triblock copolymers or highly

branched polymeric dendrimers. Block copolymers have a similar basic structure

as lipids but consist of distinct polymer chains covalently linked in a series of two

or more segments. Amphiphilic block copolymers are composed of at least one

hydrophilic block and one hydrophobic block, causing self-assembly in aqueous

solutions to nanometer-sized suprastructures. In the absence of solvent, block

copolymers can assume a wide variety of ordered morphologies such as the

lamellar phase. The transition from an isotropically disordered state to an

ordered state such as the lamellar phase is controlled by three parameters: (i)

the molecular weight of the copolymer, (ii) the mass or volume fraction ƒ of each

block and (iii) the effective interaction energy ε between monomers in the blocks

(Discher and Eisenberg, 2002). Upon addition of solvent, the lamellar phase can

Introduction

29

swell and rearrange to form rod-like or spherical structures. Typical bilayer

particles are formed, consisting of a core comprised of their insoluble part

surrounded by a corona of their soluble part (Nardin et al., 2001). The driving

forces for the self-organization are the difference in solubility of the blocks and

the constraint imposed by the chemical linkage between the blocks. Depending

on their concentration, molecular shape, hydrophobic-to-hydrophilic balance and

block-length, micelles, vesicles, cylinders or rod-like structures are formed

(Stoenescu et al., 2004). Compared to the self-assembled structures formed by

lower molecular weight amphiphilic molecules such as lipids or surfactants, block

copolymers self-assemble into significantly more stable particles. This higher

stability is due to the larger size of the hydrophobic part and the slower

dynamics of the underlying copolymer molecules caused by a higher

entanglement (Battaglia and Ryan, 2005; Meier, 2000). It is this increased

stability, along with their self-assembled nanometer-sized structures that make

block copolymers so attractive for biomedical applications such as drug delivery

(Nardin et al., 2004).

Numerous polymers have already been used for the hydrophobic block and

include inert PEE (polyethylethylene), PS (polystyrene), PDMS

(polydimethylsiloxane), PBD (polybutadiene) and the degradable PLA (polylactic

acid) and PCL (polycaprolactone). Hydrophilic blocks have been synthesized from

PEG, the negatively charged PAA (polyacrylic acid), crosslinkable PMOXA

(polymethyloxazoline) and the most common PEO (polyethylene oxide) (Discher,

2007).

Chapter 1

30

1.5 References

Abe, T., Koike, K., Ohga, T., Kubo, T., Wada, M., Kohno, K., Mori, T., Hidaka, K. and Kuwano, M. (1995) Chemosensitisation of spontaneous multidrug resistance by a 1,4-dihydropyridine analogue and verapamil in human glioma cell lines overexpressing MRP or MDR1. Br J Cancer, 72, 418-423.

Albanell, J., Codony, J., Rovira, A., Mellado, B. and Gascon, P. (2003) Mechanism of action of anti-HER2 monoclonal antibodies: scientific update on trastuzumab and 2C4. Adv Exp Med Biol, 532, 253-268.

Alderden, R., Hall, M. and Hambley, T. (2006) The Discovery and Development of Cisplatin. journal of chemical education, 83, 673-816.

Allen, T.M. and Hansen, C. (1991) Pharmacokinetics of stealth versus conventional liposomes: effect of dose. Biochim Biophys Acta, 1068, 133-141.

Anderson, L.M., Krotz, S., Weitzman, S.A. and Thimmapaya, B. (2000) Breast cancer-specific expression of the Candida albicans cytosine deaminase gene using a transcriptional targeting approach. Cancer Gene Ther, 7, 845-852.

Andresen, T.L., Jensen, S.S. and Jorgensen, K. (2005) Advanced strategies in liposomal cancer therapy: problems and prospects of active and tumor specific drug release. Prog Lipid Res, 44, 68-97.

Andriani, F., Nan, B., Yu, J., Li, X., Weigel, N.L., McPhaul, M.J., Kasper, S., Kagawa, S., Fang, B., Matusik, R.J., Denner, L. and Marcelli, M. (2001) Use of the probasin promoter ARR2PB to express Bax in androgen receptor-positive prostate cancer cells. J Natl Cancer Inst, 93, 1314-1324.

Arnold, G.S., Sasser, A.K., Stachler, M.D. and Bartlett, J.S. (2006) Metabolic biotinylation provides a unique platform for the purification and targeting of multiple AAV vector serotypes. Mol Ther, 14, 97-106.

Bagshawe, K.D., Springer, C.J., Searle, F., Antoniw, P., Sharma, S.K., Melton, R.G. and Sherwood, R.F. (1988) A cytotoxic agent can be generated selectively at cancer sites. Br J Cancer, 58, 700-703.

Barker, S.D., Coolidge, C.J., Kanerva, A., Hakkarainen, T., Yamamoto, M., Liu, B., Rivera, A.A., Bhoola, S.M., Barnes, M.N., Alvarez, R.D., Curiel, D.T. and Hemminki, A. (2003) The secretory leukoprotease inhibitor (SLPI) promoter for ovarian cancer gene therapy. J Gene Med, 5, 300-310.

Bartlett, J.S., Kleinschmidt, J., Boucher, R.C. and Samulski, R.J. (1999) Targeted adeno-associated virus vector transduction of nonpermissive cells mediated by a bispecific F(ab'gamma)2 antibody. Nat Biotechnol, 17, 181-186.

Battaglia, G. and Ryan, A.J. (2005) Bilayers and interdigitation in block copolymer vesicles. J Am Chem Soc., 127, 8757-8764.

Bilsland, A.E., Anderson, C.J., Fletcher-Monaghan, A.J., McGregor, F., Evans, T.R., Ganly, I., Knox, R.J., Plumb, J.A. and Keith, W.N. (2003) Selective ablation of human cancer cells by telomerase-specific adenoviral suicide gene therapy vectors expressing bacterial nitroreductase. Oncogene, 22, 370-380.

Bisby, R.H., Mead, C. and Morgan, C.G. (2000a) Active uptake of drugs into photosensitive liposomes and rapid release on UV photolysis. Photochem Photobiol, 72, 57-61.

Bisby, R.H., Mead, C. and Morgan, C.G. (2000b) Wavelength-programmed solute release from photosensitive liposomes. Biochem Biophys Res Commun, 276, 169-173.

Blok, M.C., van der Neut-Kok, E.C., van Deenen, L.L. and de Gier, J. (1975) The effect of chain length and lipid phase transitions on the selective permeability properties of liposomes. Biochim Biophys Acta, 406, 187-196.

Blume, G. and Cevc, G. (1993) Molecular mechanism of the lipid vesicle longevity in vivo. Biochim Biophys Acta, 1146, 157-168.

Bondurant, B., Mueller, A. and O'Brian, D. (2001) Photoinitiated destabilization of sterically stabilized liposomes. Biochimica et Biophysica acta, 1511, 113-122.

Introduction

31

Bosslet, K., Czech, J., Lorenz, P., Sedlacek, H.H., Schuermann, M. and Seemann, G. (1992) Molecular and functional characterisation of a fusion protein suited for tumour specific prodrug activation. Br J Cancer, 65, 234-238.

Boussif, O., Lezoualc'h, F., Zanta, M.A., Mergny, M.D., Scherman, D., Demeneix, B. and Behr, J.P. (1995) A versatile vector for gene and oligonucleotide transfer into cells in culture and in vivo: polyethylenimine. Proc Natl Acad Sci U S A, 92, 7297-7301.

Bremner, K.H., Seymour, L.W., Logan, A. and Read, M.L. (2004) Factors influencing the ability of nuclear localization sequence peptides to enhance nonviral gene delivery. Bioconjug Chem, 15, 152-161.

Canetta, R., Rozencweig, M. and Carter, S.K. (1985) Carboplatin: the clinical spectrum to date. Cancer Treat Rev, 12 Suppl A, 125-136.

Cheson, B.D. (1992) New antimetabolites in the treatment of human malignancies. Semin Oncol, 19, 695-706.

Cho, Y.W., Kim, J.D. and Park, K. (2003) Polycation gene delivery systems: escape from endosomes to cytosol. J Pharm Pharmacol, 55, 721-734.

Chowdhury, S., Chester, K.A., Bridgewater, J., Collins, M.K. and Martin, F. (2004) Efficient retroviral vector targeting of carcinoembryonic antigen-positive tumors. Mol Ther, 9, 85-92.

Chu, C.J., Dijkstra, J., Lai, M.Z., Hong, K. and Szoka, F.C. (1990) Efficiency of cytoplasmic delivery by pH-sensitive liposomes to cells in culture. Pharm Res, 7, 824-834.

Collins, D., Litzinger, D.C. and Huang, L. (1990) Structural and functional comparisons of pH-sensitive liposomes composed of phosphatidylethanolamine and three different diacylsuccinylglycerols. Biochim Biophys Acta, 1025, 234-242.

Cortez-Retamozo, V., Backmann, N., Senter, P.D., Wernery, U., De Baetselier, P., Muyldermans, S. and Revets, H. (2004) Efficient cancer therapy with a nanobody-based conjugate. Cancer Res, 64, 2853-2857.

Cronin, J., Zhang, X.Y. and Reiser, J. (2005) Altering the tropism of lentiviral vectors through pseudotyping. Curr Gene Ther, 5, 387-398.

Discher, D.E. (2007) Emerging applications of polymersomes in delivery: From molecular dynamics to shrinkage of tumors. Progress in Polymer Science, 32, 838-857.

Discher, D.E. and Eisenberg, A. (2002) Polymer vesicles. Science., 297, 967-973. Doyle, L.A., Yang, W., Abruzzo, L.V., Krogmann, T., Gao, Y., Rishi, A.K. and Ross, D.D.

(1998) A multidrug resistance transporter from human MCF-7 breast cancer cells. Proc Natl Acad Sci U S A, 95, 15665-15670.

Droogendijk, H.J., Kluin-Nelemans, H.J., van Doormaal, J.J., Oranje, A.P., van de Loosdrecht, A.A. and van Daele, P.L. (2006) Imatinib mesylate in the treatment of systemic mastocytosis: a phase II trial. Cancer, 107, 345-351.

Duzgunes, N., Straubinger, R.M., Baldwin, P.A., Friend, D.S. and Papahadjopoulos, D. (1985) Proton-induced fusion of oleic acid-phosphatidylethanolamine liposomes. Biochemistry, 24, 3091-3098.

Eto, Y., Gao, J.Q., Sekiguchi, F., Kurachi, S., Katayama, K., Maeda, M., Kawasaki, K., Mizuguchi, H., Hayakawa, T., Tsutsumi, Y., Mayumi, T. and Nakagawa, S. (2005) PEGylated adenovirus vectors containing RGD peptides on the tip of PEG show high transduction efficiency and antibody evasion ability. J Gene Med, 7, 604-612.

Fabre, J.W. and Collins, L. (2006) Synthetic peptides as non-viral DNA vectors. Curr Gene Ther, 6, 459-480.

Feil, H., Bae, Y., Feijen, J. and Kim, S. (1993) Effect of Comonomer Hydrophilicity and Ionization on the Lower Critical Solution Temperature of N-isopropylacrylamide Copolymers. Macromolecules, 26, 2496-2500.

Florent, J.C., Dong, X., Gaudel, G., Mitaku, S., Monneret, C., Gesson, J.P., Jacquesy, J.C., Mondon, M., Renoux, B., Andrianomenjanahary, S., Michel, S., Koch, M., Tillequin, F., Gerken, M., Czech, J., Straub, R. and Bosslet, K. (1998) Prodrugs of anthracyclines for use in antibody-directed enzyme prodrug therapy. J Med Chem, 41, 3572-3581.

Chapter 1

32

Fonseca, M.J., Storm, G., Hennink, W.E., Gerritsen, W.R. and Haisma, H.J. (1999) Cationic polymeric gene delivery of beta-glucuronidase for doxorubicin prodrug therapy. J Gene Med, 1, 407-414.

Funhoff, A.M., van Nostrum, C.F., Lok, M.C., Kruijtzer, J.A., Crommelin, D.J. and Hennink, W.E. (2005) Cationic polymethacrylates with covalently linked membrane destabilizing peptides as gene delivery vectors. J Control Release, 101, 233-246.

Gaber, M.H., Hong, K., Huang, S.K. and Papahadjopoulos, D. (1995) Thermosensitive sterically stabilized liposomes: formulation and in vitro studies on mechanism of doxorubicin release by bovine serum and human plasma. Pharm Res, 12, 1407-1416.

Gabizon, A.A. (2001) Pegylated liposomal doxorubicin: metamorphosis of an old drug into a new form of chemotherapy. Cancer Invest, 19, 424-436.

Gerasimov, O.V., Boomer, J.A., Qualls, M.M. and Thompson, D.H. (1999) Cytosolic drug delivery using pH- and light-sensitive liposomes. Adv Drug Deliv Rev, 38, 317-338.

Gollan, T.J. and Green, M.R. (2002) Redirecting retroviral tropism by insertion of short, nondisruptive peptide ligands into envelope. J Virol, 76, 3558-3563.

Gotoh, A., Ko, S.C., Shirakawa, T., Cheon, J., Kao, C., Miyamoto, T., Gardner, T.A., Ho, L.J., Cleutjens, C.B., Trapman, J., Graham, F.L. and Chung, L.W. (1998) Development of prostate-specific antigen promoter-based gene therapy for androgen-independent human prostate cancer. J Urol, 160, 220-229.

Graham, J., Mushin, M. and Kirkpatrick, P. (2004) Oxaliplatin. Nat Rev Drug Discov, 3, 11-12.

Gregoriadis, G., Wills, E.J., Swain, C.P. and Tavill, A.S. (1974) Drug-carrier potential of liposomes in cancer chemotherapy. Lancet, 1, 1313-1316.

Gu, J., Kagawa, S., Takakura, M., Kyo, S., Inoue, M., Roth, J.A. and Fang, B. (2000) Tumor-specific transgene expression from the human telomerase reverse transcriptase promoter enables targeting of the therapeutic effects of the Bax gene to cancers. Cancer Res, 60, 5359-5364.

Guo, X. and Szoka, F.C., Jr. (2001) Steric stabilization of fusogenic liposomes by a low-pH sensitive PEG--diortho ester--lipid conjugate. Bioconjug Chem, 12, 291-300.

Hafez, I.M. and Cullis, P.R. (2001) Roles of lipid polymorphism in intracellular delivery. Adv Drug Deliv Rev, 47, 139-148.

Haisma, H.J., Sernee, M.F., Hooijberg, E., Brakenhoff, R.H., vd Meulen-Muileman, I.H., Pinedo, H.M. and Boven, E. (1998) Construction and characterization of a fusion protein of single-chain anti-CD20 antibody and human beta-glucuronidase for antibody-directed enzyme prodrug therapy. Blood, 92, 184-190.

Han, H.D., Shin, B.C. and Choi, H.S. (2006) Doxorubicin-encapsulated thermosensitive liposomes modified with poly(N-isopropylacrylamide-co-acrylamide): drug release behavior and stability in the presence of serum. Eur J Pharm Biopharm, 62, 110-116.

Hatakeyama, H., Akita, H., Kogure, K., Oishi, M., Nagasaki, Y., Kihira, Y., Ueno, M., Kobayashi, H., Kikuchi, H. and Harashima, H. (2007) Development of a novel systemic gene delivery system for cancer therapy with a tumor-specific cleavable PEG-lipid. Gene Ther, 14, 68-77.

Hedley, S.J., Auf der Maur, A., Hohn, S., Escher, D., Barberis, A., Glasgow, J.N., Douglas, J.T., Korokhov, N. and Curiel, D.T. (2006) An adenovirus vector with a chimeric fiber incorporating stabilized single chain antibody achieves targeted gene delivery. Gene Ther, 13, 88-94.

Helmlinger, G., Yuan, F., Dellian, M. and Jain, R.K. (1997) Interstitial pH and pO2 gradients in solid tumors in vivo: high-resolution measurements reveal a lack of correlation. Nat Med, 3, 177-182.

Huang, Q., Zhang, X., Wang, H., Yan, B., Kirkpatrick, J., Dewhrist, M.W. and Li, C.Y. (2004) A novel conditionally replicative adenovirus vector targeting telomerase-positive tumor cells. Clin Cancer Res, 10, 1439-1445.

Introduction

33

Huch, M., Abate-Daga, D., Roig, J.M., Gonzalez, J.R., Fabregat, J., Sosnowski, B., Mazo, A. and Fillat, C. (2006) Targeting the CYP2B 1/cyclophosphamide suicide system to fibroblast growth factor receptors results in a potent antitumoral response in pancreatic cancer models. Hum Gene Ther, 17, 1187-1200.

Ishida, T., Harashima, H. and Kiwada, H. (2002) Liposome clearance. Biosci Rep, 22, 197-224.

Ishida, T., Kirchmeier, M.J., Moase, E.H., Zalipsky, S. and Allen, T.M. (2001) Targeted delivery and triggered release of liposomal doxorubicin enhances cytotoxicity against human B lymphoma cells. Biochim Biophys Acta, 1515, 144-158.

Iwai, M., Harada, Y., Tanaka, S., Muramatsu, A., Mori, T., Kashima, K., Imanishi, J. and Mazda, O. (2002) Polyethylenimine-mediated suicide gene transfer induces a therapeutic effect for hepatocellular carcinoma in vivo by using an Epstein-Barr virus-based plasmid vector. Biochem Biophys Res Commun, 291, 48-54.

Jain, R.K. (1987) Transport of molecules in the tumor interstitium: a review. Cancer Res, 47, 3039-3051.

Jain, R.K. (1990) Physiological barriers to delivery of monoclonal antibodies and other macromolecules in tumors. Cancer Res, 50, 814s-819s.

Juliano, R.L. and Ling, V. (1976) A surface glycoprotein modulating drug permeability in Chinese hamster ovary cell mutants. Biochim Biophys Acta, 455, 152-162.

Kakinuma, H., Fujii, I. and Nishi, Y. (2002) Selective chemotherapeutic strategies using catalytic antibodies: a common pro-moiety for antibody-directed abzyme prodrug therapy. Journal of immunological methods, 269, 269-281.

Kamiya, H., Ochiai, H., Harashima, H., Ito, M. and Matsuda, A. (2006) Transient expression of Drosophila melanogaster deoxynucleoside kinase gene enhances cytotoxicity of nucleoside analogs. Nucleosides Nucleotides Nucleic Acids, 25, 553-560.

Kaneko, S., Hallenbeck, P., Kotani, T., Nakabayashi, H., McGarrity, G., Tamaoki, T., Anderson, W.F. and Chiang, Y.L. (1995) Adenovirus-mediated gene therapy of hepatocellular carcinoma using cancer-specific gene expression. Cancer Res, 55, 5283-5287.

Karmali, P.P. and Chaudhuri, A. (2007) Cationic liposomes as non-viral carriers of gene medicines: resolved issues, open questions, and future promises. Med Res Rev, 27, 696-722.

Kerr, D.E., Senter, P.D., Burnett, W.V., Hirschberg, D.L., Hellstrom, I. and Hellstrom, K.E. (1990) Antibody-penicillin-V-amidase conjugates kill antigen-positive tumor cells when combined with doxorubicin phenoxyacetamide. Cancer Immunol Immunother, 31, 202-206.

Kichler, A. (2004) Gene transfer with modified polyethylenimines. J Gene Med, 6 Suppl 1, S3-10.

Kircheis, R., Wightman, L., Schreiber, A., Robitza, B., Rossler, V., Kursa, M. and Wagner, E. (2001) Polyethylenimine/DNA complexes shielded by transferrin target gene expression to tumors after systemic application. Gene Ther, 8, 28-40.

Klibanov, A.L., Maruyama, K., Torchilin, V.P. and Huang, L. (1990) Amphipathic polyethyleneglycols effectively prolong the circulation time of liposomes. FEBS Lett., 268, 235-237.

Knorr, V., Allmendinger, L., Walker, G.F., Paintner, F.F. and Wagner, E. (2007) An acetal-based PEGylation reagent for pH-sensitive shielding of DNA polyplexes. Bioconjug Chem, 18, 1218-1225.

Kong, G. and Dewhirst, M.W. (1999) Hyperthermia and liposomes. Int J Hyperthermia, 15, 345-370.

Kono, K. (2001) Thermosensitive polymer-modified liposomes. Adv Drug Deliv Rev, 53, 307-319.

Kursa, M., Walker, G.F., Roessler, V., Ogris, M., Roedl, W., Kircheis, R. and Wagner, E. (2003) Novel shielded transferrin-polyethylene glycol-polyethylenimine/DNA complexes for systemic tumor-targeted gene transfer. Bioconjug Chem, 14, 222-231.

Chapter 1

34

Li, Y., Yu, D.C., Chen, Y., Amin, P., Zhang, H., Nguyen, N. and Henderson, D.R. (2001) A hepatocellular carcinoma-specific adenovirus variant, CV890, eliminates distant human liver tumors in combination with doxorubicin. Cancer Res, 61, 6428-6436.

Liu, D., Ren, T. and Gao, X. (2003) Cationic transfection lipids. Curr Med Chem, 10, 1307-1315.

LoBuglio, A.F., Wheeler, R.H., Trang, J., Haynes, A., Rogers, K., Harvey, E.B., Sun, L., Ghrayeb, J. and Khazaeli, M.B. (1989) Mouse/human chimeric monoclonal antibody in man: kinetics and immune response. Proc Natl Acad Sci U S A, 86, 4220-4224.

Lown, J.W. (1993) Anthracycline and anthraquinone anticancer agents: current status and recent developments. Pharmacol Ther, 60, 185-214.

Lu, Y. and Steiner, M.S. (2000) Transcriptionally regulated adenoviruses for prostate-specific gene therapy. World J Urol, 18, 93-101.

Lungwitz, U., Breunig, M., Blunk, T. and Gopferich, A. (2005) Polyethylenimine-based non-viral gene delivery systems. Eur J Pharm Biopharm, 60, 247-266.

Manome, Y., Wen, P.Y., Dong, Y., Tanaka, T., Mitchell, B.S., Kufe, D.W. and Fine, H.A. (1996) Viral vector transduction of the human deoxycytidine kinase cDNA sensitizes glioma cells to the cytotoxic effects of cytosine arabinoside in vitro and in vivo. Nat Med, 2, 567-573.

Maruyama, K., Ishida, O., Takizawa, T. and Moribe, K. (1999) Possibility of active targeting to tumor tissues with liposomes. Adv Drug Deliv Rev, 40, 89-102.

Maruyama, K., Takizawa, T., Yuda, T., Kennel, S.J., Huang, L. and Iwatsuru, M. (1995) Targetability of novel immunoliposomes modified with amphipathic poly(ethylene glycol)s conjugated at their distal terminals to monoclonal antibodies. Biochim Biophys Acta, 1234, 74-80.

Maruyama, K., Unezaki, S., Takahashi, N. and Iwatsuru, M. (1993) Enhanced delivery of doxorubicin to tumor by long-circulating thermosensitive liposomes and local hyperthermia. Biochim Biophys Acta, 1149, 209-216.

Mathis, J.M., Williams, B.J., Sibley, D.A., Carroll, J.L., Li, J., Odaka, Y., Barlow, S., Nathan, C.O., Li, B.D. and DeBenedetti, A. (2006) Cancer-specific targeting of an adenovirus-delivered herpes simplex virus thymidine kinase suicide gene using translational control. J Gene Med, 8, 1105-1120.

Matsumura, Y. and Maeda, H. (1986) A new concept for macromolecular therapeutics in cancer chemotherapy: mechanism of tumoritropic accumulation of proteins and the antitumor agent smancs. Cancer Res, 46, 6387-6392.

McCart, J.A., Wang, Z.H., Xu, H., Hu, Y., Park, B., Alexander, H.R. and Bartlett, D.L. (2002) Development of a melanoma-specific adenovirus. Mol Ther, 6, 471-480.

Meier, W. (2000) Polymer nanocapsules. Chem. Soc. Rev., 29, 295-303. Merlin, J.L. (1991) In vitro evaluation of the association of thermosensitive liposome-

encapsulated doxorubicin with hyperthermia. Eur J Cancer, 27, 1031-1034. Meyer, D.L., Jungheim, L.N., Law, K.L., Mikolajczyk, S.D., Shepherd, T.A., Mackensen,

D.G., Briggs, S.L. and Starling, J.J. (1993) Site-specific prodrug activation by antibody-beta-lactamase conjugates: regression and long-term growth inhibition of human colon carcinoma xenograft models. Cancer Res, 53, 3956-3963.

Morimoto, E., Inase, N., Mlyake, S. and Yoshizawa, Y. (2001) Adenovirus-mediated suicide gene transfer to small cell lung carcinoma using a tumor-specific promoter. Anticancer Res, 21, 329-331.

Mueller, A., Bondurant, B. and O'Brien, D.F. (2000) Visible-Light-Stimulated Destabilization of PEG-liposomes. Macromolecules, 33, 4799-4804.

Nardin, C., Bolikal, D. and Kohn, J. (2004) Nontoxic block copolymer nanospheres: design and characterization. Langmuir., 20, 11721-11725.

Nardin, C., Widmer, J., Winterhalter, M. and Meier, W. (2001) Amphiphilic block copolymer nanocontainers as bioreactors. Eur. Phys. J. E, 4, 403-410.

Nettelbeck, D.M., Rivera, A.A., Balague, C., Alemany, R. and Curiel, D.T. (2002) Novel oncolytic adenoviruses targeted to melanoma: specific viral replication and cytolysis by expression of E1A mutants from the tyrosinase enhancer/promoter. Cancer Res, 62, 4663-4670.

Introduction

35

Northfelt, D.W., Martin, F.J., Working, P., Volberding, P.A., Russell, J., Newman, M., Amantea, M.A. and Kaplan, L.D. (1996) Doxorubicin encapsulated in liposomes containing surface-bound polyethylene glycol: pharmacokinetics, tumor localization, and safety in patients with AIDS-related Kaposi's sarcoma. J Clin Pharmacol, 36, 55-63.

Peng, X.Y., Won, J.H., Rutherford, T., Fujii, T., Zelterman, D., Pizzorno, G., Sapi, E., Leavitt, J., Kacinski, B., Crystal, R., Schwartz, P. and Deisseroth, A. (2001) The use of the L-plastin promoter for adenoviral-mediated, tumor-specific gene expression in ovarian and bladder cancer cell lines. Cancer Res, 61, 4405-4413.

Pereboev, A.V., Nagle, J.M., Shakhmatov, M.A., Triozzi, P.L., Matthews, Q.L., Kawakami, Y., Curiel, D.T. and Blackwell, J.L. (2004) Enhanced gene transfer to mouse dendritic cells using adenoviral vectors coated with a novel adapter molecule. Mol Ther, 9, 712-720.

Pereboeva, L., Komarova, S., Mahasreshti, P.J. and Curiel, D.T. (2004) Fiber-mosaic adenovirus as a novel approach to design genetically modified adenoviral vectors. Virus Res, 105, 35-46.

Plumb, J.A., Bilsland, A., Kakani, R., Zhao, J., Glasspool, R.M., Knox, R.J., Evans, T.R. and Keith, W.N. (2001) Telomerase-specific suicide gene therapy vectors expressing bacterial nitroreductase sensitize human cancer cells to the pro-drug CB1954. Oncogene, 20, 7797-7803.

Reichert, J.M., Rosensweig, C.J., Faden, L.B. and Dewitz, M.C. (2005) Monoclonal antibody successes in the clinic. Nat Biotechnol, 23, 1073-1078.

Sapra, P. and Allen, T.M. (2003) Ligand-targeted liposomal anticancer drugs. Prog Lipid Res, 42, 439-462.

Schepelmann, S., Ogilvie, L.M., Hedley, D., Friedlos, F., Martin, J., Scanlon, I., Chen, P., Marais, R. and Springer, C.J. (2007) Suicide gene therapy of human colon carcinoma xenografts using an armed oncolytic adenovirus expressing carboxypeptidase G2. Cancer Res, 67, 4949-4955.

Schild, H. (1992) Poly(N-isopropylacrylamide): experiment, theory and application. Progress in Polymer Science, 17, 163-249.

Senter, P.D., Schreiber, G.J., Hirschberg, D.L., Ashe, S.A., Hellstrom, K.E. and Hellstrom, I. (1989) Enhancement of the in vitro and in vivo antitumor activities of phosphorylated mitomycin C and etoposide derivatives by monoclonal antibody-alkaline phosphatase conjugates. Cancer Res, 49, 5789-5792.

Sharma, S.K., Bagshawe, K.D., Burke, P.J., Boden, J.A., Rogers, G.T., Springer, C.J., Melton, R.G. and Sherwood, R.F. (1994) Galactosylated antibodies and antibody-enzyme conjugates in antibody-directed enzyme prodrug therapy. Cancer, 73, 1114-1120.

Shimura, H., Suzuki, H., Miyazaki, A., Furuya, F., Ohta, K., Haraguchi, K., Endo, T. and Onaya, T. (2001) Transcriptional activation of the thyroglobulin promoter directing suicide gene expression by thyroid transcription factor-1 in thyroid cancer cells. Cancer Res, 61, 3640-3646.

Siders, W.M., Halloran, P.J. and Fenton, R.G. (1996) Transcriptional targeting of recombinant adenoviruses to human and murine melanoma cells. Cancer Res, 56, 5638-5646.

Smith, G.K., Banks, S., Blumenkopf, T.A., Cory, M., Humphreys, J., Laethem, R.M., Miller, J., Moxham, C.P., Mullin, R., Ray, P.H., Walton, L.M. and Wolfe, L.A., 3rd. (1997) Toward antibody-directed enzyme prodrug therapy with the T268G mutant of human carboxypeptidase A1 and novel in vivo stable prodrugs of methotrexate. J Biol Chem, 272, 15804-15816.

Snitkovsky, S. and Young, J.A. (2002) Targeting retroviral vector infection to cells that express heregulin receptors using a TVA-heregulin bridge protein. Virology, 292, 150-155.

Song, C.W., Lokshina, A., Rhee, J.G., Patten, M. and Levitt, S.H. (1984) Implication of blood flow in hyperthermic treatment of tumors. IEEE Trans Biomed Eng, 31, 9-16.

Chapter 1

36

Spratt, T., Bondurant, B. and O'Brien, D.F. (2003) Rapid release of liposomal contents upon photoinitiated destabilization with UV exposure. Biochim Biophys Acta, 1611, 35-43.

Springer, C.J., Bagshawe, K.D., Sharma, S.K., Searle, F., Boden, J.A., Antoniw, P., Burke, P.J., Rogers, G.T., Sherwood, R.F. and Melton, R.G. (1991) Ablation of human choriocarcinoma xenografts in nude mice by antibody-directed enzyme prodrug therapy (ADEPT) with three novel compounds. Eur J Cancer, 27, 1361-1366.

Stachler, M.D. and Bartlett, J.S. (2006) Mosaic vectors comprised of modified AAV1 capsid proteins for efficient vector purification and targeting to vascular endothelial cells. Gene Ther, 13, 926-931.

Stoenescu, R., Graff, A. and Meier, W. (2004) Asymmetric ABC-Triblock Copolymer Membranes Induce a Directed Insertion of Membrane Proteins. Macromol Biosci, 4, 930-935.

Stribbling, S.M., Martin, J., Pedley, R.B., Boden, J.A., Sharma, S.K. and Springer, C.J. (1997) Biodistribution of an antibody-enzyme conjugate for antibody-directed enzyme prodrug therapy in nude mice bearing a human colon adenocarcinoma xenograft. Cancer Chemother Pharmacol, 40, 277-284.

Stubbs, M., McSheehy, P.M., Griffiths, J.R. and Bashford, C.L. (2000) Causes and consequences of tumour acidity and implications for treatment. Mol Med Today, 6, 15-19.

Syrigos, K.N., Rowlinson-Busza, G. and Epenetos, A.A. (1998) In vitro cytotoxicity following specific activation of amygdalin by beta-glucosidase conjugated to a bladder cancer-associated monoclonal antibody. Int J Cancer, 78, 712-719.

Tai, C.K., Logg, C.R., Park, J.M., Anderson, W.F., Press, M.F. and Kasahara, N. (2003) Antibody-mediated targeting of replication-competent retroviral vectors. Hum Gene Ther, 14, 789-802.

Takeuchi, H., Kojima, H., Yamamoto, H. and Kawashima, Y. (2001) Evaluation of circulation profiles of liposomes coated with hydrophilic polymers having different molecular weights in rats. J Control Release, 75, 83-91.

Tannock, I.F. and Rotin, D. (1989) Acid pH in tumors and its potential for therapeutic exploitation. Cancer Res, 49, 4373-4384.

Torchilin, V.P., Levchenko, T.S., Whiteman, K.R., Yaroslavov, A.A., Tsatsakis, A.M., Rizos, A.K., Michailova, E.V. and Shtilman, M.I. (2001) Amphiphilic poly-N-vinylpyrrolidones: synthesis, properties and liposome surface modification. Biomaterials, 22, 3035-3044.

Torchilin, V.P., Omelyanenko, V.G., Papisov, M.I., Bogdanov, A.A., Jr., Trubetskoy, V.S., Herron, J.N. and Gentry, C.A. (1994) Poly(ethylene glycol) on the liposome surface: on the mechanism of polymer-coated liposome longevity. Biochim Biophys Acta, 1195, 11-20.

Uherek, C., Fominaya, J. and Wels, W. (1998) A modular DNA carrier protein based on the structure of diphtheria toxin mediates target cell-specific gene delivery. J Biol Chem, 273, 8835-8841.

Vassaux, G., Hurst, H.C. and Lemoine, N.R. (1999) Insulation of a conditionally expressed transgene in an adenoviral vector. Gene Ther, 6, 1192-1197.

Vaughan, T.J., Osbourn, J.K. and Tempest, P.R. (1998) Human antibodies by design. Nat Biotechnol, 16, 535-539.

Vrudhula, V.M., Senter, P.D., Fischer, K.J. and Wallace, P.M. (1993) Prodrugs of doxorubicin and melphalan and their activation by a monoclonal antibody-penicillin-G amidase conjugate. J Med Chem, 36, 919-923.

Wall, M.E. and Wani, M.C. (1995) Camptothecin and taxol: discovery to clinic--thirteenth Bruce F. Cain Memorial Award Lecture. Cancer Res, 55, 753-760.

Weinstein, J.N., Magin, R.L., Yatvin, M.B. and Zaharko, D.S. (1979) Liposomes and local hyperthermia: selective delivery of methotrexate to heated tumors. Science, 204, 188-191.

Introduction

37

Woodle, M.C., Engbers, C.M. and Zalipsky, S. (1994) New amphipatic polymer-lipid conjugates forming long-circulating reticuloendothelial system-evading liposomes. Bioconjug Chem, 5, 493-496.

Wu, Z., Asokan, A. and Samulski, R.J. (2006) Adeno-associated virus serotypes: vector toolkit for human gene therapy. Mol Ther, 14, 316-327.

Wurdinger, T., Verheije, M.H., Raaben, M., Bosch, B.J., de Haan, C.A., van Beusechem, V.W., Rottier, P.J. and Gerritsen, W.R. (2005) Targeting non-human coronaviruses to human cancer cells using a bispecific single-chain antibody. Gene Ther, 12, 1394-1404.

Yamamoto, M., Alemany, R., Adachi, Y., Grizzle, W.E. and Curiel, D.T. (2001) Characterization of the cyclooxygenase-2 promoter in an adenoviral vector and its application for the mitigation of toxicity in suicide gene therapy of gastrointestinal cancers. Mol Ther, 3, 385-394.

Yang, Q., Mamounas, M., Yu, G., Kennedy, S., Leaker, B., Merson, J., Wong-Staal, F., Yu, M. and Barber, J.R. (1998) Development of novel cell surface CD34-targeted recombinant adenoassociated virus vectors for gene therapy. Hum Gene Ther, 9, 1929-1937.

Yatvin, M.B., Weinstein, J.N., Dennis, W.H. and Blumenthal, R. (1978) Design of liposomes for enhanced local release of drugs by hyperthermia. Science, 202, 1290-1293.

Zalipsky, S., Hansen, C.B., Oaks, J.M. and Allen, T.M. (1996) Evaluation of blood clearance rates and biodistribution of poly(2-oxazoline)-grafted liposomes. J Pharm Sci, 85, 133-137.

Zalipsky, S., Qazen, M., Walker, J.A., 2nd, Mullah, N., Quinn, Y.P. and Huang, S.K. (1999) New detachable poly(ethylene glycol) conjugates: cysteine-cleavable lipopolymers regenerating natural phospholipid, diacyl phosphatidylethanolamine. Bioconjug Chem, 10, 703-707.

Zanta, M.A., Belguise-Valladier, P. and Behr, J.P. (1999) Gene delivery: a single nuclear localization signal peptide is sufficient to carry DNA to the cell nucleus. Proc Natl Acad Sci U S A, 96, 91-96.

Zeng, H., Wei, Q., Huang, R., Chen, N., Dong, Q., Yang, Y. and Zhou, Q. (2007) Recombinant adenovirus mediated prostate-specific enzyme pro-drug gene therapy regulated by prostate-specific membrane antigen (PSMA) enhancer/promoter. J Androl, 28, 827-835.

Zhang, M., Li, S., Nyati, M.K., DeRemer, S., Parsels, J., Rehemtulla, A., Ensminger, W.D. and Lawrence, T.S. (2003) Regional delivery and selective expression of a high-activity yeast cytosine deaminase in an intrahepatic colon cancer model. Cancer Res, 63, 658-663.

Zhang, R., Straus, F.H. and DeGroot, L.J. (2001) Adenoviral-mediated gene therapy for thyroid carcinoma using thymidine kinase controlled by thyroglobulin promoter demonstrates high specificity and low toxicity. Thyroid, 11, 115-123.

Zhou, J.H., Tang, B., Liu, X.L., He, D.W. and Yang, D.T. (2007) hTERT-targeted E. coli purine nucleoside phosphorylase gene/6-methylpurine deoxyribose therapy for pancreatic cancer. Chin Med J (Engl), 120, 1348-1352.

Zhu, Z.B., Makhija, S.K., Lu, B., Wang, M., Kaliberova, L., Liu, B., Rivera, A.A., Nettelbeck, D.M., Mahasreshti, P.J., Leath, C.A., Barker, S., Yamaoto, M., Li, F., Alvarez, R.D. and Curiel, D.T. (2004) Transcriptional targeting of tumors with a novel tumor-specific survivin promoter. Cancer Gene Ther, 11, 256-262.

Chapter 2 State of the art: a nanoreactor coming to life

39

Chapter 2:

State of the art: a nanoreactor

coming to life

Chapter 2

40

State of the art: a nanoreactor coming to life

41

2.1 Aim of the work: a novel directed enzyme-prodrug strategy

Since both antibody-directed enzyme-prodrug therapy (ADEPT) and gene-

directed enzyme-prodrug therapy (GDEPT) (Chapter 1) suffer from several

drawbacks, such as immunogenicity of the antibody-enzyme conjugate,

inefficient transfection of gene delivery vectors, unsustained gene expression,

pathogenesis of viral vectors and the risk of insertional mutagenesis, we want to

introduce a novel enzyme–prodrug system based on triblock copolymeric

vesicles.

Patrick Van Gelder, my co-promotor was working in collaboration with Prof. W.

Meier at the university of Basel where they inserted a bacterial outer membrane

protein maltoporin (LamB) (Ranquin and Van Gelder, 2004) into the triblock

copolymeric membrane of nanometer-sized vesicles composed of poly(2-

methyloxazoline)-β-poly(dimethyl siloxane)- β-poly(2-methyloxazoline) (PMOXA-

PDMS-PMOXA). They showed that maltoporin is still functional after insertion in

the membrane. Moreover, lambda phage is still able to bind to its receptor and

inject its DNA through the porin into the polymeric vesicles (Graff et al., 2002).

This clearly showed that otherwise impermeable polymeric vesicles can be

permeabilized for various solutes by incorporation of bacterial porins. Nardin et

al. also demonstrated that enzymes can be trapped inside these polymeric

vesicles without loss of function by encapsulating β-lactamase and inserting

OmpF in the membrane (Nardin et al., 2000b).

In the lab of Prof. J. Steyaert, my promoter, nucleoside hydrolases from

various origins were extensively researched for several years. These enzymes

catalyse the hydrolysis of nucleosides into the respective nucleobase and ribose

(Versees et al., 2001). They in particular studied the nucleoside hydrolase of

Trypanosoma vivax (TvNH) that is capable of hydrolysing several nucleoside

analogs into the respective nucleobase analog. Since many chemotherapeutics

used to date are nucleobases they saw the potential of TvNH to function as a

prodrug activating enzyme in cancer therapy.

Patrick Van Gelder brought these two visions together and thus the idea was

born to use such PMOXA-PDMS-PMOXA vesicles that are permeabilized by

Chapter 2

42

bacterial porins, to encapsulate prodrug-activating enzymes and use such

nanoreactors as a novel prodrug activating system. This system is likely to have

many advantages over ADEPT since the enzyme is shielded from the

environment and thus protected from proteolytic degradation and binding to

components of the immune system. In addition, each therapeutic unit holds a

high concentration of enzymes as compared to ADEPT where only one enzyme is

present per therapeutic unit.

In this project the two main goals are: validating the TvNH as a prodrug

activating enzyme and ultimately deliver a proof of principle that nanoreactors

that are permeabilized with bacterial porins and encapsulate a prodrug activating

enzyme can function as prodrug activating systems.

In this chapter the individual building blocks of these nanoreactors are

described in more detail.

2.2 PMOXA-PDMS-PMOXA triblock copolymers

Amphiphilic block copolymers are composed of at least one hydrophilic block

and one hydrophobic block, covalently linked. They are similar to low molar-mass

amphiphiles such as surfactants or lipids, and can self-assemble into a plethora

of lyotropic mesophases (Alexandridis et al., 1998; Forster and Plantenberg,

2002). At low concentrations, they self assemble into micelles of various shapes

and at higher concentrations into lyotropic liquid crystalline phases. Typical

bilayer particles are formed upon self-assembly in aqueous solutions, consisting

of a core comprised of their insoluble part surrounded by a corona of their

soluble part (Nardin et al., 2001). The driving forces for the self-organization are

the difference in solubility of the blocks and the constraint imposed by the

chemical linkage between the blocks. Depending on their concentration,

molecular shape, hydrophobic-to-hydrophilic balance and block-length, micelles,

vesicles, cylinders or rod-like structures are formed.

In this work the ABA triblock copolymer poly(2-methyloxazoline)-b-

poly(dimethyl siloxane)-b-poly(2-methyloxazoline) (PMOXA-PDMS-PMOXA) is

used which was synthesized by the group of Wolfgang Meier (Figure 1) or

purchased from Polymer Source Inc.. For a given composition of PMOXA-PDMS-

State of the art: a nanoreactor coming to life

43

PMOXA, a homogeneous liquid crystalline phase is observed in water at

concentrations above 48 wt. %. Below 48 wt. % the system shows a broad

miscibility gap where the lamellar phase coexists with excess water. The phase

behaviour of this triblock copolymer is similar to that of typical bilayer forming

lipids like lecithin and it self assembles into vesicular structures in dilute aqueous

solutions which consist of a spherically closed triblock copolymer membrane as

shown by Nardin et al (Nardin et al., 2000a).. These vesicles have a size range of

50 to 500 nm as shown by cryo-TEM. Furthermore, cryo-TEM images also

revealed a membrane thickness of 10 nm. The mechanical and visco-elastical

properties of this PMOXA-PDMS-PMOXA membrane were characterized by

applying short electric field pulses to giant free-standing membranes (Nardin,

2000). Electrical field pulses are used to charge the membrane resulting in

electric stress. Above a critical voltage the membrane will rupture and a fast

discharge across the rupture occurs. Analysis of the discharge kinetics gives

information about the strength and stability of the membrane. This experiment

indicated that PMOXA-PDMS-PMOXA membranes were considerably more stable

than conventional black lipid membranes and posses a high flexibility provided by

the PDMS middle block. Crosslinking the PMOXA-PDMS-PMOXA polymers via

methacrylate endgroups further increased the stability. The reason for this higher

stability compared to conventional lipid membranes, is the larger size of the

hydrophobic part and the slower dynamics of the underlying copolymer

molecules caused by a higher entanglement (Battaglia and Ryan, 2005). Since

the formation of membrane-like structures of the PMOXA-PDMS-PMOXA triblock

copolymers in aqueous solutions, closely resembles the behaviour of lipid

molecules, the polymer membranes can be regarded as a mimetic of biological

membranes. To test this hypothesis, Nardin et al. successfully reconstituted the

outer membrane protein OmpF in the membrane of PMOXA-PDMS-PMOXA

triblock copolymeric vesicles (Nardin et al., 2001). Although the polymer

membrane (10 nm) is two- to threefold thicker than a conventional lipid

membrane, the functionality of the membrane protein was fully preserved upon

reconstitution in the polymer membrane. This is possible due to the high

flexibility of the hydrophobic block and the conformational freedom of the

polymer molecules, which allows a block copolymer membrane to adapt to the

specific geometric and dynamic requirements of membrane proteins without

considerable loss of free energy (Pata and Dan, 2003). Currently, other

Chapter 2

44

membrane proteins have been reconstituted in the PMOXA-PDMS-PMOXA

membrane, including LamB (Graff et al., 2002), aquaporin (Stoenescu et al.,

2004) and FhuA (Nallani et al., 2006).

Finally, the PMOXA outer blocks posses similar stealth properties as PEG which

makes PMOXA-PDMS-PMOXA based vesicles very interesting for use in

biomedical applications (Woodle et al., 1994).

Figure 1: composition of PMOXA-PDMS-PMOXA polymers. In this representation the polymers were functionalised with methacrylate endgroups for polymer crosslinking via UV irradiation. X en y represent the amount of dimethyl siloxane and methyloxazoline repeats respectively. For the polymers used in this thesis, x and y are 72 and 18 respectively or 54 and 21 respectively.

2.3 Bacterial outer membrane proteins

The outer membrane protects Gram-negative bacteria against a harsh

environment. At the same time, the embedded proteins fulfil a number of tasks

that are crucial to the bacterial cell, such as solute and protein translocation, as

well as signal transduction. Unlike membrane proteins from all other sources,

integral outer membrane proteins do not consist of transmembrane α-helices,

but instead fold into antiparallel β-barrels.

The outer membrane (OM) is an asymmetrical membrane consisting of

phospholipids and mostly lipopolysaccharides (LPS) in the inner and outer

monolayer, respectively. Transport across the outer membrane is mediated by

channel forming proteins. General porins like OmpF, OmpC and PhoE allow

passive diffusion of small hydrophilic solutes with molecular weights up to 600

Da. The flux of solutes through these porins is proportional to the permeability

and the concentration gradient between the periplasmic space and the outside

State of the art: a nanoreactor coming to life

45

medium. If this concentration gradient becomes too small, the flux can only be

maintained by increasing the permeability. One possible option to do this is to

increase the number of channels. However, only small increases in flux can be

gained using this option: the theory of diffusion (Berg, 1993) tells us that only

1% covering of the cell surface with channels is sufficient to reach 50% of the

flux of a cell that is completely covered with channels. Alternatively, higher

fluxes could potentially be obtained by increasing the radius (r) of the channel

with the consequence that the channel surface increases with r2. Unfortunately,

toxins and bile salts would then easily enter the cell and the protective function

of the OM would be lost. Therefore, other porins have evolved that facilitate the

uptake of certain solutes by the presence of a specific binding site. Examples are

LamB, ScrY and Tsx, which are specific for maltodextrins, sucrose and

nucleosides respectively.

In this thesis two aspecific porins, OmpF and PhoE and one specific porin, Tsx

were used.

2.3.1. Aspecific porins OmpF and PhoE

OmpF, the receptor for bacteriophage T2, is one of the general porins and is

abundantly present in the outer membrane. PhoE on the other hand is a

phosphate-limitation induced porin that is only expressed under phosphate

starvation. Both porins allow transport of solutes up to 600 Da. OmpF is a

slightly cation selective porin with a molecular mass of 37 kDa whereas PhoE is

an anion selective porin with a molecular mass of 36.8 kDa. The transport of

solutes through both porins is osmotically regulated and therefore solely

governed by the concentration gradient across the outer membrane. They both

fold into 16 stranded β-barrels and insert in the outer membrane as homotrimers

(Cowan et al., 1992) (Figure 2). Loop L3 plays an important factor in pore

permeability because it folds back into the barrel thus forming a constriction

zone at half the height of the channel, giving it an hourglass-like shape.

Residues at this constriction zone on L3 and the opposing barrel wall contribute

to the ion-selectivity filter by creating a transverse electrostatic field. Site

directed mutagenesis revealed that lysine and arginine residues determine the

anion selectivity of PhoE together with additional charges at the mouth and

inside the barrel (Bauer et al., 1989; Benz et al., 1989). In OmpF, the cation

Chapter 2

46

selectivity is determined by three arginine residues and two aspartate residues at

the constriction zone (Phale et al., 2001). OmpF and PhoE, like all other porins

are extremely stable proteins that can resist denaturation in the presence of 5M

guanidinium chloride or 2% SDS at 70 °C. It was shown by Phale et al. (Phale et

al., 1998) that the latching loop L2 contributes strongly to this extreme stability

because it bends over the wall of an adjacent monomer. Additionally, the trimer

hydrophobic interface adds to the robustness of the trimeric porin (Phale et al.,

1998; Van Gelder et al., 1996).

Figure 2: Bacterial β-barrel porins OmpF and PhoE as seen from the plane of the membrane (OmpF) and from the top of the membrane (PhoE). Loop L3 is bend back into the barrel.

Conductance measurements revealed that both porins can exist in an open or

closed state, depending on the transmembrane voltage (Delcour, 1997). Charged

residues in the channel are most probably responsible for this voltage gating

since replacement of these residues result in a changed voltage sensitivity (Saint

et al., 1996; Van Gelder et al., 1997). Interestingly, OmpF and PhoE show an

opposite voltage dependency. Moreover, the voltage dependency of PhoE can be

changed to match that of OmpF by substituting the charged residues at the

constriction zone by those of OmpF (Samartzidou and Delcour, 1998). Therefore

it is believed that charged residues at the constriction zone are responsible for

voltage gating.

Since both OmpF and PhoE are aspecific porins that show only a slight ion

selectivity and allow transport of solutes up to 600 Da, they are suited for

L3

OmpF PhoE

State of the art: a nanoreactor coming to life

47

transport of nucleoside(analogues) and nucleobase(analogues) through the

nanoreactor membrane.

2.3.2. Specific porin Tsx

Tsx was discovered as the receptor for bacteriophage T6, hence its name Tsx.

Tsx is a substrate-specific transporter with a molecular mass of 31.5 kDa that

has low affinity (µM to mM) binding sites for its substrates which are nucleosides

and deoxynucleosides. These binding sites are saturable and allow efficient

diffusion at low concentration gradients (Benz et al., 1988; Maier et al., 1988).

Although Tsx is substrate specific, it allows diffusion of small solutes such as

serine (Heuzenroeder and Reeves, 1981; Luckey and Nikaido, 1980) and also

transports albicidin, a relatively high molecular weight (850 Da) antibiotic (Birch

et al., 1990).The importance of the Tsx protein for nucleoside uptake becomes

apparent only at low substrate concentrations (<1µM). At higher concentrations,

Tsx becomes dispensable and transport of nucleosides occurs through general

porins. In tsx mutants, the rate of uptake of adenosine and thymidine is strongly

reduced but interestingly the rate of uptake of cytidine is unaffected.

Furthermore, conductance measurements showed that the saturation constant Ks

for cytidine, 2x10-2 M is very high compared to adenosine (6x10-4 M), meaning

that the affinity of Tsx for cytidine is very low (Maier et al., 1988). This is

probably related to the fact that the expression of the tsx gene is under a

negative control of cytR and cytidine is the effector molecule of the cytR

repressor. The apparent Tsx-independent permeation of cytidine across the outer

membrane at very low concentrations may allow cytidine to alert the cell to the

presence of other exogenous nucleosides. Comparison of the in vivo Tsx

dependent transport of adenosine and adenine arabinoside indicated that Tsx

does not strongly differentiate between nucleosides with different pentose

moieties (Krieger-Brauer and Braun, 1980). Tsx does not seem to play a role in

the transport of free nucleobases or monophosphate nucleosides (Benz et al.,

1988; McKeown et al., 1976; van Alphen et al., 1978). Reconstituted Tsx forms a

very low-conductance channel of 10 pS in 1M KCl compared to 1.9 nS for OmpF

(Buehler et al., 1991), which is indicative for a narrow pore (Benz et al., 1988;

Maier et al., 1988).

Chapter 2

48

The crystal structure of Tsx in complex with nucleosides was solved in 2004 by

Ye and Van den Berg (Ye and Van Den Berg, 2004).

Figure 3: Crystal structure of the E. coli Tsx channel. Surface representation viewed from the extracellular side (A). Cut away side view showing the nucleoside binding sites Nuc0, Nuc1 and Nuc2 in thymidine soaked crystals. The aromatic residues that line the channel and that are involved in nucleoside binding are indicated in cyan.

Tsx is a monomeric protein that folds into a 12 stranded β-barrel. There are no

extracellular loops that bend back into the barrel to form an hourglass shaped

constriction. Instead the channel has a much longer constriction zone, spanning

almost the entire membrane. The cross section of the channel in the plane of the

membrane shows a keyhole-like shape (Figure 3a). Transport of the nucleosides

through the channel is achieved via a greasy slide mechanism (Van Gelder et al.,

2002): the substrate binds through hydrophobic contacts with pairs of aromatic

residues located on opposite sides of the channel and through hydrogen-bonding

interactions with ionisable residues. Binding and release of the substrate by

these weak binding sites results in the movement of the substrates through the

pore (Figure 3b). The net direction of transport is governed by the concentration

gradient of the substrate across the outer membrane. Both the nucleobase

moiety and the riboside moiety contribute to the binding.

State of the art: a nanoreactor coming to life

49

Since Tsx is able to efficiently transport nucleosides at submicromolar

concentrations, it is very suitable to permeabilise nanoreactors for nucleosides

and nucleoside analogs.

2.4 Nucleoside Hydrolase of Trypanosoma vivax

Nucleoside hydrolases (or nucleoside N-ribohydrolases, NHs) comprise a

superfamily of structurally related metalloproteins with a unique β–sheet

topology. They are glycosidases that catalyze the hydrolysis of the N-glycosidic

bond between the anomeric carbon atom of ribose and purine or pyrimidine base

to form the free nucleobase and ribose (Figure 4). They all exert a stringent

specificity towards the ribose moiety but their preference towards the nature of

the nucleobase moiety is more variable. All members of this superfamily have

the characteristic N-terminal motif DXDXXXDD. Scanning genomes for the

occurrence of this fingerprint showed that NHs are widely distributed in nature.

They have been found in bacteria (Ogawa et al., 2001; Petersen and Moller,

2001), protozoa (Cui et al., 2001; Pelle et al., 1998), yeasts (Kurtz et al., 2002),

insects (Ribeiro and Valenzuela, 2003), mesozoa (Versees et al., 2003), plants,

amphibians and fish. Interestingly they were not found in mammals. Their

metabolic role is best understood in parasitic protozoa were they are part of the

salvage pathway for the scavenging of purines from their environment. Parasitic

protozoa rely on this purine salvage pathway for their survival since they are

unable to synthesize purines de novo. These protozoan NHs have been

extensively studied since they are potential targets for treatment. Protozoan NHs

are classified into three categories based on their substrate specificity: the base-

aspecific inosine-uridine preferring NHs,(IU-NH) the purine-specific inosine-

adenosine-guanosine NHs (IAG-NH) and the 6-oxopurine specific inosine-

guanosine NHs (IG-NH).

Chapter 2

50

Figure 4: Catalytic reaction of NHs. The NH-catalysed hydrolysis of a ribonucleoside (examplified by inosine).

The NH of Trypanosoma vivax (TvNH) used in this project belongs to the IAG-

NH class of nucleoside hydrolases with a molecular mass of 37 kDa.

Trypanosoma vivax causes trypanosomiasis, or sleeping sickness, in domestic

animals and wildlife. Therefore it was well studied as a possible target in our lab.

The substrate specificity was first determined by measuring the activity for

various substrates (Table 1). A faster turnover (kcat ↑) and higher substrate

affinity (KM ↓) resulted in a higher specificity for purines as compared to

pyrimidines (Versees et al., 2001). In 2001 Versées et al. solved the crystal

structure of TvNH which gave much insight in the mechanism of IAG-NHs since it

was the first structure of an IAG-NH that was solved. They solved the structure

of TvNH and TvNH in complex with 3-deaza-adenosine, an adenosine analogue

that is an inhibitor of the TvNH and binds to the active site with high affinity

(Figure 5A). TvNH is a homodimer, with each monomer being a single domain,

globular protein. Each subunit consists of 10 β-strands, 12 α-helices and three

short 310 helices (Figure 5B). Eight of the ten β-strands form a central mixed

sheet with 7 parallel strands (β1-β6 and β10) and one antiparallel strand (β7).

The parallel strands β1-β6 and α-helices 1-7 form a motif that resembles a

Rossman fold. The first βαβαβ motif of the Rossman fold is known to be involved

in nucleotide binding. Since TvNH is specific for nucleosides instead of

nucleotides, the consensus sequences GXGXXG involved in phosphate- binding,

is replaced in TvNH by an aspartate-rich region that binds Ca2+.

The active sites are located at the C-terminal ends of the core 8-stranded β-

sheet of both subunits and is formed by strands β1, β2, β4 and β5 and helices

α1, α3, α10 and α11. A bound calcium ion is also situated in the active site. 3-

State of the art: a nanoreactor coming to life

51

deaza-adenosine is bound in the active site with its ribose interacting with the

Ca2+ and the purine base pointed towards the exit of the active-site. The bound

ribose adopts an envelope conformation whereas the purine base adopts a syn

conformation towards the ribose. When 3-deaza-adenosine is bound, the Ca2+-

ion interacts with two hydroxyl groups of the ribose. Furthermore, the enzyme

forms hydrogen bonds with all three hydroxyl groups of the ribose, explaining

the high specificity toward the ribose moiety of the substrates. The 3-deaza-

adenine leaving group is involved in extensive aromatic stacking interactions with

two tryptophan residues (Trp 83 and 260). It is this aromatic stacking that

contributes to the purine selectivity of TvNH.

Figure 5: Structure (A) and topology (B) of a TvNH subunit

The enzymatic reaction occurs via a SN1mechanism with an oxocarbenium ion

transition state. The nucleophile in this reaction is probably a Ca-bound water

molecule that is activated by Asp10 by abstraction of a proton. In many other

enzymes such as IU-NHs, phosphorylases and AMP-nucleosidases,

hydrolysis/phosphorolysis of the N-glycosidic bond in nucleosides and nucleotides

commonly involves the protonation of the leaving nucleobase concomitant with

nucleophilic attack. In TvNH however no general acid or hydrogen bond donor

could be identified and leaving group activation occurs by another mechanism:

aromatic stacking of the purine base by tryptophans 83 and 260 lowers the pka

of the leaving group by several units, allowing protonation of N-7 by a solvent

molecule (Versees et al., 2004) (Figure 6)

The aromatic stacking of the heterocyclic nucleobase is an important feature of

this enzyme, which makes TvNH promiscuous towards the nature of the

BA BA

Chapter 2

52

nucleobase moiety. In our lab several nucleoside analogues were tested as

substrate for TvNH (Table 1). These data clearly show that when the ribose

moiety is altered like in 2’-deoxy nucleoside analogues, the activity of TvNH is

greatly diminished. Alternatively, when the nucleobase moiety is altered like in

purine riboside, 6-thioguanosine..., the activity of TvNH is retained. One

exception is p-nitrophenyl riboside. This is due to the fact that p-nitrophenyl

riboside cannot be protonated, however it is susceptible to decomposition when

the ribosyl group is converted to the oxocarbenium ion, resulting in reduced but

not diminished activity of TvNH.

The nucleobase analogues of several compounds tested are known cytotoxic

agents such as 2-fluoroadenine, 2-chloroadenine, 6-methylpurine and 6-

thioguanine. Therefore TvNH is a good candidate for prodrug activation.

Figure 6: Schematic representation of the interactions in the active site of nucleoside hydrolase of T. vivax and a nucleoside.

State of the art: a nanoreactor coming to life

53

Table 1: kinetic parameters for nucleosides and nucleoside analogues of TvNH

Compound kcat (S-1) Km (µM) Kcat/Km

(s-1mM-1) Kd (µM)

Adenosine 2.58 8 322

Guanosine 2.31 2.33 991

Inosine 5.19 5.37 966

Cytidine 0.338 925 0.36

Uridine 0.022 586 0.037

Para-nitrophenylriboside 0.206 257 0.8

Xanthosine 2.35 1760 1.3

Purine riboside 3.92 3.79 1034

2’-deoxyadenosine 0.0043 23 0.19

3’-deoxyadenosine 0.00031 223 0.0014

5’-deoxyadenosine 0.030 656 0.046

7-methylguanosine 3.45 2.13 1610

3-methylcytidine 0.98 1.97 497

3-deaza-adenosine 0.2

7-deaza-adenosine 356

5’-deoxyguanosine 0.079 504 0.16 5’-deoxy-7-methylguanosine

4.33 126.9 34.1

2-fluoroadenosine 1.86 39.05 47.6

6-iodopurineriboside 2.17 <2 >1085

8-bromoadenosine

6-thioguanosine 1.77 3.56 497

2-chloroadenosine 1.97 2.70 729

Nicotinamide riboside 1.59 4.82 330

6-methylpurine riboside 4.3 < 10 > 430

2.5 Purine nucleoside analogs as prodrugs

Nucleoside analogs are one of the most common classes of drugs used to treat

cancer. Cytotoxic nucleoside analogs are antimetabolites that can be

incorporated into DNA or RNA macromolecules or inhibit enzymes involved in

nucleoside synthesis, hereby inducing apoptosis. When using nucleoside analogs

as prodrugs, it is important that the nucleoside analog prodrug is non toxic and

cannot be metabolised by endogenous enzymes to a toxic form. Furthermore,

the nucleobase analog drug should be toxic and cannot be metabolised by

endogenous enzymes to an inactive form.

Chapter 2

54

2.5.1. Deoxyadenosine analogs

The development of 2-chloro-2’ deoxyadenosine (CdA) and 2-fluoro-2’-

deoxyadenosine (FdA) derives from the understanding of the pathogenesis of

adenosine deaminase (ADA) deficiency. In this metabolic disease, 2’-

deoxyadenosine is accumulated due to the enzyme defect. 2’-deoxyadenosine is

then further metabolised by deoxycytidine kinase (dCK), nucleoside

monophospate kinase and nucleoside diphospohate kinase to form the metabolite

2’-deoxyadenosine triphosphate (dATP) which is toxic in higher concentrations

(Figure 7). Since dCK is highly upregulated in lymphocytes this results in the

immunodeficiency related to ADA-deficiency. This understanding led Carson and

co-workers to synthesize 2’-deoxyadenosine analogs that also exert a

lymphocytic activity but are no substrates of ADA and can therefore not be

neutralised by deamination (Carson et al., 1980). Of these analogs, CdA had the

most favourable therapeutic ratio with an ID50 (% inhibition of growth) of 0,003

µM, determined on the T-lymphoblastoid cell line CCRF-CEM. FdA had an ID50 of

0,15 µM (Carson et al., 1980; Parker et al., 1998). From experiments with dCK-

deficient lymphoblasts, it was clear that activation by dCK is the main route for

cytotoxicity. When dCK-deficient lymphoblasts were used, the ID50 increased a

1000 fold and a 100 fold for CdA and FdA respectively. Whereas deficiency of

deoxyadenosine kinase led to a 10 fold increase of the ID50 of CdA and had no

effect on the ID50 of FdA. Furthermore, cytotoxicity of CdA and FdA was

completely blocked by addition of deoxycytidine.

The cytotoxicity involves several mechanisms. Like dATP, they inhibit

ribonucleotide reductase, hereby reducing the pool of deoxynucleotide

triphosphates for DNA synthesis and enhancing their own cytotoxicity by self-

potentiation. They also inhibit DNA- and RNA polymerase after incorporation into

DNA and RNA respectively, hereby terminating chain elongation (Parker et al.,

1988). However, neither mechanism can account for the remarkable cytotoxicity

in resting cells where both ribonucleotide reductase activity and DNA synthesis

are extremely low (Carson et al., 1983). The most likely explanation is the

interference of CdA and FdA with DNA repair (Pettitt et al., 2000) and activation

of the caspase 9/caspase 3 death pathway by interaction with the pro-apoptotic

factor Apaf1 (Genini et al., 2000).

State of the art: a nanoreactor coming to life

55

Figure 7: metabolic pathways of 2-fluoro-2’-deoxyadenosine and 2-fluoroadenine

CdA, also known as Cladribine, is used as a therapeutic agent to treat chronic

lymphocytic leukaemia and hairy-cell leukaemia (Carrera et al., 1991). It was

discovered by Bontemps et al. that one of the major catabolites in plasma and

urine of CdA, 2-chloroadenine (CAde), is also cytotoxic (Bontemps et al., 2000).

Breakdown of CdA to CAde is probably due to acid hydrolysis in the stomach,

degradation by gut bacterial phosphorylases and by phosphorolytic cleavage in

the liver. The initial phosphorylation of CAde is catalysed by the adenine

phosphoribosyltransferase (APRT) since it is inhibited in the presence of adenine.

Further phosphorylation to 2-chloro-ATP is catalysed by adenylate kinase and

nucleoside diphosphate kinase (Bontemps et al., 2000). An identical metabolic

pathway was found for FdA (Parker et al., 1998) (Figure 7).

Although 2-chloro-ATP and 2-fluoro-ATP, like 2-chloro-dATP induce apoptosis,

this is done by a different pathway (Barbieri et al., 1998). Parker and co-workers

showed that F-ATP inhibits protein, DNA and RNA synthesis. The effect on DNA

synthesis is probably the result of a shortage of enzymes involved in DNA

synthesis. This means that 2-fluoroadenine is also cytotoxic to non proliferating

cells which was confirmed by Parker et al. using non proliferating CCRF-CEM

cells, MRC-5 cells (human lung fibroblasts) and Balb-3T3 cells (mouse embryonic

cells) (Parker et al., 1998).

In conclusion, the nucleoside analogs CdA and FdA are only toxic to

lymphocytes due to the necessary activation by deoxycitidine kinase. In contrast,

2-fluoro-2’-deoxyadenosine

2-fluoro-dAMPATP

ADPdCK

2-fluoro-dADPATP

ADPN-MPK

2-fluoro-dATPNTP

NDPN-DPK

2-fluoroadenine

PRibose-1P

E.coli PNP

2-fluoro-AMPPRPP

PPiAPRT

2-fluoro-ADPATP

ADPAdK

2-fluoro-ATPNTP

NDPN-DPK

APRT = adenine phosphoribosyltransferase

dCK = deoxycytidine kinase

N-MPK = nucleosidemonophosphate kinase

N-DPK = nucleoside diphosphatekinase

AdK = adenylate kinase

PNP = purine nucleosidephosphorylase

2-fluoro-2’-deoxyadenosine

2-fluoro-dAMPATP

ADPdCK

2-fluoro-dADPATP

ADPN-MPK

2-fluoro-dATPNTP

NDPN-DPK

2-fluoroadenine

PRibose-1P

E.coli PNP

2-fluoro-AMPPRPP

PPiAPRT

2-fluoro-ADPATP

ADPAdK

2-fluoro-ATPNTP

NDPN-DPK

APRT = adenine phosphoribosyltransferase

dCK = deoxycytidine kinase

N-MPK = nucleosidemonophosphate kinase

N-DPK = nucleoside diphosphatekinase

AdK = adenylate kinase

PNP = purine nucleosidephosphorylase

Chapter 2

56

the nucleobase analogs 2-chloroadenine and 2-fluoroadenine are toxic to other

cell types due to a different activation pathway. Another advantage is the

cytotoxicity of both 2-chloroadenine and 2-fluoroadenine on resting cells. Taken

together, this makes 2-chloro-2’-deoxyadenosine and 2-fluoro-2’-

deoxyadenosine good prodrug candidates to treat non-lymphocytic tumors.

For instance, as shown by Parker et al., 2-fluoro-2’-deoxyadenosine and

another deoxyadenosine analog 6-methylpurine deoxyribose (6-MPdR) are good

substrates for E. coli purine nucleoside phosphorylase (PNP) that cleaves them

into 2-fluoroadenine or 6-methylpurine and ribose-1-phosphate (Parker et al.,

2003) (Figure 7). In contrast, they are not cleaved by human PNP because it

cannot cleave adenosine and adenosine analogs. Parker et al. successfully

transfected D54 gliomas with the gene encoding E.coli PNP. These transfected

tumor cells were subsequently injected into the flanks of nude mice (NCr-nu) and

then treated intraperitoneally with FdA and 6-MPdR. This resulted in excellent

antitumor activity for both nucleoside analogs. 6-MPdR is a very good prodrug

candidate to be used in suicide gene therapy with E. coli PNP since its nucleobase

6-methylpurine has high bystander activity. This means that 6-methylpurine can

diffuse to neighbouring cells that do not express E. coli PNP. Gadi et al. showed

that less than 1% of PNP expressing cells is sufficient to kill an entire cell

population in vitro owing to the membrane permeability of 6-methylpurine (Gadi

et al., 2000). Parker et al. found that also in vivo tumors in which only 20 % of

the cells express E.coli PNP are destroyed by FdA and 6-MPdR due to this

bystander activity (Parker et al., 2003).

Although CdA was not tested as a prodrug in combination with E. coli PNP,

Bzowska and Kazimierczuk, showed that 2-chloro-2’-deoxyadenosine can also be

efficiently cleaved by E.coli PNP into 2-chloroadenine and ribose monophosphate

(Bzowska and Kazimierczuk, 1995).

State of the art: a nanoreactor coming to life

57

2.5.2. 6-Thioguanosine

Although little is known about the cytotoxic effects of 6-thioguanosine, the

nucleobase 6-thioguanine is a well known cytotoxic agent. It has been used to

treat acute lymphoblastic leukaemia (ALL) over the last 45 years. It is also used

to treat inflammatory bowel disease and autoimmune conditions and to reduce

organ rejection after transplantation (Al Hadithy et al., 2005). Before exerting its

cytotoxicity, 6-thioguanine needs to be activated by hypoxanthine-guanine

phosphoribosyl transferase (HGPRT) to form 6-thioguanosine monophosphate

(TGMP). TGMP can then further be metabolised to form 6-thioguanosine

triphosphate (TGTP) by guanylate kinase and nucleoside diphosphate kinase or it

can be metabolised to form 6-thio-2’-deoxyguanosine triphosphate (TdGTP) by

nucleoside diphosphate reductase and nucleoside diphosphate kinase (Figure 8).

Although TGMP is a substrate of guanylate kinase, the reaction velocity for this

substrate is only 0.04% of the velocity of the natural substrate GMP. This leads

to an accumulation of TGMP which inhibits de novo purine synthesis through

negative feedback inhibition of the enzyme phosphoribosyl amine synthetase

(McCollister et al., 1964) and progressive irreversible inhibition of inosine 5’-

phosphate dehydrogenase (Hampton, 1963). Since the binding affinity to

guanylate kinase of TGMP is comparable to that of GMP, TGMP furthermore acts

as a competitive inhibitor of guanylate kinase with a Ki of 60µM (Miech et al.,

1969). This inhibition leads to a drastic decrease in GDP and GTP concentrations.

Since GDP and GTP are required for nucleic acid synthesis and function as

coenzymes of vital enzymatic reactions, this inhibition also contributes to the

cytotoxic action of 6-thioguanine.

Another key event in the toxicity of 6-thioguanine is the incorporation of

TdGTP into DNA. Interestingly, highly lethal concentrations of 6-thioguanine do

not immediately interfere with cell growth or cell cycle progression but only after

completion of the cell cycle in which TdGTP was incorporated into the DNA. The

cells treated with 6-thioguanine are predominantly arrested in the G2 phase of

the cell cycle (Wotring and Roti Roti, 1980). Maybaum and Mandel confirmed

that this was the result of chromatid disruption (Maybaum and Mandel, 1983).

Chapter 2

58

Figure 8: Metabolic pathways of 6-thioguanine

2.6 Targeting of nanoreactors

2.6.1. Passive targeting

Targeting of drugs can be either active or passive. In 1986, Maeda and

Matsumura introduced the concept of passive targeting to solid tumors by means

of the enhanced permeability and retention effect (EPR) (Matsumura and Maeda,

1986). It is well know that the vascular permeability of tumors is enhanced due

to the action of secreted factors such as kinin. This allows macromolecules to

diffuse from the bloodstream into the tumor interstitium. Furthermore, the

lymphatic drainage system is impaired so that macromolecules are retained in

the interstitium for a prolonged time (Jain, 1987). Particles ranging from 10 to

500 nm in size are able to extravasate through these leaky blood vessels (Figure

9) and remain there for a longer period. A perfect example of this passive

targeting is pegylated lyposomal doxorubicine (Doxil®) which is a long term

circulating carrier (t1/2 ~ 3 days) that is accumulated in tumor tissue (Green and

Rose, 2006). Doxil is used to treat relapsing ovarian cancer and AIDS-related

Kaposi’s sarcoma. Compared to conventional doxorubicin, plasma levels of Doxil

are higher since doxorubicin is encapsulated in liposomes and remains in the

blood for a prolonged time. Furthermore, clearance from the bloodstream is 9

6-thioguanine

6-thioguanosine monophosphate

6-thioguanosine diphosphate

6-thioguanosine triphosphate

6-thio-deoxyguanosine diphosphate

6-thio-deoxyguanosine triphosphate

PRPP

PPi

HGPRT

ATP

ADPGK

N-DPKNTP

NDP

Thioredoxin 2SH

Thioredoxin S-S

N-DPR

NTP

NDPN-DPK

HGPRT = hypoxanthine-guanosinephosphoribosyltransferase

GK = Guanylate kinase

N-DPK = nucleosidediphosphatekinase

N-DPR = nucleosidediphosphatereductase6-thioguanine

6-thioguanosine monophosphate

6-thioguanosine diphosphate

6-thioguanosine triphosphate

6-thio-deoxyguanosine diphosphate

6-thio-deoxyguanosine triphosphate

PRPP

PPi

HGPRT

ATP

ADPGK

N-DPKNTP

NDP

Thioredoxin 2SH

Thioredoxin S-S

N-DPR

NTP

NDPN-DPK

6-thioguanine

6-thioguanosine monophosphate

6-thioguanosine diphosphate

6-thioguanosine triphosphate

6-thio-deoxyguanosine diphosphate

6-thio-deoxyguanosine triphosphate

PRPP

PPi

HGPRT

ATP

ADPGK

N-DPKNTP

NDP

Thioredoxin 2SH

Thioredoxin S-S

N-DPR

NTP

NDPN-DPK

HGPRT = hypoxanthine-guanosinephosphoribosyltransferase

GK = Guanylate kinase

N-DPK = nucleosidediphosphatekinase

N-DPR = nucleosidediphosphatereductase

State of the art: a nanoreactor coming to life

59

times less than compared to doxorubicin alone. This is the effect of pegylation of

the liposomes which decreases the uptake of the liposomes by macrophages,

predominantly kuppfer cells. Finally, tissue distribution of Doxil is 25 times less

than compared to doxorubicin alone. This means that doxil is predominantly

accumulated in tumor tissue and not in healthy tissue due to the EPR effect.

Doxorubicin is known to exhibit cardiotoxicity as an effect of the accumulation in

the heart. This cardiotoxicity is remarkably reduced when doxil formulations are

used (Safra et al., 2000).

Figure 9: Schematic overview of the enhanced permeation and retention effect (EPR). Drug molecules are represented as dark blue dots.

2.6.2. Active targeting with single domain antibodies

The first idea of active drug targeting called the ‘magic bullet’ approach was

proposed by Ehrlich in the nineteenth century. The magic bullet is composed of

two units, the drug unit and the targeting unit which is responsible for delivering

the drug to the designated site. The following substances can be used as

targeting moieties: antibodies and their fragments, lectins, lipoproteins,

hormones, saccharides, peptides, polynucleotides, folate… Monoclonal antibodies

are the most frequently used. Research shows that several body compartments

and pathologies can be targeted including components of the cardiovascular

system, the reticulo-endothelial system, lymphatic system, tumors, infarcts,

inflammations, infections and transplants. In the early years of targeted drug

therapy, the drug was directly linked to the antibody (= immunotoxin). Later on

Chapter 2

60

soluble or insoluble carriers where loaded with multiple active molecules and

conjugated to the targeting unit, thus delivering a high amount of drug per

targeting unit.

In this project, the antigen binding domain of single chain antibodies derived

from camelids is used for targeting purposes. These naturally occurring single

chain antibodies devoid of a light chain were discovered by Hamers et al. in 1993

(Hamers-Casterman et al., 1993) (Figure 10). These so called heavy-chain

antibodies (HCAb) are present in the serum of camels, dromedaries and llamas.

It should be emphasized that they posses both the conventional heterotetrameric

antibodies common to all vertebrates as well as the homodimeric heavy chain

antibodies. The relative proportion seems to vary with an average of about 50 %

of each type. Whether a conventional antibody or HCAb is raised, depends on the

antigen.

Figure 10: shematic representation of a conventional monoclonal antibody and a heavy chain antibody

HCAbs have a lower molecular weight of 100 kDa compared to conventional

antibodies because they lack the CH1 domain (Hamers-Casterman et al., 1993).

To structurally compensate for the absence of CH1, HCAbs have a longer hinge

between the variable domain called VHH and CH2. Up until now, 3 different types

of hinges have been found: a short hinge (11 amino acids) and a long hinge (35

amino acids) in dromedary and a long hinge (29 amino acids) in llama (Hamers-

Casterman et al., 1993; Vu et al., 1997). Both longer hinges contain a repeated

Pro-Xaa motif that is assumed to adopt a helical conformation that functions as a

rigid rod-like spacer (Evans et al., 1986; Roditi et al., 1989).

VL

CH3

CH1

CH2

VH

CL

CH1CL VH

VL

CH1VH

CL

VL

F(ab’) 50 kDa F(ab’)2 100kDa

Monoclonal antibody

CH3

VHH

VHH

CH2

Single chain camel antibody

Nanobody 15 kDa

VLVH

scFv 25 kDa

VL

CH3

CH1

CH2

VH

CL

CH3

CH1

CH2

VH

CL

CH1CL VH

VL

CH1CL VH

VL

CH1VH

CL

VL

F(ab’) 50 kDa F(ab’)2 100kDa

Monoclonal antibody

CH3

VHH

VHH

CH2

Single chain camel antibody

Nanobody 15 kDa

VLVH

scFv 25 kDa

State of the art: a nanoreactor coming to life

61

Because there is no light chain in HCAbs, their antigen binding site is

composed of a single domain, VHH, compared to the VH-VL paired antigen

binding site of conventional antibodies. Sequence analysis of various llama VHH

domains revealed that they all belong to the VHIII family (Muyldermans et al.,

1994; Vu et al., 1997). Although they belong to the same family, their sequences

differ more from each other than expected for members of the same family.

Conventionally, members of the same family share a nucleotide sequence

identity above 80 % whereas VHH sequence identities range from 58 to 90%.

This lower identity probably reflects a high rate of somatic mutation which can

lead to increased diversity.

To compensate for the absence of the light chain, adaptations in the VHH

domain are necessary. For instance, there are four important hydrophobic to

hydrophilic amino acid substitutions in the framework 2 region of the VHH

domain: Leu11Ser, Val37Phe, Gly44Glu, Leu45Arg/Cys and Trp47Gly (Figure

11). These residues normally interact with the VL domain. Consequently,

substituting them to more hydrophilic residues will impair the ability of the VHH

to form heterodimers with a VL domain and increase the solubility and stability of

the HCAbs (Dumoulin et al., 2002). This was confirmed by Davies and

Riechmann who analysed the effect of three of these mutations (G44E, L45R and

W47G) on a human VHIII. Non mutated VH domains aggregated at a protein

concentration above 1 mg/ml whereas the aggregation of the mutated VH was

significantly reduced (Davies and Riechmann, 1994).

Besides these VH/VHH hallmark substitutions in the framework regions, also

the hypervariable regions differ from each other. First, the average CDR3 region

of the VHH domain is about twice as long as in the VH domain. This results in a

higher antigen binding surface, hereby increasing the repertoire. Secondly, the

CDR1 region is elongated with 4 residues (27-30). In conventional antibodies

these 4 residues form the loop connecting the two β-sheets of the

immunoglobulin fold. Their hypervariability in VHHs shows their involvement in

antigen binding. Finally, the long CDR3 region is connected to the CDR1 region

via a disulfide bridge. This disulphide bond restricts the conformational flexibility

of the long CDR3 loop in the absence of a bound antigen. Immobilising the loop

in this manner will minimise the entropic penalty upon binding of the antigen.

Chapter 2

62

This long CDR3 loop has another interesting advantage. Lauwereys et al.

found that a large amount of HCAbs that were isolated after immunisation with

enzymes, are competitive enzyme inhibitors that bind to the active site of the

enzymes (Lauwereys et al., 1998). This was further investigated for HCAbs that

bind to the active site of lysozyme. This enzyme inhibition is due to the convex

shape of the VHH, predominantly formed by the CDR3 loop that protrudes in the

convex substrate binding pocket of lysozyme (De Genst et al., 2006). In

contrast, conventional antibodies that are competitive enzyme inhibitors are rare

because the antigen-binding surface is either flat or concave (Padlan, 1996).

Therefore it is clear that HCAbs recognize epitopes that are currently out of reach

for conventional antibodies.

Figure 11: schematic representation of VH and VHH sequences. Hallmark amino acid substitutions in the framework regions are shown.

Since VHHs are single domain antigen binding molecules, their isolation and

expression is more straight forward and easier as compared to the smallest

antigen binding fragment, ScFv, of conventional antibodies. In case of the ScFv

fragments which are a combination of a VH and VL domain, first a library of

cloned VH and VL domains is generated after immunisation. Since there are

several VH and VL gene families, several sets of primers are necessary to clone

the whole repertoire. Subsequently binders to a specific antigen are selected

from random combinations of these VH and VL domains. This means that large

libraries need to be screened in order to isolate the native combinations and that

FR1 FR2 FR3

CDR1 CDR3CDR2

S11 F37E44R45G47

VHH

FR1 FR3FR2

CDR1 CDR2 CDR3

L11 V37G44L45W47

VH

FR1 FR2 FR3

CDR1 CDR3CDR2

S11 F37E44R45G47

VHH

FR1 FR2 FR3

CDR1 CDR3CDR2

FR1 FR2 FR3

CDR1 CDR3CDR2

S11 F37E44R45G47

VHH

FR1 FR3FR2

CDR1 CDR2 CDR3

L11 V37G44L45W47

VH

FR1 FR3FR2

CDR1 CDR2 CDR3

FR1 FR3FR2

CDR1 CDR2 CDR3

L11 V37G44L45W47

VH

State of the art: a nanoreactor coming to life

63

many binders with lower affinity and stability will be obtained. In case of the

VHH, only one library of VHH domains needs to be generated from which binders

are selected via phage display vectors. This significantly reduces the cloning and

screening labour to isolate the native binder. Furthermore, only one set of

primers is needed to clone the whole repertoire because all VHH domains belong

to the same family. Finally, expression of the VHHs in E.coli generally results in

5-10 mg/l of VHH which is about 10 times higher than the production of ScFvs

(Arbabi Ghahroudi et al., 1997).

VHHs isolated from immunised libraries are highly specific for the target

antigen with Kd constants in the nanomolar range (Muyldermans and Lauwereys,

1999), which is similar to the affinity of most scFv, and do not cross-react with

other non-related antigens. This was demonstrated by cAb-Lys3 that was

isolated after immunisation with hen egg-white lysozyme. cAb-Lys3 binds henn

lysozyme as well as turkey lysozyme that only differs by 6 amino acid residues

but not to human lysozyme which is more distantly related (Arbabi Ghahroudi et

al., 1997).

In conclusion, VHH are distinguished from the scFv by some unique

properties: size, stability, solubility, ease of cloning a library and selecting highly

specific antigen binders, in vivo maturation and high expression levels. Therefore

it is expected that they will perform better than other antibody fragments in

numerous applications such as diagnostics, in vivo imaging, targeting and as

crystallisation aids. Since their discovery in 1993, numerous VHH have been

isolated against different antigens, for multiple purposes as shown in Table 2.

Chapter 2

64

Table 2: selected examples of VHHs and their applications.

Antigen VHH Application Ref

Variant surface glycoprotein of Trypanosomes

NbAn33 Trypanosome

targeting of Tr-Apol-I (Baral et al., 2006)

Prostate specific antigen cAbPSA-N7 and

cAbPSA-N50 Prostate cancer

diagnostics (Saerens et al., 2005)

Carcinoembryogenic antigen

cAbCEA-5 Tumor targeting

(Cortez-Retamozo

et al., 2004)

Human lysozyme cAbHUL-6 Inhibition of amyloid

fibril formation

(Dumoulin et al., 2003)

MazE cAbmaz1 Crystallisation aid (Loris et al.,

2003)

Neisseria meningitidis LPS VHH5G Inhibition of LPS mediated sepsis

(El Khattabi et al., 2006)

Epidermal growth factor receptor VIII

ORBI-83 VHH Tumor targeting (Omidfar et al., 2007)

caffeine VSA2 Caffeine detection (Ladenson

et al., 2006)

State of the art: a nanoreactor coming to life

65

2.7 References

Al Hadithy, A.F., de Boer, N.K., Derijks, L.J., Escher, J.C., Mulder, C.J. and Brouwers, J.R. (2005) Thiopurines in inflammatory bowel disease: pharmacogenetics, therapeutic drug monitoring and clinical recommendations. Dig Liver Dis, 37, 282-297.

Alexandridis, P., Olsson, U. and Lindman, B. (1998) Arecord nine different phases (four cubic, two hexagonal, and one lamellar liquid crystalline and two micellar solutions) in a ternary isothermal system of an amphiphilic block copolymer and selective solvents (water and oil). langmuir, 14, 2627-2638.

Arbabi Ghahroudi, M., Desmyter, A., Wyns, L., Hamers, R. and Muyldermans, S. (1997) Selection and identification of single domain antibody fragments from camel heavy-chain antibodies. FEBS Lett, 414, 521-526.

Baral, T.N., Magez, S., Stijlemans, B., Conrath, K., Vanhollebeke, B., Pays, E., Muyldermans, S. and De Baetselier, P. (2006) Experimental therapy of African trypanosomiasis with a nanobody-conjugated human trypanolytic factor. Nat Med, 12, 580-584.

Barbieri, D., Abbracchio, M.P., Salvioli, S., Monti, D., Cossarizza, A., Ceruti, S., Brambilla, R., Cattabeni, F., Jacobson, K.A. and Franceschi, C. (1998) Apoptosis by 2-chloro-2'-deoxy-adenosine and 2-chloro-adenosine in human peripheral blood mononuclear cells. Neurochem Int, 32, 493-504.

Battaglia, G. and Ryan, A.J. (2005) Bilayers and interdigitation in block copolymer vesicles. J Am Chem Soc., 127, 8757-8764.

Bauer, K., Struyve, M., Bosch, D., Benz, R. and Tommassen, J. (1989) One single lysine residue is responsible for the special interaction between polyphosphate and the outer membrane porin PhoE of Escherichia coli. J Biol Chem, 264, 16393-16398.

Benz, R., Schmid, A., Maier, C. and Bremer, E. (1988) Characterization of the nucleoside-binding site inside the Tsx channel of Escherichia coli outer membrane. Eur J Biochem, 176, 699-705.

Benz, R., Schmid, A., Van der Ley, P. and Tommassen, J. (1989) Molecular basis of porin selectivity: membrane experiments with OmpC- PhoE and OmpF-PhoE hybrid proteins of Escherichia coli K-12. Biochim Biophys Acta, 981, 8-14.

Berg, H. (1993) Random walks in biology. Princeton university press. Birch, R.G., Pemberton, J.M. and Basnayake, W.V. (1990) Stable albicidin resistance in

Escherichia coli involves an altered outer-membrane nucleoside uptake system. J Gen Microbiol, 136, 51-58.

Bontemps, F., Delacauw, A., Cardoen, S., Van Den Neste, E. and Van Den Berghe, G. (2000) Metabolism and cytotoxic effects of 2-chloroadenine, the major catabolite of 2-chloro-2'-deoxyadenosine. Biochem Pharmacol, 59, 1237-1243.

Buehler, L.K., Kusumoto, S., Zhang, H. and Rosenbusch, J.P. (1991) Plasticity of Escherichia coli porin channels. Dependence of their conductance on strain and lipid environment. J Biol Chem, 266, 24446-24450.

Bzowska, A. and Kazimierczuk, Z. (1995) 2-Chloro-2'-deoxyadenosine (cladribine) and its analogues are good substrates and potent selective inhibitors of Escherichia coli purine-nucleoside phosphorylase. Eur J Biochem, 233, 886-890.

Carrera, C.J., Piro, L.D., Saven, A., Beutler, E., Terai, C. and Carson, D.A. (1991) 2-Chlorodeoxyadenosine chemotherapy triggers programmed cell death in normal and malignant lymphocytes. Adv Exp Med Biol, 309A, 15-18.

Carson, D.A., Wasson, D.B., Kaye, J., Ullman, B., Martin, D.W., Jr., Robins, R.K. and Montgomery, J.A. (1980) Deoxycytidine kinase-mediated toxicity of deoxyadenosine analogs toward malignant human lymphoblasts in vitro and toward murine L1210 leukemia in vivo. Proc Natl Acad Sci U S A, 77, 6865-6869.

Chapter 2

66

Carson, D.A., Wasson, D.B., Taetle, R. and Yu, A. (1983) Specific toxicity of 2-chlorodeoxyadenosine toward resting and proliferating human lymphocytes. Blood, 62, 737-743.

Cortez-Retamozo, V., Backmann, N., Senter, P.D., Wernery, U., De Baetselier, P., Muyldermans, S. and Revets, H. (2004) Efficient cancer therapy with a nanobody-based conjugate. Cancer Res, 64, 2853-2857.

Cowan, S.W., Schirmer, T., Rummel, G., Steiert, M., Ghosh, R., Pauptit, R.A., Jansonius, J.N. and Rosenbusch, J.P. (1992) Crystal structures explain functional properties of two E. coli porins. Nature, 358, 727-733.

Cui, L., Rajasekariah, G.R. and Martin, S.K. (2001) A nonspecific nucleoside hydrolase from Leishmania donovani: implications for purine salvage by the parasite. Gene, 280, 153-162.

Davies, J. and Riechmann, L. (1994) 'Camelising' human antibody fragments: NMR studies on VH domains. FEBS Lett, 339, 285-290.

De Genst, E., Silence, K., Decanniere, K., Conrath, K., Loris, R., Kinne, J., Muyldermans, S. and Wyns, L. (2006) Molecular basis for the preferential cleft recognition by dromedary heavy-chain antibodies. Proc Natl Acad Sci U S A, 103, 4586-4591.

Delcour, A.H. (1997) Function and modulation of bacterial porins: insights from electrophysiology. FEMS Microbiol Lett, 151, 115-123.

Dumoulin, M., Conrath, K., Van Meirhaeghe, A., Meersman, F., Heremans, K., Frenken, L.G., Muyldermans, S., Wyns, L. and Matagne, A. (2002) Single-domain antibody fragments with high conformational stability. Protein Sci, 11, 500-515.

Dumoulin, M., Last, A.M., Desmyter, A., Decanniere, K., Canet, D., Larsson, G., Spencer, A., Archer, D.B., Sasse, J., Muyldermans, S., Wyns, L., Redfield, C., Matagne, A., Robinson, C.V. and Dobson, C.M. (2003) A camelid antibody fragment inhibits the formation of amyloid fibrils by human lysozyme. Nature, 424, 783-788.

El Khattabi, M., Adams, H., Heezius, E., Hermans, P., Detmers, F., Maassen, B., van der Ley, P., Tommassen, J., Verrips, T. and Stam, J. (2006) Llama single-chain antibody that blocks lipopolysaccharide binding and signaling: prospects for therapeutic applications. Clin Vaccine Immunol., 13, 1079-1086. Epub 2006 Aug 1023.

Evans, J.S., Levine, B.A., Trayer, I.P., Dorman, C.J. and Higgins, C.F. (1986) Sequence-imposed structural constraints in the TonB protein of E. coli. FEBS Lett, 208, 211-216.

Forster, S. and Plantenberg, T. (2002) From self-organizing polymers to nanohybrid and biomaterials. Angew Chem Int Ed Engl, 41, 689-714.

Gadi, V.K., Alexander, S.D., Kudlow, J.E., Allan, P., Parker, W.B. and Sorscher, E.J. (2000) In vivo sensitization of ovarian tumors to chemotherapy by expression of E. coli purine nucleoside phosphorylase in a small fraction of cells. Gene Ther, 7, 1738-1743.

Genini, D., Adachi, S., Chao, Q., Rose, D.W., Carrera, C.J., Cottam, H.B., Carson, D.A. and Leoni, L.M. (2000) Deoxyadenosine analogs induce programmed cell death in chronic lymphocytic leukemia cells by damaging the DNA and by directly affecting the mitochondria. Blood, 96, 3537-3543.

Graff, A., Sauer, M., Van Gelder, P. and Meier, W. (2002) Virus-assisted loading of polymer nanocontainer. Proc Natl Acad Sci U S A, 99, 5064-5068. Epub 2002 Mar 5026.

Green, A.E. and Rose, P.G. (2006) Pegylated liposomal doxorubicin in ovarian cancer. Int J Nanomedicine, 1, 229-239.

Hamers-Casterman, C., Atarhouch, T., Muyldermans, S., Robinson, G., Hamers, C., Songa, E.B., Bendahman, N. and Hamers, R. (1993) Naturally occurring antibodies devoid of light chains. Nature, 363, 446-448.

Hampton, A. (1963) Reactions of Ribonucleotide Derivatives of Purine Analogues at the Catalytic Site of Inosine 5'-Phosphate Dehydrogenase. J Biol Chem, 238, 3068-3074.

Heuzenroeder, M.W. and Reeves, P. (1981) The tsx protein of Escherichia coli can act as a pore for amino acids. J Bacteriol, 147, 1113-1116.

State of the art: a nanoreactor coming to life

67

Jain, R.K. (1987) Transport of molecules in the tumor interstitium: a review. Cancer Res, 47, 3039-3051.

Krieger-Brauer, H.J. and Braun, V. (1980) Functions related to the receptor protein specified by the tsx gene of Escherichia coli. Arch Microbiol, 124, 233-242.

Kurtz, J.E., Exinger, F., Erbs, P. and Jund, R. (2002) The URH1 uridine ribohydrolase of Saccharomyces cerevisiae. Curr Genet, 41, 132-141.

Ladenson, R.C., Crimmins, D.L., Landt, Y. and Ladenson, J.H. (2006) Isolation and characterization of a thermally stable recombinant anti-caffeine heavy-chain antibody fragment. Anal Chem, 78, 4501-4508.

Lauwereys, M., Arbabi Ghahroudi, M., Desmyter, A., Kinne, J., Holzer, W., De Genst, E., Wyns, L. and Muyldermans, S. (1998) Potent enzyme inhibitors derived from dromedary heavy-chain antibodies. Embo J, 17, 3512-3520.

Loris, R., Marianovsky, I., Lah, J., Laeremans, T., Engelberg-Kulka, H., Glaser, G., Muyldermans, S. and Wyns, L. (2003) Crystal structure of the intrinsically flexible addiction antidote MazE. J Biol Chem, 278, 28252-28257.

Luckey, M. and Nikaido, H. (1980) Specificity of diffusion channels produced by lambda phage receptor protein of Escherichia coli. Proc Natl Acad Sci U S A, 77, 167-171.

Maier, C., Bremer, E., Schmid, A. and Benz, R. (1988) Pore-forming activity of the Tsx protein from the outer membrane of Escherichia coli. Demonstration of a nucleoside-specific binding site. J Biol Chem, 263, 2493-2499.

Matsumura, Y. and Maeda, H. (1986) A new concept for macromolecular therapeutics in cancer chemotherapy: mechanism of tumoritropic accumulation of proteins and the antitumor agent smancs. Cancer Res, 46, 6387-6392.

Maybaum, J. and Mandel, H.G. (1983) Unilateral chromatid damage: a new basis for 6-thioguanine cytotoxicity. Cancer Res, 43, 3852-3856.

McCollister, R.J., Gilbert, W.R., Jr., Ashton, D.M. and Wyngaarden, J.B. (1964) Pseudofeedback Inhibition of Purine Synthesis by 6-Mercaptopurine Ribonucleotide and Other Purine Analogues. J Biol Chem, 239, 1560-1563.

McKeown, M., Kahn, M. and Hanawalt, P. (1976) Thymidine uptake and utilization in Escherichia coli: a new gene controlling nucleoside transport. J Bacteriol, 126, 814-822.

Miech, R.P., York, R. and Parks, R.E., Jr. (1969) Adenosine triphosphate-guanosine 5'-phosphate phosphotransferase. II. Inhibition by 6-thioguanosine 5'-phosphate of the enzyme isolated from hog brain and sarcoma 180 ascites cells. Mol Pharmacol, 5, 30-37.

Muyldermans, S., Atarhouch, T., Saldanha, J., Barbosa, J.A. and Hamers, R. (1994) Sequence and structure of VH domain from naturally occurring camel heavy chain immunoglobulins lacking light chains. Protein Eng, 7, 1129-1135.

Muyldermans, S. and Lauwereys, M. (1999) Unique single-domain antigen binding fragments derived from naturally occurring camel heavy-chain antibodies. J Mol Recognit, 12, 131-140.

Nallani, M., Benito, S., Onaca, O., Graff, A., Lindemann, M., Winterhalter, M., Meier, W. and Schwaneberg, U. (2006) A nanocompartment system (Synthosome) designed for biotechnological applications. J Biotechnol, 123, 50-59. Epub 2005 Dec 2020.

Nardin, C., Hirt, T., Leukel, J. and Meier, W. (2000a) polymerized ABA triblock copolymer vesicles. langmuir, 16, 1035-1041.

Nardin, C., Thoeni, S., Widmer, J., Winterhalter, M. and Meier, W. (2000b) Nanoreactors based on (polymerized) ABA-triblock copolymer vesicles. Chem Comm, 1433-1434.

Nardin, C., Widmer, J., Winterhalter, M. and Meier, W. (2001) Amphiphilic block copolymer nanocontainers as bioreactors. Eur. Phys. J. E, 4, 403-410.

Nardin, C., Winterhalter, M., Meier, W. (2000) Giant Free-Standing ABA Triblock Copolymer Membranes. Langmuir, 16, 7708-7712.

Ogawa, J., Takeda, S., Xie, S.X., Hatanaka, H., Ashikari, T., Amachi, T. and Shimizu, S. (2001) Purification, characterization, and gene cloning of purine nucleosidase from Ochrobactrum anthropi. Appl Environ Microbiol, 67, 1783-1787.

Chapter 2

68

Omidfar, K., Rasaee, M.J., Kashanian, S., Paknejad, M. and Bathaie, Z. (2007) Studies of thermostability in Camelus bactrianus (Bactrian camel) single-domain antibody specific for the mutant epidermal-growth-factor receptor expressed by Pichia. Biotechnol Appl Biochem, 46, 41-49.

Padlan, E.A. (1996) X-ray crystallography of antibodies. Adv Protein Chem, 49, 57-133. Parker, W.B., Allan, P.W., Hassan, A.E., Secrist, J.A., 3rd, Sorscher, E.J. and Waud, W.R.

(2003) Antitumor activity of 2-fluoro-2'-deoxyadenosine against tumors that express Escherichia coli purine nucleoside phosphorylase. Cancer Gene Ther, 10, 23-29.

Parker, W.B., Allan, P.W., Shaddix, S.C., Rose, L.M., Speegle, H.F., Gillespie, G.Y. and Bennett, L.L., Jr. (1998) Metabolism and metabolic actions of 6-methylpurine and 2-fluoroadenine in human cells. Biochem Pharmacol, 55, 1673-1681.

Parker, W.B., Bapat, A.R., Shen, J.X., Townsend, A.J. and Cheng, Y.C. (1988) Interaction of 2-halogenated dATP analogs (F, Cl, and Br) with human DNA polymerases, DNA primase, and ribonucleotide reductase. Mol Pharmacol, 34, 485-491.

Pata, V. and Dan, N. (2003) The effect of chain length on protein solubilization in polymer-based vesicles (polymersomes). Biophys J, 85, 2111-2118.

Pelle, R., Schramm, V.L. and Parkin, D.W. (1998) Molecular cloning and expression of a purine-specific N-ribohydrolase from Trypanosoma brucei brucei. Sequence, expression, and molecular analysis. J Biol Chem, 273, 2118-2126.

Petersen, C. and Moller, L.B. (2001) The RihA, RihB, and RihC ribonucleoside hydrolases of Escherichia coli. Substrate specificity, gene expression, and regulation. J Biol Chem, 276, 884-894.

Pettitt, A.R., Sherrington, P.D. and Cawley, J.C. (2000) Role of poly(ADP-ribosyl)ation in the killing of chronic lymphocytic leukemia cells by purine analogues. Cancer Res, 60, 4187-4193.

Phale, P.S., Philippsen, A., Kiefhaber, T., Koebnik, R., Phale, V.P., Schirmer, T. and Rosenbusch, J.P. (1998) Stability of trimeric OmpF porin: the contributions of the latching loop L2. Biochemistry, 37, 15663-15670.

Phale, P.S., Philippsen, A., Widmer, C., Phale, V.P., Rosenbusch, J.P. and Schirmer, T. (2001) Role of charged residues at the OmpF porin channel constriction probed by mutagenesis and simulation. Biochemistry, 40, 6319-6325.

Ranquin, A. and Van Gelder, P. (2004) Maltoporin: sugar for physics and biology. Res Microbiol, 155, 611-616.

Ribeiro, J.M. and Valenzuela, J.G. (2003) The salivary purine nucleosidase of the mosquito, Aedes aegypti. Insect Biochem Mol Biol, 33, 13-22.

Roditi, I., Schwarz, H., Pearson, T.W., Beecroft, R.P., Liu, M.K., Richardson, J.P., Buhring, H.J., Pleiss, J., Bulow, R., Williams, R.O. and et al. (1989) Procyclin gene expression and loss of the variant surface glycoprotein during differentiation of Trypanosoma brucei. J Cell Biol, 108, 737-746.

Saerens, D., Frederix, F., Reekmans, G., Conrath, K., Jans, K., Brys, L., Huang, L., Bosmans, E., Maes, G., Borghs, G. and Muyldermans, S. (2005) Engineering camel single-domain antibodies and immobilization chemistry for human prostate-specific antigen sensing. Anal Chem, 77, 7547-7555.

Safra, T., Muggia, F., Jeffers, S., Tsao-Wei, D.D., Groshen, S., Lyass, O., Henderson, R., Berry, G. and Gabizon, A. (2000) Pegylated liposomal doxorubicin (doxil): reduced clinical cardiotoxicity in patients reaching or exceeding cumulative doses of 500 mg/m2. Ann Oncol, 11, 1029-1033.

Saint, N., Lou, K.L., Widmer, C., Luckey, M., Schirmer, T. and Rosenbusch, J.P. (1996) Structural and functional characterization of OmpF porin mutants selected for larger pore size. II. Functional characterization. J Biol Chem, 271, 20676-20680.

Samartzidou, H. and Delcour, A.H. (1998) E.coli PhoE porin has an opposite voltage-dependence to the homologous OmpF. Embo J, 17, 93-100.

Stoenescu, R., Graff, A. and Meier, W. (2004) Asymmetric ABC-Triblock Copolymer Membranes Induce a Directed Insertion of Membrane Proteins. Macromol Biosci, 4, 930-935.

State of the art: a nanoreactor coming to life

69

van Alphen, L., Verkleij, A., Leunissen-Bijvelt, J. and Lugtenberg, B. (1978) Architecture of the outer membrane of Escherichia coli. III. Protein-lipopolysaccharide complexes in intramembraneous particles. J Bacteriol, 134, 1089-1098.

Van Gelder, P., de Cock, H. and Tommassen, J. (1996) Assembly of bacterial outer membrane proteins. In von Heijne, G. (ed.), Membrane protein assembly. R G Landes Company, Georgetown,TX, pp. 63-82.

Van Gelder, P., Dumas, F., Bartoldus, I., Saint, N., Prilipov, A., Winterhalter, M., Wang, Y., Philippsen, A., Rosenbusch, J.P. and Schirmer, T. (2002) Sugar transport through maltoporin of Escherichia coli: role of the greasy slide. J Bacteriol, 184, 2994-2999.

Van Gelder, P., Saint, N., Phale, P., Eppens, E.F., Prilipov, A., van Boxtel, R., Rosenbusch, J.P. and Tommassen, J. (1997) Voltage sensing in the PhoE and OmpF outer membrane porins of Escherichia coli: role of charged residues. J Mol Biol, 269, 468-472.

Versees, W., Decanniere, K., Pelle, R., Depoorter, J., Brosens, E., Parkin, D.W. and Steyaert, J. (2001) Structure and function of a novel purine specific nucleoside hydrolase from Trypanosoma vivax. J Mol Biol, 307, 1363-1379.

Versees, W., Loverix, S., Vandemeulebroucke, A., Geerlings, P. and Steyaert, J. (2004) Leaving group activation by aromatic stacking: an alternative to general acid catalysis. J Mol Biol, 338, 1-6.

Versees, W., Van Holsbeke, E., De Vos, S., Decanniere, K., Zegers, I. and Steyaert, J. (2003) Cloning, preliminary characterization and crystallization of nucleoside hydrolases from Caenorhabditis elegans and Campylobacter jejuni. Acta Crystallogr D Biol Crystallogr, 59, 1087-1089.

Vu, K.B., Ghahroudi, M.A., Wyns, L. and Muyldermans, S. (1997) Comparison of llama VH sequences from conventional and heavy chain antibodies. Mol Immunol, 34, 1121-1131.

Woodle, M.C., Engbers, C.M. and Zalipsky, S. (1994) New amphipatic polymer-lipid conjugates forming long-circulating reticuloendothelial system-evading liposomes. Bioconjug Chem, 5, 493-496.

Wotring, L.L. and Roti Roti, J.L. (1980) Thioguanine-induced S and G2 blocks and their significance to the mechanism of cytotoxicity. Cancer Res, 40, 1458-1462.

Ye, J. and Van Den Berg, B. (2004) Crystal structure of the bacterial nucleoside transporter Tsx. EMBO J, 23, 3187-3195. Epub 2004 Jul 3122.

Chapter 3 Encapsulation of therapeutic NH in functionalised nanocapsules

71

Chapter 3: Encapsulation of

therapeutic nucleoside hydrolase

in functionalised nanocapsules

Gerard Huysmans, An Ranquin, Lode Wyns, Jan Steyaert, Patrick Van Gelder

Abstract

Liposomes are introduced as encapsulating carrier for prodrug activating enzymes. Inosine- adenosine- guanosine preferring nucleoside hydrolase of Trypanosoma vivax, a potential prodrug activating enzyme, was encapsulated in functionalized dioleyl-phosphatidylglycerol/egg-phosphatidylglycerol (DOPC/EPG) liposomes. First, transport of nucleosides through general diffusion porins OmpF and PhoE was measured in swelling assays, after which fully functional nanoreactors were developed. Enzyme catalysis of p-nitrophenylriboside, a substrate analogue for nucleoside hydrolases, was significantly higher in permeabilized vesicles than in control vesicles without porins. Residual activity of control vesicles possibly resides in an interaction between the enzyme and the liposomes. This interaction was not of electrostatic nature, since it remained unaffected after the addition of high salt or after perturbation of liposome surface charge and charge density. With these vesicles, we have introduced a new strategy for prodrug therapy, combining the benefits of ADEPT and liposome targeting strategies. Part of this work was published in Journal of Controlled Release 102 (2005) 171–179

Chapter 3

72

Encapsulation of therapeutic NH in functionalised nanocapsules

73

3.1 Introduction

Cancer is a major death cause throughout the world. Current strategies as

radiotherapy and surgery have only gained limited success in the early stages of

the disease and mostly chemotherapy is necessary, because the tumor is only

detected in a later stage (Aghi et al., 2000). Most anticancer drugs are anti-

proliferative agents that display little selectivity, also damaging normal

proliferating cells and thus have a limited therapeutic index. As a consequence,

anticancer drugs are administered in suboptimal conditions, promoting therapy

failure as well as the development of drug resistant cancer cells. Recent

developments in oncogenesis have introduced more selective chemotherapeutics

or lead to direct targeting of anticancer drugs or prodrugs to the tumor site

(Allen, 2002; Torchilin, 2000).

Liposomes can carry a large load of drugs to the tumor site without the need

for covalent linkage between the carrier and the drug. Consequently, antibody

directed liposomes have been proposed as carriers to deliver drugs or prodrugs

locally to the tumor (Gregoriadis, 1995). Subsequent destabilization of the

liposome in the micro-environment of the tumor allows the release of the

therapeutics. Prodrugs can then be metabolized through the physical properties

of the micro-environment or through endogenous enzymes (Denny, 2001).

In combination with endogenous enzymes the use of prodrugs is rather

limited. Therefore intensive research is made into two promising paths to

broaden the range of usable prodrugs (Greco and Dachs, 2001; Senter and

Springer, 2001; Xu and McLeod, 2001): antibody directed enzyme prodrug

therapy or ADEPT and gene-directed enzyme prodrug therapy or GDEPT. Both

strategies are two-step processes, in which firstly an exogenous enzyme is

delivered to the tumor site and in which during the second step this enzyme is

used to locally convert a systematically administered prodrug into a toxic agent.

In ADEPT strategies, the enzyme is targeted to the tumor site in a fusion with a

monoclonal antibody; in GDEPT, the so called “suicide gene” for the enzyme is

brought directly into the cancer cells, mostly through a viral vector. As for

ADEPT, the major obstacle might be the immune response of the host to the

exogenous enzyme (Allen, 2002), hurdles in GDEPT are the membrane

Chapter 3

74

permeability for prodrugs and the lacking of an efficient gene transfer (Springer

and Niculescu-Duvaz, 2000).

With our research objectives, we introduce a complementary strategy, being

therapeutic nanoreactors. By combining the benefits of the above described

therapies, we wish to overcome current problems with ADEPT and GDEPT,

through encapsulating an exogenous enzyme in a non-immunogenic and

permeabilized carrier. As in ADEPT, the reactor will be targeted to the tumor

through a cancer specific monoclonal antibody. Encapsulation of the exogenous

enzyme makes evasion of the host immune system possible and permeabilization

will ensure that prodrugs and hydrolysed metabolites can pass through the

vesicle wall.

Initially, liposome formulations will be developed to encapsulate the enzyme.

Already a plethora of liposome encapsulated enzymes have found applications in

biomedical industry, reviewed by (Walde and Ichikawa, 2001), but

proteoliposomes in the described formulations only function as a stabilizing and

protective carrier rather than being a functional reactor shell. Permeabilization

will be obtained with prokaryotic outer membrane proteins in the nanovesicle

shell. Reconstitution of outer membrane proteins in liposomes is widely used for

functional characterization of these proteins (Nikaido, 2003) and has recently

been explored in terms of nanoreactors with acetylcholinesterase (Colletier et al.,

2002; Nasseau et al., 2001) and β-lactamase (Graff et al., 2001). In both

reactors, OmpF was functionally capable of transporting the enzymatic substrate

(acetylcholine and ampicillin, respectively) inside the reactor and of releasing the

metabolized reaction products back into the environment.

Here, we report our first findings based upon the encapsulation of the inosine–

adenosine–guanosine preferring nucleoside hydrolase (TVNH; EC 3.2.2.1) of

Trypanosoma vivax (Versees et al., 2001) in porin permeabilized liposomes

(Figure 1) and show that these reactors are significantly more active than control

reactors without porins. Nucleoside hydrolases catalyse the hydrolysis of the N-

glycosidic bond in ribonucleosides. Given that no homologues of this class of

enzymes are found in mammalian cells (Hammond and Gutteridge, 1984), these

enzymes are attractive for use in cancer prodrug therapy since endogenous

hydrolysation of the prodrug will be limited.

Encapsulation of therapeutic NH in functionalised nanocapsules

75

Figure 1: schematic representation of a therapeutic nanoreactor with a lipid shell encapsulating TVNH and functionalised with bacterial outer membrane protein OmpF. Substrate analogue 6-methyl-purine riboside, a potential prodrug, diffuses into the reactor through OmpF and is converted into toxic 6-methylpurine, which is released back into the environment through diffusion.

3.2 Materials and methods

3.2.1. Purification of T. vivax nucleoside hydrolase

Purification of the wild type TVNH was performed as described by Versées et

al. (Versees et al., 2001). Shortly, TVNH was purified in a two step purification

scheme from Escherichia coli strain WK6 (Zell and Fritz, 1987) containing the

gene coding for the enzyme in a pQE-30 expression vector (Qiagen). In a two-

step purification protocol, the presence of an N-terminal His6-tag allowed

recovery of the protein on a Ni-NTA affinity column (Qiagen). pH gradient step

elution from the affinity column was followed by gel filtration on a superdex-200

column (Amersham Bioscience). SDS polyacrylamide gel electrophoresis

confirmed the purity of the enzyme.

Chapter 3

76

3.2.2. Purification of E. coli porins OmpF and PhoE

BL21(DE3) ∆lamB ∆ompC ∆ompA ompF::Tn5 and BL21(DE3) ∆lamB ompR

∆ompA strains were used for overproduction of OmpF and PhoE, respectively, as

described by Prilipov et al. (Prilipov et al., 1998). Overproduced porins were

subsequently purified by combining detergent extraction methods according to

Agterberg et al. (Agterberg et al., 1990) and Garavito and Rosenbusch (Garavito

and Rosenbusch, 1986). After cell harvesting by centrifugation, the cell pellets

were resuspended in 2% SDS 20 mM Tris–HCl pH 8 and shaken for 1 h at 60 °C

to obtain cell lysis. Lysates were pelleted by ultracentrifugation (20,000 rpm, 30

min, 4 °C) and resuspended in 0.125% octylpolyoxyethylene (oPOE) 20 mM

NaH2PO4 pH 7.3 for pre-extraction at 37 °C for 45 min. After a second cycle of

ultracentrifugation, pellets were resuspended in the same phosphate buffer with

3% oPOE for overnight extraction at 4°C and an additional 45 min at 37°C.

Finally, 10 mM EDTA was added to the supernatant after ultracentrifugation, to

strip of lipopolysaccharides. Purified proteins were concentrated (amiconR MWCO

5000) and dialyzed (Spectra/Por MWCO 6500-8000) against 20 mM NaH2PO4 150

mM NaCl 1% oPOE pH 7.3. SDS-polyacrylamide gel electrophoresis ascertained

the purity of both porins.

3.2.3. Electrophysiology

Black lipid membrane experiments were carried out as described previously by

Van Gelder et al. (Van Gelder et al., 2000) Conductance properties of OmpF and

PhoE channels were characterized in symmetrical solutions of 1 M KCl 1 mM

CaCl2 10 mM Tris pH 7.4, using soy bean phosphatidylcholine (Sigma) bilayers

over the circular aperture in the Teflon septum. Porins inserted spontaneously

after injection. Steady-state (I/V) curves were recorded with a BLM-120 Bilayer

Membrane Amplifier (Bio-Logic) and digitalized through a Powerlab/4SP

(ADInstruments). Conductance histograms result from at least 250 channel

events.

Encapsulation of therapeutic NH in functionalised nanocapsules

77

3.2.4. Preparation of proteoliposomes and swelling assays

Production of proteoliposomes was initially described by Nikaido and co-

workers (Luckey and Nikaido, 1980; Nikaido et al., 1991; Nikaido and

Rosenberg, 1983). Briefly, dried lipids (DOPC/EPG 4:1) were dissolved in diethyl

ether and a smooth film was formed under a gentle stream of nitrogen. After

drying, vesicles were spontaneously formed upon the addition of buffer (20 mM

Hepes pH 7.0). Porins were reconstituted at a concentration of approximately 20

ng µl-1 in buffer. Subsequently, all liposomes were sonicated and dried. For

swelling assays, liposomes were resuspended in 17% dextrane (MW 40,000,

Fluka) and shaken vigorously after 1 h. Swelling assays were performed as

reported previously by Nikaido (Nikaido and Rosenberg, 1983). Liposome

swelling upon the addition of substrate was followed 1 min

spectrophotometrically at 500 nm and transport flux was determined as follows:

dtdA

Ai⎟⎠⎞

⎜⎝⎛=Φ

1

Where Φ is the transport flux, Ai the initial absorption and dA/dt the

absorption difference during the measured time. Transport of substrates was

measured under isotoneous concentration as determined with raffinose, a higher

molecular weight sugar that is unable to diffuse through the porins. Under this

concentration, the scattering signal of the liposomes remains constant

throughout the measuring time within the sensitivity of the measurement,

indicating no swelling nor shrinking of the liposomes and making visualization of

solute transport unambiguously possible. Accordingly, arabinose, ribose, glucose,

inosine and cytidine added at the isotoneous concentration were tested for

transport through porins.

3.2.5. Preparation of enzyme encapsulating proteoliposomes

For the development of enzyme encapsulating proteoliposomes, sonicated and

dried liposomes as described above were resuspended in buffer containing 0.5

mg ml-1 TVNH. Subsequently, eight freeze–thaw cycles and extrusion through

Chapter 3

78

200 nm polycarbonate filters provided uniform and unilamellar proteoliposomes.

Uniform size distribution was visualized through dynamic light scattering at 532

nm with a scattering angle of 90° (Laser-Spectroscatter 201, RiNA netzwerk RNA

technologien). Free enzyme was removed through metal affinity chromatography

(QIAexpress Ni-NTA Protein purification system, QIAGEN). Activity of the

reactors was followed spectrophotometrically at 400 nm by enzymatic hydrolysis

of 1 mM p-nitrophenylriboside in buffer.

3.2.6. Fluorescence measurements

Fluorescence of free enzyme and of free enzyme in the presence of liposomes

(20–80 µl ml-1) was followed between 300 and 400 nm with a BowmaR Series 2

Luminescence Spectrometer after excitation at 280 nm (slit width of 4 nm and

scan speed of 1nm s-1).

3.3 Results

3.3.1. Purification and functional characterization of TVNH and porins

TVNH purification resulted in a single band on Coomassie stained SDS-PAGE

(Figure 2). Although kinetic parameters were slightly different from previously

reported values (Versees et al., 2001), the enzymatic characteristics of the

protein showed that it was fully functional (Table 1) and measured differences in

kinetic parameters are most likely due to the different buffer system.

SDS extraction of porins from the cell-envelope fraction yielded highly pure

material (Figure 2). The trimeric form, which runs on SDS-PAGE at 95 kDa when

non-boiled, is converted to a 36-kDa band when heated at 100 °C for 10 min.

The double band observed in the trimer lane for both porins is due to residual

lipopolysaccharides. Full stripping of lipopolysaccharides, however, is not

necessary for our purposes.

Encapsulation of therapeutic NH in functionalised nanocapsules

79

Figure 2: SDS-page gels after Coomassie staining of (a) TVNH, (b) OmpF and (c) PhoE. Lanes 2 and 3 in (b) and (c) represent non-boiled and boiled porins, respectively. Arrows indicate monomers.

Table 1: Kinetic constants of TVNH in 20 mM Hepes pH 7

TVNH TVNH + liposomes

inosine ρ-nitrophenylriboside ρ-nitrophenylriboside

kcat (s-1) 1.226 ± 0.043 0.081 ± 0.003 0.018 ± 0.002

KM (µM) 2.63 ± 0.63 1325 ± 101 2487 ± 571

Reconstitution of the purified porins in planar bilayers or black lipid

membranes (BLM) allows to visualize (and thus ascertain) their functionality

through determination of their conductance. The channel transitions from current

traces of OmpF reconstituted in planar membranes clearly showed one

population with a unit step value of 1.66 ± 0.44 nS which is in fair agreement

with Buehler et al. (Buehler et al., 1991) (Figure 3). Closure of the channels was

observed above the critical voltage of 155 mV as reported by Saint et al. (Saint

et al., 1996). Similarly, the measurement was repeated for PhoE, resulting in a

channel conductance of 1.22 ± 11 nS and closure above 135 mV.

1 2 1 2 3 1 2 3

Chapter 3

80

Figure 3: Conductance measurements of OmpF (A) and PhoE (B). Conductance histograms result from at least 250 channel events

3.3.2. Solute transport through OmpF and PhoE

OmpF and PhoE are general diffusion porins with a molecular exclusion limit of

roughly 600 Da. There is no selectivity for the transported molecules, although

OmpF shows slight cation selectivity while PhoE favours transport of anionic

molecules (Koebnik et al., 2000). Transport efficiency through porins was

ascertained for low molecular weight sugars (arabinose, ribose and glucose) and

nucleosides (cytidine and inosine) by reconstitution of purified porins in

proteoliposomes (Figure 4). Results for both porins were compared after

normalization to diffusion of arabinose. All tested solutes were transported

efficiently, with the exception of inosine: as inosine was still moderately

transported through OmpF, practically no diffusion was observed through PhoE.

Most likely, the voluminous hypoxantine residue lies at the basis of this transport

inefficiency. Interestingly, although only different in the orientation of its

hydroxyl groups in regard to arabinose, ribose diffused more efficiently through

either OmpF or PhoE and was even transported in the absence of porins (results

not shown). As ribose is most likely the slow release product in the reaction

scheme of nucleoside hydrolases (Vandemeulebroucke et al., 2003), this efficient

diffusion comes as an unexpected advantage

Encapsulation of therapeutic NH in functionalised nanocapsules

81

Figure 4: Transport of molecular substrates through porins OmpF (■) and PhoE (□). Transport fluxes were normalized to and recalculated as a percentage arabinose transport, for which transport efficiency was stated 100%.

3.3.3. Activity of nucleoside hydrolase encapsulated in proteoliposomes

Size distributions of proteoliposomes containing 0.5 mg ml-1 TVNH with and

without reconstituted porins were determined by dynamic light scattering. No

significant difference in radius was found between the different reactors: the

control reactors without porins had a mean hydrodynamic radius of 57.28 ±

10.11 nm, while OmpF and PhoE containing reactors displayed radii of 56.90 ±

6.85 and 52.92 ± 11.55 nm, respectively.

The enzymatic activity of encapsulated TVNH was determined using p-

nitrophenyl riboside as a substrate. Due to the low solubility of p-

nitrophenylriboside, it was impossible to perform swelling assays with this

substrate, but based upon the structural analogy with cytidine (Figure 5) and its

comparable molecular weight (271 and 243.2 for p-nitrophenylriboside and

cytidine, respectively), no transport problems through either one of the porins

was expected. Comparison of the hydrolytic activity of OmpF and PhoE

permeabilized reactors showed that both proteoliposomes were equally active.

Since there was no significant difference in hydrodynamic radius of the porins

OmpF and PhoE, it is obvious that diffusion of the enzyme substrate across both

porins occurred at comparable rates.

Chapter 3

82

Figure 5: Molecular structures of cytidine (left) and p-nitrophenylriboside (right).

Controls clearly were less active than either OmpF or PhoE permeabilized

liposomes (Figure 6). The residual enzymatic activity of the control vesicles

might stem from free TVNH after the affinity chromatography separation step,

but can also occur from adsorbed enzyme at the liposome surface. However, the

higher activity of permeabilized liposomes cannot be a result of more adsorbed

enzyme. The hydrodynamic radii of control and functionalized vesicles were equal

within experimental errors and furthermore there is no scientific ground to

assume that more protein will adsorb to a porin containing liposome.

Encapsulation of therapeutic NH in functionalised nanocapsules

83

Figure 6: Activity of TVNH encapsulated in OmpF (▲) and PhoE (o) permeabilized liposomes and in non permeabilized control vesicles (■). Liposomes consisting of DOPC and EPG in a 4:1 ratio were developed. Activity was measured with 1 mM final concentration p-nitrophenolriboside in 20 mM Hepes pH 7.0 at 400nm. Error bars are too small to show.

Another plausible explanation to the hydrolyzing activity of controls is that the

liposome lipid bilayer is permeable to the more hydrophobic p-

nitrophenylriboside. This seems unlikely with the rather high molecular weight of

p-nitrophenylriboside. Nevertheless, p-nitrophenylriboside might locally

destabilize the lipid bilayer and thus initiate its transport across the bilayer. In

this respect, Walde and Marzetta (Walde and Marzetta, 1998) already reported

membrane permeability for higher molecular weight hydrophobic molecules as

substrates to chymotrypsin encapsulating liposomes. Furthermore, Chakrabarti

et al. (Chakrabarti et al., 1994) reported membrane permeability for ADP with

polynucleotide phosphorylase encapsulated in POPC-liposomes. With the

surprising transport efficiency of ribose in mind, we could suggest a favourable

interaction of the phenyl part of p-nitrophenylriboside with the lipid bilayer.

3.3.4. Interactions between TVNH and liposomes

Adsorption of encapsulated enzymes on lipid surfaces is reported numerously

in literature and was reviewed recently (Walde and Ichikawa, 2001). To

determine whether or not the TVNH interacts with the liposomes, additional

experiments were performed with empty and non functionalized control

liposomes as described for swelling assays, to which TVNH was added externally

Chapter 3

84

and co-incubated for 90 min prior to activity measurements with p-

nitrophenylriboside. Thus, inhibition of TvNH activity was observed dependent on

liposome concentration and incubation time (results not shown). Determination

of the kinetic parameters resulted in a decrease in kcat as well as an increase in

KM in comparison to earlier published results from our group (Versees et al.,

2001) (Table 1). However, it still remains unclear whether these data reveal a

real shift in kinetic parameters or whether a (partial) inactivation of part of the

enzyme results in an apparent change of the parameters.

Although we could not unambiguously determine the cause of this inhibition,

fluorescence experiments do indicate a direct or indirect interaction between the

TvNH and the liposomes. TVNH shows a characteristic strong fluorescence signal

which peaks around 340 nm due to two nearly surface exposed tryptophan

residues, responsible for aromatic stacking in the catalytic cleft. Upon the

addition of liposomes, the fluorescence signal of these tryptophans is quenched

in a strongly concentration dependent manner, suggesting shielding of these

residues from the aqueous environment (Figure 7). In contrast, fluorescence of

rhodamine in control experiments with liposomes was not quenched.

Furthermore, this quenching implicates that these interacting TVNH molecules

will be inactivated or at least severely hampered in catalytic activity, explaining

the altered kinetic parameters discussed above. Indeed, as less enzyme

molecules are catalytically active, the conversion rate will decrease, leading to a

lower kcat. An additional increase in KM could point to the above suggested

interaction between the substrate and the liposomes, facilitating substrate

transport inside the vesicles. This leads to an increase in intravesicular substrate

concentration in such a manner that the external substrate concentration is

lower, resulting in an increased apparent KM.

Encapsulation of therapeutic NH in functionalised nanocapsules

85

Figure 7: Fluorescence scan of TVNH between 300 and 400 nm in absence (a) and presence of different amounts of liposome preparations (20 (b), 40 (c) and 80 µl ml-1 (d) after 90 min incubation at 30 °C. Liposomes (20 µl ml-1) alone showed no inherent fluorescence (e). Emission spectrum was measured after excitation at 280 nm.

In an attempt to prevent this enzyme–liposome interaction, a series of

liposomes with different net charge and charge density were constructed. This

can be established in a controllable manner through varying the DOPC/EPG ratio.

Since the enzyme charge around the catalytic cleft is predominantly negative

(Versees et al., 2001), a series of more negative liposomes was developed by

increasing the EPG content. No significant changes in catalysis as compared to

previous liposome measurements were observed upon the addition of the TVNH

(Figure 8). Furthermore, the addition of salt up to 1 M is unable to abolish the

interaction (results not shown). We therefore conclude that the enzyme–

liposome interaction cannot be of electrostatic nature. Instead, a strong

hydrophobic force is found more likely to be responsible for the interaction.

To confirm this hypothesis we looked at the effect of various detergents on the

activity of TVNH. The activities were measured in the presence of detergent with

the substrate inosine in 20mM Hepes pH7.0. The enzymatic reaction was

followed spectrophotometrically at 280 nm and compared to the activity of TvNH

in absence of detergent (Table 2). Since al detergents have a severe negative

effect on the activity of TvNH, it is plausible to say that the interaction between

TvNH and liposomes is of a hydrophobic nature.

Chapter 3

86

Figure 8: Activity of TvNH in presence of DOPC:EPG liposomes after 90 min co-incubation. The enzyme/liposome ratio is in agreement as described for the development of the nanoreactors. Activities were normalized to the liposome density

Table 2: Effects of detergents on the activity of TvNH. Detergents were used at a final concentration of 2x the critical micelle concentration.

detergent % TvNH activity

remaining

Anzergent 3-10 6% Anzergent 3-12 9% SB12 20% n-nonyl-β-D-glucoside 9% n-decyl- β -D-glucoside 39% n-dodecyl- β -D-glucoside 24% n-octyl- β -D-maltoside 4% n-nonyl- β -D-maltoside 11% n-decyl- β -D-maltoside 2% n-undecyl- β -D-maltoside 7%

Encapsulation of therapeutic NH in functionalised nanocapsules

87

3.4 Discussion

This report is a pioneering step to introduce an alternative approach to

liposome based cancer treatment and a new principle for prodrug strategies.

Nucleoside hydrolase of T. vivax, a potential attractive enzyme for cancer

therapy, was encapsulated in liposomes, which were functionalized with specific

porins of E. coli. By liposome swelling assays, we could show the successful

diffusion of the substrates for these enzymes. The enzymatic activity showed

that functionalized reactors were more effective in substrate hydrolysation than

control reactors.

At this point, we were unable to establish whether the shielding of the

tryptophan residues of TvNH was a result from permanent adsorption on the

liposome surface or from the formation of liposome induced enzyme aggregates.

The latter, however, would lead to a less clear liposome concentration dependent

enzyme inactivation since at low liposome concentration the vesicle surface is

still large enough to initiate aggregation of more enzyme molecules. This might

pose problems in terms of immunogenicity.

The techniquely less demanding liposome approach was chosen here to

demonstrate this concept of functionalized nanoreactors. Further experiments

will concentrate on a novel kind of polymer based vesicles instead of natural

lipids. We recently showed that functional Tsx, a nucleoside/nucleotide specific

outer membrane protein of E. coli, could be incorporated in these polymer

vesicles. These polymer vesicles might shield the porins from the immune

system since the polymeric membrane is thicker than a lipid bilayer, and at the

same time prevent non-specific adsorption of enzymes.

This strategy may prove to be an attractive alternative to ADEPT and GDEPT,

overcoming distinct hurdles of these therapies.

Chapter 3

88

3.5 References

Aghi, M., Hochberg, F. and Breakefield, X.O. (2000) Prodrug activation enzymes in cancer gene therapy. J Gene Med, 2, 148-164.

Agterberg, M., Adriaanse, H., Lankhof, H., Meloen, R. and Tommassen, J. (1990) Outer membrane PhoE protein of Escherichia coli as a carrier for foreign antigenic determinants: immunogenicity of epitopes of foot-and- mouth disease virus. Vaccine, 8, 85-91.

Allen, T. (2002) Ligand-targeted therapeutics in anticancer therapy. nature reviews cancer, 2, 750-763.

Buehler, L.K., Kusumoto, S., Zhang, H. and Rosenbusch, J.P. (1991) Plasticity of Escherichia coli porin channels. Dependence of their conductance on strain and lipid environment. J Biol Chem, 266, 24446-24450.

Chakrabarti, A.C., Breaker, R.R., Joyce, G.F. and Deamer, D.W. (1994) Production of RNA by a polymerase protein encapsulated within phospholipid vesicles. J Mol Evol, 39, 555-559.

Colletier, J.P., Chaize, B., Winterhalter, M. and Fournier, D. (2002) Protein encapsulation in liposomes: efficiency depends on interactions between protein and phospholipid bilayer. BMC Biotechnol, 2, 9.

Denny, W. (2001) Prodrug strategies in cancer therapy. European Journal of medicinal chemistry, 36, 577-595.

Garavito, R.M. and Rosenbusch, J.P. (1986) Isolation and crystallization of bacterial porin. Methods Enzymol, 125, 309-328.

Graff, A., Winterhalter, M. and Meier, W. (2001) Nanoreactors from polymer-stabilized liposomes. Langmuir, 17, 919-923.

Greco, O. and Dachs, G.U. (2001) Gene directed enzyme/prodrug therapy of cancer: historical appraisal and future prospectives. J Cell Physiol, 187, 22-36.

Gregoriadis, G. (1995) Engineering liposomes for drug delivery: progress and problems. Trends Biotechnol, 13, 527-537.

Hammond, D.J. and Gutteridge, W.E. (1984) Purine and pyrimidine metabolism in the Trypanosomatidae. Mol Biochem Parasitol, 13, 243-261.

Koebnik, R., Locher, K.P. and Van Gelder, P. (2000) Structure and function of bacterial outer membrane proteins: barrels in a nutshell [In Process Citation]. Mol Microbiol, 37, 239-253.

Luckey, M. and Nikaido, H. (1980) Specificity of diffusion channels produced by lambda phage receptor protein of Escherichia coli. Proc Natl Acad Sci U S A, 77, 167-171.

Nasseau, M., Boublik, Y., Meier, W., Winterhalter, M. and Fournier, D. (2001) Substrate-permeable encapsulation of enzymes maintains effective activity, stabilizes against denaturation, and protects against proteolytic degradation. Biotechnol Bioeng, 75, 615-618.

Nikaido, H. (2003) Molecular basis of bacterial outer membrane permeability revisited. Microbiol Mol Biol Rev, 67, 593-656.

Nikaido, H., Nikaido, K. and Harayama, S. (1991) Identification and characterization of porins in Pseudomonas aeruginosa. J Biol Chem, 266, 770-779.

Nikaido, H. and Rosenberg, E.Y. (1983) Porin channels in Escherichia coli: studies with liposomes reconstituted from purified proteins. J Bacteriol, 153, 241-252.

Prilipov, A., Phale, P.S., Van Gelder, P., Rosenbusch, J.P. and Koebnik, R. (1998) Coupling site-directed mutagenesis with high-level expression: large scale production of mutant porins from E. coli. FEMS Microbiol Lett, 163, 65-72.

Encapsulation of therapeutic NH in functionalised nanocapsules

89

Saint, N., Lou, K.L., Widmer, C., Luckey, M., Schirmer, T. and Rosenbusch, J.P. (1996) Structural and functional characterization of OmpF porin mutants selected for larger pore size. II. Functional characterization. J Biol Chem, 271, 20676-20680.

Senter, P.D. and Springer, C.J. (2001) Selective activation of anticancer prodrugs by monoclonal antibody-enzyme conjugates. Adv Drug Deliv Rev, 53, 247-264.

Springer, C.J. and Niculescu-Duvaz, I. (2000) Prodrug-activating systems in suicide gene therapy. J Clin Invest, 105, 1161-1167.

Torchilin, V. (2000) drug targeting. European Journal of Pharmaceutical Sciences, 11, 81-91.

Van Gelder, P., Dumas, F. and Winterhalter, M. (2000) Understanding the function of bacterial outer membrane channels by reconstitution into black lipid membranes [In Process Citation]. Biophys Chem, 85, 153-167.

Vandemeulebroucke, A., Versees, W., De Vos, S., Van Holsbeke, E. and Steyaert, J. (2003) Pre-steady-state analysis of the nucleoside hydrolase of Trypanosoma vivax. Evidence for half-of-the-sites reactivity and rate-limiting product release. Biochemistry, 42, 12902-12908.

Versees, W., Decanniere, K., Pelle, R., Depoorter, J., Brosens, E., Parkin, D.W. and Steyaert, J. (2001) Structure and function of a novel purine specific nucleoside hydrolase from Trypanosoma vivax. J Mol Biol, 307, 1363-1379.

Walde, P. and Ichikawa, S. (2001) Enzymes inside lipid vesicles: preparation, reactivity and applications. Biomol Eng, 18, 143-177.

Walde, P. and Marzetta, B. (1998) Bilayer permeability-based substrate selectivity of an enzyme in liposomes. Biotechnol Bioeng, 57, 216-219.

Xu, G. and McLeod, H.L. (2001) Strategies for enzyme/prodrug cancer therapy. Clin Cancer Res, 7, 3314-3324.

Zell, R. and Fritz, H.J. (1987) DNA mismatch-repair in Escherichia coli counteracting the hydrolytic deamination of 5-methyl-cytosine residues. Embo J, 6, 1809-1815.

Chapter 4 Comparison and characterisation of PMOXA-PDMS-PMOXA vesicles

91

Chapter 4: Comparison and

characterisation of PMOXA-PDMS-PMOXA

vesicles Abstract The phase behaviour of triblock copolymers in aqueous solutions is controlled by their chemical composition, the length and structure of the individual blocks and the molecular architecture of the whole polymer. In addition the molecular weight distribution of the individual blocks has a profound effect on the phase behaviour of such systems. Small changes in these properties can lead to different phase behaviour in aqueous solutions. Therefore we screened several PMOXA-PDMS-PMOXA polymers to see which polymer batches are able to spontaneously self-assemble in to nanometer-sized vesicles. Two polymer batches that show such phase behaviour were identified. In addition we were able to produce enzymatically active nanoreactors with these batches by encapsulating TvNH in the polymeric vesicles and incorporating the porin Tsx in the polymeric membrane. The nanoreactors were analysed by DLS, AFM and TEM. We also performed a preliminary experiment to determine the interaction of such nanoreactors with macrophages by fluorescently labelling the encapsulated TvNH with Alexa®555.

Chapter 4

92

Comparison and characterisation of PMOXA-PDMS-PMOXA vesicles

93

4.1 Introduction

Block copolymers consist of hydrophilic and hydrophobic blocks and behave

similar to conventional surfactants in aqueous solutions. They self assemble into

micelles and vesicles of different size and shape and at high concentrations they

form lyotropic liquid crystalline phases. Their phase behaviour is predominantely

controlled by their chemical composition, the length and structure of the

individual blocks and the molecular architecture of the whole polymer. In

addition the molecular weight distribution of the individual blocks has a profound

effect on the phase behaviour of such systems (Nardin and Meier, 2001). The

ABA triblock copolymer poly(2-methyloxazoline)-b-poly(dimethyl siloxane)-

bpoly(2-methyloxazoline) (PMOXA-PDMS-PMOXA) can self assemble into various

structures in aqueous solutions, including nanotubes, free standing films and

vesicular structures, depending on the experimental conditions and constitution

of the polymer (Sauer and Meier, 2004). This implies that even small changes in

the block copolymer composition can lead to different phase behaviour. The

presence of impurities can also have an effect on the phase behaviour of triblock

copolymers. Thus, there is a high risk for batch variability, since it is very difficult

to precisely control the synthesis and length of both hydrophilic and hydrophobic

blocks. This is something we indeed observed with the PMOXA-PDMS-PMOXA

polymers.

In this study we show the different behaviour of different batches of

PMOXA18-PDMS72-PMOXA18 and PMOXA21-PDMS54-PMOXA21 and the ability of

these triblock copolymers to form self-assembled enzymatically active

nanoreactors that encapsulated nucleoside hydrolase of Trypanosoma vivax

(TvNH) and are permeabilized by the bacterial outer membrane channel Tsx. In a

second part we characterised nanoreactors formed by PMOXA21-PDMS54-PMOXA21

via AFM and TEM and performed a preliminary test to see whether such

nanoreactors can be taken up by macrophages, since uptake by macrophages or

other antigen presenting cells is the first step in the activation of the immune

system.

Chapter 4

94

4.2 Materials and methods

4.2.1. Purification of T. vivax nucleoside hydrolase

Purification of the wild type TVNH was performed as described by Versées

et al. (Versees et al., 2001). Shortly, TVNH was purified in a two step purification

scheme from Escherichia coli strain WK6 (Zell and Fritz, 1987) containing the

gene coding for the enzyme in a pQE-30 expression vector (Qiagen). In a two

step purification protocol, the presence of an N-terminal His6-tag allowed

recovery of the protein on a Ni-NTA affinity column (Qiagen). pH gradient step

elution from the affinity column was followed by gel filtration on a superdex-200

column (Amersham Bioscience). SDS polyacrylamide gel electrophoresis

confirmed the purity of the enzyme (chapter 3).

4.2.2. Cloning, expression and purification of Tsx

For cloning of Tsx, a pGtsx plasmid was constructed (Figure 1). The pGompf

plasmid as described by Prilipov et al. served as a template. Firstly, the gene

encoding OmpF was cut from the vector with restriction enzymes PstI and ClaI.

Secondly, the sequence 5’-

GAACACCACCACCACCACCACCTTGTTCCGCGTGGTAGTAT-3’ encoding a His6-tag

followed by a trombine cleaving sequence was ligated in the vector. Finally the

gene encoding tsx was cloned in this new vector

Comparison and characterisation of PMOXA-PDMS-PMOXA vesicles

95

pGnhistromtsx3733 bps

500

1000

15002000

2500

3000

3500

XbaIEcoRIPacI

BsiHKAIBsp1286I

BsrGIPstI

ClaIXhoIISmlIBsmI

BtgIBpu10I

NruIBsiHKAI++EcoRV

Bpu1102IBmrIXcmIVan91IOliI

BamHIXhoII

Bpu1102IEcoO109ISmlI

AccIIIXhoII

NaeINgoMIV

BanIIBsp1286I

BsaAIDraIII

PsiI

BsiHKAIBsp1286I

XhoIIXhoIISmlI

XmnIBsiHKAI

Bsp1286IBsaHI

ScaI

BpmIBmrI

AhdI

XhoIIXhoII

XhoIISmlI

XhoII

SmlIBsiHKAI

Bsp1286I

SmlI

SapIXhoII

nhistrom

Tsx insert

Figure 1: schematic representation of the pGtsx vector. The gene encoding Tsx is cloned between restriction sites ClaI and BamHI (depicted in red) and the sequence encoding the His6-tag and trombine cleaving site is cloned between PstI and ClaI (depicted in blue).

BL21(DE3) ∆lamB ompR ∆ompA strains were used for overproduction of Tsx

as described by Prilipov et al. (Prilipov et al., 1998). For this purpose, the gene

encoding Tsx was cloned in the pGompf plasmid. For purification purposes an

Nterminal 6His-tag was added. Overproduced porins were subsequently purified

by combining detergent extraction methods of Agterberg et al. (Agterberg et al.,

1990) and Garavito and Rosenbusch (Garavito and Rosenbusch, 1986). After cell

harvesting by centrifugation, the cell pellets were resuspended in 2% SDS 20

mM Tris pH 8.0 and shaken for 1 h at 60 °C to obtain cell lysis. Lysates were

pelleted by ultracentrifugation (20,000xg, 30 min, 4 °C) and resuspended in

0.125% oPOE 20 mM NaH2PO4 pH 7.3 for pre-extraction at 37 °C for 45 min.

After a second cycle of ultracentrifugation, pellets were resuspended in the same

phosphate buffer with 3% oPOE for overnight extraction at 4°C and an additional

45 min at 37°C. Finally, 10 mM EDTA was added to the supernatant after

ultracentrifugation, obtaining stripping of lipopolysaccharides. Since Tsx was

expressed with an N-terminal 6His-tag, it was further purified on a Ni-NTA

affinity column (Qiagen) and gel filtration on a superdex-200 column (Amersham

Bioscience) to remove contaminating proteins.

Chapter 4

96

4.2.3. Preparation of nanoreactors

4.2.3.1. Lamellar film rehydration method

Block copolymers and porins are mixed in ethanol to a molar ratio of 10 to 1.

This solution is then dried at the bottom of a glass tube under vacuum to remove

all the remaining solvent. Subsequently, the lamellar film is rehydrated for 24 h

at 4°C by adding PBS pH 7.4 containing 54 µM TvNH to a final polymer

concentration of 1%. This results in swelling of the lamellar film and formation of

vesicles. Finally, extrusion with a polycarbonate filter of 0.2 µm gives

monodisperse vesicles. Non encapsulated TvNH is removed by Ni-NTA affinity

chromatography due to the presence of an N-terminal His6-tag.

4.2.3.2. Solvent evaporation method

A solution of purified porin is mixed with copolymer in ethanol to the desired

molar ratio of protein to polymer (1/10). To encapsulate the enzyme in the

interior of the vesicles, the homogeneous polymer-porin solution is added drop

wise to PBS, pH 7.4 buffer containing TvNH (54 µM) to a final polymer

concentration of 1% and stirred for two hours at room temperature. During this

incubation period the nanoreactors are formed by self-assembly and ethanol is

evaporated. The solution is stirred for an additional 48 h at 4°C. Then the

resulted dispersion is repeatedly extruded through a 0.2 µM polycarbonate filter

to obtain monodisperse vesicles. The non encapsulated TvNH is removed by Ni-

NTA affinity chromatography.

4.2.3.3. Direct dispersion method

In the direct suspension method the polymer is suspended in PBS pH 7.4

containing 54 µM TvNH at a final concentration of 1% and gently stirred at 4°C

overnight to form self-assembled vesicles. To destabilize the polymer membrane

of the vesicles, O.5 % Triton X-100 is subsequently added and the suspension is

sonicated twice in a water bath sonicator for 10s. Porins are added shortly after

the sonication step and will incorporate inside the polymeric wall of the

nanoreactors. Biobeads (5g/ml) (BioRad) are added to remove the detergent

Comparison and characterisation of PMOXA-PDMS-PMOXA vesicles

97

Vesicles were repeatedly extruded through a polycarbonate filter and non

encapsulated TvNH is removed by Ni-NTA affinity chromatography..

4.2.4. DLS measurements

The size and polydispersity of the nanoreactors was determined via DLS

(laser-spectroscatter 201 by RiNA GmbH, Berlin, Germany) at 532 nm with a

scattering angle of 90°. Ten data sets were recorded and the size distribution

was analyzed using the software CONTIN.

4.2.5. Reducing sugar assay

To determine the enzymatic activity of encapsulated TvNH for guanosine in the

nanoreactors we used a reducing sugar assay as described by Parkin (Parkin,

1996). Briefly, the enzymatic reaction was stopped by adding a CuSO4–solution.

The Cu2+ is reduced to Cu+

by the reaction product ribose. This reduced Cu reacts

with neocuproin to form a complex. Color (yellow) development of this complex

was achieved by heating the solution 8 min at 95°C and the OD at 450nm was

measured. A standard curve with known ribose concentration was used to

determine the extinction coefficient under the assay conditions.

4.2.6. Atomic Force Microscopy (AFM)

Freshly cleaved mica was functionalized with 3-aminopropyltriethoxysilane

(APES) to allow adsorption of the nanoreactors to the mica chip. This was

done by placing two cups with 30 µl APES and 10 µl N,N-diisopropylethylamine,

respectively, at the bottem of a desiccator, the mica chips were placed at the

upper part of the desiccator and this was incubated for 2 h. The mica chips were

left to cure for an additional 48 h. Subsequently, 20 µl of a 10Ox diluted sample

of nanoreactors prepared with PMOXA21-PDMS54-PMOXA21 via the solvent

evaporation method was placed on the APES-mica chip. After 30 min of

incubation at room temperature, the chip was rinsed 4X with PBS to wash of non

adsorbed nanoreactors. The chip was finally wetted with PBS for scanning in

liquid.

Chapter 4

98

A Nanoscope IIIa atomic force microscope (Digital Instruments/Veeco) in

tapping mode at room temperature was used to acquire 512 x 512 pixel images.

We used an NP tip (Veeco), a 125-µm cantilever with a nominal spring constant

of 50 N m-1 and resonant frequencies in a range of 279 to 362 kHz. The scan

rate was 1.5 Hz and the scan size was 1.5 x 1.5 µm. Images were flattened prior

to analysis using the NanoScope 6.11r1 software (Digital Instruments/Veeco).

4.2.7. Transmission Electron Microscopy (TEM)

To obtain a support film for TEM measurements, carbon was coated onto

copper/formvar (polyvinyl formal) grids by evaporating a uniform sheet of carbon

on the grids using a vacuum evaporator. Samples were prepared by depositing

10 µl of undiluted nanoreactor solution (stock of 10mg polymer/ml) on the

carbon-coated side of the grids. After 30-60 seconds adsorption, the excess of

liquid was removed by touching the grid surface with a filter paper. The grid was

then washed 3 times with a drop of distilled water.

The specimen was stained by lowering the grid twice on a drop of negative

stain solution (1-2 % uranyl acetate) for 20 seconds.

Grids were analysed using a Phillips Tecnai 10 electron microscope operating

at 80 kV. Micrographs were recorded with a SIS camera.

4.2.8. Fluorescent labelling of TvNH

To fluorescently label TvNH we used the amine reactive Alexa®555 dye

(Molecular probes). This fluorescent dye has spectra that are identical to those

of the Cy3 dye with an absorption maximum at 555 nm and a fluorescence

emission maximum at 565 nm. This amine reactive succinimidyl ester forms a

stable amide bond with amine groups on the protein. Coupling of the Alexa®555

dye was performed as follows. TvNH was dialysed to 0.1 M sodium bicarbonate

buffer pH 8.3 to a final concentration of 20 mg/ml. Alexa®555 was dissolved in

DMSO to a final concentration of 10mg/ml. 1ml of the TvNH sample was mixed with

100 µl Alexa®555 and incubated for 1 h at room temperature under continuous

stirring. The reaction was stopped by adding 1.5 M of hydroxylamine pH 8.5. After

Comparison and characterisation of PMOXA-PDMS-PMOXA vesicles

99

one hour of incubation at room temperature, non bonded Alexa®555 was removed

via gelfiltration on a superdex 75 High Resolution column. The running buffer used

for the gelfiltration was Phosphate Buffered Saline (PBS) pH 7.4. The degree of

labelling (DOL) was determined by:

[ ] dyeproteinMWA

DOLε

max=

Where Amax is the absorption of the protein-dye complex at 555 nm, MW is

the molecular weight of the protein and εdye is the extinction coefficient of the

dye at 555 nm (150.000 cm-1 M-1). The protein concentration in the protein-dye

complex is determined by measuring the absorption at 280 nm and using the

Lambert-Beer law. However, a correction needs to be made for the absorbance

of the dye at 280 nm by using the following equation:

( )CFAAAprotein max280 −=

Where A280 is the absorption of the protein-dye complex at 280 nm and CF is

the correction factor of the dye at 280 nm (0.08).

4.2.9. In vitro nanoreactors-macrophages interactions

Macrophages were recruited to the peritoneum by injection of thioglycolate.

After 24 h 5.107 peritoneal exudate cells (PEC’s) were harvested. The cells

were cultured in 36 well plates at 37°C in RPMI-1640 medium (Invitrogen)

supplemented with 10 % heat-inactivated fetal calf serum, 0.03% L-glutamine,

100 mg/ml streptomycin, 100 mg/ml penicillin, 1mM nonessential amino acids

and 1 mM sodium pyruvate (Invitrogen). After 24 h of incubation the medium

was removed and 50 µl of undiluted fluorescent Alexa®555 nanoreactors was

added. As a negative control, 50 µl of labelled nanoreactors were added to Wells

without macrophages. Pictures were taken after 15 min, 45 min and 17h with an

eclipse TE2000-S microscope (Nikon) fitted with a Chroma filter for excitation

540-580nm and emission 600nm.

Chapter 4

100

4.3 Results

4.3.1. Polymer batch variability

Initial experiments were conducted using the PMOXA18-PDMS72-PMOXA18

triblock copolymer, kindly provided by Prof. W. Meier from the department of

chemistry at the university of Basel (Ranquin et al., 2005). Preparation of

nanoreactors was very reproducible until a new polymer batch was provided.

This new polymer had an identical composition to the first polymer batch.

However, results previously obtained could not be reproduced. With our first

batch of polymer, we used the lamellar film method to produce nanoreactors.

After rehydration of the polymer film, the size of vesicles ranges from 50 nm

to 1µM with a mean diameter of ± 160 nm (Figure 2 A). After extrusion a more

monodisperse vesicle population was obtained with a mean radius of ± 70nm

(Figure 2 B).

With the new polymer batch, large aggregates are formed after film

rehydration with a mean radius of ± 2.2 µm (Figure 3A). These aggregates can

not be extruded since they are too large. Sonication of these aggregates results

in smaller aggregates with a radius ranging from 160 nm to 1,2 µM (Figure 3 B).

After successive extrusion through a 0.45 µM and 0.2 µm filter, a more

monodisperse population was obtained with a mean radius of ± 240 nm (Figure

3 C). Moreover, these nanoreactors were not active, although sonication had no

effect on the activity of the TvNH enzyme (data not shown).

Comparison and characterisation of PMOXA-PDMS-PMOXA vesicles

101

A B

Figure 2: DLS measurements of Tsx permeabilized nanoreactors composed of PMOXA18-PDMS72-PMOXA18 (Basel), prepared via the lamellar film rehydration method before extrusion (A) and after extrusion (B).

Preparation via the solvent exchange method resulted in smaller aggregates

but again no activity was found for such aggregates. Finally, preparation via the

direct method resulted in the formation of relatively small aggregates with a

radius of 238 nm (Figure 4 A). Incorporation of Tsx was achieved by destabilizing

the polymer membrane with 0.05 % TritonX-100. After incorporation of Tsx,

TritonX-100 was removed via biobeads. After removal of TritonX-100 we again

obtained very large aggregates with a mean diameter of 1 µm (Figure 4 C).

Chapter 4

102

A B

C

Figure 3: DLS measurements of lamellar film preparation method after film rehydration (A), after sonication (B) and after extrusion (C).

We tested all three preparation methods for several polymer batches without

obtaining small (< 200 nm) active nanoreactors until finally we purchased

PMOXA21-PDMS54-PMOXA21 from Polymer Source Inc. This polymer has the

same hydrophilic and hydrophobic building blocks as the polymer received from

Basel but there is a small difference in the block lengths. The Basel polymer has

a hydrophilic block length of 18 and a hydrophobic block length of 72. Although

the hydrophilic blocks of the Polymer Source polymer are similar in length as the

Basel polymer, the hydrophobic block length is significantly shorter. This might

Comparison and characterisation of PMOXA-PDMS-PMOXA vesicles

103

be an advantage for incorporating porins since the polymer membrane thickness

will be more in accordance to the natural environment of porins, i.e. a

phospholipid membrane.

A B

C D

Figure 4: DLS measurements of the direct preparation method after vesicle formation(A), after incorporation of Tsx via TritonX-100 (B) and after removal of TritonX-100 via Biobeads (C). Tsx permeabilized nanoreactors composed of PMOXA21-PDMS54-PMOXA21 (Polymer Source Inc.) prepared via the solvent evaporation method (D).

Chapter 4

104

Nanoreactors with the Polymer Source polymer were prepared via the solvent

exchange method. After formation and removal of non encapsulated TvNH via Ni-

NTA, we obtained vesicles with a mean radius of ± 100 nm (Figure 4D). Since

the sample is already monodisperse, extrusion is not necessary. The activity of

the nanoreactors was measured for guanosine using a reducing sugar assay.

Guanosine concentrations ranging from 50 to 1000 µM were tested and the

obtained data was fitted to a Michaelis-Menten equation (Figure 5). With this

polymer we were able to make small (100 nm radius) and active nanoreactors.

Furthermore, these results were very reproducible. Therefore we chose to use

this polymer for our future experiments.

Figure 5: Activity of Tsx permeabilized nanoreactors that encapsulate TvNH, composed of PMOXA21-PDMS54-PMOXA21 (purchased from Polymer Source Inc.). Reaction rates were determined using a reducing sugar assay. The data were fitted to a Michaelis-Menten equation.

4.3.2. AFM and TEM images of nanoreactors

AFM images clearly show nanoreactors with low polydispersity (Figure 6 A and

B). Analysis of the cross section of the nanoreactors shows a radius of

approximately 75 nm and a height of 10-20 nm (Figure 6C). This is not in

agreement with the results of DLS measurements which showed a mean radius

of 104 nm. Similar observations were made by Meier and co-workers (personal

communication) and this discrepancy between AFM and DLS results was

attributed to the collapse of the vesicles on the mica chip.

Comparison and characterisation of PMOXA-PDMS-PMOXA vesicles

105

A B

C

Figure 6: Atomic force microscopy images. Height image (A), amplitude image (B) and nanoreactor cross sections (C). Images were analysed using the Nanoscope software.

Chapter 4

106

Figure 7: TEM images of nanoreactors. Samples were analysed using a Phillips Tecnai 10 electron microscope operating at 80 kV. Micrographs were recorded on with a SIS camera.

Nanoreactors of approximately 170 nm in diameter can be detected on TEM

images (Figure 7). This size is in good agreement with the size measured in DLS

(200 nm) (data not shown). Although the image is of low resolution, the sample

forms uniformly sized vesicles of spherical shape.

4.3.3. Fluorescently labelled Tsx-TvNH nanoreactors

The amine reactive dye Alexa®555 was used to fluorescently label the

encapsulated enzyme TvNH. After removal of the non bonded dye via

gelfiltration, the approximate number of dye molecules per protein molecule

(degree of labelling) was calculated to be 0.5. This fluorescently labelled TvNH

(5µM) was used to prepare fluorescent nanoreactors via the solvent evaporation

method. To perform in vitro macrophage uptake experiment, the fluorescent

vesicles were additionally purified on a superdex 75 HR gelfiltration column with

with PBS, pH 7.4 as running buffer (results not shown). Fluorescent vesicles with

a mean radius of ± 70 nm (Figure 8) were obtained.

Comparison and characterisation of PMOXA-PDMS-PMOXA vesicles

107

Figure 8: Analysis of the size-distributions of fluorescent Tsx-TvNH nanoreactors by DLS.

4.3.4. In vitro interaction of fluorescent nanoreactors with macrophages

By producing fluorescent nanoreactors the uptake by or binding to

macrophages can be visualised by fluorescent microscopy. As a negative control

fluorescent nanoreactors were added to wells that do not contain macrophages.

Fluorescent microscopy images of this negative control show a slight increase in

fluorescence over time at the bottom of the well (Figure 9, left panel). In

contrast, when fluorescent nanoreactors are added to cultured macrophages, the

fluorescence increases rapidly near the macrophages (Figure 9, right panel).

After 17 h of incubation, there is a strong fluorescence present on the

macrophages, indicating that nanoreactors are concentrated at the cell surface.

However no clear uptake of vesicles is seen. This concentration of vesicles at the

surface of macrophages, can be the result of nonspecific binding, specific binding

to cell surface receptors of the macrophages or entrapment of the vesicles in the

extra cellular matrix of the macrophage cell culture.

Chapter 4

108

After 15 min

After 45 min

After 17 hours

Figure 9: Fluorescent images of the interaction between macrophages and fluorescent nanoreactors. Left panel: negative control where no macrophages were cultured before adding the fluorescent nanoreactors, Right panel: macrophages were cultured and after removal of culture medium, fluorescent nanoreactors were added. Pictures were taken on an eclipse TE2000-S (Nikon) fluorescent microscope with a Chroma filter for excitation 540-580nm and emission 600nm.

Comparison and characterisation of PMOXA-PDMS-PMOXA vesicles

109

4.4 Discussion

Nanoreactors based on triblock copolymers are a completely new and

promising field. Therefore if pharmacological applications on an industrial scale

are to be envisioned it is of uttermost importance that the synthesis of the

polymers can be precisely controlled in order that the different polymer batches

have identical phase behaviour.

Given that the phase behaviour of triblock copolymers in aqueous solutions

depends on the composition, length and structure of the individual blocks as well

as on the molecular architecture of the whole polymer and the molecular weight

distribution of the individual blocks, it is very important to precisely control these

properties to assure the reproducibility of the experiments performed with them.

However, due to the nature of the synthesis of the polymer, it is very difficult to

precisely control the length and molecular weight distribution. From our

experience it is clear that small changes can have profound effect on the

behavior of the polymer in aqueous solutions. Polymers with identical

composition but from a different batch, behave very differently in aqueous

solution where one polymer batch self assembles into nanometer sized vesicles

while another results in the formation of micrometer sized aggregates. This

different behavior was observed for different production methods. The larger

aggregates could be nanotubes or free standing films since no enzymatic activity

could be detected. Several PMOXA-PDMS-PMOXA triblock copolymers with

different compositions were screened for their ability to form enzymatically active

vesicles in aqueous solutions (data not shown). Finally, a commercially available

PMOXA21-PDMS54-PMOXA21, that self-assembles into nanoreactors with a mean

diameter of 150 - 200 nm via the solvent evaporation method was found. By fine

tuning the production method, monodisperse vesicles were obtained without

extrusion. Furthermore, the production of such nanoreactors is highly

reproducible even with different batches.

The size and shape of the nanoreactors was analysed via DLS, AFM and TEM.

AFM images (height and amplitude) showed a monodisperse population.

However, the size determined via AFM does not concur with the size determined

by DLS. With AFM, a mean diameter of 150 nm and a height of 15 nm were

found whereas the mean diameter in DLS was 200 nm. This discrepancy can be

Chapter 4

110

explained by a fundamental difference in both techniques. DLS measurements

are performed in solution whereas in AFM measurements the vesicles are

adsorbed to a mica chip. Adsorption of the vesicles to the chip can lead to

deformation of the vesicles. This was also observed by the group of W. Meier and

is explained as the collapse of the vesicles on the chip which gives the vesicles a

shape comparable with a deflated ball. Also force effects of the sampling AFM tip

might contribute to deformation of the vesicles.

Although the resolution of the TEM image is very low, it also shows a spherical

shaped monodisperse population with a mean diameter of 170 nm. Since

macrophages play an important role in both innate and adaptive immunity we

investigated whether macrophages are able to engulf nanoreactors. Uptake of

nanoreactors can result in the activation of macrophages leading to the

production and secretion of effector molecules such as IL-1β, IL-6, IL-10, IL-12,

IL-18, INF-α/β and TNF-α (Martinez et al., 2008). In addition, breakdown of the

nanoreactors in the lysosome and antigen presentation at the cell surface can

activate CD4+ T-cells leading to a humane or cellular immune response.

Therefore we performed a preliminary experiment to evaluate the in vitro uptake

of fluorescent nanoreactors by macrophages. In this experiment no clear uptake

of nanoreactors was observed during a 17 h period. Since the culture medium

was removed before adding the nanoreactors, the results after 17 h have to be

interpreted with caution. It is well known that deprivation from culture medium

for more than several hours will lead to cell death. Due to the absence of a

positive control, we can not be sure that the lack of a fluorescence signal inside

the macrophages means that the reactors are not taken up by the macrophages.

However, these observations are in good agreement with the findings of Broz et

al. (Broz et al., 2007). They also used PMOXA-PDMS-PMOXA vesicles to target

macrophages. Their goal was to promote macrophage uptake of the vesicles by

targeting the scavenger receptor A1 from macrophages with polyA. Unliganded

vesicles were not taken up. Nevertheless, the uptake of nanoreactors by

macrophages demands further investigation.

Comparison and characterisation of PMOXA-PDMS-PMOXA vesicles

111

4.5 references

Agterberg, M., Adriaanse, H., Lankhof, H., Meloen, R. and Tommassen, J. (1990) Outer membrane PhoE protein of Escherichia coli as a carrier for foreign antigenic determinants: immunogenicity of epitopes of foot-and- mouth disease virus. Vaccine, 8, 85-91.

Broz, P., Marsch, S. and Hunziker, P. (2007) Targeting of vulnerable plaque macrophages with polymer-based nanostructures. Trends Cardiovasc Med., 17, 190-196.

Garavito, R.M. and Rosenbusch, J.P. (1986) Isolation and crystallization of bacterial porin. Methods Enzymol, 125, 309-328.

Martinez, F., Sica, A., Mantovani, A. and Locati, M. (2008) Macrophage activation and polarization. Front Biosci. , 13, 453-461.

Nardin, C. and Meier, W. (2001) Polymerizable amphiphilic block copolymers: From nanostructured hydrogels to nanoreactors and ultrathin films. Chimia, 55, 142- 146.

Parkin, D.W. (1996) Purine-specific nucleoside N-ribohydrolase from Trypanosoma brucei brucei. Purification, specificity, and kinetic mechanism. J Biol Chem, 271, 21713- 21719.

Prilipov, A., Phale, P.S., Van Gelder, P., Rosenbusch, J.P. and Koebnik, R. (1998) Coupling site-directed mutagenesis with high-level expression: large scale production of mutant porins from E. coli. FEMS Microbiol Lett, 163, 65-72.

Ranquin, A., Versees, W., Meier, W., Steyaert, J. and Van Gelder, P. (2005) Therapeutic Nanoreactors: Combining Chemistry and Biology in a Novel Triblock Copolymer Drug Delivery System. Nano Lett, 5, 2220-2224.

Sauer, M. and Meier, W. (2004) Polymer nanocontainers for drug delivery. In, pp. 224- 237.

Versees, W., Decanniere, K., Pelle, R., Depoorter, J., Brosens, E., Parkin, D.W. and Steyaert, J. (2001) Structure and function of a novel purine specific nucleoside hydrolase from Trypanosoma vivax. J Mol Biol, 307, 1363-1379.

Zell, R. and Fritz, H.J. (1987) DNA mismatch-repair in Escherichia coli counteracting the hydrolytic deamination of 5-methyl-cytosine residues. Embo J, 6, 1809-1815.

Therapeutic nanoreactors: combining chemistry and biology in a novel triblock copolymer drug delivery system

113

Chapter 5:

Therapeutic nanoreactors:

combining chemistry and biology in a novel

triblock copolymer drug delivery system

Ranquin An, Versées Wim, Meier Wolfgang, Steyaert Jan and Patrick Van Gelder

Abstract Triblock copolymeric nanoreactors are introduced as an alternative for liposomes as encapsulating carrier for prodrug activating enzymes. Inosine–adenosine–guanosine preferring nucleoside hydrolase of Trypanosoma vivax, a potential prodrug activating enzyme, was encapsulated in nanometer-sized vesicles constructed of PMOXA-PDMS-PMOXA triblock copolymers. The nanoreactor is functionalized by incorporation of bacterial porins, OmpF or Tsx, in the reactor wall. Efficient cleavage of three natural substrates and one prodrug, 2-fluoroadenosine by the nanoreactors, was demonstrated. Part of this work was published in Nanoletters 5(11) (2005) 2220-2224

Chapter 5

114

Therapeutic nanoreactors: combining chemistry and biology in a novel triblock copolymer drug delivery system

115

5.1 Introduction

Currently chemotherapy is one of the major strategies to treat cancer

patients. Despite its success, it is limited by several drawbacks such as low

bioavailability of the chemotoxin, low drug concentrations at the tumor site,

systemic toxicity, lack of specificity and the appearance of drug-resistant tumors.

To overcome these problems many different strategies have been developed

including improved drug formulations (e.g. liposomes) (Vail et al., 2004),

resistance modulators (Bauer et al., 2005) and antidote/toxicity modifiers

(Anderson, 2005). Selective local enzymatic activation of prodrugs is also a

possibility to increase drug concentrations in the tumor and decrease systemic

toxicity (Denny, 2001), thus improving the therapeutic index. Unfortunately

human tumors rarely express useful activating enzymes at high concentrations.

Therefore exogenous enzymes need to be used and directed to the tumor.

Directed enzyme-prodrug therapies involve two stages. In the first step the

activating enzyme is directed to the tumor. In the second step the non-toxic

prodrug is systemically administered and subsequently converted in high local

concentrations of an anticancer drug by the enzyme at the tumor site. The

targeting of the enzyme can either be mediated by antibodies, termed antibody-

directed enzyme-prodrug therapy (ADEPT) or by a gene-vector, termed gene-

directed enzyme-prodrug therapy (GDEPT). Both ADEPT and GDEPT suffer from

the same shortcomings. First, most activating enzymes are immunogenic.

Second, the efficient targeting remains an obstacle and finally, most prodrugs

are also activated by endogenous enzymes.

Several strategies have been proposed to decrease the immunogenicity of the

activating enzyme such as antibody-directed abzyme prodrug therapy (ADAPT)

(Kakinuma et al., 2002). Abzymes are catalytic antibodies that are raised against

a stable transition state analog and can be humanized to reduce immunogenicity.

Their catalytic activity however is usually 1000 fold lower compared to enzymes

that catalyze the same reaction (Avalle et al., 2000). Another way to avoid

immunogenicity is polymer-directed enzyme-prodrug therapy (PDEPT) (Satchi et

al., 2001). In this approach a polymer-drug conjugate is injected first and

Chapter 5

116

accumulates in the tumor tissue by a mechanism called enhanced permeability

and retention effect (EPR) (Matsumura and Maeda, 1986). Afterwards an

enzyme-polymer conjugate is injected that activates the prodrug at the tumor

site. The polymer shields the enzyme from its environment, thereby reducing

immunogenicity. Unfortunately, synthesis of polymer-enzyme conjugates

typically have a low yield and reduced activity of the enzyme (Ashihara et al.,

1978).

We recently proposed a novel strategy in the form of therapeutic nanoreactors

to avoid immunogenicity without loss of enzyme activity (Huysmans et al.,

2005). In this scheme the enzyme is shielded from the immune system by

encapsulation in liposomes. To make these liposomes functional, they were

permeabilized by channel forming proteins to allow diffusion of the substrate and

product but not the enzyme through the reactor wall. Unfortunately, these

reactors showed some uncontrollable characteristics, potentially leading to

immunogenicity of these liposomes. Furthermore, liposome based carriers are

unstable in blood serum and need to be grafted with PEG or other polymers to

increase the average circulation time in the body (Cattel et al., 2004).

In this study we introduce a more promising kind of nanoreactor in which the

reactor wall is composed of the amphiphilic ABA triblock copolymer, poly(2-

methyloxazoline)-poly(dimethylsiloxane)-poly(2-methyloxazoline) (PMOXA-

PDMS-PMOXA) (Figure 1). This copolymer has an average molecular weight of

8660 g/mol and self assembles to form stable vesicular structures in aqueous

solutions (Graff et al., 2002). Containers with controlled mean diameters of 200

nm can be obtained by successive extrusion. Vesicles made of this triblock

copolymer are more stable and less permeable, especially in dilute solutions,

compared to liposomes (Nardin and Meier, 2001). This is due to the higher

length of the hydrophobic block of the polymer, slower dynamics and

intermolecular steric stabilization. Furthermore, these vesicles are completely

covered with PMOXA that has similar stealth properties as PEG. Stealth

properties allow vesicles to escape clearance by the immune system because of

low adsorption of plasma proteins and low hepatosplenic uptake (Woodle et al.,

1994). This results in longer blood circulation times. Also, non specific uptake of

Therapeutic nanoreactors: combining chemistry and biology in a novel triblock copolymer drug delivery system

117

non permeabilized PMOXA-PDMS-PMOXA nanocontainers by COS-7 cells and

THP-1-derived macrophages is completely absent in vitro (Broz et al., 2005),

which makes these nanocontainers biocompatible and good candidates for in vivo

use.

Figure 1: Schematic representation of a functionalized nanoreactor build up of PMOXA-PDMS-PMOXA, permeabilized by the bacterial outer membrane protein Tsx and encapsulated with Trypanosoma vivax Nucleoside Hydrolase (TvNH).

Despite a polymeric membrane thickness of 10 nm, three times wider than a

biological lipid membrane, channel forming proteins have previously been

incorporated in PMOXA-PDMS-PMOXA vesicles without loss of function (Meier et

al., 2000). This is probably due to the high flexibility of the hydrophobic PDMS

block and the presence of shorter polymers which segregate around the protein.

A recent mean field analysis also indicated that protein incorporation only causes

a minor energy penalty in the polymeric membrane (Pata and Dan, 2003). We

permeabilized our nanoreactors by incorporation of two different bacterial outer

membrane, channel forming proteins, called porins (OmpF and Tsx) in the vesicle

wall. OmpF forms a 16-stranded transmembrane β-barrel that functions as a

molecular sieve, allowing concentration driven diffusion of solutes < 600 Da

(Koebnik et al., 2000). Tsx on the other hand forms a 12-stranded β-barrel and

allows specific transport of nucleosides and nucleotides (Ye and Van Den Berg,

Chapter 5

118

2004). Since it has a binding site for nucleosides in the interior of the channel,

rapid transport of nucleosides at very low concentrations is possible compared to

slow diffusion through the nonspecific porin OmpF. By permeabilizing the

nanoreactors with OmpF or Tsx, small substrates can be transported across the

vesicle membrane to reach the interior where they are activated by the enzyme.

Subsequently products can diffuse out of the vesicle to the exterior.

As a proof of principle for our system we chose the purine specific nucleoside

hydrolase of Trypanosoma vivax (TvNH) as prodrug activating enzyme. This

enzyme is a member of the nucleoside hydrolase superfamily that catalyses the

hydrolysis of the N-glycosidic bond of β-ribonucleoside forming the free nucleic

base and ribose (Versees and Steyaert, 2003). These enzymes are widely

distributed in nature but they are not present in mammals. Since TvNH is purine

specific, its natural substrates are inosine, adenosine and guanosine.

Crystallographic data on the T. vivax enzyme showed that the ribose is tightly

bound to the enzyme with all its hydroxyl groups involved in multiple

stereoselective H-bonds (Versees et al., 2001). This makes the enzyme highly

specific towards the ribose moiety. The nucleic base in contrast is stacked

between two tryptophans residues and forms very few specific interactions with

the enzyme. Consequently TvNH is less specific towards the nucleic base moiety.

This feature makes TvNH a promising candidate for enzyme-prodrug strategies

since many known chemotoxins are nucleobase analogs (Klopfer et al., 2004;

Parker et al., 1998).

To explore the possibility of using TvNH encapsulating nanoreactors as

prodrug activating moieties, we measured activities of such nanoreactors for

three natural substrates and one prodrug 2-fluoroadenosine.

Therapeutic nanoreactors: combining chemistry and biology in a novel triblock copolymer drug delivery system

119

5.2 Materials and methods

5.2.1. Purification of TvNH

Purification of the wild type TVNH of T. vivax was performed as described by

Versées et al. (Versees et al., 2001). Shortly, TVNH was purified in a two step

purification scheme from E. coli strain WK6 (Zell and Fritz, 1987) containing the

gene for the enzyme in a pQE-30 expression vector (Qiagen). In a two-step

purification protocol, the presence of an N-terminal His6-tag allowed recovery of

the protein on a Ni-NTA affinity column (Qiagen). pH gradient step elution from

the affinity column was followed by gel filtration on a superdex-200 column

(Amersham Bioscience). SDS polyacrylamide gel electrophoresis confirmed the

purity of the enzyme.

5.2.2. Purification of E. coli porins OmpF and Tsx

BL21(DE3) ∆lamB ompR ∆ompA strains were used for overproduction of OmpF

and Tsx as described by Prilipov et al. (Prilipov et al., 1998). For this purpose,

the gene encoding Tsx was cloned in the pGompf plasmid. For purification

purposes an N-terminal 6His-tag was added. Overproduced porins were

subsequently purified by combining detergent extraction methods of Agterberg et

al. (Agterberg et al., 1990) and Garavito and Rosenbusch (Garavito and

Rosenbusch, 1986). After cell harvesting by centrifugation, the cell pellets were

resuspended in 2% SDS 20 mM Tris pH 8.0 and shaken for 1 h at 60 °C to obtain

cell lysis. Lysates were pelleted by ultracentrifugation (20,000xg, 30 min, 4 °C)

and resuspended in 0.125% oPOE 20 mM NaH2PO4 pH 7.3 for pre-extraction at

37 °C for 45 min. After a second cycle of ultracentrifugation, pellets were

resuspended in the same phosphate buffer with 3% oPOE for overnight

extraction at 4°C and an additional 45 min at 37°C. Finally, 10 mM EDTA was

added to the supernatant after ultracentrifugation, obtaining stripping of

lipopolysaccharides. Purified proteins were concentrated (amiconR MWCO 5000)

and dialyzed (Spectra/Por MWCO 6500-8000) against 20 mM NaH2PO4 150 mM

NaCl 1% oPOE pH 7.3. Since Tsx was expressed with an N-terminal 6His-tag, it

was further purified on a Ni-NTA affinity column (Qiagen) and gel filtration on a

Chapter 5

120

superdex-200 column (Amersham Bioscience) to remove contaminating proteins.

SDS-polyacrylamide gel electrophoresis ascertained the purity of both porins.

5.2.3. Preparation of nanoreactors

To produce permeabilized nanoreactors, a 1 % (w/v) polymer (PMOXA18-

PDMS72-PMOXA18)-ethanol solution was mixed with water solubilised porins,

purificated according to Prilipov et al. (Prilipov et al., 1998). Typically 250 µl of

polymer-ethanol solution was mixed with porin solution to give a final porin

concentration of 0.01 µg/µl or 0.1 µg/µl. This results in a molecular ratio of porin

to polymer of 1:100 and 1:10 respectively. Since a vesicle with a diameter of

about 200 nm contains about 13000 triblock copolymer molecules, this results in

130 to 1300 porin molecules per vesicle (Nardin, 2000). The solution was then

dried to produce a lamellar polymer/porin film. This film was subsequently

rehydrated for several hours under continues stirring in 20 mM Hepes, pH 7.0

containing ± 50 µM of the prodrug activating enzyme TvNH. To obtain a

monodisperse nanoreactor sample, the solution was extruded several times

through a polycarbonate filter with a pore diameter of 200 nm. Since TvNH has

a 6His-tag, non encapsulated prodrug activating TvNH was removed on a Ni-NTA

affinity column (Amersham Biosciences). After each step the samples were

analyzed by DLS to determine the size of the nanoreactors and the polydispersity

of the sample (laser-spectroscatter 201 by RiNA GmbH, Berlin, Germany) at 532

nm with a scattering angle of 90°. Ten data sets were recorded and the size

distribution was analyzed using the software CONTIN.

5.2.4. Trypsin digestion

To make sure that no free TvNH was present in the nanoreactor samples, a

Trypsin digestion was performed. The activity of samples of the nanoreactors

permeabilized with Tsx before Ni-NTA affinity chromatography, the flow through

and the elution was followed spectrophotometrically at 400nm using 1mM p-

nitrophenyl riboside as substrate. This activity was measured again after

incubation at 37°C for 1hour in the presence of Trypsin (16 units).

Therapeutic nanoreactors: combining chemistry and biology in a novel triblock copolymer drug delivery system

121

5.2.5. Activity measurements of nanoreactors

The enzymatic activity of the nanoreactors for three natural substrates

inosine, adenosine and guanosine- and one prodrug, 2-fluoroadenosine was

determined using a reducing sugar assay as described by Parkin (Parkin, 1996).

Briefly, the enzymatic reaction was stopped by adding a CuSO4–solution. The

Cu2+ is reduced to Cu+ by the reaction product ribose. This reduced Cu reacts

with neocuproin to form a complex. Color development (Yellow) of this complex

was achieved by heating the solution 8 min at 95°C and the OD at 450nm was

measured. A standard curve with known ribose concentration was used to

determine the extinction coefficient under the assay conditions.

5.3 Results

5.3.1. Kinetic parameters of TvNH

To ensure the possibility to use nanoreactors encapsulating TvNH as prodrug

activators, the kinetic parameters of TvNH for four prodrugs, 2-fluoroadenosine,

2-chloroadenosine 6-methylpurine riboside and 6-thioguanosine were determined

(Table 1). We found that these prodrugs are efficiently hydrolyzed to their

cytotoxic base with the same efficiency as the natural substrates. Thus, TvNH

can be used as a prodrug activating enzyme in combination with these prodrugs.

Table 1: Kinetic parameters of TvNH for natural substrates and prodrugs

kcat (s -1) KM (µM)

Inosine 5.19 ± 0.08 5.37 ± 0.42

Adenosine 2.6 ± 0.2 8 ± 1.8

Guanosine 2.31 ± 0.11 2.33 ± 0.47

2-fluoroadenosine 1.86 ± 0.11 39.05 ± 7.99

2-chloroadenosine 1.41 ± 0.05 4.56 ± 0.91

6-methylpurine riboside

4.34 ± 0.13 < 10

Chapter 5

122

5.3.2. Preparation of nanoreactors

Size distributions of nanoreactors containing 50 µM TVNH and reconstituted

porins OmpF and Tsx were determined with dynamic light scattering before and

after extrusion through a polycarbonate filter whit a pore diameter of 200 nm.

Before extrusion, reactors with a mean diameter of 362 nm were obtained

whereas the mean diameter decreases to 200nm after the extrusion. No

significant differences were found between OmpF (Figure 2) and Tsx (results not

shown) permeabilized reactors.

Figure 2: Analysis of the size-distributions of OmpF-functionalized nanoreactors by DLS before extrusion (A) and after extrusion (B). Measurements were carried out at 90° and 532 nm on a laser-spectroscatter 201 by RiNa GmbH, Berlin, Germany. Size distributions were determined by CONTIN analysis.

Nanoreactors were subsequently purified on a Ni-NTA column to remove non-

encapsulated TvNH. DLS measurements confirmed that this purification step has

no effect on the size distribution of the nanoreactors (results not shown).

To ensure that this purification step completely removes all non-encapsulated

TvNH, a Trypsin digestion was performed. Enzymatic activity of the sample

before Ni-NTA affinity chromatography, the flow through and the elution (eluted

with 500 µM imidazol) was measured before incubation with Trypsin. This activity

was stated as 100 % activity for each sample. After incubation for 1 hour with

Trypsin, the activity of the different samples was measured again (Figure 3).

Therapeutic nanoreactors: combining chemistry and biology in a novel triblock copolymer drug delivery system

123

Figure 3: Activity measurements of nanoreactors before and after Trypsin digestion. The activity of each sample before Trypsin digestion was stated as 100 % activity. The activity was measured spectrophotometrically at 400 nm with 500 µM of p-nitrophenyl riboside.

For the sample before Ni-NTA and the elution sample, the activity is greatly

decreased by Trypsin digestion which shows that the digestion was successful.

However it was not complete since there is still 55% and 28% of the activity

present, respectively. For the purified nanoreactors (sample flow through)

digestion with Trypsin has no effect on the activity which means that the enzyme

is protected by encapsulation in the polymeric vesicles and that no non-

encapsulated TvNH was present in the sample.

5.3.3. Encapsulation efficiency of TvNH

Since TvNH is deactivated by detergents (see chapter 3), it was impossible to

determine the encapsulated enzyme concentration by dissolving the nanoreactor

and measuring the enzyme activity. Therefore the efficiency of encapsulation was

determined by comparative SDS-page analysis. Known quantities of free TvNH

(0.1-3 µg) were run on an SDS-PAGE together with an aliquot of the nanoreactor

sample (Figure 4). To assure complete disruption of the nanoreactors, 0.5 %

TritonX-100 was added to the sample. The intensity of each coomassie blue

stained TvNH band was measured with Intelligent Quantifier software (Bio Image

Systems Inc., Michigan, USA) to determine the quantity of TvNH in the

nanoreactor sample. A standard curve of band intensity versus known amount of

TvNH was calculated and used to determine the quantity of TvNH in the

nanoreactor sample. This leads to an estimated encapsulation efficiency of 15 %.

0

20

40

60

80

100

120

Before NiNta Flow through Ni-NTA Elution Ni-NTA

% a

ctiv

ity

Chapter 5

124

Two protein bands are visible in the nanoreactor sample corresponding to TvNH

and porin. The porin band however is running at a higher molecular weight than

the purified porin, probably due to strong interaction between polymer and porin.

Therefore we couldn’t use comparative SDS-page analysis to determine the

amount of incorporated porin.

Figure 4: SDS -PAGE analysis for determination of the encapsulation efficiency of TvNH. Free TvNH (0.1-3 µg) was run together with an aliquot of nanoreactor sample (X).

5.3.4. Activity of nanoreactors

To determine the activity of the nanoreactors and compare it to the free

enzyme, we used three natural substrates of TvNH: inosine, guanosine and

adenosine. Additionally we measured the activity of 2-fluoroadenosine to

evaluate the prodrug activating properties of the nanoreactors. 2-

fluoroadenosine is similar to the known prodrug 2-fluoro-2’-deoxyadenosine

which is used in combination with E.coli purine nucleoside phosphorylase (PNP)

in ADEPT and GDEPT therapies (Parker et al., 2003). Activities for all substrates

were measured using a reducing sugar assay described by Parkin et al. (Parkin,

1996)

First of all the activity of non permeabilized nanoreactors with encapsulated

TvNH was measured (data not shown). These nanoreactors showed no activity at

all which is an improvement compared to the liposomal nanoreactors (Huysmans

et al., 2005). For the permeabilized nanoreactors, the rate of product formation

V was determined for various substrate concentrations (0-1000 µM) and for porin

ratios 1:100 and 1:10 OmpF and 1:100 Tsx. These rates were fitted to a

hyperbolic curve to determine apparent kinetic constants (Figure 5) (Table 2).

For fitting purposes, at least 10 individual points were measured per experiment.

NH (µg) 3 2.5 1.5 1 0.5 0.3 0.2 0.1 X

OmpF

TvNH

NH (µg) 3 2.5 1.5 1 0.5 0.3 0.2 0.1 X

OmpF

TvNH

Therapeutic nanoreactors: combining chemistry and biology in a novel triblock copolymer drug delivery system

125

A B

C

Figure 5: Product formation rate of nanoreactors permeabilized with ratios 1:100 OmpF (B) or Tsx (C) to polymer and ratio 1:10 OmpF (A) to polymer in function of substrate concentration. The substrates used are inosine (■), adenosine (●), guanosine (▲) and 2-fluoroadenosine (▼).The data were fitted to a hyperbolic curve

Chapter 5

126

Table 2: Kinetic properties of nanoreactors permeabilized with OmpF or Tsx

The (KM)app value corresponds to the exterior substrate concentration at which

the rate of product formation (V) is half its maximal value (Vmax). Vmax divided by

the total enzyme concentration equals (kcat)app.

In all cases we saw that (KM)app > KM,enzyme. This observation indicates that the

substrate concentration is higher outside the nanoreactor as compared to inside.

Furthermore, the activity of the nanoreactors is a function of the porin to

polymer ratio in the reactor wall. When this ratio increases 10 times (from 1:100

to 1:10) for OmpF permeabilized nanoreactors, the (KM)app and (kcat)app also

increase approximately 10 times. This observation is also seen when Tsx, a

nucleoside specific transporter, is used. All data indicate that the (kcat)app value is

correlated to the amount and the nature of the porin used. It thus appears that

the activity of the nanoreactors is limited by transport at low porin

concentrations. At higher porin concentrations or when the specific transporter

Tsx is used, the (kcat)app value of the nanoreactors is higher than that of the free

enzyme. This might indicate that under these conditions, the activity of the

nanoreactors is no longer limited by transport. The exact reason for this

exceptionally high activity of the nanoreactors is as yet not understood.

Inosine Adenosine Guanosine 2-Fadenosine

kcat (s -1) 5.2 ± 0.1 2.6 ± 0.2 2.3 ± 0.1 1.9 ± 0.1 (TvNH)free

KM (µM) 5.4 ± 0.4 8.0 ± 1.8 2.3 ± 0.5 39.1 ± 8.0 (kcat)app (s -1) 2.3 ± 0.1 2.2 ± 0.1 4.3 ± 0.2 0.3 ± 0.1 1/100 OmpF

nanoreactors (KM)app (µM) 356 ± 23 602 ± 71 582 ± 57 396 ± 169 (kcat)app (s -1) 37.5 ± 1.9 12.1 ± 0.8 52.8 ± 3.1 2.9 ± 0.3 1/10 OmpF

nanoreactors (KM)app (µM) 84 ± 15 41 ± 12 125 ± 23 35 ± 19 (kcat)app (s -1) 24.1 ± 1.5 19.0 ± 1.5 21.5 ± 1.1 1.9 ± 0.1 1/100 Tsx

nanoreactors (KM)app (µM) 80 ± 18 24 ± 17 100 ± 18 38 ± 10

Therapeutic nanoreactors: combining chemistry and biology in a novel triblock copolymer drug delivery system

127

5.4 Discussion

In this study we propose a new therapeutic tool based on nanoreactors that

are composed of PMOXA18-PDMS72-PMOXA18 triblock copolymers. These

nanoreactors were functionalized by encapsulating the prodrug activating

enzyme TvNH and permeabilizing the reactor wall with bacterial membrane

porins OmpF and Tsx.

We were able to produce a monodisperse sample of nanoreactors with a mean

size of 200nm. Since particles with a size of 10-500 nm are able to extravasate

through the leaky blood vasculature of tumor tissue, this is a good size for

passive targeting to tumors via the enhanced permeability and retention effect

(Jain, 1987).

We furthermore proved that the encapsulated enzyme is completely protected

from degradation by proteases which is an enormous advantage compared to

ADEPT strategies were the prodrug activating enzyme is exposed to the

environment.

It was demonstrated that these nanoreactors can efficiently hydrolyze

different substrates including the prodrug 2-fluoroadenosine, resulting in the

release of the cytotoxic molecule, 2-fluoroadenine. Furthermore, nanoreactors

that were not permeabilized by porins showed no activity at all which is a great

improvement compared to liposomal nanoreactors (Huysmans et al., 2005). For

nanoreactors with a high porin content or when the nucleoside specific

transporter Tsx is used we found an (kcat)app that is higher than the kcat of the

free enzyme. It seems unlikely that encapsulation of the enzyme increases the

turn over rate. Therefore this high (kcat)app is probably caused by an

overestimation of the amount of encapsulated enzyme. This means that the

encapsulation efficiency is greater than calculated via the comparative SDS-PAGE

analysis (Knight and Chambers, 2003).

Evidently, these nanoreactors are flexible systems that can be used with

different enzyme/substrate combinations and targeted to different tumor tissues

or organs. Depending on the functionalities of the nanoreactors, they could be

applied in other fields than cancer therapy such as gene-, RNAi- or drug delivery,

diagnostics and in vivo imaging.

Chapter 5

128

5.5 References

Agterberg, M., Adriaanse, H., Lankhof, H., Meloen, R. and Tommassen, J. (1990) Outer membrane PhoE protein of Escherichia coli as a carrier for foreign antigenic determinants: immunogenicity of epitopes of foot-and- mouth disease virus. Vaccine, 8, 85-91.

Anderson, B. (2005) Dexrazoxane for the prevention of cardiomyopathy in anthracycline treated pediatric cancer patients. Pediatr Blood Cancer, 44, 584-588.

Ashihara, Y., Kono, T., Yamazaki, S. and Inada, Y. (1978) Modification of E. coli L-asparaginase with polyethylene glycol: disappearance of binding ability to anti-asparaginase serum. Biochem Biophys Res Commun, 83, 385-391.

Avalle, B., Friboulet, A. and Thomas, D. (2000) Catalysis by anti-idiotypic antibodies. Chem Immunol, 77, 80-88.

Bauer, K.S., Karp, J.E., Garimella, T.S., Wu, S., Tan, M. and Ross, D.D. (2005) A phase I and pharmacologic study of idarubicin, cytarabine, etoposide, and the multidrug resistance protein (MDR1/Pgp) inhibitor PSC-833 in patients with refractory leukemia. Leuk Res, 29, 263-271.

Broz, P., Benito, S.M., Saw, C., Burger, P., Heider, H., Pfisterer, M., Marsch, S., Meier, W. and Hunziker, P. (2005) Cell targeting by a generic receptor-targeted polymer nanocontainer platform. J Control Release, 102, 475-488.

Cattel, L., Ceruti, M. and Dosio, F. (2004) From conventional to stealth liposomes: a new Frontier in cancer chemotherapy. J Chemother, 16 Suppl 4, 94-97.

Denny, W. (2001) Prodrug strategies in cancer therapy. European Journal of medicinal chemistry, 36, 577-595.

Garavito, R.M. and Rosenbusch, J.P. (1986) Isolation and crystallization of bacterial porin. Methods Enzymol, 125, 309-328.

Graff, A., Sauer, M., Van Gelder, P. and Meier, W. (2002) Virus-assisted loading of polymer nanocontainer. Proc Natl Acad Sci U S A, 99, 5064-5068. Epub 2002 Mar 5026.

Huysmans, G., Ranquin, A., Wyns, L., Steyaert, J. and Van Gelder, P. (2005) Encapsulation of therapeutic nucleoside hydrolase in functionalised nanocapsules. J Control Release, 102, 171-179.

Jain, R.K. (1987) Transport of molecules in the tumor interstitium: a review. Cancer Res, 47, 3039-3051.

Kakinuma, H., Fujii, I. and Nishi, Y. (2002) Selective chemotherapeutic strategies using catalytic antibodies: a common pro-moiety for antibody-directed abzyme prodrug therapy. journal of immunological methods, 269, 269-281.

Klopfer, A., Hasenjager, A., Belka, C., Schulze-Osthoff, K., Dorken, B. and Daniel, P.T. (2004) Adenine deoxynucleotides fludarabine and cladribine induce apoptosis in a CD95/Fas receptor, FADD and caspase-8-independent manner by activation of the mitochondrial cell death pathway. Oncogene, 23, 9408-9418.

Knight, M.I. and Chambers, P.J. (2003) Problems associated with determining protein concentration: a comparison of techniques for protein estimations. Mol Biotechnol, 23, 19-28.

Koebnik, R., Locher, K.P. and Van Gelder, P. (2000) Structure and function of bacterial outer membrane proteins: barrels in a nutshell [In Process Citation]. Mol Microbiol, 37, 239-253.

Matsumura, Y. and Maeda, H. (1986) A new concept for macromolecular therapeutics in cancer chemotherapy: mechanism of tumoritropic accumulation of proteins and the antitumor agent smancs. Cancer Res, 46, 6387-6392.

Meier, W., Nardin, C. and Winterhalter, M. (2000) Reconstitution of Channel Proteins in (Polymerized) ABA Triblock Copolymer Membranes Angew Chem Int Ed Engl., 39, 4599-4602.

Nardin, C., Hirt, T., Leukel, J., Meier, W. (2000) Polymerized ABA Triblock Copolymer Vesicles. Langmuir, 16, 1035-1041.

Therapeutic nanoreactors: combining chemistry and biology in a novel triblock copolymer drug delivery system

129

Nardin, C. and Meier, W. (2001) Polymerizable amphiphilic block copolymers: From nanostructured hydrogels to nanoreactors and ultrathin films. Chimia, 55, 142-146.

Parker, W.B., Allan, P.W., Hassan, A.E., Secrist, J.A., 3rd, Sorscher, E.J. and Waud, W.R. (2003) Antitumor activity of 2-fluoro-2'-deoxyadenosine against tumors that express Escherichia coli purine nucleoside phosphorylase. Cancer Gene Ther, 10, 23-29.

Parker, W.B., Allan, P.W., Shaddix, S.C., Rose, L.M., Speegle, H.F., Gillespie, G.Y. and Bennett, L.L., Jr. (1998) Metabolism and metabolic actions of 6-methylpurine and 2-fluoroadenine in human cells. Biochem Pharmacol, 55, 1673-1681.

Parkin, D.W. (1996) Purine-specific nucleoside N-ribohydrolase from Trypanosoma brucei brucei. Purification, specificity, and kinetic mechanism. J Biol Chem, 271, 21713-21719.

Pata, V. and Dan, N. (2003) The effect of chain length on protein solubilization in polymer-based vesicles (polymersomes). Biophys J, 85, 2111-2118.

Prilipov, A., Phale, P.S., Van Gelder, P., Rosenbusch, J.P. and Koebnik, R. (1998) Coupling site-directed mutagenesis with high-level expression: large scale production of mutant porins from E. coli. FEMS Microbiol Lett, 163, 65-72.

Satchi, R., Connors, T.A. and Duncan, R. (2001) PDEPT: polymer-directed enzyme prodrug therapy. I. HPMA copolymer-cathepsin B and PK1 as a model combination. Br J Cancer, 85, 1070-1076.

Vail, D.M., Amantea, M.A., Colbern, G.T., Martin, F.J., Hilger, R.A. and Working, P.K. (2004) Pegylated liposomal doxorubicin: proof of principle using preclinical animal models and pharmacokinetic studies. Semin Oncol, 31, 16-35.

Versees, W., Decanniere, K., Pelle, R., Depoorter, J., Brosens, E., Parkin, D.W. and Steyaert, J. (2001) Structure and function of a novel purine specific nucleoside hydrolase from Trypanosoma vivax. J Mol Biol, 307, 1363-1379.

Versees, W. and Steyaert, J. (2003) Catalysis by nucleoside hydrolases. Curr Opin Struct Biol, 13, 731-738.

Woodle, M.C., Engbers, C.M. and Zalipsky, S. (1994) New amphipatic polymer-lipid conjugates forming long-circulating reticuloendothelial system-evading liposomes. Bioconjug Chem, 5, 493-496.

Ye, J. and Van Den Berg, B. (2004) Crystal structure of the bacterial nucleoside transporter Tsx. EMBO J, 23, 3187-3195. Epub 2004 Jul 3122.

Zell, R. and Fritz, H.J. (1987) DNA mismatch-repair in Escherichia coli counteracting the hydrolytic deamination of 5-methyl-cytosine residues. Embo J, 6, 1809-1815.

Chapter Nanoreactor mediated prodrug activation and killing of neuroblastoma cells

131

Chapter 6: Nanoreactor

mediated prodrug activation and killing

of neuroblastoma cells

Abstract

In this study we introduce a novel enzyme-prodrug system as an alternative to antibody-directed enzyme-prodrug therapy (ADEPT), based on a nanoreactor concept. The nanoreactor is a triblock copolymeric vesicle composed of poly(2-methyloxazoline)-poly(dimethylsiloxane)-poly(2-methyloxazoline) (PMOXA-PDMS-PMOXA) that encapsulates the prodrug activating enzyme nucleoside hydrolase of Trypanosoma vivax (TvNH). To permeabilize this vesicle for substrates and products the bacterial nucleoside specific porin Tsx, is incorporated in the polymer membrane. After screening of three compounds: 2-fluoroadenosine, 2-chloroadenosine and 6-thioguanosine we found that 6-thioguanosine is significantly less toxic than the nucleobase 6-thioguanine. The EC50’s are 9,2 and 4,8 µM for 6-thioguanosine after 48 h and 72h of incubation respectively and 3,3 and 1,4 µM for 6-thioguanine after 48 h and 72 h of incubation respectively. We further evaluated the cytotoxic effect of the nanoreactors in combination with the prodrug 6-thioguanosine on neuroblastoma cell cultures and found that 0,33 mg/ml of nanoreactors and 5 µM of 6-thioguanosine is sufficient to kill 80 % of the cell population after 72 h of incubation.

Manuscript in preparation

Chapter 6

132

Nanoreactor mediated prodrug activation and killing of neuroblastoma cells

133

6.1 Introduction

Enzyme-prodrug therapies were developed to selectively convert a relatively

non-toxic prodrug to an active drug at the tumor site. This leads to a higher local

drug concentration in the tumor thus improving the anti tumor effect.

Additionally, this lowers the systemic drug concentration hereby reducing

unwanted side effects that accompany conventional cancer chemotherapy. The

enzyme can either be delivered by an antibody-enzyme fusion protein

(antibodydirected enzyme-prodrug therapy, ADEPT) (Bagshawe et al., 2004) or

by a vector carrying the gene encoding for the exogenous enzyme (gene-

directed enzyme-prodrug therapy, GDEPT) (Dachs et al., 2005). Both therapeutic

strategies suffer from difficulties such as immunogenicity of the antibody-enzyme

conjugate in ADEPT systems or inefficient gene transfection, prolonged gene

expression, pathogenicity and immunogenicity in GDEPT strategies.

One way to reduce the immunogenicity of exogenous enzymes, predominantly

from bacterial origin (Senter and Springer, 2001) (Syrigos et al., 1998)

(Stribbling et al., 1997) that are used in enzyme-prodrug therapy is to

encapsulate the enzyme in liposomes. Thus, the enzyme is shielded from the

environment and protected from proteolysis and interaction with components of

the immune system. Such liposomes have to be grafted with polyethyleneglycol

(PEG) molecules to improve their stability in blood plasma and to increase their

circulation time. An alternative for liposome based systems is to make use of

block copolymers that are able to self-organize into nanometersized polymeric

vesicles. The polymeric vesicles are composed of the amphiphilic triblock

copolymer poly(2-methyloxazoline)-poly(dimethylsiloxane)-poly(2-

methyloxazoline) (PMOXA-PDMS-PMOXA) (chapter 2 Figure 1). It was

demonstrated by Nardin et al. that PMOXA-PDMS-PMOXA membranes are

considerably more stable than conventional phospholipid membranes and posses

a high flexibility provided by the PDMS middle block (Nardin, 2000).

Furthermore, the PMOXA outer blocks posses similar stealth properties as PEG

which makes PMOXA-PDMS-PMOXA based vesicles promising for use in

biomedical applications (Woodle et al., 1994). Previously, Meier and co-workers

Chapter 6

134

were able to reconstitute bacterial channelforming proteins such as OmpF

(Nardin et al., 2001) and LamB (Graff et al., 2002) in the triblock copolymer

shell. This way, molecules are able to enter and leave the vesicle’s interior. This

finding opened the door to the development of nanoreactors which can be used

in cancer therapy. Many different enzymes have been used in enzyme-prodrug

strategies including Escherichia cloaca β-lactamase (Cortez-Retamozo et al.,

2004), herpes simplex thymidine kinase (Iwai et al.2002; Mathis et al., 2006), E.

coli nitroreductase (Bilsland et al., 2003; Plumb eal., 2001), yeast cytosine

deaminase (Zeng et al., 2007) and E. coli purinnucleoside phosphorylase (Parker

et al., 2003; Zhou et al., 2007)

We have recently introduced a new strategy for enzyme-prodrug therapy

(Ranquin et al., 2005) (Figure 1) based on nanoreactors. The nucleoside

hydrolase of Trypanosoma vivax (TvNH), was encapsulated inside the polymeric

vesicle (Versees et al., 2001). Although Trypanosoma vivax nucleoside hydrolase

(TvNH) catalyses the hydrolysis of nucleosides and several nucleoside analogues

(Ranquin et al., 2005), it has never been tested as a prodrug activating enzyme.

To make sure substrates and products can readily diffuse in and out of

thepolymeric vesicles, Tsx, a nucleoside specific porin, was incorporated into

thepolymeric membrane. We previously showed that such vesicles were able

tohydrolyse Inosine, adenosine, guanosine and the nucleoside analog 2-

fluoroadenosine (Ranquin et al., 2005).

In this study we validate such a nanoreactor system as an alternative toADEPT

and GDEPT. We have tested the cytotoxic capacity of these nanoreactorson N2A

neuroblastoma cells. To that purpose we have screened severalnucleoside

analogs that are hydrolysed by TvNH into cytotoxic nucleobaseanalogs as

potential prodrugs. 2-fluoroadenine and 2-chloroadenine are knowncytotoxic

agents that are phosporylated by cells and exert their cytotoxic actionat the

nucleoside triphosphate level (Parker et al., 1988). Moreover, 2-fluoro-2’-

deoxyadenosine is already used as a prodrug in combination with E. coli

purinenucleoside phosphorylase in GDEPT systems (Parker et al., 2003). Since

TvNH isunable to hydrolyse 2’-deoxyadenosine and 2’-deoxyadenosine analogs,

we explored whether 2-fluoroadenosine and 2-chloroadenosine can also serve as

prodrugs in combination with TvNH. 6-thioguanine is another known toxic

compound that has been used to treat acute lymphoblastic leukaemia (ALL) over

Nanoreactor mediated prodrug activation and killing of neuroblastoma cells

135

the last 45 years. It has also been used to treat inflammatory bowel disease

andorgan rejection after transplantation. Although little is known about the

cytotoxic effects of the nucleoside 6-thioguanosine, it is also a possible prodrug

candidate in our screening.

Figure 1: Schematic representation of a Tsx permeabilized nanoreactor that encapsulates TvNH and is composed of PMOXA-PDMS-PMOXA.

6.2 Materials and Methods

6.2.1. Nucleoside analogs and nucleobase analogs

2-Fluoroadenosine and 2-chloroadenosine were purchased from Sigma and 6-

thioguanosine was purchased from Acor. To obtain the respective nucleobases, a

complete hydrolysis was performed with TvNH and followed

spectrophotometrically at 280 nm for 2-fluoroadenosine, 262 nm for 2-

chloroadenosine and 355nm for 6-thioguanosine. . The reaction was carried out

with 1µM of TvNH at 37°C in PBS pH 7.4. After complete hydrolysis, TvNH was

Chapter 6

136

removed from the solutions by filtration on a Microcon centrifugal filter device

(Millipore) with a cutt-off of 30 kDa.

6.2.2. Cytotoxicity Assay

Neuroblastoma cells (N2A) were plated at a cell density of 104 cells per well in

100 µl Dulbecco’s Modified Eagle’s Medium / Nutrient Mixture F-12 Ham’s Liquid

Medium (DMEM/F12) (GIBCO)). The cells were allowed to adhere to the plates

and were grown at 37°C. After 24 h solutes were added and the cells were grown

at 37°C for an additional 24 h-72 h. The cell viability was determined using an in

vitro toxicity test based on neutral red uptake (purchased from Sigma). The dye

neutral red (Toluylene Red) is taken up by viable cells via active transport and

stored in lysosomes. In contrast non viable cells are unable to take up this dye.

In our cytotoxicity assays, 2µl of neutral red solution was added to each well and

uptake was allowed for 45 min. at 37°C. Subsequently the medium was removed

and the wells were washed with 100 µl of Hanks Balanced Salt Solution (HBSS)

to remove residual dye. The neutral red that was taken up by viable cells was

then solubilised by adding 100 µl of solubilisation solution (50 % ethanol, 1%

Acetic acid). After 10 min of solubilisation at room temperature, the absorbance

at 540 nm of the dye was measured spectrophotometrically. Each condition was

measured fivefold. The Neutral Red uptake for untreated control cells that were

grown in medium for 48 h and 72 h was stated as 100 % cell survival. 0,5 %

sodium azide was used as a positive control and incubated for 4 h. The dye

uptake for this condition was stated as 0 % cell survival.

6.2.3. EC50 determination

To determine the EC50 values of 6-thioguanosine and 6-thioguanine, cells

were cultured as described in 6.2.2. After 24 h of growth, concentrations ranging

from 0 to 200 µM of both compounds were added to the cells. Each concentration

was measured fivefold. After 48 h and 72 h cell survival was determined using

the neutral red in vitro toxicity assay as described in 6.2.2. The survival of

untreated control cells was stated as 100 %. The obtained data were fitted with

the logistic dose response equation from the programme Origin:

Nanoreactor mediated prodrug activation and killing of neuroblastoma cells

137

( )

⎥⎥⎦

⎤

⎢⎢⎣

⎡⎟⎟⎠

⎞⎜⎜⎝

⎛+

−+=

p

xx

AAAy

0

212

1

Where A1 and A2 are the maximum and minimum survival percentages

respectively, x0 is the concentration of solute at the midpoint of the sigmoid

curve, i.e. the concentration were half of the toxic effect is observed (EC50) and p

is the power number of the sigmoid curve.

6.2.4. Swelling assays with Tsx proteoliposomes

Proteoliposomes were produced as described by Nikaido and co-workers

(Luckey and Nikaido, 1980; Nikaido et al., 1991; Nikaido and Rosenberg, 1983).

Briefly, 1,2-Dioleoyl-sn-Glycero-3-phosphocholine (DOPC) and egg phosphatidyl

glycerol (EPG) were mixed in a 4:1 ratio and dissolved in diethyl ether.

Subsequently, a smooth film was formed under a gentle stream of nitrogen. After

drying, liposomes were spontaneously formed upon the addition of Phosphate

Buffered Saline (PBS). Proteoliposomes with reconstituted Tsx, were made by

rehydration in PBS that contained purified Tsx (20 µg ml-1). Subsequently,

liposomes were sonicated for 5 seconds in a water bath sonicator. Swelling

assays were performed as described previously by Nikaido (Nikaido and

Rosenberg, 1983). Liposome swelling upon the addition of substrate was

followed for 5 min spectrophotometrically at 500 nm and the transport flux was

determined as follows:

dtdA

Ai⎟⎠⎞

⎜⎝⎛=Φ

1

Where Φ is the transport flux, Ai the initial absorption and dA/dt the

absorption difference during the measured time. Adenosine, adenine, ribose, 6-

thioguanosine and 6-thioguanine were added to the proteoliposome solution at a

final concentration of 0.8 mM. Because of the low solubility of the tested solutes,

the swelling assay was performed under non-isotoneous concentrations.

Chapter 6

138

6.2.5. Preparation of TvNH encapsulated nanoreactors

Nanoreactors permeabilized by Tsx were prepared as follows: PMOXA21-

PDMS54-PMOXA21 (Polymer Source Inc.) was dissolved in ethanol at a

concentration of 8.3 % for 1h at room temperature. Subsequently, Tsx, cloned

and purified as described in chapter 4, was added at a final concentration of

0.1%. This ethanol solution was added drop-wise to a PBS solution containing

150 µM of TvNH (expressed and purified as described in chapter 3) under

continuous stirring. The solution was stirred for an additional 72 h at 4°C. To

remove non-encapsulated TvNH, 1ml of PBS-buffered Ni-NTA resin (GE

Healthcare) was added and allowed to bind his-tagged TvNH for 24 h. The

nanoreactors were separated from the Ni-NTA resin on a PD10 column. The size

and polydispersity of the nanoreactors was determined via DLS

(laserspectroscatter 201 by RiNA GmbH, Berlin, Germany) at 532 nm with a

scattering angle of 90°. Ten data sets were recorded and the size distribution

was analyzed using the software CONTIN.

6.2.6. Enzymatic activity Assay

To determine the activity of Tsx permeabilized nanoreactors and humane

purine nucleoside phosphorylase (hPNP) the rate of hydrolysis of 6-thioguanosine

was measured spectrophotometrically at 355nm. The rate of hydrolysis of

guanosine was measured spectrophotometrically at 300 nm. The reaction was

carried out in PBS, pH 6.6 at 37°C to measure the activity of nanoreactors. To

measure the activity of hPNP, the reaction was carried out in 50 mM Potassium

phosphate buffer pH 7.0. Different concentrations of 6-thioguanosine and

guanosine ranging from 0-1 mM were tested. The obtained reaction rates were

fitted to a Michaelis-Menten equation to determine Vmax and KM :

SKSV

VM +

×= max

Nanoreactor mediated prodrug activation and killing of neuroblastoma cells

139

6.2.7. Limulus Amebocyte Lysate (LAL) assay

To determine the endotoxin concentration in the fluorescent nanoreactor

sample, a LAL assay was performed (Cambrex). This test is based on the

activation of a protoenzyme present in the Limulus Amebocyte lysate by

endotoxin. This activated protoenzyme cleaves the colourless substrate Ac-Ile-

Glu-Ala-Arg-pNA to form p-nitroaniline (pNA). The pNA release is continually

measured spectrophotometrically at 405 nm throughout the incubation period.

The concentration of endotoxin in a given sample is calculated from its reaction

time by comparison to the reaction time of solutions containing known amounts

of endotoxin standard.

A sample of 1000X diluted nanoreactors was used to determine the endotoxin

concentration. To obtain a standard curve, E. coli O55:B5 derived endotoxin

(Cambrex) was used in the following concentrations: 0, 0.005, 0.05, 0.5, 5 en 50

endotoxin units (EU)/ml. As a monitor for product inhibition or enhancement, a

positive product control (PPC) was performed by spiking the nanoreactor samples

with 50 EU. Prior to adding the LAL, the samples were heated at 37 °C for 10

min. The calculations of the EU in the nanoreactor samples and the recovery of

EU in the PPC were done by WinKQCL software (Cambrex).

6.3 Results

6.3.1. Cytotoxicity screening of prodrugs and drugs

First we compared the cytotoxicity of prodrugs versus drugs on neuroblastoma

cell lines (N2A cells). The toxic effects on this cell line of 2-fluoroadenosine, 2-

chloroadenosine, 6-thioguanosine, 2-fluoroadenine, 2-chloroadenine and 6-

thioguanine were measured at two concentrations (25 µM and 2.5 µM). A

viability assay was performed after 24 h and 48 h incubation (Figure 2). We

found that 2-fluoroadenosine is very toxic for N2A cells with dye uptake

percentages of approximately 30 and 10 % at 2.5 and 25 µM respectively after

24 h incubation and a dye uptake percentage of 5 % after 48 h incubation

(Figure 2A). Furthermore, the cytotoxicity of the nucleoside is higher than the

cytotoxicity of the nucleobase which disqualifies this molecule as a good prodrug

candidate. 2-Chloroadenosine and 2-chloroadenine show only moderate cytotoxicity

Chapter 6

140

even at the highest concentration and longer incubation periods. As for 2-

fluoroadenosine, the cytotoxicity of 2-chloroadenosine is slightly higher than the

cytotoxicity of its nucleobase. As a result, 2-cloroadenosine is not a good prodrug

candidate (Figure 2B).

In case of 6-thioguanosine, both the nucleoside and nucleobase show no

significant toxicity after 24 h of incubation. After 48 h however, the cytotoxicity of

the nucleobase 6-thioguanine is significantly increased leading to cell survival

percentages of 30 % (Figure 2C). This delayed cytotoxicity is in good agreement with

the findings of Wotring and Roti Roti (Wotring and Roti Roti, 1980). They found that

incorporation of the active metabolite 6-thioGTP into newly synthesized DNA leads to

cell cycle arrest in the G2 phase. Therefore the cytotoxic action of 6-thioguanine only

becomes apparent after completion of the cell cycle in which the 6-thioGTP was

incorporated into the DNA. 6- Thioguanosine however, does not show a significant

increase in cytotoxicity after 48 h. This means that thioguanosine is a good prodrug

candidate for enzymeprodrug therapy.

Nanoreactor mediated prodrug activation and killing of neuroblastoma cells

141

A B

05

1015202530354045

25 µM/24 h 2.5 µM/24 h 25 µM/48 h 2.5 µM/48 h

% c

ell s

urvi

val

2-fluoroadenosine2-fluoroadenine

0102030405060708090

100

25 µM/24h 2.5 µM/24h 25 µM/48h 2.5 µM/48h

% c

ell s

urvi

val

2Cl-adenosine2Cl-adenine

C

0

20

40

60

80

100

120

25 µM/24 h 2.5 µM/24 h 25 µM/48 h 2.5 µM/48 h

% c

ell s

urvi

val

6-thioguanosine6-thioguanine

Figure 2: In vitro cytotoxicity screening of candidate prodrugs. (A) compares the cytotoxicity of the prodrug 2-fluoroadenosine to the drug 2-fluoroadenine, (B) compares the prodrug 2-chloroadenosine to the drug 2-chloroadenine and (C) compares the prodrug 6-thioguanosine to the drug 6-thioguanine (C) on N2A cell cultures. Cell survival was determined after 24h and 48h via a neutral red based toxicity assay.

6.3.2. EC50 determination of 6-thioguanosine and 6-thioguanine

Given that 6-thioguanosine is a promising prodrug candidate, we determined

the EC50 for the nucleoside analog, 6-thioguanosine, and the nucleobase analog,

6-thioguanine. To this end we measured the cell viability for concentrations of

both compounds, ranging from 0-200 µM. Since the cytotoxic action of 6-

thioguanine only becomes apparent after 24h, cell survival was measured after

Chapter 6

142

48 h and 72 h of incubation. Cell survival was determined using a Neutral Red

based in vitro toxicity assay (Figure 3, Table 1).

A B

Figure 3: EC50 determination of 6-thioguanosine (A) and 6-thioguanine (B) for N2A cells. Cell survival for several concentrations ranging from 0-200 µM after 48h (■) and 72h (▲) incubation was measured via a Neutral Red based toxicity assay. The data were fitted with the logistic dose response equation in Origin, to determine the EC50 values.

Table 1: EC50 values of 6-thioguanosine and 6-thioguanine.

6-thioguanosine 6-thioguanine

48 h 72 h 48 h 72 h

EC50 9,2 ± 1,3 4,8 ± 0,5 µM 3,3 ± 0,7 µM 1,4 ± 0,2 µM

For 6-thioguanosine cell survival showed an initial plateau phase with a start

of cell death after approximately 2 and 3 µM after 48 and 72 h, respectively. In

contrast 6-thioguanine showed an immediate decrease in cell viability. The EC50

of 6-thioguanine is approximately 3 times lower than the EC50 of 6-thioguanosine

(Table 1), indicating that 6-thioguanine is more toxic to N2A cells than 6-

thioguanosine. Especially at low concentrations (0-10 µM) 6-thioguanine exerts a

high cytotoxicity in comparison to 6-thioguanosine. When used in this

concentration range, 6-thioguanosine can be used as a prodrug.

Nanoreactor mediated prodrug activation and killing of neuroblastoma cells

143

6.3.3. 6-Thioguanosine activation by human purine nucleoside phosphorylase (hPNP)

To validate 6-thioguanosine as a prodrug for enzyme-prodrug therapy, it is

important to know that it can not be activated by an endogenous enzyme. This

would lead to toxic side effects and loss of the tumor specificity of the therapy.

Since 6-thioguanosine is also toxic for neuroblastoma cells at high concentrations

we believe that it is activated by the endogenous enzyme purine nucleoside

phosphorylase (PNP). PNP catalyses the cleavage of the glycosidic bond of ribo

and deoxyribonucleosides in the presence of inorganic orthophosphate (Pi) as a

second substrate to generate the purine base and ribo(deoxyribo)-1-phosphate.

Although PNP is a ubiquitous enzyme that is present in many organisms, the

substrate specificity may differ from species to species. Human PNP (hPNP) only

hydrolyses 6-oxopurine nucleosides such as guanosine and inosine or nucleoside

analogs that have an atom of similar electron density at the 6 position (Bzowska

et al., 2000). Therefore 6-thioguanosine may be hydrolyzed by hPNP to form 6-

thioguanine. This is particularly important since hPNP is highly expressed in

lymphocytes. This is well illustrated by the PNP deficiency syndrome in which

patients suffer from severe T-cell immunodeficiency or decreased B-cell function

(Giblett et al., 1975; Markert, 1991) In our experiments we only observed

toxicity of 6-thioguanosine at concentrations > 5 or 10 µM after 48 h and 72 h of

incubation, respectively. To determine whether this cytotoxicity is caused by

hydrolysis of 6-thioguanosine by hPNP, we determined the activity of this

enzyme for 6-thioguanosine and compared it to the activity for the natural

substrate guanosine. The rate of hydrolysis was measured

spectrophotometrically at 300 nm for guanosine and 355 nm for 6-thioguanosine

in 50 mM potassium phosphate buffer pH 7.0. Substrate concentrations ranging

from 0 to 1mM were tested (Figure 4).

Chapter 6

144

A B

Figure 4: Enzymatic activity of human PNP for guanosine (A) and 6-thioguanosine (B). Reactions were carried out in 50 mM potassium phosphate buffer pH 7.0 at 37°C. The data were fitted to a Michaelis–Menten equation.

The affinity of hPNP for 6-thioguanosine (451 µM) is much lower as compared

to guanosine (72 µM), illustrated by a higher KM value for 6- thioguanosine. But

because hPNP has a high turnover rate for 6-thioguanosine, hydrolysis of 6-

thioguanosine by hPNP might lead to lymphocyte cytotoxicity.

6.3.4. Transport of 6-thioguanosine and 6-thioguanine by Tsx

To validate 6-thioguanosine as a prodrug in combination with Tsx

permeabilized nanoreactors, it is important to test if 6-thioguanosine can reach

the encapsulated TvNH allowing activation of 6-thioguanosine to 6-thioguanine.

In addition the produced 6-thioguanine has to be transported to the exterior of

the nanoreactor after formation. Therefore the Tsx mediated transport of 6-

thioguanosine and 6-thioguanine was analysed. To this end we performed

swelling assays with proteoliposomes that incorporate Tsx. Swelling was

recorded spectrophotometrically at 500 nm for 5 min to calculate the flux of

different substrates. Adenosine, which is well transported by Tsx (Benz et al.,

1988), was used as a positive control. In addition, transport of adenine and

Nanoreactor mediated prodrug activation and killing of neuroblastoma cells

145

ribose by Tsx proteoliposomes was also measured (Figure 5). Our results indicate

that

Figure 5: Transport through Tsx proteoliposomes. The transport was determined via a swelling assay under non- isotoneous concentrations. Each solute was added to the proteoliposome solution at a final concentration of 0,8 mM. Swelling was measured spectrophotometrically at 500 nm for 5 min.

6.3.5. Kinetics of Tsx-TvNH nanoreactors

Nanoreactors with a mean radius of 103 nm (as determined by DLS, data not

shown) that are permeabilized by Tsx and that encapsulate TvNH were prepared

as described above (6.2.5).

The rate of hydrolysis of 6-thioguanosine by these nanoreactors was

determined spectrophotometrically at 355 nm in a PBS buffer pH 6.6 at 37°C. 6-

Thioguanosine concentrations ranging from 0-300 µM were used in initial rate

experiments and the data were fitted to a Michaelis-Menten equation to

determine the apparent Vmax and apparent KM (Figure 6). Vmax is 1.44 (mg/ml)-1

min-1 and KM is 38 µM. The KM value for these nanoreactors is approximately 10

times higher as compared to the KM of free TvNH for 6-thioguanosine (3.56 µM).

These results are comparable to results obtained previously for inosine,

guanosine, adenosine and 2- fluoroadenosine (Ranquin et al., 2005). This higher

apparent KM value of the nanoreactors can be explained by a difference in

exterior and interior substrate concentrations. If the rate of transport is lower

than the rate of hydrolysis, the interior substrate concentration will be lower than

0

0,005

0,01

0,015

0,02

0,025

0,03Φ

(min

)-1Adenosine

Adenine

ribose

6-thioguanosine

6-thioguanine

Chapter 6

146

the exterior concentration. The interior substrate concentration however is not

known and thus the known exterior concentration is used to determine the

apparent KM value. If we assume that encapsulation of TvNH in the polymeric

nanoreactors does not alter the kinetic behaviour of the enzyme we can conclude

that the interior substrate concentration is 10 times lower than the exterior

concentration. From SDS page analysis the encapsulated enzyme concentration

used in our enzymatic activity was estimated to be 7.5 10-9 M (data not shown).

From this we estimated the kcat for the encapsulated enzyme to be 1.05 s-1 which

is comparable to the kcat value of the free TvNH for 6-thioguanosine ( 1.77 s-1 ).

Figure 6: Enzymatic activity of Tsx permeabilzed nanoreactors that encapsulate TvNH. The reaction rates were measured spectrophotometrically at 355 nm in PBS pH 6.6 at 37 °C. The data were fitted to the Michaelis-Menten equation.

6.3.6. Cytotoxicity of 6-thioguanosine activated by TvNH encapsulating nanoreactors

To explore the therapeutic potential of Tsx permeabilized nanoreactors that

encapsulate TvNH in combination with the prodrug 6-thioguanosine, we

measured the cytotoxicity of the nanoreactors in the presence and absence of

the prodrug. A LAL assay was performed to determine the endotoxin content of

the sample. De endotoxin content was < 50 EU/ml. Since we used only 5 and 10

µl of nanoreactors sample in the assay, interference of endotoxin is not

Nanoreactor mediated prodrug activation and killing of neuroblastoma cells

147

expected. Two concentrations of nanoreactors, 0.33 mg/ml and 0.66 mg/ml were

tested in combination with 5, 10, 15 and 20 µM of the prodrug 6-thioguanosine.

The cell survival was measured after 72 h of incubation. As a positive control 0.5

% sodium azide was added and incubated for 4 h. To evaluate the effect of 6-

thioguanosine in combination with the nanoreactors, the dye uptake was stated

as 100% for cells were 0.33 mg/ml or 0.66 mg/ml nanoreactors was added

without prodrug. Comparison of these conditions to the untreated control cells

shows that adding nanoreactors to the culture medium has a minor effect on cell

growth (Figure 7). From the EC50 experiments we know that 4.8 µM of 6-

thioguanosine results in 50 % cell kill after 72 h. However we calculated the time

necessary to hydrolyse 5 µM of 6-thioguanosine with 0.33 mg/ml and 0.66

mg/ml nanoreactors to be 5.4 h and 2.7 h respectively. To completely hydrolyse

20 µM of 6-thioguanosine, 7.7 h and 3.9 h are needed respectively. This means

that the cells are only exposed to 6-thioguanosine for a short term. For both

nanoreactor concentrations used, the survival percentage is already close to the

minimal value and does not decrease much by increasing the 6-thioguanosine

concentration. Furthermore, addition of 0.5 % sodium azide to untreated cells

result in similar dye uptake percentages, suggesting that even at low

concentrations of 6-thioguanosine, all cells are killed. This is to be expected

when the nanoreactor activity is similar in culture medium as in PBS since

approximately 1 µM of 6-thioguanine is sufficient to kill 50% of the cell

population. At 5 µM 6-thioguanosine, this concentration is reached by 0.33

mg/ml and 0.66 mg/ml nanoreactors after 20 min and 10 min respectively. At 20

µM 6-thioguanosine, this is reached after 5 min and 2.5 min respectively. For the

same reason, there is no difference in cytotoxicity between 0.33 mg/ml and 0.66

mg/ml nanoreactors.

Chapter 6

148

Figure 7: Cytotoxicity of 6-thioguanosine activated by Tsx permeabilized nanoreactors that encapsulate TvNH. 5, 10, 15 and 20 µM of 6-thioguanosine (6TG) was added and the survival percentage was determined after 72 h of incubation via a neutral red based toxicity assay. Cell survival was stated as 100% when only nanoreactors were added but no 6-thioguanosine.

6.4 Discussion

Enzyme-prodrug therapy systems generally involve two steps. In the first step

the prodrug- activating enzyme is directed towards tumor tissue. In the second

step, the prodrug is administered systemically. Since the prodrug-activating

enzyme is only present in tumor tissue, the cytotoxic compound will only be

formed at the site of the tumor. A good prodrug needs to have the following

characteristics: (i) it should be relatively non toxic to avoid damage to healthy

tissue accompanied by severe side effects for the patient, which is the case for

many chemotherapeutics used to date. (ii) In addition, it is very important that

the prodrug is not activated by endogenous enzymes. This will also lead to

damage to healthy tissue and the tumor specificity of the therapy will be lost.

(iii) Finally, the prodrug needs to be efficiently activated by the prodrug

activating enzyme at the site of the tumor.

Many chemotherapeutic drugs are nucleoside analogs or nucleobase analogs

that interfere with DNA synthesis or cell division by incorporating into newly

synthesized DNA or by inhibition of enzymes involved in DNA synthesis or

nucleoside metabolism.

Nanoreactor mediated prodrug activation and killing of neuroblastoma cells

149

In a first part of this study, we want to validate TvNH as a prodrug activating

enzyme for cancer therapy. TvNH, which catalysis the hydrolysis of nucleosides

into the corresponding nucleobase and ribose, can be compared to the known

prodrug activating E. coli purine nucleoside phosphorylase (EcPNP). This enzyme

catalyses a similar enzymatic reaction by cleaving the glycosidic bond of ribo-and

deoxyribonucleosides in the presence of inorganic orthophosphate (Pi) as a

second substrate to generate the purine base and ribo(deoxyribo)-1-phosphate.

EcPNP is used in combination with the 6-aminodeoxynucleoside analog 2-fluoro-

2’-deoxyadenosine.

Although TvNH is able to cleave several nucleosides and nucleoside analogs

(chapter 2 table 1) it displays little activity towards 2’-deoxyadenosine with a

kcat value that is 1000X smaller and a KM value 4 times larger as compared to

adenosine. Therefore it can not be used to hydrolyse 2-fluoro-2’-

deoxyadenosine. TvNH however can efficiently cleave 2-fluoroadenosine and 2-

chloroadenosine with a similar kcat value and a similar KM value for 2-

chloroadenosine and with a KM value 5 times larger for 2-fluoroadenosine as

compared to adenosine. Another compound, 6-thioguanosine that is efficiently

cleaved by TvNH with similar kcat and KM values as compared to guanosine,

shows great potential as a prodrug. Although little is known about the toxicity of

6-thioguanosine, the nucleobase 6-thioguanine has been used to treat acute

lymphoblastic leukaemia, inflammatory bowel disease and to reduce organ

rejection after transplantation (Al Hadithy et al., 2005).

In an initial screening, the cytotoxicity of the prodrug candidates 2-

fluoroadenosine, 2-chloroadenosine and 6-thioguanosine on neuroblastoma cell

cultures (N2A cells) was compared to the cytotoxicity of the respective

nucleobase analogs.

Our experiment showed that 2-fluoroadenosine can not be used as a prodrug

since its cytotoxicity is higher than that of the nucleobase 2-fluoroadenine

(Figure 2A). This high cytotoxicity can be explained by the action of adenosine

kinase that metabolises 2-fluoroadenosine to form 2-fluoroadenosine

monophosphate. This can be further metabolised to 2-fluoroATP by the

subsequent action of adenylate kinase and nucleoside-diphosphate kinase or to

2-fluorodATP by the subsequent action of adenylate kinase, ribonucleotide

reductase and nucleoside-diphosphate kinase. In contrast 2-Fluoroadenine is

Chapter 6

150

activated to 2-fluoroadenosine monophosphate by hypoxanthine

phosphoribosyltransferase. A difference in expression level or activity of

adenosine kinase and hypoxanthine phosphorybosyltransferase towards 2-

fluoroadenosine and 2-fluoroadenine respectively, can explain the difference in

cytotoxicity of both compounds.

On the other hand 2-chloroadenosine shows only moderate cytotoxicity

towards neuroblastoma cells (Figure 2B), suggesting that 2-chloroadenosine is

not or slowly metabolised by one of the enzymes necessary to convert it to the

toxic triphosphate metabolites; adenosine kinase, nucleoside diphosphate kinase

or ribonucleotide reductase. Given the lack of high cytotoxicity of 2-

chloroadenine, 2-chloroadenosine is not suitable as a prodrug.

6-thioguanosine showed significantly lower toxicity than 6-thioguanine (Figure

2C). The cytotoxic effect of 6-thioguanine only becomes apparent after 24h

which is in agreement with the cytotoxic mechanism of 6-thioguanine. Cells are

capable of finishing the cell cycle in which 6-thiodGTP was incorporated because

this does not interfere with DNA elongation. Incorporation of 6-thiodGTP however

introduces severe chromatid disruption which results in cell cycle arrest

predominantly in the G2 phase (Maybaum and Mandel, 1983; Wotring and Roti

Roti, 1980). Determination of the EC50 values of 6-thioguanosine and 6-

thioguanine confirms the difference in cytotoxicity between both compounds

(Figure 3). However 6-thioguanosine also display toxicity at higher

concentrations which we believe is caused by hydrolyses of 6-thioguanosine by

PNP. Humane PNP can hydrolyse 6-oxonucleosides such as inosine and

guanosine. Although we only found cytotoxicity at higher concentrations it is

important to determine whether hPNP can efficiently hydrolyze 6-thioguanosine

since hPNP is highly expressed in lymphocytes, in particular T-cells. This is

illustrated by PNP deficiency which causes severe T-cell immunodeficiency

(Giblett et al., 1975; Markert, 1991). We therefore determined the activity of

hPNP for 6-thioguanosine and compared it to the activity for guanosine (Figure

4). Although hPNP has a lower affinity for 6-thioguanosine than for guanosine,

the turnover rate for 6-thioguanosine is higher than for guanosine. This result

suggests that 6-thioguanosine could be toxic towards lymphocytes, resulting in

immunodeficiency.

Nanoreactor mediated prodrug activation and killing of neuroblastoma cells

151

In a second part of this study we want to validate the use of polymeric

nanoreactors as an alternative for ADEPT and GDEPT strategies. We introduced

the concept of nanoreactors in 2005 to confine an enzymatic reaction to the

inside of polymeric vesicles composed of the amphiphilic triblock copolymer

copolymer poly(2-methyloxazoline)-poly(dimethylsiloxane)-poly(2-

methyloxazoline) (PMOXA-PDMS-PMOXA) (chapter 2 Figure 1). This system can

be used as an alternative to ADEPT. The advantages of using such nanoreactors

are the protection of the encapsulated enzyme from proteolysis in the blood,

avoiding immunogenicity of the enzyme by shielding it from components of the

immune system and introducing a large amount of enzymes per antibody-reactor

conjugate as compared to only one enzyme molecule in an antibody-enzyme

conjugate in ADEPT. To this end we encapsulated TvNH inside the PMOXA-

PDMSPMOXA vesicles (Figure 1). The triblock copolymeric membrane is

permeabilized by incorporation of the bacterial porin Tsx. This incorporation will

allow transport of TvNH substrates and products across the polymeric membrane

since Tsx is a nucleoside specific porin.

Activity measurements of Tsx permeabilized nanoreactors for 6-thioguanosine

showed that it is efficiently hydrolyzed with a similar estimated kcat value and an

apparent KM value 10 times higher as compared to free enzyme (Figure 6). This

10 times higher KM is in agreement with earlier results obtained with inosine,

adenosine, guanosine and 2-fluoroadenosine (Ranquin et al., 2005) and can be

explained by a difference in exterior and interior substrate concentrations. If the

rate of transport is lower than the rate of hydrolysis, the interior substrate

concentration will be lower than the exterior concentration. The apparent KM

value was determined based on the exterior concentration. If we assume that

encapsulation of TvNH in the polymeric nanoreactors does not alter the kinetic

behaviour of the enzyme we can conclude from our determined KM value that

the interior substrate concentration is 10 times lower than the exterior

concentration. Thus the rate of transport is 10 times lower than the rate of

enzymatic conversion.

Finally, we used Tsx permeabilized and TvNH encapsulated nanoreactors in

combination with 6-thioguanosine in a cytotoxicity assay to determine whether

such nanoreactors can be used as a novel enzyme-prodrug system. In our

cytotoxicity experiment we stated the cell survival of cells that were only treated

Chapter 6

152

with nanoreactors as 100 % survival. Although nanoreactors have a negative

effect on cell growth, this is only a minor effect and it is independent of the

nanoreactor concentration. As a positive control we used the cytotoxin sodium

azide at a final concentration of 0.5 %. This concentration is sufficient to

completely kill a cell population within 4 h. The cell survival after treatment with

Sodium azide was therefore stated as 0%. We demonstrated that 5 µM of 6-

thioguanosine and 0.33 mg/ml nanoreactors are sufficient to almost completely

kill the neuroblastoma cell culture within 72 h (Figure 7). Although 5µM of 6-

thioguanosine is also toxic and leads to 50 % cell kill after 72 h (Figure 3), the

exposure to this high concentration of 6-thioguanosine is very short. We

calculated the time to completely hydrolyze 6-thioguanosine to 6-thioguanine to

be 20 min for 0.33 mg/ml nanoreactors. Therefore we believe that the

cytotoxicity is completely caused by the action of the produced 6-thioguanine.

These results show that Tsx-TvNH nanoreactors indeed have great potential as

a novel enzyme-prodrug system.

Nanoreactor mediated prodrug activation and killing of neuroblastoma cells

153

6.5 References

Al Hadithy, A.F., de Boer, N.K., Derijks, L.J., Escher, J.C., Mulder, C.J. and Brouwers, J.R. (2005) Thiopurines in inflammatory bowel disease: pharmacogenetics, therapeutic drug monitoring and clinical recommendations. Dig Liver Dis, 37, 282-297.

Bagshawe, K., Sharma, S. and Begent, R. (2004) Antibody-directed enzyme prodrug therapy (ADEPT) for cancer. Expert Opin Biol Ther, 4, 1777-1789.

Benz, R., Schmid, A., Maier, C. and Bremer, E. (1988) Characterization of the nucleosidebinding site inside the Tsx channel of Escherichia coli outer membrane. Eur J Biochem, 176, 699-705.

Bilsland, A.E., Anderson, C.J., Fletcher-Monaghan, A.J., McGregor, F., Evans, T.R., Ganly, I., Knox, R.J., Plumb, J.A. and Keith, W.N. (2003) Selective ablation of human cancer cells by telomerase-specific adenoviral suicide gene therapy vectors expressing bacterial nitroreductase. Oncogene, 22, 370-380.

Bzowska, A., Kulikowska, E. and Shugar, D. (2000) Purine nucleoside phosphorylases: properties, functions, and clinical aspects. Pharmacol Ther, 88, 349-425.

Cortez-Retamozo, V., Backmann, N., Senter, P.D., Wernery, U., De Baetselier, P., Muyldermans, S. and Revets, H. (2004) Efficient cancer therapy with a nanobodybased conjugate. Cancer Res, 64, 2853-2857.

Dachs, G., Tupper, J. and Tozer, G. (2005) From bench to bedside for gene-directed enzyme prodrug therapy of cancer. Anticancer drugs, 16, 349-359.

Giblett, E.R., Ammann, A.J., Wara, D.W., Sandman, R. and Diamond, L.K. (1975) Nucleoside-phosphorylase deficiency in a child with severely defective T-cell immunity and normal B-cell immunity. Lancet, 1, 1010-1013.

Graff, A., Sauer, M., Van Gelder, P. and Meier, W. (2002) Virus-assisted loading of polymer nanocontainer. Proc Natl Acad Sci U S A, 99, 5064-5068. Epub 2002 Mar 5026.

Iwai, M., Harada, Y., Tanaka, S., Muramatsu, A., Mori, T., Kashima, K., Imanishi, J. and Mazda, O. (2002) Polyethylenimine-mediated suicide gene transfer induces a therapeutic effect for hepatocellular carcinoma in vivo by using an Epstein-Barr virus-based plasmid vector. Biochem Biophys Res Commun, 291, 48-54.

Luckey, M. and Nikaido, H. (1980) Specificity of diffusion channels produced by lambda phage receptor protein of Escherichia coli. Proc Natl Acad Sci U S A, 77, 167-171. Markert, M.L. (1991) Purine nucleoside phosphorylase deficiency. Immunodefic Rev, 3, 45-81.

Mathis, J.M., Williams, B.J., Sibley, D.A., Carroll, J.L., Li, J., Odaka, Y., Barlow, S., Nathan, C.O., Li, B.D. and DeBenedetti, A. (2006) Cancer-specific targeting of an adenovirus-delivered herpes simplex virus thymidine kinase suicide gene using translational control. J Gene Med, 8, 1105-1120.

Nardin, C., Widmer, J., Winterhalter, M. and Meier, W. (2001) Amphiphilic block copolymer nanocontainers as bioreactors. Eur. Phys. J. E, 4, 403-410.

Nardin, C., Winterhalter, M., Meier, W. (2000) Giant Free-Standing ABA Triblock Copolymer Membranes. Langmuir, 16, 7708-7712.

Nikaido, H., Nikaido, K. and Harayama, S. (1991) Identification and characterization of porins in Pseudomonas aeruginosa. J Biol Chem, 266, 770-779.

Nikaido, H. and Rosenberg, E.Y. (1983) Porin channels in Escherichia coli: studies with liposomes reconstituted from purified proteins. J Bacteriol, 153, 241-252.

Parker, W.B., Allan, P.W., Hassan, A.E., Secrist, J.A., 3rd, Sorscher, E.J. and Waud, W.R. (2003) Antitumor activity of 2-fluoro-2'-deoxyadenosine against tumors that express Escherichia coli purine nucleoside phosphorylase. Cancer Gene Ther, 10, 23-29.

Chapter 6

154

Parker, W.B., Bapat, A.R., Shen, J.X., Townsend, A.J. and Cheng, Y.C. (1988) Interactio of 2-halogenated dATP analogs (F, Cl, and Br) with human DNA polymerases, DN primase, and ribonucleotide reductase. Mol Pharmacol, 34, 485-491.

Plumb, J.A., Bilsland, A., Kakani, R., Zhao, J., Glasspool, R.M., Knox, R.J., Evans, T.R and Keith, W.N. (2001) Telomerase-specific suicide gene therapy vector expressing bacteria nitroreductase sensitize human cancer cells to the pro-dru CB1954. Oncogene, 20, 779 7803.

Ranquin, A., Versees, W., Meier, W., Steyaert, J. and Van Gelder, P. (2005) Therapeutic Nanoreactors: Combining Chemistry and Biology in a Novel Triblock Copolymer Drug Delivery System. Nano Lett, 5, 2220-2224.

Senter, P.D. and Springer, C.J. (2001) Selective activation of anticancer prodrugs by monoclonal antibody-enzyme conjugates. Adv Drug Deliv Rev, 53, 247-264.

Stribbling, S.M., Martin, J., Pedley, R.B., Boden, J.A., Sharma, S.K. and Springer, C.J. (1997) Biodistribution of an antibody-enzyme conjugate for antibody-directed enzyme prodrug therapy in nude mice bearing a human colon adenocarcinoma xenograft. Cancer Chemother Pharmacol, 40, 277-284.

Syrigos, K.N., Rowlinson-Busza, G. and Epenetos, A.A. (1998) In vitro cytotoxicity following specific activation of amygdalin by beta-glucosidase conjugated to a bladder cancer-associated monoclonal antibody. Int J Cancer, 78, 712-719.

Versees, W., Decanniere, K., Pelle, R., Depoorter, J., Brosens, E., Parkin, D.W. and Steyaert, J. (2001) Structure and function of a novel purine specific nucleoside hydrolase from Trypanosoma vivax. J Mol Biol, 307, 1363-1379.

Woodle, M.C., Engbers, C.M. and Zalipsky, S. (1994) New amphipatic polymer-lipid conjugates forming long-circulating reticuloendothelial system-evading liposomes. Bioconjug Chem, 5, 493-496.

Wotring, L.L. and Roti Roti, J.L. (1980) Thioguanine-induced S and G2 blocks and their significance to the mechanism of cytotoxicity. Cancer Res, 40, 1458-1462.

Zeng, H., Wei, Q., Huang, R., Chen, N., Dong, Q., Yang, Y. and Zhou, Q. (2007) Recombinant adenovirus mediated prostate-specific enzyme pro-drug gene therapy regulated by prostate-specific membrane antigen (PSMA) enhancer/promoter. J Androl, 28, 827-835.

Zhou, J.H., Tang, B., Liu, X.L., He, D.W. and Yang, D.T. (2007) hTERT-targeted E. coli purine nucleoside phosphorylase gene/6-methylpurine deoxyribose therapy for pancreatic cancer. Chin Med J (Engl), 120, 1348-1352.

Chapter 7 Production of cAbLys-3 mutants for selective coupling to nanoreactors

155

Chapter 7: Production of cAbLys-

3 mutants for selective coupling to

nanoreactors Abstract To specifically crosslink cAbLys-3 to the surface of PMOXA-PDMS-PMOXA polymeric nanoreactors using the heterobifunctional crosslinker N-(p-maleimidophenyl)isocyanate, cAbLys-3 S17C, S134C and S141C mutants were prepared and expressed in E.coli. These cAbLys-3 mutants are still able to bind lysozyme with similar affinity as the wild-type cAbLys-3. Since the expression yield in E. coli is very low, an alternative expression system using Pichia Pastoris was introduced which improved the yield 60 fold.

Production of cAbLys-3 mutants for selective coupling to nanoreactors

156

Chapter 7

157

7.1 Introduction

It is well know that the vascular permeability of tumors is enhanced due to the

action of secreted factors such as kinin. This allows macromolecules to diffuse

from the bloodstream into the tumor interstitium. Furthermore, the lymphatic

drainage system is impaired so that macromolecules are retained in the

interstitium for a prolonged time (Jain, 1987; Matsumura and Maeda, 1986).

Therefore particles ranging from 10-500 nm, such as liposomes, can be passively

targeted to tumor tissue due to this enhanced permeation and retention effect

(EPR) (Bornmann et al., 2008; Green and Rose, 2006; Koo et al., 2005). In

addition the attachment of ligands can further improve the accumulation of

nanoparticles in tumor tissue. Monoclonal antibodies are the most frequently

used ligands for targeting to tumor tissue.

In this project we want to use the antigen binding domain of single chain

antibodies derived from camelids for targeting purposes. These naturally

occurring single chain antibodies devoid of a light chain were discovered by

Hamers et al. in 1993 (Hamers-Casterman et al., 1993). These so called heavy-

chain antibodies (HCAb) are present in the serum of camels, dromedaries and

llamas. Due to the missing light chain in HCAbs, their antigen binding site is

composed of a single domain, VHH, compared to the VH-VL paired antigen

binding site of conventional antibodies. As a result VHHs are distinguished from

the smallest antigen binding domain, scFv, of conventional antibodies by some

unique properties: size, stability, solubility, ease of cloning a library and selecting

highly specific antigen binders, in vivo maturation and high expression levels

(Arbabi Ghahroudi et al., 1997; Muyldermans and Lauwereys, 1999). Since their

discovery VHHs, now called nanobodies, have been used for multiple purposes

such as diagnostics (Saerens et al., 2005), tumor targeting (Cortez-Retamozo et

al., 2004) and crystallisation aids (Loris et al., 2003).

We have recently introduced a new strategy for enzyme-prodrug therapy

(Ranquin et al., 2005) based on nanoreactors. The nucleoside hydrolase of

Trypanosoma vivax (TvNH), that cleaves nucleoside(analogs) in

nucleobase(analogs), was encapsulated inside the polymeric vesicle to activate

nucleoside analog prodrugs. To make sure substrates and products can readily

Production of cAbLys-3 mutants for selective coupling to nanoreactors

158

diffuse in and out of the polymeric vesicles, Tsx, a nucleoside specific porin (Benz

et al., 1988) was incorporated into the polymeric membrane. Since these

nanoreactors showed promising in vitro anti-tumor activity in combination with

the prodrug 6-thioguanosine, we want to target these nanoreactors to tumor

tissue by linking a nanobody to the surface of the reactors.

The nanoreactors are composed of the triblock copolymer poly(2-

methyloxazoline)-b-poly(dimethyl siloxane)-b-poly(2-methyloxazoline) (PMOXA-

PDMS-PMOXA). The endgroup of the hydrophilic block PMOXA that is available for

nanobody coupling is a hydroxyl group. Although there is a plethora of linkers

available for the chemical crosslinking of proteins, only one commercially

available linker was found that is reactive towards hydroxyl groups: the

heterobifunctional linker N-(p-maleimidophenyl)isocyanate (PMPI) (Pierce). The

isocyanate end of PMPI can react with hydroxyl groups to form carbamate

linkages whereas the maleimide end reacts with sulfhydryls to form a thioether

bond.

As a proof of principle we want to couple cAbLys-3 to the surface of

nanoreactors. This nanobody binds to lysozyme with nanomolar range affinity

(Desmyter et al., 1996) and is well characterised. Although nanobodies posses

four highly conserved cysteine residues (which form two disulphide bridges),

these cysteine residues are buried in the three dimensional structure and

therefore unavailable for linkage to the PMPI linker. Therefore we introduced

additional accessible cysteine residues at position 17, 134 and 141 by

substitution of serine residues.

Since expression of the cAbLys-3 serine to cysteine mutants in Escherichia coli

resulted in very low yields, we additionally used expression in Pichia pastoris to

obtain high quantities of mutant cAbLys-3. P. pastoris reaches high cell densities

and the strong methanol inducible alcohol oxidase promotor can be used for

expression. Furthermore by including the α-factor signal sequence, the protein is

secreted into the medium allowing easy purification.

Chapter 7

159

7.2 Materials and Methods

7.2.1. Site directed mutagenesis of cAbLys-3

S17C, S134C and S141C point mutations were introduced to cAbLys-3 via

site-directed mutagenesis using the QuickChangeTM Site-directed Mutagenesis

KIT from Stratagene. As template for PCR reactions we used the plasmid

pHen6c-cAbLys-3 (Figure 1) (Conrath et al., 2001). The primers used for

introducing the serine to cysteine mutations are given in Table 1. After

amplification of the plasmid using these primers and PfuTurboTM DNA

polymerase, methylated, non-mutated parental DNA was digested with DpnI.

Subsequently the mutated DNA was transformed to XL1-Blue supercompetent

cells (Invitrogen) that are capable of repairing the nicks in the mutated DNA.

This results in the formation of circular plasmid DNA that contains the mutated

cAbLys-3 gene. The faithful introduction of the mutations was confirmed by DNA

sequencing (ABI prism 3100 genetic analyzer, Applied Biosystems). Finally, the

plasmid was transformed to E. coli WK6 cells for expression.

Plac operatorPelB

CAbLys-3

His-tagpHenCAbLys-3

3638 bps

AmpR

Plac operatorPelB

CAbLys-3

His-tagpHenCAbLys-3

3638 bps

AmpR

Figure 1: Schematic representation of the pHen6c-cAbLys-3 plasmid. The plasmid contains a plac operator, the gene encoding cAbLys-3 with a pelB signal sequence and a C-terminal His6-tag, an Ampicillin Resistance gene (AmpR) and Ori (not shown on the figure).

Production of cAbLys-3 mutants for selective coupling to nanoreactors

160

Table 1: Primers used for the site-directed mutagenesis of cAbLys-3 and pPicZα colony PCR. Base substitutions for serine to cysteine mutations are depicted in bold

Reverse primer 5’-ACAGGAGAGTCTCAGACACCCTCCAGCCTGCACC-3’ S17C

Forward primer 5’-GGTGCAGGCTGGAGGGTGTCTGAGACTCTCCTGT-3’

Reverse primer 5’-GATGGTGATGGTGGTGGCAGGAGACGGTGACCTG-3’ S134C

Forward primer 5’-CAGGTCACCGTCTCCTGCCACCACCATCACCATC-3’

Reverse primer 5’-CGACGGCCAGTGAATTCTAGCAGTGATGGTGATGGTGGTG-3’ S141C

Forward primer 5’-CACCACCATCACCATCACTGCTAGAATTCACTGGCCGTCG-3’

Reverse primer 5’-CCTACAGTCTTACGGTAAACG-3’ pPicZα

Forward primer 5’-GACTGGTTCCAATTGACAAGC-3’

7.2.2. Bacterial expression and purification of cAbLys-3 mutants

Large scale production of cAbLys-3 mutants, S17C, S134C and S141C was

performed in shaker flasks by growing the E. coli WK6 cells in Terrific Broth (TB)

(Table 2) supplemented with 0.1% (w/v) glucose and 100 µg/ml ampicillin until

an OD600 of 0.6–0.9 was obtained. Expression was then induced with 1 mM IPTG

for 16 hours at 37°C. After pelleting, the cells were resuspended in 0.2 M Tris-

HCL pH 8.0, 0.5 mM EDTA, 0.5 M sucrose (TES). The periplasmic proteins were

extracted by an osmotic shock by diluting the resuspended cells 3x in 50 mM

Tris-HCL pH 8.0, 0.12 mM EDTA, 0.12 M sucrose. This periplasmic extract was

subsequently loaded onto pre-equilibrated (50 mM Hepes pH 7.0, 1M NaCl) Ni-

NTA Sepharose (GE healthcare), and after washing with 50 mM Hepes pH 7.0,

1M NaCl, 10mM imidazol, the bound proteins were eluted with 50 mM Hepes pH

7.0, 1M NaCl, 1M imidazol. Finally the protein was concentrated on Vivaspin

(Vivascience) concentrators with a cut-off of 5 kDa and loaded onto a Superdex-

75 16/60 (Pharmacia) gelfiltration chromatography column, equilibrated with

phosphate buffered saline (PBS). The purity of the cAbLys-3 mutants was

evaluated by Coomassie brilliant blue stained SDS-PAGE.

Chapter 7

161

7.2.3. Expression of cAbLys-3 S17C in Pichia Pastoris

7.2.3.1. Cloning

pHen6c-cAbLys-3 S17C was digested with PstI and EcoRI. The restriction

fragment containing cAbLys-3 S17C was ligated into a PstI, EcoRI digested

pPiCZα vector (Figure 2) (Invitrogen). The ligation mixture was subsequently

transformed to electro-competent Top10 cells and subsequently grown at 37°C

for 24 h on Luria Broth (LB) plates (Table 2). To identify clones that posses

pPICZα-cAbLys-3 S17C, a colony PCR was performed with pPICZα forward and

reverse primers (Table 1). The sequence of the positive clones was confirmed by

DNA sequencing (ABI prism 3100 genetic analyzer, Applied Biosystems). Two

positive pPICZα-CAbLys-3 S17C plasmids were isolated via a Qiagen midiprep

and linearised with BstXI for transformation to P. Pastoris X-33 cells.

7.2.3.2. Transformation

Electro-competent Pichia Pastoris X-33 cells (Invitrogen) were prepared as

follows. The cells were grown on a YPD plate (Table 2) at 30 °C for 48 h to obtain

single colonies. A single colony was picked and grown in 5ml YPD for 24 h at

30°C. This overnight cell culture was used to inoculate 250 ml YPD medium.

After 18 h at 30 °C, the cells (OD600 = 1.3) were collected and washed twice with

sterile cold water. The cells were made electro-competent by adding 800 µl 1M

sorbitol.

Linearised pPICZα-cAbLys-3 S17C plasmid (10 µg) was added to 80 µl electro-

competent cells and electroporated with an E.coli pulser (BioRad) at 1.5kV. 1 ml

1 M sorbitol was added and the cells were incubated for 1 h at 30 °C.

Subsequently 200 µl 5x YPD medium was added and after 1 h incubation at

30°C, the cells were plated on YPDS plates containing 100 µg/ml, 250 µg/ml and

500 µg/ml Zeocine. Cells were grown on these plates for 2 days at 30 °C.

Production of cAbLys-3 mutants for selective coupling to nanoreactors

162

Figure 2: Schematic representation of the PicZα-cAbLys-3 S17C plasmid. This plasmid contains, a 5’AOX1 promotor region, the gene coding for cAbLys-3S17C with an α-factor signal sequence and a C-terminal His6-tag, followed by an AOX1 transcription termination region, a Zeocine Resistance gene followed by a CYC1 transcription termination region and an Ori.

7.2.3.3. Expression and purification

Single colonies from the 250 µg/ml and 500 µg/ml Zeocine plates were used

to inoculate 10 ml YPG medium (Table 2). After 48 h at 30 °C this culture was

used to inoculate three flasks of 250 ml BGY (Table 2) medium at an OD600 of

0.04. After 24 h at 30 °C, the cells were pelleted and resuspended in three flasks

of 250 ml BMY (Table 2) medium to induce expression of cAbLys-3 S17C. During

the following 36 h the cultures were re-induced 4 times with 0,5 % methanol.

After a total of 48 h induction at 30 °C, the cells were removed and NaCl and

imidazol were added to the supernatant at a final concentration of 1 M and 20

mM, respectively, and purificated as described under 7.2.2.

Chapter 7

163

Table 2: Composition of culture media used.

TB LB YPD∗ YPG YPDS∗ BGY BMY Trypton 12 g/l 10 g/l 20 g/l 20 g/l 20 g/l 20 g/l 20 g/l Yeast extract 24 g/l 5 g/l 10 g/l 10 g/l 10 g/l 10 g/l 10 g/l NaCl 5 g/l 10 g/l Glucose 20 g/l 20 g/l Glycerol 30 g/l 30 g/l Sorbitol 182.2 g/l K-phosphate pH 6.0 200 mM 200 mM

Methanol 0.5 % Biotin 0.4 mg/l 0.4 mg/l ∗ To make culture plates, 20 g/l agar was added

7.2.4. Enzyme-Linked Immuno Sorbent Assay (ELISA)

F96 Maxisorb Nunc-immunoplates were coated overnight at 4 °C with 1 µg/ml

lysozyme in PBS. To avoid nonspecific adsorption of antibodies to the

immunoplates, they were blocked for two hours at room temperature with 1%

(w/v) caseïne in PBS. Three concentrations, 0.1 µM, 1 µM and 10 µM of cAbLys-3

S17C, S134C and S141C were added to the lysozyme coated plates and

incubated for 2 hours at room temperature. To detect bound cAbLys-3, firstly,

mouse α-His6 antibody was added, followed by a goat α-mouse antibody-

peroxidase conjugate. Each antibody was incubated at room temperature for two

hours. After incubation with each antibody, the plate was washed with PBS, 0.01

% (v/v) Tween to remove unbound antibody. Signal development was achieved

by adding 0.004 % peroxidase substrate tetramethyl benzidine (TMB) and 3 %

(v/v) H2O2 in 0.1 M phosphate buffer pH 6.0, 2% (w/v) citric acid. After 30 min

the enzymatic reaction was stopped by adding 1N H2SO4. Finally colour

development was measured at 45O nm on a SpectraMax multiwell plate reader

(Molecular Devices). To determine the nonspecific binding of α-mouse antibody-

peroxidase conjugate, this antibody was added to lysozyme coated wells without

cAbLys-3 mutant antibody and to wells without mouse α-His6 antibody. As a

positive control, 10 µM cAbLys-3 was used.

Production of cAbLys-3 mutants for selective coupling to nanoreactors

164

7.3 Results

7.3.1. Production of cAbLys-3 mutants in E. coli

To target nanoreactors to tumor sites we like to couple camelid antibodies, so

called nanobodies®, to the polymers of the nanoreactor. Since the reactive group

of PMOXA is a hydroxyl group we will use the hetero-bifunctional linker PMPI

where the isocyanate end can react with hydroxyl groups and the maleimide end

is able to reacts with sulfhydryls.

Although nanobodies posses four highly conserved cysteine residues (which form

two disulphide bridges), these cysteine residues are buried in the three

dimensional structure and therefore unavailable for linkage to the PMPI linker

(Figure 3). Therefore cAbLys-3 mutants were made to introduce additional

cysteine residues for coupling to the nanoreactors. This method has an important

advantage compared to the more conventional coupling methods that use amine-

or carboxyl- reactivity. By introducing additional cysteines, the coupling is highly

specific and can be controlled whereas amine- or carboxyl- coupling is highly

unspecific due to the abundance of these groups in proteins.

Since serine to cysteine substitutions introduce only a minor structural

change, we identified two serine residues, S17 and S134 that are located far

from the cAbLys-3 antigen binding site (Figure 3). Additionally we introduced an

extra cysteine residue at position 141, at the end of the C-terminal His6-tag. This

residue is not visible in the X-ray structure.

Mutants were originally cloned and expressed in E. coli WK6 cells. Purification

of these three mutants resulted in a single band on Coomassie stained SDS-

PAGE (Figure 4).

Chapter 7

165

Figure 3: Three dimensional structure of cAbLys-3 (blue) and Lysozyme (green). Cysteine residues in cAbLys-3 are depicted in pink, ser17 and ser134 are depicted in red.

A B C D

Figure 4: SDS-PAGE analysis of purified cAbLys-3 S17C (A), cAbLys-3 S134C (B) and cAbLys-3 S141C (C), expressed in E. coli and cAbLys-3 S17C expressed in P. pastoris

We obtained 0.23 mg/l culture, 0.20 mg/l and 0.15 mg/l of cAbLys-3 S134C,

S141 and S17C respectively. This low yield is probably the result of different

codon usage in E. coli as compared to camelidae. Especially eukaryotic glycine,

arginine, lysine and leucine codons are rarely used by E.coli (Figure 5).

Production of cAbLys-3 mutants for selective coupling to nanoreactors

166

Figure 5: Codon usage of the cAbLys-3 sequence in E. coli. The Y-axis represents the relative adaptiveness of E. coli to use the codon. A relative adaptiveness smaller than 50 % is depicted in grey. The analysis was performed using Graphical Codon Usage Analyser software (www.gcua.de).

Chapter 7

167

7.3.2. Functionality of cAbLys-3 mutants

To determine whether these mutants are unaffected in antigen binding, we

performed an Enzyme-Linked Immuno Sorbent Assay (ELISA) (Figure 6). The

results clearly show that cAbLys-3 S17C and cAbLys-3 S134C are able to bind

lysozyme with similar affinities as wild type cAbLys-3. Moreover, the binding

shows a concentration dependent signal. CAbLys-3 S141C on the other hand

shows only a low binding signal as compared to the wild type nanobody.

Lysozyme + + + + + + + α-His6 + + + + + α-mouse∼peroxidase + + + + + +

cAbLys-3 S17C + cAbLys-3 S134C + cAbLys-3 S141C + cAbLys-3 +

Figure 6: ELISA to determine antigen binding capacity of cAbLys-3 mutants.

7.3.3. Production of cAbLys-3 S17C in P. Pastoris

Although we were able to produce pure cAbLys-3 S17C, S134C and S141C

that are still able to bind lysozyme in E. coli, the yield was very low. Therefore,

an alternative expression method in P. Pastoris was used to obtain larger

quantities of purified nanobody. Since cAbLys-3 S17C has the best lysozyme

binding properties this mutant was preferred for high yield expression in P.

00,1

0,20,30,4

0,50,6

0,70,8

OD450

100 nM1 µM10 µMnegative controls

Production of cAbLys-3 mutants for selective coupling to nanoreactors

168

Pastoris. By using a PicZα plasmid, the expression of cAbLys-3 S17C can be

induced by methanol since it contains an AOX1 promotor region. Due to the

presence of an α-factor signal sequence, the expressed protein is secreted into

the culture medium. Expression in P. Pastoris resulted in pure cAbLys-3 S17C

with a yield of 8.7 mg/l culture. This is approximately 60 times higher than the

yield obtained with E. coli (Figure 4).

7.4 Discussion

Targeting of enzymes and/or lipidic or polymeric vehicles to tumor sites is

mostly accomplished by chemical coupling of antibodies. In this study we want to

couple camelid antibodies to PMOXA-PDMS-PMOXA triblock copolymers. As a

model antibody we used cAbLys-3, an antibody directed against lysozyme. In

future experiments, the nanoreactor-antibody complex can than conveniently be

used to target Lewis Lung carcinoma cells (3LL cells) which express lysozyme on

their plasma membrane (personal communication)

The presence of an hydroxyl group on the polymers forced us to use the

hetero bifunctional linker PMPI which can be coupled to the polymer via its

isocyanate end and to the antibody with its maleimide group to form a thioether

bond with sulfhydryl residues.

Therefore, the solved X-ray structure of cAbLys-3 was scrutinized for possible

site-directed cysteine mutants that wouldn’t interfere with antigen binding. We

were able to construct three cAbLys-3 serine to cysteine mutants, S17C, S134C

and S141C that still can bind lysozyme (Figure 6). CAbLys-3 S17C and cAbLys-3

S134C bind lysozyme with similar affinity as the wild type cAbLys-3. CAbLys-3

S141C, however, has a lower binding capacity as compared to the wild type. This

is unexpected since residue 141 is located equally far from the antigen binding

site as residue 134. Since residue 141 is not part of the visible folded structure

but situated just behind the C-terminal His6-tag, we believe the lower signal in

the ELISA assay is caused by interference of binding of the mouse α-His antibody

to the His6-tag of cAbLys-3 S141C.

Chapter 7

169

Expression of the cAbLys-3 mutants in E. coli yields only a low amount of

protein (0.15 – 0.23 mg/l culture medium) as compared to WT cAbLys-3 (4

mg/l) which can be explained by misfolding of the nanobody. Introduction of an

additional cysteine can result in malformation of disulphide bridges and

misfolding of the protein. Formation of correct disulphide bridges is highly

controlled in the periplasmic space and involves an intricate cascade of redox

proteins. The oxidation of sulfhydryl residues is mainly catalysed by the

periplasmic chaperone DsbA which is recycled through an oxidation reaction with

the membrane protein DsbD (Kadokura and Beckwith, 2002). Failure of this

system by over expression of the target protein or by non-prokaryotic

heterologous targets might result in the misfolding and subsequent aggregated

of the latter. That part of the antibodies are indeed misfolded and aggregation is

confirmed by the presence of cAbLys-3 mutant protein in the insoluble fraction

after the periplasmic extraction (data not shown). However the yield could not be

improved by performing the periplasmic extraction in the presence of

Dithiothreitol (DTT) (data not shown). The reason for the low yield of cAbLys-3 in

general can be explained by the difference in codon usage between E. coli and

higher eukaryotes such as camelidae (Figure 5). Especially eukaryotic glycine,

arginine, lysine and leucine codons are rarely used by E.coli. Since a complete

change of these codons to the ones preferred by E. coli would be time consuming

and very costely, we used an eukaryotic expression host, P. pastoris. Although

codon usage of the cAbLys-3 sequences are far from ideal for expression in P.

pastoris, larger expression yields can be obtained through high cell densities and

strong promoters. The expression yield for cAbLys-3 S17C was improved 60 fold

by using P. pastoris.

Although preliminary experiments for the coupling of TvNH-Tsx nanoreactors

to cAbLys-3 have been performed, they were unsuccessful (data not shown).

Much effort still has to be put into fine tuning the chemical reaction and finding

an efficient assay for the detection of nanobody-coupled nanoreactors.

Production of cAbLys-3 mutants for selective coupling to nanoreactors

170

7.5 References

Arbabi Ghahroudi, M., Desmyter, A., Wyns, L., Hamers, R. and Muyldermans, S. (1997) Selection and identification of single domain antibody fragments from camel heavy-chain antibodies. FEBS Lett, 414, 521-526.

Benz, R., Schmid, A., Maier, C. and Bremer, E. (1988) Characterization of the nucleoside-binding site inside the Tsx channel of Escherichia coli outer membrane. Eur J Biochem, 176, 699-705.

Bornmann, C., Graeser, R., Esser, N., Ziroli, V., Jantscheff, P., Keck, T., Unger, C., Hopt, U.T., Adam, U., Schaechtele, C., Massing, U. and von Dobschuetz, E. (2008) A new liposomal formulation of Gemcitabine is active in an orthotopic mouse model of pancreatic cancer accessible to bioluminescence imaging. Cancer Chemother Pharmacol, 61, 395-405.

Conrath, K.E., Lauwereys, M., Galleni, M., Matagne, A., Frere, J.M., Kinne, J., Wyns, L. and Muyldermans, S. (2001) Beta-lactamase inhibitors derived from single-domain antibody fragments elicited in the camelidae. Antimicrob Agents Chemother, 45, 2807-2812.

Cortez-Retamozo, V., Backmann, N., Senter, P.D., Wernery, U., De Baetselier, P., Muyldermans, S. and Revets, H. (2004) Efficient cancer therapy with a nanobody-based conjugate. Cancer Res, 64, 2853-2857.

Desmyter, A., Transue, T.R., Ghahroudi, M.A., Thi, M.H., Poortmans, F., Hamers, R., Muyldermans, S. and Wyns, L. (1996) Crystal structure of a camel single-domain VH antibody fragment in complex with lysozyme. Nat Struct Biol, 3, 803-811.

Green, A.E. and Rose, P.G. (2006) Pegylated liposomal doxorubicin in ovarian cancer. Int J Nanomedicine, 1, 229-239.

Hamers-Casterman, C., Atarhouch, T., Muyldermans, S., Robinson, G., Hamers, C., Songa, E.B., Bendahman, N. and Hamers, R. (1993) Naturally occurring antibodies devoid of light chains. Nature, 363, 446-448.

Jain, R.K. (1987) Transport of molecules in the tumor interstitium: a review. Cancer Res, 47, 3039-3051.

Kadokura, H. and Beckwith, J. (2002) Four cysteines of the membrane protein DsbB act in concert to oxidize its substrate DsbA. Embo J, 21, 2354-2363.

Koo, O.M., Rubinstein, I. and Onyuksel, H. (2005) Role of nanotechnology in targeted drug delivery and imaging: a concise review. Nanomedicine, 1, 193-212.

Loris, R., Marianovsky, I., Lah, J., Laeremans, T., Engelberg-Kulka, H., Glaser, G., Muyldermans, S. and Wyns, L. (2003) Crystal structure of the intrinsically flexible addiction antidote MazE. J Biol Chem, 278, 28252-28257.

Matsumura, Y. and Maeda, H. (1986) A new concept for macromolecular therapeutics in cancer chemotherapy: mechanism of tumoritropic accumulation of proteins and the antitumor agent smancs. Cancer Res, 46, 6387-6392.

Muyldermans, S. and Lauwereys, M. (1999) Unique single-domain antigen binding fragments derived from naturally occurring camel heavy-chain antibodies. J Mol Recognit, 12, 131-140.

Ranquin, A., Versees, W., Meier, W., Steyaert, J. and Van Gelder, P. (2005) Therapeutic Nanoreactors: Combining Chemistry and Biology in a Novel Triblock Copolymer Drug Delivery System. Nano Lett, 5, 2220-2224.

Saerens, D., Frederix, F., Reekmans, G., Conrath, K., Jans, K., Brys, L., Huang, L., Bosmans, E., Maes, G., Borghs, G. and Muyldermans, S. (2005) Engineering camel single-domain antibodies and immobilization chemistry for human prostate-specific antigen sensing. Anal Chem, 77, 7547-7555.

Chapter 7

171

7.6 Appendix

Another way of targeting nanoreactors to specific cell types can be

accomplished by adhesins that specifically recognize sugars present on the target

cells. These sugars can be part of glycosylated proteins or of lipids such as

ceramides.

Ed Conway, from the Centrum voor Transgene Technologie en Gentherapie

(KULeuven), is studying the acute renal failure syndrome (Kindt et al., 2008). He

and his co-workers are able to relief the symptoms in a mouse model by

injection of naked plasmid DNA that encodes for the anti-apoptotic protein

survivin. However, non-targeted DNA injection results in systemic expression of

survivin, a protein that is also upregulated in tumor tissue. Therefore there is

great interest in using polymeric vesicles for directing the plasmid construct to

kidney cells.

For targeting of plasmid encapsulated vesicles, a bacterial adhesion, PapG II,

was cloned, expressed and purified. PapG II is the adhesion at the tip of the P

pilus that mediates attachment of uropathogenic Escherichia coli to the

uroepithelium of the human kidney via binding to globotriaosylceramides

(Dodson et al., 2001).

For high expression, the gene was cloned with the gateway® technology

(invitrogen). This technology is a universal cloning method that takes advantage

of the site-specific recombination of bacteriophage lambda to provide a rapid and

highly efficient way to move DNA sequences into multiple vector systems. We

isolated the PapG II encoding gene from the genomic DNA of E. coli APEC 54. For

coupling purposes a C-terminal Lys3-tag was introduced via PCR. Firstly, a gene

construct with flanking attB recombination sequences was made via PCR with

attB forward and attB reverse primers (Table 1). This construct was then

inserted into a gateway® donor vector (pDONR) via recombination, catalyzed by

BP clonase™ II. The plasmid was subsequently transformed to Ca-competent E.

coli DH5α cells and cultured on kanamycine plates. Since the pDONR vector has

a kanamycine resistance gene, cells that were not transformed with pDONR, can

not grow. Additionally, the pDONR vector contains the gene sequence of the

toxin ccdB, a toxin, between the att recombination sequences. Therefore, the

Production of cAbLys-3 mutants for selective coupling to nanoreactors

172

cells transformed with plasmids that do not contain the PapG II gene are killed

and only cells with the PapG II vector can grow. Via an identical recombination

method, the gene coding for PapGII was cloned into a gateway® destination

vector (pDEST) that contains an ampicillin resistance gene. Subsequently the

pDEST was transformed to E. coli BL21 (DE3) C43 cells and cultured on ampicillin

plates. After each transformation step a colony PCR was performed with secL A1

and secB primers (Table 1) to select positive clones and the sequence was

confirmed by DNA sequencing (ABI prism 3100 genetic analyzer, Applied

Biosystems).

Table 1: primers used for the cloning of PapG II via the gateway® technology. Recombination sequences are depicted in bold, the sequence encoding the Lys3-tag is depicted in gray

Primer Sequence attB forward primer

5’- GGGGACAAGTTTGTACAAAAAAGCAGGCTTAAGAAGGAGATATACCATG

AAAAAATGGTTCCCAGCTTTGTTATTTTC-3’ attB reverse primer

5’-GGGGACCACTTTGTACAAGAAAGCTGGGTATTACTTCTTCTTGCCGATATT

CTTAAATAAGAATAACAT -3’

secL A1 5’-CTCTCGCGT TAACGCTAGC ATGGAT-3’ secL B 5’-GTAACATCAGAGATTTTGAGACAC-3’

Large scale production of PapG II was performed in shaker flasks by growing the

bacteria in Luria Broth (LB) and 100 µg/ml ampicillin, at 28 °C, until an OD600 of

0.6–0.9 was obtained. Expression was then induced with 0,2 mM IPTG for 16

hours at 28°C. After pelleting the cells, they were resuspended in 50 mM Hepes

pH 7.0, 2.5 mM EDTA, 30 % (w/v) sucrose. The periplasmic proteins were

extracted by an osmotic shock by diluting the resuspended cells 3x in 50 mM

Hepes pH 7.0. PapG II was purified from this periplasmic extract via cation

exchange chromatography on a 30S source column (Pharmacia). The column

was equilibrated, loaded and washed with 50 mM Hepes pH 7.0. PapG II was

eluted with 50 mM Hepes pH 7.0, 1 M NaCl by applying a gradient. Finally the

protein was concentrated on Vivaspin (Vivascience) concentrators with a cut-off

of 5 kDa and loaded onto a Superdex-75 16/60 (Pharmacia) gelfiltration

chromatography column, which was equilibrated with phosphate buffered saline

(PBS). The purity of PapG II was evaluated by Coomassie brilliant blue stained

SDS-PAGE.

Chapter 7

173

Figure 1: SDS-PAGE analysis of E. coli PapG II

References Dodson, K.W., Pinkner, J.S., Rose, T., Magnusson, G., Hultgren, S.J. and Waksman, G.

(2001) Structural basis of the interaction of the pyelonephritic E. coli adhesin to its human kidney receptor. Cell, 105, 733-743.

Kindt, N., Menzebach, A., Van de Wouwer, M., Betz, I., De Vriese, A. and Conway, E.M. (2008) Protective role of the inhibitor of apoptosis protein, survivin, in toxin-induced acute renal failure. Faseb J, 22, 510-521.

Chapter 8 General discussion

175

Chapter 8:

General discussion

Chapter 8

176

8.1 General discussion

We succeeded in showing the potential of polymeric nanoreactors as a novel

prodrug activating system, and thus delivered a first proof of principal (chapter

6). This system can have a great advantage compared to conventional

antibodydirected enzyme-prodrug therapy (ADEPT). One of the biggest concerns

in ADEPT is the immunogenicity of the antibody-enzyme conjugate. This can be

diminished by altering T cell epitopes (Harding et al., 2005), but this is a very

labour intensive process. By encapsulating the enzyme in nanoreactors, it is

shielded from the environment thus protecting it from proteolytic degradation

and detection by the immune system. Moreover when using nanoreactors a high

concentration of enzyme can be targeted to the tumor per therapeutic unit

whereas only one enzyme molecule is targeted to the tumor per therapeutic unit

in ADEPT. We showed that when the enzyme is encapsulated in polymeric

vesicles that are permeabilized by porins, the enzymatic activity still remains

high (Ranquin et al., 2005).

Although we were not yet successful in attaching nanobodies to the surface of

the nanoreactors for targeting purposes, passive targeting of polymeric vesicles

is also a possibility due to the enhanced permeation and retention effect (EPR) of

leaky tumor vasculature (Matsumura and Maeda, 1986) (Jain, 1987). We did

construct the necessary cAbLys-3 cysteine mutants for attachment to the

hydroxyl endgroup of PMOXA via the heterobifunctional PMPI linker (chapter 7).

The biggest issue in the coupling of cAbLys-3 cysteine mutants to nanoreactors is

finding an accurate way for the detection of targeted nanoreactors and

separation of targeted and non-targeted reactors. There are some alternative

coupling strategies that could also be used. For instance, the streptavidin-biotin

coupling method as described by Broz et al. (2005) is a possibility. For this

purpose biotinylated PMOXA21-PDMS54-PMOXA21 polymers were already made.

However, the production method of biotinylated nanoreactors is not yet

optimized. Since biotin is a bulky group compared to the nanobodies, random

attachment might interfere with antigen binding and thus it will be important to

specifically biotinylate the nanobodies. Another option for targeting of

nanoreactors by nanobodies is to attach an α-helical membrane anchor to the C

terminus of the nanobody. This way the nanobody can be inserted into the

General discussion

177

polymeric membrane in a similar way as the porin. However approximately 50 %

of the inserted nanobodies will be pointed to the interior of the vesicles instead of

the exterior. Moreover, since the polymeric membrane (10 nm) is much thicker

than a biological membrane (4 nm) a very large linker region between anchor

and nanobody will be needed.

To evaluate the true therapeutic potential of the nanoreactors, a lot of

questions need to be answered concerning the in vivo behaviour of the

nanoreactors.

Firstly, it is important to determine whether these nanoreactors are able to

elicit an immune response. Uptake by macrophages or other antigen presenting

cells is an important first step in eliciting an immune response. Therefore we

already performed a preliminary macrophage uptake experiment with fluorescent

nanoreactors (chapter 5). Although this experiment is far from conclusive, the

lack of macrophage uptake seen in this experiment is supported by earlier work

on the same type of vesicles by Broz et al (Broz et al., 2005). Here, a poly A

ligand had to be attached to the surface of PMOXA-PDMS-PMOXA vesicles to

target the A1 scavenger receptor of macrophages and to promote macrophage

uptake. A similar experiment with fluorescent macrophages was done by De

Vocht et al. (manuscript in preparation). They used FACS analysis to determine

whether fluorescent nanoreactors are taken up by macrophages. They found that

only a small percentage (7%) of nanoreactors is associated with macrophages.

Additionally, they analysed the secretion of several cytokines to determine if

macrophages are activated in the presence of nanoreactors. Cytokine secretion

was only observed for samples that were contaminated with endotoxin, as

determined via a LAL assay (personal communication). This result also indicates

that the reactors are not taken up by macrophages and do not activate

macrophages. A second experiment with endotoxin free samples in vitro and in

vivo is on the way, to confirm previous results and determine the uptake and

activation in vivo.

Secondly, it is important to determine the biodistribution and toxicity of the

reactors that are injected intravenously or intraperitonealy since accumulation in

organs such as the hart, liver and kidney can lead to severe toxicity. Although we

6). The in vivo toxicity towards various cell types however can not be predicted.

Chapter 8

178

Since polymeric vesicles are comparable with liposomes, accumulation of

nanoreactors in the liver is expected. This accumulation might lead to liver

toxicity. For this reason, De Vocht et al. evaluated the toxicity of nanoreactors

towards hepatocytes in vitro (manuscript in preparation). They observed no

significant toxicity for nanoreactor concentrations ranging from 50-1000 µg/ml

as compared to the positive control. Although these results are promising and

polymers made from PMOXA-PDMS-PMOXA blocks are approved by the FDA for

use in contact-lens material, the in vivo toxicity towards various cell types

remains unknown as is the case for many block copolymer particles (micelles,

nanotubes …). The available published data are scattered bits and pieces mostly

concerned with in vitro tests on cell cultures. For instance, silicium containing

compounds such as siloranes were shown to have a very low genotoxic potential

on cultured mouse fibroblast cells (Kostoryz et al., 2007). Possibly similar results

will be obtained with PDMS blocks but a definite proof is still missing. It is clear

that an extensive study on the biodistribution and toxicity is needed to provide

the necessary answers.

Thirdly, it is important that nanocontainers can be removed from the body by

degradation and secretion since the PDMS middle block is not degradable.

Although we validated the nanoreactor prodrug activation system in

combination with nucleoside hydrolase of Trypanosoma vivax and 6-

thioguanosine, it can easily be used with other known prodrug activating

enzymes such as carboxipeptidase G2 (Stribbling et al., 1997), β-lactamase

(Meyer et al., 1993) (Cortez-Retamozo et al., 2004), alkaline phosphatase

(Senter et al., 1989), E. coli purine nucleoside phosphorylase (Parker et al.,

2003), etc.. Furthermore nanoreactors can also be used for purposes other than

prodrug activation. For instance, they have potential to be used in enzyme

replacement therapy. Nowadays, the enzymes used in enzyme replacement

therapy are pegylated to prolong the blood circulation (Brewerton et al., 2003).

By incorporating the enzyme in PMOXA-PDMS-PMOXA polymers blood circulation

times might be further increased resulting in less frequent injections of the

enzyme.

Nanoreactors can also be useful in diagnostics where screening of a certain

blood compound is performed with an antibody-enzyme conjugate. However, the

saw no reactor dependent toxicity towards neuroblastoma cells in vitro (chapter

General discussion

179

antibody to enzyme ratio (1:1) in these conventional methods is rather poor.

Using antibody linked nanoreactors can improve the detection signal since a high

concentration of enzymes is linked to the detecting antibody. As a result, the

detection limit of the conventional technique can be improved allowing for

disease detection in an early stage.

Finally, nanoreactors can potentionally be used for in vivo biosensing of blood

glucose levels. For the monitoring of glucose levels in diabetes patients, a

glucose sensor is used. Practically all of the reported sensors that continuously

measure glucose levels are electrochemically based and take advantage of the

reaction of glucose with oxygen. This reaction is catalyzed by glucose oxidase

and forms gluconic acid and peroxide (Wilson et al., 1992). The sensors are

multilayered devices where the enzyme is trapped between two semipermeable

membranes. The inner membrane is connected to an electrode to detect

produced radicals. Instead of using such multilayered devices, a simpler devices

based on glucose oxidase encapsulated nanoreactors, can be envisioned.

Chapter 8

180

8.2 References

Brewerton, L., Fung, E. and Snyder, F. (2003) Polyethyleenglycol-conjugated adenosine phosphorylase: development of alternative enzyme therapy for adenosine deaminase deficiency. Biochim biophys acta, 1637, 171-177.

Broz, P., Benito, S.M., Saw, C., Burger, P., Heider, H., Pfisterer, M., Marsch, S., Meier, W. and Hunziker, P. (2005) Cell targeting by a generic receptor-targeted polymer nanocontainer platform. J Control Release, 102, 475-488.

Cortez-Retamozo, V., Backmann, N., Senter, P.D., Wernery, U., De Baetselier, P., Muyldermans, S. and Revets, H. (2004) Efficient cancer therapy with a nanobodybased conjugate. Cancer Res, 64, 2853-2857.

Harding, F., Liu, A., Stickler, M., Razo, Chin, R., Faravashi, N., Viola, W., Graycar, T., Yeung, V., Aehle, W., Meijer, D., Wong, S., Rashid, M., Valdes, A. and Schellenberger, V. (2005) A beta-lactamase with reduced immunogenicity for the targeted delivery of chemotherapeutics using antibody-directed enzyme prodrug therapy. Mol Cancer Ther, 4, 1791-1800.

Jain, R.K. (1987) Transport of molecules in the tumor interstitium: a review. Cancer Res, 47, 3039-3051.

Kostoryz, E., Zhu, Q., Zhao, H., Glaros, A. and Eick, J. (2007) Assessment of cytotoxicity and DNA damage exhibited by siloranes and oxiranes in cultured mammalian cells. Mutat Res, 634, 156-162.

Matsumura, Y. and Maeda, H. (1986) A new concept for macromolecular therapeutics in cancer chemotherapy: mechanism of tumoritropic accumulation of proteins and the antitumor agent smancs. Cancer Res, 46, 6387-6392.

Meyer, D.L., Jungheim, L.N., Law, K.L., Mikolajczyk, S.D., Shepherd, T.A., Mackensen, D.G., Briggs, S.L. and Starling, J.J. (1993) Site-specific prodrug activation by antibody-beta-lactamase conjugates: regression and long-term growth inhibition of human colon carcinoma xenograft models. Cancer Res, 53, 3956-3963.

Parker, W.B., Allan, P.W., Hassan, A.E., Secrist, J.A., 3rd, Sorscher, E.J. and Waud, W.R. (2003) Antitumor activity of 2-fluoro-2'-deoxyadenosine against tumors that express Escherichia coli purine nucleoside phosphorylase. Cancer Gene Ther, 10, 23-29.

Ranquin, A., Versees, W., Meier, W., Steyaert, J. and Van Gelder, P. (2005) Therapeutic Nanoreactors: Combining Chemistry and Biology in a Novel Triblock Copolymer Drug Delivery System. Nano Lett, 5, 2220-2224.

Senter, P.D., Schreiber, G.J., Hirschberg, D.L., Ashe, S.A., Hellstrom, K.E. and Hellstrom, I. (1989) Enhancement of the in vitro and in vivo antitumor activities of phosphorylated mitomycin C and etoposide derivatives by monoclonal antibodyalkaline phosphatase conjugates. Cancer Res, 49, 5789-5792.

Stribbling, S.M., Martin, J., Pedley, R.B., Boden, J.A., Sharma, S.K. and Springer, C.J. (1997) Biodistribution of an antibody-enzyme conjugate for antibody-directed enzyme prodrug therapy in nude mice bearing a human colon adenocarcinoma xenograft. Cancer Chemother Pharmacol, 40, 277-284.

General discussion

181

Wilson, G., Zhang, Y., Reach, G., Moatti-Sirat, D., Poltout, V., Thévenot, D., Lemonrner, R. and Klein, J. (1992) Progress toward the Development of an Implantable Sensor for Glucose. Clin Chem, 38, 1613-1617.

summary

183

Summary

Directed enzyme-prodrug therapies aim to improve conventional

chemotherapy by activating a non toxic prodrug into a toxic drug only at the site

of the tumor. This considerably lowers the systemic toxicity associated with

conventional chemotherapy. Directed enzyme-prodrug therapy involves two

stages. In the first step the activating enzyme is directed to the tumor. In the

second step the non-toxic prodrug is systemically administered. Subsequently,

the prodrug is converted to a toxic drug by the prodrug activating enzyme

resulting in high local concentrations of an anticancer drug at the tumor site. The

targeting of the enzyme can either be mediated by antibodies, named antibody-

directed enzyme-prodrug therapy (ADEPT) (Bagshawe et al., 2004) or by a gene-

vector, named gene-directed enzyme-prodrug therapy (GDEPT) (Dachs et al.,

2005). Both ADEPT and GDEPT suffer from several shortcomings such as

immunogenicity of an antibody-enzyme conjugate in ADEPT or inefficient

transfection, unsustained gene expression, pathogenesis of viral vectors and the

risk of insertional mutagenesis in GDEPT.

Therefore we introduced a novel enzyme-prodrug strategy in which the

prodrug activating enzyme is not directly linked to a tumor targeting antibody

but encapsulated in a nanometer-sized vesicle, called nanoreactor. We first

evaluated a liposomal enzyme reactor (Huysmans et al., 2005). Nucleoside

hydrolase of Trypanosoma vivax (TvNH), was encapsulated in liposomes. TvNH is

able to hydrolyse nucleoside analogs into the respective nucleobase analogs and

ribose. The vesicles were permeabilised for substrates and products by inserting

porins of Escherichia coli (OmpF and PhoE) into the lipid membrane. By liposome

swelling assays, we could show the successful diffusion of the substrates of TvNH

across the lipid membrane. Functionalized reactors were more effective in

Summary

184

substrate hydrolysis than control reactors, however the control reactors do show

enzymatic activity which indicates that the liposomes are leaky.

Therefore we constructed nanometer-sized reactors that are composed of

PMOXA-PDMS-PMOXA triblock copolymers. These triblock copolymers are able to

self assemble into various aggregates such as vesicles, micelles, nanotubes and

free standing films in aqueous solutions. Their phase behaviour in aqueous

solutions depends on the composition, length and structure of the individual

blocks as well as on the molecular architecture of the whole polymer and the

molecular weight distribution of the individual blocks. Small changes in these

properties can lead to a different behaviour in aqueous solutions. We therefore

evaluated several PMOXA-PDMS-PMOXA polymers from different batches and

different compositions for their ability to self assemble into nanometer-sized

vesicles. We identified two batches that were able to do so: PMOXA18-PDMS72-

PMOXA18 (Ranquin et al., 2005) from the lab of Prof. W. Meier at the university in

Basel and the commercially available PMOXA21-PDMS54-PMOXA21 from Polymer

Source. Furthermore, we were able to make enzymatically active vesicles with

these polymers. To this end, TvNH was encapsulated in the polymeric vesicles

and the polymeric membrane was permeabilized by incorporation of OmpF or

Tsx. These reactors were able to hydrolyse several substrates such as adenosine,

guanosine, inosine, 2-fluoroadenosine and 6-thioguanosine. By fine tuning the

production method, enzymatically active nanoreactors with a mean radius of 75-

100 nm, as determined by DLS, were obtained. AFM measurements did not

confirm the size measured by DLS but this is attributed to the adsorption of the

reactors to the mica chip and the forces applied by the scanning tip which leads

to deformation of the reactors. On the other hand, TEM images confirmed the

size and monodispersity measured with DLS.

After a screening of several potential prodrugs, 6-thioguanosine was selected

as a good candidate. We found that the cytotoxic effect of 6-thioguanine on

neuroblastoma cells only becomes apparent after 24h which is in agreement with

the cytotoxic mechanism of 6-thioguanine (Maybaum and Mandel, 1983; Wotring

and Roti Roti, 1980). Determination of the EC50 values of 6-thioguanosine and 6-

thioguanine showed a difference in cytotoxicity between both compounds.

However, 6-thioguanosine also display toxicity at higher concentrations which is

attributed to the hydrolyses of 6-thioguanosine by human purine nucleoside

summary

185

phosphorylase (hPNP). This could lead to toxicity of the prodrug 6-thioguanosine

towards lymphocytes.

To validate the use of polymeric nanoreactors as an alternative for ADEPT and

GDEPT strategies we explored the possibility to use TvNH encapsulating

nanoreactors in combination with 6-thioguanosine as a novel enzyme-prodrug

therapy. To this end, the cytotoxicity of nanoreactors in the absence and

presence of 6-thioguanosine towards neuroblastoma cells was measured.

Although nanoreactors have a negative effect on cell growth, this is only a minor

effect and it is independent of the nanoreactor concentration. We demonstrated

that 5µM of 6-thioguanosine and 0,33 mg/ml nanoreactors are sufficient to

almost completely kill the neuroblastoma cell culture within 72 h.

Although macromolecules ranging from 10-500 nm in size can be targeted

passively to tumor tissue due to the enhanced permeation and retention effect

(EPR) (Matsumura and Maeda, 1986) (Jain, 1987), more specific targeting by

attaching ligands to the nanoreactors will improve targeting. In this study we

want to couple camelid antibodies to PMOXA-PDMS-PMOXA triblock copolymers.

As a model antibody we used cAbLys-3, an antibody directed against lysozyme.

The presence of a hydroxyl group on the polymers forced us to use the hetero

bifunctional linker PMPI which can be coupled to the polymer via its isocyanate

end and to the antibody with its maleimide group to form a thioether bond with

sulfhydryl residues. Cysteines present in cAbLys-3 are involved in disulphide

bridges and therefore unavailable for coupling.

Therefore, the solved X-ray structure of cAbLys-3 was scrutinized for possible

site-directed cysteine mutants that wouldn’t interfere with antigen binding. We

were able to construct three cAbLys-3 serine to cysteine mutants, S17C, S134C

and S141C that still can bind lysozyme with similar efficiency as the wild type

cAbLys-3. Since expression of the cAbLys-3 mutants in E. coli yields only a low

amount of protein (0.15 – 0.23 mg/l culture medium) as compared to WT

cAbLys-3 (4 mg/l), we used an eukaryotic expression host, Pichia pastoris.

Larger expression yields were obtained through high cell densities and strong

promoters. The expression yield for cAbLys-3 S17C was improved 60 fold by

using P. pastoris.

Although preliminary experiments for the coupling of TvNH-Tsx nanoreactors

to cAbLys-3 have been performed, they were unsuccessful (data not shown).

Summary

186

Much effort still has to be put into fine tuning the chemical reaction and finding

an efficient assay for the detection of nanobody-coupled nanoreactors.

References

Bagshawe, K., Sharma, S. and Begent, R. (2004) Antibody-directed enzyme prodrug therapy

(ADEPT) for cancer. Expert Opin Biol Ther, 4, 1777-1789. Dachs, G., Tupper, J. and Tozer, G. (2005) From bench to bedside for gene-directed enzyme

prodrug therapy of cancer. Anticancer drugs, 16, 349-359. Huysmans, G., Ranquin, A., Wyns, L., Steyaert, J. and Van Gelder, P. (2005) Encapsulation of

therapeutic nucleoside hydrolase in functionalised nanocapsules. J Control Release, 102, 171-179.

Jain, R.K. (1987) Transport of molecules in the tumor interstitium: a review. Cancer Res, 47, 3039-3051.

Matsumura, Y. and Maeda, H. (1986) A new concept for macromolecular therapeutics in cancer chemotherapy: mechanism of tumoritropic accumulation of proteins and the antitumor agent smancs. Cancer Res, 46, 6387-6392.

Maybaum, J. and Mandel, H.G. (1983) Unilateral chromatid damage: a new basis for 6-thioguanine cytotoxicity. Cancer Res, 43, 3852-3856.

Ranquin, A., Versees, W., Meier, W., Steyaert, J. and Van Gelder, P. (2005) Therapeutic Nanoreactors: Combining Chemistry and Biology in a Novel Triblock Copolymer Drug Delivery System. Nano Lett, 5, 2220-2224.

Wotring, L.L. and Roti Roti, J.L. (1980) Thioguanine-induced S and G2 blocks and their significance to the mechanism of cytotoxicity. Cancer Res, 40, 1458-1462.

summary

187

Samenvatting Gerichte enzym-prodrug therapieën hebben als doel de conventionele

chemotherapie te verbeteren door niet toxische prodrugs enzymatisch om te

zetten naar toxische drugs op de plaats van de tumor. Op deze manier worden

de neveneffecten, te wijten aan systemische toxiciteit van conventionele

chemotherapeutica, aanzienlijk verminderd. Gerichte enzym-prodrug therapieën

zijn opgebouwd uit twee fasen. In een eerste fase wordt het prodrug activerend

enzym naar de tumor gericht. In een tweede fase wordt de niet toxische prodrug

systemisch toegediend. Deze prodrug wordt dan enkel ter hoogte van de tumor

omgezet tot de toxische drug door het aanwezige enzym. Het enzym kan op

twee verschillende manieren naar de tumor gericht worden. Enerzijds door de

koppeling van het enzyme aan een antilichaam dat specifiek op tumor cellen

bindt (ADEPT) (Bagshawe et al., 2004) Anderzijds door het gen dat codeert voor

het enzym naar de tumor te brengen via een vector (GDEPT) (Dachs et al.,

2005). Beide systemen hebben te kampen met enkele problemen zoals

immunogeniciteit van de antilichaam-enzym conjugaten in ADEPT en een

inefficiënte gen transfectie, verlies van genexpressie, pathogenese van virale gen

vectoren en het risico op insertionele mutagenese in GDEPT.

Daarom introduceren we een nieuwe methode voor enzym-prodrug therapie

die gebaseerd is op het inkapselen van een prodrug activerend enzym in

nanometerschaal vesikels, die nanoreactoren genoemd worden.

Allereerst werden liposoom-enzym reactoren gemaakt en getest (Huysmans et

al., 2005). Het nucleoside hydrolase van Trypanosoma vivax (TvNH), dat

nucleoside analogen kan omzetten naar nucleobase analogen en ribose, werd

geïncapsuleerd in liposomen. De liposomes werden vervolgens gepermeabiliseerd

door de incorporatie van bacteriële porines afkomstig van E. coli (OmpF en PhoE)

in de lipiden membraan. Liposoom zweltesten toonden aan dat natuurlijke

substraten van dit enzym kunnen diffunderen door deze porines zodat ze kunnen

omgezet worden. Hoewel de TvNH ingekapselde liposomen die gepermeabiliseerd

Samenvatting

188

werden in staat zijn substraten te hydrolyseren, vertonen ook de niet

gepermeabiliseerd controle reactoren een zeker activiteit. Dit duid erop dat de

liposomen doorlaatbaar zijn voor het substraat of het enzyme .

Daarom werden vervolgens nanoreactoren gemaakt die opgebouwd zijn uit

PMOXA-PDMS-PMOXA triblok copolymeren. Deze polymeren zijn in staat om

spontaan aggregaten zoals vesikels, micellen, nanotubes en vrijstaande films te

vormen in waterige oplossingen. Het fasegedrag van zulke polymeren in waterige

oplossingen wordt bepaald door de compositie, de lengte en structuur van de

individuele blokken alsook de complete moleculaire architectuur en de distributie

van het moleculaire gewicht van de individuele blokken. Kleine veranderingen in

deze eigenschappen kunnen leiden tot een ander fasegedrag in waterige

oplossingen. Verschillende PMOXA-PDMS-PMOXA polymeren, afkomstig uit

verschillende synthese batchen en met verschillende composities werden daarom

gescreened. Op deze manier werden twee batchen geïdentificeerd die in staat

zijn om spontaan vesikels te vormen in waterige oplossingen. Bovendien zijn

deze vesicles actief wanneer TvNH wordt ingekapseld en de porines Tsx of OmpF

worden geïncorporeerd in het membraan. De eerste batch, PMOXA18-PDMS72-

PMOXA18 (Ranquin et al., 2005), werd ons geschonken door Prof. W. Meier van

de universiteit in Basel en de tweede batch, PMOXA21-PDMS54-PMOXA21, werd

aangekocht bij Polymer Source. Deze reactoren zijn instaat substraten als

inosine, adenosine, guanosine, 2-fluoroadenosine en 6-thioguanosine om te

zetten naar hun respectievelijke basen en ribose. Door de productie methode van

de nanoreactoren te verfijnen zijn we in staat nanoreactoren te maken met een

gemiddelde straal van 75-100 nm. De grootte van de partikels, die gemeten

werd met DLS, werd niet bevestigd door AFM metingen. Dit is te wijten aan de

adsorptie van de nanoreactoren aan de mica chip en de kracht die de naald

uitoefent op de partikels. TEM beelden daarentegen bevestigen wel de

afmetingen van de reactoren die gemeten werden via DLS. Bovendien tonen

deze beelden monodisperse, sferische vesikels.

Na screening van verschillende potentiële prodrugs, werd 6-thioguanosine

geselecteerd. In onze experimenten komt het cytotoxisch effect van de drug 6-

thioguanine op neuroblastoma cellen komt pas na 24 uur tot uiting. Deze

observatie is te verklaren door het cytotoxisch mechanisme van 6-thioguanine

dat in detail beschreven wordt door Maybaum en Mandel en Wotring en Roti Roti

samenvatting

189

(Maybaum and Mandel, 1983) (Wotring and Roti Roti, 1980). Door de EC50

waarden van zowel 6-thioguanosine als 6-thioguanine te bepalen is het duidelijk

dat 6-thioguanosine bij lage concentraties (~ 1µM) beduidend minder toxisch is

dan de drug 6-thioguanine. Bij hogere concentraties (~ 5µM) vertoont ook de

prodrug 6-thioguanosine cytotoxiciteit. Deze toxiciteit wordt veroorzaakt door de

omzetting van 6-thioguanosine naar 6-thioguanine door endogeen purine

nucleoside phosphorylase. Dat voornamelijk tot expressie komt in lymfocyten.

Om triblok copolymere nanoreactoren te valideren als alternatief voor ADEPT

en GDEPT, werd het cytotoxisch effect op neuroblastoma cellen van TvNH

ingekapselde nanoreactoren, gepermeabiliseerd door Tsx, in combinatie met 6-

thioguanosine nagegaan. Hiervoor werden twee verschillende concentraties

nanoreactoren gebruikt (0,33 mg/ml en 0,66 mg/ml). Beide reactor

concentraties vertoonden weinig effect op de celgroei, in afwezigheid van 6-

thioguanosine. Wanneer 6-thioguanosine wordt toegevoegd, volstaat 5 µM om de

volledige celpopulatie te vernietigen in 72 uur.

Macromoleculen met een grootte van 10-500 nm kunnen via het enhanced

permeation and retention effect (EPR) (Jain, 1987) (Matsumura and Maeda,

1986) op een passieve manier naar tumoren gericht worden. Het gebruik van

antilichamen gericht naar tumoren kan de accumulatie in de tumor nog

verhogen. In deze studie werden kameel antilichamen gebruikt (cAbLys-3) voor

de targeting van nanoreactoren naar tumoren. Omdat de eindstandige groep van

de PMOXA-PDMS-PMOXA polymeren bestaat uit een hydroxyl groep werden we

gedwongen de hetero bifunctionele linker PMPI te gebruiken. Deze linker kan

reageren met hydroxyl groepen op de polymeren en met thiol groepen op het

antilichaam. De 4 cysteine residu’s aanwezig in cAbLys-3, zijn betrokken bij

disulfide bruggen waardoor zij niet in aanmerking komen voor koppeling met de

nanoreactoren. Om deze reden werden er 3 extra cysteine residu’s

geïntroduceerd via gerichte mutagenese van serine residu’s. De serine residu’s

(S17, S134 en S141) werden gekozen om dat ze ver gelegen zijn van de antigen

binding site. Aangezien de expressie van de mutanten in E. coli zeer laag was

(0.15 - 0.23 mg/ml) werd een eukaryoot expressie systeem gebruikt, nl. de gist

Pichia pastoris. Op deze manier werd de expressie van cAbLys-3 S17C 60 maal

verhoogd. Hoewel er reeds preliminaire experimenten werden uitgevoerd om

cAbLys-3 mutanten te koppelen aan nanoreactoren, waren zij niet succesvol.

Samenvatting

190

Referenties

Bagshawe, K., Sharma, S. and Begent, R. (2004) Antibody-directed enzyme prodrug therapy (ADEPT) for cancer. Expert Opin Biol Ther, 4, 1777-1789.

Dachs, G., Tupper, J. and Tozer, G. (2005) From bench to bedside for gene-directed enzyme prodrug therapy of cancer. Anticancer drugs, 16, 349-359.

Huysmans, G., Ranquin, A., Wyns, L., Steyaert, J. and Van Gelder, P. (2005) Encapsulation of therapeutic nucleoside hydrolase in functionalised nanocapsules. J Control Release, 102, 171-179.

Jain, R.K. (1987) Transport of molecules in the tumor interstitium: a review. Cancer Res, 47, 3039-3051.

Matsumura, Y. and Maeda, H. (1986) A new concept for macromolecular therapeutics in cancer chemotherapy: mechanism of tumoritropic accumulation of proteins and the antitumor agent smancs. Cancer Res, 46, 6387-6392.

Maybaum, J. and Mandel, H.G. (1983) Unilateral chromatid damage: a new basis for 6- thioguanine cytotoxicity. Cancer Res, 43, 3852-3856.

Ranquin, A., Versees, W., Meier, W., Steyaert, J. and Van Gelder, P. (2005) Therapeutic Nanoreactors: Combining Chemistry and Biology in a Novel Triblock Copolymer Drug Delivery System. Nano Lett, 5, 2220-2224.

Wotring, L.L. and Roti Roti, J.L. (1980) Thioguanine-induced S and G2 blocks and their significance to the mechanism of cytotoxicity. Cancer Res, 40, 1458-1462.

Publications

191

Publications Dao-Thi MH, Van Melderen L, De Genst E, Buts L, Ranquin A,

Wyns L, Loris R. Crystallization of CcdB in complex with a GyrA

fragment. Acta Crystallogr D Biol Crystallogr. 2004 Jun;60(Pt

6):1132-4.

Ranquin A, Van Gelder P. Maltoporin: sugar for physics and

biology. Res Microbiol. 2004 Oct;155(8):611-6.

Huysmans G, Ranquin A, Wyns L, Steyaert J, Van Gelder P.

Encapsulation of therapeutic nucleoside hydrolase in functionalised

nanocapsules. J Control Release. 2005 Jan 20;102(1):171-9.

Ranquin A, Versées W, Meier W, Steyaert J, Van Gelder P.

Therapeutic nanoreactors: combining chemistry and biology in a

novel triblock copolymer drug delivery system. Nano Lett. 2005

Nov;5(11):2220-4.

Ranquin A, De Vocht C, Van Gelder P. Polymer-based

nanoreactors for medical applications. In: Polymer-based

nanostructures: Medical Applications. P Broz, Editor. Royal Society

of Chemistry. In press.

Ranquin A, De Vocht C, Wilkinson H, Steyaert J, Van Gelder P.

Nanoreactor mediated prodrug activation and killing of

neuroblastoma cells. Manuscript in preparation.

De Vocht C, Ranquin A, Van Ginderachter J, Vanhaecke T,

Versées W, Van Gelder P, Steyaert J. Polymeric enzyme-loaded

nanoreactors for enzyme replacement therapy. Manuscript in

Preparation.