Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Faculty of Science and Bio-engineering Sciences Department of Structural Biology Brussels Polymeric nanoreactors for use in enzyme replacement therapy Thesis submitted in fulfilment of the requirement for the degree of Doctor (Ph.D.) in Bioscience Engineering ir. Caroline De Vocht Academic year: 2009-2010 Promotor: Prof. Dr. ir. Jan Steyaert Co-promotors: Prof. Dr. ir. Wim Versées Dr. Patrick Van Gelder Print: Silhouet, Maldegem © 2010 Caroline De Vocht 2010 Uitgeverij VUBPRESS Brussels University Press VUBPRESS is an imprint of ASP nv (Academic and Scientific Publishers nv) Ravensteingalerij 28 B-1000 Brussels Tel. + 32 (0)2 289 26 50 Fax + 32 (0)2 289 26 59 E-mail: [email protected] www.vubpress.be ISBN 978 90 5487 756 1 NUR 915 / 923 Legal Deposit D/2010/11.161/080 All rights reserved. No parts of this book may be reproduced or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the author. Acknowledgments I would like to start by thanking my promotor Jan Steyaert for giving me the opportunity to do this research and for entrusting this interesting project with high biomedical relevance to me. I would also like to express my gratitude to Wim Versées for the terrific scientific guidance, for his critical view and helpful suggestions. I greatly appreciated your support and learned a lot from you. Furthermore, I would like to thank Patrick Van Gelder for his invaluable contribution to this work and for introducing me to the wonderful world of polymeric drug delivery. A special thanks also goes to An Ranquin upon whom I could always depend on for experimental guidance, for revising this manuscript and for a lot of other things as well. Furthermore, I would like to express my gratitude to all SBB-members for their help and for providing me with a nice and stimulating work-atmosphere! Some people I would like to thank in particular: Adinda, An VDM, Karo, Elke, Inge and Kim; thanks for your support and for always being there for me. I also want to thank all the JAST-desk-colleagues for the great ambiance. Lieven, Klaas and Mike also deserve a special thanks for coming to the rescue several times to help me out with the most diverse problems. Many thanks to Ronnie for helping me with the AFM experiments. I am as well very thankful to Bruno and Maria for their helpful assistance and to Nadine for taking care of all the paperwork. As it seemed very difficult for me to have a peaceful and longlasting relationship with my computer(s), a special thanks also goes out to the IT-guys Wim W and Wim C, who helped me out countless times and saved me from many heart attacks. This project would not have been possible without the expertise of numerous people from other departments sharing their knowledge with me. From the lab of Cellular and Molecular Immunology, I would like to thank Jo Van Ginderachter for his excellent support regarding the immunogenicity studies. I also want to thank Benoit Stijlemans for his help with the LAL-test and Ella Omasta for her expertise concerning working with mice. From the In vivo Cellular and Molecular Imaging lab, a special thanks goes out to Tony Lahoutte and Vicky Caveliers who were so kind to introduce me in the fascinating world of the in vivo imaging and who guided me through the biodistribution studies. I would also like to thank Cindy Peleman for her terrific assistance with the animal imaging work. Furthermore, I would like to express my gratitude to Tamara Vanhaecke, Vera Rogiers and Bart Degreef from the Department of Toxicology for their help with the toxicology studies. Last but not least, I gratefully thank Michio Hirano from the Department of Neurology of Columbia University Medical Center who gave me the opportunity to join his lab in New York for several months. It was a great experience and I learned a lot from this period abroad. Acknowledgements also go to IWT Vlaanderen, FWO Vlaanderen and the Vrije Universiteit Brussel for their financial support. Graag wil ik ook enkele vrienden bedanken die altijd en overal voor mij klaarstonden. Bedankt Anneke voor je steun en luisterend oor. Bedankt ook aan Nick, Evelien, Kia, Nele, Joris, Sara en Tim; de VUB heeft ons samengebracht en ik hoop dat onze vriendschap nog lang blijft voortbestaan. Ook Tom wil ik speciaal bedanken; tijdens de beginjaren van mijn doctoraat was je er steeds om te luisteren en advies te geven en ook nu weet ik dat ik steeds op je kan rekenen. Een welgemeende dankjewel ook voor de vriendinnen-klik uit Antwerpen; Yannique, Katelijne, Marie, Els, Mary, Yoka, Anne-Fré en Céline. Onze vele leuke activiteitjes gaven mij de energie om zo efficiënt mogelijk met mijn tijd om te springen. Bedankt Gaëlle voor de oppeppende skype-babbels, ondanks de kilometers afstand was je er steeds voor me. Bedankt Thomas voor de hulp bij het ontwerpen van de mooie cover. Bedankt Wouter om af en toe voor de nodige verstrooiing te zorgen. Bedankt Carole om steeds die zotte spring-in-’t-veld te zijn waardoor je me wist op te vrolijken wanneer nodig. En bedankt Michiel om me te besmetten met je aanstekelijke drive om ook buiten de lijnen te kleuren. Tot slot wil ik graag mijn familie bedanken voor hun onvoorwaardelijke steun; bedankt Nathalie, Stéphanie, mama en papa om er altijd voor me te zijn en me te stimuleren om in alles wat ik doe steeds het beste van mezelf te geven! Table of contents Chapter 1 Introduction 1.1 Enzyme deficiencies 1.1.1 General 1.1.2 MNGIE, mitochondrial neurogastrointestinal encephalomyopathy 1.2 Enzyme replacement therapy 1.3 Delivery of therapeutic proteins and enzymes 1.3.1 Conventional drug delivery: native and pegylated proteins 1.3.2 Nanotechnology in drug delivery 1.4 Nanoreactors for enzyme delivery 1.4.1 The nanoreactor toolbox 1.4.2 Production methods 1.4.3 Controlling the nanoreactor activity 1.4.4 Biomedical applications 1.5 References 1 3 3 4 15 16 16 18 26 26 33 34 36 40 Chapter 2 Scope of the thesis 53 Chapter 3 Production and biophysical characterization of the nanoreactors 3.1 Introduction 3.2 Materials and methods 3.2.1 Cloning, expression and purification of thymidine phosphorylase 3.2.2 Cloning, expression and purification of porins Tsx and OmpF 3.2.3 Production of nanoreactors 3.2.4 Dynamic Light Scattering 3.2.5 Spectrophotometric assay for TP activity 3.2.6 Atomic Force Microscopy 3.2.7 Detection of endotoxins: LAL assay 3.2.8 Removal of endotoxins 3.3 Results 3.3.1 Providing the nanoreactor building blocks: enzymes and porins 3.3.2 Production, size and dispersity of the nanoreactors 3.3.3 Enzymatic activity of the nanoreactors 3.3.4 Encapsulation efficiency 3.4 Conclusion and discussion 3.5 References 59 61 62 62 63 64 65 65 65 66 67 67 67 70 72 73 75 75 Chapter 4 Stability of the nanoreactors 4.1 Introduction 4.2 Materials and methods 4.2.1 Production of nanoreactors 77 79 79 79 4.2.2 Production of liposomes 4.2.3 Dynamic Light Scattering 4.2.4 Spectrophotometric assay for TP activity 4.3 Results 4.4 Conclusion and discussion 4.5 References 80 80 80 81 84 84 Chapter 5 Toxicity of the nanoreactors 5.1 Introduction 5.2 Materials and methods 5.2.1 Production of nanoreactors 5.2.2 Hepatocyte cell culture 5.2.3 LDH test 5.2.4 MTT test 5.2.5 Statistical analysis 5.3 Results 5.4 Conclusion and discussion 5.5 References 87 89 90 90 90 90 91 91 91 94 94 Chapter 6 Immunogenicity of the nanoreactors 6.1 Introduction 6.2 Materials and methods 6.2.1 Production of nanoreactors 6.2.2 Macrophage cell culture 6.2.3 In vivo inflammatory response study 6.2.4 Cytokine analysis and NO detection 6.2.5 Immunization of animals 6.2.6 Antibody detection 6.3 Results 6.3.1 Ex vivo macrophage response study 6.3.2 In vivo macrophage response study 6.3.3 Antibody response study 6.4 Conclusion and discussion 6.5 References 97 99 100 100 100 101 101 102 102 103 103 107 108 112 114 Chapter 7 Biodistribution of the nanoreactors 7.1 Introduction 7.2 Materials and methods 7.2.1 Nanoreactor synthesis and characterization 7.2.2 Radiolabeling of particles with technetium-99m 7.2.3 Radiolabeling of enzyme with iodine 7.2.4 Animal guidelines 7.2.5 Dynamic planar gamma camera imaging procedure 7.2.6 Image acquisition and processing 7.2.7 Ex vivo analysis 117 119 120 120 120 121 122 122 123 123 7.2.8 Statistical analysis 7.3 Results 7.3.1 Kinetic analysis of short-term biodistribution 7.3.2 Long-term biodistribution analysis 7.4 Conclusion and discussion 7.5 References 124 124 124 127 130 132 Chapter 8 Efficiency of the nanoreactors 8.1 Introduction 8.2 Materials and methods 8.2.1 Production of nanoreactors 8.2.2 In vitro assay and sample processing 8.2.3 HPLC analysis 8.3 Results 8.4 Conclusion and discussion 8.5 References 135 137 138 138 138 139 140 143 143 Chapter 9 Detailed evaluation of the nanoreactor build-up 9.1 Introduction 9.2 Materials and methods 9.2.1 Production of nanoreactors 9.2.2 Characterization of TPE.coli containing nanoreactors 9.2.3 Activity measurements of TvNH containing nanoreactors 9.2.4 Protease treatment 9.2.5 Surface pressure experiments 9.2.6 Fluorescent labeling of thymidine phosphorylase 9.3 Results 9.3.1 Enzyme activity versus porin/enzyme concentration 9.3.2 Protease sensitivity of nanoreactors 9.3.3 Interaction between the enzyme and the polymeric particle 9.3.4 Nanoreactors with other enzymes 9.3.5 Nanoreactors with other polymers 9.4 Conclusion and discussion 9.5 References 145 147 147 147 148 148 148 149 149 150 150 152 154 157 158 159 161 Chapter 10 General conclusion 163 Chapter 11 Summary 171 Chapter 12 Samenvatting 181 Chapter 13 Publications and scientific manifestations 191 Chapter 1 Introduction Introduction 1 Introduction 1.1 Enzyme deficiencies 1.1.1 General A lot of known genetic deficiencies in human are caused by a mutation of a crucial enzyme, leading to the accumulation of one or more of its substrates. As a result toxic concentrations of the substrates appear in the bloodstream and tissues and disturb the normal metabolic degradation pathways. Enzyme deficiencies are usually due to inherited recessive autosomal metabolic defects that give rise to a whole range of rare and fatal diseases, each with their own clinical characteristics. For example, adenosine deaminase (ADA) deficiency is one of the best studied enzyme deficiency disorders. Due to the lack of the enzyme ADA there is an accumulation of the substrates adenosine and deoxyadenosine [1]. The latter causes an increase in S-adenosylhomocysteine which is toxic to immature lymphocytes. As a result, the lymphocytes fail to mature and the immune system is severely compromised or completely absent. As such, ADA deficiency accounts for about 15% of all cases of Severe Combined Immunodeficiency (SCID). Other examples of enzyme deficiencies are Fabry disease [2], Pompe disease [3], Gaucher disease [4, 5], Hurler disease (mucopolysaccharidosis type I) [6], Lesch-Nyhan disease [7] and MNGIE (mitochondrial neurogastrointestinal encephalomyopathy) [8]; all summarized in Table 1-1. 3 Chapter 1 Table 1-1: Examples of enzyme deficiencies with their mutated enzyme and accumulating substrates. Deficiency Enzyme Substrates ADA deficiency adenosine deaminase adenosine, deoxyadenosine Fabry disease -galactosidase A trihexosylceramide Pompe disease -glucosidase glycogen Gaucher disease glucocerebrosidase glucocerebroside Hurler disease -L-iduronidase glycosaminoglycan Lesch-Nyhan disease hypoxanthine-guanine phosphoribosyltransferase hypoxanthine, guanine MNGIE thymidine phosphorylase thymidine, deoxyuridine 1.1.2 MNGIE, mitochondrial neurogastrointestinal encephalomyopathy In this study, thymidine phosphorylase deficiency was chosen as enzyme deficiency disease model system. Thymidine phosphorylase Thymidine phosphorylase (TP) is a cytosolic homodimeric enzyme that catalyzes the reversible phosphorolysis of the nucleosides thymidine and deoxyuridine into the nucleobases thymine and uracil respectively, and 2deoxyribose-1-phosphate (Figure 1-1) [9]. Figure 1-1: Reaction of the enzyme thymidine phosphorylase (TP). 4 Introduction TP is widely expressed in human tissues, including the gastrointestinal system, brain, peripheral nerve, spleen, bladder, lung and at lower levels in muscle. It is not expressed however in kidney, gall bladder, aorta and fat [10-14]. Human TP has been studied extensively and exhibits at least three functions: catalysis (as thymidine phosphorylase), angiogenesis and cell trophism [9, 15, 16]. TP is also called platelet-derived endothelial cell growth factor (PD-ECGF) and endothelial cell growth factor 1 (ECGF1) [17], because of its angiogenic properties, as well as gliostatin, to denote its inhibitory effects on glial cell proliferation [18-20]. The angiogenic and cell trophism of TP may be related to the generation of ribose as a byproduct of thymidine catabolism. Cancer researchers have studied TP extensively because its expression and activity are increased in some tumors, possibly reflecting their profuse vascularization [9]. The three-dimensional structure of thymidine phosphorylase from Escherichia coli [21], Bacillus stearothermophilus [22] and human [23] were solved previously. The enzyme exists as an S-shaped dimer containing two identical subunits with a dimeric molecular mass ranging from 90 kDa in E.coli to 110 kDa in mammals. Each subunit is composed of a small -helical domain of six helices and a large /β domain. The /β domain includes a six-stranded mixed β-sheet and a four-stranded antiparallel β-sheet. The small and large domains are connected by three peptide loops. The active site lies in a cavity between the small and the large domain. The pyrimidine binding site is located at the small α-helical domain and the dimer interface, while the phosphate binding site is located in the large α/β domain (Figure 1-2). The three peptide loops act as a hinge allowing the two domains to move between the open (inactive) and closed (active) conformations. By bringing together the active site residues, this domain movement is critical for the enzymatic activity. 5 Chapter 1 Figure 1-2: Structure of E.coli thymidine phosphorylase. The two different monomers of the S-shaped dimer are presented in green and cyan. The -domains and /-domains are defined with curly brackets. The ligand SO4 (present in the crystallization solution) binds in the phosphate pocket that is located in the large /-domain, while the thymine binds in the small α-helical domain and the dimer interface. The PDB entry 1AZY was used to construct the dimer and the ligands SO4 and thymine were modeled after superposition of the 1TPT onto the 1AZY structure. Oxygen atoms are colored in red, nitrogen in blue, sulfur in orange and carbon in pink. Mendieta et al. described a two-step SN1 mechanism for the TPE.coli phosphorolysis reaction [24]. In the first rate-limiting step the O2 of thymidine interacts with a proton from the imidazole ring of the protonated histidine 85 residue, which makes the pyrimidine base a better leaving group. The weakening of the glycosidic bond generates an oxocarbocation transition state. The second step is a fast reaction between this oxocarbocation and a phosphate dianion, which is generated by deprotonation of the incoming phosphate by the same histidine 85 residu. As such, the phosphate dianion acts as a nucleophile that gives rise to 2-deoxyribose-1-phosphate and thymine. It is clear that in this mechanism the highly conserved histidine 85 plays a crucial role. 6 Introduction Thymidine phosphorylase deficiency causes MNGIE In 1999, mutations in the TYMP gene, encoding the enzyme thymidine phosphorylase, were identified as the cause of mitochondrial neurogastrointestinal encephalomyopathy or MNGIE [8]. The disease locus was mapped to a telomeric region of the long arm of chromosome 22 [25]. The mutations drastically reduce the activity of this cytosolic enzyme, which results in systemic increases of its substrates thymidine and deoxyuridine. This leads to nucleoside pool imbalance which specifically affects the mitochondrial DNA (mtDNA), causing mtDNA instability (depletion, multiple deletions and point mutations). This in turn results in defects of the respiratory chain and mitochondrial dysfunction. The selective effect on mtDNA can be explained by three factors. First, mitochondrial dNTP pools are physically separated and are regulated independently [26-29]. Second, mitochondrial nucleotide pools are probably more vulnerable to the toxic effects of excessive thymidine than nuclear nucleotide pools (which can rely also on the de novo synthesis) (Figure 1-3) [28-30]. Third, human mitochondria probably lack an effective DNA mismatch repair system [31]. Figure 1-3: Schematic representation of thymidine metabolism in MNGIE. Due to a decreased or absent thymidine phosphorylase (TP) activity, increased thymidine is salvaged predominantly in mitochondria, especially in quiescent cells where the mitochondrial thymidine kinase (TK2) is constitutively active. This results in increased dTTP levels, nucleotide pool imbalances and mutations in mitochondrial DNA (mtDNA). Nuclear DNA (nDNA) relies more on the de novo salvage pathway and is not affected by increased thymidine levels. 7 Chapter 1 MNGIE, a mitochondrial disease Mitochondria are the powerhouses of eukaryotic cells. These organelles generate energy in the form of adenosine triphosphate (ATP) from carbohydrates, fats and proteins via oxidative phosphorylation. Mitochondria are unique organelles because they possess their own genetic material, mitochondrial DNA. This feature is due to the origin of mitochondria as prokaryotic endosymbionts of eukaryotic cells. MtDNA is a 16.6 kb, double-stranded, circular molecule. This small genome contains tightly compacted genes for 22 transfer RNAs (tRNAs), 13 polypeptides and 2 ribosomal RNAs (rRNAs). All 13 polypeptides are subunits of the oxidative phosphorylation system: 7 belong to Complex I (NADH-CoQ oxidoreductase), 1 to Complex III (CoQ-cytochrome c oxidoreductase), 3 to Complex IV (cytochrome c oxidase or COX) and 2 to Complex V (ATP synthase). These subunits are synthesized within the mitochondrion where they are assembled together with a large number of subunits encoded by the nuclear DNA (nDNA). The nDNA-encoded subunits are synthesized in the cytoplasm and afterwards transported into the mitochondrion. Figure 1-4 : Schematic representation of the mitochondrial respiratory chain. Protons are pumped from the matrix to the intermembrane space through complexes I, III, and IV, and are pumped back to the matrix through complex V to produce ATP. Coenzyme Q (CoQ) and cytochrome c (Cyt c) are electron (e−) transfer carriers. 8 Introduction Because mtDNA derives from the oöcyte, the mode of transmission of mtDNA and of mtDNA point mutations differs from Mendelian inheritance and is called maternal inheritance. Thus, a disease expressed in both sexes, but no evidence of paternal transmission is strongly suggestive of an mtDNA mutation. Moreover, because there are thousands of mitochondria in each cell, with an average of 5 mtDNAs per organelle, mutations in mtDNA result in two populations of mtDNAs, mutated and wild-type, a condition known as heteroplasmy. The phenotypic expression of a mtDNA mutation is regulated by the threshold effect; the mutant phenotype is expressed in the heteroplasmic cells only when the relative proportion of the mutant mtDNAs reaches a certain critical value. That is why a respiratory chain effect becomes only manifest in some tissues, in which the threshold of mutant mtDNAs is surpassed, and not in others [32]. About 200 mtDNA point mutations and innumerable partial deletions have been associated with human diseases, most of which affect the central and peripheral nervous system. Examples of typical mitochondrial diseases are MELAS (mitochondrial encephalomyopathy, lactic acidosis and stroke like episodes), LS (Leigh syndrome) and LHON (Leber hereditary optic neuropathy) [32]. Although caused by a mutation in nDNA (TYMP gene), MNGIE is a mitochondrial disease (but with a classical Mendelian inheritance). It is caused by defects in the intergenomic communication between the nuclear and mitochondrial genome [33]. Thus, even though MNGIE is not caused by primary mutations in the mitochondrial genome, it finally results in alterations in mtDNA (point mutations, depletions and deletions [34-37]) due to nucleotide pool imbalances. It is still uncertain how mtDNA point mutations are produced in MNGIE, but based on the sequence specificity of the alterations (T-to-C, T-to-A and TT-toAA mutations), two molecular events have been hypothesized as schematically explained in Figure 1-5: a ‘next nucleotide effect’ leading to a high level of direct misincorporation of the incorrect nucleotide [38] and ‘dislocation mutagenesis’ [39]. It was shown that direct misincorporation of a dGMP opposite a template thymine residue is a common event [40, 41]. After the 9 Chapter 1 misincorporation, high concentrations of dTTP in the mitochondria of MNGIE cells may accelerate polymerase activity across the 5’-AA template region [42] (Figure 1-5A). In addition, elevated dTMP and dTTP derived from the increased thymidine level will compromise the mtDNA polymerase exonuclease removal of the mismatched nucleotide [39, 43]. The T.G misincorporation will result in a T-to-C mutation following a subsequent round of mtDNA replication. Secondly, a dislocation mutagenesis model could account for the T-to-A and TT-to-AA mutations. According to this proposed mechanism, mutations are produced by a transient single-nucleotide misalignment in the template DNA. For example, a T residue within a homopolymeric region loops out and leads to the incorporation of dTMP opposite the adjacent A residue. If subsequent nucleotides are incorporated before extrahelical T realigns, a single T deletion will result. Alternatively, if realignment occurs before the next nucleotide is incorporated, the 3’-end T of the daughter strand will realign inappropriately with a template T, leading to a T-to-A transversion [42] (Figure 1-5B). Also depletions and deletions in mtDNA of MNGIE patients were observed. NADHdehydrogenase 5 has been identified as a hot-spot for mtDNA deletions [44]. The multiple deletions were identified in skeletal muscle, brain and kidney. Figure 1-5: Proposed mechanism for site-specific point mutations of mtDNA in MNGIE. A. Model of misincorporation and next-nucleotide effect. B. Model of dislocation mechanism with the next-nucleotide effect [42]. 10 Introduction Physiology and diagnosis The main clinical features of MNGIE are progressive external ophtalmoplegia, severe gastrointestinal dysmotility, cachexia, peripheral neuropathy, leukoencephalopathy and mitochondrial abnormalities [45]. The disease is chronic and progressive. Although the age-at-onset varies from 5 months to 43 years (most commonly the symptoms start in the second or third decade of life), the clinical manifestations are remarkably homogeneous. Gastrointestinal dysfunction is the most prominent feature and leads to progressive loss of weight (average weight loss of 15 kg) up to severe cachexia. Patients usually die of complications of their gastrointestinal problems and their critical nutritional status, with an average age of death of 37 years [46]. In almost all MNGIE patients clear evidence of mitochondrial dysfunction is found, like lactic acidosis, defects of mitochondrial respiratory chain enzymes in muscle, cytochrome c oxidase deficient muscle fibers, depletion in mtDNA and multiple mtDNA deletions [47]. In patients who are suspected of having MNGIE, the determination of TP activity in buffy coats (this is the fraction of the blood containing white blood cells and platelets) (with a spectrophotometric assay) and the determination of thymidine and deoxy-uridine levels in plasma (with a reversed phase HPLC method) are diagnostic. Decreased buffy coat TP activity of lower than or equal to 8% of the activity found in controls and increased concentrations of thymidine (more than 3 µmol/l) and deoxyuridine (more than 5 µmol/l) in plasma is sufficient to diagnose MNGIE (Table 1-2) [48, 49]. When these clinical criteria are reached, mtDNA is sequenced to identify the genetic defect that causes the TP deficiency. Apart from these typical MNGIE patients, some late-onset MNGIE patients (affected at 55 to 61 years old) were reported in which a partial loss of TP activity causes an incomplete and mild MNGIE phenotype [50, 51]. Asymptomatic TYMP mutation carriers have moderate reductions of TP activity but undetectable thymidine and deoxyuridine levels. 11 Chapter 1 Table 1-2: Thymidine and deoxyuridine concentrations in plasma (expressed in µM) and TP activity in buffy coats (expressed as nmol thymine produced/hour/mg TP) of MNGIE patients, late-onset MNGIE patients, TYMP mutation carriers and healthy controls [49, 50]. Number of persons tested Thymidine Deoxyuridine TP activity (µM) mean±SD (Range) (µM) mean±SD (Range) (nmol/h/mg) mean±SD (Range) MNGIE 25 8.6±3.4 (3.9-17.7) 14.2±4.4 (5.5-24.4) 10±15 (0-46) Late-onset MNGIE 3 1.03±0.45 (0.4-1.4) 2.9±1.9 (1.0-4.7) 84±19.5 (58-105) Heterozygotes 14 <0.05 <0.05 222±89 (90-360) Controls 20 <0.05 <0.05 634±217 (253-1000) MNGIE is a rather uncommon disease with worldwide around 130 known cases up until now (personal communication Dr. Hirano). Due to misdiagnosis (most likely as anorexia nervosa) there are probably more cases around. All MNGIE patients are of different ethnic groups and males and females are equally affected. Thymidine phosphorylase deficient mice To further characterize the disease and to better evaluate new therapies, a MNGIE mouse model was generated independently by two different groups [20, 52]. Because the murine uridine phosphorylase (UP) effectively degrades thymidine, deoxyuridine and uridine (in contrast to the human UP which catabolizes the ribonucleoside uridine, but not the deoxyribonucleosides thymidine and deoxyuridine) [53], a double knockout TP-/-UP-/- strain was generated. The TP knockout mouse was produced by inserting a DNA cassette into the pyrimidine phosphorylase consensus sequence in exon 4 of the mouse TP gene through homologous recombination in murine embryonic stem cells. The DNA cassette contains a phosphoglycerate kinase promotor and a neomycin resistance gene flanked by LoxP sites. Subsequently, these cells were microinjected into blastocysts to obtain mutant chimeric mice that generated heterozygous TP-knockout mice [52]. By crossing these TP 12 Introduction knockout mice with CRE-expressing mice the large neomycin cassette inserted into the TP gene was successfully flipped out. Afterwards these mice were backcrossed with C57BL/6 mice for 10 generations to obtain TP knockout animals with a homogenous nuclear background. In a similar way UP knockouts were produced and both strains were mated to obtain TP-/-UP-/double knockouts. TP-/-UP-/- mice were characterized extensively by Lopez et al. [52]. TP activity was undetectable in all studied tissues except liver, which had only a decreased activity. As a consequence thymidine and deoxyuridine were significantly increased by 4-65 fold in all tissues tested. In addition, brain of TP-/-UP-/- mice showed partial mtDNA depletion at 14-18 months, resulting in reduced activities of respiratory chain complex enzymes. Thus, TP-/-UP-/- mice recapitulate several features of MNGIE patients including TP deficiency, elevated thymidine and deoxyuridine levels in plasma and tissues, mtDNA depletion and respiratory chain defects. Nevertheless, the last two features could only be observed in the brain of the double knockout mice. The brain specific phenotype may be due to the shorter life span of the mice compared with humans and the relatively modest increases of nucleosides in mutant mice. First, the average age-at-onset of symptoms in MNGIE is 18.7 years; therefore the 2-year lifespan of mice may not be sufficient to allow significant accumulation of mtDNA alterations and consequent respiratory chain dysfunctions. Secondly, in contrast to humans who have undetectable amounts of thymidine and deoxyuridine in tissues and plasma, wild type mice have detectable amounts of these nucleosides; therefore TP-/-UP-/- mice show only 4-65 fold increases in nucleosides, whereas MNGIE patients show >100 fold increases [54]. The group of Dr. Hirano is now trying to enhance the mitochondrial phenotype of the TP-/-UP-/- mice by stressing them with exogenous thymidine and deoxyuridine (in drinking water). Therapeutic approaches Since systemic accumulations of thymidine and deoxyuridine are the cause of MNGIE, strategies that are able to reduce the toxic concentrations of these 13 Chapter 1 circulating nucleosides may be therapeutic [55]. Attempts to reduce thymidine concentrations in the blood of two patients with MNGIE through hemodialysis indicated that although repeated dialysis could reduce the basal level of circulating thymidine, a more continuous elimination of nucleosides is needed for a permanent reduction [48, 56]. Because in healthy individuals platelets have high TP activity [57], repeated platelet infusions were used to try to restore TP enzyme activity in MNGIE patients [58]. In vitro and in vivo studies showed that the nucleoside concentrations could be reduced to undetectable levels. Unfortunately the decrease in toxic nucleoside levels is only transient and a more effective restoration of the TP activity is required. In 2006, Dr. Hirano started to use allogeneic stem cell transplantations to provide long-term restoration of TP activity. From the 2 MNGIE patients initially treated significantly with improved allogeneic her stem quality of cell transplantations, life and 1 demonstrated patient virtually undetectable thymidine and deoxyuridine levels in plasma, while the other patient died several weeks after the transplantation [59]. Up until now, transplants were attempted in total in 12 patients. Two had problems with the conditioning regimen which include chemotherapy to partially ablate bone marrow and make space for donor cells. From the 10 patients that underwent transplants, 5 have died and 5 showed biochemical and clinical improvements (personal communication Dr. Hirano). In the cases where the transplantation failed, the patients were already severely debilitated by MNGIE and/or the quality and source of the donor stem cells were not optimal. Rejection of the donor stem cells is a well-known and intrinsic risk of stem cell transplantations in general. Since erythrocytes encapsulated with adenosine deaminase have been used successfully in a patient with adenosine deaminase deficiency [60], a carrier erythrocyte entrapped thymidine phosphorylase therapy (CEETP) was used to treat one MNGIE patient [61]. Although the plasma and urine concentrations decreased significantly, the patient’s clinical condition remained poor and she died of pneumonia 21 days after CEETP. This approach is however promising since the encapsulation of the TP prevents the formation of antibodies and 14 Introduction maintains the enzyme activity during the erythrocyte life span. In general, enzyme replacement strategies have shown their potential in other enzyme deficiencies as well. 1.2 Enzyme replacement therapy In enzyme replacement therapy (ERT) the deficient enzyme is administered intravenously in its native or stabilized form. Enzyme replacement therapy is in general a well-tolerated and safe strategy to ‘treat’ or better ‘stabilize’ enzyme deficiency diseases [62]. Because the substrates, in most cases nucleosides, equilibrate rapidly across cell membranes (by nucleoside transporters) [63], high substrate concentrations in tissues and organs lead to high plasma levels. Therefore, restoring the enzyme activity in plasma is sufficient to reverse the principal abnormalities and intravenous injections of the enzyme are administered. In most cases immunosuppressive drugs (like cyclosporine A and azathioprine) are co-administered to prevent a strong antibody response and induce immune tolerance to the therapeutic enzyme [64]. Since ERT has been successfully introduced for patients with Gaucher disease by Brady and coworkers [65], this kind of treatment has been taken into consideration for other enzyme deficiency disorders (mostly lysosomal storage diseases) as well. Clinical trials have demonstrated the clinical benefit of this therapeutic approach in Fabry disease [66], mucopolysaccharidosis type I (Hurler disease), II and VI [67] and Pompe disease [68]. In most enzyme deficiency disorders the substrate concentrations accumulate during the first years of life until a threshold is exceeded and the disease pass from a preclinical to a clinical state with physical symptoms. Thus, timing of the ERT is very important; the earlier the ERT starts, the better the response to therapy [69]. Unfortunately, good biomarkers that allow fast diagnosis and adequate judgment of the stage of the disease are currently lacking (or not used frequently because of the rarity of the disorders) and in most cases ERT is only started in the clinical state, after appearance of the symptoms [70]. 15 Chapter 1 The biggest limitation of ERT is probably the financial burden of the treatment. Large pharmacological quantities of recombinant enzyme need to be produced and an individualized approach to treatment is followed rather than a set dosage regimen [71]. As a result, average annual costs per patient are very high with estimated averages of 180 000 USD for Gaucher disease, 215 000 USD for Fabry disease and 260 000 USD for Hurler disease [72]. Importantly, ERT is not a cure that completely eliminates all signs and symptoms, but a treatment that can improve the quality of life significantly. Patients need repeated intravenous injections, probably two to four weekly and potentially lifelong, to achieve a continuous elimination of the toxic substrates. Prolonging the circulation time of the therapeutic enzyme would lead to decreased frequency of administration and decreased costs of the therapy and as such would mean a tremendous improvement for patients suffering from such a genetic disorder. 1.3 Delivery of therapeutic proteins and enzymes 1.3.1 Conventional drug delivery: native and pegylated proteins A lot of proteins, in particular hormones, blood factors, vaccines, interferons, monoclonal antibodies and enzymes, can be very useful as therapeutics in various biomedical applications. Since the approval of recombinant human insulin in 1982, many biotechnology products were shown to be effective and safe for the treatment of numerous diseases [73]. The first generation pharmaceuticals were recombinant proteins with an amino acid sequence identical to that of the native human polypeptide. These formulations however often do not possess optimal drug characteristics and suffer from low stability and solubility, very short circulation half-lives, sensitivity to proteolytic degradation and immunogenicity and toxicity issues [74]. Therefore, an increasing number of second generation protein therapeutics are engineered to reach better efficiency. 16 Introduction To improve their stability, proteolytic resistance, immunogenicity and circulation half-life, therapeutic proteins can be modified by covalently linking various poly(ethylene glycol) (PEG) molecules to their surface, a technique called pegylation [75]. PEG has a long history as a non-toxic, nonimmunogenic, hydrophilic, uncharged and non-biodegradable polymer and has been approved by the FDA (Food and Drug Administration) as ‘generally recognized as safe’. In the beginning, linear PEG reagents of 5 to 10 kDa were used. Later, branched, forked and multi-arm PEG chains and PEG dendrimers reaching molecular masses up to 60 kDa became quite common [76, 77]. Since PEG is highly flexible and can coordinate numerous water molecules, pegylated proteins have a greatly expanded hydrodynamic volume with an apparent molecular weight 5 to 10 times higher than that of a globular protein with the same molecular mass. This results in reduced kidney filtration. In addition, PEG chains exhibit stealth properties by shielding the protein from its environment, hereby protecting it against proteolytic degradation and recognition by the immune system (epitope shielding). Finally, pegylation can mask or modify protein charges which prevents interaction with charged blood components. All these features result in a prolonged circulation half-life (from 2 to 400 times longer) of the pegylated protein in comparison with the native protein [78]. The FDA has approved several pegylated proteins as therapeutics, and probably more are under investigation (Table 1-3). However, pegylation also has disadvantages. The attachment of PEG molecules to a protein can modify the interaction capabilities or active sites of the protein that are responsible for its biological function [79]. In addition, the nonimmunogenicity of PEG became questionable since specific anti-PEG antibodies were found in some clinical trials. Some reports indicate the loss of longcirculating properties when the PEG-protein conjugate is administered multiple times due to a strong anti-PEG immune response [80, 81]. Although widely used, pegylation is a rather complicated and expensive approach. In particular when the therapeutic protein is an enzyme, there is a serious risk of a decrease or a complete loss of enzyme activity. Because the rather expensive production of recombinant protein is followed by an 17 Chapter 1 inefficient coupling to the PEG chains and multiple purification steps, costs of this technology are quite high [79]. Furthermore, since pegylation includes a chemical modification, the whole developing process from bench to market needs to pass numerous trials, requiring a lot of time and money. A more convenient strategy to deliver therapeutic enzymes to the body is to encapsulate them in appropriate carrier systems. Table 1-3: Approved PEG-protein conjugates. PEG conjugate Company Disease Year to market PEG-adenosine deaminase (Adagen®) [82] Enzon Pharmaceuticals Severe Combined Immunodeficiency disease (SCID) 1990 PEG-asparaginase (Oncaspar®) [83] Enzon Pharmaceuticals Acute lymphoblastic leukemia 1994 PEG-interferon 2b (PEG-Intron®) [84] Schering-Plough Hepatitis C 2000 PEG-interferon 2a (Pegasys®) [85] Hoffman-La Roche Hepatitis C 2002 PEG-granulocyte colony stimulating factor (Filgrastim, Neulasta®) [86] Amgen Treating neutropenia during chemotherapy 2002 PEG-growth hormone receptor antagonist (Pegvisomant, Somavert®) [87] Pfizer Acromegaly 2002 1.3.2 Nanotechnology in drug delivery Nanotechnology (derived from the Greek word ‘nano’ meaning dwarf) is an emerging branch of multidisciplinary science that deals with designing tools and devices in the nanometer-size-range. Over the last decades new nanotechnology-based drug delivery systems were developed for the controlled release and/or the targeted delivery of proteins in the body [88, 18 Introduction 89]. The carrier acts as a depot for high concentrations of therapeutics and provides a solubilising and protective environment. This can be used either for local delivery, where the carrier is retained at the site of delivery, or for systemic delivery, where it circulates through the vasculature, increasing the circulation time of the therapeutic agent [90]. Liposomes Liposomes are artificial phospholipid vesicles made up of an aqueous interior surrounded by a lipid bilayer (Figure 1-6A). They are composed of one or more bilayers of natural and/or synthetic lipids like PE (phosphatidylethanolamine), DOPC (dioleoylphosphatidylcholine) or EPG (egg phosphatidylglycerol) which are relatively biocompatible and biodegradable. The vesicles are formed spontaneously when the lipids are hydrated in aqueous solutions. The major driving force behind this self-association is the amphiphilic character of the lipids which favors a micelle or vesicle conformation where the hydrophobic tails are removed from the aqueous surrounding and the hydrophilic parts are exposed into the water. Liposomes have been investigated as carriers of various pharmacologically active agents. Drugs with varying lipophilicities can be encapsulated in liposomes, either in the phospholipid membrane (strongly lipophilic drugs), in the entrapped aqueous volume (strongly hydrophilic drugs) or at the bilayer interface (drugs with intermediate lipophilicity) [91]. The sustained release of the drugs occurs gradually due to leakage of the substance or degradation of the liposome. In addition, 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. Such a triggered release can be accomplished by pHsensitive [92], thermo-sensitive [93] or photo-sensitive [94] liposomes. Some liposomal formulations have reached the market or are now entering clinical trials [95, 96]. Ambisome® (Gilead Sciences) in which the encapsulated drug is the antifungal amphotericin B [97], Myocet® (Elan Pharmaceuticals, Inc) encapsulating the anticancer agent doxorubicin [98] and Daunoxome® (Gilead Sciences) in which the drug incorporated is daunorubicin [99], are 19 Chapter 1 some examples of approved liposomal drugs. Note that plasma half-lives are around 7-10 hours for Ambisome®, 2-3 hours for Myocet® and 5 hours for Daunoxome® in comparison with less than 0.5 hours for the free drugs. In addition, several studies reported the encapsulation of enzymes inside liposomes [100]. The enzyme must be liberated from the particle before it can catalyze a particular reaction occurring outside the liposome. Enzymecontaining liposomes are under development for various biomedical applications, including enzyme replacement therapy. Table 1-4 gives an overview of enzyme containing lipid particles constructed and investigated for their use in enzyme replacement therapy. Table 1-4: Selected examples of enzyme-containing liposomes for potential use in enzyme replacement therapy [101]. Investigation Reference Amyloglucosidase Enzyme First proposal to use enzyme-containing liposomes in treatment for Pompe disease [102] Dextranase Development of a model system for a lysosomal storage disease in rat [103] -Galactosidase In vitro evaluation of the use of enzymecontaining liposomes for the treatment of cultured gangliosidosis fibroblasts [104] Glucocerebrosidase Treatment of a patient with Gaucher disease [105] -Glucuronidase In vivo investigation in mice of the effect of charged enzyme-containing liposomes [106] Liposomes protect the encapsulated molecules from rapid degradation resulting in prolonged circulation persistence. Due to their enhanced size the clearance route shifts from the kidneys (size-selective cut-off for glomerular filtration is approximately 60 kDa) to the reticuloendothelial system (RES), also called mononuclear phagocyte system (MPS) [107, 108]. RES clearance is based on the phagocytotic uptake most often by macrophages especially in the 20 Introduction liver, spleen and lymph nodes. However, when the target site is beyond the RES system, liposome uptake by macrophages and their consequent removal from circulation is one of the main disadvantages for the use of liposomes as drug delivery systems. RES does not recognize the liposomes themselves but rather the opsonins which are serum proteins bound to the surface of liposomes [109, 110]. Complement components, which play a role in host defense against invading pathogens, comprise another important system able to recognize liposomes [111]. This system acts through initiating membrane lysis and enhancing uptake by phagocytotic cells. Another limitation of liposomes is their instability in plasma due to their interaction with high (HDL) and low density (LDL) lipoproteins, since this interaction results in the premature release of the encapsulated drug into the plasma [112]. Pegylated liposomes An effective strategy to increase the circulation time of liposomes is to incorporate PEG polymers in the lipid bilayer (Figure 1-6B) [113]. These PEG chains form a protective coat on the liposome surface and decrease recognition by opsonins and uptake by the RES system [114]. Allen and coworkers [107] were the first to show the long-circulating effect of pegylated liposomes and the names ‘stealth liposomes’ (‘stealth liposomes’ is a registered trademark of Liposome Technology, Inc.) or ‘sterically stabilised liposomes’ have been given to this new class of liposomes. By reducing RES uptake, long-circulating liposomes can passively accumulate inside tissues and organs. This phenomenon, called passive targeting, is especially useful in solid tumors [115]. Due to the leaky vasculature, longcirculating liposomes are preferentially accumulated in the tumor area [116]. As such, sterically stabilized liposomes are explored to encapsulate anti-cancer drugs [117]. Some examples of stealth liposomal-based anti-cancer ® pharmaceuticals are Doxil , a pegylated liposomal formulation of doxorubicin (half-life of 1.5 to 45 hours) [118], Lipoplatin®, a pegylated liposomal formulation of cisplatin (half-life of 60-117 hours) [119] and SPI-077®, another pegylated liposomal formulation of cisplatin (half-life of 60-117 hours) 21 Chapter 1 [120]. Stealth liposomes can also be used for active targeting to specific tissues by coupling targeting moieties like ligands, antibodies or growth factors to the terminal of the PEG chain or the phospholipid head group [121-124]. Although PEG remains the standard polymer for the steric stabilization of liposomes, other polymers also possess stealth-like properties such as polyvinyl pyrrolidone [125], polyvinyl alcohol [126], polyacril amide [127] and polymethyl and polyethyl oxazoline [128, 129]. These polymers appear to be attractive alternatives for designing long-circulation liposomes. Unfortunately, stealth liposomes still have some shortcomings and are not ideal as drug delivery vehicles. Fast drug leakage due to the rather unstable character of lipidic particles in plasma [130, 131] and accelerated blood clearance remain major drawbacks of these liposomal delivery systems. The latter was reported by Ishida et al. [132, 133] and Laverman et al. [134] which show that intravenous injection of PEG-grafted liposomes may significantly alter the pharmacokinetic behavior of a second dose. Recent evidence also shows that sterically stabilized liposomes are still prone to opsonization. The dense PEG brush does delay opsonization, but selective protein deposition eventually occurs after 10 to 20 hours [135, 136]. The prolonged circulation longevity of stealth liposomes is now believed to be rather due to diminished aggregation of liposomes. Figure 1-6: Schematic representation of liposomes (A) and stealth liposomes (B). 22 Introduction Polymeric particles Owing to their low molecular weight (MW < 1 kDa), aggregation of lipids results in molecularly thin membranes that possess a high dynamical and physical softness. As a consequence, many liposome properties like stability, degradation and drug retention are difficult to control. In order to obtain more robust membranes with controllable properties, extensive efforts were made within the last decades to design polymeric vesicles. The polymeric nanoparticles described in the literature so far are constructed from block copolymers. Such a block copolymer consist of distinct polymer chains covalently linked in a series of two or more segments [137]. Amphiphilic block copolymers are composed of at least one hydrophilic block and one hydrophobic block, causing self-assembly in aqueous solutions to nanometer-sized structures. 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 weight, hydrophobic-to-hydrophilic balance and block-length, amphiphilic block copolymers can aggregate into micelles (Figure 1-7A), vesicles, cylinders or rod-like structures [138]. Compared to the selfassembled structures formed by lower molecular weight amphiphilic molecules such as lipids or surfactants, block copolymers self-assemble into significantly thicker, less permeable and mechanically more stable particles [139]. 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 [140, 141]. It is this increased stability, along with their selfassembled nanometer-sized structures that make block copolymers so attractive for biomedical applications such as drug delivery. Numerous natural and synthetic polymers have already been used for the production of drug delivery nanoparticles. Examples of polymers for the hydrophobic block include inert PEE (polyethylethylene), PS (polystyrene), PDMS (polydimethylsiloxane), the degradable PLA (polylactic acid) and PCL (polycaprolactone). Hydrophilic blocks have been synthesized from PEG 23 Chapter 1 (polyethylene glycol) or PEO (polyethylene oxide), the negatively charged PAA (polyacrylic acid) and the crosslinkable PMOXA (polymethyloxazoline) [142]. Up until now, PLGA (polylactic-co-glycolic acid) is the most frequently used copolymer, synthesized from the two monomers lactic acid and glycolic acid. This polymer is FDA approved for the use in therapeutic devices, due to its biodegradability and biocompatibility [143]. Polymeric particles can be composed of diblock copolymers (Figure 1-7B) [144], triblock copolymers (Figure 1-7C), branched polymeric dendrimers or mixed polymer-lipid composites [145]. Analogous to liposomes, stimuliresponsive polymeric particles have been developed based on the triggered release of their content by destabilization of the polymeric membrane upon specific stimuli like pH, temperature, redox potential or ultrasound [146]. Injection of polymeric particles into the blood circulation of rats and mice shows prolonged circulation times. Clearance half-lives range from 15 to 30 hours and significantly exceed the circulation times of stealth liposomes in similar dose injections [147]. Figure (A), in triblock colored 1-7: Schematic representation of proteins incorporated in polymeric micelles particles composed of diblock copolymers (B) and in particles composed of copolymers (C). Hydrophilic blocks are colored in blue, hydrophobic blocks are in grey. Various proteins have been entrapped in nanoparticles (typically between 50250 nm), either as model system, or as possible therapeutic application (Table 1-5 gives a snapshot of some examples). Enzymes, such as -galactosidase 24 Introduction entrapped in nanoparticles of polybutylcyanoacrilate (PBCA) [148], or superoxide dismutase loaded in polylactide-co-glycolide (PLGA) nanoparticles [149], were delivered to the brain. In addition, some examples are reported of nanoparticles constructed for the potential use in enzyme replacement therapy. Genta et al. first encapsulated the enzyme prolidase inside particles consisting of PLGA [150], later inside nanoparticles consisting of chitosan [151] and tested them in the enzyme therapy for prolidase deficiency. Garnacho and coworkers designed a new delivery carrier, based on acid sphingomyelinase encapsulation in polystyrene and PLGA nanocarriers targeted to an endothelial surface protein to use in enzyme replacement therapy for type B Niemann-Pick disease [152]. Nanotechnology, especially nanomedicine, is a rapidly growing field and will lead to more medical benefits in the coming years. Nevertheless, the development of biocompatible and biodegradable drug carriers possessing small particle size, high loading capacity, extended circulation time and ability to accumulate in required pathological sites in the body still has many unresolved issues and remains a challenging field of research. Table 1-5: Examples of polymeric nanoparticles used as carrier for proteins. Polymer Payload Proposed application Reference PEG-PLA catalase model system [153] chitosan FITC-BSA vaccine [154] polyacrylamide ovalbumine vaccine [155] PEG-PLA and PEGPCL paclitaxel and doxorubicin cancer therapy [156] PLGA superoxide dismutase oxidative stress [157] PLA cyanine dye imaging [158] PEO-PPS-PEO glucose oxidase stimuli responsive particles releasing their content upon exposure to oxidative environment [159] 25 Chapter 1 1.4 Nanoreactors for enzyme delivery Nanoreactors are defined as submicrometer hollow spheres that encapsulate enzymes in their aqueous interior and that incorporate channel forming proteins inside their wall, hereby permeabilizing the nanoreactor. The enzyme remains inside the particles and the substrates and products can diffuse in and out. As such, the enzymatic reaction is restricted to the inner volume of the nanocontainer, which is in contrast to the ‘classical’ enzyme delivery particles where the enzymes need to be released before they can catalyse the reaction. As a consequence, nanoreactors avoid the problems of release profile optimization or the release of therapeutics in inappropriate compartments compared with the conventional carriers. 1.4.1 The nanoreactor toolbox Polymers Although few examples are known of liposomal reactors [160-162], the majority of nanoreactors reported thus far are constructed of block copolymers. The polymer that is most commonly used is the PMOXA-PDMSPMOXA or poly(2-methyloxazoline)-poly(dimethylsiloxane)-poly(2- methyloxazoline) triblock copolymer (Figure 1-8) [163]. The main advantage of this triblock copolymer is the versatility of the self-assembled particles; 1) the shell has a hydrophilic, biocompatible low-protein-binding surface and an extremely high mechanical and thermal stability, 2) the shell can act as a biomembrane for protein reconstitution, 3) the nanovesicle can encapsulate hydrophilic substances, and 4) the shell material has enormous possibilities for molecular functionalization (like crosslinking and targeting). Also, the nontoxic and biodegradable PEG-oligo(DTO suberate)-PEG or glycol)-block-oligo(desamino-tyrosyltyrosine octyl ester poly(ethylene suberate)-block- poly(ethylene glycol) [164] and the PEO-PDMS-PMOXA or poly(ethylene oxide)-block-poly(dimethylsiloxane)-block-poly(2-methyloxazoline) as an asymmetric block copolymer [138] have been used to construct nanoreactors. 26 Introduction Figure 1-8: Structure of poly(2-methyloxazoline)-poly(dimethylsiloxane)-poly(2methyloxazoline) or PMOXA-PDMS-PMOXA triblock copolymer. Porines Nature provides a great variety of specific and non specific channel proteins which allow translocation of substrates across biological membranes. It is clear that membrane-like structures formed by amphiphilic block copolymers closely resemble typical biological membranes. Actually, it has been shown that membrane proteins can be successfully reconstituted in such artificial polymer membranes [165]. The block copolymer membranes are considerably thicker (e.g. 10 nm) than conventional lipid bilayers (3-5nm) due to the larger size of the underlying copolymer molecules. This raises the question whether the proteins can preserve their activity within a block copolymer membrane. However, the high flexibility of the hydrophobic block and the conformational freedom of the polymer molecules allows a block copolymer membrane to adapt to the specific geometric and dynamic requirements of membrane proteins without considerable loss of free energy and thus minimizing the hydrophobic mismatch [166] (Figure 1-9). This technology can be used on the one hand to study protein-membrane interactions (e.g. antimicrobial proteins like alamethicin) [167] and on the other hand to permeabilize polymeric membranes for substrates and products and make functionalized nanoreactors for drug delivery applications. 27 Chapter 1 Figure 1-9: A molecular dynamics simulation of stable insertion of a mimetic protein pore into either thin (A) or thick (B) block copolymer membranes (taken from [142]). Up to now most channels used in nanoreactor design are derived from the Gram-negative outer membrane [168]. These proteins form β-barrels and can be produced and isolated in large quantities (tens of mg/L culture) [169] and are extremely rigid (stable up to 70°C in the presence of 2% sodium dodecyl sulphate, in organic solvents and 4 M guanidinium hydrochloride). The integration of protein channels into block copolymer membranes was initially realized by Meier and coworkers. They demonstrated that the general diffusion porin OmpF (Outer membrane protein F) can be integrated in PMOXA-PDMS-PMOXA copolymer membranes [170]. Surprisingly, the functionality of the OmpF protein was fully preserved despite the artificial surrounding. This channel functions as a molecular sieve, allowing concentration-driven diffusion of molecules with a molecular weight below 600 Da. There is no selectivity for the transported molecules, although OmpF shows slight cation selectivity [160, 171]. PhoE is also a general diffusion porin but it favours transport of more anionic molecules [168]. Up until now, PhoE however has not been used to permeabilize nanoreactors. Secondly, Graff et al. reconstituted the bacterial channel forming protein LamB (or Maltoporin) in ABA-copolymer vesicles [172]. The outer membrane protein LamB is a specific transporter for maltodextrins but also serves as a receptor for λ phage and triggers the injection of DNA. Since DNA translocation was 28 Introduction observed, this study proved again that porins can preserve their activity within a completely synthetic block copolymer membrane (Figure 1-10). Figure 1-10: Schematic representation of a DNA-loaded PMOXA–PDMS–PMOXA vesicle. λ-phage binds a LamB protein, and the DNA is transferred across the block copolymer membrane. Tsx is an outer membrane transporter of Gram-negative bacteria constructed of a monomeric 12-stranded -barrel. In contrast to the general porines, Tsx allows specific transport of nucleosides and nucleotides. Since it has at least three binding sites for nucleosides in the interior of the channel, rapid transport even at low concentrations is possible [173]. The channel is shaped like a keyhole; upon binding the base moiety of the nucleoside is located in the narrow part of the keyhole, while the bulky sugar occupies the wider opening (Figure 1-11). Pairs of aromatic residues and flanking ionizable residues are involved in nucleoside binding. Nucleoside transport occurs by diffusion from one binding site to the next [174]. The direction of transport is driven by the concentration gradient across the membrane. Apparently the Tsx-dependent uptake differs slightly for the various nucleosides. Among the five nucleosides investigated, the binding increased in the following order: cytidine, guanosine, uridine, adenosine and thymidine. Despite its specificity, the Tsx pore also allows concentration-driven aspecific diffusion of small molecules unrelated to its specific substrates [175]. 29 Chapter 1 Figure 1-11: Nucleoside binding sites in the Tsx channel. 1. Surface representation showing the keyhole shape of the Tsx pore (left panel). The right panel shows a close-up of the channel with a nucleoside bound at Nuc1. 2. Stereo view of the nucleoside binding sites in thymidine-soaked crystals. The nucleosides that could be built in the density at Nuc1 and Nuc2 are shown in green. The aromatic residues that line the channel and that are involved in nucleoside binding are indicated in cyan. The front (F) and the back (B) of the barrel are indicated (taken from [174]). All previously discussed porins are rather small in diameter (7-11 Å) and molecules with a molecular weight greater than 600 Da have difficulties passing through these small channels. To overcome this, Nallani and coworkers introduced a two deletion mutant of the protein FhuA to permeabilize nanoreactors which they dubbed synthosomes. This monomeric protein is an outer membrane receptor for the uptake of iron scavenging siderophores and consists of 22 β-strands wrapped around a plug-domain. 30 Introduction Removal of this internal domain results in a channel with a large pore diameter (39-46 Å) that ensures rapid compound flux [176]. Kumar and coworkers described the incorporation of the bacterial waterchannel protein Aquaporin Z into a PMOXA-PDMS-PMOXA triblock copolymer membrane with a large hydrophobic-to-hydrophilic block ratio. The water permeability varied with the porin/polymer ratio and showed a clear channelmediated transport. Furthermore, the selectivity of the membrane for water over small solutes such as salt, glucose, urea and glycerol was demonstrated [177]. Symmetry and orientation of the membrane plays a crucial role for the insertion of integral membrane proteins into artificial membrane systems. Stoenescu and coworkers demonstrated this using an Aquaporin that was fused to a poly-his-tag. Using antibodies that are specifically directed against the his-tag, it was possible to investigate the orientation of the proteins in the artificial membrane. The asymmetric ‘ABC’ triblock copolymer membranes always favor one orientation, while there is no preferred orientation in symmetric ‘ABA’-polymer membranes [138]. Enzymes In contrast to the long list of enzymes entrapped in liposomes (eg: alkaline phosphatase, asparaginase, chymotrypsin, elastase, lysozyme, peroxidase, proteinase K, … [101]), up until now only few examples are reported where enzymes are entrapped inside the aqueous space of polymeric nanoreactors. As proof of principle for the nanoreactor-technology Meier and coworkers encapsulated the enzyme β-lactamase in PMOXA-PDMS-PMOXA particles. The enzyme is able to hydrolyze β-lactam antibiotics like ampicillin which in turn can reduce iodine to iodide. This reduction was used to monitor the activity and functionality of the enzyme inside the nanoreactor. In this nanoreactormodelsystem the enzyme is encapsulated in the aqueous interior of the nanoreactor and the substrate ampicillin is transported across the membrane via the porin OmpF [178]. As these nanoreactors clearly showed enzyme 31 Chapter 1 activity, it was evident that both the channel protein and the encapsulated enzyme remain functional after incorporation in a nanoreactor system [179, 180]. Also acetylcholinesterase, which hydrolyses the neurotransmitter acetylcholine, could succesfully be entrapped in PMOXA-PDMS-PMOXA particles whithout loss of function. This enzyme was chosen to provide an easy and accurate biosensor to detect insecticide residuals [171]. The encapsulation of a prodrug activating enzyme led the nanoreactor strategy to its first biomedical proof of principle. Prodrug activating enzymes are promising tools in cancer therapy as they can convert a non-toxic prodrug into a cytotoxic agent. In Antibody Directed Enzyme Prodrug Therapy (ADEPT), the enzyme is targeted to the tumor in a fusion with a monoclonal antibody in order to avoid systemic toxicity. Although this strategy has advantages compared to non-targeted chemotherapy, the major obstacle is the immunogenicity of the exogenous enzyme. The encapsulation of the prodrug activating enzyme nucleoside hydrolase of Trypanosoma vivax (TvNH) inside DOPC/EPG liposomes permeabilized with OmpF was a first attempt to improve this prodrug therapy [161]. Unfortunately such lipidic particles are unstable in blood serum and need to be grafted with poly(ethylene glycol) (PEG) or other polymers to increase the average circulation time. For this reason, a more promising kind of nanoreactor, composed of PMOXA-PDMS-PMOXA triblock copolymer was constructed. These nanoreactors encapsulate the prodrug activating enzyme TvNH and are further functionalized by incorporating bacterial porins in the reactor wall, thus allowing transport of solutes across the membrane. The reactors can efficiently hydrolyze different substrates including the prodrug 2-fluoroadenosine, resulting in the release of the cytotoxic molecule, 2-fluoroadenine [181]. Recently, more enzyme-containing nanoreactors are being developed for various biomedical applications (see chapter 1.4.4). 32 Introduction 1.4.2 Production methods Up until now there are 3 different methods described to prepare functional nanoreactors: the solvent evaporation method, the lamellar film rehydration method and the direct dispersion method. The reported preparation methods are only used on laboratory scale and produce a few milliliters of nanoreactor suspension. Solvent evaporation method In the solvent evaporation method, a solution of purified porin is mixed with copolymer dissolved in an appropriate solvent (like ethanol), to the desired molar ratio of protein to polymer. For encapsulation of the enzyme in the interior of the vesicles, the homogeneous polymer-porin solution is added dropwise to an aqueous solution containing the enzyme and stirred for several hours at room temperature. During this incubation period the nanoreactors are formed by self-assembly and the solvent is evaporated. Then, the resulted dispersion is repeatedly extruded through filters of a desired pore size. This leads to monodisperse particles [163]. Lamellar film rehydration method Block copolymers and porins are mixed in solvent as described in the previous method. 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 by adding an aqueous solution containing the enzyme. Finally, extrusion produces monolamellar and monodisperse vesicles [182]. Direct dispersion method In the direct dispersion method the polymers are suspended in an aqueous solution containing the enzyme and gently stirred at room temperature overnight to form self-assembled nanocompartments. Triton X-100 is subsequently added and the suspension is sonicated twice for 10 seconds to achieve destabilization of the polymeric membrane. Porins are added shortly after the sonication step and will incorporate inside the polymeric wall of the 33 Chapter 1 nanoreactors. Biobeads are added to remove residual detergent [145]. Repeated extrusion through a polycarbonate filter of desired pore size can be used to obtain uniformly sized vesicles. In the preparation methods described above, the encapsulation efficiency is usually around 10% to 30%. Therefore the non-entrapped enzyme molecules have to be separated from the enzyme-containing vesicles in a last step of the preparation procedure. This is usually carried out either by size-exclusion chromatography (e.g. using sepharose 4B), by affinity chromatography or by dialysis. If the encapsulation efficiency or entrapment efficiency (EE) is unacceptably low, freeze-thaw cycles can be used to enhance entrapment. Also, using charged polymers that have affinity for the encapsulated compound [183] or using higher concentrations of enzyme might improve encapsulation efficiency [101]. However, in such cases removing the adhering molecules from the outside of the reactor might be more difficult. 1.4.3 Controlling the nanoreactor activity To further increase the stability of nanoreactors, the polymer membrane can be stabilized by crosslinking the polymers. To this end, Nardin et al. used methacrylated PMOXA-PDMS-PMOXA polymers [170]. These polymers carry reactive polymerizable methacrylate endgroups. After formation of the nanoreactors, the polymers were crosslinked via UV irradiation. Since every polymer molecule carries two polymerizable groups, the polymerization leads to the formation of crosslinked polymer structures. As a result, the polymerized nanocontainers possess solid state properties like shape persistence and elasticity. This polymerization did not affect the shape of the reactors nor did it modify the activity of the nanoreactors. By choosing the appropriate channel forming protein, the translocation of substrates and products across the membrane can be controlled. Aspecific porins allow transport of solutes below a size threshold while specific porins transport specific solutes as discussed above. Additionally, porins like OmpF and PhoE are voltage-gating porins, which means that they can open and close 34 Introduction depending on the transmembrane voltage potential, also called the Donnan potential. When a Donnan potential of above 100 mV is applied, the porins close [184]. This mechanism has evolved in Gram-negative bacteria to protect the cell against drastic changes in the environment and the voltage-gating characteristic can be used to switch on and off the activity of nanoreactors by external stimuli. The polymer-protein hybrid membrane can be regarded as a semipermeable membrane separating the internal volume from the external solution. Nardin et al. successfully controlled the activity of nanoreactors composed of PMOXA-PDMS-PMOXA polymers through voltage-gating of the OmpF porin [170]. To apply a Donnan potential they used sodium-polystyrenesulfonate (Na-PSS), a polyectrolyte ion. When Na-PSS is added to the external solution, a Donnan potential is created since Na-PSS cannot permeate the membrane. When the Donnan potential exceeded 100 mV, they saw a complete deactivation of the nanoreactors due to the closure of OmpF. This closure was reversed by adding NaCl, hereby decreasing the Donnan potential below 100 mV. Another way to control the activity of the nanoreactors is by using an enzyme that is switched on and off by external factors such as pH and temperature. For instance, 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) [185, 186]. By choosing an enzyme that is inactive at neutral pH but active at mild acidic conditions, the activity of the nanoreactor is restricted to the extracellular matrix of solid tumors. Broz et al. demonstrated this method by encapsulating the plant-derived enzyme acid phosphatase in OmpF permeabilized nanoreactors [187]. This enzyme is pHdependent and catalyses fast dephosphorylation of various phosphate substrates at mildly acidic pH ranging from 4-6.5 while it is inactive at highly acidic, neutral or basic pH. By varying the pH of the surrounding solution, the nanoreactor was able to change its state of activity as demonstrated by the conversion of a soluble non-fluorescent substrate into a water-insoluble fluorescent dye inside the polymeric particle [187]. They successfully created 35 Chapter 1 an environment-controlled and triggerable sensor-effector nanoreactor that is only active at mildly acidic pH. 1.4.4 Biomedical applications The delivery of enzymes can be used for a wide range of purposes. Although most applications reported up until now are applications of polymeric nanocarriers without channel proteins, we expect that nanoreactors will soon be exploited as improved drug carriers for the same diagnostic and therapeutic uses. In this chapter a snapshot of reported applications of nanoreactor-based strategies are summarized. Various polymeric drug delivery systems are currently applied or under development for their use in cancer therapy. Entrapping anti-cancer drugs inside nanoparticles, allows scientists to minimize drug degradation, to prevent undesirable side effects exercised on normal cells by cytotoxic drugs and to increase drug bioavailability [188]. The use of nanoreactors in cancer therapy makes it possible to encapsulate a prodrug-activating enzyme rather than the drug itself. In this strategy the free non-toxic prodrug is administered systemically and is converted to a toxic drug by the active nanoreactors which can be targeted to the tumor, hence decreasing toxic side effects to a minimum. The encapsulation of a prodrug-activating enzyme inside polymeric nanoreactors is a revolutionary new anti-cancer therapy introduced by Ranquin et al. [181]. This work opens the door to the encapsulation of various other prodrug-activating enzymes. Because of their cell-targeting potential, nanoreactors can also be used as diagnostic tools and targeted drug delivery systems for a wide range of diseases. For example, Broz and coworkers constructed nanoreactors that may be of particular value for the detection and treatment of vulnerable plaque macrophages in atherosclerotic diseases [189]. Current diagnostic techniques for the detection of plaque macrophages are often limited by insufficient sensitivity and selectivity. Nanoreactors with macrophage-specific delivery [190] are promising candidates for improved diagnostic strategies. In addition, 36 Introduction such nanoreactors were evaluated for targeted drug delivery. Receptor-specific targeting using statin-loaded nanometer-sized triblock copolymer vesicles with macrophage targeting moieties were constructed to rupture macrophage-rich atherosclerotic plaques [191]. The formation of such plaques in the coronary arteries is the main cause of heart attacks. Polymeric vesicles can also be used as new drug carrier systems for brain delivery. Targeting drugs to the brain by crossing the blood-brain barrier (BBB) is quite a challenge. Adequate brain delivery systems must have longcirculation properties and appropriate surface characteristics to allow interactions with BBB endothelial cells. Several studies indicated that the surface modification of particles by coating them with polysorbate-80 or polyethylene glycol is effective in brain drug delivery [192, 193]. By choosing the right surface polymer, enzyme-loaded nanoreactors could have great potential in the treatment of metabolic brain diseases, such as Lesch-Nyhan syndrome. This is a neurogenetic disorder caused by the deficiency of the purine salvage enzyme hypoxanthine-guanine phosphoribosyltransferase (HPRT). Nanoreactors can also be applied as biosensors. Although using lipids instead of polymers, Vamvakaki et al. used the nanoreactor-strategy to develop a novel nano-biosensor for pesticide analysis. The enzyme acetylcholinesterase (AChE) is encapsulated in the internal environment of liposomes and substrate entrance is facilitated by the incorporation of the porin OmpF (Figure 1-12). Within the liposomes, the pH sensitive fluorescent indicator pyranine was also immobilized for the optical transduction of the enzymatic activity. Increasing amounts of pesticides lead to a decrease of the enzymatic activity for the hydrolysis of the acetylcholine and thus to a decrease in the fluorescent signal of the pH indicator. This biosensor system has been applied successfully for the detection of two widely used organophosphorus pesticides in drinking water [194]. 37 Chapter 1 Figure 1-12: Schematic diagram of the acetylcholinesterase-based liposome biosensor for pesticide detection. Microparticle systems are also developed to produce crystals under controlled conditions. Michel and coworkers used liposomes containing alkaline phophatase as the encapsulated enzyme. This enzyme converts p-nitrophenyl phosphate into phosphate and subsequently into calcium phosphate crystals, since Ca2+ ions are already present inside the vesicles [195]. In this case the substrate diffuses through the lipidic membrane without the help of channel proteins. Reconstitution of the phosphate-induced PhoE porin might be an improvement of this reactor set-up [196]. Sauer et al. reported a quiet similar system of ion carrier-assisted precipitation of calcium phosphate in giant PMOXA-PDMS-PMOXA (Lasalocid A, vesicles. Alamethicin In and this study three different ionofores N,N-dicyclohexyl-N’,N’-dioctadecyl-3- oxapentane-1,5-diamide) were used for selective or unselective calcium transport over the polymeric membrane into the intravesicular space. Calcium phosphate crystals start to grow at the inner surface of the membrane [197]. This strategy allows the inorganic particles to grow only in restricted regions and opens the route to control crystal size in biomimetic systems. In addition, antioxidant nanoreactors based on encapsulated superoxide dismutase (SOD) were constructed [198, 199]. The polymeric membrane of the nanoreactors is capable of protecting the SOD and allowing penetration of the inner space by dioxygen, where the conversion to hydrogen peroxide and oxygen is catalyzed. In this example, the level of superoxide permeability 38 Introduction obviates the necessity of inserting channel proteins. Oxidative stress has been shown to play a significant role in many disease states including arthritis, Parkinson’s disease, cancer and AIDS. Therefore, such nanoreactors open a new direction to the development of new antioxidant drugs. The nanoreactor system can also be applied as a novel method for the construction of biologically active surfaces for analytics and sensors with space-controlled reactions. Recently, biotinylated nanoreactors were immobilized on a glass substrate. The surface of the glass substrate was first structured with bovine serum albumin labeled with biotin. Exposure of biotinylated protein patterns to streptavidin followed by incubation with biotinylated particles lead to their immobilization. The model reaction chosen for enzymatic conversion inside the nanoreactors was the dephosphorylation of the fluorogenic substrate ELF97 by acid phosphatase (Figure 1-13) [200]. Figure 1-13: Schematic representation of the immobilization of a nanoreactor system on a glass surface. Acid phosphatase was encapsulated in the OmpF bearing biotinylated vesicles and immobilized on the surface of the glass structure with streptavidin (taken from [200]). 39 Chapter 1 From the above it is clear that nanoreactors possess great potential in various biomedical applications. Nevertheless, this technology is still in the infantile stage. At the moment nanoreactors are only tested in cell cultures and to our knowledge there are no published results about the behavior of this type of polymer vesicles in a real in vivo environment. Crucial points toward future use of nanoreactors in vivo are biodistribution, biodegradability, efficacy and short-term and long-term toxicity. 1.5 References [1] R. Hirschhorn, Adenosine deaminase deficiency: molecular basis and recent developments, Clin Immunol Immunopathol, 76 (1995) S219-227. [2] R.O. Brady, R. Schiffmann, Clinical features of and recent advances in therapy for Fabry disease, Jama, 284 (2000) 2771-2775. [3] N.A. van der Beek, M.L. Hagemans, A.T. van der Ploeg, A.J. Reuser, P.A. van Doorn, Pompe disease (glycogen storage disease type II): clinical features and enzyme replacement therapy, Acta Neurol Belg, 106 (2006) 82-86. [4] E. Beutler, Gaucher disease: new molecular approaches to diagnosis and treatment, Science, 256 (1992) 794-799. [5] P. Guggenbuhl, B. Grosbois, G. Chales, Gaucher disease, Joint Bone Spine, 75 (2008) 116-124. [6] H.S. Scott, S. Bunge, A. Gal, L.A. Clarke, C.P. Morris, J.J. Hopwood, Molecular genetics of mucopolysaccharidosis type I: diagnostic, clinical, and biological implications, Hum Mutat, 6 (1995) 288-302. [7] R.A. Harkness, G.M. McCreanor, R.W. Watts, Lesch-Nyhan syndrome and its pathogenesis: purine concentrations in plasma and urine with metabolite profiles in CSF, J Inherit Metab Dis, 11 (1988) 239-252. [8] I. Nishino, A. Spinazzola, M. Hirano, Thymidine phosphorylase gene mutations in MNGIE, a human mitochondrial disorder, Science, 283 (1999) 689-692. [9] N.S. Brown, R. Bicknell, Thymidine phosphorylase, 2-deoxy-D-ribose and angiogenesis, Biochem J, 334 ( Pt 1) (1998) 1-8. [10] A. Yoshimura, Y. Kuwazuru, T. Furukawa, H. Yoshida, K. Yamada, S. Akiyama, Purification and tissue distribution of human thymidine phosphorylase; high expression in lymphocytes, reticulocytes and tumors, Biochim Biophys Acta, 1034 (1990) 107-113. [11] J. Waltenberger, K. Usuki, B. Fellstrom, K. Funa, C.H. Heldin, Plateletderived endothelial cell growth factor. Pharmacokinetics, organ distribution and degradation after intravenous administration in rats, FEBS Lett, 313 (1992) 129-132. [12] P.A. Eccleston, K. Funa, C.H. Heldin, Neurons of the peripheral nervous system express thymidine phosphorylase, Neurosci Lett, 192 (1995) 137-141. [13] S.B. Fox, A. Moghaddam, M. Westwood, H. Turley, R. Bicknell, K.C. Gatter, A.L. Harris, Platelet-derived endothelial cell growth factor/thymidine 40 Introduction phosphorylase expression in normal tissues: an immunohistochemical study, J Pathol, 176 (1995) 183-190. [14] K. Matsukawa, A. Moriyama, Y. Kawai, K. Asai, T. Kato, Tissue distribution of human gliostatin/platelet-derived endothelial cell growth factor (PD-ECGF) and its drug-induced expression, Biochim Biophys Acta, 1314 (1996) 71-82. [15] L. Griffiths, I.J. Stratford, Platelet-derived endothelial cell growth factor thymidine phosphorylase in tumour growth and response to therapy, Br J Cancer, 76 (1997) 689-693. [16] F. Focher, D. Ubiali, M. Pregnolato, C. Zhi, J. Gambino, G.E. Wright, S. Spadari, Novel nonsubstrate inhibitors of human thymidine phosphorylase, a potential target for tumor-dependent angiogenesis, J Med Chem, 43 (2000) 2601-2607. [17] N.W. Barton, R.O. Brady, J.M. Dambrosia, A.M. Di Bisceglie, S.H. Doppelt, S.C. Hill, H.J. Mankin, G.J. Murray, R.I. Parker, C.E. Argoff, et al., Replacement therapy for inherited enzyme deficiency--macrophage-targeted glucocerebrosidase for Gaucher's disease, N Engl J Med, 324 (1991) 14641470. [18] F. Ishikawa, K. Miyazono, U. Hellman, H. Drexler, C. Wernstedt, K. Hagiwara, K. Usuki, F. Takaku, W. Risau, C.H. Heldin, Identification of angiogenic activity and the cloning and expression of platelet-derived endothelial cell growth factor, Nature, 338 (1989) 557-562. [19] H. Koide, T. Asai, K. Hatanaka, S. Akai, T. Ishii, E. Kenjo, T. Ishida, H. Kiwada, H. Tsukada, N. Oku, T cell-independent B cell response is responsible for ABC phenomenon induced by repeated injection of PEGylated liposomes, Int J Pharm. [20] S. Akiyama, T. Furukawa, T. Sumizawa, Y. Takebayashi, Y. Nakajima, S. Shimaoka, M. Haraguchi, The role of thymidine phosphorylase, an angiogenic enzyme, in tumor progression, Cancer Sci, 95 (2004) 851-857. [21] A. Walter, G. Kuehl, K. Barnes, G. VanderWaerdt, The vesicle-to-micelle transition of phosphatidylcholine vesicles induced by nonionic detergents: effects of sodium chloride, sucrose and urea, Biochim Biophys Acta, 1508 (2000) 20-33. [22] M.J. Pugmire, S.E. Ealick, The crystal structure of pyrimidine nucleoside phosphorylase in a closed conformation, Structure, 6 (1998) 1467-1479. [23] R.A. Norman, S.T. Barry, M. Bate, J. Breed, J.G. Colls, R.J. Ernill, R.W. Luke, C.A. Minshull, M.S. McAlister, E.J. McCall, H.H. McMiken, D.S. Paterson, D. Timms, J.A. Tucker, R.A. Pauptit, Crystal structure of human thymidine phosphorylase in complex with a small molecule inhibitor, Structure, 12 (2004) 75-84. [24] J. Mendieta, S. Martin-Santamaria, E.M. Priego, J. Balzarini, M.J. Camarasa, M.J. Perez-Perez, F. Gago, Role of histidine-85 in the catalytic mechanism of thymidine phosphorylase as assessed by targeted molecular dynamics simulations and quantum mechanical calculations, Biochemistry, 43 (2004) 405-414. [25] M. Hirano, J. Garcia-de-Yebenes, A.C. Jones, I. Nishino, S. DiMauro, J.R. Carlo, A.N. Bender, A.F. Hahn, L.M. Salberg, D.E. Weeks, T.G. Nygaard, Mitochondrial neurogastrointestinal encephalomyopathy syndrome maps to chromosome 22q13.32-qter, Am J Hum Genet, 63 (1998) 526-533. 41 Chapter 1 [26] A.J. Berk, D.A. Clayton, A genetically distinct thymidine kinase in mammalian mitochondria. Exclusive labeling of mitochondrial deoxyribonucleic acid, J Biol Chem, 248 (1973) 2722-2729. [27] R.K. Bestwick, C.K. Mathews, Unusual compartmentation of precursors for nuclear and mitochondrial DNA in mouse L cells, J Biol Chem, 257 (1982) 9305-9308. [28] R.K. Bestwick, G.L. Moffett, C.K. Mathews, Selective expansion of mitochondrial nucleoside triphosphate pools in antimetabolite-treated HeLa cells, J Biol Chem, 257 (1982) 9300-9304. [29] P. Ferraro, G. Pontarin, L. Crocco, S. Fabris, P. Reichard, V. Bianchi, Mitochondrial deoxynucleotide pools in quiescent fibroblasts: a possible model for mitochondrial neurogastrointestinal encephalomyopathy (MNGIE), J Biol Chem, 280 (2005) 24472-24480. [30] D. Bogenhagen, D.A. Clayton, Thymidylate nucleotide supply for mitochondrial DNA synthesis in mouse L-cells. Effect of 5-fluorodeoxyuridine and methotrexate in thymidine kinase plus and thymidine kinase minus cells, J Biol Chem, 251 (1976) 2938-2944. [31] D.L. Croteau, R.H. Stierum, V.A. Bohr, Mitochondrial DNA repair pathways, Mutat Res, 434 (1999) 137-148. [32] S. DiMauro, E.A. Schon, Mitochondrial disorders in the nervous system, Annu Rev Neurosci, 31 (2008) 91-123. [33] I. Nishino, A. Spinazzola, M. Hirano, MNGIE: from nuclear DNA to mitochondrial DNA, Neuromuscul Disord, 11 (2001) 7-10. [34] Y.C. Kocaefe, S. Erdem, M. Ozguc, E. Tan, Four novel thymidine phosphorylase gene mutations in mitochondrial neurogastrointestinal encephalomyopathy syndrome (MNGIE) patients, Eur J Hum Genet, 11 (2003) 102-104. [35] G. Pontarin, P. Ferraro, M.L. Valentino, M. Hirano, P. Reichard, V. Bianchi, Mitochondrial DNA depletion and thymidine phosphate pool dynamics in a cellular model of mitochondrial neurogastrointestinal encephalomyopathy, J Biol Chem, 281 (2006) 22720-22728. [36] R. Laforce, Jr., P.N. Valdmanis, N. Dupre, G.A. Rouleau, A.F. Turgeon, M. Savard, A novel TYMP mutation in a French Canadian patient with mitochondrial neurogastrointestinal encephalomyopathy, Clin Neurol Neurosurg, 111 (2009) 691-694. [37] J. Gamez, M.C. Lara, F. Mearin, C. Oliveras-Ley, N. Raguer, M. Olive, A.T. Leist, A. Perello, M. Perona, C. Cervera, A.L. Andreu, R. Marti, M. Hirano, A novel thymidine phosphorylase mutation in a Spanish MNGIE patient, J Neurol Sci, 228 (2005) 35-39. [38] G. Phear, J. Nalbantoglu, M. Meuth, Next-nucleotide effects in mutations driven by DNA precursor pool imbalances at the aprt locus of Chinese hamster ovary cells, Proc Natl Acad Sci U S A, 84 (1987) 4450-4454. [39] E.C. Butcher, E.L. Berg, E.J. Kunkel, Systems biology in drug discovery, Nat Biotechnol, 22 (2004) 1253-1259. [40] P.J. Wilderman, A.I. Vasil, Z. Johnson, M.L. Vasil, Genetic and biochemical analyses of a eukaryotic-like phospholipase D of Pseudomonas aeruginosa suggest horizontal acquisition and a role for persistence in a chronic pulmonary infection model, Mol Microbiol, 39 (2001) 291-303. 42 Introduction [41] S. Ferraris, S. Clark, E. Garelli, G. Davidzon, S.A. Moore, R.H. Kardon, R.J. Bienstock, M.J. Longley, M. Mancuso, P. Gutierrez Rios, M. Hirano, W.C. Copeland, S. DiMauro, Progressive external ophthalmoplegia and vision and hearing loss in a patient with mutations in POLG2 and OPA1, Arch Neurol, 65 (2008) 125-131. [42] Y. Nishigaki, R. Marti, W.C. Copeland, M. Hirano, Site-specific somatic mitochondrial DNA point mutations in patients with thymidine phosphorylase deficiency, J Clin Invest, 111 (2003) 1913-1921. [43] S.E. Lim, W.C. Copeland, Differential incorporation and removal of antiviral deoxynucleotides by human DNA polymerase gamma, J Biol Chem, 276 (2001) 23616-23623. [44] Y. Nishigaki, R. Marti, M. Hirano, ND5 is a hot-spot for multiple atypical mitochondrial DNA deletions in mitochondrial neurogastrointestinal encephalomyopathy, Hum Mol Genet, 13 (2004) 91-101. [45] M. Hirano, G. Silvestri, D.M. Blake, A. Lombes, C. Minetti, E. Bonilla, A.P. Hays, R.E. Lovelace, I. Butler, T.E. Bertorini, et al., Mitochondrial neurogastrointestinal encephalomyopathy (MNGIE): clinical, biochemical, and genetic features of an autosomal recessive mitochondrial disorder, Neurology, 44 (1994) 721-727. [46] M.C. Lara, M.L. Valentino, J. Torres-Torronteras, M. Hirano, R. Marti, Mitochondrial neurogastrointestinal encephalomyopathy (MNGIE): biochemical features and therapeutic approaches, Biosci Rep, 27 (2007) 151-163. [47] M. Hirano, Y. Nishigaki, R. Marti, Mitochondrial neurogastrointestinal encephalomyopathy (MNGIE): a disease of two genomes, Neurologist, 10 (2004) 8-17. [48] A. Spinazzola, R. Marti, I. Nishino, A.L. Andreu, A. Naini, S. Tadesse, I. Pela, E. Zammarchi, M.A. Donati, J.A. Oliver, M. Hirano, Altered thymidine metabolism due to defects of thymidine phosphorylase, J Biol Chem, 277 (2002) 4128-4133. [49] R. Marti, A. Spinazzola, S. Tadesse, I. Nishino, Y. Nishigaki, M. Hirano, Definitive diagnosis of mitochondrial neurogastrointestinal encephalomyopathy by biochemical assays, Clin Chem, 50 (2004) 120-124. [50] R. Marti, J.J. Verschuuren, A. Buchman, I. Hirano, S. Tadesse, A.B. van Kuilenburg, A.H. van Gennip, B.J. Poorthuis, M. Hirano, Late-onset MNGIE due to partial loss of thymidine phosphorylase activity, Ann Neurol, 58 (2005) 649652. [51] D.L. Dorset, A. Engel, M. Haner, A. Massalski, J.P. Rosenbusch, Twodimensional crystal packing of matrix porin. A channel forming protein in Escherichia coli outer membranes, J Mol Biol, 165 (1983) 701-710. [52] L.C. Lopez, H.O. Akman, A. Garcia-Cazorla, B. Dorado, R. Marti, I. Nishino, S. Tadesse, G. Pizzorno, D. Shungu, E. Bonilla, K. Tanji, M. Hirano, Unbalanced deoxynucleotide pools cause mitochondrial DNA instability in thymidine phosphorylase-deficient mice, Hum Mol Genet, 18 (2009) 714-722. [53] D. Cao, J.J. Leffert, J. McCabe, B. Kim, G. Pizzorno, Abnormalities in uridine homeostatic regulation and pyrimidine nucleotide metabolism as a consequence of the deletion of the uridine phosphorylase gene, J Biol Chem, 280 (2005) 21169-21175. [54] M.L. Valentino, R. Marti, S. Tadesse, L.C. Lopez, J.L. Manes, J. Lyzak, A. Hahn, V. Carelli, M. Hirano, Thymidine and deoxyuridine accumulate in tissues 43 Chapter 1 of patients with mitochondrial neurogastrointestinal encephalomyopathy (MNGIE), FEBS Lett, 581 (2007) 3410-3414. [55] P.F. Chinnery, J. Vissing, Treating MNGIE: is reducing blood nucleosides the first cure for a mitochondrial disorder?, Neurology, 67 (2006) 1330-1332. [56] G. la Marca, S. Malvagia, B. Casetta, E. Pasquini, I. Pela, M. Hirano, M.A. Donati, E. Zammarchi, Pre- and post-dialysis quantitative dosage of thymidine in urine and plasma of a MNGIE patient by using HPLC-ESI-MS/MS, J Mass Spectrom, 41 (2006) 586-592. [57] C. Desgranges, G. Razaka, M. Rabaud, P. Picard, F. Dupuch, H. Bricaud, The human blood platelet: a cellular model to study the degradation of thymidine and its inhibition, Biochem Pharmacol, 31 (1982) 2755-2759. [58] M.C. Lara, B. Weiss, I. Illa, P. Madoz, L. Massuet, A.L. Andreu, M.L. Valentino, Y. Anikster, M. Hirano, R. Marti, Infusion of platelets transiently reduces nucleoside overload in MNGIE, Neurology, 67 (2006) 1461-1463. [59] M. Hirano, R. Marti, C. Casali, S. Tadesse, T. Uldrick, B. Fine, D.M. Escolar, M.L. Valentino, I. Nishino, C. Hesdorffer, J. Schwartz, R.G. Hawks, D.L. Martone, M.S. Cairo, S. DiMauro, M. Stanzani, J.H. Garvin, Jr., D.G. Savage, Allogeneic stem cell transplantation corrects biochemical derangements in MNGIE, Neurology, 67 (2006) 1458-1460. [60] B.E. Bax, M.D. Bain, L.D. Fairbanks, A.D. Webster, P.W. Ind, M.S. Hershfield, R.A. Chalmers, A 9-yr evaluation of carrier erythrocyte encapsulated adenosine deaminase (ADA) therapy in a patient with adult-type ADA deficiency, Eur J Haematol, 79 (2007) 338-348. [61] N.F. Moran, M.D. Bain, M.M. Muqit, B.E. Bax, Carrier erythrocyte entrapped thymidine phosphorylase therapy for MNGIE, Neurology, 71 (2008) 686-688. [62] R.O. Brady, R. Schiffmann, Enzyme-replacement therapy for metabolic storage disorders, Lancet Neurol, 3 (2004) 752-756. [63] J.B. Rose, I.R. Coe, Physiology of nucleoside transporters: back to the future, Physiology (Bethesda), 23 (2008) 41-48. [64] E. Kakkis, T. Lester, R. Yang, C. Tanaka, V. Anand, J. Lemontt, M. Peinovich, M. Passage, Successful induction of immune tolerance to enzyme replacement therapy in canine mucopolysaccharidosis I, Proc Natl Acad Sci U S A, 101 (2004) 829-834. [65] N.W. Barton, F.S. Furbish, G.J. Murray, M. Garfield, R.O. Brady, Therapeutic response to intravenous infusions of glucocerebrosidase in a patient with Gaucher disease, Proc Natl Acad Sci U S A, 87 (1990) 1913-1916. [66] W.R. Wilcox, M. Banikazemi, N. Guffon, S. Waldek, P. Lee, G.E. Linthorst, R.J. Desnick, D.P. Germain, Long-term safety and efficacy of enzyme replacement therapy for Fabry disease, Am J Hum Genet, 75 (2004) 65-74. [67] E.D. Kakkis, J. Muenzer, G.E. Tiller, L. Waber, J. Belmont, M. Passage, B. Izykowski, J. Phillips, R. Doroshow, I. Walot, R. Hoft, E.F. Neufeld, Enzymereplacement therapy in mucopolysaccharidosis I, N Engl J Med, 344 (2001) 182-188. [68] A. Amalfitano, A.R. Bengur, R.P. Morse, J.M. Majure, L.E. Case, D.L. Veerling, J. Mackey, P. Kishnani, W. Smith, A. McVie-Wylie, J.A. Sullivan, G.E. Hoganson, J.A. Phillips, 3rd, G.B. Schaefer, J. Charrow, R.E. Ware, E.H. Bossen, Y.T. Chen, Recombinant human acid alpha-glucosidase enzyme 44 Introduction therapy for infantile glycogen storage disease type II: results of a phase I/II clinical trial, Genet Med, 3 (2001) 132-138. [69] M.A. Hamdan, M.H. Almalik, H.M. Mirghani, Early administration of enzyme replacement therapy for Pompe disease: Short-term follow-up results, J Inherit Metab Dis, (2008). [70] M. Beck, New therapeutic options for lysosomal storage disorders: enzyme replacement, small molecules and gene therapy, Hum Genet, 121 (2007) 1-22. [71] T.A. Burrow, R.J. Hopkin, N.D. Leslie, B.T. Tinkle, G.A. Grabowski, Enzyme reconstitution/replacement therapy for lysosomal storage diseases, Curr Opin Pediatr, 19 (2007) 628-635. [72] J.E. Wraith, Limitations of enzyme replacement therapy: current and future, J Inherit Metab Dis, 29 (2006) 442-447. [73] G. Walsh, Biopharmaceutical benchmarks--2003, Nat Biotechnol, 21 (2003) 865-870. [74] H. Schellekens, Immunogenicity of therapeutic proteins: clinical implications and future prospects, Clin Ther, 24 (2002) 1720-1740; discussion 1719. [75] F.M. Veronese, G. Pasut, PEGylation, successful approach to drug delivery, Drug Discov Today, 10 (2005) 1451-1458. [76] J.M. Harris, R.B. Chess, Effect of pegylation on pharmaceuticals, Nat Rev Drug Discov, 2 (2003) 214-221. [77] M.J. Roberts, M.D. Bentley, J.M. Harris, Chemistry for peptide and protein PEGylation, Adv Drug Deliv Rev, 54 (2002) 459-476. [78] P. Caliceti, F.M. Veronese, Pharmacokinetic and biodistribution properties of poly(ethylene glycol)-protein conjugates, Adv Drug Deliv Rev, 55 (2003) 1261-1277. [79] V. Gaberc-Porekar, I. Zore, B. Podobnik, V. Menart, Obstacles and pitfalls in the PEGylation of therapeutic proteins, Curr Opin Drug Discov Devel, 11 (2008) 242-250. [80] X. Wang, T. Ishida, H. Kiwada, Anti-PEG IgM elicited by injection of liposomes is involved in the enhanced blood clearance of a subsequent dose of PEGylated liposomes, J Control Release, 119 (2007) 236-244. [81] T. Ishida, X. Wang, T. Shimizu, K. Nawata, H. Kiwada, PEGylated liposomes elicit an anti-PEG IgM response in a T cell-independent manner, J Control Release, 122 (2007) 349-355. [82] Y. Levy, M.S. Hershfield, C. Fernandez-Mejia, S.H. Polmar, D. Scudiery, M. Berger, R.U. Sorensen, Adenosine deaminase deficiency with late onset of recurrent infections: response to treatment with polyethylene glycol-modified adenosine deaminase, J Pediatr, 113 (1988) 312-317. [83] M.L. Graham, Pegaspargase: a review of clinical studies, Adv Drug Deliv Rev, 55 (2003) 1293-1302. [84] Y.S. Wang, S. Youngster, M. Grace, J. Bausch, R. Bordens, D.F. Wyss, Structural and biological characterization of pegylated recombinant interferon alpha-2b and its therapeutic implications, Adv Drug Deliv Rev, 54 (2002) 547570. [85] P. Bailon, A. Palleroni, C.A. Schaffer, C.L. Spence, W.J. Fung, J.E. Porter, G.K. Ehrlich, W. Pan, Z.X. Xu, M.W. Modi, A. Farid, W. Berthold, M. Graves, Rational design of a potent, long-lasting form of interferon: a 40 kDa branched 45 Chapter 1 polyethylene glycol-conjugated interferon alpha-2a for the treatment of hepatitis C, Bioconjug Chem, 12 (2001) 195-202. [86] O.B. Kinstler, D.N. Brems, S.L. Lauren, A.G. Paige, J.B. Hamburger, M.J. Treuheit, Characterization and stability of N-terminally PEGylated rhG-CSF, Pharm Res, 13 (1996) 996-1002. [87] P.J. Trainer, W.M. Drake, L. Katznelson, P.U. Freda, V. Herman-Bonert, A.J. van der Lely, E.V. Dimaraki, P.M. Stewart, K.E. Friend, M.L. Vance, G.M. Besser, J.A. Scarlett, M.O. Thorner, C. Parkinson, A. Klibanski, J.S. Powell, A.L. Barkan, M.C. Sheppard, M. Malsonado, D.R. Rose, D.R. Clemmons, G. Johannsson, B.A. Bengtsson, S. Stavrou, D.L. Kleinberg, D.M. Cook, L.S. Phillips, M. Bidlingmaier, C.J. Strasburger, S. Hackett, K. Zib, W.F. Bennett, R.J. Davis, Treatment of acromegaly with the growth hormone-receptor antagonist pegvisomant, N Engl J Med, 342 (2000) 1171-1177. [88] S.D. Caruthers, S.A. Wickline, G.M. Lanza, Nanotechnological applications in medicine, Curr Opin Biotechnol, 18 (2007) 26-30. [89] S. Sandhiya, S.A. Dkhar, A. Surendiran, Emerging trends of nanomedicine: an overview, Fundam Clin Pharmacol, 23 (2009) 263-269. [90] Y. Shi, L.C. Li, Current advances in sustained-release systems for parenteral drug delivery, Expert Opin Drug Deliv, 2 (2005) 1039-1058. [91] G. Gregoriadis, Drug entrapment in liposomes, FEBS Lett, 36 (1973) 292296. [92] C.J. Chu, J. Dijkstra, M.Z. Lai, K. Hong, F.C. Szoka, Efficiency of cytoplasmic delivery by pH-sensitive liposomes to cells in culture, Pharm Res, 7 (1990) 824-834. [93] K. Kono, Thermosensitive polymer-modified liposomes, Adv Drug Deliv Rev, 53 (2001) 307-319. [94] R.H. Bisby, C. Mead, C.G. Morgan, Active uptake of drugs into photosensitive liposomes and rapid release on UV photolysis, Photochem Photobiol, 72 (2000) 57-61. [95] P. Goyal, K. Goyal, S.G. Vijaya Kumar, A. Singh, O.P. Katare, D.N. Mishra, Liposomal drug delivery systems--clinical applications, Acta Pharm, 55 (2005) 1-25. [96] M.L. Immordino, F. Dosio, L. Cattel, Stealth liposomes: review of the basic science, rationale, and clinical applications, existing and potential, Int J Nanomedicine, 1 (2006) 297-315. [97] P.R. Veerareddy, V. Vobalaboina, Lipid-based formulations of amphotericin B, Drugs Today (Barc), 40 (2004) 133-145. [98] D.S. Alberts, F.M. Muggia, J. Carmichael, E.P. Winer, M. Jahanzeb, A.P. Venook, K.M. Skubitz, E. Rivera, J.A. Sparano, N.J. DiBella, S.J. Stewart, J.J. Kavanagh, A.A. Gabizon, Efficacy and safety of liposomal anthracyclines in phase I/II clinical trials, Semin Oncol, 31 (2004) 53-90. [99] T.M. Allen, F.J. Martin, Advantages of liposomal delivery systems for anthracyclines, Semin Oncol, 31 (2004) 5-15. [100] G. Gregoriadis, Enzyme entrapment in liposomes, Methods Enzymol, 44 (1976) 218-227. [101] P. Walde, S. Ichikawa, Enzymes inside lipid vesicles: preparation, reactivity and applications, Biomol Eng, 18 (2001) 143-177. [102] G. Gregoriadis, P.D. Leathwood, B.E. Ryman, Enzyme entrapment in liposomes, FEBS Lett, 14 (1971) 95-99. 46 Introduction [103] C.M. Colley, B.E. Ryman, The use of a liposomally entrapped enzyme in the treatment of an artificial storage condition, Biochim Biophys Acta, 451 (1976) 417-425. [104] G.C. Reynolds, H.J. Baker, R.H. Reynolds, Enzyme replacement using liposome carriers in feline Gm1 gangliosidosis fibroblasts, Nature, 275 (1978) 754-755. [105] P.E. Belchetz, J.C. Crawley, I.P. Braidman, G. Gregoriadis, Treatment of Gaucher's disease with liposome-entrapped glucocerebroside: betaglucosidase, Lancet, 2 (1977) 116-117. [106] L.D. Steger, R.J. Desnick, Enzyme therapy. VI: Comparative in vivo fates and effects on lysosomal integrity of enzyme entrapped in negatively and positively charged liposomes, Biochim Biophys Acta, 464 (1977) 530-546. [107] T.M. Allen, C. Hansen, J. Rutledge, Liposomes with prolonged circulation times: factors affecting uptake by reticuloendothelial and other tissues, Biochim Biophys Acta, 981 (1989) 27-35. [108] T. Ishida, H. Harashima, H. Kiwada, Liposome clearance, Biosci Rep, 22 (2002) 197-224. [109] Y.J. Kao, R.L. Juliano, Interactions of liposomes with the reticuloendothelial system. Effects of reticuloendothelial blockade on the clearance of large unilamellar vesicles, Biochim Biophys Acta, 677 (1981) 453461. [110] J.H. Senior, Fate and behavior of liposomes in vivo: a review of controlling factors, Crit Rev Ther Drug Carrier Syst, 3 (1987) 123-193. [111] J. Szebeni, The interaction of liposomes with the complement system, Crit Rev Ther Drug Carrier Syst, 15 (1998) 57-88. [112] T.M. Allen, A study of phospholipid interactions between high-density lipoproteins and small unilamellar vesicles, Biochim Biophys Acta, 640 (1981) 385-397. [113] A.L. Klibanov, K. Maruyama, V.P. Torchilin, L. Huang, Amphipathic polyethyleneglycols effectively prolong the circulation time of liposomes, FEBS Lett., 268 (1990) 235-237. [114] G. Blume, G. Cevc, Molecular mechanism of the lipid vesicle longevity in vivo, Biochim Biophys Acta, 1146 (1993) 157-168. [115] D. Papahadjopoulos, T.M. Allen, A. Gabizon, E. Mayhew, K. Matthay, S.K. Huang, K.D. Lee, M.C. Woodle, D.D. Lasic, C. Redemann, et al., Sterically stabilized liposomes: improvements in pharmacokinetics and antitumor therapeutic efficacy, Proc Natl Acad Sci U S A, 88 (1991) 11460-11464. [116] A. Gabizon, D. Papahadjopoulos, Liposome formulations with prolonged circulation time in blood and enhanced uptake by tumors, Proc Natl Acad Sci U S A, 85 (1988) 6949-6953. [117] L. Cattel, M. Ceruti, F. Dosio, From conventional to stealth liposomes: a new frontier in cancer chemotherapy, Tumori, 89 (2003) 237-249. [118] A.A. Gabizon, Y. Barenholz, M. Bialer, Prolongation of the circulation time of doxorubicin encapsulated in liposomes containing a polyethylene glycol-derivatized phospholipid: pharmacokinetic studies in rodents and dogs, Pharm Res, 10 (1993) 703-708. [119] T. Boulikas, G.P. Stathopoulos, N. Volakakis, M. Vougiouka, Systemic Lipoplatin infusion results in preferential tumor uptake in human studies, Anticancer Res, 25 (2005) 3031-3039. 47 Chapter 1 [120] E.S. Kim, C. Lu, F.R. Khuri, M. Tonda, B.S. Glisson, D. Liu, M. Jung, W.K. Hong, R.S. Herbst, A phase II study of STEALTH cisplatin (SPI-77) in patients with advanced non-small cell lung cancer, Lung Cancer, 34 (2001) 427-432. [121] T.M. Allen, I. Ahmad, D.E. Lopes de Menezes, E.H. Moase, Immunoliposome-mediated targeting of anti-cancer drugs in vivo, Biochem Soc Trans, 23 (1995) 1073-1079. [122] O.P. Medina, Y. Zhu, K. Kairemo, Targeted liposomal drug delivery in cancer, Curr Pharm Des, 10 (2004) 2981-2989. [123] P. Sapra, T.M. Allen, Ligand-targeted liposomal anticancer drugs, Prog Lipid Res, 42 (2003) 439-462. [124] K. Maruyama, O. Ishida, T. Takizawa, K. Moribe, Possibility of active targeting to tumor tissues with liposomes, Adv Drug Deliv Rev, 40 (1999) 89102. [125] V.P. Torchilin, T.S. Levchenko, K.R. Whiteman, A.A. Yaroslavov, A.M. Tsatsakis, A.K. Rizos, E.V. Michailova, M.I. Shtilman, Amphiphilic poly-Nvinylpyrrolidones: synthesis, properties and liposome surface modification, Biomaterials, 22 (2001) 3035-3044. [126] X. Mu, Z. Zhong, Preparation and properties of poly(vinyl alcohol)stabilized liposomes, Int J Pharm, 318 (2006) 55-61. [127] V.P. Torchilin, M.I. Shtilman, V.S. Trubetskoy, K. Whiteman, A.M. Milstein, Amphiphilic vinyl polymers effectively prolong liposome circulation time in vivo, Biochim Biophys Acta, 1195 (1994) 181-184. [128] M.C. Woodle, C.M. Engbers, S. Zalipsky, New amphipatic polymer-lipid conjugates forming long-circulating reticuloendothelial system-evading liposomes, Bioconjug Chem, 5 (1994) 493-496. [129] M.C. Woodle, Controlling liposome blood clearance by surface-grafted polymers, Adv Drug Deliv Rev, 32 (1998) 139-152. [130] C. Kirby, G. Gregoriadis, Plasma-induced release of solutes from small unilamellar liposomes is associated with pore formation in the bilayers, Biochem J, 199 (1981) 251-254. [131] A. Sharma, U.S. Sharma, Liposomes in drug delivery: progress and limitations, International Journal of Pharmaceutics, 154 (1997) 123-140. [132] T. Ishida, R. Maeda, M. Ichihara, K. Irimura, H. Kiwada, Accelerated clearance of PEGylated liposomes in rats after repeated injections, J Control Release, 88 (2003) 35-42. [133] T. Ishida, M. Harada, X.Y. Wang, M. Ichihara, K. Irimura, H. Kiwada, Accelerated blood clearance of PEGylated liposomes following preceding liposome injection: effects of lipid dose and PEG surface-density and chain length of the first-dose liposomes, J Control Release, 105 (2005) 305-317. [134] P. Laverman, M.G. Carstens, O.C. Boerman, E.T. Dams, W.J. Oyen, N. van Rooijen, F.H. Corstens, G. Storm, Factors affecting the accelerated blood clearance of polyethylene glycol-liposomes upon repeated injection, J Pharmacol Exp Ther, 298 (2001) 607-612. [135] N. Dos Santos, C. Allen, A.M. Doppen, M. Anantha, K.A. Cox, R.C. Gallagher, G. Karlsson, K. Edwards, G. Kenner, L. Samuels, M.S. Webb, M.B. Bally, Influence of poly(ethylene glycol) grafting density and polymer length on liposomes: relating plasma circulation lifetimes to protein binding, Biochim Biophys Acta, 1768 (2007) 1367-1377. 48 Introduction [136] S.M. Moghimi, J. Szebeni, Stealth liposomes and long circulating nanoparticles: critical issues in pharmacokinetics, opsonization and proteinbinding properties, Prog Lipid Res, 42 (2003) 463-478. [137] D.E. Discher, A. Eisenberg, Polymer vesicles, Science., 297 (2002) 967973. [138] R. Stoenescu, A. Graff, W. Meier, Asymmetric ABC-Triblock Copolymer Membranes Induce a Directed Insertion of Membrane Proteins, Macromol Biosci, 4 (2004) 930-935. [139] A. Taubert, A. Napoli, W. Meier, Self-assembly of reactive amphiphilic block copolymers as mimetics for biological membranes, Curr Opin Chem Biol, 8 (2004) 598-603. [140] W. Meier, Polymer nanocapsules, Chem. Soc. Rev., 29 (2000) 295-303. [141] G. Battaglia, A.J. Ryan, Bilayers and interdigitation in block copolymer vesicles, J Am Chem Soc., 127 (2005) 8757-8764. [142] D.E. Discher, Emerging applications of polymersomes in delivery: From molecular dynamics to shrinkage of tumors, Progress in Polymer Science, 32 (2007) 838-857. [143] J. Panyam, V. Labhasetwar, Biodegradable nanoparticles for drug and gene delivery to cells and tissue, Adv Drug Deliv Rev, 55 (2003) 329-347. [144] L.G. Ding J, Water-Soluble Hollow Nanospheres as Potential Drug Carriers, J. Phys. Chem. B, 102 (1998) 6107-6113. [145] T. Ruysschaert, A.F. Sonnen, T. Haefele, W. Meier, M. Winterhalter, D. Fournier, Hybrid nanocapsules: interactions of ABA block copolymers with liposomes, J Am Chem Soc, 127 (2005) 6242-6247. [146] O. Onaca, R. Enea, D.W. Hughes, W. Meier, Stimuli-responsive polymersomes as nanocarriers for drug and gene delivery, Macromol Biosci, 9 (2009) 129-139. [147] P.J. Photos, L. Bacakova, B. Discher, F.S. Bates, D.E. Discher, Polymer vesicles in vivo: correlations with PEG molecular weight, J Control Release, 90 (2003) 323-334. [148] L. Hasadsri, J. Kreuter, H. Hattori, T. Iwasaki, J.M. George, Functional protein delivery into neurons using polymeric nanoparticles, J Biol Chem, 284 (2009) 6972-6981. [149] M.K. Reddy, L. Wu, W. Kou, A. Ghorpade, V. Labhasetwar, Superoxide dismutase-loaded PLGA nanoparticles protect cultured human neurons under oxidative stress, Appl Biochem Biotechnol, 151 (2008) 565-577. [150] I. Genta, P. Perugini, F. Pavanetto, K. Maculotti, T. Modena, B. Casado, A. Lupi, P. Iadarola, B. Conti, Enzyme loaded biodegradable microspheres in vitro ex vivo evaluation, J Control Release., 77 (2001) 287-295. [151] C. Colonna, B. Conti, P. Perugini, F. Pavanetto, T. Modena, R. Dorati, P. Iadarola, I. Genta, Ex vivo evaluation of prolidase loaded chitosan nanoparticles for the enzyme replacement therapy, Eur J Pharm Biopharm, 70 (2008) 58-65. [152] C. Garnacho, R. Dhami, E. Simone, T. Dziubla, J. Leferovich, E.H. Schuchman, V. Muzykantov, S. Muro, Delivery of acid sphingomyelinase in normal and niemann-pick disease mice using intercellular adhesion molecule1-targeted polymer nanocarriers, J Pharmacol Exp Ther, 325 (2008) 400-408. 49 Chapter 1 [153] T.D. Dziubla, A. Karim, V.R. Muzykantov, Polymer nanocarriers protecting active enzyme cargo against proteolysis, J Control Release, 102 (2005) 427-439. [154] F. Chen, Z.R. Zhang, F. Yuan, X. Qin, M. Wang, Y. Huang, In vitro and in vivo study of N-trimethyl chitosan nanoparticles for oral protein delivery, Int J Pharm, 349 (2008) 226-233. [155] S.M. Standley, I. Mende, S.L. Goh, Y.J. Kwon, T.T. Beaudette, E.G. Engleman, J.M. Frechet, Incorporation of CpG oligonucleotide ligand into protein-loaded particle vaccines promotes antigen-specific CD8 T-cell immunity, Bioconjug Chem, 18 (2007) 77-83. [156] F. Ahmed, R.I. Pakunlu, A. Brannan, F. Bates, T. Minko, D.E. Discher, Biodegradable polymersomes loaded with both paclitaxel and doxorubicin permeate and shrink tumors, inducing apoptosis in proportion to accumulated drug, J Control Release., 116 (2006) 150-158. Epub 2006 Jul 2020. [157] M.K. Reddy, V. Labhasetwar, Nanoparticle-mediated delivery of superoxide dismutase to the brain: an effective strategy to reduce ischemiareperfusion injury, Faseb J, 23 (2009) 1384-1395. [158] R. Tong, V.J. Coyle, L. Tang, A.M. Barger, T.M. Fan, J. Cheng, Polylactide nanoparticles containing stably incorporated cyanine dyes for in vitro and in vivo imaging applications, Microsc Res Tech. [159] A. Napoli, M.J. Boerakker, N. Tirelli, R.J. Nolte, N.A. Sommerdijk, J.A. Hubbell, Glucose-oxidase based self-destructing polymeric vesicles, Langmuir, 20 (2004) 3487-3491. [160] M. Nasseau, Y. Boublik, W. Meier, M. Winterhalter, D. Fournier, Substrate-permeable encapsulation of enzymes maintains effective activity, stabilizes against denaturation, and protects against proteolytic degradation, Biotechnol Bioeng, 75 (2001) 615-618. [161] G. Huysmans, A. Ranquin, L. Wyns, J. Steyaert, P. Van Gelder, Encapsulation of therapeutic nucleoside hydrolase in functionalised nanocapsules, J Control Release, 102 (2005) 171-179. [162] M. Yoshimoto, S. Wang, K. Fukunaga, D. Fournier, P. Walde, R. Kuboi, K. Nakao, Novel immobilized liposomal glucose oxidase system using the channel protein OmpF and catalase, Biotechnol Bioeng, 90 (2005) 231-238. [163] C. Nardin, Hirt, T., Leukel, J., Meier, W., Polymerized ABA Triblock Copolymer Vesicles, Langmuir, 16 (2000) 1035-1041. [164] C. Nardin, D. Bolikal, J. Kohn, Nontoxic block copolymer nanospheres: design and characterization, Langmuir., 20 (2004) 11721-11725. [165] W. Meier, C. Nardin, M. Winterhalter, Reconstitution of Channel Proteins in (Polymerized) ABA Triblock Copolymer Membranes Angew Chem Int Ed Engl., 39 (2000) 4599-4602. [166] V. Pata, N. Dan, The effect of chain length on protein solubilization in polymer-based vesicles (polymersomes), Biophys J, 85 (2003) 2111-2118. [167] D.D. Vijayan K, Lal J, Janmey P, Goulian M, Interactions of MembraneActive Peptides with Thick, Neutral, Nonzwitterionic Bilayers, J. Phys. Chem. B, 109 (2005) 14356-14364. [168] R. Koebnik, K.P. Locher, P. Van Gelder, Structure and function of bacterial outer membrane proteins: barrels in a nutshell Mol Microbiol, 37 (2000) 239-253. 50 Introduction [169] A. Prilipov, P.S. Phale, P. Van Gelder, J.P. Rosenbusch, R. Koebnik, Coupling site-directed mutagenesis with high-level expression: large scale production of mutant porins from E. coli, FEMS Microbiol Lett, 163 (1998) 6572. [170] C. Nardin, J. Widmer, M. Winterhalter, W. Meier, Amphiphilic block copolymer nanocontainers as bioreactors, Eur. Phys. J. E, 4 (2001) 403-410. [171] M. Winterhalter, Hilty, C., Bezrukov, S.M., Nardin, C., Meier, W., Fournier, D., Controlling membrane permeability with bacterial porins: applications to encapsulated enzymes, Talanta, 55 (2001) 965-971. [172] A. Graff, M. Sauer, P. Van Gelder, W. Meier, Virus-assisted loading of polymer nanocontainer, Proc Natl Acad Sci U S A, 99 (2002) 5064-5068. Epub 2002 Mar 5026. [173] C. Maier, E. Bremer, A. Schmid, R. Benz, 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 (1988) 2493-2499. [174] J. Ye, B. Van Den Berg, Crystal structure of the bacterial nucleoside transporter Tsx, EMBO J, 23 (2004) 3187-3195. Epub 2004 Jul 3122. [175] R. Benz, A. Schmid, C. Maier, E. Bremer, Characterization of the nucleoside-binding site inside the Tsx channel of Escherichia coli outer membrane, Eur J Biochem, 176 (1988) 699-705. [176] M. Nallani, S. Benito, O. Onaca, A. Graff, M. Lindemann, M. Winterhalter, W. Meier, U. Schwaneberg, A nanocompartment system (Synthosome) designed for biotechnological applications, J Biotechnol, 123 (2006) 50-59. Epub 2005 Dec 2020. [177] M. Kumar, M. Grzelakowski, J. Zilles, M. Clark, W. Meier, Highly permeable polymeric membranes based on the incorporation of the functional water channel protein Aquaporin Z, Proc Natl Acad Sci U S A, 104 (2007) 20719-20724. [178] C. Nardin, S. Thoeni, J. Widmer, M. Winterhalter, W. Meier, Nanoreactors based on (polymerized) ABA-triblock copolymer vesicles, Chem Comm, (2000) 1433-1434. [179] C. Nardin, W. Meier, Polymerizable amphiphilic block copolymers: From nanostructured hydrogels to nanoreactors and ultrathin films, Chimia, 55 (2001) 142-146. [180] C. Nardin, W. Meier, Hybrid materials from amphiphilic block copolymers and membrane proteins, J Biotechnol, 90 (2002) 17-26. [181] A. Ranquin, W. Versees, W. Meier, J. Steyaert, P. Van Gelder, Therapeutic Nanoreactors: Combining Chemistry and Biology in a Novel Triblock Copolymer Drug Delivery System, Nano Lett, 5 (2005) 2220-2224. [182] T. Ruysschaert, M. Germain, J.F. Gomes, D. Fournier, G.B. Sukhorukov, W. Meier, M. Winterhalter, Liposome-based nanocapsules, IEEE Trans Nanobioscience, 3 (2004) 49-55. [183] J.P. Colletier, B. Chaize, M. Winterhalter, D. Fournier, Protein encapsulation in liposomes: efficiency depends on interactions between protein and phospholipid bilayer, BMC Biotechnol, 2 (2002) 9. [184] H. Samartzidou, A.H. Delcour, E.coli PhoE porin has an opposite voltagedependence to the homologous OmpF, Embo J, 17 (1998) 93-100. [185] I.F. Tannock, D. Rotin, Acid pH in tumors and its potential for therapeutic exploitation, Cancer Res, 49 (1989) 4373-4384. 51 Chapter 1 [186] M. Stubbs, P.M. McSheehy, J.R. Griffiths, Causes and consequences of acidic pH in tumors: a magnetic resonance study, 13-30. [187] P. Broz, S. Driamov, J. Ziegler, N. Ben-Haim, S. Marsch, W. Meier, P. Hunziker, Toward Intelligent Nanosize Bioreactors: A pH-Switchable, ChannelEquipped, Functional Polymer Nanocontainer, Nano Lett, 6 (2006) 2349-2353. [188] V.P. Torchilin, Micellar nanocarriers: pharmaceutical perspectives, Pharm Res., 24 (2007) 1-16. Epub 2006 Nov 2016. [189] P. Broz, S. Marsch, P. Hunziker, Targeting of vulnerable plaque macrophages with polymer-based nanostructures, Trends Cardiovasc Med., 17 (2007) 190-196. [190] P. Broz, S.M. Benito, C. Saw, P. Burger, H. Heider, M. Pfisterer, S. Marsch, W. Meier, P. Hunziker, Cell targeting by a generic receptor-targeted polymer nanocontainer platform, J Control Release, 102 (2005) 475-488. [191] P. Broz, N. Ben-Haim, M. Grzelakowski, S. Marsch, W. Meier, P. Hunziker, Inhibition of Macrophage Phagocytotic Activity by a Receptortargeted Polymer Vesicle-based Drug Delivery Formulation of Pravastatin, J Cardiovasc Pharmacol, 51 (2008) 246-252. [192] P. Calvo, B. Gouritin, H. Chacun, D. Desmaele, J. D'Angelo, J.P. Noel, D. Georgin, E. Fattal, J.P. Andreux, P. Couvreur, Long-circulating PEGylated polycyanoacrylate nanoparticles as new drug carrier for brain delivery, Pharm Res., 18 (2001) 1157-1166. [193] A. Ambruosi, H. Yamamoto, J. Kreuter, Body distribution of polysorbate80 and doxorubicin-loaded [14C]poly(butyl cyanoacrylate) nanoparticles after i.v. administration in rats, J Drug Target., 13 (2005) 535-542. [194] V. Vamvakaki, N.A. Chaniotakis, Pesticide detection with a liposomebased nano-biosensor, Biosens Bioelectron, 22 (2007) 2848-2853. [195] M. Michel, M. Winterhalter, L. Darbois, J. Hemmerle, J.C. Voegel, P. Schaaf, V. Ball, Giant liposome microreactors for controlled production of calcium phosphate crystals, Langmuir, 20 (2004) 6127-6133. [196] P. Van Gelder, H. De Cock, J. Tommassen, Detergent-induced folding of the outer-membrane protein PhoE, a pore protein induced by phosphate limitation, Eur J Biochem, 226 (1994) 783-787. [197] M. Sauer, T. Haefele, A. Graff, C. Nardin, W. Meier, Ion-carrier controlled precipitation of calcium phosphate in giant ABA triblock copolymer vesicles, Chem Commun (Camb), (2001) 2452-2453. [198] F. Axthelm, O. Casse, W.H. Koppenol, T. Nauser, W. Meier, C.G. Palivan, Antioxidant nanoreactor based on superoxide dismutase encapsulated in superoxide-permeable vesicles, J Phys Chem B, 112 (2008) 8211-8217. [199] O. Onaca, D.W. Hughes, V. Balasubramanian, M. Grzelakowski, W. Meier, C.G. Palivan, SOD Antioxidant Nanoreactors: Influence of Block Copolymer Composition on the Nanoreactor Efficiency, Macromol Biosci. [200] M. Grzelakowski, O. Onaca, P. Rigler, M. Kumar, W. Meier, Immobilized protein-polymer nanoreactors, Small, 5 (2009) 2545-2548. 52 Chapter 2 Scope of the thesis Scope of the thesis 2 Scope of the thesis Due to their low bioavailability, many naturally occurring enzymes cannot be used in their native form as therapeutics in enzyme replacement. A large variety of carrier systems have been developed to protect and deliver enzymes, based on lipids, polymers or conjugates. Nevertheless, to solve the specific pathological situation in enzyme deficiencies, a long-lasting restoration of enzyme activity is required and the delivery of the enzyme needs further improvement. We suggest this can be accomplished by developing enzyme loaded nanoreactors based on biocompatible, biodegradable and non-toxic polymer systems. Due to their adapted size, stealth surface properties and multi-functionality, they can cope with the complexity of the in vivo biological conditions. The scope of this work was to develop new enzyme-containing polymeric nanoreactors and to investigate the therapeutic potential of these nanoreactors as drug delivery vehicles in enzyme replacement therapy. As proof of principle, we focused on the enzyme deficiency MNGIE (mitochondrial neurogastrointestinal encephalomyopathy), which is caused by a deficiency of the enzyme thymidine phosphorylase. This model system was chosen because 1) the lab has experience in working with enzymes of the purine and pyrimidine salvage pathway, 2) a mouse model of this disease is available, 3) there is proof of a clear correlation between the amount of toxic substrates in plasma/urine and the pathology of the disease and 4) there is still no therapy available for this devastating disease. The nanoreactors investigated in this study are constructed of the amphiphilic triblock copolymer poly(methyloxazoline)-poly(dimethylsiloxane)- poly(methyloxazoline) (PMOXA-PDMS-PMOXA). The main advantage of this triblock copolymer is the versatility of the self-assembled particles; 1) the shell has a hydrophilic, biocompatible low-protein-binding surface and an extremely high mechanical and thermal stability, 2) the shell can act as a biomembrane for protein reconstitution, 3) the nanovesicle can encapsulate hydrophilic 55 Chapter 2 substances, and 4) the shell material has enormous possibilities for molecular functionalization. The amphiphilic ® formulated by Ciba-Vision PMOXA-PDMS-PMOXA was initially as a material for contact lenses, because of its biocompatibility and low-protein binding capacity. In addition, the potential of the PMOXA polymer as outer shell in drug delivery particles is reported (and patented) by Zalipsky and Woodle since this polymer seems to possess stealth properties comparable to the PEG polymer. In this study the porin Tsx was selected to permeabilize the polymeric membrane, allowing substrates and products to go in and out of the nanoparticle. This nucleoside specific porin is known to be very efficient in transporting nucleosides and nucleobases even at low concentrations. As such, the rather low but toxic concentrations of the substrates thymidine and deoxyuridine in the plasma of MNGIE patients can be transported across the polymeric membrane into the nanocapsule where they can be degraded into their corresponding products. All together, the thymidine phosphorylase containing nanoreactors, schematically represented in Figure 2-1, serve both to protect the enzyme from degradation and to allow the enzyme to act in situ. As such, these nanoparticles could have advantages as enzyme delivery vehicles for the use in enzyme replacement therapy of MNGIE. The aim of this thesis was to construct such enzyme containing nanoreactors and to evaluate them as potential drug delivery particles. Most of the aforementioned studies start from the assumption that the PMOXA-PDMS-PMOXA nanoparticles are more stable than liposomes and it is often implied that those particles are non-immunogenic and inert towards macrophages. However, until now systematic analysis to unequivocally support these claims is lacking. Therefore, this study not only focuses on the construction of new thymidine phosphorylase containing nanoreactors, but also on the investigation of their stability, toxicity, immunogenicity, biodistribution and efficacy, not only in vitro but also in vivo. Up until now, there is a clear lack of in vivo studies of different polymeric particles. The 56 Scope of the thesis available published data are scattered bits and pieces mostly focusing on in vitro tests on cell cultures. Hence, this work is of high relevance for the further exploitation of the PMOXA-PDMS-PMOXA particles as drug delivery devices. Figure 2-1: Schematic representation of the envisioned thymidine phosphorylase (TP) containing PMOXA-PDMS-PMOXA nanoreactor with integrated Tsx porins within the polymeric membrane. 57 Chapter 3 Production and biophysical characterization of the nanoreactors Production and biophysical characterization of the nanoreactors 3 Production and biophysical characterization of the nanoreactors 3.1 Introduction The construction and characterization of new polymeric nanometer-sized bioreactors, so called nanoreactors, as a new approach for enzyme replacement therapy for MNGIE is investigated in this study. The therapeutic enzyme, thymidine phosphorylase, is encapsulated in polymeric particles constructed of the amphiphilic triblock copolymer poly(2-methyloxazoline)block-poly(dimethylsiloxane)-block-poly(2-methyloxazoline) (PMOXA-PDMS- PMOXA). PMOXA was chosen as hydrophilic outer block for the nanoreactors because this polymer possesses protein-repellent and stealth properties, shielding the encapsulated enzyme from aggression by external agents [1, 2]. The nanoparticles are permeabilized for substrates and products by integrating bacterial channel proteins in their polymeric wall, a strategy introduced by Wolfgang Meier and co-workers [3]. They showed that channel forming proteins can be incorporated without loss of function inside a polymeric wall, despite a membrane thickness three times wider than a biological lipid membrane [4]. This results in reactors where the enzymatic reaction is restricted to the inner volume of the polymeric nanocontainer. This is a relatively new concept that could give advantages over the conventional drug delivery systems for the delivery of enzymes. Recently, other PMOXA-PDMS-PMOXA nanoreactors were constructed for their use in various biomedical applications like prodrug cancer therapy [5], prevention of heart attack [6] and development of bio-sensors [7]. However, the use of such nanoreactors as enzyme delivery devices in enzyme replacement therapy has never been investigated before. For enzyme replacement applications the circulation time of active nanoreactors in the bloodstream is extremely important. The longer the particles remain active in 61 Chapter 3 the bloodstream, the less frequently administrations need to be performed. Even a rather small improvement in the circulation half-life would mean a tremendous progress for patients suffering from a fatal enzyme deficiency disorder (by improving their quality of life as well as by reducing the high costs of the therapy). In this context it is important to note that the nanoreactors investigated in this study are developed as proof of principle for the concept of using enzyme containing PMOXA-PDMS-PMOXA nanoreactors for enzyme replacement purposes. Since the choice of working with the enzyme thymidine phosphorylase was predominantly made on practical considerations, the results must be placed in a broader perspective, knowing that the nanoreactors still have many tunable properties. In this first chapter the construction and biophysical characterization of new thymidine phosphorylase containing nanoreactors is described. Size, dispersity, morphology, activity, encapsulation efficiency and pyrogenicity are important characteristics for the use of nanoparticles as drug delivery systems in general and are thus the first parameters to determine. 3.2 Materials and methods 3.2.1 Cloning, expression and purification of thymidine phosphorylase The gene encoding the Escherichia coli C thymidine phosphorylase (TPE.coli) was cloned via the HindIII-BamHI restriction sites in the pQE30 vector (Qiagen). The protein was expressed in E.coli WK6 cells containing the pQE30TPE.coli construct. Overnight cultures were used to inoculate baffled flasks with Terrific Broth and 50 mg/ml ampicillin and were incubated at 37°C while shaking until the cells reached an optical density at 600 nm between 0.6-0.8. The medium was then cooled to 28°C and expression of the recombinant protein was induced by adding isopropyl β-D-1-thiogalactopyranoside (IPTG) (0.5 mM final concentration). The next morning cells were harvested and resuspended in 20 mM Tris pH 7.5 with 1 M NaCl, 10 mM imidazole and the protease inhibitors leupeptine (1 µg/ml) and 4-(2-aminoethyl) benzenesulfonyl fluoride hydrochloride (AEBSF) (0.1 mg/ml). The cells were lysed via a 62 Production and biophysical characterization of the nanoreactors passage through the cell crack (Atlas Copco) and the disrupted cell suspension was cleared via centrifugation (30 minutes at 15000 rpm). The presence of an N-terminal His6-tag allows a two-step purification scheme, consisting of a NiNTA affinity chromatography (Qiagen) followed by a gelfiltration on a Superdex 200 16/90 column (Amersham Biosciences) (in 20 mM Hepes, 150 mM NaCl and 1 mM CaCl2 pH 7.5). For the recombinant expression of human TP (TPhuman) two plasmids, calTP and GroESL, were kindly provided by Dr. Krosky (SAIC Frederick, Inc). The calTP plasmid (derived from the pCAL-n-EK vector from Stratagene) contains an open reading frame encoding the TPhuman with an N-terminal calmodulinbinding peptide-tag. The GroESL plasmid encodes for the chaperons GroES and GroEL which enhance correct folding of the human enzyme. Both plasmids were expressed in E.coli BL21 cells which were induced with IPTG after 2 hours growth at 37°C and allowed to grow further for 4 hours at 20°C. After cell harvesting, the pellet was resuspended in lysisbuffer containing 2.9 mM CaCl2, 2.7 mM KCl, 1.5 mM KH2PO4, 0.5 mM MgCl2, 200 mM NaCl, 8.1 mM Na2HPO4, 1 mM DTT and 0.1% NP40 and lysed via 3 passages through a French Press (Beun de Ronde Serlabo) at 11000 psi. Due to the calmodulin-binding peptidetag the protein could be purified via calmodulin-sepharose affinity chromatography (Stratagene). Finally, a gelfiltration step on a Superdex 200 16/90 column (Amersham Biosciences) was performed (in 50 mM Tris-HCl Ph 8.0, 150 mM NaCl and 1 mM DTT). 3.2.2 Cloning, expression and purification of porins Tsx and OmpF OmpF was cloned into the pGOmpF vector and expressed into a BL21(DE3) ΔlamB ΔompC ΔompA OmpF::Tn5 strain as described by Prilipov et al. [8]. After cell harvesting by centrifugation, the cell pellets were resuspended in a 20 mM Tris-HCl pH 8 lysisbuffer containing 2% SDS and shaken for 1 hour at 60°C to obtain cell lysis. Lysates were pelleted by ultracentrifugation and resuspended in 20 mM NaH2PO4 pH 7.3 + 0.125% oPOE for pre-extraction at 37°C for 45 minutes. After a second cycle of centrifugation, pellets were resuspended in the same phosphate buffer with 3% oPOE for extraction at 63 Chapter 3 37°C for 45 minutes. After centrifugation and concentration, a gel filtration on a Superdex 200 16/90 column (Amersham Biosciences) was performed (in 20 mM NaH2PO4, 150 mM NaCl and 1% oPOE pH 7.3) to remove contaminating proteins. Tsx was cloned into a pGTsx vector, which was derived from the pGOmpF vector by introducing the sequences for an N-terminal His6-tag and a thrombin cleaving site into the PstI site. BL21(DE3) ΔlamB ompR ΔompA strain was used for overproduction of Tsx as described by Prilipov et al. [8]. The Tsx porin was purified by detergent extraction out of the bacterial membrane by using octylpolyoxyethylene (oPOE). After cell harvesting by centrifugation, the cell pellets were resuspended in 20 mM Tris-HCl pH 8.0, 150 mM NaCl, 5% w/v sucrose and 2 mM EDTA. The cells were lysed via a passage through the cell crack (Atlas Copco). Lysates were pelleted by centrifugation and resuspended in 50 mM Tris-HCl pH 8.0 + 0.125% oPOE for pre-extraction at 37°C for 45 minutes. After a second cycle of centrifugation, pellets were resuspended in 50 mM Tris-HCl pH 8.0 with 250 mM NaCl and 3% oPOE for extraction at 37°C for 45 minutes. Since Tsx was expressed with a His6-tag, it was further purified on a Ni-NTA affinity column (Qiagen). Finally, a gel filtration on a Superdex 200 16/90 column (Amersham Biosciences) was performed in a buffer containing 50 mM Tris-HCl pH 7.3, 250 mM NaCl and 1% oPOE to remove contaminating proteins. 3.2.3 Production of nanoreactors PMOXA (poly 2-methyloxazoline)20 - PDMS (poly dimethylsiloxane)54 - PMOXA20 triblock copolymers with a molecular weight of 7484 g/mol were purchased from Polymer Source Inc. (Montreal, Canada). The thymidine phosphorylase encapsulating nanoreactors (TP-NRs) were constructed via the solvent evaporation method [9]. First, 20 mg of the PMOXA-PDMS-PMOXA polymer was dissolved in 300 µl ethanol and porin was added at the desired molecular ratio of protein to polymer. Secondly, this solution was dripped in PBS containing 100 µM of thymidine phosphorylase (resulting in a final concentration of 10 mg polymer/ml) and stirred for 1 hour at room 64 Production and biophysical characterization of the nanoreactors temperature in order to evaporate the ethanol. Afterwards, continuous stirring for several days at 4°C results in self-assembled nanoreactors. If necessary, successive extrusion through a polycarbonate membrane (Avestin) with pore diameter of 400 nm and 200 nm consecutively was performed. Finally, the non-encapsulated enzymes were removed via Ni-NTA affinity chromatography for TPE.coli or via calmodulin affinity chromatography for TPhuman. 3.2.4 Dynamic Light Scattering The size and polydispersity of the nanoreactors was determined via Dynamic Light Scattering (DLS) using a laser-spectroscatter (RiNA GmbH, Berlin, Germany) at 532 nm with a scattering angle of 90°. Ten data sets were recorded and the autocorrelation curve, size distribution and counts of each data set were analyzed via the CONTIN software. 3.2.5 Spectrophotometric assay for TP activity The concentrations of thymidine phosphorylase and thymidine stock solutions were determined using the following extinction coefficients: 280 mM cm for TPE.coli, 280 -1 -1 for thymidine. All kinetic experiments were performed in 200 mM mM cm = 28.99 mM cm -1 = 24.41 -1 nm -1 nm -1 for TPhuman and 267 nm = 9.65 potassium phosphate pH 6.8 at 37°C. Product formation was determined spectrophotometrically using the difference in absorption substrate thymidine and the product thymine (∆290 nm between -1 the -1 = -1 mM cm ). To measure the enzymatic activity of the nanoreactors, 900 µM thymidine was mixed with 20 µl nanoreactors (in a cuvet with a total volume of 500 µl) and the decrease in absorption was measured at 290 nm using a Cary 100Bio UV visible spectrophotometer (Varian). 3.2.6 Atomic Force Microscopy Freshly cleaved mica was coated with APTES (3-aminopropyltriethoxy silane) by the evaporation method [10]. In short, 30 µl of APTES was dropped into a plastic cap and 10 µl of N,N-diisopropylethylamine in another cap in a dessicator filled with argon. Freshly cleaved mica was mounted at the top of 65 Chapter 3 the dessicator and left for the reaction to proceed for 1-2 hours. Next, the cap with APTES was removed and purged with argon for 2 minutes. The mica sheets were cured for 1-2 days. The nanoreactors prepared via the solvent evaporation method were deposited onto the APTES-mica and left to adsorb for 30 minutes. They were gently washed with pure water (resistivity was over 18 MΩ.cm). Images were obtained with a multimode AFM equipped with a Nanoscope IIIa controller (Veeco, Santa Barbara, CA) using a glass fluid cell. All measurements were made with standard Veeco NP-S silicon nitride cantilevers with lengths of 200 µm and nominal spring constants of 0.06 N/m. AFM measurements were performed in pure water using tapping mode at a scan rate around 1.5-2 Hz. 3.2.7 Detection of endotoxins: LAL assay The level of bacterial endotoxins, especially lipopolysaccharide (LPS), in the protein or nanoreactor suspensions were determined with the Limulus ameboecyte lysate assay (LAL-assay) according to the manufacturer’s instructions (Cambrex, Inc.). The principle of the test is based on the fact that bacteria cause intravascular coagulation in the horseshoe crab, Limulus polyphemus. The agent responsible for the clotting phenomena resides in the crab’s ameboecytes and a lysate of these ameboecytes (Limulus Ameboecyte Lysate or LAL) can therefore be used for the detection of endotoxins. The Kinetic-QCL® LAL assay is a quantitative, kinetic assay for the detection of Gram negative bacterial endotoxins. The samples and standards are first incubated with the LAL reagent and placed in a 96 well plate. During this incubation, the bacterial endotoxins present in the samples will catalyze the activation of a pro-enzyme present in the LAL reagent to an activated enzyme. Secondly, a colorless substrate is added which will be consumed by the activated enzyme thereby releasing the chromophore p-nitroaniline. The plate is placed in an incubating plate reader and monitored over time (via WinKQCL® Software) for the appearance of the yellow color. 66 Production and biophysical characterization of the nanoreactors 3.2.8 Removal of endotoxins A first technique used to remove endotoxins from the porin samples is a MonoP chromatofocusing. Chromatofocusing is a chromatography method where proteins with the same pI are focused in high concentration and resolved from other proteins that have different pI in a pH gradient matter. In this case, the endotoxins present in the sample are washed away from the Mono-P column during a slow pH gradient, hereafter the porin itself is eluted from the column at a pH = pI. A start buffer of 0.025 M bis-Tris/HCl pH 6.2 and an eluent Polybuffer 74/HCl pH 3.4 (1/10 diluted) was applied so that the porin is eluted after 2/3 of the pH gradient. A Mono-P column (Tricorn) was used at a flow rate of 1 ml/min. Secondly, polymyxin B affinity chromatography (Sigma) was used to remove endotoxins from the samples. Since polymyxin B has affinity for the lipid A part of LPS it is able to separate the porin from the LPS fraction. After equilibrating the resin with endotoxin-free 0.1 M ammonium bicarbonate buffer pH 8.0, the sample was loaded and eluted at a flow rate of 0.2 ml/min. The last method applied to remove endotoxins is the use of Prosep Remtox beads (Millipore). A small amount of the beads were mixed with the sample and incubated for 1.5 hour at 4°C while mixing slowly. Afterwards, the beads were separated from the samples via centrifugation. 3.3 Results 3.3.1 Providing the nanoreactor building blocks: enzymes and porins Thymidine phosphorylase (TP) is a cytosolic dimeric enzyme that catalyzes the reversible phosphorolysis of the nucleosides thymidine and deoxyuridine into the nucleobases thymine and uracil. Two different enzymes were initially purified: the Escherichia coli TP (TPE.coli) and the human TP (TPhuman). The TPE.coli enzyme was expressed in high levels in E.coli WK6 cells and purified via Ni-NTA affinity chromatography followed by a gelfiltration. A quasi pure protein 67 Chapter 3 could be obtained with a subunit molecular weight of around 52 kDa as determined with sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) analysis and a yield of 13 mg/l culture medium (Figure 3-1). Pure TPhuman was obtained with a subunit molecular weight of around 54 kDa and a yield of 3 mg/l culture medium via a calmodulin-affinity-chromatography followed by a gelfiltration (Figure 3-1). The catalytic parameters kcat, KM and kcat/KM of the purified enzymes were determined via initial rate experiments. The initial rate of conversion of different concentrations of thymidine was measured spectrophotometrically at 290 nm in 200 mM potassium phosphate pH 6.8 at 37°C. The results were fitted with the Michaelis-Menten equation (Origin software) to determine the catalytic parameters of the enzymes, which are summarized in Table 3-1. Table 3-1: Catalytic parameters of purified thymidine phosphorylases. kcat (s-1) KM (µM) kcat/KM (s-1 µM-1) TPE.coli 413.6 ± 71.1 643.2 ± 268.7 0.64 ± 0.29 TPhuman 16.9 ± 1.1 82.5 ± 19.7 0.2 ± 0.05 Initially two different porins, OmpF and Tsx, were chosen for characterization as potential transporters in our nanoreactors. Both were recombinantly expressed and purified. OmpF is a trimeric protein of which each monomer forms a 16-stranded transmembrane -barrel that functions as a molecular sieve, allowing concentration driven diffusion of solutes smaller than 600 Da [11]. Tsx on the other hand is a monomer that forms a 12-stranded -barrel and allows specific transport of nucleosides and nucleotides [12]. Since Tsx has specific binding sites for nucleosides in the interior of the channel, rapid transport of nucleosides at low concentrations is possible compared to slow diffusion through the non-specific porin OmpF. Detergent extraction with oPOE of the porins from the cell envelope fraction yielded highly pure material. The Tsx porin is observed on SDS-PAGE as a band corresponding well to the molecular weight of the protein (31 kDa) (Figure 3-1) and was produced with a yield of 8 mg/l culture medium. OmpF is observed on SDS-PAGE as a band 68 Production and biophysical characterization of the nanoreactors which corresponds well with the expected subunit molecular weight of 37 kDa (Figure 3-1). For this porin, a yield of 5 mg/l culture medium was obtained. Figure 3-1: Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) of the purified proteins. The protein molecular weight markers used are the SM1811 from Fermentas (left gel), the Mark 12 from Invitrogen (middle gel) and the SM0431 from Fermentas (right gel). Since the porins are extracted from the outer bacterial membrane it is expected that the samples contain residual lipopolysaccharides. However, full stripping of lipopolysaccharides is needed since we aim to use the porins as building blocks (incorporated inside the polymeric membrane) for nanoreactors that we want to administer in vivo. LPS induce strong immune responses after injection in mammals and such unwanted side-effects need to be avoided in our study [13]. Therefore, the amount of endotoxin in the purified Tsx and OmpF samples was determined via the LAL assay. This is the most sensitive and FDA-recommended test to detect endotoxins in therapeutic samples [14]. With the LAL test the amount of endotoxin in the sample is calculated in endotoxin units (EU). The maximum amount of EU allowed to be administered in vivo without inducing immune responses is 10 EU per mouse. After purification, the Tsx porin has an endotoxin load lower than 0.05 EU/ml. As such, for this porin no further purification is needed. The endotoxin load measured in purified OmpF samples is however very high (above the detection 69 Chapter 3 range of the assay, > 50 EU/ml). endotoxins from the porin Three different methods to remove the were compared: polymyxin B affinity chromatography, Mono-P chromatofocusing and Prosep Remtox beads. After polymyxin B affinity chromatography, the LAL test still displays too high amounts of endotoxin units (> 50 EU/ml). In addition, not only the LPS but the entire OmpF-LPS aggregates are sticking to the column and the major amount of porin was only released from the polymyxin beads after regenerating the column. This indicates that the affinity between polymyxin B and OmpF is too high to use this technique for the removal of LPS from the porin sample. Secondly, incubating the OmpF sample with Prosep Remtox beads was used to remove the LPS. Despite its easy and fast procedure the endotoxin load after incubation with Remtox beads is still higher than 50 EU/ml. In a last attempt, a Mono-P chromatofocusing was performed to remove LPS from the OmpF samples. After loading the samples at pH 6.2 a pH gradient was introduced and the OmpF porin was eluted at a pH around 4.6 which corresponds well with the pI of the porin. The pooled fractions were analyzed with the LAL test and showed an endotoxin amount of 1.774 EU/ml which is low enough for in vivo use. In conclusion, the nucleoside specific porin Tsx has a higher yield and doesn’t need extra purification steps to remove endotoxins. Moreover, earlier work describes the efficient transport of thymidine through these porins [15]. In addition, the successful incorporation of the porin Tsx into a PMOXA-PDMSPMOXA polymeric membrane, with evidence of protein functionality, was proved previously [5]. Therefore, Tsx was the preferred porin to use for nanoreactor construction. 3.3.2 Production, size and dispersity of the nanoreactors The amphiphilic copolymer PMOXA20-PDMS54-PMOXA20 was used to construct thymidine phosphorylase containing nanoreactors (TP-NRs). TP-NRs were prepared via the solvent evaporation method as described above, which included extensive stirring for several days. If the stirring was performed in a 70 Production and biophysical characterization of the nanoreactors continuous and homogeneous matter, the nanoparticle sample spontaneously has a high degree of monodispersity (as measured with DLS) and no extrusion was needed. Otherwise, successive extrusion through a filter with pore diameter of 400 nm and 200 nm was performed. Finally, the non-encapsulated enzymes were removed via affinity chromatography. The samples were analyzed by dynamic light scattering to determine the size and the polydispersity of the nanoreactors. Nanoreactors with a mean size around 200 nm diameter were obtained with low batch-to-batch variability (Figure 3-2). Figure 3-2: Size distribution curve (log-scale) of free thymidine phosphorylase in blue and thymidine phosphorylase containing nanoreactors in red. The mean radius of the nanoreactors is around 100 nm. The TP-NRs were further analyzed via Atomic Force Microscopy (AFM). As shown in Figure 3-3 the nanoreactors appear as relatively monodisperse and round particles. Nevertheless, the adsorption of the nanoreactors onto APTES resulted in a larger base area. This behaviour has also been observed for vesicle-bilayer complexes [16]. The images indicate that the polymeric particles are very flexible: the upper part of the particle is easily deformed by the tip interaction during scanning. Increasing the tip force on the nanoreactor also resulted in reduced height of the particles. The adsorption and tip-sample interaction during scanning resulted in significantly lower dimensions of the nanoreactors obtained with AFM than ensemble measurements with DLS [17]. 71 Chapter 3 Analysis of the cross section of the nanoreactors shows a diameter of approximately 100 nm and a height of 10-20 nm (Figure 3-3C), while DLS measurements of the same particles showed a mean radius of 104 nm. Figure 3-3: AFM image of enzyme filled nanoreactors: (A) Tapping mode height image (scale bar 300 nm). (B) Zoomed-in 3D-image of the nanoreactor indicated by the arrow on image (A). (C) Cross-sectional analysis of 3 nanoreactors; blue colored section corresponds to the nanoreactor marked with 2 blue "+" - signs and the white crosssection line in image (A), red curve corresponds to the 2 marked nanoreactors in the middle of image (A). The distance between the "+" - signs on figure (A) is indicated by the vertical lines on figure (C). 3.3.3 Enzymatic activity of the nanoreactors The in vitro enzymatic activity of the TP-NRs was confirmed spectrophotometrically by measuring the decrease in absorption at 290 nm due to thymidine consumption. The thymidine concentration used is 900 µM which is above the KM of the enzyme. As such, the enzyme can work in theory close to its maximal turnover efficiency. However, this concentration corresponds to the substrate concentration outside the particles. Since the activity of the nanoreactors is probably limited by the transport of substrate 72 Production and biophysical characterization of the nanoreactors and product over the membrane, the substrate concentration inside the particle that is available for the enzyme is probably lower. TP-NRs induce, in comparison to control samples (PBS and empty nanoparticles), a decrease in absorption due to the conversion of thymidine to thymine. These results show that the nanoreactors possess thymidine phosphorylase activity. However, in comparison to the same amounts of free enzyme, the enzyme activity of the nanoreactors is rather low. Figure 3-4: Activity measurements showing the change in absorption as a function of time measured at 290 nm. For each measurement 900 µM thymidine was mixed with 20 µl nanoparticles or control samples (in a cuvet with a total volume of 500 µl) and the reaction was followed for 10 minutes at 37°C. 3.3.4 Encapsulation efficiency The efficiency of encapsulation (i.e. the amount of TPE.coli found in the nanoparticles after separation versus the starting amount of protein) was determined by comparative SDS-PAGE analysis. Known quantities of free TPE.coli (0.5 µg, 1 µg, 2 µg, 4 µg, 8 µg, 12 µg, 24 µg) were run on an SDSPAGE together with an aliquot (20 µl) of two different nanoreactor samples. The nanoparticles were first disrupted completely by boiling them for 15 minutes in the presence of SDS-containing loading buffer. The intensity of each coomassie blue stained TPE.coli band was measured with Quantity One 73 Chapter 3 Software (ChemiDocXRS, Biorad). The amount of TPE.coli incorporated in the nanoreactor sample was determined via a standard curve that plots the protein amount versus band intensity of known amounts of protein. The overall enzyme concentration in the sample is 10 µM (0.5 mg/ml) (Figure 3-5). As such, an enzyme encapsulation efficiency of around 10% is found. Take care, 10 µM is the overall concentration in the sample, the local enzyme concentration inside the particles is higher (ideally 183 molecules per particle*). Figure 3-5: SDS-PAGE analysis of two TP-NR samples. The nanoparticles with enzyme and porin (NP+TP+Tsx) show encapsulation of TPE.coli (±52 kDa) and incorporation of porin Tsx (±31 kDa). The nanoreactor-sample without porins (NP+TP-Tsx) shows only an analogues enzyme band. Known concentrations of TPE.coli (lane 1 = 0.5 µg; lane 2 = 1 µg; lane 3 = 2 µg; lane 4 = 4 µg; lane 5 = 8 µg; lane 6 = 12 µg; lane 7 = 24 µg) serve as standard to calculate the encapsulation efficiency of the enzyme inside the nanoreactors. The protein molecular weight marker used is SMO431 from Fermentas. * The volume of the nanoreactors with an internal radius of 90 nm is 3.05 x 10-18 l. With an initial enzyme concentration of 100 µM and the assumption that there is no attraction or repulsion of the enzyme to the particles, but a simple encapsulation of the specific volume, there will be ideally 183 molecules TP inside each nanoreactor. 74 Production and biophysical characterization of the nanoreactors 3.4 Conclusion and discussion New thymidine phosphorylase containing nanoreactors composed of PMOXAPDMS-PMOXA triblock copolymers were constructed. The nanoreactors are permeabilized with the bacterial membrane protein Tsx. We were able to produce monodisperse nanoreactors with a mean size of 200 nm. This is the ideal size if we intend long-circulating particles [18]. Furthermore, we demonstrated that the enzyme thymidine phosphorylase can be encapsulated inside the nanoreactors with an encapsulation efficiency around 10%, comparable to enzyme encapsulation in liposomes, with a high batch-to-batch reproducibility. In addition, these nanoreactors are able to convert thymidine into thymine and thus show enzymatic activity. Although rather low, the activity is significantly above the background signal. Moreover, the pyrogenicity of the porins by means of endotoxin contamination was addressed and purification steps were introduced if necessary to ensure the absence of LPS in the porin and concomitant nanoreactor samples. All together, the physicochemical properties of the nanoreactors are promising and further steps in the preclinical characterization of the particles can be carried out. 3.5 References [1] M.C. Woodle, C.M. Engbers, S. Zalipsky, New amphipatic polymer-lipid conjugates forming long-circulating reticuloendothelial system-evading liposomes, Bioconjug Chem, 5 (1994) 493-496. [2] S. Zalipsky, C.B. Hansen, J.M. Oaks, T.M. Allen, Evaluation of blood clearance rates and biodistribution of poly(2-oxazoline)-grafted liposomes, J Pharm Sci, 85 (1996) 133-137. [3] C. Nardin, J. Widmer, M. Winterhalter, W. Meier, Amphiphilic block copolymer nanocontainers as bioreactors, Eur. Phys. J. E, 4 (2001) 403-410. [4] M. Winterhalter, Hilty, C., Bezrukov, S.M., Nardin, C., Meier, W., Fournier, D., Controlling membrane permeability with bacterial porins: applications to encapsulated enzymes, Talanta, 55 (2001) 965-971. [5] A. Ranquin, W. Versees, W. Meier, J. Steyaert, P. Van Gelder, Therapeutic Nanoreactors: Combining Chemistry and Biology in a Novel Triblock Copolymer Drug Delivery System, Nano Lett, 5 (2005) 2220-2224. [6] P. Broz, N. Ben-Haim, M. Grzelakowski, S. Marsch, W. Meier, P. Hunziker, Inhibition of Macrophage Phagocytotic Activity by a Receptor-targeted Polymer 75 Chapter 3 Vesicle-based Drug Delivery Formulation of Pravastatin, J Cardiovasc Pharmacol, 51 (2008) 246-252. [7] M. Grzelakowski, O. Onaca, P. Rigler, M. Kumar, W. Meier, Immobilized protein-polymer nanoreactors, Small, 5 (2009) 2545-2548. [8] A. Prilipov, P.S. Phale, P. Van Gelder, J.P. Rosenbusch, R. Koebnik, Coupling site-directed mutagenesis with high-level expression: large scale production of mutant porins from E. coli, FEMS Microbiol Lett, 163 (1998) 6572. [9] C. Nardin, Hirt, T., Leukel, J., Meier, W., Polymerized ABA Triblock Copolymer Vesicles, Langmuir, 16 (2000) 1035-1041. [10] M. Bezanilla, M. Srinivas, D. Laney, Y. Lyubchenko, H. Hansma, Adsorption of DNA to mica, silylated mica, and minerals : characterization by atomic force microscopy, Langmuir 11 (1995) 655-659 [11] R. Koebnik, K.P. Locher, P. Van Gelder, Structure and function of bacterial outer membrane proteins: barrels in a nutshell Mol Microbiol, 37 (2000) 239253. [12] J. Ye, B. Van Den Berg, Crystal structure of the bacterial nucleoside transporter Tsx, EMBO J, 23 (2004) 3187-3195. Epub 2004 Jul 3122. [13] A.J. Ulmer, H. Flad, T. Rietschel, T. Mattern, Induction of proliferation and cytokine production in human T lymphocytes by lipopolysaccharide (LPS), Toxicology, 152 (2000) 37-45. [14] J.B. Hall, M.A. Dobrovolskaia, A.K. Patri, S.E. McNeil, Characterization of nanoparticles for therapeutics, Nanomedicine (Lond), 2 (2007) 789-803. [15] R. Benz, A. Schmid, C. Maier, E. Bremer, Characterization of the nucleoside-binding site inside the Tsx channel of Escherichia coli outer membrane, Eur J Biochem, 176 (1988) 699-705. [16] S. Kumar, J.H. Hoh, Direct visualization of vesicle-bilayer complexes by atomic force microscopy, Langmuir, 16 (2000) 9936-9940. [17] Y. Ebenstein, Nahum, E., Banin, U., Tapping Mode Atomic Force Microscopy for Nanoparticle Sizing: Tip-Sample Interaction Effects, Nano Lett, 2 (2002) 945-950. [18] S.M. Moghimi, A.C. Hunter, J.C. Murray, Long-circulating and targetspecific nanoparticles: theory to practice, Pharmacol Rev, 53 (2001) 283-318. 76 Chapter 4 Stability of the nanoreactors Stability of the nanoreactors 4 Stability of the nanoreactors 4.1 Introduction Polymeric nanoreactors functionalized by encapsulating the enzyme thymidine phosphorylase and permeabilized by incorporating the membrane protein Tsx in the reactor wall were constructed. The nanoreactors have unique physical properties which make them promising particles for enzyme delivery. In theory, the stability of such polymeric particles should be better than the stability of liposomes [1, 2]. This is due to the higher molecular weight and the lower flexibility of the underlying block copolymers. However, the interaction of PMOXA-PDMS-PMOXA nanoreactors with plasma proteins and other blood components is still unknown. Such interactions may influence uptake and clearance and hence potentially affect distribution and delivery of the therapeutic enzyme [3]. Nanoparticle-protein-binding occurs almost instantaneously once the particle enters biological fluid and the physical properties of the drug delivery particle may significantly change. Ideally, the integrity of the nanoreactors is not disturbed in the complex biological environment. As such, understanding the nanoreactor hematocompatibility is an important step during the preclinical characterization of the nanoreactors as enzyme delivery vehicles. Therefore, in this chapter the nanoreactors are characterized in terms of their in vitro stability in serum, thereby mimicking the in vivo situation in the bloodstream. 4.2 Materials and methods 4.2.1 Production of nanoreactors The TPE.coli (thymidine phosphorylase from E.coli) encapsulating nanoreactors (TP-NRs) were constructed via the solvent evaporation method as described in detail in section 3.2.3 [4]. 79 Chapter 4 4.2.2 Production of liposomes The production of liposomes was initially described by Nikaido and coworkers [5, 6]. Briefly, DOPC/ePG lipids dissolved in chloroform (Avanti) were mixed in a ratio of 4:1 and the solution was dried under a gentle stream of nitrogen until a smooth film was formed. Afterwards, 200 µl high quality diethyl ether was added and this solution was incubated for 10 minutes at room temperature in order to dissolve the film completely. Next, the lipid solution was dried while shaking in a 50°C water bath. Drying was completed by incubating the tubes for 30 minutes in a dessicator. Hydration of this lipid film was performed by adding 1 ml of PBS containing the TPE.coli enzyme (100 µM) and the Tsx porin (30 µg) and the vesicles were allowed to swell for at least 30 minutes. Next, the solution was gently sonicated in a water bath sonicator (about 5 seconds) and a homogenous and turbid solution was obtained. Subsequently, eight freeze-thaw cycles and extrusion through a 200 nm polycarbonate filter provided uniform and unilammellar proteoliposomes. The non-encapsulated enzyme was further removed via Ni-NTA affinity chromatography. 4.2.3 Dynamic Light Scattering The size and polydispersity of the nanoreactors was determined via Dynamic Light Scattering (DLS) using a laser-spectroscatter (RiNA GmbH, Berlin, Germany) at 532 nm with a scattering angle of 90°. Ten data sets were recorded and the auto correlation curve, size distribution and counts of each data set were analyzed via the CONTIN software. 4.2.4 Spectrophotometric assay for TP activity All kinetic experiments were performed in 200 mM potassium phosphate pH 6.8 at 37°C. Product formation was determined spectrophotometrically using the difference in absorption between the substrate thymidine and the product thymine (∆290 nm = -1 mM-1cm-1). To measure the enzymatic activity of TPE.coli inside nanoreactors and liposomes, 900 µM thymidine was mixed with 40 µl of 80 Stability of the nanoreactors the nanoreactor/serum solution or 40 µl of the liposome/serum solution. For the activity measurement of free TPE.coli incubated in serum, 20 µl of a 1000x dilution was added to 900 µM thymidine. The decrease in absorption was measured at 290 nm during 10 minutes using a Cary 100Bio UV visible spectrophotometer (Varian). Every measurement was performed twice and the mean decrease in absorption was calculated. 4.3 Results The stability of the thymidine phosphorylase containing nanoreactors (TP-NRs) in physiological conditions is a very important requirement if we aim the intravenous administration of the particles for enzyme replacement purposes. For stability studies, the permeabilized TP-NRs were incubated for several days in 50% naïve mouse serum at 37°C. As a reference, the same experiments were done with PBS instead of blood serum. At different time points both the size distribution (by DLS) and the enzyme activity (290 nm) was measured. Also, liposomes with encapsulated TPE.coli and permeabilized with the porin Tsx (TP-LIPs) were constructed, analogous to previously reported liposomal reactors [7, 8]. TP-LIPs and free TPE.coli were used as control samples. DLS measurements show that the morphology and size of the TP-NRs in mouse serum remains stable over time (Figure 4-1). Although the polydispersity of the nanoreactor samples increases slightly, we can conclude that the nanoreactors are still intact after 4 days of incubation at 37°C. As a negative control, detergent (TritonX-100) was added to the same sample after 96 hours of incubation. This results in a clear shift of the DLS pattern to lower sizes, demonstrating the destruction of the particles. The same results were obtained when the reactors are incubated in PBS (data not shown), although in this case a more narrow dispersity window is found due to the absence of serum proteins. 81 Chapter 4 Figure 4-1: DLS distribution patterns (intensity in function of radius for ten consecutive measurements) of TP-NRs incubated in 50% naive mouse serum after (A) 0 h and (B) 96 h at 37°C. (C) As negative control, detergent TritonX-100 was added to the sample after 96 h of incubation, resulting in a destruction of the particles. To measure the enzyme activity of the TP-NRs, TP-LIPs and free TP incubated in serum, an aliquot was taken after the corresponding times and used for the spectrophotometric measurement as described above. The enzyme activity of the TP-LIPs increases significantly during their incubation period at 37°C (Figure 4-2). Such an increase in apparent activity of the liposomeencapsulated enzyme is expected if enzyme is leaking out of the liposomes, as previously reported [9, 10]. Indeed, for the encapsulated enzymes the netto rate of substrate conversion is limited by the rate of diffusion of the substrate through the porins. Hence, release of the enzyme to the medium leads to the relief of diffusion barriers with concomitant higher relative rates of substrate consumption. In this context it is noteworthy that the activity values plotted in figure 4-2 represent activities relative to the value at the zero-time-point for each formulation and that hence at all times the absolute activity of the encapsulated enzyme preparations remains below the activity of the free enzyme. In contrast to the TP-LIPs, no increase in activity is observed for the TP-NRs. Thus, the enzyme is contained inside the polymeric particles without leakage. This is a clear advantage for their in vivo use because the release of a foreign enzyme will elicit an immune response. Note that the enzyme activity of the TP-NRs decreases slightly over time. After three days half of the initial activity remained. This decrease in activity cannot be due to the degradation 82 Stability of the nanoreactors of the enzyme, because the activity of the free enzyme in serum remains constant. One possible hypothesis for this decrease in activity could be a partial obstruction of the porins by serum proteins. This can be due to the artificial and static in vitro setting and has to be clarified further in the future. In the control experiments where the free TP, TP-NRs and TP-LIPs are incubated in PBS the same results were obtained only no leakage of enzyme out of the liposomes was detected as no increase in relative activity was observed (data not shown). These results confirm once again that upon interaction with serum proteins the liposomes become leaky, while the nanoreactors remain their integrity and keep the enzyme encapsulated and protected. Figure 4-2: Residual activity of TP-NRs (●), TP-LIPs (■) and free TP (▲) during their incubation in 50% mouse serum. For each enzyme formulation the relative activity scaled towards its zero-time-point activity is plotted as function of time. The data represent at least three independent experiments. 83 Chapter 4 4.4 Conclusion and discussion Here, we present for the first time stability measurements of nanometer-sized PMOXA-PDMS-PMOXA reactors. We can conclude that, although a slight decrease in enzymatic activity was observed over time, the TP-NRs are stable and not leaky at 37°C in serum for several days. The study clearly shows that the nanoreactors are more suited for in vivo use than analogous particles constructed of lipids. Independently from us and at the same time, Litvinchuk and co-workers also investigated the stability of PMOXA-PDMS-PMOXA nanoparticles (without porins) [11]. They as well find that the nanoparticles are considerably more stable than conventional liposomes. In plasma 50% of the encapsulated calcein was released after 48 hours. This release was however attributed to the presence of sodium azide that was added to the blood plasma as sterilization agent. Due to the more robust nature of polymeric nanoreactors compared to liposomes, their application as enzyme delivery vehicles is promising. In the next chapters, important in vivo properties of the nanoreactors (like toxicity, immunogenicity and biodistribution) will be addressed in order to further evaluate the nanoreactors as enzyme replacement particles. 4.5 References [1] D.E. Discher, Emerging applications of polymersomes in delivery: From molecular dynamics to shrinkage of tumors, Progress in Polymer Science, 32 (2007) 838-857. [2] J.C. Lee, H. Bermudez, B.M. Discher, M.A. Sheehan, Y.Y. Won, F.S. Bates, D.E. Discher, Preparation, stability, and in vitro performance of vesicles made with diblock copolymers, Biotechnol Bioeng, 73 (2001) 135-145. [3] P. Aggarwal, J.B. Hall, C.B. McLeland, M.A. Dobrovolskaia, S.E. McNeil, Nanoparticle interaction with plasma proteins as it relates to particle biodistribution, biocompatibility and therapeutic efficacy, Adv Drug Deliv Rev, 61 (2009) 428-437. [4] C. Nardin, Hirt, T., Leukel, J., Meier, W., Polymerized ABA Triblock Copolymer Vesicles, Langmuir, 16 (2000) 1035-1041. 84 Stability of the nanoreactors [5] M. Luckey, H. Nikaido, Diffusion of solutes through channels produced by phage lambda receptor protein of Escherichia coli: inhibition by higher oligosaccharides of maltose series, Biochem Biophys Res Commun, 93 (1980) 166-171. [6] H. Nikaido, Proteins forming large channels from bacterial and mitochondrial outer membranes: porins and phage lambda receptor protein, Methods Enzymol, 97 (1983) 85-100. [7] B. Chaize, J.P. Colletier, M. Winterhalter, D. Fournier, Encapsulation of enzymes in liposomes: high encapsulation efficiency and control of substrate permeability, Artif Cells Blood Substit Immobil Biotechnol, 32 (2004) 67-75. [8] G. Huysmans, A. Ranquin, L. Wyns, J. Steyaert, P. Van Gelder, Encapsulation of therapeutic nucleoside hydrolase in functionalised nanocapsules, J Control Release, 102 (2005) 171-179. [9] S.J. Comiskey, T.D. Heath, Serum-induced leakage of negatively charged liposomes at nanomolar lipid concentrations, Biochemistry, 29 (1990) 36263631. [10] T. Ishida, H. Harashima, H. Kiwada, Liposome clearance, Biosci Rep, 22 (2002) 197-224. [11] S. Litvinchuk, Z. Lu, P. Rigler, T.D. Hirt, W. Meier, Calcein release from polymeric vesicles in blood plasma and PVA hydrogel, Pharm Res, 26 (2009) 1711-1717. 85 Chapter 5 Toxicity of the nanoreactors Toxicity of the nanoreactors 5 Toxicity of the nanoreactors 5.1 Introduction An important prerequisite of the new enzyme-containing nanoreactors is that they may not provoke toxic effects on healthy cells and tissues. Although polymers made of PMOXA-PDMS-PMOXA blocks are approved by the FDA for use in contact-lens material, the toxicity of the polymers for in vivo use is still not determined. To date, there are only a few studies investigating the toxic effects of nanomaterials and no clear guidelines are presently available to quantify these effects [1]. However, it is very important to investigate the toxicity of the nanoreactors as early as possible in their development to avoid clinical trial failure due to the toxic effects of these new drug compounds. Since hepatocytes are a rich source of enzymes involved in xenobiotic biotransformations, the liver is the predominant target for xenobiotic-induced toxicity. For economical and ethical reasons, the use of primary hepatocytes in culture has gained interest as an alternative to in vivo toxicity studies [2]. The importance of short-term hepatocyte cultures in the characterization of acute hepatotoxicity during early drug development is now widely recognized [3]. In contrast, long-term models which can address sub-chronic and chronic hepatotoxicity are not yet available. In this chapter, the potential toxic effect of the thymidine phosphorylase containing nanoreactors is evaluated in an in vitro model derived from rat hepatocytes. 89 Chapter 5 5.2 Materials and methods 5.2.1 Production of nanoreactors The TPE.coli (thymidine phosphorylase from E.coli) encapsulating nanoreactors (TP-NRs) were constructed via the solvent evaporation method as described in detail in section 3.2.3 [4]. 5.2.2 Hepatocyte cell culture Male Sprague Dawley rats (around 250 g) were kept under controlled environmental conditions (12-hour light–dark cycle) and fed a standard diet (Animalabo A04 and water ad libitum). Experiments were carried out in accordance to the regulations of the local ethical committee for animal experiments of the Vrije Universiteit Brussel (VUB). Hepatocytes were isolated (viability > 80% as assessed by trypan blue dye exclusion) and plated at a density of 0.57 × 105 cells/cm2 at 37°C, in an atmosphere of 5% CO2 and 95% air and 100% relative humidity [5]. They were cultured in Williams medium E (GIBCO 22551) supplemented with different antibiotics (50 µg/ml kanamycin monosulphate, 10 µg/ml sodiumampicillin, 7.3 IU/ml sodiumbenzylpenicillin, 50 µg/ml streptomycin sulphate) and additives (2 mM L-glutamine, 5 µg/ml insulin, 7 ng/ml glucagon, 25 µg/ml hydrocortisone-sodium succinate). Cells were allowed to attach to the plastic substratum for 4 hours. Then, serumcontaining medium was removed and fresh, serum-free culture medium supplemented with different concentrations of nanoreactors (50 µg/ml, 100 µg/ml, 500 µg/ml and 1000 µg/ml) was added to the cultures. The medium was further renewed daily. 5.2.3 LDH test The leakage of the enzyme lactate dehydrogenase (LDH) was measured using the Merckotest (VWR International, Leuven, Belgium). Briefly, a mixture of pyruvate and NADH is added to the supernatant and/or cell debris. The decrease in NADH concentration is measured spectrophotometrically at 340 90 Toxicity of the nanoreactors nm and is an amount for the LDH activity. The LDH leakage was calculated in percentage by the ratio: (100 x LDH activity in supernatant)/(LDH activity in (supernatant + cells)). 5.2.4 MTT test The cytotoxicity in the MTT-assay was determined by measuring the reduction of water soluble yellow MTT (3-(4,5-dimethyl-2-thiazolyl)-2,5- diphenyltetrazolium bromide, Sigma) to water insoluble dark blue MTT formazan crystals in living cells, at 570 nm. The results are calculated as relative values (%) to the negative control, whereas untreated control is set to be 100% viable. 5.2.5 Statistical analysis Results are expressed as the mean ± standard deviation of at least three independent experiments. Statistical analysis was performed using a one-way ANOVA test, followed by a Bonferroni correction to study differences amongst individual mean values. P-values < 0.05 are considered significant. 5.3 Results Since the liver is the predominant organ in which biotransformation of foreign compounds takes place, it is the ultimate organ to test xenobiotic–induced toxicity [6, 7]. Therefore, possible toxic effects of the TP-NRs on hepatocytes were investigated. As such, hepatocytes were isolated from male outbred Sprague-Dawley rats by a two-step collagenase perfusion method [8] and cultured in Williams medium E as described previously [5]. After 4 hours of cell adhesion, serum-containing medium was removed and fresh, serum-free culture medium supplemented with different concentrations of TP-NRs (50 µg/ml, 100 µg/ml, 500 µg/ml and 1000 µg/ml) was added to the hepatocyte cultures. Cytotoxicity was tested by cellular morphology, membrane leakage of lactate dehydrogenase (LDH assay) and mitochondrial function (MTT assay) as a 91 Chapter 5 function of culture time. The lactate dehydrogenase (LDH) release assay and the 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyltetrazolium bromide (MTT) reduction assay have been widely used for evaluating cell viability in culture. The LDH assay measures the release of the intracellular enzyme LDH upon damage of the plasma membrane. Therefore, the amount of LDH release corresponds to the amount of death cells. On the other hand, the MTT assay measures the conversion of MTT into purple-colored MTT formazan by the redox activity of living cells and a decrease in cellular MTT reduction corresponds to cell damage. Medium alone and medium with addition of PBS were used as negative controls. SuperFasLigand™ (human recombinant; Alexis Biochemicals) in combination with cycloheximide (Sigma), two known apoptosis-inducers, acted as the positive control [9, 10]. The results show no significant difference in LDH leakage between the various conditions up to 48 hours of incubation (Figure 5-1). Only after prolonged exposure and at higher doses of TP-NRs (96 hours for the 500 µg/ml and 72 hours for the 1000 µg/ml concentrations), elevated levels of LDH leakage were measured. This cytotoxic effect was confirmed by visual inspection of the morphology of the hepatocytes: after prolonged incubation with TP-NRs at high doses, the hepatocyte-morphology became apoptotic in comparison to healthy untreated cells (data not shown). However, TP-NRs at lower doses (50 µg/ml and 100 µg/ml) did not affect the viability of the hepatocytes throughout culture time. Indeed, the hepatocytes treated with TP-nanoreactors at these concentrations still looked healthy after 96 h of incubation (Figure 5-2). The cytotoxic effect of nanoreactors was also tested via the MTT-test. Unfortunately this assay displayed false positive results. An increase in viability was observed after incubation with TP-nanoreactors, and changes in concentration or incubation time had no significant influence on the viability (data not shown). Such inconsistencies were also reported by others [11, 12] and could be due to an interference of the MTT-formazan with the nanoparticles. 92 Toxicity of the nanoreactors Figure 5-1: LDH leakage of hepatocytes incubated with different concentrations of TPNRs as a function of time. Medium with or without TP-NRs was renewed daily. Medium and PBS serve as negative controls, a combination of SuperFasLigand with cycloheximide is used as the positive control. The data represent at least three independent experiments. Statistical analysis were performed using a one-way ANOVA test, followed by a Bonferroni post hoc test. (*p < 0.05 and ***p < 0.001 compared to medium and PBS control values). Figure 5-2: Morphology of cultured hepatocytes incubated for 96 h with (A) medium as negative control, (B) SuperFasLigand-cycloheximide as positive control, (C) TP-NRs at 50 µg/ml and (D) TP-NRs at 100 µg/ml. All images were acquired with constant microscope settings (magnification = 100x). 93 Chapter 5 5.4 Conclusion and discussion Overall, we can conclude that the thymidine phosphorylase encapsulating nanoreactors of concentrations up to 100 µg polymer/ml have no acute cytotoxic effect on hepatocytes. This is evidenced by the maintenance of the cellular morphology and the absence of LDH leakage of hepatocytes during their incubation with nanoreactors. It is moreover questionable if concentrations of 100 μg/ml and higher reflect biological relevant conditions. Indeed, the nanoreactors will circulate in the bloodstream and local concentrations at liver tissue will probably never reach these levels. In conclusion, our results demonstrate that enzyme containing PMOXA-PDMSPMOXA nanoreactors up to 100 µg/ml do not affect the viability of hepatocytes under the experimental conditions of this study. These results are an important step in the evaluation of the nanoreactors as drug delivery devices and opens the door to further examination of their in vivo behavior. In addition, since little information is available on nanoparticle toxicity in general, these data are also a valuable contribution to the toxicity risk assessment knowledge of nanomaterials. 5.5 References [1] S.M. Hussain, K.L. Hess, J.M. Gearhart, K.T. Geiss, J.J. Schlager, In vitro toxicity of nanoparticles in BRL 3A rat liver cells, Toxicol In Vitro, 19 (2005) 975-983. [2] G. Eisenbrand, B. Pool-Zobel, V. Baker, M. Balls, B.J. Blaauboer, A. Boobis, A. Carere, S. Kevekordes, J.C. Lhuguenot, R. Pieters, J. Kleiner, Methods of in vitro toxicology, Food Chem Toxicol, 40 (2002) 193-236. [3] T. Vanhaecke, V. Rogiers, Hepatocyte cultures in drug metabolism and toxicological research and testing, Methods Mol Biol, 320 (2006) 209-227. [4] C. Nardin, Hirt, T., Leukel, J., Meier, W., Polymerized ABA Triblock Copolymer Vesicles, Langmuir, 16 (2000) 1035-1041. [5] T. Henkens, P. Papeleu, G. Elaut, M. Vinken, V. Rogiers, T. Vanhaecke, Trichostatin A, a critical factor in maintaining the functional differentiation of primary cultured rat hepatocytes, Toxicol Appl Pharmacol, 218 (2007) 64-71. [6] V. Rogiers, A. Vercruysse, Rat hepatocyte cultures and co-cultures in biotransformation studies of xenobiotics, Toxicology, 82 (1993) 193-208. 94 Toxicity of the nanoreactors [7] E.F. Brandon, C.D. Raap, I. Meijerman, J.H. Beijnen, J.H. Schellens, An update on in vitro test methods in human hepatic drug biotransformation research: pros and cons, Toxicol Appl Pharmacol, 189 (2003) 233-246. [8] P. Papeleu, T. Vanhaecke, T. Henkens, G. Elaut, M. Vinken, S. Snykers, V. Rogiers, Isolation of rat hepatocytes, Methods Mol Biol, 320 (2006) 229-237. [9] S. Nagata, P. Golstein, The Fas death factor, Science, 267 (1995) 14491456. [10] K. Ito, N. Kiyosawa, K. Kumagai, S. Manabe, N. Matsunuma, T. Yamoto, Molecular mechanism investigation of cycloheximide-induced hepatocyte apoptosis in rat livers by morphological and microarray analysis, Toxicology, 219 (2006) 175-186. [11] J.M. Worle-Knirsch, K. Pulskamp, H.F. Krug, Oops they did it again! Carbon nanotubes hoax scientists in viability assays, Nano Lett, 6 (2006) 1261-1268. [12] T. Laaksonen, H. Santos, H. Vihola, J. Salonen, J. Riikonen, T. Heikkila, L. Peltonen, N. Kumar, D.Y. Murzin, V.P. Lehto, J. Hirvonen, Failure of MTT as a toxicity testing agent for mesoporous silicon microparticles, Chem Res Toxicol, 20 (2007) 1913-1918. 95 Chapter 6 Immunogenicity of the nanoreactors Immunogenicity of the nanoreactors 6 Immunogenicity of the nanoreactors 6.1 Introduction The enzyme containing nanoreactors constructed in this study have unique physicochemical properties which make them promising candidates for drug delivery. However, the application of the nanoreactors as intravenous enzyme delivery vehicles depends on avoiding rapid elimination from the systemic circulation by cells of the immune system. Nanoparticle recognition and uptake by immune cells may occur through different pathways [1]. The common initial step is the adsorption of opsonins to the surface of the nanoparticles. Opsonins are blood proteins that bind to the particle, facilitate the recognition by special immune cells and/or signal immune cells to ingest or destroy the particle. Nanoparticles whose surfaces are not shielded to prevent adsorption of opsonins are removed from the bloodstream within seconds [2]. The protein binding to the surface of nanoparticles depends on their surface characteristics like size, charge, hydrophobicity and composition [3]. In this context, the polymethyloxazoline (PMOXA) polymer was chosen as outer block for the nanoreactors since it has good stealth properties, able to evade binding of opsonins [4, nanoreactors 5]. However, constructed of the immunogenicity PMOXA-PDMS-PMOXA of multi-component triblock copolymers, enzymes and porins has never been investigated before. In theory, the proteins are protected from the environment since the porins are embedded inside the thick polymeric membrane while the enzymes are encapsulated inside the inner space of the particles. Consequently, it is very important to test the interaction of the thymidine phosphorylase containing nanoreactors with the immune system. As nanotechnology-based products are relatively new drug candidates, there are currently no guidelines or harmonized standards for assessing their immunogenicity. The main immunogenic properties addressed here are the inflammatory and antigenic potency of the nanoparticles. These two key 99 Chapter 6 questions will not reveal the complete immunogenic picture, but will provide essential insight in the behavior of the nanoreactors in a complex in vivo environment. Important preclinical parameters to examine in this context are; endotoxin load, uptake by macrophages, cytokine production, nitric oxide production and nanoparticle-specific antibody responses [6]. Therefore, this chapter describes the investigation of the above mentioned parameters in order to evaluate the nanoreactors further as potential enzyme delivery particles. 6.2 Materials and methods 6.2.1 Production of nanoreactors The TPE.coli (thymidine phosphorylase from E.coli) encapsulating nanoreactors were constructed via the solvent evaporation method as described in detail in section 3.2.3 [7]. In addition to the nanoreactors containing enzyme and porin (NP+TP+Tsx), other nanoreactor-compositions were prepared and used in parallel as control samples; nanoparticles with enzyme without porin (NP+TPTsx) and empty nanoparticles (NP-TP-Tsx). 6.2.2 Macrophage cell culture For the in vitro macrophage studies, 5 C57BL/6 mice were each administered with 3 ml thioglycolate solution intraperitoneally (ip). After 4 days, mice were euthanized and peritoneal macrophages were collected by flushing the peritoneal cavity with 10 ml sucrose solution. The cell suspension was centrifuged at 1400 rpm for 8 min at 4°C and resuspended in complete RPMI 1640 medium (supplemented with 10% heat-inactivated Fetal Calf Serum, 100 mg/ml penicillin, 100 mg/ml streptomycin, 0.03% L-glutamine, 1 mM nonessential amino-acids, 1 mM Na-pyruvate and 0.02 mM mercapto-ethanol). These thioglycolate-elicited peritoneal macrophages were plated out at 106 cells/ml and incubated at 37°C. After 4 hours of adhesion the non-adherent cells were washed away and half of the adherent cells (> 98% macrophages) were pre-stimulated with 100 U/ml interferon-γ 100 (IFNγ) to prime the Immunogenicity of the nanoreactors macrophages. The following day different concentrations of TP-NRs (50 µg polymer/ml, 100 µg polymer/ml and 500 µg polymer/ml) were added to the macrophage cultures and duplicate samples were prepared. After 24 h, 48 h and 72 h of incubation at 37°C, the cell culture supernatants were harvested and stored at -20°C until they were analyzed for their nitric oxide and cytokine load. 6.2.3 In vivo inflammatory response study To study the inflammatory potency of the nanoparticles in vivo, naïve C57BL/6 mice (5 mice per experimental group) received a intraperitoneal injection of 100 µl TP-NRs (50 µg polymer/mouse). Ip injection of PBS and a sub-lethal dose of LPS (50 µg/mouse) were used as negative and positive control, respectively. After 6 hours mice were euthanized and peritoneal macrophages and blood samples were collected. Peritoneal macrophages were cultured at 37°C and after 48 hours the supernatant was analyzed for the detection of the inflammatory cytokines IL-6, IL-1β and TNF- and for NO. Similarly, the cytokines were evaluated in the serum. 6.2.4 Cytokine analysis and NO detection Concentrations of TNF- (DuoSet ELISA Development System, R&D Systems), IL-6 and IL-1 (Pharmingen, BD Biosciences) in serum and cell supernatants were determined by sandwich ELISA as recommended by the suppliers. The amount of nitric oxide (NO) in the cell supernatants was measured by detecting the more stable NO-derivative nitrite (NO2-). Nitrite quantification in cell supernatants was assayed by a standard Griess reaction [8]. Therefore, 100 μl of cell-free supernatants were added to an equal volume of Griess reagent (0.5% sulfanilamide and 0.05% N-1 naphthylethylenediamine hydrochloride in 2.5% phosphoric acid). After a 10-minute incubation at room temperature absorbance at 540 nm was recorded. A standard curve was generated with known concentrations of NaNO2 in culture medium. All reagents for the Griess reaction were obtained from Sigma. 101 Chapter 6 6.2.5 Immunization of animals Naive C57BL/6 mice were split into 6 groups each with 4 animals per experimental group. Mice in group 1 received an intravenous (iv) injection of 200 µl sterile PBS, mice in group 2 received an iv injection of 200 µl sterile PBS containing NP+TP+Tsx (final concentration of 666 µg polymer/mouse, 60 µg TP/mouse and 5 µg Tsx/mouse), mice in group 3 received an iv injection of 200 µl NP+TP-Tsx (final concentration of 666 µg polymer/mouse and 60 µg TP/mouse), mice in group 4 received an iv injection of 200 µl NP-TP-Tsx (final concentration of 666 µg polymer/mouse), mice in group 5 received an iv injection of 200 µl sterile PBS containing free TPE.coli (final concentration of 60 µg TP/mouse) and mice in group 6 received an intraperitoneal (ip) injection of 200 µl sterile PBS containing free TPE.coli bound to the adjuvant Alum (final concentration of 60 µg TP/mouse). Imject® Alum (Pierce) is a solution of aluminium hydroxide and magnesium hydroxide and was mixed with TPE.coli according to the manufacturer instructions. After the initial injection, 2 boost injections were performed with every time ten days in between each injection. Several hours before each injection, approximately 100 µl blood was collected via orbital blood extraction of each animal and blood was stored at 4°C. The samples were then centrifuged and serum was stored at -20°C until it was further analyzed. 6.2.6 Antibody detection Immunized mice sera were analyzed for specific anti-TP or anti-Tsx antibodies by ELISA. Briefly, microtiter plates (MaxiSorp®, Nunc) were each coated with 1 µg/ml TPE.coli or Tsx in 0.1 M NaHCO3 pH 8.2 at 4°C overnight. The wells were washed and residual protein binding sites were blocked by adding 200 µl/well of 10% fetal calf serum in PBS for 2 hours at room temperature. Afterwards serial dilutions of sera in blocking buffer were added and incubated overnight at 4°C. Bound specific antibodies were detected by using a goat anti-mouse antibody conjugated to Horse Radish Peroxidase (HRP) (Sigma). 102 Immunogenicity of the nanoreactors After adding the substrate tetramethylbenzidine (diluted in a substrate-buffer containing H2O2) the reaction was stopped after three minutes with 50 µl of 0.1 N H2SO4 and the optical density was measured at 405 nm. In between each step the wells were washed extensively with PBS containing 0.05% Tween. Uncoated plates were treated in exactly the same way as controls to detect background signal. 6.3 Results 6.3.1 Ex vivo macrophage response study An important prerequisite of enzyme delivery devices for the use in enzyme replacement therapy is that they have long circulation times within the bloodstream and do not alert the immune system or induce overt inflammation. An important factor in the clearance of foreign compounds from the circulation and the initiation of inflammatory responses is their recognition by the mononuclear phagocyte system, most notably macrophages [9]. Macrophages are cells that stimulate other immune cells and remove foreign materials by ingesting them through a process known as phagocytosis. The rapid sequestration of intravenously injected nanoreactors from the blood by macrophages would be problematic for their application as long circulating devices in enzyme replacement therapy. Therefore, it is very important that TP-NRs are not recognized by macrophages and are immunologically inert. To investigate this, the impact of TP-NRs on the inflammatory status of macrophages was tested. For ex vivo studies the effect of nanoreactors on thioglycolate-elicited peritoneal macrophages isolated from C57BL/6 mice was analyzed [10]. Cells were plated in complete RPMI medium and incubated at 37°C. After 4 hours of adhesion, non-adherent cells were washed away, leaving almost pure (> 98%) macrophages in the plate. Half of these cells were pre-stimulated with 100 U/ml of the known macrophage activating cytokine interferon- (IFN) to prime the macrophages [11]. The following day, different concentrations of TP-NRs (50 µg polymer/ml, 100 µg polymer/ml and 500 µg polymer/ml) were added 103 Chapter 6 to the macrophage cultures and duplicate samples were prepared to assess the induction of inflammatory mediators. Plain polymeric nanoparticles without porin and without enzyme (NP-TP-Tsx) and nanoreactors without porin but with enzyme (NP+TP-Tsx) were used as controls and added at the same concentrations as the complete TP-NRs (NP+TP+Tsx). The free bacterial enzyme TPE.coli (5 µM) was added to investigate the shielding effectiveness of encapsulation. PBS was the negative control, while the known pro- inflammatory bacterial compound lipopolysaccharide (LPS) (100 ng/ml) served as positive control. In this context it is very important to note that all the nanoparticle preparations used in this experiment need to be clean of LPS, in order to exclude LPS dependent immune responses [12]. Given that Tsx is a Gramnegative bacterial outer membrane protein, the risk of LPS contamination, a potent endotoxin, is considerable. Therefore, nanoreactor samples used in this study were routinely checked by the LAL-assay to determine the LPS content (see also Chapter 3). Only the samples with an endotoxin unit lower than 50 EU/ml were used. After 24 h, 48 h and 72 h the cell culture supernatants were harvested and analyzed via ELISA for the presence of the inflammatory cytokines tumor necrosis factor- (TNF-) and interleukin-6 (IL-6) [13]. These cytokines are produced by activated macrophages and are the principal mediators of the acute phase immune response. Another indicator for inflammatory macrophage activation is nitric oxide (NO) production [14]. The results show that none of the nanoparticles tested (NP-TP-Tsx, NP+TPTsx, NP+TP+Tsx) induce NO, TNF- or IL-6 production by naïve macrophages (Figure 6-1). Interestingly, the free enzyme significantly provoked the secretion of all three inflammatory mediators. This means that the encapsulation of the enzyme in polymer particles indeed protects the protein from the immune system. Additionally, these results unequivocally indicate that the polymers possess good stealth properties towards peritoneal 104 Immunogenicity of the nanoreactors macrophages and hence are not recognized by naïve macrophages as foreign molecules. However, on IFN- primed macrophages, complete TP-NRs (NP+TP+Tsx) dosedependently elevate the amounts of NO, TNF- and IL-6 produced (Figure 61). Since no significant responses were found in the other conditions (NP-TPTsx, NP+TP-Tsx) we can conclude that the induction of inflammatory mediators is porin-dependent. Importantly however, TP-NR-induced TNF- levels and IL-6 levels are much lower as compared to the positive control and the free enzyme. This indicates that even on pre-activated macrophages the inflammatory effect of the nanoreactors is very limited; again highlighting that encapsulation efficiently shields the inflammatory potency of free TP. 105 Chapter 6 Figure 6-1: Macrophage response to different compositions of nanoparticles after 24 h, 48 h and 72 h of incubation. PBS was the negative control, while free TPE.coli (5 µM) and LPS (100 ng/ml) served as positive controls. (A) NO production (measured as the more stable NO-derivative NO2-) of naive macrophages, (B) NO production (measured as the more stable NO-derivative NO2-) of IFN- primed macrophages, (C) TNF- production of naive macrophages, (D) TNF- production of IFN- primed macrophages, (E) IL-6 production of naive macrophages, (F) IL-6 production of IFN- primed macrophages. 106 Immunogenicity of the nanoreactors 6.3.2 In vivo macrophage response study The inflammatory potency of the nanoparticles was also investigated in vivo. Therefore, naive C57BL/6 mice (5 mice per experimental group) received a intraperitoneal injection of 100 µl TP-NRs (final concentration of 500 µg polymer/ml). Intraperitoneal injection of PBS and a sub-lethal dose of LPS (50 µg/mouse) were used as negative and positive control, respectively. After 6 h mice were euthanized and peritoneal macrophages and blood samples were collected. Peritoneal macrophages were cultured at 37°C and after 48 h the supernatant was analyzed for the detection of the inflammatory cytokines IL6, IL-1β and TNF- and for NO. Similarly, the cytokines were evaluated in the serum. Only low amounts of IL-6, IL-1β and TNF- could be found in macrophage-supernatant and in serum of mice treated with TP-NRs, while high concentrations were detected in mice injected with LPS (Table 6-1). Table 6-1: Concentrations of NO2- and inflammatory cytokines produced by peritoneal macrophages (PECs) and serum after ip administration of PBS, TP-NRs (500 µg/ml NP+TP+Tsx) and LPS (50 µg/mouse) in mice. (*n.e. = not evaluated) NO2- produced by PECs (µM) NO2- produced in serum (µM) TNF- produced by PECs (pg/ml) TNF- produced in serum (pg/ml) IL-6 produced by PECs (pg/ml) IL-6 produced in serum (pg/ml) IL-1 produced by PECs (pg/ml) IL-1 produced in serum (pg/ml) PBS TP-NRs LPS 0.0 1.5 >100 n.e.* n.e.* n.e.* 147.5 81.9 560.2 0.0 0.0 324.0 7298.3 7668.8 28666.0 98.6 45.5 38811.0 7.6 0.0 1164.1 0.0 0.0 183.5 107 Chapter 6 6.3.3 Antibody response study Other important components of the immune system that are involved in the removal of pathogens from the circulation are antibodies. The formation of antibodies against antigenic epitopes of the nanoreactors (and subsequent antigen-binding) can contribute to altered nanoparticle biodistribution in the form of rapid clearance. The antibody-coated particle can either directly be recognized by phagocytes via binding to Fc-receptors or they can activate the complement system leading to complement receptor-mediated phagocytosis. The complement system consists of different proteins (opsonins) that bind to the antigen-antibody complex. Subsequently, this system triggers a cascade of biochemical cleaving reactions resulting in the removal of the antigen from the circulation. Since the nanoreactors contain bacterial proteins, there is a realistic chance that upon contact with the blood, production of antibodies against these foreign proteins is induced. Nevertheless, as long as the nanoreactors remain stable in the bloodstream, the proteins are in theory shielded by the polymer and not (completely) exposed to the environment. The longer the nanoreactors can escape from recognition by the immune system, the longer they can provide their enzymatic activity in situ and the more promising their use in enzyme replacement therapy will be. As such, the investigation of antibody responses against the nanoreactors is an important issue in evaluating their non-immunogenicity and shielding capacity. For this study, again different formulations of nanoparticles were prepared: nanoparticles with enzyme and porin (NP+TP+Tsx), nanoparticles with enzyme without porin (NP+TP-Tsx) and empty nanoreactors (NP-TP-Tsx). The size (via DLS), enzyme activity (via spectrophotometer) and encapsulation efficiency (via comparative SDS-PAGE analysis) of the nanoreactors was determined as previously described. The nanoparticles containing enzyme are around 200 nm in diameter, are enzymatically active and show an encapsulation efficiency of 10% resulting in a final enzyme concentration of 0.6 µg/µl. The empty nanoparticles are considerably smaller, with a diameter around 100 nm, and 108 Immunogenicity of the nanoreactors obviously lack enzymatic activity. In addition, all samples were confirmed to be endotoxin-free (via the LAL-test) before administration in vivo. PBS was used as negative control and free TP and TP adsorbed to the adjuvant Alum were used as positive controls. This last control was used because adding an adjuvant to the antigen stimulates an improved immune response compared to the antigen alone [15]. 6 groups of C57BL/6 mice (4 mice per experimental group) were immunized with the different nanoreactor formulations and control samples, as schematically represented in Figure 6-2. At different time points blood samples were collected and the sera of the immunized animals were analyzed for the presence of specific antibodies against the enzyme TP or the porin Tsx using ELISA. Figure 6-2: Schematic representation of the followed immunization scheme and the protocol of the antibody response study. The results clearly show elevated anti-TP antibody levels in mice immunized with nanoparticles containing TP (Figure 6-3). More surprisingly, the amount of specific TP antibodies found in mice immunized with the nanoreactors was even higher than the antibody titer detected in mice treated with the free TP. Figure 6-4 demonstrates visibly that the responses to the TP containing nanoparticles are stronger than the responses to the free TP. They are even in 109 Chapter 6 the same range than the responses to the free enzyme in complex with the adjuvant Alum. These results are rather unexpected and demonstrate that the bacterial enzyme is not protected against antibody response by the nanoparticle at all. Although the enzyme is strongly associated with the particle, it is clearly not shielded from the environment in the nanoreactor samples used for this study. Hence, the particle appears to act as an adjuvant in presenting the enzyme to the immune system. The adjuvant effect of these nanoreactors is in strong contrast with the envisioned shielding effect and questions their composition and thereby their therapeutic potential. Although the antibody response to the enzyme is the most important characteristic to evaluate at this point, the antibody response to the porin was also measured in the sera of the immunized mice. The level of anti-Tsx antibodies detected in mice immunized with nanoparticles containing Tsx (NP+TP+Tsx) are slightly higher than the response measured in the mice immunized with the formulations lacking Tsx (Figure 6-4 and Figure 6-5). Nevertheless, this increase is rather low and does not significantly differ from the anti-Tsx response in mice that did not receive the antigen at all. Note that the amount of Tsx present in the administered dose of nanoreactors is considerably lower than the amount of TP (5 µg Tsx to 60 µg TP per mouse). However, from these data we can conclude that in response to a relevant dose of nanoreactors, the specific antibody response to the porin is rather low. 110 Immunogenicity of the nanoreactors Figure 6-3: ELISA results of TP specific antibody titers measured in sera of mice immunized with different formulations of nanoreactors. PBS acts as negative control, TP and TP + Alum act as positive controls. Nanoreactors containing the enzyme TP elicit higher antibody titers than free enzyme. The antibody response to the nanoreactors is in the same range than the response to the enzyme in complex with Alum adjuvant. Error bars represent mean + standard deviation (n = 4). Figure 6-4: ELISA results of serum samples after the second boost injection (30 days after the first injection). The upper horizontal row of plates are the uncoated plates, the middle row are the plates coated with Tsx and the lowest row are the plates coated with TP. The sera of the mice immunized with six different formulations (6 vertical rows) were added in different dilutions (in duplicate) and the presence of specific anti-TP or anti-Tsx antibodies was detected. 111 Chapter 6 Figure 6-5: ELISA results of Tsx specific antibody titers measured in sera of mice immunized with different formulations of nanoreactors. Nanoreactors containing the porin Tsx elicit slight increased antibody titers. Error bars represent mean + standard deviation (n = 4). 6.4 Conclusion and discussion The data from the ex vivo and in vivo macrophage activation studies show that the effects of TP-NRs on macrophage activation and concomitant inflammation are minor. Only a slight pro-inflammatory effect was seen on strongly IFN- primed macrophages. Importantly, no inflammatory effect is detected in vivo. Since there is only a limited diffusion of particles from the peritoneum to the bloodstream, the level of cytokines present in the serum after intraperitoneal injection of the nanoreactors is rather an indirect indication of the potent inflammatory response. However, strong inflammatory responses in the peritoneum are usually reflected in the blood as well. Since the ip injected nanoparticles do not induce high cytokine levels in the local peritoneal macrophages nor in the serum, we can safely conclude that the nanoreactors are immunologically inert towards macrophages. Clearly, the stealthy polymers are able to protect the enzyme from recognition by these 112 Immunogenicity of the nanoreactors phagocytotic cells. The slight inflammatory effect on IFN- primed macrophages in vitro is probably due to surface exposed parts of the bacterial porins that are recognized by macrophages. Previously, it was reported that the porins are mainly burried in the polymeric membrane that is two- to threefold thicker than conventional lipid bilayers [16, 17]. The immunogenic response to the porins can probably be minimized by choosing thicker PMOXA side-blocks that shield the incorporated porins better from the environment. These results, demonstrating that the TP-NRs induce no acute innate immune response after intraperitoneal injection, are a first important step in the immunologic characterization of the nanoparticles. These observations are in good agreement with the findings of Broz and coworkers [18] who also reported the absence of non-specific binding of PMOXA-PDMS-PMOXA nanoreactors to macrophages. Only when a specific ligand was attached to the nanoparticles, able to bind to the scavenger receptor A1 of the macrophages, receptor-specific binding and uptake could be demonstrated. This low nonspecific binding supports the stealth properties of the carrier and makes them promising candidates for use in drug delivery. In a next phase, the nanoreactors were administered intravenously and antibody responses were analyzed in the sera after several boost injections. Surprisingly, the antibody response to the nanoreactor-encapsulated-enzyme is higher than to the free enzyme. These results were rather unforeseen as they show the opposite outcome than expected. The nanoreactors even seem to act as an adjuvant increasing the level of specific anti-TP antibody production. Because it is generally accepted that physical association of antigen and adjuvant is required for adjuvant activity, the observed effect is probably not due to the release of free enzyme by fast degradation of the nanoreactors. More likely, the adjuvant effect of the nanoreactors is caused by nanoreactor-associated enzymes that are not nicely encapsulated inside the particles but rather exposed to the environment. As such, the nanoreactors present the enzyme as repeats on their surface acting as an antigenpresenting platform, thereby provoking higher antibody responses. A number of other studies also described adjuvant properties of nanoparticles [19-23]. 113 Chapter 6 For example, polymethylmethacrylate nanoparticles induce long-lasting antibody titers in HIV whole virus vaccine in mice. The antibody response was even 100 times higher than when a traditional adjuvant was used [24]. Obviously, for the development of vaccines the adjuvant effect is beneficial. However, since strong antibody responses will cause fast elimination from the circulation, such effect is not desirable for enzyme replacement therapy purposes. Furthermore, the antibody response will become stronger after several boost injections, hereby devastating the therapeutic effect of life-long needed enzyme replacement therapy. Taken together, the nanoreactors do not provoke acute inflammatory responses on macrophages, but they do induce a specific antibody response against the enzyme. These observations limit the potential use of these particular nanoreactors as enzyme delivery particles for enzyme replacement purposes significantly. Comprehensive studies to reveal the detailed composition of the nanoreactors are necessary to further understand the critical parameters that determine their antigenic and even adjuvant properties. 6.5 References [1] B.S. Zolnik, A. Gonzalez-Fernandez, N. Sadrieh, M.A. Dobrovolskaia, Nanoparticles and the immune system, Endocrinology, 151 458-465. [2] D.E. Owens, 3rd, N.A. Peppas, Opsonization, biodistribution, and pharmacokinetics of polymeric nanoparticles, Int J Pharm, 307 (2006) 93-102. [3] M.A. Dobrovolskaia, S.E. McNeil, Immunological properties of engineered nanomaterials, Nat Nanotechnol, 2 (2007) 469-478. [4] M.C. Woodle, C.M. Engbers, S. Zalipsky, New amphipatic polymer-lipid conjugates forming long-circulating reticuloendothelial system-evading liposomes, Bioconjug Chem, 5 (1994) 493-496. [5] S. Zalipsky, C.B. Hansen, J.M. Oaks, T.M. Allen, Evaluation of blood clearance rates and biodistribution of poly(2-oxazoline)-grafted liposomes, J Pharm Sci, 85 (1996) 133-137. [6] M.A. Dobrovolskaia, P. Aggarwal, J.B. Hall, S.E. McNeil, Preclinical Studies To Understand Nanoparticle Interaction with the Immune System and Its Potential Effects on Nanoparticle Biodistribution, Mol Pharm, (2008). [7] C. Nardin, Hirt, T., Leukel, J., Meier, W., Polymerized ABA Triblock Copolymer Vesicles, Langmuir, 16 (2000) 1035-1041. 114 Immunogenicity of the nanoreactors [8] L.C. Green, D.A. Wagner, J. Glogowski, P.L. Skipper, J.S. Wishnok, S.R. Tannenbaum, Analysis of nitrate, nitrite, and [15N]nitrate in biological fluids, Anal Biochem, 126 (1982) 131-138. [9] A.U. Daniels, F.H. Barnes, S.J. Charlebois, R.A. Smith, Macrophage cytokine response to particles and lipopolysaccharide in vitro, J Biomed Mater Res, 49 (2000) 469-478. [10] I. Catelas, A. Petit, R. Marchand, D.J. Zukor, L. Yahia, O.L. Huk, Cytotoxicity and macrophage cytokine release induced by ceramic and polyethylene particles in vitro, J Bone Joint Surg Br, 81 (1999) 516-521. [11] H.W. Murray, G.L. Spitalny, C.F. Nathan, Activation of mouse peritoneal macrophages in vitro and in vivo by interferon-gamma, J Immunol, 134 (1985) 1619-1622. [12] V.M. Hitchins, K. Merritt, Decontaminating particles exposed to bacterial endotoxin (LPS), J Biomed Mater Res, 46 (1999) 434-437. [13] A. Prabhu, C.E. Shelburne, D.F. Gibbons, Cellular proliferation and cytokine responses of murine macrophage cell line J774A.1 to polymethylmethacrylate and cobalt-chrome alloy particles, J Biomed Mater Res, 42 (1998) 655-663. [14] P. Tripathi, P. Tripathi, L. Kashyap, V. Singh, The role of nitric oxide in inflammatory reactions, FEMS Immunol Med Microbiol, 51 (2007) 443-452. [15] H.F. Stills, Jr., Adjuvants and antibody production: dispelling the myths associated with Freund's complete and other adjuvants, Ilar J, 46 (2005) 280293. [16] C. Nardin, W. Meier, Hybrid materials from amphiphilic block copolymers and membrane proteins, J Biotechnol, 90 (2002) 17-26. [17] V. Pata, N. Dan, The effect of chain length on protein solubilization in polymer-based vesicles (polymersomes), Biophys J, 85 (2003) 2111-2118. [18] P. Broz, S.M. Benito, C. Saw, P. Burger, H. Heider, M. Pfisterer, S. Marsch, W. Meier, P. Hunziker, Cell targeting by a generic receptor-targeted polymer nanocontainer platform, J Control Release, 102 (2005) 475-488. [19] B. Slutter, L. Plapied, V. Fievez, M.A. Sande, A. des Rieux, Y.J. Schneider, E. Van Riet, W. Jiskoot, V. Preat, Mechanistic study of the adjuvant effect of biodegradable nanoparticles in mucosal vaccination, J Control Release, 138 (2009) 113-121. [20] S. Ribeiro, S.G. Rijpkema, Z. Durrani, A.T. Florence, PLGA-dendron nanoparticles enhance immunogenicity but not lethal antibody production of a DNA vaccine against anthrax in mice, Int J Pharm, 331 (2007) 228-232. [21] T. Uto, X. Wang, K. Sato, M. Haraguchi, T. Akagi, M. Akashi, M. Baba, Targeting of antigen to dendritic cells with poly(gamma-glutamic acid) nanoparticles induces antigen-specific humoral and cellular immunity, J Immunol, 178 (2007) 2979-2986. [22] J. Kreuter, Nanoparticles as adjuvants for vaccines, Pharm Biotechnol, 6 (1995) 463-472. [23] A.K. Jain, A.K. Goyal, P.N. Gupta, K. Khatri, N. Mishra, A. Mehta, S. Mangal, S.P. Vyas, Synthesis, characterization and evaluation of novel triblock copolymer based nanoparticles for vaccine delivery against hepatitis B, J Control Release, 136 (2009) 161-169. 115 Chapter 6 [24] F. Stieneker, J. Kreuter, J. Lower, High antibody titres in mice with polymethylmethacrylate nanoparticles as adjuvant for HIV vaccines, Aids, 5 (1991) 431-435. 116 Chapter 7 Biodistribution of the nanoreactors Biodistribution of the nanoreactors 7 Biodistribution of the nanoreactors 7.1 Introduction The delivery of therapeutic proteins to the body is a challenge due to their limited circulation time, which prevents long-term therapeutic efficacy. During the past decades many efforts were made to develop appropriate carrier systems to enhance the size and thereby the circulation half-life of therapeutic drugs. Nanoparticles are attractive drug delivery candidates because they are large enough to avoid fast clearance by the kidneys (size-selective cut-off for glomerular filtration is approximately 60 kDa) [1] and small enough to prevent removal from the circulation by simple filtration of the first capillary barrier encountered. Nevertheless, the clearance behavior and tissue distribution of intravenously injected nanoparticles is also greatly influenced by their surface characteristics. Many conventional nanoparticles are rapidly removed from the bloodstream due to clearance by macrophages of the mononuclear phagocyte system (MPS), also called the reticuloendothelial system (RES). In particular Kupffer cells, which are specific macrophages from liver and spleen, are known to be involved in recognizing and eliminating nanoparticles [2]. This fast clearance strongly prevents their application in controlled drug delivery. However, surface engineering can lead to particles that can evade, at least to some extent, MPS uptake and thereby provoke prolonged residence in the blood. Coating the nanoparticle surface with a suitable hydrophilic polymer such as poly(ethylene glycol) (PEG) has been shown to confer long circulation properties [3]. The presence of the hydrophilic coating is thought to sterically stabilize the particles against opsonization and phagocytosis [4]. Except from PEG, also other hydrophilic polymers are found to be effective steric stabilizers. Amongst them are the oxazoline based polymers which are claimed by Zalipsky and Woodle to possess stealth properties to the same extend as PEG [5]. It has been shown that the poly(methyl-oxazoline) or PMOXA polymer, conjugated to a lipid, is able to convey long circulation and low 119 Chapter 7 hepatosplenic uptake [6]. Despite these initial encouraging results, it is not verified further whether the usefulness of the PMOXA polymer also extends to other applications. In this chapter, the in vivo behavior of the new thymidine phosphorylase containing PMOXA-PDMS-PMOXA nanoreactors is investigated. The biodistribution of such complex nanoparticle formulations (including polymer, enzyme and porine) is not yet known and it is very important to unravel their in vivo fate. This is not only of high relevance for their potential use in enzyme replacement therapy, but also for the further exploitation of the PMOXA-PDMSPMOXA particles as drug delivery devices in general. 7.2 Materials and methods 7.2.1 Nanoreactor synthesis and characterization The TPE.coli (thymidine phosphorylase from E.coli) encapsulating nanoreactors (TP-NRs) were constructed via the solvent evaporation method as described in detail in section 3.2.3 [7]. Radiolabeling was achieved, either by labeling the nanoparticles directly after their production process, or by incorporating radiolabeled enzyme, as will be explained in the next sections. 7.2.2 Radiolabeling of particles with technetium-99m The labeling of the nanoreactors with technetium (99mTc) was performed after their production, on the hydroxyl end-groups of the polymer, according to Wunderlich et al. [8]. Briefly, 200 µl of the nanoreactor suspension (6.6 mg polymer/ml) was mixed with 20 µl of freshly prepared SnCl2 dissolved in 0.05 M HCl (0.1 mg/ml). After shaking for 15 minutes (at room temperature), 200 µl of 99m TcO4 (± 2 mCi or 74 MBq) was added and the vial was shaken again for 15 minutes. 99m Tc pertechnetate was obtained from a 99 Mo/99mTc radionuclide generator (Drytec, GE Healthcare). Stannous chloride is needed for the reduction of pertechnetate to create reactive technetium. After the radiolabeling of the nanoreactors, the radiolabeling yield and stability was 120 Biodistribution of the nanoreactors determined by instant thin layer chromatography on silica gel strips (ITLC-SG) (Pall Life Sciences) in physiologic H2O (0.9% NaCl) as solvent. This quality control was performed directly after and several hours after the radiolabeling. A drop of the labeled sample was deposited on one end of the strip and the strip was placed vertically in a beaker containing the solvent. The solvent was allowed to migrate to the end of the strip. Subsequently, the strips were removed and cut in half, and the radioactivity of each section was measured in a gamma counter. Free unreacted 99m Tc will migrate to the solvent front while 99m Tc-labeled nanoreactor stays at the application point. The labeling-efficiency was calculated in percentage as: (counts at application point / (counts at application point + counts at solvent front)) * 100. An aliquot of the particles was analyzed (after the radioactive decay) via DLS (size) and SDS-PAGE (encapsulation efficiency) in order to confirm their integrity and quality after labeling. 7.2.3 Radiolabeling of enzyme with iodine The enzyme thymidine phosphorylase was labeled with Iodine-123 (I123) and Iodine-125 (I125) according to the standard iodination protocol for proteins using iodogen as oxidizing agent [9, 10]. Iodogen vials were prepared by dissolving 1 mg iodogen (1,3,4,6-tetrachloro-3,6-di-phenylglycouril) in 1 ml dichloromethane. 100 µl of this solution was added per glass vial and the solvent was allowed to evaporate until a dry film at the bottom of the vial was obtained. The iodogen vials were stored at -20°C until further use. For the labeling, 1 ml PBS and ±1 mCi I123 (GE Healthcare, UK) or I125 (MP Biomedicals LLC, California) was added to an iodogen coated vial. For labeling with I123 some cold NaI was added (5 µl of a 0.2 mg/ml stock solution) in order to lower the specific activity of the radionuclide. As such, the specific activity is in the same order of magnitude as the I125 specific activity. In a next step, 200 µl enzyme solution (1 mg TP dissolved in PBS) was added and the vials were gently agitated at 5-min time intervals over a period of at least 1 hour. Afterwards, the reaction was stopped by adding 10 µl Na2SO3 (0.15 mg/ml) and the sample was transferred to a clean glass vial. Quality control was 121 Chapter 7 performed via solid phase extraction by loading an aliquot of the labeled sample onto a C18 seppak column (Waters). After elution with PBS, the labeling-efficiency was calculated by comparing the amount of radioactivity on the column (containing the labeled protein) with the flow through (containing the free iodine). The labeled enzyme solution was purified further via several passages through an Ag-filter (Millipore), which selectively captures the free iodine. 7.2.4 Animal guidelines Wistar rats and C57BL/6 mice (both from Harlan, Horst, The Netherlands) were used nanoparticles to study and the labeled biodistribution of free Minimum enzyme. labeled enzyme-containing three animals per experimental group were used. During the entire study, the animals were housed in stainless steel cages with sawdust bedding. They were kept at an average room temperature of 24°C, a relative humidity of 50% and a 12-hour day/night cycle. The study protocol was approved by the Ethics Committee for animal studies of the Vrije Universiteit Brussel. Guidelines of the National Institute of Health Principles of Laboratory Animal Care were followed. 7.2.5 Dynamic planar gamma camera imaging procedure For in vivo dynamic planar gamma camera imaging studies, the animals were anesthetized by intraperitoneal injection of pentobarbital (50 mg/kg bodyweight) (CEVA, Brussels, Belgium). The animals were placed on the collimator surface of a gamma camera and a tail vein intravenous (iv) injection of 99m Tc labeled nanoparticles (TP-NRs-99mTc, ±10 mCi/animal) (half-life = 6 hours; energy = 141 keV) or I123 labeled enzyme (TP-I123, ±1 mCi/animal) (half-life = 13 hours; energy = 159 keV) was performed (n = 3 per tracer). Images were acquired for 65 minutes. Directly after the imaging period, the animals were sacrificed by a lethal dose of sodium pentobarbital. 122 Biodistribution of the nanoreactors 7.2.6 Image acquisition and processing Images were acquired in prone position using a gamma camera (e.cam180, Siemens Medical Systems) mounted with a low energy high resolution collimator. A total of 480 frames were acquired in two steps: 300 frames of 1 s followed by 180 frames of 20 s. The energy window was set at 20% around 141 keV for 99m Tc and 20% around 159 keV for I123. All images were processed using AMIDE software [11]. Regions of interest (ROIs) were drawn around the liver, spleen, lung, heart (left ventricle) and bladder. Also, one ROI encompassing the total body was specified. All images were analyzed using the same ROIs. Average counts per minute (cpm) were measured in all frames and corrected for injected activity based on the average cpm in the total body ROI for each rat. Time activity curves were generated for the different organs. 7.2.7 Ex vivo analysis For long-term ex vivo biodistribution studies, an iv injection of I125 labeled nanoreactors (NP-TP-I125) or I125 labeled free enzyme (TP-I125) (±100 µCi) (half-life = 59.6 days; energy = 20 to 40 keV) was performed. The animals were anesthetized via inhalation of isoflurane during the iv injection of the labeled compounds. Bloodsamples were collected using microcapillaries at different time-points after injection to obtain a time-activity blood curve. The animals were sacrificed 1 hour or 3 days after the injection by intravenous administration of a lethal dose of pentobarbital (CEVA, Brussels, Belgium). Major organs/tissues (liver, spleen, lung, heart, kidney and small intestines) and blood were collected and weighted. The radioactivity content of the blood samples and the dissected organs and tissues was measured in a gamma counter. Tissue and organ uptake was calculated as percentage of injected activity (IA) per tissue or organ corrected for decay. 123 Chapter 7 7.2.8 Statistical analysis The amount of NP-TP-I125 and TP-I125 in the blood was expressed as average ± standard deviation (in % injected activity/total blood volume or %IA/TBV) and compared with a Mann-Whitney test using Graphpad Prism. The test was considered statistically significant when the P-value was less than 0.05. 7.3 Results 7.3.1 Kinetic analysis of short-term biodistribution A short-term biodistribution analysis of the nanoreactors and the free enzyme was performed using in vivo dynamic planar gamma camera imaging with the following constructs: TP-NRs-99mTc and TP-I123. The labeling-efficiency of the TP-NRs-99mTc construct was 98%. The stability of the labeled particles was tested in serum and their size (diameter around 200 nm) was confirmed via DLS. For the TP-I123 construct, a labeling-efficiency of more than 90% was obtained after the iodination reaction. Three rats per experimental group received an iv injection of TP-NRs-99mTc or TP-I123 followed by dynamic planar imaging during 65 minutes. During the imaging period a clear difference in biodistribution between the enzyme containing nanoreactors and the free enzyme was observed. The free enzyme was almost completely cleared by the hepatobiliary system into the intestines and to a lesser extent by the kidneys. A slight signal was also found in the thyroid gland, indicating a fraction of free I123 released from the enzyme. In contrast, the TP containing nanoreactors were trapped in a large scale in the liver and the spleen (Figure 7-1). These results were confirmed by analyzing the time activity curves for the different organs (Figure 7-2). These curves were obtained by processing the images using the AMIDE software. 124 Biodistribution of the nanoreactors Figure 7-1: Gamma-camera images of a rat 5, 10, 20, 40 and 60 minutes after iv injection of TP-I123 and TP-NRs-99mTc. Each image was scaled to its own maximum pixel value. 125 Chapter 7 Figure 7-2: Kinetic plots of the % injected activity (% IA) over time in different organs after iv injection of TP-I123 and TP-NRs-99mTc. The accumulation of the nanoreactors in liver and spleen suggests interaction with the mononuclear phagocyte system (MPS). However, the radioactive signal in these organs is rather high: 60% for liver and 10% for spleen of the total injected activity. It is possible that the obtained biodistribution pattern of the nanoparticles is influenced by the presence of colloidal particles of tin 126 Biodistribution of the nanoreactors oxides which can be formed by the reduction of technetium by stannous chloride during the radiolabeling process (and which cannot be resolved and quantified via ITLC). Such interference of colloidal tin was also reported by others [12] and is correlated to accumulations in liver and spleen due to the absorbance of the nanoreactors to the surface of the colloidal aggregates. As such, the organ uptake values can be slightly over-estimated. To circumvent these problems efforts were made to label the nanoreactors by encapsulating the labeled enzyme TP-I123. Unfortunately, this was practically not achievable due to the too short half-life of this radioligand in comparison to the time needed for the production process of the nanoreactors. 7.3.2 Long-term biodistribution analysis To perform long-term biodistribution studies, the enzyme was linked to the radioisotope I125, which has a half-life of 59.6 days. A labeling-efficiency around 80% was obtained after labeling the enzyme with I125. Purification on an Ag-filter resulted in a labeling-efficiency high enough (> 90%) for in vivo use. Subsequently, the labeled enzyme was used to construct nanoreactors. The difference in biodistribution between the free TP (TP-I125) and the TP containing nanoreactors (NP-TP-I125) was investigated 1 hour and 3 days after a single bolus iv injection in rats. Since the energy of the gamma radiation emitted from I125 is too low for gamma-camera detection, the biodistribution was analyzed ex vivo via dissection of the major organs and tissues followed by gamma counting. Also blood samples were collected at different time-points and analyzed for their radioactive load. The results indicate that the biodistribution of free and encapsulated TP differs significantly in Wistar rats (Figure 7-3). Already 1 hour after iv injection, a considerable higher value for the nanoreactors is observed in the liver and the spleen compared to the free enzyme. After 3 days, the difference in hepatosplenic uptake between the two formulations is even more pronounced. Importantly, no significant difference in blood-load was observed between TPI125 and NP-TP-I125. 127 Chapter 7 Figure 7-3: Biodistribution of TP-I125 (blue) and NP-TP-I125 (red) 1 hour (upper graph) and 3 days (lower graph) after iv injection in Wistar rats. Three animals per experimental group were analyzed and means and standard deviations are displayed in % injected activity (IA) per organ or per tissue. 128 Biodistribution of the nanoreactors The blood-pool activity was assessed at different time-points after the injection and the blood-clearance curves (Figure 7-4) indicate that the nanoreactors are in the end as rapidly cleared from the circulation as the free enzyme. During the first hour after administration, the blood values are consistently higher for the nanoreactors compared to the free enzyme (p < 0.05). However, from 1 hour onwards, the two blood curves are very similar, underlining the fast clearance of the nanoreactors from the blood circulation. Figure 7-4: The blood-clearance curves of TP-I125 (dashed blue line) and NP-TP-I125 (full red line). Rats were given a single bolus iv injection. The % injected activity (IA) per total body volume (TBV) was plotted over time. The TBV was calculated as 7% of the total body weight. Three animals per experimental group were analyzed and means and standard deviations are plotted. The result is statistically significant, because the P value based on a Mann-Whitney test, is 0.023 (< 0.05). In order to investigate whether similar results could be observed in another species, the long-term biodistribution study was repeated in C57BL/6 mice. Three days after iv injection of TP-I125 and NP-TP-I125, the main organs and tissues were analyzed for their radioactive load using a gamma counter. As shown in Figure 7-5, exactly the same results could be obtained as in the rats. Also in mice, the liver and spleen values of NP-TP-I125 are significantly higher than those for TP-I125, while no difference in blood-pool activity was observed. 129 Chapter 7 Figure 7-5: Biodistribution of TP-I125 (blue) and NP-TP-I125 (red) 3 days after iv injection in C57BL/6 mice. Three animals per experimental group were analyzed and means and standard deviations are displayed in % injected activity (IA) per organ or per tissue. 7.4 Conclusion and discussion In this chapter, we describe the in vivo biodistribution of the new thymidine phosphorylase containing nanoreactors (TP-NRs). By encapsulating the enzyme in polymeric nanoreactors, we aim to increase the blood circulation half-life of the therapeutic enzyme, which is an important prerequisite for their use in enzyme replacement therapy. First, we investigated the short-term biodistribution using in vivo gamma camera imaging. Taken together, the imaging studies with the I123 labeled enzyme revealed an expected fast clearance of the free enzyme via the hepatobiliary system. The 99m Tc labeled nanoreactors however were also cleared rapidly from the blood circulation and were predominantly trapped in the liver and the spleen. This suggests uptake of the nanoreactors by the mononuclear phagocyte system (MPS). The absolute uptake values in liver and spleen might however be somewhat overestimated due to the formation of colloidal oxidative 130 99m Tc species formed Biodistribution of the nanoreactors during the labeling procedure. Nevertheless, the technetium labeling of the nanoreactors is in general a rather mild process used to label other nanoparticles as well [13]. 123 encapsulation of the I Attempts to image the nanoreactors via labeled enzyme failed due to practical reasons. From the ex vivo biodistribution studies with I125 labeled nanoreactors and I125 labeled enzyme in rats, again, a clear difference in biodistribution between the free enzyme and the nanoreactor-associated enzyme was observed. The nanoreactors accumulated mainly in liver and spleen, while the values in these organs were clearly lower for the free enzyme. Here, the absolute organ uptake values are considerably lower in comparison with the 99m Tc experiment. This is most likely caused by inevitable release of the radionuclide by in vivo de-iodination and possibly also a partial release of the enzyme from the particle. Also here, no difference in blood circulation half-life could be demonstrated between the free enzyme and the nanoreactor-associated enzyme. Moreover, these results could be confirmed in C57BL/6 mice. All together, these data indicate that the nanoreactors do not possess the ideal size and/or surface characteristics to evade hepatosplenic uptake and to prolong the circulation time of the incorporated enzyme significantly. As such, the potential advantageous effect of the nanoreactors for use in enzyme replacement therapy is strongly compromised. It has been emphasized that the clearance behavior and tissue distribution of intravenously injected drug carriers are greatly influenced by their surface characteristics and their size. The surface characteristics determine whether the nanoparticles are recognized and taken up by phagocytotic cells of the MPS. Since many nanoparticles do not possess ideal shielding properties, different approaches have been developed to circumvent rapid hepatosplenic uptake. One approach is to administer large doses of placebo carriers in an effort to impair the phagocytic capacity of macrophages, thereby allowing subsequently administered particles to remain in systemic circulation for prolonged periods [14]. Another strategy is transient apoptotic destruction of liver and spleen macrophages by prior administration of gadolinium chloride [15]. These strategies, although successful in experimental models, have however little 131 Chapter 7 justification in practice as they suppress the essential defense system of the body. The size and deformability of nanoparticles plays another critical role in their clearance by the liver and spleen. Ideally, the size of an engineered longcirculating particle should not exceed 200 nm. If larger, the particle must be deformable enough to bypass splenic filtration [16]. In addition, there is also a clear-cut relationship between particle size and the extent to which they reach the hepatic parenchyma. As a rough approximation, the size of the particles should be in the range of 120 to 200 nm in diameter to substantially avoid particle trapping in the liver. To conclude, our nanoreactors of 200 nanometer size covered with the stealthy polymer PMOXA had promising characteristics to act as longcirculating devices in theory. Nevertheless, in practice we have observed that they are not able to evade hepatosplenic uptake and to prolong circulation times in the blood significantly. Whether this behavior is due to their surface characteristics (complex core of porin/PMOXA-PDMS-PMOXA), their size (just too big), their rigidity or to a combination of several parameters requires further investigation. 7.5 References [1] A. Edwards, B.S. Daniels, W.M. Deen, Ultrastructural model for size selectivity in glomerular filtration, Am J Physiol, 276 (1999) F892-902. [2] E. Sadauskas, H. Wallin, M. Stoltenberg, U. Vogel, P. Doering, A. Larsen, G. Danscher, Kupffer cells are central in the removal of nanoparticles from the organism, Part Fibre Toxicol, 4 (2007) 10. [3] A.L. Klibanov, K. Maruyama, V.P. Torchilin, L. Huang, Amphipathic polyethyleneglycols effectively prolong the circulation time of liposomes, FEBS Lett., 268 (1990) 235-237. [4] D.D. Lasic, F.J. Martin, A. Gabizon, S.K. Huang, D. Papahadjopoulos, Sterically stabilized liposomes: a hypothesis on the molecular origin of the extended circulation times, Biochim Biophys Acta, 1070 (1991) 187-192. [5] M.C. Woodle, C.M. Engbers, S. Zalipsky, New amphipatic polymer-lipid conjugates forming long-circulating reticuloendothelial system-evading liposomes, Bioconjug Chem, 5 (1994) 493-496. [6] S. Zalipsky, C.B. Hansen, J.M. Oaks, T.M. Allen, Evaluation of blood clearance rates and biodistribution of poly(2-oxazoline)-grafted liposomes, J Pharm Sci, 85 (1996) 133-137. [7] C. Nardin, Hirt, T., Leukel, J., Meier, W., Polymerized ABA Triblock Copolymer Vesicles, Langmuir, 16 (2000) 1035-1041. 132 Biodistribution of the nanoreactors [8] G. Wunderlich, T. Gruning, B.R. Paulke, A. Lieske, J. Kotzerke, 99mTc labelled model drug carriers - labeling, stability and organ distribution in rats, Nucl Med Biol, 31 (2004) 87-92. [9] T.B. Schulz, R. Jorde, H.L. Waldum, P.G. Burhol, Preparation of 125Ilabeled gastric inhibitory polypeptide, using water-insoluble 1,2,4,6tetrachloro-3 alpha,6 alpha-diphenylglycoluril (Iodo-gen) as the oxidizing agent, Scand J Gastroenterol, 17 (1982) 379-382. [10] G.B. Saha, J. Whitten, R.T. Go, Conditions of radioiodination with iodogen as oxidizing agent, Int J Rad Appl Instrum B, 16 (1989) 431-433. [11] A.M. Loening, S.S. Gambhir, AMIDE: a free software tool for multimodality medical image analysis, Mol Imaging, 2 (2003) 131-137. [12] T. Banerjee, A.K. Singh, R.K. Sharma, A.N. Maitra, Labeling efficiency and biodistribution of Technetium-99m labeled nanoparticles: interference by colloidal tin oxide particles, Int J Pharm, 289 (2005) 189-195. [13] M. Stevanovic, T. Maksin, J. Petkovic, M. Filipic, D. Uskokovic, An innovative, quick and convenient labeling method for the investigation of pharmacological behavior and the metabolism of poly(DL-lactide-co-glycolide) nanospheres, Nanotechnology, 20 (2009) 335102. [14] S.M. Moghimi, S.S. Davis, Innovations in avoiding particle clearance from blood by Kupffer cells: cause for reflection, Crit Rev Ther Drug Carrier Syst, 11 (1994) 31-59. [15] M.J. Hardonk, F.W. Dijkhuis, C.E. Hulstaert, J. Koudstaal, Heterogeneity of rat liver and spleen macrophages in gadolinium chloride-induced elimination and repopulation, J Leukoc Biol, 52 (1992) 296-302. [16] S.M. Moghimi, C.J. Porter, I.S. Muir, L. Illum, S.S. Davis, Non-phagocytic uptake of intravenously injected microspheres in rat spleen: influence of particle size and hydrophilic coating, Biochem Biophys Res Commun, 177 (1991) 861-866. 133 Chapter 8 Efficiency of the nanoreactors Efficiency of the nanoreactors 8 Efficiency of the nanoreactors 8.1 Introduction MNGIE patients are characterized by thymidine phosphorylase (TP) gene mutations in association with drastically reduced TP activity and strikingly elevated levels of the substrates thymidine and deoxyuridine in plasma. Circulating levels of these substrates range from 4 to 24 µmol/l in MNGIE patients, while their levels are undetectable in healthy persons [1, 2]. To characterize the disease and to evaluate new therapies, a MNGIE mouse model was generated by our collaborator Dr. Hirano and his team from the Department of Neurology at Columbia University Medical Center. Beside the enzyme TP, the murine uridine phosphorylase (UP) is also able to convert thymidine. Therefore, a double knock-out TP-/-UP-/- strain was constructed in order to establish a MNGIE mouse model [3]. These mice develop several, but not all, of the typical MNGIE features, such as TP deficiency, elevated thymidine and deoxyuridine levels, mtDNA depletion and respiratory chain defects in the brain. The relatively mild phenotype in the animal model in comparison to the human disease is probably due to a combination of two factors. First, the levels of nucleosides in tissues of the knock-out animals are only modestly increased (4- to 65-fold elevated) compared to the more than 100-fold increases of both deoxyuridine and thymidine in tissues of MNGIE patients. Second, the short lifespan of the animals may not be sufficient to develop severe alterations of mitochondrial DNA. In order to intensify the biochemical abnormalities and augment mitochondrial DNA alterations, the TP/- UP-/- mice are treated with exogenous thymidine, deoxyuridine, or both (supplemented in drink water). Nevertheless, since the concentrations of the toxic substrates in the knock-out mice are significantly higher than the untraceable low levels in naïve mice, the TP-/-UP-/- mice are a powerful model system to evaluate the potential effect of our TP containing nanoreactors as enzyme replacement devices. The effect of TP administrations in TP-/-UP-/- mice 137 Chapter 8 has not yet been investigated before, neither in a native, pegylated nor controlled release form. The only reported attempt of using enzyme replacement to treat MNGIE was performed by Moran and coworkers who reported the use of an erythrocyte entrapped thymidine phosphorylase to treat one MNGIE patient [4]. Although the substrate concentrations in the plasma decreased significantly, the patient died several weeks after the therapy due to pneumonia. Except from this preliminary and single case study, no further investigations in this area were performed. In this chapter, the efficiency of the nanoreactors as potential new enzyme delivery devices for the enzyme replacement therapy of MNGIE is evaluated. The aim of this study was to investigate whether the thymidine phosphorylase containing nanoreactors are able to reduce the toxic substrate concentrations of thymidine and deoxyuridine in vitro and in vivo using a TP-/-UP-/- mouse model. 8.2 Materials and methods 8.2.1 Production of nanoreactors The TPE.coli (thymidine phosphorylase from E.coli) encapsulating nanoreactors (TP-NRs) were constructed via the solvent evaporation method as described in detail in section 3.2.3 [5]. 8.2.2 In vitro assay and sample processing Plasma samples of a MNGIE patient (100 µl) were incubated with TP-NRs (25 µl containing 12 µg TP) for 30 minutes at 37°C. Free thymidine phosphorylase (25 µl containing 12 µg TP) and PBS (25 µl) were used as positive and negative control respectively. Subsequently, the samples were treated with perchloric acid (PCA) to a final concentration of 0.35 M to precipitate the proteins. After a short incubation on ice, the samples were centrifuged (10 minutes, 17900 g) and the supernatant was collected. This supernatant was analyzed immediately or stored at -20°C until further use. 138 Efficiency of the nanoreactors 8.2.3 HPLC analysis Separation of nucleosides was carried out through High Pressure Liquid Chromatography (HPLC) with gradient elution and UV detection, according to a previously described method with modifications [6-8]. A reversed-phase C18 column (Alltima C18, 100 Å pore size, 5 µm particle size, 250 mm x 4.6 mm, catalogue number 88056) protected by a guard column was used to separate the sample components. Injected samples (50 µl) were eluted over 60 minutes through a methanol-gradient buffered mobile phase. The gradient was obtained by employing a combination of 3 eluents; eluent A: 20 mM potassium phosphate pH 5.6, eluent B: methanol and eluent C: water. The elution gradient was programmed as described in Table 8-1. The column was kept at 30°C during the entire experiment and the absorption of the eluate was measured at 267 nm. Table 8-1: Gradient conditions of three eluents applied to separate nucleosides in plasma samples via HPLC. Eluent A = 20 mM potassium phosphate pH 5.6; eluent B = methanol; eluent C = water. Time (min) Flow (ml/min) % eluent A % eluent B % eluent C 0 1.5 100 0 0 5 1.5 100 0 0 25 1.5 82.6 17.4 0 26 1.5 0 0 100 30 1.5 0 0 100 31 1.5 0 100 0 35 2.0 0 100 0 45 2.0 0 100 0 46 1.5 0 100 0 47 1.5 0 0 100 50 1.5 0 0 100 51 1.5 100 0 0 60 1.5 100 0 0 139 Chapter 8 8.3 Results To investigate the in vitro efficiency of the TP containing nanoreactors, plasma of a MNGIE patient was incubated with the nanoreactors and the nucleoside concentrations were subsequently determined by HPLC. PBS and free TP were used as negative and positive control respectively. Quantification of thymidine and deoxyuridine levels was based on external standards which were treated in exact the same way as the plasma samples. The area under the peaks was correlated to the amount of thymidine or deoxyuridine via a corresponding standard curve. The identification of the peaks was based upon retention time and always confirmed by treatment of a second aliquot of each sample with a large excess of thymidine phosphorylase to eliminate the thymidine and deoxyuridine peaks. The results show a clear thymidine and deoxyuridine peak in untreated (only PBS was added) MNGIE plasma samples (Figure 8-1). From the standard curves (data not shown) the concentrations of thymidine and deoxyuridine were calculated as 8.0 µM and 13.1 µM respectively. These results correspond well to previous reported data (8.6 µM thymidine and 14.2 µM deoxyuridine as mean values from 25 MNGIE patients) [2]. After treatment with free TP, the thymidine peak disappears (< 0.05 µM) and the deoxyuridine peak decreases to a concentration of 0.6 µM (Figure 8-2). As expected the free enzyme is able to decrease the substrate concentrations efficiently in this in vitro setup. Moreover, we observe an increase of the thymine peak, which confirms substrate-product conversion. However, after treatment with the TP-NRs, only a slight decrease in substrate concentrations could be observed. Concentrations of 6.8 µM thymidine and 11.1 µM deoxyuridine were detected (Figure 8-3). Thus, the nanoreactors are not able to decrease the toxic substrate amounts significantly. Nevertheless, the nanoreactor formulation contains a significant amount of enzyme as verified by SDS-PAGE analysis (data not shown) and the same amount of enzyme was added to the plasma sample as was used in the free enzyme condition. 140 Efficiency of the nanoreactors In conclusion, the nanoreactor associated enzyme is not active enough to convert the substrates efficiently. Therefore, the in vivo experiments in the TP/- UP-/- mice were cancelled. Figure 8-1: HPLC diagram of plasma of a MNGIE patient pre-treated with PBS. The relevant peaks (thymidine, deoxyuridine and thymine) are identified by their typical retention times. 141 Chapter 8 Figure 8-2: HPLC diagram of plasma of a MNGIE patient pre-treated with free TP. The relevant peaks are identified by their typical retention times. Figure 8-3: HPLC diagram of plasma of a MNGIE patient pre-treated with TP-NRs. The relevant peaks (thymidine, deoxyuridine and thymine) are identified by their typical retention times. 142 Efficiency of the nanoreactors 8.4 Conclusion and discussion Unfortunately, the TP containing nanoreactors, as presented in their current composition, are not efficient in reducing the toxic substrate concentrations in MNGIE plasma in vitro. Therefore, the in vivo experiments in the TP-/-UP-/mice were cancelled at this point. Although the enzyme is clearly associated to the nanoparticle, it is not able to efficiently convert the substrates into its products. Probably, the enzyme is stuck into or onto the particle in a nonactive conformation. As such, these nanoreactors fail in their role of being active enzyme carriers and cannot be used in enzyme replacement therapy. In order to produce better enzyme containing nanoreactors, the construction and composition of the nanoreactors need to be further scrutinized critically. 8.5 References [1] M.C. Lara, M.L. Valentino, J. Torres-Torronteras, M. Hirano, R. Marti, Mitochondrial neurogastrointestinal encephalomyopathy (MNGIE): biochemical features and therapeutic approaches, Biosci Rep, 27 (2007) 151-163. [2] R. Marti, A. Spinazzola, S. Tadesse, I. Nishino, Y. Nishigaki, M. Hirano, Definitive diagnosis of mitochondrial neurogastrointestinal encephalomyopathy by biochemical assays, Clin Chem, 50 (2004) 120-124. [3] L.C. Lopez, H.O. Akman, A. Garcia-Cazorla, B. Dorado, R. Marti, I. Nishino, S. Tadesse, G. Pizzorno, D. Shungu, E. Bonilla, K. Tanji, M. Hirano, Unbalanced deoxynucleotide pools cause mitochondrial DNA instability in thymidine phosphorylase deficient mice, Hum Mol Genet, (2008). [4] N.F. Moran, M.D. Bain, M.M. Muqit, B.E. Bax, Carrier erythrocyte entrapped thymidine phosphorylase therapy for MNGIE, Neurology, 71 (2008) 686-688. [5] C. Nardin, Hirt, T., Leukel, J., Meier, W., Polymerized ABA Triblock Copolymer Vesicles, Langmuir, 16 (2000) 1035-1041. [6] R.A. Hartwick, S.P. Assenza, P.R. Brown, Identification and quantitation of nucleosides, bases and other UV-absorbing compounds in serum, using reversed-phase high-performance liquid chromatography. I. Chromatographic methodology, J Chromatogr, 186 (1979) 647-658. [7] R.A. Hartwick, A.M. Krstulovic, P.R. Brown, Identification and quantitation of nucleosides, bases and other UV-absorbing compounds in serum, using reversed-phase high-performance liquid chromatography. II. Evaluation of human sera, J Chromatogr, 186 (1979) 659-676. [8] R. Marti, Y. Nishigaki, M. Hirano, Elevated plasma deoxyuridine in patients with thymidine phosphorylase deficiency, Biochem Biophys Res Commun, 303 (2003) 14-18. 143 Chapter 9 Detailed evaluation of the nanoreactor build-up Detailed evaluation of the nanoreactor build-up 9 Detailed evaluation of the nanoreactor build-up 9.1 Introduction Although the first characterization of the thymidine phosphorylase containing nanoreactors looked promising, the latest data indicate that the activity and in vivo behavior of the nanoreactors is not like expected/hoped (see chapters 6, 7 and 8). Indeed, we observed a specific antibody response against the enzyme, hepatosplenic uptake and low enzymatic activity. These observations suggest that the nanoreactor composition is not like it used to be and indicate that the enzyme is not encapsulated properly in an active form inside the particles. Such a configuration is however an absolute prerequisite in order to shield the protein from proteolytic degradation and to allow the enzyme to catalyze the substrates that are transported to the inner cavity of the nanoparticles. The hampered enzyme activity and specific immune responses are clearly unwanted properties that limit the use of these nanoreactors in enzyme replacement therapy. Especially for this specific purpose, nanoreactors with better immunogenic and catalytic properties are needed. This chapter summarizes the efforts that were made to unravel the exact nanoreactor composition and to try to resolve the problems encountered. 9.2 Materials and methods 9.2.1 Production of nanoreactors The TPE.coli (thymidine phosphorylase from E.coli) containing nanoreactors were constructed via the solvent evaporation method as described in detail in section 3.2.3 [1]. Nanoreactors with varying porin concentrations (100 µg, 150 µg and 300 µg Tsx) were constructed and analyzed. As control sample, nanoparticles with enzyme without porin were used. In addition, nanoreactors 147 Chapter 9 with other enzymes; the human thymidine phosphorylase (TPhuman), the commercial Escherichia coli thymidine phosphorylase (TPE.coli,Sigma from Sigma) and the Trypanosoma vivax nucleoside hydrolase (TvNH), were prepared in exactly the same way and used as reference samples. 9.2.2 Characterization of TPE.coli containing nanoreactors The TPE.coli containing nanoreactors were analyzed via DLS (to determine their size and dispersity), via SDS-PAGE analysis (to determine the amount of enzyme and porin incorporated in the nanoreactors) and via spectrophotometry (to determine the enzymatic activity) as described in detail in chapter 3. 9.2.3 Activity measurements of TvNH containing nanoreactors The enzymatic activity (i.e. hydrolysis of nucleosides) of the TvNH containing nanoreactors was determined with inosine as substrate making use of a reducing sugar colorimetric assay as described by Parkin and Ranquin [2, 3]. Briefly, 150 µl nanoreactors were added to 1 mM inosine and the enzymatic reaction was stopped at different time-points by adding a CuSO4 solution. The Cu2+ is reduced to Cu+ by the reaction product ribose. This reduced Cu+ can than react 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 was measured at 450 nm. A standard curve with known ribose concentrations was used to determine the concentration of ribose formed in the nanoreactortreated samples under the assay conditions. 9.2.4 Protease treatment In order to investigate their proteolytic resistance, the nanoreactors were treated with proteinase K according to Axthelm et al. [4]. Proteinase K is a protease with a broad specificity and therefore very efficient in digesting proteins. 1 mg/ml proteinase K (stock solution of 20 mg/ml, Roche) was added to the nanoreactors and the mixture was incubated for 4 h at 37°C. Before and 148 Detailed evaluation of the nanoreactor build-up after the proteinase treatment, the protein content, size and activity of the samples were analyzed via SDS-PAGE, DLS and spectrophotometry respectively. 7.2.5 Surface pressure experiments For surface pressure experiments a MicroTrough S (Kibron Inc.) was used. The microwell-trough was filled with the aqueous sub-phase (3 ml), either PBS or PBS containing 1 M NaCl. A PMOXA-PDMS-PMOXA monolayer was formed by adding 2 µl of polymer dissolved in ethanol (10 mg/ml) at the air-liquid surface. After the formation of a stable monolayer with a surface pressure around 20 mN/m, proteins were added to a final concentration of 0.08 µM with a microsyringe at the bottom of the trough. Continuous recording of the surface pressure was performed. Prior to experiments the trough was thoroughly cleaned and the surface pressure detection wire probe was cleaned by flaming. The experiments were performed at room temperature. For protection of impurities the trough was housed in a Plexiglas cabinet. 9.2.6 Fluorescent labeling of thymidine phosphorylase To fluorescently label the enzyme TPE.coli, we used the amine reactive Alexa®555 dye (Invitrogen). The spectra of Alexa®555 (excitation/emission maxima ~555/565 nm) are an almost exact match to those of the Cy3 dye. The amine reactive succinimidyl ester of Alexa®555 forms a stable amide bond with amine groups on the protein. Before coupling the Alexa®555 dye, the enzyme was dialyzed to 0.1 M sodium bicarbonate buffer pH 8.3 and concentrated to a final concentration of 20 mg/ml. Alexa®555 was dissolved in DMSO to a final concentration of 10 mg/ml. Then, 1 ml of the TP 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 hydroxylamine pH 8.5. The non-bonded Alexa®555 was subsequently removed via gelfiltration on a superdex 75 high resolution column (Amersham Biosciences) with PBS as running buffer. 149 Chapter 9 9.3 Results 9.3.1 Enzyme activity versus porin/enzyme concentration In order to identify the cause of the low enzymatic activity of the nanoreactors, kinetic experiments were performed using nanoreactors with two different porin concentrations (150 µg and 300 µg Tsx). Nanoparticles encapsulating TPE.coli without porin Tsx were used as controls. The conversion of 900 µM thymidine was followed spectrophotometrically at 290 nm for 10 minutes. Theoretically, when more porin is added, the overall transport of the substrate across the polymeric membrane occurs faster and thus, the enzymatic activity of the nanoreactors should be higher. This was previously described by Ranquin et al. for TvNH containing nanoreactors [3]. For the TPE.coli containing nanoparticles however, an opposite trend is observed. Namely, the enzymatic activity of the nanoreactors increases with decreasing porin concentrations. Even more surprisingly, control samples without reconstituted channel proteins are also enzymatically active and their enzymatic activity is higher than the activity of the particles containing the membrane protein Tsx (Figure 9-1). Clearly, these results show that transport of thymidine across the membrane is not necessary for enzymatic activity. The presence of Tsx even seems to have a negative effect on the enzymatic activity. Furthermore, SDS-PAGE analysis shows that the encapsulation efficiency of TPE.coli decreases with increasing Tsx concentrations. The protein band representing the enzyme becomes slightly smaller when more Tsx is added (Figure 9-1). 150 Detailed evaluation of the nanoreactor build-up Figure 9-1: Left: Thymidine conversion catalyzed by TPE.coli containing nanoreactors with different porin concentrations measured as the decrease in optical density (OD) at 290 nm in function of time. Right: SDS-PAGE analysis of the nanoreactors showing the amount of TPE.coli (52 kDa) and Tsx (31 kDa) present in the samples. The protein molecular weight marker used is Mark 12 (Invitrogen). Subsequently, the initial rate of conversion of different concentrations of thymidine by nanoreactors containing enzyme TPE.coli and porin Tsx (100 µg) was measured spectrophotometrically at 290 nm (in 200 mM potassium phosphate pH 6.8 at 37°C). The results were fitted to the Michaelis-Menten equation (Origin software) to determine the catalytic parameters of the TPcontaining nanoreactors (Figure 9-2). As such, an apparent kcat,app of 0.073 ± 0.003 s-1 and a KM,app of 760 ± 69 µM were obtained. In comparison to the catalytic parameters of the free enzyme (kcat = 413.6 ± 71.1 s-1 and KM = 643.2 ± 268.7 µM ; see chapter 3), the enzymatic activity of the nanoreactors is extremely low. These results are also not in agreement with the results obtained by Ranquin et al. [3], where the KM,app increased tenfold, indicating that transport over the membrane was the rate limiting step. In case of the TP encapsulating nanoreactors, the KM,app is in the same range as the KM of the free enzyme. This again suggests that transport of thymidine is not necessary for enzymatic conversion. However, roughly estimated, the catalytic turnover 151 Chapter 9 of the enzymes is reduced by a factor 104. This finding probably reflects that most enzyme molecules are present in a catalytically inactive conformation. Figure 9-2: Initial product formation rate of nanoreactors (composed of enzyme TPE.coli and 100 µg Tsx) as a function of thymidine concentration. The data were fitted to a hyperbolic curve using Origin software to determine the apparent catalytic parameters kcat,app and KM,app. From these experiments we conclude that since the porins are not required for activity, the enzyme is not situated inside the particles. The enzyme is probably located or associated into or onto the polymeric membrane. Such association or partial insertion of enzymes in the membrane also explains the observed low activity. 9.3.2 Protease sensitivity of nanoreactors If the TP enzymes are associated with the polymeric membrane and located at least partially at the outside of the particle, they are prone to degradation by proteases. 152 Detailed evaluation of the nanoreactor build-up To verify whether the enzyme is solvent exposed and can be attacked by proteases, proteinase K was added to the nanoreactors (NP+TP+Tsx) and the mixture was incubated for 4 hours at 37°C. Nanoparticles with enzyme without porine (NP+TP-Tsx) and free enzyme (TP) were used as control samples. The SDS-PAGE analysis of the samples before and after proteinase K treatment demonstrates that the nanoreactor associated enzyme (either with or without porin) is not protected against proteolytic degradation (Figure 9-3). No protein band could be detected in the proteinase K treated samples. The nanoparticle associated enzyme is degraded to the same extend as the free enzyme. Also, no enzymatic activity could be measured after proteinase K treatment in none of the samples (data not shown). Analysis of the size distribution of the nanoparticles by DLS reveals that the particles are not completely degraded after proteinase K treatment (Figure 9-4). However, a small population at lower sizes could be observed. This indicates that the nanoreactors undergo slight compositional changes and are not completely unaffected by the protease treatment. Figure 9-3: SDS-PAGE analysis of TP containing nanoreactors and free TP before and after proteinase K (prot K) treatment. Lane 1: NP+TP+Tsx; lane 2: NP+TP+Tsx without prot K after 4 h incubation at 37°C; lane 3: NP+TP+Tsx with prot K after 4 h incubation at 37°C. Lane 4: NP+TP-Tsx; lane 5: NP+TP-Tsx without prot K after 4 h incubation at 37°C; lane 6: NP+TP-Tsx with prot K after 4 h incubation at 37°C. Lane 7: free TP without prot K after 4 h incubation at 37°C; lane 8: free TP with prot K after 4 h incubation at 37°C. The protein molecular weight marker used is SM0431 (Fermentas). 153 Chapter 9 Figure 9-4: DLS distribution patterns of nanoreactors (A) before and (B) after proteinase K treatment. The intensity was plotted for ten consecutive measurements as a function of the radius. 9.3.3 Interaction between the enzyme and the polymeric particle The motivation of this study was to test whether the enzyme TPE.coli has affinity for the polymer and tends to stick to the polymeric membrane. This could explain why straightforward encapsulation of the enzyme inside the aqueous space of the nanoparticles is prevented. The enzyme TPhuman is used as comparative sample in this study. The interaction of monolayers of PMOXAPDMS-PMOXA (at air-water interface) with the enzymes TPE.coli and TPhuman was investigated by surface pressure studies. Changes of the surface pressure upon addition of the protein, indicate interactions between protein and polymer. The porin Tsx acts as a positive control for these experiments, because it has already been proven that this porin interacts with such a polymeric membrane [3]. If diluted to a concentration below the critical micellar concentration of the detergent used for porin extraction and storage, the porin is forced to incorporate inside the hydrophobic block of the membrane in order to shield its hydrophobic surfaces and avoid precipitation. 154 Detailed evaluation of the nanoreactor build-up At physiological salt concentrations (PBS), our results show beside a clear interaction with Tsx, also an interaction with TPE.coli. No changes in surface pressure were observed after adding TPhuman (Figure 9-5A). However, the observed changes are rather surprising, since interactions are normally associated with an increase in surface pressure rather than a decrease. The behavior of PMOXA-PDMS-PMOXA triblock copolymers at the air-water interface has previously been reported by Haefele and coworkers [5]. Different phases of monolayer arrangement were identified ranging from an expanded conformation at large molecular areas to mushroom and brush conformations by stretching the chains due to space limitations. Probably, upon interaction, the polymers reorganize from a mushroom conformation into a more expanded conformation resulting in a decrease in surface pressure. The same experiments were therefore performed in high salt conditions (1 M NaCl), where the polymers are probably directly organized in the expanded conformation and where the expected increase in surface pressure was observed each time the porin Tsx was added (Figure 9-5B). In this condition, the interaction of Tsx is characterized by an initial strong insertion into the polymer layer followed by a slow decrease of the surface pressure. This might hint to a possible rearrangement of the polymer phase. In response to the enzymes TPE.coli and TPhuman also an increase in surface pressure was detected, although the kinetics were different as compared to the Tsx interaction (Figure 9-5B). This suggests no real incorporation is occurring, but rather an association of the enzymes with the PMOXA blocks in the solution. The surface pressure experiments are however difficult to interpret and it is questionable whether these results are really indicative for relevant interactions between the triblock copolymer and the enzymes. Moreover, the monolayer approach deviates from fully hydrated membrane systems like nanoreactors, where the polymers are probably organized in different conformations and behave in a different way than at air-water surfaces. 155 Chapter 9 Figure 9-5: Surface pressure measurements of monolayers composed of PMOXA-PDMSPMOXA triblock copolymers in PBS (A) and in high salt solution (B). Arrows indicate the addition of 0.08 µM protein to the solution. The surface pressure changes are recorded over time. The association of the enzyme TP with the polymeric particle was further investigated using fluorescently labeled enzyme. By making use of a TPAlexa®555, we want to distinguish between association occurring by binding of the enzyme to the outside of the particles (which should be able to take place after nanoparticle synthesis) and association by (partial) incorporation of the enzyme into the polymeric wall during the production process. Premade empty nanovesicles were mixed with free TP-Alexa®555 in order to verify whether the enzyme interacts with the surface of the nanoparticles. After the elimination of unbound/unassociated enzyme by Ni affinity chromatography, the sample did not show any fluorescence signal. In contrast, when using the fluorescently labeled enzyme to produce nanoreactors, an estimated amount of 10% of the fluorescent signal was retained in the nanoreactor sample after purification on a Ni affinity column. This is in close agreement with the previously determined ‘encapsulation’ efficiencies using SDS-PAGE analysis (see section 3.3.4). Nevertheless, from the above we now know that no real encapsulation occurs, but rather an association of the enzyme to the particles. In conclusion, since 156 Detailed evaluation of the nanoreactor build-up no strong non-specific interaction and binding of the enzyme to the surface of the particles is detected, the association of the enzyme takes place during the production process of the nanoreactors. The exact localization of the enzymes in the particle and the precise nanoparticle/enzyme conformation is however not clear at the moment. Unfortunately, the resolution of the fluorescence microscope is not high enough to analyze the exact localization of the enzyme onto the 200 nm-sized particles in more detail. 9.3.4 Nanoreactors with other enzymes Next, we wanted to investigate whether the problems we encounter with TPE.coli incorporation into the polymer wall of the nanoreactors are enzyme specific. Indeed, previous studies reported the successful encapsulation of other enzymes inside similar PMOXA-PDMS-PMOXA nanoreactors. For example -lactamase and nucleoside hydrolase from Trypanosoma vivax (TvNH) were used and their activity was preserved after incorporation in the nanoreactors [3, 6]. To this end, we used our polymers and production procedure to incorporate different enzymes. As test cases TvNH and other sources of TP; human TP (TPhuman) and TPE.coli from a commercial source (Sigma, hereafter called TPE.coli,Sigma) were used. Strikingly, we were not able to make active nanoreactors with none of these enzymes. Also for these enzyme-reactor systems, the enzymes lose the majority of their activity during the production process. For example, figure 9-6 shows the results of the TvNH containing nanoreactors. In its active conformation, this enzyme is able to cleave nucleosides into the nucleobase and the ribose. The activity of the enzyme can be followed by a reducing sugar assay that measures the ribose production. The results demonstrate that also here the nanoreactors without porins are the most active ones (Figure 9-6 left). Moreover, the amount of enzyme associated to the particle decreases when more porin was added (Figure 9-6 right). In conclusion, the problems encountered are clearly not enzyme specific. Since the same protocol and conditions were used as described in the paper by Ranquin et al. [3], the TvNH containing nanoreactors constructed 157 Chapter 9 here exhibit a totally different behavior. This seems to indicate that the problems are rather polymer dependent. Figure 9-6: Left: Inosine conversion catalyzed by TvNH containing nanoreactors with different porin concentrations measured as the increase in product (ribose) formation in function of time. Right: SDS-PAGE analysis of the nanoreactors showing the amount of TvNH (37 kDa) and Tsx (31 kDa) present in the samples. The protein molecular weight marker used is SM0431 (Fermentas). 9.3.5 Nanoreactors with other polymers The previous experiments suggest that failure to encapsulate enzymes in the interior of the polymeric nanoparticles is not enzyme dependent but rather polymer dependent. From the beginning of this study we worked with the same batch of commercially available PMOXA20-PDMS54-PMOXA20 triblock copolymers (from Polymer Source Inc.), in order to reduce batch-to-batch variations. Since it is very difficult to precisely control the synthesis and length of both hydrophilic and hydrophobic blocks of block copolymers, there is a high risk for polymer batch variability. However, it is possible that, due to longterm storage, the polymer degraded during the proceedings of this study. Therefore, we ordered and examined fresh polymer batches from different sources in order to investigate whether different, preferentially better, nanoreactor properties can be obtained. To this end, 3 batches of commercial 158 Detailed evaluation of the nanoreactor build-up PMOXA20-PDMS54-PMOXA20 polymers from Polymer Source Inc. and 1 batch of PMOXA20-PDMS42-PMOXA20 polymers, kindly provided by Prof. Meier from the Department of Chemistry from Basel University, were used. Although the commercial and non-commercial polymers show a slight difference in hydrophobic block length, the hydrophobic-to-hydrophilic ratio is still very similar and we expect both triblock copolymers to form nanoreactors. After visual inspection of the various polymer batches, already clear differences in the morphology, texture and color of the polymer powder could be observed. The construction of TPE.coli containing nanoparticles with the different PMOXA-PDMS-PMOXA polymer batches, resulted in nanoparticles of slightly different sizes and turbidity (data not shown). Unfortunately, we were not able to obtain more active nanoreactors with any of the polymer batches available. 9.4 Conclusion and discussion Since the TP containing nanoreactors showed very low enzymatic activity and specific immune responses against the enzyme, we performed various experiments in order to pinpoint the exact problem. First, we were able to show, much to our surprise, that incorporation of the nucleoside specific porin Tsx in the nanoreactor wall is not necessary for enzymatic conversion of thymidine by the TP enzyme. Curiously, the highest enzymatic activity was found for nanoreactors that have no Tsx incorporated. So, instead of improving enzymatic activity by incorporating a nucleoside specific transporter in the polymeric membrane, the enzymatic activity is decreased as a function of Tsx incorporation. This is very much in contrast to other findings [3, 7] where enzymatic activity is only obtained after permeabilization of the reactor wall and enzymatic activity increases proportional to porin concentration. Further, we demonstrated that although our nanoreactors contain a relatively high amount of enzyme, they show very poor activity. The apparent catalytic turnover of the enzyme containing nanoreactors was drastically decreased in comparison to the free enzyme. Moreover, they are sensitive to proteolytic 159 Chapter 9 degradation. From these experiments it is clear that the enzyme is not nicely encapsulated inside the particles like it is supposed to be, but that the enzyme is rather associated to the outside of the particle in a non-active conformation. Since no strong attraction and binding of the enzyme to the particle surface and polymeric membrane was observed, the association of the enzyme to the particles is occurring during the nanoreactor production process. Thereby, the enzyme is incorporated into and/or onto the polymeric wall, losing its functional integrity. These data also support the adjuvant effect of the nanoreactors as described in chapter 6. Most likely, the problems are rather polymer-dependent than enzymedependent, because attempts to encapsulate other enzymes encountered the same inconsistencies. Even when exactly the same conditions were used to construct previously reported TvNH containing nanoreactors, encapsulation remained unsuccessful. In addition, analyzing the structure of the enzyme TP gave us an idea of the amount of hydrophobic residues on the surface of the enzyme. Such hydrophobic patches can cause aggregation (cf. membrane proteins) and prevent encapsulation of the enzyme inside the aqueous cavity of the nanoreactors. Nevertheless, we could not detect significant hydrophobic patches on the surface of the TPE.coli. As such, we suggest that the TP enzyme is at least theoretically perfectly suited for encapsulation. Unfortunately, thus far, we were not able to resolve the problems using other polymer batches of different sources. Further in depth investigation is needed to identify the problems encountered and to make better enzyme containing nanoreactors with high reproducibility. It is possible that the polymers underwent transitions during the enrolment of this study due to long-term storage. Air humidity can be the cause of polymer degradation. As a result, the polymers form rather nanospheres that have the same size as vesicles, but are not hollow. In such case, encapsulation of the enzyme is not possible and it is realistic that the enzyme is incorporated in the polymeric matrix forming protein/polymer aggregates instead of vesicles. The presence of impurities in the polymer batch can also cause problems in producing good reproducible nanoreactors. 160 Detailed evaluation of the nanoreactor build-up All together, we suggest that polymer degradation, poor polymer purity and batch-to-batch variability of the polymer stocks are the main causes of our problems. Nanoreactors based on triblock copolymers are a completely new and promising field. Therefore, if pharmacological applications on an industrial scale are envisioned it is of uttermost importance that the synthesis of the polymers can be precisely controlled in order to obtain identical nanoreactors of different polymer batches. Further investigation is necessary to better control the polymer chemistry in order to construct nanoreactors exhibiting exactly the same properties and behaving in precisely the same way every time a new batch of reactors is made from a new batch of polymers. 9.5 References [1] C. Nardin, Hirt, T., Leukel, J., Meier, W., Polymerized ABA Triblock Copolymer Vesicles, Langmuir, 16 (2000) 1035-1041. [2] D.W. Parkin, Purine-specific nucleoside N-ribohydrolase from Trypanosoma brucei brucei. Purification, specificity, and kinetic mechanism, J Biol Chem, 271 (1996) 21713-21719. [3] A. Ranquin, W. Versees, W. Meier, J. Steyaert, P. Van Gelder, Therapeutic Nanoreactors: Combining Chemistry and Biology in a Novel Triblock Copolymer Drug Delivery System, Nano Lett, 5 (2005) 2220-2224. [4] F. Axthelm, O. Casse, W.H. Koppenol, T. Nauser, W. Meier, C.G. Palivan, Antioxidant nanoreactor based on superoxide dismutase encapsulated in superoxide-permeable vesicles, J Phys Chem B, 112 (2008) 8211-8217. [5] T. Haefele, K. Kita-Tokarczyk, W. Meier, Phase behavior of mixed Langmuir monolayers from amphiphilic block copolymers and an antimicrobial peptide, Langmuir, 22 (2006) 1164-1172. [6] C. Nardin, S. Thoeni, J. Widmer, M. Winterhalter, W. Meier, Nanoreactors based on (polymerized) ABA-triblock copolymer vesicles, Chem Comm, (2000) 1433-1434. [7] M. Grzelakowski, O. Onaca, P. Rigler, M. Kumar, W. Meier, Immobilized protein-polymer nanoreactors, Small, 5 (2009) 2545-2548. 161 Chapter 10 General conclusion General conclusion 10 General conclusion In this study we propose a new therapeutic tool for enzyme delivery in enzyme replacement therapy, based on nanoreactors composed of PMOXA-PDMSPMOXA triblock copolymers. As a proof of principle we chose to develop nanoreactors for the potential use in enzyme replacement therapy of MNGIE. This is a severe disorder caused by the deficiency of the crucial enzyme thymidine phosphorylase (TP). At the moment, allogeneic stem cell transplantation is the only available therapy for MNGIE. Unfortunately, the treatment is associated with significant morbidity and mortality. Furthermore, for 30-to-40% of the patients, no matching related or unrelated donors are available. Therefore, a safer and more universal treatment for TP-deficiency would be beneficial to MNGIE patients and enzyme replacement therapy could provide a promising alternative strategy. The aim of this study was to develop and evaluate self-assembling PMOXAPDMS-PMOXA nanoreactors that are filled with active thymidine phosphorylase and are permeabilized for small molecules by incorporating a bacterial porin Tsx in the reactor wall. The nucleoside specific Tsx porin should allow the transport of the substrates and products across the particle membrane, resulting in nanometer-sized bioreactors. Such nanoreactors are significantly different from the classical drug delivery particles, because the enzyme doesn’t need to be released from the particle to fulfill its biological function. They represent a new and innovative concept in the area of drug delivery research. Especially for the delivery of enzymes, there is an urgent need for better delivery systems, because their catalytic activity is easily impaired by the smallest conformational change or environmental attack. Although nanoreactors have been suggested already for the therapeutic delivery of enzymes [1-5], detailed studies concerning their stability, toxicity, immunogenicity, biodistribution and in vivo efficacy are lacking. These properties are however of crucial importance for the further exploitation and evaluation of such nanoreactors as therapeutic enzyme delivery devices. 165 Chapter 10 For this study, we prepared stable TP-containing nanoreactors by the solvent evaporation method, based on previously described successful nanoreactor production protocols as reported by the group of Wolfgang Meier [6-9]. The production process resulted in monodisperse particles with a diameter of approximately 200 nm, as determined by DLS. Suspension in blood serum had no immediate effect on vesicle stability and no leakage of the enzyme from the particles was detected. On the contrary, enzyme leakage was observed when TP was encapsulated in liposomes, showing that our polymeric nanoparticles exhibit higher mechanical stability. In addition, no stimulation of phagocytes was detected when nanoreactors were incubated with naïve macrophages in vitro. On IFN- primed macrophages, a minor porin dependent activation was observed upon incubation with the nanoreactors. If needed, this response could possibly be circumvented by substituting the bacterial-derived Tsx nucleoside pore with a human homologue, such as the eENT or VDAC porin for example [10]. Subsequent in vivo experiments demonstrated however that the bacterial porins are sufficiently shielded by the polymer, since no acute inflammatory effects on macrophages could be detected after intraperitoneal injection of the nanoreactors. Moreover, proliferating liver cells are unaffected when cultured for an extended time with relevant concentrations of nanoreactors. These effects are comparable to those obtained with PEGmodified particles (i.e. sterically stabilized liposomes) [11]. Taken together, the nanoreactors exhibit high stability and low acute toxicity and immunogenicity, making them promising towards in vivo applications at a first view. Unfortunately, further investigation revealed that the nanoreactors induce a strong enzyme-specific antibody response after repeated iv injections in mice. This implies that the enzyme is not shielded from the environment as stated before. Moreover, hepatosplenic uptake of the nanoreactors prevents longterm blood circulation in living mice and rats. In addition, the enzymatic activity of the TP-containing nanoreactors seemed insufficient to significantly reduce the thymidine levels in the plasma of MNGIE patients. Taken together, these characteristics are major drawbacks and limit the use of this type of 166 General conclusion nanoreactors in enzyme replacement therapy. The impaired enzymatic activity of the nanoreactors was unexpected and not in agreement with previous reports on PMOXA-PDMS-PMOXA nanoreactors [2, 4, 5], suggesting an incorrect nanoreactor composition. Moreover, the specific anti-TP immune response indicates that the enzyme is exposed to the environment. This prompted us to re-evaluate the nanoreactor build-up. The subsequent experiments unequivocally indicated that the enzyme indeed is not encapsulated properly inside the particles, but tends to be associated to the surface of the particles in a predominantly inactive conformation. The exact cause of the disturbed encapsulation is not elucidated completely yet and needs further investigation. At our opinion, the polymers are probably not able to form hollow particles and instead, homogenic dense polymer-matrices are formed that incorporate both the enzyme and the porin. This incorporation renders the enzyme inactive, probably due to conformational changes. All together, we can conclude that although the polymer-based nanoreactors present a promising new tool in biomedical applications, the TP-containing nanoreactors as presented in this study do not possess ideal properties for use in enzyme replacement therapy. Nevertheless, this investigation revealed important information on the stability, toxicity and immunogenicity of such nanoreactors and was of high relevance for the further exploitation of PMOXAPDMS-PMOXA triblock copolymer based particles as drug delivery devices. Although we were not successful in generating highly active nanoreactors, we could demonstrate overall low immunogenicity and toxicity of the PMOXAPDMS-PMOXA particles. Moreover, we were the first to provide in vivo data on PMOXA-PDMS-PMOXA nanoparticles, elucidating the potential of this type of vesicles for other applications. Since our nanoreactors seem to be very effective in inducing a specific immune response against the associated enzyme and are particularly stable under physiological conditions, the nanoparticles could for example be tested for their use in vaccination strategies. Future studies need to focus more on the reproducibility of nanoreactor preparation and the improvement of polymer quality. Beside the problems we 167 Chapter 10 encountered here with the use of PMOXA-PDMS-PMOXA triblock copolymers, other studies also report clear batch-to-batch differences (personal communication with Prof. Meier and Dr. Ranquin). Given that the phase behavior of triblock copolymers in aqueous solutions depends on the composition, length, structure and molecular weight distribution of the individual blocks as well as on the molecular architecture of the whole polymer, we believe it is very important to better 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. Fine tuning the production method of the polymers and the nanoreactors will be a key step in the exploitation of triblock based nanoreactors as drug delivery devices. In addition, other polymers than PMOXA-PDMS-PMOXA may prove to be more effective for the production of enzyme-containing nanoreactors. Such polymers should ideally be biodegradable, which is not the case for the PDMS middle block of our triblock copolymer. Overall, the development of enzyme-containing nanoparticles for use in enzyme replacement therapy is a challenge, because life-time administrations of appropriate enzyme formulations are required to restore the enzyme activity. Therefore, the ideal enzyme replacement delivery device needs to possess long blood circulation times, high activity and low immunogenicity properties. Recent studies underline the difficulty of constructing nanocarriers that provide long circulation together with low immunogenicity [12-14]. The long circulation property can be obtained by grafting stealthy polymers like PEG to the outer core of the particle in order to sterically stabilize the particle and prevent aggregation [15]. However, repeated administrations of such sterically stabilized particles result in the formation of PEG specific (or drug specific) antibodies [16, 17]. Immunogenic aspects are less important if only a single bolus injection is required like it is the case for particular cancer therapy purposes. This is probably the reason why nanoparticle-mediated drug delivery has booked its first successes and is entering clinical trials for such applications [18-20]. 168 General conclusion Nevertheless, the development of controlled drug delivery systems remains a challenging field of research. In addition to the highly innovative and engineered drug delivery strategies making use of complex combinations of drugs, carriers and ligands, other drug delivery systems focus more on the way nature provides solutions for efficient delivery of compounds to the body (like drugs loaded on erythrocytes [21] and drugs coupled to human serum albumin [22]). However, each strategy has its own advantages and disadvantages and every application needs its own optimization. At this stage there are no simple predictors that will guarantee the best delivery strategy for a given protein. The design and optimization of carriers represent an important step in the development of new protein delivery approaches. Both the physical (such as compartmentalization) and size, the shape, chemical mechanical (such as properties pegylation, and polymer composition and ligand coupling) properties of the delivery systems need to be further optimized to better control their behavior in a biological context. Undoubtedly, further scientific research will lead to improved protein therapeutics that will provide more efficacious drugs and favorable clinical outcomes. References [1] C. Nardin, S. Thoeni, J. Widmer, M. Winterhalter, W. Meier, Nanoreactors based on (polymerized) ABA-triblock copolymer vesicles, Chem Comm, (2000) 1433-1434. [2] A. Ranquin, W. Versees, W. Meier, J. Steyaert, P. Van Gelder, Therapeutic Nanoreactors: Combining Chemistry and Biology in a Novel Triblock Copolymer Drug Delivery System, Nano Lett, 5 (2005) 2220-2224. [3] P. Broz, S. Driamov, J. Ziegler, N. Ben-Haim, S. Marsch, W. Meier, P. Hunziker, Toward Intelligent Nanosize Bioreactors: A pH-Switchable, ChannelEquipped, Functional Polymer Nanocontainer, Nano Lett, 6 (2006) 2349-2353. [4] F. Axthelm, O. Casse, W.H. Koppenol, T. Nauser, W. Meier, C.G. Palivan, Antioxidant nanoreactor based on superoxide dismutase encapsulated in superoxide-permeable vesicles, J Phys Chem B, 112 (2008) 8211-8217. [5] M. Grzelakowski, O. Onaca, P. Rigler, M. Kumar, W. Meier, Immobilized protein-polymer nanoreactors, Small, 5 (2009) 2545-2548. [6] C. Nardin, J. Widmer, M. Winterhalter, W. Meier, Amphiphilic block copolymer nanocontainers as bioreactors, Eur. Phys. J. E, 4 (2001) 403-410. [7] A. Graff, M. Sauer, P. Van Gelder, W. Meier, Virus-assisted loading of polymer nanocontainer, Proc Natl Acad Sci U S A, 99 (2002) 5064-5068. Epub 2002 Mar 5026. 169 Chapter 10 [8] P. Broz, S.M. Benito, C. Saw, P. Burger, H. Heider, M. Pfisterer, S. Marsch, W. Meier, P. Hunziker, Cell targeting by a generic receptor-targeted polymer nanocontainer platform, J Control Release, 102 (2005) 475-488. [9] M. Nallani, S. Benito, O. Onaca, A. Graff, M. Lindemann, M. Winterhalter, W. Meier, U. Schwaneberg, A nanocompartment system (Synthosome) designed for biotechnological applications, J Biotechnol, 123 (2006) 50-59. Epub 2005 Dec 2020. [10] A. Deniaud, C. Rossi, A. Berquand, J. Homand, S. Campagna, W. Knoll, C. Brenner, J. Chopineau, Voltage-dependent anion channel transports calcium ions through biomimetic membranes, Langmuir, 23 (2007) 3898-3905. [11] X. Shan, C. Liu, Y. Yuan, F. Xu, X. Tao, Y. Sheng, H. Zhou, In vitro macrophage uptake and in vivo biodistribution of long-circulation nanoparticles with poly(ethylene-glycol)-modified PLA (BAB type) triblock copolymer, Colloids Surf B Biointerfaces, 72 (2009) 303-311. [12] H. Koide, T. Asai, K. Hatanaka, K. Shimizu, M. Yokoyama, T. Ishida, H. Kiwada, N. Oku, [Elucidation of accelerated blood clearance phenomenon caused by repeat injection of PEGylated nanocarriers], Yakugaku Zasshi, 129 (2009) 1445-1451. [13] S.M. Moghimi, J. Szebeni, Stealth liposomes and long circulating nanoparticles: critical issues in pharmacokinetics, opsonization and proteinbinding properties, Prog Lipid Res, 42 (2003) 463-478. [14] S.M. Moghimi, A.C. Hunter, Capture of stealth nanoparticles by the body's defences, Crit Rev Ther Drug Carrier Syst, 18 (2001) 527-550. [15] A.L. Klibanov, K. Maruyama, V.P. Torchilin, L. Huang, Amphipathic polyethyleneglycols effectively prolong the circulation time of liposomes, FEBS Lett., 268 (1990) 235-237. [16] T. Ishida, X. Wang, T. Shimizu, K. Nawata, H. Kiwada, PEGylated liposomes elicit an anti-PEG IgM response in a T cell-independent manner, J Control Release, 122 (2007) 349-355. [17] H. Koide, T. Asai, K. Hatanaka, S. Akai, T. Ishii, E. Kenjo, T. Ishida, H. Kiwada, H. Tsukada, N. Oku, T cell-independent B cell response is responsible for ABC phenomenon induced by repeated injection of PEGylated liposomes, Int J Pharm. [18] A. Surendiran, S. Sandhiya, S.C. Pradhan, C. Adithan, Novel applications of nanotechnology in medicine, Indian J Med Res, 130 (2009) 689-701. [19] S.D. Caruthers, S.A. Wickline, G.M. Lanza, Nanotechnological applications in medicine, Curr Opin Biotechnol, 18 (2007) 26-30. [20] V. Wagner, A. Dullaart, A.K. Bock, A. Zweck, The emerging nanomedicine landscape, Nat Biotechnol., 24 (2006) 1211-1217. [21] B.E. Bax, M.D. Bain, L.D. Fairbanks, A.D. Webster, P.W. Ind, M.S. Hershfield, R.A. Chalmers, A 9-yr evaluation of carrier erythrocyte encapsulated adenosine deaminase (ADA) therapy in a patient with adult-type ADA deficiency, Eur J Haematol, 79 (2007) 338-348. [22] W.J. Gradishar, Albumin-bound paclitaxel: a next-generation taxane, Expert Opin Pharmacother, 7 (2006) 1041-1053. 170 Chapter 11 Summary Summary 11 Summary Many known genetic deficiencies in human are caused by a mutation in a gene encoding a crucial enzyme, leading to the accumulation of its substrates. As a result, toxic amounts of the substrates appear in the bloodstream, disturbing the normal metabolic neurogastrointestinal degradation pathways. encephalomyopathy is MNGIE such a or mitochondrial unique autosomal- recessive disorder associated with mitochondrial DNA alterations [1]. In 1998, Nishino et al. mapped the disease locus and identified a deficiency in the enzyme thymidine phosphorylase (TP) as the cause of the disorder [2]. Lossof-function mutations in the TP gene lead to systemic accumulations of the substrates thymidine and deoxyuridine [3]. Current therapies for MNGIE are still in the developmental stage and include platelet infusion [4] and allogeneic stem cell transplantation [5]. Unfortunately, these treatments are far from ideal and suffer from the lack of sustained TP activity and serious safety risks [6]. A more reliable and generally well-tolerated strategy to treat enzyme deficiencies is enzyme replacement therapy [7, 8]. In enzyme replacement therapy the deficient enzyme is administered intravenously in its native or stabilized form [9]. Direct administration of native enzymes is usually limited due to rapid protein elimination from the circulation [10]. To improve their stability, proteolytic resistance, immunogenicity and circulation half-life, therapeutic proteins can be modified by covalently linking various poly(ethylene glycol) (PEG) molecules to their surface, a technique called pegylation [11]. However, the pegylation process is complex, expensive, and often compromises enzyme activity [12, 13]. A more convenient strategy to deliver therapeutic enzymes to the body is to encapsulate them in appropriate carrier systems [14]. In a first attempt, proteins were trapped in liposomes [15]. Unfortunately, these lipidic carrier systems suffer from rapid leakage of the protein, poor stability and fast clearance by the mononuclear phagocyte system [16, 17]. By grafting PEG on the liposome surface, a new generation of long-circulating or stealth liposomes 173 Chapter 11 was introduced [18]. However, fast drug leakage and the accelerated blood clearance [19] remain major drawbacks of these lipidic delivery systems. In order to obtain more robust membranes with controllable properties, extensive efforts were made within the last decade to design polymeric vesicles [20-23]. The aim of this study was to develop and evaluate polymeric nanometer-sized reactors for use in enzyme replacement therapy, with MNGIE as proof of concept. Therefore, the therapeutic enzyme, thymidine phosphorylase, is encapsulated in polymeric particles constructed of the amphiphilic triblock copolymer PMOXA-PDMS-PMOXA (poly(2-methyloxazoline)-block- poly(dimethylsiloxane)-block-poly(2-methyloxazoline)) (Figure 11-1). The nanoparticles are permeabilized for substrates and products by integrating bacterial channel proteins (or porins) in their polymeric wall, a strategy introduced by Meier and co-workers [24]. The nucleoside specific porin Tsx was selected as channel-forming protein for the nanoreactors, since it contains specific nucleoside binding sites. In this way, the Tsx porin allows the efficient transport of the substrates and products through the capsule wall. This results in reactors where the enzymatic reaction is restricted to the inner volume of the polymeric nanocontainer. Figure 11-1: Schematic representation of the envisioned thymidine phosphorylase encapsulating nanoreactor. 174 Summary We constructed thymidine phosphorylase encapsulating nanoreactors (TP-NRs) according to the method of Ranquin et al. [25]. Monodisperse nanoreactors with a mean diameter around 200 nm were obtained. SDS-PAGE analysis revealed an enzyme encapsulation efficiency of ±10% and showed a clear incorporation of the porin Tsx. The in vitro enzymatic activity of the TP-NRs was confirmed spectrophotometrically by measuring a decrease in absorption at 290 nm due to thymidine consumption. The stability of the permeabilized TP-NRs was investigated by incubating them for several days in mouse serum at 37°C. TP containing liposomes were used as control samples. Although a slight decrease in enzymatic activity was observed over time, the TP-NRs are stable and not leaky at 37°C in serum for several days. The high mechanical stability of the nanoreactors indicate that they are potentially more suited for in vivo use than analogous particles constructed of lipids. Next, possible toxic effects of the TP-NRs on hepatocytes were investigated. Hepatocytes were isolated from rats and cultured with different concentrations of TP-NRs. Cytotoxicity was tested by cellular morphology and membrane leakage of lactate dehydrogenase (LDH assay) as a function of time. The results showed no significant influence on the viability of the hepatocytes up to 48 h of incubation. Only after prolonged exposure and at higher doses of TP-NRs, elevated levels of LDH leakage were measured. This low cytotoxic effect was confirmed by visual inspection of the morphology of the hepatocytes. Further, the impact of TP-NRs on the inflammatory status of macrophages was tested ex vivo and in vivo. For ex vivo studies the effect of nanoreactors on isolated peritoneal macrophages was analyzed. The results showed that the nanoparticles do not induce inflammatory cytokine production by naïve macrophages. Interestingly, the free enzyme significantly provoked the secretion of all inflammatory mediators tested. This means that the encapsulation of the enzyme in polymer particles indeed protects the protein from the immune system. On IFN-γ primed macrophages, a slight porindependent effect was observed. Subsequently, the inflammatory potency of the nanoparticles was tested in vivo. Only low amounts of cytokines could be found in macrophage-supernatant and in serum of mice treated with an 175 Chapter 11 intraperitoneal injection of TP-NRs, while high concentrations were detected in control mice injected with the highly inflammatory compound LPS. Hence, in a physiologically relevant in vivo setting, the TP-NRs do not provoke acute inflammatory responses. In a next phase, the nanoreactors were administered intravenously and enzyme-specific antibody responses were analyzed in the sera after several boost injections. Surprisingly, the antibody response to the nanoreactor-encapsulated enzyme was higher than to the free enzyme. The nanoreactors even seem to act as an adjuvant, increasing the level of specific anti-TP antibody production. This effect is not desirable for their use in enzyme replacement therapy, because such antibody responses will cause fast elimination of the TP-NRs from the blood circulation. Moreover, the antibody response will only become stronger after several boost injections, hereby devastating the therapeutic effect of life-long needed enzyme replacement therapy. Furthermore, the biodistribution of the nanoreactors was investigated by comparing the in vivo fate of radiolabeled enzyme-containing nanoreactors with radiolabeled free enzyme. The results indicated that the biodistribution of free and encapsulated TP differs significantly. The nanoreactors accumulated mainly in liver and spleen, while the concentrations in these organs were clearly lower for the free enzyme. No significant difference in blood circulation half-life could be demonstrated between the free enzyme and the nanoreactorassociated enzyme. These data indicate that the nanoreactors do not possess the ideal size and/or surface characteristics to evade hepatosplenic uptake and to prolong the blood circulation time of the incorporated enzyme significantly. To investigate whether the nanoreactors are able to decrease the toxic substrate concentrations in MNGIE patients, the TP-NRs were incubated with plasma of a MNGIE patient at 37°C. The thymidine and deoxyuridine concentrations were determined via HPLC. The results showed that the nanoreactors are not efficient enough to convert the high amount of substrates to normal concentrations. At that point, a detailed re-evaluation of the nanoreactor build-up was performed. The results revealed that the enzyme is not encapsulated properly inside the inner cavity of the nanoreactors, but is 176 Summary rather associated to the outside of the particle in a non-active conformation. This explains the enzyme specific antibody response and low enzymatic activity of the particles. Unfortunately, thus far, we were not able to resolve the encapsulation problems. Further in depth investigation is needed to identify the problems encountered and to make better enzyme containing nanoreactors with high reproducibility. In conclusion, we constructed thymidine phosphorylase containing nanoparticles and evaluated their potential as new enzyme delivery vehicles. All together, we can conclude that although the results potentiate a broad new class of technologically useful polymer-based nanoreactors, the TP-NRs as presented in this study do not possess ideal properties for use in enzyme replacement therapy. Nevertheless, this investigation revealed important information on the stability, toxicity and immunogenicity of such nanoreactors and was of high relevance for the further exploitation of these particles as drug delivery devices. References [1] I. Nishino, A. Spinazzola, M. Hirano, MNGIE: from nuclear DNA to mitochondrial DNA, Neuromuscul Disord, 11 (2001) 7-10. [2] I. Nishino, A. Spinazzola, M. Hirano, Thymidine phosphorylase gene mutations in MNGIE, a human mitochondrial disorder, Science, 283 (1999) 689-692. [3] M.L. Valentino, R. Marti, S. Tadesse, L.C. Lopez, J.L. Manes, J. Lyzak, A. Hahn, V. Carelli, M. Hirano, Thymidine and deoxyuridine accumulate in tissues of patients with mitochondrial neurogastrointestinal encephalomyopathy (MNGIE), FEBS Lett, 581 (2007) 3410-3414. [4] M.C. Lara, B. Weiss, I. Illa, P. Madoz, L. Massuet, A.L. Andreu, M.L. Valentino, Y. Anikster, M. Hirano, R. Marti, Infusion of platelets transiently reduces nucleoside overload in MNGIE, Neurology, 67 (2006) 1461-1463. [5] M. Hirano, R. Marti, C. Casali, S. Tadesse, T. Uldrick, B. Fine, D.M. Escolar, M.L. Valentino, I. Nishino, C. Hesdorffer, J. Schwartz, R.G. Hawks, D.L. Martone, M.S. Cairo, S. DiMauro, M. Stanzani, J.H. Garvin, Jr., D.G. Savage, Allogeneic stem cell transplantation corrects biochemical derangements in MNGIE, Neurology, 67 (2006) 1458-1460. [6] P.F. Chinnery, J. Vissing, Treating MNGIE: is reducing blood nucleosides the first cure for a mitochondrial disorder?, Neurology, 67 (2006) 1330-1332. [7] T.A. Burrow, R.J. Hopkin, N.D. Leslie, B.T. Tinkle, G.A. Grabowski, Enzyme reconstitution/replacement therapy for lysosomal storage diseases, Curr Opin Pediatr, 19 (2007) 628-635. 177 Chapter 11 [8] W.R. Wilcox, M. Banikazemi, N. Guffon, S. Waldek, P. Lee, G.E. Linthorst, R.J. Desnick, D.P. Germain, Long-term safety and efficacy of enzyme replacement therapy for Fabry disease, Am J Hum Genet, 75 (2004) 65-74. [9] N.W. Barton, R.O. Brady, J.M. Dambrosia, A.M. Di Bisceglie, S.H. Doppelt, S.C. Hill, H.J. Mankin, G.J. Murray, R.I. Parker, C.E. Argoff, et al., Replacement therapy for inherited enzyme deficiency--macrophage-targeted glucocerebrosidase for Gaucher's disease, N Engl J Med, 324 (1991) 14641470. [10] D.T. Achord, F.E. Brot, C.E. Bell, W.S. Sly, Human beta-glucuronidase: in vivo clearance and in vitro uptake by a glycoprotein recognition system on reticuloendothelial cells, Cell, 15 (1978) 269-278. [11] J.M. Harris, R.B. Chess, Effect of pegylation on pharmaceuticals, Nat Rev Drug Discov, 2 (2003) 214-221. [12] M.S. Hershfield, PEG-ADA replacement therapy for adenosine deaminase deficiency: an update after 8.5 years, Clin Immunol Immunopathol, 76 (1995) S228-232. [13] V. Gaberc-Porekar, I. Zore, B. Podobnik, V. Menart, Obstacles and pitfalls in the PEGylation of therapeutic proteins, Curr Opin Drug Discov Devel, 11 (2008) 242-250. [14] T.M. Allen, P.R. Cullis, Drug delivery systems: entering the mainstream, Science, 303 (2004) 1818-1822. [15] P. Walde, S. Ichikawa, Enzymes inside lipid vesicles: preparation, reactivity and applications, Biomol Eng, 18 (2001) 143-177. [16] M. Yamauchi, K. Tsutsumi, M. Abe, Y. Uosaki, M. Nakakura, N. Aoki, Release of drugs from liposomes varies with particle size, Biol Pharm Bull, 30 (2007) 963-966. [17] A. Gabizon, R. Chisin, S. Amselem, S. Druckmann, R. Cohen, D. Goren, I. Fromer, T. Peretz, A. Sulkes, Y. Barenholz, Pharmacokinetic and imaging studies in patients receiving a formulation of liposome-associated adriamycin, Br J Cancer, 64 (1991) 1125-1132. [18] A.L. Klibanov, K. Maruyama, V.P. Torchilin, L. Huang, Amphipathic polyethyleneglycols effectively prolong the circulation time of liposomes, FEBS Lett., 268 (1990) 235-237. [19] P. Laverman, M.G. Carstens, O.C. Boerman, E.T. Dams, W.J. Oyen, N. van Rooijen, F.H. Corstens, G. Storm, Factors affecting the accelerated blood clearance of polyethylene glycol-liposomes upon repeated injection, J Pharmacol Exp Ther, 298 (2001) 607-612. [20] D.E. Discher, Emerging applications of polymersomes in delivery: From molecular dynamics to shrinkage of tumors, Progress in Polymer Science, 32 (2007) 838-857. [21] D.E. Discher, A. Eisenberg, Polymer vesicles, Science., 297 (2002) 967973. [22] R. Gref, Y. Minamitake, M.T. Peracchia, V. Trubetskoy, V. Torchilin, R. Langer, Biodegradable long-circulating polymeric nanospheres, Science, 263 (1994) 1600-1603. [23] J. Panyam, V. Labhasetwar, Biodegradable nanoparticles for drug and gene delivery to cells and tissue, Adv Drug Deliv Rev, 55 (2003) 329-347. 178 Summary [24] C. Nardin, S. Thoeni, J. Widmer, M. Winterhalter, W. Meier, Nanoreactors based on (polymerized) ABA-triblock copolymer vesicles, Chem Comm, (2000) 1433-1434. [25] A. Ranquin, W. Versees, W. Meier, J. Steyaert, P. Van Gelder, Therapeutic Nanoreactors: Combining Chemistry and Biology in a Novel Triblock Copolymer Drug Delivery System, Nano Lett, 5 (2005) 2220-2224. 179 Chapter 12 Samenvatting Samenvatting 12 Samenvatting Heel wat beschreven genetische afwijkingen in de mens hebben tot gevolg dat een levensnoodzakelijk enzym ontbreekt of slecht functioneert. Hierdoor accumuleren hoge, vaak toxische concentraties van de substraten, wat de normale stofwisseling verstoort en aanleiding geeft tot een brede waaier aan pathologieën. Mitochondriaal neurogastrointestinaal encephalomyopatie of MNGIE is zo een unieke autosomaal-recessieve aandoening die gepaard gaat met wijzigingen in het mitochondriale DNA [1]. In 1998 ontdekte Nishino et al. dat een deficiëntie in het enzym thymidine phosphorylase (TP) de oorzaak is van deze ziekte [2]. Mutaties in het TP-gen leiden tot een geheel of gedeeltelijk functieverlies van het enzym en veroorzaken systemische accumulaties van de substraten thymidine en deoxyuridine [3]. De beschreven therapieën voor MNGIE, zoals transfusie van bloedplaatjes [4] en allogene stamcel transplantaties [5], staan nog in de kinderschoenen. Daarenboven zijn deze behandelingen verre van ideaal en hebben ze heel wat tekortkomingen zoals een onvoldoende lange werking van de enzymactiviteit en ernstige veiligheidsrisico’s [6]. Een veiligere en algemeen aanvaarde strategie voor de behandeling van enzymdeficiënties is de enzymvervangingstherapie [7, 8]. In de enzymvervangingstherapie wordt het deficiënt enzym intraveneus toegediend in zijn natieve of gestabiliseerde vorm [9]. In de praktijk is de intraveneuze toediening van het deficiënt enzym in zijn natieve vorm echter weinig efficiënt, aangezien eiwitten een korte verblijfstijd in het lichaam hebben, waardoor het doseringsinterval te kort wordt voor praktisch gebruik [10]. Om het doseringsinterval te vergroten en de stabiliteit en de proteaseresistentie te verhogen, worden therapeutische eiwitten vaak covalent gekoppeld aan poly-ethyleenglycol (PEG) moleculen, een proces dat pegylatie wordt genoemd [11]. Desalniettemin is pegylatie een complexe en dure strategie waarbij het enzym vaak een deel van zijn activiteit verliest [12, 13]. Het afleveren van therapeutische enzymen met behulp van passende dragersystemen is een betere en meer algemeen toepasbare strategie [14]. 183 Chapter 12 Zo worden eiwitten vaak ingekapseld in liposomen [15]. Spijtig genoeg vertonen ook deze lipide partikels nadelen, zoals een snelle vrijgave van het enzym, lage stabiliteit en snelle klaring door het mononucleair fagocytsysteem [16, 17]. Door liposomen te construeren die PEG moleculen op hun oppervlak dragen, ontstond een nieuwe generatie van lang-circulerende (of ‘stealth’) liposomen [18]. Toch blijven de snelle vrijlating van het enzym en de snelle klaring belangrijke tekortkomingen van deze afleveringssystemen [19]. Vandaar werden in het afgelopen decennium aanzienlijke inspanningen geleverd om polymere partikels te vervaardigen met beter controleerbare en afstembare eigenschappen [20-23]. De bedoeling van deze studie was om polymere nanometer-schaal reactors te construeren en deze te evalueren voor hun gebruik in enzym- vervangingstherapie, met MNGIE als modelsysteem. Daarom werd het therapeutisch enzym, thymidine phosphorylase, ingekapseld in polymere partikels bestaande uit het triblok copolymeer PMOXA-PDMS-PMOXA (poly(2methyloxazoline)-block-poly(dimethylsiloxane)-block-poly(2-methyloxazoline)) (Figuur 12-1). De wand van de partikels wordt gepermeabiliseerd voor de substraten en producten door een porievormend membraaneiwit (ook porine genoemd) in de polymeerwand te integreren. Deze strategie werd voor het eerst geïntroduceerd door de groep van Meier [24]. Het nucleoside-specifiek porine Tsx werd geselecteerd als kanaal-vormend eiwit omdat het over specifieke bindingsplaatsen voor nucleosiden beschikt. Op deze manier zorgt het Tsx voor een efficiënt transport van de substraten en de producten doorheen de polymere wand. Dit resulteert in reactoren waar de enzymatische reactie beperkt is tot de binnenste caviteit van de polymere nanocontainer. De thymidine phosphorylase-bevattende nanoreactors (TP-NRs) werden vervaardigd volgens Ranquin et al. [25]. Monodisperse nanoreactors met een gemiddelde diameter van 200 nm werden bekomen. SDS-PAGE analyse demonstreerde een TP-inkapselefficiëntie van ±10% en toonde aan dat ook het porine Tsx duidelijk geïncorporeerd zat in de nanoreactor wand. De in vitro enzymatische activiteit werd bevestigd via spectrofotometrie. De stabiliteit van de TP-NRs 184 werd onderzocht door ze enkele dagen te incuberen in Samenvatting muizenserum bij 37°C. TP-bevattende liposomen werden gebruikt als controlestaal. Ondanks een lichte daling in enzymatische activiteit, bleken de TP-NRs erg stabiel bij 37°C en vertoonden ze geen snelle vrijlating van het enzym. Hiermee toont deze studie aan dat de TP-NRs beter geschikt zijn voor in vivo gebruik dan analoge partikels vervaardigt uit lipiden. Vervolgens werd onderzocht of de TP-NRs toxisch zijn voor levercellen. Verschillende concentraties TP-NRs werden aan rat hepatocyten toegevoegd en hun effect op de leefbaarheid van de levercellen werd onderzocht via een lactaat dehydrogenase test (LDH test) en via een microscopisch onderzoek van de cellulaire morfologie. Er werd geen significant effect op de leefbaarheid van de cellen vastgesteld gedurende een incubatieperiode van 48 uur. Enkel na langere tijd van blootstelling en bij hogere concentraties aan TP-NRs werd een verhoogde LDH-lekkage vastgesteld. Deze lage toxiciteit werd bevestigd door een visuele inspectie van de levercellen. Figuur 12-1: Schematische voorstelling van de beoogde thymidine phosphorylase bevattende nanoreactoren. 185 Chapter 12 Vervolgens werd de impact van de TP-NRs op de inflammatoire status van macrofagen onderzocht zowel ex vivo als in vivo. Voor de ex vivo studies werd het effect van de nanoreactors op geïsoleerde peritoneale macrofagen geanalyseerd. De resultaten toonden aan dat er geen inflammatoire cytokines werden geproduceerd door naïeve macrofagen als respons op de TP-NRs. Nochtans induceerde het vrij enzym wel duidelijk de secretie van inflammatoire cytokines. Dit betekent dat de inkapseling van het enzym in de partikels inderdaad bescherming biedt tegen inflammatoire immune responsen. Op IFN- gestimuleerde macrofagen werd een licht porineafhankelijk effect gemeten. Verder werd het inflammatoir karakter van de TPNRs ook in vivo getest. Enkel lage cytokine hoeveelheden werden gedetecteerd in het macrofaag supernatant en het serum van muizen die een intraperitoneale injectie kregen van TP-NRs. In de controle muizen, waar de sterk inflammatoire component LPS werd toegediend, kon wel een hoge cytokine respons teruggevonden worden. In een volgende fase werden de TPNRs herhaaldelijk intraveneus toegediend en werd de specifieke antilichaam respons op het enzym geanalyseerd. De antilichaamrespons op de nanoreactors was tot onze verbazing groter dan de respons op het vrij enzym. De nanoreactors tegenstelling gedragen tot wat we zich dus als verwachtten adjuvans een en hoge veroorzaken specifieke in anti-TP antilichaamrespons. Dit effect is geenszins voordelig voor hun gebruik in enzymvervangingstherapie aangezien dit aanleiding geeft tot een snelle eliminatie van de TP-NRs uit de bloedcirculatie. Bovendien zal de antilichaamrespons steeds groter worden na meerdere injecties, wat het therapeutisch effect van de levenslang noodzakelijke enzymvervangingstherapie teniet zal doen. Verder werd de biodistributie van de TP-NRs onderzocht en vergeleken met die van het vrij enzym aan de hand van radiogelabelde TP-NRs en radiogelabeld vrij TP. De resultaten toonden aan dat de in vivo distributie van de partikels en het vrij enzym significant verschillen. De nanoreactoren zijn vooral terug te vinden in de lever en de milt, terwijl de waarden in deze organen duidelijk lager zijn voor het vrij enzym. Belangrijk is ook dat er geen duidelijk verschil 186 Samenvatting in bloedcirculatietijden tussen de TP-NRs en het vrij TP vastgesteld kon worden. Deze data wijzen erop dat de nanoreactors geen ideale eigenschappen bezitten om te ontsnappen aan opname door het mononucleair fagocyt systeem in de lever en milt. Hierdoor zijn de TP-NRs ook niet in staat de halfwaardetijd van het enzym in het bloed aanzienlijk te verhogen. Tot slot werd onderzocht of de TP-NRs in staat zijn de toxische concentraties van de substraten in het bloed van MNGIE patiënten te verlagen. Daarvoor werden de TP-NRs geïncubeerd met plasma van een MNGIE patiënt, waarna de thymidine- en deoxyuridineconcentraties bepaald werden via HPLC. Spijtig genoeg werd vastgesteld dat de activiteit van de TP-NRs niet voldoende is om de substraatconcentraties tot een normaal niveau te verlagen. Deze eerder onverwachte resultaten dwongen ons de nanoreactorcompositie in vraag te stellen en een grondige evaluatie van de nanoreactorsamenstelling door te voeren. De uitgevoerde experimenten brachten aan het licht dat het enzym niet mooi ingekapseld zit, maar eerder geassocieerd is aan de buitenkant van het partikel, waardoor het zich in een niet-actieve conformatie bevindt. Dit verklaard meteen ook de sterke antilichaamrespons tegen het enzym. Tot dusver werd er nog geen oplossing gevonden voor dit probleem. Verder onderzoek is nodig om de oorzaak van de inkapselingsproblemen te identificeren en betere nanoreactors te vervaardigen. In conclusie werden er in deze studie enzym bevattende nanopartikels gemaakt en werd hun therapeutisch potentieel geëvalueerd in het kader van de enzymvervangingstherapie voor MNGIE. Hoewel nieuwe polymeer- gebaseerde nanoreactors geconstrueerd werden, moet besloten worden dat de TP-NRs, zoals hier beschreven, niet de ideale eigenschappen vertonen voor gebruik in belangrijke enzymvervangingstherapie. informatie vergaard over Desalniettemin de werd stabiliteit, heel wat toxiciteit en immunogeniciteit van deze nanoreactoren en is dit werk van aanzienlijk belang voor de verdere exploitatie van zulke partikels als enzymafleveringssystemen. 187 Chapter 12 Referenties [1] I. Nishino, A. Spinazzola, M. Hirano, MNGIE: from nuclear DNA to mitochondrial DNA, Neuromuscul Disord, 11 (2001) 7-10. [2] I. Nishino, A. Spinazzola, M. Hirano, Thymidine phosphorylase gene mutations in MNGIE, a human mitochondrial disorder, Science, 283 (1999) 689-692. [3] M.L. Valentino, R. Marti, S. Tadesse, L.C. Lopez, J.L. Manes, J. Lyzak, A. Hahn, V. Carelli, M. Hirano, Thymidine and deoxyuridine accumulate in tissues of patients with mitochondrial neurogastrointestinal encephalomyopathy (MNGIE), FEBS Lett, 581 (2007) 3410-3414. [4] M.C. Lara, B. Weiss, I. Illa, P. Madoz, L. Massuet, A.L. Andreu, M.L. Valentino, Y. Anikster, M. Hirano, R. Marti, Infusion of platelets transiently reduces nucleoside overload in MNGIE, Neurology, 67 (2006) 1461-1463. [5] M. Hirano, R. Marti, C. Casali, S. Tadesse, T. Uldrick, B. Fine, D.M. Escolar, M.L. Valentino, I. Nishino, C. Hesdorffer, J. Schwartz, R.G. Hawks, D.L. Martone, M.S. Cairo, S. DiMauro, M. Stanzani, J.H. Garvin, Jr., D.G. Savage, Allogeneic stem cell transplantation corrects biochemical derangements in MNGIE, Neurology, 67 (2006) 1458-1460. [6] P.F. Chinnery, J. Vissing, Treating MNGIE: is reducing blood nucleosides the first cure for a mitochondrial disorder?, Neurology, 67 (2006) 1330-1332. [7] T.A. Burrow, R.J. Hopkin, N.D. Leslie, B.T. Tinkle, G.A. Grabowski, Enzyme reconstitution/replacement therapy for lysosomal storage diseases, Curr Opin Pediatr, 19 (2007) 628-635. [8] W.R. Wilcox, M. Banikazemi, N. Guffon, S. Waldek, P. Lee, G.E. Linthorst, R.J. Desnick, D.P. Germain, Long-term safety and efficacy of enzyme replacement therapy for Fabry disease, Am J Hum Genet, 75 (2004) 65-74. [9] N.W. Barton, R.O. Brady, J.M. Dambrosia, A.M. Di Bisceglie, S.H. Doppelt, S.C. Hill, H.J. Mankin, G.J. Murray, R.I. Parker, C.E. Argoff, et al., Replacement therapy for inherited enzyme deficiency--macrophage-targeted glucocerebrosidase for Gaucher's disease, N Engl J Med, 324 (1991) 14641470. [10] D.T. Achord, F.E. Brot, C.E. Bell, W.S. Sly, Human beta-glucuronidase: in vivo clearance and in vitro uptake by a glycoprotein recognition system on reticuloendothelial cells, Cell, 15 (1978) 269-278. [11] J.M. Harris, R.B. Chess, Effect of pegylation on pharmaceuticals, Nat Rev Drug Discov, 2 (2003) 214-221. [12] M.S. Hershfield, PEG-ADA replacement therapy for adenosine deaminase deficiency: an update after 8.5 years, Clin Immunol Immunopathol, 76 (1995) S228-232. [13] V. Gaberc-Porekar, I. Zore, B. Podobnik, V. Menart, Obstacles and pitfalls in the PEGylation of therapeutic proteins, Curr Opin Drug Discov Devel, 11 (2008) 242-250. [14] T.M. Allen, P.R. Cullis, Drug delivery systems: entering the mainstream, Science, 303 (2004) 1818-1822. [15] P. Walde, S. Ichikawa, Enzymes inside lipid vesicles: preparation, reactivity and applications, Biomol Eng, 18 (2001) 143-177. 188 Samenvatting [16] M. Yamauchi, K. Tsutsumi, M. Abe, Y. Uosaki, M. Nakakura, N. Aoki, Release of drugs from liposomes varies with particle size, Biol Pharm Bull, 30 (2007) 963-966. [17] A. Gabizon, R. Chisin, S. Amselem, S. Druckmann, R. Cohen, D. Goren, I. Fromer, T. Peretz, A. Sulkes, Y. Barenholz, Pharmacokinetic and imaging studies in patients receiving a formulation of liposome-associated adriamycin, Br J Cancer, 64 (1991) 1125-1132. [18] A.L. Klibanov, K. Maruyama, V.P. Torchilin, L. Huang, Amphipathic polyethyleneglycols effectively prolong the circulation time of liposomes, FEBS Lett., 268 (1990) 235-237. [19] P. Laverman, M.G. Carstens, O.C. Boerman, E.T. Dams, W.J. Oyen, N. van Rooijen, F.H. Corstens, G. Storm, Factors affecting the accelerated blood clearance of polyethylene glycol-liposomes upon repeated injection, J Pharmacol Exp Ther, 298 (2001) 607-612. [20] D.E. Discher, Emerging applications of polymersomes in delivery: From molecular dynamics to shrinkage of tumors, Progress in Polymer Science, 32 (2007) 838-857. [21] D.E. Discher, A. Eisenberg, Polymer vesicles, Science., 297 (2002) 967973. [22] R. Gref, Y. Minamitake, M.T. Peracchia, V. Trubetskoy, V. Torchilin, R. Langer, Biodegradable long-circulating polymeric nanospheres, Science, 263 (1994) 1600-1603. [23] J. Panyam, V. Labhasetwar, Biodegradable nanoparticles for drug and gene delivery to cells and tissue, Adv Drug Deliv Rev, 55 (2003) 329-347. [24] C. Nardin, S. Thoeni, J. Widmer, M. Winterhalter, W. Meier, Nanoreactors based on (polymerized) ABA-triblock copolymer vesicles, Chem Comm, (2000) 1433-1434. [25] A. Ranquin, W. Versees, W. Meier, J. Steyaert, P. Van Gelder, Therapeutic Nanoreactors: Combining Chemistry and Biology in a Novel Triblock Copolymer Drug Delivery System, Nano Lett, 5 (2005) 2220-2224. 189 Chapter 13 Publications and scientific manifestations Publications and scientific manifestations 13 Publications and scientific manifestations 13.1 Publications De Vocht C., Ranquin A., Willaert R., Van Ginderachter J., Vanhaecke T., Rogiers V., Versées W., Van Gelder P. and Steyaert J., Assessment of stability, toxicity and immunogenicity of new polymeric nanoreactors for use in enzyme replacement therapy of MNGIE. J. Contr. Release, 137 (3), 246-254, 2009. Ranquin A., De Vocht C. and Van Gelder P., Polymer-based nanoreactors for medical applications. In: Polymer-based nanostructures: Medical Applications. Broz P., Editor. Royal Society of Chemistry. 315-332, 2010. Ranquin A., De Vocht C., Wilkinson H., Steyaert J. and Van Gelder P., Prodrug activation by enzyme encapsulating triblock copolymer nanoreactors. In submission. 13.2 Scientific manifestations 11th European Symposium on Controlled Drug Delivery, April 2010, Egmond aan Zee, The Netherlands, publication in abstract book. 13th International Symposium on Purine and Pyrimidine Metabolism in Man, June 2009, Stockholm, Sweden, oral presentation. SBE’s Fourth International Conference on Bioengineering and Nanotechnology, July 2008, Dublin, Ireland, poster presentation. Knowledge for Growth, June 2008, Gent, Belgium, oral presentation. Mini-symposium on polymeric drug delivery systems, April 2008, Brussels, Belgium, oral presentation. 193 Chapter 13 VIB Seminar 2008, March 2008, Blankenberge, Belgium, poster presentation. Recent Advances in Drug Delivery Systems, February 2007, Salt Lake City, Utah, USA, poster presentation. Symposium of the Belgian Society of Biochemistry, December 2006, Gembloux, Belgium, poster presentation. NSTI Nanotech, Mai 2006, Boston, Massachusetts, USA, poster Belgium, poster presentation. VIB Seminar presentation. 194 2006, March 2006, Blankenberge,