UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Biophysical characterization of lipid nanoparticles containing nucleic acid polymers as produced by microfluidic… Leung, Alex Kar-Kei 2014

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2014_september_leung_alex.pdf [ 6.22MB ]
Metadata
JSON: 24-1.0165976.json
JSON-LD: 24-1.0165976-ld.json
RDF/XML (Pretty): 24-1.0165976-rdf.xml
RDF/JSON: 24-1.0165976-rdf.json
Turtle: 24-1.0165976-turtle.txt
N-Triples: 24-1.0165976-rdf-ntriples.txt
Original Record: 24-1.0165976-source.json
Full Text
24-1.0165976-fulltext.txt
Citation
24-1.0165976.ris

Full Text

BIOPHYSICAL CHARACTERIZATION OF LIPID NANOPARTICLES CONTAINING NUCLEIC ACID POLYMERS AS PRODUCED BY MICROFLUIDIC MIXING   by   Alex Kar-Kei Leung   B.Sc. (Hons.), The University of Calgary, 2007    A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF  THE REQUIREMENTS FOR THE DEGREE OF   DOCTOR OF PHILOSOPHY  in   The Faculty of Graduate and Postdoctoral Studies  (Biochemistry and Molecular Biology)    THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)   August 2014   ©Alex Kar-Kei Leung, 2014  ii  Abstract Lipid nanoparticles (LNP) are currently the most advanced delivery systems for enabling siRNA to be used for therapeutic applications.  However, the structure of these siRNA-LNP systems has not been well defined previously.  The objective of this thesis is to determine the structure of nucleic acid-LNP systems produced by a novel microfluidic mixing technique and to use this structural understanding to develop systems with improved gene silencing efficacy.   The first part of the thesis focuses on determining the structure of siRNA-LNP systems produced by microfluidic mixing and the effects of varying lipid components on the structure and encapsulation properties.  SiRNA-LNP were formulated using an ionizable cationic lipid, distearoylphosphatidylcholine, cholesterol and a polyethylene glycol-lipid.  Cryo-TEM of siRNA-LNP produced by microfluidic mixing exhibit a solid, electron-dense core with siRNA encapsulation efficiency close to 100%.  Molecular dynamics modeling indicates that the core of the particle consists of periodic aqueous compartments containing siRNA.  The ability of the lipid mixture to adopt non-bilayer phases seems to be crucial for the encapsulation of siRNA.  The microfluidic mixing technology was also extended to the encapsulation of plasmid DNA and mRNA.  These results provide an understanding of the structure and the mechanism of formation for siRNA-LNP produced by microfluidic mixing. Since it is clear that formation of siRNA-LNP by microfluidic mixing does not require any bilayer-forming lipids, it is possible to generate particles with high cationic lipid content and bilayer-destabilizing lipids without compromising particle stability.  The last part of the thesis aims to improve the gene silencing efficacy of siRNA-LNP by enhancing the endosomolytic properties of the LNP.  This was attempted by two way: first by increasing the cationic lipid content of the LNP and second, by the incorporation of bilayer-destabilizing "helper" lipids in iii  the formulation.  A novel "helper" lipid dioleoyl-four amino butyric acid (DOFAB) was synthesized for this purpose.  In contrary to the hypothesis, both increasing cationic lipid content and the incorporation of "helper" lipids in siRNA-LNP formulations led to decreased in vitro gene silencing activity.  This leads to new questions as to role of cationic lipids and "helper" lipids play in the intracellular release of siRNA.  iv  Preface All biophysical experiments, including formulation of siRNA-LNP, characterization and in vitro experiments were performed by myself.  Cryogenic transmission electron microscopy of siRNA-LNP was performed by myself with assistance from Bradford Ross and Garnet Martens in the  BioImaging Facility at The University of British Columbia.  Molecular dynamics simulation of the structure of siRNA-LNP in Chapter 2 was performed by Dr. D. Peter Tieleman, Dr. Svetlana Baoukina and Elham Afshinmanesh from The University of Calgary.  Plasmid DNA and MC 53 messenger RNA used in Chapter 3 were prepared by Dr. Paulo J.C. Lin and Dr. Qing Wang, respectively, from The University of British Columbia.  The novel reverse-headgroup lipid, dioleoyl-four amino butyric acid (DOFAB), used in Chapter 4 was synthesized by Dr. Josh Zaifman from The University of British Columbia. Experimental designs, ideas and data analysis were performed by myself with important contributions from Dr. Pieter R. Cullis, Dr. Ismail M. Hafez, Dr. Igor Zhigaltsev, Dr. Yuen Yi C. Tam, Dr. Paulo J.C. Lin and Mr. Sam Chen.  I was responsible for writing the entire thesis with the exception of the section involving the molecular dynamics simulation in Chapter 2, which was written by Dr. D. Peter Tieleman and the section involving the synthesis of DOFAB, which was written by Dr. Josh Zaifman.  Dr. Pieter R. Cullis and Dr. Yuen Yi C. Tam was responsible for editing this thesis. Chapter 2 of this thesis regarding the structure of siRNA-LNP produced by microfluidic mixing has been published in a peer-reviewed journal.   Alex K.K. Leung, Ismail M. Hafez, Svetlana Baoukina, Nathan M. Belliveau, Igor V. Zhigaltsev, Elham Afshinmanesh, D. Peter Tieleman, Carl L. Hansen, Michael J. Hope and Pieter R. Cullis. 2012. Lipid nanoparticles v  containing siRNA synthesized by microfluidic mixing exhibit an electron-dense nanostructured core.  Journal of Physical Chemistry C 116: 18440-18450  vi  Table of Contents Abstract .......................................................................................................................................... ii Preface ........................................................................................................................................... iv Table of Contents ......................................................................................................................... vi List of Tables ................................................................................................................................. x List of Figures ............................................................................................................................... xi List of Abbreviations ................................................................................................................. xiv Acknowledgements .................................................................................................................... xix Chapter 1: Introduction ............................................................................................................... 1 1.1 Lipids in Biomembranes ....................................................................................................... 1 1.1.1 Lipid diversity in biomembranes .................................................................................... 1 1.1.2 Major classes of membrane lipids .................................................................................. 2 1.2 The Endocytic Pathway ......................................................................................................... 8 1.2.1 Molecular machinery of clathrin-mediated endocytosis ................................................ 8 1.2.2 Molecular itineraries along the clathrin-mediated endocytic pathway ......................... 10 1.2.3 Anionic lipids of the endocytic pathway ...................................................................... 14 1.3 Lipid Polymorphism and Membrane Fusion ....................................................................... 15 1.3.1 Phase behaviour of different lipids ............................................................................... 15 1.3.2 Role of lipid polymorphism in the fusion of lipid membranes..................................... 18 1.3.3 Phase behaviour of mixtures of cationic and anionic lipid ........................................... 21 1.4 Liposomes and Their Application in Drug Delivery ........................................................... 22 1.4.1 Early developments of liposomes ................................................................................. 22 1.4.2 Liposomes as drug delivery vehicles ............................................................................ 25 1.5 Delivery of Nucleic Acid-Based Therapeutics.................................................................... 29 1.5.1 Formulation of plasmid nanoparticles .......................................................................... 29 1.5.2 RNA interference .......................................................................................................... 32 1.5.3 siRNA as therapeutics .................................................................................................. 33 1.5.4 Lipid components of siRNA-lipid nanoparticle (LNP) systems .................................. 35 1.5.5 Formulation of siRNA-lipid nanoparticles by extrusion .............................................. 41 1.5.6 Formulation of siRNA-lipid nanoparticles by T-tube in-line mixing........................... 45 vii  1.5.7 Formulation of siRNA-lipid nanoparticles by microfluidic mixing ............................. 45 1.5.8 Solid-core lipidic systems in nature.............................................................................. 48 1.6  Thesis Objective ................................................................................................................. 49 Chapter 2: Lipid Nanoparticles Containing siRNA Synthesized by Microfluidic Mixing Exhibit an Electron-Dense Nanostructured Core .................................................................... 53 2.1 Introduction ......................................................................................................................... 53 2.2 Materials and Methods ........................................................................................................ 54 2.2.1 Materials ....................................................................................................................... 54 2.2.2 Preparation of POPC/cholesterol bilayer vesicles. ....................................................... 55 2.2.3 Preparation of siRNA-LNP systems. ............................................................................ 55 2.2.4 Cryo-TEM. ................................................................................................................... 57 2.2.5 RNase protection assay. ................................................................................................ 57 2.2.6 31P-NMR studies. .......................................................................................................... 58 2.2.7 FRET membrane fusion studies. .................................................................................. 58 2.2.8. Sucrose density gradient centrifugation. ..................................................................... 59 2.2.9. Computer simulation of siRNA-LNP systems. ........................................................... 60 2.3 Results ................................................................................................................................. 60 2.3.1 Microfluidic mixing allows highly efficient encapsulation of siRNA in siRNA-LNP systems over a wide range of siRNA/lipid charge ratios ....................................................... 60 2.3.2 LNP systems exhibit an electron dense core structure as indicated by cryo-TEM ...... 62 2.3.3 LNP containing cationic lipid exhibit limit sizes consistent with the formation of inverted micellar structures in the LNP interior both in the presence and absence of siRNA ............................................................................................................................................... 64 2.3.4 Encapsulated siRNA is immobilized in the siRNA-LNP system ................................. 67 2.3.5 Encapsulated siRNA is fully protected from degradation by external RNase A .......... 69 2.3.6 Encapsulated siRNA is complexed with internalized cationic lipid............................. 70 2.3.7 LNP siRNA systems have a different density than aqueous core bilayer vesicles ....... 72 2.3.8 Simulation results indicate that LNP siRNA systems exhibit a nanostructured core ... 74 2.4 Discussion ........................................................................................................................... 80 Chapter 3: Structural Properties of Lipid Nanoparticles Containing Nucleic Acid Polymers Formulated by Microfluidic Mixing.......................................................................................... 83 viii  3.1 Introduction ......................................................................................................................... 83 3.2 Materials and Methods ........................................................................................................ 84 3.2.1 Materials ....................................................................................................................... 84 3.2.2 Encapsulation of nucleic acids ..................................................................................... 85 3.2.3 Cryo-TEM .................................................................................................................... 86 3.3 Results ................................................................................................................................. 87 3.3.1 Visualization of LNP containing siRNA conjugated to gold nanoparticles supports the nanostructured core model for siRNA-LNP systems ............................................................ 87 3.3.2 The nanostructure of LNP systems produced by microfluidic mixing is dependent on lipid composition ................................................................................................................... 89 3.3.3 LNP siRNA systems containing higher levels of siRNA exhibit reduced lamellar structure. ................................................................................................................................ 92 3.3.4 Reduced LNP siRNA encapsulation efficiencies observed at high cationic lipid contents can be improved by incorporation of DOPE ........................................................... 94 3.3.5 Microfluidic mixing can be used to encapsulate longer nucleic acid polymers such as plasmids and mRNA into LNP systems ................................................................................ 99 3.4 Discussion ......................................................................................................................... 102 Chapter 4: Influence of Helper Lipids and High Cationic Lipid Contents on the Transfection Potency of siRNA-Lipid Nanoparticles Formulated by Microfluidic Mixing..................................................................................................................................................... 107 4.1 Introduction ....................................................................................................................... 107 4.2. Materials and Methods ..................................................................................................... 108 4.2.1. Materials .................................................................................................................... 108 4.2.2 Cell culture ................................................................................................................. 109 4.2.3 TNS assay ................................................................................................................... 109 4.2.4 31P NMR spectroscopy ............................................................................................... 110 4.2.5 Encapsulation of nucleic acids ................................................................................... 110 4.2.6. FRET membrane fusion studies. ............................................................................... 111 4.2.7 Gene knockdown ........................................................................................................ 112 4.2.8 Confocal microscopy .................................................................................................. 113 4.3 Results ............................................................................................................................... 113 ix  4.3.1 LNP siRNA systems containing DOPE and high cationic lipid contents are not effective gene silencing agents ............................................................................................ 113 4.3.2 Design of a new class of helper lipids ........................................................................ 117 4.3.3 DOFAB is zwitterionic at physiological pH............................................................... 119 4.3.4 DOFAB exhibits strong bilayer-destabilizing capabilities in mixture of with DOPC and cholesterol. .................................................................................................................... 121 4.3.5 Incorporation of DOFAB into lipid nanoparticles does not affect encapsulation efficiencies ........................................................................................................................... 123 4.3.6 LNP formulated with DOFAB have enhanced fusion capability ............................... 126 4.3.7 LNP containing DOFAB are not as effective in mediating gene silencing in vitro ... 128 4.4. Discussion ........................................................................................................................ 131 Chapter 5: Summarizing Discussion and Future Directions ................................................ 134 Bibliography .............................................................................................................................. 143    x  List of Tables Table 2.1  Numbers of nearest neighbors in the first coordination shell for each molecule in the LNP core. ...................................................................................................................................... 78   xi  List of Figures Figure 1.1  Common glycerophospholipids in the cell membrane. ................................................ 6 Figure 1.2  Structure of ceramide and sphingomyelin .................................................................... 7 Figure 1.3  Structure of cholesterol................................................................................................. 7 Figure 1.4  Molecular iternary through the clathrin-mediated endocytic pathway....................... 13 Figure 1.5  Polymorphic phase behaviour of lipids according to the molecular shape hypothesis........................................................................................................................................................ 17 Figure 1.6  The "stalk" model of membrane fusion ...................................................................... 20 Figure 1.7  Encapsulation of small molecule drugs in response to transmembrane pH gradient . 28 Figure 1.8  Encapsulation of nucleic acids with spontaneous vesicle formation (SVF) by ethanol dilution .......................................................................................................................................... 31 Figure 1.9  Evolution of cationic lipids ........................................................................................ 40 Figure 1.10  Encapsulation of nucleic acid by the "stabilized anti-sense lipid particle" (SALP) method........................................................................................................................................... 43 Figure 1.11  Encapsulation of nucleic acid by the preformed vesicle (PFV) method .................. 44 Figure 1.12  Encapsulation of siRNA by microfluidic mixing ..................................................... 47 Figure 1.13  Lipid components of siRNA-LNP system formulated by microfluidic mixing ....... 52 Figure 2.1  siRNA encapsulation efficiency of siRNA-LNP systems prepared by microfluidic mixing at various siRNA/lipid ratios ............................................................................................ 61 Figure 2.2  LNP containing DLin-KC2-DMA exhibit electron dense cores both in the presence and absence of encapsulated siRNA as indicated by cryo-TEM .................................................. 63 Figure 2.3  LNP exhibit limit sizes consistent with inverted micelle structure in presence and absence of siRNA ......................................................................................................................... 66 xii  Figure 2.4  Encapsulated siRNA is immobilized on the NMR timescale ..................................... 68 Figure 2.5  siRNA encapsulated in LNP is fully protected from external RNase ........................ 69 Figure 2.6  Cationic lipid is associated with internalized siRNA in siRNA-LNP systems .......... 71 Figure 2.7  The density of siRNA-LNP systems is consistent with a hydrophobic lipid core as indicated by density gradient ultracentrifugation ......................................................................... 73 Figure 2.8  Self-assembly from a random configuration (a) into a building block (b) for a lipid nanoparticle (LNP)........................................................................................................................ 76 Figure 2.9  A lipid nanoparticle (LNP) contains irregular water-filled cavities separated by bilayer membranes, with nucleic acids bound to membrane surfaces .......................................... 77 Figure 2.10  Spatial density distributions for selected molecular groups around DSPC and DLin-KC2-DMA .................................................................................................................................... 79 Figure 3.1  Cryo-TEM of LNP produced with gold nanoparticle-conjugated siRNA.................. 88 Figure 3.2  Cryo TEM micrographs of LNP produced with different cationic lipid saturation and total cationic lipid content ............................................................................................................. 91 Figure 3.3  LNP formulated with 20% DOKC2-DMA exhibit a different proportion of multilamellar structures depending on siRNA content ................................................................. 93 Figure 3.4  Effects of cationic lipid and phospholipid contents on the siRNA encapsulation efficiency of lipid nanoparticle systems ....................................................................................... 97 Figure 3.5  LNP-siRNA systems with high cationic lipid contents maintain solid core structures for both DSPC and DOPE............................................................................................................. 98 Figure 3.6  Influence of cationic lipid content and phospholipid species on the plasmid and mRNA encapsulation efficiencies of lipid nanoparticle systems ............................................... 100 xiii  Figure 3.7  LNP containing plasmid DNA and mRNA exhibit blebs when DSPC is present but not when DOPE is substituted for DSPC .................................................................................... 101 Figure 4.1  Comparison of the knockdown of the androgen receptor (AR) in LNCap cells using LNP composed with 80% cationic lipid and DOPE ................................................................... 115 Figure 4.2  Influence of lipid content on the cellular uptake of DLin-KC2-DMA siRNA LNP 116 Figure 4.3  Proposed mechanism of the membrane disruptive effects of reverse-headgroup lipids..................................................................................................................................................... 118 Figure 4.4  Determination of the pKa’s of DOFAB using the TNS Assay ................................ 120 Figure 4.5  DOFAB exhibits strong bilayer-destabilizing abilities in mixtures with DOPC and cholesterol ................................................................................................................................... 122 Figure 4.6  siRNA encapsulation efficiency of LNP siRNA systems containing various amounts of DOFAB ................................................................................................................................... 125 Figure 4.7  Incorporation of DOFAB into LNP systems increases the fusion capability ........... 127 Figure 4.8  Knockdown of the androgen receptor (AR) in LNCaP cells with either DOPE or DOFAB as the helper lipid ......................................................................................................... 129 Figure 4.9  Influence of the helper lipid on the cellular uptake of DLin-KC2-DMA siRNA LNP..................................................................................................................................................... 130 Figure 5.1  Assembly of siRNA-LNP by microfluidic mixing ................................................... 140 Figure 5.2  The conical lipid, DOPE, increases the propensity of the lipid mixture to form inverted micellar structures, leading to the encapsulation of siRNA.......................................... 141 Figure 5.3  Cryo TEM of gold nanoparticles encapsulated in LNP by microfluidic mixing. .... 142    xiv  List of Abbreviations 31P NMR phosphorus-31 nuclear magnetic resonance Ac cross-sectional area at the ends of the acyl chains of the lipid molecule Ago2 Argonaute 2 Ah optimal cross-sectional area per lipid headgroup AP-2 adaptor protein 2 apoA1 apolipoprotein A1 apoE E apolipoprotein E ApoB apolipoprotein B apoB-100 apolipoprotein B-100 ARH autosomal recessive hypercholesterolemia CG  coarse-grained CHEMS cholesteryl hemisuccinate CNV choroidal neovascurization cryo-TEM cryogenic transmission microscopy Dab2 disabled-2 DLinKC2-DMA 2,2-dilinoleyl-4-(2-dimethylaminoethyl)-[1,3]-dioxolane DNA deoxyribonucleic acid DODAC N,N-dioleoyl-N,N-dimethylammonium chloride DODAP 1,2-dioleoyl-3-dimethylammonium propane DOFAB dioleoyl-four amino butyric acid DOKC2-DMA 2,2-dioleoyl-4-(2-dimethylaminoethyl)-[1,3]-dioxolane DOPE 1, 2 dioleoyl-sn-glycero-3-phosphoethanolamine DOPS 1,2-dioleoyl-sn-glycero-3-phosphoserine xv  DOTMA N-[1-(2,3-dioleyloxy) propyl]-N,N,N-trimethylammonium chloride DPPC 1,2-dipalmitoyl-sn-glycero-3-phosphocholine DSPC 1,2-distearoyl-sn-glycero-3-phosphocholine DSPE 1,2-distearoyl-sn-glycero-3- phosphoethanolamine DUPA 2-[3-(1,3-dicarboxypropyl)-ureido]pentanedioic acid EDTA Ethylenediaminetetraacetic acid EEA1 early endosome antigen 1 eGFP enhanced green fluorescent protein EGFR epidermal growth factor receptor EPR enhanced permeation and retention ER endoplasmic reticulum ESCRT endosomal sorting complexes required for transport F fluorescent Intensity FCHO Fer/Cip4 homology domain-only FID free induction decays Fmax maximum fluorescent intensity Fo initial fluorescent intensity FRET fluorescent resonance energy transfer FYVE Fab1, YOTB, Vac1, EEA1 domain gp41 glycoprotein 41 GPCR G-protein coupled receptors HEPES 4-(2-Hydroxyethyl) piperazine-1-ethanesulfonic acid HIV-1 human immunodeficiency virus 1 HRP horseradish peroxidase HRS hepatocyte growth-factor-regulated tyrosine kinase substrate xvi  HSC70 heat shock cognate 70 i.v. intravenous IDL intermediate-density lipoprotein LBPA lysobisphosphatidic acid lc length of the acyl chain LDL low-density lipoprotein LDLR low-density lipoprotein receptors LNCap androgen-sensitive human prostate adenocarcinoma LNP lipid nanoparticles LUV large unilamellar vesicle lyso-PC lyso-phosphatidylcholine M6PR mannose-6-phosphate receptor MES 2-(N-Morpholino) ethanesulfonic acid MLV multilamellar vesicle mRNA messenger RNA MVB multivesicular bodies NBD-PE 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine-N-(7-nitro-2-1,3-benzoxadiazol-4-yl) NP-40 Nonidet P-40 (Igepal CA-630) NSF N-ethylmaleimide-sensitive factor PA phosphatidic acid PBS phosphate buffered saline PBSCM PBS supplemented with 1 mM MgCl2 and 0.1 mM CaCl2 PC phosphatidylcholine PE phosphatidylethanolamine xvii  PEG polyethylene glycol PEG-c-DMA N-[(methoxy polyethylene glycol 2000 carbamyl]-1,2 dimyristyloxlpropyl-3-amine PEI polyethylenimine PFV pre-formed vesicle PG phosphatidylglycerol PI phosphatidylinositol PI(3)P phosphatidylinositol 3-phosphate PI(3,4,5)P3   phosphatidylinositol (3,4,5)-triphosphate PI(3,5)P2 phosphatidylinositol 3,5-bisphosphate PI(4)P phosphatidylinositol 4-phosphate PI(4,5)P2 phosphatidylinositol 4,5-bisphosphate PLGA and poly(lactic-co-glycolic acid) (PLGA) POPC 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine PS phosphatidylserine PSMA prostate specific membrane antigen PTGS post-transcription gene silencing RDF radial distribution function RES reticuloendothelial system Rh-PE 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine-N-(lissamine rhodamine B sulfonyl) RISC RNA-induced gene silencing complex RNA ribonucleic acid RNAi RNA interference RPMI Roswell Park Memorial Institute medium xviii  RSV respiratory syncytical virus S lipid shape parameter SALP stabilized anti-sense lipid particle SARS Severe acute respiratory syndrome SDS Sodium dodecyl sulfate siRNA short-interfering RNA siRNA-LNP siRNA-lipid nanoparticles SIV Simian immunodeficiency virus SM sphingomyelin SNALP stable nucleic acid lipid particle SNAP soluble NSF attachment proteins SNARE soluble NSF attachment protein receptors SPLP stabilized plasmid-lipid particles SVF spontaneous vesicle formation TH lamellar-to-hexagonal HII transition temperature TNS 2-(p-toluidino)-6-napthalene sulfonic acid v volume per lipid molecule v-ATPase vacuolar-type H+-ATP hydrolases VEGF Vascular endothelial growth factor VLDL very-low density lipoprotein   xix  Acknowledgements I would like to thank my supervisor Dr. Pieter R. Cullis for giving me an opportunity to work in his laboratory.  Pieter introduced me to the exciting field of lipid nanoparticle delivery technologies and has taught me a lot not only in the field of lipids but also on how to be a successful scientist.  Working with Pieter is a real inspiration.  I would also like to thank my committee members, Dr. Marcel Bally and Dr. Franck Duong for their support and valuable advice throughout my studies. Members of the Cullis lab are extremely helpful and supportive through-out the years and I owe many thanks to them.  Many thanks to Dr. Chris Tam and Dr. Mick Hope for their valuable experimental advice and for making sure the lab runs smoothly.  Ms. Cayetana Schluter has been extremely helpful in organizing meetings with Pieter and for making sure everyone has a lab bench to work.  Thank you to Dr. Ying Tam, Dr. Paulo Lin, Dr. Justin Lee, Mr. Andrew Cottle, Mr. Jayesh Kulkarni, Ms. Joslyn Quick and Mr. Sam Chen for making the lab such a fun place to work.  I would also like to thank Dr. Barb Mui, Dr. Chen Wan, Dr. Genc Basha, Dr. Igor Zhigaltsev, Dr. Ismail Hafez, Dr. Kaley Wilson, Dr. Kim Wong, Dr. Rob Fraser, Dr. Theresa Allen, Ms. Mina Ordobadi, Mr. Raymond Pan, Mr. Oleg Sannikov, Ms. Yan Liu for their support, assistance and helpful advice. Many thanks to my fellow graduate students in the Department of Biochemistry and Molecular Biology for their support and for making the Department such a fun place to be.  Special thanks to Ms. Colette Chiu for keeping my sanity in check and for not increasing the price of her psychiatric booth despite rampant inflation in recent years. xx  I would like to thank my parents, Wilson and Alice, and my brothers, Leslie and Louis, for their love and support.  Thank you for all the years of support, for looking after me and for letting me know that home is always the best place to be. A very special thank you to Aleeza for everything.  Thank you for offering a place of comfort when it is needed.  Thank you for letting me know what is the most important thing in life.  Thank you for that wonderful smile that tells me everything is going to be okay. 1  Chapter 1: Introduction 1.1 Lipids in Biomembranes The lipid bilayer membrane is an indispensable structural component of biological membranes and is essential to life on Earth. It is difficult to imagine how life could have evolved on Earth without the controlled interior environment provided by a lipid membrane.  For example, it is likely that synthesis and polymerization of the first genetic materials, such as DNA or RNA, required concentration of chemical precursors in the compartment provided by a lipid bilayer (Luisi et al., 1999).  It has been suggested that the compartmentalization provided by a lipid envelope facilitated the high concentrations necessary to promote these reactions (Deamer and Oro, 1980).  Besides separating the "living" interior from the external environment, lipid membranes also surround organelles  within the cell in which specialized reactions take place.  For example, the membrane of the biosynthetic endoplasmic reticulum (ER) provides a platform for the synthesis of hydrophobic membrane proteins (Hegde and Keenan, 2011; Shao and Hegde, 2011).  The membrane of the lysosome provides a barrier to segregate degradative enzymes, such as acid hydrolases, from the cytosol (Luzio et al., 2007).  The lipid bilayer can also maintain an electrochemical gradient generated by proton pumps which is used in mitochondria to generate the ATP vital to sustain life. Interestingly, drugs that are weak bases can be accumulated into lipid vesicles exhibiting a pH gradient (Mayer et al., 1986a; Nichols and Deamer, 1976).  The use of pH gradients for the encapsulation of drugs in liposomal delivery systems is discussed in more detail in Section 1.3.2.  1.1.1 Lipid diversity in biomembranes Biological membranes contain a huge diversity of lipids, possibly as many as 10,000 to 100,000 different lipid species in a cell (Wenk, 2010).  Although it is unclear why so many 2  different lipid species are present, it is probable that they regulate diverse cellular functions.  Different vesicles along the endocytic pathway have unique composition of phosphatidylinositols (van Meer et al., 2008).  Lipids influence the function of many proteins; β-adrenergic receptors and caveolin preferentially interact with cholesterol (Hanson et al., 2008; Murata et al., 1995).  The function of respiratory chain complexes III and IV in the mitochondria requires the presence of cardiolipin in the mitochondrial membrane whereas the activation of rhodopsin in photoreceptor rod outer segments is dependent on the presence of omega-3 fatty acids (Niu et al., 2004; Pfeiffer et al., 2003).  Lipid components can regulate the localization of membrane proteins, membrane proteins are proposed to be sorted to the plasma membrane based on their association with sphingolipids and cholesterol (Simons and Ikonen, 1997).  Also, membrane proteins have been hypothesized to preferentially reside in membranes with a thickness matching that of their transmembrane domains (Kaiser et al., 2011). Lipids such as ceramides, diacylglycerols and phosphatidylinositols have important roles in cellular signalling pathways (Almena and Mérida, 2011).  Finally, many different lipids are known to adopt a variety of non-bilayer phases that have been hypothesized to modulate membrane fusion, as discussed in more detail in Section 1.2 (Cullis et al., 1986).  1.1.2 Major classes of membrane lipids Glycerophospholipids.  Glycerophospholipids are the most common lipid species found in eukaryote cell membranes.  They consist of a glycerol backbone attached to fatty acid chains via ester linkages in the sn-1 and sn-2 positions (Coskun and Simons, 2011).  The fatty acid chains can be variable in length and saturation, with palmitic acid (C16:0) and oleic acid (C18:1 cis Δ9) being the most common in plasma membranes such as the erythrocyte membrane.  The sn-3 position of the glycerol backbone is modified by various phosphate-containing headgroups to 3  produce phosphatidylcholine (PC), phosphatidylethanolamine (PE), phosphatidylserine (PS), phosphatidylglycerol (PG), phosphatidic acid (PA) and phosphatidylinositol (PI)  (Figure 1.1).  Phosphatidylcholine is the main lipid of most cellular membranes and accounts for approximately 50% of all phospholipids in the cell (van Meer, 2005).  The next most common glycerophospholipid in the membrane is PE, which is mainly located in the cytoplasmic leaflet of the plasma membrane.  It possesses an overall conical molecular shape because of its relatively small polar headgroup and adopts inverted non-bilayer phases such as the hexagonal HII phase in isolation.  It has been hypothesized that non-bilayer lipids such as PE impose a curvature stress on the bilayer membrane facilitating membrane fusion, budding, and modulate the localization and activities of membrane proteins (Chernomordik and Kozlov, 2008).  Another phospholipid that normally resides in the cytosolic leaflet of the plasma membrane is the negatively-charged PS; however, it appears on the exoplasmic leaflet of eukaryotic cells during apoptosis and blood coagulation mediated by platelets (Fadok et al., 1998; Lentz, 2003).  Glycerophospholipids also have a wide diversity of signalling functions and the prime examples are PI and PA (English, 1996; Di Paolo and De Camilli, 2006).  Many species of PI serve as molecular landmarks for various membranes, such as PI(4,5)P2 and PI(3,4,5)P3  on the plasma membrane, PI(3)P on early endosomes, PI(3,5)P2 on late endosomes and PI(4)P on the trans-Golgi network (van Meer et al., 2008; Di Paolo and De Camilli, 2006). Sphingolipids.  Sphingolipids use sphingosine as the backbone (Coskun and Simons, 2011).  Ceramide is formed by attaching a fatty acid moiety to the sphingosine backbone via an amide linkage (Fahy et al., 2005) (Figure 1.2).  Ceramide can be further modified into a variety of different sphingolipids.  For example, sphingomyelin, a major component of plasma membranes, is created when a phosphocholine is attached to ceramide. It has been suggested to segregate 4  with cholesterol to form lipid rafts enriched with certain membrane proteins (Edidin, 2003; van Meer et al., 2008).  Various carbohydrates can also be attached to ceramide, with the most common ones being glucose and galactose, forming glucosyl-ceramide and galactosylceramide respectively (van Meer, 2005).  Many of these glycosphingolipids, along with sphingomyelin, reside on the outer leaflet of the plasma membrane.  Sphingolipids are involved in a wide variety of cellular signalling events including apoptosis (ceramide and sphingosine), cell differentiation (ceramide) and inflammatory responses (spingosine-1-phosphate) (Hannun and Obeid, 2008). Sterols.  Sterols make up a class of non-polar lipids characterized by a four fused-ring structure - three 6-membered and one five-membered rings (Fahy et al., 2005).  They are synthesized by condensation of isoprene units (Fahy et al., 2005).  Sterols have a wide range of physiological functions, including hormone signalling and lipid metabolism.  The most abundant sterol, cholesterol, accounts for 20 to 25% of lipid molecules in the plasma membrane (Ikonen, 2008).  Its planar four ring hydrophobic region and a small, hydroxyl polar headgroup modulates the physical properties of biological membranes (Figure 1.3).  Cholesterol preferentially interacts with relatively saturated lipids, such as phosphatidylcholines and sphingolipids, in membranes.  The precise nature of this interaction is still unclear, but the “umbrella model” suggests that the choline headgroup of sphingolipid or PC shields cholesterol’s hydrophobic region from water (Huang and Feigenson, 1999; van Meer et al., 2008).  Membranes comprised of sphingolipids and saturated PCs can form a solid, gel phase at physiological temperature, the presence of cholesterol fluidizes gel phase lipids to form the so-called liquid-order (lo) phase.  It has been proposed that lo phases formed from sphingolipids and cholesterol laterally segregate in biomembranes to form lipid rafts (de Almeida et al., 2003).  These lipid rafts are hypothesized to play a role in the lateral localization of membrane proteins by forming thicker membranes that 5  can accommodate proteins with longer transmembrane domains (Coskun and Simons, 2011; Simons and Ikonen, 1997).     6    Figure 1.1  Common glycerophospholipids in the cell membrane.  Typical glycerophospholipids consist of two fatty acids (green) esterfied to the sn-1 and sn-2 positions of glycerol (pink).  A negatively charge phosphate (yellow) is attached to the sn-3 position of glycerol.  Glycerophospholipids are further modified by attachment of different headgroups (blue).  Cardiolipin is commonly found in the mitochondria and contains 2 phosphatidic acids bridged by a glycerol moiety.     7   Figure 1.2  Structure of ceramide and sphingomyelin Sphingolipids contain sphingosine (pink) as the backbone of the structure.  Ceramide has a fatty acid (green) attached onto the sphingosine backbone via an amide linkage.  Ceramide is a basic component of other sphingolipids.  Sphingomyelin for example consists of a negatively charged phosphate (yellow) along with a positively charged choline (blue) attached onto ceramide         Figure 1.3  Structure of cholesterol Cholesterol is the major sterol of mammalian cell membrane.  It can account for up to 25% of lipid molecules in the plasma membrane.  It consists for 4 large hydrophobic rings with a relatively small, polar hydroxyl headgroup.  8  1.2 The Endocytic Pathway Endocytosis is the process by which cells use to internalize a variety of molecules and cell-surface receptors.  There are numerous endocytic pathways utilized by cells, including the classical clathrin-mediated endocytosis, caveolin-dependent endocytosis, phagocytosis, macropinocytosis and numerous other pathways that are independent of both clathrin and caveolin (Doherty and McMahon, 2009; Mayor and Pagano, 2007).  Recent evidence suggests that LNP enters hepatocytes through the classical clathrin-mediated endocytosis by binding to serum apoE and interacting with low-density lipoprotein receptors (LDLR) on the cell surface (Akinc et al., 2010).  The following section will discuss various aspects of clathrin-mediated endocytosis and how it may impact the intracellular release of siRNA delivered by LNP systems. 1.2.1 Molecular machinery of clathrin-mediated endocytosis Clathrin-mediated endocytosis can be separated into two categories: constitutive and ligand-induced pathways (Benmerah and Lamaze, 2007).  In constitutive pathway, which includes the LDLR and transferrin receptor pathways, the receptor is internalized regardless of whether ligand is bound onto the receptor.  In contrast, cell surface receptors in the ligand-induced pathways, which includes the epidermal growth factor receptor (EGFR) and G-protein coupled receptors (GPCR) pathways, are only internalized when ligands are bound.  Despite this difference, both pathways share some common molecular machineries.  The first stage in the endocytic pathway is the invagination of the plasma membrane and the formation of a clathrin-coated pit containing the receptor to be internalized (McMahon and Boucrot, 2011).  Recent evidence suggests that initial formation of the clathrin-coated requires Fer/Cip4 homology domain-only (FCHO) proteins as the knockdown of the FCHO proteins using RNAi results in a complete loss of the clathrin-coated pits on the plasma membrane and greatly reduced the 9  internalization of transferrin, LDL and epidermal growth factor (Henne et al., 2010).  The F-BAR domain of FCHO proteins was shown to bind to and causes extensive tubulation in liposomes containing phosphatidylinositol(4,5)P2, an anionic lipid commonly found on the inner leaflet of the plasma membrane, thus providing an force for the initial generation of the clathrin-coated pit on the planar plasma membrane (Henne et al., 2010).  FCHO proteins also recruit other scaffold proteins, such as eps15 and intersectin 1, which then recruit adaptor protein 2 (AP-2).  As its name suggests, AP-2 serves as an adaptor for the recruitment of proteins required for the formation of clathrin-coated pits and it links the receptor cargo to clathrin (Doherty and McMahon, 2009).  Knockdown of AP-2 levels significantly reduces the number of the clathrin-coated pits associated with the plasma membrane (Motley et al., 2003).  AP-2 also involves in the selection of the receptor cargo by binding to cargo-specific adaptor proteins.  For example, AP-2 binds to β-arrestin to recruit GPCR to the clathrin-coated pits and ARH (autosomal recessive hypercholesterolemia) and Dab2 (disabled-2) link LDLR to AP-2 for subsequent internalization (Benmerah and Lamaze, 2007; Ferguson et al., 1996; He et al., 2002; Morris and Cooper, 2001).  Clathrin is then recruited to the plasma membrane by AP2 and self-polymerizes into a combination of hexagons and pentagons, forming cage around the clathrin-coated pit.  Other proteins recruited to the clathrin-coated pit, such as epsin, amphiphysin and endophilins, work to increase the membrane curvature at the pit and clathrin serves to stabilize that curvature (Ford et al., 2002; Hinrichsen et al., 2006; McMahon and Gallop, 2005).   The final detachment of the clathrin-coated vesicle from the plasma membrane requires the action of the dynamin.  Dynamin is recruited to the neck of the budding vesicle and cause scission of the clathrin-coated vesicle from the plasma membrane in a GTP-dependent manner.  Studies have indicated that dynamin undergoes conformational changes upon GTP hydrolysis 10  and self-assemble into stacks of helical rings (Hinshaw and Schmid, 1995).  Upon binding to anionic liposomes, these helical rings are shown to cause constriction and tubulation of liposomes and it is implied that such constriction produced at the neck of the clathrin-coated pit induces fission of the plasma membrane which releases the clathrin-coated vesicle (Roux et al., 2006; Sweitzer and Hinshaw, 1998). 1.2.2 Molecular itineraries along the clathrin-mediated endocytic pathway Upon release from the plasma membrane, the clathrin-coated vesicle sheds its clathrin with the help of the ATPase heat shock cognate 70 (HSC70) and auxilin (Schlossman et al., 1984; Ungewickell et al., 1995).  This newly formed endocytic vesicle then fuses with other endocytic vesicles and with pre-existing sorting endosomes in a process controlled by Rab5, Early Endosome Antigen-1 (EEA1) and soluble NSF attachment protein receptors, known as SNAREs (Lawe et al., 2002).  The membrane of the sorting endosome contains many tubuluar protrusions and it is believed that receptors destined to return to the plasma membrane are sorted to these tubules due to the increased surface area-to-volume ratio of these regions (Figure 1.4; Maxfield and McGraw, 2004).  These receptors are either delivered directly to the plasma membrane from the sorting endosome or first being delivered to the endocytic recycling compartment (Maxfield and McGraw, 2004).   Cargo destined for degradation, such as LDL, are sorted to the lumen of the endosome (Maxfield and McGraw, 2004).  The lumen of sorting endosome has a pH of around 6.0 and this mildy acidic pH faciliates the release of most ligands from its receptors, which includes LDL and its receptor LDLR (Casey et al., 2010; Davis et al., 1987; Mukherjee et al., 1997).  These released ligands remain in the lumen of the sorting endosome and are delivered to the late endosome as the organelle matures (Figure 1.4).  Membrane receptors are sent to the late 11  endosome for downregulation following ubiquitylation of its cytoplasmic domain.  One example is the epidermal growth factor receptor (EGFR).  The ubiquitylated receptor is linked to clathrin through the help of hepatocyte growth-factor-regulated tyrosine kinase substrate (HRS) which is believed to retain the receptor on the endosomal membrane (Raiborg et al., 2002, 2006).  They are then transported into lumenal vesicles with the help of ESCRT (endosomal sorting complexes required for transport) complexes which invaginates the endosomal membrane into forming internal vesicles containing the receptor to be degraded (Babst et al., 2002; Gruenberg and Stenmark, 2004).  These intralumenal vesicles are enriched with the anionic phospholipid lysobisphatidic acid (LBPA), which will be discussed in more detail in the following section.   The resulting multivesicular bodies (MVB) are 300 to 400 nm in diameter with numerous intralumenal vesicles and through of the action of vacuolar-type H+-ATP hydrolases (v-ATPase) the lumenal pH is now approximately 5.5 (Casey et al., 2010; Gruenberg, 2001).  Multivesicular bodies, often referred to as late endosomes in some literature, differ from early sorting endosomes in that it does not contain any receptors to be recycled back to the plasma membrane and contains an increased level of lysosomal hydrolases (Saftig and Klumperman, 2009).  Lysosomal acid hydrolases are delivered to the late endosome by the mannose-6-phosphate receptor pathway. Mannose-6-phosphate is attached onto newly synthesized lysosomal hydrolase which is then bound onto mannose-6-phosphate receptor (M6PR) in the trans-Golgi network.  The M6PR then delivers the hydrolase to the endosome where the bound hydrolase is released into the lumen and the M6PR is recycled back to the trans-Golgi.  Lysosome can be distinguished from other endocytic compartments in that it lacks M6PR, rich in lysosomal membrane proteins, such as LAMP1 and LAMP2 and its electron-dense appearance when observed with transmission electron microscopy (Luzio et al., 2007; Saftig and Klumperman, 12  2009; Schröder et al., 2007).  Electron microscopy has provided evidence that cargo are delivered from the late endosomal MVB to the lysosome through direct fusion between the two organelles (Bright et al., 1997, 2005; Futter et al., 1996).  The fusion between late endosomes and lysosomes are mediated by an array of different proteins which includes N-ethylmaleimide-sensitive factor (NSF), soluble NSF attachment proteins (SNAP), Rab GTPases, and SNAREs (Mullock et al., 1998; Pryor et al., 2004).   It is important to note that transport the endocytic pathway is not strictly unidirectional as MVB and lysosomes can fuse with the plasma membrane, resulting in the cellular release of exosomes and lysosomal enzymes, respectively, to the extracellular milieu (Figure 1.4; Rodríguez et al., 1997; Théry et al., 2002; Trajkovic et al., 2008).  A recent study by Sahay et al (2013) suggests that these exocytic events may severely limit the efficiency siRNA delivered by lipid nanoparticles as a majority (~70%) of the internalized siRNA is transported out of the cell.  It is apparent that the itinerary taken by siRNA-LNP systems will greatly affect its gene knockdown efficacy.   13   Figure 1.4  Molecular iternary through the clathrin-mediated endocytic pathway. The first stage of the endocytic pathway involves the formation of a clathrin-coated pit containing the receptor to be internalized.  Scission from the plasma membrane is mediated through the GTPase dynamin (green circles).  The resulting endocytic vesicle then loses its clathrin (red lines) coat and fuses with pre-existing sorting endosomes or other endocytic vesicles.  Receptors destined to return to the plasma membrane are either delivered directly to the plasma membrane or being delivered to endocytic recycling compartment first.  Cargo destined for degradation are sorted to the lumen of the sorting endosome and proceed to the multivesicular bodies.  Multivesicular bodies can either deliver cargo to the lysosome for further degradation or fuse with the plasma membrane, resulting in the release of exosomes (purple circles).  Interactions between the endocytic pathway and the biosynthetic pathway (endoplasmic reticulum and Golgi complex) are omitted in this diagram for simplicity.  14  1.2.3 Anionic lipids of the endocytic pathway Different species of the negatively charged phosphatidylinositols (PI) are present in different endocytic compartments.  The major PI species on the cytoplasmic leaflet of the plasma membrane is PI(4,5)P2, which has important roles in the formation of the clathrin-coated pit by recruiting AP-2, epsin and dynamin to the plasma membrane (Ford et al., 2002; Gaidarov and Keen, 1999).  The early endosomal membrane resembles the plasma membrane in lipid composition, but as the early endosome matures, PI(3)P begins to accumulate in the membrane through phosphorylation of PI by PI(3) kinase or sequential dephosphorylation of PI(3,4,5)P3 (Di Paolo and De Camilli, 2006).  PI(3)P plays an important part in the dynamics of early endosomes.  Many major early endosomal proteins, including EEA1, recognize PI(3)P (Ellson et al., 2002; Kutateladze, 2006).  The interaction between PI(3)P and the FYVE domain of EEA1 has been shown to be important in the fusion of early endosomes (Mills et al., 1998). In addition to PI, lysobisphosphatidic acid (LBPA) also has an important role in the endocytic pathway, especially in the late endosome.  LBPA normally accounts for less than 1% of total phospholipids in the cell; however, it represents approximately 15% of phospholipids in the late endosome and 70% of phospholipids in the intraluminal vesicular bodies (Kobayashi et al., 1998, 2002).  LBPA has been shown to promote the formation of multivesicular liposomes in response to a pH gradient similar to that found in the endosome and is hypothesized to play a role in the formation of multivesicular bodies in the late endosome (Matsuo et al., 2004).  The structure of LBPA is unusual in that it possesses two glycerol groups, each attached to a single acyl chain.  This lipid is of special interest because of the role it may play in the intracellular release of nucleic acids delivered by LNP systems containing cationic lipids.  It has been proposed that cationic lipid/nucleic acid complexes enter the cell through endocytosis and 15  cationic lipids facilitate the release of the nucleic acid cargo by forming membrane-disrupting non-bilayer phases with the anionic lipids of the endosome (Hafez et al., 2001).  PI may not be available to interact with the internalized cationic lipids in the lumen of endosomes as it is primarily resides on the cytoplasmic leaflet of the endosomal membrane.  In contrast, the LBPA on the internal vesicular membranes of late endosomes could interact with the internalized cationic lipid/nucleic acid complex and form membrane disruptive ion pairs (Hafez et al., 2001).   1.3 Lipid Polymorphism and Membrane Fusion 1.3.1 Phase behaviour of different lipids The structural preferences of pure or mixed lipid systems is dependent on many factors such as acyl chain unsaturation, temperature, headgroup size, hydration state and protonation of the headgroup.  The effects of all these factors contribute to the molecular shape properties of lipids (Cullis et al., 1986).  The molecular shape of lipids is described by a dimensionless shape parameter, S, which is defined as v/Ahlc , where v is the volume per lipid molecule, Ah is the optimal cross-sectional area per lipid headgroup and lc is the length of the acyl chain (Cullis et al., 1986).  An additional parameter, Ac, which is the cross-sectional area at the ends of the acyl chains of the lipid molecule, can be introduced to illustrate the molecular shape of lipids.  When Ah/Ac is greater than 1, S will also be greater than 1, and the lipid has an inverted cone shape.  When Ah/Ac and S are roughly equal to 1, the molecular shape of the lipid is a cylinder.  When Ah/Ac is less than 1, the lipid adopts a cone shape.  Then the polymorphic phase behaviour of lipids becomes a matter of geometry in order to satisfy the packing constraints imposed by these different molecular shapes (Figure 1.5; Cullis et al., 1986).  The easiest way for cylindrically-shaped lipids to organize themselves is in a bilayer structure, whereas cone-shaped lipids form 16  structures such as the hexagonal HII phase and lipids with an inverted cone shape will assemble into micellar structures. The molecular shape model offers a useful tool for predicting and rationalizing the phase behaviour of different lipids.  For example, the difference in the lamellar-to-hexagonal HII transition temperature (TH) between distearoyl-PE (DSPE, TH = 100oC) and dioleoyl-PE (DOPE, TH = 10oC) can be explained by the fact that the acyl chains need to melt to give the molecule a more pronounced cone shape before the hexagonal HII phase can be adopted.  The smaller headgroup of PE compared to that of PC means that PE has a more conical shape, therefore explaining the stronger HII phase-forming propensity of PE as compared to PC (Cullis et al., 1986).  Deprotonation of the PE headgroup at pH 9 increases electrostatic repulsion in the headgroup resulting in a larger effective headgroup area of the lipid that allows PE to adopt the bilayer structure (Cullis and de Kruijff, 1978a).  Protonation of the anionic PS and PA reduces their effective headgroup areas leading to the formation of hexagonal HII phase (Farren et al., 1983; Hope and Cullis, 1980).  Calcium has similar effects on lipid systems containing acidic phospholipids and cardiolipin (Cullis et al., 1978a; Tilcock and Cullis, 1981).  Calcium promotes hexagonal HII phases in these mixtures by neutralizing the negative charges in the lipid headgroups, therefore decreasing the effective headgroup areas.  Dehydration is another method to decrease the effective lipid headgroup area by removing the shell of water surrounding the headgroup.  For example, PC can be forced to adopt hexagonal HII structures by dehydration at high temperature (Luzzati et al., 1968).   The molecular shape model of lipid polymorphism is further supported by the observation that inverted cone-shaped lipids such as lyso-PC and detergents can stabilize the cone-shaped DOPE in a bilayer configuration (Madden and Cullis, 1982). 17     Figure 1.5  Polymorphic phase behaviour of lipids according to the molecular shape hypothesis. The shape of a lipid molecule is dictated by a ratio of the effective cross-sectional area of the headgroup (Ah) to that of the acyl-chain region (Aa).  An Ah/Aa ratio greater than 1 gives an overall inverted cone shape and the lipids are expected to assemble into structures with a positive curvature, such as a micelle, upon hydration.  Bilayer forming lipids have roughly equal Ah and Aa, giving the lipid an overall cylindrical shape with no instrinsic curvature.  Cone-shaped lipids have a smaller headgroup compared to the area of acyl chain region, resulting in an Ah/Aa ratio of less than 1.  These lipids have a instrinsic negative curvature, which can assembled into various inverted phases upon hydration, such as hexagonal HII phase, inverted micelles and cubic phases (not shown).   18  1.3.2 Role of lipid polymorphism in the fusion of lipid membranes Current models of membrane fusion involves three distinct stages: membrane adhesion, hemifusion and pore formation (Chernomordik and Kozlov, 2008).  In membrane adhesion, membranes must come to close proximity with each other.  A force is required to drive membranes closer than 3 nm of each other because of the repulsion from the shell of water molecules that are tightly associated with the lipid headgroup (Burger, 2000).  It has been shown that direct dehydration of the lipid-water interface can promote membrane fusion of artificial lipid vesicles (MacDonald, 1985).  The attractive force driving two lipid vesicles together could arise from the hydrophobic interior of the membrane that is exposed in membrane defects (Burger, 2000).  Membrane defects can come from local curvature differences in the membrane created by differences in lipid composition and insertion of proteins or peptides (Bentz, 2000; Blumenthal et al., 2003).  Once the membrane overcomes the membrane adhesion step, fusion proceeds to a hemifused state.  This hemifusion state is characterized by a stalk formed from the fused outer monolayers of the two membranes (Chernomordik and Kozlov, 2008; Siegel, 1993, 1999).  The stalk is later expanded and ruptures into forming the fusion pore (Figure 1.6; Chernomordik and Kozlov, 2008).  The fusion pore creates a connection between the respective aqueous compartments of the two vesicles.  The formation of a fusion pore can be observed by measuring the mixing of aqueous contents of the two vesicles or by electrophysiological techniques such as patch-clamp (Blumenthal et al., 2003).   It is clear that during membrane fusion, lipids must deviate from the bilayer conformation and adopt intermediates with a high degree of curvature (Figure 1.5 and Figure 1.6).  There is considerable evidence to support a role for non-bilayer forming lipids in mediating membrane fusion.  Lipids inducing negative curvature that promote hexagonal HII phase, other factors such 19  as calcium, high temperature, dehydration and fusogenic lipids with an intrinsic conical molecular shape, are known to promote the fusion of unilamellar vesicles (Cullis et al., 1986).  The addition of calcium has been shown to induce fusion of vesicles composed of acidic phospholipids such as PS and PA in much the same way calcium induces hexagonal HII phases in acidic phospholipids (Hope et al., 1983; Papahadjopoulos et al., 1990; Tilcock and Cullis, 1981).  Also, membrane fusion can be promoted by the addition of cone-shaped lipids such as DOPE, which preferentially adopts the hexagonal HII phase (Ellens et al., 1989).  On the contrary, inverted cone shaped lipids inducing positive curvature such as lysoPC can inhibit fusion in model membrane systems and in biomembranes (Chernomordik, 1996; Chernomordik et al., 1995).   Fusion in biological membranes is modulated by fusion proteins.  Many fusion proteins enhance the negative curvature of membranes to promote membrane fusion.  Fusion peptides from influenza virus and Simian immunodeficiency virus (SIV) can lower the TH of lipids (Epand and Epand 1994; Epand et al 1994).  In addition, the fusion peptide from influenza can also induce the formation of a non-bilayer cubic phase in previously bilayer lipids (Colotto and Epand 1997).  Also, phospholipase C mediates cellular membrane fusion through the formation of diacylglycerol that may promote the formation of cubic phases in the membrane (Nieva et al 1995; Bananez et al 1997).  Some fusion peptides require the presence of cone-shaped lipids for function.  For example, the action of a fusion peptide derived from gp41 of HIV-1 requires PE (Martins et al 1993).   20   Figure 1.6  The "stalk" model of membrane fusion The fusion between two lipid membranes is hypothesized to begin with close contact between the two membrane driven by attractive forces from the hydrophobic interior of the membranes.  A hemifused stalk intermediate is formed as the outer monolayers of the two membranes became continuous.  The stalk expands into a transmembrane contact which eventually ruptures into a fusion pore.  21  1.3.3 Phase behaviour of mixtures of cationic and anionic lipid The use of cationic lipids for transfection of cultured cells was first described in 1987 by Felgner et al.  It was found that cationic lipids interact with anionic nucleic acid polymers such as plasmids to form a transfection competent “lipoplex” that is taken into the cell through endocytosis (Friend et al., 1996).  The role of cationic lipids in the delivery of nucleic acids will be discussed in more detail in Section 1.5.4.  The polymorphic phase behaviour of cationic lipids is of interest because cationic lipids have been shown to induce non-bilayer phases when mixed with anionic lipids, potentially leading to destabilization of the endosomal membrane and release of nucleic acid from the endosome (Hafez and Cullis, 2001).  It was determined that equimolar mixtures of cationic lipids and anionic lipids commonly found in cellular membranes readily adopt the hexagonal HII phase (Hafez et al., 2001).  This ability to form hexagonal HII structures can be rationalized using the molecular shape argument.  Electrostatic interactions between the headgroups of the cationic and anionic lipids effectively reduce their cross-sectional area.  This leads to a cone-shaped ion pair that readily adopts the hexagonal HII phase.  A related phenomenon was observed for cationic and anionic surfactants which adopt micellar conformation in isolation but assemble into bilayer when mixed together (Kaler et al., 1989).  Systems containing the cationic lipid N,N-dioleoyl-N,N-dimethylammonium chloride (DODAC) and the anionic lipid cholesteryl hemisuccinate (CHEMS) adopt the hexagonal HII phase when the mixture exhibits a net neutral surface charge (Hafez et al., 2000).  However, the system adopts a bilayer conformation when it exhibits either a net positive or negative charge by altering the pH or the cationic-to-anionic lipid ratio (Hafez et al., 2000).  Ionization of the lipid headgroups increases hydration and reduces the formation of ion pairs between lipid headgroups.  This consequently allows both lipids to adopt their native bilayer conformation.  In addition, 22  studies have also suggested that the association of cationic and anionic lipids facilitates the intracellular release of DNA as it displaces the DNA from the cationic lipoplex (Xu and Szoka, 1996; Zelphati and Szoka, 1996). It has been suggested that the formation of non-bilayer structures and the displacement of nucleic acid caused by the interaction between cationic lipids in a lipoplex and endosomal anionic lipids compose the main mechanism by which cationic lipid-based delivery systems release its nucleic acid cargo into the cytosol (Hafez et al., 2001; Li and Szoka, 2007). 1.4 Liposomes and Their Application in Drug Delivery 1.4.1 Early developments of liposomes The first description of liposomes  used negative staining electron microscopy to show that lecithin (egg yolk PC) assembles into concentric lamellae when hydrated (Bangham and Horne, 1964).  Later it was reported that these lipid vesicles can maintain a concentration gradient of ions, which can be disrupted by the addition of detergents (Bangham et al., 1965a, 1965b).  This demonstrated that these lipid vesicles constitute an enclosed bilayer system containing an aqueous interior and can serve as a basic model membrane for studying the physical properties of lipids in cell membranes (Papahadjopoulos and Watkins, 1967).  Studies on the properties of these model membranes showed that their physical properties such as osmotic swelling and ability to trap solutes including ions and glucose, were similar to those of cell membranes (Sessa and Weissmann, 1968).  These lipid vesicles were first termed "Bangasomes" , however the term "liposomes" was adopted subsequently (Sessa and Weissmann, 1968).  One of the first applications of liposomes as model cell membranes was the study on the effects of anesthetics.  Bangham and colleagues demonstrated that liposomes exposed to anesthetics were more permeable to ions (Johnson et al., 1973).  The discovery of liposomes demonstrated that the lipid 23  bilayer serves as a permeability barrier in cell membranes and proved to be a valuable model system for the study of cell membrane properties. The first liposomes prepared by Bangham were simple aqueous dispersion of lecithin and contained multiple concentric lipid bilayers (Bangham and Horne, 1964).  Many techniques have been developed subsequently to prepare liposomes with a single bilayer membrane and an aqueous interior for use as a model cell membrane system.  One of the earliest methods to produce unilamellar vesicles was through sonication of an aqueous dispersion of phospholipids (Huang, 1969; Johnson and Bangham, 1969; Johnson et al., 1971).  This method of preparation typically produces liposomes with a single bilayer membrane and a diameter of about 25 nm.  Liposomes can also be produced without sonication.  Lipids can be first dissolved in ethanol and then rapidly injected into a vigorously stirring aqueous solution (Batzri and Korn, 1973).  Particles are similar to those prepared by sonication, with a single bilayer membrane and a diameter of approximately 26 nm.  However, the resulting preparation typically contains about 7.5% of residual ethanol, which has to be removed by dialysis.  Both preparation methods produce very small particles with a large surface area-to-volume ratio that limits the encapsulation of various solutes.  In 1978, Szoka and Papahadjopoulous developed the reverse-phase evaporation method in order to combat this issue (Szoka Jr and Papahadjopoulos, 1978).  Mixture of lipids is first dissolved in organic solvent such as chloroform or ether.  Aqueous buffer is then added directly to this lipid/organic solvent mixture, forming an opaque suspension of lipids.  The organic solvent is removed using a rotary evaporator, resulting in an aqueous suspension of lipid vesicles.  This preparation produces vesicles larger than 100 nm in diameter and allows higher encapsulation efficiencies of various solutes such as sucrose, sodium and carboxyfluorescein. 24  The development of the extrusion technique significantly improved our ability to efficiently produce large unilamellar vesicles (LUV) and encapsulate solutes in these vesicles.  The original extrusion technique uses a relatively low pressure to push an aqueous suspension of multilamellar vesicles (MLV) sequentially through polycarbonate membranes with decreasing pore size (Olson et al., 1979).  This method was subsequently improved by extrusion of MLV under higher pressures using a purpose-built high pressure extruder, which eliminated the need for sequentially extrusions and resulted in a technique that rapidly (< 10 min) and reproducibly resulted in the formation of LUV (Hope et al., 1985).  With regard to mechanism, bilayer-forming lipids spontaneously assemble into MLV on hydration.  These MLV are micron sized particles with variable lamellarity.  During extrusion, the MLV are forced through the pores of the polycarbonate membrane leading to rupture and formation of unilamellar vesicles with sizes that reflect the pore size of the polycarbonate membrane of the lipid membrane (Hunter and Frisken, 1998; Mui et al., 2003).  The size and lamellarity of the vesicles can be improved by repeated extrusion of MLV through the polycarbonate membrane (Mayer et al., 1986b).  The size of the resulting particle is dependent on the pore size of the polycarbonate membrane and can range from 60 to 100 nm with a narrow size distribution (Mayer et al., 1986b).  The lamellarity of the liposomes can be determined using 31P NMR by comparing the phospholipid phosphorus signal intensity before and after the addition of a membrane impermeable line-broadening reagent to the external medium.  The phosphorus signal intensity will decrease by an amount proportional to the fraction of phospholipid exposed to the external medium.  Using this technique, it was shown that vesicles produced by rapid extrusion become unilamellar after approximately 5 passes through the polycarbonate membrane.  In addition, the resulting particle 25  can maintain an electrochemical gradient which has important ramifications in the development of encapsulation methods for various compounds in the liposome (Mayer et al., 1985, 1986a). 1.4.2 Liposomes as drug delivery vehicles The potential for using liposomes as drug carriers was recognized soon after their discovery in the early 1960s.  Early studies on liposome manufacturing techniques often focused on the encapsulation of various solutes (Sessa and Weissmann, 1968).  Gregoriadis was one of the first to recognize the drug delivery potential of liposomes and applied it for enzyme replacement therapy (Gregoriadis et al., 1971).  Dry lipid was hydrated with aqueous buffer containing the enzyme amyloglucosidase or albumin; however, the encapsulation efficiencies for both proteins were extremely poor, less than 10% of the original proteins were encapsulated.  The hydration of dry lipid film using aqueous buffers containing antibiotics such as penicillin and actinomycin also resulted in very poor encapsulation of the antibiotics (Gregoriadis, 1973).   The major breakthrough for improving drug encapsulation efficiencies in liposomes came with the development the transmembrane pH gradient loading technique to entrap weakly basic drugs.  It was previously demonstrated that weakly basic compounds such as catecholamines can be concentrated inside of liposomes exhibiting an acidic interior (Nichols and Deamer, 1976).  It was then shown (Mayer et al., 1986c) that drugs that are weak bases, such as doxorubicin, can also be accumulated into liposomes with an acidic interior. The encapsulation of drugs in response to a transmembrane pH gradient can be readily explained.  At neutral pH, weakly basic drugs are a mixture of the protonated, membrane-impermeable form DH+ and the deprotonated, membrane-permeable form D.  The deprotonated form of the drug D can readily diffuse across the membrane.  Once inside the acidic interior of the liposome, where the internal pH is lower than the pKa of the drug, the drug becomes protonated and trapped inside the liposome (Figure 26  1.7A).  At equilibrium the drug concentration gradient will mirror the proton gradient, thus a pH gradient of 3 units will result in a thousand fold higher concentration of drug inside the liposome as compared to the exterior medium. The loading of weakly basic drugs can also be accomplished using ammonium sulfate, which is an alternative method to generate a transmembrane pH gradient (Bolotin et al., 1994).  Ammonium sulfate is first encapsulated inside the liposome and the external buffer exchanged to establish an ammonium sulfate gradient.  An equilibrium of positively charged ammonium ion (NH4+) and neutral ammonia (NH3) exists inside the liposome.  Since the lipid membrane is highly permeable to the uncharged NH3, one proton remains behind as NH3 travels down its concentration gradient out of liposome, resulting in acidification of the liposome interior.  Weakly basic drugs can then be encapsulated using this pH gradient (Figure 1.7B).  Loading in response to pH gradients can result in interior concentrations of drug that are so high that they form nanocrystals inside the liposomes, leading to a characteristic "coffee bean" appearance in electron micrograph (Abraham et al., 2005).   The use of an ionophore offers a third method to generate the pH gradient needed for the encapsulation of weakly basic drugs (Fenske et al., 1998).  A transmembrane gradient of divalent cations such as Mn2+ or Mg2+ is first generated by preparing LUV in solutions of MnSO4 or MgSO4 followed by exchange of the external buffer with a sucrose solution.  After generating the divalent cation gradient, the weakly basic drug, the ionophore A 23187 and EDTA are added to the LUV.  The ionophore couples the transport of one divalent cation out of the LUV with the import of two protons, resulting in acidification of the interior of the vesicle (Fenske et al., 1998).  The exported divalent cations are chelated by EDTA while the drug enters the vesicles following the pH gradient generated (Figure 1.7C).  This method showed very high levels of 27  encapsulation (>80%) for both vincristine and ciprofloxacin with drug retention properties similar to other pH gradient loading methods (Fenske et al., 1998). The goal of using liposomes as drug delivery vehicles is to enhance the potency of drugs and reduce their side effects by increasing their delivery to the sites of disease while avoiding the healthy tissues.  Many studies have noted that the vasculature at sites of disease such as tumours, sites of infection and sites of inflammation is “leakier” than in healthy tissues (Skinner et al., 1990; Steinberg et al., 1990).  For example, the neovasculature formed in tumours lacks the smooth muscle wall of normal vasculature and is relatively permeable to particles as large as 200 nm diameter or more (Yuan et al., 1994, 1995).  As a result, long-circulating liposomes with a diameter of 100 nm preferentially escape in the region of tumours which can result in large increases in the amount of drug that is delivered to the tumour. This phenomenon is called the enhanced permeation and retention (EPR) effect (Maeda, 2001; Maeda et al., 2000).  It is important to note that the encapsulated drug is not available to target cells and must be released from the liposome to exert its effect.  For certain drugs the rate of drug release strongly influences the anticancer potency of the formulation. This is particularly true of cell cycle-specific drugs such as vincristine (Mayer et al., 1993; Webb et al., 1995). The rate of drug release can be controlled changing the lipid composition of the liposome (Charrois and Allen, 2004) or by varying the drug-to-lipid ratio (Johnston et al., 2006) to achieve the optimal therapeutic activity.    28   Figure 1.7  Encapsulation of small molecule drugs in response to transmembrane pH gradient  (A) In the standard pH gradient method, lipid films are first hydrated with buffer of low pH (typically citrate buffer, pH 4), followed by extrusion, generating unilamellar vesicles (LUV).  A pH gradient is generated by exchanging the external buffer with a buffer of higher pH (typically HEPES, pH 7.5).  Weakly basic drugs enter the vesicle in their neutral form but are protonated in the acidic interior of the vesicle and become trapped.  (B) In the ammonium sulfate method, a transmembrane ammonium sulfate gradient is first generated by hydrating and extruding the lipid in a buffer containing ammonium sulfate, followed by an exchange of the external buffer.  The interior of the vesicle is acidified as the neutral NH3 exits the vesicle, leaving behind a proton in the process.  This pH gradient can then be used for the encapsulation of weakly basic drugs.  (C)  In the ionophore method, a transmembrane gradient of Mg2+ or Mn2+ is first generated.  The ionophore A23187 couples the export of Mg2+ with the import of 2 protons, therefore acidifying the interior of the vesicle.  A chelator of Mg2+, such as EDTA, is usually required to remove the Mg2+ as it is transported out of the vesicle.  29  1.5 Delivery of Nucleic Acid-Based Therapeutics 1.5.1 Formulation of plasmid nanoparticles The use of cationic lipid to mediate cellular transfection of DNA was first describe by Felgner et al. in 1987 using the permanently positively charged cationic lipid, N-[1-(2,3-dioleyloxy) propyl]-N,N,N-trimethylammonium chloride (DOTMA).  The formulation used was a simple mixture of DOTMA and DOPE, forming a complex (lipoplex) with plasmid DNA.  This study was followed by the synthesis of a series of cationic lipids with varying in vitro transfection activities (Felgner et al., 1994).  Lipoplexes are often microns in diameter, are unstable and difficult to reproduce. They are often formulated with excess cationic charge to promote not only interaction with the plasmid DNA but also to facilitate association with the negatively charged cell surface to facilitate uptake by endocytosis (Stamatatos et al., 1988).  Incorporation of helper lipids such as DOPE is thought to facilitate the release of plasmid DNA into the cytoplasm due to its propensity to adopt the non-bilayer, membrane lytic hexagonal HII phase.  Despite intensive efforts, however, lipoplexes have not proven useful for in vivo applications. The chief reason for this is that they are highly toxic due to their positive charge and are removed quickly from the circulation by the fixed and free macrophages of the mononuclear phagocyte system due to their size and charge. As a result they do not distribute efficiently to target cells such as tumour cells.   These difficulties have led to attempts to encapsulate plasmid DNA within an enclosed bilayer membrane that exhibits low surface charge.  One such techniques is the detergent dialysis method (Wheeler et al., 1999).  In this method, plasmid DNA is mixed with the cationic lipid DODAC, DOPE and a PEG-lipid in the presence of the detergent octylglucoside.  Removal of the detergent by dialysis facilitates the formation of "stabilized plasmid-lipid particles" (SPLP) 30  which appear to be single lipid bilayer vesicles with plasmid DNA encapsulated inside (Tam et al., 2000).  This detergent dialysis method results in “stable plasmid-lipid vesicles” (SPLP) of 70 nm in diameter with an encapsulation efficiency of over 70% that exhibit the long circulation lifetimes required to access target tissues such as tumours.  However, this process is difficult reproduce and is not scalable.  An alternative method to generate SPLP is the spontaneous vesicle formation (SVF) by ethanol dilution method (Jeffs et al., 2005).  Plasmid DNA was prepared in an acidic buffer, while cationic lipid, DSPC, cholesterol and PEG-lipid are dissolved in ethanol.  The plasmid and lipid mixtures are then combined using a T-tube mixer (Figure 1.8).  Vesicles form spontaneously as the lipids precipitate from solution as the polarity of the medium is raised. Residual ethanol in the final mixture is then removed by dialysis.  This method results in particles of approximately 100 nm in diameter with multiple concentric bilayer membranes and an encapsulation efficiency of over 80% (Jeffs et al., 2005).  This method has also been used for the encapsulation of siRNA and the resulting particles are termed "stable nucleic acid lipid particle" (SNALP).  The encapsulation of siRNA will be discussed in more detail in Section 1.5.6.    31   Figure 1.8  Encapsulation of nucleic acids with spontaneous vesicle formation (SVF) by ethanol dilution This method is also commonly referred as the "T-tube" method.  It consists of the mixing of lipids dissolved in ethanol with nucleic acid in aqueous buffer using a T-tube connector.  The flow of the two streams of fluid is controlled with a peristaltic pump.  Vesicle are formed as the ethanol is diluted below the solubility limit of the lipid.  The ethanol content is further diluted upon exiting the T-tube connector to stabilized the resulting particles.  Unencapsulated nucleic acid can be removed by using an ion exchange column but is often unnecessary as the encapsulation efficiencies can exceed 80%.  Residual ethanol is removed by diafiltration against phosphate buffered saline (PBS).  This method was applied for both the encapsulation of plasmid DNA (stabilized plasmid-lipid particles, SPLP), and antisense oligonucleotides or siRNA (stable nucleic acid lipid particle, SNALP).  32  1.5.2 RNA interference In the late 1980s, plant biologists realized that the expression of antisense RNA in transgenic plants led to gene silencing (Ecker and Davis, 1986; Matzke and Matzke, 2004).  At around the same time, researchers working to improve the colour pigments of petunia flower realized that overexpression of the enzyme responsible for the synthesis of colour pigments, chalcone synthase, led to a decrease in colour pigments due to a decreased expression of the enzyme (Napoli et al., 1990).  Later it was discovered that this decrease in gene expression occurs at the post-transcriptional level through increased turnover of messenger RNA (mRNA) and the phenomenon was termed post-transcription gene silencing (PTGS) (Van Blokland et al., 1994).  It has been suggested that PTGS in plants originated as a defense mechanism to protect plants from viral infection (Ratcliff et al., 1997).  In 1998, Andrew Fire and Craig Mello published their seminal paper on RNA interference (RNAi) and clearly demonstrated that double stranded RNA mediates sequence specific gene silencing in Caenorhabditis elegan (Fire et al., 1998).  It was later demonstrated that similar pathways are found in most eukaryotes including mammals (Elbashir et al., 2001) and mice (McCaffrey et al., 2002).  These discoveries not only provided a powerful new tool to study genes with unknown function but also unleashed a whole new opportunity in the field of medicine as it is now possible, from a theoretical standpoint, to specifically silence any disease-causing gene such as oncogenes in cancer.  Andrew Fire and Craig Mello were later awarded the Nobel Prize in Physiology or Medicine in 2006 for "their discovery of RNA interference - gene silencing by double-stranded RNA". RNAi is an endogenous pathway in eukaryotic cells that uses RNA molecules to specifically catalyze degradation of complementary mRNA sequences, possibly as a mechanism to protect against pathogens such as viruses.  When long double-stranded RNA is introduced into cells, the 33  enzyme Dicer cleaves the RNA into shorter fragments of approximately 21-23 nucleotides long known as short-interfering RNA (siRNA; Bernstein et al., 2001).  Dicer is a large 200 kDa nuclease of the RNaseIII family and is specific for double-stranded RNA (Filipowicz et al., 2005).  The siRNA is then loaded into a protein complex called RNA-induced gene silencing complex (RISC) (Rand et al., 2004).  RISC is a complex of many different proteins of which Argonaute 2 (Ago2) is responsible for unwinding the siRNA and degradation of the sense strand (Matranga et al., 2005; Rand et al., 2005).  The single-stranded antisense strand allows the RISC complex to actively seek out mRNA complementary to the antisense strand of the siRNA (Ameres et al., 2007).  Once bound, Ago2 mediates the cleavage of the mRNA at a position corresponding to between nucleotide 10 and 11 from the 5' end of the antisense strand (Tomari and Zamore, 2005).  The RISC complex can then seek out and degrade additional mRNA, leading to remarkable silencing of the target gene (Hutvágner and Zamore, 2002). 1.5.3 siRNA as therapeutics It is theoretically possible to use siRNA to specifically block the synthesis of proteins responsible for a certain disease; however, the actual translation of siRNA into clinical use has been hampered by several major difficulties.  First, "naked" siRNA is unstable and is rapidly degraded by serum nucleases (Choung et al., 2006; Layzer et al., 2004).  Second, systemically administered siRNA is rapidly cleared from the circulation by the kidneys and fails to accumulate at the site of disease.  Finally, even if siRNA penetrates to target tissue, it cannot penetrate the cell membrane to enter the cytoplasm (de Fougerolles et al., 2007; Whitehead et al., 2009).  Delivery vehicles are therefore important to overcome these difficulties and realize the potential of siRNA as a therapeutic. 34  It is important to note that siRNA can be delivered locally to some tissues without the use of a sophisticated delivery system.  It has been shown that siRNA administered intranasally inhibits infection by respiratory syncytical virus (RSV) in the lungs (Bitko et al., 2005).  In addition, intratracheally administered siRNA against the SARS coronavirus can alleviate SARS-like symptoms in mice (Li et al., 2005).  The eye is another popular tissue for local administration of siRNA.  Intravitreal injection of siRNA against VEGF and VEGF receptor has been shown to decrease the extent of neovascularization in animal models of laser-induced retinal and choroidal neovascularization (CNV) (Shen et al., 2006; Tolentino et al., 2004).  Local administration of siRNA has also shown gene silencing activities in the brain.  Intraventricular injection of siRNA has been shown to mediate gene silencing of eGFP, dopamine transporter and serotonin transporter in the brains of mice (Thakker et al., 2004, 2005).  Despite the successes from the above examples, many tissues in the body are inaccessible by local administration and systemic delivery to these tissues requires the use of an appropriate delivery system. One of the simplest strategies is the direct conjugation of cholesterol onto siRNA.  Intravenous administration of cholesterol-conjugated siRNA against apolipoprotein B (ApoB) has led to the silencing of ApoB mRNA in the liver, decreased plasma level of ApoB protein and a reduction in serum cholesterol levels (Soutschek et al., 2004).  The use of 2'-O-methyl sugar modification and a partial phosphothioate backbone on the siRNA protect it from degradation by serum nucleases.  Cationic polymers such as polyethylenimine (PEI) are commonly used in combination with siRNA to generate transfection-competent polyplexes.  PEI is a synthetic polymer with a very high cationic charge density.  Polyplexes of PEI and siRNA against VEGF have shown intratumoural localization and silencing of VEGF after intravenous administration (Schiffelers et al., 2004).  Other synthetic polymers such as chitosan, cyclodextrin and 35  poly(lactic-co-glycolic acid) (PLGA) are also commonly used carriers for in vivo delivery of siRNA (Hu-Lieskovan et al., 2005; Patil et al., 2010; Pillé et al., 2006; Yuan et al., 2011).   Lipid nanoparticles (LNP) are currently the most advanced delivery systems for systemic delivery of siRNA with several formulations of siRNA-LNP already in clinical trials (Allen and Cullis, 2013).  It is currently the most potent system for gene silencing in hepatocytes with an effective dose of 0.005 mg siRNA/kg body weight in murine models and 0.03 mg siRNA/kg body weight in non-human primates following a single intravenous injection (Jayaraman et al., 2012).  The following sections discuss various formulation methods for siRNA-LNP systems in more detail. 1.5.4 Lipid components of siRNA-lipid nanoparticle (LNP) systems Cationic lipids are an essential component for the efficient encapsulation of siRNA, lipid formulations without cationic lipids result in poor encapsulation of nucleic acids (Maclachlan, 2007).  The positive charges of cationic lipids facilitate the encapsulation of the anionic siRNA via electrostatic interactions.  Cationic lipid can also provide a positive surface charge to the LNP, which promotes binding of the LNP with the negatively charged cell surface.  It has been shown that cell surface proteoglycan and sialic acids interact with cationic liposomes and facilitate cellular uptake of the liposomes via the endocytic pathway (Mislick and Baldeschwieler, 1996; Mounkes et al., 1998; Stamatatos et al., 1988; Wrobel and Collins, 1995).  As mentioned above, cationic lipids of the LNP are hypothesized to form membrane disruptive ion pairs with anionic lipids of the endosome thereby facilitating intracellular release of siRNA (Hafez et al., 2001).  However, positively charged liposomes are rapidly eliminated by the mononuclear phagocyte system (MPS; Litzinger, 1997) and most cationic lipids are highly toxic (Audouy et al., 2002; Liebert et al., 2000; Scheule et al., 1997; Zhang et al., 2005).   36  Ionizable cationic lipids with primary, secondary of tertiary amines in the headgroup and apparent pKa values less than 7 have been developed to deal with these problems. The original ionizable cationic lipid used was 1,2-dioleoyl-3-dimethylammonium propane (DODAP) (Maurer et al., 2001; Semple et al., 2001) which was chosen because it exhibits a pKa of 6.6 (Bailey and Cullis, 1994).  DODAP was used to encapsulate antisense oligonucleotides using a pre-formed vesicle (PFV) (see Section 1.5.5) approach, the resulting nanoparticles exhibited the longer circulation lifetimes following i.v. administration expected for a lipid particle with little surface charge (Semple et al., 2001).  It has been found that significant improvements in the potency of lipid nanoparticle (LNP) formulations of siRNA can be achieved by varying the chemical structure of the ionizable cationic lipid.  This is noted in Figure 1.9 where the progression from DODAP to DODMA to DLinKDMA to DLinKC2DMA to DLinMC3DMA is shown.  These lipids were identified using an in vivo screening model employing LNP siRNA systems containing siRNA to silence Factor VII (FVII), one of the proteins in the clotting cascade.  Factor VII is made in hepatocytes and secreted into the circulation, thus the potency of the siRNA-LNP systems can be monitored by measuring FVII levels in the circulation 24 h after i.v. administration of the LNP.  Remarkably, the potency of siRNA-LNP systems improved from ~10 mg siRNA/kg body weight to result in 50% gene silencing for LNP containing DODAP to 0.005 mg siRNA/kg body weight for the LNP containing DLinMC3DMA (Jayaraman et al., 2012; Semple et al., 2010).  The improvements in potency have been attributed to a need to use lipids with optimized pKa values and maximized bilayer destabilizing capabilities in the presence of anionic lipids.  The optimized pKa for DLinMC3 is 6.4.  In this thesis the primary lipid used is DLinKC2-DMA, which was the most potent lipid available at the time this thesis was initiated.  37  The type of PEG-lipid can also influence the potency of the LNP siRNA particle. The PEG moiety provides a hydrophilic steric barrier on the surface of the LNP.  PEG-lipids are incorporated into siRNA-LNP formulations to prevent aggregation of the particle during the formulation process (Maurer et al., 2001).  Without the PEG-lipid coating it is impossible to produce small LNP siRNA systems with diameters of 100 nm or less, large micron sized aggregates are formed.  However, the PEG coating on the LNP can also inhibit association with the plasma membrane of target cells, indicating a need to remove the coating before cellular uptake can proceed.  One potential solution is to use acid-labile PEG-lipid where the PEG moiety dissociates from the lipid anchor at low pH (Choi et al., 2003; Guo and Szoka, 2001; Shin et al., 2003).  However, the pH required is too low to exist in the extracellular medium (pH < 5).  Another approach is to use a PEG-lipid that dissociates from the LNP following i.v. administration.  The length of the lipid chain anchor on PEG-lipids influences its retention on liposomes (Parr et al., 1994), recently it has been shown that PEG-lipids with short (C14:0) acyl chains dissociate from LNP in vivo halftime of approximately 1 h (Mui et al., 2013), whereas PEG-lipids with longer (C18) acyl chains exhibit dissociation rates of days or longer. This offers the possibility of employing LNP stabilized by PEG-lipids with short acyl chains that rapidly dissociate following i.v. administration, allowing the siRNA-LNP systems to become potent transfection agents.  In this thesis we primarily employ PEG-lipids containing C14:0 acyl chains as these are the systems that are the most potent in vivo.  In addition to cationic lipids and PEG-lipids, LNP formulations of siRNA typically contain other structural lipids.  The most commonly used lipids are cholesterol and a saturated phosphatidylcholine such as distearoyl-PC (DSPC) or dipalmitoyl-PC (DPPC).  The reasons why these lipids are required is not clear.  Cholesterol is included because of historical data for 38  bilayer liposomal systems indicating that liposomes consisting of PC alone accumulate cholesterol from serum components (Rodrigueza et al., 1993) and thus using systems containing approximately equimolar cholesterol are in better equilibrium with their surroundings. The need for DSPC is counter-intuitive as the presence of DSPC would be expected to mitigate against fusion with the endosomal membrane following uptake into target cells due to its cylindrical shape.  However, as discussed in Chapter 4, elimination of DSPC compromised the gene silencing potency of siRNA-LNP systems.  A final point to note regarding the LNP systems invested in this thesis is that there is no targeting information on the LNP exterior that would be expected to facilitate uptake into target cells such as the hepatocytes in the liver.  There has been extensive work to incorporate targeting ligands with LNP, targeting moieties such as antibody fragments and small molecule ligands have been coupled onto the surface of LNP in hopes to increase cellular uptake of the LNP in target tissues by receptor-mediated endocytosis (Allen and Cullis, 2013; Sapra and Allen, 2003).  It has been reported that liposomes with monoclonal antibody fragment against HER2 showed better cellular uptake in HER2-overexpressing breast cancer xenografts as compared to non-targeted liposomes (Kirpotin et al., 2006).  Also, antibody-bearing liposomes containing antisense oligonucleotides against viral mRNA has showed better efficacy against viral infection as compared to the same liposome without targeting antibodies (Leonetti et al., 1990).   In addition to using antibodies as the targeting moiety, it has been demonstrated recently that the cellular uptake and in vitro gene-silencing activity of siRNA-LNP can be improved by conjugating a small molecule targeting ligand against the Na+/K+ ATPase onto the PEG moiety of PEG-lipid (Tam et al., 2013).  Typically the targeting moiety is conjugated to the distal end of the PEG moiety on PEG-lipids and the resulting targeting lipid can be added along with the rest 39  of the lipid components during the formulation process.  Alternatively, the targeting lipid is added to stable, pre-formed liposomes in a process called post-insertion (Iden and Allen, 2001).  The use of post-insertion technique prevents the targeting lipid from interfering with the formulation process and provides a relatively simple method for preparing targeted LNP formulations. In the case of the siRNA-LNP systems characterized here, the remarkable potency for gene silencing in hepatocytes following i.v. administration can be attributed to association with apolipoprotein E following dissociation of the PEG-lipid coating. The presence of ApoE on the LNP facilitates uptake into hepatocytes via the LDL and scavenging receptors (Akinc et al., 2010).   40   Figure 1.9  Evolution of cationic lipids Early cationic lipids, such as N-[1-(2,3-dioleyloxy) propyl]-N,N,N-trimethylammonium chloride (DOTMA) and N,N-dioleoyl-N,N-dimethylammonium chloride (DODAC), contain a permanently positive quaternary amine headgroup which causes the LNP to be rapidly eliminated by the MPS.  Ionizable cationic lipids with tertiary amines in the headgroup have an apparent pKa less than 7 which allowed encapsulation of nucleic acids to be performed at acidic pH (pH 4.0) while the resulting LNP will exhibit a net neutral charge at physiological pH.   41  1.5.5 Formulation of siRNA-lipid nanoparticles by extrusion Many of the techniques used for the encapsulation of siRNA were first developed for the encapsulation of plasmid DNA or antisense oligonucleotides.  In the "stabilized antisense lipid particle" (SALP) method, a solution of the lipids in ethanol is added to an aqueous solution of nucleic acid at pH 4.0 (Semple et al., 2001).  The relatively low pH of the buffer ensures that the ionizable cationic lipid, 1,2-dioleoyl-3-dimethylammonium propane (DODAP, pKa 6.6), is positively charged for interaction with the nucleic acid.  The vesicles are subsequently extruded through polycarbonate membranes to ensure uniform size distribution as mentioned in Section 1.5.1.  Dialysis in buffer at pH 7.0 removes residual ethanol in the mixture and deprotonates the ionizable cationic lipid, creating a relatively neutral system with prolonged circulation lifetimes (Figure 1.10).  Encapsulation efficiencies of over 70% and particle diameter of about 100 nm were achieved with this method (Semple et al., 2001).  Interestingly, the resulting particles are mainly unilamellar at a low nucleic acid-to-lipid ratio, but adopt a multilamellar configuration at higher nucleic acid-to-lipid ratios.  It was suggested that nucleic acid bridges between individual lipid vesicles, promoting the formation of multilamellar structures. A variation of the ethanol-drop method is the preformed vesicle (PFV) method (Maurer et al., 2001).  Preformed vesicles containing DODAP, cholesterol, DSPC and a PEG-lipid at low pH, typically pH 4, where the DODAP is protonated are prepared by extrusion in the presence of up to 40% ethanol (Figure 1.11).  Extrusion is usually very rapid in the presence of ethanol.  A solution of nucleic acid at pH 4.0 is then slowly added to the ethanol-destabilized liposomes to avoid particle aggregation and the mixture is then incubated for 1 hour to allow encapsulation of the nucleic acid.  Ethanol is removed by dialysis after the encapsulation step and the pH of the solution is adjusted to 7.0 to create a charge neutral system.  This method generated 42  multilamellar particles of approximately 100 nm in diameter with nucleic acid encapsulation efficiency of over 80% (Maurer et al., 2001).  It is suggested that the attachment of nucleic acid onto the surface of the liposome creates an adhesion point between liposomes, facilitating the formation of the multilamellar vesicles.  This also explains the high nucleic acid encapsulation efficiency achieved with this method as the nucleic acid appears to be entrapped between lamellae of the multilamellar particle.    43    Figure 1.10  Encapsulation of nucleic acid by the "stabilized anti-sense lipid particle" (SALP) method This method involves the dropwise addition of an ethanolic solution of lipids into an aqueous buffer containing nucleic acid.  The drop in ethanol concentration facilitates the formation of vesicle as the ethanol concentration is now below the solubility limit of the lipids.  The vesicles are subsequently extruded, followed by dialysis to remove residual ethanol and to raise the pH of the buffer.  Encapsulation efficiencies are typically in the range of 70%.  The resulting particles are approximately 80 to 140 nm in diameter and show a range of lamellarity, depending on the initial nucleic acid to lipid ratio.     44   Figure 1.11  Encapsulation of nucleic acid by the preformed vesicle (PFV) method Lipid vesicles are first produced by dropwise addition of an ethanolic solution lipids into aqueous buffer, followed by extrusion.  An aqueous solution of nucleic acid is added slowly to the resulting vesicle solution, which usually contains approximately 40% ethanol.  The mixture is incubated for 1 hour at 37oC to allow encapsulation to take place.  Residual ethanol is removed by dialysis and the pH of the solution is raised to 7.0.  This method usually results in multilamellar particle of approximately 100 nm in diameter with an encapsulation efficiency of approximately 80%.   45  1.5.6 Formulation of siRNA-lipid nanoparticles by T-tube in-line mixing The spontaneous vesicle formation (SVF) by ethanol dilution described previously for the encapsulation of plasmid DNA was later adopted for the encapsulation of siRNA and is commonly referred to as the "T-tube" method (Section 1.5.1).  As with the encapsulation of plasmid DNA, mixtures of lipids dissolved in ethanol are mixed with siRNA dissolved in an aqueous buffer using a T-tube mixer (Figure 1.8).  The flow of the two streams of fluid is controlled with a peristaltic pump.  The method uses low amounts of PEG-lipid, typically in the range of 1-5% (mol% lipid), and results in particles ranging from 70 to 80 nm in diameter (Judge et al., 2006; Zimmermann et al., 2006).  The encapsulation of siRNA is highly efficient with encapsulation efficiencies routinely exceeds 90%.  This method has become the scale-up method of choice for companies such as Alnylam Pharmaceuticals, Tekmira Pharmaceuticals and more recently Acuitas Therapeutics (formerly Alcana Technologies).  Several LNP formulations currently under clinical trial are produced with this method including ALN-TTR02 from Alnylam Pharmaceuticals and TKM-PLK1 from Tekmira Pharmaceuticals.  Limitations of the T-tube mixer formulation process are that it is difficult to apply to laboratory scale formulations due to the high flow rates required to achieve rapid mixing, and the mixing rate is limited which results in an inability to formulate LNP with lipid compositions that require very rapid mixing to achieve stable systems. 1.5.7 Formulation of siRNA-lipid nanoparticles by microfluidic mixing The Cullis laboratory has recently developed a version of the in line mixing method employing a microfluidic herring-bone micromixer (Belliveau et al., 2012).  The herring-bone micromixer has two inlets, one for an ethanolic mixture of lipids and the other for a buffered solution of siRNA at pH 4.0.  A dual syringe pump is used to drive the two streams of fluid into 46  the herring-bone micromixer (Figure 1.12).  As the two streams of fluid meet at the herring-bone micromixer, they fold and wrap around each other, exponentially decreasing the diffusion length between the two streams.  This allows for rapid mixing of the two streams of fluid on a millisecond timescale.  The rapid decrease in solvent polarity results in the precipitation of lipids in accordance to their solubility.  The results are consistent with the initial condensation of inverted micelles of cationic lipids around the siRNA which serve as nucleating structures for the rest of the lipids to assemble into a lipid nanoparticle (LNP).  The results are also consistent with the PEG-lipid, as the last component to be deposited on the nascent LNP, providing the outer shell of the stabilized particle.  For example, the size of the particle is freely adjustable between 20 to 100 nm in diameter simply by altering the PEG-lipid content of the formulation.  In comparison with the PFV technique, microfluidic mixing results in higher encapsulation efficiency (> 90%), smaller particles and permits small scale production with little loss due to dead volume.  In addition it allows for well-defined and reproducible mixing between the solutions of lipids and siRNA, resulting in lower batch-to-batch variation (Belliveau et al., 2012).  Both LNP produced by T-tube in-line mixing and LNP produced by microfluidic mixing exhibit an electron dense core as detected by cryo-transmission microscopy (cryo-TEM).  Such features are inconsistent with normal bilayer lipid structures and the structural features of LNP will be explored in more detail in Chapter 2. The following section will discuss other “solid core” systems that have been previously characterized.    47    Figure 1.12  Encapsulation of siRNA by microfluidic mixing This method employs a herring bone micromixer to facilitate the mixing lipids dissolved in ethanol and siRNA in aqueous buffer.  The flow of the two streams of fluid are controlled with a dual syringe pump.  The herring bone micromixer exponentially increases the surface area between the two streams of fluid, resulting in rapid mixing on a millisecond timescale.  Residual ethanol is removed by dialysis and the pH of the solution is raised to 7.0.  This method results in over 90% siRNA encapsulation efficiencies.  Particle size is adjustable from 20 to 50 nm by varying the PEG-lipid content from 1% to 5%.      48  1.5.8 Solid-core lipidic systems in nature The body utilizes natural solid core systems for the transport of lipids in the form of lipoproteins.  Lipoproteins are assemblies of different apolipoproteins and lipids that serve as transporters for phospholipids, triglycerides, cholesterol and cholesterol ester in circulation.  Cryo-transmission electron microscopy (cryo-TEM) of very-low density lipoprotein (VLDL), intermediate-density lipoprotein (IDL) and low-density lipoprotein (LDL) show that they exhibit electron-dense solid core structures (van Antwerpen et al., 1999; Spin and Atkinson, 1995). Three-dimensional reconstruction of LDL from cryo-TEM reveals a hydrophobic core of apolar lipids, such as cholesterol ester and triglyceride, surrounded by a layer of phospholipids and apolipoprotein B-100 (apoB-100) (Orlova et al., 1999).  This organization of apolar lipids with apolipoprotein and other polar lipids has prompted research in utilizing LDL and LDL-mimicking systems as potential delivery systems for the transport of lipophilic drugs.  One example is Lipofundin, which is a colloidal fat emulsion of egg lecithin and triglyceride used for parenteral nutrition.  In cryo-TEM, it appears as electron-dense solid core particles of variable sizes along with the presence of multilamellar vesicles, presumably a result of excess lecithin in the formulation (Kuntsche et al., 2011).  The solid-core structure of colloidal fat emulsions is very similar to that observed for lipoprotein particles (van Antwerpen et al., 1999; Kuntsche et al., 2011).  In addition, mimetic lipoproteins particles constructed from recombinant apolipoprotein A1 (Apo A1) or apolipoprotein E (Apo E) along with cholesterol-conjugated siRNA have shown hepatic siRNA delivery and gene silencing following systemic administration (Nakayama et al., 2012). It has recently been demonstrated that solid-core nanoemulsions of triglyceride and phospholipid can be constructed by microfluidic mixing (Zhigaltsev et al., 2012).  The formation 49  of such solid-core systems is hypothesized to be driven by the condensation of lipid components as the polarity of the solvent rapidly increases when the ethanol stream is mixed with the aqueous stream.  The core of the particle is comprised of the hydrophobic lipid, triolein, which is expected to be the first lipid to condense out of solution forming a nucleation centre for the more polar lipid, 1-palmitoyl,2-oleoyl phosphatidylcholine (POPC), to coat the surface of the particle as it reaches its solubility limit.  The size of the resulting particle represents the smallest stable structure that can be assembled from the lipid components and is a function of the ratio of the core hydrophobic lipid to the more polar coating lipid. 1.6  Thesis Objective  Lipid nanoparticles are currently the most advanced delivery systems for the systemic administration of siRNA.  Despite recent progress in formulation techniques of siRNA-LNP, the structure of these systems remain largely unknown.  The overall objective of this thesis is to determine the structure of siRNA-LNP using the microfluidic mixing technique and apply this structural knowledge to prepare systems with improved gene silencing efficacy.  The siRNA-LNP system described in this thesis is prepared with the cationic lipid DLinKC2-DMA, DSPC, cholesterol and the PEG-lipid, PEG-c-DMA (Figure 1.13) using the microfluidic mixing technology described above (Figure 1.12).  This main objective of the thesis is further divided into three chapters, each with a specific aim.  In Chapter 2, the structure of siRNA-LNP systems produced by microfluidic mixing is investigated using cryo-TEM, a variety of biophysical techniques and molecular dynamics simulations.  Cryo-TEM and siRNA encapsulation measurements of siRNA-LNP systems produced by microfluidics gave the first indications that these systems are not bilayer, vesicular systems.  Other biophysical experiments including, 31P NMR, fluorescent resonance energy 50  transfer (FRET) membrane fusion assay and density measurement all suggest siRNA-LNP systems produced by microfluidic mixing contain a hydrophobic core.  Lastly, molecular dynamics simulation provides a model of siRNA-LNP and indicates the core of the particle consist of inverted micelles of lipids, some of which contain siRNA.  The mechanism leading to the formation of such solid core system will also be discussed. Chapter 3 will explore the mechanism leading to the formation of the solid-core interior of siRNA-LNP by adjusting various components of particle, including cationic lipid and siRNA content.  The solid-core model of siRNA-LNP generated by molecular dynamics simulation in Chapter 2 is first validated with the use of gold nanoparticle-conjugated siRNA.  The solid-core siRNA-LNP system is shown to be part of a continuum of structures that can be either multilamellar or inverted micellar depending on the cationic lipid and siRNA content.  The role of cationic and neutral, non-bilayer lipids in the encapsulation of siRNA is also explored.  Also, this Chapter extends the microfluidic mixing technique to the encapsulation of longer nucleic acids, such as plasmid DNA and messenger RNA (mRNA), demonstrating the versatility of the microfluidic mixing technique and the fact that solid core structures are also observed for these larger payloads. Chapter 4 focuses on improving the gene silencing efficacy of siRNA-LNP by attempting to enhance the endosomolytic properties of the LNP.  In the first attempt, the cationic lipid content of the LNP was increased to 80 mol% as it is thought that cationic lipids facilitate endosomal escape by forming membrane-disrupting ion pairs with anionic lipids of the endosome.  However increasing cationic lipid content led to decreased in vitro gene silencing activity.  The next approach incorporated bilayer-destabilizing "helper" lipids in the formulation in an attempt to enhance the endosomolytic properties of the LNP.  A novel "helper" lipid dioleoyl-four amino 51  butyric acid (DOFAB) was synthesized for this purpose.  Biophysical characterization of DOFAB suggests this novel lipid possesses bilayer-destabilizing properties and LNP containing DOFAB displays enhanced fusion capability as determined by the lipid mixing assay.  However, in vitro gene silencing results were disappointing as LNP containing DOFAB performed more poorly than particles comprised with the bilayer-stabilizing DSPC.  The role of cationic lipids and "helper" lipids in the intracellular release of siRNA is a point of discussion in this Chapter.  52   Figure 1.13  Lipid components of siRNA-LNP system formulated by microfluidic mixing Formulations of siRNA-LNP described in this thesis consists of the ionizable cationic lipid 2,2-dilinoleyl-4-(2-dimethylaminoethyl)-[1,3]-dioxolane (DLin-KC2-DMA), the phospholipid 1,2-distearyl-phosphatidylcholine (DSPC), cholesterol, and the PEG-lipid N-[(methoxy polyethylene glycol 2000 carbamyl]-1,2-dimyristyloxlpropyl-3-amine (PEG-c-DMA).       53  Chapter 2: Lipid Nanoparticles Containing siRNA Synthesized by Microfluidic Mixing Exhibit an Electron-Dense Nanostructured Core  2.1 Introduction Lipid nanoparticles (LNP) are the leading delivery systems for enabling the therapeutic potential of siRNA for systemic applications (Semple et al., 2010; Zimmermann et al., 2006).  LNP siRNA systems containing optimized ionizable cationic lipids can exhibit remarkable in vivo potencies at doses as low as 0.02 mg siRNA/kg body weight for silencing liver (hepatocyte) target genes in rodents following intravenous (i.v.) injection (Semple et al., 2010).  These systems are relatively non-toxic, leading to therapeutic indices in mice approaching 1000, indicating potential clinical utility.  Despite this progress, the structure of these siRNA-LNP systems is unclear.  Some models of siRNA-LNP suggest a bilayer vesicle structure of the LNP with siRNA on the inside in an aqueous interior (Jeffs et al., 2005), however other observations suggest that such models may be incorrect.  For example, recent cryo-transmission electron microscopy (cryo-TEM) studies of siRNA-LNP systems formed by mixing an ethanol stream containing lipid with an aqueous stream containing siRNA using a T-tube mixing system results in LNP that have electron-dense cores (Crawford et al., 2011) rather than the less dense aqueous cores observed for vesicular systems (Zhigaltsev et al., 2006).  In addition, formulation of siRNA-LNP systems using the T-tube mixer (Abrams et al., 2010) can result in siRNA encapsulation efficiencies above 70%, an observation that is difficult to reconcile with bilayer vesicular structure.  This is because encapsulation depends on the presence of cationic lipid and it would therefore be expected that the maximum encapsulation efficiency should be approximately 50% for bilayer systems, assuming that the cationic lipid is equally distributed on both sides of the bilayer. 54  In this work we characterize the structure of siRNA-LNP systems produced using a rapid microfluidic mixing technology (Zhigaltsev et al., 2012) by employing a variety of biophysical assays as well as in-silico simulations.  Cryo-TEM studies show that these siRNA-LNP systems exhibit an electron-dense core (in contrast to bilayer vesicle systems), and that “limit size” systems can be generated at high PEG-lipid contents that are consistent with the ability of siRNA to stimulate formation of inverted micelles in association with cationic lipid.  Fluorescent energy resonance transfer (FRET), 31P NMR and RNase digestion studies show that encapsulated siRNA is associated with internalized cationic lipid, is effectively immobilized on the NMR timescale and is fully protected from external RNase.  Density gradient studies show that the density of siRNA-LNP systems can vary from being significantly less dense than bilayer vesicle systems to exhibiting increased densities at higher siRNA contents.  Taken together, these experimental results suggest that siRNA resides in inverted micelles within the overall siRNA-LNP structure.  These results are consistent with molecular modeling studies that indicate these siRNA-LNP systems have a nanostructured core consisting of a periodic arrangement of aqueous compartments, some of which contain siRNA duplexes. 2.2 Materials and Methods 2.2.1 Materials   The lipids 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC), 1,2-distearoyl-sn-glycero-3-phosphocholine (DSPC), 1, 2 dioleoyl-sn-glycero-3-phosphoserine (DOPS), 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine-N-(7-nitro-2-1,3-benzoxadiazol-4-yl) (NBD-PE), and 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine-N-(lissamine rhodamine B sulfonyl) (Rh-PE)  were obtained from Avanti Polar Lipids (Alabaster, AL).  4-(2-Hydroxyethyl) piperazine-1-ethanesulfonic acid (HEPES), cholesterol, Sephadex G-50, tetrasodium EDTA, Trizma Base and 55  xylene cyanol were obtained from Sigma-Aldrich (St. Louis, MO).  Triton X-100 and 2-(N-Morpholino) ethanesulfonic acid (MES) were obtained from BDH (Westchester, PA).  Ammonium acetate, boric acid, sodium acetate and sodium chloride were obtained from Fisher Scientific (Fair Lawn, NJ). Bovine pancreatic RNase A was purchased from Applied Biosystems/Ambion (Austin, TX). De-ionized formamide was obtained from GIBCO BRL (Grand Island, NY).  Bromophenol Blue was obtained from BioRad (Hercules, CA).  The phosphodiester siRNA used in this study was obtained from Invitrogen (Carlsbad, CA) and phosphorothioate siRNA were 25-mer blunt-end duplexes (sense - 5’ GCCUUAACUUUGGUGAUCAAGGAUA 3’) and were obtained from Dharmacon (Lafayette, CO).  The ionizable cationic lipid 2,2-dilinoleyl-4-(2-dimethylaminoethyl)-[1,3]-dioxolane (DLin-KC2-DMA) and N-[(methoxy polyethylene glycol 2000 carbamyl]-1,2-dimyristyloxlpropyl-3-amine (PEG-c-DMA) were obtained from AlCana Technologies Inc., Vancouver, BC. Cholesterol E total cholesterol assay kit was obtained from Wako Diagnostics (Richmond, VA).  Quant-it RiboGreen RNA assay kit was obtained from Molecular Probes (Eugene, OR). 2.2.2 Preparation of POPC/cholesterol bilayer vesicles. Bilayer POPC/cholesterol (1:1; mol/mol) vesicles were prepared by hydration of a dried lipid film with PBS and the dispersion was then freeze-thawed five times using liquid nitrogen.  Unilamellar vesicles were then prepared by extruding the frozen and thawed lipid suspension 10 times through two stacked 80 nm pore size polycarbonate filters. 2.2.3 Preparation of siRNA-LNP systems. LNP were prepared by mixing appropriate volumes of lipid stock solutions in ethanol buffer with an aqueous phase containing siRNA duplexes employing a microfluidic micro-mixer as 56  described elsewhere (Zhigaltsev et al., 2012).  For the encapsulation of siRNA, the desired amount of siRNA was dissolved in 25 mM sodium acetate, pH 4.0.  Equal volumes of the lipid in ethanol and the siRNA in water were combined in the microfluidic micro-mixer using a dual-syringe pump (Model S200, KD Scientific, Holliston, MA) to drive the solutions through the micro-mixer at a combined flow rate of 2 ml/min (1 ml/minute for each syringe). A herringbone micromixer (Stroock et al., 2002) was employed.  The mixed material was diluted into an equal volume of 25 mM sodium acetate buffer, pH 4.0, upon leaving the micro-mixer outlet thus reducing the ethanol content to 25%.  The lipid mixture was then dialyzed for 4 hours against 1000 volumes of 50 mM MES/50 mM sodium citrate buffer (pH 6.7) followed by an overnight dialysis against 1000 volumes of 1X phosphate buffered saline, pH 7.4 (GIBCO, Carlsbad, CA) using Spectro/Por dialysis membranes (molecular weight cutoff 12000 – 14000 Da, Spectrum Laboratories, Rancho Dominguez, CA).  The mean diameter of the LNP after dialysis was 41.3 ± 14.9 nm as determined by dynamic light scattering (number mode; NICOMP 370 Submicron Particle Sizer, Santa Barbara, CA). Lipid concentrations were determined by measuring total cholesterol using the Cholesterol E enzymatic assay from Wako Chemicals USA (Richmond, VA). Encapsulation efficiency was calculated by determining unencapsulated siRNA content by measuring the fluorescence upon the addition of RiboGreen (Molecular Probes, Eugene, OR) to the siRNA-LNP (Fi) and comparing this value to the total siRNA content that is obtained upon lysis of the LNP by 1% Triton X-100 (Ft): % encapsulation = (Ft – Fi)/Ft x 100. Limit size particles were prepared by mixing 20 mM DLin-KC2-DMA/PEG-c-DMA (90/10, mol%) dissolved in ethanol with five volumes of 25 mM sodium acetate, pH 4.0 containing siRNA.  Mixing of the two streams are accomplished using the above mentioned herring micro-mixer device with a flow-rate of 0.5 mL/min for the lipid/ethanol stream and 2.5 mL/min for the 57  aqueous stream.  Ethanol was removed by first diluting the lipid mixture with sodium acetate buffer, pH 4.0 and then removing the ethanol-containing buffer with an Amicon Ultra, 10,000 MWCO, regenerated cellulose concentrator (Millipore, Billerica, MA).  The process was repeated five times to ensure all residual ethanol was removed. 2.2.4 Cryo-TEM. Cryo-TEM samples were prepared by applying 3 µL siRNA-LNP at 10-20 mg/ml total lipid to a standard electron microscopy grid with a perforated carbon film.  Excess liquid was removed from the grid by blotting and then the grid was plunge-frozen in liquid ethane to rapidly freeze the sample using a Vitrobot system (FEI, Hillsboro, Oregon).  Images were taken under cryogenic conditions (~88K) at a magnification of 50000x with an AMT HR CCD side mount camera. Samples were loaded with a Gatan 70 degree cryo-transfer holder in an FEI G20 Lab6 200kV TEM (FEI, Hillsboro, OR) under low dose conditions with an under-focus of 4-6 µm to enhance image contrast. Experiments were performed at the University of British Columbia Bio-imaging Centre (Vancouver, BC).  Particle diameters were measured from the micrographs with the aid of ImageJ (National Institute of Health, Bethesda, MD). Average diameters and standard deviations were calculated from more than 150 particles. 2.2.5 RNase protection assay. Factor VII siRNA was encapsulated with LNP formulations containing 40% DLin-KC2-DMA, 11.5% DSPC, 47.5% cholesterol and 1% PEG-c-DMA (mol %).  1.0 µg of siRNA (entrapped in LNP) was incubated with 0.05 µg bovine pancreatic RNase A (Ambion, Austin, TX) in 50 µL of 20 mM HEPES (pH 7.0) at 37oC for 1 hour.  At the end of the incubation, a 10 µL aliquot of the reaction mix was added to 30 µL FA dye (de-ionized formamide, TBE, PBS, xylene cyanol, bromophenol blue, triton X-100) to halt the RNase reaction.  Gel electrophoresis 58  was performed using 20% native polyacrylamide gel and nucleic acids were visualized by staining with SYBR-Safe (Invitrogen, Carlsbad, CA). 2.2.6 31P-NMR studies. Proton-decoupled 31P NMR spectra were obtained using a Bruker AVII 400 spectrometer operating at 162 MHz. Free induction decays (FID) corresponding to ~104 scans were obtained with a 15 μs, 55-degree pulse with a 1 s interpulse delay and a spectral width of 64 kHz. An exponential multiplication corresponding to 50 Hz line broadening was applied to the FID prior to Fourier transformation. The sample temperature was regulated using a Bruker BVT 3200 temperature unit. Measurements were performed at 25oC. Experiments were performed at the Centre for the Drug Research and Development, Vancouver, BC.  2.2.7 FRET membrane fusion studies. Fusion between siRNA-LNP and anionic DOPS bilayer vesicles was assayed by employing a fluorescence resonance energy transfer lipid mixing assay (Hafez et al., 2000; Struck et al., 1981).  Labeled DOPS vesicles containing NBD-PE and Rh-PE (1 mol % each) were prepared by hydration of the lipid in a thin film with 20mM HEPES buffer at pH 7.0 followed by 10 extrusions through two stacked 100 nm pore size polycarbonate filters (Nuclepore) using the Extruder (Northern Lipids, Vancouver, BC).  LNP composed of 40% DLinKC2-DMA, 11.5% DSPC, 47.5% cholesterol, 1% PEG-c-DMA were prepared with an siRNA-to-total lipid ratio (D/L ratio, wt/wt) of 0, 0.06 and 0.24. A D/L ratio of 0.24 represents a charge ratio of negative (siRNA - phosphate) to positive (cationic lipid - amine) of one. Lipid mixing experiments were conducted as previously described (Hafez et al., 2000).  Unlabeled LNP were added to a stirred cuvette containing NBD-PE/Rh-PE labeled DOPS vesicles at a 2:1 lipid molar ratio (200 μM LNP:100 μM DOPS vesicles) in 2 mL of 10 mM ammonium acetate, 10 mM MES, 10 mM 59  HEPES and 130 mM NaCl equilibrated to pH 5.5.  Fluorescence of NBD-PE was monitored using 465 nm excitation, and 535 nm emission using a Perkin-Elmer LS-55 fluorimeter with a 1 x 1 cm cuvette under continuous low speed stirring.  Lipid mixing was monitored for approximately 10 min, after which 20 μL of 10% Triton X-100 was added to disrupt all lipid vesicles, representing infinite probe dilution (0.1% Triton X-100 vol/vol final).  Lipid mixing was expressed as a percentage of infinite probe dilution determined using the equation: % lipid mixing = (F – Fo) / (Fmax – Fo) x 100, where F is the fluorescence intensity measured by the assay, Fo is the initial fluorescence intensity of NBD-PE/Rh-PE/DOPS vesicles and Fmax is the maximum fluorescence intensity at infinite probe dilution after the addition of Triton X-100 (0.1% v/v final). 2.2.8. Sucrose density gradient centrifugation. Solutions of 1%, 2.5%, 5%, 10% and 15% sucrose (wt/vol) were prepared in distilled water and used to make 10 mL step gradient.  Gradients were prepared successively overlaying 2 mL of less concentrated sucrose on top of the more concentrated ones.  SiRNA-LNP and POPC/cholesterol vesicles were prepared as described above.  Five hundred μL of sample was applied to the gradient and was centrifuged at 39000 rpm in a Beckman SW41 swing bucket rotor using a Beckman Coulter Optima LE-80K Ultracentrifuge (Brea, CA) for 18 hours.  These conditions results in an average centrifugal force of approximately 190000 x g in the middle of the tube.  After centrifugation, 500 uL fractions were successively removed from the top.  Lipid content from each fraction was determined using the Cholesterol E enzymatic assay from Wako Chemicals USA (Richmond, VA) described above.     60  2.2.9. Computer simulation of siRNA-LNP systems. Detailed description of the methods used for the computer simulation of siRNA-LNP system has been published elsewhere (Leung et al., 2012).  Briefly, the LNP containing nucleic acids was formed by first simulating the self-assembly of a smaller building block using models of lipids and nucleic acids, then creating a large particle by spatial translations of the building block, coating the large particle with a polymer-grafted lipid. The building block was self-assembled starting from random configurations in a small system (Figure 2.8A). Self-assembly resulted in a periodic nanostructure with enclosed water compartments (Figure 2.8B). The water molecules were removed, and the building block was multiplied 3x3x3 times to obtain a large nanoparticle, retaining a small spacing (<1 nm) between the translated unit cells. This nanoparticle was coated with a polymer-grafted lipid. The coating layer contained the polymer lipid in random conformations distributed within a thin slab (~ 3 nm), and was placed in close proximity (~ 1 nm) at all facets of the nanoparticle. This system was then re-solvated (water placed inside and outside of LNP) in a large simulation box.  2.3 Results 2.3.1 Microfluidic mixing allows highly efficient encapsulation of siRNA in siRNA-LNP systems over a wide range of siRNA/lipid charge ratios   As noted above, high siRNA encapsulation levels for siRNA in LNP systems are inconsistent with bilayer structure, where a maximum encapsulation efficiency of 50% would be expected. Here we characterized the encapsulation efficiencies of siRNA-LNP systems over a range of siRNA/cationic lipid charge ratios using the microfluidic mixing process as described in Methods and employing the lipid mixture DLinKC2-DMA/DSPC/Chol/PEG-lipid (40/11.5/47.5/1; mol/mol). The resulting siRNA-LNP systems exhibited diameters of 61  approximately 50 nm (number mode) as measured by dynamic light scattering. Essentially complete (~ 95% as indicated using the Ribo-Green assay) encapsulation was achieved over a wide range of siRNA/cationic charge ratios, including charge ratios as high as 1.25 (Figure 2.1). These observations suggest that the microfluidic mixing technique for formulating siRNA allows somewhat more efficient encapsulation than the T-tube mixing technique where trapping efficiencies of ~70% have been noted (Abrams et al., 2010).   Figure 2.1  siRNA encapsulation efficiency of siRNA-LNP systems prepared by microfluidic mixing at various siRNA/lipid ratios SiRNA-LNP with the lipid composition DLin-KC2-DMA/DSPC/Chol/PEG-lipid (40/11.5/47.5/1; mol/mol) were prepared using microfluidic mixing as described in Methods.  Measurements of trapping efficiency were made as indicated in Methods.  Data points and error bars represent the average encapsulation efficiencies and standard deviations, respectively, measured from three independently prepared LNP formulations.  Inset is a schematic representation of a bilayer vesicle with an equal distribution of cationic lipids between the two leaflets of membrane.  Such vesicle is expected to have a maximum siRNA encapsulation efficiency of 50%.  62  2.3.2 LNP systems exhibit an electron dense core structure as indicated by cryo-TEM  In the next set of experiments the cryo-TEM characteristics of the siRNA-LNP systems produced by microfluidic mixing were investigated for the DLinKC2-DMA/DSPC/Chol/PEG-lipid (40/11.5/47.5/1; mol/mol) lipid composition.  As shown in Figure 2.2 A and B, siRNA-LNP systems prepared at siRNA/total lipid ratios of 0.06 and 0.24, corresponding to siRNA-to-cationic charge ratios of 0.25 and 1, exhibited an electron dense core similar to that observed for siRNA-LNP systems formulated using the T-tube apparatus (Crawford et al., 2011).  The electron dense siRNA-LNP structure contrasts with the less dense interior of a vesicle system with an aqueous interior generated from POPC/cholesterol (1:1; mol/mol) (Figure 2.2C) and is visually similar to the electron dense interior exhibited by cryo-TEM of a colloidal fat emulsion as reported by Kuntsche et al.  The siRNA-LNP formulation employed in Figure 2.2A contains siRNA at a 0.06 siRNA/lipid (wt/wt) ratio which corresponds to an siRNA-to-cationic lipid charge ratio of 0.25.  As a result, when the siRNA-LNP is formulated at pH 4.0, approximately 75% of the cationic lipid is not complexed to siRNA in the LNP.  It is therefore of interest that the siRNA-LNP observed in Figure 2.2A exhibit an electron dense interior with no evidence of an internal aqueous core.  This suggests that the cationic lipid may contribute to the electron dense interior even when not complexed to siRNA.  In order to determine whether this is the case, LNP systems with the same lipid composition but no siRNA were formulated employing the microfluidic process and characterized by cryo-TEM. As shown in Figure 2.2D, an electron dense core was observed in the absence of siRNA, indicating that cationic lipids such as DLinKC2-DMA, possibly in combination with DSPC and cholesterol, contribute to electron dense structures in the LNP interior. 63   Figure 2.2  LNP containing DLin-KC2-DMA exhibit electron dense cores both in the presence and absence of encapsulated siRNA as indicated by cryo-TEM LNP were prepared by microfluidic mixing employing a herringbone mixer as indicated in Methods. POPC/cholesterol (50/50; mol/mol) bilayer vesicles were prepared by extrusion through polycarbonate filters with 80 nm pore size. (A): Cryo-TEM micrograph obtained from siRNA-LNP with lipid composition DLinKC2-DMA/DSPC/Chol/PEG-lipid (40/11.5/47.5/1; mol/mol) and siRNA at a siRNA/lipid ratio of 0.06, wt/wt., corresponding to an siRNA/cationic lipid charge ratio of 0.25. (B) LNP with the same lipid composition as for (A) but prepared at an siRNA/lipid ratio of 0.24 wt/wt, corresponding to an siRNA/cationic lipid charge ratio of 1. (C): Cryo-TEM micrograph of POPC/cholesterol (1:1) vesicles. (D): LNP with the same lipid composition as for (A) and (B) but prepared in the absence of siRNA.  64  2.3.3 LNP containing cationic lipid exhibit limit sizes consistent with the formation of inverted micellar structures in the LNP interior both in the presence and absence of siRNA The structures that could give rise to the electron dense cores detected by cryo-TEM are of interest.  In the absence of siRNA it may be proposed that the cationic lipid, in association with a counter-ion, adopts an inverted structures such as inverted micelles, consistent with the propensity of these lipids for highly curved “inverted” structures such as the hexagonal HII phase in mixtures with anionic lipids (Hafez and Cullis, 2001; Hafez et al., 2001).  These structures are inverted in the sense that the polar headgroups are oriented toward interior aqueous cores with diameters as small as 3 nm (Tinker and Pinteric, 1971).  The actual equilibrium radius of an inverted micelle could be larger as dictated by the intrinsic or spontaneous radius of curvature of the constituent lipids (Lafleur et al., 1996).  In turn, assuming a bilayer thickness of 4 nm this would suggest that LNP systems composed of pure cationic lipid should exhibit limit sizes with diameters in the range of 11 nm or larger, which is essentially a bilayer surrounding an aqueous interior with diameter as small as 3 nm.  Alternatively, in the presence of siRNA, it is logical to suppose that the limit size particle consists of a distorted inverted micelle of cationic lipid surrounding the siRNA oligonucleotide.  In turn, this would suggest a limit size system as small as 14  nm diameter, assuming that the siRNA contained in this inverted micelle is surrounded by an inner monolayer of cationic lipid within an outer monolayer of surface lipid and that the dimensions of the siRNA are 2.6 nm in diameter and 5.8 nm in length (Rosenberg et al., 1976). Here we explored whether limit size particles compatible with such structures could be generated using PEG-lipid as the surface lipid. In this regard a vesicle containing an internal aqueous core of 3 nm diameter has an outside-to-inside surface area ratio of 6.8 (assuming a bilayer thickness of 4 nm), indicating that the outer monolayer requires the presence of lipids that 65  provide an interfacial area approximately 7 times larger than the inner monolayer area. Assuming that the interfacial area for a lipid such as DLin-KC2-DMA is similar to 1, 2 dioleoyl-sn-glycero-3-phosphocholine (DOPC; 0.7 nm2; Alwarawrah et al., 2010), it is straightforward to show that approximately 10 mol% PEG2000-lipid (surface area per molecule 36 nm2; Soong and Macdonald, 2007) would required to coat inverted micelles composed of DLin-KC2DMA with an aqueous core 3 nm in diameter. We therefore examined the limit size LNP that could be achieved for a DLin-KC2-DMA/PEG-c-DMA system (90:10, mol:mol) produced by the microfluidic mixing process.  The size determined from cryo-TEM micrographs (Figure 2.3A) was 14.7 ± 6.9 nm, consistent with an ability of the cationic lipid to form inverted micellar structures with interior aqueous diameters in the range of 8 nm.  On the other hand, the limit size of LNP siRNA systems formulated at an siRNA-to-cationic lipid charge ratio of one resulted in limit size systems of 22.7 ± 6.1 nm in diameter (Figure 2.3B), consistent with the presence of inverted micelles with interior diameters of approximately 15 nm consisting of cationic lipid complexed to internalized siRNA.     66   Figure 2.3  LNP exhibit limit sizes consistent with inverted micelle structure in presence and absence of siRNA Limit size LNP were prepared by microfluidic mixing as indicated in Methods. (A): Cryo-TEM micrograph obtained from LNP with lipid composition DLin-KC2DMA/PEG-lipid (90/10; mol/mol) in the absence of siRNA.  (B): LNP with the same composition as for (A) but prepared with siRNA at an siRNA-to-cationic lipid charge ratio of 1.   (C): Size distribution of LNPs in Figure 2A and (D): Size distribution of LNPs in Figure 2B. Particle diameters are determined with the aid of Image J (NIH, Bethesda, MD) and average diameters are calculated from over 150 particles.  67  2.3.4 Encapsulated siRNA is immobilized in the siRNA-LNP system If the siRNA is complexed to cationic lipid and localized in an inverted micelle inside the LNP, it would be expected to be less mobile than if freely tumbling in the aqueous interior of a bilayer vesicle system.  The mobility of the siRNA can be probed using 31P NMR techniques.  In particular, it would be expected that limited motional averaging would be possible for complexed siRNA, leading to very broad “solid state” 31P NMR resonances due to the large chemical shift anisotropy of the phosphate phosphorus (DiVerdi et al., 1981).  If, on the other hand, the siRNA is able to freely tumble in an aqueous environment, rapid motional averaging would be expected to lead to narrow, readily detectable, 31P NMR spectra.  In order to avoid conflicting 31P NMR signals arising from phosphorus in phospholipids and siRNA, phosphorothioate siRNA, which gives a 31P NMR signal that is shifted downfield from the normal phosphate resonance, was used to ascertain the motional environment of the siRNA.   As shown in Figure 2.4A, the 31P NMR signal from free phosphorothioate siRNA is a doublet peak, shifted approximately 56 ppm downfield from the phosphate resonance (Eckstein and Jovin, 1983; Maurer et al., 2001).  The doublet structure may be attributed to the presence of Rp and Sp isomers of the siRNA strands (Hopkins et al., 1989).  For siRNA-LNP systems with the lipid composition DLin-KC2-DMA/DSPC/Chol/PEG-c-DMA (40/11.5/47.5/1 mol %) and containing siRNA (0.06 siRNA/lipid; wt/wt) no 31P NMR signal was observable for the encapsulated siRNA (Figure 2.4B), consistent with immobilization within the LNP core.  Detergent solubilization of the LNP particle using sodium dodecyl sulphate (SDS; 1% vol/vol) resulted in recovery of the siRNA signal (Figure 2.4C).   68   Figure 2.4  Encapsulated siRNA is immobilized on the NMR timescale  (A): 31P signal from free (phosphothioate) siRNA.  Note that phosphothioate siRNA, which gives rise to a 31P NMR resonance ~56 ppm downfield of the phosphodiester peak, was used to avoid overlap with the 31P NMR signal arising from the DSPC phosphorus. (B): 31P NMR spectrum of phosphothioate siRNA encapsulated at a siRNA/lipid ratio of 0.06 (wt/wt) in LNP containing DLinKC2-DMA /DSPC/Chol/PEG-lipid (40/11.5/47.5/1; mol/mol). (C): 31P NMR signal arising from the same sample as (B) after the addition of 1% SDS to solubilize the particle. The spectra depicted were obtained from 15000 transients as described under Methods.  69  2.3.5 Encapsulated siRNA is fully protected from degradation by external RNase A SiRNA sequestered in the LNP core should be fully protected from degradation by externally added RNase.  LNP siRNA systems with the lipid composition DLin-KC2-DMA/DSPC/Chol/PEG-c-DMA (40/11.5/47.5/1, mol %) were incubated with bovine pancreatic RNase A to determine the protection of encapsulated siRNA.  As shown in Figure 2.5, gel electrophoresis indicates that free siRNA is degraded, while the siRNA encapsulated in the LNP formulated by the microfluidic method is completely protected.  Addition of the detergent Triton X-100 dissolves the LNP, releases the siRNA, and results in siRNA degradation in the presence of RNase.    Figure 2.5  siRNA encapsulated in LNP is fully protected from external RNase siRNA was either employed in the free form or encapsulated in LNP containing DLinKC2-DMA /DSPC/Chol/PEG-lipid (40/11.5/47.5/1; mol/mol) at an siRNA/lipid ratio of 0.06 (wt/wt).  Encapsulation was performed using the microfluidic mixer as indicated in Section 2.2.3. The integrity of the siRNA was challenged with 1 μg/mL of bovine pancreatic RNase A.  5% Triton X-100 was added to solubilize the LNP.  Gel electrophoresis was performed on 20% native polyacrylamide gel and siRNA visualized by staining with SYBR-Safe.  70  2.3.6 Encapsulated siRNA is complexed with internalized cationic lipid As indicated above, the electron dense core of the siRNA-LNP systems may be suggested to consist of encapsulated siRNA complexed to cationic lipid and the remaining lipid (cationic lipid, phospholipid, cholesterol and PEG-lipid) is either present in the core in inverted micellar or related nanostructures, or is resident on the LNP exterior.  It would then be expected that at high siRNA contents corresponding to siRNA-to-cationic lipid charge ratios of 1, where all the cationic lipid is complexed with internalized siRNA, little or no cationic lipid would be present on the external monolayer.  In order to test whether this is the case a fluorescence resonance energy transfer (FRET) assay for exterior cationic lipid was developed.  This assay, which is essentially a membrane fusion assay, utilized negatively charged bilayer lipid vesicles composed of dioleoylphosphatidylserine (DOPS) that contained the FRET pair, NBD-PE and Rh-PE at 1 mol% each.  LNP siRNA systems consisting of DLin-KC2-DMA/DSPC/Chol/PEG-lipid (40/11.5/47.5/1 mol %) were added to the DOPS vesicles and incubated at pH 5.5.  The pKa of DLinKC2-DMA is 6.7 (Semple et al., 2010) and thus more than 90% of the DLinKC2-DMA on the outside of the LNP will be charged at pH 5.5, potentially promoting fusion with negatively charged DOPS LNP.  As indicated elsewhere (Hafez et al., 2000; Struck et al., 1981) fusion is observed as an increase in the NBD-PE fluorescence at 535 nm as the NBD-PE and Rh-PE probes become diluted following lipid mixing. As shown in Figure 2.6, when the LNP systems contained no siRNA, substantial fusion was observed, consistent with a considerable proportion of the DLin-KC2-DMA residing on the outer monolayer of the LNP system.  When the LNP systems contained siRNA at a siRNA-to-total lipid ratio of 0.06 (wt/wt), which corresponds to an siRNA-to-cationic charge ratio of 0.25, however, the extent of fusion was reduced (Figure 2.6), and for siRNA-LNP systems prepared 71  with an siRNA-to-cationic lipid charge ratio of one little or no fusion was observed, indicating that little or no DLin-KC2-DMA was present on the LNP siRNA exterior.  This supports the hypothesis that in LNP with high siRNA content all the cationic lipid is complexed with siRNA and sequestered in the LNP interior.   Figure 2.6  Cationic lipid is associated with internalized siRNA in siRNA-LNP systems The amount of external cationic lipid in siRNA-LNP systems was assayed as a function of the siRNA phosphate-to-cationic lipid charge ratio using the FRET lipid mixing assay described in Methods.  Three LNP systems DLin-KC2-DMA/DSPC/Chol/PEG-lipid (40/11.5/47.5/1; mol/mol) were prepared at charge ratios of 0 (solid line), 0.25 (dotted line) and 1 (dash line).  The lipid mixing assay was performed at pH 5.5 to ensure that essentially all external DLin-KC2-DMA was positively charged.  The reaction was initiated by injecting the LNP (at t=30 s) into a stirred cuvette containing the anionic DOPS/NBD-PE/Rh-PE (98:1:1 molar ratio) vesicles.  72  2.3.7 LNP siRNA systems have a different density than aqueous core bilayer vesicles If the siRNA-LNP systems exhibit a lipid core, they would be expected to exhibit a different density as compared to vesicles with an aqueous core.  In the absence of siRNA where the interior consists of inverted micelles of cationic lipid the density should be less than vesicular systems, and the density should increase as more siRNA is encapsulated.  As shown in Figure 2.7, when empty LNPs without siRNA and LNP containing siRNA at 0.06 siRNA-to-total lipid ratio are centrifuged on a 1 – 15% sucrose step gradient as described under Methods, the LNP remained on the top of the gradient (Figure 2.7).  In a parallel experiment, POPC/cholesterol (1:1; mol/mol) bilayer vesicles were centrifuged on an identical gradient and the vesicles distributed as a broad peak centered at around fraction 7 of the gradient.  Increasing siRNA content in the LNP results in an increase in density as LNP containing siRNA at a 0.24 siRNA-to-total lipid ratio (corresponding to a 1:1 siRNA/cationic lipid charge ratio) were much denser, exhibiting a peak centered at fraction 15 of the column.  In an additional experiment (data not shown) empty LNPs and LNP containing siRNA at 0.06 siRNA-to-lipid ratio were introduced first, at the bottom of the centrifuge tube, and a 1-10% sucrose step gradient layered on top.  After centrifugation, the LNP were found to have redistributed to the top of the gradient.  These results provide strong evidence that siRNA-LNP systems contain a lipid core with a density dependent on the amount of siRNA encapsulated.    73   Figure 2.7  The density of siRNA-LNP systems is consistent with a hydrophobic lipid core as indicated by density gradient ultracentrifugation A 1-15% sucrose step gradient was used as described in Section 2.2.8. Fractions (500 µL) were successively removed from the top of gradient following centrifugation at 190,000 g for 18 h and were assayed for cholesterol in POPC/cholesterol bilayer vesicles (open circles), empty LNP system (filled squares), siRNA-LNP systems at a siRNA/cationic lipid charge ratios of 0.25 and 1 (filled triangles and filled diamonds, respectively).   74  2.3.8 Simulation results indicate that LNP siRNA systems exhibit a nanostructured core The first step in computer modeling was to simulate the self-assembly of a putative unit cell for a lipid nanoparticle.  To this end a mixture of DLin-KC2-DMA, DSPC, cholesterol and nucleic acids was placed in a small box in a random configuration (Figure 2.8A).  The self-assembly was performed in several independent simulations at different hydration levels to explore possible structures formed in the mixture.  We also simulated self-assembly of a medium-sized system containing a polymer-grafted lipid in addition to other components.  Comparing resulting structures, we found that a small system at low hydration (Figure 2.8B) was similar to the core of the self-assembled medium sized system and selected it as a building block for a larger LNP.  The large LNP was constructed by multiplying the building block, coating the resulting structure with a PEG-lipid layer, and re-solvating the system (see Section 2.2.9 and Leung et al., 2012).  Small spacing between the unit cells and between the polymer layer and the nanoparticle was allowed to adjust the water contents upon equilibration of the LNP.   During the equilibration (~ 1 µs), the water compartments closed and the PEG-grafted lipids adsorbed on its surface with polymer chains oriented towards the solution.  On the simulation time, the LNP gradually transformed into a smooth rounded capsule with an averaged diameter of approximately 44 nm (Figure 2.9A).  The outer layer of the LNP is constituted by a roughly homogeneous coating of PEG-lipids.  Inside the LNP, irregular water filled compartments of diameters ranging from approximately 3 to 9 nm are separated by bilayer membranes with DNAs bound to the membrane surface (Figure 2.9 B-D).  The structure of the LNP core resembles an inverted hexagonal HII phase distorted at a high hydration level (Marrink and Mark, 2004).  The volume of water trapped inside the LNP constitutes approximately 6 waters per lipid which corresponds to 0.10 μL/μmol lipid, in good agreement with experimental data (data not shown). 75  Given the significant fraction of lipids on the LNP surface (~ 1/4), we expect this number to be slightly higher for a larger size LNP.  In the LNP core, the numbers of nearest neighbors in the first coordination shell are given in Table 2.1.  There are strong preferential interactions between the phosphates of the nucleic acid and positively charged groups of DLin-KC2-DMA lipids.  Increased numbers of DLin-KC2-DMA - DSPC neighbors originate from favorable interactions of the positively charged group of the cationic lipid with the negatively charged phosphate groups of DSPC.  These interactions lead to formation of molecular pairs which manifest in the radial distribution function (RDF) as a pronounced second peak at a distance of ~ 1nm (data not shown).  Spatial density distributions for selected groups are shown in Figure 2.10.  The densities of headgroups are higher in the LNP core (Figure 2.10A) than on the LNP surface (Figure 2.10B).  This is expected from a high negative curvature of the water-filled compartments inside the LNP, and the presence of polymer-grafted lipids on the surface.  The distribution of DLin-KC2-DMA around DSPC lipids is shifted towards the DNA phosphates, and cholesterol is distributed somewhat away from DLin-KC2-DMA (Figure 2.10A).  Due to a high cationic lipid to DNA ratio, not all cationic lipids in the LNP core are bound to DNA.  These free (Figure 2.10C) and bound (Figure 2.10D) cationic lipid fractions have a noticeably different 3D neighborhood.  A somewhat heterogeneous surrounding of DSPC and cholesterol aligned along the linker of the free lipid becomes for the bound lipid displaced away and to the side by DNA phosphates.     76    Figure 2.8  Self-assembly from a random configuration (a) into a building block (b) for a lipid nanoparticle (LNP) A mixture of DLin-KC2-DMA, DSPC and cholesterol (576 DLin-KC2-DMA lipids, 144 DSPC lipids and 576 cholesterol molecules; 44/11/44; mol/mol) is placed in a small simulation box at a low hydration level, see text. DLin-KC2-DMA is shown in yellow, cholesterol in pink, DSPC in grey, lipid polar moiety in cyan, nucleic acids (12 bp duplex DNA) in red, water not shown for clarity.   77    Figure 2.9  A lipid nanoparticle (LNP) contains irregular water-filled cavities separated by bilayer membranes, with nucleic acids bound to membrane surfaces (A) side view, (B,C) cross-section, and (D) zoom-in views. Cationic lipid DLin-KC2-DMA  is shown in yellow, cholesterol in pink, DSPC in grey, lipid polar moiety in cyan, PEG-lipid in violet, nucleic acids (duplex  DNA) in red, water not shown for clarity. The lipid composition was DLin-KC2-DMA/DSPC/cholesterol/PEG-lipid (4:1:4:1; mol/mol) and DNA to lipid ratio ~ 0.05 wt/wt.    78  Table 2.1  Numbers of nearest neighbors in the first coordination shell for each molecule in the LNP core. Estimated numbers in the absence of preferential interactions are shown in brackets. Neighbors for DNA are averaged over the entire LNP. The numbers are calculated for DNA phosphates, positively charged moiety of DLin-KC2-DMA, the cholesterol polar group and the DSPC phosphate group (or choline group in the case of DNA neighbors) by integrating the RDFs over the first maximum.  # of neighbors for molecule: DLin-KC2-DMA cholesterol DSPC DNA phosphates DLin-KC2-DMA 0.76 (0.78) 0.57 (0.78) 0.43 (0.2) 0.84 cholesterol 0.58 (0.51) 0.43 (0.51) 0.14 (0.13) 0.14 DSPC 1.70 (1.10) 0.51 (1.10) 0.27 (0.28) 0.24 DNA (phosphate) 2.59 (1.43) 0.43 (1.43) 0.19 (0.36) - 79    Figure 2.10  Spatial density distributions for selected molecular groups around DSPC and DLin-KC2-DMA Spatial density distributions for selected molecular groups around DSPC in the LNP core (a) and on the LNP surface (b), and DLin-KC2-DMA in the LNP core free (c) and bound to DNA (d). The surfaces of constant number densities are plotted at the values corresponding approximately to the average density in the first coordination shell. In the molecular representations, DSPC headgroup is shown as dark grey, glycerol-ester region as light grey, and hydrocarbon chains as transparent grey beads; DLin-KC2-DMA headgroup is shown as yellow, linker as light yellow, and chains as transparent yellow beads. In the density distributions, DNA phosphates are colored in semi-transparent red, DLin-KC2-DMA headgroup in cyan and linker in yellow, cholesterol polar group in pink, DSPC headgroup in dark grey and glycerol-ester in light grey.  80  2.4 Discussion The model for siRNA-LNP structure developed by molecular simulation (Figure 2.9) is fully consistent with the experimental results presented here.  The model shows that encapsulated siRNA is located in internalized distorted inverted micelles complexed to cationic lipid, with remaining lipids organized around small internal aqueous compartments.  Such organization could account for the increased electron density as compared to bilayer vesicle systems as evidenced by cryo-TEM studies.  This organization is also consistent with the ability of particles composed of DLinKC2-DMA and PEG-lipid to adopt structures as small as 14 nm diameter in the absence of siRNA and 23 nm diameter in the presence of siRNA, as well as the lack of external cationic lipid in siRNA-LNP systems prepared at high siRNA-to-cationic lipid charge ratios and the essentially complete protection of encapsulated siRNA from external RNase.  The modeling results provide interesting insight into the location and role of DSPC. Previously DSPC has been included to provide increased stability to the preformed vesicles used in the preformed vesicle formulation method (Semple et al., 2001) and has been thought to play a stabilizing role in the LNP structure formed.  The results presented here suggest that DSPC could also be forming ion pairs between the DSPC phosphate group and cationic lipid headgroup, leaving the DSPC choline function to associate with siRNA phosphates.  Such interactions have been suggested by other investigators (Lee et al., 2009).  A final point is that, as suggested by the molecular modeling results, it is likely that cholesterol is distributed roughly homogeneously in the LNP core and surface, given its ability to partition into both lamellar and inverted lipid structures such as the hexagonal HII phase (Cullis and de Kruijff, 1979). A major advantage of the model presented in Figure 2.9 is that it can explain how siRNA encapsulation efficiencies approaching 100% can be achieved during the formulation process.  81  Specifically, during the microfluidic synthesis process the following steps are achieved: first, rapid mixing between the aqueous phase containing siRNA and the ethanol phase containing cationic lipid, second the association of cationic lipid with siRNA to form hydrophobic nucleating structures, and third, as the polarity of the medium increases further, coating of the nucleating structures by remaining lipids as they reach their solubility limits in the ethanol/water system, forming the final siRNA-LNP structure.  In this picture it is clear how essentially all of the siRNA could be incorporated into LNP systems at non-saturating siRNA loading, leading to complete encapsulation. There are, however, other structures that could be compatible with high encapsulation efficiencies. In previous work (Maurer et al., 2001; Semple et al., 2001), we have encapsulated antisense oligonucleotides into LNP systems containing ionizable cationic lipids with encapsulation efficiencies of up to 70% using the PFV method.  The LNP formed are small multilamellar vesicles where the oligonucleotide is apparently located at the lamellar interfaces.  These systems contained higher levels of the bilayer forming lipid DSPC and lower levels of cationic lipid than employed here.  In order to show that the systems we observe here are part of a continuum of structures that are sensitive to the proportions (and types) of lipid components, we examined the cryo-TEM morphology of an siRNA-LNP system containing a larger proportion of DSPC and a lower proportion of cationic lipid (DLin-KC2-DMA/DSPC/Chol/PEG-lipid in the proportions 20/31.5/47.5/1; mol/mol, as opposed to 40/11.5/47.5/1; mol/mol).  This system also exhibited high encapsulation efficiencies above 90% but also showed evidence of multilamellar structure, as shown in Figure 3.2. It would therefore appear that the internal structure of siRNA-LNP systems changes from multilamellar to inverted micellar as the proportion of bilayer-forming species is reduced, as may be expected based on 82  polymorphic phase preferences of component lipids.  The effects of cationic lipid and siRNA content have on the structure of the siRNA-LNP system will be discussed in more detail in Chapter 3. It is likely that the results presented here for siRNA-LNP systems produced by microfluidic mixing will also extend to siRNA-LNP systems composed of the same or similar lipids but prepared by other in-line mixing protocols such as the T-tube mixer, as long as the mixing rates are sufficiently fast. In this regard it may be noted that siRNA-LNP systems containing ionizable cationic lipids prepared by T-tube mixing show the distinctive electron-dense core by cryo-TEM (Crawford et al., 2011).   83  Chapter 3: Structural Properties of Lipid Nanoparticles Containing Nucleic Acid Polymers Formulated by Microfluidic Mixing  3.1 Introduction In the previous Chapter it was demonstrated that LNP produced by microfluidic mixing contain a solid-core interior as observed by cryo-TEM.  Experimental data and molecular modeling indicated that the solid core interior of the LNP contains inverted micellar structures, some of which contain siRNA.  There are three areas that require further investigation.  First, are there additional ways in which the inverted micellar structure can be confirmed?  Second, previous work using a related formulation technique to encapsulate antisense oligonucleotides using different lipid compositions revealed an LNP structure that consisted of concentric bilayers rather than a uniformly electron dense interior (Maurer et al., 2001).  This raises the question as to the structure of LNP oligonucleotide systems when a higher proportion of “bilayer” components are present.  Third, can the microfluidic formulation technique be extended to longer nucleic acid polymers such as mRNA and plasmid DNA?  In this Chapter, the structural features of LNP containing gold nanoparticles, LNP containing a range of lipid compositions and LNP containing mRNA and plasmid DNA are examined.  It is shown that encapsulation of gold nanoparticles containing siRNA coupled to the gold surface results in a “currant bun” morphology as detected by cryo-TEM, consistent with the inverted micellar model for siRNA encapsulation previously proposed.  Second, it is shown that as the amount of bilayer lipid in LNP siRNA systems is increased, an increasing proportion of bilayer lipid structure is observed in the LNP siRNA systems without compromising siRNA encapsulation efficiency, however, as the proportion of bilayer lipid is reduced below 10 mol% the encapsulation efficiency is reduced significantly.  Third, it is shown that the same 84  microfluidic mixing technique employed for encapsulation of siRNA can also be applied to mRNA and DNA, resulting in LNP systems with an electron-dense core. These results are discussed with regard to LNP siRNA structure, the role of DSPC in LNP structure and the potential utility of LNP systems for delivery of mRNA and plasmid DNA.  3.2 Materials and Methods 3.2.1 Materials The lipids 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine (DOPE) and 1,2-distearoyl-sn-glycero-3-phosphocholine (DSPC) were obtained from Avanti Polar Lipids (Alabaster, AL).  Cholesterol, sodium acetate and sodium citrate tribasic were obtained from Sigma-Aldrich (St. Louis, MO).  2-(N-Morpholino)ethanesulfonic acid (MES) was obtained from BDH (Westchester, PA).    The ionizable cationic lipid 2,2-dilinoleyl-4-(2-dimethylaminoethyl)-[1,3]-dioxolane (DLin-KC2-DMA) and N-[(methoxy polyethylene glycol 2000 carbamyl]-1,2-dimyristyloxlpropyl-3-amine (PEG-c-DMA) were obtained from AlCana Technologies Inc., (Vancouver, BC). Cholesterol E Total Cholesterol assay kit was obtained from Wako Diagnostics (Richmond, VA).  Quant-it RiboGreen RNA assay kit was obtained from Molecular Probes (Eugene, OR). 3.2.2 Encapsulation of siRNA-conjugated gold nanoparticle SiRNA-conjugated gold nanoparticles (siRNA-gold NP) were obtained from Alnylam Pharmaceuticals (Cambridge, MA).  The molar ratio of conjugated siRNA to gold nanoparticle is 6.3.  LNP were prepared by mixing appropriate volumes of lipid stock solutions in ethanol buffer with an aqueous phase containing siRNA-gold NP employing a microfluidic micro-mixer as described previously.  Appropriate amount siRNA-gold NP were dissolved in 25 mM sodium acetate buffer, pH 4.0 to achieve a final siRNA-to-lipid charge ratio of 0.25.  Lipid dissolved in 85  ethanol and 3 volumes of nucleic acids in buffer were combined in the microfluidic micro-mixer using a dual-syringe pump (PHD Ultra, Harvard Apparatus, Holliston, MA), effectively reducing the ethanol concentration to 25% upon leaving the micro-mixer.  Flow rates were set at 0.5 ml/minute for the lipid/ethanol stream and 1.5 ml/minute for the siRNA/aqueous stream to drive the solutions through the micro-mixer at a combined flow rate of 2 ml/min.  A herringbone micro-mixer (Stroock et al., 2002) was employed.  The lipid mixture was then dialyzed for 4 hours against 1000 volumes of 50 mM MES/50 mM sodium citrate buffer (pH 6.7) followed by an overnight dialysis against 1000 volumes of 1x phosphate buffered saline, pH 7.4 (GIBCO, Carlsbad, CA) using Spectro/Por dialysis membranes (molecular weight cutoff 12000 – 14000 Da, Spectrum Laboratories, Rancho Dominguez, CA). 3.2.3 Encapsulation of nucleic acids LNP were prepared by mixing appropriate volumes of lipid stock solutions in ethanol buffer with an aqueous phase containing nucleic acids employing a microfluidic micro-mixer as described elsewhere (Zhigaltsev et al., 2012).  All nucleic acids were dissolved in 25 mM sodium acetate buffer, pH 4.0.  Lipid dissolved in ethanol and 3 volumes of nucleic acids in buffer were combined in the microfluidic micro-mixer using a dual-syringe pump (PHD Ultra, Harvard Apparatus, Holliston, MA), effectively reducing the ethanol concentration to 25% upon leaving the micro-mixer.  Flow rates were set at 0.5 ml/minute for the lipid/ethanol stream and 1.5 ml/minute for the siRNA/aqueous stream to drive the solutions through the micro-mixer at a combined flow rate of 2 ml/min.  A herringbone micro-mixer (Stroock et al., 2002) was employed.  The lipid mixture was then dialyzed for 4 hours against 1000 volumes of 50 mM MES/50 mM sodium citrate buffer (pH 6.7) followed by an overnight dialysis against 1000 volumes of 1x phosphate buffered saline, pH 7.4 (GIBCO, Carlsbad, CA) using Spectro/Por 86  dialysis membranes (molecular weight cutoff 12000 – 14000 Da, Spectrum Laboratories, Rancho Dominguez, CA).   The mean diameter of the LNP after dialysis was determined by dynamic light scattering (number mode; Zetasizer Nano ZS, Malvern Instruments Inc., Westborough, MA). Lipid concentrations were determined by measuring total cholesterol using the Cholesterol E enzymatic assay from Wako Chemicals USA (Richmond, VA). Encapsulation efficiency was calculated by determining unencapsulated siRNA content by measuring the fluorescence upon the addition of RiboGreen (Molecular Probes, Eugene, OR) to the siRNA-LNP (Fi) and comparing this value to the total siRNA content that is obtained upon lysis of the LNP by 1% Triton X-100 (Ft): % encapsulation = (Ft – Fi)/Ft x 100. 3.2.4 Cryo-TEM Cryo-TEM samples were prepared by applying 3 µL LNP at 10-20 mg/ml total lipid to a standard electron microscopy grid with a perforated carbon film. Excess liquid was removed from the grid by blotting and then the grid was plunge-frozen in liquid ethane to rapidly freeze the sample using a Vitrobot system (FEI, Hillsboro, Oregon). Images were taken under cryogenic conditions (~88K) at a magnification of 50000x with an AMT HR CCD side mount camera. Samples were loaded with a Gatan 70 degree cryo-transfer holder in an FEI G20 Lab6 200kV TEM (FEI, Hillsboro, OR) under low dose conditions with an under-focus of 4-6 µm to enhance image contrast. Experiments were performed at the University of British Columbia Bio-imaging Centre (Vancouver, BC).  Particle diameters were measured from the micrographs with the aid of ImageJ (National Institute of Health, Bethesda, MD). Average diameters and standard deviations were calculated from more than 100 particles.  87  3.3 Results 3.3.1 Visualization of LNP containing siRNA conjugated to gold nanoparticles supports the nanostructured core model for siRNA-LNP systems Previous work from this laboratory has demonstrated that siRNA-LNP systems produced by microfluidic mixing of siRNA in aqueous solution with a lipid mixture containing cationic lipids dissolved in ethanol exhibit an electron-dense core when viewed using cryo-transmission electron microscopy (cryo-TEM) that is inconsistent with a bilayer membrane structure.  Molecular dynamic simulations suggest that these “solid-core” LNP contain a nanostructured core with periodic aqueous compartments containing the encapsulated siRNA (Leung et al., 2012).  In order to further validate this model, LNP systems containing siRNA chemically coupled to gold nanoparticles to enhance contrast was investigated.  The siRNA-gold nanoparticle size was 15.5 nm as measured by dynamic light scattering (number mode) and the LNP siRNA-gold systems were produced using microfluidic mixing employing the lipid mixture DLin-KC2-DMA/DSPC/Chol/PEG-lipid (40/11.5/47.5/1; mol/mol) at an siRNA/lipid charge ratio of 0.25 (Leung et al., 2012). The resulting LNP systems had an average size of 68.0 ± 8.9 nm as measured by cryo-TEM and exhibited a solid core interior featuring intensely electron-dense dots corresponding to the entrapped gold nanoparticles (Figure 3.1A).  This “currant bun” morphology provides strong support for the nanostructured core model suggested previously by experimental results and by molecular modeling (Leung et al., 2012). As shown in Figure 3.1A, individual LNP contain varying amounts of encapsulated gold nanoparticles.  If the gold particles are distributed randomly it would be expected that the number of gold particles per LNP would vary linearly with the LNP diameter. As shown in 88  Figure 3.1B there is an approximately linear correlation between the size of LNP and the number of gold nanoparticles encapsulated (Figure 3.1B, R2 = 0.623).    Figure 3.1  Cryo-TEM of LNP produced with gold nanoparticle-conjugated siRNA (A) LNP prepared with DLinKC2-DMA/DSPC/Chol/PEG-lipid (40/11.5/47.5/1; % mol) at a siRNA to lipid charge ratio of 0.25.  (B)  The number of gold nanoparticles in individual LNP was counted manually and was correlated to the size of the corresponding LNP measured from the cryo-TEM using Image J.    89  3.3.2 The nanostructure of LNP systems produced by microfluidic mixing is dependent on lipid composition The electron dense core of siRNA-LNP systems composed of DLin-KC2-DMA/DSPC/Chol/PEG-lipid (40/11.5/47.5/1; mol/mol) has been attributed to the ability of the cationic lipid/DSPC/cholesterol mixture to form inverted micellar nanostructures, some of which contain siRNA, inside the LNP.  Previous work has also shown that the electron dense core is independent of the presence of siRNA, indicating that the cationic lipid/DSPC/cholesterol lipid system can form inverted micellar nanostructures with an aqueous interior in the absence of siRNA.  In addition, other work from this laboratory has shown that LNP formed using a different mixing technique but containing similar lipids (with a higher proportion of DSPC) can exhibit a multilamellar nanostructured core as visualized by cryo-TEM (Maurer et al., 2001).  It is therefore of interest to determine the influence of lipid composition on LNP nanostructure for LNP formed using the microfluidic mixing process to characterize the continuum of structures possible. As shown in Figure 3.2A, LNP formed from DLinKC2-DMA/DSPC/Chol/PEG-lipid (40/11.5/47.5/1; mol/mol) exhibits the characteristic electron dense interior as observed by cryo-TEM.  However, if the amount of cationic lipid is reduced by 20 mol% and the proportion of (bilayer-forming) DSPC increased by 20 mol% (corresponding to a lipid composition DLinKC2-DMA/DSPC/Chol/PEG-lipid (20/31.5/47.5/1; mol/mol)) lamellar structures begin to emerge in the outermost layer of the LNP, while the interior remained solid-core (Figure 3.2B). This behavior could suggest that more hydrophobic lipid aggregates may first serve as nucleating particles, possibly containing a higher proportion of DLin-KC2-DMA, that are subsequently coated with lipids containing a higher proportion of DSPC, leading to bilayer structures. 90  An alternative way of increasing the propensity of component lipids to adopt a lamellar organization is to decrease the unsaturation of acyl chains in component lipids.  In particular, the linoleic acid composition of DLinKC2-DMA may be expected to contribute to a pronounced “cone” shape driving formation of inverted micellar structure.  Decreasing the unsaturation will reduce this structural preference and enhance the formation of bilayer structures (Cullis et al., 1986).  We therefore examined the morphology of LNP systems formed from the oleic acid analogue of DLin-KC2-DMA, namely DO-KC2-DMA.  As shown in Figure 3.2C, LNP containing 40% DO-KC2-DMA exhibit an electron dense interior, similar to LNP containing DLin-KC2-DMA.  However, when the DO-KC2-DMA is reduced to 20 mol% and the DSPC content increased to 31.5 mol%, a significant portion of the LNP exhibit morphology consistent with bilayer nanostructures (Figure 3.2D).  It may be noted that the proportion of LNP exhibiting putative bilayer morphology is greater for LNP containing DOKC2-DMA and 31.5% DSPC  (Figure 3.2D) than in the LNP containing DLinKC2-DMA and 31.5% DSPC (Figure 3.2B) consistent with a decreased ability of the DOKC2-DMA to induce inverted micellar structures.     91   Figure 3.2  Cryo TEM micrographs of LNP produced with different cationic lipid saturation and total cationic lipid content (A): LNP prepared with DLinKC2-DMA/DSPC/Chol/PEG-lipid (40/11.5/47.5/1; mol/mol).  (B)  LNP with lipid composition DLinKC2-DMA/DSPC/Chol/PEG-lipid (20/31.5/47.5/1; mol/mol).  (C)  LNP with lipid composition DOKC2-DMA/DSPC/Chol/PEG-lipid (40/11.5/47.5/1; mol/mol) and (D) DOKC2-DMA/DSPC/Chol/PEG-lipid (20/31.5/47.5/1; mol/mol).   92  3.3.3 LNP siRNA systems containing higher levels of siRNA exhibit reduced lamellar structure.  The siRNA content of siRNA-LNP systems may also be expected to influence LNP structure. In particular, it may be expected that siRNA will induce inverted micellar structure in combination with unsaturated cationic lipids due to charge association between the negatively charged oligonucleotide and the positively charged lipids.  In order to test whether this is the case the structure of LNP composed of DO-KC2-DMA/DSPC/Chol/PEG-lipid (20/31.5/47.5/1; mol/mol) that contained siRNA at siRNA-to-cationic lipid charge ratios of 0, 0.25 and 1 were investigated employing cryo-TEM.  As the siRNA/lipid charge ratio is increased, the proportion of particles displaying bilayer structure in the population of LNP decreases from nearly 90% when no siRNA is present to less than 10% at a siRNA/cationic lipid charge ratio of 1 (Figure 3.3).  This suggests that electrostatic interactions between cationic lipids and the siRNA provide a strong driving force for the formation of inverted micellar structures surrounding siRNA duplexes, leading to the uniformly electron dense core of the LNP.    93   Figure 3.3  LNP formulated with 20% DOKC2-DMA exhibit a different proportion of multilamellar structures depending on siRNA content (A) Cryo TEM of LNP composed of DOKC2-DMA/DSPC/Chol/PEG-c-DMA (20/31.5/47.5/1 mol %) formulated in the absence of siRNA.  (B) LNP with the same lipid composition (A) but with siRNA at a siRNA/lipid charge ratio of 0.25.  (C)  LNP with the same lipid composition as in (A) and (B) but prepared with siRNA at siRNA/lipid charge ratio of 1.  (D)  Particles from cryo-TEM of each sample were counted manually and visually classified into multilamellar or solid core particles.     94  3.3.4 Reduced LNP siRNA encapsulation efficiencies observed at high cationic lipid contents can be improved by incorporation of DOPE The ionizable cationic lipids contained in the siRNA-LNP systems described here serve two functions, first to allow efficient encapsulation of siRNA and second to facilitate efficient endosomal escape following uptake into target cells.  As shown elsewhere (Jayaraman et al., 2012) the gene silencing potency of LNP siRNA systems depends crucially on the pKa of the ionizable cationic lipid, a feature that has been attributed to the need to generate positively charged lipids (at endosomal pH values) that can combine with endogenous anionic lipids to disrupt the endosome and release encapsulated siRNA.  In this context it would appear potentially advantageous to increase the proportion of cationic lipid in the LNP to achieve maximal endosomal escape and gene silencing potency.  We therefore investigated the formulation properties of LNP containing increasing proportions of DLinKC2-DMA in preparations containing DLinKC2-DMA/DSPC/Cholesterol/PEG-c-DMA. The LNP were prepared with DSPC and PEG-c-DMA held constant at 11.5% and 1% respectively, while the DLinKC2-DMA content was increased from 50% to 80% with cholesterol content decreased proportionally.  All formulations were prepared at a constant siRNA/cationic lipid charge ratio of 0.25.  Surprisingly, siRNA encapsulation efficiencies dropped dramatically as the DLinKC2-DMA content was increased beyond 70% (Figure 3.4A) and this effect was even more pronounced when the formulation contained higher percentages (2.5% or 5%) of PEG-lipid. As noted elsewhere (Belliveau et al., 2012), higher PEG-lipid contents lead to smaller LNP as the PEG-lipid is located preferentially on the LNP exterior, leading to higher surface to volume ratios for higher PEG contents and correspondingly smaller sizes. The sizes of LNP siRNA systems prepared at 1, 2.5 and 5 mol% PEG lipid were 46.2 nm, 31.1 nm and 26.9 nm, 95  consistent with previous observations. The reduction in trapping efficiencies at higher PEG-lipid contents could arise, at least in part, due to the larger surface to volume ratios of the smaller LNP systems, leading to a larger proportion of cationic lipid on the surface where any associated siRNA would be exposed.   However, this cannot be the complete picture because for a given PEG-lipid content an increase in cationic lipid content can dramatically decrease trapping efficiencies without a significant change in size.  For example, for a DLinKC2-DMA formulation containing 5% PEG-lipid, increasing the DLinKC2-DMA content from 50% to 80% decreases siRNA encapsulation efficiencies from 80% to zero (Figure 3.4A), however the LNP size did not change significantly.   Another reason for low encapsulation efficiencies at high cationic lipid contents could be that the packing properties of components lipids at higher cationic lipid contents do not facilitate encapsulation.  For example, it can be estimated that the dimensions of the siRNA duplexes are 2.6 nm in diameter and 5.8 nm in length, corresponding to a surface area of approximately 31 nm2.  If each of the cationic lipids associated with the siRNA are associated with a negative charge one would expect 42 lipids per siRNA.  Although the area per molecule (Am) of the ionizable cationic lipids is unknown, it would not be expected to be greater than observed for unsaturated PCs which have Am values of approximately 0.6 nm2 (Alwarawrah et al., 2010) suggesting that the 42 cationic lipids may not cover the surface of the siRNA, requiring additional lipids, such as DSPC or cholesterol to form the inverted micellar shell.  At the high cationic lipid contents the cholesterol content is correspondingly low.  DSPC would not be expected to pack well in the inverted micellar environment surrounding the siRNA and therefore packing constraints may inhibit encapsulation at higher cationic lipid contents.  In this regard a lipid such as DOPE, which has a pronounced cone shape compatible with hexagonal HII 96  organization (Cullis et al., 1986) may provide improved packing properties.  We therefore examined the siRNA trapping efficiencies achieved at high cationic lipid contents when DSPC was replaced by DOPE.  As shown in Figure 3.4A, the siRNA encapsulation efficiencies of LNP formulations containing DOPE in place of DSPC were considerably improved.  For formulations with 1% PEG-c-DMA, the substitution of DOPE completely prevented the drop in encapsulation efficiencies at high cationic lipid content (Figure 3.4A).  Substantial improvements were also observed for formulations containing either 2.5% or 5% PEG-c-DMA.   It is of interest to characterize the structural features of the LNP siRNA systems at high cationic lipid contents.  The structure of LNP prepared with 80% DLinKC2-DMA containing either DSPC or DOPE using cryo-TEM was examined by cryo-TEM, however replacing DSPC with DOPE did not alter the structure of the resulting LNP (Figure 3.5).  Each LNP still possesses a solid, electron-dense core, similar to LNP prepared at lower cationic lipid content.  The average size of the LNP increases as cationic lipid content was increased (Figure 3.4B). This is consistent with the idea that DLinKC2-DMA preferentially locates to the interior of the particle, as the ratio of surface lipid-to-interior lipid is decreased, the size of the LNP increases.    97   Figure 3.4  Effects of cationic lipid and phospholipid contents on the siRNA encapsulation efficiency of lipid nanoparticle systems (A) LNP were formulated using 1% (blue), 2.5% (red) and 5% (green) PEG-c-DMA with varying amounts of DLinKC2-DMA and cholesterol.  Formulations were prepared using either 11.5% DSPC (solid lines) or 11.5% DOPE (dash lines).  All formulations were prepared at a siRNA/lipid charge ratio of 0.25.  Measurements of encapsulation efficiency were made using the RiboGreen Assay.  (B)  Particle size of the corresponding LNP formulations in (A).  Particle size was reported as the number-weighted mean measured using the Malvern ZetaSizer.  98   Figure 3.5  LNP-siRNA systems with high cationic lipid contents maintain solid core structures for both DSPC and DOPE (A, B, C)  Cryo-TEM of LNP composed of DLin-KC2-DMA/DSPC/cholesterol/PEG-c-DMA at a ratio of: (A) 80/11.5/7.5/1 (mol %), (B) 80/11.5/6/2.5 (mol %) and (C) 80/11.5/3.5/5 (mol %).  (D, E, F) Cryo-TEM of LNP with the same composition as (A), (B) and (C), respectively, but DSPC was replaced with DOPE.  All LNP-siRNA systems were formulated at an siRNA/lipid charge ratio of 0.25.  99  3.3.5 Microfluidic mixing can be used to encapsulate longer nucleic acid polymers such as plasmids and mRNA into LNP systems  It is of interest to determine whether the microfluidic mixing method for formulating LNP siRNA systems can be extended to encapsulation of mRNA and plasmids.  Representative plasmid (PCI-neo-eGFP, 6 kb) and mRNA (1.7 kb, expressing firefly luciferase) were chosen and the encapsulation using the lipid mixture DLin-KC2-DMA/phospholipid/cholesterol/PEG-c-DMA and the same microfluidic mixing protocol as employed for siRNA was determined.  The encapsulation efficiencies for each polynucleotide were compared at 2 different cationic lipid concentrations, 50% versus 80% DLin-KC2-DMA with either DSPC or DOPE in the formulation Figure 3.6.  Both plasmids and mRNA can be efficiently encapsulated with 50% DLinKC2-DMA LNP systems, achieving over 70% encapsulation of plasmid DNA and 90% of mRNA (Figure 3.6).  Similar to the encapsulation of siRNA, increasing the DLin-KC2-DMA amount to 80% had a negative impact on the encapsulation efficiency.  Incorporation of DOPE in place of DSPC into the high cationic lipid formulation resulted in a limited improvement in encapsulation.  An interesting feature of cryo-electron micrographs of plasmid- or mRNA-containing LNP prepared with DSPC is the “bleb” structure appearing on approximately 51% of plasmid-LNP and 20% of mRNA-LNP (Figure 3.7).  When the same formulation was prepared with DSPC replaced by DOPE, the protrusion disappeared and the LNP became a spherical, solid core particle.  The fact that the protrusion is only observed in the presence of DSPC suggests that in formulations containing plasmid or mRNA, DSPC can be segregated from other lipid components, possibly forming a membrane around the less electron-dense protrusion.  This leaves the lipids that are more compatible with inverted micellar or related structures, such as 100  DLinKC2-DMA and cholesterol, to form the electron-dense portion of the particle in association with the nucleic acids.  This interpretation is supported by the fact that when DOPE, which favours the hexagonal HII phase structure in isolation, is substituted for DSPC no blebs are observed.    Figure 3.6  Influence of cationic lipid content and phospholipid species on the plasmid and mRNA encapsulation efficiencies of lipid nanoparticle systems (A) Plasmid encapsulation efficiency of LNP prepared with DLin-KC2-DMA/phospholipid/cholesterol/PEG-c-DMA at a molar ratio of 50/11.5/37.5/1 or 80/11.5/7.5/1.  Formulations were prepared using either DSPC or DOPE as the phospholipid.  Measurements of encapsulation efficiency were made using the RiboGreen Assay.  (B)  Number-based average particle size of the plasmid-LNP as measured by dynamic light scattering.  (C)  Messenger RNA encapsulation efficiency of LNP prepared with the same lipid composition as in (A).  All formulations were prepared at a lipid/mRNA charge ratio of 4.  Measurements of encapsulation efficiency were made using the RiboGreen Assay.  (D)  Number-based average particle size of the mRNA-LNP as measured by dynamic light scattering.  101   Figure 3.7  LNP containing plasmid DNA and mRNA exhibit blebs when DSPC is present but not when DOPE is substituted for DSPC  (A, B)  Cryo-TEM of LNP-plasmid DNA prepared with the lipid composition DLinKC2-DMA/phospholipid/Chol/PEG-c-DMA (50/11.5/37.5/1 mol %) with the phospholipid being either (A) DSPC or (B) DOPE.  (C, D)  Cryo-TEM of LNP-plasmid DNA prepared with the lipid composition DLinKC2-DMA/phospholipid/Chol/PEG-c-DMA (80/11.5/7.5/1 mol%) with the phospholipid being either (C) DSPC or (D) DOPE.  All LNP-plasmid DNA systems were prepared at a plasmid DNA/lipid charge ratio of 0.17.  (E to H)  Cryo-TEM of LNP with the same lipid composition as (A) to (D) respectively, but prepared with mRNA at a mRNA/lipid charge ratio of 0.25.   102  3.4 Discussion The results presented here show that incorporation of gold nanoparticles in LNP siRNA systems results in a “currant bun” morphology, that LNP siRNA systems can exist in a continuum of bilayer and uniformly electron dense structures depending on the lipid and siRNA content, that siRNA encapsulation efficiencies are compromised at high cationic lipid contents and that the microfluidic mixing technique can be employed to efficiently encapsulate both mRNA and plasmid DNA.  There are four areas that warrant further discussion, namely the utility of LNP containing gold nanoparticles both to support the nanostructured core model of LNP siRNA systems as well as their potential utility in their own right; second the relation between the structural preferences of component lipids and LNP morphology; third the role of lipid packing as it relates to siRNA trapping efficiencies; and finally the structure and utility of LNP systems containing mRNA and plasmid DNA. We discuss these areas in turn.  The currant bun appearance of the gold nanoparticles encapsulated in the LNP systems (Figure 3.1) showing the gold-conjugated siRNA embedded in the interior of the LNP fully supports the nanostructured core model presented elsewhere (Belliveau et al., 2012). In addition, these results demonstrate efficient encapsulation of gold nanoparticles within LNP systems using the microfluidic formulation technique.  These gold-containing LNP could have considerable utility for cell and tissue imaging of LNP and contents due to the high contrast between gold nanoparticles and surrounding environments as well as the relatively low toxicity of colloidal gold preparations (Connor et al., 2005; Hainfeld et al., 2006).  In addition, due to the efficiency with which gold nanoparticles can absorb optical light leading to local heating, the potential for triggered release of LNP contents is apparent (Paasonen et al., 2007).   103    The presented studies clearly show that the morphology of LNP formulations of nucleic acid polymers produced by microfluidic mixing is highly dependent on the proportions of component lipids as well as the amount and type of nucleic acid being encapsulated. The appearance of a proportion of apparently bilayer structure as the amount of DSPC is increased or the unsaturation of the cationic lipid is decreased suggests that the uniformly electron dense interior of LNP siRNA systems observed for high (40 mol% or more) contents of DLinKC2-DMA and low (11.5 mol%) proportions of DSPC (Leung et al., 2012) represents one extreme of LNP-oligonucleotide organization with an interior consisting of inverted micelles, whereas the other extreme is LNP containing concentric bilayers as observed for a lipid composition containing low (20 mol%) levels of cationic lipid and high (25 mol%) levels of DSPC (Semple et al., 2001).  As shown here at low levels of cationic lipid (20 mol%) and high (31.5 mol%) levels of DSPC, significant bilayer structure in the LNP siRNA systems is observed in combination with electron-dense cores.  The differences between the low cationic lipid and high DSPC contents observed here as compared to previous work (Semple et al., 2001) can be attributed to the different mode of sample preparation, the use of a more saturated cationic lipid and the entrapment of antisense oligonucleotides as compared to siRNA.  In particular, the enhanced electronegativity of double-stranded siRNA as compared to single stranded DNA may be expected to drive formation of inverted structure more effectively.  The reduced siRNA encapsulation efficiencies observed at high cationic lipid contents have interesting implications with regard to the packing properties of component lipids.  The observation that trapping efficiencies decrease dramatically when the cationic lipid content is increased to 80 mol% and the cholesterol content is decreased to 7.5 mol% is counter-intuitive as it may be expected that more positive charge should accommodate more siRNA during the 104  encapsulation process.  However it is apparent that high levels of cationic lipid, in combination with reduced levels of cholesterol, result in poor encapsulation properties.  The fact that siRNA encapsulation efficiencies can be rescued by substituting DOPE for DSPC argues for a role of DOPE in forming the inverted micelles containing siRNA.    The ability of microfluidic mixing approaches to enable the efficient encapsulation of mRNA and plasmid DNA in LNP systems clearly provides well-defined delivery systems of considerable interest for promotion of gene expression as opposed to gene silencing.  It will be of interest to determine whether such systems exhibit the same in vivo apoE-dependent transfection capabilities as LNP siRNA systems for hepatocytes and neurons (Akinc et al., 2010; Rungta et al., 2013).  With regard to the structural features of LNP containing mRNA and plasmid DNA, the appearance of apparently bilayer “blebs” on the LNP systems containing DSPC suggests the segregation of bilayer-preferring lipids such as DSPC in such regions.  This may also be suggested by a calculation of amount of cationic lipid required to encapsulate mRNA or plasmid DNA in an LNP particle. For example, for a 5 kb plasmid, encapsulation of a single plasmid in an LNP will require association with 10,000 cationic lipids. Assuming a LNP density of 0.9 g/mL each 90 nm LNP plasmid system would contains three plasmids. The charge association of cationic lipid and plasmid may be expected to segregate the cationic lipid from DSPC leading to the blebbing phenomena observed. Such segregation may also occur when DOPE is present, however the preference of DOPE for inverted structures would suggest enhanced ability to reside in the hydrophobic core of the LNP, eliminating the bleb structures.  Encapsulation of plasmid DNA can also be achieved using the detergent dialysis method (Tam et al., 2000; Wheeler et al., 1999) or the "T-tube" method (Jeffs et al., 2005).  The detergent dialysis method is not scalable whereas the T-tube mixing process is currently the 105  preferred process for large scale manufacturing of LNP siRNA systems. However, the high flow rate (~ 1 mL/s) to required to achieve rapid mixing of the ethanol and aqueous streams (Jeffs et al., 2005) using the "T-tube" method makes the method difficult to employ for small sample volumes and does not allow production of LNP at high cationic lipid contents or sizes below 50 nm  (Belliveau et al., 2012).  The herring-bone micromixer creates more reproducible and controllable mixing of two fluids at a much lower flow rate (~ 2 mL/min). For plasmid-LNP systems the nuclear envelope is the major barrier to the translocation of plasmid DNA into the nucleus and efficient transgene expression is only observed in rapidly dividing cell as the plasmid DNA can only enter the nucleus when the nuclear envelope disintegrates during mitosis (Bally et al., 1999; Mortimer et al., 1999).  Messenger RNA-LNP systems eliminate the need for nuclear translocation (Hecker, 2013; Zou et al., 2010).  The therapeutic potential of mRNA was demonstrated by the fact that chemically modified mRNA can mediate the expression of the target protein, SP-B, in the lungs following intratracheal administration in SP-B deficient mice (Kormann et al., 2011).  However many tissues are inaccessible by local administration and the LNP mRNA systems described here should allow systemic delivery with particular potency for tissues such as the liver (hepatocytes). Recently, it was demonstrated that chemically modified mRNA can be encapsulated inside of preformed cationic vesicles with the help of protamine to condense the mRNA (Wang et al., 2013).  Our results demonstrate that the encapsulation of mRNA in LNP systems can be accomplished with ease by using microfluidic mixing eliminating the need for nucleic acid-condensing agents such as protamine.  Microfluidic mixing to produce LNP mRNA systems results in particles of 70 nm in diameter with an mRNA encapsulation efficiency of over 90% and are expected to display biodistribution and pharmacokinetic properties similar to that of siRNA-LNP systems.  The 106  ability to systemically deliver mRNA has considerable potential for the treatment of cancer and a variety of genetic disorders. In summary the results presented in this work provide additional support for the nanostructured core model of LNP siRNA systems, provide new LNP nanogold, plasmid and mRNA systems that could have significant potential as contrast and therapeutic agents, and provide insight into the roles that the different lipid components play in determining LNP structure. It is anticipated that these findings will materially aid in the production of more potent LNP systems for treatment of disease.      107  Chapter 4: Influence of Helper Lipids and High Cationic Lipid Contents on the Transfection Potency of siRNA-Lipid Nanoparticles Formulated by Microfluidic Mixing 4.1 Introduction The components of the LNP systems investigated in this thesis are ionizable cationic lipid, DSPC, cholesterol and PEG-lipid.   Detailed studies have been conducted to develop optimized cationic lipids and the PEG-lipids have been developed that exhibit appropriate dissociation rates to allow uptake into target cells in vivo.  The role of cholesterol has received less attention, however cholesterol is well-known to exchange into LNP systems in vivo (Rodrigueza et al., 1993) and thus inclusion of cholesterol during the formation of the LNP is more likely to represent the equilibrium particle composition.  The component that has received little attention is DSPC, which was included as a “structural” component in early studies (Semple et al., 2001) but whose actual role is quite undefined.  Here the importance of DSPC is investigated in the context of the use of helper lipids in transfection processes.  In addition, the influence of higher levels of cationic lipid is characterized and the impact of introducing a novel “reversed headgroup” lipid into the LNP formulations is studied. Helper lipids are typically neutral lipids used to increase the transfection efficiency of cationic lipid-DNA complexes (Felgner et al., 1994; Mok et al., 1999).  The improvement in transfection efficiency has been attributed to their ability in promoting non-bilayer phases that are thought to be crucial in the disruption of endosomal membranes (Koltover et al., 1998).  The most commonly used "helper" lipid in lipid-mediated transfection is the zwitterionic phospholipid dioleoyl-phosphatidylethanolamine (DOPE) (Farhood et al., 1995).   It is well established that DOPE induces bilayer to hexagonal HII phase transition in lipids that would normally adopt the bilayer phase (Cullis and de Kruijff, 1978b; Cullis et al., 1978b).  The small 108  cross-sectional area of the ethanolamine headgroup in DOPE give the lipid a overall conical molecular structure that is capable of introducing negative curvature to an otherwise planar membrane (Cullis et al., 1986).  This "helper" activity is not restricted to DOPE as other lipids that adopt or promote non-bilayer structures can be included in the lipid formulation to improve transfection efficiency (Hirsch-Lerner et al., 2005). Previous studies have used a maximum of 50 mol% of cationic lipid due to difficulties in achieving reasonable siRNA encapsulation efficiencies for higher cationic lipid contents.  As detailed in the previous Chapter, by using DOPE in place of DSPC cationic lipid contents as high as 80 mol% can be achieved, offering the possibility of determining whether higher cationic lipid contents can improve potency.   Finally, the properties of a novel "reverse-headgroup" lipid with strong bilayer-destabilizing abilities as a helper lipid in the LNP formulation are investigated.  By replacing the bilayer-stabilizing distearoyl-phosphatidylcholine (DSPC) with a more fusogenic "reverse headgroup" lipid, the resulting LNP is expected to possess greater endosomolytic capability.  The results presented in this study help to unravel the roles of cationic lipid and zwitterionic phospholipids play in the intracellular release of siRNA from LNP systems. 4.2. Materials and Methods 4.2.1. Materials The lipids 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine (DOPE) and 1,2-dioleoyl-sn-glycero-3-phosphoserine (DSPC), 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine-N-(7-nitro-2-1,3-benzoxadiazol-4-yl) (NBD-PE), and 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine-N-(lissamine rhodamine B sulfonyl) (Rh-PE) were obtained from Avanti Polar Lipids (Alabaster, AL).  4-(2-Hydroxyethyl) piperazine-1-ethanesulfonic acid (HEPES), and cholesterol, were 109  obtained from Sigma-Aldrich (St. Louis, MO).  Triton X-100 and 2-(N-Morpholino)ethanesulfonic acid (MES) were obtained from BDH (Westchester, PA).  Ammonium acetate, sodium acetate and sodium chloride were obtained from Fisher Scientific (Fair Lawn, NJ).  The ionizable cationic lipid 2,2-dilinoleyl-4-(2-dimethylaminoethyl)-[1,3]-dioxolane (DLin-KC2-DMA) and N-[(methoxy polyethylene glycol 2000 carbamyl]-1,2-dimyristyloxlpropyl-3-amine (PEG-c-DMA) were obtained from AlCana Technologies Inc., (Vancouver, BC). Cholesterol E Total Cholesterol assay kit was obtained from Wako Diagnostics (Richmond, VA).  Quant-it RiboGreen RNA assay kit was obtained from Molecular Probes (Eugene, OR). 4.2.2 Cell culture The androgen-sensitive human prostate adenocarcinoma (LNCap) cells were cultured in RPMI medium supplemented with 10% fetal bovine serum.  All cells were incubated at 37oC with 5% CO2.  All cell culture reagents were obtained from Invitrogen (Burlington, ON). 4.2.3 TNS assay Approximately 8 mg of the lipid DOFAB was dissolved in anhydrous ethanol and then injected a vortexing test tube containing 25 mM sodium acetate buffer at pH 4.0 to achieve a final lipid concentration of 2 mg/mL and ethanol concentration of 40%.  The lipid suspension was dialyzed in 25 mM sodium acetate buffer at pH 4.0 overnight to completely removed the ethanol.  The pKa of DOFAB was determined using the fluorescence probe, TNS, which increases in fluorescence quantum yield upon binding to lipid membrane.  The DOFAB solution was diluted to 100 μM lipid in 2 mL of a poly-buffered solution containing 10 mM ammonium acetate, 10 mM MES, 10 mM HEPES, 130 mM of NaCl, where the pH ranged from 2 to 11.  TNS was added into the cuvette at a final concentration of 1 μM just prior to the commencement 110  of measurement.  Fluorescence was monitored using a Perkin Elmer LS-55 spectrofluorometer at excitation wavelength of 321 nm and emission wavelength at 445 nm.  All readings were subtracted by the reading obtained at the highest pH, which represents the baseline fluorescence, followed by normalization to the reading obtained at the lowest pH.  The final pH of the solutions was determined after each measurement.  All measurements were done in replicates. 4.2.4 31P NMR spectroscopy Lipid films were hydrated in the NMR tube with 100 mM HEPES, 50 mM NaCl (pH 7.0) to achieve a final concentration of 25 mM phospholipids.  The samples were freeze-thawed 5 times with vigorous vortexing in between cycles.  Samples were stored at –20 °C until data acquisition.  Proton decoupled 31P NMR spectra were obtained using a Bruker AVII 400 spectrometer operating at 162MHz.   Free induction decay (FID) corresponding to 4000 scans were obtained with a 15 μs, 55o pulse with a 1 s interpulse delay and a spectral width of 64 kHz.  An exponential multiplication corresponding to 50 Hz of line broadening was applied to the FID prior to Fourier transformation.  The sample temperature was regulated using a Bruker BVT 3200 temperature unit. 4.2.5 Encapsulation of nucleic acids LNP were prepared by mixing appropriate volumes of lipid stock solutions in ethanol buffer with an aqueous phase containing nucleic acids employing a microfluidic micro-mixer as described elsewhere (Zhigaltsev et al., 2012).  All nucleic acids were dissolved in 25 mM sodium acetate buffer, pH 4.0.  Lipid dissolved in ethanol and 3 volumes of nucleic acids in buffer were combined in the microfluidic micro-mixer using a dual-syringe pump (PHD Ultra, Harvard Apparatus, Holliston, MA), effectively reducing the ethanol concentration to 25% upon leaving the micro-mixer.  Flow rates were set at 0.5 ml/minute for the lipid/ethanol stream and 1.5 111  ml/minute for the siRNA/aqueous stream to drive the solutions through the micro-mixer at a combined flow rate of 2 ml/min.  A herringbone micro-mixer (Stroock et al., 2002) was employed.  The lipid mixture was then dialyzed for 4 hours against 1000 volumes of 50 mM MES/50 mM sodium citrate buffer (pH 6.7) followed by an overnight dialysis against 1000 volumes of 1x phosphate buffered saline, pH 7.4 (GIBCO, Carlsbad, CA) using Spectro/Por dialysis membranes (molecular weight cutoff 12000 – 14000 Da, Spectrum Laboratories, Rancho Dominguez, CA).   The mean diameter of the LNP after dialysis was determined by dynamic light scattering (number mode; Zetasizer Nano ZS, Malvern Instruments Inc., Westborough, MA). Lipid concentrations were determined by measuring total cholesterol using the Cholesterol E enzymatic assay from Wako Chemicals USA (Richmond, VA). Encapsulation efficiency was calculated by determining unencapsulated siRNA content by measuring the fluorescence upon the addition of RiboGreen (Molecular Probes, Eugene, OR) to the siRNA-LNP (Fi) and comparing this value to the total siRNA content that is obtained upon lysis of the LNP by 1% Triton X-100 (Ft): % encapsulation = (Ft – Fi)/Ft x 100. 4.2.6. FRET membrane fusion studies. Fusion capabilities between LNP produced with various helper lipids and anionic DOPS bilayer vesicles was assayed by employing a fluorescence resonance energy transfer lipid mixing assay (Hafez et al., 2000; Struck et al., 1981).  Lipid mixing experiments were conducted as previously described (Leung et al., 2012).  Briefly, LNP composed of 40% DLinKC2-DMA, 47.5% cholesterol, 1% PEG-c-DMA and 11.5% of the helper lipids, DSPC, DOPE or DOFAB were prepared in the absence of siRNA.  Labeled DOPS vesicles containing NBD-PE and Rh-PE (1 mol % each) were prepared by hydration of the lipid in a thin film with 20 mM HEPES buffer 112  at pH 7.0 followed by 10 extrusions through two stacked 100 nm pore size polycarbonate filters (Nuclepore).    Unlabeled LNP were mixed with fluorescent DOPS vesicles at a 2:1 lipid molar ratio (200 μM LNP:100 μM DOPS vesicles) in a cuvette containing 2 mL of 10 mM ammonium acetate, 10 mM MES, 10 mM HEPES and 130 mM NaCl equilibrated to pH 5.5.  Fluorescence of NBD-PE was monitored using 465 nm excitation, and 535 nm emission using a Perkin-Elmer LS-55 fluorimeter under continuous low speed stirring.  Triton X-100 was to added to cuvette to final concentration of 0.1% (v/v) after approximately 10 minutes to disrupt all lipid vesicles, representing infinite probe dilution.  Lipid mixing was expressed as a percentage of infinite probe dilution determined using the equation: % lipid mixing = (F – Fo) / (Fmax – Fo) x 100, where F is the fluorescence intensity measured by the assay, Fo is the initial fluorescence intensity of NBD-PE/Rh-PE/DOPS vesicles and Fmax is the maximum fluorescence intensity at infinite probe dilution after the addition of Triton X-100. 4.2.7 Gene knockdown Cells were seeded at 200,000 cells/well in 6-well cell culture plates with 2 mL of medium per well and were allowed to grow overnight.  Fresh medium containing the desired concentration of LNP was added the following day.  The cells were treated continuously for 48 hours.  Protein was extracted with lysis buffer (0.5% deoxycholic acid, 1% NP-40 and 0.1% SDS in 1x PBS supplemented with protease inhibitor cocktail) by incubating on ice for 30 minutes.  Total protein (10 μg) was immunoblotted with antibodies against the androgen receptor (Santa Cruz Biotechnology, Santa Cruz, CA) and β-actin (AbCam, Cambridge, MA).  Antigen-antibody complexes were detected using Millipore Immobilon Western Chemiluminescent HRP Substrate (Billerica, MA).   113  4.2.8 Confocal microscopy Cells were seeded onto poly-D-lysine cover slips (BD BioSciences, Bedford, MA) at 200,000 cells/well in 6-well cell culture plates.  Fresh medium containing the desired concentration of LNP was added the following day and cells were treated continuously for 48 hours.  Cells were washed once with PBSCM (1x PBS supplemented with 1 mM MgCl2 and 0.1 mM CaCl2) and fixed with 3% paraformaldehyde for 15 minutes.  Cells were then washed once with PBSCM and cell nuclei are stained with TO-PRO-3 iodide (Invitrogen, Eugene, OR) in 0.1% Triton X-100 in PBS for 15 minutes.  Cover slips were washed extensively with PBS and water to reduce background.  Cover slips were mounted onto glass slides and examined with confocal microscopy (TCS SP8, Leica Microsystems, Concord, ON).  Fluorochromes were excited at 561 nm (DiI) and 633 nm (TO-PRO-3) and images were collected with a 62x oil-immersion objective lens. 4.3 Results 4.3.1 LNP siRNA systems containing DOPE and high cationic lipid contents are not effective gene silencing agents Previous work from this laboratory has shown that LNP siRNA systems composed of DLinKC2-DMA/DSPC/Cholesterol/PEG-c-DMA/DiI (40/11.5/47.3/1/0.2; mol %) exhibit excellent (> 90%) trapping efficiencies but if the proportion of cationic lipid is increased to 80% (reducing the cholesterol content in proportion) to achieve the lipid composition DLinKC2-DMA/DOPE/Cholesterol/PEG-c-DMA (80/11.5/7.5/1/; mol %) the encapsulation efficiency decreases dramatically.  However, substitution of DOPE for DSPC substantially improves siRNA encapsulation efficiencies.  It is of interest to determine whether such systems exhibit improved transfection and gene silencing capabilities as compared to the systems containing 40 114  mol% cationic lipid.  The in vitro gene silencing potency of LNP systems composed of DLinKC2-DMA/DSPC/Cholesterol/PEG-c-DMA/DiI (40/11.5/47.3/1/0.2; % mol), DLinKC2-DMA/DOPE/Cholesterol/PEG-c-DMA/DiI (40/11.5/47.3/1/0.2; % mol) and DLinKC2-DMA/DOPE/Cholesterol/PEG-c-DMA/DiI (80/11.5/7.5/1/0.2; % mol) were therefore compared using an in vitro assay.  Androgen-sensitive human prostate adenocarcinoma (LNCaP) cells were treated continuously for 48 hours with LNP containing either siRNA against the androgen receptor (AR) or control (FVII) siRNA.  Surprisingly, formulations containing 40% DLinKC2-DMA and the DOPE in place of DSPC performed worse than the formulation containing DSPC (Figure 4.1).  Further, LNP containing 80% DLinKC2-DMA and DOPE were no more effective than the systems containing only 40% DLinKC2-DMA (Figure 4.1).  These results are unexpected, as replacement of DSPC by DOPE and increasing cationic lipid content would be expected to increase the endosomolytic properties of the LNP. These results could be accounted for if the uptake levels of LNP systems containing DOPE were substantially less than for systems containing DSPC. Cellular uptake of the LNP systems was therefore examined by monitoring the intracellular fluorescence of the lipid label DiI using confocal microscopy.  Uptake was measured as fluorescence intensity per cell normalized to that of the untreated control.  Confocal images were taken after 24 hours of continuous incubation.  While the decreased efficacy of the 40% DLinKC2-DMA, DOPE formulation can be explained by the decrease in uptake, the results did not explain the decreased gene silencing activity of the 80% DLinKC2-DMA system as it clearly showed the highest uptake among the three formulations tested (Figure 4.2).  The increased uptake of 80% DLinKC2-DMA can be attributed 115  to the enhanced attraction to the negatively charged cell surface with more cationic charge on the LNP.  The question remains as to why this does not translate into better gene silencing activity.    Figure 4.1  Comparison of the knockdown of the androgen receptor (AR) in LNCap cells using LNP composed with 80% cationic lipid and DOPE SiRNA against the androgen receptor (siAR) was encapsulated with DLin-KC2-DMA/DSPC/cholesterol/PEG-c-DMA/DiI (40/11.5/47.3/1/0.2 mol%), DLin-KC2-DMA/DOPE/cholesterol/PEG-c-DMA/DiI (40/11.5/47.3/1/0.2 mol%) and DLin-KC2-DMA/DOPE/cholesterol/PEG-c-DMA/DiI (80/11.5/7.3/1/0.2 mol %) at a siRNA/lipid charge ratio of 0.25.  SiRNA against Factor VII was encapsulated with the same lipid formulations as negative controls.  Cells were treated continuously for 48 hours and protein expression was analyzed by immunoblotting.    116    Figure 4.2  Influence of lipid content on the cellular uptake of DLin-KC2-DMA siRNA LNP (A) Untreated cells were imaged as a negative control.  LNCap cells were treated with AR siRNA encapsulated with (B) DLin-KC2-DMA/DSPC/cholesterol/PEG-c-DMA/DiI (40/11.5/47.3/1/0.2 mol%), (C) DLin-KC2-DMA/DOPE/cholesterol/PEG-c-DMA/DiI (40/11.5/47.3/1/0.2 mol %) or (D)  DLin-KC2-DMA/DOPE/cholesterol/PEG-c-DMA/DiI (80/11.5/7.3/1/0.2 mol %) continuously for 24 hours at a lipid dose of 30 μg/mL.  Nuclei were stained with DRAQ-5 shown in blue and the LNP marker DiI is shown in green.  (E) Cellular uptake was assessed using Leica Application Suite Advance Fluorescence software (Leica Microsystems, Concord, ON) by quantifying DiI fluorescence per cell.  All fluorescent intensities were normalized to that of the untreated control.   117  4.3.2 Design of a new class of helper lipids  The above experiments indicate that LNP-siRNA systems containing DOPE are inferior in mediating gene knockdown in vitro compared to the same system containing the bilayer-stabilizing lipid DSPC despite the non-bilayer preferences of DOPE.  In an attempt to develop an improved “helper” lipid a new class of lipids was designed.  The design principle was focused on the ability to induce non-bilayer phases in the bilayer-forming lipid phosphatidylcholine (PC), which accounts for greater than 50% of phospholipids found in cellular membranes (van Meer, 2005).  Zwitterionic phospholipids such as PC have a negatively charged phosphate close to the glycerol backbone, while the positively charged choline is located at the terminus of the headgroup (Figure 4.3B).  It was thought that "reverse-headgroup" lipids that have a reverse charge orientation in the headgroup region as compared to zwitterionic lipids such as PC (Figure 4.3B) might cause the lipid to combine with zwitterionic lipids in such a way that the positive and negative charges in the headgroups are apposed, leading to charge neutralization.  This charge neutralization should then decrease the combined cross-sectional area of the headgroup region, leading to the formation of an ion pair that adopts a cone-shaped compatible with non-bilayer phases such as the hexagonal HII phase (Figure 4.3A).  This design is based on the proposed mechanism of action for cationic lipids, which has been shown to form non-bilayer phases upon addition to anionic lipids (Hafez et al., 2001; Semple et al., 2010).  The prototypic lipid, dioleoyl-four amino butyric acid (DOFAB), contains a terminal carboxylic acid group and a secondary amine close to the acyl chain region (Figure 4.3B).  It is expected that the carboxylic acid and secondary amine would have an apparent pKa of 4 and 8 respectively, which are typical for these functional groups attached onto a lipid moiety.  This 118  allows DOFAB to be zwitterionic at physiological pH and have the maximum ability to bind to the zwitterionic headgroup of PC and induces non-bilayer phases.   Figure 4.3  Proposed mechanism of the membrane disruptive effects of reverse-headgroup lipids (A) Reverse-headgroup lipids have an opposite charge orientation in the headgroup region as compared to zwitterionic phospholipids such as phosphatidylcholine.  On their own, both phosphatidylcholine and reverse-headgroup lipids have a cylindrical molecular structure and supports formation of a stable lipid bilayer.  When combined, electrostatic charge interaction in the headgroup region causes the cross-sectional area of the combined headgroup to decrease.   The resulting ion pair adopts a "cone"-shaped molecular structure which promotes the formation of non-bilayer phases and therefore disrupts the bilayer membrane.  (B) Structure of the reverse-headgroup lipid DOFAB and the phospholipid DOPC at pH 7, indicated are the locations of the charged moieties in the headgroup region.  119  4.3.3 DOFAB is zwitterionic at physiological pH The pKa values of DOFAB were measured by titrating the fluorescence of 2-(p-toluidino)-6-napthalene sulfonic acid (TNS) from pH 2 to 11.  The anionic TNS molecule fluoresces upon association with positively charge membrane and is a probe frequently used for examining surface charges on membranes (Jayaraman et al., 2012).  Pure DOFAB was hydrated with 25 mM sodium acetate at pH 4.0 leading to a milky dispersion.  The carboxylic acid moiety in the headgroup is almost certainly protonated at this pH, giving an overall positive charge in the headgroup region, preventing aggregation of the liposomes.  The titration curve of DOFAB showed two inflection points at 4.7 and 9.1, representing the pKa values of the carboxylic acid and secondary amine moieties, respectively (Figure 4.4).  This indicates that DOFAB is zwitterionic and net neutral at physiological pH values and becomes progressively more positive charged as the pH is lowered or more negatively charged if the pH is raised.     120    Figure 4.4  Determination of the pKa of DOFAB using the TNS Assay Aqueous suspensions of the DOFAB lipids were added to buffer containing 1 μM TNS, 10 mM HEPES, 10 mM MES, 10 mM ammonium acetate and 140 mM NaCl with the pH ranging from 2.0 to 10.5.  Fluorescence was monitored using an excitation wavelength of 380 nm and an emission wavelength of 475 nm.  Fluorescence intensities were normalized to the maximum intensity.  TNS titration was conducted in duplicates with the reported pKa being the average of duplicate measurements.   121  4.3.4 DOFAB exhibits strong bilayer-destabilizing capabilities in mixture of with DOPC and cholesterol. The ability of DOFAB to induce non-bilayer structures in mixtures of DOPC and cholesterol was investigated using 31P NMR.  Cholesterol was included in the mixture as it is the major sterol in cellular membranes (van Meer et al., 2008) and is a major component of the most potent siRNA-LNP systems.  An equimolar mixture of DOFAB/DOPC/cholesterol (1:1:1, mol/mol) was hydrated in 100 mM HEPES, 50 mM NaCl (pH 7.0) and the polymorphic phase preferences monitored using 31P NMR.  As shown in Figure 4.5, the DOFAB/DOPC/cholesterol system at 25°C exhibited 31P NMR spectra that have the characteristic asymmetric lineshape of phospholipids arranged in a bilayer conformation with a high field peak and a low field shoulder (Cullis and de Kruijff, 1979; McLaughlin et al., 1975). The superimposed isotropic lineshape indicates the presence of non-bilayer phases, such as inverted micellar and cubic phases (Cullis and de Kruijff, 1976, 1979; McLaughlin et al., 1975).  Increasing the temperature of the lipid mixture to 37oC produced a dramatic change in the 31P NMR spectrum, which exhibited a low-field peak and high-field shoulder that is diagnostic for the presence of hexagonal HII phase (Figure 4.5).  At 60oC the mixture completely adopts the hexagonal HII phase.  This result suggests DOFAB is a potent bilayer-destabilizing lipid, capable of inducing hexagonal HII phase when mixed with equimolar levels of DOPC and cholesterol.    122   Figure 4.5  DOFAB exhibits strong bilayer-destabilizing abilities in mixtures with DOPC and cholesterol Proton-decoupled 31P NMR spectra of an equimolar mixture of DOFAB/DOPC/Cholesterol at pH 7.  At 25oC, the mixture is mainly in the bilayer conformation with a significant isotropic signal indicating the presence of non-bilayer phases such as inverted micelles.  At the physiological temperature of 37oC, the majority of lipids is switched over to the inverted hexagonal HII phase with a portion of it adopting other non-bilayer phases that generates an isotropic signal.  By 60oC, all lipids adopt the hexagonal HII phase.  123  4.3.5 Incorporation of DOFAB into lipid nanoparticles does not affect encapsulation efficiencies The ionizable moieties in the headgroup of DOFAB serve two purposes: first, as described above, it potentially allows DOFAB to interact with the headgroups of zwitterionic phospholipids such as PC to induce non-bilayer structures that may facilitate the intracellular release of siRNA. Second, based on the titration curve of DOFAB (Figure 4.4), the lipid headgroup will exhibit a positive charge at the formulation pH of 4.0, and thus would be expected to maintain or enhance siRNA encapsulation properties.  The formulation properties of LNP-siRNA prepared from DLinKC2-DMA/DSPC/Cholesterol/PEG-c-DMA (40/11.5/47.5/1 mol%) were compared to formulations containing DLinKC2-DMA/DOFAB/Cholesterol/PEG-c-DMA (40/11.5/47.5/1 mol%).  Both formulations were prepared at a constant siRNA/lipid charge ratio of 0.25.  The replacement of DSPC with DOFAB in the formulation lead to slight drop in the encapsulation efficiency from 89% to 83% (Figure 4.6).  The size of the resulting LNP containing DOFAB is 54.8 nm, approximately 10 nm larger than the DSPC-containing particle (45.5 nm).   In order to determine whether LNP systems could exhibit good encapsulation properties at higher DOFAB contents the amount of DOFAB was increased from 11.5% to 59% (with cholesterol decreased proportionally) and the encapsulation properties characterized. DLinKC-DMA and PEG-c-DMA contents were held constant at 40% and 1% respectively.  All formulations were prepared at a siRNA/lipid charge ratio of 0.25.  As shown in Figure 4.6, the encapsulation efficiency of LNP prepared with DOFAB remained over 80% for all DOFAB concentrations tested, ranging from 11.5% to 59%.  The size of the LNP as measured by dynamic light scattering (number mode) ranged from 51.3 nm to 61.6 nm for all the DOFAB 124  concentrations tested and within the size limit for cellular uptake (Figure 4.6, Huang et al., 2010).  This demonstrated that the total fusogenic lipid content of LNP, which is sum of the cationic lipid DLinKC2-DMA and DOFAB content, can be as high as 99% without compromising encapsulation efficiency.    125    Figure 4.6  siRNA encapsulation efficiency of LNP siRNA systems containing various amounts of DOFAB LNP systems without DOFAB were prepared with the lipid composition DLinKC2-DMA/DSPC/Chol/PEG-lipid (40/11.5/47.5/1; mol%).  DOFAB was included in the formulation by first replacing the DSPC with DOFAB and then proportionally replacing the cholesterol as more DOFAB was added.  Molar percentage of fusogenic lipid is the total percentage of DLinKC2-DMA and DOFAB in the formulation.  Measurements of encapsulation efficiency were made using the RiboGreen Assay.  Particle size was reported as the number-weighted mean measured using the Malvern ZetaSizer.  126  4.3.6 LNP formulated with DOFAB have enhanced fusion capability It is of interest to determine whether replacing the bilayer-stabilizing DSPC with DOFAB result in an  increase in the fusogenicity of the LNP.  The FRET lipid mixing assay (Hafez et al., 2000; Struck et al., 1981) was used to compare the fusogenicity of LNP formulations containing DSPC, DOPE and DOFAB. DOPE is a well documented helper lipid used in numerous gene delivery formulations (Farhood et al., 1995; Koltover et al., 1998).  Three different formulations of LNP were prepared containing DLinKC2-DMA/helper lipid/cholesterol/PEG-c-DMA (40/11.5/47.5/1 mol %), with the helper lipid being either DSPC, DOPE or DOFAB.  The target vesicle used this assay is an anionic vesicle composed of DOPS/NBD-PE/Rh-PE (98/1/1 mol %), in which the NBD-PE and Rh-PE function as FRET pairs.  At pH 5.5, over 90% of DLinKC2-DMA is positively charged (Leung et al., 2012).  Fusion of the cationic LNP with the anionic DOPS vesicles result in dilution of lipids in the DOPS vesicles, increasing the distance between the NBD-PE and Rh-PE, and decreasing the extent of fluorescent energy transfer between the two lipids.  The extent of fusion between the LNP and the target vesicle is observed as a recovery of NBD-PE fluorescence.  As shown in Figure 4.7, replacing the 11.5% of DSPC in the formulation with either DOPE or DOFAB greatly enhanced the ability of the LNP to fuse with the anionic target vesicle.  Increasing the DOFAB content in the formulation to 41.5% by decreasing the cholesterol content proportionally did not lead to any further increase in fusion (data not shown).  This suggests that cholesterol may play a role in facilitating fusion of LNP as cholesterol is known is facilitate the lamellar-hexagonal transition in mixtures of bilayer and non-bilayer preferring lipids (Cullis et al., 1978b; Tilcock et al., 1982).  These results support the hypothesis that DOFAB is a fusogenic, zwitterionic lipid and its incorporation in the LNP 127  formulation is may enhance its ability to disrupt the endosome, promoting intracellular release of the siRNA cargo.  Figure 4.7  Incorporation of DOFAB into LNP systems increases the fusion capability LNP systems containing 11.5% DSPC (blue solid line), DOPE (red solid line) and DOFAB (green solid line) were prepared while keeping DLinKC2-DMA, cholesterol and PEG-lipid contents constant at 40%, 47.5% and 1%, respectively.  The lipid mixing assay was performed at pH 5.5 to ensure that essentially all external DLinKC2-DMA was positively charged. The reaction was initiated by injecting the LNP (at t=30 s) into a stirred cuvette containing the anionic DOPS/NBD-PE/Rh-PE (98/1/1 mol%) vesicles.   128  4.3.7 LNP containing DOFAB are not as effective in mediating gene silencing in vitro It is expected the incorporation of DOFAB in LNP-siRNA systems would increase the efficacy of gene silencing by enhancing the system’s ability to disrupt the endosome.  We therefore compared the in vitro gene silencing efficacy between LNP systems using either DSPC, DOPE or DOFAB as helper lipids.  Formulation containing DOPE was included to serve as a comparison.  SiRNA-LNP systems were formulated with the lipid composition DLinKC2-DMA/helper lipid/cholesterol/PEG-c-DMA (40/11.5/47.5/1 mol %), at a siRNA/lipid charge ratio of 0.25.  Androgen-sensitive human prostate adenocarcinoma (LNCap) cells were treated continuously for 48 hours with LNP encapsulated with either siRNA against the androgen receptor (AR) or control siRNA against hepatocyte Factor VII.  As shown earlier, formulations containing DOPE are not as effective in mediating gene silencing in vitro as compared to formulation containing DSPC (Figure 4.8).  The gene silencing efficacy of LNP containing DOFAB are also not as effective as formulations contain DSPC and its efficacy is similar to that of the DOPE formulation  (Figure 4.8). Cellular uptake of the LNP systems was examined by monitoring the intracellular fluorescence of the lipid label DiI with confocal microscopy.  Uptake was measured as fluorescence intensity per cell normalized to that of the untreated control.  Confocal images taken after 24 hours of continuous treatment correlated with the gene silencing data as LNP systems containing DSPC clearly showed higher cellular uptake than systems containing either DOPE and DOFAB (Figure 4.9).  This suggests that the observed weaker ability of the DOFAB formulation in mediating gene silencing was partially contributed by decreased cellular uptake of the DOFAB-containing LNP as compared to LNP containing DSPC.   129    Figure 4.8  Knockdown of the androgen receptor (AR) in LNCaP cells with either DOPE or DOFAB as the helper lipid SiRNA against the androgen receptor was encapsulated with DLinKC2-DMA/helper lipid/Cholesterol/PEG-c-DMA 40/11.5/47.5/1; mol/mol), with the helper lipid being DSPC, DOPE or DOFAB.  Cells were treated continuously for 48 hours and protein expression was analyzed by immunoblotting.     130   Figure 4.9  Influence of the helper lipid on the cellular uptake of DLin-KC2-DMA siRNA LNP LNCap cells were treated with AR siRNA encapsulated with DLin-KC2-DMA/helper lipid/cholesterol/PEG-c-DMA/DiI (40/11.5/47.3/1/0.2 mol%) continuously for 24 hours at a lipid dose of 30 μg/mL. (A) Untreated cells were imaged as a negative control.  Cells treated with LNP containing the helper lipid (B) DSPC, (C) DOPE, or (D) DOFAB.  Nuclei were stained with DRAQ-5 shown in blue and DiI fluorescence from LNP is shown in green.  (E) Cellular uptake was assessed using Leica Application Suite Advance Fluorescence software (Leica Microsystems, Concord, ON) by quantifying DiI fluorescence per cell.  All fluorescent intensities were normalized to that of the untreated control   131  4.4. Discussion The results presented in this study raised some intriguing questions regarding the role of cationic lipids and "helper" lipids in mediating intracellular release of encapsulated siRNA.  Two points concern the reduced efficacy observed for LNP containing DOPE rather than DSPC and the fact that higher levels of cationic lipid do not lead to enhanced potency.  The results of Figure 4.8 clearly show that substituting DOPE for DSPC in siRNA-LNP systems dramatically reduces the gene silencing potency of systems containing 40 mol% cationic lipid. As indicated in Figure 4.9 the substitution of DOPE does reduce LNP uptake levels by approximately a factor of two, however the gene silencing potency is reduced by at least a factor of 5  (Figure 4.8).  This observation challenges currently accepted dogma regarding the role of DOPE in transfection reagents, where it is commonly thought that the presence of DOPE, which favours non-bilayer membrane lytic conformations, contributes to destabilization of the endosomal membrane.  It is possible that substitution of DOPE for DSPC affects the fate of LNP following uptake.  For example, a recent study on the role of endocytic recycling have on efficiency of intracellular siRNA delivery shows that approximately 70% of internalized LNP siRNA eventually exits the cell through a mechanism dependent on the protein Niemann-Pick Type C-1 (NPC1) (Sahay et al., 2013).  It is possible that LNP containing DOPE are subjected to different recycling pathways, leading to increased exocytosis of LNP.   The results indicating that increased levels of cationic lipid do not lead to enhanced potency are also surprising.  As shown in Figure 4.2, increasing the amount of DLinKC2-DMA from 40 mol% to 80 mol% dramatically improves uptake into the LNCaP cells, but does not substantially improve gene silencing potency (Figure 4.1).  For example, the cellular uptake of the 80 mol% DLinKC2-DMA were more than 15 folds greater than the containing 40 mol% DLinKC2-DMA, 132  however, the dose required to achieve close to complete knockdown was similar to the formulation containing just 40% DLinKC2-DMA.  When compared to the 40% DLinKC2-DMA containing DSPC, the cellular uptake was about 7 folds higher for the 80% cationic lipid formulation but the dose required to achieve complete knockdown is approximately 5 times higher.  This argues against the postulated role of ionizable cationic lipids that are proposed to combine with endogenous anionic lipids to form membrane lytic ion pairs (Hafez and Cullis, 2001; Semple et al., 2010) as it would be logical to suppose that more cationic lipid should lead to greater destabilization. It is possible that the amount of cationic lipid in a single LNP is sufficient to combine with all the anionic lipid present in a given endosome, in which case no advantage would be expected for higher cationic lipid contents.  Alternatively, it is possible that recycling pathways to export the LNP from the cell may be enhanced or that the siRNA in the LNP containing 80% DLinKC2-DMA remains bound to the cationic lipids and is not bioavailable.  A recent study with LNP-siRNA systems suggests that approximately only 1 to 2% of the total internalized siRNA is released into the cytosol (Gilleron et al., 2013).  In this context, a slight increase in the association between the siRNA and the cationic lipid would further decrease the amount of siRNA available in the cytosol and significantly reducing the efficacy of gene silencing. These and other possibilities are under active investigation.  The final area of discussion concerns the novel “reversed headgroup” lipid DOFAB.  It is clear from the results presented that DOFAB offers no substantial improvements over DOPE or DSPC as a helper lipid.  The most dramatic improvements in potency of LNP siRNA systems has come from optimizing cationic lipids with much less emphasis on the development of neutral "helper" lipids (Akinc et al., 2009).  A successful "helper" lipid should have (1) the ability to promote the formation of non-bilayer phases that would mediate endosomal disruption; (2) must 133  not affect the formulation properties of the LNP such as encapsulation efficiency, particle size, and stability; (3) must be charge neutral at physiological pH to ensure prolonged circulation lifetime; and lastly (4) must not affect the cellular uptake of the LNP.  It is clear that DOFAB satisfied all of the above requirements except for reducing cellular uptake of LNP.  It has been shown that the uptake of nanoparticles is dependent on the protein corona that are adsorbed onto the surface of the nanoparticle (Lesniak et al., 2012).  For example, it was demonstrated that the uptake LNP into hepatocytes is dependent on the binding of apolipoprotein E (apoE E) onto the surface of the LNP (Akinc et al., 2010).  It is possible that the replacement of DSPC with DOFAB has altered the surface properties of the LNP, resulting in lower adsorption of proteins that facilitate uptake.   The results presented in this Chapter pose more questions than answers.  First, although it is commonly believed that “non-bilayer” lipids such as DOPE assist in transfection processes, the results presented here show that DOPE inhibits LNP siRNA potency as compared to the bilayer lipid DSPC.  Second, while it is thought that cationic lipids assist in the intracellular release of siRNA, increasing the cationic lipid in the formulation does not necessarily translate into improved gene silencing.  Finally, the design and synthesis of a lipid, DOFAB, that may be suggested to improve on the properties of DSPC and DOPE by virtue of an “inverted” headgroup structure has not led to enhanced potency.  These studies clearly suggest new directions to achieve the next generation of LNP siRNA systems with improved potency properties.  134  Chapter 5: Summarizing Discussion and Future Directions The results presented in the thesis clearly suggests that siRNA-LNP systems have an internal solid hydrophobic core comprised of lipids and inverted micelles of cationic lipids containing the encapsulated siRNA.  This raises two questions worthy of further discussion.  The first question relates to the mechanism behind the formation of such solid core structures and factors that dictates the encapsulation of nucleic acid polymers.  The second area of discussion concerns with applying these structural information in constructing a class of LNP carrier systems with novel applications.     Lipoproteins are natural solid core systems used for the transport of lipids in the body.  Three-dimensional reconstruction of LDL from cryo-TEM revealed a solid interior composed of hydrophobic lipids, such as cholesterol ester and triglyceride, surrounded by a layer of more polar phospholipid (Orlova et al., 1999).  The appearance of lipoproteins is very similar to that of siRNA-LNP as observed with cryo-TEM and may offer insights to the mechanism behind the formation of the solid core interior of siRNA-LNP systems.  The mechanism leading to the solid core interior of LNP systems is a function of the polarity of lipid components and their solubility in ethanol.  As the ethanolic stream of lipids combines with the acidic buffer, the positively charged cationic lipid forms an inverted micelle around the siRNA due to electrostatic interaction (Belliveau et al., 2012).  This inverted micelle of cationic lipid and siRNA will condense out of solution first as it is the most hydrophobic entity in the mixture with the hydrophobic acyl chains of the lipid exposed to the aqueous environment (Figure 5.1).  The near complete encapsulation of siRNA supports the hypothesis that nucleates of siRNA with cationic lipid are the first entity to condense out of solution and form a nucleation center for the rest of lipids as the polarity increases  (Belliveau et al., 2012).  The most hydrophilic lipid of the 135  mixture, PEG-c-DMA, would be expected to be the last lipid to come of solution and coat the surface of the particle.  According to this model, the size of the particle should be a function of the ratio of surface lipids to core lipids as it dictates the surface-to-volume ratio of the resulting particle.  This has been shown to be case as increasing the amount of the surface lipid, PEG-c-DMA, from 1% to 5% (mol %) corresponds to a decrease of particle diameter from approximately 50 nm to 20 nm (Belliveau et al., 2012).  This model in turn provides an explanation for the poor nucleic acid encapsulation properties of formulations containing higher contents of cationic lipid.  The model proposed above clearly points to the role of inverted micelle formation during the encapsulation of siRNA in LNP systems and this can be used to explain the siRNA encapsulation efficiencies of systems composed high cationic lipid content.  As illustrated in Chapter 3, the encapsulation of siRNA decreases with increasing cationic lipid content in the formulation, while the incorporation of the cone-shaped lipid, DOPE, in the formulation rescues the encapsulation efficiency.   It is important to keep in mind that formulations prepared with 40% DLin-KC2-DMA contains 47.5% of cholesterol.  Cholesterol has the ability to facilitate the formation of hexagonal HII phase in mixtures containing cone-shaped lipids (Cullis et al., 1978b).  In addition, cholesterol has been shown to promote hexagonal HII phase formation in mixtures containing cationic and anionic lipids (Hafez et al., 2001).  During the encapsulation of siRNA, cholesterol serves to facilitate the formation of an inverted micelle of cationic lipids (Figure 5.2).  In formulations containing 80% DLin-KC2-DMA, the cholesterol content is simultaneously decreased to 7.5%.  This leads to an decreased ability to form an inverted micelle around the siRNA and results in decreased encapsulation efficiency.  In support of this argument, replacing the bilayer-forming 136  DSPC with the cone-shaped lipid DOPE in the formulation increases the siRNA encapsulation efficiency by promoting the formation of inverted micelles around the siRNA (Figure 5.2). The proposed structure of LNP formulated by microfluidic mixing also makes it an ideal candidate for the encapsulation of hydrophobic drugs, such as taxanes.  Clinically viable formulations of taxanes have been hampered by its low water solubility (Zhigaltsev et al., 2010).  Current formulations of taxanes, such as paclitaxel (Taxol) and docetaxel (Taxotere), involves the use of surfactants to increase the solubility of the drugs, however, toxicity associated with these surfactant is a major concern (van Zuylen et al., 2001).  Hydrophobic drugs, such as docetaxel, can be conjugated with a lipid moiety and be encapsulated in the hydrophobic interior of LNP in much the same way triolein can be assembled into solid-core LNP by using microfluidic mixing (Zhigaltsev et al, unpublished results).  The lipid-conjugated docetaxel can be dissolved in ethanol along with a surface lipid such as PC and the ethanolic stream of lipid/docetaxel can be combined with an aqueous phase using microfluidic mixing.  As the concentration of ethanol is diluted below the solubility limit of the docetaxel-lipid conjugate, it condenses out of solution which will then be coated by layer of PC as it reaches its solubility limit.  The encapsulated docetaxel can then be liberated from LNP and reverted back to its active form by the action of lipases in the body.  This offers several major advantages to previous formulations of hydrophobic drugs, such as Taxol and Taxotere.  First, lipid carriers are relatively non-toxic compared to surfactant and encapsulation of the drug inside of LNP carrier system minimizes toxicity due to systemic exposure to the free drug.  This is demonstrated by the fact that the maximum tolerated dose of docetaxel encapsulated in liposome is more than three times higher than that of Taxotere (Zhigaltsev et al., 2010).  The second advantaged relates to the size of the resulting particle produced by microfluidic mixing.  It is theoretically possible 137  to construct solid-core systems of approximately 20 nm in diameter by microfluidics, similar to limit size system comprised of POPC and triolein.  It is expected that the smaller particle would have better tumour penetration as compared to larger vesicular systems (80 - 100 nm in diameter) resulting in greater potency (Huo et al., 2013).  Lastly, in comparing particles of similar size, solid-core particles offer higher drug loading capacity compared to vesicular systems as the interior of the particle can theoretically be entirely comprised of the condensed drug (Zhigaltsev, personal communication).  Gold nanoparticle-conjugated siRNA was used to validate the solid-core model of siRNA-LNP in thesis, however, gold-containing LNP also have potential for therapeutic use.  Gold is a good absorber of x-ray radiation and gold nanoparticles has been shown to provide a therapeutic advantage in tumour-bearing mice after radiotherapy (Hainfeld et al., 2004).  However, the use gold nanoparticles in radiation medicine suffers from poor accumulation of gold nanoparticles in tumours and rapid excretion through the kidneys following intravenous injection (Hainfeld et al., 2004).  One of the methods to improve on the biodistribution and cellular uptake of gold nanoparticles is to couple the particles with lipid-based delivery systems such as lipid nanoparticles (Chithrani et al., 2010).  Formulations of gold nanoparticles with liposome have been prepared by several different methods: (1) reverse-phase evaporation with gold nanoparticles suspended in the organic phase (Hong et al., 1983; Paasonen et al., 2007); (2) physical adsorption of gold nanoparticles onto the surface of liposomes (Kojima et al., 2008); (3) post-insertion of gold nanoparticle-conjugated phospholipids onto preformed liposomes (Chithrani et al., 2010) and (4) embedding gold nanoparticle into dry lipid film followed by hydration (Park et al., 2006).  A microfluidic mixing method for encapsulating gold nanoparticle within LNP was developed by utilizing lipids used in formulations of siRNA-LNP and 138  commercially available negatively-charged gold nanoparticles.  Microfluidic mixing offers the most straight-forward method for the encapsulation of gold nanoparticles as it only involve rapid mixing of lipid dissolved in ethanol with gold nanoparticles suspended in an acidic aqueous buffer.  Cationic lipids would encapsulated the negatively-charged gold nanoparticles in a manner similar to the encapsulation of siRNA.  Cryo-TEM of the resulting particle clearly shows the gold nanoparticles imbedded within the solid core interior of the LNP (Figure 5.3).  It is expected that LNP carrier system would improve the accumulation of the gold nanoparticle to tumour sites by the EPR effect and facilitate cellular uptake of the gold nanoparticle through the endocytic pathway.  This technology would have dramatic impact in improving the therapeutic potential of radiotherapy. In addition to its use as X-ray absorber in radiation medicine, metallic nanoparticles such as gold nanoparticles can also be used for generating thermo-sensitive LNP.  Metallic nanoparticle embedded in a lipid bilayer can be excited by light radiation or magnetic field to heat the bilayer membrane to its gel-to-liquid crystal phase transition temperature (TM) where the permeability of the membrane is the greatest.  Liposomes with gold nanoparticles embedded in its membrane have shown to release its aqueous content following irradiation with ultra-violet light (Paasonen et al., 2007).  This has tremendous application in the delivery of small molecule drugs as the release of drugs can now be controlled by localized exposure to radiation.  Thermo-sensitive LNP can be generated using the above mention scheme for the encapsulation of hydrophobic drugs with the addition of cationic lipids to facilitate the encapsulation of the negatively-charged gold nanoparticle.  Ultra-violet light can be used to excite the gold nanoparticle embedded inside of the LNP.  The heat that is generated could cause the LNP disassemble, releasing its drug payload.  In addition to gold, other metallic nanoparticles such as iron oxide were utilized in 139  generating thermo-sensitive liposomes (Amstad et al., 2011).  In the case of iron oxide, an alternating magnetic field can be used to excite iron oxide nanoparticles embedded in the lipid membrane of the vesicle to induce leakage of aqueous content (Amstad et al., 2011).  The iron oxide nanoparticle can be modified with various functional groups such as carboxylic acids or silica that render the surface negatively-charged to facilitate encapsulation in LNP containing cationic lipids (Laurent et al., 2008).  Drug-loaded LNP containing metallic nanoparticles are expected to only release its cargo in the tumour upon localized irradiation and would provide a therapeutic advantage by reducing systemic exposure to the encapsulated drug due to passive leakage.   140   Figure 5.1  Assembly of siRNA-LNP by microfluidic mixing Lipids reach its solubility limit and start to come out of solution as the ethanol in the lipid stream is being diluted by the aqueous buffer.  The acidic sodium acetate buffer protonates the ionizable cationic lipids (blue) and forms an inverted micelle around the siRNA (red) due to electrostatic interaction.  As the polarity of the solvent increases further, inverted micelles begin to aggregate and the surface is coated by a layer of PEG-lipids (purple) as it reach its solubility limit thereby forming the siRNA-LNP.    141   Figure 5.2  The conical lipid, DOPE, increases the propensity of the lipid mixture to form inverted micellar structures, leading to the encapsulation of siRNA. Encapsulation of siRNA (red) requires the lipids to adopt an inverted micelle around the siRNA.  Cholesterol (orange) facilitates the formation of an inverted micelle around the siRNA in formulations containing 50% cationic lipid (blue).  As the cholesterol is removed from the formulation to accommodate an increasing amount of cationic lipid (up to 80 mol%), the conical lipid DOPE is required to drive the formation the inverted micelle around the siRNA.    142   Figure 5.3  Cryo TEM of gold nanoparticles encapsulated in LNP by microfluidic mixing. Commercially available anionic gold nanoparticles (Pelco NanoXact Gold 5 nm, Ted Pella Inc., Redding, CA) were encapsulated in LNP prepared with DLinKC2-DMA/DSPC/Chol/PEG-lipid (40/11.5/47.5/1; % mol).  The four panels illustrates different structures of LNP within the same preparation of gold-LNP.        143  Bibliography Abraham, S.A., Waterhouse, D.N., Mayer, L.D., Cullis, P.R., Madden, T.D., and Bally, M.B. (2005). The liposomal formulation of doxorubicin. Methods Enzymol. 391, 71–97. Abrams, M.T., Koser, M.L., Seitzer, J., Williams, S.C., DiPietro, M.A., Wang, W., Shaw, A.W., Mao, X., Jadhav, V., Davide, J.P., et al. (2010). Evaluation of efficacy, biodistribution, and inflammation for a potent siRNA nanoparticle: effect of dexamethasone co-treatment. Mol. Ther. 18, 171–180. Akinc, A., Goldberg, M., Qin, J., Dorkin, J.R., Gamba-Vitalo, C., Maier, M., Jayaprakash, K.N., Jayaraman, M., Rajeev, K.G., Manoharan, M., et al. (2009). Development of lipidoid-siRNA formulations for systemic delivery to the liver. Mol. Ther. 17, 872–879. Akinc, A., Querbes, W., De, S., Qin, J., Frank-kamenetsky, M., Jayaprakash, K.N., Jayaraman, M., Rajeev, K.G., Cantley, W.L., Dorkin, J.R., et al. (2010). Targeted delivery of RNAi therapeutics with endogenous and exogenous ligand-based mechanisms. Mol. Ther. 18, 1357–1364. Allen, T.M., and Cullis, P.R. (2013). Liposomal drug delivery systems: from concept to clinical applications. Adv. Drug Deliv. Rev. 65, 36–48. De Almeida, R.F.M., Fedorov, A., and Prieto, M. (2003). Sphingomyelin/phosphatidylcholine/cholesterol phase diagram: boundaries and composition of lipid rafts. Biophys. J. 85, 2406–2416. Almena, M., and Mérida, I. (2011). Shaping up the membrane: diacylglycerol coordinates spatial orientation of signaling. Trends Biochem. Sci. 36, 593–603. Alwarawrah, M., Dai, J., and Huang, J. (2010). A molecular view of the cholesterol condensing effect in DOPC lipid bilayers. J. Phys. Chem. B 114, 7516–7523. Ameres, S.L., Martinez, J., and Schroeder, R. (2007). Molecular basis for target RNA recognition and cleavage by human RISC. Cell 130, 101–112. Amstad, E., Kohlbrecher, J., Müller, E., Schweizer, T., Textor, M., and Reimhult, E. (2011). Triggered release from liposomes through magnetic actuation of iron oxide nanoparticle containing membranes. Nano Lett. 11, 1664–1670. 144  Van Antwerpen, R., La Belle, M., Navratilova, E., and Krauss, R.M. (1999). Structural heterogeneity of apoB-containing serum lipoproteins visualized using cryo-electron microscopy. J. Lipid Res. 40, 1827–1836. Audouy, S.A.L., de Leij, L.F.M.H., Hoekstra, D., and Molema, G. (2002). In vivo characteristics of cationic liposomes as delivery vectors for gene therapy. Pharm. Res. 19, 1599–1605. Babst, M., Katzmann, D.J., Estepa-Sabal, E.J., Meerloo, T., and Emr, S.D. (2002). Escrt-III: an endosome-associated heterooligomeric protein complex required for mvb sorting. Dev. Cell 3, 271–282. Bailey, A.L., and Cullis, P.R. (1994). Modulation of membrane fusion by asymmetric transbilayer distributions of amino lipids. Biochemistry 33, 12573–12580. Bally, M.B., Harvie, P., Wong, F.M.P., Kong, S., Wasan, E.K., and Reimer, D.L. (1999). Biological barriers to cellular delivery of lipid-based DNA carriers. Adv. Drug Deliv. Rev. 38, 291–315. Bangham, A.D., and Horne, R.W. (1964). Negative staining of phospholipids and their structural modification by surface-active agents as observed in the electron microscope. J. Mol. Biol. 8, 660–668. Bangham, A.D., Standish, M.M., and Watkins, J.C. (1965a). Diffusion of univalent ions across the lamellae of swollen phospholipids. J. Mol. Biol. 13, 238–252. Bangham, A.D., Standish, M.M., and Weissmann, G. (1965b). The action of steroids and streptolysin S on the permeability of phospholipid structures to cations. J. Mol. Biol. 13, 253–259. Batzri, S., and Korn, E.D. (1973). Single bilayer liposomes prepared without sonication. Biochim. Biophys. Acta 298, 1015–1019. Belliveau, N.M., Huft, J., Lin, P.J., Chen, S., Leung, A.K., Leaver, T.J., Wild, A.W., Lee, J.B., Taylor, R.J., Tam, Y.K., et al. (2012). Microfluidic Synthesis of Highly Potent Limit-size Lipid Nanoparticles for In Vivo Delivery of siRNA. Mol. Ther. Nucleic Acids 1, e37. Benmerah, A., and Lamaze, C. (2007). Clathrin-coated pits: vive la différence? Traffic 8, 970–982. Bentz, J. (2000). Membrane fusion mediated by coiled coils: a hypothesis. Biophys. J. 78, 886–900. 145  Bernstein, E., Caudy, A.A., Hammond, S.M., and Hannon, G.J. (2001). Role for a bidentate ribonuclease in the initiation step of RNA interference. Nature 409, 363–366. Bitko, V., Musiyenko, A., Shulyayeva, O., and Barik, S. (2005). Inhibition of respiratory viruses by nasally administered siRNA. Nat. Med. 11, 50–55. Van Blokland, R., Van der Geest, N., Mol, J.N.M., and Kooter, J.M. (1994). Transgene-mediated suppression of-chalcone synthase expression in Petunia hybrida results from an increase in RNA turnover. Plant J. 6, 861–877. Blumenthal, R., Clague, M.J., Durell, S.R., and Epand, R.M. (2003). Membrane fusion. Chem. Rev. 103, 53–69. Bolotin, E.M., Cohen, R., Bar, L.K., Emanuel, N., Ninio, S., Danilo, D.L., and Barenholza, Y. (1994). Ammonium sulfate gradients for efficient and stable remote loading of amphipathic weak bases into liposomes and ligandoliposomes. J. Liposome Res. 4, 455–479. Bright, N.A., Reaves, B.J., Mullock, B.M., and Luzio, J.P. (1997). Dense core lysosomes can fuse with late endosomes and are re-formed from the resultant hybrid organelles. J. Cell Sci. 110 ( Pt 1, 2027–2040. Bright, N.A., Gratian, M.J., and Luzio, J.P. (2005). Endocytic delivery to lysosomes mediated by concurrent fusion and kissing events in living cells. Curr. Biol. 15, 360–365. Burger, K.N. (2000). Greasing membrane fusion and fission machineries. Traffic 1, 605–613. Casey, J.R., Grinstein, S., and Orlowski, J. (2010). Sensors and regulators of intracellular pH. Nat. Rev. Mol. Cell Biol. 11, 50–61. Charrois, G.J.R., and Allen, T.M. (2004). Drug release rate influences the pharmacokinetics, biodistribution, therapeutic activity, and toxicity of pegylated liposomal doxorubicin formulations in murine breast cancer. Biochim. Biophys. Acta 1663, 167–177. Chernomordik, L. (1996). Non-bilayer lipids and biological fusion intermediates. Chem. Phys. Lipids 81, 203–213. Chernomordik, L. V, and Kozlov, M.M. (2008). Mechanics of membrane fusion. Nat. Struct. Mol. Biol. 15, 675–683. 146  Chernomordik, L., Chanturiya, a, Green, J., and Zimmerberg, J. (1995). The hemifusion intermediate and its conversion to complete fusion: regulation by membrane composition. Biophys. J. 69, 922–929. Chithrani, D.B., Dunne, M., Stewart, J., Allen, C., and Jaffray, D.A. (2010). Cellular uptake and transport of gold nanoparticles incorporated in a liposomal carrier. Nanomedicine 6, 161–169. Choi, J.S., MacKay, J.A., and Szoka, F.C. (2003). Low-pH-sensitive PEG-stabilized plasmid-lipid nanoparticles: preparation and characterization. Bioconjug. Chem. 14, 420–429. Choung, S., Kim, Y.J., Kim, S., Park, H.-O., and Choi, Y.-C. (2006). Chemical modification of siRNAs to improve serum stability without loss of efficacy. Biochem. Biophys. Res. Commun. 342, 919–927. Connor, E.E., Mwamuka, J., Gole, A., Murphy, C.J., and Wyatt, M.D. (2005). Gold nanoparticles are taken up by human cells but do not cause acute cytotoxicity. Small 1, 325–327. Coskun, U., and Simons, K. (2011). Cell membranes: the lipid perspective. Structure 19, 1543–1548. Crawford, R., Dogdas, B., Keough, E., Haas, R.M., Wepukhulu, W., Krotzer, S., Burke, P. a, Sepp-Lorenzino, L., Bagchi, A., and Howell, B.J. (2011). Analysis of lipid nanoparticles by Cryo-EM for characterizing siRNA delivery vehicles. Int. J. Pharm. 403, 237–244. Cullis, P.R., and de Kruijff, B. (1976). 31P NMR studies of unsonicated aqueous dispersions of neutral and acidic phospholipids. Effects of phase transitions, p2H and divalent cations on the motion in the phosphate region of the polar headgroup. Biochim. Biophys. Acta 436, 523–540. Cullis, P.R., and de Kruijff, B. (1978a). The polymorphic phase behaviour of phosphatidylethanolamines of natural and synthetic origin. A 31P NMR study. Biochim. Biophys. Acta 513, 31–42. Cullis, P.R., and de Kruijff, B. (1978b). Polymorphic phase behaviour of lipid mixtures as detected by 31P NMR - evidence that cholesterol may destabilize bilayer structure in membrane systems containing phosphatidylethanolamine. Biochim. Biophys. Acta 507, 207–218. Cullis, P.R., and de Kruijff, B. (1979). Lipid polymorphism and the functional roles of lipids in biological membranes. Biochim. Biophys. Acta 559, 399–420. 147  Cullis, P.R., Verkleij, A.J., and Ververgaert, P.H.J.T. (1978a). Polymorphic phase behaviour of cardiolipin as detected by 31P NMR and freeze-fracture techniques. Effects of calcium, dibucaine and chlorpromazine. Biochim. Biophys. Acta 513, 11–20. Cullis, P.R., van Dijck, P.W., de Kruijff, B., and de Gier, J. (1978b). Effects of cholesterol on the properties of equimolar mixtures of synthetic phosphatidylethanolamine and phosphatidylcholine. A 31P NMR and differential scanning calorimetry study. Biochim. Biophys. Acta 513, 21–30. Cullis, P.R., Hope, M.J., and Tilcock, C.P. (1986). Lipid polymorphism and the roles of lipids in membranes. Chem. Phys. Lipids 40, 127–144. Davis, C.G., Goldstein, J.L., Sudhof, T.C., Anderson, R.G.W., Russell, D.W., and Brown, M.S. (1987). Acid-dependent ligand dissociation and recycling of LDL receptor mediated by growth growth factor homology region. Nature 326, 760–765. Deamer, D.W., and Oro, J. (1980). Role of lipids in prebiotic structures. BioSystems 12, 167–175. DiVerdi, J.A., Opella, S.J., Ma, R.-I., and Kallenbach, N.R. (1981). 31P NMR of DNA in eukaryotic chromosomal complexes. Biochem. Biophys. Res. Commun. 102, 885–890. Doherty, G.J., and McMahon, H.T. (2009). Mechanisms of endocytosis. Annu. Rev. Biochem. 78, 857–902. Ecker, J.R., and Davis, R.W. (1986). Inhibition of gene expression in plant cells by expression of antisense RNA. Proc. Natl. Acad. Sci. U. S. A. 83, 5372–5376. Eckstein, F., and Jovin, T.M. (1983). Assignment of resonances in the phosphorus-31 nuclear magnetic resonance spectrum of poly[d(A-T)] from phosphorothioate substitution. Biochemistry 22, 4546–4550. Edidin, M. (2003). Lipids on the frontier: a century of cell-membrane bilayers. Nat. Rev. Mol. Cell Biol. 4, 414–418. Elbashir, S.M., Harborth, J., Lendeckel, W., Yalcin, A., Weber, K., and Tuschl, T. (2001). Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells. Nature 411, 494–498. Ellens, H., Siegel, D.P., Alford, D., Yeagle, P.L., Boni, L., Lis, L.J., Quinn, P.J., and Bentz, J. (1989). Membrane fusion and inverted phases. Biochemistry 28, 3692–3703. 148  Ellson, C.D., Andrews, S., Stephens, L.R., and Hawkins, P.T. (2002). The PX domain: a new phosphoinositide-binding module. J. Cell Sci. 115, 1099–1105. English, D. (1996). Phosphatidic acid: a lipid messenger involved in intracellular and extracellular signalling. Cell. Signal. 8, 341–347. Fadok, V.A., Bratton, D.L., Frasch, S.C., Warner, M.L., and Henson, P.M. (1998). The role of phosphatidylserine in recognition of apoptotic cells by phagocytes. Cell Death Differ. 5, 551–562. Fahy, E., Subramaniam, S., Brown, H.A., Glass, C.K., Merrill, A.H., Murphy, R.C., Raetz, C.R.H., Russell, D.W., Seyama, Y., Shaw, W., et al. (2005). A comprehensive classification system for lipids. J. Lipid Res. 46, 839–861. Farhood, H., Serbina, N., and Huang, L. (1995). The role of dioleoyl phosphatidylethanolamine in cationic liposome mediated gene transfer. Biochim. Biophys. Acta 1235, 289–295. Farren, S.B., Hope, M.J., and Cullis, P.R. (1983). Polymorphic phase preferences of phosphatidic acid: a 31P and 2H NMR study. Biochem. Biophys. Res. Commun. 111, 675–682. Felgner, J.H., Kumar, R., Sridhar, C.N., Wheeler, C.J., Tsai, Y.J., Border, R., Ramsey, P., Martin, M., and Felgner, P.L. (1994). Enhanced gene delivery and mechanism studies with a novel series of cationic lipid formulations. J. Biol. Chem. 269, 2550–2561. Felgner, P.L., Gadek, T.R., Holm, M., Roman, R., Chan, H.W., Wenz, M., Northrop, J.P., Ringold, G.M., and Danielsen, M. (1987). Lipofection: a highly efficient, lipid-mediated DNA-transfection procedure. Proc. Natl. Acad. Sci. U. S. A. 84, 7413–7417. Fenske, D.B., and Cullis, P.R. (2008). Liposomal nanomedicines. Expert Opin. Drug Deliv. 5, 25–44. Fenske, D.B., Wong, K.F., Maurer, E., Maurer, N., Leenhouts, J.M., Boman, N., Amankwa, L., and Cullis, P.R. (1998). Ionophore-mediated uptake of ciprofloxacin and vincristine into large unilamellar vesicles exhibiting transmembrane ion gradients. Biochim. Biophys. Acta 1414, 188–204. Ferguson, S.S.G., Downey, W.E., Colapietro, A., Barak, L.S., Ménard, L., and Caron, M.G. (1996). Role of β-Arrestin in mediating agonist-promoted G protein-coupled receptor internalization. Science. 271, 363–366. 149  Filipowicz, W., Jaskiewicz, L., Kolb, F.A., and Pillai, R.S. (2005). Post-transcriptional gene silencing by siRNAs and miRNAs. Curr. Opin. Struct. Biol. 15, 331–341. Fire, A., Xu, S., Montgomery, M.K., Kostas, S.A., Driver, S.E., and Mello, C.C. (1998). Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans. Nature 391, 806–811. Ford, M.G.J., Mills, I.G., Peter, B.J., Vallis, Y., Praefcke, G.J.K., Evans, P.R., and McMahon, H.T. (2002). Curvature of clathrin-coated pits driven by epsin. Nature 419, 361–366. De Fougerolles, A., Vornlocher, H., Maraganore, J., and Lieberman, J. (2007). Interfering with disease: a progress. Nat. Rev. Drug Discov. 6, 443–453. Friend, D.S., Papahadjopoulos, D., and Debs, R.J. (1996). Endocytosis and intracellular processing accompanying transfection mediated by cationic liposomes. Biochim. Biophys. Acta 1278, 41–50. Futter, C.E., Pearse, A., Hewlett, L.J., and Hopkins, C.R. (1996). Multivesicular endosomes containing internalized EGF-EGF receptor complexes mature and then fuse directly with lysosomes. J. Cell Biol. 132, 1011–1023. Gaidarov, I., and Keen, J.H. (1999). Phosphoinositide-AP-2 interactions required for targeting to plasma membrane clathrin-coated pits. J. Cell Biol. 146, 755–764. Gilleron, J., Querbes, W., Zeigerer, A., Borodovsky, A., Marsico, G., Schubert, U., Manygoats, K., Seifert, S., Andree, C., Stöter, M., et al. (2013). Image-based analysis of lipid nanoparticle-mediated siRNA delivery, intracellular trafficking and endosomal escape. Nat. Biotechnol. 31, 638–646. Gregoriadis, G. (1973). Drug entrapment in liposomes. FEBS Lett. 36, 292–296. Gregoriadis, G., Leathwood, P.D., and Ryman, B.E. (1971). Enzyme entrapment in liposomes. FEBS Lett. 14, 95–99. Gruenberg, J. (2001). The endocytic pathway: a mosaic of domains. Nat. Rev. Mol. Cell Biol. 2, 721–730. Gruenberg, J., and Stenmark, H. (2004). The biogenesis of multivesicular endosomes. Nat. Rev. Mol. Cell Biol. 5, 317–323. 150  Guo, X., and Szoka, F.C. (2001). Steric stabilization of fusogenic liposomes by a low-pH sensitive PEG--diortho ester--lipid conjugate. Bioconjug. Chem. 12, 291–300. Hafez, I.M., and Cullis, P.R. (2001). Roles of lipid polymorphism in intracellular delivery. Adv. Drug Deliv. Rev. 47, 139–148. Hafez, I.M., Ansell, S., and Cullis, P.R. (2000). Tunable pH-sensitive liposomes composed of mixtures of cationic and anionic lipids. Biophys. J. 79, 1438–1446. Hafez, I.M., Maurer, N., and Cullis, P.R. (2001). On the mechanism whereby cationic lipids promote intracellular delivery of polynucleic acids. Gene Ther. 8, 1188–1196. Hainfeld, J.F., Slatkin, D.N., and Smilowitz, H.M. (2004). The use of gold nanoparticles to enhance radiotherapy in mice. Phys. Med. Biol. 49, N309–N315. Hainfeld, J.F., Slatkin, D.N., Focella, T.M., and Smilowitz, H.M. (2006). Gold nanoparticles: a new X-ray contrast agent. Br. J. Radiol. 79, 248–253. Hannun, Y.A., and Obeid, L.M. (2008). Principles of bioactive lipid signalling: lessons from sphingolipids. Nat. Rev. Mol. Cell Biol. 9, 139–150. Hanson, M.A., Cherezov, V., Griffith, M.T., Roth, C.B., Jaakola, V.-P., Chien, E.Y.T., Velasquez, J., Kuhn, P., and Stevens, R.C. (2008). A specific cholesterol binding site is established by the 2.8 A structure of the human beta2-adrenergic receptor. Structure 16, 897–905. He, G., Gupta, S., Yi, M., Michaely, P., Hobbs, H.H., and Cohen, J.C. (2002). ARH is a modular adaptor protein that interacts with the LDL receptor, clathrin, and AP-2. J. Biol. Chem. 277, 44044–44049. Hecker, J.G. (2013). Synthetic Messenger RNA and Cell Metabolism Modulation. In Synthetic Messenger RNA and Cell Metabolism Modulation: Methods and Protocols, P.M. Rabinovich, ed. (Totowa, NJ: Humana Press), pp. 73–88. Hegde, R.S., and Keenan, R.J. (2011). Tail-anchored membrane protein insertion into the endoplasmic reticulum. Nat. Rev. Mol. Cell Biol. 12, 787–798. Henne, W.M., Boucrot, E., Meinecke, M., Evergren, E., Vallis, Y., Mittal, R., and McMahon, H.T. (2010). FCHo proteins are nucleators of clathrin-mediated endocytosis. Science. 328, 1281–1284. 151  Hinrichsen, L., Meyerholz, A., Groos, S., and Ungewickell, E.J. (2006). Bending a membrane: how clathrin affects budding. Proc. Natl. Acad. Sci. U. S. A. 103, 8715–8720. Hinshaw, J.E., and Schmid, S.L. (1995). Dynamin self-assembles into rings suggesting a mechanism for coated vesicle budding. Nature 374, 190–192. Hirsch-Lerner, D., Zhang, M., Eliyahu, H., Ferrari, M.E., Wheeler, C.J., and Barenholz, Y. (2005). Effect of “helper lipid” on lipoplex electrostatics. Biochim. Biophys. Acta 1714, 71–84. Hong, K., Friend, D.S., Glabe, C.G., and Papahadjopoulos, D. (1983). Liposomes containing colloidal gold are useful probe of liposome-cell interactions. Biochim. Biophys. Acta 732, 320–323. Hope, M.J., and Cullis, P.R. (1980). Effects of divalent cations and pH on phosphatidylserine model membranes: A 31P NMR study. Biochem. Biophys. Res. Commun. 92, 846–852. Hope, M.J., Walker, D.C., and Cullis, P.R. (1983). Ca2+ and pH induced fusion of small unilamellar vesicles consisting of phosphatidylethanolamine and negatively charged phospholipids: a freeze fracture study. Biochem. Biophys. Res. Commun. 110, 15–22. Hope, M.J., Bally, M.B., Webb, G., and Cullis, P.R. (1985). Production of large unilamellar vesicles by a rapid extrusion procedure. Characterization of size distribution, trapped volume and ability to maintain a membrane potential. Biochim. Biophys. Acta 812, 55–65. Hopkins, S., Furman, P.A., and Painter, G.R. (1989). Investigation of the stereochemical course of DNA synthesis catalysed by human immunodeficiency virus type 1 reverse transcriptase. Biochem. Biophys. Res. Commun. 163, 106–110. Huang, C. (1969). Studies on phosphatidylcholine vesicles. Formation and physical characteristics. Biochemistry 8, 344–352. Huang, J., and Feigenson, G.W. (1999). A microscopic interaction model of maximum solubility of cholesterol in lipid bilayers. Biophys. J. 76, 2142–2157. Huang, L., Sullenger, B., and Juliano, R. (2010). The role of carrier size in the pharmacodynamics of antisense and siRNA oligonucleotides. J. Drug Target. 18, 567–574. Hu-Lieskovan, S., Heidel, J.D., Bartlett, D.W., Davis, M.E., and Triche, T.J. (2005). Sequence-specific knockdown of EWS-FLI1 by targeted, nonviral delivery of small interfering RNA inhibits tumor growth in a murine model of metastatic Ewing’s sarcoma. Cancer Res. 65, 8984–8992. 152  Hunter, D.G., and Frisken, B.J. (1998). Effect of extrusion pressure and lipid properties on the size and polydispersity of lipid vesicles. Biophys. J. 74, 2996–3002. Huo, S., Ma, H., Huang, K., Liu, J., Wei, T., Jin, S., Zhang, J., He, S., and Liang, X.-J. (2013). Superior penetration and retention behavior of 50 nm gold nanoparticles in tumors. Cancer Res. 73, 319–330. Hutvágner, G., and Zamore, P.D. (2002). A microRNA in a multiple-turnover RNAi enzyme complex. Science. 297, 2056–2060. Iden, D.L., and Allen, T.M. (2001). In vitro and in vivo comparison of immunoliposomes made by conventional coupling techniques with those made by a new post-insertion approach. Biochim. Biophys. Acta 1513, 207–216. Ikonen, E. (2008). Cellular cholesterol trafficking and compartmentalization. Nat. Rev. Mol. Cell Biol. 9, 125–138. Jayaraman, M., Ansell, S.M., Mui, B.L., Tam, Y.K., Chen, J., Du, X., Butler, D., Eltepu, L., Matsuda, S., Narayanannair, J.K., et al. (2012). Maximizing the potency of siRNA lipid nanoparticles for hepatic gene silencing in vivo. Angew. Chem. Int. Ed. Engl. 51, 8529–8533. Jeffs, L.B., Palmer, L.R., Ambegia, E.G., Giesbrecht, C., Ewanick, S., and MacLachlan, I. (2005). A Scalable, Extrusion-Free Method for Efficient Liposomal Encapsulation of Plasmid DNA. Pharm. Res. 22, 362–372. Johnson, S.M., and Bangham, A.D. (1969). Potassium permeability of single compartment liposomes with and without valinomycin. Biochim. Biophys. Acta 193, 82–91. Johnson, S.M., Bangham, A.D., Hill, M.W., and Korn, E.D. (1971). Single bilayer liposomes. Biochim. Biophys. Acta 233, 820–826. Johnson, S.M., Miller, K.W., and Bangham, A.D. (1973). The opposing effects of pressure and general anaesthetics on the cation permeability of liposomes of varying lipid composition. Biochim. Biophys. Acta 307, 42–57. Johnston, M.J.W., Semple, S.C., Klimuk, S.K., Edwards, K., Eisenhardt, M.L., Leng, E.C., Karlsson, G., Yanko, D., and Cullis, P.R. (2006). Therapeutically optimized rates of drug release can be achieved by varying the drug-to-lipid ratio in liposomal vincristine formulations. Biochim. Biophys. Acta 1758, 55–64. 153  Judge, A.D., Bola, G., Lee, A.C.H., and MacLachlan, I. (2006). Design of noninflammatory synthetic siRNA mediating potent gene silencing in vivo. Mol. Ther. 13, 494–505. Kaiser, H., Or, A., Róg, T., Nyholm, T.K.M., Chai, W., and Feizi, T. (2011). Lateral sorting in model membranes by cholesterol- mediated hydrophobic matching. Proc. Natl. Acad. Sci. U. S. A. 108, 16628–16633. Kaler, E.W., Murthy, A.K., Rodriguez, B.E., and Joseph, A.N. (1989). Spontaneous vesicle formation in aqueous mixtures of single-tailed surfactants. Science. 245, 1371–1374. Kirpotin, D.B., Drummond, D.C., Shao, Y., Shalaby, M.R., Hong, K., Nielsen, U.B., Marks, J.D., Benz, C.C., and Park, J.W. (2006). Antibody targeting of long-circulating lipidic nanoparticles does not increase tumor localization but does increase internalization in animal models. Cancer Res. 66, 6732–6740. Kobayashi, T., Stang, E., Fang, K.S., de Moerloose, P., Parton, R.G., and Gruenberg, J. (1998). A lipid associated with the antiphospholipid syndrome regulates endosome structure and function. Nature 392, 193–197. Kobayashi, T., Beuchat, M.-H., Chevallier, J., Makino, A., Mayran, N., Escola, J.-M., Lebrand, C., Cosson, P., Kobayashi, T., and Gruenberg, J. (2002). Separation and characterization of late endosomal membrane domains. J. Biol. Chem. 277, 32157–32164. Kojima, C., Hirano, Y., Yuba, E., Harada, A., and Kono, K. (2008). Preparation and characterization of complexes of liposomes with gold nanoparticles. Colloids Surf. B. Biointerfaces 66, 246–252. Koltover, I., Salditt, T., Radler, J.O., and Safinya, C.R. (1998). An Inverted Hexagonal Phase of Cationic Liposome-DNA Complexes Related to DNA Release and Delivery. Science. 281, 78–81. Kormann, M.S.D., Hasenpusch, G., Aneja, M.K., Nica, G., Flemmer, A.W., Herber-Jonat, S., Huppmann, M., Mays, L.E., Illenyi, M., Schams, A., et al. (2011). Expression of therapeutic proteins after delivery of chemically modified mRNA in mice. Nat. Biotechnol. 29, 154–157. Kuntsche, J., Horst, J.C., and Bunjes, H. (2011). Cryogenic transmission electron microscopy (cryo-TEM) for studying the morphology of colloidal drug delivery systems. Int. J. Pharm. 417, 120–137. Kutateladze, T.G. (2006). Phosphatidylinositol 3-phosphate recognition and membrane docking by the FYVE domain. Biochim. Biophys. Acta 1761, 868–877. 154  Lafleur, M., Bloom, M., Eikenberry, E.F., Gruner, S.M., Han, Y., and Cullis, P.R. (1996). Correlation between lipid plane curvature and lipid chain order. Biophys. J. 70, 2747–2757. Laurent, S., Forge, D., Port, M., Roch, A., Robic, C., Vander Elst, L., and Muller, R.N. (2008). Magnetic iron oxide nanoparticles: synthesis, stabilization, vectorization, physicochemical characterizations, and biological applications. Chem. Rev. 108, 2064–2110. Lawe, D.C., Chawla, A., Merithew, E., Dumas, J., Carrington, W., Fogarty, K., Lifshitz, L., Tuft, R., Lambright, D., and Corvera, S. (2002). Sequential roles for phosphatidylinositol 3-phosphate and Rab5 in tethering and fusion of early endosomes via their interaction with EEA1. J. Biol. Chem. 277, 8611–8617. Layzer, J.M., Mccaffrey, A.P., Tanner, A.K., Huang, Z.A.N., and Kay, M.A. (2004). In vivo activity of nuclease-resistant siRNAs In vivo activity of nuclease-resistant siRNAs. RNA 10, 766–771. Lee, H., de Vries, A.H., Marrink, S.-J., and Pastor, R.W. (2009). A coarse-grained model for polyethylene oxide and polyethylene glycol: conformation and hydrodynamics. J. Phys. Chem. B 113, 13186–13194. Lentz, B.R. (2003). Exposure of platelet membrane phosphatidylserine regulates blood coagulation. Prog. Lipid Res. 42, 423–438. Leonetti, J.P., Machy, P., Degols, G., Lebleu, B., and Leserman, L. (1990). Antibody-targeted liposomes containing oligodeoxyribonucleotides complementary to viral RNA selectively inhibit viral replication. Proc. Natl. Acad. Sci. U. S. A. 87, 2448–2451. Lesniak, A., Fenaroli, F., Monopoli, M.P., Christoffer, A., Dawson, K.A., and Salvati, A. (2012). Effects of the Presence or Absence of a Protein Corona on Silica Nanoparticle Uptake and Impact on Cells. ACS Nano 6, 5845–5857. Leung, A.K.K., Hafez, I.M., Baoukina, S., Belliveau, N.M., Zhigaltsev, I. V, Afshinmanesh, E., Tieleman, D.P., Hansen, C.L., Hope, M.J., and Cullis, P.R. (2012). Lipid Nanoparticles Containing siRNA Synthesized by Microfluidic Mixing Exhibit an Electron-Dense Nanostructured Core. J. Phys. Chem. C 116, 18440–18450. Li, W., and Szoka, F.C. (2007). Lipid-based nanoparticles for nucleic acid delivery. Pharm. Res. 24, 438–449. 155  Li, B., Tang, Q., Cheng, D., Qin, C., Xie, F.Y., Wei, Q., Xu, J., Liu, Y., Zheng, B., Woodle, M.C., et al. (2005). Using siRNA in prophylactic and therapeutic regimens against SARS coronavirus in Rhesus macaque. Nat. Med. 11, 944–951. Liebert, M.A., Tousignant, J.D., Gates, A.M.Y.L., Ingram, L.A., Johnson, C.L., Nietupski, J.B., Cheng, S.H., Eastman, S.J., and Scheule, R.K. (2000). Comprehensive Analysis of the Acute Toxicities Induced by Systemic Administration of Cationic Lipid: Plasmid DNA Complexes in Mice. Hum. Gene Ther. 11, 2493–2513. Litzinger, D.C. (1997). Limitations of cationic liposomes for antisense oligonucleotide delivery in vivo. J. Liposome Res. 7, 51–61. Luisi, P.L., Walde, P., and Oberholzer, T. (1999). Lipid vesicles as possible intermediates in the origin of life. Curr. Opin. Colloid Interface Sci. 4, 33–39. Luzio, J.P., Pryor, P.R., and Bright, N.A. (2007). Lysosomes: fusion and function. Nat. Rev. Mol. Cell Biol. 8, 622–632. Luzzati, V., Wicki, T.G.-K., and Tardieu, A. (1968). Polymorphism of Lecithin. Nature 218, 1031–1034. MacDonald, R.I. (1985). Membrane fusion due to dehydration by polyethylene glycol, dextran, or sucrose. Biochemistry 24, 4058–4066. Maclachlan, I. (2007). Liposomal Formulations for Nucleic Acid Delivery. In Antisense Drug Technologies, 2nd Ed., (Taylor & Francis Group, LLC), pp. 237–270. Madden, T.D., and Cullis, P.R. (1982). Stabilization of bilayer structure for unsaturated phosphatidylethanolamines by detergents. Biochim. Biophys. Acta 684, 149–153. Maeda, H. (2001). The enhanced permeability and retention (EPR) effect in tumor vasculature: the key role of tumor-selective macromolecular drug targeting. Adv. Enzyme Regul. 41, 189–207. Maeda, H., Wu, J., Sawa, T., Matsumura, Y., and Hori, K. (2000). Tumor vascular permeability and the EPR effect in macromolecular therapeutics: a review. J. Control. Release 65, 271–284. Marrink, S.-J., and Mark, A.E. (2004). Molecular view of hexagonal phase formation in phospholipid membranes. Biophys. J. 87, 3894–3900. 156  Matranga, C., Tomari, Y., Shin, C., Bartel, D.P., and Zamore, P.D. (2005). Passenger-strand cleavage facilitates assembly of siRNA into Ago2-containing RNAi enzyme complexes. Cell 123, 607–620. Matsuo, H., Chevallier, J., Mayran, N., Le Blanc, I., Ferguson, C., Fauré, J., Blanc, N.S., Matile, S., Dubochet, J., Sadoul, R., et al. (2004). Role of LBPA and Alix in multivesicular liposome formation and endosome organization. Science 303, 531–534. Matzke, M.A., and Matzke, A.J.M. (2004). Planting the seeds of a new paradigm. PLoS Biol. 2, E133. Maurer, N., Wong, K.F., Stark, H., Louie, L., McIntosh, D., Wong, T., Scherrer, P., Semple, S.C., and Cullis, P.R. (2001). Spontaneous entrapment of polynucleotides upon electrostatic interaction with ethanol-destabilized cationic liposomes. Biophys. J. 80, 2310–2326. Maxfield, F.R., and McGraw, T.E. (2004). Endocytic recycling. Nat. Rev. Mol. Cell Biol. 5, 121–132. Mayer, L.D., Bally, M.B., Hope, M.J., and Cullis, P.R. (1985). Uptake of antineoplastic agents into large unilamellar vesicles in response to a membrane potential. Biochim. Biophys. Acta 816, 294–302. Mayer, L.D., Bally, M.B., Hope, M.J., and Cullis, P.R. (1986a). Techniques for encapsulating bioactive agents into liposomes. Chem. Phys. Lipids 40, 333–345. Mayer, L.D., Hope, M.J., and Cullis, P.R. (1986b). Vesicles of variable sizes produced by a rapid extrusion procedure. Biochim. Biophys. Acta 858, 161–168. Mayer, L.D., Bally, M.B., and Cullis, P.R. (1986c). Uptake of adriamycin into large unilamellar vesicles in response to a pH gradient. Biochim. Biophys. Acta 857, 123–126. Mayer, L.D., Nayar, R., Thies, R.L., Boman, N.L., Cullis, P.R., and Bally, M.B. (1993). Identification of vesicle properties that enhance the antitumour activity of liposomal vincristine against murine L1210 leukemia. Cancer Chemother. Pharmacol. 33, 17–24. Mayor, S., and Pagano, R.E. (2007). Pathways of clathrin-independent endocytosis. Nat. Rev. Mol. Cell Biol. 8, 603–612. McCaffrey, A.P., Meuse, L., Pham, T.-T.T., Conklin, D.S., Hannon, G.J., and Kay, M. A. (2002). RNA interference in adult mice. Nature 418, 38–39. 157  McLaughlin, A.C., Cullis, P.R., Hemminga, M.A., Hoult, D.I., Radda, G.K., Ritchie, G.A., Seeley, P.J., and Richards, R.E. (1975). Application of 31P NMR to model and biological membrane systems. FEBS Lett. 57, 213–218. McMahon, H.T., and Boucrot, E. (2011). Molecular mechanism and physiological functions of clathrin-mediated endocytosis. Nat. Rev. Mol. Cell Biol. 12, 517–533. McMahon, H.T., and Gallop, J.L. (2005). Membrane curvature and mechanisms of dynamic cell membrane remodelling. Nature 438, 590–596. Van Meer, G. (2005). Cellular lipidomics. EMBO J. 24, 3159–3165. Van Meer, G., Voelker, D.R., and Feigenson, G.W. (2008). Membrane lipids: where they are and how they behave. Nat. Rev. Mol. Cell Biol. 9, 112–124. Mills, I.G., Jones, A.T., and Clague, M.J. (1998). Involvement of the endosomal autoantigen EEA1 in homotypic fusion of early endosomes. Curr. Biol. 8, 881–884. Mislick, K.A., and Baldeschwieler, J.D. (1996). Evidence for the role of proteoglycans in cation-mediated gene transfer. Proc. Natl. Acad. Sci. U. S. A. 93, 12349–12354. Mok, K.W., Lam, A.M.I., and Cullis, P.R. (1999). Stabilized plasmid-lipid particles: factors influencing plasmid entrapment and transfection properties. Biochim. Biophys. Acta 1419, 137–150. Morris, S.M., and Cooper, J.A. (2001). Disabled-2 colocalizes with the LDLR in clathrin-coated pits and interacts with AP-2. Traffic 2, 111–123. Mortimer, I., Tam, P., MacLachlan, I., Graham, R.W., Saravolac, E.G., and Joshi, P.B. (1999). Cationic lipid-mediated transfection of cells in culture requires mitotic activity. Gene Ther. 6, 403–411. Motley, A., Bright, N.A., Seaman, M.N.J., and Robinson, M.S. (2003). Clathrin-mediated endocytosis in AP-2-depleted cells. J. Cell Biol. 162, 909–918. Mounkes, L.C., Zhong, W., Cipres-Palacin, G., Heath, T.D., and Debs, R.J. (1998). Proteoglycans Mediate Cationic Liposome-DNA Complex-based Gene Delivery in Vitro and in Vivo. J. Biol. Chem. 273, 26164–26170. Mui, B., Chow, L., and Hope, M.J. (2003). Extrusion technique to generate liposomes of defined size. Methods Enzymol. 367, 3–14. 158  Mui, B.L., Tam, Y.K., Jayaraman, M., Ansell, S.M., Du, X., Tam, Y.Y.C., Lin, P.J., Chen, S., Narayanannair, J.K., Rajeev, K.G., et al. (2013). Influence of Polyethylene Glycol Lipid Desorption Rates on Pharmacokinetics and Pharmacodynamics of siRNA Lipid Nanoparticles. Mol. Ther. Nucleic Acids 2, e139. Mukherjee, S., Ghosh, R.N., and Maxfield, F.R. (1997). Endocytosis. Physiol. Rev. 77, 759–803. Mullock, B.M., Bright, N. a, Fearon, C.W., Gray, S.R., and Luzio, J.P. (1998). Fusion of lysosomes with late endosomes produces a hybrid organelle of intermediate density and is NSF dependent. J. Cell Biol. 140, 591–601. Murata, M., Peränen, J., Schreiner, R., Wieland, F., Kurzchalia, T. V, and Simons, K. (1995). VIP21/caveolin is a cholesterol-binding protein. Proc. Natl. Acad. Sci. U. S. A. 92, 10339–10343. Nakayama, T., Butler, J.S., Sehgal, A., Severgnini, M., Racie, T., Sharman, J., Ding, F., Morskaya, S.S., Brodsky, J., Tchangov, L., et al. (2012). Harnessing a physiologic mechanism for siRNA delivery with mimetic lipoprotein particles. Mol. Ther. 20, 1582–1589. Napoli, C., Lemieux, C., and Jorgensen, R. (1990). Introduction of a Chimeric Chalcone Synthase Gene into Petunia Results in Reversible Co-Suppression of Homologous Genes in trans. Plant Cell 2, 279–289. Nichols, J.W., and Deamer, D.W. (1976). Catecholamine uptake and concentration by liposomes maintaining pH gradients. Biochim. Biophys. Acta 455, 269–271. Niu, S.-L., Mitchell, D.C., Lim, S.-Y., Wen, Z.-M., Kim, H.-Y., Salem, N., and Litman, B.J. (2004). Reduced G protein-coupled signaling efficiency in retinal rod outer segments in response to n-3 fatty acid deficiency. J. Biol. Chem. 279, 31098–31104. Olson, F., Hunt, C.A., Szoka, F.C., Vail, W.J., and Papahadjopoulos, D. (1979). Preparation of liposomes of defined size distribution by extrusion through polycarbonate membranes. Biochim. Biophys. Acta 557, 9–23. Orlova, E. V, Sherman, M.B., Chiu, W., Mowri, H., Smith, L.C., and Gotto, A.M. (1999). Three-dimensional structure of low density lipoproteins by electron cryomicroscopy. Proc. Natl. Acad. Sci. U. S. A. 96, 8420–8425. Paasonen, L., Laaksonen, T., Johans, C., Yliperttula, M., Kontturi, K., and Urtti, A. (2007). Gold nanoparticles enable selective light-induced contents release from liposomes. J. Control. Release 122, 86–93. 159  Di Paolo, G., and De Camilli, P. (2006). Phosphoinositides in cell regulation and membrane dynamics. Nature 443, 651–657. Papahadjopoulos, D., and Watkins, J.C. (1967). Phospholipid model membranes II. Permeability properties of hydrated liquid crystals. Biochim. Biophys. Acta 135, 639–652. Papahadjopoulos, D., Nir, S., and Dfizgfines, N. (1990). Molecular Mechanisms of Calcium-Induced Membrane Fusion. J. Bioenerg. Biomembr. 22, 157–179. Park, S.-H., Oh, S.-G., Mun, J.-Y., and Han, S.-S. (2006). Loading of gold nanoparticles inside the DPPC bilayers of liposome and their effects on membrane fluidities. Colloids Surf. B. Biointerfaces 48, 112–118. Parr, M.J., Ansell, S.M., Choi, L.S., and Cullis, P.R. (1994). Factors influencing the retention and chemical stability of poly(ethylene glycol)-lipid conjugates incorporated into large unilamellar vesicles. Biochim. Biophys. Acta 1195, 21–30. Patil, Y.B., Swaminathan, S.K., Sadhukha, T., Ma, L., and Panyam, J. (2010). The use of nanoparticle-mediated targeted gene silencing and drug delivery to overcome tumor drug resistance. Biomaterials 31, 358–365. Pfeiffer, K., Gohil, V., Stuart, R.A., Hunte, C., Brandt, U., Greenberg, M.L., and Schägger, H. (2003). Cardiolipin stabilizes respiratory chain supercomplexes. J. Biol. Chem. 278, 52873–52880. Pillé, J.-Y., Li, H., Blot, E., Bertrand, J.-R., Pritchard, L.-L., Opolon, P., Maksimenko, a, Lu, H., Vannier, J.-P., Soria, J., et al. (2006). Intravenous delivery of anti-RhoA small interfering RNA loaded in nanoparticles of chitosan in mice: safety and efficacy in xenografted aggressive breast cancer. Hum. Gene Ther. 17, 1019–1026. Pryor, P.R., Mullock, B.M., Bright, N.A., Lindsay, M.R., Gray, S.R., Richardson, S.C.W., Stewart, A., James, D.E., Piper, R.C., and Luzio, J.P. (2004). Combinatorial SNARE complexes with VAMP7 or VAMP8 define different late endocytic fusion events. EMBO Rep. 5, 590–595. Raiborg, C., Bache, K.G., Gillooly, D.J., Madshus, I.H., Stang, E., and Stenmark, H. (2002). Hrs sorts ubiquitinated proteins into clathrin-coated microdomains of early endosomes. Nat. Cell Biol. 4, 394–398. Raiborg, C., Wesche, J., Malerød, L., and Stenmark, H. (2006). Flat clathrin coats on endosomes mediate degradative protein sorting by scaffolding Hrs in dynamic microdomains. J. Cell Sci. 119, 2414–2424. 160  Rand, T.A., Ginalski, K., Grishin, N. V, and Wang, X. (2004). Biochemical identification of Argonaute 2 as the sole protein required for RNA-induced silencing complex activity. Proc. Natl. Acad. Sci. U. S. A. 101, 14385–14389. Rand, T.A., Petersen, S., Du, F., and Wang, X. (2005). Argonaute2 cleaves the anti-guide strand of siRNA during RISC activation. Cell 123, 621–629. Ratcliff, F., Harrison, B.D., and Baulcombe, D.C. (1997). A Similarity Between Viral Defense and Gene Silencing in Plants. Science. 276, 1558–1560. Rodríguez, A., Webster, P., Ortego, J., and Andrews, N.W. (1997). Lysosomes behave as Ca2+-regulated exocytic vesicles in fibroblasts and epithelial cells. J. Cell Biol. 137, 93–104. Rodrigueza, W. V, Pritchard, P.H., and Hope, M.J. (1993). The influence of size and composition on the cholesterol mobilizing properties of liposomes in vivo. Biochim. Biophys. Acta 1153, 9–19. Rosenberg, J.M., Seeman, N.C., Day, R.O., and Rich, A. (1976). RNA double helices generated from crystal structures of double helical dinucleoside phosphates. Biochem. Biophys. Res. Commun. 69, 979–987. Roux, A., Uyhazi, K., Frost, A., and De Camilli, P. (2006). GTP-dependent twisting of dynamin implicates constriction and tension in membrane fission. Nature 441, 528–531. Rungta, R.L., Choi, H.B., Lin, P.J.C., Ko, R.W., Ashby, D., Nair, J., Manoharan, M., Cullis, P.R., and Macvicar, B.A. (2013). Lipid Nanoparticle Delivery of siRNA to Silence Neuronal Gene Expression in the Brain. Mol. Ther. Nucleic Acids 2, e136. Saftig, P., and Klumperman, J. (2009). Lysosome biogenesis and lysosomal membrane proteins: trafficking meets function. Nat. Rev. Mol. Cell Biol. 10, 623–635. Sahay, G., Querbes, W., Alabi, C., Eltoukhy, A., Sarkar, S., Zurenko, C., Karagiannis, E., Love, K., Chen, D., Zoncu, R., et al. (2013). Efficiency of siRNA delivery by lipid nanoparticles is limited by endocytic recycling. Nat. Biotechnol. 31, 653–658. Sapra, P., and Allen, T.M. (2003). Ligand-targeted liposomal anticancer drugs. Prog. Lipid Res. 42, 439–462. Scheule, R.K., St. George, J.A., Bagley, R.G., Marshall, J., Kaplan, J.M., Akita, G.Y., Wang, K.X., Lee, E.R., Harris, D.J., Jiang, C., et al. (1997). Basis of pulmonary toxicity associated with cationic lipid-mediated gene transfer to the mammalian lung. Hum. Gene Ther. 8, 689–707. 161  Schiffelers, R.M., Ansari, A., Xu, J., Zhou, Q., Tang, Q., Storm, G., Molema, G., Lu, P.Y., Scaria, P. V, and Woodle, M.C. (2004). Cancer siRNA therapy by tumor selective delivery with ligand-targeted sterically stabilized nanoparticle. Nucleic Acids Res. 32, e149. Schlossman, D.M., Schmid, S.L., Braell, W.A., and Rothman, J.E. (1984). An enzyme that removes clathrin coats: purification of an uncoating ATPase. J. Cell Biol. 99, 723–733. Schröder, B., Wrocklage, C., Pan, C., Jäger, R., Kösters, B., Schäfer, H., Elsässer, H.-P., Mann, M., and Hasilik, A. (2007). Integral and associated lysosomal membrane proteins. Traffic 8, 1676–1686. Semple, S.C., Klimuk, S.K., Harasym, T.O., Dos Santos, N., Ansell, S.M., Wong, K.F., Maurer, N., Stark, H., Cullis, P.R., Hope, M.J., et al. (2001). Efficient encapsulation of antisense oligonucleotides in lipid vesicles using ionizable aminolipids: formation of novel small multilamellar vesicle structures. Biochim. Biophys. Acta 1510, 152–166. Semple, S.C., Akinc, A., Chen, J., Sandhu, A.P., Mui, B.L., Cho, C.K., Sah, D.W.Y., Stebbing, D., Crosley, E.J., Yaworski, E., et al. (2010). Rational design of cationic lipids for siRNA delivery. Nat. Biotechnol. 28, 172–176. Sessa, G., and Weissmann, G. (1968). Phospholipid spherules (liposomes) as a model for biological membranes. J. Lipid Res. 9, 310–318. Shao, S., and Hegde, R.S. (2011). Membrane protein insertion at the endoplasmic reticulum. Annu. Rev. Cell Dev. Biol. 27, 25–56. Shen, J., Samul, R., Silva, R.L., Akiyama, H., Liu, H., Saishin, Y., Hackett, S.F., Zinnen, S., Kossen, K., Fosnaugh, K., et al. (2006). Suppression of ocular neovascularization with siRNA targeting VEGF receptor 1. Gene Ther. 13, 225–234. Shin, J., Shum, P., and Thompson, D.H. (2003). Acid-triggered release via dePEGylation of DOPE liposomes containing acid-labile vinyl ether PEG-lipids. J. Control. Release 91, 187–200. Siegel, D.P. (1993). Energetics of intermediates in membrane fusion: comparison of stalk and inverted micellar intermediate mechanisms. Biophys. J. 65, 2124–2140. Siegel, D.P. (1999). The modified stalk mechanism of lamellar/inverted phase transitions and its implications for membrane fusion. Biophys. J. 76, 291–313. Simons, K., and Ikonen, E. (1997). Functional rafts in cell membranes. Nature 387, 569–572. 162  Skinner, S.A., Tutton, P.J.M., and Brien, P.E.O. (1990). Microvascular architecture of experimental colon tumors in the rat. Cancer Res. 50, 2411–2417. Soong, R., and Macdonald, P.M. (2007). PEG molecular weight and lateral diffusion of PEG-ylated lipids in magnetically aligned bicelles. Biochim. Biophys. Acta 1768, 1805–1814. Soutschek, J., Akinc, A., Bramlage, B., Charisse, K., Constien, R., Donoghue, M., Elbashir, S., Geick, A., Hadwiger, P., Harborth, J., et al. (2004). Therapeutic silencing of an endogenous gene by systemic administration of modified siRNAs. Nature 432, 173–178. Spin, J.M., and Atkinson, D. (1995). Cryoelectron Microscopy of Low Density Lipoprotein in Vitreous Ice. Biophys. J. 68, 2115–2123. Stamatatos, L., Leventis, R., Zuckermann, M.J., and Silvius, J.R. (1988). Interactions of cationic lipid vesicles with negatively charged phospholipid vesicles and biological membranes. Biochemistry 27, 3917–3925. Steinberg, F., Konerding, M.A., and Streffer, C. (1990). The vascular architecture of human xenotransplanted tumors: histological, morphometrical, and ultrastructural studies. J. Cancer Res. Clin. Oncol. 116, 517–524. Stroock, A.D., Dertinger, S.K.W., Ajdari, A., Mezic, I., Stone, H. a, and Whitesides, G.M. (2002). Chaotic mixer for microchannels. Science. 295, 647–651. Struck, D.K., Hoekstra, D., and Pagano, R.E. (1981). Use of resonance energy transfer to monitor membrane fusion. Biochemistry 20, 4093–4099. Sweitzer, S.M., and Hinshaw, J.E. (1998). Dynamin undergoes a GTP-dependent conformational change causing vesiculation. Cell 93, 1021–1029. Szoka Jr, F., and Papahadjopoulos, D. (1978). Procedure for preparation of liposomes with large internal aqueous space and high capture by reverse-phase evaporation. Proc. Natl. Acad. Sci. U. S. A. 75, 4194–4198. Tam, P., Monck, M., Lee, D., Ludkovski, O., Leng, E.C., Clow, K., Stark, H., Scherrer, P., Graham, R.W., and Cullis, P.R. (2000). Stabilized plasmid-lipid particles for systemic gene therapy. Gene Ther. 7, 1867–1874. Tam, Y.Y.C., Chen, S., Zaifman, J., Tam, Y.K., Lin, P.J.C., Ansell, S., Roberge, M., Ciufolini, M. a, and Cullis, P.R. (2013). Small molecule ligands for enhanced intracellular delivery of lipid nanoparticle formulations of siRNA. Nanomedicine 9, 665–674. 163  Thakker, D.R., Natt, F., Hüsken, D., Maier, R., Müller, M., van der Putten, H., Hoyer, D., and Cryan, J.F. (2004). Neurochemical and behavioral consequences of widespread gene knockdown in the adult mouse brain by using nonviral RNA interference. Proc. Natl. Acad. Sci. U. S. A. 101, 17270–17275. Thakker, D.R., Natt, F., Hüsken, D., van der Putten, H., Maier, R., Hoyer, D., and Cryan, J.F. (2005). siRNA-mediated knockdown of the serotonin transporter in the adult mouse brain. Mol. Psychiatry 10, 782–789, 714. Théry, C., Zitvogel, L., and Amigorena, S. (2002). Exosomes: composition, biogenesis and function. Nat. Rev. Immunol. 2, 569–579. Tilcock, C.P.S., and Cullis, P.R. (1981). The polymorphic phase behaviour of mixed phosphatidyserine-phosphatidylethanolamine model systems as detected by 31P-NMR. Effects of divalent cations and pH. Biochim. Biophys. Acta 641, 189–201. Tilcock, C.P., Bally, M.B., Farren, S.B., and Cullis, P.R. (1982). Influence of cholesterol on the structural preferences of dioleoylphosphatidylethanolamine-dioleoylphosphatidylcholine systems: a phosphorus-31 and deuterium nuclear magnetic resonance study. Biochemistry 21, 4596–4601. Tinker, D.O., and Pinteric, L. (1971). On the identification of lamellar and hexagonal phases in negatively stained phospholipid-water systems. Biochemistry 10, 860–865. Tolentino, M.J., Brucker, A.J., Fosnot, J., Ying, G., Wu, I.-H., Malik, G., Wan, S., and Reich, S.J. (2004). Intravitreal injection of vascular endothelial growth factor small interfering RNA inhibits growth and leakage in a nonhuman primate, laser-induced model of choroidal neovascularization. Retina 24, 660. Tomari, Y., and Zamore, P.D. (2005). Perspective: machines for RNAi. Genes Dev. 19, 517–529. Trajkovic, K., Hsu, C., Chiantia, S., Rajendran, L., Wenzel, D., Wieland, F., Schwille, P., Brügger, B., and Simons, M. (2008). Ceramide triggers budding of exosome vesicles into multivesicular endosomes. Science. 319, 1244–1247. Ungewickell, E., Ungewickell, H., Holstein, S.E.H., Lindner, R., Prasad, K., Barouch, W., Martin, B., Greene, L.E., and Eisenberg, E. (1995). Role of auxilin in uncoating clathrin-coated vesicles. Nature 378, 632–635. 164  Wang, Y., Su, H.-H., Yang, Y., Hu, Y., Zhang, L., Blancafort, P., and Huang, L. (2013). Systemic delivery of modified mRNA encoding herpes simplex virus 1 thymidine kinase for targeted cancer gene therapy. Mol. Ther. 21, 358–367. Webb, M.S., Harasym, T.O., Masin, D., Bally, M.B., and Mayer, L.D. (1995). Sphingomyelin-cholesterol liposomes significantly enhance the pharmacokinetic and therapeutic properties of vincristine in murine and human tumour models. Br. J. Cancer 72, 896–904. Wenk, M.R. (2010). Lipidomics: new tools and applications. Cell 143, 888–895. Wheeler, J.J., Palmer, L., Ossanlou, M., MacLachlan, I., Graham, R.W., Zhang, Y.P., Hope, M.J., Scherrer, P., and Cullis, P.R. (1999). Stabilized plasmid-lipid particles: construction and characterization. Gene Ther. 6, 271–281. Whitehead, K.A., Langer, R., and Anderson, D.G. (2009). Knocking down barriers: advances in siRNA delivery. Nat. Rev. Drug Discov. 8, 129–138. Wrobel, I., and Collins, D. (1995). Fusion of cationic liposomes with mammalian cells occurs after endocytosis. Biochim. Biophys. Acta 1235, 296–304. Xu, Y., and Szoka, F.C. (1996). Mechanism of DNA release from cationic liposome/DNA complexes used in cell transfection. Biochemistry 35, 5616–5623. Yuan, F., Leunig, M., Huang, S.K., Berk, D.A., Papahadjopoulos, D., and Jam, R.K. (1994). Mirovascular permeability and interstitial penetration of sterically stabilized (stealth) liposomes in a human tumor xenograft microvascular permeability and interstitial penetration of sterically stabilized (stealth) liposomes in a human tumor xenograft. Cancer Res. 54, 3352–3356. Yuan, F., Dellian, M., Fukumura, D., Leunig, M., Berk, D.A., Torchilin, V.P., and Jain, R.K. (1995). Vascular Permeability in a Human Tumor Xenograft: Molecular Size Dependence and Cutoff Size. Cancer Res. 55, 3752–3756. Yuan, X., Naguib, S., and Wu, Z. (2011). Recent advances of siRNA delivery by nanoparticles. Expert Opin. Drug Deliv. 8, 521–536. Zelphati, O., and Szoka, F.C. (1996). Mechanism of oligonucleotide release from cationic liposomes. Proc. Natl. Acad. Sci. U. S. A. 93, 11493–11498. Zhang, J.-S., Liu, F., and Huang, L. (2005). Implications of pharmacokinetic behavior of lipoplex for its inflammatory toxicity. Adv. Drug Deliv. Rev. 57, 689–698. 165  Zhigaltsev, I. V, Maurer, N., Edwards, K., Karlsson, G., and Cullis, P.R. (2006). Formation of drug-arylsulfonate complexes inside liposomes: a novel approach to improve drug retention. J. Control. Release 110, 378–386. Zhigaltsev, I. V, Winters, G., Srinivasulu, M., Crawford, J., Wong, M., Amankwa, L., Waterhouse, D., Masin, D., Webb, M., Harasym, N., et al. (2010). Development of a weak-base docetaxel derivative that can be loaded into lipid nanoparticles. J. Control. Release 144, 332–340. Zhigaltsev, I. V, Belliveau, N., Hafez, I., Leung, A.K.K., Huft, J., Hansen, C., and Cullis, P.R. (2012). Bottom-up design and synthesis of limit size lipid nanoparticle systems with aqueous and triglyceride cores using millisecond microfluidic mixing. Langmuir 28, 3633–3640. Zimmermann, T.S., Lee, A.C.H., Akinc, A., Bramlage, B., Bumcrot, D., Fedoruk, M.N., Harborth, J., Heyes, J.A., Jeffs, L.B., John, M., et al. (2006). RNAi-mediated gene silencing in non-human primates. Nature 441, 111–114. Zou, S., Scarfo, K., Nantz, M.H., and Hecker, J.G. (2010). Lipid-mediated delivery of RNA is more efficient than delivery of DNA in non-dividing cells. Int. J. Pharm. 389, 232–243. Van Zuylen, L., Verweij, J., and Sparreboom, a (2001). Role of formulation vehicles in taxane pharmacology. Invest. New Drugs 19, 125–141.   

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0165976/manifest

Comment

Related Items