Design, elaboration, characterization and evaluation of niosome formulations for gene delivery to retina and brain

Design, elaboration, characterization and evaluation of niosome formulations for gene delivery to retina and brain Mohamed Ahmed Ahmed Mashal Vitoria-Gasteiz 2020 NanoBioCel Grupo de Micro y Nano Tecnologías, Biomateriales y Células Design, elaboration, characterization and evaluation of niosome formulations for gene delivery to retina and brain Mohamed Ahmed Ahmed Mashal NanoBiocel Group, Laboratory of Pharmaceutics University of the Basque Country (UPV/EHU) Faculty of Pharmacy Vitoria-Gasteiz, 2020 (cc)2020 MOHAMED AHMED AHMED MASHAL (cc by 4.0) Design, elaboration, characterization and evaluation of niosome formulations for gene delivery to retina and brain Mohamed Ahmed Ahmed Mashal NanoBiocel Group, Laboratory of Pharmaceutics University of the Basque Country (UPV/EHU) Faculty of Pharmacy Vitoria-Gasteiz, 2020 ACKNOWLEDGMENTS First and to the greatest extent, praises and thanks to the Allah, the Almighty, for His showers of blessings during the way of my research work . It is said that “If you're not thankful to people, you're not thankful to God”. Therefore, I have to say that the completion of this thesis would have been impossible without the love, support, guidance and encouragement of many individuals. I would like to begin by expressing my deepest gratitude to my dearest supervisor Prof. Dr. Jose Luis Pedraz for giving me the opportunity to work on this thesis in his laboratory. He taught me the preciseness in science and the good practice in the lab. I feel myself privileged for being able to work under his experienced supervision towards accomplishing this research successfully. Honestly, I’ve always considered prof. Dr. Pedraz and Angela as my family in Spain. Dr. Pedraz, I can't thank you enough for all that you have done for me! It is with great pleasure that I am expressing my thanks and appreciations to my dearest supervisor, Prof. Dr. Gustavo Puras, to whom I am immensely indebted. I truly thank him for being so enthusiastic about the work, for his great scientific guidance, reviewing the thesis manuscript and for his decent constructive comments. I can’t express enough thanks to my soul mate, wife and co-worker, Dr. Noha Attia, for what no words can describe. Noha, without your dedications, support, and patience, obtaining my Ph.D. would have been totally impossible. Research is an exciting journey of discovery. Thus, I extend my gratitude to all coauthors for their sincere invaluable support. The human, alongside with the scientific support from all members of the team was a continuous feeling throughout the whole journey in the lab. Lastly, I own my warmest thanks to my dear ones, my Mom (Layla) and all my family members. ACKNOWLEDGMENT FOR THE FINANCIAL SUPPORT This project was supported by the Basque Country Government (CGIC10/172), Spanish Ministry of Education (Grant CTQ2017-84415-R, MAT2015-69967-C3-1R), the Generalitat de Catalunya (2014/SGR/624), and the Instituto de Salud Carlos III (CB06_01_0019, CB06_01_1028). The authors also wish to thank the intellectual and technical assistance from the ICTS “NANBIOSIS”, more specifically by the Drug Formulation Unit (U10) of the CIBER in Bioengineering, Biomaterials, and Nanomedicine (CIBER-BBN) at the University of Basque Country (UPV/EHU). Technical and human support provided by SGIker (UPV/EHU) is gratefully acknowledged. ACKNOWLEDGMENT TO THE EDITORIALS Authors would like to thank the editorials for granting permission to reuse their previously published articles in this thesis. Mashal Mohamed, et al. "Retinal gene delivery enhancement by lycopene incorporation into cationic niosomes based on DOTMA and polysorbate 60."Journal of Controlled Release 254 (2017): 55-64. Mashal Mohamed, et al. "Non-viral vectors based on cationic niosomes as efficient gene delivery vehicles to central nervous system cells into the brain."International journal of pharmaceutics 552.1-2 (2018): 48-55. Mashal, Mohamed, et al. "Gene delivery to the rat retina by non-viral vectors based on chloroquine-containing cationic niosomes."Journal of Controlled Release 304 (2019): 181-190. Man cannot remake himself without suffering, for he is both the marble and the sculptor. "Man, The Unknown" Alexis Carrel Chapter1 4 (4). This was a major achievement and allowed long stretches of DNA to be rapidly and accurately sequenced. In 1983, the polymerase chain reaction (PCR) technique was discovered by Kary Mullis, a breakthrough that enabled scientists to rapidly amplify DNA (5). 1.1.2. Clinical trials and commercial products of gene therapy After decades of research, cell and gene therapies are moving from bench lab to bedside. Since the year 1989 (first gene therapy clinical trial), more than 2600 clinical trials for gene therapy have been approved globally till the year 2017 (Figure 1) (6). Figure 1. The number of gene therapy clinical trials approved worldwide 1989-2017 (6). In 1990 the 4-year-old DeSilva with severe combined immune deficiency (SCID) became the first patient to undergo gene therapy in the United States. This trial was authorized by the NIH Recombinant DNA Advisory Committee (RAC) and the Food and Drug Administration (FDA). Functional adenosine deaminase (ADA) genes 144 12814 37 38 67 51 82 68 117 9610898 85101112 117 90 120 81 92 87102 125 135 169 118 132 0 20 40 60 80 100 120 140 160 180 Chapter1 5 were transferred by retroviral vectors into the cultured T cells. The T lymphocytes with the corrected gene were reinfused back into the patient about 12 days after blood was drawn. Also, at the NIH Clinical Center in1991, Cynthia a 9-years-old girl underwent another trial. Unfortunately, in 1999 the 18-year-old Jesse Gelsinger with inherited enzyme deficiency was the first victim of gene therapy after 4 days of injection with a genetically altered adenovirus into his liver. This was a direct reason for severe handicap in the research field of this technology (7). The field of gene therapy was shaken again in 2002, when a three-year-old boy with SCID treated in a French trial developed leukemia. Therefore, gene therapy clinical trials were halted again (8). Nevertheless, Gendicine (recombinant human p53 adenovirus) was approved in 2003 by the China Food and Drug Administration (CFDA) as a first gene therapy product to treat head and neck cancer (9). In 2007, British doctors performed the world’s first gene therapy operation to treat Leber Congenital Amaurosis due to RPE65 mutations via ocular subretinal injection of adeno-associated virus (AAV) gene vector (10). In 2012, the first gene therapy drug approved by the European Medicines Agency (EMA) was Glybera (alipogene tiparvovec). Glybera is an AAV vector engineered to express lipoprotein lipase for the treatment of lipoprotein lipase deficiency (11). Figure 2. Distribution countries of approval gene therapy drugs (12) Chapter1 6 The year of 2017 is considered the distinguished starting point for gene therapy in the United States. In August 2017, kymriah® was approved by the FDA as the first gene therapy product to be marketed for treatment of acute lymphoblastic leukemia. In the same year, Yescarta®, was approved by FDA for treating large B-cell lymphoma. Furthermore, by December 2017, the FDA-approved Luxturna® became the first in vivo gene therapy drug to treat Leber congenital amaurosis LCA (a rare inherited eye disease). Recently, Zolgensma® became the first gene therapy approved by FDA in 2019 to treat children less than two years of age with spinal muscular atrophy (SMA), a leading hereditary cause of infant mortality. Figure 3. Timeline of gene therapy development (12) Such recently approved drugs for human use based on gene, along with advances on revolutionary genome editing technologies, suggest that gene therapy can become a regular medical option into the clinical practice. Chapter1 7 One of the organs that is considered an enticing gene therapy target is the eye, because of its accessibility and its immune privilege. The vision process is highly complex, and requires coordination of numerous components in both eye and brain. Moreover, the retina of the eye represents a part of the central nervous system (CNS) and is actually a brain tissue. Although the RPE65 gene therapy (via viral carriers) has paved the way for the treatment of retinal diseases. Some other genes are too large to be carried into the retina via viral carriers, such as the gene causing Stargardt’s hereditary maculopathy. The Retinostat® is a gene therapy that is currently in phase I clinical trial for the wet subtype of age-related macular degeneration (AMD). It inhibits blood vessel growth by expressing angiostatin and endostatin proteins. On the other side, the sFLT gene (carried on AAV vector) was injected subretinal to block the vascular endothelial growth factor (VEGF) (13). Table 1 depicts some gene therapy clinical trials for eye disorders at different stages. In a CNS clinical trial (NCT01454596), the gene for epidermal growth factor receptor (EGFRvIII) was incorporated in retrovirus to treat patients with glioblastoma. A retroviral vector transporting a chimeric antigen receptor (CAR) for the EGFRvIII tumor antigen, can be used to mediate genetic transfer of CAR with high efficiency (14). Table 2 elaborates more on the gene therapy clinical trials for CNS disorders. Table 1. Gene therapy clinical trials on eye diseases (15). Chapter1 8 Table 2. Gene therapy clinical trials on CNS diseases (15). 1.1.3. Gene therapy strategies Currently, various gene therapy strategies do exist. Such strategies could be categorized into: (1) gene replacement for the monogenic diseases, (2) gene addition to treat acquired diseases, (3) gene editing to introduce targeted changes in host genome, and (4) alteration of gene expression by targeting RNA. The treatment may take place outside of the body (ex vivo) or inside the body (in vivo). Modified viruses or other vectors are used as gene delivery systems To deliver the gene into the genome into the cells. The significant potential of using plasmids for gene therapy has been recognized since 1990. Selecting the right composition of a plasmid is fundamental to ensure the success of gene therapy. There are many factors and elements to be considered when choosing the plasmid backbone such as: cloning or expression safety, plasmid size, antibiotic resistance, restriction sites in multiple cloning site (MCS), promoter, terminator, ribosome binding site (RBS) sequence, or protein modifications (adding a tag or a fusion protein to the plasmid to further understand the function of a specific gene) (16). Both figure 4 and table 3 describe different elements of the plasmid. Chapter1 9 Figure 4. A representative map of the main features of a plasmid vector. (17) Table 3. Plasmid elements and their description. (17) Plasmids are simpler and cheaper to make, ship and store compared with viral and RNA-based vectors and have a much longer shelf life. The modular nature of plasmids also allows enables simple molecular cloning, making them easy to manipulate and design for therapeutic use. In addition, the plasmids can be distributed repeatedly, unlike viruses. Most plasmid DNA preparations include many topological plasmid variations, including supercoiled (the preferred topology), but also the undesirable linear and open circular forms of the plasmid. In order to deliver their Chapter1 10 payload, Plasmids require vectors, physical forces, or advanced modifications for uptake and nuclear localization. Since plasmids are non-replicating episomes, the expression of transgenes is transient and diluted by cell division. The dinucleotides of unmethylated cytosine-phosphate-guanine (CpG) are more widespread in bacterial DNA than in mammalian DNA. They have the ability to be recognized by the mammalian immune system via toll-like receptor (TLR)-9, and are potentially precipitating not only in transgene silencing, but also in immune response (18). RNA precursors are essential targets in genetic therapies. Antisense oligonucleotides (ASOs) are single-stranded (ss) DNA or RNA sequences that can be used for silencing overexpressed proteins in toxic “gain of function” diseases. They can target RNA for degradation, preventing the translation of a specific RNA into protein and altering the splicing of pre-mRNA. ssDNA complementary to mRNA could be used to block the translation of specific mRNA (19). Aptamers are short, single-stranded DNA or RNA (ssDNA or ssRNA) molecules that can bind to a specific target, including proteins, peptides, carbohydrates, small molecules, toxins, and even living cells. Aptamers are selected from a large oligonucleotide library through a process called SELEX (Sequential Evolution of Ligands by Exponential Enrichment) (20). Recently, the gene editing tool CRISPR/Cas9 (Clustered Regularly Interspaced Short Palindromic Repeats/ CRISPR-associated protein 9) was used (figure 5). Cas9 is an RNA-guided DNA endonuclease enzyme that uses CRISPR sequences as a guide to identify and cleave specific strands of DNA complementing the CRISPR sequence in order to mediate genome alteration with high precision. Unlike the previous gene editing techniques such as Transcription Activators-Like Effective Nucleases (TALENS) this technology is simple, easy to use and inexpensive. However, they have limitations, such as narrow targeting range and potential for off-target mutagenesis (21). But with higher precision, one of the approaches based on the CRISPR technique is prime editing technology tool, which both specifies the target location and encodes the desired edit. The prime technique writes new genetic information directly Chapter1 11 into a specified DNA site using a catalytically impaired Cas9 endonuclease fused to an engineered reverse transcriptase, programmed with a prime editing guide RNA (pegRNA). It rewrites DNA by cutting just one strand to add, delete, or substitute base pairs that can edit more genetic mutation forms than genome-editing approaches like CRISPR-Cas9 (22). RNA interference (RNAi) is formulated and presented as a small double-stranded (ds) RNA intended to down-regulate its target transcript (i.e., mRNA) and protein. Nonetheless, for long-term knockdown, a more robust strategy is to use another RNAi route, by delivering synthetic microRNA (miRNA) (24). Additionally, short interfering RNAs (siRNAs) can downregulate a selected mRNA gene that appears to degrade through the RNA-interfering silencing complex (RISC). The platforms siRNA and ASO are both nucleic acids containing antisense strand designed to identify a target mRNA. Though, ASOs have one strand while siRNAs have two, which can reduce cost and simplify delivery (25). In 2018, Alnylam launched the ONPATTRO® (patisiran) drug which has been approved by the FDA as the first siRNA-based therapy for the rare hereditary disease transthyretin-mediated amyloidosis in adult patients (26). Chapter1 12 Figure 5. CRISPR/Cas9-mediated DNA cleavage and simplified repair mechanism (23). 1.2.Gene delivery vectors Current gene delivery vehicles, namely vectors, are classified mainly into two classes: the viral and the non-viral vectors. 1.2.1.Viral vectors Viruses are attractive gene-delivery vehicles due to their ability for efficient nucleic acid delivery to specific cell type while escaping host cell immune surveillance. Moreover, they may exploit the cellular machinery to encourage their replication (27). Viral delivery system is based on a virus with nucleic acid coated by capsid protein and in many cases further by an envelope structure. One or more viral structural genes are deleted to deactivate the virus from spread in the host organism (28). Viral vectors had been designed for temporary short-term and permanent long-term expression, and can be conveyed by both RNA and DNA viruses with either single-stranded or double- Chapter1 13 stranded genomes (29). However, many disadvantages and safety issues are related to the use of viral vectors such as complex and expensive production, limited carrying capacity, broad tropism, oncogenicity, mutagenicity, immunogenicity and inflammatory responses (30). Recently, chimeric viral vectors have been developed in a trial to overcome the limitations of each viral species, by combining suitable features of two or more different viruses into one (31). The retroviral capsid is an enveloped shell of protein with size of 80-100 nm. It has an average packaging capacity of 8 kb. Retroviral vector integrates in the host cell genome, by using reverse transcriptase in a stable and permanent manner. Replicationcompetent and replication-defective are the two types of retroviral vectors. Retrovirus are mainly applied in studies for tissue repair and engineering due to their capability of infecting dividing cells without developing any immunogenic viral proteins (32). Lentiviruses, subtype of retrovirus, are able to integrate into non-dividing cells. The lentiviral genome is similar to other retroviruses; however, it contains six other genes -two regulatory genes and four accessory genesthat code for proteins essential for viral replication, binding, infection, and release. Lentivirus common example is the human immunodeficiency virus type 1 (HIV-1). Lentiviral vectors seem to be less mutagenic than their retroviral counterparts (33). Adenoviruses are icosahedral, non-enveloped, ds DNA viruses with a size of 70 nm, and a genome size of about 36 kb. They are the most widely used viral vectors able to overcome the limitations of other viral vectors as the retrovirus Among the 50 different serotypes of adenovirus, the types 2 and 5 are the most frequently used ones (34). The packaging capacity of dsDNA adenoviruses is about 7.5 kb of foreign DNA with short-term episomal expression with broad range of host cells. In addition, there is a possibility to hybridize adenovirus vectors with sleeping beauty transposase system for chromosomal integration or with CRISPR/Cas9 nuclease for gene editing (35). The AAV is a small (22 nm in diameter) nonpathogenic parvovirus with a nonenveloped, icosahedral capsid. Its genome is composed of a linear, ss DNA. It is used Chapter1 20 Solid lipid nanoparticles (SLN) were presented in 1990 as a substitute delivery system to liposomes, emulsions and polymeric nanoparticles. SLNs are composed of solid fat which is dispersed in an aqueous phase in the presence of surfactants to enhance its stability. The proper selection of lipids and surfactants can affect the particle size, stability, loaded molecules and behaviors of release. Their lipid components are solid at both body and ambient temperature (57). Nanostructured lipid carriers (NLCs) are spherical structures with a mixed solid and liquid matrix, having an aqueous core surrounded by a lipid bilayer. There are three major types of NLC: cationic, neutral, and targeting-modified NLC. Cationic NLCs can be used as carriers for negatively charged substance, including proteins, polypeptides, oligonucleotides, RNAs and DNAs (58) NLC have many clinical applications because of superior biocompatibility, high biodegradability and low immunogenicity. They are used in the delivery of nucleic acids including distinct miRNA molecules for cancer gene therapy. Niosomes are self-assembled non-ionic surfactant vesicles. They are composed of three main components: (1) a non-ionic surfactant such as polysorbates; (2) a neutral helper lipid such as cholesterol, squalene and lycopene; (3) a cationic lipid such as DOTMA. The incorporation of non-ionic molecules in niosomes reduces the undesirable toxicity of cationic lipids showing better cellular viability profiles compared to their corresponding anionic or cationic counterparts (59). Niosomes have first emerged as a vesicle delivery system in the 70s in the field of cosmetics industry. Thanks to their capability to encapsulate both hydrophobic and hydrophilic drugs, niosomes are reported as potential carriers for the delivery of drugs such as doxorubicin, vaccines, insulin, siRNA. They have many applicable therapeutic effects (e.g. antiAlzheimer, anti-cancer, anti-oxidant, anti-diabetics and anti-microbial) and can be administrated via various methods, as intravenously, orally (e.g., Flurbiprofen); ocular (e.g., Chloramphenicol, Acetazolamide, Fluconazole), and topically (e.g., Erythromycin, Minoxidil, Rofecoxib) (60). Niosomes are highly stable, yet slightly leakier than liposomes. In comparison to liposomes, niosomes can be formulated at a lower cost, longer stability and less toxicity. Moreover, researchers highlighted on the Chapter1 21 biocompatible, biodegradable pharmaceutics and low immunogenic features of their components (non-ionic surfactants, helper lipids as well as charged molecules). Niosomes have been widely used as oligonucleotide carriers. The formulation of niosomes, by the method of solvent emulsification-evaporation technique, managed to deliver pCMS-EGFP plasmid to the retina and brain. The results proved that niosomes could protect DNA from degradation and introduced good trafficking pattern with good chemical and physical stability and relatively smaller sizes. In addition, niosomes can also be used as vectors in DNA vaccines which provide a simple, stable and costeffective solution compared with liposomes. Niosomes can also assist as a delivery system for targeting stem cells (60). Lipopolysomes are multifunctional nanocarriers composed by combining polymersomes and liposomes. Lipopolysomes have the ability for simultaneous encapsulation of hydrophobic and hydrophilic molecules (61). The vector resulted by co‐formulation of plasmid DNA and lipopolysomes is called lipopolyplexes. In nonviral gene delivery systems this term (lipopolyplexes) conclude a diverse component of lipids and polymers co-formulated with genetic materials. One advantage of lipopolyplexes vehicles is that they have the potential to be targeted to specific cell types by attaching peptide targeting ligands on the surface, thus increasing both the transfection efficiency and selectivity for disease targets such as cancer cells (62). Chapter1 22 Figure 9. Different vectors (A) and vector/DNA complexes (B) containing cationic lipid. 1.4.Gene delivery barriers On their way to the target cells in vivo, gene delivery systems must circumvent several extracellular and intracellular barriers (portrayed in figure 10). Therefore, the design of innovative optimal non-viral vectors is a great challenge to surpass these hurdles. 1.4.1. Extracellular barriers Whatever the administration route is (e.g., inhalation, intramuscular injection, gavage, intravascular injection, oral, etc.), gene delivery vector will be inevitably in contact with the extracellular environment. Basically, these extracellular barriers Chapter1 23 include lipid bilayer membrane, many blood components and endothelial barriers. Additional extracellular barriers need to be overcome by the vehicle when specific tissues need to be targeted, such as the brain, the eye or the lungs (37). The control of unwanted immune responses is the way to success of gene therapy strategies. Gene vectors are potentially able to trigger immune responses as lymphocytes. Moreover, enzymes such as lipases and nucleases can degrade the nano-formulation and the genetic payload and therefore interfere with transfection efficiency (63). The viral and non-viral vectors have been shown to induce an immune response. However, this activation has been most associated with viral vectors. The injection of cationic lipoplexes in the circulation cause the release of TNFα and IFNγ into the serum as inflammatory response chemokines. This may be as a result of unmethylated CpG motifs on the plasmid DNA and the subsequent recognition by TLRs (64).Extracellular barrier to be overcome strongly depend on the organ to be treated as well as the route of administration. In the systemic circulation, blood flows to every organ and tissue in the body. Generally, intravenous administration is the most studied route of administration of non-viral gene delivery systems. Once gene vehicles are introduced into the circulation, they are subject to enzymatic degradation, serum proteins inactivation, complementmediated clearance and reticuloendothelial system recognition (66). Formed blood elements (erythrocytes, leukocytes, and platelets) and serum proteins ( albumin, immunoglobulins, and fibronectin) have a negative surface charge which interacts with the net positive charge of non-viral vector/DNA complex leading to aggregation or dissociation of the complex (67). The delivery of non-viral gene vector to the lung is usually hampered by the pulmonary architecture, the presence of respiratory secretions (mucus and lung surfactant), the clearance mechanisms, and the activation of the immune system. According to the lung disease (e.g., cystic fibrosis, asthma, emphysema and lung cancer), the target cells can vary from epithelial cells, alveolar cells, macrophages, respiratory stem cells or endothelial cells (68). The respiratory secretions bind to the Chapter1 24 complexes and sterically obstruct their way to the target cells limiting their diffusions and effectiveness. Figure 10. Extracellular and intracellular barriers, adapted from (65) The blood-brain barrier (BBB) represents a huge obstacle upon systemic delivery of non-viral vector/DNA complexes into the brain. BBB is composed of brain microvascular endothelial cells, pericytes, astrocytes, tight junctions, and basal laminae (69). Many approaches exploit the receptor-mediated uptake of molecules such as transferrin, lactoferrin and insulin to cross the BBB, since receptors of those molecules Macrophage . Endonucleases Systemicdelivery Extracellularbarriers PlasmaproteinsComplex Nucleus Nuclearpore Transcription Endosome Endocytosis DNAplasmid Intracellularbarriers Extracellularmatrix Extravasation Chapter1 25 are expressed on many cell types, including neurons and the capillary endothelial cells of the BBB. Another approach uses peptidomimetic monoclonal antibodies (known as a molecular Trojan horse) to target specific receptors on the BBB and induce receptormediated transcytosis of the non-viral vector into the CNS. Other strategy includes transient mechanical disruption of the BBB and RNAi-mediated knockdown of tight junction proteins (70). Local administration to the brain, either by injection or by infusion is investigated in many pre-clinical studies. Intranasal delivery is another noninvasive means to deliver non-viral gene carriers to the brain, with the ability to pass the BBB and transfect and express the encoded proteins (71). With regard to the eye, it is a highly compartmentalized and immune-privileged organ that offers interesting advantages as a gene therapy target (72). However, relevant biological barriers such as cornea, sclera, aqueous humor, blood-retinal barriers (BRB), choroidal and conjunctival blood flow, lymphatic clearance, and tear dilution need to be deeply considered. The BRB, which is composed of tight epithelial junctions, limits the delivery of non-viral vector/DNA complexes to the retina via systemic administration. Two strategies were suggested to overcome such BRB, (1) by using vectors smaller than 100 nm, and (2) by the use of ligand-targeted vectors that recognize specific receptors in the BRB (37). 1.4.2. Intracellular barriers Gene delivery systems have to overcome many intracellular barriers extending from cell surface to nuclear entry for successful gene therapy. The non-viral vector/DNA complex enter the cells either by ligand-receptor binding interaction (receptor-mediated endocytosis) or by charge-mediated interactions with proteoglycans on cell membranes. Thus, the vesicles are susceptible to a cascade processes of complex uptake, endolysosomal escape, trafficking to the nucleus, vehicle unpacking and nuclear entry (73). Being up taken into cells is not the only limiting barrier, however it is one of the most limiting steps affecting non-viral vehicle efficiency. Endocytosis is a vesiclemediated process that can be mediated by five main endocytic pathways: clathrin- Chapter1 26 mediated endocytosis (CME), caveolae-mediated endocytosis (CvME), clathrincaveolae-independent endocytosis, macropinocytosis and phagocytosis (74). Lysosomes are membrane-bound intracellular organelles with an acidic pH (45). They have an essential role for degradation and recycling of macromolecules delivered by endocytosis, phagocytosis, and autophagy (75). The ability of many nonviral vehicles to deliver nucleic acid efficiently may be attributed to their strong buffering capacity (pH ranges from 5 to 7). Such strong buffering capacity prevents the acidification of endosomes by acting as ‘proton sponges’ (48). Productive gene transfer requires DNA to eventually cross the nuclear envelope through nuclear pore complexes (NPCs) before initiation of transcription. In the absence of cell division, the intact nuclear envelope impedes the entry of carrierplasmid complexes (76). Various approaches were used to improve plasmids’ nuclear targeting, such as: complexation of plasmids with peptides, proteins, ligands, polymers, and inclusion of transcription factor-binding sites (77). As well, nuclear localization sequence (NLS) peptides can be directly bound to the DNA in order to promote its transport to the nucleus by the importins (78). Chapter1 27 Figure 11. Mechanisms of cellular uptake of non-viral vector/DNA complex. 1.5.Stability of formulations Chemical non-viral vectors are liquid formulations which upon storage are susceptible to stability problems. In general, nanoparticles are extremely unstable due to their high surface energy and they tend to change themselves or react with substances to reach a relatively stable state. Changes during storage and/or shipping requires preparation of freshly lipid/DNA complexes before every single use. Nevertheless, an acceptable stability is essential for pharmaceutical development and commercialization. Generally, physical stability affects the shelf-life of non-viral vector that is related to many parameters, such as: uniformity of size distribution, PDI, zeta potential, lamellar changes, aggregations and/or fusion (79). The change in such parameters may be considered as a function of pH, temperature, buffer concentration, ionic strength, storage time, etc. (80). Micropinocytosis Phagocytosis Clathrinandcaveolae- independent endocytosis Clathrin-dependent endocytosis Caveolae-dependent endocytosis Cellmembrane Macropinosome Lysosome Clathrincoatedvesicle Phagosome Caveosome Lateendosome Nucleus Unpacking Nuclearimport Non-viralvector/DNA complex Dynamin Clathrin Caveolin endosome Chapter1 28 The physical stability study upon storage at different temperature (mainly at 25°C, 4°C, -20°C and -80°C) is one of the basic stability studies. Freezing could alter the membrane structure, and consequently the shape, of Lipofectamine 2000® vesicles and enhance their gene delivery action without compromising cell viability in many cell lines (81). Generally, the changes in physical characteristics by freezing of vesicles can lead to decrease or increase in transfection efficiency. The storage at subzero temperature may be a good substitute for storage at 4°C due to hydrolytic degradation caused by excess bulk water of aqueous formulation leading to less stable formulation on long-term storage. Nonetheless, cryopreservation at -80°C was used for the storage of liposomes, its use for DNA complexes may be unlogic as the addition of cryoprotectants as DMSO and sugars may affect the transfection efficacy and viability of the DNA complexes. In addition, cryopreservation had a damaging effect at the molecular levels specially on the DNA molecules (82). The biological stability of chemical non-viral vectors refers to the interaction of DNA with different components in the biological system. Better understanding of their interactions is essential to establish specific design criteria. The aggregation of DNA by serum proteins and degradation by DNase enzymes are essential obstacles in the stability of DNA vectors. Interestingly, PEG coating tends to provide the protection of DNA against the serum degradation hazards (83). 1.6.References 1. Wirth T, Parker N, Ylä-Herttuala S. History of gene therapy. Gene. 2013;525(2):162-9. 2. Friedmann T, Roblin R. Gene therapy for human genetic disease? Science. 1972;175(4025):949-55. 3. Kelly Jr TJ, Smith HO. A restriction enzyme from Hemophilus influenzae: II. Base sequence of the recognition site. Journal of molecular biology. 1970;51(2):393409. 4. Totomoch-Serra A, Marquez MF, Cervantes-Barragán DE. Sanger sequencing as a first-line approach for molecular diagnosis of Andersen-Tawil syndrome. F1000Research. 2017;6. 5. Bartlett JM, Stirling D. A short history of the polymerase chain reaction. PCR protocols: Springer; 2003. p. 3-6. 6. Ginn SL, Amaya AK, Alexander IE, Edelstein M, Abedi MR. Gene therapy clinical trials worldwide to 2017: An update. The journal of gene medicine. 2018;20(5):e3015. Chapter1 29 7. Marshall E. Gene therapy death prompts review of adenovirus vector. Science. 1999;286(5448):2244-5. 8. Check E. Second cancer case halts gene-therapy trials. Nature Publishing Group; 2003. 9. Zhang W-W, Li L, Li D, Liu J, Li X, Li W, et al. The first approved gene therapy product for cancer Ad-p53 (Gendicine): 12 years in the clinic. Human gene therapy. 2018;29(2):160-79. 10. Bainbridge JW, Smith AJ, Barker SS, Robbie S, Henderson R, Balaggan K, et al. Effect of gene therapy on visual function in Leber's congenital amaurosis. New England Journal of Medicine. 2008;358(21):2231-9. 11. Ylä-Herttuala S. Endgame: glybera finally recommended for approval as the first gene therapy drug in the European union. Molecular Therapy. 2012;20(10):18312. 12. Shahryari A, Saghaeian Jazi M, Mohammadi S, Razavi Nikoo H, Nazari Z, Hosseini ES. Development and Clinical Translation of Gene Therapy for Genetic Disorders. Frontiers in genetics. 2019;10:868. 13. Bright Focus Foundation. Gene therapy for eye disease 2018, Octobar 24 [Available from: https://www.brightfocus.org/macular/article/gene-therapy-eyedisease. 14. Eskilsson E, Røsland GV, Solecki G, Wang Q, Harter PN, Graziani G, et al. EGFR heterogeneity and implications for therapeutic intervention in glioblastoma. Neuro-oncology. 2017;20(6):743-52. 15. Alliance for regenerative medicine. Select Near-Term Clinical Trial Milestones & Data Readouts: 2019+ 2019, March [Available from: https://alliancerm.org/. 16. genomecompiler. how to choose the perfect vector for your molecular biology experiment 2016, March 21 [Available from: http://www.genomecompiler.com/how-to-choose-the-perfectvector/. 17. addgene. Plasmids 101: What is a plasmid? 2014, jan 14 [Available from: https://blog.addgene.org/plasmids-101-what-is-a-plasmid. 18. Hardee CL, Arévalo-Soliz LM, Hornstein BD, Zechiedrich L. Advances in nonviral DNA vectors for gene therapy. Genes. 2017;8(2):65. 19. Weiler J, Hunziker J, Hall J. Anti-miRNA oligonucleotides (AMOs): ammunition to target miRNAs implicated in human disease? Gene therapy. 2006;13(6):496. 20. Pang X, Cui C, Wan S, Jiang Y, Zhang L, Xia L, et al. Bioapplications of cellSELEX-generated aptamers in cancer diagnostics, therapeutics, theranostics and biomarker discovery: a comprehensive review. Cancers. 2018;10(2):47. 21. Ran FA, Hsu PD, Wright J, Agarwala V, Scott DA, Zhang F. Genome engineering using the CRISPR-Cas9 system. Nature protocols. 2013;8(11):2281. 22. Anzalone AV, Randolph PB, Davis JR, Sousa AA, Koblan LW, Levy JM, et al. Search-and-replace genome editing without double-strand breaks or donor DNA. Target. 2019;5(3). 23. Lipocalyx. VIROMER® CRISPR FOR RNP DELIVERY 2019, Octobar, [Available from: https://viromer-transfection.com/crispr. 24. Keeler AM, Elmallah MK, Flotte TR. Gene therapy 2017: progress and future directions. Clinical and translational science. 2017;10(4):242-8. Chapter2 36 Chapter2 37 Gene therapy approach aims to treat both inherited and acquired diseases by delivering a therapeutic genetic material or its regulatory elements to target cells. The future of gene therapy depends mainly on the success to design optimal vector. Viral and non-viral vectors have been used as gene delivery carriers. In spite of the fact that viral vectors have the advantage of high gene transfection, the use of viral vectors is limited due to their side effects. Thus, non-viral vectors, have been developed and applied in gene therapy for their advantages, such as its safety, high gene capacity, stability, chemical design flexibility, and low immunogenic response. Therefore, the main objective of this thesis to design novel niosome formulations, containing novel helper lipid, able to deliver genes to eye and brain, safely and effectively. 1. To study the effect of lycopene, as natural helper lipid, in niosome formulation based on cationic lipid (DOTMA) and non-ionic surfactant (polysorbate 60), to boost transfection efficiency in retinal pigment epithelial (RPE-19) cells, without compromising cell viability and using it to transfect rat retina in vivo. 2. To investigate the-lycopene containing-niosomes to transfect NT2 cells, primary cortical culture as well as brain cortex of rats, as safe and efficient nonviral vectors to deliver DNA into the CNS to face many neurological disorders. 3. To ensure the role and importance of helper molecule to turn on transfection efficiency. Niosomes formulation was probed with different cationic lipid {2,3di (tetradecyloxy) propan-1-amine (hydrochloride salt)} and different helper molecule (chloroquine diphosphate) to transfect rat retinal cells. the incorporation of chloroquine within nano formulations, rather than as a cotreatment of the cells, could open a new avenue for in vivo retinal gene delivery. Chapter2 38 Chapter2 39 Chapter 3 Retinal gene delivery enhancement by lycopene incorporation into cationic niosomes based on DOTMA and polysorbate 60 Chapter3 38 Chapter3 39 Retinal gene delivery enhancement by lycopene incorporation into cationic niosomes based on DOTMA and polysorbate 60 Mohamed Mashala,1, Noha Attiaa,b,1, Gustavo Purasa,c, Gema Martínez-Navarretec,d, Eduardo Fernándezc,d, Jose Luis Pedraza,c, a NanoBioCel Group, Laboratory of Pharmaceutics, School of Pharmacy, University of the Basque Country (UPV/EHU), Paseo de la Universidad 7, 01006 Vitoria-Gasteiz, Spain b Histology and Cell Biology Department, Faculty of Medicine, University of Alexandria, Alexandria, Egypt c Networking Research Centre of Bioengineering, Biomaterials and Nanomedicine (CIBER-BBN), VitoriaGasteiz, Spain d Neuroprothesis and Neuroengineering Research Group, Miguel Hernández University, Elche, Spain Journal of Controlled Release 254 (2017) 55–64 ABSTRACT The present study aimed to evaluate the incorporation of the natural lipid lycopene into niosome formulations based on cationic lipid DOTMA and polysorbate 60 non-ionic surfactant to analyze the potential application of this novel formulation to deliver genetic material into the rat retina. Both niosomes with and without lycopene were prepared by the reverse phase evaporation method and physicochemically characterized in terms of size, zeta potential, polydispersity index and capacity to condense, release and protect the DNA against enzymatic digestion. In vitro experiments were performed in ARPE-19 cells after complexion of niosomes with pCMS-EGFP plasmid at appropriate cationic lipid/DNA ratios. At 18/1 mass ratio, nioplexes containing lycopene had nanometric size, positive zeta potential, low polydispersity and were able to condense, release and protect DNA. Percentage of transfected cell was around 35% without compromising cell viability. The internalization pathways studies revealed a preference to caveolae mediated endocytosis and macropinocytosis, which could circumvent lysosomal degradation. Both subretinal and intravitreal administrations to the rat retina showed that nioplexes were able to transfect efficiently the outer segments of the retina, which offer reasonable hope for the treatment of many inherited retinal diseases by a safe non-viral vector formulation after the less invasive intravitreal administration. Keywords: Niosomes, Lycopene, Gene therapy, Retina, Non-viral vectors, Nanotechnology Chapter3 40 3.1.Introduction The abnormal expression or activity of numerous retinal proteins has been linked to the pathogenesis of several blinding retinal disorders with a genetic background, such as Leber congenital amaurosis [1], age-related macular degeneration [2] or retinitis pigmentosa [3]. Unfortunately, most of these devastating conditions do not have effective treatment at the moment. Although novel approaches, such as enzyme/protein replacement and stem cell-based therapies have shown recently promising results, gene therapy is by far the most well-developed field of research for the treatment of both inherited and acquired retinal disorders [4,5]. The unique anatomical and histological features of the eye provide both benefits and challenges for the progress in gene-based ocular therapeutics [6]. In the last decade, many viral and non-viral gene delivery approaches have been developed for the treatment of many retinal pathologies [7,8]. Compared with their counterparts, non-viral vectors have attracted great attention as safer alternative to deliver genetic material, since they can circumvent many safety issues that are still associated with viral gene delivery systems, such as immunogenicity, mutagenicity and oncogenic effects [9]. Consequently, the use of non-viral vectors in clinical trials has increased since 2004, while that of viral vector has decreased significantly [10]. Actually, cationic lipids and cationic polymers are the most commonly used non-viral vectors [10,11]. However, to date, one of the main problems that non-viral formulations have to face, in order to reach the clinical practice, is their limited transfection efficiency. Therefore, research activity on this area merits special attention for the scientific community [12]. As drug delivery system, niosomes have received growing attention by time for being osmotically active and chemically stable formulations [13]. Besides, when it comes to easy handling and low toxicity, they are considered quite advantageous over the well-known liposomes [14]. However, their use as gene delivery systems has been poorly studied, although some recent results have revealed their appealing properties to transfect efficiently brain and retinal cells in rats [15–17]. Chapter3 41 Niosomes, for gene delivery purposes, are self-assembled vesicular nano carrier systems composed typically by non-ionic surfactant, “helper” and cationic lipids [18]. The non-ionic “electrically neutral” surfactants enhance the stability of niosome formulations [19]. Additionally, cationic lipids form complexes by electrostatic interactions upon the addition of negatively charged genetic material [15,16], and “helper” lipids have a marked influence on both the physicochemical and biological properties of niosome gene carriers [15,17,20]. Recently, it has been reported on the literature the flattering properties of the “helper” lipid squalene (a natural lipid that belongs to the terpenoid family) in cationic niosome gene delivery formulations. Therefore, we decided to investigate the effect that lycopene, another natural lipid, could have on a niosome formulation based on cationic lipid N-[1-(2,3-dioleoyloxy)propyl]-N,N,N-trimethylammonium chlor-ide (DOTMA) and non-ionic surfactant polysorbate 60. Lycopene is a carotenoid that contains 40 carbon and 56 hydrogen atoms (Fig. 1-C). Classically, it is known to be one of the most potent natural antioxidants that mediate cytoprotective, immunomodulatory and anticancer activities [21]. Fig. 1. Chemical structures of the cationic lipid, N-[1-(2,3-dioleoyloxy)propyl]- N,N,N-trimethylammonium chloride (DOTMA) (A), Polysorbate 60 (B), and Lycopene (C). Chapter3 42 Additionally, lycopene can be found at high concentration levels in the eye, where it has shown both anti-inflammatory and anti-angiogenic effects [22]. We designed two novel niosome vector formulations for retinal gene delivery purposes based on the same cationic lipid N-[1-(2,3-dioleoy-loxy)propyl]-N,N,Ntrimethylammonium chloride (DOTMA) and the same non-ionic surfactant polysorbate 60, in the absence/presence of lycopene (DP60 and DP60L, respectively). Both niosomes were elabo-rated by the solvent emulsification-evaporation technique and com-pared in terms of particle size, polydispersity index (PDI) and zeta potential. Upon the addition of the reported pCMS-EGFP plasmid at different cationic lipid/DNA ratios (w/w), nioplexes were obtained and characterized by size, PDI, morphology, and the ability to condense, release and protect the DNA from enzymatic digestion. In vitro experiments were performed to compare the behavior of both vectors in ARPE-19 cells regarding their cellular uptake, transfection efficiency, viability, and internalization mechanism. Following the in vitro characterization, the most promising formulation was administered to rat eyes via intravitreal and subretinal injection in order to evaluate transfection efficiency by confocal microscopy in both whole-mount and sagittal cross sections of the retina. 3.2.Materials and methods 3.2.1. Production of cationic niosomes Niosomes were elaborated with slight modifications of the previously described reverse phase evaporation method [23]. Briefly, 5 mg of cationic lipid DOTMA D (Avanti Polar Lipids Inc., Alabama, USA) and 26 mg of polysorbate 60 P60 (0.5%, w/v, Sigma-Aldrich, Madrid, Spain) with/without 1 mg lycopene L (Sigma-Aldrich, Madrid, Spain) were dissolved in 1 ml of organic solvent, dichloromethane (Panreac, Barcelona, Spain). The emulsions were obtained by sonication of such organic phase with 5 ml milliQ water for 30 s at 45 W (Branson Sonifier 250®, Branson Ultrasonics Corporation, Danbury, USA). Dichloro-methane was removed from emulsions by evaporation under magnetic agitation for 2 h leaving the cationic nanoparticles in the aqueous medium. The corresponding molar ratios of both DP60 and DP60L formulations were 1:4 and 1:4:0.4, respectively. Chapter3 43 3.2.2.Plasmid propagation and elaboration of nioplexes pCMS-EGFP plasmid (5541 bp, Plasmid Factory, Bielefeld, Germany), was propagated in Escherichia coli DH5-α and purified using the Qiagen endotoxin-free plasmid purification Maxi-prep kit (Qiagen, California, USA) according to the manufacturer's instructions. The purified plasmid DNA was quantified by measuring absorbance at 260 nm in a NanoDrop® Spectrophotometer (Thermo Fisher Scientific Inc. Denver, USA). The purity of the plasmid was verified by agarose gel electrophoresis (Bio-Rad, Madrid, Spain) in Tris Borate-EDTA buffer, pH 8.0 (TBE buffer). DNA bands were detected using GelRed™ (Bio-Rad, Madrid, Spain) to stain DNA, and images were observed with a ChemiDoc™ MP Imaging System (Bio-Rad, Madrid, Spain). The stock solution of pCMS-EGFP plasmid (0.5 mg/ml) was estimated to be around 0.14 μM. Both DP60 and DP60L nioplexes (niosome/DNA complexes) were elaborated by mixing an appropriate volume of a stock solution of pCMS-EGFP plasmid (0.5 mg/ml) with different volumes of the niosome suspensions (1 mg cationic lipid/ml) to obtain different cationic lipid/DNA mass ratios (w/w). The mixture was left for 30 min at room temperature to enhance electrostatic interaction between the cationic niosomes and the negatively charged plasmid. 3.2.3.Characterization of niosomes/nioplexes Particle size and polydispersity index (PDI) were determined by dynamic light scattering (DLS) with Zetasizer Nano ZS (Malvern Instruments, UK). Determination of zeta potential by Laser Doppler Velocimetry (LDV) was carried out with the same instrument, where samples were dispersed in a 0.1 mM NaCl solution. Particle size, reported as hydrodynamic diameter, was obtained by cumulative analysis. All measurements were carried out in triplicate. Chapter3 50 reagent Lipofectamine™2000 significantly decreased to 82% (p < 0.05). The micrographs obtained in (Fig. 3B) demonstrated that transfected ARPE-19 cells maintained a normal morphology with both nioplexes even with high cationic lipid/DNA mass ratios. Fig. 2. Physicochemical characterization of nioplexes. A) Effect of cationic lipid/DNA mass ratio (w/w) on both particle size (bars) and zeta potential (lines). Each data point represents the mean ± SD (n = 3). TEM of DP60 (B1) and DP60L nioplexes (B2) at ratio of 18/1 cationic lipid/DNA mass ratio (w/w). Scale bar = 500 nm. Binding, SDSinduced release and protection of DNA at different cationic lipid/DNA mass ratios (w/w) of nioplexes based on both DP60 (C1) and DP60L (C2) visualized by agarose electrophoresis. Lanes 1–3 correspond to uncomplexed DNA; lanes 4–6, cationic lipid/DNA mass ratio 6/1; lanes 7–9, cationic lipid/DNA mass ratio 12/1; lanes 10–12, cationic lipid/DNA mass ratio 18/1; lanes 13–15, cationic lipid/DNA mass ratio 22/1. Nioplexes were treated with SDS (lanes 2, 5, 8, 11 and 14) and DNase I + SDS (lanes 3, 6, 9, 12 and 15). OC: open circular form, SC: supercoiled form. 3.3.3.Cell uptake studies Chapter3 51 Nioplexes at cationic lipid/DNA mass ratio of best transfection, 18/ 1, were used to determine uptake percentage in ARPE-19 cells. Fig. 4-A features the percentage of FITC-positive cells quantified by flow cytometry. In general, the uptake percentage increased over time for both formulations. Additionally, DP60 uptake values (50.5%, 62.4%; and 75.5%) were significantly higher (p < 0.05) than values of DP60L uptake (5.8%, 11.1% and 23%) at the time points tested, respectively. Fig. 3. In vitro transfection efficiency and cell viability in ARPE-19 cells at 72 h posttransfection. (A) Flow cytometry-based evaluation of the percentage of EGFP-positive cells (bars) and percentage of viable cells (lines) at different cationic lipid/DNA mass ratios (w/w). Values represent mean ± SD (n = 3). (*P < 0.05 vs. Lipofectamine™2000 transfection). (#P < 0.05 vs. Lipofectamine™2000 viability). (B) Overlay of fluorescence and phase-contrast micrographs of ARPE-19 cells 72 h post-transfection at different cationic lipid/DNA mass ratios (w/w). Scale bar = 100 μm. Confocal micrographs were obtained after 2 and 4 h of incubation to visualize the progress of complex internalization over time. Fig. 4-B showed that both nioplexes were homogeneously distributed within the cytoplasm 2 h of incubation. However, after 4 h of incubation, cytoplasmic aggregates of DP60 complexes were discerned compared to their DP60L counterparts that maintained their relative homogeneous cytoplasmic distribution. Chapter3 52 Fig. 4. Uptake of FITC-labeled nioplexes in ARPE-19 cells. Both DP60 and DP60L at a mass ratio of 18/1 (w/w). (A) Percentage of FITC-positive cells. Data represent mean ± SD (n = 3). *p < 0.05. (B) Fluorescence micrographs of ARPE-19 cells at 2 h and 4 h of incubation with FITC-labeled DP60 and DP60L nioplexes (green). Nuclei stained with Dapi (blue). Original magnification 63×. Scale bar = 20 μm. (For inter-pretation of the references to color in this figure legend, the reader is referred to the web version of this article.) 3.3.4. Cell internalization studies Fig. 5 illustrates that DP60 nioplexes (green) highly co-localized with each one of the three different endocytosis markers used (red), resulting in yellowish signals that were quantified by Mander's overlap coefficient. Values of M ≥ 0.6 indicated a positive co-localization. However, DP60L complexes, co-localized mainly with cholera toxin (M = 0.76 ± 0.04) and dextran (M = 0.76 ± 0.07) endocytosis markers, whereas no positive co-localization was observed with transferrin (M = 0.53 ± 0.03). Chapter3 53 Fig. 5. Three-channel overlay RGB images of ARPE-19 cells showing nioplexes with FITC-labeled pCMS-EGFP (green) and one of the endocytosis markers in red (AlexaFluor® 555-Cholera Toxin, AlexaFluor® 546-Transferrin or AlexaFluor® 594dextran). Presence of yellow/orange color represents the overlay of an endocytic marker and nioplexes. (M = Mander's overlap coefficient). Original magnification 63×, Scale bar = 25 μm. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) 3.3.5.Histology and immunofluorescence analysis of EGFP expression in vivo After 72 h of intravitreal administration of DP60L nioplexes, EGFP expression in the rat retina was analyzed by CLSM (Fig. 6). The analysis of whole-mount preparations (Fig. 6-A, B) revealed native EGFP-expression in some ganglion cells containing NeuN immunoreactivity (red color) as well as in some cells in contact with a typical microglial-like morphology, while no fluorescence was detected in control retina (data not shown). Fig. 7 shows vertical retinal sections where EGFP expression was detected in both ganglion cell layer (GCL) labeled with anti-NeuN (red color), and outer segments (OS) of photoreceptors with anti-Recoverin, a marker of photoreceptors Chapter3 54 (red color) after intravitreal (Fig. 7-A, B) and subretinal injection (Fig. 7-C, D), regardless the route of administration. Both positive controls of rat retinae transfected with Lipofectamine™ 2000 and negative controls of non-transfected retinae can be observed in the Supplementary data section (Fig. S1). 3.4.Discussion Due to its appealing chemical structure, the commercially available cationic lipid DOTMA has been widely used for gene delivery applications [25]. As shown in Fig. 1A, its structure is composed of a polar head-group, two non-polar hydrophobic chains, a linker and a back-bone, which classically are known as the four domains that rule gene transfection process [26]. We combined DOTMA with the non-ionic surfactant polysorbate 60, in a niosome formulation at a molar ratio of 1:4 respectively, in order to enhance cell tolerance [27] and provide a steric barrier to avoid aggregation [28]. It has been reported on the literature that the presence of PEG chains in the chemical structure of polysorbates (Fig. 1-B) provides physicochemical stability to lipid formulations [29], conserves effectiveness over time and boosts transfection efficiency [30]. Compared with polysorbate 80, another polysorbate that has been widely used in the elaboration of niosome formulations for gene delivery applications, [15–17,31] polysorbate 60 could offer some important advances. For instance, the lack of double bonds in the hydrocarbon chains (Fig. 1-B) could provide low permeability of the vesicles, and therefore better stability of niosome membranes [29]. Additionally, compared with other hydrophilic surfactants such as polysorbates 80, 40 or 20, the low hydrophiliclipophilic balance (HLB) value of polysorbate 60 (14.9) could help to solubilize lycopene more efficiently [31]. The addition of the natural and non-polar lipid lycopene (Fig. 1-C) into niosome bilaminar membrane could increase its fluidity, disturb membrane packing, and consequently vesicle susceptibility to environmental stresses [32]. Chapter3 55 Fig. 6. Immunohistochemical study of EGFP expression in retinal whole-mount preparations 3 days after intravitreal administration of DP60L nioplexes. Partial colocalization of EGFP (green color) with NeuN-positive ganglion cells (red) was observed in the ganglion cell layer (GCL) (red). EGFP expression was observed as well in some cells with typical microglial morphology in GCL (NeuN-negative cells). Nuclei were stained with Hoechst 33342 (blue). Scale bars: 10 μm. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Once elaborated by the reverse phase elaboration method, both DP60 and DP60L niosomes showed appropriate size (in the nanometric scale) and PDI values (below 0.5) for gene delivery purposes (Table 1). High positive ZP values (> +25 mV) ensure longlasting stability [33], spontaneous electrostatic interaction with DNA, as well as binding of the resulting nioplexes to the negatively charged components of the cell membrane prior to cellular uptake [26]. To elaborate nioplexes, we added pCMS-EGFP reporter plasmid to both niosome formulations at different cationic lipid/DNA mass ratios, since otherwise, the complex assembly process could be slowed down [34]. The slight changes discerned in the size of nioplexes (100–150 nm, Fig. 2-A), at the mass ratios studied, might be due to the delicate balance of different events involved in the multistep self-assembled complex formation, such as: electrostatic interaction, further membrane merging, lipid mixing Chapter3 56 and aggregate growth [34]. Regarding the ZP values, the gradual increase of superficial charge along with cationic lipid/DNA ratios (w/w) suggests the capacity of cationic niosomes to bind to and neutralize the negatively charged phosphate groups of DNA [17]. Lycopene addition reduced ZP value of DP60L nioplexes, compared to DP60, at all ratios studied. This fact could be explained by the perturbation of the lipid membrane bilayer, which could dissipate the electrical potential (Fig. 2-A) [35]. In any case, both formulations could function as gene delivery carriers, since the positively charged complexes could interact electrostatically with the anionic cell coat, inducing early steps of the endocytosis process [36]. The high positive ZP value of both DP60 and DP60L nioplexes, especially at 18/1 mass ratio (42 and 27 mV, respectively), could ensure the discrete morphology and absence of aggregates observed by TEM micrographs (Fig. 2-B) [[37].]. Among other factors that can influence on the transfection process, the electrostatic interactions between the negatively charged phosphate groups of the DNA and the positively charged amine groups of the cationic niosomes merits special attention [15–17,38]. We observed by agarose gel electrophoresis assay that at all cationic lipid/DNA ratios tested, both niosomes were able to condense, release and protect the DNA from enzymatic digestion (Fig. 2-C1and C2). Once we evaluated that our formulations were biotechnologically suitable for gene delivery purposes, we proceeded to evaluate their biological performance in ARPE-19 cells. ARPE-19 cell line has a normal karyotype and has functional and structural properties similar to retinal pigment epithelia (RPE) in vivo, expresses RPE-specific markers, hence it is considered a suitable transfection model to investigate our vectors' effectiveness and safety before its application in vivo [39]. It has been reported that the non-ionic nature of surfactants makes niosomes well tolerated by cells [40]. Our results in Fig. 3 show higher cell viability values in cells transfected with both nioplexes when compared with cells transfected with Lipofectamine™ 2000. Additionally, we observed under the fluorescence microscope that cells transfected with both nioplexes maintained their normal morphology, even at high cationic lipid/DNA ratios (Fig. 3- Chapter3 57 B). Although the percentage of transfected cells with DP60L niosomes at 18/1 mass ratio was significantly lower than that obtained with commercially available Lipofectamine™ 2000, our niosomes formulation was better tolerated by ARPE-19 cells. Therefore, it could be an interesting alternative to Lipofectamine™ 2000, since some authors have reported damage on the retina associated to the in vivo administration of Lipofectamine™ 2000 in the eye [41]. Regarding the transfection efficiency, the lipid composition is considered a primary limiting factor that affects to this process [25]. We clearly observed in Fig. 3 the impact that lycopene had on transfection efficiency in ARPE-19 cells, since values were clearly higher when lycopene was present in the niosome formulation. Although the exact mechanism of lycopene action has not yet been fully elucidated, some authors suggest the existence of a lycopene receptor and/or transporter in the nuclear membrane of cells [42]. Additionally, other study has documented the capacity of lycopene to modulate transcription [43]. Such effect could be either by direct interactions with transcription factors such as nuclear factor-kappa, or by indirect modifications of transcriptional activity. In any case, further research is still needed to determine the exact mechanism [44]. The ascending transfection percentages obtained by DP60L at high mass ratios might be attributed, partially, to the triggering effect of free niosomes absorbed on to the cell membrane. At higher cationic lipid/DNA mass ratios, there is a large excess of cationic lipid to DNA, therefore a population of free niosomes is expected [45]. This free cationic lipid could prolong cellular retention or decreases degradation rate of DNA [25]. Nonetheless, additional experiments are needed to elucidate the detailed mechanisms involved. Chapter3 58 Fig. 7. Confocal fluorescence micrographs of retinal cross-sections after 3 days of intravitreal (A,B) and subretinal (C,D) administration of DP60L nioplexes. After intravitreal injections, EGFP fluorescence was observed in glial cells in the GCL, and in the outer segments of the photoreceptors. Localization of EGFP after subretinal injections was detected as well in OS of photoreceptors (stained with recoverin, in red) and some microglial cells. Cell nuclei were counterstained with Hoechst 33342 (blue). Scale bars: 20 μm. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Chapter3 59 To determine whether enhanced internalization was among the effects that lycopene incorporation could have in niosome formulations, we studied the percentage of cellular uptake of both DP60 and DP60L formulations at the mass ratio of best transfection efficiency (18/1) in ARPE-19 cells at different times (Fig. 4). Surprisingly, flow cytometry studies showed that lycopene addition clearly reduced the percentage of cellular uptake at all times studies when compared to DP60 formulation (Fig. 4-A). Such reduction in the cellular uptake could probably be due to the lower zeta potential of DP60L formulation compared with DP60 (Fig. 2-A). Additionally, CLSM studied (Fig. 4-B), excluded mere electrostatic adherence of cationic nioplexes to the negatively charged surface of ARPE-19 cell, since a clear intracellular distribution of both nioplexes was observed in the case of both formulations. In any case, the cytoplasmatic distribution of both nioplexes showed a different behavior. Whereas DP60L nioplexes maintained a homogeneous distribution in the cytoplasm over the time, DP60 nioplexes showed some aggregates at 4 h. The differences observed in the cytoplasmatic distribution of both formulations could suggest different internalization pathways. Therefore, and motivated by the differences observed between both formulation in terms of transfection efficiency and cellular uptake, we next studied the cellular trafficking of both nioplexes at the mass ratio of best transfection efficiency (18/1). Three of the pathways most employed in the uptake processes were assayed; clathrin-mediated endocytosis (CME), caveola-mediated endocytosis (CvME) and macropinocytosis (Fig. 5) [9]. The performance of non-viral vectors is known to be clearly affected by their distinct cellular internalization pathway, taking into account the variable effectiveness of every pathway in the release of DNA into the cytoplasm, which is one of the critical steps in the eventual transgene expression [34]. Although there is not a clear consensus in the scientific community, it is widely accepted that the endolysosomal fate is the hallmark feature of CME [37,46]. In the other hand, CvME and macropinocytosis are widely related to non-acidic and non-digestive routes of cellular uptake [47,48]. Therefore, our results observed in Fig. 5 suggest that internalization of DP60L via these last pathways could be advantageous Chapter3 66 [35] W.S. Sung, I.S. Lee, D.G. Lee, Damage to the cytoplasmic membrane and cell death caused by lycopene in Candida albicans, J. Microbiol. Biotechnol. 17 (2007) 1797–1804. [36] F. Sakurai, R. Inoue, Y. Nishino, A. Okuda, O. Matsumoto, T. Taga, F. Yamashita, Y. Takakura, M. Hashida, Effect of DNA/liposome mixing ratio on the physico-chemical characteristics, cellular uptake and intracellular trafficking of plasmid DNA/cationic liposome complexes and subsequent gene expression, J. Control. Release 66 (2000) 255–269. [37] L.M. Bareford, P.W. Swaan, Endocytic mechanisms for targeted drug delivery, Adv. [38] E. Ojeda, G. Puras, M. Agirre, J. Zárate, S. Grijalvo, R. Pons, R. Eritja, G. Martinez-Navarrete, C. Soto-Sánchez, E. Fernandez, Niosomes based on synthetic cationic lipids for gene delivery: the influence of polar head-groups on the transfection efficiency in HEK-293, ARPE-19 and MSC-D1 cells, Org. Biomol. Chem. 13 (2015) 1068–1081. [39] A.V. Kuznetsova, A.M. Kurinov, M.A. Aleksandrova, Cell models to study regulation of cell transformation in pathologies of retinal pigment epithelium, J. Ophthalmol. 2014 (2014) 801787. [40] K. Aramaki, J. Yamada, Y. Tsukijima, T. Maehara, D. Aburano, Y. Sakanishi, K. Kitao, Formation of bilayer membrane and niosomes by doubletailed poly-glyceryl-type nonionic surfactant, Langmuir 31 (2015) 10664–10671. [41] S. Kachi, Y. Oshima, N. Esumi, M. Kachi, B. Rogers, D.J. Zack, P.A. Campochiaro, Nonviral ocular gene transfer, Gene Ther. 12 (2005) 843–851. [42] A. Liu, N. Pajkovic, Y. Pang, D. Zhu, B. Calamini, A.L. Mesecar, R.B. van Breemen, Absorption and subcellular localization of lycopene in human prostate cancer cells, Mol. Cancer Ther. 5 (2006) 2879–2885. [43] P. Palozza, R.E. Simone, A. Catalano, M.C. Mele, Tomato lycopene and lung cancer prevention: from experimental to human studies, Cancers (Basel) 3 (2011) 2333–2357. Chapter3 67 [44] Y. Sharoni, M. Danilenko, N. Dubi, A. Ben-Dor, J. Levy, Carotenoids and transcription, Arch. Biochem. Biophys. 430 (2004) 89–96. [45] J.G. Smith, T. Wedeking, J.H. Vernachio, H. Way, R.W. Niven, Characterization and in vivo testing of a heterogeneous cationic lipid-DNA formulation, Pharm. Res. 15 (1998) 1356–1363. [46] A. El-Sayed, H. Harashima, Endocytosis of gene delivery vectors: from clathrin-dependent to lipid raft-mediated endocytosis, Mol. Ther. 21 (2013) 1118– 1130. [47] A. Ferrari, V. Pellegrini, C. Arcangeli, A. Fittipaldi, M. Giacca, F. Beltram, Caveolae-mediated internalization of extracellular HIV-1 tat fusion proteins visualized in real time, Mol. Ther. 8 (2003) 284–294. [48] S. Xiang, X. Zhang, Cellular Uptake Mechanism of Non-viral Gene Delivery and Means for Improving Transfection Efficiency, INTECH Open Access Publisher, 2013. [49] K.I. Berns, The Parvoviruses, Springer Science & Business Media, 2013. [50] M. Almasieh, A.M. Wilson, B. Morquette, J.L. Cueva Vargas, A. Di Polo, The molecular basis of retinal ganglion cell death in glaucoma, Prog. Retin. Eye Res. 31 (2012) 152–181. [51] T.F. Martens, K. Remaut, H. Deschout, J.F. Engbersen, W.E. Hennink, M.J. van Steenbergen, J. Demeester, S.C. De Smedt, K. Braeckmans, Coating nanocarriers with hyaluronic acid facilitates intravitreal drug delivery for retinal gene therapy, J. Control. Release 202 (2015) 83–92. Chapter3 68 Chapter4 69 Chapter 4 Non-viral vectors based on cationic niosomes as efficient gene delivery vehicles to central nervous system cells into the brain Chapter4 70 Chapter4 71 Non-viral vectors based on cationic niosomes as efficient gene delivery vehicles to central nervous system cells into the brain Mohamed Mashala,1, Noha Attiaa,b,1, Cristina Soto-Sánchezc,d, Gema Martínez-Navarretec,d, Eduardo Fernándezc,d, Gustavo Purasa,c,⁎, José Luis Pedraza,c,⁎ a NanoBioCel Group, Laboratory of Pharmaceutics, School of Pharmacy, University of the Basque Country (UPV/EHU), Paseo de la Universidad 7, 01006 Vitoria-Gasteiz, Spain b Histology and Cell Biology Department, Faculty of Medicine, University of Alexandria, Alexandria, Egypt c Biomedical Research Networking Centre in Bioengineering, Biomaterials and Nanomedicine (CIBER-BBN), Spain d Neuroprothesis and Neuroengineering Research Group, Miguel Hernández University, Elche, Spain Abstract Development of safe and efficient non-viral vectors to deliver DNA into the CNS represents a huge challenge to face many neurological disorders. We elaborated niosomes based on DOTMA cationic lipid, lycopene “helper” lipid and polysorbate 60 as non-ionic surfactants for gene delivery to the CNS. Niosomes, and their corresponding nioplexes obtained after the addition of the pCMS-EGFP plasmid, were characterized in terms of size, charge, morphology and capacity to condense, release and protect DNA. In vitro experiments were performed in NT2 cells to evaluate transfection efficiency, viability, cellular uptake and intracellular distribution. Additionally, transfection in primary cortex cells were performed prior to brain administration into rat cerebral cortex. Data obtained showed that nioplexes exhibited not only adequate physicochemical properties for gene delivery applications, but also relevant transfection efficiencies (17%), without hampering viability (90%). Interestingly, In vivo experiments depicted promising protein expression in both cortical glial cells and blood vessels. Keywords: Central nervous system, Gene therapy, Non-viral vectors, Niosomes, Cationic lipids Chapter4 72 4.1.Introduction Gene therapy concept relies on the introduction of genetic material into target cells to modify protein expression for therapeutic purposes (Pezzoli et al., 2012). Over the time has emerged as a promising strategy for the treatment of many diseases. From a practical point of view, the nervous system transfection represents a huge challenge to address many devastating neurological disorders such as Parkinson or Alzheimer diseases (Nobre and Almeida, 2011) that are difficult to treat with traditional pharmacology approaches, mainly due to the brain physical barriers that drugs need to overcome after systemic administration and the complexity of the system (Nagabhushan Kalburgi et al., 2013). However, nowadays, gene therapy clinical trials for neurological disorders are still few in number, mainly due to the lack of safe and suitable approaches to deliver genetic material to targets cells (Yin et al., 2014). Non-viral vectors have received increasing attention thanks to their flattering properties such as; easy elaboration, low cytotoxicity, immune tolerance and lack of oncogenic effects. Additionally, it is possible to use larger DNA inserts (Mansouri et al., 2004; Tang and Szoka, 1997). Therefore, a notable shift of preclinical studies has occurred from viral to non-viral vectors in various applications, including neurological disorders (Peluffo et al., 2015). However, the expression of genes delivered via such vectors is typically low, which justifies the need for further research in this topic. Among non-viral vectors, niosomes have recently emerged as promising gene delivery systems (Attia et al., 2017; Ochoa et al., 2014; Ojeda et al., 2016c; Puras et al. 2014, 2015). Basically, niosomes are nonphospholipid liposome like vesicles, with a bilayer structure (Ojeda et al., 2016a). Compared to liposomes, niosomes are recognized for their low cost and superior chemical and storage stabilities (Rajera et al., 2011). The typical components of a niosome formulation include the cationic lipid, which electrostatically interacts with negatively charged genetic material to form complexes, known as nioplexes (Agirre et al., 2015). The “helper” lipid, which enhances physicochemical properties of the formulation and the intracellular disposition of the complexes (Ojeda et al., 2016b), and the non-ionic surfactants, which increase the stability of the formulation and avoid aggregation between vesicles (Huang et al., 2011). Among cationic lipids, DOTMA has been widely used to elaborate lipid nanoparticles for gene delivery applications, mainly Chapter4 73 due to its high aqueous solubility and its appealing chemical structure, which includes a polar head-group, a linker, a backbone and a hydrophobic domain (Rezaee et al., 2016). Lycopene is a natural carotenoid known for its potent antioxidant properties (Krishnamoorthy et al., 2013). Recently, has been used as “helper” lipid to enhance transfection efficiency in retina (Mashal et al., 2017). Polysorbates contain polyethylene glycol (PEG) chains that improve transfection efficiency of liposome formulations (Meyer et al., 1998). Therefore, polysorbates are classically incorporated into niosome for gene delivery applications as non-ionic surfactants. 4.2.Material and methods 4.2.1. Preparation of niosomes and nioplexes The reverse phase emulsification evaporation technique was used for elaboration of niosomes as previously described (Mashal et al., 2017). Briefly, the cationic lipid DOTMA (0.1% w/v, Avanti Polar Lipids Inc. Alabama, USA), non-ionic surfactant Polysorbate 60 (0.5% w/ v, Sigma Aldrich, Madrid, Spain) and helper lipid Lycopene (0.02% w/ v, Sigma Aldrich Madrid, Spain) were mixed at a molar ratio of 1/4/0.4 and dissolved in 1 ml of the organic solvent dichloromethane (Panreac, Barcelona, Spain). The chemical structure of such components can be observed on Fig. 1. Next, 5 ml miliQ water were added and the emulsion was sonicated for 30 s at 45 W (Branson Sonifier 250, Danbury, USA). After 2 h of magnetic evaporation, the organic phase was removed, leaving DP60L cationic niosomes suspended in the aqueous phase at a 1 mg/ml cationic lipid DOTMA concentration. Nioplexes were formed by mixing a stock solution of pCMS-EGFP plasmid (0.5 mg/ml) with niosome suspension. Under gentle pipetting, different cationic lipid/DNA (w/w) ratios were elaborated. The mixture was left for 30 min at room temperature. 4.2.2. Size and zeta potential measurement The hydrodynamic diameter and zeta potential of both niosomes and nioplexes, were measured by Dynamic Light Scattering (DLS) and Laser Doppler Velocimetry (LDV) using Zetasizer Nano ZS (Malvern Instrument, UK). Particle size reported as Chapter4 74 hydrodynamic diameter was obtained by Z-average. All measurements were carried out in triplicates. Only data that met the quality criteria were included in the study. 4.2.3. Cryo-TEM analysis The morphology of formulations was revealed by Cryo-TEM analysis and samples were examined by a TEM, TECNAI G2 20 TWIN (FEI), operating at an accelerating voltage of 200 KeV in a bright-field and low-dose image mode (Ojeda et al., 2015). Digital images were acquired with digital camera. Fig. 1. Chemical composition of DP60L cationic niosomes. (A) Cationic lipid DOTMA-Cl, (B) Polysorbate 60 and (C) Lycopene. 4.2.4. Agarose gel electrophoresis studies The ability of the niosomes to condense, release and protect plasmid DNA against enzymatic digestion was assayed by agarose gel electrophoresis. Nioplexes were analyzed at different cationic lipid/DNA (w/ w) ratios (200 ng of DNA/well). The agarose gel (0.8%, w/v) was immersed in a Tris-acetate-EDTA buffer and exposed for 30 min to 120 V. DNA bands were stained with GelRed® and observed with a ChemiDoc® MP Imaging System. SDS solution (3%) and DNase I enzyme (1 U DNase I/2.5 µg DNA) were added to the samples to evaluate the release and protection, Chapter4 75 respectively. The integrity of DNA in each sample was compared to a control of untreated DNA. 4.2.5. NT2 cell culture and in vitro transfection NT2 cells (ATCC®−CRL,1973) were cultivated in complete medium, Dulbecco’s Modified Eagle’s Medium (DMEM), enriched with 10% fetal bovine serum (FBS) and antibiotics (100 U/ml penicillin and 100 μg/ml streptomycin, Gibco®, Life Technologies S.A., Madrid, Spain). Before transfection, NT2 cells were seeded in 24well plates at an initial density of 8 ×104 cells/well and allowed to grow to 70–80% confluence. Then, the medium was replaced with serum-free Opti-MEM (Gibco®, Life Technologies S.A., Madrid, Spain), and cells were exposed to nioplexes (1.25 μg of pCMS-EGFP/well). After 4 h of incubation, transfection medium was removed and refreshed with complete medium. Cells were allowed to grow for 24 h until being analyzed by flow cytometry (FACSCalibur, BD, San Jose, USA). Positive control (Lipofectamine® 2000, Gibco®, Life Technologies S.A., Madrid, Spain) was prepared following the manufacturer’s protocol. 4.2.6. Cellular uptake and intracellular distribution studies NT2 cells were cultured as previously mentioned. The regular growth media was removed and cells were exposed to nioplexes (prepared with FITC-labeled pCMSEGFP plasmids, DareBio. Madrid, Spain). After 4 h of incubation, the transfection medium was removed and cells were washed with PBS, trypsinized and analyzed by FACSCalibur flow cytometer. 10.000 events were collected and analyzed for each sample. Each sample was analyzed in triplicate. To determine the intracellular distribution of internalized nioplexes, cells were seeded on coverslips (24-well plates) and exposed to the nioplexes. The intracellular distribution of nioplexes was analyzed by confocal laser scanning microscopy (CLSM; Olympus Fluoview 500). 4.2.7. Endocytosis mechanism Chapter4 82 Fig. 4. Cellular uptake studies in NT2 cells 4 h post incubation. (A) Flow cytometry dot-plots (SSC-H and FL1) of control cells without treatment (A1), naked DNA (A2), DP60L nioplexes at 14/1 (w/w) ratio (A3), and lipofectamine 2000® (L2K) at 2/1(w/w) ratio (A4). (B) Flow cytometry measurement of NT2 cells treated with FITC-labeled formulations. Error bars represent SD (n = 3). *p < 0.05. (C) Fluorescence microscopy images of NT2 cells after 4 h of incubation with FITC-labeled DP60L nioplexes at 14/1 (w/w) ratio. Cells were stained with DAPI-fluoromount G (blue). White arrows indicate nanoparticles around the nucleus. (Scale bar =25 μm). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) Chapter4 83 Fig. 5. Internalization studies. (A) Confocal microscopy images showing intracellular distribution in NT2 cells of nioplexes at 14/1 ratio labelled with FITC-pCMS EGFP (green) plasmid and AlexaFluor 555-Cholera Toxin (A1), AlexaFluor 546-Transferrin (A2), AlexaFluor 594-Dextran (A3), and LysoTracker Red DND-99 markers (A4), all in red. (M1 = Mandeŕs overlap coefficient). Scale bars =25 µm. (B) Effect of endocytic inhibitors on the cellular uptake of FITC-labeled nioplexes. Data were normalized to uptake without inhibitor. (C) Agarose gel electrophoresis DNA released profiles of lipoplexes and nioplexes incubated with or without PS. (D) Quantification of released DNA in agarose gel electrophoresis. mentioned endocytosis mechanisms were simultaneously inhibited by prior treatment with methyl cyclodextrin (28%). However, MPC inhibition by wortmannin, as an inhibitor of MPC, had less effect on cellular uptake reduction, since uptake values where higher, around 80%. The capacity of DNA compacted with both DP60L niosomes and lipofectamine® 2000 to escape from the late endosome, simulated by PS micelles (167 nm, and −65 mV), was analyzed by agarose gel electrophoresis assay (Fig. 5-C). As noticed in Fig. 5-D, in the presence of PS micelles, the amount of DNA released from lipoplexes was around 20%, while in the case of nioplexes, approximately, double amount of DNA, 40%, was released. Without the previous incubation with the anionic PS micelles, the released DNA was almost 0% in the case Chapter4 84 of lipoplexes, and about 30% with DP60L nioplexes. Additionally, the pH titration curve (Supplementary data, Fig. 1) provided evidence of the buffering capacity of DP60L formulation, since pH value, slightly decreased from 4.1 to 1.7, after the addition of 1.400 µL of HCl 0.1 M. 4.1.1. Primary cortical neuron and in vivo gene expression studies Fig. 6. Primary cortical culture and in vivo gene expression of EGFP carried by DP60 nioplexes at 14/1 mass ratio (w/w). A1 and A2, transfection of primary neuronal cell cultures 24 h post transfection. NeuN-positive neurons (red) and nuclei counterstained with Hoechst 33342 (blue) (Scale bars =20 and 40 μm, respectively). B, in vivo gene expression of pCMS-EGFP 72 h after intracortical administration of nioplexes. Nuclei are shown in blue (Hoechst), neurons in red (NeuN+), and EGFP expression (GFP+) in green. White arrow points a blood vessel. Scale bar 40 μm. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) Chapter4 85 EGFP-expressing neural cell morphology were discerned on primary cortical cultures transfected with DP60L nioplexes at 14/1 mass ratio (w/w) in Fig. 6-A1. However, none of them were NeuN+ (Fig. 6-A2). 72 h after mice intracranial injection of nioplexes, immunochemistry of the cryopreserved sections showed NeuN− (non-neuronal) cells with glial morphology, expressing EGFP throughout dendritic processes. Interestingly, cells in the wall of some vessels were EGFP+ as well. 4.4.Discussion Gene therapy directed to the CNS represents a great challenge to bring normal gene copies and correct mutant gene deficiencies in many neurodegenerative diseases (Maguire et al., 2014). However, in order to apply this promising technology into the regular clinic practice, safe and efficient gene carrier systems are needed. Compared with viral vector, non-viral vectors have several important advantages. For instance, they are easier and cheaper to produce, and there is no preexisting immunity to these vectors. Additionally, they are not derived from pathogens, and consequently show less safety concerns. However, their transfection efficiency is lower than their virus-based counterparts. Therefore, research on this topic merits special attention. Based on the flattering properties showed recently by niosomes to transfect retinal cells after both intravitreal and subretinal administration (Mashal et al., 2017; Puras et al., 2014; Ochoa et al., 2014), we elaborated cationic niosomes based on DOTMA, polysorbate 60 and lycopene to deliver DNA into the brain cortex of rats after cerebral cortex administration. DOTMA is a highly water-soluble quaternary ammonium salt lipid that has been extensively used for gene delivery purposes, due to its ability to condense DNA (Zhang et al., 2012). Lycopene is a natural carotenoid that when used as “helper” lipid, enhances transfection efficiency in retinal cells (Mashal et al., 2017). Polysorbate 60 is the major component of the niosome formulation, and its poly(ethylene glycol) structure has been reported that increases in vivo transfection efficiency in CNS (Tang et al., 2003; Yao et al., 2012). Prior to perform any biological study, we characterized in Fig. 2 our formulation in terms of size, superficial charge, morphology and capacity to condense, release and protect DNA against enzymatic digestion. Although there is no a general rule about the Chapter4 86 optimal particle size of formulations for each gene delivery application, it is generally accepted that this parameter clearly influences the final performance, and that in case of niosomes, is affected by the cationic lipid/DNA mass ratio (Puras et al., 2015). Size of DP60L niosomes was around 100 nm. Interestingly, when DNA was incorporated to obtain nioplexes at 6/1 ratio, size increased up to 154 nm. However, at higher ratios, size decreased, probably due to the electrostatic interactions that condense DNA more efficiently, and therefore, decreased PDI values as well. Additionally, differences observed in PDI values could be due to different DNA topologies found in nioplexes (Cherng et al., 1999). Regarding ZP values, we found a positive correlation with the cationic lipid/DNA ratio, which suggests that cationic DOTMA binds and neutralize the negatively charged DNA (Paecharoenchai et al., 2012). Under TEM observation, niosomes appeared spherical, while nioplexes at 14/1 mass ratio showed heterogeneous shapes, but mostly elongated. Interestingly, at this ratio, spherical niosomes were also discerned (Fig. 2-B2), which could accumulate on the surface of cell membranes and enhance DNA delivery (Smith et al., 1998; Song and Liu, 1998). To further analyze electrostatic interactions, we performed a gel retardation assay (Fig. 2-C), since an optimum balance is required for efficient gene delivery (Paecharoenchai et al., 2012). Despite the incomplete DNA condensation observed, which could be affect by different thermodynamic factors, kinetics mixing, or by the lycopene incorporation in to the niosome formulation or the high aqueous solubility of DOTMA (Mahato, 2005), all cationic lipid/DNA ratios analyzed, were able to protect plasmid DNA against enzymatic digestion. Once physicochemical properties of nioplexes were analyzed, next, we performed in vitro studies to evaluate, initially, both transfection efficiency and viability in NT2 cells. These cells represent an interesting model to study the efficiency of gene delivery vectors into CNS due to their capacity to differentiate into bot neuronal and glial cells (Agirre et al., 2015). Unlike other teratocarcinoma cell lines, the NT2 cells depict an exclusive commitment to a neural lineage when exposed to retinoic acid (RA). Therefore, it has been considered as a promising human cell source in studies of cell in vivo therapeutic applications in many neurodegenerative diseases such as Parkinson disease (Cacciotti et al., 2017). Accordingly, we considered NT2 cells as an interesting Chapter4 87 model to study the efficiency of gene delivery vectors into CNS. Additionally, these cells represent a promising platform in cell-based gene delivery as they could be genetically-modified and then transplanted (Tinsley and Eriksson, 2004). Data obtained in Fig. 3 revealed that transfection efficiency increased in proportion to the cationic lipid/DNA mass ratio, reaching the peak at 14/1 ratio (17% of cells were transfected). Although the percentage of transfected cells were inferior to those obtained with commercially available lipifectamine®2000 (17% and 37%, respectively), viability values were higher (90% and 83%, respectively). The reported low cytotoxicity of nioplexes represents an appealing feature for potential further in vivo applications, since the in vivo use of lipifectamine®2000 is discouraged due to its cytotoxicity (Yang et al., 2014). To better understand the transfection process mediated by DP60L nioplexes in NT2 cells, we evaluated the cellular uptake and the intracellular trafficking of nioplexes, since those two factors clearly influence on the final performance of gene delivery carriers (Puras et al., 2015). Cellular uptake values were compared with lipifectamine®2000. As observed in Fig. 4, lipoplexes obtained with commercially available lipofectamine®2000, were more efficiently internalized than DP60L nioplexes (around 85% and 60%, respectively), which could explain the higher percentage of EGFP expression in NT2 cells transfected with lipofectamine®2000. Differences observed in cellular uptake between both formulations could be due to particular topologies of complexes or specific interactions of the complexes with cell membrane lipids (Cherng et al., 1999). In any case, percentage of transfected cells with DP60L nioplexes (17%, Fig. 3) were clearly inferior to the percentage of positive cells for FITC-labeled DP60L nioplexes (60%, Fig. 4-B), which suggest the influence of other biological events, such as intracellular trafficking or endosomal scape, in the final EGFP expression (Cardarelli et al., 2016). Therefore, we analyzed three of the most employed cellular internalization pathways such as clathrin-mediated endocytosis (CME), caveolae-mediated endocytosis (CvME) and macropinocytosis (MPC). Although there is not a clear consensus regarding the most efficient endocytosis pathway, the release of DNA into the cytoplasm, and therefore, the final performance, is clearly affected by the cellular uptake process (Nam et al., 2009). The results observed in Fig. 5 suggested that DP60L nioplexes were internalized, mainly, by CvME and CME, while MPC had Chapter4 88 much less participation in the cellular uptake process. It is generally accepted that both CvME and CME are endocytosis routes that transfer genetic material to late endosomes/lysosomes, where the acidic environment degrades the DNA, making transfection process inefficient (Agirre et al., 2015). Therefore, the observation of nioplexes in the late endosome (Fig. 5-A4) might explain the relatively low transfection efficiency values (17%) observed (Fig. 3), despite the fact that high number of NT2 cells (60%) captured the complexes (Fig. 4). Nonetheless, this hypothesis needs further verification since the endosomal escape capacity of nioplexes, if present, could evade degradation. Therefore, in next experiments, we evaluated the ability of nioplexes to escape from degradation in lysosomes. Several particle dependent endosomal escape mechanisms have been reported in the literature, being the proton-sponge mechanism one of the most widely described (Varkouhi et al., 2011). Therefore, we analyzed the pH-buffering capacity of cationic DP60L niosomes. As observed (Supplementary data, Fig. 1), the incorporation of both cationic DOTMA and lycopene lipids into the niosome formulation, increased the pH-buffering capacity upon titration with 0.1 M HCl compared with niosomes elaborated only with polysorbate 60, which could suggest that those lipids could increase the proton sponge effect, and therefore, the endosomal scape capacity of DP60L niosomes. Another widely proposed endosomal scape mechanism consists on the destabilization of the endosomal membrane by electrostatic interactions between the cationic nanoparticles and the anionic lipids of the late endosome membrane, which could allow the DNA release to the cytoplasm (Varkouhi et al., 2011). To evaluate this endosomal scape mechanism, we added DP60L nioplexes to anionic micelles made with phosphatidyl serine (PS), that simulated the endosomal compartment, and the DNA release from PS micelles was evaluated in an agarose gel electrophoresis assay (Agirre et al., 2015). As observed in Fig. 5, about 40% of DNA was released form DP60L nioplexes, in the presence of PS, micelles. However, only 30% of DNA was released without previous incubation with the anionic PS micelles, which demonstrates the capacity of the DP60L nioplexes to release DNA once they contact the endosomal lipid bilayer membrane. In the case of lipoplexes based on lipofectamine®2000, all DNA was condensed with the formulation and up to 20% was released in the presence of PS micelles. Chapter4 89 Next, and prior to perform in vivo studies, we evaluated transfection efficiency of DP60L nioplexes in primary cortical cultures of rat embryos, since primary cells normally express their tissue-specific receptors, and mimic in vivo conditions, where different kind of neurons and glial cells are mixed to set up neuronal-glial networks. In these conditions, we observed that, apparently, only glial cells, expressed EGFP (Fig. 6-A1). This assumption was further confirmed by lack of NeuN+-immunoreactivity in EGFP expressing cells (Fig. 6-A2). Such preferential transfection of glial cells could be attributed to their higher mitotic and/or phagocytic activities (Schafer and Stevens, 2013). Interestingly, in in vivo experiments performed by direct intracranial injection of nioplexes, again, NeuN negative cells (neuroglia and cells in blood vessel wall) were the only ones transfected by DP60L nioplexes (Fig. 6-B). These results reveal the incapacity of nioplexes to transfect neuron cells, probably due to the impaired uptake and/or intracellular trafficking (Bergen et al., 2008). Therefore, DP60L nioplexes could be of great interest to transfect glial cells in the CNS in glia-related neurological disorders. Glial cells constitute over 70% of the total cell population in the CNS, and they play a pivotal role for the normal development and function of nervous tissue (Alvarez et al., 2013; Fields and Stevens-Graham, 2002). Their perturbation is associated with several neurological disorders such as; stroke, multiple sclerosis, epilepsy, Alzheimer’s and Parkinson’s diseases (Barres, 2008; Milligan and Watkins, 2009). Therefore, the preferential transfection of glial cells could be of great importance in future applications in glia-related neurological disorders. Additionally, cells in the wall of some blood vessels were also transfected with DP60L nioplexes (Fig. 6-B, white arrow). Transfection at this level could be of great relevance in cerebrovascular diseases, such as; stroke, transient ischemic attacks, subarachnoid hemorrhage or vascular dementia, just to name a few. 4.5.Conclusion In summary, we conclude that non-viral vector formulations based on niosome nanoparticles, where DOTMA is the cationic lipid, lycopene the “helper” lipid and polysorbate 60 the non-ionic surfactant, presents suitable physicochemical properties Chapter4 90 for gene delivery applications in terms of size, superficial charge, polydispersity, or capacity to protect genetic material against enzymatic digestion. In addition, such formulation was able to transfect efficiently NT2 cultured cells, were both clathrin and caveolae-mediated endocytosis pathways predominated over macropinocytosis, exhibiting endosomal scape properties that could explain the high protein expression levels observed. Promising results obtained in both primary cortical cultures of rat embryos and in in vivo conditions after intracranial injection open the door for future application of such niosomes as efficient gene delivery tools for therapeutic treatment of some degenerative as well as malignant CNS disorders. 4.6.Acknowledgements This project was supported by the Basque Country Government (GIC15/85), Spanish Grant MAT 2015 -69976-C3-1, SAF2013-42347-R, and by Research Chair “Bidons Egara”. The authors also wish to thank the intellectual and technical assistance from the ICTS “NANBIOSIS”, more specifically by the Drug Formulation Unit (U10) of the CIBER in Bioengineering, Biomaterials, and Nanomedicine (CIBER-BBN) at the University of Basque Country (UPV/EHU). Technical and human support provided by SGIker (UPV/EHU) is gratefully acknowledged 4.7. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.ijpharm.2018.09.038. Chapter4 91 Supplementary figure 1. pH buffering capacity of assay of both DP60L and P60 niosomes. 4.8.References Agirre, M., Ojeda, E., Zarate, J., Puras, G., Grijalvo, S., Eritja, R., Garcia del Cano, G., Barrondo, S., Gonzalez-Burguera, I., Lopez de Jesus, M., Salles, J., Pedraz, J.L., 2015. New insights into gene delivery to human neuronal precursor NT2 cells: a comparative study between lipoplexes, nioplexes, and polyplexes. Mol. Pharm. 12, 4056–4066. https://doi.org/10.1021/acs.molpharmaceut.5b00496. Alvarez, J.I., Katayama, T., Prat, A., 2013. Glial influence on the blood brain barrier. Glia 61, 1939–1958. https://doi.org/10.1002/glia.22575. Attia, N., Mashal, M., Grijalvo, S., Eritja, R., Zarate, J., Puras, G., Pedraz, J.L., 2017. Stem cell-based gene delivery mediated by cationic niosomes for bone regeneration. Nanomedicine S1549-9634(17)30200-9 [pii]. Barres, B.A., 2008. The mystery and magic of glia: a perspective on their roles in health and disease. Neuron 60, 430–440. https://doi.org/10.1016/j.neuron.2008.10.013. Bergen, J.M., Park, I.K., Horner, P.J., Pun, S.H., 2008. Nonviral approaches for neuronal delivery of nucleic acids. Pharm. Res. 25, 983–998. https://doi.org/10.1007/s11095007-9439-5. Cacciotti, I., Ceci, C., Bianco, A., Pistritto, G., 2017. Neuro-differentiated Ntera2 cancer stem cells encapsulated in alginate beads: first evidence of biological Chapter 5 98 Chapter 5 99 Gene delivery to the rat retina by non-viral vectors based on chloroquinecontaining cationic niosomes Mohamed Mashala,1, Noha Attiaa,b,c,1, Gema Martínez-Navarreted,e, Cristina Soto-Sánchezd,e, Eduardo Fernándezd,e, Santiago Grijalvod,f, Ramón Eritjad,f, Gustavo Purasa,d, , Jose Luis Pedraza,d, a NanoBioCel Group, Laboratory of Pharmaceutics, School of Pharmacy, University of the Basque Country (UPV/EHU), Paseo de la Universidad 7, 01006 Vitoria-Gasteiz, Spain b Histology and Cell Biology Department, Faculty of Medicine, University of Alexandria, Alexandria, Egypt c Department of Basic Sciences, The American University of Antigua-College of Medicine, Coolidge, Antigua and Barbuda d Networking Research Centre of Bioengineering, Biomaterials and Nanomedicine (CIBER-BBN), Vitoria-Gasteiz, Spain e Neuroprothesis and Neuroengineering Research Group, Miguel Hernández University, Elche, Spain f Institute of Advanced Chemistry of Catalonia (IQAC-CSIC), Spain Journal of Controlled Release 304 (2019) 181–190 ABSTRACT The incorporation of chloroquine within nano formulations, rather than as a cotreatment of the cells, could open a new avenue for in vivo retinal gene delivery. In this manuscript, we evaluated the incorporation of chloroquine diphosphate into the cationic niosome formulation composed of poloxamer 188, polysorbate 80 non-ionic surfactants, and 2,3-di (tetradecyloxy) propan-1-amine (hydrochloride salt) cationic lipid, to transfect rat retina. Niosome formulations without and with chloroquine diphosphate (DPP80, and DPP80-CQ, respectively) were prepared by the reverse phase evaporation technique and characterized in terms of size, PDI, zeta potential, and morphology. After the incorporation of the pCMS-EGFP plasmid, the resultant nioplexes -at different cationic lipid/DNA mass ratioswere further evaluated to compact, liberate, and secure the DNA against enzymatic digestion. In vitro procedures were achieved in ARPE-19 cells to assess transfection efficacy and intracellular transportation. Both nioplexes formulations transfected efficiently ARPE-19 cells, although the cell viability was clearly better in the case of DPP80-CQ nioplexes. After subretinal and intravitreal injections, DPP80 nioplexes were not able to transfect the rat retina. However, chloroquine containing vector showed protein expression in many retinal cells, depending on the administration route. These data provide new insights for retinal gene delivery based on chloroquine-containing niosome non-viral vectors. Keywords: Niosomes, Non-viral vector, Gene therapy, Retina, Chloroquine Chapter 5 100 5.1.Introduction Retinal degeneration is a devastating ocular pathology caused by functional impairment of genes related mainly to the phototransduction process, the structure and metabolism of the retinal cells, and the maturation process of the mRNA needed to synthetize specific proteins. One of the most promising alternatives to treat retinal disorders like age-related macular degeneration [1], Leber congenital amaurosis (LCA) [2] retinitis pigmentosa [3] or choroideremia consists on the delivery of a normal copy of the mutated genes to the affected cells by means of gene therapy technology [4]. Since the success of first RPE65-gene-replacement trials for LCA type-2, further clinical trials of gene therapy have been conducted for other devastating retinal disorders [5]. In most of those clinical trials, viral-vectors have been used to deliver the genetic material. Among them, adeno-associated virus (AAV) stand out for their safety profile [6]. In fact, Luxturna, the first gene therapy-based medicine approved by the FDA in 2017 for the treatment of mutations in RPE65 gene linked to retinitis pigmentosa and Stargardt disease, is based on such AAV. However, the limited carrying capacity of genetic material, around 5 kb, hampers their ap-plication to deliver genes that over pass such size to the retina. For instance, ABCA4 and MYO7A genes, whose mutations can be related to Stargardt disease and Usher Syndrome Type 1B, respectively [7], have a size of around 7 kb. Therefore, the use of non-viral vectors represents an interesting alternative, since the size of DNA that can be inserted in some of these formulations is theoretically unlimited [8,9]. Despite their limited transfection efficiency and transient gene ex-pression, nonviral vectors have emerged as a promising alternative to deliver genetic material. Some of the main advantages of such gene delivery systems, in addition to their higher carrying capacity, include their low cost of production or their capacity to be easily modified in order to enhance their performance [10,11]. Hence, the research ac-tivity related to the design and characterization of novel non-viral vector formulations for gene delivery has considerably increased [12]. Cationic niosomes are self-assembled vesicular nanovehicles, similar to liposomes, with encouraging properties for gene delivery applications. To mention a few, the chemical structure of niosomes makes it possible to provide more stable and less cytotoxic formulations at a low cost [13]. The Chapter 5 101 amphiphilic nature of non-ionic surfactants enable niosomes to trap both hydrophobic and hydrophilic compounds [14].The cationic part here is the hydrochloride salt of the cationic lipid 2,3-di (tetra-decyloxy) propan-1-amine (D). Such cationic lipid contains the four pivotal components that manage the gene transfection process: a polar head, a backbone, a linker, and two non-polar tails [15]. One of the key limiting steps in the transfection process is the endosomal escape. Chloroquine is a known endosomal disrupting molecule and lysosomotropic agent that can cross the blood retinal barrier. Although chloroquine has been used in vitro as a pre-treatment of cultured cells to facilitate gene delivery [16], this study will be the first -to the best of our knowledgeto apply a chloroquine-containing nio-some formulation in gene delivery setting. The incorporation of one or more of the materials at the molecular level, within the nano-formulation, can dramatically affect the transfection process under in vitro and in vivo conditions [12]. Thereafter, scientists may face a great challenge in the near future to test a library of different materials that can be incorporated within the gene delivery vehicles. Based on the aforementioned (D) cationic lipid, two niosome vehicles were formulated for retinal gene delivery with two non-ionic surfactants [polysorbate 80 (P80) and poloxamer 188 (P)], in the absence/presence of chloroquine (CQ), referred as DPP80 and DPP80-CQ, respectively (Fig. 1). The two vehicles were prepared by the emulsification/solvent evaporation system and characterized in terms of particle size, polydispersity index (PDI) and zeta potential. Then, the reporter pCMS-EGFP plasmid was added at different weight ratios of cationic lipid to obtain nioplexes. Such DPP80/DPP80-CQ nioplexes were further characterized by size, PDI, charge, morphology, and the capability to compact, liberate and protect the DNA from digestive enzymes. In vitro comparative studies of both vehicles in ARPE-19 cells were achieved respecting their cellular uptake, transfection efficiency, viability and internalization mechanism. Finally, the two formulations were administered to rat eyes via intravitreal and subretinal injections, as a probe of concept, to estimate transfection efficiency by confocal microscopy. 5.2.Materials and methods Chapter 5 102 5.2.1. Production of cationic niosomes The synthesis of the hydrochloride salt form of the cationic lipid 2,3-di (tetradecyloxy) propan-1-amine (D) was performed as described in the literature, with slight modifications of the laboratory protocol [17]. Niosomes were composed by modified reverse-phase evaporation approach [18]. Concisely, 5 mg (0.1% w/v) of the lipid was dispersed in 1 ml of dichloromethane (organic phase). Subsequently, 5 ml milliQ water containing 12.5 mg (0.25% w/v) poloxamer 188 (P) (Sigma-Aldrich, Madrid, Spain), 12.5 mg (0.25% w/v) polysorbate 80 (P80) (Sigma-Aldrich, Madrid, Spain) and 2.5 mg (0.05% w/v) chloroquine diphosphate (CQ) (Sigma-Aldrich, Madrid, Spain) were added to the organic phase. Fig. 1. Chemical structure of the components of DPP80 and DPP80-CQ niosomes. (A) Polysorbate80, (B) Chloroquine diphosphate salt, (C) Poloxamer188 and (D) cationic lipid (DTPA-Cl). The emulsions were obtained by sonication of the mixture for 50 s at 45 W (Branson Sonifier 250®, Danbury, USA). Di-chloromethane was eliminated from emulsions by dissipation under magnetic stirring for 2 h, rendering the cationic niosomes in the aqueous medium. The molar ratios of both DPP80 and DPP80-CQ formulations were, 1.9/0.3/1.9 and 1.9/0.3/1.9/1, respectively. Chapter 5 103 5.2.2. Plasmid propagation and elaboration of nioplexes The protocols for propagation, purification, and quantification of pCMS-EGFP plasmid (5541 bp, Plasmid Factory, Bielefeld, Germany), have been described previously [12]. The nioplexes (niosome/DNA complexes) of both DPP80 and DPP80CQ were formed by mixing an adequate volume of pCMS-EGFP plasmid stock solution (0.5 mg/ml) with various amounts of the niosome suspension (1 mg cationic lipid/ ml) to get different cationic lipid/DNA mass ratios (w/w). To enhance the electrostatic interaction, the nioplexes mixture was allowed to settle for 30 min at room temperature. 5.2.3. Characterization of niosomes/nioplexes Dynamic light scattering (DLS) technique was used to estimate both particle size and polydispersity index (PDI) measurements (Malvern Zetasizer Nano ZS, UK). Particle size, was determined by cumulative analysis of the recorded hydrodynamic diameter. Laser Doppler Velocimetry (LDV) was used to assess zeta potential (ZP). Samples were dispersed in a 0.1 mM NaCl solution. Triple measurements were carried out for all samples. The morphology of both niosomes and nioplexes was estimated by transmission electron microscopy (TEM). Shortly, onto glow-discharged carbon coated grids, 5 μl sample was allowed to adhere on the surface for 60 s. Samples were examined under TEM, Tecnai G2 20 Twin (FEI, Eindhoven, The Netherlands). In a bright-field image mode, the operation was done with an accelerating voltage of 200 keV. Digital images were captured by an Olympus SIS Morada di-gital camera. Niosomes' ability to compact, liberate and safeguard DNA from enzymatic digestion was assessed by agarose gel electrophoresis studies. Nioplexes samples (200 mg of plasmid DNA/20 μl) were com-pared to naked (uncomplexed) DNA. The agarose gel (0.8% w/v) was immersed in a Tris–acetate–EDTA buffer, and the DNA samples were run on the gel for 30 min at 120 V. Next, agarose gel was stained with GelGreen®. A digital ChemiDoc™ MP Imaging System (Bio-Rad, Madrid, Spain) was used for band observation. 20 μl of a 2% SDS solution (Sigma-Aldrich, Madrid, Spain) was added to the samples to estimate the liberation of DNA from both DPP80 and DPP80-CQ vehicles at different cationic lipid/DN mass ratios. In addition, DNase I (Sigma-Aldrich, Madrid, Spain) was added at a concentration of 1 unit of DNase I/2.5 μg DNA to Chapter 5 104 evaluate the protective ability of both vehicles against enzymatic digestion. Then, the samples were incubated at 37 °C for 30 min and a 2% SDS solution was added to evaluate if released the DNA from the vehicles wad protected from the enzymatic digestion. 5.2.4. In vitro transfection experiments ARPE-19 cells, purchased from the American Type Culture Collection (ATCC, CRL 2302®), were seeded in 24-well plates at a density of 8 × 104 cells/well, with 500 μl of complete medium, formed of D-MEM/F-12 containing 10% fetal bovine serum (Gibco®, California, USA). Then, at a confluence level of 70–80%, the media was removed, and cells were exposed to nioplexes at different cationic lipid/DNA mass ratios (w/w) (1.25 μg DNA/well) dispersed in serum free Opti-MEM® solution (Gibco®, California, USA) at 37 °C for 4 h. Subsequently, transfection medium was removed, and cells were thoroughly washed 3 times with PBS. Then, cells were cultured in 1 ml of complete medium and allowed to grow for further 72 h until fluorescence microscopy imaging (Nikon TSM) and flow cytometry analysis (FACSCalibur™, BD Biosciences, USA) were done. FL1 (530/30) was used to detect EGFP-expressing transfected cells, and FL3 (670) was used to detect Propidium Iodidestained dead/dying cells. Untransfected cells were used as negative control for all experiments, while cells transfected with Lipofectamine™ 2000 (Invitrogen, California, USA), according to manufacturer's protocol, were considered as positive controls. 10.000-gated events were acquired and analyzed using Flowing Software 2.5.1. Data represent the mean (± SD) of three independent experiments, each of them performed in triplicate. 5.2.5. Cellular uptake and endocytosis mechanism studies FITC-labeled (pCMS-EGFP) plasmid (DareBio, Madrid, Spain) was used instead of pCMS-EGFP plasmid to estimate the cellular uptake of the vehicles. The same protocol described in the previous 5.2.4 section, was used to incubate and maintain ARPE-19 cells, and to evaluate cellular uptake. After removal of the transfection medium and multiple washes of the plates with PBS, the cells were assayed Chapter 5 105 by FACSCalibur flow cytometer. The negative control cells were transfected with naked DNA, and the percentage of FITC-positive cells represented the cellular uptake values. Each specimen was assayed in triplicate. Different up-take inhibitors were used to estimate the endocytosis mechanism of vehicles. Genistein, chlorpromazine hydrochloride, methyl-β-cyclo-dextrin and wortmannin were used as inhibitors for caveolae-mediated endocytosis (CvME), clathrin-mediated endocytosis (CME), both (CvME and CME) and macropinocytosis (MPC), respectively. Nioplexes at 10/1 cationic lipid/DNA mass ratio were complexed with pCMS-EGFP plasmid, and followed the same protocol described in the previous 5.2.4 section to transfect ARPE19 cells. Prior to the addition of nioplexes, cells were incubated with either 200 μM genistein for 30 min, or with 5 mM methyl-β-cyclodextrin, 50 nM wortmannin, or with 5 μg/ml chlorpromazine hydrochloride for 60 min. Cells were incubated with serumfree Opti-MEM® solution for 4 h at 37 °C. Subsequently, cells were carefully washed with PBS after removal of the transfection medium. Then, complete medium was added, and cells were incubated to grow for a further 72 h until flow cytometer analysis was done to determine the transfection efficiency. Each specimen was analyzed in triplicate. 5.2.6. Buffering capacity of niosomes The buffering capacity of both DPP80 and DPP80-CQ niosomes was assayed by volumetric analysis. Briefly, 10 ml formulation samples were titrated with aliquots of 100 μl 0.1 M HCl solution, and the changes in pH value were monitored by a pH meter (CRISON, GLP 21, Barcelona, Spain). 5.2.7. In vivo studies Intravitreal (4 μl containing 100 ng of pDNA) and subretinal injection (1 μl containing 25 ng of pDNA) of both DPP80 and DPP80-CQ nioplexes suspension were performed into four adult female Sprague–Dawley rats (6–7 weeks old, 200–300 g weight) per formulation. Experiments were done according to the Spanish and European Union regulations for the use of animals in research and the Association for Research in Vision and Ophthalmology (ARVO) statement, as de-scribed in the Chapter 5 106 literature [12]. To deliver nioplexes to the subretinal space, a bent 33-gauge needle was introduced through a sclerotomy (1–2 mm) posterior to ora serrata and in a tangential direction toward the posterior retinal pole along the subretinal space. Successful administration was confirmed by the appearance of a partial retinal detachment by direct ophthalmoscopy of the eye fundus through the operating microscope (Zeiss OPMI® pico; Carl Zeiss Meditec GmbH, Jena, Germany). The untreated right eyes were injected only with the vehicles and served as negative controls. Rats were sacrificed and perfused with 4% paraformaldehyde (PFA) after 72 h and eyes were removed, opened at the cornea and immersed in PFA. For whole mounts, retina was dissected from the eyecup and flattened onto Superfrost glass slides (Superfrost Plus, Fisher Scientific). For cryosections, the eyes were cryoprotected in sucrose and embedded in Tissue-Tek® OCT (optimum cutting temperature). The eyes were cryosectioned at 16 μm and transferred directly onto microscope slides (Superfrost, Fisher Scientific). For immunohistochemistry, whole mounts or retinal sections were washed and blocked with 10% bovine serum albumin and 0,05% triton in PBS for 1 h (cryosections) or 2 h (whole mounts). Both sections and whole mounts were incubated overnight at 4 °C with primary antibodies: rabbit anti-NeuN (Merck Millipore, MA, USA), rabbit anti-recoverin (Merck Millipore, MA, USA), rabbit anti-Protein kinase C (PKC, Santa Cruz Biotechnology) and rabbit anti-GFAP (Santa Cruz Biotechnology). Samples were rinsed and incubated with Alexa Fluor 555 donkey anti rabbit (Thermofisher Scientific) and counterstained with Hoechst 33342 (Thermofisher Scientific). Finally, whole mounts and sections were mounted with antifade mounting meédium and evaluated with a Leica TCS SPE spectral confocal microscope (Leica Microsystems GmbH, Wetzlar, Germany). 5.2.8. Statistical analysis INSTAT program (GraphPad Software, San Diego, CA, USA) was used to perform the statistical analysis. Differences between groups at significance levels of Chapter 5 107 95% were calculated by the ANOVA and the Student's t-test. In all cases, P values < .05 were regarded as significant. Normal distribution of samples was assessed by the Kolmogorov-Smirnov test and the homogeneity of the variance by the Levene test. Numerical data were presented as mean ± SD. Table 1 Physical features of both DPP80 and DPP80-CQ niosomes in terms of size (nm), Polydispersity index (PDI), and zeta potential (mV). The values exemplify the mean ± SD (n = 3). Size (nm) Zeta potential (mV) PDI DPP80 90.41 ± 0.65 44.3 ± 1.48 0.42 ± 0.01 DPP80-CQ 118.18 ± 1.46 28.9 ± 7.73 0.13 ± 0.02 5.3.Results 5.3.1. Characterization of niosomes/nioplexes Both niosome vehicles were prepared by mixing [polysorbate 80 (P80) and poloxamer 188 (P)] non-ionic surfactants and cationic lipid (D). In the absence/presence of chloroquine (CQ), niosomes were referred as DPP80, or as DPP80CQ, respectively (Fig. 1). Both niosomes were prepared by the emulsification/solvent evaporation method and were characterized in terms of particle size, zeta potential (ZP) and polydispersity index (PDI) as shown in Table 1. The incorporation of chloroquine into the DPP80 niosome formulation increased the size of those niosomes from 90 to118 nm. Moreover, upon chloroquine addition, the ZP values decreased remarkably to 29 mV in DDP80-CQ niosomes compared to 44 mV in DPP80 niosomes. Interestingly, the addition of chloroquine also decreased the PDI value from 0.42 in DPP80 formulation to 0.13 in DPP80-CQ formulation. Fig. 2 illustrates the physicochemical characterization of DPP80 and DPP80-CQ nioplexes. In Section 2-A, the size and ZP values of both nioplexes at different ratios Chapter 5 114 to polycation-DNA complexes enhanced the expression level of the delivered genes in both in vitro and in vivo conditions at doses below the known toxicity levels [22]. P80 has been reported to act as a co-emulsifier along with P, in drug and gene delivery endeavors [23]. Moreover, the encouraging properties of P80 create a steric barrier that evades the aggregation of nano-vesicles, enhances the cell tolerance [11], and improves transfection efficiency due to the presence of polyethylene glycol (PEG) chains in its structure [24]. However, the ability P to form network structures might be more suitable than P80, if used with water-soluble cationic lipids, to enhance their flexibility and durability [25]. In such case, a mixture of two specific types of non-tensioactive molecules could provide a synergistic enhancement of nano-vesicle stabilization [26]. Regarding the cationic lipid, the high solubility of the D-Cl salt enhances biodistribution of lipid/plasmid complexes, and therefore, transfection efficiency [27]. However, in a previous study, we observed that the solubility of cationic lipid can dramatically shift the transfection results according to the type of the cells and the way of formulation. In that study, the DTPA cationic lipid (non-salt form) succeeded to transfect retinal cells in vitro conditions [11], while in such mentioned study, the salt form failed to transfect retinal cells in vivo. Interestingly, the same formulation with the same salt form of cationic lipid (DPP80) succeeded to transfect cerebral cortical cells in vivo [28]. Strikingly, both salt/non-salt forms of the cationic lipid were able to transfect ARPE-19 cells in vitro conditions. In any case, the non-salt form was superior in terms of transfection and cell viability. This contradiction emphasizes the lack of correlation between the in vivo and in vitro transfection conditions and the importance of the formulation at physical level. To emphasize the impact of chloroquine, DPP80 and DPP80-CQ niosomes were elaborated and compared. The characterization data of both niosomes were analyzed (Table 1). The incorporation of chloroquine slightly increased the size of niosomes by about 28 nm, and reduced both PDI (about 69% decrease) and ZP (about 34% decrease). Drug/gene delivery vehicles are generally favored by small poly-dispersity values [29]. The positive ZP values (> +25 mV) detected for both niosomes would reflect a Chapter 5 115 potentially long-lasting stability. Once the niosomes were characterized in terms of size, PDI, and zeta potential, nioplexes were elaborated with the pCMS-EGFP plasmid at various cationic lipid/DNA mass ratios by adding the reporter plasmid to the niosomes and not the opposite to ensure proper condensation process [30]. Fig. 5. Retinal cross sections micrographs obtained by confocal microscopy 3 days post subretinal injection of DPP80-CQ nioplexes (A–D). EGFP protein was observed mainly in GCL (white arrows), photoreceptors (yellow arrows) and RPE cells (blue arrows). Retinal sections were stained with antibodies against NeuN(A), recoverin (B, D) and protein kinase C (C). The cell nuclei were counterstained with Hoechst 33342. GCL, ganglion cell layer; INL, inner nuclear layer; ONL, outer nuclear layer; OS, photoreceptor outer segment Scale bars: 20 μm. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) The ZP of DPP80 nioplexes was clearly lower when compared to the same niosomes without chloroquine (Fig. 2-A). On the other hand, ZP of chloroquinecontaining nioplexes (DPP80-CQ) oscillated within a narrower range (19–25 mV) in comparison to DPP80-CQ niosomes (29 mV). Generally, the compaction of DNA is improved when 90% of the charge is compensated in an aqueous solution [31]. Strikingly, at 8/1 and 10/1 mass ratios for DPP80 and DPP80-CQ, respectively, ZP values fluctuated within a narrow range (23–27 mV) which represents a small reduction Chapter 5 116 in ZP for DPP80-CQ compared to ZP of niosomes (29 mV). This suggests a spontaneous electrostatic inter-action of pDNA with DPP80-CQ niosomes at 10/1 mass ratio which could be explained by a direct interaction of chloroquine with pDNA. Regarding PDI values of nioplexes, an obvious effect of chloroquine addition at all ratios studied above 4/1 was observed, as PDI values decreased in comparison to DPP80 formulation (Supplementary Table 1). The electron micrographs illustrated a discrete, almost spherical morphology and absence of aggregates in DPP80 complexes (Fig. 2-B1). By contrast, DPP80-CQ nioplexes appeared as clusters of multilamellar planar structures that form string-like colloidal aggregates (Fig. 2-B2). The lamellar spacing was around 5.5–6 nm, suggesting that the pDNA strands were complexed with the cationic lipid bilayers [15]. Similarly, many mixtures of neutral lipids (as DOPC and DOPE), along with cationic lipids (as DOTAP), extensively used for gene delivery purposes, are known to form lamellar complexes with DNA [32]. The agarose gel retardation assay showed that both niosomes, at all studied cationic lipid/DNA ratios, were able to condense, release and protect the DNA from enzymatic digestion (Fig. 2-C). Of note, the relatively lower DNA condensation, observed by the chloroquine-containing formulation (Fig. 2-C2), did not hamper the release or the protection of the condensed DNA, which is of utmost importance during the transfection process. Any change in condensation efficiency might affect the pattern and topology of spatial DNA configuration. Even more, the state of DNA condensation can be affected by both the type and the content of the surfactant or other additives as chloroquine. Therefore, the fine-tuning of such molecules could be of importance to unveil the mechanism of condensation of different types of DNA molecules within different nano-vesicles. Even at high concentrations of chloroquine, 100 μg/ml, ARPE19 cells appeared healthy with good viability, despite the appearance of many vacuoles in the cytoplasm (Supplementary Fig. 1). The transfection efficiency in vitro, ARPE-19 cells, fluctuated within a small range in both vectors at all mass ratios studied (Fig. 3). However, the cell viability was in favor of DPP80-CQ (Fig. 3-A). Noteworthy, chloroquine inhibits lysosomal enzymes by increasing the pH of the lysosomes and disturbing their fusion with autophagosomes, thus inhibits autophagy [33]. Moreover, Chapter 5 117 chloroquine and its autophagy inhibiting derivative, hydroxychloroquine, are both FDA-approved agents [34]. According to the cell type or the state of stress, autophagy might protect or promote cell death in the eye [35]. Fig. 6. Confocal fluorescence micrographs of whole mount (A, B) and cross-sections (C, D) of the retina 3 days after intravitreal administration of DPP80-CQ nioplexes. EGFP expression can be observed in both GCL (A, C and D, white arrows) and INL (B and C, yellow arrows). Interestingly, some protein expression was also observed in OPL (C, blue arrows). Whole mount and retinal sections were stained with NeuN (AD). The cell nuclei were counterstained with Hoechst 33342 (blue). GCL, Ganglion cell layer; INL, inner nuclear layer; ONL, outer nuclear layer; OPL, outer plexiform layer. Scale bars: 20 μm. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) This mutable nature of autophagy might be the reason for the increased cell viability observed with DDP80-CQ formulation in comparison with its chloroquine-free Chapter 5 118 counterpart, DPP80. Generally, cell viability and metabolism of ARPE-19 cells are relatively unaffected by the concentrations of chloroquine between 10 and 30 μg/ml, though affected in a dosage-dependent fashion afterward [36]. To analyze whether the enhanced cell internalization of nioplexes was among the effects that chloroquine could have on niosome formulations, we determined the percentage of ARPE-19 cell uptake of both DPP80 and DPP80-CQ formulations at the mass ratios of best transfection efficiency, 8/1 for DPP80 and 10/1 for DPP80-CQ (Fig. 4). Interestingly, flow cytometry studies showed that chloroquine incorporation had an insignificant effect on the percentage of cellular uptake when compared to the DPP80 formulation (Fig. 4-A). Such observation is most probably due to the indifferent surface charge of both nioplexes at the aforementioned mass ratios (22.5 ± 7.3 and 25.3 ± 2.5 for DPP80 and DPP80-CQ, respectively, P > .05). The similar uptake percentages in such ratios could justify the unaltered transfection results depicted pre-viously (Fig. 3-A). The transfection efficiency can be markedly affected by the mechanism of endocytosis. Consequently, we studied three of the most active cellular internalization pathways: clathrin-mediated endocytosis (CME), caveolae-mediated endocytosis (CvME) and macro-pinocytosis (MPC). The results observed in Fig. 4-B suggested that DPP80-CQ nioplexes were internalized mainly by MPC, while CvME and CME had less participation in the cellular uptake process. Due to its ability to internalize larger structures, macropinocytosis pathway has been proposed to mediate the internalization of non-viral gene delivery vehicles [37]. Moreover, MPC is considered as the major pathway responsible for DNA transfection in certain cell types [38]. In contrary, DPP80 nioplexes were internalized mainly by CME and, to a lesser extent, by MPC, while CvME had a much less participation in the cellular uptake process. However, the minor fluctuation in transfection efficiency between the two nioplexes could be due to limited variations between the two main different mechanisms of internalization (CME for DPP80 and MPC for DPP80-CQ). The delivery of genetic material by CvME and CME passes through late endosomes/lysosomes, which in-creases the hazards of DNA degradation and lowers the transfection efficiency [39]. So, an expected trivial effect of CvME and especially CME pathways could explain the high percentages of EGFP expression in ARPE cells by both nioplexes (Fig. 3), compared to lipofectamine® 2000 Chapter 5 119 (approximately, 80% and 75% of lipofectamine®2000 for DPP80 and DPP80-CQ, respectively). Afterwards, we analyzed the pH-buffering capacity of both niosomes (Fig. 4C). The incorporation of chloroquine into the niosome formulation increased the pHbuffering capacity upon titration with 0.1 M HCl, compared to the niosomes elaborated without chloroquine (at pH values > 2). Though, there was no change in the buffering capacity when the pH was < 2 for both niosomes. Chloroquine might induce endosomal and lysosomal escape via the proton sponge effect [40]. This result could suggest that chloroquine-containing formulation could increase the proton sponge effect, and therefore, the endosomal escape capacity of DPP80-CQ niosomes. However, as the predominant mechanism of internalization for DPP80-CQ was neither CvME nor CME, the impact of the proton sponge effect of chloroquine on the transfection efficiency was insignificant. Based on the previously mentioned physicochemical and in vitro biological results, we were enthusiastic to perform a preliminary in vivo study to evaluate the transfection efficiency of our formulations, DPP80-CQ in particular, in rat retinae after both subretinal (Fig. 5), and intravitreal injections (Fig. 6). Subretinal injection is a well-known clinical route to deliver genetic/drug material to the back of the eye. In addition, it enables direct contact of the injected nucleic acids with the outer retinal layers, photoreceptors and RPE cells. Noteworthy, clinical trials to treat many inherited retinal diseases such as LCA type 2 used the subretinal injection route [41]. However, it is less desirable than the intravitreal route due to the possible complications; such as retinal detachment or the localized side effects around the site of injection. Generally, IV injection is more widely applicable in the clinical practice due to its ability to deliver genetic material to a larger retinal surface, in addition to less surgical trauma compared to the SR route [42]. Surprisingly, DPP80 did not induce any transfection to retinal cells in vivo after both subretinal or intravitreal injections (Supplementary Fig. 2), whereas the Chapter 5 120 chloroquine-containing formulation, DPP80-CQ did (Figs. 5 and 6). The lack of correlation between in vitro and in vivo transfection results has been widely reported as it is a context-dependent matter [12]. Based on previous physicochemical and in vitro biological results, we were enthusiastic to perform a preliminary in vivo study to evaluate the transfection efficiency of our formulations, DPP80-CQ in particular, in rat retinae after subretinal (Fig. 5) and intravitreal injections (Fig. 6). Subretinal injection is a well-known clinical viable route to deliver genetic material to the eye. It enables direct contact of the injected nucleic acids with the outer retinal layers, photoreceptors and RPE cells. Noteworthy, clinical trials to treat many inherited retinal diseases such as LCA type 2 use subretinal injection [41]. However, it is less desirable than the IV route due to the possible complications such as retinal detachment or the localized effect around the site of injection. Generally, intravitreal injection is more widely applicable in the clinical practice due to its ability to deliver genetic materials to a larger retinal surface and advantages of less surgical trauma compared to the SR route [42]. Subretinal administration allows direct contact of genetic material with RPE cells and outer layer of the retina. Although this route of administration is highly effective to locally transfect cells close to the site of the injection, the occasionally observed side effects, related to this invasive route, such as retinal detachment, hemorrhages or alterations in RPE cells can hamper its practice [43]. In any case, subretinal injections have been widely used on clinical trials to treat some devastating genetic disorders of the retina reporting excellent outcomes [44]. In addition, the recently FDA/EMA-approved Luxturna medicine to deliver healthy copies of the RPE65 gene to the retina is administered by subretinal injection. In our in vivo experiments, after subretinal administration of nioplexes, we observed localized EGFP expression, mainly in some photoreceptor and RPE cells, close to the injection site. Transfection at this level can have clinical relevance because mutations of > 200 genes in RPE cells/photoreceptors are related to relevant ge-netic Chapter 5 121 disorders of the retina such as; Leber congenital amaurosis, retinitis pigmentosa, and Stargardt disease, to name just a few ones [45]. Compared to subretinal injection, intravitreal injection represents an interesting alternative to deliver genetic material to the back of the eye, and therefore to access retinal structure. It is a less invasive route, more easily to perform, and higher doses can be delivered [46]. Consequently, large retinal surfaces can be transfected by this route of administration [47]. When we administered 4 μl of DPP80-CQ nioplexes by intravitreal injection, the inner layers of the retina (GCL and INL) were mainly transfected as observed in Fig. 6. (white and yellow arrows, respectively). Transfection at this level can be of clinical relevance in treatment of devastating ocular pathologies that compromise the function of ganglion cells as glaucoma [48]. Interestingly, EGFP expression was also discerned in the OPL (Fig. 6-C, blue arrows) which suggests that nioplexes partially diffused, not only through the vitreous where they were administered, but also through the inner retinal layers until reach the OPL. Transfection of the outer layers of the retina by intravitreal administration of non-viral vectors represents a great challenge for the scientific community, since can avoid the subretinal injections and the corresponding side effects commonly associated to such injection. Unfortunately, chloroquine, like other endolytic agents, has been found to be cytotoxic in several pre-clinical or clinical trials [49]. Chloroquine passes the bloodretinal barrier and is toxic to the retina. Nevertheless, such retinal toxicity is related to large doses and long-term use of chloroquine [50]. In this study, at 10/1 cationic lipid /DNA mass ratio, the final concentration of chloroquine was only 25 μg/ml which did not induce any significant cytotoxicity in accordance with Chen et al, [36]. The affinity of retinal cells to the modified salt form of the cationic lipid, in addition to the favorable properties of P and P80, along with the effect of chloroquine, raise the possibility to target different retinal cell layers safely and effectively after both subretinal and intravitreal administrations. Chapter 5 122 5.5.Conclusions The addition of chloroquine to a niosome formulation retained its functionality in vitro, but most importantly, enhanced its transfection ability in vivo. This work highlights the use of chloroquine as a built-in component in the gene delivery vehicles to evade its toxicity and to provide new insights into the future of retinal gene therapy. 5.6.Acknowledgements and disclosures This project was supported by the Basque Country Government (CGIC10/172), Spanish Ministry of Education (Grant CTQ2017-84415-R, MAT2015-69967-C3-1R), the Generalitat de Catalunya (2014/SGR/ and the Instituto de Salud Carlos III (CB06_01_0019, CB06_01_1028). The authors also wish to thank the intellectual and technical assistance from the ICTS “NANBIOSIS”, more specifically by the Drug Formulation Unit (U10) of the CIBER in Bioengineering, Biomaterials, and Nanomedicine (CIBER-BBN) at the University of Basque Country (UPV/EHU). Technical and human support provided by SGIker (UPV/EHU) is acknowledged. 5.7. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.jconrel.2019.05.010. Chapter 5 123 Supplementary fig.1. (A) ARPE-19 cell viability after 4h of incubation with different concentrations of chloroquine diphosphate. Phase contrast microscopy of ARPE-19 cells showing control cells (Ctrl) and cells with vacuolated cytoplasm (CQ) after their treatment with chloroquine diphosphate 100ug/ml. Supplementary fig. 2. Confocal fluorescence micrographs of (A) retinal whole mount after 3 days of intravitreal administration of DPP80 nioplexes and of (B) retinal crosssections after 3 days of subretinal administration of DPP80 nioplexes. There is no GFP expression detected in different retinal layers. Scale bars: 20 μm. Chapter 6 130 Chapter 6 131 Chapter 6 General discussion Chapter 6 132 Chapter 6 133 Presently, cationic niosomes, as non-viral gene delivery carrier, have become an important tool to deliver both genetic macromolecules and drug molecules. Development of efficient non-viral gene delivery systems could reduce the time and save expenses for coming new market therapies. In addition, they would be significantly safer than their viral counterparts. Development of safe and efficient nonviral vectors to deliver DNA into the CNS represents a huge challenge to face many neurological disorders. Despite its peripheral location, the retina or neural layer of the eye, is actually a part of the central nervous system. In the current study, we have designed, prepared and characterized niosome formulations based on different cationic lipids and various helper molecules. In vitro studies were conducted to evaluate transfection efficiency, viability and internalization mechanism in ARPE-19 and NT2 cells. Subsequently, their in vivo application was evaluated in both retina and brain. 6.1 Lycopene enhances the efficacy of cationic niosomes based on DOTMA and polysorbate 60 for retinal gene delivery purposes Due to its appealing chemical structure, the commercially available cationic lipid DOTMA has been used widely for gene delivery applications (1). Fig. 1. Chemical structures of the cationic lipid, N-[1-(2,3-dioleoyloxy)propyl]-N,N,Ntrimethylammonium chloride (DOTMA) (A), Polysorbate 60 (B), and Lycopene (C). Chapter 6 134 As shown in Fig. 1-A, its structure is composed of a polar head-group, two non-polar hydrophobic chains, a linker and a back-bone, which classically are known as the four domains that rule gene transfection process (2). We combined DOTMA with the non-ionic surfactant polysorbate 60, in a niosome formulation at a molar ratio of 1:4 respectively, in order to enhance cell tolerance (3) and provide a steric barrier to avoid aggregation (4). It has been reported on the literature that the presence of PEG chains in the chemical structure of polysorbates (Fig. 1-B) provides physicochemical stability to lipid formulations (5), conserves effectiveness over time and boosts transfection efficiency (6). Table 1 Physical characterization of DP60 and DP60L niosomes regarding particle size (nm); Polydispersity index (PDI), and Zeta potential (mV). Data represent mean ± SD (n = 3). Particle Size (nm) PDI Zeta potential (mV) DP60 niosome 66.49 ± 1.17 0.46±0.02 45.30±1.57 DP60L niosome 101.60 ± 2.48 0.44±0.02 33.80±1.13 Compared with polysorbate 80, another polysorbate that has been widely used in the elaboration of niosome formulations for gene delivery applications, [15–17,31] polysorbate 60 could offer some important advances. Chapter 6 135 Fig. 2. Physicochemical characterization of nioplexes. A) Effect of cationic lipid/DNA mass ratio (w/w) on both particle size (bars) and zeta potential (lines). Each data point represents the mean ± SD (n = 3). TEM of DP60 (B1) and DP60L nioplexes (B2) at ratio of 18/1 cationic lipid/DNA mass ratio (w/w). Scale bar = 500 nm. Binding, SDSinduced release and protection of DNA at different cationic lipid/DNA mass ratios (w/w) of nioplexes based on both DP60 (C1) and DP60L (C2) visualized by agarose electrophoresis. Lanes 1–3 correspond to uncomplexed DNA; lanes 4–6, cationic lipid/DNA mass ratio 6/1; lanes 7–9, cationic lipid/DNA mass ratio 12/1; lanes 10–12, cationic lipid/DNA mass ratio 18/1; lanes 13–15, cationic lipid/DNA mass ratio 22/1. Nioplexes were treated with SDS (lanes 2, 5, 8, 11 and 14) and DNase I + SDS (lanes 3, 6, 9, 12 and 15). OC: open circular form, SC: supercoiled form. For instance, the lack of double bonds in the hydrocarbon chains (Fig. 1-B) could provide low permeability of the vesicles, and therefore better stability of niosome membranes (5). Additionally, compared with other hydrophilic surfactants such as polysorbates 80, 40 or 20, the low hydrophilic-lipophilic balance (HLB) value of polysorbate 60 (14.9) could help to solubilize lycopene more efficiently (7). The addition of the natural and non-polar lipid lycopene (Fig. 1-C) into niosome bilaminar membrane could increase its fluidity, disturb membrane packing, and consequently vesicle susceptibility to environmental stresses (8). Chapter 6 136 Fig. 3. In vitro transfection efficiency and cell viability in ARPE-19 cells at 72 h posttransfection. (A) Flow cytometry-based evaluation of the percentage of EGFP-positive cells (bars) and percentage of viable cells (lines) at different cationic lipid/DNA mass ratios (w/w). Values represent mean ± SD (n = 3). (*P < 0.05 vs. Lipofectamine™2000 transfection). (#P < 0.05 vs. Lipofectamine™2000 viability). (B) Overlay of fluorescence and phase-contrast micrographs of ARPE-19 cells 72 h post-transfection at different cationic lipid/DNA mass ratios (w/w). Scale bar = 100 μm. Once elaborated by the reverse phase elaboration method, both DP60 and DP60L niosomes showed appropriate size (in the nanometric scale) and PDI values (below 0.5) for gene delivery purposes (Table 1). High positive ZP values (> +25 mV) ensure longlasting stability (9), extempore electrostatic reciprocal action with DNA, along with binding of the nioplexes to the negatively charged units of the cell membrane previous to cellular uptake (2). Chapter 6 137 Fig. 4. Uptake of FITC-labeled nioplexes in ARPE-19 cells. Both DP60 and DP60L at a mass ratio of 18/1 (w/w). (A) Percentage of FITC-positive cells. Data represent mean ± SD (n = 3). *p < 0.05. (B) Fluorescence micrographs of ARPE-19 cells at 2 h and 4 h of incubation with FITC-labeled DP60 and DP60L nioplexes (green). Nuclei stained with Dapi (blue). Original magnification 63×. Scale bar = 20 μm. (For inter-pretation of the references to color in this figure legend, the reader is referred to the web version of this article.) To elaborate nioplexes, we added pCMS-EGFP reporter plasmid to both niosome formulations at different cationic lipid/DNA mass ratios, since otherwise, the complex assembly process could be slowed down (10). The slight changes discerned in the size of nioplexes (100–150 nm, Fig. 2-A), at the mass ratios studied, might be due to the delicate balance of different events involved in the multistep self-assembled complex formation, such as: electrostatic interaction, further membrane merging, lipid mixing and aggregate growth (10). Regarding the ZP values, the gradual increase of superficial charge along with cationic lipid/DNA ratios (w/w) suggests the capacity of cationic Chapter 6 138 niosomes to bind to and neutralize the negatively charged phosphate groups of DNA (11). Lycopene addition reduced ZP value of DP60L nioplexes, compared to DP60, at all ratios studied. This fact could be explained by the perturbation of the lipid membrane bilayer, which could dissipate the electrical potential (Fig. 2-A) (12). Fig. 5. Three-channel overlay RGB images of ARPE-19 cells showing nioplexes with FITC-labeled pCMS-EGFP (green) and one of the endocytosis markers in red (AlexaFluor® 555-Cholera Toxin, AlexaFluor® 546-Transferrin or AlexaFluor® 594dextran). Presence of yellow/orange color represents the overlay of an endocytic marker and nioplexes. (M = Mander's overlap coefficient). Original magnification 63×, Scale bar = 25 μm. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) In any case, both formulations could function as gene delivery carriers, since complexes with positive charge could interact electrostatically with the anionic cell coat, inducing early steps of the endocytosis process (13). The high positive ZP value of both DP60 and DP60L nioplexes, especially at 18/1 mass ratio (42 and 27 mV, Chapter 6 139 respectively), could ensure the discrete morphology and absence of aggregates observed by TEM micrographs (Fig. 2-B) (14). Among other factors that can command transfection success, the van der Waals interactions between phosphate groups of the DNA (negatively charged) and amine groups of the cationic niosomes (positively charged) deserve special attention (11, 15, 16). We observed by agarose gel electrophoresis assay that at all cationic lipid/DNA ratios tested, both niosomes were capable to condense, release and protect the DNA from enzymatic digestion (Fig. 2C1and C2). Once we evaluated that our nano-formulations were biotechnologically fitting for gene delivery purposes, we proceeded to evaluate their biological performance in ARPE-19 cells. ARPE-19 cell line has a normal karyotype and has functional and structural properties similar to retinal pigment epithelia (RPE) in vivo, expresses RPE-specific markers, hence it is considered a suitable transfection model to investigate our vectors' effectiveness and safety before its application in vivo (17). It has been reported that the non-ionic nature of surfactants makes niosomes well tolerated by cells (18). Our results in Fig. 3 show higher cell viability values in cells transfected with both nioplexes when compared with cells transfected with Lipofectamine™ 2000. Additionally, we observed under the fluorescence microscope that cells transfected with both nioplexes maintained their normal morphology, even at high cationic lipid/DNA ratios (Fig. 3B). Although the percentage of transfected cells with DP60L niosomes at 18/1 mass ratio was significantly lower than that obtained with commercially available Lipofectamine™ 2000, our niosomes formulation was better tolerated by ARPE-19 cells. Therefore, it could be an interesting alternative to Lipofectamine™ 2000, since some authors have reported damage on the retina associated to the in vivo administration of Lipofectamine™ 2000 in the eye (19). Regarding the transfection efficiency, the lipid composition is considered a primary limiting factor that affects to this process (1). We clearly observed in Fig. 3 the impact that lycopene had on transfection efficiency in ARPE-19 cells, since values were clearly higher when lycopene was present in the niosome formulation.