Polysaccharide functionalised span nanoparticles as gene delivery systems. Application in the treatment of colorectal liver metastasis

TeTF de doVUPSBNFOUP PD en Investigación e Desenvolvemento de Medicamentos Inés Fernández Piñeiro Polysaccharide functionalised span nanoparticles as gene delivery systems. ApplicationintheUSFBUNFOU PGDPMPSFDUBMMJWFSNFUBTUBTJT Santiago de Compostela 2017 TESE DE DOUTORAMENTO POLYSACCHARIDE FUNCTIONALISED SPAN NANOPARTICLES AS GENE DELIVERY SYSTEMS. APPLICATION IN THE TREATMENT OF COLORECTAL LIVER METASTASIS Inés Fernández Piñeiro ESCOLA DE DOUTORAMENTO INTERNACIONAL PROGRAMA DE DOUTORAMENTO EN INVESTIGACIÓN E DESENVOLVEMENTO DE MEDICAMENTOS SANTIAGO DE COMPOSTELA 2017 DECLARACIÓN DO AUTOR/A DA TESE POLYSACCHARIDE FUNCTIONALISED SPAN NANOPARTICLES AS GENE DELIVERY SYSTEMS. APPLICATION IN THE TREATMENT OF COLORECTAL LIVER METASTASIS D./Dna. Inés Fernández Piñeiro Presento a miña tese, seguindo o procedemento axeitado ao Regulamento, e declaro que: 1) A tese abarca os resultados da elaboración do meu traballo. 2) De selo caso, na tese faise referencia ás colaboracións que tivo este traballo. 3) A tese é a versión definitiva presentada para a súa defensa e coincide coa versión enviada en formato electrónico. 4) Confirmo que a tese non incorre en ningún tipo de plaxio doutros autores nin de traballos presentados por min para a obtención doutros títulos. En Santiago de Compostela, 15 de Decembro de 2017 Asdo. AUTORIZACIÓN DOS DIRECTORES DA TESE POLYSACCHARIDE FUNCTIONALISED SPAN NANOPARTICLES AS GENE DELIVERY SYSTEMS. APPLICATION IN THE TREATMENT OF COLORECTAL LIVER METASTASIS D. Alejandro Sánchez Barreiro D. Iker Badiola Etxaburu INFORMA/N: Que a presente tese, correspóndese co traballo realizado por Dna. Inés Fernández Piñeiro, baixo a miña dirección, e a utorizo a súa presentación , considerando que reúne os r equisitos esixidos no R egulamento de Estudos de Doutoramento da USC, e que como director desta non incorre nas causas de abstención establecidas na Lei 40/2015. En Santiago de Compostela, 15 de Decembro de 2017 Asdo. Asdo. En memoria de Begoña Seijo Abstract Resumen Abstract / Resumen 17 Abstract The main objective of this work has been the design of a new polysaccharide functionalised nanosystem as a non-viral microRNA vector for the treatment of liver metastasis from colorectal cancer. For this purpose, we modified our previously developed span nanoparticles using polysaccharides with endothelial targeting properties. Thus, we have developed a xanthan gum-functionalised nanosystem able to associate, protect from degradation and deliver in vitro a model plasmid without compromising cell viability. Moreover, these nanoparticles showed a remarkable shortand long-term stability at different temperatures, both in suspension and as a lyophilised product. The xanthan gum cover targeted the nanoparticles, and therefore, the plasmid delivery to endothelial cells of liver, kidney and lung in vivo. In addition, span nanoparticles were functionalised with the glycosaminoglycans chondroitin sulfate and hyaluronic acid. The resulting systems were efficiently loaded with a model plasmid and characterised in terms of physicochemical, stability, DNA protection, cytotoxicity and transfection properties. Once selected the chondroitin sulfate-functionalised nanoparticles for further studies, we evaluated their clinical potential associating microRNA in the treatment of a murine model of colorectal cancer liver metastasis. First, these nanoparticles demonstrated to successfully target liver sinusoidal endothelial cells (LSECs) and deliver miR-20a, which has been found to be downregulated in tumour-activated LSECs. Finally, the administration of these miR-20a loaded nanoparticles demonstrated to reduce the tumour volume by 80% and the LSECs infiltration into tumour foci by 70% in a murine model of colorectal liver metastasis. Therefore, these results provide an in vivo proof-of-concept of the clinical potential of the developed nanoparticles as a novel microRNA-based therapeutic strategy for cancer treatment. Keywords: nanoparticles, polysaccharides, microRNA, gene therapy, cancer Inés Fernández Piñeiro 18 Resumen El principal objetivo de este trabajo ha sido el diseño de un nuevo nanosistema funcionalizado con polisacáridos como vector no viral de microRNA para el tratamiento de metástasis hepáticas de cáncer colorrectal. Con tal finalidad se han modificado las nanopartículas de span previamente desarrolladas utilizando polisacáridos con propiedades de direccionamiento al endotelio vascular. Así, se ha desarrollado un nanosistema funcionalizado con goma xantana que es capaz de asociar, proteger de la degradación y liberar in vitro un plásmido modelo, sin comprometer la viabilidad celular. Asimismo, estas nanopartículas mostraron una notable estabilidad a corto y largo plazo durante su almacenamiento a diferentes temperaturas, tanto en suspensión como en su forma liofilizada. La cubierta de goma xantana direccionó las nanopartículas y, por lo tanto, la entrega del plásmido a las células endoteliales de hígado, riñón y pulmón in vivo. Además, las nanopartículas de span se funcionalizaron con los glucosaminoglicanos condroitín sulfato y ácido hialurónico. Un plásmido modelo fue asociado eficazmente a estas nanopartículas y los sistemas resultantes se caracterizaron en términos de propiedades fisicoquímicas, de estabilidad, de protección del DNA, de citotoxicidad y de transfección. Una vez seleccionadas las nanopartículas funcionalizadas con condroitín sulfato para la evaluación de su potencial clínico, se emplearon las mismas asociando microRNA en el tratamiento de un modelo murino de metástasis hepática de cáncer colorrectal. En un primer paso, estas nanopartículas demostraron direccionar la entrega de miR-20a -cuya expresión se encuentra disminuida en este tipo de procesos-, a las células endoteliales del sinusoide hepático. Finalmente, la administración de estas nanopartículas asociando miR-20a demostró reducir el volumen de las metástasis en un 80% y la infiltración de células endoteliales en los focos metastáticos en un 70% en un modelo animal de metástasis hepática de cáncer colorrectal. Por lo tanto, estos resultados proporcionan una prueba de concepto in vivo del potencial clínico de las nanopartículas desarrolladas como estrategia terapéutica a base de microRNA para el tratamiento del cáncer. Palabras clave: nanopartículas, polisacáridos, microRNA, terapia génica, cáncer Resumen in extenso Resumen in extenso 21 Introducción 1. Terapia génica La terapia génica consiste en la transferencia de ácidos nucleicos a células específicas para evitar los efectos indeseados causados por genes aberrantes o disfuncionales y con el objetivo final de tratar o aliviar una patología. El material genético empleado puede ser tanto un plásmido de ácido desoxirribonucleico (DNA), empleado para compensar un déficit de producción causado por un gen defectuoso o ausente, como reguladores post-transcripcionales en forma de pequeñas secuencias de ácido ribonucleico (RNA) o DNA (oligonucleótidos antisentido), utilizados para suprimir o alterar la expresión de un gen a nivel posttranscripcional. Las estrategias actualmente utilizadas en terapia génica pueden ser clasificadas en términos generales en terapias de reemplazo génico y de adición génica, aplicables a enfermedades causadas por uno o diversos defectos genéticos respectivamente, así como en terapia de modificación del RNA. Dentro de esta última se incluyen el silenciamiento genético mediado por una molécula de RNA (RNA interferente pequeño, siRNA, o microRNA) que se une a un RNA mensajero complementario, inhibiendo su traducción o provocando su degradación; así como la reprogramación del splicing del RNA, es decir la inclusión o exclusión de un exón durante la maduración del RNA mensajero (mRNA) para producir una proteína funcional, mediante el uso de oligonucleótidos antisentido. Aunque la terapia génica supone una alternativa terapéutica para numerosas enfermedades, la mayoría de los ensayos clínicos y estudios realizados actualmente están dirigidos al tratamiento del cáncer. No obstante, debe tenerse en cuenta que el cáncer engloba un conjunto de enfermedades multifactoriales que implica alteraciones genéticas tanto en las propias células cancerígenas como en las células adyacentes, es decir, en el conocido como microambiente tumoral. Por lo tanto, están surgiendo numerosas dianas terapéuticas dentro de este microambiente, como es el caso de las células endoteliales, como alternativa al tratamiento exclusivo de las células cancerosas Los ácidos nucleicos son moléculas hidrofílicas, cargadas negativamente y de alto peso molecular; características que dificultan su paso a través de barreras biológicas y membranas celulares. Además, son moléculas lábiles susceptibles de degradación enzimática por Inés Fernández Piñeiro 22 endonucleasas una vez administradas al organismo. Por lo tanto, se hace necesaria la incorporación del material genético a vehículos que aseguren el transporte a su lugar de acción, el traspaso de barreras biológicas y la protección frente a la degradación enzimática. Estos vehículos se pueden clasificar en vectores virales y vectores no virales. Los vectores virales se unen a las células diana e introducen su material genético como parte de su proceso de replicación, presentando una alta eficacia de transfección. Con respecto a los vectores no virales se caracterizan por ser más seguros y fáciles de escalar, así como por poseer una mayor capacidad de carga génica, que los vectores virales. Entre los vectores no virales destacan los nanosistemas para el transporte de material genético. 2. Polímeros naturales para la administración de moléculas bioactivas en general y material genético en particular Los polímeros naturales han despertado mucho interés en el ámbito de la administración de material genético debido a la idoneidad de sus propiedades, puesto que son sustancias biocompatibles, renovables y fácilmente modificables. Dentro de los polímeros naturales, los polisacáridos son unos de los más empleados en el diseño de sistemas transportadores de moléculas activas. Entre los motivos de esta gran aplicación destacan la capacidad de algunos polisacáridos para actuar como moléculas de direccionamiento a receptores o tejidos específicos, así como para aportar protección estérica a los sistemas transportadores reduciendo el reconocimiento de los mismos por parte del sistema reticuloendotelial y, por lo tanto, prolongando su tiempo de circulación en sangre. Los polisacáridos son polímeros formados por más de 10 monosacáridos unidos mediante enlaces glucosídicos. Son heterogéneos en estructura y composición química, pudiendo ser obtenidos de diversas fuentes naturales, como algas (p. ej. alginato), plantas (p. ej. pectinas, celulosa y ciclodextrinas) y animales (p. ej. quitosano, ácido hialurónico y condroitín sulfato). 3. Endotelio vascular El endotelio forma el revestimiento celular interno de los vasos sanguíneos. Sin embargo, las células endoteliales (CEs) tienen funciones muy diversas más allá de simplemente aportar un revestimiento a las paredes de los vasos sanguíneos. Estas CEs muestran una enorme heterogeneidad estructural, funcional y genética, que difiere a lo largo del árbol vascular y de los diferentes órganos y tejidos para adaptarse a requisitos específicos. Resumen in extenso 23 El papel del endotelio vascular en la progresión de diversas enfermedades, como es el caso del cáncer, es cada vez más evidente. Se ha descubierto que, en el microambiente tumoral, el endotelio muestra características específicas. Así, aunque las CEs están en estado quiescente durante la mayor parte de su vida adulta, pueden pasar a un estado proliferativo y migratorio en respuesta a estímulos angiogénicos procedentes de las células del microambiente tumoral. Así, la progresión del cáncer está regulada por una comunicación cruzada entre las células cancerosas y células no cancerosas del microambiente tumoral. Además de participar en la comunicación cruzada con las células tumorales, estas células no cancerosas son genéticamente más estables que las células cancerígenas. Por lo tanto, orientar las terapias antineoplásicas hacia estas células presenta un especial atractivo como estrategia alternativa para vencer al cáncer. Se han desarrollado diferentes aproximaciones terapéuticas dirigidas a tratar el potencial angiogénico de las CEs, pero sin embargo, el desarrollo de terapias antiangiogénicas sigue siendo limitado debido al escaso entendimiento que se tiene de sus beneficios, así como de los efectos secundarios y de las resistencias causadas por esta estrategia terapéutica. Sin embargo, teniendo en cuenta las diferencias moleculares y de expresión génica entre las CEs quiescentes y las activadas por el tumor, el direccionamiento de las terapias hacia el tratamiento de las CEs a un nivel transcripcional surge como una alternativa terapéutica de gran interés. Así, unas de las estrategias terapéuticas que están siendo actualmente exploradas para el tratamiento del cáncer son las dirigidas a regular la expresión de microRNAs en las CEs del microambiente tumoral. 4. Metástasis hepáticas El hígado es el órgano más frecuentemente afectado por metástasis en la mayoría de las neoplasias prevalentes, siendo las metástasis hepáticas mucho más comunes que los tumores hepáticos primarios. Esta gran susceptibilidad del hígado a las metástasis se debe, por una parte, a sus características estructurales y hemodinámicas que favorecen la interacción y retención de células tumorales, y por otra parte, a su capacidad regenerativa y de supresión inmune que propician un ambiente favorable para el crecimiento tumoral. El hígado tiene un suministro dual de sangre constituido por la artería hepática y la vena porta, drenando ambas en los sinusoides hepáticos, los cuales representan la red capilar en el hígado. Las células endoteliales del sinusoide hepático, o LSECs de “liver sinusoidal endothelial cells”, suponen el 50% de las Inés Fernández Piñeiro 30 con las nanopartículas, en contraposición con la reducción del 20% observada cuando esta molécula fue administrada en su forma libre. Asimismo, se consiguió reducir en un 70% la presencia de LSECs infiltradas en el tumor en comparación con los controles, demostrando el potencial terapéutico de las nanopartículas desarrolladas (Cap. III, Fig. 8, 9 y 10). Discusión 1. Funcionalización de nanopartículas de span con polímeros naturales En este trabajo nos hemos centrado en la funcionalización con polímeros naturales de nanopartículas lipídicas a base de monooleato de sorbitán, también conocido como span 80 (SP). Además de SP, un surfactante no iónico, estos nanosistemas incluyen en su composición la amina grasa oleilamina (OA), la cual fue incorporada para facilitar la asociación de los ácidos nucleicos a través de interacciones electrostáticas con las cargas negativas del material genético. Los polímeros naturales escogidos para funcionalizar estas nanopartículas se han seleccionado teniendo en cuenta el objetivo final de esta tesis, el cual ha sido el desarrollo de un nanosistema transportador de microRNA capaz de dirigirse a las células endoteliales del sinusoide hepático (LSECs) como nueva terapia en el tratamiento de metástasis hepática de cáncer colorrectal. Así, los polisacáridos goma xantana (XG), condrotín sulfato (CS) y ácido hialurónico (HA) se incorporaron al sistema buscando modular sus propiedades fisicoquímicas y su comportamiento biológico. De acuerdo con las propiedades fisicoquímicas de las nanopartículas resultantes de la funcionalización, podemos suponer que los grupos aniónicos de XG, CS y HA se disponen en la superficie de las nanopartículas. Esta cubierta externa cargada negativamente resulta de gran interés a la hora de evitar interacciones inespecíficas con componentes biológicos, así como para incrementar la estabilidad del sistema y para direccionar el mismo a células o regiones específicas del organismo. Adicionalmente, la disminución observada en el tamaño de las nanopartículas tras la funcionalización con los polisacáridos podría ser explicada por un incremento en el empaquetamiento de los diferentes componentes debido a interacciones electrostáticas y a la subsiguiente reticulación física de los mismos. Además, esta cubierta polisacarídica ha demostrado proteger eficazmente el plásmido pEGFP frente a la degradación enzimática por DNasas, que son enzimas presentes en los tejidos y fluídos biológicos, Resumen in extenso 31 prerrequisito para asegurar la entrega intacta de los ácidos nucleicos asociados a las nanopartículas en su lugar de acción. Por otra parte, estas nanopartículas solucionan otro aspecto crítico en el desarrollo y escalado a la industria farmacéutica de las nanomedicinas, como es la baja estabilidad de los nanosistemas y su tendencia a perder sus propiedades fisicoquímicas durante el almacenamiento. De este modo, hemos desarrollado sistemas lipídicos funcionalizados con polisacáridos que destacan por su estabilidad a corto y largo plazo, tanto en suspensión como en forma liofilizada. Este comportamiento podría explicarse por diferencias estructurales con otros nanosistemas lipídicos conocidos por su baja estabilidad, como los liposomas, puesto que los sistemas desarrollados en este trabajo muestran una estructura sólida no vesicular. 2. Evaluación de la toxicidad, capacidad de transfección y biodistribución de las nanopartículas desarrolladas Los estudios de viabilidad celular realizados in vitro con las tres formulaciones desarrolladas, mostraron una disminución en la viabilidad a las concentraciones más altas ensayadas de las nanopartículas. Sin embargo, este efecto no puede ser explicado por la toxicidad de los polímeros en sí, puesto que diversos estudios han descrito la biocompatibilidad y seguridad de XG, CS y HA. En consecuencia, esta toxicidad podría atribuirse a una mayor internalización celular de las nanopartículas mediada por la cubierta polisacarídica. Asimismo, no se observaron signos de toxicidad sistémica después de la administración intravenosa de las nanopartículas a ratones, exceptuando los infiltrados leucocitarios observados en el hígado de los animales tratados con nanopartículas de XG. Esta reacción inflamatoria leve podría ser debida tanto a la actividad inmunomoduladora de XG como a una respuesta inmune a la sobreexpresión a que dan lugar las nanopartículas de GFP, efectos que se encuentran descritos en la literatura. Igualmente, el efecto protector y estabilizador de la cubierta de polisacáridos quedó demostrado una vez que se pudo verificar la elevada capacidad de las nanopartículas de XG para transfectar la línea celular A549 en presencia de suero, el cual es una de las principales barreras a superar por los sistemas transportadores de material genético en el organismo. Por otra parte, se confirmó que tanto las nanopartículas funcionalizadas con XG como con CS pueden escapar de la captura por parte de las células de Kupffer y, por lo tanto, estos resultados Inés Fernández Piñeiro 32 apoyarían la teoría de que la funcionalización polisacarídica de las nanopartículas de span puede ayudar al sistema a evitar la eliminación por parte del sistema reticuloendotelial. Tras la administración sistémica de las nanopartículas de XG a ratones, la expresión de GFP se observó tanto en las LSECs como en las células endoteliales de los vasos sanguíneos de hígado, pulmón y riñón. La internalización de las nanopartículas por parte de las LSECs puede estar mediada por receptores de manosa, teniendo en cuenta la presencia de residuos manosa en la estructura de XG y la sobreexpresión de estos receptores en las LSECs. Sin embargo, este receptor no está presente en las células endoteliales de los vasos sanguíneos, por lo que consideramos que el direccionamiento de las nanopartículas a estas células podría ser explicado por una interacción con los dominios lipídicos caveola o mediante una unión no específica a sitios catiónicos de la membrana celular, sin la necesidad de la intervención específica de un receptor. En el caso de las nanopartículas funcionalizadas con CS y HA, la expresión de GFP se encuentra más delimitada a las LSECs después de su administración sistémica. De esta forma, se aprovechó la afinidad de ambos polisacáridos por el receptor de hialuronano (HARE), así como la afinidad del CS por el receptor de manosa, siendo ambos receptores expresados por las LSECs. Debido a la mayor expresión de GFP observada en los hígados tratados con nanopartículas de CS, se decidió continuar la investigación con esta formulación. 3. Aplicación de las nanopartículas de condroitín sulfato cargadas con microRNA en el tratamiento de la metástasis hepática de cáncer colorrectal El hígado es uno de los órganos más frecuentemente metastatizados debido a su posición anatómica y su arquitectura histológica. La transformación fenotípica de las LSECs es una de las etapas más importantes de la progresión de la metástasis hepática. Por lo tanto, tras demostrar la desregulación en la expresión de miR-20a en LSECs activadas tumoralmente, y debido a la inestabilidad de las moléculas de miRNA y a su incapacidad para cruzar barreras biológicas, se decidió incorporar esta molécula bioactiva a las nanopartículas de span funcionalizadas con CS para que ésta sea transportada específicamente a las LSECs en un modelo murino de metátasis hepática de cáncer colorrectal y, así, explorar su potencial terapéutico en el tratamiento de esta patología. En un primer paso, hemos demostrado la capacidad de las nanopartículas de span funcionalizadas con CS para asociar la molécula activa miR-20a, evidenciando la versatilidad Resumen in extenso 33 del sistema desarrollado para incorporar ácidos nucleicos de diverso tamaño. Asimismo, la capacidad de proteger y transportar específicamente esta molécula a las LSECs in vivo ha quedado demostrada tras la administración sistémica de estas nanopartículas asociando miR20a marcado fluorescentemente. La estrategia para restaurar el fenotipo normal de LSECs sanas mediante la entrega de miR20a utilizando nanopartículas de span funcionalizadas con CS demostró ser un éxito, puesto que se consiguió reducir la presencia de LSECs infiltradas en los focos tumorales en un 70%. Esto es indicativo de que las nanopartículas consiguieron transportar y restaurar los niveles de miR-20a en LSECs, disminuyendo la neoangiogénesis y el soporte vascular a los tumores. Otro resultado que supone el espaldarazo o la confirmación de la eficacia de la estrategia terapéutica diseñada es la reducción del 80% en el área ocupada por las metástasis que se consiguió con las nanopartículas en comparación con el 20% de reducción observado cuando el miR-20a se administró en su forma libre. Conclusiones Por lo tanto, las siguientes conclusiones pueden ser extraídas de este trabajo: 1. Es posible modular las propiedades fisicoquímicas de las nanopartículas de span mediante la incorporación a su composición de diversos polisacáridos naturales. 2. Los nanosistemas desarrollados muestran una destacable estabilidad tanto en suspensión como en su forma liofilizada. 3. Las nanopartículas desarrolladas han demostrado su capacidad para actuar como sistemas transportadores del plásmido pEGFP. 4. La cubierta polisacarídica permite modular la biodistribución del sistema. 5. Las nanopartículas funcionalizadas con condroitín sulfato y asociando miR-20a han permitido obtener una prueba de concepto in vivo que confirma el gran potencial terapéutico de la estrategia diseñada en el tratamiento de la metástasis hepática de cáncer colorrectal. Introduction Introduction 37 1. Gene therapy Gene therapy consists of the transference of nucleic acids to specific cells to avoid the undesirable effects caused by aberrant or malfunctioning genes with the aim to treat or alleviate a pathological condition. The genetic material can be genes encoded in deoxyribonucleic acids (DNA) molecules, or post-transcriptional regulators such as ribonucleic acids (RNA), including small interfering RNA (siRNA) and microRNA, and small DNA sequences, known as antisense oligonucleotides. Thus, DNA is used to compensate a production deficit caused by a defective or absent gene; while RNA and antisense oligonucleotides are used for gene suppression or alteration at a post-transcriptional level. Both siRNA, miRNA and antisense oligonucleotides consist of small strands of DNA or RNA that can hybridize with specific pre-mRNA or mature mRNA sequences and suppress or regulate gene expression (1, 2). Gene therapy offers new treatment possibilities for both acquired and hereditary diseases where conventional therapies are less effective, such as cancer, AIDS, cystic fibrosis, arthritis, peripheral vascular diseases and neurodegenerative disorders (3). Thus, more than 64% of all gene therapy trials worldwide are aiming at the treatment of cancer, and cancer gene therapy represents the predominant field of basic research, as it is a major global health problem accounting, annually, for more than eight million deaths worldwide. Although gene therapy was quickly adapted for cancer therapy (4) it should be taken into account that cancer is not caused by just a single alteration, but a complex, multifactorial disease involving changes in the genome, both in cancer cells and in the surrounding tissue. Recently, tumour microenvironment and, specially, tumour endothelial cells have emerged as alternative therapeutic targets for gene therapy. To fully understand the basis of this novel therapeutic approach, more specific aspects of endothelium physiology, functions and its role in cancer progress are described in further detail in section 3 and 4 of this introduction. 1.1. Gene therapy strategies Gene manipulation strategies employed in current human gene therapy can be broadly classified into gene replacement, gene addition and gene expression alteration therapies (5). -Gene replacement therapy: therapeutic approach for treating monogenic diseases which are caused by a single gene defect. Most of clinical gene therapy developments focus on this strategy, due to its simplicity and, also, to the ease of obtaining animal models for monogenic Inés Fernández Piñeiro 38 human diseases. The first gene therapy product to reach the market, called Glybera (Table 1), belongs to this category. It consists of a recombinant adeno-associated virus for treating lipoprotein lipase deficiency by replacing with a functional lipoprotein lipase gene. -Gene addition: therapeutic approach for complex disorders such as cancer and heart diseases which involve multiple genes deregulation and environmental factors. Thus, a deep understanding of these diseases mechanisms and the use of gene addition to supplement a therapeutic agent is required, since gene replacement is not feasible for these disorders. -Gene expression alteration targeting RNA: RNA can be an intermediate (mRNA) or final (microRNA) gene product with diverse functions in controlling gene expression. Two commonly utilized gene therapy strategies based on RNA biology are gene silencing and reprogramming mRNA splicing. -Gene silencing by RNA interference: gain-of-toxicity mutations lead to the production of toxic gene products which require a gene silencing strategy by RNA interference (RNAi). RNAi is a RNA-based gene silencing process which was discovered in 1998 by Fire and Mello (6), consisting in the binding of a molecule of RNA to a complementary mRNA and the consequent mRNA degradation or protein synthesis inhibition. The triggering RNA molecule can be a small interfering RNA (siRNA), which can also be used as the precursor short hairpin RNA (shRNA), or a microRNA. MicroRNAs seem to be safer, causing less cellular toxicity, and can target more than one mRNA and, therefore, modify the expression of different proteins. -Reprogramming messenger RNA splicing by antisense oligonucleotides (AONs): Many diseases are caused by mutations in genes that lead to disrupt in reading frame of pre-mRNAs. Thus, when this pre-mRNAs undergo splicing to form mature mRNAs composed of exons to be translated into proteins, the disruption in the reading frame of premRNA produce an altered mRNA and, therefore, an abnormal protein. AONs are designed to induce exon skipping or inclusion to modify this RNA splicing and produce a functional protein. 1.2. Gene delivery systems Nucleic acid delivery must overcome numerous barriers and obstacles before its therapeutic effect can be exerted. Naked nucleic acids are quickly degraded by nucleases in the Introduction 39 body and cleared via renal excretion. In addition to their instability, their hydrophilic nature, negative charge and high molecular weight prevent nucleic acids from penetrating biological barriers (7-9). These properties make necessary the development and inclusion of nucleic acids into gene delivery systems. There are two main approaches for gene delivery into target cells: viral and non-viral vectors. Viral vectors bind to target cells and introduce their genetic materials into the host cell as part of their replication process, referred as “transduction”. As non-viral vectors, physical and chemical approaches, among others, have been used for genetic transfer, which is referred as “transfection”. Non-viral vectors are safer and easier to be modified and scaled-up, as well as amenable to carry larger gene payloads, compared with viral vectors. However, they have a lower transfection efficiency (2, 10). -Non-viral vectors: the non-viral strategies for nucleic acid delivery can be divided in physical and chemical methods. -Physical mediated gene transfer: nucleic acids are delivered by compressed air or fluid (gene gun), or using ultrasound, which can force the genetic material into the target cell. Another approach is based in electroporation, consisting of membrane disruption with high-voltage electrical pulses. These techniques consist of applying a physical force that creates transient membrane holes and facilitate nucleic acid transfer to the cell nucleus (3). -Chemical mediated gene transfer: several chemicals have been used to protect and help nucleic acids to cross cell membranes, mainly condensing nucleic acids into nanocarriers. These nanocarriers can be grouped into inorganic, polymer-based and lipid-based systems, depending on the chemical used to condense the genetic material. Potential benefits of nanocarriers for gene delivery are their reduced toxicity, efficient packing and stabilization of high contents of genetic material, simple manufacturing, low cost and high versatility (11). -Viral vectors: viruses consist of ribonucleic acid (RNA) or deoxyribonucleic acid (DNA) surrounded by a protective protein coat (viral capsid) which helps the virus to attach to specific host cell receptors. Some viruses may also have a lipid bilayer envelope derived from the host cell’s membrane, and an outer layer of viral envelope made of glycoprotein. Vectors based on gammaretroviruses, adenoviruses, adeno-associated virus, herpes simplex virus and lentivirus are among the most widely used viral vectors. Although viral vectors are thought to Inés Fernández Piñeiro 46 2.2. Methods of preparation of polysaccharide nanoparticles Different methods have been utilized in the preparation of polysaccharide nanoparticles and selection of the method is greatly dependent on the nature of the therapeutic molecule (21, 22, 30-36). Briefly, techniques used in the preparation of nanoparticles loaded with thermosensitive or less stable substances such nucleic acids may be broadly classified as described below. -Emulsion cross-linking: An aqueous polysaccharide solution is emulsified with an oil phase that contains an appropriate concentration of surfactant. A cross-linker is added to the emulsion leading to the gelation of the emulsion droplets where the polymer is dissolved. This method has been used for the preparation of chitosan, alginate and dextran nanoparticles, among others. -Desolvation technique: polysaccharides aggregate forming nanoparticles by desolvation through the addition of desolvating agents like alcohols or salts, inducing a coacervation effect. This effect is observed because water-salt or alcohol interactions are more favourable than those occurring between the water and the polysaccharide. -Polyelectrolyte complexation: this is a method based on mixing aqueous solutions of two polymers carrying opposite charges. It has been extensively used to prepare natural polysaccharides-based particles and hydrogels. It is a good method to prepare nanoparticles for its mild preparation conditions and simple procedures, and it can avoid destroying the structure and property of the bioactive molecules. -Ionic cross-linking or ionic gelation: involves electrostatic interactions between charged polysaccharides and small ions of opposite charges in an aqueous environment, under mild conditions. -Self-assembly: polysaccharides grafted with lipid moieties of amphiphilic character, spontaneously form self-aggregates in aqueous solution. Consequently, polymeric micelles with core-shell structure are formed. -Nanoprecipitation: diffusion of an aqueous polymer solution in a water-miscible nonsolvent, resulting in the instantaneous formation of nanoparticles. This technique has been applied for chitosan, cyclodextrins and dextrin nanoparticles, among others. Introduction 47 3. Vascular endothelium 3.1. Endothelial cells heterogeneity The endothelium forms the inner cellular lining of blood vessels. Endothelial cells (ECs) have very distinct and unique functions besides merely providing a lining for vessel walls. Moreover, ECs show a huge heterogeneity, differing among the vascular tree and different organs and tissues to adapt to specific requirements. This heterogeneity can be explained by the enormous variety of tissue microenvironments to which ECs are exposed, although some properties are epigenetically fixed and not dependent on the surrounding environment. Vascular heterogeneity comprises morphologic, functional and molecular differences. Arterial and venous vessels possess distinct morphological and physiological characteristics. Arterial vessels present a layer of muscle cells and they deliver oxygen and nutrients from the heart to various tissues, where they branch into capillaries. Then, capillaries converge to form venous vessels that deliver deoxygenated blood back to the heart. ECs of capillaries acquire specific characteristics and functions regarding the surrounding tissue. For example, endothelium of the central nervous system forms the blood-brain barrier characterised by tight junctions, while endothelium of endocrine glands, pancreas, intestine and kidney are highly permeable due to the presence of pore-like fenestrations. Moreover, the molecular profile of ECs is often organ-specific due to the tissue-specific signal specializing them. Traditionally, three types of endothelial cells are recognised: continuous, discontinuous and fenestrated cells (Figure 1) (37, 38). -Continuous endothelium shows a basement membrane and lack of fenestrations which make it almost impermeable. It is found in arteries, veins, and capillaries of the brain, skin, heart and lung. -Discontinuous endothelium is present in the liver and bone marrow and it is characterised by the presence of large fenestrations without diaphragms and basement membrane. This endothelium also contains large circular pores within individual cells. -Fenestrated endothelium is observed in locations that are characterised by increased filtration or transendothelial transport such as exoand endocrine glands, gastric and intestinal mucosa and kidney glomeruli. It is characterised by the presence of fenestrations, which consist of transcellular pores with diaphragm and a basement membrane. Inés Fernández Piñeiro 48 Figure 1. Types of vascular endothelium (37). 3.2. Endothelium functions Vascular ECs have very distinct and unique functions, including a wide range of homeostatic functions (37-41). The main vascular endothelium functions are summarised below. -Blood haemostasis: the endothelium has an important role in maintaining blood haemostasis. ECs prevent thrombosis and maintain blood fluidity by means of different anticoagulant and procoagulant factors which are differentially expressed across the vascular tree. They can also regulate coagulation by regulating the expression of binding sites for anticoagulant and procoagulant factors on the cell surface. -Platelet and leukocyte interaction: platelet adhesion to and leukocyte rolling on the endothelium represent the initial stage of a multistep process leading to extravasation of white blood cells to sites of inflammation or infection, to platelet-leukocyte interaction and aggregation on a thrombogenic surface, and finally to vascular occlusion. Trafficking of leukocytes from blood to underlying tissue involves a multistep adhesion cascade that includes initial attachment, rolling, arrest and transmigration. This transmigration can occur between Introduction 49 ECs (paracellular pathway) or through ECs (transcellular pathway), and it takes place primarily in postcapillary venules. -Regulation of vascular tone and growth: endothelium secretes a variety of regulatory substances such as vasoactive substances, growth inhibitors, contraction-inducing factors. Deregulation of this endothelial-dependent system is involved in many cardiovascular diseases, such as hypertension and atherosclerosis. -Cell proliferation and angiogenesis: the endothelium is also involved in blood vessel formation, which is dependent upon signals exchanged between ECs and surrounding cells. Mature ECs are normally in a quiescent state, and neovascularization occurs predominantly by angiogenesis from pre-existing vessels or, to a lesser extent, by vasculogenesis or blood vessels formation by endothelial progenitor cells. -Permeability: the endothelium regulates the transport of fluids and solutes into and out of the blood. Such transport takes place primarily in the capillaries, the major exchange vessels of the circulation. Fluids and small solutes move passively across the endothelium via the paracellular route, whereas macromolecules use a transcellular route. This transcellular transport takes place by two different mechanisms: endocytosis and transcytosis. The endocytic pathway can occur via a nonspecific process or through a receptor-mediated endocytosis. The later, also called clathrin-mediated endocytosis, is performed by scavenger receptors which are particularly expressed in liver sinusoidal ECs. Clathrin-mediated endocytosis is responsible for uptake of macromolecules such as low density lipoprotein (LDL), transferrin, albumin and advanced glycosylation end products. In addition to endocytosis, macromolecules move across the endothelium through transcytosis, which is mediated by caveolae and vesiculo-vacuolar organelles (VVOs). Caveolae are membrane-bound vesicles that, with exception of liver sinusoids, are present in ECs to a greater extent than clathrin-coated pits, mainly in capillary endothelium. VVOs are focal collections of membrane-bound vesicles and vacuoles, and they are most commonly observed in venular endothelium than in capillaries. ECs were until recently considered to be just a lining for vessel walls, but it is now realised that ECs have important functions and perturbations of these functions are involved in many pathological conditions. Endothelial dysfunction is characterised by reduced vasodilation, a proinflammatory state and prothrombotic properties. It is associated with most forms of Inés Fernández Piñeiro 50 cardiovascular disease, such as hypertension, coronary artery disease, chronic heart failure, peripheral vascular disease, diabetes, chronic kidney failure, cancer and severe viral infections. 3.3. Role of endothelium in cancer Endothelial cells are in a quiescent state most of their adult life, but they can turn to a proliferative and migrative state after angiogenic stimuli from cancer cells. Diverse works suggest that angiogenesis is not only determined by cytokines and growth factors, but also endothelial cells metabolism interfere in this process. Thus, targeting endothelial metabolism is possible to target cancer progress (42). Cancer progression and metastasis is regulated by cross-talk between cancer cells and noncancer cells of tumour microenvironment. Tumour microenvironment is composed by nonmalignant cells which supports and promote tumour progression, including macrophages, fibroblasts, endothelial cells, infiltrating immune cells and extracellular matrix (43). These cellcommunication promotes tumour growth, angiogenesis, drug resistance, invasion and provides cancer cells with stem cell-like properties and epithelial-to-mesenchymal transition phenotypes. Moreover, these non-cancer cells are genetically more stable than cancer cells, thus targeting these cells can serve as an effective strategy to defeat cancer. Among non-cancer cells involved in tumour microenvironment, blood vessels endothelial cells play an important role. 3.3.1. Abnormalities in tumour endothelium Endothelium in tumour microenvironment differs from normal endothelium in respect of morphological, functional and genetic characteristics (44). There is evidence that these abnormalities in the tumour endothelium contribute to tumour growth and metastasis and, therefore, determining the biological basis underlying these abnormalities will facilitate the development of new antineoplasic therapies. Tumour ECs are more dilated and tortuous than normal ECs, with excessive branching, chaotic flow patterns and increased permeability to macromolecules through increased fenestrations and widened intercellular gaps, along with a high proliferative rate. These tumour ECs phenotypic heterogeneity is determined by diverse factors. The first factor to considered is the vascular bed of origin and the tumour microenvironment where ECs are exposed to abnormal conditions such as low pH, hypoxia, variable blood flow, hypoglycaemia, and growth factor and cytokines release by tumour cells and stromal cells. Moreover, epigenetic factors Introduction 51 play an important role in phenotypic heterogeneity. Thus, some site-specific epigenetic treads will be retained by ECs regardless of tumour microenvironment. In addition, there is some evidence that genetic alterations of tumour ECs may also influence phenotype heterogeneity. Thus, the presence of bone marrow-derived cells, such as endothelial progenitor cells, or the genetically instability of some tumour ECs may also have an impact on phenotype (41, 44). Some of these tumour endothelium abnormalities are summarised below. -Defective endothelial monolayer: tumour endothelial cells have an irregular shape and size, with fragile cytoplasmatic projections which may penetrate the vessel lumen creating small intercellular gaps in the vessel wall. -Large intercellular openings and holes: transcellular holes, fenestrations, channels and larger openings in the tumour blood vessels are probably for haemorrhage and plasma leakage observed in most tumours. -Abnormal sprouts: tumour vessels have thin cytoplasmic projections extending across the vessel lumen. The origin of these sprouts may be the oxygen seeking tip cells in hypoxic regions of the tumour microenvironment. -Altered gene expression: showing diverse patterns from different tumour types and stages of progression. For example, miRNAs are deregulated in many types of cancer and they have been demonstrated to influence the progression of the disease through alteration of the tumour microenvironment. Through miRNA manipulation in tumour endothelial cells, cancer cells are able to promote their angiogenic potential. Moreover, miRNAs have been demonstrated to be direct transferred between cancer cells and tumour microenvironment cells in exosomes or through direct cell contact. 3.3.2. Targeting endothelial cells in cancer therapy Diverse therapeutic approaches that target ECs angiogenic potential have been developed and reached the market. For example, anti-VEGF monoclonal antibody bevacizumab was the first commercially available angiogenesis inhibitor. It was approved to be used alone for glioblastoma that has not improved with other treatments and to be used in combination with other drugs to treat metastatic colorectal cancer, some non-small cell lung cancers, and metastatic renal cell cancer. Since then, other antiangiogenic drugs have been approved, Inés Fernández Piñeiro 52 including receptor tyrosine kinase inhibitors such as sorafenib, sunitinib, pazopanib, regorafenib and axitinib, or the VEGF-trap fusion protein aflibercept (45). However, the use of antiangiogenic therapy is still limited due to poor understanding of the benefit, as well as the side effects and drug resistances caused by this therapeutic (41, 46). More than 100 clinical trials have been conducted worldwide with antiangiogenic drugs, but the survival benefits of antiangiogenic drugs have been modest so far in the clinic. Surprisingly, the majority of patients stop responding to this therapy or do not respond at all, and this controversy has been highlighted lately by some preclinical studies suggesting that antiangiogenic drugs may lead to a more aggressive and invasive tumour phenotype (47). Nevertheless, targeting ECs at a transcriptional level has arisen as a therapeutic alternative. Considering that tumour ECs differ from normal ECs also at molecular level, this provides the scientific rationale for transcriptional targeting strategies. The distinct characteristics of tumour ECs make them an excellent target for gene therapy. As previously mentioned, one way ECs promote cancer progression is by altering the expression of microRNAs. Thus, therapeutic strategies aimed to regulate microRNA expression are being explored in preclinical studies (48). 4. Liver metastasis 4.1. Liver endothelium The liver receives 15 to 20% of the cardiac output and it has a dual blood supply: the hepatic artery, which delivers well oxygenated blood, and the portal vein, which delivers poorly oxygenated, nutrient-rich blood. As shown in Figure 2, both drain into the hepatic sinusoids, which represent the capillary network in the liver. After circulating in the sinusoids, blood empties into hepatic venules and ultimately in the hepatic vena cava. Liver endothelium is characterised by its heterogeneity. From portal vein to portal venules, the endothelial cells become spindle shaped, nonfenestrated, and possess short microvilli. At the transition point between the terminal portal venule and hepatic sinusoid, the ECs are smooth and large and contain many actin fibres, and together with Ito cells, they can control blood flow. In the case of the blood delivered by hepatic artery, the blood flow is regulated by a precapillary Introduction 53 sphincter consisting of tall ECs and smooth muscle cells at the junction between the hepatic artery and sinusoid (49). Figure 2. Liver blood circulation (49). 4.1.1. Liver sinusoidal endothelial cells Liver sinusoidal endothelial cells (LSECs) comprise 50% of the non-parenchymal cells of the liver. They form a discontinuous and fenestrated endothelium which functions as a selective sieve for fluids, solutes and particles from blood to hepatocytes via the space of Disse (Figure 3). The fenestrations have approximate diameters of 100 to 150 nm and provide open channels between the sinusoidal blood and the subendothelial space of Disse. Sieving plays an important role in lipoprotein metabolism, which finally occurs in the hepatocytes. LSECs also contribute to vasomotor tone, organogenesis and liver regeneration (50). In addition, LSECs also function as scavengers, eliminating soluble waste macromolecules from portal venous blood by receptor-mediated endocytosis, such as hyaluronan, acetylated low-density lipoprotein (LDL), denatured albumin, glycation end products and ovalbumin. Some of the receptors involve in this scavenger function are the mannose receptor, the hyaluronan receptors HARE and stabilin-2. Inés Fernández Piñeiro 54 Another function of LSECs is related to immunity, since they take up antigens through scavenger and mannose receptor and present them to lymphocytes via major histocompatibility complex molecules. In contrast with professional antigen presenting cells, LSECs antigen presentation to immune system results in immune tolerance rather than enhanced immunity. Thus, this tolerance prevents response to innocuous oral antigens. Figure 3. Liver sinusoidal endothelial cells characteristics and functions (50). The liver is one of the greatest leukocyte margination sites in the body. This leukocyte trafficking is mainly mediated by the sinusoids, and the migration mechanisms differs from those of other vascular beds. Thus, leukocyte accumulation and adhesion are not preceded by rolling and PECAM-1/CD-31 is not necessary. 4.2. Metastasizing to the liver The liver is the most frequently afflicted organ by metastasis for the majority of prevalent malignancies, and liver metastases are much more common than primary hepatic tumours (51). The liver is the main site of metastatic disease and a major cause of death from gastrointestinal malignancies such as colon, gastric and pancreatic carcinomas, as well as melanoma, breast Introduction 55 cancer and sarcomas (52). The main factors related to the great susceptibility of liver to metastases are: -Architectural and hemodynamic features: dual, slow and tortuous liver-specific microcirculation, along with the expression by non-parenchymal cells of surface molecules that facilitate attachment of circulating tumour cells, promote tumour cells hepatic retention. Moreover, LSECs fenestrations allow direct access of tumour cells to the basement membrane. -Regenerative capabilities: self-renewal and reconstruction capability of the liver can create a favourable environment for tumour growth. -Regional immune suppression: tolerant microenvironment permissive to foreign tumour cell survival and growth due to its constant exposure to inflammatory stimuli from the gut. 4.3. Liver metastatic tumour microenvironment The metastatic process to the liver involves four interrelated phases (51): 5. Tumour-infiltrating microvascular phase, which involves tumour cell arrest by LSECs that leads to tumour cell death or extravasation. 6. Interlobular pre-angiogenic micrometastasis phase, during which host stromal cells are recruited into avascular micrometastases. 7. Angiogenic micrometastasis phase, in which tumours become vascularised through several possible interactions with the microenvironment. 8. Growth phase that leads to the establishment of the macrometastasis. The hepatic metastatic tumour microenvironment is highly dynamic and is regulated by interactions between cellular and non-cellular components of the environment. These interactions are bidirectional, and communication occurs through soluble signalling factors, such as cytokines, chemokines and growth factor; receptor-mediated cell-cell and cellextracellular matrix contacts, and proteolytic enzymes. Each hepatic cell type is involved in different metastases phases and capable of playing tumouricidal and tumour promoting roles. -LSECs: the tumour-endothelial cell interaction determines the progression of the process. LSECs can play tumoricidal and tumour progression-promoting activities, and this role is determined indirectly by cytokines produced by Kupffer cells or directly by interactions with Annex I Nanocarriers for microRNA delivery in cancer medicine Annex I. Nanocarriers for microRNA delivery in cancer treatment 65 Abstract The number of deaths caused by cancer is expected to increase partly due to the lack of selectivity and undesirable systemic effects of current treatments. Advances in the understanding of microRNA (miRNA) functions and the ideal properties of nanosystems have brought increasing attention to the application of nanomedicine to cancer therapy. This review covers the different miRNA therapeutic strategies and delivery challenges for its application in cancer medicine. Current trends in inorganic, polymeric and lipid nanocarrier development for miRNA replacement or inhibition are summarized. To achieve clinical success, in-depth knowledge of the effects of the promotion or inhibition of specific miRNAs is required. To establish the dose and the length of treatment, it will be necessary to study the duration of gene silencing. Additionally, efforts should be made to develop specifically targeted delivery systems to cancer cells to reduce doses and unwanted effects. In the near future, the combination of miRNAs with other therapeutic approaches is likely to play an important role in addressing the heterogeneity of cancer. Keywords: cancer therapy, gene silencing, microRNA, nanocarriers, RNA interference Inés Fernández Piñeiro 66 1. Introduction 1.1. Limitations of current cancer therapies Cancer consists of a group of diseases characterized by uncontrolled division of abnormal cells that can invade and spread to other organs to form metastases. Neoplastic diseases exhibit distinctive capabilities such as proliferative signalling, evading growth suppressors, resisting cell death, enabling replicative immortality, inducing angiogenesis, and activating invasion and metastasis (1). Cancer is the second cause of death in developed countries and is associated with aging of the population and lifestyle (2). The number of deaths caused by cancer was estimated to be 8.2 million in 2012, corresponding to 13% of all deaths. Although approximately two-thirds of cancer cases are cured due to advances in diagnosis, health care and treatment, the number of deaths is expected to increase. The lack of selective delivery of current treatments and the consequent systemic toxicity are among the major reasons for this trend (3). Radiotherapy and surgery are the most effective treatments for local and non-metastatic tumours, whereas chemotherapy, hormone and biological therapies are currently used for the treatment of metastatic cancers (4). However, despite significant progress in cancer treatments, the low selectivity, undesirable systemic effects and dose-limiting toxicity of current treatments make them nonspecific and non-ideal therapeutic approaches (5). 1.2. Nanomedicine in cancer treatment Recent advances in nanomedicine applied to cancer therapeutics may help to overcome the existing limitations of antineoplastic drugs. Nanomedicine includes the design and development of nanoscopic delivery vehicles and diagnostic agents. These delivery systems can improve drug stability, increase the circulation time and selectively accumulate at the tumour site (6, 7). Their nanometric size facilitates accumulation preferably at the tumour site due to the enhanced permeability and retention effect (EPR). The EPR effect is based on the presence of fenestrated blood vessels in the tumour, leading to extravasation of nanocarriers through a passive mechanism. Additionally, active targeting can be achieved through functionalization of the nanosystems with specific ligands for receptors on target cells. Furthermore, current progress in understanding the molecular pathways and functions of cancer have enabled the identification of new targets and the development of novel therapeutic Annex I. Nanocarriers for microRNA delivery in cancer treatment 67 strategies. Hence, the finding of specifically altered signalling networks in cancer cells and the ideal properties of nanosystems have brought increasing attention to the application of nanomedicine to cancer treatment. A wide range of materials has been employed in the synthesis of nanocarriers and can be grouped into inorganic, polymer and lipid-based materials. In this review, we will cover nanosystems for microRNA (miRNA) delivery developed to date for cancer therapy. 2. MicroRNA 2.1. MicroRNA mechanism RNA interference (RNAi) is an evolutionarily conserved method of gene expression regulation. RNAi is based on a post-transcriptional pathway triggered by a double-stranded RNA (dsRNA) that leads to sequence-specific silencing of a messenger RNA (mRNA) (8). RNAi was discovered by Fire and Mello in 1998 (9) and included endogenous (miRNA) or exogenous (siRNA and shRNA) RNAs. Gene silencing occurs when the double-stranded RNA molecules incorporate into the RNA-induced silencing complex (RISC). Then, the guide strand or anti-sense strand guides RISC to the complementary or near-complementary region of the target mRNA (10). miRNAs are only partially complementary to their target mRNAs and result in their degradation or translational inhibition, whereas siRNAs and shRNAs bind completely to and cleave the complementary strand (11-13). Toxicity associated with siRNA and shRNA overexpression is being debated; however, miRNA therapeutics seem to be safer and do not compromise the gene knockdown efficacy (14-16). Functional miRNAs are produced from the cleavage of pre-miRNAs in the cytoplasm. Mature miRNAs are 20-23 base pair double-stranded molecules comprised of a guide and a passenger strand that is released after loading into RISC, as shown in Figure 1. Due to the ability of the miRNA to inhibit gene expression by partial complementarity to the mRNA, one miRNA can bind to different mRNAs and thus affect the expression of multiple genes. Inés Fernández Piñeiro 68 Figure 1. Gene silencing mechanism of miRNAs. The long primary miRNA transcripts (pri-miRNAs) obtained from the transcription of miRNA genes are cleaved by the Drosha-DGCR8 complex in the nucleus to form pre-miRNA hairpins. These pre-miRNAs are exported by Exportin 5 into the cytoplasm and cleaved by the DicerTRBP complex into miRNA duplexes, which are incorporated into the miRNAinduced silencing complex (miRISC). AGO, which is a component of the RISC, releases and degrades the passenger strand while RISC is guided by the guide strand to the target mRNA, resulting in translation repression or mRNA degradation. 2.2. MicroRNAs and cancer The use of miRNAs for cancer therapy is based in the finding that miRNA expression is deregulated in cancer tissues and the ability of miRNAs to target multiple genes and alter cancer phenotypes (17, 18). Cancers are complex diseases involving deregulated expression of Annex I. Nanocarriers for microRNA delivery in cancer treatment 69 multiple genes, whereas miRNAs can modulate different disease pathways and increase the chances of eliminating the cancer. Moreover, distinctive miRNA expression profiles have been associated with specific cancer types, allowing for the discrimination and identification of poorly differentiated tumours (19, 20). Thus, miRNAs have shown relevant clinical utility for cancer therapeutics and diagnosis. 2.3. MicroRNA therapeutic strategies: sense and antisense microRNAs In neoplastic diseases, miRNAs can be downregulated when they function as tumour suppressors or overexpressed when they function as oncogenes. Hence, two therapeutic approaches are currently being used to modulate miRNA functions: restoring miRNA activity using a synthetic miRNA and inhibiting the function of a miRNA through anti-miRNA oligonucleotides. In situations where miRNAs are down-regulated, replacement therapy with miRNA mimics is used to restore miRNA levels and their tumour suppressive properties. Because the objective of this replacement therapy is to accomplish biological functions that are identical to the endogenous miRNAs, miRNA mimics should be loaded onto RISC to silence their target mRNAs. For this reason, double-stranded miRNA mimics are preferred over single-stranded mimics because the duplex structure has been found to facilitate RISC loading and thereby enhance the gene silencing efficacy (21). In the case of overexpressed oncogenic miRNAs, the most widely used strategy is based on the use of miRNA antagonists to inhibit miRNA expression. The most common antisense approach is the use of single-stranded oligonucleotides that are partially or completely complementary to the target miRNA. The complementary binding of the antagonist to the endogenous miRNA prevents its processing by RISC. These antisense oligonucleotides (known as anti-miRNAs) are chemically modified to increase their binding affinity for the miRNA (22). Accordingly, miRNAs could act as therapeutic agents or as therapeutic targets (23). 2.4. Challenges in microRNA delivery Similar to other therapeutic oligonucleotides, miRNA delivery is a major challenge because naked miRNAs are quickly degraded by nucleases and cleared via renal excretion. Moreover, RNA administration may induce innate immune responses, leading to unwanted toxicities. In addition to their instability and toxicity, their hydrophilic nature, negative charge Inés Fernández Piñeiro 70 and high molecular weight prevent nucleic acids from crossing cell membranes. Once inside cells, miRNA should achieve endosomal escape and compete with endogenous RNAi pathways (23-25). Because miRNAs are designed to target multiple pathways, they may cause off-target gene silencing that can induce toxic effects. Finally, insufficiency or saturation of the enzymes involved in miRNA processing may lead to inefficient gene silencing efficacy. In this case, the use of mature miRNAs can avoid saturation of the processing enzymes (23). To overcome these hurdles, diverse strategies and delivery systems are currently being studied. Traditionally, delivery systems for nucleic acids are classified as viral and non-viral vectors, but this review will be limited to non-viral vectors for miRNA therapy with a focus on nanocarriers. 3. MicroRNA Delivery 3.1. Chemical modifications and oligonucleotide conjugates One of the approaches used to overcome the poor stability and immune responses of miRNAs is based on chemical modifications. These can reduce the off-target effects associated with miRNAs. The principal strategies are based on i) ribose 2’-OH group modification, ii) locked nucleic acids (LNA), iii) backbone modifications and iv) peptide nucleic acids (PNA) (26). i. The ribose 2’-OH group is easily attacked by nucleases. Substitution of the ribose 2’- OH group for 2’-O-methyl, 2’-O-fluoro or 2’-methoxyethyl has been shown to increase the binding affinity, stability and gene silencing activity of anti-miRNAs (27). ii. LNAs are antisense oligonucleotides based on ribose-modified RNA nucleotides with an extra bridge connecting the 2' oxygen and 4' carbon to create a stable “locked ring conformation”. LNAs exhibit increase stability and a reduced immune response but may lead to a reduction in activity (17). iii. The most widely used strategy among backbone modifications is phosphorotioate modifications in which one of the non-bridging phosphate oxygen atoms is replaced with a sulphur atom. This modification has been demonstrated to increase nuclease resistance but shows a short circulation life and low binding affinity (28) iv. PNA consists of uncharged oligonucleotide analogs in which the sugar-phosphodiester backbone has been replaced by N-(2-aminoethyl)-glycine units. The lack of charge of Annex I. Nanocarriers for microRNA delivery in cancer treatment 71 PNA eliminates the need for transfection reagents and enhances oligonucleotide stability with low toxicity (29). Another strategy consists of conjugation with small transport domains, such as aptamers and cell-penetrating peptides. Aptamers are small single-stranded oligonucleotides with a particular three-dimensional structure that can bind to specific surface receptors and act as drug delivery agents. Conjugation of miRNAs to aptamers has been used to specifically target the nucleic acid to cells expressing the ligands recognized by the aptamer (30). For example, Rohde et al. used the ubiquitously expressed transferrin receptor as an aptamer for miR-126 delivery (31). Conjugation to cell-penetrating peptides is intended to enable crossing of cell and endosomal membranes (32). For example, Fabani et al. conjugated a PNA anti-miR-122 to the cell-penetrating peptide penetratin to deliver the oligonucleotide in vitro (33). Although these modifications and strategies improve miRNA stability and the pharmacokinetic profile and reduce the immune response, increased toxicity and reduced activity have been observed. Consequently, the high doses of modified miRNAs required result in off target effects and saturation of the silencing machinery. Hence, the lack of efficient and specific delivery of miRNAs to tumour cells remains a significant challenge. 3.2. Nanocarriers Nanocarriers have recently become popular for miRNA delivery in an attempt to enhance cellular uptake and pharmacological effectiveness and reduce toxicity. Non-viral vectors have the advantage of being highly versatile; thus, it is possible to modify the systems with targeting ligands or polyethylene glycol (PEG) molecules to achieve site-specific delivery or prolong the circulation time of the system. Nanocarriers are safe and require simple manufacturing; moreover, they are characterized by their low immunogenicity, low cost and versatility. Table 1 summarizes miRNA nanocarries that, to the best of our knowledge, have been tested in vivo, while Table 2 lists some of the functions of the miRNAs here presented. Inés Fernández Piñeiro 78 the crosslinker ethylene glycol dimethacrylate (EGDMA) and the neutral monomer acrylamide (AAM), have also been used. Liu et al. synthesized anti-miR-21-loaded nanocapsules in which anti-miR-21 molecules were encapsulated within a thin network of polymer shells (58). The crosslinker molecules were degradable under the acidic conditions of the endosomes, leading to miRNA release. The system was able to suppress tumour growth after intratumoural administration in mice bearing glioma U87 subcutaneous tumours. Functionalized polycarbonates have also been used as polymers for miRNA delivery systems. Mittal et al. developed gemcitabine-conjugated cationic copolymers complexed with miR-205 to form micelle-like polyplexes. miR-205 is a tumour suppressor miRNA that was shown to chemosensitize gemcitabine-resistant MIA PaCa-2R pancreatic cancer cells (59). The micelles were formed from an amphiphilic copolymer containing a PEG hydrophilic segment and a polypropylene carbonate hydrophobic chain with different cationic pendent groups. These micelleplexes were demonstrated to be effective in reverting the chemo-resistance, invasion and metastasis of gemcitabine-resistant pancreatic cancer cells in vitro. In vivo studies showed a reduction in the tumour growth rate and tumour weight in a pancreatic cancer xenograft model treated with the miR-205-gemcitabine complexes. Choi et al. developed an engineered nanocarrier for RNA delivery composed of a hydrophilic shell of the CD44-targeting ligand hyaluronan, a hydrophobic core of 5β-cholanic acid as a drug reservoir, a RNA binding site of the artificial phosphate receptor Zn(II)- dipicolylamine (DPA/Zn) and a calcium phosphate layer that increased the protection and pHdependent RNA release. The system demonstrated the ability to efficiently deliver miR-34a to a colon carcinoma cell line and tumour mouse model after intravenous administration (60). This nanosystem does not contain a cationic substance to bind the nucleic acids and thus achieves higher specific accumulation and lower toxicity than conventional gene transfection systems. Liu et al. synthesized gelatinases-stimuli nanoparticles by polymerization of poly(Ԑcaprolactone) (PCL) as a hydrophobic core and PEG as a hydrophilic corona connected by a tumour-specific gelatinases-cleavable peptide (61). This system demonstrated the ability to synergistically deliver docetaxel and miR-200c, which is known to improve sensitivity of cancer cells to chemotherapy, in vitro and in vivo, thereby suppressing tumour growth in a xenograft gastric cancer mouse model. Annex I. Nanocarriers for microRNA delivery in cancer treatment 79 3.2.2.2. Natural polymers i. Chitosan Chitosan is one of the most studied natural polymers for gene delivery. It is a biocompatible, biodegradable and safe polysaccharide that has been explored as a miRNA delivery system. Deng et al. co-encapsulated miR-34a and doxorubicin into nanocomplexes composed of hyaluronic acid and chitosan (62). The authors reported efficient co-delivery of miR-34a and doxorubicin into the breast cancer cell line MDA-MB-231, thereby achieving synergistic effects on tumour suppression. Santos-Carballal et al. studied the effect of the degree of acetylation and the molecular weight of chitosan on miRNA delivery. The authors showed that polyplexes formed of chitosan and miR-145 were efficiently transfected into MCF-7 breast cancer cells in vitro (63). ii. Peptides and proteins Peptide and protein-based carriers have been widely explored for nucleic acid delivery because they can condense nucleotides through electrostatic interactions with the positive charges of the amino acids. Moreover, amino acids contribute to endosomal escape, targeted delivery and bioreversible polyplex stabilization of the system. Nanosystems containing peptides and proteins are presented below. Wang et al. prepared miR-34a self-assembled nanocomplexes composed of cationic protamine sulphate and negatively charged hyaluronic acid (HA) (64). HA was used due to its specific binding ability for the overexpressed CD44 receptors of tumours. miR-34a was efficiently delivered into both triple-negative breast cancer cells and a xenograft mouse breast cancer model in vivo. Hao et al. constructed a nanosystem for tumour suppressors miR-15a and miR-16-1 delivery based on an aptamer-conjugated atelocollagen (65).The RNA aptamer A10-3.2 was used to target prostate-specific membrane antigen (PSMA) on the cell surface of prostate cancer cells. Atelocollagen is a type I collagen, which is a fibrous protein involved in maintaining the morphology of tissues and organs. This system effectively delivered both miRNAs to PSMAoverexpressing pancreatic cancer cells in vitro and in vivo. Inés Fernández Piñeiro 80 Song et al. studied the efficacy of the antimiR-21/R3V6 peptide complex in the induction of apoptosis in C6 and A172 glioblastoma cells (66). The system was demonstrated to be more efficient than PEI and Lipofectamine and exhibited less cytotoxicity. 3.2.3. Lipid-based systems Lipids are among the most widely studied materials for nucleic acid delivery. They can be grouped into cationic lipids, ionisable lipids and helper lipids. 3.2.3.1. Cationic lipids Cationic lipids can interact with miRNAs to form lipoplexes. Generally, cationic lipids are formed by a hydrophilic head and a hydrophobic chain that will determine the transfection efficacy and toxicity of the system. Among the lipid-based systems, the most widely used cationic lipids for RNA delivery are 1,2-di-O-octadecenyl-3-trimethylammonium propane (DOTMA) and 1,2-dioleoyloxy-3-trimethylammonium propane (DOTAP). Wang et al. used the cationic lipid DOTMA to complex miR-122, which is a liver-specific miRNA associated with cancer and other hepatic diseases, and the unsaturated fatty acid oleic acid as a helper lipid in liposomal nanoparticles (67). DOTMA-oleic acid lipid nanoparticles demonstrated hepatic delivery of miR-122 when the system was administered intravenously. Wu et al. condensed miR-133b, a potential tumour suppressor, with DOTMA to form lipoplexes for the treatment of non-small cell lung cancer. This system achieved much higher lung accumulation and induction of miR-133b expression than standard transfection agent siPORT NeoFX complexes (68). The same authors successfully tested the delivery efficiency of miR-29b, which has been found to be involved in cell proliferation and apoptosis, by the lipoplexes in a xenograft mouse model of non-small cell lung cancer and achieved significant inhibition of tumour growth (69). Lee et al. engineered liposomes composed of the cationic lipids DOTAP and pegylated 1,2-distearoyl-sn-glycero-3-phosphoethanolamine (DSPE-PEG) containing cyanuric groups to chemically conjugate the protein ephrin-A1 for delivery of the let-7a miR to lung cancer cells (70). The ephrin-A1 protein is a well-known ligand of the overly expressed EphA2 receptor of lung cancer cells. This carrier inhibited effective cell proliferation, migration and tumour growth of malignant pleural mesothelioma and non-small cell lung cancer cell lines. Annex I. Nanocarriers for microRNA delivery in cancer treatment 81 Zhang et al. associated a histidine-rich antimicrobial peptide ([D]-H6L9) to the surface of DOTAP-soybean phospatidylcholine-DSPE-PEG2000 liposomes (71). The liposomes successfully associated and delivered anti-miR-10b and paclitaxel in vitro and in vivo, thereby delaying tumour growth and reducing lung metastases in a murine metastatic mammary tumour model. The histidines of [D]-H6L9 helped the system escape endosomes/lysosomes due to their protonation under acidic pH and their consequent membrane lytic effect Biocompatible and biodegradable ethylphosphocoline lipids have also been studied as cationic substances (72). Passadouro et al. developed 1-palmitoyl-2-oleoyl-sn-glycero-3ethylphosphocholine (EPOC)-cholesterol cationic liposomes associated with albumin and demonstrated their ability to inhibit miR-21, miR-221, miR-10b and miR-222 in pancreatic cancer cells and their synergistic antitumour effect when co-administered with sunitinib (73). Shi et al. utilized the cationic lipid dimethyldioctadecylammonium bromide (DDAB) to associate miR-34a in solid lipid nanoparticles. The solid lipid nanoparticles were composed of glycerol monostearate, cholesterol, soy phosphatidylcholine and DDAB and were able to deliver miR-34a alone (74) and simultaneously with paclitaxel in vivo (75). These nanoparticles demonstrated passive targetability to lung metastasis of the murine B16F10-CD44+ melanoma model and higher inhibition of tumour growth than the single drug-loaded solid lipid nanoparticles. 3.2.3.2. Ionizable lipids and helper lipids Early cationic lipids, such as DOTMA and DODAC, contain a permanently positive quaternary amine, whereas ionizable cationic lipids, such as 1,2-dioleoyl-3dimethylammonium propane (DODAP) and 1,2-dioleyloxy-N,N-dimethyl-3-aminopropane (DODMA), have tertiary amine positive charges at acidic pH but neutral charges at physiological pH (76). Moreover, the incorporation of helper and neutral lipids, such as cholesterol and saturated phosphatidylcholines (PC), has been used to increase system stability and transfection efficiency. Daige et al. used liposomes to deliver miR-34a into mice with orthotopic Hep3B and HuH7 liver cancer xenografts, causing tumour regression in both models (77). These registered liposomes (known as SMARTICLES®) are formed of the lipids 1-palmitoyl-2-oleoylsnglycero-3-phosphocholine (POPC), 1,2-dioleoyl-snglycero-3-phosphoethanolamine Inés Fernández Piñeiro 82 (DOPE), cholesteryl hemisuccinate (CHEMS), and cholesteryl-4-([2-(4morpholinyl)ethyl]amino)-4-oxoburanoate (MOCHOL) (78). The outstanding property of this system is its amphoteric behaviour because the particles are cationic during elaboration and anionic at physiological pH to facilitate movement across physiological membranes. Costa et al. developed glioblastoma-targeted stable nucleic acid lipid particles (SNALPs) consisting of liposomes coupled with chlorotoxin (CTX), which is a peptide that acts as a specific marker for gliomas. The liposomes formed of DODAP, 1,2-distearoyl-sn-glycero-3phosphocholine (DSPC), C16 Ceramide PEG2000 and 1,2-distearoyl-sn-glycero-3phosphatidylethanolamine (DSPE)-PEG-Maleimide were loaded with locked nucleic acidmodified anti-miR-21. The authors demonstrated tumour accumulation of the system after systemic injection and enhanced anti-tumoural activity after combination with orally administered sunitinib (79). Liu et al. studied the effect of co-treatment of miR-506, which is known to regulate proliferation and therapy response of cancer cells, loaded into 1,2-dioleoyl-sn-glycero-3phosphocholine (DOPC) liposomes and the antineoplastics cisplatin and olaparib in an orthotopic ovarian cancer model (80). The intraperitoneal administration of this therapy resulted in the reduction of tumour growth through miR-506 inhibition of the epithelial-tomesenchymal transition. Moreover, PEG lipids are incorporated into liposomes to prolong the circulation time, stabilize the system and reduce immunostimulation (17). However, pegylation can reduce cellular uptake of the system. To overcome this problem, D-α-tocopheryl polyethylene glycol succinate (TPGS) formed by vitamin E and PEG combines the advantages of PEG but promotes cellular uptake. Dai et al. reported the systemic delivery of paclitaxel and the let-7b miRNA loaded in micelles composed of an amphiphilic copolymer made from PEG 5000, D-αtocopherol (vitamin E) and the cationic moiety diethylentriamine (81). The system demonstrated marked potentiation due to the antiproliferative activity of paclitaxel and the let7b miRNA both in vitro and in vivo in mice bearing KRAS mutant tumour xenografts. 3.2.3.3. Lipopolyplexes Lipopolyplexes are systems composed of both lipids and polymers to address the limitations and combine the advantages of lipid-based and polymer-based systems. Chen et al. Annex I. Nanocarriers for microRNA delivery in cancer treatment 83 developed DOTAP liposome-polycation-hialuronic acid nanoparticles modified with the monoclonal antibody GC4 single chain variable fragment (scFv) for the delivery of miR-34a and siRNA to B16F10 melanoma lung metastasis in a murine model (82). The GC4 scFvtargeted system consists of a cationic liposome of DOTAP and cholesterol encapsulating a complex composed of HA, siRNA or miRNA and protamine. The intravenous injection of the system resulted in the inhibition of the tumour load in the lungs, particularly when antibodytargeted nanoparticles containing miRNA and siRNA were applied simultaneously. Moreover, the presence of the pegylated-targeting scFv reduced the toxicity of the system. Wang et al. developed near-infrared (NIR) laser-activated nanoparticles containing ammonium bicarbonate that could escape endosomes/lysosomes when heated due to conversion of encapsulated ammonium bicarbonate to gases (83). The nanoparticles consisted of a waterin-oil-in-water structure where indocyanine green (with a NIR laser-activated photothermal effect), miR-34a and ammonium bicarbonate were encapsulated in the hydrophilic core dispersed in the hydrophobic layer formed by PLGA, pluronic F127 and the phospholipid dipalmitoyl phosphatidylcoline (DPPC). Moreover, a more external hydrophilic layer contained hyaluronic acid and chitosan-pluronic F127. This thermal responsive system demonstrated cytosolic delivery of miR-34a both in vitro and in vivo in prostate cancer stem cells. 3.2.4. Clinical trials Despite the huge attention received by this field, most of the current miRNA non-viral delivery systems for cancer treatment are still in preclinical studies. The first miRNA vector to enter the clinic (MRX34) was a miR-34-loaded liposome for the treatment of solid tumours and hematologic malignancies in a multicentre (84) phase I clinical trial (https://www.clinicaltrials.gov/ct2/show/NCT01829971). A novel delivery system based on non-living nano-sized cells (85) known as EDVTM nanocells has recently entered phase I clinical trials for malignant pleural mesothelioma and non-small cell lung cancer treatment. These nanoparticles contain miR-16 and miR-15 and are targeted with an anti-epidermal growth factor receptor monoclonal antibody (https://clinicaltrials.gov/ct2/show/NCT02369198). Inés Fernández Piñeiro 84 4. Future directions The advances achieved in understanding miRNA functions and their roles in cancer have generated great expectations for miRNA therapeutics. The awarding of the Nobel Prize to Andrew Fire and Craig Mello in 2006 for the discovery of double-stranded RNA interference described in 1998 (9) attracted considerable attention and investment in the field. However, in vivo miRNA delivery has been challenging due to the physicochemical properties of miRNAs and biological barriers. The lack of rapid RNAi-based therapeutic development has limited the enthusiasm of pharmaceutical companies. Nevertheless, the development of non-viral systems for miRNA delivery has undergone a considerable boost in recent years, providing solutions for many of the challenges hindering miRNA therapeutic success. New opportunities and solutions for miRNA therapeutics have been created from progress in nanotechnology, which should reawaken the interest of pharmaceutical companies in the field (86). The ultimate goal of miRNA delivery research is to achieve an efficient delivery system that enables the translation of RNAi therapeutics into the clinic. Although there is evidence of a miRNA antineoplastic effect in several in vitro and in vivo preclinical studies, only two miRNA delivery systems for cancer treatment are currently undergoing clinical trials. To achieve clinical success, in-depth knowledge of miRNA biology is required. Because one miRNA can regulate the expression of multiple genes through different signalling pathways, all effects that promote or inhibit specific miRNAs should be completely recognized and characterized. To establish the dose and the length of treatment, it will also be necessary to study the duration of gene silencing. Efforts should also be made to develop targeted delivery systems. Delivery systems should be able to target specific cancer cells. Targeted therapies may reduce therapeutic doses and prevent possible toxic effects in other cells. For this reason, novel findings concerning cancer cells and the tumour microenvironment will play crucial roles in the emergence of new targets and targeting moieties for nanosystem functionalization. In addition to targeted delivery systems, new non-invasive imaging techniques should be developed. Monitorization of the in vivo distribution of the therapeutic nanocarrier may help optimize and predict treatment efficacy. Annex I. Nanocarriers for microRNA delivery in cancer treatment 85 Although the field has come a long way since the discovery of RNAi over a decade ago, miRNA therapy is still in its infancy. Cancer is a heterogeneous disease in which many pathways are simultaneously altered. Thus, the delivery of multiple miRNAs at the same time may be a more effective way to fight cancer in the near future. In addition to its potential as single therapy, the combination of miRNAs with other therapeutic approaches is likely to focus researchers’ attention as a means to address the heterogeneity of cancer. Table 1. Recent miRNA nanocarriers that have been tested in vivo Delivery system Targeted miRNA Therapeutic strategy Cancer type Route Ref. Inorganic systems Anti-GD2-conjugated silica nanoparticles miR-34a Replacement Neuroblastoma Intravenous (37) Polymer-based systems PEI polyplexes miR-145 and miR-33a Replacement Colon carcinoma Intravenous and intratumoural (44) Polyarginine-modified PEI polyplexes miR-145 Replacement Prostate cancer Intravenous (46) Polyurethane-short branch PEI polyplexes miR-145 Replacement Lung adenocarcinoma Intratumoural (47) PLGA-PEG nanoparticles MiR-122 Replacement Colon carcinoma Intravenous (49) Penetratin-coated PLGA nanoparticles MiR-155 Inhibition Lymphoma Intratumoural (52) PAMAM dendrimers embedded in a dextran hydrogel MiR-205 and miR-221 Replacement and inhibition Breast cancer Intratumoural (56) PLA-b-PDMAEMA dendrimers MiR-21 Inhibition Glioma Intratumoural (57) Methacrylamide/Ethylene glycol dimethacrylate/Acrylamide nanocapsules MiR-21 Inhibition Glioma Intratumoural (58) Polypropylene carbonate-PEG micelles MiR-205 Replacement Prostate cancer Intratumoural (59) Hyaluronan/Cholanic acid/Zn-dipicolylamine nanoparticles MiR-34a Replacement Colon carcinoma Intravenous (60) Polycaprolactone-PEG nanoparticles MiR-200c Replacement Gastric cancer Intravenous (61) Hyaluronic acid/Chitosan nanoparticles MiR-34a Replacement Breast cancer Intravenous (62) Hyaluronic acid/Protamine sulphate nanocapsules MiR-34a Replacement Breast cancer Intravenous (64) Inés Fernández Piñeiro 86 Aptamer-conjugated atelocollagen nanoparticles MiR-15a and miR-16-1 Replacement Breast cancer Intravenous (65) Lipid-based nanosystems DOTMA/Cholesterol/TPGS lipoplexes MiR-29b Replacement Non-small cell lung cancer Intravenous (69) DOTAP/SPC/DSPE-PEG lipoplexes MiR-10b Inhibition Breast cancer Intravenous (71) DDAB nanoparticles MiR-34a Replacement Melanoma lung metastasis Intravenous (74, 75) MOCHOL/CHEMS/DOPE/POPC lipoplexes MiR-34a Replacement Liver cancer Intravenous (77) Chlorotoxin-targeted stable nucleic acid lipid particles MiR-21 Inhibition Glioma Intravenous (79) DOPC lipoplexes MiR-206 Replacement Ovarian cancer Intraperitoneal (80) PEG 5000-D-α-tocopherol-diethylenetriamine micelles Let-7b Replacement Non-small cell lung cancer Intravenous (81) GC4-ScFv-modified DOTAP liposomes-polycationhialuronic acid nanoparticles MiR-34a Replacement Melanoma lung metastasis Intravenous (82) Hyaluronic-DPPC-PLGA-Pluronic F127 nanoparticles MiR-34a Replacement Prostate cancer Intravenous (83) This table provides examples of some of the most recent miRNA delivery systems that have demonstrated therapeutic efficacy in animal models of cancer. Anti-GD2: disialoganglioside GD2-antibody; PEI: polyethylenimine; PLGA: poly(lactic-co-glycolic acid); PEG: polyethylene glycol; PMAM: poly(amidoamine); PLA-b-PDMAEMA: polylactic acid polymethylaminoethyl methacrylate copolymer; DOTMA: 1,2-di-O-octadecenyl-3-trimethylammonium propane; TPGS: D-α-Tocopheryl polyethylene glycol succinate; DOTAP: 1,2dioleoyloxy-3-trimethylammonium propane; SPC: soybean phosphatidylcholine; DSPE-PEG: 1,2-Distearoyl-sn-glycero-3phosphoethanolamine; DDAB: dimethyldioctadecylammonium bromide; MOCHOL: cholesteryl-4-([2-(4-morpholinyl)ethyl]amino)-4oxoburanoate; CHEMS: cholesteryl hemisuccinate; DOPE: 1,2-dioleoyl-snglycero-3-phosphoethanolamine; POPC: 1-palmitoyl-2-oleoylsnglycero-3-phosphocholine; DOPC: 1,2-Dioleoyl-sn-glycero-3-phosphocholine; GC4-scFv: monoclonal antibody GC4 single-chain variable fragment; DPPC: dipalmitoyl phosphatidylcoline. Annex I. Nanocarriers for microRNA delivery in cancer treatment 87 Inés Fernández Piñeiro 94 44. Ibrahim AF, Weirauch U, Thomas M, Grunweller A, Hartmann RK, Aigner A. MicroRNA replacement therapy for miR-145 and miR-33a is efficacious in a model of colon carcinoma. Cancer Res. 2011 Aug 1;71(15):5214-24. 45. Che HL, Lee HJ, Uto K, Ebara M, Kim WJ, Aoyagi T, et al. Simultaneous Drug and Gene Delivery from the Biodegradable Poly(epsilon-caprolactone) Nanofibers for the Treatment of Liver Cancer. J Nanosci Nanotechnol. 2015 Oct;15(10):7971-5. 46. Zhang T, Xue X, He D, Hsieh JT. A prostate cancer-targeted polyarginine-disulfide linked PEI nanocarrier for delivery of microRNA. Cancer Lett. 2015 Sep 1;365(2):156-65. 47. Chiou GY, Cherng JY, Hsu HS, Wang ML, Tsai CM, Lu KH, et al. Cationic polyurethanesshort branch PEI-mediated delivery of Mir145 inhibited epithelial-mesenchymal transdifferentiation and cancer stem-like properties and in lung adenocarcinoma. J Control Release. 2012 Apr 30;159(2):240-50. 48. Gao S, Tian H, Guo Y, Li Y, Guo Z, Zhu X, et al. miRNA oligonucleotide and sponge for miRNA-21 inhibition mediated by PEI-PLL in breast cancer therapy. Acta Biomater. 2015 Oct;25:184-93. 49. Wang TY, Choe JW, Pu K, Devulapally R, Bachawal S, Machtaler S, et al. Ultrasoundguided delivery of microRNA loaded nanoparticles into cancer. J Control Release. 2015 Apr 10;203:99-108. 50. Devulapally R, Sekar TV, Paulmurugan R. Formulation of Anti-miR-21 and 4Hydroxytamoxifen Co-loaded Biodegradable Polymer Nanoparticles and Their Antiproliferative Effect on Breast Cancer Cells. Mol Pharm. 2015 Jun 1;12(6):2080-92. 51. Devulapally R, Sekar NM, Sekar TV, Foygel K, Massoud TF, Willmann JK, et al. Polymer nanoparticles mediated codelivery of antimiR-10b and antimiR-21 for achieving triple negative breast cancer therapy. ACS Nano. 2015 Mar 24;9(3):2290-302. 52. Babar IA, Cheng CJ, Booth CJ, Liang X, Weidhaas JB, Saltzman WM, et al. Nanoparticlebased therapy in an in vivo microRNA-155 (miR-155)-dependent mouse model of lymphoma. Proc Natl Acad Sci U S A. 2012 Jun 26;109(26):E1695-704. Annex I. Nanocarriers for microRNA delivery in cancer treatment 95 53. Cheng CJ, Saltzman WM. Polymer nanoparticle-mediated delivery of microRNA inhibition and alternative splicing. Mol Pharm. 2012 May 7;9(5):1481-8. 54. Ananta JS, Paulmurugan R, Massoud TF. Nanoparticle-Delivered Antisense MicroRNA21 Enhances the Effects of Temozolomide on Glioblastoma Cells. Mol Pharm. 2015 Dec 7;12(12):4509-17. 55. Wang S, Zhang J, Wang Y, Chen M. Hyaluronic acid-coated PEI-PLGA nanoparticles mediated co-delivery of doxorubicin and miR-542-3p for triple negative breast cancer therapy. Nanomedicine. 2015 Dec 19;12:411-20. 56. Conde J, Oliva N, Atilano M, Song HS, Artzi N. Self-assembled RNA-triple-helix hydrogel scaffold for microRNA modulation in the tumour microenvironment. Nat Mater. 2015 Dec 7;15:353-63. 57. Qian X, Long L, Shi Z, Liu C, Qiu M, Sheng J, et al. Star-branched amphiphilic PLA-bPDMAEMA copolymers for co-delivery of miR-21 inhibitor and doxorubicin to treat glioma. Biomaterials. 2014 Feb;35(7):2322-35. 58. Liu C, Wen J, Meng Y, Zhang K, Zhu J, Ren Y, et al. Efficient delivery of therapeutic miRNA nanocapsules for tumor suppression. Adv Mater. 2015 Jan 14;27(2):292-7. 59. Mittal A, Chitkara D, Behrman SW, Mahato RI. Efficacy of gemcitabine conjugated and miRNA-205 complexed micelles for treatment of advanced pancreatic cancer. Biomaterials. 2014 Aug;35(25):7077-87. 60. Choi KY, Silvestre OF, Huang X, Min KH, Howard GP, Hida N, et al. Versatile RNA interference nanoplatform for systemic delivery of RNAs. ACS Nano. 2014 May 27;8(5):455970. 61. Liu Q, Li RT, Qian HQ, Wei J, Xie L, Shen J, et al. Targeted delivery of miR-200c/DOC to inhibit cancer stem cells and cancer cells by the gelatinases-stimuli nanoparticles. Biomaterials. 2013 Sep;34(29):7191-203. Inés Fernández Piñeiro 96 62. Deng X, Cao M, Zhang J, Hu K, Yin Z, Zhou Z, et al. Hyaluronic acid-chitosan nanoparticles for co-delivery of MiR-34a and doxorubicin in therapy against triple negative breast cancer. Biomaterials. 2014 May;35(14):4333-44. 63. Santos-Carballal B, Aaldering LJ, Ritzefeld M, Pereira S, Sewald N, Moerschbacher BM, et al. Physicochemical and biological characterization of chitosan-microRNA nanocomplexes for gene delivery to MCF-7 breast cancer cells. Sci Rep. 2015 Sep 1;5:13567. 64. Wang S, Cao M, Deng X, Xiao X, Yin Z, Hu Q, et al. Degradable hyaluronic acid/protamine sulfate interpolyelectrolyte complexes as miRNA-delivery nanocapsules for triple-negative breast cancer therapy. Adv Healthc Mater. 2015 Jan 28;4(2):281-90. 65. Hao Z, Fan W, Hao J, Wu X, Zeng GQ, Zhang LJ, et al. Efficient delivery of micro RNA to bone-metastatic prostate tumors by using aptamer-conjugated atelocollagen in vitro and in vivo. Drug Deliv. 2015 Jul 17:1-8. 66. Song H, Oh B, Choi M, Oh J, Lee M. Delivery of anti-microRNA-21 antisenseoligodeoxynucleotide using amphiphilic peptides for glioblastoma gene therapy. J Drug Target. 2015 May;23(4):360-70. 67. Wang X, Yu B, Ren W, Mo X, Zhou C, He H, et al. Enhanced hepatic delivery of siRNA and microRNA using oleic acid based lipid nanoparticle formulations. J Controlled Release. 2013 12/28;172(3):690-8. 68. Wu Y, Crawford M, Yu B, Mao Y, Nana-Sinkam SP, Lee LJ. MicroRNA delivery by cationic lipoplexes for lung cancer therapy. Mol Pharm. 2011 Aug 1;8(4):1381-9. 69. Wu Y, Crawford M, Mao Y, Lee RJ, Davis IC, Elton TS, et al. Therapeutic Delivery of MicroRNA-29b by Cationic Lipoplexes for Lung Cancer. Mol Ther Nucleic Acids. 2013 Apr 16;2:e84. 70. Lee HY, Mohammed KA, Kaye F, Sharma P, Moudgil BM, Clapp WL, et al. Targeted delivery of let-7a microRNA encapsulated ephrin-A1 conjugated liposomal nanoparticles inhibit tumor growth in lung cancer. Int J Nanomedicine. 2013;8:4481-94. Annex I. Nanocarriers for microRNA delivery in cancer treatment 97 71. Zhang Q, Ran R, Zhang L, Liu Y, Mei L, Zhang Z, et al. Simultaneous delivery of therapeutic antagomirs with paclitaxel for the management of metastatic tumors by a pHresponsive anti-microbial peptide-mediated liposomal delivery system. J Control Release. 2015 Jan 10;197:208-18. 72. Montis C, Sostegni S, Milani S, Baglioni P, Berti D. Biocompatible cationic lipids for the formulation of liposomal DNA vectors. Soft Matter. 2014 Jun 28;10(24):4287-97. 73. Passadouro M, Pedroso de Lima MC, Faneca H. MicroRNA modulation combined with sunitinib as a novel therapeutic strategy for pancreatic cancer. Int J Nanomedicine. 2014 Jul 3;9:3203-17. 74. Shi S, Han L, Gong T, Zhang Z, Sun X. Systemic delivery of microRNA-34a for cancer stem cell therapy. Angew Chem Int Ed Engl. 2013 Apr 2;52(14):3901-5. 75. Shi S, Han L, Deng L, Zhang Y, Shen H, Gong T, et al. Dual drugs (microRNA-34a and paclitaxel)-loaded functional solid lipid nanoparticles for synergistic cancer cell suppression. J Control Release. 2014 Nov 28;194:228-37. 76. Leung AK, Tam YY, Cullis PR. Lipid nanoparticles for short interfering RNA delivery. Adv Genet. 2014;88:71-110. 77. Daige CL, Wiggins JF, Priddy L, Nelligan-Davis T, Zhao J, Brown D. Systemic delivery of a miR34a mimic as a potential therapeutic for liver cancer. Mol Cancer Ther. 2014 Oct;13(10):2352-60. 78. Tolcher AW, Rodrigueza WV, Rasco DW, Patnaik A, Papadopoulos KP, Amaya A, et al. A phase 1 study of the BCL2-targeted deoxyribonucleic acid inhibitor (DNAi) PNT2258 in patients with advanced solid tumors. Cancer Chemother Pharmacol. 2014 Feb;73(2):363-71. 79. Costa PM, Cardoso AL, Custodia C, Cunha P, Pereira de Almeida L, Pedroso de Lima MC. MiRNA-21 silencing mediated by tumor-targeted nanoparticles combined with sunitinib: A new multimodal gene therapy approach for glioblastoma. J Control Release. 2015 Jun 10;207:31-9. Inés Fernández Piñeiro 98 80. Liu G, Yang D, Rupaimoole R, Pecot CV, Sun Y, Mangala LS, et al. Augmentation of response to chemotherapy by microRNA-506 through regulation of RAD51 in serous ovarian cancers. J Natl Cancer Inst. 2015 May 20;107(7):10.1093/jnci/djv108. Print 2015 Jul. 81. Dai X, Fan W, Wang Y, Huang L, Jiang Y, Shi L, et al. Combined Delivery of Let-7b MicroRNA and Paclitaxel via Biodegradable Nanoassemblies for the Treatment of KRAS Mutant Cancer. Mol Pharm. 2015 Dec 23;13:520-33. 82. Chen Y, Zhu X, Zhang X, Liu B, Huang L. Nanoparticles modified with tumor-targeting scFv deliver siRNA and miRNA for cancer therapy. Mol Ther. 2010 Sep;18(9):1650-6. 83. Wang H, Agarwal P, Zhao S, Yu J, Lu X, He X. A Near-Infrared Laser-Activated "Nanobomb" for Breaking the Barriers to MicroRNA Delivery. Adv Mater. 2016 Jan;28(2):347-55. 84. Bouchie A. First microRNA mimic enters clinic. Nat Biotechnol. 2013 Jul;31(7):577. 85. MacDiarmid JA, Brahmbhatt H. Minicells: versatile vectors for targeted drug or si/shRNA cancer therapy. Curr Opin Biotechnol. 2011 Dec;22(6):909-16. 86. Conde J, Artzi N. Are RNAi and miRNA therapeutics truly dead? Trends Biotechnol. 2015 Mar;33(3):141-4. Objectives Objectives 101 Objectives The main objective of this work has been the design of polysaccharide functionalised span nanoparticles as gene delivery systems and, more specifically, the design of nanoparticles based on sorbitan monooleate functionalised with natural polymers as non-viral microRNA vectors to treat colorectal liver metastasis. This main objective can be articulated into the following specific objectives: 1. Functionalisation of span nanoparticles with different natural polymers to provide specific targeting properties to liver sinusoidal endothelial cells. 2. Characterisation of the developed nanoparticles in terms of physicochemical properties and stability. 3. Evaluation of the in vitro ability of the developed nanoparticles to act as non-viral gene delivery systems able to associate, protect from degradation and effectively deliver into cells a model plasmid DNA. 4. Evaluation of the in vitro and in vivo toxicity of the developed nanoparticles and their in vivo biodistribution. 5. Evaluation of the potential of the most suitable nanosystem in the treatment of colorectal cancer metastasis to the liver in an animal model, using microRNA-20a as a therapeutic molecule. Chapter I Xanthan gum-functionalised span nanoparticles for gene targeting to endothelial cells Inés Fernández Piñeiro 110 2.7. Proliferation assay The influence of XG nanoparticles on cell proliferation was evaluated by ki-67 expression. Coverslips were placed in a 24 well-plate. HUVECs were seeded over the coverslips at a density of 4x104 cells per well and incubated overnight. Then, XG nanoparticles were incubated with the cells at a concentration of 384 μg/ml for 3 hours. Cells were fixed with paraformaldehyde 4% and blocked with 10% goat serum. The coverslips were incubated with mouse antihuman Ki-67 primary antibody at a concentration of 10 μg/ml overnight at 4°C. The coverslips were incubated with antimouse secondary antibody and DAPI for 1 hour at room temperature. Finally, the coverslips were mounted on glass slides, and ki-67 expression was detected by fluorescence microscopy. The images were analysed through image processing software ImageJ, and the percentage of ki-67 positive cells was calculated from the images. 2.8. Transfection of HUVECs For transfection experiments, HUVECs were seeded in a 24-well plate at a density of 4x104 cells per well and incubated overnight. The medium was replaced with Opti-MEM and plasmidloaded NPs were added at a dose of 1 μg of plasmid per well and incubated with the cells for 2 hours. Lipofectamine 2000 was used as a positive control according to the manufacturer’s instructions. After 2 hours, the medium was replaced for serum supplemented medium. The expression of GFP was detected 24 hours post-transfection using an inverted microscope (AxioObserver.Z1 Inverted Microscope, Zeiss, Germany). 2.9. Animals Balb/c mice (6to 8-week-old males) were obtained from Charles River Laboratories Spain S.A. (Barcelona, Spain). All procedures were approved by the Ethical Committee for Animal Experimentation (CEEA) of the University of the Basque Country (EHU/UPV) in accordance with institutional, national and international guidelines regarding the protection and care of animals used for scientific purposes (CEBA/237/2012/BADIOLA ETXABURU). Mice were kept in the animal facility of EHU/UPV and had access to standard chow and water ad libitum. 2.10. In vivo toxicity and biodistribution For in vivo toxicity and biodistribution studies, 200 μl of pEGFP-loaded nanoparticles in 5% glucose was administered intravenously to mice. For this purpose, 4 mice were used per Chapter I. Xanthan gum-functionalised span nanoparticles for gene targeting to endothelial cells 111 condition and a 5% solution of glucose was injected into control mice. After 24 hours, the animals were sacrificed, blood samples were collected by cardiac puncture and major organs were fixed in paraformaldehyde and embedded in paraffin. For biochemical analysis, serum was isolated from blood samples, and the hepatic function was analysed using the aspartate aminotransferase (AST) levels using an AST colorimetric assay. For histopathological examination, sections (4 μm) of liver, spleen, lung and kidney were cut and stained with Haematoxylin and Eosin (H&E). Representative sections were observed and examined using an optical microscope (Nikon Eclipse E-100). To study the biodistribution of the nanoparticles after systemic administration, GFP expression was evaluated in major organs. For this purpose, sections of paraffin-embedded liver, spleen, lung and kidney were cut into 4-μm slices and deparaffinized and rehydrated through xylenes and graded ethanol solutions to water. After deparaffinisation and rehydration, the sections were placed in a citrate buffer solution for antigen retrieval. The sections were blocked with 10% normal goat serum for 10 minutes at room temperature and incubated overnight at 4°C with mouse antihuman GFP primary antibody at 1:100 dilution. The slides were incubated with antimouse secondary antibody at 1:1000 dilution and DAPI for 1 hour at room temperature. Finally, the slides were sealed with coverslips, and GFP expression was detected by fluorescence microscopy. 2.11. Statistical analysis All experimental measurements were collected in triplicate. The values are expressed as the mean ± standard deviation (SD). Statistically significant differences between the treatments were evaluated by analysis of variance (ANOVA) and Fisher’s least significant difference (LSD) using GraphPad Prism, with alpha < 0.05 as the probability of error. 3. Results 3.1. Nanoparticles characterisation and DNase I protection ability As shown in Table 1, XG-functionalised NPs showed an average size of 146.8 nm and a negative zeta potential of -48 mV. Incorporating pEGFP into the system did not significantly change the NP size. However, the Z potential increased slightly to -40.6 mV. The nanometric size and population homogeneity of the system was confirmed by TEM images (Figure 2). Inés Fernández Piñeiro 112 These images show NP morphology characterised by a central core and an external shell. In addition, the efficacy of plasmid association by the NPs was corroborated by agarose gel electrophoresis (Figure 3A) because no free DNA band was observed. Moreover, XG nanoparticles successfully protected pEGFP from DNase I degradation, as shown in Figure 3B, as the pEGFP band appears intact after incubation of XG-pEGFP with DNase I and subsequent plasmid release with SDS. Table 1. Size and Z potential of blank and pEGFP-loaded nanoparticles. Formulation Size (nm) PdI ζ Potential (mV) SP-OA-XG 146.8 ± 7.4 0,150 ± 0.036 -48.0 ± 1.5 SP-OA-XG (200 μg/ml pEGFP) 153.5 ± 17.8 0.102 ± 0.036 -40.6 ± 4.6 Figure 1. XG nanoparticle schematic representation, Figure 2. TEM images of blank XG nanoparticles. Chapter I. Xanthan gum-functionalised span nanoparticles for gene targeting to endothelial cells 113 Figure 3. (A) Agarose gel electrophoresis showing efficacy of pEGFP association with XG nanoparticles. (B) Agarose gel electrophoresis after DNase protection assay: (1) free pEGFP, (2) free pEGFP + SDS, (3) free pEGFP + DNase I + SDS, (4) XG-pEGFP + SDS, and (5) XG-pEGFP + DNase I + SDS. 3.2. Stability of suspended and lyophilised nanoparticles XG nanoparticles showed remarkable stability in suspension, with minor size increments after 3 months of storage at 4°C, RT and 37°C (Figure 4A). However, a 3-fold size increase was observed for samples stored for 12 months at 37°C, from 145.7 nm to 441.8 nm. Z potential showed a slight tendency to increase in samples kept at RT and at 4°C and RT after 3 and 12 months of storage, respectively (Figure 4B). In contrast, the Z potential was stable for samples stored at 37°C. In addition to stability in suspension, XG nanoparticles were able to maintain plasmid association during 12 months at all temperatures assayed (Figure 4C). Figure 4. Stability of XG nanoparticles stored in suspension at 4°C, room temperature,(RT) and 37°C. After 3 and 12 months, size (A) and Z potential (B) were Inés Fernández Piñeiro 114 measured. *p<0.05 and **p<0.01 vs. samples at initial time (t=0). Plasmid association was corroborated by agarose gel electrophoresis after 12 months of storage (C). The system was subjected to lyophilisation, and its stability in the absence of water was studied. After a preliminary study, XG nanoparticles were lyophilised using trehalose as a lyoprotectant. Samples were resuspended in water, and NP physicochemical properties were compared to their properties before lyophilisation. As shown in Figure 5A, the system displayed resistance to lyophilisation, with a small size increase and Z potential decrease. Moreover, the nanoparticles successfully maintained plasmid binding (Figure 5B). Although some size increases were observed after the resuspension of samples kept at RT and 37°C, especially after 12 months, the sizes obtained were only 1.5 times bigger than the initial size. No statistically significant differences were seen for Z potential. Additionally, lyophilised NPs conserved plasmid association after 12 months of storage (Figure 6C). Figure 5. (A) Size and Zeta potential of fresh and lyophilised XG nanoparticles. *p<0.05 vs. size at t=0. ##p<0.01 vs. Z potential at t=0. (B) Agarose gel electrophoresis demonstrates XG nanoparticles maintain plasmid association after electrophoresis (lanes 2, 3 and 4). Free pEGFP was used as control (lane 1). Chapter I. Xanthan gum-functionalised span nanoparticles for gene targeting to endothelial cells 115 Figure 6. Stability of lyophilised XG nanoparticles stored at 4°C, room temperature (RT) and 37°C. After 3 and 12 months, NPs were resuspended and size (A) and Z potential (B) were measured. *p<0.05 and **p<0.01 vs. samples at initial time (t=0). Plasmid association was corroborated by agarose gel electrophoresis after 12 months of storage (C). 3.3. Cell viability and proliferation The effect of XG nanoparticles on HUVEC cell viability was studied by XTT assay, considering the metabolic activity of the cells. HUVECs were exposed to increasing concentrations of XG nanoparticles for 2 hours. Cell viability was determined and, as shown in Figure 7, remained unchanged until the highest nanoparticle concentration of 768 μg/ml. Although cell viability decreased to 30%, this concentration is much higher than the one required to effectively transfect cells in vitro. Figure 7. Cell viability of HUVECs measured by XTT after 2 hours of incubation with XG nanoparticles. Inés Fernández Piñeiro 116 XG nanoparticles cytotoxicity was further studied in HUVECs using the ki-67 assay to analyse cell proliferation. For this purpose, cells were exposed to 384 μg/ml of XG nanoparticles for 3 hours. This concentration was higher than that required to effectively transfect the cells in vitro. Then, a ki-67 immunocytochemistry assay was performed on control and XG-treated cells and images were obtained by fluorescence microscopy. The images obtained were processed through ImageJ, and ki-67 positive XG-treated cells were calculated and compared to the control cells (Figure 8B). As shown in Figure 8, there was no significant difference in ki-67 expression between control cells and XG-treated cells, indicating that the nanoparticles do not induce growth arrest. Figure 8. Images (A) and graph (B) showing ki-67 expression in control and XGtreated cells. 3.4. In vitro transfection The ability of XG nanoparticles to transfect pEGFP was evaluated in HUVEC. Cells were incubated with pEGFP-XG nanoparticles for 2 hours, and GFP expression was evaluated by fluorescence microscopy after 24 hours. As shown in Figure 9, the system efficiently transfected the primary cell line at a dose of 1 μg of plasmid, corresponding to a nanoparticle concentration of 223.6 μg/ml. Chapter I. Xanthan gum-functionalised span nanoparticles for gene targeting to endothelial cells 117 Figure 9. HUVEC were efficiently transfected with pEGFP-loaded NPs using doses of 1 μg of plasmid per well. Lipofectamine was used as positive control, with doses of 0.5 μg of plasmid per well. 3.5. In vivo toxicity To assess the safety profile of systemically administered XG functionalised nanoparticles associated with pEGFP, the aspartate aminotransferase (AST) serum levels were analysed. For this purpose, 200 μl of pEGFP-XG nanoparticles (4.03 mg/ml) was intravenously injected in a 5% solution of glucose, and this same solution without nanoparticles was injected into the control mice. As shown in Figure 10A, no significant differences were observed in AST levels of control and XG-treated mice, indicating that there was no significant liver damage caused by the nanoparticles. Moreover, the toxicity of the nanoparticles to the major mouse organs including kidney, liver, lung and spleen, was evaluated by haematoxylin and eosin (H&E) staining (Figure 10B). Some inflammatory cell infiltration was observed in the liver and surrounding blood vessels, indicating that the XG nanoparticles may induce a mild inflammatory response in the liver. No significant pathology was observed in the other organs compared to controls. Inés Fernández Piñeiro 118 Figure 10. In vivo XG nanoparticle toxicity was evaluated by aspartate aminotransferase (AST) serum levels (A) and histopathological examination of major organs sections stained with haematoxylin and eosin (B), after intravenous administration of the nanoparticles to mice. The black arrow shows a small lymphocyte aggregate in the liver of an animal treated with XG nanoparticles. 3.6. Biodistribution The biodistribution of XG nanoparticles was evaluated after an intravenous tail injection of pEGFP-loaded nanoparticles to mice. Mice were sacrificed 24 hours after administration, and GFP expression was evaluated in major organs. First, endothelium targeting specificity was evaluated in the liver. As shown in Figure 11, after immunostaining of Kupffer cells, no colocalization was observed between Kupffer cells and GFP staining. This indicates that the nanoparticles specifically target endothelial cells and escape the phagocytic functions of Kupffer cells as part of the reticuloendothelial system (RES). The biodistribution of XG nanoparticles in other organs was evaluated. As shown in Figure 12, GFP expression was found in the endothelial cells of blood vessels in the kidney, liver and lung and in LSECs. Chapter I. Xanthan gum-functionalised span nanoparticles for gene targeting to endothelial cells 119 Figure 11. Liver sections showing GFP expression in green and Kupffer cells labelled with F4/80 antibody in red. Figure 12. Immunohistochemistry of major organs sections showing GFP expression (green) in endothelial cells 24 hours after XG-pEGFP intravenous administration to mice. Mice injected with 5% glucose solutions were used as controls (red blood cells autofluorescence in green). Cell nuclei are stained in blue with DAPI. Inés Fernández Piñeiro 126 10. Rajendran P, Rengarajan T, Thangavel J, Nishigaki Y, Sakthisekaran D, Sethi G, et al. The Vascular Endothelium and Human Diseases. International Journal of Biological Sciences. 2013 10/07;9(10):1057-69. 11. Aird WC. Phenotypic heterogeneity of the endothelium: I. Structure, function, and mechanisms. Circ Res. 2007 Feb 2;100(2):158-73. 12. Aird WC. Phenotypic heterogeneity of the endothelium: II. Representative vascular beds. Circ Res. 2007 Feb 2;100(2):174-90. 13. Aird WC. Endothelial Cell Heterogeneity. Cold Spring Harbor Perspectives in Medicine. 2012 01;2(1):a006429. 14. Hood JD, Bednarski M, Frausto R, Guccione S, Reisfeld RA, Xiang R, et al. Tumor regression by targeted gene delivery to the neovasculature. Science. 2002 Jun 28;296(5577):2404-7. 15. Ma Z, Zhang J, Alber S, Dileo J, Negishi Y, Stolz D, et al. Lipid-mediated delivery of oligonucleotide to pulmonary endothelium. Am J Respir Cell Mol Biol. 2002 Aug;27(2):1519. 16. Cicha I. Strategies to enhance nanoparticle-endothelial interactions under flow. Journal of Cellular Biotechnology. 2016 12/05/2016;1(2):191-208. 17. Lin A, Sabnis A, Kona S, Nattama S, Patel H, Dong JF, et al. Shear-regulated uptake of nanoparticles by endothelial cells and development of endothelial-targeting nanoparticles. J Biomed Mater Res A. 2010 Jun 1;93(3):833-42. 18. Gupta A, Gupta RK, Gupta GS. Targeting cells for drug and gene delivery: Emerging applicationsof mannans and mannan binding lectins. Journal of Scientific and Industrial Research. 2009 03/03/2009;68(6):465-83. 19. Scholzen T, Gerdes J. The Ki-67 protein: from the known and the unknown. J Cell Physiol. 2000 Mar;182(3):311-22. Chapter I. Xanthan gum-functionalised span nanoparticles for gene targeting to endothelial cells 127 20. Schuch RA, Oliveira TL, Collares TF, Monte LG, Inda GR, Dellagostin OA, et al. The Use of Xanthan Gum as Vaccine Adjuvant: An Evaluation of Immunostimulatory Potential in BALB/c Mice and Cytotoxicity In Vitro. Biomed Res Int. 2017;2017:3925024. 21. Ishizaka S, Sugawara I, Hasuma T, Morisawa S, Moller G. Immune responses to xanthan gum. I. The characteristics of lymphocyte activation by xanthan gum. Eur J Immunol. 1983 Mar;13(3):225-31. 22. Yang C, Hao F, He J, Lu T, Klein RL, Zhao LR, et al. Sequential Adeno-Associated Viral Vector Serotype 9-Green Fluorescent Protein Gene Transfer Causes Massive Inflammation and Intense Immune Response in Rat Striatum. Hum Gene Ther. 2016 Jul;27(7):528-43. 23. Baens M, Noels H, Broeckx V, Hagens S, Fevery S, Billiau AD, et al. The dark side of EGFP: defective polyubiquitination. PLoS One. 2006 Dec 20;1:e54. 24. Zhang F, Hackett NR, Lam G, Cheng J, Pergolizzi R, Luo L, et al. Green fluorescent protein selectively induces HSP70-mediated up-regulation of COX-2 expression in endothelial cells. Blood. 2003 Sep 15;102(6):2115-21. 25. Liu HS, Jan MS, Chou CK, Chen PH, Ke NJ. Is green fluorescent protein toxic to the living cells? Biochem Biophys Res Commun. 1999 Jul 14;260(3):712-7. 26. Goto H, Yang B, Petersen D, Pepper KA, Alfaro PA, Kohn DB, et al. Transduction of green fluorescent protein increased oxidative stress and enhanced sensitivity to cytotoxic drugs in neuroblastoma cell lines. Mol Cancer Ther. 2003 Sep;2(9):911-7. 27. Sadauskas E, Wallin H, Stoltenberg M, Vogel U, Doering P, Larsen A, et al. Kupffer cells are central in the removal of nanoparticles from the organism. Part Fibre Toxicol. 2007 Oct 19;4:10. 28. Rinkenauer AC, Press AT, Raasch M, Pietsch C, Schweizer S, Schworer S, et al. Comparison of the uptake of methacrylate-based nanoparticles in static and dynamic in vitro systems as well as in vivo. J Control Release. 2015 Oct 28;216:158-68. Inés Fernández Piñeiro 128 29. Frohlich E. The role of surface charge in cellular uptake and cytotoxicity of medical nanoparticles. Int J Nanomedicine. 2012;7:5577-91. 30. Poelstra K, Prakash J, Beljaars L. Drug targeting to the diseased liver. J Control Release. 2012 Jul 20;161(2):188-97. 31. Melgert BN, Olinga P, Jack VK, Molema G, Meijer DK, Poelstra K. Dexamethasone coupled to albumin is selectively taken up by rat nonparenchymal liver cells and attenuates LPS-induced activation of hepatic cells. J Hepatol. 2000 Apr;32(4):603-11. 32. Kamps JA, Morselt HW, Swart PJ, Meijer DK, Scherphof GL. Massive targeting of liposomes, surface-modified with anionized albumins, to hepatic endothelial cells. Proc Natl Acad Sci U S A. 1997 Oct 14;94(21):11681-5. 33. Lee R, Channon KM, Antoniades C. Therapeutic strategies targeting endothelial function in humans: clinical implications. Curr Vasc Pharmacol. 2012 Jan;10(1):77-93. 34. Sun X, Rossin R, Turner JL, Becker ML, Joralemon MJ, Welch MJ, et al. An assessment of the effects of shell cross-linked nanoparticle size, core composition, and surface PEGylation on in vivo biodistribution. Biomacromolecules. 2005 Sep-Oct;6(5):2541-54. 35. Howard M, Zern BJ, Anselmo AC, Shuvaev VV, Mitragotri S, Muzykantov V. Vascular targeting of nanocarriers: perplexing aspects of the seemingly straightforward paradigm. ACS Nano. 2014 May 27;8(5):4100-32. 36. Voigt J, Christensen J, Shastri VP. Differential uptake of nanoparticles by endothelial cells through polyelectrolytes with affinity for caveolae. Proc Natl Acad Sci U S A. 2014 Feb 25;111(8):2942-7. 37. Rasente RY, Imperiale JC, Lazaro-Martinez JM, Gualco L, Oberkersch R, Sosnik A, et al. Dermatan sulfate/chitosan polyelectrolyte complex with potential application in the treatment and diagnosis of vascular disease. Carbohydr Polym. 2016 Jun 25;144:362-70. Chapter I. Xanthan gum-functionalised span nanoparticles for gene targeting to endothelial cells 129 38. Lamichhane SP, Arya N, Ojha N, Kohler E, Shastri VP. Glycosaminoglycan-functionalized poly-lactide-co-glycolide nanoparticles: synthesis, characterization, cytocompatibility, and cellular uptake. Int J Nanomedicine. 2015 Jan 19;10:775-89. 39. Rouzet F, Bachelet-Violette L, Alsac JM, Suzuki M, Meulemans A, Louedec L, et al. Radiolabeled fucoidan as a p-selectin targeting agent for in vivo imaging of platelet-rich thrombus and endothelial activation. J Nucl Med. 2011 Sep;52(9):1433-40. 40. Verma A, Stellacci F. Effect of surface properties on nanoparticle-cell interactions. Small. 2010 Jan;6(1):12-21. Chapter II Development and characterisation of chondroitin sulfateand hyaluronic acid-incorporated sorbitan ester nanoparticles as gene delivery systems Chapter II. Development and characterisation of chondroitin sulfateand hyaluronic acid... 133 Abstract Glycosaminoglycans (GAGs) are natural polymers that are broadly used in gene delivery systems to increase stability as well as decrease toxicity and nonspecific interactions, thereby increasing transfection efficiency. In this work, we propose sorbitan ester-based lipid nanoparticles functionalised with the GAGs chondroitin sulfate (CS) and hyaluronic acid (HA) as gene delivery systems. For this purpose, we describe the design and evaluation of these nanosystems loaded with plasmid DNA, including an evaluation of their physicochemical characteristics, stability properties, ability to protect and efficiently transfect cells with Enhanced Green Fluorescent Protein plasmid (pEGFP) in vitro, and biocompatibility both in vitro and in vivo. We confirm that molecules with high biological value and targeting potential, such as HA and CS, can be successfully incorporated into our recently developed sorbitan esterbased nanoparticles and that this incorporation leads to effective stabilisation of both nanosystems as well as protects plasmid DNA. We demonstrated that the aforementioned incorporation of HA and CS enables long-term stability of the nanosystems in both liquid and lyophilised states, which is a remarkable property that can aid in their transfer to industry. The ability of these functionalised nanosystems to transfect the A549 cell line without compromising cell viability was also shown, as well as their innocuous safety profile in vivo. Thus, we provide valuable evidence of the suitable properties and potential of these hybrid nanoparticles as gene delivery systems. Inés Fernández Piñeiro 134 1. Introduction Gene therapy consists of the delivery of nucleic acids to specific cells to replace or silence malfunctioning genes with the aim of curing or favourably altering pathological progression (1). Deregulation of gene expression is associated with the origin of multiple diseases. Thus, recent advances in identification of the molecular basis of numerous inherited genetic disorders have made them susceptible to gene therapy treatment. Although gene therapy can be applied to treatment and prevention of multiple pathologies, such as infectious diseases, Parkinson’s disease or cystic fibrosis, most efforts and clinical trials have been focused on cancer therapy (2). Cancer comprises a group of diseases with multifactorial origin. Despite considerable advances in the field of cancer therapy, chemical-based therapeutics, radiotherapy and surgery have been insufficient in addressing the diverse cancer aetiologies, leading to a high rate of treatment failure. Gene-based therapies could solve this problem by targeting specific pathways that are altered in cancer while avoiding most of the side effects associated with conventional therapies (3, 4). However, gene delivery to cancer cells remains a major challenge. Nucleic acid delivery must overcome numerous barriers and obstacles before its therapeutic effect can be exerted. Nucleic acids are negatively charged macromolecules with a hydrophilic nature, characteristics that explain their poor ability to penetrate biological barriers and their susceptibility to degradation by enzymes in the body. These properties make necessary the development and inclusion of nucleic acids into gene delivery systems. Non-viral delivery systems have been extensively studied due to their safety, low cost and scalability in comparison with viral vectors (4-6). The most common traditional non-viral systems can be classified as lipidic, polymeric or polymer-lipid hybrid systems (3). Among them, cationic lipids have been broadly used for gene delivery due to their high efficacy in transfecting a variety of cell types with different types of nucleic acids and their reproducibility and ease of use (7). However, their positive charge is related to high cytotoxicity, nonspecific uptake and short blood circulation time due to interaction with negatively charged serum proteins and eventual phagocytosis by the reticuloendothelial system (RES). Diverse strategies have been applied in an attempt to shield the positive surface charge and avoid these side effects. Pegylation is known to reduce nonspecific interactions of nanosystems; however, this strategy might decrease cellular uptake and Chapter II. Development and characterisation of chondroitin sulfateand hyaluronic acid... 135 gene release (8-10). Alternatively, surface modification of nanocarriers with anionic macromolecules, such as glycosaminoglycans (GAGs), may lengthen their circulation time by reducing complement activation and thus recognition by the RES (11, 12). Therefore, it has been demonstrated that incorporation of GAGs into gene delivery systems increases the stability of the carriers and decreases toxicity and nonspecific interactions, thereby increasing their transfection efficiency in vivo (13-18). GAGs have also been shown to influence intracellular routing of nanosystems by improving endosomal escape and localisation at the nuclear periphery (19). In our study, the natural anionic polymers chondroitin sulfate (CS) and hyaluronic acid (HA) were employed to stabilise and functionalise the system by taking advantage of their suitable properties for gene delivery. They are the main components of the extracellular matrix and are considered biocompatible and biodegradable polymers, and thus, they are widely used in the biomedical field (20-23). CS and HA are mucopolysaccharides naturally present in the body, and they are composed of glucuronic acid and Nacetylglucosamine repeats connected via a β-1-4 linkage. A sulfate group in at least one of the CS side groups is the primary structural difference between them. HA is present in connective tissue, vitreous humour and synovial joint fluid, while CS has a role in wound healing and chondrogenesis (24). Despite the large number of lipid-based nanosystems described in the literature, only a few have reached the market (25). One of the limitations of lipid nanosystems is their low stability in aqueous suspensions, which compromises their transfection efficiency (26). Thus, more stable systems are necessary to scale production from basic research to the pharmaceutical industry. In addition, advances in characterisation and manufacturing of the vectors are needed for successful therapy. In this work, we propose sorbitan ester-based lipid nanoparticles functionalised with the natural polymers CS or HA as gene delivery systems. For this purpose, we describe the design and evaluation of these nanosystems loaded with plasmid DNA, including their physicochemical characterisation, their stability properties upon storage at different temperatures in both a suspension and as a lyophilised product, their ability to protect the delicate bioactive molecule associated with them, their transfection properties and their biocompatibility both in vitro and in vivo. Inés Fernández Piñeiro 142 Figure 5. Agarose gel electrophoresis of free pEGFP and pEGFP-loaded HA and CS nanoparticles stored at 4°C, room temperature (RT) and 37°C. After a preliminary screening of different sugars and their concentrations, 10% glucose and 10% sucrose were applied for lyophilisation of CS and HA nanoparticles, respectively. As shown in Figure 6A, the size of pEGFP-loaded NPs was slightly but significantly (p<0.01) increased from approximately 120 nm to 150 nm for both formulations after lyophilisation, while no significant difference in zeta potential was observed. Similar results were obtained for blank NPs, with small increases in size (120 nm to 130 for CS and 135 nm to 160 nm for HA) and zeta potential (-40 mV to -37 mV for CS and -32 mV for HA) for both formulations. Agarose gel electrophoresis was performed to evaluate the association with the plasmid after the freeze-drying process. As indicated by the free DNA bands observed in Figure 6B, lyophilisation decreased the ability of CS NPs to bind pEGFP, whereas the DNA binding property of the HA NPs was not modified. Figure 6A. Size and zeta potential of pEGFP-loaded NPs before and after lyophilisation. Error bars represents SD (n=3). **p<0.01 vs. pre-lyophilisation samples. Fig. 6B. Agarose gel electrophoresis of free pEGFP (1) and pEGFP-loaded Chapter II. Development and characterisation of chondroitin sulfateand hyaluronic acid... 143 CS (2, 3, 4) and HA (5, 6, 7) NPs after lyophilisation. Some free DNA bands are observed for CS nanoparticles. NPs were lyophilised and stored at different temperatures to determine the stability of the NPs in the absence of water. After 12 months of storage, the largest difference in the size of the lyophilised NPs was observed when stored at 37°C, increasing from approximately 140 nm to 180 nm for CS and from 150 nm to 180 nm for HA NPs (p<0.01) (Figure 7). The same pattern was observed for zeta potential, which decreased from -34 mV to -49 mV for both formulations. Figure 7. Size and zeta potential of resuspended CS and HA NPs after lyophilisation and storage for 3 and 12 months at 4°C, room temperature (RT) and 37°C. Error bars represent SD (n=3). *p<0.05 and **p<0.01 vs. samples at the initial time (t=0). 2.4. DNase protection assay To demonstrate that the NPs protect DNA from endonucleases, DNase I was incubated with free DNA and NP-associated DNA, and DNA was subsequently released from the NPs using SDS. Figure 8 demonstrates the presence of DNA bands corresponding with DNA associated with CS and HA NPs (lanes 3 and 4, respectively) after incubation with DNase I, evidencing the absence of degradation. In contrast, the absence of any band in lane 2 demonstrates the ability of DNase I to degrade free DNA. SDS, which is a surfactant used for Inés Fernández Piñeiro 144 DNA-NP disassembly, was responsible for the smeared DNA bands because it interferes with the electrophoretic mobility of DNA. Figure 8. Agarose gel electrophoresis of free pEGFP and pEGFP released from nanoparticles after 1 hour of incubation with DNase I and subsequent pEGFP-NP disassembly mediated by SDS: Free pEGFP (2) and pEGFP-loaded CS (3) and HA (4) nanoparticles in the presence of DNase I. The controls include free pEGFP (1) and pEGFP-loaded CS (5) and HA (6) nanoparticles in the absence of DNase I. 2.5. Cell viability Figure 9 shows A549 cell viability after incubation with CS and HA NPs at increasing concentrations determined with an XTT assay. Accordingly, the XTT assay results do not show a significant decrease in cell viability until the highest nanoparticle concentration tested (768 μg/ml; 58.63% for CS and 83.63% for HA). There is almost no difference in the effect of both formulations on cell viability, with a slightly higher decrease in cell viability observed with CS at 768 μg/ml being the only significant difference. Figure 9. A549 cell viability determined with an XTT assay after incubation with increasing concentrations of CS and HA nanoparticles. Error bars represent SD (n=3). *p<0.05. Chapter II. Development and characterisation of chondroitin sulfateand hyaluronic acid... 145 2.6. In vitro transfection A549 cells were transfected with different doses of pEGFP-loaded NPs, and GFP expression was observed after 24 hours using fluorescence microscopy. As a positive control, Lipofectamine was used with 1 μg of pEGFP. As shown in Figure 10, efficient transfection can be achieved with both formulations. Figure 10. A549 cells expressing GFP 24 hours after transfection with pEGFP-loaded CS and HA nanoparticles at two different plasmid doses and NP concentrations. Lipofectamine was used as the positive control. 2.7. In vivo toxicity The in vivo toxicity of pEGFP-loaded NPs was assessed after iv administration to mice by monitoring hepatic function through analysis of aspartate aminotransferase (AST) levels. As shown in Figure 11, comparable AST serum levels were found for CS NP-, HA NPand nontreated animals, suggesting the absence of undesirable effects attributed to the developed nanosystems. To further corroborate the safety of the systems, their cytotoxic effects in the major mouse organs, including the kidney, liver, lung and spleen, were evaluated by performing histopathological studies. No obvious damage was observed in the samples analysed, indicating that the NPs are not toxic to major organs. These results indicate that CS and HA NPs are well tolerated delivery systems with non-immunogenic and biocompatible properties. Inés Fernández Piñeiro 146 Figure 11A. Aspartate aminotransferase (AST) serum levels in mice 24 hours after iv injection with pEGFP-loaded CS and HA nanoparticles. Fig. 11B. Histopathological studies using H&E staining of kidney, liver, lung and spleen sections from mice systemically treated with pEGFP-loaded CS and HA nanoparticles. 3. Discussion This study was designed to explore the potential of the natural polymers HA and CS to optimise the properties of our previously developed prototype of sorbitan ester-based nanoparticles. These polysaccharides possess properties that make them suitable for gene delivery systems, such as biodegradability, stability, low toxicity and low immunogenicity, and they can be easily modified. In addition, it has been described that CS and HA can enhance Chapter II. Development and characterisation of chondroitin sulfateand hyaluronic acid... 147 gene transfection efficiency and improve the stability of NP systems (23, 27). However, incorporation of such components is not an easy task because they are hydrophilic macromolecules (carbohydrates with repeating units of saccharides joined by glycosidic linkages) that possess properties that are very different from those of the selected sorbitan ester, sorbitan monooleate. Consequently, our procedure to develop these polymer-lipid hybrid systems required a specific ´bridge´ among the different moieties and therefore a strategy to develop such a bridge. In this study, we resorted to incorporation of a hydrophobic cationic moiety able to satisfy three conditions: i) easily incorporated with the lipophilic sorbitan monooleate component; ii) developed electrostatic interactions with HA and CS polyanions to incorporate them into the final nanosystem; and iii) effectively associated with DNA as the bioactive molecule through electrostatic interactions. After a judicious selection of components, the selected material to build such a bridge was oleylamine (OA). In our previous work, we demonstrated the solid structure of the sorbitan ester-based nanoparticles, in which there is a gradient of flexibility from a lipophilic rigid core to a hydrophilic flexible surface (28). In the case of the above-described self-assembled nanosystems associated with DNA we can assume that, once developed, the anionic polymers HA and CS tend to localise in the external and hydrophilic region of the nanoparticles, thus decorating the nanosystems (Figure 1). This would explain the change in the nanoparticle surface charge from positive (29) to negative after the polyanion incorporation. In addition, the size of the nanoparticles decreased with CS and HA incorporation (from 200 nm to 130-150 nm), which could be explained by an increase in the polymer packing due to electrostatic interactions and crosslinking of the different components. In addition, we observed that incorporation of pDNA into the systems did not interfere with the size and surface charge (Table 1). Taking into account that such DNA compaction could also be helpful for providing desired properties as gene delivery carriers, such as DNA protection against enzymatic degradation, we decided to evaluate this stabilising effect. As we have demonstrated, the developed nanosystems offer efficient protection of pDNA from DNAse I degradation (Figure 8). The above-described effect of the incorporation of HA and CS could also be of interest in terms of stability. In this respect, it should be considered that delivery systems based on lipid components are characterised by their low stability in aqueous suspension, a property that limits their use and clinical success. However, we provide evidence that the developed nanosystems have considerable time and temperature stability in aqueous suspension (Figure 4). This Inés Fernández Piñeiro 148 behaviour could be explained by structural differences compared with other lipid-based nanosystems, such as vesicular systems (i.e., liposomes), which are known for their stability limitations. We only observed increases in size and zeta potential after storage at 37°C, a result that could be explained by hydrolysis of the sorbitan monooleate molecules, resulting in free fatty acids that can partially neutralise the positive charge of OA, leading to a decrease in the superficial charge and to an increase in the size. This effect could also be a consequence of Ostwald ripening, with growth of the largest particles at the expense of the smallest ones (30). Concerning stability against the lyophilisation process, we demonstrated the ability of the developed nanosystems to resist the lyophilisation process and maintain their physical properties (Figure 6), an ability of great value for shipping and long-term storage. Sugars are known to maintain lipid nanosystem stability during freeze-drying, and two different theories to explain this behaviour have been proposed. One of them is the water replacement theory, in which the sugars replace water between the polar head groups of the lipids. The second theory is known as the vitrification model and it assumes that the sugars form a glass matrix during freezing, with high viscosity and low mobility, which prevents lipid systems from aggregation and protects them from damage by ice crystals (31). Taking this into consideration, we hypothesise that the differences observed in DNA leakage among CS and HA NPs after lyophilisation could be explained by the different sugars used as lyoprotectants (glucose versus sucrose) (Figure 6A). In addition, it is possible that differences in the NP packing could interfere with the sugar disposition among the NP components. From previous work (29) and previous studies, we know that non-functionalised sorbitan monooleate-oleylamine nanoparticles do not interfere with cell viability, even when higher concentrations were tested on the same cell line. Different studies have shown that incorporation of HA and CS into delivery systems did not increase toxicity in vitro (32, 33) or even decreased the toxicity of the systems (13, 14, 16, 34). Another study showed an increase in the cytotoxicity of gemcitabine-loaded liposomes when HA was incorporated into the system and explained that this effect was due to an increase in the cellular uptake of the liposomes mediated by the CD44 receptor (35). Considering this, we hypothesise that the higher cytotoxicity observed with the developed nanosystems after incorporation of HA and CS (Figure 9) in comparison with the non-functionalised NPs is a simple consequence of the higher uptake rate of the former. In any case, high transfection efficiencies were achieved at low nanosystem doses without compromising cell viability (Figure 10), which allowed us to Chapter II. Development and characterisation of chondroitin sulfateand hyaluronic acid... 149 conclude that the developed nanosystems are characterised by an appropriate efficacy-toxicity balance in vitro. On the basis of these data, we decided to test the in vivo toxicity of the developed nanosystems (Figure 11). No signs of toxicity or inflammation were observed after systemic administration of the developed nanosystems to mice, therefore also demonstrating their innocuity in vivo, which further supports the high clinical potential of the developed nanosystems. 4. Conclusions Herein, we confirm that molecules with high biological value and targeting potential, such as HA and CS, can be successfully incorporated into our recently developed sorbitan esterbased nanoparticles and that such incorporation leads to effective stabilisation of both nanosystems and protection of a delicate bioactive molecule associated with them, such as DNA. Concretely, we have demonstrated that the aforementioned incorporation of HA and CS supports long-term stability of the nanosystems in both liquid and lyophilised states, which is a remarkable property that may aid their transfer to industry. The ability of these functionalised nanosystems to transfect the A549 cell line without compromising cell viability was also shown, as well as their innocuous safety profile in vivo. Thus, we provide valuable evidence of the suitable properties and potential of these hybrid nanoparticles as gene delivery systems. Inés Fernández Piñeiro 150 References 1. Naldini L. Gene therapy returns to centre stage. Nature. 2015 Oct 15;526(7573):351-60. 2. Collins M, Thrasher A. Gene therapy: progress and predictions. Proceedings of the Royal Society B: Biological Sciences. 2015 04/21;282(1821):20143003. 3. Pahle J, Walther W. Vectors and strategies for nonviral cancer gene therapy. Expert Opin Biol Ther. 2016;16(4):443-61. 4. Zaimy MA, Saffarzadeh N, Mohammadi A, Pourghadamyari H, Izadi P, Sarli A, et al. New methods in the diagnosis of cancer and gene therapy of cancer based on nanoparticles. Cancer Gene Ther. 2017 Jun;24(6):233-43. 5. Swami R, Singh I, Khan W, Ramakrishna S. Diseases originate and terminate by genes: unraveling nonviral gene delivery. Drug Deliv Transl Res. 2013 Dec;3(6):593-610. 6. Yin H, Kanasty RL, Eltoukhy AA, Vegas AJ, Dorkin JR, Anderson DG. Non-viral vectors for gene-based therapy. Nat Rev Genet. 2014 Aug;15(8):541-55. 7. Ding W, Wang F, Zhang J, Guo Y, Ju S, Wang H. A novel local anti-colorectal cancer drug delivery system: negative lipidoid nanoparticles with a passive target via a size-dependent pattern. Nanotechnology. 2013 Sep 20;24(37):375101,4484/24/37/375101. Epub 2013 Aug 21. 8. Schlenk F, Grund S, Fischer D. Recent developments and perspectives on gene therapy using synthetic vectors. Ther Deliv. 2013 Jan;4(1):95-113. 9. Li L, Song L, Yang X, Li X, Wu Y, He T, et al. Multifunctional "core-shell" nanoparticlesbased gene delivery for treatment of aggressive melanoma. Biomaterials. 2016 Dec;111:12437. 10. Harris TJ, Green JJ, Fung PW, Langer R, Anderson DG, Bhatia SN. Tissue-specific gene delivery via nanoparticle coating. Biomaterials. 2010 Feb;31(5):998-1006. 11. Almalik A, Day PJ, Tirelli N. HA-coated chitosan nanoparticles for CD44-mediated nucleic acid delivery. Macromol Biosci. 2013 Dec;13(12):1671-80. Chapter II. Development and characterisation of chondroitin sulfateand hyaluronic acid... 151 12. Raemdonck K, Martens TF, Braeckmans K, Demeester J, De Smedt SC. Polysaccharidebased nucleic acid nanoformulations. Adv Drug Deliv Rev. 2013 Aug;65(9):1123-47. 13. Yin H, Zhao F, Zhang D, Li J. Hyaluronic acid conjugated beta-cyclodextrinoligoethylenimine star polymer for CD44-targeted gene delivery. Int J Pharm. 2015 Apr 10;483(1-2):169-79. 14. Urbiola K, Sanmartin C, Blanco-Fernandez L, Tros de Ilarduya C. Efficient targeted gene delivery by a novel PAMAM/DNA dendriplex coated with hyaluronic acid. Nanomedicine (Lond). 2014 Dec;9(18):2787-801. 15. Ran R, Liu Y, Gao H, Kuang Q, Zhang Q, Tang J, et al. Enhanced gene delivery efficiency of cationic liposomes coated with PEGylated hyaluronic acid for anti P-glycoprotein siRNA: a potential candidate for overcoming multi-drug resistance. Int J Pharm. 2014 Dec 30;477(12):590-600. 16. Imamura M, Kodama Y, Higuchi N, Kanda K, Nakagawa H, Muro T, et al. Ternary complex of plasmid DNA electrostatically assembled with polyamidoamine dendrimer and chondroitin sulfate for effective and secure gene delivery. Biol Pharm Bull. 2014;37(4):552-9. 17. He Y, Cheng G, Xie L, Nie Y, He B, Gu Z. Polyethyleneimine/DNA polyplexes with reduction-sensitive hyaluronic acid derivatives shielding for targeted gene delivery. Biomaterials. 2013 January 2013;34(4):1235-45. 18. Kim EJ, Cho HJ, Park D, Kim JY, Kim YB, Park TG, et al. Antifibrotic effect of MMP13encoding plasmid DNA delivered using polyethylenimine shielded with hyaluronic acid. Mol Ther. 2011 Feb;19(2):355-61. 19. Naik RJ, Sharma R, Nisakar D, Purohit G, Ganguli M. Exogenous chondroitin sulfate glycosaminoglycan associate with arginine-rich peptide-DNA complexes to alter their intracellular processing and gene delivery efficiency. Biochim Biophys Acta. 2015 Apr;1848(4):1053-64.