Functionalization of dendrimers for improved gene delivery to mesenchymal stem cell

Disease, injury, and age problems compromise human quality of life and continuously motivate the search for new and more efficacious therapeutic approaches. The field of Tissue Regeneration and Engineering has greatly evolved over the last years, mainly due to the combination of the important advances verified in Biomaterials Science and Engineering with those of Cell and Molecular Biology. In particular, a new and promising area arose – Nanomedicine – that takes advantage of the extremely small size and especial chemical and physical properties of Nanomaterials, offering powerful tools for health improvement. Research on Stem Cells, the self-renewing progenitors of body tissues, is also challenging to the medical and scientific communities, being expectable the appearance of new and exciting stem cell-based therapies in the next years. The control of cell behavior (namely, of cell proliferation and differentiation) is of key importance in devising strategies for Tissue Regeneration and Engineering. Cytokines, growth factors, transcription factors and other signaling molecules, most of them proteins, have been identified and found to regulate and support tissue development and regeneration. However, the application of these molecules in long-term regenerative processes requires their continuous presence at high concentrations as they usually present short half-lives at physiological conditions and may be rapidly cleared from the body. Alternatively, genes encoding such proteins can be introduced inside cells and be expressed using cell’s machinery, allowing an extended and more sustained production of the protein of interest (gene therapy). Genetic engineering of stem cells is particularly attractive because of their self-renewal capability and differentiation potential. For Tissue Regeneration and Engineering purposes, the patient’s own stem cells can be genetically engineered in vitro and, after, introduced in the body (with or without a scaffold) where they will not only modulate the behavior of native cells (stem cell-mediated gene therapy), but also directly participate in tissue repair. Cells can be genetically engineered using viral and non-viral systems. Viruses, as a result of millions of years of evolution, are very effective for the delivery of genes in several types of cells, including cells from primary sources. However, the risks associated with their use (like infection and immunogenic reactions) are driving the search for non-viral systems that will efficiently deliver genetic material into cells. Among them, chemical methods that are promising and being investigated use cationic molecules as carriers for DNA. In this case, gene delivery and gene expression level remain relatively low when primary cells are used. The main goal of this thesis was to develop and assess the in vitro potential of polyamidoamine (PAMAM) dendrimers based carriers to deliver genes to mesenchymal stem cells (MSCs). PAMAM dendrimers are monodispersive, hyperbranched and nanospherical molecules presenting unique characteristics that make them very attractive vehicles for both drug and gene delivery. Although they have been explored for gene delivery in a wide range of cell lines, the interaction and the usefulness of these molecules in the delivery of genes to MSCs remains a field to be explored. Adult MSCs were chosen for the studies due to their potential biomedical applications (they are considered multipotent cells) and because they present several advantages over embryonic stem cells, such as easy accessibility and the inexistence of ethical restrictions to their use. This thesis is divided in 5 interconnected chapters. Chapter I provides an overview of the current literature concerning the various non-viral systems investigated for gene delivery in MSCs. Attention is devoted to physical methods, as well as to chemical methods that make use of polymers (natural and synthetic), liposomes, and inorganic nanoparticles as gene delivery vectors. Also, it summarizes the current applications of genetically engineered mesenchymal stem cells using non-viral systems in regenerative medicine, with special focus on bone tissue regeneration. In Chapter II, the potential of native PAMAM dendrimers with amine termini to transfect MSCs is evaluated. The level of transfection achieved with the dendrimers is, in a first step, studied using a plasmid DNA (pDNA) encoding for the β-galactosidase reporter gene. The effect of dendrimer’s generation, cell passage number, and N:P ratio (where N= number of primary amines in the dendrimer; P= number of phosphate groups in the pDNA backbone) on the level of transfection is evaluated, being the values always very low. In a second step, a pDNA encoding for bone morphogenetic protein-2, a protein that is known for its role in MSCs proliferation and differentiation, is used. The BMP-2 content produced by transfected cells is evaluated by an ELISA assay and its effect on the osteogenic markers is analyzed through several classical assays including alkaline phosphatase activity (an early marker of osteogenesis), osteocalcin production, calcium deposition and mineralized nodules formation (late osteogenesis markers). Results show that a low transfection level is enough to induce in vitro osteogenic differentiation in MSCs. Next, from Chapter III to Chapter V, studies are shown where several strategies are adopted to change the interaction of PAMAM dendrimers with MSCs cell membrane and, as a consequence, to enhance the levels of gene delivery. In Chapter III, generations 5 and 6 of PAMAM dendrimers are surface functionalized with arginine-glycine-aspartic acid (RGD) containing peptides – experiments with dendrimers conjugated to 4, 8 and 16 RGD units were performed. The underlying concept is that by including the RGD integrin-binding motif in the design of the vectors and by forming RGD clusters, the level of transfection will increase as MSCs highly express integrins at their surface. Results show that cellular uptake of functionalized dendrimers and gene expression is enhanced in comparison with the native dendrimers. Furthermore, gene expression is dependent on both the electrostatic interaction established between the dendrimer moiety and the cell surface and the nanocluster RGD density. In Chapter IV, a new family of gene delivery vectors is synthesized consisting of a PAMAM dendrimer (generation 5) core randomly linked at the periphery to alkyl hydrophobic chains that vary in length and number. Herein, the idea is to take advantage of both the cationic nature of the dendrimer and the capacity of lipids to interact with biological membranes. These new vectors show a remarkable capacity for internalizing pDNA, being this effect positively correlated with the –CH2– content present in the hydrophobic corona. Gene expression is also greatly enhanced using the new vectors but, in this case, the higher efficiency is shown by the vectors containing the smallest hydrophobic chains. Finally, chapter V reports the synthesis, characterization and evaluation of novel gene delivery vectors based on PAMAM dendrimers (generation 5) conjugated to peptides with high affinity for MSCs membrane binding - for comparison, experiments are also done with a peptide with low affinity binding properties. These systems present low cytotoxicity and transfection efficiencies superior to those of native dendrimers and partially degraded dendrimers (Superfect®, a commercial product). Furthermore, with this biomimetic approach, the process of gene delivery is shown to be cell surface receptor-mediated. Overall, results show the potential of PAMAM dendrimers to be used, as such or modified, in Tissue Regeneration and Engineering. To our knowledge, this is the first time that PAMAM dendrimers are studied as gene delivery vehicles in this context and using, as target, a cell type with clinical relevancy. It is shown that the cationic nature of PAMAM dendrimers with amine termini can be synergistically combined with surface engineering approaches, which will ultimately result in suitable interactions with the cytoplasmic membrane and enhanced pDNA cellular entry and gene expression. Nevertheless, the quantity of pDNA detected inside cell nucleus is always very small when compared with the bigger amount reaching cytoplasm (accumulation of pDNA is evident in the perinuclear region), suggesting that the main barrier to transfection is the nuclear membrane. Future work can then be envisaged based on the versatility of these systems as biomedical molecular materials, such as the conjugation of PAMAM dendrimers to molecules able to bind nuclear membrane receptors and to promote nuclear translocation.

Comparação e validação de vários métodos de mapeamento cortical para o planeamento pré-cirúrgico por Imagem Funcional de Ressonância Magnética

Na análise funcional de imagens do cérebro podem utilizar-se diferentes métodos na identificação de zonas de activação. Tem havido uma evolução desde o método de correlação [19], para outros métodos [9] [14] até o método baseado no modelo linear generalizado que é mais comum ser utilizado hoje e que levou ao pacote de software SPM [15]. Deve-se principalmente à versatilidade que o método tem em realizar testes com diferentes objectivos. Têm sido publicados alguns estudos comparativos. Poucos têm sido quantitativos [20] e quando o são, o número de métodos testados é reduzido[22]. Há muitos estudos comparativos do ponto de vista da estatística envolvida (da matemática) mas que têm em geral apenas ns académicos. Um objectivo deste estudo é comparar os resultados obtidos por diferentes métodos. É de particular interesse averiguar o comportamento de cada método na fronteira do local de activação. As diferenças serão avaliadas numericamente para os seguintes métodos clássicos: t de Student, coeficiente de correlação e o modelo linear generalizado. Três novos métodos são também propostos - o método de picos de Fourier, o método de sobreposição e o método de amplitude. O segundo pode ser aplicado para o melhoramento dos métodos de t de Student, coe ciente de correlação e modelo linear generalizado. Ele pode no entanto, também manter-se como um método de análise independente. A influência exercida em cada método pelos parâmetros pertinentes é também medida. É adoptado um conjunto de dados clínicos que está amplamente estudado e documentado. Desta forma elimina-se a possibilidade dos resultados obtidos serem interpretados como sendo específicos do caso em estudo. Há situações em que a influência do método utilizado na identificação das áreas de activação de imagens funcionais do cérebro é crucial. Tal acontece, por exemplo, quando um tumor desenvolve-se perto de uma zona de activação responsável por uma função importante . Para o cirurgião tornase indispensável avaliar se existe alguma sobreposição. A escolha de um dos métodos disponíveis poderá ter infuência sobre a decisão final. Se o método escolhido for mais conservador, pode verificar-se sobreposição e eliminar-se a possibilidade de cirurgia. Porém, se o método for mais restritivo a decisão final pode ser favorável à cirurgia. Artigos recentes têm suportado a ideia de que a ressonância magnética funcional é de facto muito útil no processo de decisão pré-operatório [12].O segundo objectivo do estudo é então avaliar a sobreposição entre um volume de activação e o volume do tumor. Os programas informáticos de análise funcional disponíveis são variados em vários aspectos: na plataforma em que funcionam (macintosh, linux, windows ou outras), na linguagem em que foram desenvolvidos (e.g. c+motif, c+matlab, matlab, etc.) no tratamento inicial dos dados (antes da aplicação do método de análise), no formato das imagens e no(s) método(s) de análise escolhido(s). Este facto di culta qualquer tentativa de comparação. À partida esta poderá apenas ser qualitativa. Uma comparação quantitativa implicaria a necessidade de ocorrerem três factos: o utilizador tem acesso ao código do programa, sabe programar nas diferentes linguagens e tem licença de utilização de software comercial (e.g. matlab). Sendo assim foi decidido adoptar uma estratégia unificadora. Ou seja, criar um novo programa desenvolvido numa linguagem independente da plataforma, que não utilize software comercial e que permita aplicar (e comparar quantitativamente) diferentes métodos de análise funcional. A linguagem escolhida foi o JAVA. O programa desenvolvido no âmbito desta tese chama-se Cérebro.