No description available.

A mecânica celular jaz nas propriedades materiais da membrana plasmática, fundamentalmente uma bicamada fosfolipídica com espessura de dimensões moleculares. Além de forças elásticas, tal material bidimensional também experimenta tensões viscosas devido ao seu comportamento fluido (presumivelmente newtoniano) na direção tangencial. A despeito da notável concordância entre teoria e experimentos biofísicos sobre a geometria de membranas celulares, ainda não se faz presente um método computacional para simulação de sua (real) dinâmica viscosa governada pela lei de Boussinesq-Scriven. Assim sendo, introduzimos uma formulação variacional mista de três campos para escoamentos viscosos de superfícies fechadas curvas. Nela, o fluido circundante é levado em conta considerando-se uma restrição de volume interior, ao passo que uma restrição de área corresponde à inextensibilidade. As incógnitas são a velocidade, o vetor curvatura e a pressão superficial, todas estas interpoladas com elementos finitos lineares contínuos via estabilização baseada na projeção do gradiente de pressão. O método é semi-implícito e requer a solução de apenas um único sistema linear por passo de tempo. Outro ingrediente numérico proposto é uma força que mimetiza a ação de uma pinça óptica, permitindo interação virtual com a membrana, onde a qualidade e o refinamento de malha são mantidos por remalhagem adaptativa automática. Extensivos experimentos numéricos de dinâmica de relaxação são apresentados e comparados com soluções quasi-analíticas. É observada estabilidade temporal condicional com uma restrição de passo de tempo que escala como o quadrado do tamanho de malha. Reportamos a convergência e os limites de estabilidade de nosso método e sua habilidade em predizer corretamente o equilíbrio dinâmico de compridas e finas elongações cilíndricas (tethers) que surgem a partir de pinçamentos membranais. A dependência de forma membranal com respeito a uma velocidade imposta de pinçamento também é discutida, sendo que há um valor limiar de velocidade abaixo do qual um tether não se forma de início. Testes adicionais ilustram a robustez do método e a relevância dos efeitos viscosos membranais quando sob a ação de forças externas. Sem dúvida, ainda há um longo caminho a ser trilhado para o entendimento completo da mecânica celular (há de serem consideradas outras estruturas tais como citoesqueleto, canais iônicos, proteínas transmembranares, etc). O primeiro passo, porém, foi dado: a construção de um esquema numérico variacional capaz de simular a intrigante dinâmica das membranas celulares. / Cell mechanics lies on the material properties of the plasmatic membrane, fundamentally a two-molecule-thick phospholipid bilayer. Other than bending elastic forces, such a two-dimensional interfacial material also experiences viscous stresses due to its (presumably Newtonian) surface fluid tangential behaviour. Despite the remarkable agreement on membrane shapes between theory and biophysical experiments, there is no computational method for simulating its (actual) viscous dynamics governed by the Boussinesq- Scriven law. Accordingly, we introduce a mixed three-field variational formulation for viscous flows of closed curved surfaces. In it, the bulk fluid is taken into account by means of an enclosed-volume constraint, whereas an area constraint stands for the membranes inextensible character. The unknowns are the velocity, vector curvature and surface pressure fields, all of which are interpolated with linear continuous finite elements by means of a pressure-gradient-projection scheme. The method is semi-implicit and it requires the solution of a single linear system per time step. Another proposed ingredient is a numerical force that emulates the action of an optical tweezer, allowing for virtual interaction with the membrane, where mesh quality and refinement are maintained by adaptively remeshing. Extensive relaxation experiments are reported and compared with quasi-analytical solutions. Conditional time stability is observed, with a time step restriction that scales as the square of the mesh size. We discuss both convergence and the stability limits of our method, its ability to correctly predict the dynamical equilibrium of the tether due to tweezing. The dependence of the membrane shape on imposed tweezing velocities is also addressed. Basically, there is a threshold velocity value below which the tethers shape does not arise at first. Further tests illustrate the robustness of the method and show the significance of viscous effects on membranes deformation under external forces. Undoubtedly, there is still a long way to track toward the understanding of celullar mechanics (one still has to account other structures such as cytoskeleton, ion channels, transmembrane proteins, etc). The first step has given, though: the design of a numerically robust variational scheme capable of simulating the engrossing dynamics of fluid cell membranes.

No description available.

Les ARNm sont des molécules centrales pour la régulation des gènes, aidant à convertir l'information génétique stockée dans l'ADN en protéines fonctionnelles. En tant que polymère simple brin, mesurant des centaines à des milliers de nucléotides, les ARNm peuvent former des structures secondaires et tertiaires étendues formant des particules appelés ribonucléoprotéines messagères (RNPm) en s’assemblant avec des protéines. L'organisation 3D des (pré-)RNPm influence de nombreux aspects de leur métabolisme, incluant la régulation de leur maturation, de leur export et de leur traduction dans le cytoplasme. Malgré leur importance, notre compréhension de l'organisation structurelle des (pré-)RNPm in vivo, et des principes qui la régissent est minime. Au cours de ma thèse, j'ai analysé l'organisation des (pré-)mRNP en développant une vision centrée sur l'ARN. Pour cela, j'ai mis en place une approche combinant l'hybridation in situ d'ARN monomoléculaire (smFISH) avec la microscopie à illumination structurée (SIM) et l'ai utilisée pour étudier l'organisation des mRNP dans le noyau et le cytoplasme. Nos résultats suggèrent que l'organisation (pré-)mRNP varie à différents stades de sa vie. Nous montrons que l'empaquetage (pré-)mRNP commence de manière co-transcriptionnelle, avec des introns organisés en conformations compactes. Cette organisation est modifiée au cours de la transcription au fur et à mesure que la polymérase se déplace le long du gène, assemblant finalement un intron avec les extrémités à proximité l’une de l’autre, d'une manière dépendante du spliceosome, suggérant que l'organisation co-transcriptionnelle des introns pourrait être critique pour déterminer son excision. Une fois libérés, les mRNP ont une organisation linéaire compacte dans le nucléoplasme et éventuellement une conformation en tige. L'organisation d’un mRNP dans le cytoplasme est influencée par sa traduction. Alors que la traduction ouvre les mRNP, la séparation des extrémités de l'ARNm, l'inhibition de la traduction et la libération de ribosomes, ou le recrutement dans les granules de stress, donnent aux mRNP une structure très compacte. Fait intéressant, nous trouvons rarement des ARNm avec les extrémités 5' et 3' à proximité, ce qui suggère que la traduction en boucle fermée n'est pas un état universel pour tous les ARNm en cours de traduction. Ensemble, nos résultats fournissent une image essentielle de l'organisation du mRNP dans les cellules et souligne le rôle important de la conformation du RNPm dans la régulation de la traduction et de la maturation d’une RNPm. / mRNAs act as the central molecules in gene regulation, helping convert the genetic information stored in the DNA to functional proteins. As a single-stranded polymer, hundreds to thousands of nucleotides in length, mRNAs can form extensive secondary and tertiary structures and, together with proteins, are packaged into assemblies called messenger ribonucleoproteins (mRNPs). The 3D organisation of (pre-)mRNPs influences many aspects of what happens to them, including regulating their processing, export and translation in the cytoplasm. Despite their significance, our understanding of the structural organisation of (pre-)mRNPs in vivo is minimal, as is our comprehension of the principles that govern it. During my PhD, I have developed an RNA-centric view on (pre-)mRNP organisation. For this, I have established an approach combining single-molecule RNA in situ hybridisation (smFISH) with structured illumination microscopy (SIM) and used it to study mRNP organisation in the nucleus and cytoplasm. Our results suggest that (pre-)mRNP organisation is altered at various stages during its lifetime. We show that (pre-)mRNP packaging starts co-transcriptionally, with introns organised into compact conformations. This organisation is altered during the course of transcription as the polymerase travels along the gene, finally assembling an intron with the ends in proximity in a spliceosome dependent manner, suggesting that co-transcriptional intron organisation could be critical in determining its excision. Once released, mRNPs have a compact linear organisation in the nucleoplasm and possibly a rod-like conformation. mRNP organisation in the cytoplasm is influenced by its translational status. While translation opens up mRNPs, separating the ends of the mRNA, translation inhibition and release of ribosomes, or recruitment to stress granules result in mRNPs having a highly compact structure. Interestingly, we rarely find mRNAs with the 5’ and 3’ ends in proximity, suggesting that closed-looped translation is not a universal state for all translating mRNAs. Together, our results provide a unique and essential view of mRNP organisation in cells and reveal important insight into the role of mRNP conformation in regulating translation and mRNP processing.