O formato, posição e alinhamento corretos dos componentes oculares são definidos através de uma série de mudanças morfológicas complexas durante sua embriogênese. A retina se origina de células da vesícula óptica do tubo neural enquanto as células precursoras do cristalino surgem do ectoderma que reveste o ápice da vesícula óptica. Este ectoderma é delimitado molecularmente como pre-placoidal, sofre uma série de eventos morfogênicos durante o seu desenvolvimento inicial para formar o placóide do cristalino e, posteriormente, a vesícula do cristalino. O placóide do cristalino surge a partir do espessamento do ectoderma pre-placoidal. Posteriormente, o placóide invagina para formar a vesícula do cristalino. Durante a invaginação do placóide, as células da ectoderme que circundam o placóide (células periplacodais) também se movem para fechar a abertura do cristalino que invaginou e reconstruir o ectoderma da superfície. Aqui, nos concentramos em dois processos do desenvolvimento do olho. Nós investigamos o papel da matriz extracelular no espessamento do placóide do cristalino e a dinâmica da emissão de protrusões de membrana pelas células periplacodais durante a invaginação do placóide. A matriz extracelular desempenha papel relevante na morfogênese placodal. Por exemplo, a Fibronectina na matriz extracelular entre a vesícula óptica e o ectoderma pré-placoidal é necessária para a formação de placóide do cristalino. No entanto, a dinâmica da arquitetura de Fibronectina durante a formação do placóide é desconhecida. Assim, nosso primeiro objetivo aqui foi investigar a arquitetura da Fibronectina e da Laminina, dois importantes componentes da matriz extracelular, durante o espessamento do placóide do cristalino através de imagens confocais em 3D. Nossos dados sugerem que um padrão de Fibronectina e Laminina difuso e pontuado é restrito à região do placóide. Este padrão é mantido durante o espessamento e invaginação do placóide. Encontramos um padrão similar de Laminina na região do placóide de embrião de camundongo, sugerindo a conservação desta arquitetura neste contexto. Também demonstramos que a inibição mediada por Noggin (inibidor da sinalização de BMP), que interrompe o desenvolvimento do olho, afeta a organização da Fibronectina e da Laminina, sugerindo que a sinalização de BMP regula a organização da matriz extracelular durante o desenvolvimento do placóide do cristalino. Nosso segundo objetivo foi analisar a emissão de 5 protrusões celulares finas por células periplacodais correlacionando com o movimento de invaginação. Aqui, nós investigamos a dinâmica e composição do citoesqueleto dessas protrusões para entender sua função durante o desenvolvimento do olho. Observamos uma grande quantidade de protrusões em células periplacodais de embriões de galinha e de camundongo. Nossos resultados de quantificação com protrusões de embriões de galinha não mostraram correlação entre comprimento e direção de emissão ou com meia-vida. Nós também analisamos a diversidade na composição do citoesqueleto, uma vez que encontramos protrusões positivas para Cofilina e Tubulina. Estes dados sugerem uma população heterogênea de protrusões finas de membrana periplacodais. Finalmente, também identificamos essas protrusões em outras superfícies ectodérmicas de embriões de galinha e de camundongo, sugerindo que elas desempenham um papel no desenvolvimento de ectoderme superficial. / The correct shape, position and alignment of optic components are defined through a series of complex morphological changes during the embryogenesis of the eye. The retina originates from the neural tube´s optic vesicle while the lens precursor cells arise from the ectoderm that overlie the apex of the optic vesicle. This ectoderm is molecularly delimited as preplacodal and undergoes a series of morphogenic events during its initial development to form the lens placode and subsequently the lens vesicle. The lens placode arises from the thickening of the pre-placodal ectoderm. Subsequently, the placode invaginates to form the vesicle of the lens. During the invagination of the placode, the ectodermal cells that surround the placode (peri-placodal cells) also move to close the opening of the lens that invaginated and reconstruct the surface ectoderm. Here we focus on two processes of eye development. We investigated the role of the extracellular matrix in the lens placode thickening and the dynamics of the emission of membrane protrusions by the peri-placodal cells during the lens placode invagination. The extracellular matrix plays a relevant role in placodal morphogenesis. For example, Fibronectin in the extracellular matrix between the optic vesicle and the preplacodal ectoderm is required for the formation of lens placode. However, the dynamics of the Fibronectin architecture during placode formation is unknown. Thus, our first aim here was to investigate the architecture of Fibronectin and Laminin, two important components of the extracellular matrix, during thickening of lens placode through confocal 3D images. Our data suggest that both Fibronectin and Laminin present a diffuse and punctate pattern restricted to the placodal region. This pattern is maintained during thickening and invagination of the placode. We found a similar pattern of Laminin in the placodal region of the mouse embryo, suggesting the conservation of this architecture in this context. We also demonstrate that Noggin-mediated inhibition of BMP signalling, which disrupts the development of the eye, affects the organization of Fibronectin and Laminin, suggesting that BMP signalling regulates the organization of the extracellular matrix during the lens placode development. Our second objective was to analyse the emission of thin cellular protrusions by peri-placodal cells correlating with the lens invagination movement. Here we investigated the dynamics and 3 composition of the cytoskeleton of these protrusions to understand their function during the development of the eye. We observed a large number of protrusions in peri-satellite cells of chicken and mouse embryos. Our quantification results with chicken embryo protrusions showed no correlation between length and direction of emission or with half-life. We also analysed the diversity in the composition of the cytoskeleton, and we found protrusions positive for Cofilin and Tubulin. These data suggest a heterogeneous population of periplacodal protrusions. Finally, we have also identified these protrusions on other ectodermal surfaces of chicken and mouse embryos, suggesting that they play a role in the development of surface ectoderm.

Gewebezellen sammeln ständig Informationen über die mechanischen Eigenschaften ihrer Umgebung, indem sie aktiv an dieser ziehen. Diese Kräfte werden an Zell-Matrix-Kontakten übertragen, die als Mechanosensoren fungieren. Jüngste Experimente mit Zellen auf elastischen Substraten zeigen, dass Zellen sehr empfindlich auf Veränderungen der effektiven Steifigkeit ihrer Umgebung reagieren, die zu einer Reorganisation des Zytoskeletts führen können. In dieser Arbeit wird ein theoretisches Model entwickelt, um die Selbstorganisation von Zellen in weichen Materialien vorherzusagen. Obwohl das Zellverhalten durch komplexe regulatorische Vorgänge in der Zelle gesteuert wird, scheint die typische Antwort von Zellen auf mechanische Reize eine einfache Präferenz für große effektive Steifigkeit der Umgebung zu sein, möglicherweise weil in einer steiferen Umgebung Kräfte an den Kontakten effektiver aufgebaut werden können. Der Begriff Steifigkeit umfasst dabei sowohl Effekte, die durch größere Härte als auch durch elastische Verzerrungsfelder in der Umgebung verursacht werden. Diese Beobachtung kann man als ein Extremalprinzip in der Elastizitätstheorie formulieren. Indem man das zelluläre Kraftmuster spezifiziert, mit dem Zellen mit ihrer Umgebung wechselwirken, und die Umgebung selbst als linear elastisches Material modelliert, kann damit die optimale Orientierung und Position von Zellen vorhergesagt werden. Es werden mehrere praktisch relevante Beispiele für Zellorganisation theoretisch betrachtet: Zellen in externen Spannungsfeldern und Zellen in der Nähe von Grenzflächen für verschiedene Geometrien und Randbedingungen des elastischen Mediums. Dafür werden die entsprechenden elastischen Randwertprobleme in Vollraum, Halbraum und Kugel exakt gelöst. Die Vorhersagen des Models stimmen hervorragend mit experimentellen Befunden für Fibroblastzellen überein, sowohl auf elastischen Substraten als auch in physiologischen Hydrogelen. Mechanisch aktive Zellen wie Fibroblasten können auch elastisch miteinander wechselwirken. Es werden daher optimale Strukturen als Funktion von Materialeigenschaften und Zelldichte bzw. der Geometrie der Zellpositionen berechnet. Schließlich wird mit Hilfe von Monte Carlo Simulationen der Einfluss stochastischer Störungen auf die Strukturbildung untersucht. Das vorliegende Model trägt nicht nur zu einem besseren Verständnis von vielen physiologischen Situationen bei, sondern könnte in Zukunft auch für biomedizinische Anwendungen benutzt werden, um zum Beispiel Protokolle für künstliche Gewebe im Bezug auf Substratgeometrie, Randbedingungen, Materialeigenschaften oder Zelldichte zu optimieren. / Adherent cells constantly collect information about the mechanical properties of their extracellular environment by actively pulling on it through cell-matrix contacts, which act as mechanosensors. In recent years, the sophisticated use of elastic substrates has shown that cells respond very sensitively to changes in effective stiffness in their environment, which results in a reorganization of the cytoskeleton in response to mechanical input. We develop a theoretical model to predict cellular self-organization in soft materials on a coarse grained level. Although cell organization in principle results from complex regulatory events inside the cell, the typical response to mechanical input seems to be a simple preference for large effective stiffness, possibly because force is more efficiently generated in a stiffer environment. The term effective stiffness comprises effects of both rigidity and prestrain in the environment. This observation can be turned into an optimization principle in elasticity theory. By specifying the cellular probing force pattern and by modeling the environment as a linear elastic medium, one can predict preferred cell orientation and position. Various examples for cell organization, which are of large practical interest, are considered theoretically: cells in external strain fields and cells close to boundaries or interfaces for different sample geometries and boundary conditions. For this purpose the elastic equations are solved exactly for an infinite space, an elastic half space and the elastic sphere. The predictions of the model are in excellent agreement with experiments for fibroblast cells, both on elastic substrates and in hydrogels. Mechanically active cells like fibroblasts could also interact elastically with each other. We calculate the optimal structures on elastic substrates as a function of material properties, cell density and the geometry of cell positioning, respectively, that allows each cell to maximize the effective stiffness in its environment due to the traction of all the other cells. Finally, we apply Monte Carlo simulations to study the effect of noise on cellular structure formation. The model not only contributes to a better understanding of many physiological situations. In the future it could also be used for biomedical applications to optimize protocols for artificial tissues with respect to sample geometry, boundary condition, material properties or cell density.

The mouse pancreas is a mixed exocrine and endocrine glandconsisting of three lobular compartments: the splenic, duodenal and gastric lobes. During embryogenesis, the pancreas forms from two progenitor populations located on the dorsal and ventral side of the primitive gut tube. These anlagen are brought in close proximity as the gut elongates and rotates, and fuse to form a single organ. The splenic and duodenal lobes develop from the dorsal and ventral anlagen, respectively. In the adult pancreas, exocrine tissue secretes digestive enzymes intothe gut lumen to support nutrient uptake. The endocrine Islets of Langerhans are scattered throughout the exocrine tissue and aid in regulation of energy homeostasis through the secretion of hormones. One of the key players in energy homeostasis is the pancreatic ß-cell, which is the most abundant cell type of the islets. The β-cells regulates blood glucose levels through the action of insulin. Conditions where this regulation does not function properly are gathered under the common name of Diabetes mellitus. Type 1 diabetes (T1D) is characterized by insulin deficiency due to autoimmune destruction of the ß-cells. Using recently developed protocols for optical projection tomography (OPT) whole-organ imaging, we have revealed new spatial and quantitative aspects on ß-cell mass dynamics and immune infiltration during the course of T1D development in the non-obese diabetic (NOD) mouse model. We show that although immune infiltration appears to occur asynchronously throughout the organ, smaller islets, mainly located in the periphery of the organ, preferentially loose their ß-cells during early stages of disease progression. Larger islets appear more resistant to the autoimmune attack and our data indicate the existence of a compensatory proliferative capacity within these islets. We also report the appearance of structures resembling tertiary lymphoid organs (TLOs) in association with the remaining islets during later phases of T1D progression. OPT has already proven to be a useful tool for assessments of ß-cellmass in the adult mouse pancreas. However, as with other techniques, previous protocols have relied on a tedious degree of manual postivacquisition editing. To further refine OPT-based assessment of pancreatic ß-cell mass distribution in the murine pancreas, we implemented a computational statistical approach, Contrast-Limited Adaptive Histogram Normalisation (CLAHE), to the OPT projection data of pancreata from C57Bl/6 mice. This methodology provided increased islet detection sensitivity, improved islet morphology and diminished subjectivity in thresholding for reconstruction and quantification. Using this approach, we could report a substantially higher number of islets than previously described for this strain and provide evidence of significant differences in islet mass distribution between the pancreatic lobes. The gastric lobe stood out in particular and contained a 75% higher islet density as compared to the splenic lobe. Although the development of the early pancreatic buds has been relatively well studied, later morphogenetic events are less clear and information regarding the formation of the gastric lobe has largely been missing. Using OPT we have generated a quantitative three-dimensional road map of pancreatic morphogenesis in the mouse. We show that the gastric lobe forms as a perpendicular outgrowth fromthe stem of the dorsal pancreas at around embryonic day (e) 13.5, which grows into a mesenchymal domain overlaying the pyloric sphincter and proximal part of the glandular stomach. By analyzing mutant mice with aberrant spleen development, we further demonstrate that proper formation of the gastric lobe is dependent on the initial formation of the closely positioned spleen, indicating a close interplay between pancreatic and splenic mesenchyme during development. Additionally, we show that the expression profile of markers for pancreatic multipotent progenitors within the pancreas is heterogenous with regards to lobular origin. Altogether, our studies regarding the morphogenesis and adult constitution of the mouse pancreas recognize lobular heterogeneities that add important information for future interpretations of this organ.

RNA polymerase II (Pol II) is the enzyme responsible for the synthesis of all mRNAs in eukaryotic cells. As the central component of the eukaryotic transcription machinery, Pol II is the final target of regulatory pathways. While the role for different Pol II associated proteins, co-activators and general transcription factors (GTFs) in regulation of transcription in response to different stimuli is well studied, a similar role for some subunits of the core Pol II is only now being recognized. The studies reported in this thesis address the role of the fourth largest subunit of Pol II, Rpb4, in transcription and stress response using Saccharomyces cerevisiae as the model system. Rpb4 is closely associated with another smaller subunit, Rpb7 and forms a dissociable complex (Edwards et al., 1991). The rpb4 null mutant is viable but is unable to survive at extreme temperatures (>34ºC and <12ºC) (Woychik and Young, 1989). This mutant has also been shown to be defective in activated transcription and unable to respond properly in several stress conditions (Pillai et al., 2001; Sampath and Sadhale, 2005). In spite of wealth of available information, the exact role of Rpb4 remains poorly understood. In the present work, we have used genetic, molecular and biochemical approaches to understand the role of Rpb4 as described in four different parts below: i) Studies on Genetic and Functional Interactions of Rpb4 with SAGA/TFIID Complex to Confer Promoter- Specific Transcriptional Control To carry out transcription, Pol II has to depend on several general transcription factors, mediators, activators, and co-activators and chromatin remodeling complexes. In the present study, we tried to understand the genetic and functional relationship of Rpb4 with some of the components of transcription machinery, which will provide some insight into the role of Rpb4 during transcription. Our microarray analysis of rpb4∆ strain suggests that down regulated genes show significant overlap with genes regulated by the SAGA complex, a complex functionally related to TFIID and involved in regulation of the stress dependent genes. The analysis of combination of double deletion mutants of either the SAGA complex subunits or the TFIID complex with rpb4∆ showed that both these double mutants are extremely slow growing and show synthetic growth phenotype. Further studies, including microarray analysis of these double mutants and ChIP (chromatin immunoprecipitation) of Rpb4 and SAGA complex, suggested that Rpb4 functions together with SAGA complex to regulate the expression of stress dependent genes. ii) Study of Genome Wide Recruitment of Rpb4 and Evidence for its Role in Transcription Elongation Biochemical studies have shown that Rpb4 associates sub-stoichiometrically with the core RNA polymerase during log phase but whether recruitment of Rpb4 is promoter context dependent or occurs only at specific stage of transcription remains largely unknown. Having discovered that Rpb4 can recruit on both TFIID and SAGA dominated promoters, it was important to study the genome wide role of Rpb4. Using ChIP on chip experiments, we have carried out a systematic assessment of genome wide binding of Rpb4 as compared to the core Pol II subunit, Rpb3. Our analysis showed that Rpb4 is recruited on coding regions of most transcriptionally active genes similar to the core Pol II subunit Rpb3 albeit to a lesser extent. This extent of Rpb4 recruitment increased on the coding regions of long genes pointing towards a role of Rpb4 in transcription elongation of long genes. Further studies showing transcription defect of long and GC rich genes, 6-azauracil sensitivity and defective PUR5 gene expression in rpb4∆ mutant supported the in vivo evidence of the role of Rpb4 in transcription elongation. iii) Genome Wide Expression Profiling and RNA Polymerase II Recruitment in rpb4∆ Mutant in Non-Stress and Stress Conditions Structural studies have suggested a role of Rpb4/Rpb7 sub-complex in recruitment of different factors involved in transcription (Armache et al., 2003; Bushnell and Kornberg, 2003). Though only few studies have supported this aspect of Rpb4/Rpb7 sub-complex, more research needs to be directed to explore this role of Rpb4/Rpb7 sub-complex. To study if Rpb4 has any role in recruitment of Pol II under different growth conditions, we have studied genome wide recruitment of Pol II in the presence and absence of Rpb4 during growth in normal rich medium as well as under stress conditions like heat shock and stationary phase where Rpb4 is shown to be indispensable for survival. Our analysis showed that absence of Rpb4 results in overall reduced recruitment of Pol II in moderate condition but this reduction was more pronounced during heat shock condition. During stationary phase where overall recruitment of Pol II also goes down in wild type cells, absence of Rpb4 did not lead to further decrease in overall recruitment. Interestingly, increased expression levels of many genes in the absence of Rpb4 did not show concomitant increase in the recruitment of Pol II, suggesting that Rpb4 might regulate these genes at a post-transcriptional step. iv) Role of Rpb4 in Pseudohyphal Growth The budding yeast S. cerevisiae can initiate distinct developmental programs depending on the presence of various nutrients. In response to nitrogen starvation, diploid yeast undergoes a dimorphic transition to filamentous pseudohyphal growth, which is regulated through cAMP-PKA and MAP kinase pathways. Previous work from our group has shown that rpb4∆ strain shows predisposed pseudohyphal morphology (Pillai et al., 2003), but how Rpb4 regulates this differentiation program is yet to be established. In the present study, we found that disruption of Rpb4 leads to enhanced pseudohyphal growth, which is independent of nutritional status. We observed that the rpb4∆/ rpb4∆ cells exhibit pseudohyphae even in the absence of a functional MAP kinase and cAMP-PKA pathways. Genome wide expression profile showed that several downstream genes of RAM signaling pathway were down regulated in rpb4∆ cells. Our detailed genetic analysis further supported the hypothesis that down regulation of RAM pathway might be leading to the pseudohyphal morphogenesis in rpb4∆ cells.

A hallmark feature of Dictyostelium development is the establishment and maintenance of precise cell-type proportions. In the case of D. discoideum, roughly 20% of the cells that aggregate form the stalk while the remaining 80% form the spores. In order to identify genes involved in cell-type proportioning Jaiswal et al. (2006) carried out random insertional mutagenesis (REMI) of the D. discoideum genome. This led to the identification of a novel gene, which was named trishanku (triA). A knock-out of triA did not show any defects during growth and early development but multiple defects later during development. To understand the reasons for the multiple developmental defects in the absence of triA, I looked at the genomic organization and the pattern of expression of the triA gene. In silico analysis points to the presence of more than one consensus D. discoideum promoter sequence upstream to exons1 and 2, raising the possibility that the triA gene could code for more than one transcript. Northern blot analysis confirms this prediction and provides evidence for the presence of two transcripts: triA1-2-3 (~ 2.9 kb, containing exons 1+2+3) and triA2-3 (~ 2 kb, containing exons 2+3). Both transcripts have exons 2 and 3 in common. In triA- cells, the REMI cassette is inserted in exon 2, which is common to both transcripts; thus, the absence of triA results in the lack of both. The transcripts are absent in vegetative cells but expressed during development. triA2-3 is expressed earlier, by 3h, while triA1-2-3 is expressed later, by 9h, and both remain till the end of development. triA2-3 and triA1-2-3 are differentially regulated by different aspects of the extracellular environment which include mode of development of cells (solid substratum versus shaken suspension), the presence of a high level of extracellular cAMP and formation of stable cell-cell contacts. The expression of triA2-3 and triA1-2-3 in triA- cells, one at a time under a constitutive promoter (Actin15 promoter), suggests that the two transcripts have both specific as well as overlapping functions in the cell. The triA2-3 transcript can specifically restore spore forming efficiency and stalk thickness, while the triA1-2-3 transcript can rescue the stream break up defect. Both the transcripts can rescue the sub-terminal position of the sorus, spore shape and spore viability. To address the question of stream break-up during mid to late aggregation in triA- cells, I have looked at the cell adhesion profile of triA- cells and compared it with the wild type (Ax2). triA- cells show transient disaggregation in buffer and a 2h delay in agglutination in presence of buffer with 10mM EDTA. This aberrant cell adhesion profile seen in triA- cells is in accordance with the expression pattern of genes encoding known cell adhesion molecules. triA- cells also overproduce an extracellular factor which significantly decreases the aggregate size of both Ax2 and triA-. The nature of the extracellular factor overproduced by in triA- cells is currently unknown, but it is not the same as cell-counting factor which is overproduced by smlA null cells. To look at the mis-expression of cell type-specific genes, I have monitored the movement of prestalk cells into the prespore region and vice versa in both Ax2 and triA- slugs. My studies show that there is extensive movement of prestalk cells into the prespore region and of prespore cells into the prestalk region in triA- slugs, which is absent in Ax2 slugs. Also, cells that move into the ‘wrong’ region show a change their cell fate (transdifferentiate) appropriate to the new location; whether transdifferentiation precedes or succeeds cell movement is not yet clear. Transdifferentiation is observed to a certain extent in Ax2 slugs, but only after prolonged migration; triA- slugs show enhanced transdifferentiation even in the absence of migration. To find out the possible reason(s) for the formation of a sub-terminal spore mass in the absence of triA, I have checked whether the defect lies in the ability of the prespore cells to rise up the stalk or in the ability of the upper cup (cells present above the spore mass contributed by a subset of prestalk cells and anterior like-cells) to pull the spore mass to the top. To see which of the two reasons could be responsible for the formation of a sub-terminal spore mass in triA-, I carried out transplantation experiments where the anterior one-fourth region of an Ax2 or triA- slug is grafted to the posterior four-fifth region of a triA- or Ax2 slug and the morphology of the fruiting body is observed. My studies show that the sub-terminal position of the spore mass in triA- is not due to an inability of the prespore cells to rise to the top but to a defect in the upper cup. The upper cup in triA- remains motile but is unable to remain attached to the prespore mass during culmination. It detaches, rises up the stalk and is present at the tip of the stalk. Mixing a minority of triA- cells (20%) with an excess of Ax2 (80%) results in an upper up formed by Ax2 alone. In this situation, the wild type upper cup is able to lift the triA- prespore mass to the top. Thus, the presence of triA (a prespore-specific gene) is essential for the proper functioning of the upper cup cells (which belong to the prestalk class) in order to enable prespore cells to ascend to the top of the stalk.

No description available.

Five image processing algorithms are proposed to measure the average orientation, eccentricity and size of cells in images of biological tissue. These properties, which can be embodied by an elliptical 'composite cell', are crucial for biomechanical tissue models. To automatically determine these properties is challenging due to the diverse nature of the image data, with tremendous and unpredictable variability in illumination, cell pigmentation, cell shape and cell boundary visibility. One proposed algorithm estimates the composite cell properties directly from the input tissue image, while four others estimate the properties from frequency domain data. The accuracy and stability of the algorithms are quantitatively compared through application to a wide variety of real images. Based on these results, the best algorithm is selected.

This thesis presents a whole-embryo finite element model of neurulation -- the first of its kind. An advanced, multiscale finite element approach is used to capture the mechanical interactions that occur across cellular, tissue and whole-embryo scales. Cell-based simulations are used to construct a system of constitutive equations for embryonic tissue fabric evolution under different scenarios including bulk deformation, cell annealing, mitosis, and Lamellipodia effect. Experimental data are used to determine the parameters in these equations. Techniques for obtaining images of live embryos, serial sections of fixed embryo fabric parameters, and material properties of embryonic tissues are used. Also a spatial-temporal correlation system is introduced to organize and correlate the data and to construct the finite element model. Biological experiments have been conducted to verify the validity of this constitutive model. A full functional finite element analysis package has been written and is used to conduct computational simulations. A simplified contact algorithm is introduced to address the element permeability issue. Computational simulations of different cases have been conducted to investigate possible causes of neural tube defects. Defect cases including neural plate defect, non-neural epidermis defect, apical constriction defect, and convergent extension defect are compared with the case of normal embryonic development. Corresponding biological experiments are included to support these defect cases. A case with biomechanical feedbacks on non-neural epidermis is also discussed in detail with biological experiments and computational simulations. Its comparison with the normal case indicates that the introduction of biomechanical feedbacks can yield more realistic simulation results.

Five image processing algorithms are proposed to measure the average orientation, eccentricity and size of cells in images of biological tissue. These properties, which can be embodied by an elliptical 'composite cell', are crucial for biomechanical tissue models. To automatically determine these properties is challenging due to the diverse nature of the image data, with tremendous and unpredictable variability in illumination, cell pigmentation, cell shape and cell boundary visibility. One proposed algorithm estimates the composite cell properties directly from the input tissue image, while four others estimate the properties from frequency domain data. The accuracy and stability of the algorithms are quantitatively compared through application to a wide variety of real images. Based on these results, the best algorithm is selected.

This thesis presents a whole-embryo finite element model of neurulation -- the first of its kind. An advanced, multiscale finite element approach is used to capture the mechanical interactions that occur across cellular, tissue and whole-embryo scales. Cell-based simulations are used to construct a system of constitutive equations for embryonic tissue fabric evolution under different scenarios including bulk deformation, cell annealing, mitosis, and Lamellipodia effect. Experimental data are used to determine the parameters in these equations. Techniques for obtaining images of live embryos, serial sections of fixed embryo fabric parameters, and material properties of embryonic tissues are used. Also a spatial-temporal correlation system is introduced to organize and correlate the data and to construct the finite element model. Biological experiments have been conducted to verify the validity of this constitutive model. A full functional finite element analysis package has been written and is used to conduct computational simulations. A simplified contact algorithm is introduced to address the element permeability issue. Computational simulations of different cases have been conducted to investigate possible causes of neural tube defects. Defect cases including neural plate defect, non-neural epidermis defect, apical constriction defect, and convergent extension defect are compared with the case of normal embryonic development. Corresponding biological experiments are included to support these defect cases. A case with biomechanical feedbacks on non-neural epidermis is also discussed in detail with biological experiments and computational simulations. Its comparison with the normal case indicates that the introduction of biomechanical feedbacks can yield more realistic simulation results.