Includes bibliographical references. / This report starts by describing the background research and work that had already been done on the UCT scanning tunneling microscope (STM). This system is being developed in the Department of Electrical Engineering at UCT. It goes on to describe the continuation of the research work that was done for this dissertation on the STM at UCT. The work was originally started by Dr. Tapson for his PhD (1994). and continued by the author for his MTech degree in ) 997 and 1998. The work was temporary discontinued from May 2000 till August 2002 to enable the author to work as a contract engineer at the Institute of Physics in Basel, Switzerland to learn more about the construction of probe microscopes. The new work evolved around the need to implement scanning tunneling potentiometry (STP) capability in the new STM. This capability should give the end-user the capability of looking at the sub-surface structure of any material on a sub-micron scale. The basic STP function must be implemented in two dimensions in the plane of the specimen. The STM tip is then used as a highly localized voltmeter to sense what the potential distribution is at that point on the surface. The potential information that is obtained is then used to plot two images of the potential distribution over the surface in the X and Y directions. The topographic information is obtained in the usual way from the STM scan. This method gives three collocated imagesas the result and a better understanding of the surface structure is obtained in this way. The penetration depth of the potential scan can be varied by adjusting the frequency of the applied AC signal in the X and Y directions. This use of the skin effect should allow the end user to obtain slices of the surface at various penetration levels of the specimen. These slices will give a picture of what happens from the surface up to a certain penetration depth. The interpretation of these images could be very difficult because the skin effect does not stop at a defined penetration depth. Only the 3 dB point is defined, which means that sub surface structures below the 3 dB point will also have an influence on the obtained image. During the course of the research new hardware and scanning software was implemented to enable the error-free acquisition of new data. This entailed splitting the existing XY controller into three separate parts namely a Communications interface, and two STP measurement boards. This was suggested as one of the conclusions of the MTech thesis results. The PC software stayed the same but for a change in the array size, that holds theacquired data. This was again changed after the work experience in Basel and is explained in chapter 6.

Le cerveau de drosophile est constitué entre autres des mushroom body, siège de la mémoire et de l’apprentissage. Cette structure est composée de différents types de neurones, parmi lesquels les neurones /. Ces neurones se présentent sous une forme orthogonale, avec l’axone qui se divise en une branche dorsale : la branche et une branche médiale : la branche . Le but de cette étude est de comprendre les mécanismes et voies de signalisation mis en jeu lors du développement de ces neurones.Chez la drosophile, la protéine APPL (Amyloïd Precursor Protein-Like) est l’homologue de la protéine APP humaine, connue pour son implication dans la maladie d’Alzheimer chez l’homme. Cette pathologie est caractérisée par une dégénérescence neuronale entraînant des défauts cognitifs et mnésiques. Malgré les nombreuses études focalisées sur la fonction pathologique d’APP durant les dernières décennies, peu de choses sont actuellement connues sur les fonctions physiologiques de cette protéine et notamment pendant le développement. C’est dans cette optique que nous avons étudié la fonction d’APPL et son interaction avec différentes protéines lors du développement des mushroom body. La protéine APPL a été identifiée comme étant un co-récepteur de la voie PCP (Planar Cell Polarity), permettant la régulation de la croissance axonale. Lors du développement, APPL permet le recrutement et l’activation de la protéine ABL (Abelson Tyrosine Kinase), qui phosphoryle DSH (dishevelled) et ainsi active la voie de signalisation permettant la croissance axonale.Le premier volet de cette thèse porte sur la régulation de l’activité ABL lors du développement des neurones /. S’il est établi qu’APPL permet une régulation positive de l’activité kinase d’ABL, je montre ici que la protéine HTT (huntingtine) permet de réguler négativement cette activité. Cette protéine HTT est impliquée dans la maladie de Huntington chez l’homme, une autre pathologie neurodégénérative. Ces travaux démontrent qu’HTT régule le niveau de phosphorylation d’ABL et par conséquent son activité. Le deuxième volet de cette thèse porte sur l’interaction d’APPL avec la protéine ARM (armadillo), homologue de la -caténine, lors du développement des neurones /. Je démontre que cette interaction est indépendante du rôle d’APPL dans la voie PCP. Je démontre aussi que cette interaction entre APPL et ARM est dépendante uniquement de la fonction d’ARM dans la dynamique du cytosquelette d’actine.Enfin le troisième volet de cette thèse porte sur la création de nouveaux allèles mutants pour Appl grâce à la technique du CRISPR-CAS9. La production de ces allèles permet d’avancer d’une part un possible rôle du gène voisin vnd (ventral nervous system defective) dans le développement des mushroom body, et d’autre part une interaction génétique entre Appl et vnd. / In the drosophila brain, mushroom bodies are involved in olfactory memory and learning. This structure is composed of different types of / neurons. These neurons form an orthogonal structure, with the branch projecting dorsally and the branch projecting medially. The aim of this study is to understand mechanisms and pathways involved during the development of these neurons.The drosophila APPL protein (Amyloïd Precursor Protein-Like) is the homologue of the human APP, known to be involved in Alzheimer’s disease. This pathology is characterized by neuronal degeneration inducing cognitive and memory defects. In spite of the numerous studies focused on the pathological function of APP during the last decades, few things are known on the physiological functions of this protein and more particularly during the development. This is from this perspective that we studied the APPL function and its interaction with proteins during the mushroom bodies development.The APPL protein was identified as a co-receptor of the PCP pathway (Planar Cell Polarity), involved in the axonal growth regulation. During the development, APPL allows the recruitment and the activation of the ABL protein (Abelson Tyrosine Kinase), which phosphorylates DSH (dishevelled) and so activates the axonal growth pathway.The first part investigates the regulation of ABL activity during the / neuron development. If it’s already established that APPL regulates positively the kinase activity of ABL, I show here that the HTT protein (Huntingtin) allows a negative regulation of ABL activity. In human, HTT is involved in the Huntington’s disease, another neurodegenerative disorder. This thesis work shows that HTT regulates the phosphorylation level of ABL, and therefore its activity.The second part investigates the interaction between APPL and ARM (armadillo), the homologue of the human -catenin, during the development of the / neurons. I show that this interaction is independent of the APPL function in the PCP pathway. Moreover, this interaction between APPL and ARM involves the actin cytoskeleton dynamic function of ARM, and not its Wnt pathway function.The third and last part presents new mutant alleles of APPL obtained with the CRISPR-CAS9 technique. The creation and analysis of these new alleles lead us to propose that vnd (ventral nervous system defective), neighbor gene of Appl, is also involved in / neurons development, and can interact genetically with Appl.

Drosophila es un modelo versátil para entender las bases genéticas de la señalización celular. En particular, el ojo compuesto de Drosophila proporciona un tejido ideal para estudiar los mecanismos de integración de señales que dirigen la formación de la red de neuronas fotorreceptoras. La vía de Ras/MAPK está involucrada en la determinación de todos los fotorreceptores, pero es particularmente necesaria para la especificación del fotorreceptor R7, ya que su ausencia determina la transformación del R7 en una célula no neural. De manera que durante el desarrollo del ojo, Ras media la decisión entre la diferenciación neural y no neural. Sin embargo, se desconoce el perfil de genes activados por Ras. Con el objetivo de describir este perfil, hemos realizado microarrays comparando diferentes alelos del receptor tirosina quinasa Sevenless (Sev) que participa en la especificación del R7 mediante la activación de la vía de Ras / MAPK. Uno de los genes que responden a la activación de Sev es el gen amyloid protein precursor-like (Appl). El incremento de la expresión de Appl detectado en los microarrays, también se confirmó por hibridación in situ. En un estudio detallado de localización de la proteína, observamos que aunque está presente en todos los fotorreceptores, hay más proteína Appl en R7 y R8, que en los otros. Para evaluar si Ras es necesario para la activación de Appl, realizamos clones con el alelo dominante negativo del gen Epithermal Growth factor Receptor (DERDN). En estos clones no se detectó Appl, mientras que los clones del alelo constitutivamente activo de Ras; RasV12 dieron lugar a la sobreexpresión de Appl. En cualquier otro tejido, RasV12 no produjo expresión ectópica, indicando que Ras es necesario y suficiente para la expresión de Appl aunque esta regulación es dependiente del dominio del ojo. Para ver si el factor de transcripción de la vía; Pnt, es capaz de unirse a Appl para activar su expresión realizamos clones con el alelo de pérdida de función pntΔ88. Estos clones mostraron que la falta de pnt tiene como resultado la ausencia de Appl. En segundo lugar evaluamos la posible unión directa de Pnt a Appl mediante transgénicos de regiones supuestamente enhancers de Appl unido a un promotor mínimo y al gen reportero lacZ. Ninguna de las construcciones fue capaz de dirigir la expresión de lacZ en ojo, sin embargo dos de ellas fueron capaces de dirigirla en cerebro, sugiriendo que estas dos secuencias actúan como unidades reguladoras de la expresión de Appl. Ya que no produjeron expresión en ojo, decidimos sensibilizar la respuesta de las construcciones a Ras, incrementando la actividad de la vía mediante clones RasV12. En este nuevo contexto, los dos ETS que tenían expresión en cerebro, fueron capaces de dirigir la expresión de lacZ. Además la unión de pnt a Appl se confirmó mediante experimentos de InmunoPrecipitación de la Cromatina (CHIP). Ras en ojo es responsable de la determinación de casi todos los fotorreceptores, sin embargo observamos que mientras que la determinación no está afectada en mutantes Appld, sí lo está el funcionamiento del R7. Este fotorreceptor es el único capaz de ver la luz ultravioleta. Mediante experimentos de comportamiento observamos que las moscas Appld tienen disminuida la capacidad de discernir la luz UV. Esta reducción es debida en parte porque el 2% de los axones de R7 de moscas Appld no llegan a hacer la sinapsis. Para tratar de incrementar los efectos observados y poder describir con mayor claridad la función de Appl en ojo, testamos la habilidad de los mutantes Appld de discernir el UV, en combinación con mutantes heterocigotos de proteínas descritas en el proceso de guía de axones. La perdida de función de Appl combinada con el mutante heterocigoto del gen neurotactina (nrt), produjo un claro deterioro de la capacidad del R7 para discernir UV y de los axones pare llegar a hacer la sinapsis demostrando que Appl es necesario para la correcta función del R7. / In a genome wide expression profile search for genes that characterize the Drosophila R7 photoreceptor specification we found Appl, the ortholog of human APP and a key factor in the pathogenesis of Alzheimer’s disease. We analyzed Appl expression in the eye imaginal disc and found that is highly accumulated in R7 photoreceptor cells. The R7 photoreceptor is responsible for UV light detection. To explore the link between high expression of Appl and R7 function, we have analyzed Appl null mutants and found reduced preference for UV light, likely due to mistargeted R7 axons. Moreover, axon mistargeting and inappropriate light discrimination are enhanced in combination with neurotactin mutants. R7 differentiation is triggered by the inductive interaction between R8 and R7 precursors, which results in a burst of Ras1/MAPK activated by the tyrosine kinase receptor Sevenless. Thus, we have studied whether Ras1/MAPK is responsible for the high Appl expression. Inhibition of Ras1 signaling leads to reduced Appl expression, whereas constitutive activation drives ectopic Appl expression. We show that Appl is directly regulated by the Ras/MAPK pathway through a mechanism mediated by PntP2, an ETS transcription factor that specifically binds ETS sites in the Appl regulatory region. Also, the zebrafish appb expression increased after ectopic fgfr activation in the neural tube of zebrafish embryos, suggesting a conserved regulatory mechanism.

This is an ethnographic case-study of research and development practices taking place in an interdisciplinary project between education and computer sciences. The Ensemble-project, part of the Technology Enhanced Learning programme (2008-12), has studied case-based learning in a number of diverse settings in Higher Education, working to develop semantic technologies for supporting that learning. Focussing on one of the six research settings, the discipline of archaeology, the current study has had three purposes. By opening up to scrutiny the practices of research and development, it has firstly sought to understand how a shared research question is answered in practice when divergent research approaches are brought to bear upon it. Secondly, the study has followed the emergence of a piece of semantic technology through these practices. The third aim has been to assess the advantages and disadvantages of Actor-Network Theory (ANT) in studying unfolding, open-ended processes in real time. Through critical ethnographic participation, multiple ethnographic research methods, and by drawing on ANT as theoretical practice, the study has shown the precarious and unpredictable nature of research and development work, the political nature of research methods and how multiple realities can be produced using them, and the need for technology development to flexibly respond to changing circumstances. We have also seen the mutual adoption and extension of practices by the two strands of the project into each others’ domains, and how interdisciplinary tensions resolved, while they did not disappear, through pragmatic changes within the project. The study contributes to the interdisciplinary fields of Science and Technology Studies (STS) where studies on the ‘soft sciences’, such as education, are few, and a new field of Studies in Social Science and Humanities (SSH) which is emerging alongside and from within the STS. Interdisciplinary endeavours between fields pertaining largely to the natural and the social sciences respectively have not been studied commonly within either field.