In earlier research in our lab, the manipulation of microtubules on gold patterned silicon wafers was achieved by E-beam lithography, Poly (ethylene glycol) self assembled monolayers (PEG-SAMs) and electrophoresis. To develop a technique for delicate single microtubule manipulation, further studies need to be done on PEG-SAMs and electrophoresis. As a foundation of this goal, we examined the electric field in an aqueous solution between two planar electrodes and the compatibility of the antifouling property of PEG-SAMs with the electric field. For this purpose, the distribution of microbeads was analyzed using a Boltzmann distribution. The amount of adsorbed microtubules on a PEG-SAM was examined to test the compatibility of the antifouling property of a PEG-SAM with concomitant exposure to electric field. It is shown that the product of the electric field and the effective charge of the microbead does not have a linear relation with the applied electric potential but an exponentially increasing function with respect to the potential. The antifouling property of the PEG-SAM was not retained after an exposure to the electric field.

This work describes the results of experimental study of the flow of soft objects (microgels) through microchannels. This work was carried with the intention of building a fundamental biophysical model for the flow of neutrophil cells in microcirculatory system. In Chapter 1 we give a summary of the literature describing the flow of cells and “model cells” in microchannels. Paramount to this we developed methods to modify microchannels fabricated in poly(dimethyl siloxane) (PDMS). Originally, these microchannels could not be used to mimic biological microenvironments because they are hydrophobic and have rectangular cross-sections. We designed a method to create durable protein coatings in PDMS microchannels, as outlined in Chapter 3. Surface modification of the channels was accomplished by a two-step approach which included (i) the site-specific photografting of a layer of poly(acrylamide) (PAAm) to the PDMS surface and (ii) the bioconjugation of PAAm with the desired protein. This method is compatible with different channel geometries and it exhibits excellent longevity under shear stresses up to 1 dyn/cm. The modification was proven to be successful for various proteins of various molecular weights and does not affect protein activity. The microchannels were further modified by modifying the cross-sections in order to replicate cardiovascular flow conditions. In our work, we transformed the rectangular cross-sections into circular corss-sections. Microchannels were modified by polymerizing a liquid silicone oligomer around a gas stream coaxially introduced into the channel, as outlined in Chapter 3. We demonstrated the ability to control the diameter of circular cross-sections of microchannels. The flow behaviour of microgels in microchannels was studied in a series of experiments aimed at studying microgel flow (i) under electrostatic interactions (Chapter 4), (ii) binding of proteins attached to the microgel and the microchannel (Chapter 5) and (iii) under the conditions of varying channel geometry (Chapter 6). This work overall present’s new methods to study the flow of soft objects such as cells, in the confined geometries of microchannels. Using these methods, variables can be independently probed and analyzed.

Les cellules sont capables de détecter les distributions spatiales de protéines et ainsi de migrer ou s’étendre dans la direction appropriée. Une compréhension de la réponse cellulaire aux modifications de ces distributions spatiales de protéines est essentielle pour l’avancement des connaissances dans plusieurs domaines de recherches tels que le développement, l’immunologie ou l’oncologie. Un exemple particulièrement complexe est le guidage d’axones se déroulant pendant le développement du système nerveux. Ce dernier nécessite la présence de plusieurs distributions de molécules de guidages étant attractives ou répulsives pour connecter correctement ce réseau complexe qu’est le système nerveux. Puisque plusieurs indices de guidage collaborent, il est particulièrement difficile d’identifier la contribution individuelle ou la voie de signalisation qui est déclenchée in vivo, il est donc nécessaire d’utiliser des méthodes pour reproduire ces distributions de protéines in vitro. Plusieurs méthodes existent pour produire des gradients de protéines solubles ou liées aux substrats. Quelques méthodes pour produire des gradients solubles sont déjà couramment utilisées dans plusieurs laboratoires, mais elles limitent l’étude aux distributions de protéines qui sont normalement sécrétées in vivo. Les méthodes permettant de produire des distributions liées au substrat sont particulièrement complexes, ce qui restreint leur utilisation à quelques laboratoires. Premièrement, nous présentons une méthode simple qui exploite le photoblanchiment de molécules fluorescentes pour créer des motifs de protéines liées au substrat : Laser-assisted protein adsorption by photobleaching (LAPAP). Cette méthode permet de produire des motifs de protéines complexes d’une résolution micrométrique et d’une grande portée dynamique. Une caractérisation de la technique a été faite et en tant que preuve de fonctionnalité, des axones de neurones du ganglion spinal ont été guidés sur des gradients d’un peptide provenant de la laminine. Deuxièmement, LAPAP a été amélioré de manière à pouvoir fabriquer des motifs avec plusieurs composantes grâce à l’utilisation de lasers à différentes longueurs d’onde et d’anticorps conjugués à des fluorophores correspondants à ces longueurs d’onde. De plus, pour accélérer et simplifier le processus de fabrication, nous avons développé LAPAP à illumination à champ large qui utilise un modulateur spatial de lumière, une diode électroluminescente et un microscope standard pour imprimer directement un motif de protéines. Cette méthode est particulièrement simple comparativement à la version originale de LAPAP puisqu’elle n’implique pas le contrôle de la puissance laser et de platines motorisées, mais seulement d’envoyer l’image du motif désiré au modulateur spatial. Finalement, nous avons utilisé LAPAP pour démontrer que notre technique peut être utilisée dans des analyses de haut contenu pour quantifier les changements morphologiques résultant de la croissance neuronale sur des gradients de protéines de guidage. Nous avons produit des milliers de gradients de laminin-1 ayant différentes pentes et analysé les variations au niveau du guidage de neurites provenant d’une lignée cellulaire neuronale (RGC-5). Un algorithme pour analyser les images des cellules sur les gradients a été développé pour détecter chaque cellule et quantifier la position du centroïde du soma ainsi que les angles d’initiation, final et de braquage de chaque neurite. Ces données ont démontré que les gradients de laminine influencent l’angle d’initiation des neurites des RGC-5, mais n’influencent pas leur braquage. Nous croyons que les résultats présentés dans cette thèse faciliteront l’utilisation de motifs de protéines liées au substrat dans les laboratoires des sciences de la vie, puisque LAPAP peut être effectué à l’aide d’un microscope confocal ou d’un microscope standard légèrement modifié. Cela pourrait contribuer à l’augmentation du nombre de laboratoires travaillant sur le guidage avec des gradients liés au substrat afin d’atteindre la masse critique nécessaire à des percées majeures en neuroscience. / Cells are able to sense spatial distribution of proteins and accordingly migrate or extend in the appropriate direction. Understanding cellular responses to modifications in molecular spatial distributions is essential for advances in several fields such as development, immunology and oncology. A particularly complex example is axonal guidance that occurs during the development of the nervous system, which relies on distributions of attractive and repulsive guidance molecules to correctly wire this intricate network. Since several guidance cues collaborate to development of the nervous system, it is particularly difficult to assess the individual contribution of each cue and the signaling cascade each trigger in vivo; therefore methods to reproduce those distributions individually in vitro are necessary to study in detail the effect of each guidance cue. Several methods exist to produce graded distributions of protein that are either soluble or substrate-bound. A few methods making solution gradients are already widely used in several laboratories to perform experiments with the guidance cues that are normally diffusing in vivo. However, current methods allowing the fabrication of substrate-bound gradients are quite complex, which restrict their use to a few laboratories. First, we present a straightforward method exploiting photobleaching of a fluorescently tagged molecule using a visible laser to generating substrate-bound protein patterns: Laser-assisted protein adsorption by photobleaching (LAPAP). This method allows producing complex patterns of protein with micron spatial resolution and high dynamic range. An extensive characterization of the technique was performed and as proof of functionality, axons from dorsal root ganglions cells were guided on laminin peptide gradients. Secondly, LAPAP was improved in order to produce multicomponent patterns by using lasers at different wavelengths and antibodies conjugated to fluorophores corresponding to these wavelengths. Moreover, to speed-up the fabrication process and simplify the device, we developed widefield illumination LAPAP which uses a spatial light modulator, a light emitting diode and a standard microscope to directly print patterns. This patterning method is relatively simple compared to the original LAPAP setup, since it does not involve controlling the laser power or a motorized stage, but only sends an image of the desired pattern to a spatial light modulator. Finally, we used LAPAP to show how it could be used in automated high-content screening assays to quantify the morphological changes resulting from axon growth on gradients of guidance proteins. We produced thousands of laminin-1 gradients of different slopes and analyzed the variations in neurite guidance of neuron-like cells (RGC-5). An image analysis algorithm was developed to process bright field microscopy images, detecting each cell and quantifying the soma centroid and the initiation, terminal and turning angles of the maximal neurite. This data showed that laminin gradients influence the initiation angle of neurite extension of RGC-5, but does not contribute to its turning. We believe that the results presented in this thesis will facilitate the use of substrate- bound protein patterning in typical life science laboratories, since a confocal microscope or a slightly modified standard microscope is the only specialized equipment needed to fabricate patterns by LAPAP. This could increase the number of laboratories working with substrate-bound protein patterns in order to reach the critical mass necessary for major advances in neuroscience.

Bio-microdevices are miniaturised devices based on biologically derived components (e.g., DNA, proteins, and cells) combined or integrated with microfabricated substrates. These devices are of interest for numerous applications, ranging from drug discovery, to environmental monitoring, to tissue engineering. Before a bio-microdevice can be fully developed, specific fabrication issues need to be addressed. One of the most important is the spatial immobilization of selected biomolecules in specific micro-areas of the device. Among the biomolecules of interest, the controlled immobilization of proteins to surfaces is particularly challenging due to the complexity of these macromolecules and their tendency to lose bioactivity during the immobilization step. The present Thesis reports on three novel micropatterning techniques for the spatial immobilization of proteins with bioactivity retention and improved read-out of the resulting micropatterns. The technologies developed are based on three different micropatterning approaches, namely 1) direct-writing UV laser microablation (proLAB), 2) a novel microcontact printing method (�CPTA) and 3) a replica molding method combined with bead selfassembly (BeadMicroArray). The first two technologies, proLAB and �CPTA, are an implementation of existing techniques (laser ablation and �CP, respectively), whereas the third, i.e., the BeadMicroArray, is a totally new technique and type of patterning platform. 'ProLAB' is a technology that uses a micro-dissection tool equipped with a UV laser (the LaserScissors�) for ablating a substrate made of a layer of ablatable material, gold, deposited over a thin polymer layer. The latter layer is transparent to the laser but favours protein adsorption. In the present work microchannels were chosen as the structure of interest with the aim of arranging them in 'bar-codes', so to create an 'information-addressable' microarray. This platform was fabricated and its application to specific antigen binding demonstrated. The second technique that was developed is a microstamping method which exploits the instability of a high-aspect ratio rubber stamp fabricated via soft-lithography. The technique is denominated microcontact printing trapping air (�CPTA) since the collapsing of a rubber stamp made of an array of micro-pillars over a plane glass surface resulted in the formation of a large air gap around the entire array. The method can be successfully employed for printing micro-arrays of proteins, maintaining biological activity. The technique was compared with robotic spotting and found that microarrays obtained with the �CPTA method were more homogeneous and had a higher signal-tonoise ratio. The third technique developed, the BeadMicroArray, introduces a totally new platform for the spatial addressable immobilization of proteins. It combines replica molding with microbead self-assembling, resulting in a platform where diagnostic beads are entrapped at the tip of micropillars arranged in a microarray format. The fabrication of the BeadMicroArray involves depositing functional microbeads in an array of V-shaped wells using spin coating. The deposition is totally random, and conditions were optimised to fill about half the array during spin coating. After replica molding, the resulting polymer mold contains pyramid-shaped posts with beads entrapped at the very tip of the post. Thanks to the fabrication mode involved, every BeadMicroArray fabricated contains a unique geometric code, therefore assigning a specific code to each microarray. In the present work it was demonstrated that the functionality of the beads after replica molding remains intact, and that proteins can be selectively immobilized on the beads, for instance via biorecognition. The platform showed a remarkable level of selectively which, together with an efficient blocking towards protein non-specific adsorption, lead to a read-out characterized by a very good signal-to-noise. Also, after recognition, a code was clearly visible, therefore showing the encoding capacity of this unique microarray.

Many physical systems break symmetry in their evolution. Biophysical systems, such as cells, developing organisms, and even entire populations, are no exception. Developing neurons represent a striking example of a biophysical system that breaks symmetry: neurons cultured in vitro begin as cell bodies with several tendrils (“neurites”) growing outward. A few days later, these same neurons invariably have the same new morphology: exactly one of the neurites (the “axon”) has grown hundreds of microns in length, while the others (the “dendrites”) are much shorter and are more branched. Previous work has shown that any of the neurites can become the axon, and so neurons must break symmetry during their development. The mechanisms underlying neuronal symmetry breaking and axon specification have recently attracted attention, with multiple groups proposing biophysical models to explain the phenomena. In this thesis, we perform the first analytical comparisons of these models by conducting multiple phenotypic and morphological studies of neurite growth in developing neurons. Studying neurite dynamics is technically challenging because neurites have unpredictable morphologies. In Chapter 4, we study neurite competition and neuronal symmetry breaking in hundreds of neurons by optically patterning micron-wide stripes to which the neurons adhere, and on which they grow exactly two neurites. We then use our measurements to test the accuracy of the models in the simple case when a neuron has exactly two neurites. In Chapter 5, we no longer constrain neuronal morphology. One characteristic of symmetry breaking systems is how the system’s complexity affects the symmetry breaking. We find that a majority of the models predict that neurons with more neurites break symmetry much slower than neurons with fewer neurites. Experimentally, we find that neurons with different neurite counts break symmetry at the same rate, consistent with previous observations. We then determine why the models disagree in their predictions, and rectify the models using our own experimental data. In particular, we find that neurons with higher neurite counts have higher concentrations of key proteins involved in symmetry breaking, so that neurons, regardless of neurite count, can break symmetry at the same rate. / Physics

Les cellules sont capables de détecter les distributions spatiales de protéines et ainsi de migrer ou s’étendre dans la direction appropriée. Une compréhension de la réponse cellulaire aux modifications de ces distributions spatiales de protéines est essentielle pour l’avancement des connaissances dans plusieurs domaines de recherches tels que le développement, l’immunologie ou l’oncologie. Un exemple particulièrement complexe est le guidage d’axones se déroulant pendant le développement du système nerveux. Ce dernier nécessite la présence de plusieurs distributions de molécules de guidages étant attractives ou répulsives pour connecter correctement ce réseau complexe qu’est le système nerveux. Puisque plusieurs indices de guidage collaborent, il est particulièrement difficile d’identifier la contribution individuelle ou la voie de signalisation qui est déclenchée in vivo, il est donc nécessaire d’utiliser des méthodes pour reproduire ces distributions de protéines in vitro. Plusieurs méthodes existent pour produire des gradients de protéines solubles ou liées aux substrats. Quelques méthodes pour produire des gradients solubles sont déjà couramment utilisées dans plusieurs laboratoires, mais elles limitent l’étude aux distributions de protéines qui sont normalement sécrétées in vivo. Les méthodes permettant de produire des distributions liées au substrat sont particulièrement complexes, ce qui restreint leur utilisation à quelques laboratoires. Premièrement, nous présentons une méthode simple qui exploite le photoblanchiment de molécules fluorescentes pour créer des motifs de protéines liées au substrat : Laser-assisted protein adsorption by photobleaching (LAPAP). Cette méthode permet de produire des motifs de protéines complexes d’une résolution micrométrique et d’une grande portée dynamique. Une caractérisation de la technique a été faite et en tant que preuve de fonctionnalité, des axones de neurones du ganglion spinal ont été guidés sur des gradients d’un peptide provenant de la laminine. Deuxièmement, LAPAP a été amélioré de manière à pouvoir fabriquer des motifs avec plusieurs composantes grâce à l’utilisation de lasers à différentes longueurs d’onde et d’anticorps conjugués à des fluorophores correspondants à ces longueurs d’onde. De plus, pour accélérer et simplifier le processus de fabrication, nous avons développé LAPAP à illumination à champ large qui utilise un modulateur spatial de lumière, une diode électroluminescente et un microscope standard pour imprimer directement un motif de protéines. Cette méthode est particulièrement simple comparativement à la version originale de LAPAP puisqu’elle n’implique pas le contrôle de la puissance laser et de platines motorisées, mais seulement d’envoyer l’image du motif désiré au modulateur spatial. Finalement, nous avons utilisé LAPAP pour démontrer que notre technique peut être utilisée dans des analyses de haut contenu pour quantifier les changements morphologiques résultant de la croissance neuronale sur des gradients de protéines de guidage. Nous avons produit des milliers de gradients de laminin-1 ayant différentes pentes et analysé les variations au niveau du guidage de neurites provenant d’une lignée cellulaire neuronale (RGC-5). Un algorithme pour analyser les images des cellules sur les gradients a été développé pour détecter chaque cellule et quantifier la position du centroïde du soma ainsi que les angles d’initiation, final et de braquage de chaque neurite. Ces données ont démontré que les gradients de laminine influencent l’angle d’initiation des neurites des RGC-5, mais n’influencent pas leur braquage. Nous croyons que les résultats présentés dans cette thèse faciliteront l’utilisation de motifs de protéines liées au substrat dans les laboratoires des sciences de la vie, puisque LAPAP peut être effectué à l’aide d’un microscope confocal ou d’un microscope standard légèrement modifié. Cela pourrait contribuer à l’augmentation du nombre de laboratoires travaillant sur le guidage avec des gradients liés au substrat afin d’atteindre la masse critique nécessaire à des percées majeures en neuroscience. / Cells are able to sense spatial distribution of proteins and accordingly migrate or extend in the appropriate direction. Understanding cellular responses to modifications in molecular spatial distributions is essential for advances in several fields such as development, immunology and oncology. A particularly complex example is axonal guidance that occurs during the development of the nervous system, which relies on distributions of attractive and repulsive guidance molecules to correctly wire this intricate network. Since several guidance cues collaborate to development of the nervous system, it is particularly difficult to assess the individual contribution of each cue and the signaling cascade each trigger in vivo; therefore methods to reproduce those distributions individually in vitro are necessary to study in detail the effect of each guidance cue. Several methods exist to produce graded distributions of protein that are either soluble or substrate-bound. A few methods making solution gradients are already widely used in several laboratories to perform experiments with the guidance cues that are normally diffusing in vivo. However, current methods allowing the fabrication of substrate-bound gradients are quite complex, which restrict their use to a few laboratories. First, we present a straightforward method exploiting photobleaching of a fluorescently tagged molecule using a visible laser to generating substrate-bound protein patterns: Laser-assisted protein adsorption by photobleaching (LAPAP). This method allows producing complex patterns of protein with micron spatial resolution and high dynamic range. An extensive characterization of the technique was performed and as proof of functionality, axons from dorsal root ganglions cells were guided on laminin peptide gradients. Secondly, LAPAP was improved in order to produce multicomponent patterns by using lasers at different wavelengths and antibodies conjugated to fluorophores corresponding to these wavelengths. Moreover, to speed-up the fabrication process and simplify the device, we developed widefield illumination LAPAP which uses a spatial light modulator, a light emitting diode and a standard microscope to directly print patterns. This patterning method is relatively simple compared to the original LAPAP setup, since it does not involve controlling the laser power or a motorized stage, but only sends an image of the desired pattern to a spatial light modulator. Finally, we used LAPAP to show how it could be used in automated high-content screening assays to quantify the morphological changes resulting from axon growth on gradients of guidance proteins. We produced thousands of laminin-1 gradients of different slopes and analyzed the variations in neurite guidance of neuron-like cells (RGC-5). An image analysis algorithm was developed to process bright field microscopy images, detecting each cell and quantifying the soma centroid and the initiation, terminal and turning angles of the maximal neurite. This data showed that laminin gradients influence the initiation angle of neurite extension of RGC-5, but does not contribute to its turning. We believe that the results presented in this thesis will facilitate the use of substrate- bound protein patterning in typical life science laboratories, since a confocal microscope or a slightly modified standard microscope is the only specialized equipment needed to fabricate patterns by LAPAP. This could increase the number of laboratories working with substrate-bound protein patterns in order to reach the critical mass necessary for major advances in neuroscience.

This work describes the results of experimental study of the flow of soft objects (microgels) through microchannels. This work was carried with the intention of building a fundamental biophysical model for the flow of neutrophil cells in microcirculatory system. In Chapter 1 we give a summary of the literature describing the flow of cells and “model cells” in microchannels. Paramount to this we developed methods to modify microchannels fabricated in poly(dimethyl siloxane) (PDMS). Originally, these microchannels could not be used to mimic biological microenvironments because they are hydrophobic and have rectangular cross-sections. We designed a method to create durable protein coatings in PDMS microchannels, as outlined in Chapter 3. Surface modification of the channels was accomplished by a two-step approach which included (i) the site-specific photografting of a layer of poly(acrylamide) (PAAm) to the PDMS surface and (ii) the bioconjugation of PAAm with the desired protein. This method is compatible with different channel geometries and it exhibits excellent longevity under shear stresses up to 1 dyn/cm. The modification was proven to be successful for various proteins of various molecular weights and does not affect protein activity. The microchannels were further modified by modifying the cross-sections in order to replicate cardiovascular flow conditions. In our work, we transformed the rectangular cross-sections into circular corss-sections. Microchannels were modified by polymerizing a liquid silicone oligomer around a gas stream coaxially introduced into the channel, as outlined in Chapter 3. We demonstrated the ability to control the diameter of circular cross-sections of microchannels. The flow behaviour of microgels in microchannels was studied in a series of experiments aimed at studying microgel flow (i) under electrostatic interactions (Chapter 4), (ii) binding of proteins attached to the microgel and the microchannel (Chapter 5) and (iii) under the conditions of varying channel geometry (Chapter 6). This work overall present’s new methods to study the flow of soft objects such as cells, in the confined geometries of microchannels. Using these methods, variables can be independently probed and analyzed.

Exploiting the functional diversity of proteins for fundamental research and biotechnological applications requires their functional organization into micro- and nanostructures while preserving their functional integrity to the highest possible level. My PhD research aimed to establish generic techniques based on photolithography which could be used to control the spatial as well as temporal organization of recombinantly expressed proteins on surfaces. My thesis describes in detail four strategies that I developed for achieving this goal. In the first approach a photo-induced Fenton reaction was used to selectively destroy tris(nitrilotriacetic acid) (tris-NTA) moieties on a surface. UV-irradiation through a photomask allowed localized photo-destruction and targeting of His-tagged proteins to non-irradiated regions. Photo-destruction could also be achieved by scanning selected regions with the UV laser of a confocal laser scanning microscope (CLSM) thus allowing flexible creation and modification of protein patterns. The second strategy was based on the photosensitive nitroveratryloxycarbonyl (NVOC) protection group, which was used to cage amine groups on a surface. Sequential uncaging by UV-irradiation through a photomask followed by reactions with biotin and coenzyme A was used to pattern streptavidin and ybbR-tagged proteins into microstructures. In the third approach a photo-fragmentable Histidine peptide was used to block tris-NTA surfaces against binding of His-tagged proteins. UV-irradiation through a photomask or by using a UV laser in a CLSM cleaved the peptide into short fragments which quickly dissociated from the surface due to loss in multivalency. His-tagged proteins could be efficiently targeted into irradiated regions even from a complex cell lysate. Sequential uncaging and immobilization allowed the construction of multiplexed protein patterns with a high degree of temporal control. The fourth strategy used combined peptide tags comprising of a His-tag as well as a Halo- or ybbR-tag to achieve rapid covalent immobilization of recombinant fusion proteins on surfaces functionalized with specific ligands. In combination with a photo-fragmentable histidine peptide as described above, stable spatio-temporal organization of proteins carrying these combined tags was possible. The techniques developed in this thesis enabled the photolithographical micropatterning of recombinant proteins carrying specific peptide or protein tags on surfaces in a functional manner. Owing to the generic nature of immobilization strategies, coupled with the ease of patterning, highly versatile applications of these methods both in fundamental research as well as bio-technological and analytical applications can be envisioned.

We developed new methods for controlling and studying microtubules (MTs) outside the complex workings of the living cell. Several surface treatments for preventing MT fouling on surfaces were analyzed and, for the first time, a self-assembled monolayer (SAM) was developed which prevented MT adsorption in the absence of passivating proteins. The morphology and thickness of the SAM was measured to determine the mechanism of formation and origin of the MT-resistant behavior. The SAM was integrated into electron beam lithography for patterning and manipulating MTs using electrophoresis. Reversible MT adsorption and patterning and alignment of single MTs were achieved. We characterized the mechanism for the MT migration under electric field with a focus on the electrodynamics of the flow cell and the forces acting on the MT, along with the time dependence of the process.

A major challenge in nanotechnology is the spatially controlled transport of cargo on the nanometer scale. The use of a nanoscale transport system based on molecular motors and filaments of the cytoskeleton proved as a promising approach to this problem. Therefore, the objective of this work was to pattern planar surfaces with motor proteins in a way that allows controlled and guided movement of microtubule-shuttles. The first part of the work was in particular focused on generating nanometer–sized tracks of motor proteins on unstructured surfaces. Specifically, microtubules themselves were used as biological templates for the stamping and alignment of motor proteins. Compared to other soft lithography techniques like microcontact printing this approach circumvented protein denaturation due to drying and conformational changes caused by mechanical stress. Given the large persistence length of microtubules their encounters with the boundaries of the nanotracks are limited to shallow approach angles. This way, the generated structures proved very efficient for the guiding of microtubules without topographical barriers. Topography-free guiding, as demonstrated in this work, is expected to significantly ease the design and fabrication of microtubule-transport systems and opens up the possibility to transport cargo of unlimited size, i.e. without any constraints by the dimensions of topographic guiding channels. Moreover, the biotemplated patterning is a promising tool for in vitro studies on the individual and cooperative action of motor proteins. In particular it might be helpful for the reconstitution of complex subcellular machineries in synthetic environments. As an example, microtubule-microtubule sliding by the biomolecular motor ncd has been shown to induce directional sliding between antiparallel microtubules and static cross-linking between parallel ones. The second part of the work explored an in-situ patterning technique for motor proteins to enable user-defined pattern designs, and investigated the achievable resolution. Photothermal patterning, based on localized light-to-heat conversion combined with a thermoresponsive polymer layer, was presented as a novel method. Specifically, the conformation of poly(N-isopropylacrylamide) (PNIPAM) molecules in aqueous solution was switched between the swollen state at T < 30°C (protein-repelling conformation) to the collapsed state at T > 33°C (protein-binding conformation) by optical signals of visible light to generate heat in a highly-localized manner. Upon heating of a light-absorbing layer on the substrate, the surface-grafted PNIPAM molecules collapsed locally and allowed motor proteins in solution to bind in the illuminated areas. To confirm the successful patterning of kinesin-1 molecules and their functionality microtubule-based gliding motility assays were performed. It was shown that the microtubules bind to the patterned kinesin-1 molecules and are transported exclusively in the patterned areas. While the achieved pattern sizes were currently in the range of ten micrometers, finite element modeling (implemented in COMSOL) showed that increased optical intensities possibly combined with cooling of the sample allow to significantly scale down the pattern dimensions. The produced patterns can be reversibly activated and deactivated at high and low temperature, respectively. Moreover, sequential patterning of multiple kinds of proteins on the same surface will be possible in a similar way without the need for specific linker molecules or elaborate surface preparation. Another advantage of the method is the use of visible light, which is versatile as any wavelength can be applied. In addition visible light is in comparison to other UV-based photopatterning techniques non-damaging to proteins. / Der räumlich kontrollierte Transport von nanoskaligen Objekten ist eine große Herausforderung auf dem Gebiet der Nanotechnologie. Ein auf molekularen Motoren und Filamenten des Zellskeletts basierendes Nanotransportsystem hat sich dabei als ein viel versprechender Ansatz erwiesen. Das Ziel der vorgelegten Arbeit war es daher, ebene Oberflächen so mit Motorproteinen zu strukturieren, dass eine kontrollierte und geführte Bewegung von Mikrotubuli-Transportern ermöglicht wird. Der erste Teil der Arbeit war insbesondere darauf fokussiert, Motorprotein-Spuren im Nanometerbereich zu erzeugen. Im zweiten Teil der Arbeit wurde eine Strukturierungsmethode zur Realisierung von benutzerdefinierten Musterdesigns untersucht und die erreichbare Auflösung bestimmt. Für die Nanometerstrukturierung von Oberflächen mit funktionalen Motorproteinen wurde ein neuer Ansatz demonstriert. Mit der Anwendung von Biotemplaten, wie hier der Mikrotubuli, konnte ein hoch-lokalisiertes und orientiertes Anbinden von Proteinen an Oberflächen sowie gleichzeitig geringer Proteindenaturierung erreicht werden. Durch spezifisches Stempeln beziehungsweise Binden von Motoren wurden Muster aus funktionellen Proteinen mit hoher Oberflächendichte hergestellt. Die erzeugten Motor-Spuren haben gezeigt, dass Nanometerstrukturierungen möglich sind und ohne topographische Barrieren zu zuverlässiger Führung von Mikrotubuli führen können. Bisher konnte die nicht-topographische Strukturierung von Oberflächen mit Kinesin-1-Motoren nur im Mikrometerbereich demonstriert werden. Wegen der hohen Steifigkeit der Mikrotubuli war die thermische Energie des Systems in diesen Fällen nicht ausreichend, um die führende Spitze der Mikrotubuli zurück auf das Gebiet mit den strukturierten Motoren zu biegen. Dieses Problem wird durch die kleine Breite der hier demonstrierten Motor-Nanospuren verhindert, da das Auftreffen der Mikrotubuli mit den Grenzlinien auf extrem flache Winkel begrenzt ist. Interessanterweise haben sich Spuren des nicht-prozessiven Motors Kinesin-14 für das Führen und den Transport im Nanometerbereich als noch zuverlässiger herausgestellt als Kinesin-1-Spuren. Es ist zu erwarten, dass nicht-topographisches Führen, wie es in dieser Arbeit gezeigt wurde, das Design und die Herstellung von Mikrotubuli-Transportsystemen deutlich vereinfacht und die Möglichkeit eröffnet, Cargo mit unlimitierter Größe, d.h. ohne Einschränkungen durch die Abmessungen der topographischen Führungskanäle, zu transportieren. Zusätzlich ist die biotemplierte Strukturierung ein viel versprechendes Werkzeug um das individuelle und das kooperative Arbeiten von Motorproteinen in vitro untersuchen und komplexe subzelluläre Maschinerien in synthetischer Umgebung rekonstituieren zu können. Dies wurde am Beispiel des gerichteten Gleitens des biomolekularen Motors Kinesin-14 gezeigt, der ein gerichtetes Gleiten zwischen antiparallelen Mikrotubuli und statisches Vernetzen zwischen parallelen Mikrotubuli hervorruft. Mit dem Ansatz des biotemplierten Strukturierens ist es jedoch nicht einfach möglich, benutzerdefinierte Spuren zu erzeugen. Daher wurde die photothermische Proteinstrukturierung als eine neue Methode für die frei programmierbare, hochauflösende und schnelle Herstellung von strukturierten Proteinoberflächen eingeführt. Auf diese Weise wurden Kinesin-1-Muster durch licht-induziertes Heizen einer licht-absorbierenden Substratschicht erzeugt. Die thermisch schaltbaren poly(N-isopropylacrylamid) (PNIPAM) Moleküle auf der Oberfläche kollabierten lokal und erlaubten es den Motorproteinen, in den beleuchteten Gebieten aus der Lösung an die Oberfläche zu binden. Die Bewegung gleitender Mikrotubuli bestätigte anschließend die erfolgreiche Strukturierung der Kinesin-1-Motoren und deren Funktionalität, da die Mikrotubuli an die strukturierten Motoren banden und ausschließlich in den strukturierten Gebieten transportiert wurden. Neben der Proteinstrukturierung wurde die lokalisierte Licht-zu-Wärme-Umwandlung kombiniert mit einer thermisch schaltbaren Polymerschicht auch für die lokale Aktivierung von Kinesin-1-Motoren auf der Oberfläche genutzt. Ein Vorteil der photothermischen Proteinstrukturierung ist die Möglichkeit, sichtbares Licht zu verwenden, da jede beliebige Wellenlänge angewendet werden kann und sichtbares Licht, im Vergleich zu anderen UV-basierten Photostrukturierungsmethoden, Proteine nicht schädigt. Modellierungen mit Hilfe der Finite-Elemente-Methode (implementiert in der Software COMSOL) haben gezeigt, dass die Lichtintensität und die Oberflächentemperatur speziell eingestellt werden müssen, um definierte Strukturgrößen zu erzielen. Während die derzeitig erzeugten Muster Größen im Bereich von zehn Mikrometern hatten, könnten durch höhere optische Intensitäten kombiniert mit Kühlung der Probe die Größenordnungen signifikant reduziert werden. Die reale experimentelle Auflösung wird jedoch auch von der Schaltcharakteristik des Polymers und der Proteinbindungsdynamik abhängen. Die hergestellten Muster können reversibel bei hohen beziehungsweise niedrigen Temperaturen aktiviert und deaktiviert werden. Zusätzlich können auf die gleiche Weise verschiedene Proteinsorten sequentiell auf einer Oberfläche strukturiert werden, ohne dass spezifische Bindungsmoleküle oder aufwändige Oberflächenpräparationen notwendig wären. Die Möglichkeit, Proteine reversibel an die Oberfläche zu binden, um geschriebene Muster wieder löschen zu können, wäre eine Weiterentwicklung und würde die Anwendungsmöglichkeiten der photothermischen Strukturierungsmethode erweitern. Außerdem würden optisch schaltbare Polymere das direkte Strukturieren von Motoren mit Licht ermöglichen und daher die Methode vereinfachen.