F-Aktin spielt eine wichtige Rolle bei der Steuerung der endothelialen Barrierefunktion. In dieser Arbeit verwendeten wir Colchicin, Vinca-Alkaloide (Vinblastin, Vincristin) und Paclitaxel um Mikrotubulussysteme (MT) auszulenken und den Effekt auf die Permeabilität zu untersuchen. Endothelzellen wurden auf Polycarbonatfiltermembranen gepflanzt und einem kontinuierlichen hydrostatischen Druck von 10 cm H2O ausgesetzt. Die Exposition von Endothelzell-Monolayern gegenüber Colchicin und Vinca-Alkaloiden führte innerhalb von 60 100 Minuten zeit- und dosisabhängig zu einem fünf zehnfachen Anstieg der hydraulischen Konduktivität. Dagegen war nach MT-Stabilisation durch Paclitaxel keine Permeabilitätszunahme festzustellen. Doppelimmunfluoreszenz-Mikroskopie zeigte, dass die MT-Depolymerisation durch Colchicin und Vinca Alkaloide zu F-Aktin-Umverteilung, Stressfaserbildung und Zellretraktionen mit ausgeprägter parazellulärer Lücken-Bildung führt. Diese Phänomene wurden durch Kombinationen von Vinblastin und Paclitaxel deutlich abgeschwächt. Die fluorometrische Messung des intrazellulären F-Aktins nach MT-Depolymerisation durch Vinblastin resultierte in einer signifikanten Zunahme der Aktinfilamente. Auf der anderen Seite resultierte F-Aktin Abbau durch Cytochalasin D und Clostridium difficile (TcdB-10463) morphologisch nicht in einer Veränderung von MT-Strukturen. Dabei zeigten in Interzellularbrücken gelegene MT-Filamente Kolokalisation mit F-Aktin Fragmenten. Unsere Ergebnisse demonstrieren, dass MT-Systeme an der Regulation der endothelialen Barriere beteiligt sind. Darüber hinaus verdeutlichen die Resultate eine enge Bindung von MT- und Aktin-Filamenten innerhalb endothelzellulärer Adhäsionskontakte. / The endothelian cytosceleton plays an important role in the regulation of endothelial permeability via cellular actin filaments. We tested the effect of agents known to perturb cellular microtubules on the permeability of endothelial cell monolayers. The agents chosen were colchicine, the vinca alkaloids vinblastine and vincristine and paclitaxel. Cell monolayers were prepared on polycarbonate filter membranes and exposed to a continuous hydrostatic pressure of 10 cm H2O. Colchicine and the vinca alkaloids caused a five to tenfold increase in the hydraulic conductivity of the monolayers within 60 100 min. The effect was dose and time dependent. The microtubule stabilizer paclitaxel caused no increase in permeability. Double-immunofluorescence microscopy showed that microtubule depolymerisation was associated with certain morphological features such as inter-endothelial gaps, cell retraction, f-actin reorganisation and some stressfibre appearance. These phenomena were significantly reduced when vinblastine and paclitaxel were combined. Measurement of intracellular f-actin following microtubule inhibition with vinblastine showed a significant increase in endothelial actin filaments. No changes in microtubule structures were seen when actin filaments were perturbed with cytochalasin D and Clostridium difficile (TcdB-10463). However, in this case the intercellular bridges showed that microtubules were co-localised with fragments of actin filaments from neighbouring cells. Our data demonstrate that microtubules are important for the regulation of endothelial permeability. Moreover, our results support evidens of binding between microtubules and actin filaments within endothelial cell adhesion contacts.

A major challenge in nanotechnology is the controlled transport of cargo on the nanometer scale. A promising approach to this problem is the use of molecular motors of the cellular cytoskeleton. The aim of this work was to develop a method to characterize the behavior of filamentous nanoshuttles – specifically of motor protein-driven microtubules – in three dimensions (3-D). The main requirements to meet were low impact on the nanotransport system, high spatial and temporal resolution, and versatility. Furthermore, this method was intended to be used to address open questions in the field of nanotransport. In particular, it was firstly attempted to characterize cargo transport in a system currently favored by most studies in the field, where nanoshuttles are powered by the microtubule motor best understood so far – the plus-end-directed kinesin-1. Secondly, the goal was to further the understanding of potential counter-players of kinesin-1 in nanotransport applications - the much less well understood microtubule minus-end-directed motor proteins 22S dynein and the kinesin-14 non-claret disjunctional (ncd). A novel method to study the linear forward motion as well as the axial motion of filamentous nanoshuttles, which are driven by motors of the cell cytoskeleton, has been introduced. The method uses fluorescence interference-based 3-D nanometer tracking of quantum dots as optical probes that are attached to the nanoshuttles. While other recently reported 3-D tracking techniques based on dual-focus imaging offer similar sensitivity, the method here can be easily performed on any standard epi-fluorescence microscope, even with arc lamp illumination, and additionally holds the potential to retrieve absolute height values. It is strongly suggested that the ease of use might help to spread this valuable and versatile tool for a variety of applications, including studies of interactions between single molecules or even intramolecular changes. Specifically, 3-D tracking has been used to visualize and analyze the rotation of microtubules around their longitudinal axis when they are propelled on a motor protein-coated surface. This geometry called gliding assay is currently favored for most proof-of-principle studies that investigate the use of biomolecular motors for transport of nanoscale cargo with the goal to assemble and manipulate nanostructures. The suitability of the method has been proven for kinesin-1 gliding assays, where knowledge of properties of both, microtubules and kinesin-1, allowed a very precise prediction of microtubule rotation, which was matching the actual measured values very well. The microtubule rotation in kinesin-1 gliding assays has turned out to be robust against the attachment of small cargo in the shape of quantum dots (diameter ∼20 nm), but also against the reduction of electrostatic interactions between microtubules and kinesin-1 by cleavage of the tubulin E-hook. The situation was dramatically different when large cargo (beads with diameter of ∼3 µm) was attached to microtubules. In this case, filament rotation was stopped, but otherwise the impact on motility was surprisingly low. In particular, the velocity of the gliding microtubules only decreased to a negligible degree. This shows that in principle microtubules driven by processive motors like kinesin-1 can make flexible, responsive and effective molecular shuttles for nanotransport applications. In addition, the results might indicate that in vivo kinesin-1 molecules, which transport cargo along microtubules, can likewise flexibly respond to an axial force by deviating from their path parallel to the protofilament axes. Two microtubule minus-end-directed motors that might be employed to counteract kinesin-1 in engineered nanotransport systems are dynein and ncd. Both motors have been found to be capable of generating torque causing short-pitched microtubule rotation in gliding motility assays. The results for 22S dynein helped to resolve controversial findings of earlier reports about the ability of 22S dynein to generate torque. However, it turned out difficult to establish conditions where the movement of the dynein-driven nanoshuttles was homogeneous and reproducible. In contrast, motility in ncd gliding assays looks much more promising. The obtained results supported previous reports of torque generation by ncd. Moreover, a strong dependence of rotational pitches of gliding microtubules on ATP concentration was found. The reason could be that ncd motors in the nucleotide-free microtubule-bound state impede the forward movement of gliding microtubules stronger than the axial motion. To fully understand the nature of this effect, further research is required. Most likely, this will substantially contribute to the understanding of ncd function in vivo. Furthermore, the possibility of tuning the rotation of microtubules acting as nanoshuttles might provide a means to increase control of processes like cargo-loading and unloading. / Eine große Herausforderung auf dem Gebiet der Nanotechnologie ist der kontrollierte und präzise Transport von nanoskaligen Objekten. Der Einsatz von molekularen Motoren des zellulären Zytoskeletts hat sich dabei als vielversprechender Ansatz erwiesen. Ziel der hier vorgelegten Arbeit war die Entwicklung einer Methode, um das Verhalten von filamentartigen Nanotransportern - speziell von Mikrotubuli, die durch Motorproteine über Oberflächen bewegt werden - in drei Dimensionen (3-D) zu charakterisieren. Die Hauptkriterien waren dabei eine geringe Störung des zu untersuchenden Systems, hohe räumliche und zeitliche Auflösungen sowie die generelle Anwendbarkeit für Einzelmolekülstudien. Ein weiteres Ziel war es, die entwickelte Methode zur Beantwortung offener Fragen bezüglich des Nanotransports mittels Zytoskelett-basierter Motoren einzusetzen. Insbesondere sollte das System aus Mikrotubuli und dem Motorprotein Kinesin-1, welches für die meisten aktuellen Studien zum Thema Nanotransport herangezogen wird, untersucht werden. Schließlich sollten neue Erkenntnisse über weniger gut erforschte Motorproteine, speziell über 22S Dynein und das Kinesin-14 „Non-claret disjunctional“ (Ncd), gewonnen werden. Beide Motoren könnten in Nanotransportsystemen als Gegenspieler von Kinesin-1 agieren. In der vorliegenden Arbeit wird eine neuartige, auf Fluoreszenz-Interferenz basierende 3-D Nanometertrackingmethode beschrieben. Auf deren Grundlage wird es möglich, die Bewegung von einzelnen fluoreszenten Partikeln nahe einer reflektierenden Oberfläche mit einer Genauigkeit im Nanometerbereich zu verfolgen. Im Vergleich zu anderen kürzlich vorgestellten 3-D Techniken, welche auf bifokaler optischer Mikroskopie basieren und ähnliche Genauigkeiten zulassen, ist die hier vorgestellte Methode mit deutlich geringerem Aufwand auf der Basis eines herkömmlichen Epi-Fluoreszenzmikroskops umsetzbar. Dabei kann die Fluoreszenzanregung wahlweise mit einer Bogenlampe oder einem Laser erfolgen. Weiterhin besteht die Möglichkeit, nicht nur Differenzwerte (wie bei bifokaler Mikroskopie), sondern absolute Werte in der Höhendimension zu messen. Im Ergebnis wurde ein mit geringem Aufwand umsetzbares, gleichwohl hochgradig genaues und vielseitig einsetzbares Werkzeug geschaffen, welches ideal für Studien der Interaktionen von Einzelmolekülen oder auch intramolekularer Dynamik geeignet ist. Mit Hilfe der hier vorgestellten 3-D Trackingmethode wurden die Rotationen von Mikrotubuli um ihre Längsachse während des Gleitens auf mit Motorproteinen besetzten Oberflächen analysiert. Diese Geometrie wird derzeit bevorzugt in Studien eingesetzt, welche den Einsatz von biomolekularen Motoren für den Transport von nanoskaligen Objekten untersuchen und das Ziel verfolgen, Nanostrukturen zu erzeugen und zu manipulieren. Die Ergebnisse zu Rotationen von Mikrotubuli, welche über mit Kinesin-1 besetzte Oberflächen bewegt werden, sind konsistent mit (i) der Eigenschaft von Kinesin-1 sich entlang der Protofilamente von Mikrotubuli zu bewegen und (ii) der Superhelixstruktur von in vitro rekonstituierten Mikrotubuli. Dies belegt die Eignung der Methode für die Charakterisierung von Nanotransportsystemen. Die Rotation von Mikrotubuli, welche durch Kinesin-1 angetrieben werden, hat sich sowohl beim Transport von kleinen Objekten in Form von Quantum Dots (Durchmesser ca. 20 nm) als auch bei der Reduktion elektrostatischer Wechselwirkungen zwischen Kinesin-1 und Mikrotubuli durch Verdau der Tubulin-C-Termini als stabil erwiesen. Ein vollkommen anderes Bild ergab sich für den Transport von großen Objekten (Durchmesser ca. 3 µm). In diesem Fall wurde die Rotation der Filamente angehalten. Unerwarteterweise war jedoch die Vorwärtsbewegung der Mikrotubuli und insbesondere deren Geschwindigkeit kaum betroffen. Dies zeigt, daß Mikrotubuli, welche von prozessiven Motoren wie Kinesin-1 angetrieben werden, das Potential zu responsiven, flexiblen und effektiven molekularen Shuttles besitzen. Außerdem weisen die Ergebnisse darauf hin, daß Kinesin-1-Moleküle, welche in vivo Frachten entlang von Mikrotubuli transportieren, auf seitwärts gerichtete Kräfte reagieren können, indem sie von ihrem intrinsisch vorgegebenen Pfad parallel zur Protofilamentachse des Mikrotubulus abweichen. Zwei Motoren, die sich im Gegensatz zu Kinesin-1 in Richtung des Minus-Endes von Mikrotubuli bewegen, sind 22S Dynein und Ncd. Sie sind somit als Gegenspieler von Kinesin-1 in Nanotransportsystemen prädestiniert. Beide Motoren können, ebenso wie Kinesin-1, die Translokation von Mikrotubuli über Oberflächen sowie damit verbundene Rotationen von Mikrotubuli verursachen. Im Gegensatz zu Kinesin-1 tritt die Rotation unabhängig von einer Superhelixstruktur der Mikrotubuli auf. Die Ergebnisse für 22S Dynein lösen Widersprüche zwischen früheren Studien auf, indem sie belegen, daß dieser Motor Rotationen von Mikrotubuli erzeugen kann. Jedoch scheint es unter Verwendung von 22S Dynein nicht möglich zu sein, Bedingungen zu schaffen, unter welchen sich Mikrotubuli in geeigneter Weise als Nanoshuttles homogen und reproduzierbar bewegen. Der Einsatz von Ncd ist hier deutlich erfolgversprechender. Die in diesem Falle erlangten Erkenntnisse bezüglich der Erzeugung von Rotationen von Mikrotubuli decken sich mit früheren Studien. Ein bislang unbekannter, bemerkenswerter Effekt ist dabei ein Rückgang in der Länge der Rotationsperioden mit sinkender ATP-Konzentration. Die mit dem heutigen Wissensstand über den mechanochemischen Zyklus von Ncd konsistente Erklärung ist, daß Ncd-Motoren im nukleotidfrei an Mikrotubuli gebundenen Zustand die Vorwärtskomponente der Bewegung von gleitenden Mikrotubuli stärker hemmen als die Rotationskomponente. Möglicherweise kann die sich hieraus ergebende Möglichkeit der Regulierung der Rotation von Mikrotubuli dazu eingesetzt werden, das Be- und Entladen von Nanoshuttles zu steuern.

Microtubules are a part of the cell cytoskeleton that performs different functions, such as providing the mechanical support for the shape of a cell, acting as tracks along which the motor protein move organelles from one part of the cell to another, or the forming mitotic spindle during the cell division. The microtubules are dynamic structures, namely they can grow and shrink. The phase of microtubule growth alternates with the phase of shrinkage that results in the dynamic microtubule network in the cell. However, to form stable and spatially well-defined structures, such as a mitotic spindle, the cell needs to control this stochastic process. This is done by microtubule-associated proteins (MAPs). One class of MAPs is the proteins of XMAP216/Dis1 family, which are microtubule polymerases. The founding member of this family is X. laevis XMAP215. XMAP215 is a processive polymerase acting on the microtubule plus end. XMAP215 binds either directly or reaches the microtubule plus end by the diffusion along the microtubule lattice. Being at the microtubule plus-end XMAP215 stays there transiently and helps to incorporate up to 25 tubulin dimers into microtubule lattice before it dissociates and, therefore, it processively tracks the growing microtubule end during polymerization. There are two hypothesis of microtubule assembly promotion: (i) XMAP215 repeatedly releases an associated tubulin dimer into the microtubule growing plus end or (ii) structurally stabilizes a polymerized tubulin intermediate at the growing plus end and, therefore, preventing depolymerization events. The first way results into the increase of on-rate of tubulin dimers at the microtubule end, whereas the second way results into the decrease of off-rate of tubulin dimers at the microtubule end. Here, I show the study of the mechanism of microtubule growth acceleration by XMAP215 and the dependence of XMAP215 polymerization activity on the applied force. To answer these questions, I investigated the addition of tubulin dimers to the plus end of the microtubule by XMAP215 and how this addition depends on the applied force. XMAP215 remains at the microtubule end for several rounds of tubulin addition surfing both growing and shrinking microtubule ends. Therefore, if one could track the position of the XMAP215 molecules at the very tip of a microtubule with sufficient resolution, it would provide the information about the dynamics of the microtubule end. The technique, which can detect the position of the object of interest with high spatial and temporal resolution in addition to being able to exert a force, is an optical trap. A calibrated optical trap not only provides a good measure of displacement but also enables force measurements. To monitor the position of the molecules of interest, the molecules of interest are usually attached to a microsphere. Hence, I tethered XMAP215 to a microsphere held by an optical trap, and used XMAP215 as a handle to interact with the microtubule tip. When the microtubule grows, the XMAP215 coated microsphere will move in the optical trap and this movement can be detected with high temporal and spatial resolution. My work demonstrates that cooperatively working XMAP215 molecules can not only polymerize microtubule but also harness the energy of microtubule polymerization or depolymerization to transport some cargo. There is an evidence that orthologues of XMAP215 in budding yeasts, fission yeasts and Drosophila localize on the kinetochores. Therefore, the ability of the bearing some load during microtubule polymerization could be potentially important for the XMAP215 functioning during cell division. I also showed the influence of external force applied to the XMAP215 molecules. Pointing toward microtubule growth, a force of 0.5 pN applied to the microtubule tip-coupled XMAP215-coated microsphere increases XMAP215 polymerization activity. However, the force of the same magnitude but applied against microtubule growth does not affect XMAP215 polymerization activity. This result can be explained by the fact, that the force acting in the direction of microtubule growth constrains XMAP215 to be at the very microtubule tip. Hence, XMAP215 can not diffuse away from plus-end and there is higher chance to incorporate tubulin dimers into the microtubule plus-end. The on- and off-rate of tubulin dimers at the microtubule end are both decreased when the external force applied either in direction of microtubule growth or opposite to it. The external force affects the off-rate slightly stronger than on-rate of tubulin dimer. Taking together, my study gives new insights into the mechanism of microtubule polymerization by XMAP215 and shows some novel properties of this protein.

Asymmetric positioning of the mitotic spindle in C. elegans embryos is mediated by force-generating complexes that are anchored at the plasma membrane and that pull on microtubules growing out from the spindle poles. Although asymmetric distribution of the force generators is thought to underlie asymmetric positioning of the spindle, the number and location of the force generators has not been well defined. In particular, it has not been possible to visualize individual force generating events at the cortex. We discovered that perturbation of the acto-myosin cortex leads to the formation of long membrane invaginations that are pulled from the plasma membrane toward the spindle poles. Several lines of evidence show that the invaginations, which also occur in unperturbed embryos though at lower frequency, are pulled by the same force generators responsible for spindle positioning. Thus, the invaginations serve as a tool to localize the sites of force generation at the cortex and allow us to estimate a lower limit on the number of cortical force generators within the cell.

Molecular motors, highly-efficient biological nano-machines, hold the potential to be employed for a wide range of nanotechnological applications. Towards this end, kinesin, dynein or myosin motor proteins are commonly surface-immobilized within engineered environments in order to transport cargo attached to cytoskeletal filaments. Being able to flexibly control the direction of filament motion – in particular on planar, non-topographical surfaces – has, however, remained challenging. Here, we demonstrate the applicability of a UV-laser-based ablation technique to programmably generate highly-localized patterns of functional kinesin-1 motors with different shapes and sizes on PLL-g-PEG-coated polystyrene surfaces. Straight and curved motor tracks with widths of less than 500 nm could be generated in a highly-reproducible manner and proved to reliably guide gliding microtubules. Though dependent on track curvature, the characteristic travel lengths of the microtubules on the tracks significantly exceeded earlier predictions. Moreover, we experimentally verified the performance of complex kinesin-1 patterns, recently designed by evolutionary algorithms, for controlling the global directionality of microtubule motion on large-area substrates.

During embryogenesis, a single cell gives rise to a multi-cellular embryo through successive rounds of cell division. As cells become smaller, cellular organelles adapt their sizes accordingly. The size of the mitotic spindle—the microtubule-based structure controlling these divisions—is particularly important as it determines the distance over which chromosomes are segregated. To perform its function properly, spindle size scales with cell size. However, we still lack a mechanistic understanding of the underlying microtubule-based processes that regulate spindle scaling. In this thesis, I combined quantitative microscopy and laser ablation in zebrafish embryos and Xenopus laevis egg extract encapsulated in oil droplets. My measurements revealed the influence of microtubule length dynamics, transport, and nucleation on cell size-dependent spindle scaling. Strikingly, I discovered a hierarchical regulation of spindle size. In large cells, microtubule nucleation exclusively scales spindle size relative to cell size by changing the number of microtubules within the spindle. In small cells, microtubule dynamics fine-tune spindle size by modulating microtubule length. To understand the mechanism of spindle scaling, I proposed a theoretical model based on a limiting number of microtubule nucleators and microtubule-associated proteins that regulate microtubule length. The transition from nucleation- to dynamics-based scaling requires that microtubule number and the number of microtubule-associated proteins that promote microtubule growth scale differently with cell size. This can be achieved by sequestering an inhibitor of microtubule nucleation to the cell membrane, which is consistent with my measurements of microtubule nucleation. The differential regimes of spindle scaling modulated by microtubule nucleation and dynamics imply a gradual change in spindle architecture, which may ensure faithful chromosome segregation by spindles of all sizes.

A major challenge in nanotechnology is the controlled transport of cargo on the nanometer scale. A promising approach to this problem is the use of molecular motors of the cellular cytoskeleton. The aim of this work was to develop a method to characterize the behavior of filamentous nanoshuttles – specifically of motor protein-driven microtubules – in three dimensions (3-D). The main requirements to meet were low impact on the nanotransport system, high spatial and temporal resolution, and versatility. Furthermore, this method was intended to be used to address open questions in the field of nanotransport. In particular, it was firstly attempted to characterize cargo transport in a system currently favored by most studies in the field, where nanoshuttles are powered by the microtubule motor best understood so far – the plus-end-directed kinesin-1. Secondly, the goal was to further the understanding of potential counter-players of kinesin-1 in nanotransport applications - the much less well understood microtubule minus-end-directed motor proteins 22S dynein and the kinesin-14 non-claret disjunctional (ncd). A novel method to study the linear forward motion as well as the axial motion of filamentous nanoshuttles, which are driven by motors of the cell cytoskeleton, has been introduced. The method uses fluorescence interference-based 3-D nanometer tracking of quantum dots as optical probes that are attached to the nanoshuttles. While other recently reported 3-D tracking techniques based on dual-focus imaging offer similar sensitivity, the method here can be easily performed on any standard epi-fluorescence microscope, even with arc lamp illumination, and additionally holds the potential to retrieve absolute height values. It is strongly suggested that the ease of use might help to spread this valuable and versatile tool for a variety of applications, including studies of interactions between single molecules or even intramolecular changes. Specifically, 3-D tracking has been used to visualize and analyze the rotation of microtubules around their longitudinal axis when they are propelled on a motor protein-coated surface. This geometry called gliding assay is currently favored for most proof-of-principle studies that investigate the use of biomolecular motors for transport of nanoscale cargo with the goal to assemble and manipulate nanostructures. The suitability of the method has been proven for kinesin-1 gliding assays, where knowledge of properties of both, microtubules and kinesin-1, allowed a very precise prediction of microtubule rotation, which was matching the actual measured values very well. The microtubule rotation in kinesin-1 gliding assays has turned out to be robust against the attachment of small cargo in the shape of quantum dots (diameter ∼20 nm), but also against the reduction of electrostatic interactions between microtubules and kinesin-1 by cleavage of the tubulin E-hook. The situation was dramatically different when large cargo (beads with diameter of ∼3 µm) was attached to microtubules. In this case, filament rotation was stopped, but otherwise the impact on motility was surprisingly low. In particular, the velocity of the gliding microtubules only decreased to a negligible degree. This shows that in principle microtubules driven by processive motors like kinesin-1 can make flexible, responsive and effective molecular shuttles for nanotransport applications. In addition, the results might indicate that in vivo kinesin-1 molecules, which transport cargo along microtubules, can likewise flexibly respond to an axial force by deviating from their path parallel to the protofilament axes. Two microtubule minus-end-directed motors that might be employed to counteract kinesin-1 in engineered nanotransport systems are dynein and ncd. Both motors have been found to be capable of generating torque causing short-pitched microtubule rotation in gliding motility assays. The results for 22S dynein helped to resolve controversial findings of earlier reports about the ability of 22S dynein to generate torque. However, it turned out difficult to establish conditions where the movement of the dynein-driven nanoshuttles was homogeneous and reproducible. In contrast, motility in ncd gliding assays looks much more promising. The obtained results supported previous reports of torque generation by ncd. Moreover, a strong dependence of rotational pitches of gliding microtubules on ATP concentration was found. The reason could be that ncd motors in the nucleotide-free microtubule-bound state impede the forward movement of gliding microtubules stronger than the axial motion. To fully understand the nature of this effect, further research is required. Most likely, this will substantially contribute to the understanding of ncd function in vivo. Furthermore, the possibility of tuning the rotation of microtubules acting as nanoshuttles might provide a means to increase control of processes like cargo-loading and unloading. / Eine große Herausforderung auf dem Gebiet der Nanotechnologie ist der kontrollierte und präzise Transport von nanoskaligen Objekten. Der Einsatz von molekularen Motoren des zellulären Zytoskeletts hat sich dabei als vielversprechender Ansatz erwiesen. Ziel der hier vorgelegten Arbeit war die Entwicklung einer Methode, um das Verhalten von filamentartigen Nanotransportern - speziell von Mikrotubuli, die durch Motorproteine über Oberflächen bewegt werden - in drei Dimensionen (3-D) zu charakterisieren. Die Hauptkriterien waren dabei eine geringe Störung des zu untersuchenden Systems, hohe räumliche und zeitliche Auflösungen sowie die generelle Anwendbarkeit für Einzelmolekülstudien. Ein weiteres Ziel war es, die entwickelte Methode zur Beantwortung offener Fragen bezüglich des Nanotransports mittels Zytoskelett-basierter Motoren einzusetzen. Insbesondere sollte das System aus Mikrotubuli und dem Motorprotein Kinesin-1, welches für die meisten aktuellen Studien zum Thema Nanotransport herangezogen wird, untersucht werden. Schließlich sollten neue Erkenntnisse über weniger gut erforschte Motorproteine, speziell über 22S Dynein und das Kinesin-14 „Non-claret disjunctional“ (Ncd), gewonnen werden. Beide Motoren könnten in Nanotransportsystemen als Gegenspieler von Kinesin-1 agieren. In der vorliegenden Arbeit wird eine neuartige, auf Fluoreszenz-Interferenz basierende 3-D Nanometertrackingmethode beschrieben. Auf deren Grundlage wird es möglich, die Bewegung von einzelnen fluoreszenten Partikeln nahe einer reflektierenden Oberfläche mit einer Genauigkeit im Nanometerbereich zu verfolgen. Im Vergleich zu anderen kürzlich vorgestellten 3-D Techniken, welche auf bifokaler optischer Mikroskopie basieren und ähnliche Genauigkeiten zulassen, ist die hier vorgestellte Methode mit deutlich geringerem Aufwand auf der Basis eines herkömmlichen Epi-Fluoreszenzmikroskops umsetzbar. Dabei kann die Fluoreszenzanregung wahlweise mit einer Bogenlampe oder einem Laser erfolgen. Weiterhin besteht die Möglichkeit, nicht nur Differenzwerte (wie bei bifokaler Mikroskopie), sondern absolute Werte in der Höhendimension zu messen. Im Ergebnis wurde ein mit geringem Aufwand umsetzbares, gleichwohl hochgradig genaues und vielseitig einsetzbares Werkzeug geschaffen, welches ideal für Studien der Interaktionen von Einzelmolekülen oder auch intramolekularer Dynamik geeignet ist. Mit Hilfe der hier vorgestellten 3-D Trackingmethode wurden die Rotationen von Mikrotubuli um ihre Längsachse während des Gleitens auf mit Motorproteinen besetzten Oberflächen analysiert. Diese Geometrie wird derzeit bevorzugt in Studien eingesetzt, welche den Einsatz von biomolekularen Motoren für den Transport von nanoskaligen Objekten untersuchen und das Ziel verfolgen, Nanostrukturen zu erzeugen und zu manipulieren. Die Ergebnisse zu Rotationen von Mikrotubuli, welche über mit Kinesin-1 besetzte Oberflächen bewegt werden, sind konsistent mit (i) der Eigenschaft von Kinesin-1 sich entlang der Protofilamente von Mikrotubuli zu bewegen und (ii) der Superhelixstruktur von in vitro rekonstituierten Mikrotubuli. Dies belegt die Eignung der Methode für die Charakterisierung von Nanotransportsystemen. Die Rotation von Mikrotubuli, welche durch Kinesin-1 angetrieben werden, hat sich sowohl beim Transport von kleinen Objekten in Form von Quantum Dots (Durchmesser ca. 20 nm) als auch bei der Reduktion elektrostatischer Wechselwirkungen zwischen Kinesin-1 und Mikrotubuli durch Verdau der Tubulin-C-Termini als stabil erwiesen. Ein vollkommen anderes Bild ergab sich für den Transport von großen Objekten (Durchmesser ca. 3 µm). In diesem Fall wurde die Rotation der Filamente angehalten. Unerwarteterweise war jedoch die Vorwärtsbewegung der Mikrotubuli und insbesondere deren Geschwindigkeit kaum betroffen. Dies zeigt, daß Mikrotubuli, welche von prozessiven Motoren wie Kinesin-1 angetrieben werden, das Potential zu responsiven, flexiblen und effektiven molekularen Shuttles besitzen. Außerdem weisen die Ergebnisse darauf hin, daß Kinesin-1-Moleküle, welche in vivo Frachten entlang von Mikrotubuli transportieren, auf seitwärts gerichtete Kräfte reagieren können, indem sie von ihrem intrinsisch vorgegebenen Pfad parallel zur Protofilamentachse des Mikrotubulus abweichen. Zwei Motoren, die sich im Gegensatz zu Kinesin-1 in Richtung des Minus-Endes von Mikrotubuli bewegen, sind 22S Dynein und Ncd. Sie sind somit als Gegenspieler von Kinesin-1 in Nanotransportsystemen prädestiniert. Beide Motoren können, ebenso wie Kinesin-1, die Translokation von Mikrotubuli über Oberflächen sowie damit verbundene Rotationen von Mikrotubuli verursachen. Im Gegensatz zu Kinesin-1 tritt die Rotation unabhängig von einer Superhelixstruktur der Mikrotubuli auf. Die Ergebnisse für 22S Dynein lösen Widersprüche zwischen früheren Studien auf, indem sie belegen, daß dieser Motor Rotationen von Mikrotubuli erzeugen kann. Jedoch scheint es unter Verwendung von 22S Dynein nicht möglich zu sein, Bedingungen zu schaffen, unter welchen sich Mikrotubuli in geeigneter Weise als Nanoshuttles homogen und reproduzierbar bewegen. Der Einsatz von Ncd ist hier deutlich erfolgversprechender. Die in diesem Falle erlangten Erkenntnisse bezüglich der Erzeugung von Rotationen von Mikrotubuli decken sich mit früheren Studien. Ein bislang unbekannter, bemerkenswerter Effekt ist dabei ein Rückgang in der Länge der Rotationsperioden mit sinkender ATP-Konzentration. Die mit dem heutigen Wissensstand über den mechanochemischen Zyklus von Ncd konsistente Erklärung ist, daß Ncd-Motoren im nukleotidfrei an Mikrotubuli gebundenen Zustand die Vorwärtskomponente der Bewegung von gleitenden Mikrotubuli stärker hemmen als die Rotationskomponente. Möglicherweise kann die sich hieraus ergebende Möglichkeit der Regulierung der Rotation von Mikrotubuli dazu eingesetzt werden, das Be- und Entladen von Nanoshuttles zu steuern.

Proteins have been optimized by evolution for billions of years to work on a nanometer scale. Therefore, they are extremely promising for nanotechnological applications. Cytoskeletal filaments propelled by surface-attached motor proteins have been recently established as versatile transport platforms for nano-sized cargo in molecular sorting and nano-assembly devices. However, in this gliding motility setup, cargo and motors share the filament lattice as a common substrate for their activity. Therefore, it is important to understand the influence of cargo-loading on transport properties. By performing single molecule stepping assays on biotinylated microtubules, it was shown that kinesin-1 motors first stop and then detach when they encounter a streptavidin obstacle on their path along the microtubule. Consequently, the deceleration of streptavidin coated microtubules in gliding assays could be attributed to an obstruction of kinesin-1's path on the microtubule rather than to "frictional" streptavidin-surface interactions. The insights gained by studying kinesin-1's behavior at obstacles were then used to demonstrate a novel sensing application: Using a mixture of two distinct microtubule populations that each bind a different kind of protein, the presence of these proteins was detected via speed changes in the respective microtubule populations. In future applications, this detection scheme could be combined with other recent advancements in the field, creating highly integrated lab-on-a-chip devices that use microtubule based transport to detect, sort and concentrate analytes. It has been envisioned that the kinesin-1-microtubule system could be used for even more complex appliances like nano-assembly lines. However, currently available control mechanisms for kinesin-1 based transport are not precise enough. Therefore, improved temporal control mechanisms for kinesin-1 were investigated: Using a polymer that changes its size in solution with temperature, starting and stopping of gliding microtubules was demonstrated. In combination with local heating by light, this effect could be used to control the gliding of single microtubules. Finally, a strategy to create photo-switchable kinesin-1 was developed and tested for feasibility using molecular modeling.

Microtubules are a part of the cell cytoskeleton that performs different functions, such as providing the mechanical support for the shape of a cell, acting as tracks along which the motor protein move organelles from one part of the cell to another, or the forming mitotic spindle during the cell division. The microtubules are dynamic structures, namely they can grow and shrink. The phase of microtubule growth alternates with the phase of shrinkage that results in the dynamic microtubule network in the cell. However, to form stable and spatially well-defined structures, such as a mitotic spindle, the cell needs to control this stochastic process. This is done by microtubule-associated proteins (MAPs). One class of MAPs is the proteins of XMAP216/Dis1 family, which are microtubule polymerases. The founding member of this family is X. laevis XMAP215. XMAP215 is a processive polymerase acting on the microtubule plus end. XMAP215 binds either directly or reaches the microtubule plus end by the diffusion along the microtubule lattice. Being at the microtubule plus-end XMAP215 stays there transiently and helps to incorporate up to 25 tubulin dimers into microtubule lattice before it dissociates and, therefore, it processively tracks the growing microtubule end during polymerization. There are two hypothesis of microtubule assembly promotion: (i) XMAP215 repeatedly releases an associated tubulin dimer into the microtubule growing plus end or (ii) structurally stabilizes a polymerized tubulin intermediate at the growing plus end and, therefore, preventing depolymerization events. The first way results into the increase of on-rate of tubulin dimers at the microtubule end, whereas the second way results into the decrease of off-rate of tubulin dimers at the microtubule end. Here, I show the study of the mechanism of microtubule growth acceleration by XMAP215 and the dependence of XMAP215 polymerization activity on the applied force. To answer these questions, I investigated the addition of tubulin dimers to the plus end of the microtubule by XMAP215 and how this addition depends on the applied force. XMAP215 remains at the microtubule end for several rounds of tubulin addition surfing both growing and shrinking microtubule ends. Therefore, if one could track the position of the XMAP215 molecules at the very tip of a microtubule with sufficient resolution, it would provide the information about the dynamics of the microtubule end. The technique, which can detect the position of the object of interest with high spatial and temporal resolution in addition to being able to exert a force, is an optical trap. A calibrated optical trap not only provides a good measure of displacement but also enables force measurements. To monitor the position of the molecules of interest, the molecules of interest are usually attached to a microsphere. Hence, I tethered XMAP215 to a microsphere held by an optical trap, and used XMAP215 as a handle to interact with the microtubule tip. When the microtubule grows, the XMAP215 coated microsphere will move in the optical trap and this movement can be detected with high temporal and spatial resolution. My work demonstrates that cooperatively working XMAP215 molecules can not only polymerize microtubule but also harness the energy of microtubule polymerization or depolymerization to transport some cargo. There is an evidence that orthologues of XMAP215 in budding yeasts, fission yeasts and Drosophila localize on the kinetochores. Therefore, the ability of the bearing some load during microtubule polymerization could be potentially important for the XMAP215 functioning during cell division. I also showed the influence of external force applied to the XMAP215 molecules. Pointing toward microtubule growth, a force of 0.5 pN applied to the microtubule tip-coupled XMAP215-coated microsphere increases XMAP215 polymerization activity. However, the force of the same magnitude but applied against microtubule growth does not affect XMAP215 polymerization activity. This result can be explained by the fact, that the force acting in the direction of microtubule growth constrains XMAP215 to be at the very microtubule tip. Hence, XMAP215 can not diffuse away from plus-end and there is higher chance to incorporate tubulin dimers into the microtubule plus-end. The on- and off-rate of tubulin dimers at the microtubule end are both decreased when the external force applied either in direction of microtubule growth or opposite to it. The external force affects the off-rate slightly stronger than on-rate of tubulin dimer. Taking together, my study gives new insights into the mechanism of microtubule polymerization by XMAP215 and shows some novel properties of this protein.

Optical tweezers are a sensitive position and force transducer widely employed in physics and biology. In a focussed laser, forces due to radiation pressure enable to trap and manipulate small dielectric particles used as probes for various experiments. For sensitive biophysical measurements, microspheres are often used as a handle for the molecule of interest. The force range of optical traps well covers the piconewton forces generated by individual biomolecules such as kinesin molecular motors. However, cellular processes are often driven by ensembles of molecular machines generating forces exceeding a nanonewton and thus the capabilities of optical tweezers. In this thesis I focused, fifirst, on extending the force range of optical tweezers by improving the trapping e fficiency of the probes and, second, on applying the optical tweezers technology to understand the mechanics of molecular motors. I designed and fabricated photonically-structured probes: Anti-reflection-coated, high-refractive-index, core-shell particles composed of titania. With these probes, I significantly increased the maximum optical force beyond a nanonewton. These particles open up new research possibilities in both biology and physics, for example, to measure hydrodynamic resonances associated with the colored nature of the noise of Brownian motion. With respect to biophysical applications, I used the optical tweezers to study the mechanics of single kinesin-8. Kinesin-8 has been shown to be a very processive, plus-end directed microtubule depolymerase. The underlying mechanism for the high processivity and how stepping is affected by force is unclear. Therefore, I tracked the motion of yeast (Kip3) and human (Kif18A) kinesin-8s with high precision under varying loads. We found that kinesin-8 is a low-force motor protein, which stalled at loads of only 1 pN. In addition, we discovered a force-induced stick-slip motion, which may be an adaptation for the high processivity. Further improvement in optical tweezers probes and the instrument will broaden the scope of feasible optical trapping experiments in the future.