In this work the mechanical properties of model cellular and subcellular processes have been studied using self-assembled biomimetic systems of the actin cortex of the cytoskeleton. For this purpose, microscopy based magnetic colloidal force transducers (magnetic tweezers) were designed and image analysis algorithms developed. The first part examines the transport properties of the processive motor protein myosin V. In single molecule experiments, an average step size of 36.5 nm of the molecule is calculated from a statistical analysis of the movement data. This value coincides with the helix pseudo repeat of the actin filament myosin V moves along. Using magnetic colloidal force transducers, the detachment forces of single actively moving myosin V molecules from actin filaments could be measured for forces parallel and perpendicular to the movement direction. For forces applied perpendicular to the movement direction, the unbinding forces show an increase with increasing force rate, for which a general theoretical model for the dynamic unbinding of a moving motor protein was developed. This model yields an analytic relation of unbinding force and force rate, which was used to compute an internal length of 11 nm of the binding potential and an equilibrium dissociation constant of 0.3 s−1 of the predominant kinetic state of the motor protein. To estimate an error for the values for the two potential parameters from the limited data set, a Monte carlo like method of fitting a randomly modified data set was developed. For comparison, a second method to determine the internal length of the binding potential and the equilibrium dissociation constant, from a combined rupture force probability density of all rupture events, was developed. This yielded equivalent results. The determined dissociation constant is remarkably smaller than expected from the duty ratio of the individual binding sites of the molecule. This leads to the conclusion, that a strain mediated interaction between the two actin binding sites exists, which results in longer movements of the molecule and a super-processive behavior. Additionally the movement of single myosin V molecules under forward and backward load was observed. The result that the velocity of myosin V does not change up to a load of 900 fN can be combined with previous studies to the picture that in an apparent 8 internal strain coupling between the two active domains during the movement of the molecule, the leading head is blocked by the stress the trailing head imposes on it. The second part of this work deals with self-assembled actin networks on three dimensionally micro-structured surfaces. The structures used were arrays of micro pillars made by photo lithography and etching techniques from Silicon, photo resin or PDMS. Actin filaments could selectively be polymerized from the tops of these pillars, so that after cross linking of the filaments a freely suspended quasi two dimensional actin network was obtained on top of the pillar array. This network structure served as biomimetic model for the quasi two dimensional membrane-associated actin cortex in cells. The motor protein myosin V was used to transport polystyrene nano-beads on the network structure and thus mimic inner-cellular transport processes. With magnetic colloidal force transducers local elastic properties of the actin network were measured and an approximate Young modulus of 1 N/m2 for the network was determined. The micro pillar arrays were also used to immobilize single actin filaments and bundles on them and calculate their elastic properties from the observation of their thermal undulations. Two different methods for the data analysis were used which both yielded equivalent values for the persistence length of 15 - 17 µm. The Young modulus of actin as a material property was calculated taking into account the helical structure of the filament, which leads to a value of 3.1 · 108N/m2. Using quantitative fluorescence analysis, the number of actin filaments in actin - filamin bundles could be determined. The persistence length of these bundles was analyzed in dependence on the number of filaments in the bundle. This showed a quadratic dependence, as for a homogeneous rigid rod with the same Young modulus as actin. In the third part of this work, the viscoelastic properties of vesicles that contained a self assembled internal actin cortex were examined. Actin filaments that were polymerized inside phospholipid vesicles form a thin cortex structure of mostly parallel filaments bound to the phospholipid bilayer membrane when electrostatically attracted to it. Magnetic beads were placed 9 from the outside onto substrate adhered vesicles and pulled perpendicular to the substrate. To observe the movement of beads in three dimensions, a three dimensional particle tracking algorithm was established. The bead movement in response to a periodic step force of 0.3 up to 5 pN was analyzed as the creep compliance function J(t), parallel and perpendicular to the vesicle surface, in the time domain. Using a numeric procedure the creep data J(t) was converted into the stress relaxation function G(t). This procedure showed the existence of separated time scales in the stress relaxation so that a phenomenological viscoelastic model could be used to fit the creep compliance data. The use of two response times yielded good fit results and a set of four parameters (two viscous and two elastic) was used to describe the creep response of the system. The two response times were separated by one order of magnitude. All four parameters showed a linear dependence on actin surface density. Furthermore the influence of photochemically induced damage on the actin filaments and a work hardening of the cortex of actin vesicles are discussed. The analysis of membrane fluctuations, using the magnetic bead as a marker, led to the determination of a bending rigidity of roughly 1000 kBT and the shear rigidity of 0.1 µN/m the vesicles. The comparison of these two elastic parameters leads to the conclusion that in the anisotropic actin cortex deformation modes longitudinal to the filament orientation are preferred over transverse ones. To visualize the strain field in the vesicle surface, small non-magnetic beads were additionally attached to the outside of the vesicles. By simultaneously observing the movement of a magnetic and many non-magnetic beads, the strain field inside the membrane-cortex composite sheet could be mapped. From this strain field in the two dimensional actin cortex the Young modulus of the actin - membrane composite sheet was determined to be 179 N/m2 In a fourth part evaporating droplets of entangled actin networks have been studied. At certain wetting conditions of actin solutions on glass, the triple interface line between glass, actin solution and air is pinned during a first period of the evaporation process. This leads to a radially outward flow. At the triple line, actin is accumulated in a dense structure. When outward flowing filaments encounter this structure, they are compressed against it and show 10 bending and folding with characteristic bending contour length. To analyze these bending contour length, a filament contour tracing algorithm was developed. Observed bending contours differ significantly from buckling contours expected from calculations of buckling under hydrodynamic forces. Comparison of the network mesh size with the thermal undulation amplitudes and the characteristic buckling time with the relaxation time of the thermal modes show that the bending process observed is rather a amplification of accessible thermal modes than a classical Euler buckling instability. The dominant amplified mode is determined by the characteristic time of the bending process and the mode relaxation time of the filament. In the last part phospholipid vesicles, embedded in actin-myosin networks, have been used as force sensors. Motor active actin-myosin networks have been shown to exhibit a sol-gel transition, due to a percolation, and internal movements of filaments in the moment of ATP depletion. Vesicles embedded in and electrostatically coupled to such actin myosin networks show strong deformations from their initial shape during the sol-gel transition. This deformation happens simultaneously to a drastic increase in the viscoelastic moduli of the network, which were measured simultaneously in the same sample using magnetic bead micro-rheometry. An image processing algorithm to analyze shapes of deforming vesicles was established. Thus the local deformation energy of the actin myosin network contraction was calculated. A three dimensional reconstruction method of the vesicle shape showed that the surface area of the vesicle stays constant during the transition, but the volume decreases by 25 %. The comparison of the vesicle shapes during the transition with phase diagrams of vesicle equilibrium shapes show that although the obtained shape seems to have the form of a stomatocyte, the deformation can not be described in terms of a area - volume shape transformation of vesicles. This leads to the assumption that the vesicle is deformed by locally percolating actin structures in the network. An approximate deformation force acting on a small part of the vesicle is calculated for the entire deformation process. The maximum force exerted on the vesicle during the percolation is shown to be around 100 pN, which is a reasonable value for myosin assemblies.     «

In this work the mechanical properties of model cellular and subcellular processes have been studied using self-assembled biomimetic systems of the actin cortex of the cytoskeleton. For this purpose, microscopy based magnetic colloidal force transducers (magnetic tweezers) were designed and image analysis algorithms developed. The first part examines the transport properties of the processive motor protein myosin V. In single molecule experiments, an average step size of 36.5 nm of the molecule...     »

In der vorliegenden Arbeit wurden die mechanischen Eigenschaften von Modellsystemen zellulärer und subzellulärer Prozesse mit Hilfe selbstorganisierter biomimetischer Modelle des Aktinzytoskeletts untersucht. Hierzu wurden neue Instrumente der magnetischen kolloidalen Kraftmikroskopie entworfen und neue Bildverarbeitungstechniken entwickelt. Im ersten Teil der Arbeit wurde das Transportverhalten des Motorproteins Myosin V in Einzelmolekülexperimenten untersucht. Dabei wurde durch statistische Analyse der Bewegungstrajektorien einzelner Moleküle eine mittlere Schrittweite des Moleküls von 36,5 nm bestimmt, was der Pseudowiederholungslänge des Aktinfilaments entspricht. Mit Hilfe der magnetischen Kraftmikroskopie konnten die Dissoziationskräfte einzelner Myosin V Moleküle von Aktinfilamenten bestimmt werden. Für die dabei beobachtete Ratenabhängigkeit der Dissoziationskräfte wurde ein theoretisches Modell entwickelt, welches die Abhängigkeit erklärt und die Bestimmung zweier charakteristischer Parameter des Bindungspotentials erlaubt. Die Abweichung der so bestimmten Dissoziationsrate in Abwesenheit externe Kräfte (0.3 s-1) von dem theoretisch erwarteten Wert (1,57 s-1) erlaubt den Schluss, dass eine interne Kopplung der beiden Bindungsdomainen des Myosin V Moleküls an das Aktinfilament existiert. In einem zweiten Teil wurde die Selbstorganisation komplexer Aktinstrukturen auf mikrostrukturierten Säulensubstraten und deren mechanische Eigenschaften mit Hilfe magnetischer Kraftmikroskopie und Undulationsspektroskopie untersucht. Durch Quervernetzung mit dem Aktinbindungsprotein Filamin bilden sich Aktinnetzwerke, welche die Geometrie des Substrates nachempfinden und Young Module in der Größenordung von 1 Pa aufweisen. Diese Netzwerke wurden als Transportwege für mit Myosin V beschichtete Kugeln benützt. Durch Undulationsspektroskopie von Aktinfilamenten und Bündeln konnten deren Persistenzlängen bestimmt werden Dies ergab für Aktinfilamente Werte von 15 – 17 µm und für Aktinbündel eine quadratische Abhängigkeit der Persistenzlänge von der Anzahl der Filamente. In einem dritten Teil wurden Vesikel mit innerem Aktinkortex als mechanische Zellmodelle untersucht. Mit Hilfe der magnetischen Kraftmikroskopie und einem dreidimensionalen Kugelnachverfolgungsalgorithmus konnten die Kriechkomplianz und die daraus berechnete Spannungsrelaxation der Vesikel bestimmt werden. Beide Größen zeigen eine Separation der Zeitskalen, was auf zwei getrennte Mechanismen mechanischer Eigenschaften hinweist. Die Analyse thermischer Fluktuationen der Vesikelmembran ergab Werte für die Biegesteifigkeit der Vesikel in der Ordnung von 1000 kBT. Eine Messung des Verzerrungsfeldes an der Vesikeloberfläche ergab einen Wert von ca. 180 Pa für das Young Modul des Membran – Kortex Materials. In einem vierten Teil wurden eintrocknende Tropfen von Lösungen von Aktinfilamenten betrachtet in denen ein radialer Fluß entsteht welcher einzelne Filamente gegen die Grenzfläche drückt und zu einer Biegung und Faltung der Filamente führt. Durch eine Analyse der Filamentkonturen konnte gezeigt werden, dass der Biegungsprozess nicht einer mechanischen Instabilität durch eine Kompression, sondern einer Verstärkung bereits existenter Undulationsmoden entspricht. Im letzten Teil wurden Vesikel als Kraftsensoren in Perkolationsstrukturen aktiver Aktin – Myosinnetzwerke verwendet. Durch eine Formanalyse und eine dreidimensionale Rekonstruktion der Vesikelformen konnte gezeigt werden, dass die beobachtete Deformation der Vesikel nicht einer Gleichgewichtsformumwandlung entspricht, sondern durch inhomogene Perkolation des Netzwerks mit Kompressionskräften von bis zu 100 pN entspricht.     «

In der vorliegenden Arbeit wurden die mechanischen Eigenschaften von Modellsystemen zellulärer und subzellulärer Prozesse mit Hilfe selbstorganisierter biomimetischer Modelle des Aktinzytoskeletts untersucht. Hierzu wurden neue Instrumente der magnetischen kolloidalen Kraftmikroskopie entworfen und neue Bildverarbeitungstechniken entwickelt. Im ersten Teil der Arbeit wurde das Transportverhalten des Motorproteins Myosin V in Einzelmolekülexperimenten untersucht. Dabei wurde durch statistische An...     »