Recent and Ongoing Research Projects
Probing cell structure and function using micropatterned substrates
Eukaryotic cells respond both structurally and functionally to spatial and chemical cues in the extracellular environment. The emergence of technologies for patterning proteins from length scales of millimeters down to nanometers now presents the opportunity to understand what types of cues are important, and how responses to specific cues are mediated. We are working on a several different approaches to this problem, initially focusing on model cell systems such as fibroblasts, endothelial cells and melanoma cells. This work involves a highly interdisciplinary collaboration with Dr. Lew Romer (cell biology) and Dr. David Havilands (nanostructure physics) laboratories. In collaboration with Mats Ulfendahl’s laboratory at the Karolinska Insitute we are also applying these approaches to study neuronal systems. This is a new research area, and the first papers are now making their way through review. Details will follow when they are in press.
Back to Top
-----------------------
Molecular patterning technology development
We have established two new modalities for patterning proteins. One is based a novel microfluidic technology from BioForce Nanoscience, which allows for direct writing of compositionally complex patterns on a micrometer length scale. This approach works very much like a miniaturized quill tip pen that is simply inked with a protein solution, which is then printed onto almost any type of surface. The second is an approach, in collaboration with the Haviland Group, is based on electron beam lithography that allows for protein patterns approaching feature sizes of 50 nm, approximately the dimensions of a single fibronectin molecule on a surface. Together these approaches provide powerful tools for studying cell surface interactions.
Back to Top
-----------------------
Cellular nanomechanics
The mechanical properties of cells are fundamentally important to a number of biological activities. Eukaryotic cells are soft visco-elastic bodies with a Young’s modulus ranging from ~1 kPa to ~100 kPa. It is clear that the cytoskeleton plays a central role in determining the mechanics, but the connection between the molecular constituents of a cell and its mechanical properties remains poorly understood. We have worked on the mechanics of three cellular systems, motor neurons, endothelial cells and outer hair cells. Most of our work has been on the cytoskeletal components of motor neurons, in particular nanomechanical measurements using atomic force microscopy of neurofilaments and microtubules (with their associated MAPs). The principle discovery to come from this work is that both neurofilaments and microtubules are composed of proteins that have segments that are intrinsically or natively disordered. These disordered segments form what are effectively small nanomechanical springs, which may play important mechanical roles in neurons. This work is now continuing with high resolution simulations of neurofilaments in collaboration with Mark Stevens at Sandia National Laboratory. We would also like to begin to connect the molecular scale measurements with whole cell mechanical measurements. Atomic force microscopy has also been used to examine the nanomechanics of vascular endothelial cells. Here we have shown that the AFM can be used to “see” mechanical organization in living cells at a length scale of ~100 nm. This is a critical length scale for bridging properties of molecules with those of whole cells. With Mats Ulfendahl’s laboratory we have also examined the nanomechanical properties of outer hairs cells from the cochlea. These cells are thought to be involved in tuning auditory reception by changes in mechanical properties. One of the most surprising findings here has been that indentations of OHCs at rates over 100 µm per second are almost completely non-hysteretic, indicating very small viscous effects for those types of displacements. This is compares to for example MDCK cells, which have appreciable viscous effects at a few micrometers per second.
Back to Top
-----------------------
Biophysics of intrinsically disordered proteins
The finding that disordered proteins might play a mechanical role in axons lead us to ask what other roles might disordered proteins play, and what makes a protein disordered? To address the latter question we have developed a disorder predictor based on a support vector machine. The support vector machine is a learning type algorithm that can be trained to discriminate members of two groups from each other given some set of input measurements. It is qualitatively similar to a neural net, but has the advantage that the SVM provides a quantitative measure of the importance of the different input parameters to the final output. This provide some insight into what properties of proteins are most important for determining whether or not they will be disordered. One central finding from that work is that based on sequence composition alone allows for 87% percent accurate sorting of ordered from disordered protein, similar to the best available neural net based programs. Further, using reduced amino acid sets suggests that the balance between hydrophilic and hydrophobic amino acids is a key determinant of disorder propensity. Notably a few hydrophobic amino acids such as tryptophan have extremely high weight vectors, and thus contribute more to the balance between order and disorder than any of the hydrophilic amino acids. Most recently we have performed a two dimensional analysis combining disorder score the sequence complexity, in order to see if subclasses of disordered proteins can be identified. The full results from this analysis shows that there are characteristic differences in the disorder-complexity distribution between different functional classes of proteins. They also suggest a new approach to discovering previously unappreciated relationships between proteins.
Back to Top
-----------------------
Mechanically based biosensing
AFM cantilevers have been modified to perform highly sensitive detection of biomolecules by measuring binding induced stress changes at the cantilever surface, or changes in resonant properties of the cantilevers. However, cantilevers present a number of practical challenges and despite compelling advantages, such as label free detection, their use is still restricted to a small number of highly specialized laboratories. We are presently collaborating with Angela Zapata, Marc Weinberg and Jeff Borenstein at Draper Laboratory to develop a sensitive and versatile platform to mechanically based sensing of biomolecules that is also robust and useful.
Back to Top
-----------------------
Molecular mechanisms of membrane fusion
What started as an AFM project to measure fusion forces turned into a simulation project, in collaboration with Mark Stevens and Tom Woolf, to understand the molecular details of fusion between two lipid bilayers (in the form of two liposomes). These simulations where based on coarse grain models where some of the molecular details is simplified in order to achieve longer time and length scales needed to achieve fusion. The simulations produced some interesting and surprising findings. To begin, fusion did not initiate at the point of closest approach, but at the edge of the flat contact surface between two liposomes. In retrospect this makes sense, since the membrane at the edge will be most strained. Further, the initial steps in fusion involved the splaying of lipids in the outer leaflet of one bilayer such that the lipid molecule was shared between the two liposomes. This was unstable, but when more than one lipid splayed two lipids tended stabilized each other and nucleate the formation of a stalk that resulted in hemifusion. This work is ongoing with experimental work also being planned.
Back to Top
-----------------------
Other Projects
We also have a number of other projects in collaboration with other groups…More details later.
|
