RESEARCH

Mapping cell internals with engulfed beads. Collaboration with Prof Alexander Bershadsky, WIS Dept of Molecular Cell Biology

We used optical tweezers to place protein-treated beads on the surface of multinuclear giant cells, and then tracked their motion. We expected to see a centripetal movement of the plasma membrane, but instead found that the beads were engulfed by the cells if placed near the periphery, while beads placed toward the cell center remained fixed in place. Beads that arrive near the cell center after being engulfed at the edge wander continuously nearby the nuclei. [publication page]

top

Anomalous diffusion in microtubule networks and in microtubule-dependent cellular motion. Collaboration with Prof Rony Granek, Ben Gurion University.

Classical Brownian motion of a particle in a simple solution executes thermal diffusion that follows the Stokes-Einstein relation, ⟨x²⟩=2dDt, where: where ⟨x²⟩ is the mean square displacement and D is the diffusion constant. For systems with a broad spectrum of relaxation times, e.g. polymers, diffusion may be "anomalous" in that , ⟨x²⟩~tγ, where γ≠1. Using small beads bound to microtubules at single points, we found that thermal undulations generate motion with a sub-diffusive exponent γ=3/4. This scaling can be derived for a semi-flexible polymer as a superposition of modes with wavenumber-dependent damping. We analyzed the wandering movements of beads that were engulfed into fibroblast cells and reached the perinuclear region. Surprisingly, the mean square displacement scaled with γ=3/2, an entirely unexpected exponent. This behavior depended on intimate contact with a surrounding microtubule network. A consistent explanation is that in order to move through the net, even when driven by motor proteins, the bead must push the surrounding microtubules out of the way. The drag force they impose is effectively time-dependent. Following the indication from isolated microtubules, we can expect that, η~t1/4. In the mean square displacement, then, on short times then we see randomly-oriented ballistic motion with scaling reduced by two factors of η from γ=2 to γ=3, while at long times we observed a diffusive motion with scaling γ=1/2. [publication page-I][publication page-II][publication page-III]

top

Nucleocytoplasmic transport mechanism

The cells of all higher organisms are distinguished by a nucleus, which encases the genome. Protein synthesis, on the other hand, takes place in the cytoplasm. The mRNA codes that are transcribed from the DNA genes and edited in the nucleus must be exported, while all the proteins needed in the nucleus must be imported. Proteins intended for nuclear entry contain in their primary sequence an appropriate peptide, i.e. a nuclear localization signal. This signal interacts with a receptor protein that ushers it through the nuclear pore. In fact this system of pore and receptor is able to pump the signal-carrying protein cargos to a much higher concentration in the nucleus than that in the cytoplasm. This led to a concept of vectorial transport at the nuclear pore, where the receptor carries the cargo only in one direction and then is recycled to the cytoplasm with a competing substrate. By reconstituting this process in a cell-free system based on in vitro nuclear reconstitution in Xenopus egg extract, we were able to make quantitative kinetic measurements on the accumulation of a fluorescent protein. In this system the cytosol serves as a thermodynamic reservoir for all components, so interpretation is greatly simplified. We found that the protein uptake follows simple first-order kinetics and therefore saturates at a specific value. By titrating the concentration in the cytosol, we found that the nuclear concentration follows a Michaelis-Menten dependence that is a hallmark of receptor-mediated transport. Most strikingly, we found that at saturation there is a bidirectional flux of protein through the nuclear pore mediated by the same receptor that has till now been implicated only in protein import. This implies that localization according to the signal peptide is relevant to the protein population rather than the individual molecule. [publication page]

top

Nuclear import of DNA

We set up an assay to test nuclear import of DNA, based on cell-free nuclei reconstituted in vitro. We observed accumulation of fluorescently-labeled lambda phage DNA inside. We then attached individual lambda phage molecules to small polystyrene beads, mixed these with reconstituted nuclei, and measured the uptake kinetics of single DNA molecules using optical tweezers. [publication page]

top

Agrobacterium tumefaciens

Our interest in nuclear import in general, and nuclear import of DNA in particular, led us to Agrobacterium as a successful model for gene delivery to eukaryotic cells, in this case to plant cells. Its infection process involves expression in the plant of genes encoded on a single-stranded “transfer-DNA”, or T-DNA, that arrives from the bacterium. Once in the plant cytoplasm the T-DNA interacts with a bacterial virulence protein, VirE2, that protects it from endonucleases and adapts it for nuclear import. We have studied the structure of the VirE2-ssDNA complex assembled in vitro using electron microscopy and three-dimensional reconstruction techniques. We have also studied the regulation of VirE2-ssDNA interaction by VirE1, a small chaperone of VirE2. [publication page-I][publication page-II]

top

Integration of nucleocytoplasmic transport with microtubule-based centripetal delivery

The canconical architecture of animal cells, with microtubules radiating from a centrosome nearby the nucleus, suggests one mechanism for nuclear entry that begins with motor-driven movement along the microtubules. Indeed many virus employ such a strategy to promote nuclear entry of their genomes. We wanted to test whether possession of a nuclear localization signal alone is sufficient to invoke motor-mediated movement along microtubules. To this end we adopted the Agrobacterium VirE2-ssDNA complex as a large traceable object with a well-defined nuclear import function. In fact, wild-type VirE2 nuclear import is specific to plants, but a small mutation generates a nuclear localization signal that is recognized in animal cells. We compared both the movements of DNA complexes prepared with wild-type and mutated VirE2 in cytoskeletal networks reconstituted in Xenopus egg extracts. Using a statistical analysis of many traces we could clearly distinguish a difference in their movements. This difference was sensitive to microtubule depolymerization and to inhibitors of the dynein motor protein, but not to inhibitors of kinesin. In a radially-oriented array, minus end-directed motion would carry the particles toward the cell center.[publication page]

top

Optical tweezers for maskless surface patterning.

Optical tweezers are widely used in biophysical studies. We thought to use them to nucleate and direct chemical reactions. Although the basic principle of operation may differ, we found that the same instrument is able to write patterns of noble metals, transition metal dichalcogenide compounds, and solid carbon, all deposited from clear, homogeneous liquid precursor solutions. Most surprisingly, the carbon patters contained a high proportion of tetragonally bonded (sp3), diamond-like carbon. [publication page-I][publication page-II]

top

Dynamic Force Spectroscopy of Importin beta – RanGTP interaction Collaboration with Prof. Ziv Reich, WIS Dept of Biological Chemistry

We used the atomic force microscope to perform single-molecule interaction studies of the nucleocytoplasmic transport receptor importin beta with the regulatory protein Ran. We found evidence for two states of binding to Ran in the GTP form (using the non-hydrolyzable GTP analog GppNHp), with differing strengths of adhesion. External mechanical force shifts the population to one of these states at the expense of the other.

top