The demo below shows a solution of F-actin
in the nematic liquid crystalline phase, which looks transparent by naked
eye (upper left) but contain beautiful birefringence patterns when viewed
through a set of crossed polarizers (lower left). The remaining images show
that the solution becomes opaque when actin filaments form large bundles
induced by polylysine (center, up and down), and the aggregation can be
reversed by addition of ATP (right panels). The later properties will be
addressed later in this guided tour.
We discovered a few years ago that actin filaments at very high concentrations, such as 30 mg/ml, selfassemble to form giant granules as shown below. These intriguing granules have been called tactoids in the literature. We are using a number of optical and electron microscopy techniques to understand the detail structure of these tactoids, which hopefully will help understand the mechanism of such formations. We will potentially also look into biological systems that may demonstrate similar granular structures (such as actin patches in yeast and dense bodies in smooth muscle cells)
Some preliminary results have been publishd
as a meeting proceeding (Tang, J. X., R. Oldenbourg, P. G. Allen, and P.
A. Janmey. 1997. Tactoidal granules in concentrated actin gels: a solidlike
state of protein filaments. proceedings of the material research society
fall meeting, Symposium K: materials science of the cell. Click Here for a Full Description )
We have recently demonstrated the polyelectrolyte
nature of F-actin and have applied the basic concept of counterion condensation
(Manning, 1978; Oosawa, 1971) to explain the mechanism of cation-induced
actin bundle formation by a simple analogy to DNA condensation (Tang and
Janmey, 1996; Tang et al., 1996). The nonspecific nature of cation-induced
bundle formation has been further confirmed by the similar bundle formations
of several other anionic biopolymers including microtubules, bacteriophage
fd, tobacco mosaic virus, etc. Click here for a figure of large bundles.
One striking confirmation of the paradigm relating the phenomena of actin bundling and DNA condensation is the observation of actin rings, induced by a salt containing a high concentration of divalent cations, such as 50 mM MgCl2, or multivalent cations. These actin rings are similar to DNA toroids induced by cobalt hexamine or polyamines, which have been studied extensively by researchers interested in the physical chemistry of DNA. Below are snapshots of two undulating actin rings
Lateral aggregates of charged biological filaments can also be induced by other mechanisms besides electrostatic interactions, such as by high osmotic pressure. We have shown experimentally distinguished features between these two types of bundle formation, both of which are expected to play important roles in the cellular environment. Click here for a pair of images to see the resemblance between the two types of F-actin bundles.
The rheological properties of crosslinked networks of flexible polymer chains are well predicted by the classic theories of elasticity (Ferry, 1980; Flory, 1969), but theory of viscoelasticity for an entangled but uncrosslinked network of semiflexible filaments is under current development (Everaers et al., 1999; Gittes and MacKintosh, 1998). One representative experimental network of semiflexible filaments is a solution of F-actin. The large persistence length and filament length on the order of ten microns enable this unique kind of protein filaments to form a tightly entangled network at a concentration as low as 1 mg/ml. Much of the recent theoretical development has been aimed at predicting the rheological behavior of actin networks (MacKintosh et al., 1995; Maggs, 1997; Morse, 1998). In addition to their relevance to condensed matter physics, these studies combined with recent extensive experimental work on the gel strength of F-actin, provide a basis of understanding the role this essential protein plays in cell functions and morphology (Janmey et al., 1990; Janmey et al., 1994; Wachsstock et al., 1993). Our recent study along this line has helped resolve some puzzaling descrepencies in the literature. Additionally, we have performed rheologic measurements with other biopolymers in order to further test the existing theories.
Very recently, my graduate student Chungwein Lee has started a new project to measure the change in osmotic pressure due to actin polymerization. Using a homemade osmometer from a Japanese collaborator, Professor Tadanao Ito of Kyoto University, we have detected a remarkable drop in osmolarity accompanying actin polymerization. Currently, the custom designed osmometer is being duplicated and a systematic and extensive investigation will follow this new initiative.
Many actin binding and crosslinking proteins have been identified to play important cytoskeletal roles. The most notable one is perhaps the Arp2/3 complex which has generated enormous current interests among researchers in the cytoskeletal field. The Arps2/3 is a 7-protein complex that are capable of nucleating an actin filament from the pointed end, anchoring a nacent filament to an existing one and thus forming a wide angle branch. This process is now thought to be regulated by factors through the cytoskeletal signal transduction pathways.
Interestingly enough, a recent in vitro study by myself and colleagues (PDF file) has shown a peculiar type of actin filament branches via oxidation of a small percentage of actin. The figure below shows a set of fluorescence images of such crosslinks, which resamble to some extend the branched actin network found in vivo.
Two other classes of actin binding proteins have been widely studied. The focal adhesion proteins, including talin and vinculin, which are highly expressed in the adhesion sites of a cell to its substrate. These proteins not only help anchor cells to its substrate, but regulate shape changes and motility though a cascade of signaling events.
The other class of actin crosslinking proteins include filamin and a-actinin which exist in both muscle and nonmuscle forms. These proteins are traditionally thought to crosslink actin filaments into either large bundles or more complicated mashwork under the cellular membrane. Our past effort on studying these proteins focused on examining their effect on rheologic properties of actin network. Additional work will focus on their effects on osmotic properties of actin network.
Recently, it has been recognized that many
actin binding proteins pertain either one or two domains homologous to a
32k dalton smooth muscle actin binding protein calponin. It has also been
known for some time that calponin bundles actin filaments in vitro. In light
of the fact that calponin is a highly cationic protein, we proposed a few
years ago that calponin bundles F-actin primarily by an electrostatic mechanism.
We have demonstrated a number of features of calponin-F-actin interaction
in support of this hypothesis. Click
here for a visual demonstration of one of these features, namely the
extreme sensitivity of calponin-actin bundles to solution ionic strength.
Collaborations
This is an ongoing collaboration with Dr. Gerard Wong in the X-ray laboratory headed by Professor Cyrus Safinia at UCSB. Visit their home page for more information.
This project is an ongoing collaboration with Dr. Jyotsana Lal, staff scientist at the Argonne National Laboratory. We are currently planning to perform a systematic set of measurements with variable extent of deuteration in order to precisely determine the filament diameter of F-actin in solution, as well as the thickness of its counterion layer, at various ionic conditions. Visit the homepage of ANL for operation of their neutron scattering facilities, etc.
Our biophysics research lab is starting to collaborate with the cytoskeletal research group, including labs of Bill Saxton, professor of biology; Beth Raff, professor of biology; and Claire E. Walczak, assistant professor of medical sciences, IU medical school. We are also seeking potential collaborations with the labs of Alan Bender, associate professor of biology and David Delake, associate professor of medical sciences. Our approach to these collaborative efforts is to develop novel microscopy and imaging tecniques to suit not only our own needs of structure determination of cytoskeletal protein filaments, but also the needs of these molecular biology labs. Our current thoughts include developement of the following two setups:
1. High sensitivity epifluorescence microscope with low background total internal reflection technique (LBTIR), which allows visualization of single macromolecules with appropriate fluorescence tags. This cutting edge technology has been recently developed, and has witnessed rapid recognition in the biological community, especially among those interested in cell dynamics and motility. We are seeking external funding in order to develope such a setup, which until now have been used by only a few laboratories in the world.
2. Imaging protein filaments and complexes with an atomic force microscope (AFM). Mike Hosek, a staff scientist in the IU physics department has developed an AFM for imaging metal surfaces. We have now started a collaborative effort in an attempt to image actin filaments and actin bundles using the existing AFM. We will image other protein complexes of interests to our collaborating biology labs once the instrument has been modified and tested to achieve sufficient resolution.
All research projects are funded through NSF and NIH grants. Please click here to see annual reports for NSF grant: [year 1] [year 2]