The polymerization process forms filaments of MinD at one end of the cell, which sequester MinC from the middle of the cell, thus leaving the bacterial cell division protein FtsZ to do its job. An additional component, MinE, forms a cap on the MinCD crescent so that the inhibitor, MinC, cannot reach the central FtsZ.
These proteins must inhibit division at both ends, and they do so by oscillating from one end of the cell to the other every 50 seconds. In several existing models, self-assembly is a key part of this oscillation. Joe Lutkenhaus (University of Kansas, Kansas City, KS) has recently seen self-assembly of MinD on lipid vesicles, with diffraction patterns suggesting a regular structure.
But the Tufts team is the first to visualize MinD filaments directly. Assembly was dependent on ATP (MinD is an ATPase), and filaments were much longer and thicker when vesicles were added. MinE bound to the filaments and increased their length and width even further, but also led to bundle disassembly that was dependent on ATP hydrolysis. As the bundles disassembled, they frayed preferentially at one end.“MinE is clearly doing two things at one time—bundling the MinD filaments and turning them over,” says RayChaudhuri. In its negative role, MinE may both halt the progression of polymerization toward the cell center and begin chewing away at the existing filaments. As MinD is liberated from one end, it may polymerize at the only MinE-free site: the other end of the cell. These new MinD polymers then attract MinE, and the cycle begins again.
Although MinD is not quite a bacterial version of the microtubule, RayChaudhuri can see parallels. He plans to use motor-like assays to test whether MinE's depolymerization activity is polar. And he says that, like microtubules, “this system clearly evolved as a mechanism to search and explore cellular space.” ▪