MinD (green) moves across an artificial membrane and is chased, caught, and detached by MinE (red), leaving protein-free troughs (dark bands) in their wake.
KRUSE/AAAS
In the run-up to bacterial cell division, oscillatory movements of Min proteins occlude both poles of the cell. MinD initially attaches at one end, but MinE binds and drives it off. MinD shuttles to the other end and rebinds, only to be driven off again by MinE, and so the cycle repeats. MinD carries MinC along with it, and MinC prevents membrane binding of a key contractile protein called FtsZ. With MinC guarding both ends of the cell, FtsZ can bind only at the middle.
Theoretical models had suggested that MinD and E might drive the oscillations themselves. To test this, Martin Loose, Petra Schwille (Technical University of Dresden, Germany), Karsten Kruse, and colleagues placed MinD on a lipid bilayer and then added MinE. Within an hour, periodic bands of MinD and E appeared, separated by protein-free troughs. The bands marched slowly across the membrane, mimicking the oscillations seen in the bacterial cell. “These two proteins alone are sufficient to create this spatiotemporal structure,” says Kruse.
And FtsZ alone is sufficient to constrict the cell. FtsZ links to a membrane tether, and in conjunction with almost a dozen other proteins, forms a contractile ring. Because the other proteins are mainly required for remodeling the cell wall, Masaki Ozawa, David Anderson, and Harold Erickson (Duke University, Durham, NC) wondered whether FtsZ could constrict the cell by itself. The team added FtsZ with a membrane-targeting sequence to purified liposomes. In the presence of GTP, FtsZ formed bands on the inside surfaces and constricted liposomes almost to the point of pinching off. The incomplete division may have been due to the extreme thickness of the liposome walls, Erickson says.
The work by the two teams brings closer a cell-free system for studying the events of cytokinesis. 
