The trouble is, he says, “people weren't looking early enough.” They had seen the final result but not the movement itself. Now, this team gets reorientation going with LPA before initiating migration with serum. That allows them to catch the nucleus moving backward even as the MTOC stays put.
The GTPase Cdc42 and its target MRCK were necessary and sufficient for nuclear movement. These proteins prompt actin polymerization at the front of the cell and thus actin retrograde flow, whose timing and speed match that of nuclear movement. The team favors an actin conveyor-belt model for nuclear movement, but has not ruled out a bulldozer model.
The MTOC stayed fixed in the center of the cell thanks to a group of proteins including dynein, Par6 and PKCζ; without them it wandered backward with the nucleus. How the MTOC is centered is a mystery—some have suggested that the pull of motors on microtubules is proportional to the length of the microtubules.