As a minus end–directed motor, dynein walks along astral microtubules toward the spindle pole, and thus could reel in the spindle during anaphase. The favored model for this movement predicts that dynein first binds to the bud cortex, where it captures passing microtubules. From there it could pull the spindle toward the bud. However, at least in budding yeast, significant amounts of dynein have not been found at the cortex. Nonetheless, according to Pellman, “the model is so intuitively appealing that the supposition is that [dynein] must be there even though we cannot see it.” But his results now suggest that this is not the case.
The group looked at endogenous levels of dynein in budding yeast. Even a triple-GFP tagged version of dynein was not found at the cortex, but rather on microtubule plus ends. This localization depended on the yeast homologues of CLIP-170 and LIS1, proteins implicated in dynein function in mammalian systems. Dynein was also seen on the spindle poles, its expected end-point after its motility is activated. Disturbing dynein activity, by mutating either its ATPase domain or by inactivating the dynactin complex, caused the motor to accumulate at plus ends rather than moving to the spindle poles.Based on the localization results, Pellman suggests that “the microtubule brings its own tethering and motor device out to the membrane.” He proposes that dynein is inactive on plus ends until it is activated by local recruitment to membrane domains, as has been shown for at least one class of kinesins. Local transfer to the membrane (and thus activation) could be done by Num1p, a pleckstrin homology–containing protein, as dynein again accumulated at plus ends in the absence of Num1p. ▪