The phosphorylated version (light blue) of D-TACC (dark blue) helps focus microtubule nucleation at centrosomes.

Microtubule-stabilizing proteins are common at the growing plus end of filaments, but Barros et al. (page 1039) have evidence that such proteins may work at the minus end as well.Barros et al., and papers from Kinoshita et al. on page 1047, and Peset et al. on page 1057, aim to understand what controls the shift from a radial network of interphase microtubules to the bipolar astral and spindle microtubules (MTs) required for mitosis.

Previous studies showed that TACC (transforming acidic coiled-coil) proteins work with Msps, also called XMAP215, to stabilize MTs during mitosis. The details of the interactions and mechanism have been unclear.

The authors found that TACC binds to Msps in a one-to-one ratio. The complex had greater affinity for MTs than did either protein alone and more effectively blocked MT depolymerization.

With a mix of mutation analyses and in vitro assays, the researchers showed that Aurora A phosphorylates TACC and that phospho-TACC is restricted to the centrosome. In the absence of phospho-TACC, spindle microtubules were relatively normal but centrosomal microtubules were shorter than normal or entirely absent. Thus, each group concluded that the phosphorylation of TACC by Aurora A is a key factor in shifting MT polymerization toward the centrosome.

Peset et al. and Kinoshita et al. hypothesize that phospho-TACC/Msps works to stabilize nascent MTs at the centrosome. For example, concentrating the phospho-TACC at the minus end may somehow help load it onto the filaments and subsequently increase plus end stability.

Looking in fly embryos, however, Barros et al. saw that phospho-TACC concentrates just slightly away from the centrosome where the minus ends of the MTs reside, as well as in the centrosome itself. Thus, they propose that phospho-TACC, with its strictly limited localization, actually works on the minus ends themselves.