Abstract

Cytokinesis is the process by which the cell divides into two daughter cells and is driven by the formation and ingression of an actin-myosin contractile ring, which is under the control of RhoA. The plane of division is dictated by the position of the mitotic spindle. In particular, central spindle-associated Ect2 (a GEF) generates active RhoA to form the contractile ring in the overlying cortex. We recently found that the conserved scaffold protein anillin interacts with Ect2 at the cortex, in addition to its many interactions with actin, myosin, septins and RhoA (and others). Anillin’s interaction with Ect2 may anchor central spindle microtubules at the cortex to activate RhoA in a discrete zone, in a feedback manner. Consistent with this, depletion of anillin or Ect2 results in the loss of cortically-localized central spindle microtubules. Furthermore, disrupting the central spindle by inhibiting Aurora B or by MKLP1 depletion results in the broad localization of contractile ring proteins. In addition, depletion of both MKLP1 and anillin prevents the establishment of an equatorial plane as contractile ring components spread around the entire cell.
These studies suggest that there is a strong relationship between anillin and microtubules, which may help define the division plane. We found that there is a strong negative relationship between astral microtubules and anillin’s cortical localization. Furthermore, we observed that anillin localization switches between the cortex and stable microtubules
iv
depending on the cell cycle stage or condition (i.e. drug treatment). In the absence of Rho signaling (Ect2 or RhoA RNAi), anillin localizes to central spindle microtubules. Anillin also strongly localizes to taxol-stabilized microtubules. This suggests that RhoA activation coupled with microtubule instability switches anillin from the microtubules to the cortex. Disrupting central spindle microtubules by inhibiting Aurora B or by MKLP1 depletion in taxol-stabilized cells causes anillin to re-locate to the cortex, and differs between discrete and broad depending on the treatment. Using deletion constructs, we mapped the possible microtubule-localizing region on anillin to the C-terminus. We hypothesize that there could be multiple sites where one could be Ect2-dependent to anchor cortical central spindle microtubules and the other could be Ect2-independent. We are still determining if the second interaction is direct. Recently, C. elegans ANI-1/anillin was shown to localize to astral microtubules where it removes myosin from the cortex and polarizes the early embryo. In addition, multiple Rho effectors, including mDia2, relocate to central spindle microtubules in the absence of Rho signalling in human cells. We propose that interacting with microtubules could be a conserved mechanism to keep Rho effectors off the cortex to help define cortical domains.