In eukaryotic cells, the mitotic spindle is positioned perpendicular to the axis of cell division during mitosis. Reporting in EMBO Journal, Gachet et al. now propose a mechanism for coordinating spindle orientation and cell division in fission yeast.

The authors used a strain in which the α2-tubulin gene atb2 was tagged with green fluorescent protein (GFP) to study spindle dynamics in vivo. They had previously shown in vitro that by disrupting the actin cytoskeleton (using the actin inhibitor latrunculin), spindle orientation was inefficient and that anaphase onset — when the sister chromatids separate — was delayed. These findings were now confirmed in live cells.

When analysing the behaviour of the astral microtubules that spring from the two spindle pole bodies, Gachet et al. noted that spindle orientation normally requires the simultaneous binding of astral microtubules from both poles to a defined region on opposite sides of the medial cell cortex — which they named the astral microtubule interaction zone (AMIZ). However, in latrunculin-treated cells, the astral microtubules did not seem to contact the cell cortex and spindle positioning was inefficient.

Is astral-microtubule binding to the AMIZ essential for spindle orientation? All the data seem to point that way. The analysis of a mutant strain that is defective in astral-microtubule nucleation from spindle pole bodies showed that spindle orientation was inefficient and anaphase onset was delayed. A high proportion of misorientated spindles with a corresponding delay in anaphase onset were also seen in an actin-mutant strain, as well as in latrunculin-treated cells.

These data also confirmed that an actin-containing complex at the medial cell cortex, which is necessary for astral-microtubule interactions, is important for spindle rotation and anaphase onset. Moreover, the authors showed that astral microtubules interact with actin at the medial cell cortex via two type-V myosins, Myo51 and Myo52, both of which are necessary for spindle rotation.

To examine more precisely where astral microtubules interact with the cell cortex, the authors used a GFP fusion of Cdc15, a protein that locates exclusively to the cytokinetic actomyosin ring (CAR) in mitosis. This ring structure forms in the central plane of the dividing cell and coincides with the centre of the AMIZ. When being constricted, CAR directs the assembly of the division septum. Gachet et al. showed that astral microtubules first contacted the AMIZ and then seemed to move towards the CAR.

The anillin homologue Mid1 is required for the placement of the division septum in the centre of the cell and, therefore, for the central positioning of the CAR. In the absence of Mid1, spindles rotate but fail to stabilize, which delays anaphase onset.

So, Gachet et al. suggest a two-part mechanism for spindle positioning in fission yeast — spindle orientation requires a rotational force that is provided by the simultaneous binding of astral microtubules with the AMIZ at opposite sides of the cell cortex and, once correctly aligned, the interaction of astral microtubules from at least one spindle pole body with CAR stabilizes this position.