(A) Okumura et al. built a system in which a light-sensitive protein module (AsLOV2; lilac: SsrA; blue) was expressed at the cortex (light grey), allowing the induced recruitment of nano (dark grey) fused to either NuMA (purple; top) or dynein (green; bottom). When a specific region in the cell cortex (gold circle) is illuminated, NuMA-nano binds to the light-sensitive module and is recruited to the cortex, which leads to the spindle moving (gold arrow) towards the illuminated region (top). When the same experiment is repeated by directly targeting dynein to the cortex with light, the spindle does not move (bottom). Thus, anchoring NuMA to the cortex is sufficient for spindle pulling, but anchoring dynein is not. (B) Okumura et al. also examined how wild type (WT) NuMA (left) and a mutant NuMA (right) are organized at the cortex. When the system was illuminated, WT NuMA formed clusters at the cortex, which led to the spindle moving (gold arrow) towards the illuminated region (top). However, when mutant NuMA (purple with orange star) that does not self-assemble is used, neither clustering nor spindle movement (bottom) occur, indicating that NuMA’s ability to cluster and ‘concentrate’ force at the cortex is required for spindle pulling. For simplicity the light-sensitive module is not shown in (B).