Dynein heavy chain asymmetry is lost in DYNLL1-, CHICA-, and HMMR-depleted cells. (A) GFP-dynein heavy chain (DYNC1H1) cells were transfected with control or CHICA 3′-UTR siRNA duplexes for 24 h, then left untransfected or further transfected with CHICA wild-type or the TQT mutant form for a further 48 h. The cells were then imaged using an Ultraview Vox spinning-disk confocal system. Images of metaphase cells were taken from a single z-plane at one time point. Exposure times were 100 msec for GFP-DYNC1H1 or mCherry-tagged CHICA using 50 or 30% laser power, respectively. (B) The cortical localization of dynein was scored and is plotted in the bar graph. Error bars show the SEM (n = 10 in each of two independent experiments). (C) A model or DYNLL1 function in spindle positioning and orientation. DYNLL1 is recruited to spindle poles by the triple TQT motifs in the CHICA-HMMR complex. HMMR contains an N-terminal microtubule-binding domain capable of targeting the complex to spindle microtubules. Because loss of DYNLL1 results in additional dynein at the cortex and DYNLL1 does not accumulate at the cell cortex, this model proposes that DYNLL1 is inhibitory to dynein targeting to the cortex. When a spindle pole carrying the DYNLL1–CHICA–HMMR complex approaches the cortex, this results in increased local concentration of DYNLL1 and removal of dynein from the cell cortex. Ultimately, this leads to a reduction in pulling forces on the mitotic spindle, which then fails to rotate to the correct position in response to extrinsic positioning cues.