Two cancer cell lines utilize Myosin 10 and the kinesin HSET differentially to maintain mitotic spindle bipolarity

Abstract

Cancer cells often undergo mitosis possessing more than two centrosomes. To avoid a multipolar mitosis, the consequences of which are typically aneuploidy induced senescence, they must cluster their extra centrosomes to create a pseudo-bipolar spindle. Such supernumerary centrosome clustering (SNCC) requires Myosin 10 (Myo10) and the pole-focusing kinesin HSET. We showed recently that Myo10 promotes SNCC in HeLa cells by promoting retraction fiber-based cell adhesion, and that it further supports spindle bipolarity by preventing the generation of extra spindle poles via pericentriolar material (PCM) fragmentation. Here we quantified the contribution that Myo10 and HSET make individually and together to SNCC and PCM/pole integrity in HeLa cells and in MDA-MB-231 cells, which differ from HeLa in being more dependent on SNCC and less dependent on retraction fiber-based cell adhesion. As expected, knockdown of Myo10 and HSET individually increased the frequency of multipolar spindles in both cell types. Their effects were surprisingly not additive, however. For HeLa and MDA-MB-231 cells undergoing mitosis with more than two centrosomes, the defect in SNCC was almost entirely responsible for their multipolar phenotype following knockdown of either Myo10 or HSET. For HeLa and MDA-MB-231 cells undergoing mitosis with two centrosomes, PCM/pole fragmentation was the primary cause of multipolar spindles following HSET knockdown. Unlike HeLa, however, MDA-MB-231 cells exhibited very little PCM/pole fragmentation following Myo10 knockdown. This difference may be due to the smaller role that Myo10 plays in retraction fiber-based adhesion in MDA-MB-231. Finally, we show that HSET knockdown disrupts retraction fiber formation and organization, which may explain why the defects in double knockdown cells were not significantly greater than in HSET knockdown cells. These and other results can inform efforts to target these two motor proteins to selectively kill cancer cells by increasing their frequency of multipolar divisions.

Fig 1. While Myo10 also localizes to the tips of mitotic retraction fibers in MDA-MB-231 cells, these structures are less organized and exhibit less colocalization between Myo10 and active

β1 integrin than retraction fibers in HeLa cells. (A1-A6) Representative bottom section images of interphase MDA-MB-231 cells stained for actin (Phalloidin) and Myo10. (B1-B4) Representative images (B1-B3: bottom sections; B4: middle section) of metaphase MDA-MB-231 cells stained for actin (Phalloidin), Myo10 and DNA (DAPI). (C1-C4) Representative images (C1-C3: bottom sections; C4: middle section) of metaphase HeLa cells stained for actin (Phalloidin), Myo10 and DNA (DAPI). (D) Representative images of a metaphase MDA-MB-231 cell stained for actin (Phalloidin), Myo10, open, active β1 integrin (9EG7), and DNA (DAPI) showing the degree to which Myo10 puncta at the tips of retraction fibers exhibit a signal for open, active β1 integrin (the arrows in the insets mark Myo10-positive retraction fiber tips that also exhibit the integrin signal). (E) Same as (D) except a representative metaphase HeLa cell. (F) Percent co-localization between the signals for Myo10 and open, active integrin (9EG7) at the tips of retraction fibers (from 15 cells each). The mag bar in A1 is 10 µm, and the mag bars in B1, C1, D and E are 5 µm.

Fig 2. Endogenous HSET localizes in HeLa cells to the nucleus during interphase, to spindle poles during prophase, and to the spindle from metaphase through telophase.

Representative images of Halo-Myo10 KI HeLa cells stained for HSET and DNA (DAPI) at interphase (A1-A5), prophase (B1-B5), metaphase (C1-C5), and telophase (D1-D5). Shown are middle sections and bottom sections. The white arrows in B3 mark the positions of the two spindle poles. All mag bars are 10 µm.

Fig 3. siRNA-mediated knockdown efficiencies for Myo10 and HSET individually and together in HeLa and MDA-MD-231 cells.

(A) Representative Western blot and knockdown efficiencies for Myo10 (Light Blue) and HSET (Peach) in HeLa cells from four independent experiments (see also S1 Table). (B) Same as (A) but for MDA-MB-231 cells and from three independent experiments. See also S1 Table.

Fig 4. Knockdown of Myo10 and HSET individually increases the frequency of multipolar spindles in both HeLa and MDA-MB-231 cells, but their effects are not additive.

(A) Percent of metaphase HeLa cells exhibiting bipolar, semi-polar, multipolar, or non-bipolar spindles in cells treated with control non-targeting siRNA (White), Myo10 siRNA (Pink), HSET siRNA (Tan), or both Myo10 and HSET siRNAs (Brown). (B) Same as (A) except for MDA-MB-231 cells. See also S2 Table.

Fig 5. Causes of spindle multipolarity in HeLa cells depleted of Myo10, HSET or both.

(A) Percent of multipolar HeLa cells treated with control non-targeting siRNA, Myo10 siRNA, HSET siRNA, or both Myo10 and HSET siRNAs that contained 2 centrosomes (Pink) or more than 2 centrosomes (Red). (B) Percent of multipolar HeLa cells with >2 centrosomes and treated with siRNAs as in (A) that exhibited centrosome de-clustering only (Blue), centrosome de-clustering plus acentriolar foci (Purple), centrosome de-clustering plus centriole disengagement (Green), or centrosome de-clustering plus acentriolar foci and centriole disengagement (Yellow). (C) Percent of multipolar HeLa cells with 2 centrosomes and treated with siRNAs as in (A) that exhibited acentriolar foci only (Purple), centriole disengagement only (Green) or acentriolar foci plus centriole disengagement (Yellow) (note that the control non-targeting siRNA was not scored because only ~3% of multipolar cells treated with this siRNA exhibited 2 centrosomes). (D) The efficiency of supernumerary centrosome clustering in HeLa cells for each siRNA (see text for details). All results are from four independent experiments. See also S4 Table.

Fig 6. Causes of spindle multipolarity in MDA-MB-231 cells depleted of Myo10, HSET or both.

(A) Percent of multipolar MDA-MB-231 cells treated with control non-targeting siRNA, Myo10 siRNA, HSET siRNA, or both Myo10 and HSET siRNAs that contained 2 centrosomes (Pink) or more than 2 centrosomes (Red). (B) Percent of multipolar MDA-MB-231 cells with >2 centrosomes and treated with siRNAs as in (A) that exhibited centrosome de-clustering only (Blue), centrosome de-clustering plus acentriolar foci (Purple), centrosome de-clustering plus centriole disengagement (Green), or centrosome de-clustering plus acentriolar foci and centriole disengagement (Yellow). (C) Percent of multipolar MDA-MB-231 cells with 2 centrosomes and treated with siRNAs as in (A) that exhibited acentriolar foci only (Purple), centriole disengagement only (Green) or acentriolar foci plus centriole disengagement (Yellow). (D) The efficiency of supernumerary centrosome clustering in MDA-MB-231 cells for each siRNA (see text for details). All results are from four independent experiments. See also S5 Table.

Fig 7. HeLa cells are more susceptible to PCM fragmentation than MDA-MB-231 cells following KD of either Myo10 or HSET.

Shown is the percent of total multipolar HeLa and MDA-MB-231 cells treated with control non-targeting siRNA, Myo10 siRNA, HSET siRNA, or both Myo10 and HSET siRNAs where the multipolar phenotype was due solely to acentriolar foci created by PCM fragmentation. The results are from four independent experiments. See also S6 Table.

Fig 8. Depletion of either Myo10 or HSET reduces the spindle pole content of Kizuna.

(A) Shown is the relative intensity of Kizuna staining at metaphase spindle poles in HeLa cells that had been treated with control non-targeting siRNA, Myo10 siRNA, HSET siRNA, or both Myo10 and HSET siRNAs and then stained for α-tubulin, Kizuna and DAPI. (B) Same as (A) except in MDA-MB-231 cells. Note that all of the samples were processed in parallel (i.e., fixing and staining) and imaged in parallel (i.e., identical imaging parameters), and that only bipolar cells were scored. See Materials and Methods for additional details. (C) Western blots showing the total cellular content of Kizuna in each of the samples. See also S11 Table.