BAP1 loss induces mitotic defects in mesothelioma cells through BRCA1-dependent and independent mechanisms

Abstract

The tumour suppressor BRCA1-associated protein 1 (BAP1) is the most frequently mutated cancer gene in mesothelioma. Here we report novel functions for BAP1 in mitotic progression highlighting the relationship between BAP1 and control of genome stability in mesothelioma cells with therapeutic implications. Depletion of BAP1 protein induced proteasome-mediated degradation of BRCA1 in mesothelioma cells while loss of BAP1 correlated with BRCA1 loss in mesothelioma patient tumour samples. BAP1 loss also led to mitotic defects that phenocopied the loss of BRCA1 including spindle assembly checkpoint failure, centrosome amplification and chromosome segregation errors. However, loss of BAP1 also led to additional mitotic changes that were not observed upon BRCA1 loss, including an increase in spindle length and enhanced growth of astral microtubules. Intriguingly, these consequences could be explained by loss of expression of the KIF18A and KIF18B kinesin motors that occurred upon depletion of BAP1 but not BRCA1, as spindle and astral microtubule defects were rescued by re-expression of KIF18A and KIF18B, respectively. We therefore propose that BAP1 inactivation causes mitotic defects through BRCA1-dependent and independent mechanisms revealing novel routes by which mesothelioma cells lacking BAP1 may acquire genome instability and exhibit altered responses to microtubule-targeted agents.

Fig. 1. BAP1 regulates BRCA1 protein expression in cell lines and patient tumours.

A MSTO-211H and NCI-H2452 cells were either mock-depleted or depleted with siRNAs against BRCA1 or BAP1 for 72 h, or by induction of an shBRCA1 sequence for 48 h. Cell lysates were analysed by Western blot with antibodies against BRCA1, BAP1 and α-tubulin. B MSTO-211H and NCI-H2452 cells were either mock-depleted or depleted with siRNAs against BAP1 for 72 h before being either untreated or treated with 20 µM MG132 for 6 h as indicated. Cell lysates were analysed by Western blot with antibodies against BRCA1, BAP1 and α-tubulin. C Representative immunohistochemical staining of BAP1 (left panels) and BRCA1 (right panels) on malignant mesothelioma patient tissues from the MEDUSA cohort (n = 26). Representative images from samples scored as positive or negative are shown at low (upper panels) and high (lower panels) magnification Scale bars are 500 µm in top panels and 100 µm in lower panels. D Pie chart showing the percentage of human mesothelioma tumours that stained negative or positive for protein expression of BAP1 and BRCA1. The p value indicates the significance of correlation for expression. E Cell extracts prepared from the primary human mesothelioma cell lines indicated were analysed by Western blot with antibodies against BRCA1, BAP1 and α-tubulin.

Fig. 2. BAP1 depletion leads to mitotic progression defects and loss of SAC integrity.

A MSTO-211H and NCI-H2452 were either mock-depleted or depleted of BAP1 or BRCA1 as described in Fig. 1A. Cells were fixed and stained with Hoechst 33258 to observe the DNA and histograms show the percentage of mitotic cells in the different mitotic phases indicated. B Cells treated and stained as in A, were analysed to measure the percentage of metaphase cells with misaligned chromosomes (means ± S.D.) based on n = 3 independent experiments (60 mitotic cells for each condition). C, D The percentage of interphase cells treated as in A that were multinucleated (C) or contained micronuclei (D) were scored. E Cell lysates treated as in A were analysed by Western blot with antibodies against of MAD2L1 and α-tubulin. F Cells treated as in A were processed for immunofluorescence microscopy with BUBR1 antibodies and DNA stained with Hoechst 33258. Histograms show the mean intensity of BUBR1 on metaphase chromosomes following BAP1 or BRCA1 depletion relative to that observed upon mock depletion. Data in A, B, C, D and F are expressed as means ± S.D. (n = 3).

Fig. 3. BAP1 or BRCA1 loss induces multipolar spindles and centrosome amplification.

A MSTO-211H and NCI-H2452 cells were either mock-depleted or depleted of BAP1 or BRCA1 as described in Fig. 1A before analysis by immunofluorescence microscopy with antibodies against γ-tubulin (green) and α-tubulin (red). Merge images include DNA stained with Hoechst 33258 (blue). B Histogram shows the percentage of cells with multipolar spindles. C MSTO-211H:shBRCA1 and NCI-H2452 cells were either mock-depleted or depleted of BAP1 or BRCA1 as described in Fig. 1A before analysis by immunofluorescence microscopy with antibodies against γ-tubulin (red) and C-Nap1 (green). Merge images include DNA stained with Hoechst 33258 (blue). D Histograms show the percentage of mitotic cells with amplified centrosomes. Scale bars in A and C, 6 μm.

Fig. 4. Loss of BAP1 but not BRCA1 causes reduced centrosome volume.

A MSTO-211H and NCI-H2452 cells were either mock-depleted or depleted of BAP1 or BRCA1 as described in Fig. 1A before analysis by immunofluorescence microscopy with antibodies against γ-tubulin (green) and CDK5RAP2 (red). Merge images include DNA (blue) stained with Hoechst 33258. Magnified views of centrosomes are shown. B Dot plots show the γ-tubulin and CDK5RAP2 volume in mock and siBAP1 depletion in MSTO-211H and NCI-H2452 interphase cells taken from A. C Cells treated as in A were stained with γ-tubulin (green) and CDK5RAP2 (red) antibodies. D Dot plots represent the γ-tubulin and CDK5RAP2 volume in metaphase cells taken from C. Volume measurements were performed using Imaris 3D on n = 30 cells. Scale bars in A and C, 5 μm.

Fig. 5. Loss of BAP1 but not BRCA1 causes increased spindle microtubule length.

A MSTO-211H and NCI-H2452 cells were either mock-depleted or depleted of BAP1 or BRCA1 as described in Fig. 1A before analysis by immunofluorescence microscopy with antibodies against γ-tubulin (green) and α-tubulin (red); DNA (blue) is stained with Hoechst 33258. Merge images include lines drawn from pole-to-pole from which measurements of spindle length were made (yellow). B Dot plots show the spindle length based on images shown in A. Quantification of spindle lengths is taken from three independent experiments. For each condition, 30 mitotic cells were scored. C Detachment of poles from spindle microtubules was observed in some cells depleted of BAP1, but not in mock- or BRCA1-depleted cells. Cells were stained as in A. Magnified images (right hand panels). D Cells depleted of BAP1 or BRCA1 as in A were then incubated with STLC for 6 h to generate monopolar spindles. Cells were stained with α-tubulin antibodies (green); DNA is stained with Hoechst 33258 (red). Histogram shows the length of microtubules quantified from the centre of the monopolar aster to the microtubule tips. E Representative images of cells treated and stained as in D. Scale bars in A, C and E, 8 μm.

Fig. 6. BAP1 depletion leads to loss of KIF18A and KIF18B.

A MSTO-211H and NCI-H2452 cells were either mock-depleted or depleted of BAP1 or BRCA1 as described in Fig. 1A before analysis by immunofluorescence microscopy with antibodies against α-tubulin (green) and KIF18A (red). Merge images include DNA (blue) stained with Hoechst 33258. B Histograms show KIF18A intensity at kinetochores relative to mock-depleted cells (100%) quantified from images shown in A. C MSTO-211H cells were mock-depleted or depleted of BAP1 for 48 h prior to transfection of GFP-KIF18A for 24 h as indicated. Cells were fixed and stained for GFP (red) and α-tubulin (green). Merge images include DNA (blue) stained with Hoechst 33258. D Histogram shows the spindle length as quantified from images shown in C. E MSTO-211H cells were mock-depleted or depleted of BAP1 for 48 h prior to transfection of GFP-KIF18B for 24 h. Astral microtubule (MT) volumes were determined in the area from the pole towards the cell cortex away from the main spindle using Imaris 3D. F MSTO-211H and NCI-H2452 cells were either mock-depleted or depleted with the siRNAs indicated against BAP1 for 72 h before extracts were prepared and analysed by Western blot with antibodies against BAP1, KIF18A, KIF18B and α-tubulin. Scale bars in A and C, 5 μm.

Fig. 7. BRCA1-dependent and independent mitotic functions of BAP1.

The schematic model illustrates how loss of BAP1 promotes mitotic errors and genome instability through BRCA1-dependent and -independent mechanisms. Loss of BAP1 expression is shown to lead to loss of BRCA1 on the left and loss of KIF18A and KIF18B on the right. Mitotic phenotypes associated with each pathway are indicated.