BRCA1-dependent ubiquitination of gamma-tubulin regulates centrosome number

Abstract

Proper centrosome duplication and spindle formation are crucial for prevention of chromosomal instability, and BRCA1 plays a role in this process. In this study, transient inhibition of BRCA1 function in cell lines derived from mammary tissue caused rapid amplification and fragmentation of centrosomes. Cell lines tested that were derived from nonmammary tissues did not amplify the centrosome number in this transient assay. We tested whether BRCA1 and its binding partner, BARD1, ubiquitinate centrosome proteins. Results showed that centrosome components, including gamma-tubulin, are ubiquitinated by BRCA1/BARD1 in vitro. The in vitro ubiquitination of gamma-tubulin was specific, and function of the carboxy terminus was necessary for this reaction; truncated BRCA1 did not ubiquitinate gamma-tubulin. BRCA1/BARD1 ubiquitinated lysines 48 and 344 of gamma-tubulin in vitro, and expression in cells of gamma-tubulin K48R caused a marked amplification of centrosomes. This result supports the notion that the modification of these lysines in living cells is critical in the maintenance of centrosome number. One of the key problems in understanding the biology of BRCA1 has been the identification of a specific target of BRCA1/BARD1 ubiquitination and its effect on mammary cell biology. The results of this study identify a ubiquitination target and suggest a biological impact important in the etiology of breast cancer.

FIG. 1.

Expression of BIF caused centrosome amplification in several tissue culture cell lines derived from mammary tissue. (A) The indicated cell lines either were infected with a control adenovirus expressing either GFP or LacZ (a, c, e, g, i, k, m, and o) or were infected with an adenovirus expressing BIF at an equal multiplicity of infection (b, d, f, h, j, l, n, and p). Twenty-four to 27 h after infection, cells were fixed and stained for DNA content (DAPI; blue) and with an antibody specific to γ-tubulin (a to n) (red). Centrosomes were also stained with an antibody specific to centrin (o and p) (red). (B) The cells from the experiments for which results are shown in panel A were counted, and the percentage of cells with amplified (more than two) centrosomes were scored. Results of one representative experiment are shown. The number of cells with more than two centrosomes was determined at least twice for each cell line, and at least 100 cells were counted in each experiment. The data were collected from γ-tubulin-stained cells for the first nine experimental pairs and from centrin-stained cells for the last pair on the right (Hs578T, centrin).

FIG. 2.

Reduction in BRCA1 protein levels by siRNA leads to amplified centrosomes in Hs578T cells. (A) Western blots for BRCA1 protein 48 h after transfection with BRCA1-directed or control siRNA. The same filter was restained with an antibody specific for the large subunit of RNA polymerase II (Pol II) in order to evaluate equal loading of samples. (B) Cells with more than two centrosomes were scored by fluorescence microscopy as for Fig. 1B. The histogram shows the percentage of cells with amplified centrosomes. (C) Cells were fixed and stained with an antibody specific to γ-tubulin (as for Fig. 1A) 48 h after transfection with BRCA1-directed siRNA (BRCA1-b oligonucleotide). Examples of mitotic Hs578T cells (a and b) and U2OS cells (c) are shown.

FIG. 3.

HU abolishes BIF-associated centrosome amplification in MCF10A cells. MCF10A cells were treated with 2 mM HU 3 h after infection with adBIF or adLacZ. Cells were fixed and stained for γ-tubulin at 12, 18, and 24 h postinfection (hpi). The histogram shows the percentage of cells with abnormal centrosome numbers.

FIG. 4.

BRCA1/BARD1-dependent ubiquitination of centrosome proteins. (A) Partially purified centrosomes were tested as a substrate for the BRCA1/BARD1 ubiquitin ligase activity. Ubiquitination reaction mixtures contained the centrosome fraction, E1, UbcH5c, biotinylated ubiquitin, full-length BRCA1/BARD1, and centrosomes as indicated. Ubiquitinated protein products were visualized by HRP-conjugated streptavidin and chemiluminescence. (B) Reaction products as in panel A were analyzed by Western blotting with anti-γ-tubulin. (C) In vitro ubiquitination reaction products, as above, were analyzed by Western blotting with anti-α-tubulin and anti-ɛ-tubulin as indicated. (D) In vitro ubiquitination reaction products containing either γ-tubulin or centrin-1, each expressed and purified from bacteria, were analyzed by Western blotting with anti-γ-tubulin or anti-centrin as indicated.

FIG. 5.

The in vitro modification of γ-tubulin by BRCA1/BARD1 is monoubiquitination. (A) Individual components of the ubiquitination reaction mixtures in Fig. 4B were singly omitted as indicated. Reaction products were analyzed by Western blotting with anti-γ-tubulin. (B) Ubiquitination reactions were carried out as for Fig. 4B; reaction products were affinity purified by magnetic streptavidin beads (lanes 3 and 4) and analyzed by Western blotting with anti-γ-tubulin. (C) Biotinylated ubiquitin, six-His-tagged ubiquitin, purified bovine ubiquitin, and six-His-tagged ubiquitin K48R were added to the ubiquitination reaction mixtures as indicated. Reaction products were analyzed by Western blotting with anti-γ-tubulin.

FIG. 6.

The function of the carboxy terminus of BRCA1 is necessary for ubiquitination of γ-tubulin in vitro. (A) Schematic of BRCA1, RNA helicase A, BIF, and control peptide. (B) Increasing amounts of purified BIF (lanes 3 and 4) or the same amounts of control peptide (lanes 5 and 6) were added to ubiquitination reaction mixtures. Reaction products were analyzed by Western blotting with anti-γ-tubulin. (C) BRCA1(1-1863)/BARD1, BRCA1(1-1852)/BARD1, BRCA1(1-1527)/BARD1, BRCA1(1-1000)/BARD1, and BRCA1(1-500)/BARD1 were analyzed by protein gel and visualized by staining with Coomassie brilliant blue. (D) The BRCA1/BARD1 complexes containing 15 nM (each) BRCA1 and BARD1, with BRCA1 of varying lengths, shown in panel C, were added to ubiquitination reaction mixtures as indicated and analyzed by Western blotting with anti-γ-tubulin.

FIG. 7.

BRCA1/BARD1 ubiquitinates lysines of γ-tubulin required for regulation of centrosome duplication. (A) Schematic of the ubiquitinated lysines of γ-tubulin found by mass spectroscopy. (B) Expression levels of Hs578T (a) or U2OS (b) cells transfected with epitope-tagged wild-type γ-tubulin (WT), γ-tubulin K344R, γ-tubulin K48R, or the double mutant γ-tubulin K48R K344R, as indicated, were analyzed by Western blotting with anti-HA. (C) Hs578T (a) or U2OS (b) cells were transfected with plasmids expressing the wild-type or mutant γ-tubulin and GFP-centrin. At 42 and 72 h posttransfection, cells were fixed and stained with DAPI. The centrosome content of cells was determined by counting the pinpoint GFP-centrin spots, and the percentage of cells with more than two centrosomes was determined. Each experiment was performed twice, and at least 100 cells were counted for each transfection. The histograms show the percentage of cells with extra centrosomes. (D) Examples of cells expressing GFP-centrin and the indicated γ-tubulin protein are shown with normal (a and c) or abnormal (b and d) centrosome content.