Silencing of Tubulin Binding Cofactor C Modifies Microtubule Dynamics and Cell Cycle Distribution and Enhances Sensitivity to Gemcitabine in Breast Cancer Cells

Microtubules are highly dynamic structures composed of α/β-tubulin heterodimers and are involved in many cellular functions such as cell motility, cell division, intracellular transport, and cellular architecture (1). Microtubule dynamics are highly regulated in the cell by a variety of stabilizing and destabilizing molecules (2). Owing to their intense dynamicity during mitosis, microtubule constitute important targets for anticancer drugs (3). Microtubule are necessary for the execution and timing of mitosis and for the determination of the overall duration of the cell cycle (4). Microtubule nucleation involves centrosomes that play additional roles such as organization of cytosolic microtubules, formation of spindle microtubules during mitosis, and axonemes formation in ciliogenesis (5–7). In the absence of centrosomes, the spindle is assembled but cell cleavage is interrupted (8, 9). Centrosome duplication occurs during S-phase of the cell cycle and is regulated throughout the cell cycle to control their microtubule nucleation potential (7). In some breast cancer tumor cells, an accumulation of supernumerary centrosomes has been reported (10).

Formation of α/β-tubulin heterodimers requires a folding pathway that involves many protein chaperones and cofactors (11). Cytosolic chaperonins are highly expressed during cell growth especially from G₁‐S transition to early S-phase and are essential for the folding of actin, tubulin, and other cytosolic proteins (12). The 5 tubulin binding cofactors (TBC) A to E, act at different levels of this pathway to generate polymerizable heterodimers (13). Many of these cofactors have also been shown to play other roles in regulating microtubule stability and dynamics (14). TBCD present in the cytoplasm functions as a centrosomal protein critical to the recruitment of the γ-TuRC, Microtubule growth, and mitotic spindle formation. TBCD centrosomal binding partners remain unknown and further studies are necessary to elucidate the regulatory mechanisms behind the recruiting role of TBCD during interphase and mitosis (15). Little is known about the role of TBCs in cancer. It has been shown that the inhibition of TBCA in MCF7 and HeLa cells modified the structure of microtubules and caused cell cycle arrest in G₁ (16). TBCC is an essential and key player required for the formation of the tubulin heterodimers by promoting the assembly of the α- and β-tubulin peptides. A study on POR gene in Arabidopsis thaliana showed that it encodes an ortholog of TBCC and plays an important role in releasing competent polymerizable heterodimers (17, 18). Conversely, there are no data in the literature about the relationship between TBCC and the cell cycle or response to anti-microtubule agents in breast cancer cells. Here we report the consequences of TBCC silencing in the breast cancer line MCF7, in terms of tubulin pool distribution and microtubule dynamics, cell cycle distribution, and also tumorigenic and sensitivity to chemotherapeutic agents in vivo.

Human TBCC (NM_003192) was amplified with from hTerT-HME-1 human mammary epithelium cells (American Type Culture Collection) and the purified amplicon was cloned into pcDNA6 in the antisense orientation (designated as pcDNA6/C−) using EcoRI (Fermentas) as described previously (19).

MCF7 human mammary adenocarcinoma (American Type Culture Collection; no authentification was done by the authors) were cultured and transfected with pcDNA6/C− or empty pcDNA6 as previously described (19). Selection of stable transfectants was done with blasticidin (20 μg/mL; KN-1004, Euromedex). Three clones representative of each population were obtained: MP6.1, MP6.2, and MP6.3 correspond to control clones and MC-1, MC-2, and MC-3 correspond to low TBCC clones.

Protein extraction, Western blot, and densitometry quantification were done as described previously (19). The antibodies used were anti–βIII-tubulin (clone Tuj1, 1/2,500; Covalab), anti-p53 (clone DO7, 1/1,000; DAKO), anti-cyclin E (CC05, 1/500; Calbiochem Oncogene), anti-cyclin B1 (sc-245, 1/500; Santa Cruz Biotechnology), anti-p21 (sc-397, 1/500; Santa Cruz Biotechnology), anti-CDK2 (sc-748, 1/500; Santa Cruz Biotechnology), anti-Cdc2 p34 (sc-56261, 1/500; Santa Cruz Biotechnology), anti–β-actin (clone AC-15, 1/5,000), anti–α-tubulin (clone DM1A, 1/1,000), and anti–β-tubulin (clone 2.1, 1/1,000) from Sigma Aldrich. Polyclonal antibodies against TBCC (1/800) and TBCD (1/3,000) were generously provided by N. Cowan (New York University Medical Center) and that against Arl2 (1/1,000) was generously provided by R. Kahn (Emory University School of Medicine).

Cells (20 × 10⁶) were harvested, lysed in 200 μL of buffer (100 mmol/L Pipes, pH 6.7, 1 mmol/L EGTA, and 1 mmol/L MgSO₄) and polymerizable tubulin (PT), non-polymerizable tubulin (NPT), and microtubule fractions were obtained as previously described (19).

Cells (3 × 10⁵) were seeded in 6-well plate with circular glasses of 24 mm in their bottom and transfected with the pAcGFP1-tubulin vector (Clontech) using lipofectin (Invitrogen) following the manufacturer's instructions. Analysis of microtubule dynamics was done by time-lapse microscope as previously described (19). The experiment was conducted twice on 50 microtubules of the 2 cell lines MP6.1 and MC-1.

Cells were processed for immunofluorescence as previously described (19). Antibodies used were directed toward β-tubulin 1:100 (clone 2.1; Sigma Aldrich), or TBCC 1:30 (Abnova) or γ-tubulin 1:50 (clone GTU-88, Sigma Aldrich) followed by secondary fluorescein isothiocyanate (FITC)-antibody (Dako). DNA staining was done using diaminido-phenyl-indol (DAPI; Roche).

Cells were incubated with 10 μmol/L of BrdU (B5002; Sigma Aldrich) for 5, 15, 30, and 60 minutes. Cells were then collected and labeled with anti-BrdU (347583; Becton Dickinson) and propidium iodide for microscopic analysis following the manufacturer's instructions.

Cells (3 × 10⁵) were seeded in 35-mm cell culture dish in culture medium maintained at 37°C in a 5% CO₂ atmosphere, time-lapse microscope and mitotic analysis were done as previously described (19).

Cell proliferation was estimated using nonradioactive BrdU-based cell proliferation assay (Roche) according to the manufacturer's protocol after 48, 72, 96, and 168 hours. Cells (5 × 10³ cells) were seeded in 96-well plastic plates and exposed cells were incubated for 1 week with either 0.5 nmol/L or 1 nmol/L or 3 nmol/L of gemcitabine (Lilly). MC− and MP6 nomenclatures refer to an average value of the 3 clones of each of the batch populations.

The IC₅₀ of gemcitabine was defined as the concentration inhibiting proliferation to a level equal to 50% of that of controls. IC₅₀ values were determined by methylthiazoletetrazolium assay from concentration–effect curves generated using Microsoft Excel as previously described (20).

Desalted duplex siRNA targeting TBCC 5′-CUGAGCAACUGCACGGUCA-3′ and scrambled sequences were designed by Sigma-Aldrich. The siRNAs (200 nmol/L) were transfected into 25 × 10⁴ MCF7 cells using oligofectamine (Invitrogen) according to the manufacturer's protocol on 2 consecutive days. Western blot and flow cytometry experiment were done on the third day after transfection.

Cells were collected, incubatedfor 1 hour at 4°C with propidium iodide (0.05 mg/mL) solution containing Nonidet-P40 (0.05%), and analyzed using a FACS Calibur flow cytometer (BD Biosciences Europe) as previously described (19). For the siRNA's transfected clones, the cell cycle was studied 48 hours after the first transfection. Synchronization of cells at the beginning of S-phase was done by a double-thymidine block as previously described by Whitfield and colleagues (21).

Female CB17/severe combined immunodeficient (SCID) mice, purchased from Charles River Laboratories, were bred under pathogen-free conditions at the animal facility of our institute and treated as previously described (19). Animals were treated in accordance with the European Union guidelines for laboratory animal care and use, and the study was approved by the local animal ethical committee. All mice used were 5 to 6 weeks old at the time of cells' injections. Mice were divided into 6 groups of 6 mice injected with 6 clones. A total of 3 × 10⁶ cells were injected subcutaneously in mice with 50% matrigel (BD Biosciences). The 6 mice were divided into 2 groups of treated and untreated mice. In the treated groups, gemcitabine (120 mg/kg) was injected intraperitoneally on days 1, 4, 7, and 10. Tumor volumes were measured twice per week and computed with the formula: 4/3 × (3.14 × r³), where r is the mean radius.

The statistical significance of the in vivo and in vitro data was determined with a Student's t test. P < 0.05 and P < 0.001 indicate a statistically significant and highly significant, respectively.

Three clones obtained from MCF7 stably transfected with the empty vector were designated MP6.1, MP6.2, and MP6.3 and used as controls. Three stable clones of MCF7 cells with reduced expression levels of the protein TBCC (MC−) were established and named MC-1, MC-2, and MC-3 in increasing order of expression of TBCC (Fig. 1A). In some experiments, average results of the 3 low TBCC clones and the 3 control clones are globally represented in figures as “MC−” and “MP6.” Of note, the decrease of TBCC levels in MC− cells remained limited, suggesting that TBCC is a protein essential for cell survival and that a minimal content is required.

We have investigated the effect of low TBCC content on the protein content of total α- and β-tubulins in MC− cells. No difference was observed for the total α-tubulins between the MC− and MP6 cells, although the total β-tubulins content was slightly increased in MC− clones (Fig. 1A). The βIII-tubulin isotype content proved to be heterogeneous among control clones, precluding definitive conclusions about the impact of TBCC on this protein. We investigated the expression levels of TBCD and ADP-ribosylation factor-like 2 (Arl2), which are also involved in tubulin folding, and found that Arl2 was decreased in the 3 MC− clones whereas TBCD was not modified (Fig. 1A).

The different pools of tubulins were investigated in MP6.1 and MC-1 cells, the latter having been chosen because they have the lowest content of TBCC among the available clones. MC-1 cells contained a significantly larger pool of PT (confirmed both for α- and β-tubulin) although the nonpolymerizable pool seemed to be slightly reduced (Fig. 1B).

The distribution of β-tubulin in the cytoplasm was not affected by the modification in expression of TBCC in MC-1 cells, as evidenced by immunofluorescence conducted using an antibody directed against β-tubulin (Fig. 2A). Immunofluorescence with an antibody directed against TBCC showed diffuse cytoplasmic staining both in MC-1 and MP6.1 cells (Fig. 2B). However, we found a larger proportion of stained cells in MP6.1 cells than in MC-1 cells. No difference in γ-tubulin staining was observed (Fig. 2C).

The effects of TBCC silencing on microtubule dynamic instability were compared in MP6.1 and MC-1 cells 48 hours after transient transfection with pAcGFP1-α-tubulin. Dynamic parameters that were significantly modified included the mean rates, lengths of growth and shortening, and the mean duration in pause. Mean duration of growth or shortening as the frequencies of rescue and catastrophe were not significantly different between the 2 types of cells. The overall dynamicity was increased by 44% in MC-1 cells in comparison to control cells (Table 1).

The proliferation rates of MC− and MP6 cells in vitro were evaluated using the BrdU incorporation assay up to 1 week. MC− cells were found to proliferate slightly yet significantly faster than MP6 cells (Fig. 3A). To investigate the in vivo growth capacity of MCF7 models, we injected the 6 clones MP6.1, MP6.2, MP6.3, MC-1, MC-2, and MC-3 into SCID mice and monitored their in vivo tumor formation. Although all clones generated tumors, the tumors derived from MC− clones grew significantly faster (P < 0.0001) than those derived from the MP6 clones (Fig. 3B).

Cell cycle analysis by propidium iodide in the 6 clones (Fig. 4A) showed that MC− cells had a significantly higher percentage of cells in S-phase (20.04 ± 0.68) as compared with MP6 cells (16.7 ± 0.28; P < 0.05). This observation was confirmed when MP6.1 cells were transiently exposed to siRNA directed against TBCC, and displayed profound reduction of TBCC content (Fig. 4B). In cells exposed to anti-TBCC siRNA, we also observed a significant decrease in the G₂-M fraction.

We have investigated the effect of low TBCC level on cell cycle–related proteins in MC-1 cells at baseline and after a 24-hour exposure to 1 nmol/L gemcitabine. At the basal level, cyclin-dependent kinase 1 (CDK1) and cyclin B, involved in the G₂-M transition checkpoint, and CDK2 and cyclin E, involved in the S-phase checkpoint, were less expressed in MC− than in MP6.1 cells. In addition, P21, an inhibitor of cyclins/CDKs and its activator P53 were also downregulated in MC-1 cells with respect to MP6.1 cells (Fig. 4C). After exposure to gemcitabine, MP6.1 cells responded by increasing the expression levels of all checkpoint-related proteins whereas the gemcitabine sensitive MC-1 cells responded by increasing the levels of CDK1 and cyclin B but not CDK2 or cyclin E. P53 and P21 were not increased in MC-1 cells as in the case of MP6.1 cells (Fig. 4C). These results were confirmed by quantification (Fig. 4D).

Given the impact of TBCC on cell cycle distribution, we assessed the response of TBCC models to cell cycle–targeting agents, such as S-phase and G₂-M phase–targeting compounds. We investigated the in vitro response of the cells to gemcitabine in vitro, an S-phase–specific antimetabolite. We incubated our cells for 1 week with 0.5, 1, and 3 nmol/L of gemcitabine and found that at these 3 concentrations, MC− cells presented significantly higher sensitivity to gemcitabine compared with the MP6 cells as shown by reduced BrdU incorporation (Fig. 5A). When we assayed the cytotoxicity of gemcitabine by methylthiazoletetrazolium after 72 hours exposure, the IC₅₀ values for the MC−, and MP6 cells were found to be 0.04 and 1 μmol/L respectively (Fig. 5B). MC− cells were 25-fold more sensitive to gemcitabine than the control cell line. To investigate the in vivo response of MC− and MP6 clones to gemcitabine, mice injected with the 6 different clones were divided into control groups and groups exposed to gemcitabine. The growth of the clones in control groups presented the same profile as in the first experiment (Fig. 3B). Furthermore, the antitumor activity of gemcitabine was markedly stronger in mice injected with MC− clones than in mice injected with MP6 clones (Fig. 5B). The results presented as MP6 and MC− are the mean values of the volumes of the tumors formed by the 3 clones of each group.

In vivo analysis of sensitivity to paclitaxel did not show a clear difference between MC− and MP6 xenografts (Supplementary Fig. 1).

Analysis by time-lapse microscopy showed that the total duration of mitosis was significantly shorter in MC-1 cells than in MP6.1 cells (Fig. 6A). This difference was mainly attributable to shorter prophase-metaphase duration, whereas there was no difference in the duration of anaphase-telophase. Once released from a double-thymidine block, synchronized MP6.1 and MC-1 cells progressed differently through the cell cycle. MC-1 cells completed S-phase in 8 hours whereas MP6.1 cells required 12 hours (Fig. 6B). This was also observed with short pulse incubations with BrdU, in which MC-1 cells presented higher percentages of cells in S-phase starting 5 minutes till 1 hour (Fig. 6C).

Research on microtubules in cancer cell biology and sensitivity to microtubule-binding agents has essentially focused on polymerized tubulin (microtubules) or the equilibrium between PT dimers and microtubules (22). Our results show for the first time that the content of the tubulin chaperone protein TBCC has a profound influence on tubulin pools and microtubule dynamics. We found that the NPT pool varied proportionately to TBCC content. In coherence with this observation, cells with reduced TBCC content displayed increased microtubule dynamics. These results are compared with our previous observations that cells with increased TBCC content contain less PT and display reduced microtubule dynamics (19). Overall, these data underline the importance of fine-tuning essential chaperone activities such as that of TBCC and the consequences of small differences in the content of these key proteins on the microtubule cytoskeleton. It is likely that the relatively small differences in TBCC content relative to controls observed in our antisense clones were because of the fact that profound depletion of TBCC is incompatible with cell survival.

Reduced TBCC content in stably transfected cells was associated with a larger proportion of cells in S-phase of the cell cycle, more rapid passage through S-phase, and shorter mitosis due to shorter prophase-metaphase. During prometaphase, the dynamicity of spindle microtubules is important to probe the cytoplasm and attach to chromosomes at their kinetochores (23). It is therefore possible that increased microtubule dynamicity contributed to reduced duration of mitosis in cells with reduced TBCC content. Of note, silencing of TBCC with siRNA, which was more pronounced than that observed in stable clones, also increased the proportion of cells in S-phase while significantly reducing the proportion of cells in G₂-M phase. Analysis of cellular checkpoints proteins showed a reduced content of several key cell cycle regulators. Cyclin B and CDK1 have been shown to be important regulators of mitotic spindle function (24, 25). The reduced level of p21 observed in cells with low TBCC content is likely to be involved in the increased passage in S-phase. In the scope of this study, we have not determined whether altered tubulin pools and/or enhanced microtubule dynamics were mechanistically involved in the downregulation of cell cycle checkpoints. Given the physical interactions previously shown between microtubules and several cell cycle regulators (26–28), it is possible that enhanced microtubule dynamics favors the turnover and degradation of these cell cycle regulators. Alternatively, TBCC may have properties other than the folding of tubulin peptides and may influence cell cycle checkpoints independently of the microtubule cytoskeleton. TBCB, another chaperone involved in tubulin peptide folding, has been described to bind to and activate Pak1, a p21-activated kinase involved in the regulation of microtubule dynamics (29).

Silencing of TBCC in MCF7 was associated with enhanced tumor growth in vivo compared with control cells. This is coherent with our previous observation that increased TBCC content reduced tumor growth in vivo and that TBCC expression was inversely correlated with in vitro invasive capacity in a panel of 13 human breast cancer cell lines (19). Although enhanced tumorigenesis is in keeping with the observation that MC− cells proliferate more rapidly in vitro, it is possible that various mechanisms contribute to this phenotype. Of note, cells in which TBCC was silenced also contained less Arl2 protein, involved in tubulin folding (30). We have previously shown that cells with reduced Arl2 generated more aggressive tumors in SCID mice, therefore suggesting that Arl2 modifications in TBCC-silenced cells may contribute to the tumor aggressivity phenotype (31).

Reduction of TBCC content was associated with enhanced sensitivity to gemcitabine, an S-phase–specific compound, but not to microtubule-targeted agents, both in vivo and in vitro. Increased sensitivity to gemcitabine could be explained by the fact that reduced TBCC content was associated with a higher percentage of cells in S-phase. To our knowledge, this is the first report describing a relationship between a cytoskeletal-related protein and sensitivity to gemcitabine, a compound which is widely used in the treatment of solid tumors. This observation raises the possibility that the level of expression of TBCC in tumor samples could be correlated with sensitivity to gemcitabine in the clinic. Studies of TBCC in samples of gemcitabine-treated patients are thus warranted. Conversely, MC− cells did not show altered sensitivity to the anti-microtubule agents vinorelbine or paclitaxel in vitro or to paclitaxel in vivo in spite of enhanced microtubule dynamics (Supplementary Fig. 1). This lack of differential sensitivity may be because of the similar proportion of cells in G₂-M phase of the cell cycle. This observation is in contrast to that of Goncalves and colleagues (32) who reported increased microtubule dynamicity in paclitaxel-resistant cells. In the models described by these authors, however, dynamicity was more strongly increased than in our model, with differences of 57% to 167% in comparison with controls. It is possible that sensitivity to microtubule-targeting agents is dependent both on the variation of specific parameters of microtubule dynamics, such as growth, shortening or catastrophe, and on the degree of variation of this parameter.

In conclusion, our results suggest that TBCC protein content significantly influence tubulin pools, microtubule dynamics, cell cycle distribution, mitotic duration, and proliferation and tumorigenicity. Additional studies are required to determine the role of altered microtubule dynamics on cell cycle distribution and tumor aggressivity, and the mechanism through which TBCC silencing results in altered cell cycle checkpoints. While confirming the major role of the microtubule cytoskeleton in cancer cell biology, our results also emphasize the need to take into account all stages of the microtubule life cycle, including the very initial stages of tubulin peptides folding.

C. Dumontet received a major commercial research grant from SANOFI.

R. Hage-Sleiman benefits from financial support from the Lebanese CNRS.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.