The pcDNA6/V5-His A plasmid was used to clone the 1 kb human TBCC cDNA (NM_003192) extracted from hTerT-HME-1 human mammary epithelium cells (ATCC). The mRNA was extracted using Trizol reagent (Invitrogen, Cergy Pontoise, France) following the manufacturer's instructions. Reverse transcription into cDNA was then done using Moloney leukaemia virus reverse transcriptase (Invitrogen, Cergy Pontoise, France) for 1 hour at 37°C as described in the manufacturer's manual. The cDNA obtained was then amplified by the full length TBCC forward GCCAATATGGAGTCCGTCAG and reverse CAACTGCTTAGTCCCACTGGA primers using a high fidelity polymerase according to the manufacturer's instructions (Jena Bioscience, Germany). The PCR conditions were 30 cycles of denaturation at 94°C for 1 min, annealing at 60°C for 1 min and elongation at 72°C for 1 min. Purified amplicon was subcloned into pGEMTeasy and thereafter into pcDNA6 in the sense orientation (designated as pcDNA6/C+) using EcoRI (Fermentas, France).

MCF7 (human mammary adenocarcinoma, ATCC) cells and transfectants were grown in DMEM supplemented with penicillin (200 UI/ml), streptomycin (200 μg/ml) and fetal bovine serum (10%) at 37°C in a humidified atmosphere containing 5% CO₂. Cells were transfected with pcDNA6/C+ or empty pcDNA6 using lipofectin (Invitrogen, Cergy Pontoise, France) following the manufacturer's instructions. The stable transfectants were obtained through blasticidin selection (20 μg/ml) (KN-1004, Euromedex, France). Cloning of the populations of cells was performed for each of the two batch populations and 3 clones representative of each population were selected for further characterization, on the basis of their differential levels of expression of the protein TBCC. The clones designated MC+1, MC+2 and MC+3 represent the clones overexpressing TBCC. The clones designated MP6.1, MP6.2 and MP6.3 represent the control clones.

Protein extraction and western blot analysis were performed as described previously 21. The antibodies used were anti β III-tubulin (clone Tuj1, 1/2500; Covalab, Lyon, France), anti p53 (clone DO7, 1/1000; Dako, Denmark), anti β-actin (clone AC-15, 1/5000), anti α-tubulin (clone DM1A, 1/1000), anti β-tubulin (clone 2.1, 1/1000), anti tyrosinated α-tubulin (clone TUB-1A2, 1/1000) and acetylated α-tubulin (clone 6-11B-1, 1/1000) from Sigma Aldrich (St Quentin Fallavier, France). The polyclonal antibodies against TBCC (1/800) and TBCD (1/3000) were generously provided by N. Cowan (New York University Medial Center, USA) and those against Arl2 (1/1000) and Glu-tubulin (1/1000) were generously provided by R. Kahn (Emory University School of Medicine, Atlanta, USA) and L. Lafanechère (Centre de Criblage pour des Molécules Bio-Actives, Grenoble), respectively. Expression levels of the proteins were standardized against the β-actin.

A desalted duplex siRNA targeting TBCC 5'-CUGAGCAACUGCACGGUCA-3' and its corresponding scrambled sequence were designed by Sigma-Aldrich (St Quentin Fallavier, France). The siRNAs (200 nM) were transfected into 25 × 10⁴ of MCF7, MP6.1 or MC+1 cells using oligofectamine (Invitrogen, Cergy Pontoise, France) according to the manufacturer's protocol on two consecutive days. Protein analyses by western blot or flow cytometry experiment were done on the third day after transfection.

Cell proliferation was estimated using the methylthiazoletetrazolium (MTT) and BrdU assay. For the MTT test, cells (13 × 10⁴) were seeded in a 6-well plate and incubated at 37°C. Every 24 h and for one week, MTT (3 mg) was added to each well of each plate. After 2 hrs of incubation at 37°C, supernatants were removed, formazan crystals solubilized with 3 ml of isopropanol-HCl-H₂O (90:1:9, v/v/v) and plates scanned. The absorbance was measured spectrophotometrically with a microplate reader (Labsystem Multiskanner RC) and the MTT values were obtained as subtraction of absorbances read at 540 and 690 nm wavelengths. The nonradioactive BrdU-based cell proliferation assay (Roche, Basel, Switzerland) was performed according to the manufacturer's protocol. Treated and untreated cells (5 × 10³ cells per well) were seeded in a 96-well plastic plate and the assay was performed after 48, 72, 96 and 168 hours. Treated cells were exposed to either 0.5 nM or 1 nM of gemcitabine (Lilly, IN, USA) for one week. BrdU incorporation into the DNA was determined by measuring the absorbance at 450 on an ELISA plate reader.

Cells were incubated 24 h with either 10 nM Paclitaxel (Bristol-Myers Squibb, New York, USA) or 1 nM vinorelbine (Pierre Fabre medicaments, Boulogne, France). Treated and untreated cells were then collected and incubated 1 hour at 4°C with propidium iodide (0.05 mg/ml) solution containing Nonidet-P40 (0.05%). Cells were analyzed using a FACS Calibur flow cytomoter (BD Biosciences Europe, Erembodegem, Belgium) and cell cycle distribution was determined using Modfit LT 2.0™ software (Veritysoftware Inc, Topsham, USA). For the siRNA's transfected clones, the cell cycle was studied 48 h after the first transfection.

Cells (3 × 10⁵) were seeded in a 35 mm cell culture dish, placed in culture medium maintained at 37°C in a 5% CO₂ atmosphere and observed using an inverted time lapse microscope (Olympus IX50) at the Centre Commun de Quantimétrie (Université Claude Bernard Lyon, France). Images were acquired every 2 minutes for 24 hours using a numerical CFW-1308M 1360X1024 camera (Scion, Frederick, USA) driven by ImageJ software (NIH, Bethesda, USA). 30 complete mitoses were analysed for each of the MP6.1 and MC+1 clones using ImageJ software (NIH, Bethesda, USA).

Cells (MP6.1 and MC+1) exposed or not to 10 nM of paclitaxel for 24 hours were fixed by 4% paraformaldehyde during 15 minutes at room temperature and permeabilized using a PBS-Triton X-100 0.1% solution. Non specific sites were blocked using a solution containing 0.1% bovine serum albumin and 1% fetal calf serum. Cells were incubated with either a 1:100 of an antibody against β-tubulin (clone 2.1, Sigma Aldrich) or a 1:30 dilution of a monoclonal antibody against TBCC (Abnova, Taiwan) followed by a secondary FITC-antibody (Dako, Denmark). DNA staining was performed using diaminido-phenyl-indol (DAPI) (Roche, Manheim, Germany).

Images were obtained using a laser scanning confocal TCS Sp2 DMRXA microscope x63 objective (Leica Microsystems; Wetzlar, Germany) at the Centre Commun de Quantimétrie (Université Claude Bernard Lyon, France).

Cells (20 × 10⁶) were harvested and lysed in 200 μl of buffer (100 mM Pipes, pH 6.7, 1 mM EGTA, and 1 mM MgSO₄) by two freeze-thaw cycle. Lysed cells were centrifuged at 12,000 × g for 15 minutes at 4°C. The supernatant was then ultracentrifuged (100,000 × g for 1 h at 20°C) and separated into a supernatant containing "soluble tubulins" and a pellet containing "microtubules". The microtubule fraction was resuspended in 100 μl of lysis buffer and 100 μl of the supernatant were incubated with 1 mM of GTP at 35°C for 30 minutes to allow tubulin polymerization then ultracentrifuged at 50,000 × g for 45 minutes at 35°C. The resulting pellet contained the "polymerizable tubulin" (PT) heterodimers and the supernatant contained "nonpolymerizable tubulin" (NPT) heterodimers, included tubulin peptides complexed with tubulin binding cofactors. The different fractions of tubulins were run on silver stained gels following manufacturer's recommendations (Amersham Biosciences AB, Sweden). After coloration, the single band observed at 55 kDa for the polymerizable tubulin heterodimers confirmed the success of purification (data not shown). The experiment was performed in triplicate using the two cell lines MP6.1 and MC+1. Densitometric quantification of western blots was performed with ImageJ software (NIH, USA).

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. The glasses containing the cells were placed in culture medium maintained at 37°C in a 5% CO₂ atmosphere in a time lapse inverted microscope (Olympus IX50) at the Centre Commun de Quantimétrie (Lyon, Université Claude Bernard, France). Cells were imaged with a numerical CFW-1308M 1360X1024 camera (Scion, Frederick, USA) driven by imageJ software (NIH, Bethesda, USA) using a 40× oil immersion lens (Zeiss, Göttingen, Germany). 30 pictures of the cell's microtubules were taken at 4 seconds intervals. The positions of the plus-ends of individual MT in peripheral lamellar regions of cells were tracked over time using ImageJ^® software and graphed using Microsoft^® Excel spreadsheet as position versus time to generate a 'life-history plot' for each MT. Growth and shortening rates and durations were derived by regression analysis. A difference of >0.5 μm between any two consecutive points was considered as a growth or shortening event. Transitions into depolymerisation or shortening are termed catastrophes, and transitions from shortening to growth or pause are called rescue. The catastrophe and rescue frequencies per unit time were calculated by dividing respectively the number of transitions from growth and pause to shortening and the number of transitions form shortening and pause to growth by the sum of the time in growth and pause. Dynamicity represents total tubulin exchange at the MT end and was calculated by dividing the sum of total length grown and shortened by the MT life span. The experiment was performed twice on 50 microtubules of the two clones MP6.1 and MC+1.

Female CB17/SCID mice purchased from Charles River Laboratories (Arbresle, France) were bred under pathogen-free conditions at the animal facility of our institute. Animals were treated in accordance with the European Union guidelines and French laws for the laboratory animal care and use. The animals were kept in conventional housing. Access to food and water was not restricted. All mice used were 5 to 6 weeks old at the time of cells injections. This study was approved by the local animal ethical committee. Mice were divided into six groups of six mice each which corresponds to the injections of MP6.1, MP6.2, MP6.3, MC+1, MC+2 and MC+3 cells. 3 × 10⁶ cells were injected subcutaneously in mice with 50% matrigel (BD Biosciences, Belgium). The six mice were divided into two groups of treated and untreated mice. In the treated groups, paclitaxel was injected intraperitoneally in a dose of 10 mg/kg on the same day and a week after. Mice were weighed and the tumor size was measured twice per week with an electronic caliper. The volume was then computed by considering the tumor as a sphere with the formula 4/3 (3.14 × r³), r as the mean radius. Animals were euthanized either when one of the diameters of the tumor exceeded 17 mm, or if any potential suffering of the animal was observed or if weight loss exceeded 10%.

Six stable clones of MCF7 cells overexpressing the protein TBCC (designated MC+) were established and characterized compared to clones of control cells transfected with the empty vector (designated as MP6). Among these, three clones with high levels of TBCC, with respect to three clones of control cells, were named MC+1, MC+2 and MC+3 in decreasing order of expression and further explored (Figure 1). Three clones obtained from MCF7 transfected with the empty vector were designated MP6.1, MP6.2 and MP6.3 and used as controls.

We have investigated the effect of TBCC overexpression on the protein content of TBCD and ADP-ribosylation factor-like 2 (Arl2) which are two essential partners of TBCC in the folding pathway of α/β-tubulin dimers. TBCD was increased in the MC+1 clone but not in the two other clones MC+2 and MC+3. Conversely Arl2 was increased in the three clones of MC+ with respect to the three control clones of MP6 (Figure 1). Finally, we studied the expression level of the tumor suppressor protein p53 and found that it was slightly increased in the three clones of MC+ cells when compared to MP6 cells (Figure 1).

When we tested by MTT assay the proliferation rate of the MC+1 and MP6.1 cells in vitro, we found that the MC+1 cells had a slightly higher proliferation rate in vitro than the control cells at 120 and 144 hours but not at earlier timepoints. This result was visually confirmed by comparing the number of dividing cells stained in blue due to the formation of formazan crystals by their active mitochondria. These observations were confirmed by quantification of the absorbance optical density and the curve plotted (Figure 2A) shows the significant differences between these two cell lines. These results were confirmed on the MC+2, MC+3, MP6.2 and MP6.3 clones. When we tested the proliferation rate of MC+1 and MP6.1 cells using the BrdU incorporation assay, we found no significant difference in the proliferation rate between the two clones. This was confirmed on the other clones MC+2, MC+3, MP6.2 and MP6.3 (Figure 2B).

To explore if the difference in proliferation rates is correlated to a difference in the cell cycle distribution, we investigated the percentage of cells in S and G2-M phases by propidium iodide and flow cytometry in MC+1 and MP6.1 cells (Figure 2C). We observed that MC+1 cells had a lower percentage of cells in S-phase 11.7 ± 2.6 as compared to MP6.1 cells 16.7 ± 0.78, a significant decrease of 30%, p < 0.05 (Figure 2C). This difference was also accompanied by a higher percentage of cells in G2-M phase 45 ± 3.9 as compared to the MP6.1 30 ± 1.27, a significant increase of 50%, p < 0.05 (Figure 2C). This experiment was performed on the MP6.2, MP6.3, MC+2 and MC+3 clones and the above results were confirmed.

In order to correlate these phenotypic behaviours to the increased content of TBCC protein, we inhibited the protein expression of TBCC by transient transfection using siRNA targetting TBCC in MP6.1 and MC+1 cells. The siRNA targetting TBCC we designed was validated by western blot on MCF7 cells. This sequence inhibited TBCC (-48%, p < 0.01) compared to the siRNA scrambled (data not shown) and was used to inhibit TBCC in MC+1 and MP6.1. In both the MC+1 and MP6.1 cells, the inhibition of TBCC protein caused a significant increase in percentage of cells in S-phase and a significant decrease in percentage of cells in G2-M phase (Figure 2C).

To explain the increased G2-M percentage observed in MC+1 cells, we investigated the durations of prophase-metaphase, anaphase-telophase and complete mitosis in comparison to controls. We found that MC+1 cells took more time to complete mitosis (71 minutes as compared to 57 minutes in MP6.1 cells). This longer time was due mainly to longer anaphase-telophase duration as 37 minutes compared to 20 minutes in MP6.1 and not to a difference in the prophase-metaphase (Figure 2D). The cells overexpressing TBCC proceed slowly through the mitosis which can be an explanation for the high percentage of cells in G2-M phase of cell cycle.

In order to study the impact of TBCC on tumorigenesis of breast cancer cells, we injected the six clones MP6.1, MP6.2, MP6.3, MC+1, MC+2 and MC+3 into SCID mice and monitored their in vivo progression and tumor formation. The control clones were able to form tumors of 200 to 250 mm³one month later. However, the MC+ clones presented lower capacities of growth in vivo with the tumors of volumes ranging form 1 to 50 mm³ (Figure 3).

The impact of TBCC overexpression on the cytoplasmic distribution of TBCC was studied by immunofluorescence on MP6.1 and MC+1 cells. In a first step we observed that the distribution of β-tubulins in the cytoplasm was not affected by the overexpression of TBCC. This was done by immunofluorescence using an antibody against β-tubulin and the images show similar cytoplasm staining between MP6.1 and MC+1 cells (Figure 4A). TBCC was similarly distributed in both normally and highly-expressing cells. The staining was found to be cytoplasmic however it was not observed in all cells of each cell line (Figure 4B). We found more stained cells in MC+1 cells than in control cells. After overlaying fluorescent images of TBCC with DAPI images we observed that the cells that were strongly stained for TBCC were cells in mitosis. This result suggested that TBCC was more highly expressed in cells undergoing mitosis. This was confirmed (Figure 4C) in MC+1 cells exposed to paclitaxel 10 nM for 24 hours. Paclitaxel is known to stabilize microtubules and causes cell arrest in metaphase. We observed by immunofluorescence microscopy that the cells blocked in mitosis were strongly stained with anti-TBCC antibody.

We studied the impact that TBCC overexpression might have on the total α and β-tubulins as well as on different posttranslational modifications of α-tubulin such as acetylated, tyrosinated and detyrosinated (Glu) α-tubulins and finally on the βIII-tubulin isotype. We observed no significant difference in the expression of total α-tubulin and tyrosinated α-tubulin between the clones overexpressing TBCC and the controls (Figure 5A). Regarding Glu α-tubulin and the βIII-tubulin, we observed heterogeneity of expression among both the clones of MP6 and MC+ which made impossible to conclude of a simple profile related to TBCC status (Figure 5A). However, concerning β-tubulin and acetylated α-tubulin we observed an increase in expression in the MC+ clones with respect to the MP6 clones.

In addition, we investigated the different pools of soluble polymerizable αβ-heterodimers (PT) and soluble non-polymerizable tubulins (NPT) as well as the microtubule tubulins (MT) in MC+1 and MP6.1 cells. The different pools of tubulins were obtained after fractionation of MP6.1 and MC+1 lysates by series of ultracentrifugations. We have observed that the modification of the TBCC level in MCF7 cells did not alter the total pool of α-tubulin protein although it increased the total pool of β-tubulin (Figure 5B). The overexpression of TBCC protein strongly increased the NPT fraction and strongly decreased the PT fraction, while it had only a minor effect on the tubulin content of microtubules which was slightly decreased. The profile of expression of the NPT and PT fractions was similar for both the α-tubulins and the β-tubulins but was less marked in the case of α-tubulins (Figure 5B). Quantification of expression levels of α and β-tubulins in the different fractions confirmed the results observed in the immunoblots (Figure 5C).

The effects of overexpression of TBCC on microtubule dynamic instability were determined in MC+1 cells 48 h after transient transfection with pAcGFP1-α tubulin. The microtubules in MC+1 cells grew and shortened more slowly and for shorter lengths than those in MP6.1 cells and presented significantly less dynamicity. The parameters for dynamic microtubules in MP6.1 and MC+1 cells were determined (Table 1). The mean growth rate was significantly decreased from 16.41 ± 1.99 μm/min in MP6.1 cells to 10.54 ± 0.37 μm/min in MC+1 cells, corresponding to a decrease of 36%, p < 0.05. The mean shortening rate and the mean lengths of individual growth and shortening were slightly decreased. The mean shortening duration was slightly increased compared to the mean growth duration and mean pause duration that were significantly increased from 0.15 ± 0.02 min and 0.25 ± 0.01 min in control cells to 0.22 ± 0.01 min and 0.37 ± 0.02 min in MC+1 cells, an increase of 47%, p < 0.05 and 48%, p < 0.05, respectively. The MC+1 cells presented a slower growth movement than the control cells which is the cause of a slow growth rate and a lower dynamicity. The frequencies of catastrophe and rescue are important parameters in studying the dynamicity of microtubules. The catastrophe frequency is the frequency with which the microtubules switch from either pause or growth to shortening. The rescue frequency is the frequency with which the microtubules switch from shortening to either growth or pause. As seen above, the MC+1 cells presented a shortening rate similar to that of the control cells but with a decreased growth rate. This suggests that the microtubules shortened in MC+1 cells are either taking longer time to grow (high growth duration) or staying in a pause phase (high pause duration). The time-based catastrophe frequency was slightly increased in MC+1 cells unlike the rescue frequency that was increased significantly by 54%, p < 0.05 as compared to control cells. This can be explained by the fact that the microtubules of MC+1 cells have tiny movements of growth and shortening for short distances (less than 0.5 μm) not counted in the growing and shortening events and in the overall dynamicity of the microtubule. The dynamicity was significantly decreased from 8.80 ± 1.02 μm/min in MP6.1 to 6.09 ± 1.06 μm/min in MC+1 cells. This decrease of 31%, p < 0.05 is representative of less overall distances travelled by the microtubule either in a growing or in a shortening event for 2 minutes.

Since TBCC is a major protein involved in the proper folding pathway of tubulins into microtubules, we investigated the response of the TBCC overexpressing breast cancer cells to antimicrotubule agents. These agents block the cell cycle in G2-M phase. We incubated MC+1 and MP6.1 cells 24 hours with non-toxic doses of 10 nM paclitaxel or 1 nM vinorelbine, and observed that the cells overexpressing TBCC were more sensitive to the G2-M blockage caused by these treatments with respect to the MP6.1 cells. MC+1 cells presented 79 and 83% of cells in G2-M after paclitaxel and vinorelbine exposure, respectively. The control cells were also blocked but to a maximum percentage of 55% (Figure 6A). The percentage of MC+1 and MP6.1 cells in the Sub-G0 phase of the cell cycle after 24 hours of treatment with 10 nM of paclitaxel and 1 nM of vinorelbine increased. However this increase was not statistically significant and was slightly higher in MC+1 cells than in MP6.1 cells (Figure 6B). In addition to antimicrotubule agents, we investigated the response of the cells to gemcitabine, an S-phase specific antimetabolite and we found that MC+1 cells were less sensitive to gemcitabine than the MP6.1 cells, significantly at the dose of 1 nM (Figure 6C). These experiments were performed on the MP6.2, MP6.3, MC+2 and MC+3 clones and the above results were confirmed.

We injected the six clones MP6.1, MP6.2, MP6.3, MC+1, MC+2 and MC+3 into SCID mice and monitored their in vivo progression and tumor formation. The control clones were able to form tumors unlike the MC+1, MC+2 and MC+3 clones that presented lower capacities of growth in vivo. When treated with paclitaxel, mice injected with MC+ clones revealed higher sensitivity to the treatment compared to mice injected with MP6 clones (Figure 6D).