Human breast cancer cell lines BCap37, BC-DS and BC-TS, were cultured in RPMI-1640 medium supplemented with 10% fetal bovine serum (FBS), and frozen in liquid nitrogen. After thawing, experiments were done within two weeks. Female, aged 5–6 weeks, athymic nude (nu/nu) mice were purchased from Shanghai SLAC Animal Facility. All animal care and experiments were conducted according to Zhejiang University Animal Care Committee guidelines.

BCap37, BC-DS and BC-TS cells were sub-cultured into 6-cm dishes for 48 h to reach logarithmic growth phase. Cells were first observed and photographed under an inverted microscope. Then Giemsa staining (Jiangcheng Biotech, Nanjing, China) was carried out.

Cell lines was plated into ten 6-cm dishes with a density of 4×10⁴ cells/dish. Three cell counts for each cell line were made every 24 h for 10 days. Cell growth curves were made with cell number for ordinate and time for abscissa. Doubling time (T_d) was calculated based on the formula: T_d = T × lg2/lg (N₁/N₀). N₁ (N₀) stands for the cell number at T₁ (T₀) time during logarithmic growth phase. T = T₁ − T₀.

Cells were seeded at the amount of 5×10³/well at 96-well tissue culture plates. After 12 h of incubation, a series of drug concentration gradients were added to the wells, 6 repeats for one concentration. Sixty-nine hours later, MTT solution was added. Another 3 h later, the medium containing MTT was replaced with 150 µl of DMSO in each well to dissolve the formazan crystals. The absorbance was detected at 560 nM using a microplate reader (Bio-Rad, Sunnyvale, CA, USA).

To establish human breast xenografts, BCap37, BC-DS and BC-TS (0.2 ml PBS containing 1×10⁶ cells) were injected into the right armpits of the homozygous nude athymic mice (female, 5–6-weeks old). Each cell line had two groups. The control groups were treated with PBS, while the treatment groups were treated with vinorelbine (5 mg/kg, intraperitoneal injection). The injections were repeated every 6 days for 6 injection cycles. Width (a) and length (b) of the tumors were measured every 3 days. Tumor volume was calculated with the formula: V = (π/6) × ab². When the mice were terminated, the tumor tissues were removed and weighted. Data are representative of two separate experiments.

Cellular proteins were prepared with a protein lysis buffer (Beyotime, Haimen, China) and its concentration was measured by BCA protein assay kit (KeyGen Biotech, Nanjing, China). Equal samples (15 µl containing 45 µg) of protein were separated on 6–10% SDS-PAGE gels and then transferred to polyvinylidene difluoride membranes. Then these membranes were blocked for 1 h at room temperature with 5% non-fat dry milk in Tris-buffered saline (150 mM NaCl, 20 mM Tris-HCl, pH 7.5). Membranes were washed in PBST and respectively incubated with anti-MDR1, anti-tubulin primary antibodies (Santa Cruz Biotechnology, Santa Cruz, CA, USA) at 4°C. After overnight incubation, the membranes were washed with PBST, and incubated with horseradish peroxidase-conjugated goat anti-mouse IgG followed by enhanced chemiluminescent staining using the ECL system. Tubulin was used for normalization of protein loading.

Three cell lines were treated with 20 nM vinorelbine for 48 h and then harvested by trypsinization. After centrifugation, cells were fixed in 70% ethanol at 4°C overnight and then resuspended in propidium iodide staining solution containing 20 mg/ml propidium iodide and 0.5 mg/ml RNase in PBS at room temperature for 30 min before analysis by flow cytometry. Flow cytometric analysis was performed with a Beckman Coulter flow cytometer (Beckman Coulter, Miami, FL, USA) with an excitation at 488 nm and an emission at 630 nm (15).

Three cell lines were treated with 20 nM vinorelbine for 48 h and then harvested by trypsinization. After centrifugation, cells were washed with PBS and incubated in the dark for 10 min at room temperature in 100 µl binding buffer (10 mM HEPES, pH 7.4; 140 mM NaCl; 2.5 mM CaCl₂) (Beyotime) containing Annexin V-FITC (40 µl/ml) and PI (1 µg/ml). After incubation, 400 µl binding buffer was added to each sample and cells were kept on ice (16). Flow cytometric analysis was performed with a Beckman Coulter flow cytometer (Beckman Coulter). The 488-nm laser was used for excitation and FITC was detected in FL-1 by a 525/30-bp filter while PI was detected in FL-2 by a 575/30-bp filter (16).

Migration assays were performed in a 24-well Transwell chamber (Corning, Cambridge, MA, USA). Three cell lines were harvested at logarithmic growth phase and plated into upper chambers (0.2 ml serum-free medium containing 1×10⁵ cells). The lower parts of the chambers were filled with 0.5 ml of RPMI-1640 medium containing 10% FBS. After 24 h of incubation, the migration cells were stained and enumerated.

Each cell line had four groups. The control groups were incubated with Rhodamine 123 (10 µg/ml) for 30 min. In treatment 1, 3 h of incubation was added with 5 µM verapamil before Rhodamine 123. Treatment 2 was followed by 2 h of incubation with RPMI-1640 medium based on treatment 1. Treatment 3 was followed by 2 h of incubation with 5 µM verapamil based on treatment 1. After washing with 4°C PBS, and harvesting by trypsinization, intracellular Rhodamine 123 fluorescence intensity was determined with Coulter Epics V instrument (Beckman Coulter, Fullerton, CA, USA).

Data are presented as mean ± standard error of three independent experiments. Two-sided Student's t-test was used to determine the statistical difference between various experimental and control groups. Differences were considered statistically significant at p<0.05 (17).

Two vinorelbine-resistant sublines BC-DS and BC-TS, were successfully established from the human breast cancer cell line BCap37, with different 'two-stage screening methods'. BC-DS cells were selected based on continuous exposure to vinorelbine using a dose-stepwise incremental strategy (Fig. 1A). In the adaptation stage, BCap37 cells were exposed to vinorelbine from 2 to 50 nM (2, 5, 10, 20, 50 nM) for 72 h step by step. In the consolidation stage, previously selected cells were continuously cultured in medium containing 50 nM vinorelbine. When they multiplied normally, we had established the BC-DS cell line.

BC-TS were selected based on a strategy of pulsed exposure to vinorelbine with time-stepwise increments (Fig. 1B). In the adaptation stage, BCap37 cells were exposed to 50 nM vinorelbine for 1 to 48 h (1, 2, 4, 12, 24, 48 h) step by step. In the consolidation stage, previously selected cells were exposured ten times to 50 nM vinorelbine for 72 h. The resulting cell line was named as BC-TS.

Morphological characterizations of BC-DS and BC-TS are different from the parental BCap37 cells (Fig. 1C and D). Bcap37 cells grew closely and had clearly demarcated colony edges while BC-DS and BC-TS had larger cell size and the cells grew loosely with various shapes.

Biological characterization of breast cancer cells may change during the establishment of drug-resistant sublines. Thus, we examined the growth rate of BCap37, BC-DS and BC-TS both in vitro and in vivo (Fig. 2A and B). Based on the data from in vitro growth assays, the doubling time of BCap37, BC-DS and BC-TS were 34.1±2.3, 43.8±3.3 and 35.0±3.9 h, respectively. Thus, BC-DS had slower proliferation rate compared to BCap37. While BC-TS had exactly the same proliferation rate as BCap37, however, its adaptability became worse, as it needed more time to reach logarithmic growth phase.

To study the proliferation ability of BCap37, BC-DS and BC-TS cells in vivo, we further established tumor xenograft models with homozygous nude athymic mice. In vivo, BCap37 still has the most aggressive proliferation based on the in vitro data, and BC-TS needed the longest time to adapt (Fig. 2B).

To analyze the migratory activity of BCap37, BC-DS and BC-TS cells, Transwell assay was conducted (Fig. 2C and D). We counted the migrated cells of these three cell lines. Compared with parental BCap37 cells, there were 35.3% more cells migrating successfully in BC-DS, but 48.4% less in BC-TS. The results indicate that the migration ability of BC-DS was enhanced, but that of BC-TS was attenuated.

BC-DS and BC-TS cell lines were selected as vinorelbine-resistant. We used MTT assay to examine their sensitivity to vinorelbine in vitro. The IC₅₀ value of 72 h vinorelbine exposure for BCap37, BC-DS and BC-TS was 2.3±0.4, 729±100 and 120±21 nM, respectively (Fig. 3A and Table I). Thus, BC-DS and BC-TS were about 317-fold and 52-fold more resistant to vinorelbine than the parental BCap37.

To observe if BC-DS and BC-TS had stable vinorelbine-resistant characterizations, we cultured the three cell types in vinorelbine-free medium and detected the IC₅₀ values every 30 days. As presented in Fig. 3B, the IC₅₀ values at 72 h vinorelbine exposure decreased markedly in BC-DS and BC-TS with time. It took about 90 days for them to lose the vinorelbine-resistance.

We also used flow cytometric analyses, which indicated that both BC-DS and BC-TS were much more resistant to vinorelbine-induced cell cycle arrest and apoptosis. The cell cycle of BCap37 was obviously arrested in 20 nM vinorelbine for 48 h, while BC-DS and BC-TS were slightly affected in the same conditions (Fig. 3C). Annexin V/PI assay showed the percentage of apoptotic cells in BCap37 was 23.34, while it was only 5.34 and 8.33 for BC-DS and BC-TS, respectively (Fig. 3D).

Sensitivity of BC-DS and BC-TS to vinorelbine was also observed in vivo. Vinorelbine had dramatic inhibiting effect of tumor growth on BCap37 (Fig. 4A and B), but little on BC-DS (Fig. 4C and D) and BC-TS (Fig. 4E and F). Corresponding tissue sections were stained with H&E or for the proliferation marker Ki-67. Compared to BCap37 cells, fewer BC-DS and BC-TS cells exhibited vacuolization and apoptotic features (Fig. 4G), but more BC-DS and BC-TS cells were Ki-67 positive (Fig. 4H), which proved their resistance to vinorelbine in vivo.

Acquired multidrug resistance (MDR) is the main mechanism of chemotherapeutic drug resistance. We next examined their sensitivity to other chemotherapeutic agents including paclitaxel, doxorubicin, methotrexate, 5-fluorouracil and cisplatin. The MTT assay showed BC-DS and BC-TS exhibited significantly higher resistance than BCap37 to vinorelbine, paclitaxel, doxorubicin and cisplatin (Table I). As to 5-fluorouracil, there was a slight increase, but not significant in drug resistance on both BC-DS and BC-TS, whereas to methotrexate, BC-DS became more sensitive while BC-TS stayed the same. These findings suggested that BC-DS and BC-TS may represent two distinct MDR phenotypes.

Multidrug transporter P-gp could induce multidrug resistance after exposure to any drug tested (18). To determine whether P-gp was one of the main reasons responding for multidrug resistance of BC-DS and BC-TS in our study, we detected its expression in three cell lines through western blotting. Compared with BCap37, BC-DS had a remark able increase in P-gp expression while it was slight for BC-TS (Fig. 5A). This finding may explain the ability of BC-DS cells to tolerate a much higher concentration of vinorelbine than BC-TS.

Furthermore, to investigate whether intracellular drug accumulation was significantly decreased in BC-DS and BC-TS, Rhodamine 123 was used as a molecular probe in drug efflux assay. Verapamil is a calcium channel blocker and also a P-gp inhibitor that can reverse MDR (19). According to the assay (Fig. 5B), no significant change in Rhodamine 123 retention was observed in BCap37 cells with or without verapamil co-treatment. On the contrary, verapamil significantly inhibited Rhodamine 123 efflux in both BC-DS and BC-TS cell lines. Furthermore, quantity of Rhodamine 123 changed more in BC-DS than BC-TS before and after verapamil co-treatment, which indirectly indicated greater expression of P-gp in the BC-DS cell line.

Interestingly, there were also unknown bands observed around 250 kDa, the concentration of which was similar to P-gp, the expression was the most in BC-DS and the least in BCap37 (Fig. 5A), temporarily, it was named M250. The tight connection between M250 and P-gp strongly indicated M250 to be a potential tumor resistance-associated protein similar to P-gp.