The inhibitory effects of lobaplatin, or in combination with gemcitabine on triple-negative breast cancer cells in vitro and in vivo

Cell culture and reagents

Human TNBC cells MDA-MB-231 and MDA-MB-468 (Shanghai Cell Bank at Chinese Academy of Sciences) are both basal subtypes. MDA-MB-231 cells were derived from the pleural effusion of a 51-year-old Caucasian female patient with metastatic breast adenocarcinoma, and MDA-MB-468 cells were isolated from the pleural fluid of metastatic cancer in a 51-year-old black woman with breast cancer [14], [15], [16]. They were cultured in Leibovitz’s L15 medium (BasalMedia, Shanghai, China) supplemented with 10% fetal bovine serum (FBS) (Wisent, USA), 100 U/mL penicillin, and 100 μg/mL streptomycin at 37 °C in a 100% air atmosphere. GEM (Hansen, Jiangsu, China), doxorubicin hydrochloride (HISUN, Zhejiang, China), and epirubicin hydrochloride (pudepharma, Shanxi, China) were dissolved in 0.9% NaCl, while Cisplatin (Hansen, Jiangsu, China), LBP (Hainan Changan International Pharmaceutical Co.Ltd, China), Carboplatin, and Oxaliplatin were dissolved in the complete medium and stored at – 80 °C.

Cell viability

Cellular viability assay was performed by CCK-8 (Dojindo, Japan) assay. Exponentially growing cells were seeded into 96-well plates and treated with the above anticancer drugs alone or in combination on the second day. After 48 h, 10 µL of CCK-8 was added to each well and incubated at 37 °C. The optical density of cells was determined by a multifunctional microwell detector (Bio-Tek, USA) at 450 nm and expressed as absorbance values. Relative to blank-well values, values in experimental wells and control wells were normalized.

Synergy assay

The combination index (CI) was used to evaluate the synergy of LBP and other chemotherapy drugs. This method was originally described by Chou and Talalay [17, 18]. To calculate CI values, the software CompuSyn was used, taking the entire shape of the growth inhibition curve into account for calculating whether a combination is synergistic, additive, or antagonistic [19, 20]. CI<0.90 indicates synergy, 0.91≤CI<1.10 indicates additive and CI≥1.11 indicates the antagonism of the two-drug combination.

Cellular apoptosis analysis

TNBC cells (2–3 × 10⁵ cells) were seeded in 12-well plates overnight and treated with LBP, GEM, or LBP plus GEM for 48 h. Samples were prepared using Annexin V-FITC Apoptosis Detection Kit (Nanjing Fcmacs Biotechnology, China) according to the manufacturer’s instructions, then detected using a BD ACCURI C6 flow cytometer (BD Biosciences, USA) and analyzed by FlowJo 10.4 software (FlowJo, USA). Early apoptotic cells were Annexin-V+/PI- and late apoptotic cells were Annexin-V+/PI+.

Bliss independent test was used to detect whether LBP and GEM synergistically promote apoptosis. Suppose the apoptosis rate of LBP and GEM alone at different concentrations was W_a and W_b. W_ab is calculated by the Bliss formula (W_ab=W_a + W_b – W_aW_b). The observed apoptosis rate of LBP plus GEM combination (W_{ab, o}) is compared with W_ab. If W_{ab, o}>W_ab, the combination treatment is thought to be Synergy; if W_{ab, o}=W_ab, the combination treatment is Independent; and if W_{ab, o}<W_ab, the combination is Antagonism [21], [22], [23].

Cell cycle progression analysis

TNBC cells (2–3 × 10⁵ cells) were seeded in 12-well plates overnight and then, the next night, treated for 24 or 48 h. Cells were harvested and resuspended in ice-cold 70% ethanol and fixed overnight at –20 °C. After washing with PBS, 20 µg/mL⁻¹ RnaseA solution (TaKaRa, Japan) was added and incubated at 37 °C for 30 min. Then add PI staining solution (KeyGEN BioTECH, Jiangsu, China), incubate for 20 min in the dark, use a flow cytometer to detect immediately, and analyze the cell cycle distribution by FlowJo 10.4 software.

Immunofluorescence staining

MDA-MB-468 cells (6 × 10⁵ cells) were seeded in 6-well plates containing sterile coverslips. After treatment, cells were washed with PBS, fixed in 4% paraformaldehyde for 10 min, and permeabilized with 0.1% Triton X-100/PBS for 15 min. Block with 5% bovine serum albumin (BSA) (BioFroxx, Germany) for 1 h and incubate with anti-γ-H2A.X antibody (CST, USA) diluted (1:400) in blocking buffer for 2 h at room temperature. Then incubate goat anti-rabbit IgG-iFluor 488 fluorescent secondary antibodies (1:1,000, Fcmacs) followed by DAPI staining solution (Beyotime, Shanghai, China) in the dark. Fluorescence was viewed with an Olympus Fluoview FV3000 confocal microscope system (Olympus, Center Valley, PA) using a 60 × objective and the results were semi-quantitatively analyzed with Image software.

Western blotting

Cell protein was extracted by radioimmunoprecipitation assay (RIPA) lysis buffer (Beyotime, Shanghai, China). 40 µg of total protein was electrophoresed at constant pressure (80–100 V) on a 10% SDS-polyacrylamide gel, and then the proteins were transferred to 0.2 µm PVDF membrane (Bio-Rad, Hercules) by electroporation at 350 mA. The membrane was blocked with blocking buffer for 1 h, then incubated with the specific primary antibody (1:1,000) overnight at 4 °C, and horseradish peroxidase (HRP)-linked anti-mouse/rabbit secondary antibody (1:2,000) for 1 h. Specific first antibodies: anti-GAPDH (60004-1-lg) and anti-dCK (17758-1-AP) antibodies were bought from proteintech (China), anti-cleaved caspase 3 (#9661), and anti-PARP (#9532) antibodies were bought from Cell Signaling Technology (Danvers, MA, USA). Horseradish peroxidase (HRP)-linked anti-mouse and anti-rabbit secondary antibody were bought from Fcmacs and Beyotime, respectively. Immunoreactive bands were imaged in the minichemTM chemiluminescence imaging system (SAGECREATION, China) and the intensities of bands were quantified by ImageJ software.

TNBC xenograft mouse model

All mouse experiments were performed with the approval of the Institutional Animal Care and Use Committee of Nanjing University. Forty SPF-grade female athymic BALB/c nude mice of four-week-old were purchased from GemPharmatech (Jiangsu, China) and stabilized for one week. MDA-MB-468 (1 × 10⁷ cells) was inoculated into the right armpit of the mice. Record body weight and tumor size (V=length×width²×½) every 2 days, when the tumor grew to 80–100 mm³, the mice were randomly divided into four groups (n=4) and injected (i. p.) with 0.9% NaCl (Control), 20 mg/kg GEM followed by 0.9% NaCl (GEM), 0.9% NaCl followed by 6 mg/kg LBP (LBP), or GEM followed by LBP (LBP + GEM) once a week for 6 treatment cycles. After anesthesia, the periocular blood was collected for routine blood tests (mindray BC-5,500, Shenzhen, China), then the mice were sacrificed, and tumor tissues were removed, photographed, and weighed. The relative tumor volume (RTV)=Vt/V0, where Vt is the volume on each measurement and V0 is the volume on initial treatment. T/C(%)=(Mean RTV of the treated group)/(Mean RTV of the control group) × 100%, which is expressed for the therapeutic effect of a given compound [24, 25].

Statistical analysis

All data were obtained from at least three independent experiments and expressed as mean ± standard deviation (mean ± SD). Prism 7.0 software was mainly used for statistical analysis and graphing. Difference comparison was determined via Student’s t-test or analysis of variance (ANOVA) followed by Tukey’s test using SPSS software (IBM, Armonk, NY, USA). p<0.05 was considered the minimal level of significance.

Inhibitory effect of LBP on TNBC cells

Platinum drugs are common clinical chemotherapeutic drugs. To compare the inhibitory effect of three generations of platinum drugs on TNBC cells, we treated MDA-MB-468 and MDA-MB-231 cells with cisplatin, carboplatin, oxaliplatin, and LBP. The inhibitory effect of platinum drugs on cell proliferation was determined by CCK-8 assays. The 50% inhibitory concentration (IC50) of cisplatin, carboplatin, oxaliplatin, and LBP on MDA-MB-468 cells was 6.71, 103.8, 35.48, and 7.45 µM, respectively, while the inhibitory rate of the four drugs at 25 µM was about 89.5 ± 0.2%, 16.0 ± 1.4%, 41.2 ± 1.5%, and 92.9 ± 0.1%, respectively (Figure 1A). The IC50 of cisplatin, carboplatin, oxaliplatin and LBP on MDA-MB-231 cells was 22.9, 54.37, 47.86, and 22.38 µM, respectively, while the inhibitory rate of 25 µM cisplatin, oxaliplatin, and LBP on MDA-MB-231 cells was about 51.3 ± 5.5%, 42.0 ± 1.1%, and 50.3 ± 1.3%, respectively. 25 µM carboplatin had almost no inhibitory effect on MDA-MB-231 cells (Figure 1A). Thus, we concluded that LBP exhibited a similar inhibitory effect to cisplatin, but was superior to that of carboplatin and oxaliplatin in both cell lines.

Figure 1:

The inhibitory effects of LBP on TNBC cells. (A) The inhibition curves of three generations of platinum drugs and the bar graphs corresponding to the inhibition rates of the four drugs at 25 µM for 48 h. (B) Apoptosis detected by flow cytometry for 48 h. (C) Distribution of different phrases in the cell cycle detected by flow cytometry for 24 h. Data are represented as the mean ± SD. *p<0.05, **p<0.01, ***p<0.001.

Next, we observed the apoptotic promotion of LBP on the two cell lines using flow cytometry. LBP significantly promoted the apoptosis of MDA-MB-468 cells in a time- and dose-dependent manner (Figure 1B). Consistent with the CCK8 results, LBP promoted the apoptosis of MDA-MB-468 cells more strongly than of MDA-MB-231 cells. For example, 12.5 µM LBP induced apoptosis in 53.93 ± 4.18% of MDA-MB-468 cells, which was higher than the apoptosis rate of MDA-MB-231 cells (44.36 ± 1.88%) at 50 µM LBP. The cell cycle test showed that the percentage of the two cell lines in S phase increased greatly after being treated with LBP, from 21.13 ± 0.55% (the control) to 46.07 ± 0.4% (5 µM LBP) for MDA-MB-468 cells, and from 37.73 ± 0.81% (the control) to 69.47 ± 1.19% (25 µM LBP) for MDA-MB-231 cells (Figure 1C), indicating that DNA synthesis was blocked by LBP. The effect of cell cycle arresting by LBP on MDA-MB-468 cells was also greater than on MDA-MB-231 cells. Thus, we concluded that MDA-MB-468 cells were more sensitive to LBP than MDA-MB-231 cells.

Synergistic inhibition efficacy of LBP and GEM on TNBC cells

The combination of anti-cancer drugs can improve treatment effectiveness and reduce drug resistance. Therefore, we combined LBP with the common anticancer chemotherapy drugs, doxorubicin, epirubicin, docetaxel, and GEM to explore which drug performed the best synergistic inhibition efficacy with LBP. We first measured the inhibitory effect on cell growth of these drugs and plotted a curve to calculate the 25, 50, and 75% inhibitory concentrations (IC25, IC50, and IC75) using Prism 7.0 software (Figure 2, Tables 1 and 2). Then, we combined LBP with different drugs at the IC25, IC50, and IC75, and calculated CI values using CompuSyn software to detect whether there was a synergistic effect. The CI values are shown in Tables 3 and 4. A CI value less than 0.9 indicated a synergistic effect between the two drugs. The smaller the CI value, the more significant the synergistic effect was. LBP plus GEM had a remarkable synergistic effect on MDA-MB-468 and MDA-MB-231 cells and the CI value was significantly lower than that of other drug combinations. When 15 µM of LBP plus 0.38 µM GEM (IC25) were added to MDA-MB-468 cells, the CI value was ∼0.33, which was the lowest CI value (Table 3), thus, indicating the strongest synergy. Moreover, when LBP was combined with 0.25 µM GEM, its anti-proliferation effect was greatly improved. The IC50 reduced by ∼10-fold from 7.68 ± 1.84 µM (LBP alone) to 0.92 ± 1.51 µM (LBP + GEM) (Figure 3A).

Figure 2:

The inhibition curves of doxorubicin, epirubicin, docetaxel, and GEM for 48 h.

Figure 3:

The synergistic effects of LBP and GEM on MDA-MB-468 cells for 48 h. (A) The IC50 of LBP alone or in combination with 0.25 µM GEM. (B) Cell morphology changes observed using light microscopy with a 10 × objective. (C) The proportion of apoptosis determined by flow cytometry. (D) Distribution of different phrases in the cell cycle detected by flow cytometry. The data are represented as the mean ± SD. *p<0.05, **p<0.01, ***p<0.001, ^nsp>0.05.

In the following study, concentrations of 2.5 or 5 µM LBP combined with 0.25 µM GEM were used to investigate the synergistic effects. Under the microscope, we observed that the cell morphology was damaged more significantly by the LBP and GEM combination, compared with that of LBP alone (Figure 3B). In addition, LBP + GEM effectively increased apoptosis in a dose- and time-dependent manner (Figure 3C). The apoptosis rate of LBP + GEM at (2.5 + 0.25) µM and (5 + 0.25) µM for 48 h was 22.04 ± 1.93% and 48.37 ± 5.2%, respectively, and for 72 h was 37.93 ± 0.59% and 80.2 ± 6.72%, respectively, which was higher than cells treated with LBP or GEM alone. 2.5 or 5 µM LBP or 0.25 µM GEM alone for 48 h was 6.21 ± 1.40%, 11.23 ± 5.65%, and 13.73 ± 1.29%, respectively, and for 72 h was 11.66 ± 1.24%, 51.63 ± 9.27% and 9.57 ± 0.71%, respectively. Using the Bliss Independence Model, we observed that LBP and GEM had an excellent synergistic effect to promote apoptosis. The apoptosis rate predicted by the model for 2.5 and 5 µM LBP plus GEM for 48 h was 19.09 ± 1.48% and 23.41 ± 4.34%, respectively, and for 72 h was 20.11 ± 1.11% and 56.26 ± 7.26%, which were significantly lower than the corresponding observed values (p<0.05). However, LBP in combination with 0.25 µM GEM did not enhance its arresting effect on the cell cycle (Figure 3D).

Effects of LBP and GEM on TNBC cell death

Levels of γ-H2AX produced by the phosphorylation of H2AX can reflect the degree of DNA damage. To observe DNA damage induced by GEM, LBP, or combinations thereof, the fluorescence intensity of MDA-MB-468 cells labeled with FITC-γ-H2A antibodies was detected using confocal laser scanning microscopy (Figure 4A, B). The results revealed that the fluorescence of cells treated with 0.25 µM GEM alone was only marginally more intense than the control. The fluorescence intensity of cells treated with 2.5 or 5 µM LBP was much higher than that of the control. However, the difference between treatments of LBP alone and in combination with GEM was not observed. Those results indicated that LBP could cause nuclear damage to MDA-MB-468 cells, but no synergistic effect with GEM was observed at GEM concentrations of up to 0.25 µM.

Figure 4:

Effects of LBP and GEM on MDA-MB-468 cell death. (A) Image of the nucleus obtained using a confocal microscope with a 60 × objective for 24 h. FITC-γ-H2A antibodies were used to label DNA damage, while DAPI was used to label nuclei. (B) The mean fluorescence intensity of γ-H2AX for each group analyzed by ImageJ software. (C) Detection of cleaved PARP and Caspase3 by western blotting in TNBC cells treated with LBP alone. (D) Detection of cleaved PARP and Caspase3 in MDA-MB-468 cells treated with LBP plus GEM. (E) Detection of dCK by western blotting in MDA-MB-468 cells treated with GEM, LBP, or in combination. The data are represented as the mean ± SD. *p<0.05, **p<0.01, ***p<0.001, ^nsp>0.05.

To investigate the mechanism of LBP on cell death, the apoptotic proteins were detected by western blotting. After MDA-MB-468 cells were treated with LBP alone at 12.5 or 25 µM, cleavage-Caspase3 and cleavage-PARP were visible and in a dose-and time-dependent manner while that was not observed in MDA-MB-231 cells treated with 50 or 75 µM LBP (Figure 4C). Thus, LBP alone promoted the apoptosis of MDA-MB-468 cells through the classical apoptosis pathway. While adding 0.25 µM GEM to LBP, the level of cleavage-Caspase3 and cleavage-PARP was not elevated (Figure 4D), indicating that 0.25 µM GEM enhanced the apoptosis-promoting effect of LBP on MDA-MB-468 cells probably not through the classical apoptotic pathway. Deoxycytidine kinase (dCK) can convert GEM to GEM monophosphate for subsequent active metabolites, which are incorporated into DNA and block DNA synthesis. dCK deficiency is frequently described as one of the important causes of tumor resistance to GEM in vitro or in vivo [10]. Besides, cisplatin was found to be able to increase dCK activity in HCT116 cells [26]. Therefore, we hypothesized whether LBP also improved the activity of dCK to make GEM play a better role in promoting MDA-MB-468 cell death so that they showed synergistic in promoting cell death. Using western blotting, we found that dCK protein levels of MDA-MB-468 cells could be upregulated by LBP treatment (Figure 4E), which might explain why there was such a synergy between LBP and GEM.

Effects of LBP and GEM in TNBC xenograft mouse models

The xenograft mouse model of MDA-MB-468 was established to test the in vivo anti-tumor effect of LBP alone or in combination with GEM. When the tumor grew to 80–100 mm³, the mice were randomly divided into four groups (n=4), the Control group, the GEM (20 mg/kg) group, the LBP (6 mg/kg) group, and the LBP + GEM (6 mg/kg + 20 mg/kg) group, of which the regimen-administration schedule is as the below (Figure 5A). Because the MDA-MB-468 model exhibited weak tumorigenesis and slow growth, the treatment lasted ∼6 weeks. We observed that mice in three drug-treated groups lost weight to varying degrees following each dose, but regained the lost weight within several days (Figure 5B). At the end of the treatment, significant differences in tumor size were observed between the control group (567.6 ± 126.2 mm³) and LBP group (302.7 ± 131.6 mm³) (p<0.05), which indicated that 6 mg/kg LBP could effectively inhibit tumor growth in vivo. GEM treatment alone (529.5 ± 97.7 mm³) had a weak inhibitory effect on the growth of tumors in vivo compared with the control group (p>0.05). When LBP was combined with GEM, the synergistic effect was slight and not statistically significant (302.7 ± 131.6 mm³ in the LBP group vs. 207.7 ± 83.94 mm³ in the LBP + GEM group; p>0.05). Similar results were also observed for tumor weight (Figure 5C–E). Thus, we concluded that LBP alone had an excellent ability to inhibit the growth of MDA-MB-468 cells in vivo, but the effect was not significantly enhanced when combined with GEM. Routine blood tests showed that there was no significant difference between the treated groups and the control group (Table 5), indicating that the dosing schedule was tolerable and safe for mice.

Figure 5:

Effects of LBP and GEM in TNBC xenograft mouse models. (A) The regimen-administration schedule: BALB/c tumor-bearing nude mice were randomly divided into four groups (n=4), the control group, the GEM (20 mg/kg) group, the LBP (6 mg/kg) group, and the LBP + GEM (6 mg/kg + 20 mg/kg) group. Each group was injected once a week for 6 treatment cycles. (B) The curves of body weights during administration. (C) The curves of tumor growth between groups. (D) Image of tumors following therapy. (E) Comparisons of tumor weights between groups. The data are represented as the mean ± SD. *p<0.05, **p<0.01, ***p<0.001, ^nsp>0.05.