Loss of plakoglobin promotes decreased cell-cell contact, increased invasion, and breast cancer cell dissemination in vivo

Abstract

Introduction: The majority of deaths from breast cancer are a result of metastases; however, little is understood about the genetic alterations underlying their onset. Genetic profiling has identified the adhesion molecule plakoglobin as being three-fold reduced in expression in primary breast tumors that have metastasized compared with nonmetastatic tumors. In this study, we demonstrate a functional role for plakoglobin in the shedding of tumor cells from the primary site into the circulation.

Methods: We investigated the effects of plakoglobin knockdown on breast cancer cell proliferation, migration, adhesion, and invasion in vitro and on tumor growth and intravasation in vivo. MCF7 and T47D cells were stably transfected with miRNA sequences targeting the plakoglobin gene, or scramble vector. Gene and protein expression was monitored by quantitative polymerase chain reaction (qPCR) and Western blot. Cell proliferation, adhesion, migration, and invasion were measured by cell counting, flow cytometry, and scratch and Boyden Chamber assays. For in vivo experiments, plakoglobin knockdown and control cells were inoculated into mammary fat pads of mice, and tumor growth, shedding of tumor cells into the bloodstream, and evidence of metastatic bone lesions were monitored with caliper measurement, flow cytometry, and microcomputed tomography (μCT), respectively.

Results: Plakoglobin and γ-catenin expression were reduced by more than 80% in all knockdown cell lines used but were unaltered after transfection with the scrambled sequence. Reduced plakoglobin resulted in significantly increased in MCF7 and T47D cell proliferation in vitro and in vivo, compared with control, with significantly more tumor cells being shed into the bloodstream of mice bearing plakoglobin knockdown tumors. In addition, plakoglobin knockdown cells showed a >250% increase in invasion through basement membrane and exhibited reduced cell-to-cell adhesion compared with control cells.

Conclusion: Decreased plakoglobin expression increases the invasive behavior of breast cancer cells. This is the first demonstration of a functional role for plakoglobin/γ-catenin in the metastatic process, indicating that this molecule may represent a target for antimetastatic therapies.

Figure 1

Comparative expression of adhesion molecules. Relative expression of (A) plakoglobin and (B) e-cadherin compared with GAPDH ± SEM for T47D, HeLa, MDA-MB-231, MDA-MB-436, and MCF7 cells, as assessed with real-time PCR. (C) Western blots showing protein expression of γ-catenin, e-cadherin, and RNA polymerase from MCF7, T47D, HeLa, and MDA-MB-231 cells.

Figure 2

Relative expression of plakoglobin and e-cadherin compared with GAPDH ± SEM before and after siRNA knockdown. (A) Scramble sequence or miRNA cassette 2 in MCF7 cells; (B) scramble sequence or miRNA cassette 3 in MCF7 cells; and (C) scramble sequence or miRNA cassette 2 in T47D cells. (D) Western blots showing γ-catenin and e-cadherin expression after transfection with scramble sequence or miRNA cassettes 2 and 3. (E) Immunohistochemical staining for γ-catenin, e-cadherin, and β-catenin (green). In the control cells, γ-catenin, e-cadherin, and β-catenin are expressed on the cell surface clearly demarcating the cell-cell junctions. In the knockdown lines, γ-catenin staining is reduced, and e-cadherin and β-catenin are detected in the nucleus and the cytoplasm.

Figure 3

Growth curves of control and knockdown cells. Mean ± SEM for (A) MCF7 and (B) T47D control and plakoglobin knockdown cells. (C, D) Relative expression of plakoglobin over the 96-hour time course of the cell-proliferation experiment in control and knockdown MCF7 and T47D cells, respectively. (E, F) The effects of plakoglobin expression on tumor cell adhesion. (E) Spheroid formation and (F) the percentage of cells adhering to a confluent monolayer of cells of an identical cell type after 2 hours of co-culture. Data shown are expressed as mean ± SEM (*P < 0.005, by one-way ANOVA followed by the Dunnett two-sided multiple comparison test).

Figure 4

Effects of plakoglobin expression on tumor cell invasion, migration, and expression of the tumor-suppressor gene NM23-H1. (A, B) Histograms of mean ± SEM percentage of MCF7 (A) and T47D (B) tumor cell invasion through a basement membrane matrix (Matrigel)-coated membrane compared with an uncoated membrane. Histograms showing mean ± SEM percentage of wound closure by MCF7 cells 24, 48, and 72 hours after scratch (C), and T47D cells 24 and 48 hours after scratch (D). (E) Expression of the tumor-suppressor gene, NM23-H1, in control and plakoglobin-knockdown MCF7 and T47D cells. (*P < 0.005, by one-way ANOVA followed by the Dunnett two-sided multiple-comparison test).

Figure 5

Effects of plakoglobin expression on tumor growth in vivo. Ten Balb/c nude mice per group were inoculated with 1 × 10⁵ MCF7 control or MCF7 3C-3 cells (A) or 1 × 10⁵ T47D control or T47D 2A-4 (B) into the fifth and tenth mammary fat pads. Tumors were measured twice per week, and all animals were killed 32 (MCF7) or 34 (T47D) days after tumor implantation. Data show mean ± SEM tumor volume over time. (C) Plakoglobin expression in excised MCF7 control and 3C-3 tumors. (D) Plakoglobin expression in excised T47D control and 2A-4 tumors. PCR data are shown as a percentage of GAPDH. (*P < 0.05, by one-way ANOVA, followed by the Dunnett two-sided multiple comparison test).

Figure 6

Effects of plakoglobin expression on tumor cell shedding into the bloodstream. (A) Numbers of circulating tumor cells per milliliter of blood isolated from five mice bearing control MCF7 cells and five mice with MCF7 3C3 tumors. Tumor cells were isolated from mice with equal tumor volume (1 cm³ per mouse), and non-tumor-bearing mice were used as a control. Cells from each group were sorted into one well of a 96-well plate, and (B) the numbers of cells isolated from the circulation with clonogenic potential are shown. (*P < 0.005, compared with MCF7 control cells, by one-way ANOVA followed by the Dunnett two-sided multiple-comparison test).