Article Figures & Data
Figures
- Fig. 1.
EphA2 overexpression in breast cancer specimens. A, EphA2 protein levels were assessed by immunohistochemical staining of formalin-fixed, paraffin-embedded specimens of malignant (top) or benign (bottom) breast specimens (bar, 40 μm). The insets are higher magnification images (bar, 10 μm). Note that the nonimmunoreactive cytoplasm of benign epithelium (arrowheads) contrasts with the strong and cytoplasmic immunoreactivity of malignant cells. B, the results of immunohistochemical staining of benign and malignant mammary tissues with EphA2-specific antibodies are shown. The staining intensity and fraction of cells staining positive was evaluated as described in “Materials and Methods.” Statistical analyses revealed differences in EphA2 staining of benign and malignant samples (P < 1 × 10−6).
- Fig. 2.
EphA2 overexpression in malignant cells decreases ligand binding. A, whole cell lysates from cell models of nontransformed mammary epithelia (Lanes 1–3) or aggressive breast cancers (Lanes 4–8) were resolved by SDS-PAGE. Western blot analysis was performed using EphA2-specific (D7) antibodies. As a loading control, the membranes were stripped and reprobed with antibodies specific for β-catenin. The relative electrophoretic mobility of standards is shown on the right. B, MCF-10A cells were cotransfected with pBABE-Puro and either pNeoMSV (Vector) or pNeoMSV-EphA2 (EphA2) and treated in the presence or absence of 0.5 μg/ml EA1. Western blot analysis of whole cell lysates resolved by SDS-PAGE was performed using EphA2-specific antibodies (top). To control for sample loading, the membranes were stripped and reprobed with β-catenin antibodies (bottom). The P-Tyr content of immunoprecipitated EphA2 was determined by Western blot analyses with P-Tyr-specific antibodies (PY-20 and 4G10; middle). The blots were then stripped and reprobed with EphA2-specific antibodies as a loading control (clone D7; data not shown). C, monolayers of vector- or EphA2-transfected MCF-10A cells were stained with EphA2-specific antibodies (clone D7). Note that EphA2 was enriched within sites of cell-cell contact in vector-transfected controls but was diffusely distributed in EphA2-transfected cells. Bar, 50 μm.
- Fig. 3.
EphA2 overexpression induces malignant transformation but is reversed by ligand binding. A, to measure anchorage-independent cell growth and survival, 1 × 104 vector- or EphA2-transfected MCF-10A cells were suspended in soft agar in the presence or absence of 0.5 μg/ml EA1. After 7 days, colony formation was scored microscopically, and clusters containing at least 3 cells were defined as a positive colony. MCFEphA2 cells demonstrated significant increases in anchorage-independent growth (∗, P < 4 × 10−7), whereas EA1 treatment significantly blocks the growth of MCFEphA2 cells (∗∗, P < 5 × 10−6). B, the phenotype of control and EphA2-transformed MCF-10A cells was evaluated after incubation atop polymerized Matrigel ± 0.5 μg/ml EA1 (bottom panels) or an appropriately matched vehicle (PBS; top panels). Whereas control MCF-10A cells were organized into spherical colonies, MCFEphA2 cells displayed a stellate growth pattern in Matrigel that mimicked the behavior of aggressive breast cancer cells (MDA-MB-231; data not shown). Note that treatment with 0.5 μg/ml EA1 caused the phenotype of MCFEphA2 cells to be indistinguishable from that of control MCF-10A cells.
- Fig. 4.
EphA2 overexpression conveys tumorigenic potential. A–C, MCFEphA2 cells were implanted s.c. into the right craniolateral thorax (axilla) of athymic (nu/nu) mice. A, within 4 days, the implanted cells formed palpable masses (arrowheads). B, the histological appearance of the tumor revealed that these masses were almost entirely composed of moderately differentiated and invasive tumor cells (arrowheads) that formed dysplastic tubules with fluid-filled lumens (L). C, neoplastic cells (arrowheads) invaded adjacent skeletal muscle fibers (m). D, MCFEphA2 cells injected i.v. into the tail vein of athymic mice formed emboli in the lung within large- and medium-sized vessels (arrowheads). Histological examination of pulmonary tumor thrombi in athymic mice revealed that tumor cells partially to totally obstructed intravascular spaces but did not invade the vessel wall. Bar, 40 μm. E, tumorigenesis by EphA2-transformed MCF-10A cells (or vector-transfected controls) was evaluated after s.c. or tail vein injection. The significance of tumor formation was estimated to be P < 1.3 × 10−7 as determined by χ2 analyses.
Tables
- Table 1
Summary of EphA2-overexpressing MCF-10A cells
The subcellular localization and P-Tyr content of EphA2 in control (vector) and EphA2-overexpressing MCF-10A cells are summarized. Also shown is a summary of the behaviors of the two cell types as measured in the presence of absence of soluble ligand (0.5 μg/ml EA1). Subcellular localization P-Tyr content Soft agar colonization Behavior in 3dRBMa Vector Cell-cell contacts High Negative Benign EphA2 Membrane ruffles Low High Aggressive (stellate) EphA2+EA1 Membrane ruffles High Low Benign -
a 3dRBM, 3-dimensional reconstituted basement membrane.
-
Volume 61, Issue 5
