ErbB2 requires integrin alpha5 for anoikis resistance via Src regulation of receptor activity in human mammary epithelial cells

Abstract

ErbB2, a receptor tyrosine kinase highly expressed in many tumors, is known to inhibit apoptotic signals. Overexpression of ErbB2 causes anoikis resistance that contributes to luminal filling in three-dimensional mammary epithelial acinar structures in vitro. Given that integrins and growth factor receptors are highly interdependent for function, we examined the role of integrin subunits in ErbB2-mediated survival signaling. Here, we show that MCF-10A cells overexpressing ErbB2 upregulate integrin alpha5 via the MAP-kinase pathway in three-dimensional acini and found elevated integrin alpha5 levels associated with ErbB2 status in human breast cancer. Integrin alpha5 is required for ErbB2-mediated anoikis resistance and for optimal ErbB2 signaling to the Mek-Erk-Bim axis as depletion of integrin alpha5 reverses anoikis resistance and Bim inhibition. Integrin alpha5 is required for full activation of ErbB2 tyrosine phosphorylation on Y877 and ErbB2 phosphorylation is associated with increased activity of Src in the absence of adhesion. Indeed, we show that blocking elevated Src activity during cell detachment reverses ErbB2-mediated survival and Bim repression. Thus, integrin alpha5 serves as a key mediator of Src and ErbB2-survival signaling in low adhesion states, which are necessary to block the pro-anoikis mediator Bim, and we suggest that this pathway represents a potential novel therapeutic target in ErbB2-positive tumors.

Fig. 1.

Integrin α5 expression is upregulated in ErbB2-overexpressing cells. (A) RNA was prepared from attached MCF-10A cells infected with retrovirus of empty vector (pBabe), wild-type ErbB2 (NeuN) or constitutively active ErbB2 (NeuT). qRT-PCR was used to compare mRNA levels for integrins α2, α5 and α6. Values were normalized to cyclophilin expression and represented as mean ± s.e.m of fold change for at least three individual experiments (**P<0.05, ***P<0.01 by Students t-test). (B) Lysates from MCF-10A cells expressing pBabe empty vector, NeuN or NeuT were analyzed by immunoblotting with indicated antibodies. pBabe or NeuN cells were placed in morphogenesis assays and (C) cell lysates collected at indicated times and protein analyzed by immunoblotting or (D) cells were fixed and immunostained at day 8 with antibodies for anti-integrin α5 (green) and integrin α6 (red), and nuclei counterstained with DAPI (blue).

Fig. 2.

Integrin α5 is elevated in ErbB2 positive breast cancers. (A) Normal and tumor tissue from human breast tissue microarrays were immunohistochemically stained with an anti-integrin α5 antibody (brown) and counterstained with hematoxylin (blue) and the expression of integrin α5 was analyzed. Representative images from six tissue sections are shown (magnification 20×). (B) Immunohistochemistry was used to determine integrin α5 expression, which was classified as described in the Materials and Methods section. From 94 cases, 64 out of 70 ErbB2-positive cases were also found to stain positive for integrin α5. A two-tailed Pearson's correlation test was used to analyze the relationship between integrin α5 expression and ErbB2 status. Results were considered to be significant at P<0.05.

Fig. 3.

Integrin α5 expression is required for ErbB2-mediated inhibition of luminal apoptosis during morphogenesis. (A) At 48 hours after initial transfection with siRNA oligonucleotides against integrin α5 or luciferase control, MCF-10A-NeuN cells were placed in 3D morphogenesis assay for the indicated times followed by cell lysis and immunoblotting using indicated antibodies. (B) Transfected NeuN cells were cultured in 3D morphogenesis assay for 10 or 12 days, and then fixed and stained with the activated (cleaved) caspase-3 antibody (red, apoptosis marker). Nuclei were counterstained with DAPI (blue). (C) Acini positive for activated (cleaved) caspase-3 were counted and the fold change in caspase-3 positivity per experiment was plotted. Acini containing two or more activated caspase-3 cells were scored as positive (n=3, >100 acini/experiment; mean ± s.e.m. of the fold change. **P<0.05, ***P<0.01 by Students t-test).

Fig. 4.

Integrin α5 is required for ErbB2-mediated anoikis resistance, Bim inhibition and anchorage independence. (A) At 48 hours after initial transfection with siRNA oligonucleotides against integrin α5 or luciferase control, MCF-10A-NeuN were harvested as attached cells, or placed in suspension for 48 hours and then analyzed for apoptosis using DNA-fragmentation ELISA. Values represent the mean ± s.e.m. of the fold change of A₄₀₅ for at least three independent experiments (**P<0.05 by Students t-test when comparing results for the 48-hour suspension cells). (B) In parallel experiments, transfected cells were lysed from monolayer culture or from 448-hour suspension culture and proteins analyzed by immunoblotting. (C) MCF-10A cells stably overexpressing activated ErbB2 (NeuT) were transfected with luciferase control or integrin α5 siRNA oligonucleotides. At 48 hours after initial transfection cells were placed in soft agar plates for 21 days and then stained with INT Violet. Colony numbers were determined and the percentages normalized to that of control siRNA. The values shown represent averages of normalized values (n=3, mean ± s.e.m.; **P<0.001, ***P<0.05 by Students t test).

Fig. 5.

Integrin α5 is required for ErbB2-mediated signaling. (A-B) At 48 hours after initial transfection with luciferase control or integrin α5 siRNA oligos, MCF-10A-NeuN cells were placed in suspension for indicated times before lysis. Proteins were analyzed by immunoblotting with indicated antibodies.

Fig. 6.

Integrin regulation of ErbB2-mediated survival is associated with Src activation in suspension and morphogenesis. (A-C) NeuN-expressing cells were cultured in 3D morphogenesis assay and treated on day 8 with DMSO vehicle control or PP2 (10 μM). (A) At days 10 and 12, acini were fixed and stained with antibody against activated caspase-3 (red). Nuclei were counterstained with DAPI (blue). Representative confocal images for day 12 are shown, and (B) acini positive for activated caspase-3 were counted at days 10 and 12 following treatment and the percent of total acini are plotted (n=3, mean ± s.d., >100 acini/experiment; **P<0.001, ***P<0.0001 by Students t-test). (C) Cell lysates were prepared on day 12 and proteins analyzed by immunoblotting. (D,E) MCF-10A-NeuN cells were infected with lentivirus carrying shRNA constructs against Src or scrambled sequence control. (D) Cells were lysed after 48-hour suspension culture, and proteins analyzed by immunoblotting or (E) were stained with annexin V and propidium iodide, followed by FACS analysis (n=3). Black bars represent annexin V stain (early apoptosis), gray bars represent propidium iodide stain (late apoptosis). Histograms are plotted as mean ± s.e.m, *P<0.05 by Students t-test.

Fig. 7.

ErbB2 regulates integrin α5 expression via the Mek-Erk pathway during morphogenesis. (A) MCF-10A-NeuN cells were cultured in 3D morphogenesis assay and treated on day 4 with DMSO vehicle control, U0126 (10 μM) or LY294002 (50 μM). At day 6, acini were lysed and proteins analyzed by immunoblotting with indicated antibodies. (B) MCF-10A cells stably overexpressing Mek2-DD, Myr-Akt or pBabe control vector were cultured in 3D morphogenesis assays and lysed at day 4, day 6 or day 8. Proteins were analyzed by immunoblotting.