doi: 10.1038/srep46659.
Aptamer-mediated impairment of EGFR-integrin αvβ3 complex inhibits vasculogenic mimicry and growth of triple-negative breast cancers
Affiliations
- PMID: 28425453
- PMCID: PMC5397976
- DOI: 10.1038/srep46659
Item in Clipboard
Aptamer-mediated impairment of EGFR-integrin αvβ3 complex inhibits vasculogenic mimicry and growth of triple-negative breast cancers
Sci Rep.
.
Abstract
Current treatment options for triple-negative breast cancers (TNBCs) is limited by the absence of well-defined biomarkers, excluding a targeted therapy. Notably, epidermal growth factor receptor (EGFR) is overexpressed in a great proportion of TNBCs and is a negative prognostic factor. In clinical trials, however, existing EGFR inhibitors showed disappointing outcome. Oligonucleotide aptamers are a valid alternative to antibodies for diagnostic and therapeutic uses. Here, we prove that, when applied to aggressive TNBC cell lines with unique stem-like plasticity, the anti-EGFR CL4 aptamer, but not erlotinib or cetuximab, prevents the vasculogenic mimicry (VM) capability of the cells and destroys previously formed channels in three-dimensional culture. Notably, we found that CL4 impairs the matrix-induced integrin αvβ3 interaction with EGFR and integrin αvβ3-dependent cell adhesion. Consistently, the aptamer strongly inhibits VM and tumor growth in a xenograft TNBC model. These data suggest that in TNBC cells, EGFR may cooperate with integrin αvβ3 to regulate integrin binding to extracellular ligands required for VM, and EGFR-targeting by CL4 aptamer may counteract this event. Overall, we demonstrate a novel mechanism of action for CL4 related with integrin αvβ3-EGFR interaction, that may help to develop new oligonucleotide-based strategy addressing unmet need for TNBCs therapy.
Conflict of interest statement
The authors declare no competing financial interests.
Figures
(a) MDA-MB-231 and BT-549 cells were serum-starved for 18 hours and then left untreated or stimulated with 20 ng/ml EGF in the absence or in the presence of 10 μmol/l erlotinib or 200 nmol/l CL4 or CL4Sc, for 15 minutes, as indicated. Cell lysates were immunoblotted with anti-pEGFR and anti-EGFR antibodies. Equal loading was confirmed by immunoblot with anti-vinculin antibody. Values below the blot indicate the ratio of pEGFR to total EGFR signal levels, normalized to the respective vinculin signal level, and reported as relative to EGF stimulated cells in the presence of CL4Sc, arbitrarily set to 1 (labeled with asterisk). (b) MDA-MB-231 and BT-549 cells were seeded on Matrigel monolayer in the absence or in the presence of 200 nmol/l CL4 or CL4Sc or 10 μmol/l erlotinib for 24 hours. (c) BT-549 cells were seeded on Matrigel, in the presence of 200 nmol/l CL4 or CL4Sc aptamers, for 4 hours. Note that just 4 hours CL4-treatment is sufficient to prevent VM channels formation. (d) Lysates from MDA-MB-231 and BT-549 cells grown on Matrigel for 24 hours, in the presence of CL4Sc or CL4, as in (b), were immunoblotted with anti-PARP antibody. Vinculin was used as a loading control. The 24 hours-treatment of BT-549 cells with 10 nmol/l trabectedin, an antitumor drug, was used as a positive control of PARP cleavage. (e) BT-549 cells were grown on Matrigel for 24 hours and then treated with 200 nmol/l CL4 or CL4Sc for 4 hours. Destruction of preformed tubes was determined as the percentage of intact loops of CL4-treated cells compared with control. (b,c,e) Cells were photographed by phase-contrast microscopy. Representative photographs of at least five independent experiments were shown. Magnification 10×, scale bar = 200 μm. Bars depict means ± SD. ***P < 0.001 relative to CL4Sc-treatment.
(a) BT-549 cells, seeded on Matrigel in the presence of 200 nmol/l CL4 or CL4Sc for 24 hours, were stained with anti-VE-cadherin antibody, visualized by fluorescence microscopy and photographed; nuclei were stained with DAPI. Magnification 20×, scale bar = 100 μm. (b–f) BT-549 cells were left untreated or treated as in (a) and mRNA levels of the indicated VM-genes were determined by RT-qPCR. Bars depict means ± SD of three independent experiments. ***P < 0.001; **P < 0.01; *P < 0.05; one-way ANOVA followed by Tukey’s multiple comparison test. No statistically significant variations among CL4Sc- and mock-treatment was obtained.
(a) BT-549 cells grown in 2D or on Matrigel in the absence or in the presence of 200 nmol/l CL4 or CL4Sc or 10 μmol/l erlotinib for 24 hours were fixed, permeabilized and labelled with anti-αvβ3 LM609 (red) and anti-EGFR (green) antibodies. Co-localization results appear yellow in the merged images. Nuclei were stained with DAPI. All digital images were captured at the same setting to allow direct comparison of staining patterns (Magnification 63×, 0.7× digital zoom). Scale bar = 20 μm. White square indicate the area showed in insets. Arrowheads indicate some co-localization points between EGFR and integrin αvβ3. (b) Equal amounts of lysates from BT-549 cells grown in 2D or on Matrigel in the presence of 200 nmol/l CL4 or CL4Sc were directly subjected to Western blotting or prior immunoprecipitated with anti-EGFR antibody. Filter was cut in two pieces that were immunoblotted with anti-EGFR and anti-integrin β3 antibodies, stripped, rejoined and immunoblotted with anti-integrin αv antibody. Dashed line indicates the boundary between the two pieces of the filter. Vinculin was used as a loading control. Molecular weights of indicated proteins are reported (left). The bands were quantified by densitometric analysis and the amount of αv or β3 co-immunoprecipitated with EGFR relative to immunoprecipitated EGFR levels is reported. Values are shown relative to CL4Sc control, arbitrarily set to 1 (right). The data shown here represent three experiments exhibiting similar effects. Note that no significant change in αv, β3 and EGFR levels was observed in total lysates following CL4 treatment.
(a,b) Cells were mock-treated or pretreated with 200 nmol/l CL4 or CL4Sc, 10 μmol/l erlotinib, 100 μmol/l cetuximab or 10 μg/ml anti-αvβ3 LM609, as indicated, and then subjected to the adhesion assay on vitronectin-coated plates. Results are expressed as percentage of adherent cells considering the mock-treated control cells as 100%. In b (insert), lysates from NIH3T3 cells were immunoblotted with anti-integrin β3, anti-EGFR and α-tubulin antibodies. (c,d) Representative phase-contrast images of BT-549 cells grown on Matrigel monolayer for the indicated times in the absence or in the presence of 200 nmol/l CL4 or 100 μmol/l cetuximab (c) or 10 μg/ml anti-αvβ3 LM609 (d). Magnification 10×, scale bar = 200 μm. Tube formation ability was determined as the percentage of reduction in loop formation of treated cells compared with mock-treated cells. Note that in the presence of CL4 treatment no loops were observed. Bars depict means ± SD of three independent experiments. ***P < 0.001; **P < 0.01; *P < 0.05 relative to mock-treated cells; one-way ANOVA followed by Tukey’s multiple comparison test (a–c) or student’s t-test (d).
(a) Mice bearing MDA-MB-231 xenografts were injected intravenously with CL4 or CL4Sc, used as a negative control, at the times indicated by the arrows. Day 0 marks the start of treatments. Tumor growth was monitored by calipers over times and experimental raw data (expressed as fold increase) were interpolated with no curve fitting or regression analysis. ***P < 0.001; **P < 0.01; *P < 0.05 relative to CL4Sc (n = 6). (b) Tumor volume was measured by high-frequency ultrasound at day 14. Example shows analysis in one representative animal from each treatment group. (c) Mice body weight was measured at the indicated days and the mean weight of each group is shown. (d) Representative sections of tumors from CL4Sc and CL4 groups were stained with H&E (upper panels) and Ki-67 (lower panels) antibody. Ki-67 proliferation index was calculated as the percentage of Ki-67 positive cells/total cell count for 10 randomly selected 40× microscopic fields (n = 3 individual tumors, ***P < 0.001 relative to CL4Sc). Magnification 20×, scale bar = 100 μm. In (a,c,d) error bars depict means ± SD.
(a,b) In vivo imaging and quantification of IntegriSense in tumor-bearing mice. (a) Representative volume renderings taken at the same color gating from CL4Sc- and CL4-treated mice injected with IntegriSense at day 21 (upper panels). Representative images of the single tumors excised from CL4Sc- and CL4-treated mice after in vivo imaging (lower panels). (b) The amount of fluorescence (pmoles) was quantified in specific ROIs encompassing the tumor in the animal and normalized to tumor volume (cm3). Error bars depict means ± SD. P = 0.0091 (n = 4). (c) Representative sections of tumors (CL4Sc group) were stained with anti-integrin β3 (red) and anti-EGFR (green) antibodies and analysed by confocal microscopy. Nuclei were stained with DAPI. Co-localization results appear yellow in the merged image. Magnification 63×, scale bar = 10 μm. (d–f) Equal amounts of lysates from recovered tumors were immunoprecipitated with anti-integrin αvβ3 LM609 antibody (d) or anti-EGFR antibody (e) and immunoblotted with the indicated antibodies. Total lysates were immunoblotted with anti-EGFR, anti-integrin αv and anti-integrin β3 antibodies, as indicated. Equal loading was confirmed by immunoblot with anti-Ku-80 antibody (f). Molecular weights of indicated proteins are reported. Representative data are shown from one of three independent experiments. In d, the histogram reports the amount of EGFR co-immunoprecipitated with integrin αvβ3 relative to αv levels. Values are shown relative to CL4Sc control, arbitrarily set to 1. (g,h) Representative sections of tumors harvested from CL4Sc and CL4 groups were stained with integrin β3 (g) and PAS/CD31 (h) as indicated. (Magnification 40×, scale bar = 50 μm; Magnification 20×, scale bar = 100 μm).
By binding to EGFR, CL4 aptamer impairs integrin αvβ3-EGFR interaction, causing inhibition of integrin binding to matrix and, in turn, VM.
References
Publication types
MeSH terms
Substances
LinkOut - more resources
Full Text Sources
Other Literature Sources
Research Materials
Miscellaneous
