Vascular Endothelial Growth Factor Receptor-3 and Focal Adhesion Kinase Bind and Suppress Apoptosis in Breast Cancer Cells

Focal adhesion kinase (FAK) is a 125-kDa nonreceptor protein tyrosine kinase that is localized to focal adhesions, which are contact sites between cells and their extracellular matrix (1). FAK has been shown to be overexpressed in breast, colon, and thyroid cancers, as well as sarcomas (2–5). One of the functions of the high levels of FAK in tumors is to promote survival by suppressing apoptosis (1, 6–9). We have shown that interruption of FAK function leads to a loss of adhesion and viability and causes apoptosis (10–12).

The COOH terminus of FAK (FAK-CD) has been extensively studied for its regulation of FAK function (10, 13, 14). Prior work from our lab has shown that displacement of FAK from focal adhesions by a dominant-negative gene construct, FAK-CD, led to dephosphorylation and degradation of FAK, resulting in apoptotic cell death in breast cancer cells (10). We hypothesize that FAK-CD causes apoptosis by displacement of binding partners to FAK, but the important binding partner(s) remains unknown.

To identify peptide sequences from potential binding partners of FAK-CD, we used phage display methodology and found a peptide-binding partner with homology to vascular endothelial growth factor receptor-3 (VEGFR-3). This suggested a potential interaction between FAK and VEGFR-3. In fact, these two tyrosine kinases have been indirectly linked via common binding partners (6, 8). For example, activation of FAK results in the autophosphorylation of Tyr³⁹⁷ that creates a site of interaction with the Src homology-2 (SH2) domain containing proteins (15). One of these proteins, Src, is necessary for FAK activation (8) and interacts with FAK as a survival factor in colon cancer (16). Src has also been shown to act upstream of VEGFR-3, activating the receptor during times of oxidative stress (17). In addition, VEGFR-3 and FAK share similar binding partners that contain other SH2 and SH3 domains that include growth factor receptor binding protein 2, Src homology and collagen, SHP-2, p85 subunit of phosphatidylinositol 3-kinase, and integrins (17, 18).

VEGFR-3 is a receptor tyrosine kinase that is involved with lymphangiogenesis and has been detected in the cytoplasm of tumor cells, lymphatic endothelia adjacent to cancer cells, and in carcinoma metastases to lymph nodes (19, 20). VEGFR-3 signaling is required for lymphatic vessel growth and is regulated by its ligands, VEGF-C and VEGF-D (21). Overexpression of VEGF-C and VEGF-D in mouse tumor models leads to increased intratumoral lymphangiogenesis with a significant increase in tumor metastasis to regional lymph nodes and distant metastasis (22, 23), linking VEGFs and tumor metastasis.

This report is the first to show that VEGFR-3 and FAK are binding partners. In addition, we have shown that a 12-amino-acid VEGFR-3 peptide displaced FAK from its location in the focal adhesions, resulting in decreased proliferation, cellular detachment, and apoptosis in breast cancer cells.

Cell culture. BT474, BT20, MCF7, MCF10A, and mouse embryo fibroblasts (MEF) FAK^+/+ and FAK^−/− (gift from S. Hanks, Department of Cell and Developmental Biology, Vanderbilt University, Nashville, TN) cells were maintained according to the American Type Culture Collection (Manassas, VA) protocols.

Phage display. Peptide-binding partners to FAK-CD were isolated by phage display. FAK-CD was isolated with Baculoviral system (Invitrogen, Carlsbad, CA) and was coated to a plate and used as a target with a phage library containing random 12-amino-acid peptides (New England Biolabs, Beverly, MA). After three rounds of panning, recovered phages were sequenced to identify potential peptides that bind FAK. An isolated peptide (Peptide-35) that bound FAK was found to be homologous to VEGFR-3 by BLAST search.

Tissue samples and immunohistochemistry assays. Formalin-fixed, paraffin-embedded tumor blocks were obtained from the University of Florida Tissue Bank. VEGFR-3 immunohistochemistry was done using polyclonal antibody, VEGFR-3 (1:100; Chemicon International, Temecula, CA), to the NH₂-terminal region of VEGFR-3. FAK immunohistochemistry was done with the FAK-specific monoclonal antibody 4.47 (Upstate, Lake Placid, NY). Slides were incubated for 60 minutes with either anti-VEGFR-3 polyclonal antibody or anti-FAK antibody after deparaffinization, heat-induced epitope recovery, Citra retrieval (BioGenex, San Ramon, CA), and blocking with Sniper (Biocare Medical, Concord, CA). The slides were washed with TBS and incubated with goat-anti-rabbit Mach2 horseradish peroxidase polymer (Biocare) for 30 minutes. The chromogenic reaction was done with 3,3′-diaminobenzidine (Vector Laboratories, Burlingame, CA).

Immunohistochemistry scoring for VEGFR-3 and FAK. A single board-certified pathologist scored breast tissue sections for VEGFR-3 and FAK expression, based on a scoring system that measured intensity (0, none; 1, borderline; 2, weak; 3, moderate; 4, strong), percent positive cells (0-100), and cellular localization (nucleus, cytoplasm, membrane, or combination).

Plasmids and transfection. The mammalian expression vector encoding HA-GFP-FRNK has been described previously (13). A 2,037-bp region (21-2057 nucleotides) from the 5′ end of VEGFR-3 that included the region homologous to Peptide-35 was cloned using TOPO cloning into pcDNA 3.1D/V5-His-TOPO vector (Invitrogen). HEK293 cells were plated at a density of 2 × 10⁶ per 100-mm culture plate and allowed to attach for 24 hours and then transfected by using Effectene (Qiagen, Valencia, CA) according to the manufacturer's protocol.

Reagents and antibodies. AV3 peptide, a 12-amino-acid VEGFR-3 sequence (WHWRPWTPCKMF), homologous to Peptide-35 was coupled to a TAT sequence (YGRKKRRQRRR) to allow cellular penetration. For a control, we used a randomly scrambled peptide containing the same 12 amino acids as AV3. Antibodies against M13 (Amersham Biosciences-Pharmacia, Piscataway, NJ); FAK (Upstate and Santa Cruz Biotechnology, Santa Cruz, CA); glutathione S-transferase (GST), α-tubulin, β-actin, and MOPC-21 (Sigma, St. Louis, MO); VEGFR-3 (Santa Cruz Biotechnology; Chemicon; Lab Vision, Fremont, CA; and Zymed, South San Francisco, CA); green fluorescent protein (GFP; Clontech, Palo Alto, CA), procaspase-3 and paxillin (BD Biosciences-Transduction Laboratories, San Jose, CA); and Rhodamine Red, Texas Red, FITC, and Alexa Fluor 488–conjugated secondary antibodies (Molecular Probes, Eugene, OR) were used for immunoprecipitation, immunoblotting, and immunohistochemistry.

Expression of recombinant GST fusion proteins. GST fusion proteins containing focal adhesion targeting (FAT) domain of FAK (GST-FAT) and NH₂ terminus of VEGFR-3 were engineered by PCR. The fusion proteins were expressed in Escherichia coli bacteria by incubation with 0.2 mmol/L isopropyl β-d-galactopyranoside, isopropyl-l-thio-B-d-galactopyranoside for 6 hours at 37°C. The bacteria were lysed by sonication, and the fusion proteins were purified with Glutathione Sepharose 4B beads (Amersham Biosciences, Piscataway, NJ).

Expression and isolation of baculovirus-expressed FAK-CD protein. PFastBac HT donor plasmid with full-length FAK-CD was transformed into DH10Bac competent cells (Life Technologies, Bethesda Research Laboratories). After transformation and transposition of the FAK-CD DNA into a bacmid DNA, colonies containing recombinant bacmids were identified by disruption of the LacZ gene inside the bacmid DNA. The isolated recombinant bacmid DNAs from white LacZ⁻ colonies were used for transfection of Sf9 insect cells. The 6× His-tagged baculoviral FAK-CD protein was isolated from High five, H5 insect cells with the Bac-to-Bac Baculovirus system buffers on Ni-NTA resin columns following the manufacturer's protocol (Invitrogen). Isolated His-tagged FAK-CD protein was analyzed by Western blot with anti-His and anti-FAK (C-20) antibodies.

In vitro binding and pull down. For pull-down experiments, a 600-bp VEGFR-3 region (1235-1829) containing Peptide-35 homologous sequence was subcloned into pGEX4T vector. The GST-VEGFR-3 fusion protein was purified and isolated onto glutathione-Sepharose beads as previously described (see above; ref. 24). GST-paxillin (smaller fragment of paxillin that binds FAK; gift from M. Schaller, Department of Cell and Developmental Biology, University of North Carolina, Chapel Hill, NC) and GST were also isolated onto beads as above. For in vitro binding assay, cell lysates from BT474 cells were precleared with GST protein (10 μg), and the cleared supernatants (500 μg) were incubated for 1 hour at 4°C with 10 μg of GST fusion protein immobilized on glutathione agarose beads. The beads were washed thrice with cold PBS. The bound proteins were analyzed by Western blotting. Pure baculoviral FAK-CD protein (500 ng) was isolated and used to determine direct binding using the above protocol.

Phage overlay, immunoprecipitation, and Western blotting. Cell lysates were prepared as previously described (24). For immunoprecipitation, 500 μg of total protein were precleared with protein A/G-agarose beads (Calbiochem, La Jolla, CA) at 4°C for 1 hour and then incubated with 5 to 10 μL of antibody overnight followed by a 2-hour incubation with protein A/G-agarose beads at 4°C. Precipitates were washed thrice with cold PBS, and beads were resuspended in SDS-PAGE sample loading buffer, boiled for 5 minutes, and resolved by SDS-PAGE. Proteins were transferred to polyvinylidene difluoride membrane, probed with appropriate phage or antibody, and detected by chemiluminescence (Perkin-Elmer, Life Sciences, Inc., Boston, MA). For phage overlay experiments GST, GST-FAT (853-1052 amino acids of FAK) and baculoviral FAK-CD proteins were used.

Immunofluorescent staining. Cells were immunostained with FAK, paxillin, and VEGFR-3 antibodies and then evaluated for colocalization with a Leica confocal microscope and the MRC-1024 confocal laser scanning system. Immunofluorescent staining was also used to evaluate displacement of FAK from the focal adhesions after peptide treatment and evaluated with Zeiss microscope as previously described (12).

Detachment assay. Cells were treated with AV3 and control peptides at 7.7 and 15.5 μmol/L for 6 hours. Detached and attached cells were counted by hemocytometer.

Cell proliferation assay. Cells were treated with peptides for 6 hours at 7.7 and 15.5 μmol/L. The 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium tetrazolium compound from Promega (Madison, WI) was then added, and the cells were incubated at 37°C for 1 hour. The 96-well plate was read at 490 nm with a plate reader to determine viability.

Apoptosis assay. Cells were treated with peptides at 7.7 and 15.5 μmol/L for 24 hours and then harvested. Hoechst 33342 was added, and the specimens were mounted on glass coverslips as previously described (25). The slides were viewed under the Zeiss microscope for apoptotic nuclei. We calculated the percent apoptosis by dividing the number of apoptotic cells by the total number of cells (n = 300).

Statistical analysis. Student's t test was used to determine significance. Data was significant if P < 0.05.

Identification of binding partners to FAK-CD. We initially used phage display to identify binding partners to the regulatory region, FAK-CD. We identified two peptides/phages that bound to FAK-CD. Peptide-32 bound to FAK-CD and contained the LD paxillin-binding motif that validated the phage display. Peptide-35 also bound to FAK-CD and had homology with VEGFR-3 that was revealed by Blast search. Binding between phage-35 and FAK-CD was detected by phage overlay (Fig. 1A). As controls, the overlay was probed with anti-FAK-C20 that recognized FAK-CD but not GST (Fig. 1B) and then probed with anti-GST that recognized GST but not FAK-CD (Fig. 1C). Next, to narrow down the region of binding between phage-35 and FAK-CD, we did phage overlay experiment with the COOH-terminal part of FAK-CD (853-1052 amino acids), GST-FAT protein (Fig. 1D), representing the FAT domain of FAK. We clearly show that phage-35 binds GST-FAT and FAK-CD but not GST. Thus, phage-35 binds the COOH-terminal part of FAK-CD, the FAT domain, which has been shown is involved in apoptotic signaling (14).

The binding between FAK-CD and phage-35 and the homology of phage-35 to VEGFR-3 led us to investigate the interaction between VEGFR-3 and FAK.

VEGFR-3 is overexpressed and present in the cytoplasm of cancer cells. VEGFR-3 is a membrane-bound receptor, and its overexpression has been shown in gastric and ovarian cancer (19, 20). However, the expression of VEGFR-3 in breast cancer cells has not been well established. In this study, we have analyzed the expression of VEGFR-3 and FAK in invasive breast cancer and normal tissues (Fig. 2A). Breast tissue specimens were stained with FAK 4.47 monoclonal antibody (top) and with an antibody that recognized the extracellular domain of VEGFR-3 (bottom). Both the normal and invasive breast cancer cells expressed VEGFR-3 in the cytoplasm and nucleus and FAK was expressed in the cytoplasm in normal and tumor cells (Fig. 2A). However, normal breast tissue showed borderline/weak VEGFR-3 expression in the cytoplasm and nucleus, whereas the breast carcinoma showed stronger VEGFR-3 expression in both the cytoplasm and nucleus. FAK expression was increased in the breast tumors and that correlated with VEGFR-3 overexpression in the tumor samples (Fig. 2A). We analyzed seven breast cancer samples for FAK and VEGFR-3 expression, and four matched normal breast tissues. We found overexpression of VEGFR-3 in 3 of 7 (43%) tumors and overexpression of FAK in 4 of 7 (57%) tumors. Normal tissues had weak/borderline expression of VEGFR-3 in 100% of cases, and weak/borderline FAK expression was detected in 75% of cases.

We compared this data with VEGFR-3 expression in breast cancer cell lines (BT474, BT20, and MCF7), normal breast cells (MCF10A), and in MEFs that were FAK^−/− or FAK^+/+ (Fig. 2B) and found overexpression of VEGFR-3 in breast cancer cell lines. Together, these results suggest that VEGFR-3 is overexpressed in invasive breast carcinoma and that VEGFR-3 is present in the cytoplasm of tumor cells, indicating the capability of VEGFR-3 to interact with FAK.

Direct association between VEGFR-3 and FAK in vitro and in vivo. To show direct interaction of FAK and VEGFR-3, we isolated, as a GST-VEGFR-3 fusion protein, a 600-bp (1235-1829 bp) region of VEGFR-3, containing the 12-amino-acid VEGFR-3 sequence (WHWRPWTPCKMF), homologous to Peptide-35. This VEGFR-3 fusion protein contains amino acids 412-610 from the NH₂-terminal domain of VEGFR-3. We used this fusion protein in pull-down experiments to determine whether there was a physical interaction between VEGFR-3 and FAK. As shown in Fig. 3A, endogenous FAK from BT474 breast cancer cell lysate bound GST-VEGFR-3 but not GST alone. As a positive control, we used a known FAK binding partner, paxillin (6, 8), and in the pull-down assay, as expected, FAK did bind to GST-paxillin (Fig. 3A). To determine direct binding of VEGFR-3 to the COOH terminus of FAK, we next used purified baculoviral FAK-CD in a pull-down experiment with GST-VEGFR-3. FAK-CD directly bound to GST-VEGFR-3 but not to GST (Fig. 3B). As controls, the GST-VEGFR-3 construct was recognized by an anti-VEGFR-3 antibody (Fig. 3B), and proteins were confirmed by Coomassie staining (Fig. 3C). These results showed direct binding of FAK-CD and the VEGFR-3 fragment, containing amino acids 412-610 from the NH₂-terminal domain of VEGFR-3, in vitro.

We then sought to confirm binding of FAK and VEGFR-3 in vivo by performing immunoprecipitation experiments after cotransfection of these molecules. We prepared lysates from HEK293 cells that were cotransfected with a V5-tagged 2,037-bp (21-2057 bp) region of VEGFR-3 (V5-VEGFR-3) and GFP-FAK-CD, or V5-VEGFR-3 and GFP (Fig. 4A). The 2,000-bp region of VEGFR-3 (1-679 amino acids) contained the homologous sequence that bound FAK-CD by phage display (above). Immunoprecipitation experiments were done using an anti-V5 antibody and then probing with anti-FAK (C20) and anti-GFP antibodies. We showed a strong interaction between VEGFR-3 and FAK-CD after probing with anti-FAK (C20) or with the anti-GFP antibody that recognizes the GFP-FAK-CD construct (Fig. 4A). To confirm the effectiveness of the coimmunoprecipitation, an anti-VEGFR-3 antibody showed equal levels of V5-VEGFR-3 (Fig. 4A). In addition, we narrowed down the binding region in FAK-CD and did coimmunoprecipitation of V5/VEGFR-3 with two deletion constructs of FAK-CD (700-733 amino acid deletion and larger 700-852 amino acid deletion), containing the FAT domain (Fig. 4B). We have shown that the NH₂-terminal fragment of VEGFR-3, encoding first 679 amino acids, binds deletion variants of FAK-CD as well as GFP-FAK-CD (Fig. 4B). Thus, 1 to 679 amino acids of VEGFR-3 can bind the COOH-terminal domain of FAK-CD, the FAT-domain of FAK (853-1052 amino acids).

In the next series of immunoprecipitation experiments, we tested binding of endogenous VEGFR-3 and FAK with lysates prepared from BT474 breast cancer cells, which contain high levels of VEGFR-3. We then did a coimmunoprecipitation experiment with an antibody against VEGFR-3 (Fig. 4C). We showed a strong interaction between VEGFR-3 and FAK in BT474 cells. An interaction between VEGFR-3 and FAK was not detected with isotype control, MOPC-21 antibody. To further confirm the specificity of this binding, we used FAK-null and FAK-positive MEFs that express VEGFR-3 for further coimmunoprecipitation experiments (Fig. 4D). Coimmunoprecipitation showed binding between VEGFR-3 and FAK only in FAK-positive MEFs but not in FAK-null MEFs (Fig. 4D). Therefore, these results show that VEGFR-3 specifically binds to FAK in vivo.

VEGFR-3 and FAK colocalize. The in vivo binding between VEGFR-3 and FAK was further shown by dual immunofluorescent confocal microscopy (Fig. 5A). FAK was revealed with Texas Red–conjugated secondary antibody, and VEGFR-3 was revealed with Alexa 488–conjugated secondary antibody. After dual immunostaining, the red FAK staining was mostly localized at the focal adhesions and cytoplasm, whereas the green VEGFR-3 staining was detected throughout the cytoplasm and nucleus (Fig. 5A,, top). We confirmed focal adhesion staining by colocalization of FAK and the focal adhesion marker protein, paxillin (Fig. 5A,, bottom). FAK and paxillin mainly colocalized in the focal adhesions and cytoplasm. Importantly, both FAK and VEGFR-3 colocalized in the cytoplasm, further showing the physical interaction between these kinases (Fig. 5A).

AV3 peptide causes displacement of FAK from the focal adhesions. To determine the biological effect of expressing the short 12 amino-acid fragment (AV3) of VEGFR-3 in cells, we first did immunofluorescent staining for FAK after treating MEF FAK-positive and FAK-negative cells. The effect of AV3 was FAK specific, because AV3 did not affect FAK^−/− cells (data not shown). However, treatment with AV3 for 2 hours effectively displaced FAK from the focal adhesions in FAK^+/+ cells, whereas the control-scrambled peptide showed no displacement (Fig. 5B,, top). At the same time, AV3 and the control peptide did not displace paxillin from the focal adhesions in MEF FAK^+/+ cells after 2 hours of treatment (Fig. 5B , bottom).

Next, we confirmed the specific effect of AV3 on FAK displacement from focal adhesions in breast cancer cells BT474 after 2 and 24 hours of treatment. Cells were treated with AV3 or control-scrambled peptide for 2 hours and then stained for FAK. We showed that FAK was displaced from the focal adhesions after AV3 treatment but not after treatment with the control-scrambled peptide (Fig. 5C). Cells treated with AV3 and control-scrambled peptide for 24 hours were immunostained for FAK, VEGFR-3 (Fig. 5D), and paxillin (data not shown). We found that even after 24 hours of treatment with scrambled peptide, FAK was still present at focal adhesions, and there was no change in VEGFR-3 distribution (Fig. 5D,, top). AV3 treatment caused FAK displacement from the focal adhesions and decreased the cytoplasmic fraction of VEGFR-3 (Fig. 5D,, bottom) but did not affect paxillin localization (data not shown). We found that colocalization of FAK and VEGFR-3 significantly decreased after AV3 treatment but not the scrambled peptide treatment (Fig. 5D , right). These results show that the AV3 peptide directly and specifically affected FAK localization in breast cancer cells.

AV3 peptide causes increased detachment, decreased proliferation, and increased apoptosis in breast cancer cells. Because displacement of FAK from the focal adhesions of breast cancer cells is associated with detachment and apoptosis (10), we next sought to determine the biological significance of disrupting the FAK/VEGFR-3 interaction with AV3. We treated a number of breast cancer cell lines with a TAT-conjugated AV3 peptide. BT474 breast cancer cells that expressed a high level of VEGFR-3 were treated with the AV3 or control-scrambled peptides for 6 hours, harvested, and counted for cell detachment. The AV3-treated BT474 cells showed increased dose-dependent detachment when compared with control-scrambled peptide (Fig. 6A). The effect on detachment was not cell line specific, because dose-dependent detachment was also seen in BT20 and MCF7 breast cancer cell lines after AV3 treatment. Moreover, a minimal effect on detachment was seen in MCF10A normal breast cells after either peptide treatment (Fig. 5A), suggesting a “therapeutic window” between normal and tumor cells. There was also a decrease in cell proliferation in BT474 cells when treated with the AV3 peptide when compared with the control peptide (Fig. 6B). Again, the effect was not cell line specific, because a decrease in cell proliferation was also seen in BT20 and MCF7 breast cancer cells after AV3 treatment compared with control peptide. Moreover, proliferation was similar with either peptide treatment in normal MCF10A breast cells (Fig. 6B). AV3-induced detachment and decreased proliferation in BT474 breast cancer cells also corresponded to a significant level of apoptosis after AV3 treatment (Fig. 6C). Apoptosis was also biochemically confirmed by increased procaspase-3 cleavage after AV3 treatment in BT474 cells but not in normal MCF10A breast cells (Fig. 6D). The procaspase-3 cleavage was confirmed by densitometry. The ratio of the cleaved procaspase-3/β-actin was 2-fold greater in the BT474 lysate treated with AV3 (0.35) when compared with control peptide (0.73). Furthermore, neither peptide caused a significant level of procaspase-3 cleavage and apoptosis in normal MCF10A breast cells. The AV3 peptide affects detachment, proliferation, and apoptosis more significantly in breast cancer cells than in normal breast cells, showing a therapeutic window with the AV3 peptide treatment.

This study is the first to show that VEGFR-3 interacts with FAK, specifically with the COOH terminus of FAK, in different cell types. In addition, these results have identified a specific sequence of VEGFR-3 (AV3) that binds to FAK. Introduction of this binding peptide into breast cancer cells disrupted FAK localization and resulted in breast cancer cell detachment and apoptosis, showing a biological effect of the peptide itself.

We have shown that AV3 binds specifically to the FAT domain, a region in FAK-CD that is involved with focal adhesion localization and apoptosis (6, 8), and AV3 specifically displaces FAK from the focal adhesions. This mechanism of FAK displacement and apoptosis with the AV3 peptide is similar to the effect of FAK-CD on breast cancer cells, as our group has shown previously (10). These results emphasize the importance of the COOH terminus of FAK in survival signaling and the potential for targeting this region to induce tumor cell apoptosis.

VEGFR-3, while linked to lymphangiogenesis, has been shown to be overexpressed in tumor cells (19, 20) and has been seen in the cytoplasm, along nuclear and cellular membranes (26). In addition, we have shown that VEGFR-3 is located in the cytoplasm and nucleus of invasive breast cancer and normal breast tissues, and that tumor cells overexpress VEGFR-3 more than do normal breast cells. How VEGFR-3 is internalized has yet to be determined, but the cytoplasmic localization of VEGFR-3 provides a rationale for its interaction with FAK. In addition to the direct interaction that we have shown, FAK and VEGFR-3 share many binding partners that allow for indirect interaction via the SH2 domain (6, 8, 17, 18). Thus, we propose that the binding of these two kinases suppress apoptosis in tumor cells.

The AV3 peptide seemed to disrupt the binding between FAK-CD and VEGFR-3, introduction of the AV3 peptide into breast cancer cells led to specific displacement of FAK from the focal adhesions, which is consistent with prior work from our group (10, 11). Importantly, the AV3 peptide caused detachment and apoptosis in breast cancer cells when compared with the control-scrambled peptide. In addition, AV3 peptide had a specific effect on colocalization of FAK and VEGFR-3. Finally, there seems to be a therapeutic window, because the peptide affected cancer cells more than normal breast cells. Taken together, these results could be used as a basis for the development of novel molecular therapeutics that target the signaling between FAK and VEGFR-3 and cause apoptosis in breast cancer.

Grant support: National Cancer Institute grant CA-65910 (W.G. Cance).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Dr. Jinli Chang for help with phage display; Jennifer Vella, Richard Finch, and Richard Wilson for technical assistance; Dr. Alfred Chung and the Protein Chemistry Core Facility at the University of Florida for the synthesis of the necessary peptides; the confocal microscope facility of the University of Florida Shands Cancer Center for the skillful technical assistance of Douglas E. Smith, The Molecular Tissue Bank at University of Florida for providing samples and help with immunohistochemistry and Dr. Nicole Massoll (Department of Pathology) for scoring tissue specimens.