|Home | About | Journals | Submit | Contact Us | Français|
Breast cancer is one of the most commonly diagnosed malignancies in women. Despite the remarkable success of mammography screening and use of adjuvant systemic therapy, it is estimated that approximately 200,000 new diagnoses will be made this year and 40,000 deaths will occur due to this disease (American Cancer Society). Angiogenesis, the growth of vessels from pre-existing microvasculature, is an essential component of tumor progression and has emerged as a therapeutic modality for anti-angiogenic therapies in cancer.
Here we report in vitro and in vivo findings with a 20 amino acid peptide belonging to the collagen IV family, modified to facilitate possible translation to clinical applications. The two cysteines in its natural peptide progenitor were replaced by L-α-amino-n-butyric acid, a non-natural amino acid. The modified peptide was tested in vitro using endothelial cells and in vivo using mouse orthotopic breast cancer xenograft model with MDA-MB-231 human breast cancer cells. This modified peptide demonstrated no significant changes in activity from the parent peptide; however, because it lacks cysteines, it is more suitable for clinical translation. We also investigated its efficacy in combination with a commonly used chemotherapeutic agent paclitaxel; the inhibition of tumor growth by the peptide was similar to that of paclitaxel alone, but the combination did not exhibit any additional inhibition. We have performed further characterization of the mechanism of action (MOA) for this peptide to identify its target receptors, enhancing its translation potential as an antiangiogenic, non-vascular endothelial growth factor (VEGF) targeting agent for therapy in breast cancer.
Breast cancer is one of the most commonly diagnosed malignancies in women. Despite the remarkable success of mammography screening and use of adjuvant systemic therapy it is estimated that in United States approximately 200,000 new diagnoses will be made this year and 40,000 deaths will occur due to this disease (American Cancer Society). Thus, there continues to be an unmet need for novel therapeutic approaches especially for particular types of cancers such as triple negative breast cancer that are not amenable to endocrine or HER2-targeted therapy.
Tumor growth and progression are dependent on angiogenesis, the growth of microvessels from pre-existing microvasculature and this process has been validated as a therapeutic target. Angiogenesis is controlled by a balance among endogenous proand anti-angiogenic factors, and involves numerous processes such as endothelial cell proliferation, invasion, migration and adhesion and interactions among several cell types such as endothelial cells, pericytes and smooth muscle cells.1–3 During the last two decades a number of endogenous angiogenesis inhibitors have been identified, some being large macromolecules such as endostatin and thrombospondin, while others are shorter fragments contained in molecules such as tumstatin peptide, a 20 amino acid fragment contained in the noncollagenous domain of collagen IV molecule.4–8
Research, investigating angiogenesis as a therapeutic approach to cancer, led to the development and the approval of bevacizumab (Avastin, Roche/Genentech), an anti-VEGF monoclonal antibody, as a cancer therapeutic in 2004. However, in breast cancer bevacizumab only improved progression-free survival, but not overall survival9–11 and recently it has been withdrawn as a treatment option for metastatic breast cancer. Other agents, e.g., targeting VEGF receptors, either extracellularly with monoclonal antibodies or intracellularly with tyrosine kinase inhibitors, are currently in use or in clinical trials.12–14 Yet, despite the promise of anti-VEGF therapeutics, targeting VEGF alone may be insufficient for eradication of malignant tumors and alternatives should be developed to complement the current VEGF-based therapies with strategies that target other angiogenic factors.1 Using a bioinformatics approach our laboratory has discovered a number of endogenous peptides with anti-angiogenic potential.15 Considering the diversity of these endogenous peptides, we can envision a therapeutic approach based on a cocktail of peptides, which target different aspects of the angiogenic process that would ultimately lead to effective tumor inhibition. However, such an approach will require access to multiple peptides which have been optimized for clinical applications.
Peptides have been investigated as potential therapeutic agents in recent years, as demonstrated by over 20 FDA approved peptide drugs8,16 and over 300 therapeutic peptides under development for a variety of applications including cancer. Peptides as drugs have certain advantages compared with small molecules due to their specificity for targets and low toxicity. However, the properties that make them desirable drug candidates also make them metabolically unstable leading to overall poor bioavailability. These unfavorable properties can be overcome by using peptide mimetics, which retain the desired characteristics (e.g., target specificity) but circumvent many problems (e.g., rapid proteolysis). Such mimetics can be generated by replacing natural amino acids with non-natural amino acids or by using more resistant backbones. The process of peptide optimization involves the deconstruction of the original peptide by developing a structure-activity relationship to identify the minimal sequence required for activity (a pharmacophore), followed by reassembly, which can involve replacement of amino acids with structurally similar but more metabolically stable natural amino acids, non-natural amino acids or D-amino acids or the use of protease resistant backbones to generate new more translatable peptides. There are several examples of this approach that have been employed for generation of peptide drug candidates, e.g., ABT-510 (Abbott Laboratories), a 7-mer mimetic peptide derived from the mal II domain of the thrombospondin protein (TSP-1) whose stability and bioactivity have been enhanced by introducing terminal end modifications and non-natural amino acid substitutions; the peptide has demonstrated anti-angiogenic properties and is in clinical trials for several types of cancer.17 Another approach is the introduction of L-α-amino-n-butyric acid as a substitute for structurally similar amino acids.18 Peptides containing multiple cysteines can be difficult to synthesize and are often unstable due to their ease of oxidation, thus they are replaced by other amino acids such as serine or by non-natural amino acids that are structurally similar but more stable such as L-α-amino-n-butyric acid.19–25
In this study we present the biological activity of a biomimetic peptide modified from an endogenous sequence derived from the α5 fibril of collagen IV, whose anti-angiogenic activity was previously demonstrated in MDA-MB-231 ER/PR/HER2 triple negative breast cancer xenograft model.26 We generated a second generation peptide that is more stable and maintains its biological activity thus creating a potential drug candidate for translation to clinical applications.
The parent peptide, NH2-LRRFSTMPFMFCNINNVCNF-amide, denoted SP2000 was identified previously in our laboratory through a bioinformatics method based on sequence alignment of conserved domains.15 It was demonstrated to have activity in vivo by inhibiting the growth of breast and lung cancer xenografts.26,29 To facilitate translational applications of the compound, we substituted the cysteines in SP2000 by L-α-amino-n-butyric acid with the goal of developing a biomimetic peptide which retains biological activity with increased stability. To test if the modified compound (SP2012) performs comparably to the parent compound SP2000, we investigated its activity in proliferation and migration assays shown in Figure 1. In proliferation assay, Figure 1A, the modified sequence demonstrated an IC50 of 32.9 ± 4.7 µM in comparison to the original sequence which has an IC50 of 5.77 ± 0.49 µM (p < 0.05). Figure 1B shows the results of the performance of the two peptides in a migration assay. In this case the modified peptide, SP2012, inhibited 96.5 vs. 90.3% of the cells for SP2000 at a concentration of 50 µM (p < 0.05).
Based on these two functional assays, which demonstrated the activity of the modified compound, we continued to characterize the activity of the modified sequence in additional in vitro assays before proceeding to in vivo testing.
Typically migration potential is measured by using modified Boyden chamber assays which require the cells to respond to a chemoattractant gradient while having to migrate through a microporous surface. However, in the in vivo setting cells could be initially exposed to gradients of chemoattractants (i.e., VEGF) followed by migration under constant concentration of stimulant. Also, in vivo the cells do not necessarily have to migrate through porous barriers, thus we investigated the activity of our compounds using a wound healing type assay. The results are illustrated in Figure 2A and B. The peptide demonstrates a dose response, completely inhibiting migration of the cells at 50 µM. Similar activity was also observed on MDA-MB-231 tumor cells, with inhibition of 80% at 50 µM.
Adhesion is an important step in the processes of proliferation, migration and tube formation, thus we assessed if the peptide has an effect on adhesion. Figure 2C illustrates strong inhibition of adhesion on HUVEC and MDA-MB-231 cells. In comparison to control, the adhesion of HUVEC is almost completely inhibited (96%), which also correlates with the strong activity of the compound in the migration assay. The peptide also exhibits a strong activity in inhibiting the adhesion of tumor cells, 75% inhibition at 50 µM. The IC50 for the optimized peptide inhibition in adhesion on HUVEC is 14.3 ± 4.05 µM.
Endothelial cells have the ability to form tubes in vitro when cultured on Matrigel.27 This assay recapitulates key aspects of angiogenesis in vivo: adhesion, migration and alignment, and it has been used as a screening tool for angiogenic and anti-angiogenic compounds.30
Figure 3 illustrates the activity of the peptide in inhibiting tube formation on HUVEC and MEC. At 10 µM the peptide completely inhibits the formation of tubes on Matrigel (Fig. 3B). The peptide also has the ability to disrupt established tubes, giving an even stronger indication for its application as an antiangiogenic agent (Fig. 3C and D). Figure 3E–H demonstrate the inhibition potential of the peptide on the MEC. In this case complete inhibition is achieved at 100 µM and very strong inhibition at 50 µM. We show the results of inhibition of high concentrations because the effect of complete inhibition is qualitatively obvious and does not require quantification.
Typically angiogenesis research is conducted using HUVEC as a cell model, however recently there has been a move toward using MEC because they are considered a better model of the angiogenesis process as these are the cells that will be affected by the anti-angiogenic treatments.31 In an effort to fully characterize the activity of our compound we tested its activity using MEC. The results are shown in Figure 4A–C showing the dose responses on inhibition of proliferation, adhesion and migration respectively. The compound is active in all assays with IC50 of 82.9 ± 7.4; 0.71 ± 0.21 and 1.88 ± 0.81 µM in proliferation, adhesion and migration respectively. Figure 4D illustrates the activity in the wound healing migration assay with representative images of the migration in Figure 4E, showing complete inhibition of migration at 100 and 50 µM.
Identification of the receptor is required for understanding the mechanism of action. Previously, we have demonstrated via antibody neutralization studies that the activity of the parent compound SP2000 can be blocked by antibodies against the αVβ1 or αVβ3 integrins, and thus we hypothesized that this would be true for the new modified sequence, since the cysteines have been substituted by similar amino acids and the in vitro activity has not been compromised.15 We now characterize the interaction in more depth using different methods since antibody neutralization has its limitations. First, receptor pulldown assays were performed with cellular extracts which where incubated and crosslinked with a custom probe to identify interacting proteins. The probe was generated by covalently linking the peptide SP2012 to a reagent consisting of a tri-functional backbone via a NHS reaction. The resulting peptide probe contains a biotin moiety and a UV crosslinkable moiety. The probe was mixed with cellular extracts to allow the peptide to bind to its cellular target followed by crosslinking of the complexes formed by long wave UV (365 nm). The protein crosslinked to the peptide probe was shown to be biotinylated by protein gel blot using streptavidin HRP which detects the biotin present on the probe and its identity confirmed with specific antibodies against β1 integrin (Fig. 5A, lane 1). The pull down complex did not show any staining for β3 and the unmodified probe (quenched probe without SP2012 peptide) did not crosslink to β1 integrin (results not shown).
To confirm the specificity of the interaction between SP2012 and β1 integrin we knocked down the β1 integrin by transfecting a β1 specific shRNA into HUVEC and repeated the experiment; in this case the probe was not able to isolate the β1 complex (Fig. 5A, lane 2). Lanes 3 and 4 of Figure 5A show that the β1 levels of expression were knocked down approximately 48% in comparison to the control. β-actin was used as a control to confirm specific knockdown of β1 integrin (data not shown).
Furthermore we attempted to identify the complex between the probe and the cellular receptor on whole cells, not cellular extracts. This was accomplished by using a proximity ligation assay to show that two proteins are in close association on the cell membrane. Using this assay first we showed that the αV and the β1 proteins are in a complex (close proximity) on the membrane of HUVEC (Fig. 5C). This cellular complex should be abundantly present on HUVEC and this is clearly illustrated by the abundant red staining. Cellular nuclei are counterstained with DAPI. Next we used a modification of the proximity ligation assay to show that the SP2012 probe generated above binds to the αVβ1 integrin present on HUVEC cell membrane. For this the SP2012 probe was first allowed to bind the HUVEC that had been plated on tissue culture plates and subsequently crosslinked by long range UV light as above. Using standard immunohistochemistry techniques and anti-biotin and anti-αV antibodies we demonstrated that biotin, which is attached to SP2012, and αV are in close proximity on the cell surface (Fig. 5D). However, if the αV antibody is replaced with antibodies against another integrin subunit we still observe complex formations. This is due to the fact that integrins exist at focal adhesions in close proximity and this still leads to complex formation (although the level of staining appears to be different). Despite this assay limitation, if we consider the three experiments together (pulldown, pulldown of the knockdowns and proximity assay) we can conclude with confidence that our probe is binding to the β1 subunit which is highly abundant on endothelial cells as the αVβ1 integrin. Additional confirmation of specificity is provided by an experiment in which we used anti-biotin and anti-PCNA (proliferating cell nuclear antigen) antibodies. We did not observe any complexes on the cell surfaces, which is expected because PCNA is a nuclear protein, demonstrating the specificity of the interaction of SP2012 with the integrins (data not shown).
Furthermore we investigated the effect of the peptide treatment on the total amount of VEGF receptor 2, the main signaling receptor in the VEGF pathways. The protein gel analysis of the cell extracts treated with peptide for 24 h demonstrated a reduction in total VEGFR2 levels at high concentration (100 µM) while lower concentrations had no effect (Fig. 5B).
Based on the in vitro performance of the modified sequence we tested its activity in orthotopic breast cancer xenografts using triple-negative MDA-MB-231 human breast cancer cells in immunodeficient SCID mice. The results shown in Figure 6A demonstrate that the effect of SP2012 on tumor growth is indistinguishable from the effect of the parent sequence SP2000. Figure 6B represents the quantification of microvascular density in tumors. The SP2012 compound reduces the microvascular density by 80 vs. 60% for SP2000 (p < 0.05). The reduction in microvascular density serves not only as a biomarker indicating efficacy of treatment, it also indicates that our peptide acts on the proposed target in vivo by acting on endothelial cells and reducing their capability to form new microvasculature. A scrambled sequence of the parent peptide (SP2000) was tested in vivo in a previous experiment and showed no statistical difference between the control group (solubilization vehicle) and scrambled peptide.26
Often angiogenic therapy, such as bevacizumab, is administered in combination with chemotherapy. To assess how the activity of the peptide alone compares with the activity of a common chemotherapeutic agent (i.e., paclitaxel) we compared the in vivo activity of our peptide alone and in combination with paclitaxel.11 Peptide treatment was as efficient as the chemotherapy treatment alone indicated by the absence of statistical difference between peptide treatment alone and paclitaxel alone (Fig. 7A). Toxicity was monitored throughout the experiment by monitoring animal weight changes and also staining organs post mortem; no toxicity was observed in the animals receiving peptide treatment. Peptide treatment showed no apparent toxicity in the organ staining (results not shown) while one mouse in the paclitaxel group died while under treatment. Microvascular density was quantified and despite no apparent increase in tumor growth suppression, microvascular density was further decreased by the combination treatment (p < 0.05) (Fig. 7B and C). This further decrease in microvascular density of the combination treatment could be due in part to the anti-angiogenic properties of the paclitaxel,32 or perhaps due to some synergy between the peptide and paclitaxel.
Angiogenesis has been unequivocally demonstrated to play a role in cancer development, both in primary tumors and in metastasis. Thus it has become the focus of cancer drug development and it has led to the development of anti-angiogenic therapies. Targeting angiogenesis using extracellular approaches such as antibodies targeting VEGF (bevacizumab), the VEGF receptor (ramucirumab), or intracellular approaches targeting tyrosine kinases involved in angiogenesis-related signal transduction (sunitinib, sorafenib, pazopanib or vatalanib) have been studied intensively. However, it appears that targeting a single aspect of angiogenesis is insufficient to stop cancer progression, thus new efforts are being directed at developing combination therapies33,34 targeting multiple aspects of angiogenesis or combining it with chemotherapy. However, one concern of this multi-targeting therapeutic approach is toxicity; each component by itself has a level of toxicity due to its off-target effects thus combining multiple agents could lead to a significant increase in toxicity and possibly the introduction of new off-target effects. Peptide drugs are being developed as one novel approach to overcome some of these issues. Most therapeutic peptides are short, typically less than 30 amino acids, and highly specific to their targets with low toxicity. It is anticipated that a combination of multiple peptidebased agents will generate strong therapeutic effects without significantly increasing toxicity. One limitation of peptides is their short half-life because of their susceptibility to proteolysis and thus reduced bioavailability. Thus peptidomimetics, modification of peptides with non-natural amino acids or protease resistant backbones, are under intensive development.
In this study we introduce an anti-angiogenic peptidomimetic agent and its application for the treatment of triple-negative breast cancer, which lack the expression of receptors like estrogen, progesterone, and HER2 resulting in a higher risk breast cancer phenotype amenable only to chemotherapy interventions.35 Thus agents effective in this setting are of therapeutic interest. Also, because our agent targets endothelial cells and the process of angiogenesis, rather than a particular cancer cell receptor, it is conceivable that these new molecules will be active in other breast cancer phenotypes and even other types of solid tumors.
We previously discovered a potent anti-angiogenic peptide derived from collagen IV and in this study we introduce a more stable and druggable compound by substituting the cysteines with L-α-amino-n-butyric acid, protecting the peptide against oxidation and proteolysis. The modified compound SP2012 retains anti-angiogenic activity in endothelial proliferation and migration assays; however, the anti-proliferative activity is somewhat reduced while anti-migratory activity is enhanced as evidenced by the corresponding values of IC50. We also tested its activity in a series of alternate complementary assays and on microvascular endothelial cells to characterize its activity in more detail. One in vitro assay that is regarded a good indicator of in vivo activity is tube formation. The compound inhibited tube formation (10 µM) and also was able to initiate the disruption of pre-existing tubes within one hour. In addition, the peptide demonstrates a strong anti-tumorigenic activity by inhibiting tumor cell adhesion and migration. Based on these results we tested its activity in vivo using triple-negative orthotopic breast tumor xenografts. The inhibition in tumor growth was significant compared with control and the microvasculature was also significantly reduced demonstrating an anti-angiogenic mechanism of action. Moreover, the peptide performed comparably to the common chemotherapeutic drug paclitaxel, thus suggesting that it might be a potential potent monotherapy agent. Surprisingly, the combination of the two agents did not show any additional tumor growth suppression although the tumors had significantly less microvasculature in animals treated with the combination. Experiments are underway to understand this anomaly and also to test combinations with other chemotherapeutic agents.
Advancement of new drugs to the clinic is highly dependent on a detailed characterization of the mechanism of action (MOA). Previously, in a preliminary investigation using neutralizing monoclonal antibodies against the αVβ1 and αVβ3 integrins we demonstrated a significant reduction in the in vitro activity of the parent peptide SP2000.15 These results showed a significant reduction in the activity of the peptide; however function blocking antibodies experiments have their own limitations as the binding epitope might be different or the peptide which is much smaller in size and could perhaps still bind the receptor in spite of the antibody binding. Thus, in the present study using more specific assays we confirmed through specific pulldown of the β1 integrin subunit from cell lysates that the modified peptide retained the β1 integrin target. Also using proximity ligation assays on whole cells we showed binding of the peptide to the integrin complex on the cell surface. We have not shown definitively which unit of the complex is involved in binding; however, the results of three types of experiments in combination indicate that the binding occurs via the β1 subunit.
Previous preclinical and clinical studies investigated the application of short peptides that target integrins and thus will inhibit the angiogenesis process as well.36 Their effect was minimal in clinical settings and it was postulated that this effect is due to biphasic dose response of the treatment; at high concentration the treatment is anti-angiogenic characterized by a decrease in VEGFR2 levels while at low concentration the treatment is proangiogenic by increasing the level of VEGFR2 levels.37 In this study we demonstrated that the peptide targets a specific integrin αVβ1 rather than multiple integrins via the RGD domain, hence its effect is much more specific. In addition, we showed that the peptide has a monotonic dose response on the expression of VEGEFR2, which was considered problematic in previous studies. There is a difference in the concentration at which the peptide elicits a response on the VEGFR2 expression levels (100 µM) and the integrin mediated activity such as inhibition of migration and adhesion (25 µM). We would like to suggest that this difference might actually be expected; the adhesion and migration activity is directly related to the binding of the integrin and thus we expect that this effect can be achieved at fairly low concentrations. The downregulation of the VEGF receptor is a much more complex and distant interaction. The activity is mediated via the integrin binding and requires activation of multiple signaling molecules which will eventually feedback and generate an effect upon the receptor expression levels; thus we expect that such effects require a higher peptide concentration. Thus, this peptide provides an effective method of targeting angiogenesis via a specific pathway. Moreover, integrins containing β1 subunits, such as α6β1 are overexpressed in multiple cancers and thus peptide treatment should also exhibit an inhibitory effect on tumor cells.
In conclusion, we have developed a biomimetic peptide for anti-angiogenic and anti-tumorigenic treatment of breast cancer and demonstrated that it has strong activity in vitro and in vivo as a single agent or in combination with chemotherapy. Further studies of bioavailability and in vivo pharmacokinetic properties of this peptide will be performed to compare its properties to the cysteine-containing parent peptide. This peptide can be further optimized to make it less susceptible to proteolysis by making other non-natural or D-amino acid replacements. Very importantly considering the large diversity of other anti-angiogenic peptides we have identified, we can envision a drug discovery approach that would combine peptides targeting different aspects of cancer progression and development, which can be optimized through a process similar to the one presented here, resulting in a very specific and effective therapeutic cocktail for tumor inhibition.
Human umbilical vein endothelial cells (HUVEC) and Microvascular Endothelial cells (MEC) were purchased from Lonza. The cells were grown and maintained according to the manufacturer's recommendation using Endothelial Basal Media (EBM-2) supplemented with the Bullet Kit (EGM-2 or EGM-2MV respectively) from Lonza. All experiments were performed with cells of passage 2–7. Breast cancer cells, MDA-MB-231, were generously provided by Dr. Zaver Bhujwalla who provided us with the following details about the cell line: MDA-MB-231 breast cancer cells were purchased from the American Type Culture Collection (ATCC) and used within 6 mo of obtaining them from ATCC; the cell line was tested and authenticated by ATCC by two independent methods; the ATCC cytochrome C oxidase I PCR assay and short tandem repeat profiling using multiplex PCR.
The cells were propagated in RPMI-1640 medium (Gibco, 11875093) supplemented with 10% FBS and antibiotics (1% penicillin/streptomycin). All cells were maintained under standard conditions of 37°C and 5% CO2.
The peptides, denoted as SP2000 and SP2012 were synthesized using solid phase synthesis and were supplied as TFA salts (New England Peptide). The supplier provided product characterization (MALDI-TOF and HPLC traces) as proof of MW and purity accuracy. The peptides were of >95% purity. Peptides were solubilized in 10% DMSO (for in vivo assays) and 5% DMSO (for in vitro assays) and water due to their hydrophobic profile. The pH of solubilized peptides was checked and found to be around pH 7. In vivo controls tested the effect of the vehicle and in the in vitro experiments the DMSO percent was maintained at non-toxic threshold (determined by toxicity curves of DMSO on cells) with a final DMSO percentage (<0.2%) which was used as control in all in vitro experiments.
Proliferation assays were performed using a colorimetric proliferation reagent WST-1 (Roche, 11644807001). Briefly, we plated 2,000 cells/well in 96-well plates and allowed to adhere overnight, and added peptides in fully supplemented media the next day. After 3 d of incubation the media was replaced with serum-free EBM-2 media containing WST-1 reagent, and incubated for 4 h per manufacturer's recommendations. This method provides a direct measure of viable cells; the end measurement, the absorbance at 450 nm is a measure of the amount of formazan dye which is the consequence of the cleavage of the tetrazolium salt WST-1 by the mitochondrial succinate-tetrazolium reductase. Results were read with a Victor V fluorescence plate reader (Perkin Elmer). Dose response curves of percent live cells (in comparison to untreated cells but incubated in complete media with 0.2% DMSO) were created.
The inhibition of migration was determined using a real time migration assay system based on electrical impedance (RT-CIM, ACEA Biosciences). Briefly, 45,000 cells (HUVEC or MEC) per well were added with or without peptides to the top compartment of a CIM plate (Roche, 05665817001) which is separated from the bottom plate by a microporous (8 µm) polycarbonate membrane. The top of the membrane was coated with fibronectin (20 µg/ml) before the cells were added. Media with or without chemoattractant (i.e., fully supplemented EBM-2) was added to the bottom compartment of the chamber and the plate was incubated at 37°C for 20 h. The sensors integrated on the bottom side of the membrane allow for immediate and continuous monitoring of cells as they move. The RT-CIM technology allows for easy quantification of cell migration by monitoring the cell index derived from the measured impedances.
The inhibition activity of the peptide was assessed using E-plates (Roche, 05232368001). The cells (25,000 cells/well) were plated in the presence or absence of the peptide on E-plates and their adhesion over time (3 h) was monitored by measuring changes in the electrical impedance, which is a direct measure of the cells adhering on the electrodes. We calculate IC50 values from dose curves.
Migration inhibition was also investigated in a wound healing type assay. This assay was performed using the Oris Pro Migration assay (Platypus Technologies, CMA 1.101). Briefly, 25,000 cells/well in full media were added to the 96-well plate containing stoppers to block migration of cells to the center region of the wells. Cells were allowed to adhere for 4 h, after which the stoppers were removed. Cells were washed one time with PBS and fully supplemented media with or without compound was added to the wells. After 18 h cells were stained with calcein AM (0.5 µg/ml) (Invitrogen) and the cells that migrated to the center of the well were quantified by reading fluorescence using a Victor V plate reader (Perkin Elmer) and also imaged using a Nikon microscope (Eclipse T-100); images were acquired with the CCD Sensicam mounted on a Nikon microscope (Cooke Company). The detection of the cells that migrated into the previously restricted region is possible due to the addition of a detection mask at the bottom of the plate, which obstructs from measurement cells that did not migrate.
Compounds were also tested for their ability to inhibit tube formation, a process critical in angiogenesis. We followed an established protocol described in Arnaoutva et al.27 briefly, 50 µL/well of Matrigel (BD Biosciences, 356231) was plated in a cold 96-well plate and incubated at 37°C for 30 min for polymerization. 15,000 cells/well were added to the top of the gel and incubated in complete media in the presence or absence of peptide for 19 h. Images were captured using the CCD Sensicam mounted on a Nikon microscope (Eclipse T-100).
The putative cellular receptor target was identified by crosslinking a custom made peptide probe with endothelial cell extracts and determining the identity of the crosslinked proteins by protein gel blotting. Cellular extracts were generated from HUVECs lysed for 2 h using lysing buffer [150 mM NaCl, 1 mM EDTA, 100 µL/ml Protease Inhibitors (Sigma, P2714-1BTL), 10 µL/ml Phosphatase inhibitors (Sigma, P5726-1ML and P0044-1ML) and 1% Triton]. Brief spinning (14,000 g/15 min) was used to eliminate cellular debris. Protein concentration was determined using the Bradford Assay (BioRad, 5000201). The custom probe was created by using a trifunctional backbone (Pierce, 33073). This construct contains a biotinylated arm, a crosslinkable moiety, and an amine reactive arm. First, the probe was incubated with peptide in the presence of an 8:1 excess of DTT for 1 h at room temperature with agitation. Overnight dialysis at 4°C was used to eliminate excess probe. The bi-functional probe was incubated with cell extract (1:100) for 45 min at room temperature with agitation. The complexes formed were covalently crosslinked by exposure to UV (365 nm) for 8 min. The complexes were purified using Monomeric Avidin (Pierce, 20228) columns and separated by SDS PAGE. Proteins were then transferred to nitrocellulose membranes and probed with streptavidin and antibodies against β1 integrin (Cell Signaling, 4706).
β1 integrin was knocked down using an shRNA against the human β1 gene (Sigma, SHCLND-NM_002211). Briefly, HUVECs were plated in a 24-well plate at 75,000 cells/well and incubated for 72 h with the plasmid containing the shRNA in the presence of transfection reagent, Mirus, Trans IT-2020 (MIR 5404). Cells extracts were prepared as described above and used for receptor pulldown as described above.
Cells were cultured at a density of 25,000/well in gelatin coated 8-well chamber slides (NUNC, 177402) the night prior to the experiment. Cells were fixed in 1% paraformaldehyde for 15 min at 37°C, and then blocked with 100 mg/ml BSA for 30 min at 37°C. Cells were incubated with the same UV crosslinkable probe described above for 1 h at 37°C followed by crosslinking (8 min, at 365 nm on ice). Incubation with rabbit anti-αV and mouse anti-β1 or rabbit anti-αV and mouse anti-biotin antibodies was performed overnight at 4°C. Next, the cells were incubated with proximity antibodies (antimouse and anti-rabbit antibodies containing plus and minus probes of DNA used for amplification) (Olink, 9210) for 2 h at 37°C. Following hybridization and ligation of probes that are in close proximity the complex was amplified via rolling circle replication and detected via a fluorescent probe. The complexes were visualized by epifluorescence using a Nikon microscope and images acquired with the CCD Sensicam.
Animals were housed and treated according to the approved animal protocol of the Institutional Care and Use Committee at Johns Hopkins Medical Institution (JHMI). Orthotopic breast tumors were initiated in SCID mice using MDA-MB-231 cells. 2 × 106 cells per 100 µL aliquot of single cell suspension were injected in the breast mammary fat pad. Tumors reached volumes of 75–100 mm3 in approximately 14–21 d. Mice were randomized and arranged in groups (8 mice per group) with similar tumor volumes (no statistical difference among averages) and treatment was commenced. Peptides were administered once per day intraperitoneally (i.p.) at doses of 10 mg/kg.28 The dose was chosen based on our previous work using the parent compound and the same in vivo model.26 Paclitaxel was administered, i.p. at 5 mg/kg once a week, and to maintain the daily treatment regime PBS was administered on the non-treatment days. Tumors were measured every fourth day using calipers and the tumor volume was calculated using the formula V = ab2/2, where a is the larger and b is the smaller diameter.
Following the sacrifice of animals, tumors and organs (liver, kidney and lung) were excised and stored in IHC zinc fixative (BD Biosciences) for 10–14 d and processed by the JHMI Immunohistochemistry Core Facility. Briefly, tissues were embedded in paraffin and 5 µm sections were collected from the central cross-sectional area of the tissues. Following deparaffinization CD31 staining was performed overnight (BD PharMingen). Hematoxylin-Eosin (H&E) staining was also performed to assess toxicity. CD31 was quantified using FRiDA software (Johns Hopkins University). Pixels representative of CD31 staining were isolated from the image using masking and subsequently counted. We counted the number of pixels per image of three images per sample and compared across conditions.
Statistical significance (p < 0.05) within each experiment was assessed using standard statistical assessments such as Student's t-test and ANOVA accompanied by Dunnett's test if we were comparing different sets of data to one group or Tukey pair comparison when we were comparing multiple groups.
The authors thank all members of the laboratory for insightful discussions and review of the manuscript. Special thanks to Ping He for his help with the completion of several experiments. We are grateful to ACEA Biosciences for the use of the Personal RT-CIS system which allowed us to perform migration and adhesion experiments and in particular to Dr. Yama Abassi for detailed technical discussions. This work was supported by the National Institutes of Health grants R21 CA131931 and R01 CA138264.
No potential conflicts of interest were disclosed.