Advertisement
Models and Technologies

Improvement of Stability and Efficacy of C16Y Therapeutic Peptide via Molecular Self-Assembly into Tumor-Responsive Nanoformulation

Yanping Ding, Tianjiao Ji, Ying Zhao, Yinlong Zhang, Xiaozheng Zhao, Ruifang Zhao, Jiayan Lang, Xiao Zhao, Jian Shi, Saraswati Sukumar and Guangjun Nie
DOI: 10.1158/1535-7163.MCT-15-0484 Published October 2015

Abstract

Peptide therapeutics hold great promise for the treatment of cancer due to low toxicity, high specificity, and ease of synthesis and modification. However, the unfavorable pharmacokinetic parameters strictly limit their therapeutic efficacy and clinical translation. Here, we tailor-designed an amphiphilic chimeric peptide through conjugation of functional 3-diethylaminopropyl isothiocyanate (DEAP) molecules to a short antitumor peptide, C16Y. The ultimate DEAP–C16Y peptides self-assembled into spherical nanostructures at physiologic conditions, which dissociated to release individual peptide molecules in weakly acidic tumors. DEAP–C16Y peptides showed negligible cytotoxicity but impaired vascular endothelial cell migration and tubule formation by inactivation of the focal adhesion kinase and PI3K–Akt pathways, as well as tumor cell invasion by decreasing invadopodia formation. Compared with C16Y, the systemically administered DEAP–C16Y nanostructures exhibited superior stability, thereby allowing prolonged treatment interval and resulting in significant decreases in microvessel density, tumor growth, and distant metastasis formation in orthotopic mammary tumor models. Through encapsulation of hydrophobic doxorubicin, DEAP–C16Y nanostructure served as a smart carrier to achieve targeted drug delivery and combination therapy. Our study, for the first time, demonstrates that a simple nanoformulation using a functional antitumor peptide as the building block can show innate antitumor activity and also provide a nanoplatform for combination therapy, opening a new avenue for the design of antitumor nanotherapeutics. Mol Cancer Ther; 14(10); 2390–400. ©2015 AACR.

Introduction

Angiogenesis and metastasis are two essential hallmarks of cancer (1). Various growth factors, receptors, and extracellular matrix (ECM) components coordinate to induce tumor vessel sprouting for sustained neoplastic growth (2). Tumor neovasculature also provides a primary pathway for tumor cell dissemination that depends on the invasive capability of tumor cells. Therapeutic targeting of tumor angiogenesis and tumor cell invasion shows great potential for preventing cancer progression (3, 4).

Bioactive peptides are promising therapeutic agents due to low toxicity, high specificity, high tissue penetration, and ease of synthesis and modification (5, 6). Multiple antitumor peptides have been identified, mostly derived from natural proteins (7, 8). A peptide that competes for laminin-1 binding, C16Y (DFKLFAVYIKYR), inhibits angiogenesis and tumor growth by targeting integrin αvβ3 and α5β1 (9, 10). The major obstacle to peptide application is the short half-life due to the enzymatic degradation and rapid excretion (5). Strategies for increasing the stability and activity of peptide drugs are required to improve their therapeutic outcome.

Nanostructures with adequate size, shape, and surface properties can improve the pharmacokinetics and bioavailability of drugs and realize combination therapy (11). Amphiphilic peptides with distinct hydrophobic and hydrophilic segments can self-assemble into particular nanostructures in aqueous solutions (12). Peptide self-assemblies have been widely reported as nanocarriers to deliver antitumor drugs (13–18). A peptide antagonist of CXCR4 was demonstrated to form self-assembled nanoparticles that exhibited innate activity against tumor metastasis (19). Inspired by the peptide self-assembly concept, we sought to develop, for the first time, a tumor-responsive nanoformulation using an antitumor peptide as a building block and other functional moieties as responsive elements. Such agents would have intrinsic antitumor activity and simultaneously serve as nanocarriers for combination therapy.

In the present study, using C16Y as a building block, we designed an amphiphilic chimeric peptide, which self-assembled into a nanostructure with a hydrophobic core and a hydrophilic shell at the physiologic pH. In weakly acidic tumor tissue, this nanostructure dissociated to release peptide molecules that subsequently targeted cells expressing integrin αvβ3 and α5β1. The peptide nanostructure exhibited superior stability and suppressed tumor angiogenesis, growth, and metastasis. Through encapsulation of chemotherapeutic drugs, the peptide nanostructure intelligently achieved targeted drug delivery and combination therapy.

Materials and Methods

Peptide synthesis

3-diethylaminopropyl isothiocyanate (DEAP) was purchased from Sigma-Aldrich. Peptides C16Y (DFKLFAVYIKYR), DEAP–C16Y (DEAP4-K3-L8-G2-C16Y-G), a fragment of human serum albumin (HSAF, DELRDEGKASSAKQ), and DEAP–HSAF (DEAP4-K3-L8-G2-HSAF-G) were synthesized and analyzed using high-performance liquid chromatography and electrospray ionization mass spectrometry by GL Biochem. According to the quality reports provided by the manufacturer, the purity was more than 95% for C16Y and HSAF, and more than 90% for DEAP–C16Y and DEAP–HSAF.

Critical micelle concentration

DEAP–C16Y (0–200 μmol/L) was incubated with pyrene (24 μg/L) in PBS (pH 7.4) for 1 hour. The fluorescence emission spectrum of each solution was recorded using an excitation wavelength of 334 nm by an F-4600 fluorescence spectrophotometer (Hitachi, Japan). The intensity ratio of the first (370–373 nm) to third (381–384 nm) vibronic bands versus concentration was plotted to calculate the critical micelle concentration (CMC; ref. 20).

Preparation and characterization of peptide nanostructures

DEAP–C16Y (1 mg) and DEAP–HSAF (1 mg) were each dissolved in 10 μL of DMSO, followed by separate dilution into 1 mL of PBS (pH 7.4) under ultrasonication (600 W) for 2 minutes. After incubation for 2 hours, the morphology was examined by transmission electron microscopy (TEM; Tecnai G2 F20 U-TWIN, FEI), and size/polydispersity index (PDI) was analyzed by dynamic light scattering (DLS, Zetasizer Nano ZS90; 20).

Cell lines and animals

Human umbilical vein endothelial cells (HUVEC) were a kind gift from Xiyun Yan and were originally purchased from CellSystems Biotechnology Vertrieb GmbH (Troisdorf, Germany). The breast cancer cell lines MDA-MB-231 and 4T1 were purchased from National Platform of Experimental Cell Resources for Science-Technology. Cell lines used were those frozen within 6 months of purchase from the cell bank (authenticated using short tandem repeat DNA profiling analysis). These cell lines were not authenticated independently. HUVECs were cultured in RPMI-1640 medium (WISENT), containing 10% FBS (WISENT) and endothelial cell growth supplement (Macgene). MDA-MB-231 cells were maintained by DMEM (WISENT) supplemented with 10% FBS. 4T1 cells were cultured in RPMI-1640 medium containing 10% FBS. BALB/c nude mice and BALB/c mice (female, 6–8 weeks) were purchased from Vital River Laboratory Animal Technology (Beijing, China). All animal studies were approved by the Institutional Animal Care and Use Committee of Peking University.

DEAP–C16Y nanoprobe

DEAP–C16Y (0.5 mg), tetramethylrhodamine-5-isothiocyanate (TRITC, 0.1 mg) and black hole quencher (BHQ-1, 0.1 mg) were dissolved in 10 μL of DMSO and diluted into 1 mL of PBS (pH 7.4) under ultrasonication (600 W) for 2 minutes. After incubation for 2 hours, the sample was centrifuged at 10,000 × g for 5 minutes to isolate the aqueous phase. After the pH was adjusted, the fluorescence intensity was detected using the F-4600 fluorescence spectrophotometer (excitation, 555 nm and emission, 580 nm).

Tumor inoculation

Each BALB/c nude mouse or BALB/c mouse was implanted in one mammary fat pad with MDA-MB-231 (5 × 106) or 4T1 (1 × 106) cells in 100 μL of a mixture of PBS and Matrigel (1:1, v/v), respectively. Tumor volume was calculated as length × width2/2.

In vivo imaging of the DEAP–C16Y nanoprobe

DEAP–C16Y nanoprobe was i.v. injected to mice bearing MDA-MB-231 tumors with average sizes of 200 mm3. Fluorescence images were captured using a Maestro in vivo imaging system (Cambridge Research and Instrumentation).

Cytotoxicity

Cells in 96-well plates (50% confluence) were treated with peptides in culture medium containing 2% FBS for 24 hours. Cell viability was evaluated using the cell counting kit-8 (CCK-8; Dojindo Laboratories; ref. 21).

Endothelial cell function

For scratch migration, confluent HUVECs were scratched and treated with peptides for 18 hours. The wound closure width (W) was determined using the formula W0 h − W18 h. Relative wound closure width was calculated by normalizing mean wound closure width in each group to that in the PBS-treated group. For Transwell migration, HUVECs (2 × 104) in 150 μL of RPMI-1640 medium containing peptides were seeded into Millicell culture inserts hanging in a 24-well plate. After incubation for 1 hour, cells were chemoattracted by 850 μL of 5% FBS/RPMI-1640 with peptides for 6 hours. After fixation with 4% paraformaldehyde and staining with 0.1% crystal violet/PBS, cells that migrated across the filters were captured. For tubule formation, HUVECs were seeded onto a Matrigel-coated 24-well plate (5 × 104/well) and incubated with peptides in 2% FBS/RPMI-1640 for 6 hours. Tubule length was calculated using AngioSys 2.0 Image Analysis Software (TCS Cellworks). For these assays, three identical replicates were performed for each treatment and three random fields in each replicate were captured by a light microscopy (AMG EVOS xl core; Life Technologies; ref. 22).

Tumor cell invasion

MDA-MB-231 cells (5 × 104) were suspended in 150 μL of DMEM containing peptides and seeded into a Matrigel-coated Millicell chamber. DMEM supplemented with 5% FBS and peptides with identical concentration to those in Millicells were used as chemoattractant for 24 hours (23). Following fixation and crystal violet staining, migrated cells were captured by light microscopy. Three identical replicates were performed and three fields of each treatment were assessed for quantification.

Gelatin degradation

Coverslips were coated with 50 μg/mL poly-L-lysine/PBS for 20 minutes, activated by 0.5% glutaraldehyde/PBS for 15 minutes, and rinsed in 0.2% FITC-gelatin (AnaSpec) for 10 minutes. MDA-MB-231 cells (3 × 104) were seeded onto each coverslip and treated with peptides in 5% FBS/DMEM for 12 hours. After incubation with TRITC-phalloidin (Cytoskeleton) for 40 minutes and Hoechst 33342 (Life Technology) for 5 minutes, cells were detected by confocal microscopy (LSM710; Carl Zeiss; ref. 24).

Western blot analysis

Proteins in the gels were transferred to polyvinylidene difluoride membranes (Merck Millipore) that were subsequently blocked with 5% non-fat milk and incubated with appropriate primary and secondary antibodies. Signals were developed using the SuperSignal West Pico chemiluminescent substrate (Thermo Scientific; ref. 22).

Immunohistochemical staining

Tumor sections (7 μm) were deparaffinized, rehydrated, and incubated with 0.3% H2O2 in methanol. After antigen retrieval in 10 mmol/L citrate buffer (pH 6.0) at 95°C for 15 minutes, sections were blocked with 5% goat serum/PBS for 1 hour and incubated with primary antibody at 4°C overnight followed by biotinylated secondary antibody at room temperature for 1 hour and horseradish peroxidase–conjugated streptavidin at 37°C for 30 minutes. DAB was used for color development. Sections were counterstained with hematoxylin (25).

Statistical analysis

Data were analyzed by the Student t test for comparison of two groups and one-way ANOVA followed by the post hoc test for multiple groups.

Results

Design and characterization of the C16Y nanoformulation

Because C16Y (DFKLFAVYIKYR) is an integrin-targeted peptide (10), we used it to engineer an amphiphilic peptide for the creation of peptide self-assembling nanostructures with antitumor properties. The hydrophobic molecule DEAP has the pKb near 6.8 (20, 26, 27), which is in the range of the tumor extracellular pH (∼6.7–7.1; ref. 28). Thus, we constructed an amphiphilic peptide DEAP–C16Y, DEAP4-K3-L8-G2-DFKLFAVYIKYR-G (Fig. 1A). Four functional DEAP molecules and eight leucine residues were engineered at the N-terminus to constitute the hydrophobic segment. Three lysine residues provided primary amino groups for DEAP conjugation. An extra glycine residue was designed at the C-terminus to eliminate terminal charges, which could interfere with the self-assembly process. DEAP–C16Y had a CMC of 1.78 μmol/L (6.43 mg/L) in PBS at pH 7.4 (Fig. 1B), above which concentration this peptide can form a stable nanostructure. TEM examination showed that DEAP–C16Y (1 mg/mL) formed spherical nanostructures at pH 7.4 (simulating physiologic pH), which turned to a disassembled state when the pH was adjusted to 6.8 (simulating tumor pH; Fig. 1C). The diameters of most DEAP–C16Y nanostructures were uniformly distributed around 30 nm at pH 7.4 (PDI = 0.373) whereas approximately 7 nm at pH 6.8 (Fig. 1D). These results demonstrated that DEAP–C16Y can self-assemble into a regular nanostructure at physiologic pH, which dissociated in weakly acidic environment.

Figure 1.
Figure 1.

Design and characterization of the DEAP–C16Y nanoformulation. A, schematic structure of the DEAP–C16Y peptide. B, DEAP–C16Y has a CMC of 1.78 μmol/L. C, TEM images of DEAP–C16Y nanostructures at pH 7.4 and 6.8; scale bars, 50 nm. D, the size distribution of DEAP–C16Y nanostructures at pH 7.4 and 6.8 analyzed by DLS.

Responsiveness of DEAP–C16Y nanostructures to the tumor microenvironment

We further evaluated the responsiveness of DEAP–C16Y nanostructures to the acidic tumor microenvironment. A DEAP–C16Y nanoprobe was constructed by loading a hydrophobic fluorescence probe TRITC and its specific quencher BHQ-1 into the DEAP–C16Y nanostructure. We assumed that the nanoprobes were quenched at physiologic pH due to the proximity of fluorescent dyes and quenchers, which dissociated and transformed to the actively fluorescent state in the weakly acidic tumor tissue (Fig. 2A). The nanoprobe exhibited similar morphology and particle size to DEAP–C16Y nanostructures at pH 7.4, as well as structural collapse when the local pH was changed from pH 7.4 to pH 6.5 (Supplementary Fig. S1A). This was associated with a 4-fold increase in fluorescence intensity (Supplementary Fig. S1B, Fig. 2B), confirming that the nanoprobes were only activated at acidic pHs. A mixture of naked TRITC and BHQ-1 could not form nanostructures at pH 7.4 (Supplementary Fig. S1A) and its fluorescence intensity remained constant at different pHs (Supplementary Fig. S1C), suggesting that the fluorescence intensity of free dyes does not change with variations in pH. Next, we injected the nanoprobes intravenously into tumor-bearing mice and measured in vivo fluorescence at different time intervals (Fig. 2C and D). A distinct fluorescent signal was observed in the tumor 1 hour after administration, and its intensity steadily increased up to 3 hours. Thereafter, signal intensity gradually decreased and the distribution of the signal became less focal due to the metabolism of the released fluorescent dye. These observations illustrated that DEAP–C16Y nanostructures specifically disassembled in the weakly acidic tumor tissue.

Figure 2.
Figure 2.

Construction and imaging of the DEAP–C16Y nanoprobe. A, schematic design of the DEAP–C16Y nanoprobe. B, relative fluorescence intensity of the DEAP–C16Y nanoprobe at different pHs in PBS. The intensity at each pH was normalized to that of the TRITC and BHQ-1 mixture in PBS (pH = 6.5). C, in vivo imaging of DEAP–C16Y nanoprobes that were i.v. injected to tumor-bearing mice at different time intervals (right). Mice injected with PBS were used as the background control (left). D, quantification of the photon counts in tumors based on four replicates.

The antiangiogenic effects of DEAP–C16Y

As DEAP–C16Y is a modified form of C16Y, we next determined whether DEAP–C16Y retained the antiangiogenic activity. To exclude the possible toxicity induced by DEAP molecules, we designed a control peptide by conjugating identical DEAP molecules and amino acid residues to a fragment of human serum albumin (DEAP–HSAF: DEAP4-K3-L8-G2-DELRDEGKASSAKQ-G, Supplementary Fig. S2A), because peptides derived from the natural protein albumin may have low possibility to induce additional physiologic or pathologic responses. Meanwhile, the HSAF peptide has identical N-terminal amino acid residue with that of C16Y, and 50% of the amino acids in C16Y were also contained in the HSAF peptide. DEAP–HSAF formed nanostructures similar to those of DEAP–C16Y at pH 7.4, which also dissociated at pH 6.8 (Supplementary Fig. S2B). To investigate the antiangiogenic activity of DEAP–C16Y, we detected three major events of tumor angiogenesis: vascular endothelial cell viability, migration, and tubule formation. All peptides had a little effect on the viability of HUVECs, even at a concentration of 100 μmol/L (Fig. 3A). Interestingly, compared with DEAP–HSAF (50 μmol/L), both DEAP–C16Y (50 μmol/L) and C16Y (50 μmol/L) significantly retarded HUVEC migration as evidenced by the results from scratch migration (reduction: DEAP–C16Y, 49%; C16Y, 33%) and Transwell migration (reduction: DEAP–C16Y, 93%; C16Y, 89%; Fig. 3B and C and Supplementary Fig. S2C). The significant variation in the inhibition efficiency between the two assays indicated that DEAP–C16Y not only inhibited cell motility but also greatly disturbed cell adhesion. Moreover, both DEAP–C16Y and C16Y suppressed endothelial tubule network formation more than 80% (Fig. 3D and Supplementary Fig. S2D). DEAP–C16Y exhibited more dramatic effects than C16Y (Fig. 3B–D), likely due to the increased stability of the DEAP–C16Y peptide after the N-terminal protection (29). These results strongly suggested that DEAP–C16Y possessed more potent antiangiogenesis activity than C16Y. To investigate the molecular mechanism, we examined the downstream signaling pathways of integrin α5β1 and αvβ3 involved in endothelial cell migration (30). Phosphorylation of focal adhesion kinase (FAK), PI3K, and protein kinase B (Akt) were suppressed whereas extracellular signal–regulated kinases (ERK1/2) was not affected after DEAP–C16Y treatment (Fig. 3E), indicating that DEAP–C16Y inhibited endothelial cell function by hampering focal adhesion and the PI3K–Akt pathway.

Figure 3.
Figure 3.

Effects of DEAP–C16Y on vascular endothelial cells. A, viability of HUVECs treated by peptides at different concentrations for 24 hours. B, motility of HUVECs treated by peptides for 18 hours in a scratch migration assay. Relative wound closure width was calculated by normalizing mean wound closure width in each treatment to that in the PBS group. C, migration of HUVECs treated by peptides in a Transwell migration assay. Number of migrated cells in each group was quantified. D, tubule formation of HUVECs treated by C16Y or DEAP–C16Y. Total network length in each treatment group was quantified. E, Western blot analysis of integrin downstream molecules. Serum-starved HUVECs were pretreated with C16Y or DEAP–C16Y (50 μmol/L) for 30 minutes, and stimulated by 5% FBS for 30 minutes. The intensity of the bands on the Western blot analysis was analyzed by ImageJ, and the intensity ratio of each phosphorylated band to total protein band was calculated and quantified on the basis of three replicates (right). For above experiments, cells treated with PBS or DEAP–HSAF (50 μmol/L) served as negative controls; *, P < 0.05; **, P < 0.01 compared with the PBS group.

Effect of DEAP–C16Y on tumor cells

Because integrin α5β1 and αvβ3 are also highly expressed on the plasma membrane of tumor cells (31), we investigated the influence of DEAP–C16Y on tumor cells. Neither DEAP–C16Y nor C16Y affected the viability of MDA-MB-231 cells (Fig. 4A); however, both peptides (50 μmol/L) prevented tumor cell invasion in contrast with DEAP–HSAF that had no effect (DEAP–C16Y, 80%; C16Y, 66%; Fig. 4B). The inhibitory effect of DEAP–C16Y was consistently greater than that of C16Y. The formation of invadopodia is regarded as critical for tumor cell invasion, which involves polymerization of the actin cytoskeleton, cell adhesion, and ECM degradation (32). Thus, we studied invadopodia formation in MDA-MB-231 cells 12 hours post DEAP–C16Y treatment using the gelatin-degradation assay. Evident invadopodia were represented as an overlap of F-actin with gelatin clearing areas within merged images detected by confocal microscopy (Fig. 4C and Supplementary Fig. S2E). Obvious degraded areas were observed in both PBS and DEAP–HSAF–treated groups. In contrast, marginal gelatin degradation appeared when cells were incubated with either DEAP–C16Y or C16Y, suggesting that DEAP–C16Y could disturb the invadopodia formation. The structural organization of invadopodia requires cortactin, whereas membrane type 1 matrix metalloproteinase (MT1-MMP) is a key invadopodia enzyme for ECM degradation (33). The expression of cortactin and MT1-MMP was decreased by DEAP–C16Y (Fig. 4D), supporting that DEAP–C16Y inhibited invadopodia formation and tumor cell invasion.

Figure 4.
Figure 4.

Effects of DEAP–C16Y on tumor cells. A, viability of MDA-MB-231 cells treated with peptides for 24 hours. B, invasiveness of MDA-MB-231 cells treated by peptides. Migrated cells were quantified in each treatment group. C, gelatin degradation of MDA-MB-231 cells. MDA-MB-231 cells were seeded onto coverslips coated with FITC-gelatin (green) and incubated with peptides (50 μmol/L) in DMEM/5% FBS for 12 hours. Cells were immunostained by TRITC-conjugated phallodin for F-actin (red) and Hoechst 33342 for nucleus (blue); scale bars, 20 μm. D, Western blot analysis of invadopodia-related proteins. Serum-starved MDA-MB-231 cells were pretreated with peptides (50 μmol/L) for 30 minutes, stimulated by 5% FBS for 12 hours, and immunoblotted with indicated antibodies. Serum-starved cells and cells cultured in 10% FBS served as controls. The intensity of each band was analyzed by ImageJ, and intensity ratio of cortactin or MT1-MMP band to GAPDH band in each treatment was calculated, normalized to that in the PBS group, and quantified on the basis of three replicates; **, P < 0.01 compared with the PBS group.

Antitumor efficacy of DEAP–C16Y nanoformulation

As DEAP–C16Y nanostructures dissociated at weakly acidic pH, we hypothesized that systemically injected DEAP–C16Y nanostructures: (i) will maintain nanostructure in the circulation; (ii) will have an increased circulation time compared with C16Y; (iii) will accumulate in the tumor tissue due to the enhanced permeability and retention (EPR) effect; and (iv) will dissociate into peptide monomers in response to the acidic extracellular pH. Compared with administration of C16Y peptide, this strategy will allow more specific and efficient peptide accumulation at tumor sites and reduce the chance of degradation and clearance from systemic circulation, thereby enhancing the antitumor efficacy. To verify this hypothesis, we first compared the stability of DEAP–C16Y nanostructures and C16Y in blood circulation (25). Fluorescent molecules Cy5.5 were conjugated to the amino residues of C16Y (containing three amino residues) or DEAP–C16Y (containing two amino residues). Female BALB/c mice were i.v. injected with Cy5.5, Cy5.5–C16Y, or Cy5.5–DEAP–C16Y nanostructures. At different time points, blood samples were collected and their fluorescence intensity was measured using an in vivo imaging system. The fluorescence signal of Cy5.5–DEAP–C16Y preparation was greatly prolonged compared with that from Cy5.5–C16Y or Cy5.5 (Fig. 5A), indicating the increased circulation time of DEAP–C16Y nanostructures. To examine the tissue accumulation and distribution of DEAP–C16Y nanostructures, we injected the Cy5.5–DEAP–C16Y preparation intravenously to tumor-bearing mice. After 12 hours, tumors and major organs were removed for ex vivo imaging. The fluorescence signal accumulated mainly in the tumor, with some in the liver, kidney, and lung (Supplementary Fig. S3A). Next, we evaluated the antitumor effect of DEAP–C16Y nanostructures using the MDA-MB-231 xenograft model. When tumors grew to approximately 100 mm3 (on day 7), mice were divided into five groups (n = 5/group) and treated in one of the following ways: PBS, DEAP–HSAF nanostructures every third day, C16Y each day, DEAP–C16Y nanostructures every other day (DEAP–C16Y-2d), or DEAP–C16Y nanostructures every third day (DEAP–C16Y-3d) via the tail vein injection. The peptide concentration (6.5 μmol/kg) was determined on the basis of a previous study and was equivalent to 0.2 mg C16Y per mouse (9). This dosage also ensured the nanostructure at a stable state as its concentration in mouse blood circulation was much higher than the CMC of DEAP–C16Y. Mice were allowed to live for 8 days beyond the treatment period of 12 days. From the tumor growth curves and final tumor weights, both DEAP–C16Y and C16Y slowed-down tumor growth (Supplementary Fig. S3B and S3C). Compared with daily treatment of C16Y, DEAP–C16Y-2d was more effective in preventing tumor growth, whereas DEAP–C16Y-3d achieved similar antitumor effects. Immunohistochemical analysis of the endothelial cell marker CD31 in tumor sections revealed that DEAP–C16Y and C16Y decreased microvessel density, and this paralleled the inhibition of tumor growth (Supplementary Fig. S3D and S3E).

Figure 5.
Figure 5.

Antitumor effects of DEAP–C16Y nanostructures in vivo. A, blood clearance rate of DEAP–C16Y nanostructures. Cy5.5 molecules were conjugated to equal amounts of C16Y or DEAP–C16Y in PBS (pH 7.4). Cy5.5–DEAP–C16Y nanostructures were obtained by incubation at room temperature for 3 hours and dialysis for removal of free Cy5.5 molecules. BALB/c mice were i.v. injected with Cy5.5, Cy5.5–C16Y, or Cy5.5–DEAP–C16Y nanostructures. At 0, 1, 3, 5, 8, 12, and 24 hours, blood samples were collected and their fluorescence intensity was examined by in vivo imaging system. The fluorescence intensity at each time point was normalized to that at 0 hours. Three replicates were performed for quantification. B, growth curves of 4T1 tumors. Mice bearing 4T1 tumors were treated with DEAP–HSAF nanostructures, C16Y or DEAP–C16Y nanostructures for indicated period. Tumor volume was calculated every other day. C, weight of 4T1 tumors on day 27 when mice were sacrificed. D, quantification of microvessel density in 4T1 tumors analyzed by immunohistochemical staining of CD31. E, staining of lung sections from above 4T1-bearing mice by H&E or PCNA. T indicates metastatic foci with high density of positively stained nuclei and PCNA; scale bars, 200 μm. Number of metastasis foci per lung was quantified in F. And lung weight was quantified in G; *, P < 0.05; **, P < 0.01 compared with the DEAP–HSAF group.

To further validate the antitumor effects of DEAP–C16Y nanostructures, we used a syngeneic mammary tumor model by implanting highly metastatic 4T1 cells into mouse mammary fat pads. Mice received similar treatments to those bearing xenograft tumors (n = 5/treatment), and consistent inhibition trends of DEAP–C16Y nanostructures on tumor growth, tumor weights, and tumor vessels were observed (Fig. 5B–D). Because DEAP–C16Y suppressed tumor cell invasion, we further examined metastasis formation in the lung, a preferential site of metastasis for 4T1 tumors. Lung sections were stained by either hematoxylin and eosin (H&E) or proliferating cell nuclear antigen (PCNA) to visualize highly proliferating tumor regions in the lung. Apparent metastases were observed in DEAP–HSAF–treated mice, whereas C16Y or DEAP–C16Y significantly decreased the number of metastatic foci (C16Y, 72%; DEAP–C16Y-2d, 84%; DEAP–C16Y-3d, 71%; Fig. 5E and F). DEAP–C16Y-2d showed most potency in metastasis inhibition, which was likely caused by the prevention of tumor cell invasion and neovasculature formation. Moreover, lung weights of DEAP–C16Y– or C16Y-treated mice were consistently lower than those of DEAP–HSAF–treated mice (Fig. 5G). These results demonstrated that DEAP–C16Y nanostructures possessed improved stability and antitumor capacity compared with that of C16Y, and efficiently suppressed angiogenesis, tumor growth, and distant metastasis.

To reveal the possible adverse effects of DEAP–C16Y nanoformulation, we treated normal BALB/c mice with C16Y or DEAP–C16Y nanostructures for 14 days (n = 5/treatment). Morphology of major tissues was analyzed by H&E staining, and no obvious changes were observed after each treatment compared with tissues resected from untreated mice (Supplementary Fig. S4A). Importantly, there was no body weight loss in any drug-formulation–treated mice (Supplementary Fig. S4B).

Combination therapy of DEAP–C16Y and doxorubicin

Antiangiogenesis strategies are clinically used in combination with chemotherapy. Thus, we examined the therapeutic effect of a combination of DEAP–C16Y and doxorubicin. DEAP–C16Y or DEAP–HSAF coassembled with hydrophobic doxorubicin into spherical nanostructures (DEAP–C16Y–Dox and DEAP–HSAF–Dox) in PBS at pH 7.4, which disassembled at pH 6.8 (Supplementary Fig. S5A). Because DEAP–C16Y nanostructure consists of a hydrophobic core and a hydrophilic shell, we predicted that DEAP–C16Y and doxorubicin would coassemble by encapsulating hydrophobic doxorubicin into the core. The encapsulation and loading efficiency of doxorubicin is summarized in Supplementary Table S1. Drug release profiles demonstrated that the rate of doxorubicin release from DEAP–C16Y–Dox increased as the pH was decreased (Supplementary Fig. S5B). Next, we investigated the therapeutic efficacy of DEAP–C16Y–Dox using the 4T1 mammary tumor model. Compared with DEAP–HSAF, DEAP–C16Y, hydrophobic doxorubicin or DEAP–HSAF–Dox, DEAP–C16Y–Dox showed greater inhibition in tumor growth based on both tumor volume and tumor weight (Fig. 6A and B), suggesting that DEAP–C16Y–Dox combined the antiangiogenic effect of DEAP–C16Y and the cytotoxic effect of doxorubicin. Furthermore, staining of lung sections with H&E or PCNA showed that DEAP–C16Y–Dox decreased the number of metastasis foci by 92% (Fig. 6C and Supplementary Fig. S5C), which was more dramatic than that by each monotherapy (DEAP–C16Y: 71%; doxorubicin: 41%). This anti-metastatic effect was also confirmed by the measurement of lung weight (Supplementary Fig. S5D). DEAP–HSAF–Dox was more effective on suppressing tumor growth and lung metastasis than free doxorubicin, possibly due to the improved stability and bioavailability of DEAP—HSAF–Dox as a result of the EPR effect and the pH-responsive drug release. Altogether, these investigations demonstrated that the combination therapy of DEAP–C16Y and doxorubicin could achieve more profound inhibitory effect on tumor growth and metastasis.

Figure 6.
Figure 6.

Antitumor efficacy of DEAP–C16Y–Dox nanostructures. A, mice bearing 4T1 tumors were treated with indicated formulations every third day (peptide concentration: 6.5 μmol/kg; doxorubicin concentration: 5 μmol/kg). Tumor volume was calculated every other day. B, weight of 4T1 tumors in A on day 27. C, number of metastatic foci per lung from mice in A was quantified; *, P < 0.05; **, P < 0.01 compared with the DEAP–HSAF group. D, proposed antitumor mechanism of DEAP–C16Y nanostructures.

Discussion

Although multiple nanoformulations have been engineered for cancer treatment, most previous studies emphasize using multifunctional nanomaterials as drug vehicles. Our present work demonstrates a novel strategy for developing antitumor nanoformulations by using short therapeutic peptides. The amphiphilic peptide DEAP–C16Y exhibited a pH-responsive assembly/disassembly behavior that enabled it to remain intact and stable in the circulation, yet facilitated its dissociation to release functional peptides in tumor tissues. By binding to integrin receptors, DEAP–C16Y impaired endothelial cell function by inactivation of the FAK and PI3K–Akt pathways, and tumor cell invasion by decreasing the invadopodia formation. All these mechanisms ultimately resulted in the antitumor activity of DEAP–C16Y nanostructures (Fig. 6D). This represents a novel approach to enhancing the stability and bioavailability of therapeutic peptides, problems that are widely recognized as major challenges of peptide-based therapeutics for clinical applications.

At physiologic pH, DEAP–C16Y assembles into uniform micelles with an average diameter around 30 nm. This is a favorable property as nanostructures less than 100 nm have superior penetration and accumulation ability in tumors (34). When DEAP–C16Y nanoprobes were administered to tumor-bearing mice, a fluorescent signal was selectively located at the tumor site. This was possibly resulted from the EPR effect of the nano-sized peptide particles, and the fluorescence from separation of TRITC and its quencher following their liberation in the acidic tumor microenvironment. With tunable size and pH sensitivity, DEAP–C16Y nanoformulation offers great promise for targeted tumor therapy.

In our study, both C16Y and DEAP–C16Y exhibited negligible toxicity to endothelial cells and breast cancer cells, in contrast with a previous study reporting that C16Y inhibited ovarian cancer cell growth (35). This may simply reflect a cell-type difference. On the other hand, like others, we found that C16Y had no effect on the viability of HUVECs (9, 35). Compared with C16Y, DEAP–C16Y showed greater potency in the suppression of endothelial cell migration, tubule formation, and tumor cell invasion. Protection of the N-terminus of C16Y by DEAP molecules and leucine residues likely enhanced the peptide stability according to the N-end rule (29), and this in turn led to increased efficacy. It is also possible that some novel mechanisms elicited by the conjugation of chimeric molecules, such as binding to additional receptors, may also be operating. Moreover, our work, for the first time, discovered that DEAP–C16Y inhibited tumor cell invasion and invadopodia formation, which was in agreement with the former finding that the laminin-derived peptide C16 facilitated the invadopodia formation (36).

DEAP–C16Y had a more profound inhibitory effect on angiogenesis, tumor growth, and metastasis than free C16Y. Because DEAP–C16Y had low toxicity to tumor cells, the antitumor effect of DEAP–C16Y was most likely attributable to the angiogenesis blockade, which slows tumor growth but does not induce regression. Inhibition of tumor blood vessel formation and growth has the added benefit of providing reduced opportunities for tumor metastasis, and following DEAP–C16Y treatment, lung metastasis was dramatically decreased. On the basis of the tissue histology, DEAP–C16Y nanoformulation had no overt adverse effects on major tissues, a desirable feature of any therapeutics.

Besides their intrinsic antitumor actions, DEAP–C16Y nanostructures can carry hydrophobic chemotherapeutic drugs, resulting in combination therapy. Compared with each monotherapy, DEAP–C16Y–Dox achieved more significant antitumor efficacy, possibly due to the selective release and accumulation of doxorubicin and peptides in the tumor tissue.

In summary, we have constructed a pH-responsive peptide nanostructure that showed an innate capacity to inhibit tumor angiogenesis and invasion. This concept can be extended to investigate its application to other therapeutic peptides with distinct targets, and to explore the opportunities to use different peptides for combination therapy. On the basis of the Israelachvili's surfactant number theory, the ultimate morphology of an amphiphilic peptide is correlated with the length and volume of the hydrophobic tail(s) and the area of the hydrophilic head group (37). Amphiphilic peptides with larger hydrophilic head groups tend to form such spherical nanostructures. This study will help realize the goal of developing specific and effective tumor therapies that have limited side effects on normal tissues.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Authors' Contributions

Conception and design: Y. Ding, T. Ji, Y. Zhao, G. Nie

Development of methodology: Y. Ding, T. Ji, Y. Zhao, G. Nie

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): Y. Ding, T. Ji, Y. Zhang, Xiaozheng Zhao, R. Zhao, J. Lang, J. Shi

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): Y. Ding, T. Ji, Y. Zhang, Xiaozheng Zhao, R. Zhao, J. Lang, G. Nie

Writing, review, and/or revision of the manuscript: Y. Ding, T. Ji, Xiao Zhao, S. Sukumar, G. Nie

Study supervision: G. Nie

Grant Support

This work was supported by the grants from MoST 973 (2012CB934004 and 2011CB933400; to G. Nie), the National Natural Science Foundation of China (31325010; to G. Nie, 31300822; to Y. Ding, 51203032 and 21373067; to Y. Zhao), and the Key Research Program of the Chinese Academy of Sciences (KGZD-EW-T06; to G. Nie).

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.

Acknowledgments

The authors thank Professor Xiyun Yan from Institute of Biophysics, Chinese Academy of Sciences for providing HUVECs. The authors express special thanks to Professor Gregory Jon Anderson for revising the article and Professor Lajos Balogh for providing suggestions.

Footnotes

  • Note: Supplementary data for this article are available at Molecular Cancer Therapeutics Online (http://mct.aacrjournals.org/).

  • Received June 10, 2015.
  • Revision received July 22, 2015.
  • Accepted July 22, 2015.

References

  1. 1.↵
    1. Hanahan D,
    2. Weinberg RA
    . Hallmarks of cancer: the next generation. Cell 2011;144:646–74.
    OpenUrlCrossRefPubMed
  2. 2.↵
    1. Weis SM,
    2. Cheresh DA
    . Tumor angiogenesis: molecular pathways and therapeutic targets. Nat Med 2011;17:1359–70.
    OpenUrlCrossRefPubMed
  3. 3.↵
    1. Chung AS,
    2. Lee J,
    3. Ferrara N
    . Targeting the tumour vasculature: insights from physiological angiogenesis. Nat Rev Cancer 2010;10:505–14.
    OpenUrlCrossRefPubMed
  4. 4.↵
    1. Valastyan S,
    2. Weinberg RA
    . Tumor metastasis: molecular insights and evolving paradigms. Cell 2011;147:275–92.
    OpenUrlCrossRefPubMed
  5. 5.↵
    1. Vlieghe P,
    2. Lisowski V,
    3. Martinez J,
    4. Khrestchatisky M
    . Synthetic therapeutic peptides: science and market. Drug Discov Today 2010;15:40–56.
    OpenUrlCrossRefPubMed
  6. 6.↵
    1. Kaspar AA,
    2. Reichert JM
    . Future directions for peptide therapeutics development. Drug Discov Today 2013;18:807–17.
    OpenUrlCrossRefPubMed
  7. 7.↵
    1. Monk BJ,
    2. Poveda A,
    3. Vergote I,
    4. Raspagliesi F,
    5. Fujiwara K,
    6. Bae DS,
    7. et al.
    Anti-angiopoietin therapy with trebananib for recurrent ovarian cancer (TRINOVA-1): a randomised, multicentre, double-blind, placebo-controlled phase 3 trial. Lancet Oncol 2014;15:799–808.
    OpenUrlCrossRefPubMed
  8. 8.↵
    1. Rosca EV,
    2. Koskimaki JE,
    3. Rivera CG,
    4. Pandey NB,
    5. Tamiz AP,
    6. Popel AS
    . Anti-angiogenic peptides for cancer therapeutics. Curr Pharm Biotechnol 2011;12:1101–16.
    OpenUrlCrossRefPubMed
  9. 9.↵
    1. Ponce ML,
    2. Hibino S,
    3. Lebioda AM,
    4. Mochizuki M,
    5. Nomizu M,
    6. Kleinman HK
    . Identification of a potent peptide antagonist to an active laminin-1 sequence that blocks angiogenesis and tumor growth. Cancer Res 2003;63:5060–4.
    OpenUrlAbstract/FREE Full Text
  10. 10.↵
    1. Kim H,
    2. Csaky KG
    . Nanoparticle-integrin antagonist C16Y peptide treatment of choroidal neovascularization in rats. J Control Release 2010;142:286–93.
    OpenUrlCrossRefPubMed
  11. 11.↵
    1. Wang J,
    2. Sui M,
    3. Fan W
    . Nanoparticles for tumor targeted therapies and their pharmacokinetics. Curr Drug Metab 2010;11:129–41.
    OpenUrlCrossRefPubMed
  12. 12.↵
    1. Cui H,
    2. Webber MJ,
    3. Stupp SI
    . Self-assembly of peptide amphiphiles: from molecules to nanostructures to biomaterials. Biopolymers 2010;94:1–18.
    OpenUrlCrossRefPubMed
  13. 13.↵
    1. Sundar S,
    2. Chen Y,
    3. Tong YW
    . Delivery of therapeutics and molecules using self-assembled peptides. Curr Med Chem 2014;21:2469–79.
    OpenUrlCrossRefPubMed
  14. 14.↵
    1. Soukasene S,
    2. Toft DJ,
    3. Moyer TJ,
    4. Lu H,
    5. Lee HK,
    6. Standley SM,
    7. et al.
    Antitumor activity of peptide amphiphile nanofiber-encapsulated camptothecin. ACS Nano 2011;5:9113–21.
    OpenUrlCrossRefPubMed
  15. 15.↵
    1. Standley SM,
    2. Toft DJ,
    3. Cheng H,
    4. Soukasene S,
    5. Chen J,
    6. Raja SM,
    7. et al.
    Induction of cancer cell death by self-assembling nanostructures incorporating a cytotoxic peptide. Cancer Res 2010;70:3020–6.
    OpenUrlAbstract/FREE Full Text
  16. 16.↵
    1. Zha RH,
    2. Sur S,
    3. Stupp SI
    . Self-assembly of cytotoxic peptide amphiphiles into supramolecular membranes for cancer therapy. Adv Healthc Mater 2013;2:126–33.
    OpenUrlCrossRefPubMed
  17. 17.↵
    1. Zhang P,
    2. Cheetham AG,
    3. Lin YA,
    4. Cui H
    . Self-assembled Tat nanofibers as effective drug carrier and transporter. ACS Nano 2013;7:5965–77.
    OpenUrlCrossRefPubMed
  18. 18.↵
    1. Han K,
    2. Chen S,
    3. Chen WH,
    4. Lei Q,
    5. Liu Y,
    6. Zhuo RX,
    7. et al.
    Synergistic gene and drug tumor therapy using a chimeric peptide. Biomaterials 2013;34:4680–9.
    OpenUrlCrossRefPubMed
  19. 19.↵
    1. Tarasov SG,
    2. Gaponenko V,
    3. Howard OM,
    4. Chen Y,
    5. Oppenheim JJ,
    6. Dyba MA,
    7. et al.
    Structural plasticity of a transmembrane peptide allows self-assembly into biologically active nanoparticles. Proc Natl Acad Sci U S A 2011;108:9798–803.
    OpenUrlAbstract/FREE Full Text
  20. 20.↵
    1. Zhao Y,
    2. Ji T,
    3. Wang H,
    4. Li S,
    5. Zhao Y,
    6. Nie G
    . Self-assembled peptide nanoparticles as tumor microenvironment activatable probes for tumor targeting and imaging. J Control Release 2014;177:11–9.
    OpenUrlCrossRefPubMed
  21. 21.↵
    1. Wang H,
    2. Zhao Y,
    3. Wu Y,
    4. Hu YL,
    5. Nan K,
    6. Nie G,
    7. et al.
    Enhanced anti-tumor efficacy by co-delivery of doxorubicin and paclitaxel with amphiphilic methoxy PEG-PLGA copolymer nanoparticles. Biomaterials 2011;32:8281–90.
    OpenUrlCrossRefPubMed
  22. 22.↵
    1. Ding Y,
    2. Song N,
    3. Liu C,
    4. He T,
    5. Zhuo W,
    6. He X,
    7. et al.
    Heat shock cognate 70 regulates the translocation and angiogenic function of nucleolin. Arterioscler Thromb Vasc Biol 2012;32:e126–34.
    OpenUrlAbstract/FREE Full Text
  23. 23.↵
    1. Wang X,
    2. Song X,
    3. Zhuo W,
    4. Fu Y,
    5. Shi H,
    6. Liang Y,
    7. et al.
    The regulatory mechanism of Hsp90alpha secretion and its function in tumor malignancy. Proc Natl Acad Sci U S A 2009;106:21288–93.
    OpenUrlAbstract/FREE Full Text
  24. 24.↵
    1. Yang J,
    2. Song X,
    3. Chen Y,
    4. Lu XA,
    5. Fu Y,
    6. Luo Y
    . PLCγ1-PKCγ signaling-mediated Hsp90α plasma membrane translocation facilitates tumor metastasis. Traffic 2014;15:861–78.
    OpenUrlCrossRefPubMed
  25. 25.↵
    1. Ji T,
    2. Ding Y,
    3. Zhao Y,
    4. Wang J,
    5. Qin H,
    6. Liu X,
    7. et al.
    Peptide assembly integration of fibroblast-targeting and cell-penetration features for enhanced antitumor drug delivery. Adv Mater 2015;27:1865–73.
    OpenUrlCrossRefPubMed
  26. 26.↵
    1. Park SY,
    2. Baik HJ,
    3. Oh YT,
    4. Oh KT,
    5. Youn YS,
    6. Lee ES
    . A smart polysaccharide/drug conjugate for photodynamic therapy. Angew Chem Int Ed Engl 2011;50:1644–7.
    OpenUrlCrossRefPubMed
  27. 27.↵
    1. Baik HJ,
    2. Oh NM,
    3. Oh YT,
    4. Yoo NY,
    5. Park SY,
    6. Oh KT,
    7. et al.
    3-Diethylaminopropyl-bearing glycol chitosan as a protein drug carrier. Colloids Surf B Biointerfaces 2011;84:585–90.
    OpenUrlCrossRefPubMed
  28. 28.↵
    1. Webb BA,
    2. Chimenti M,
    3. Jacobson MP,
    4. Barber DL
    . Dysregulated pH: a perfect storm for cancer progression. Nat Rev Cancer 2011;11:671–7.
    OpenUrlCrossRefPubMed
  29. 29.↵
    1. Bachmair A,
    2. Finley D,
    3. Varshavsky A
    . In vivo half-life of a protein is a function of its amino-terminal residue. Science 1986;234:179–86.
    OpenUrlAbstract/FREE Full Text
  30. 30.↵
    1. Avraamides CJ,
    2. Garmy-Susini B,
    3. Varner JA
    . Integrins in angiogenesis and lymphangiogenesis. Nat Rev Cancer 2008;8:604–17.
    OpenUrlCrossRefPubMed
  31. 31.↵
    1. Desgrosellier JS,
    2. Cheresh DA
    . Integrins in cancer: biological implications and therapeutic opportunities. Nat Rev Cancer 2010;10:9–22.
    OpenUrlCrossRefPubMed
  32. 32.↵
    1. Sibony-Benyamini H,
    2. Gil-Henn H
    . Invadopodia: the leading force. Eur J Cell Biol 2012;91:896–901.
    OpenUrlCrossRefPubMed
  33. 33.↵
    1. Artym VV,
    2. Zhang Y,
    3. Seillier-Moiseiwitsch F,
    4. Yamada KM,
    5. Mueller SC
    . Dynamic interactions of cortactin and membrane type 1 matrix metalloproteinase at invadopodia: defining the stages of invadopodia formation and function. Cancer Res 2006;66:3034–43.
    OpenUrlAbstract/FREE Full Text
  34. 34.↵
    1. Bourzac K
    . Nanotechnology: carrying drugs. Nature 2012;491:S58–S60.
    OpenUrlCrossRefPubMed
  35. 35.↵
    1. Shinagawa A,
    2. Yoshida Y,
    3. Kurokawa T,
    4. Horiuchi Y,
    5. Tsuyoshi H,
    6. Orisaka M,
    7. et al.
    The potent peptide antagonist to angiogenesis, C16Y, and cisplatin act synergistically in the down-regulation of the Bcl-2/Bax ratio and the induction of apoptosis in human ovarian cancer cells. Int J Oncol 2011;39:1359–64.
    OpenUrlPubMed
  36. 36.↵
    1. Nascimento CF,
    2. de Siqueira AS,
    3. Pinheiro JJ,
    4. Freitas VM,
    5. Jaeger RG
    . Laminin-111 derived peptides AG73 and C16 regulate invadopodia activity of a human adenoid cystic carcinoma cell line. Exp Cell Res 2011;317:2562–72.
    OpenUrlCrossRefPubMed
  37. 37.↵
    1. Israelachvili J
    . Soft and Biological Structures. Intermolecular and surface forces. 3rd ed. Boston: Academic Press; 2011. p. 535–76.
View Abstract
Back to top
Molecular Cancer Therapeutics: 14 (10)
October 2015
Volume 14, Issue 10

Sign up for alerts

View this article with LENS

Open full page PDF
Article Alerts
Email Article
Citation Tools
Improvement of Stability and Efficacy of C16Y Therapeutic Peptide via Molecular Self-Assembly into Tumor-Responsive Nanoformulation
Yanping Ding, Tianjiao Ji, Ying Zhao, Yinlong Zhang, Xiaozheng Zhao, Ruifang Zhao, Jiayan Lang, Xiao Zhao, Jian Shi, Saraswati Sukumar and Guangjun Nie
Mol Cancer Ther October 1 2015 (14) (10) 2390-2400; DOI: 10.1158/1535-7163.MCT-15-0484
del.icio.us logo Digg logo Reddit logo Twitter logo CiteULike logo Facebook logo Google logo Mendeley logo

Jump to section

Advertisement