Characterization of the synthesized naked and peptide-conjugated GNPs

GNPs were prepared as described previously [19, 20]. Characterization of GNPs using the UV–Vis spectra (400–700 nm) showed an absorption band shift towards higher wavelengths (red shift), indicating the increase of GNP size after surface modification. The synthesized GNPs showed a strong absorption at λmax ~ 516 nm, which is attributable to globular GNPs with diameters lower than 20 nm [31, 32]. Addition of MU/MUA to the surface of GNPs led to a notable red shift from λmax ~ 516 nm to 522 nm. After conjugation of VGB3 to the modified GNPs, a considerable shift from λmax ~ 522 nm to around ~ 550 nm and a new single absorption band were emerged at λmax ~ 280 nm, which the later can be associated with the tryptophan residue in the structure peptide (Fig. 2a). Based on measurements by UV standard curve, we have detected no unconjugated peptide fraction in the reaction mixture.

According to the DLS graphs (Fig. 2b), the mean diameter of GNPs was 16 nm and the size distribution was between 7–40 nm, whereas the average diameter of GNP-peptide was 30, indicating a slightly more variation than free GNP. The value of Zeta potential, an indicator of dispersion stability and the tendency of GNPs to aggregate in solution, indicated that GNPs were negatively charged, which is due to the citrate ions. Immobilization of the positively-charged peptide molecules onto highly-negatively charged (− 51.4 ± 0.88 mv) GNPs decreased the negative charge to − 23.9 ± 0.55 mv (Fig. 2c).

Fig. 2

The characterization of the synthesized naked and conjugated GNPs. a UV–Vis absorption spectra of GNPs before and after modifications. b DLS curves of GNPs and GNP–VGB3 related to hydrodynamic diameters. c The zeta potentials of GNPs and GNP–VGB3. d FT-IR spectra of GNP and VGB3 and their conjugations in the 400–4000 cm−1 region. e AFM of GNP and GNP–VGB3. (1, 2) AFM analysis depicted the mean diameter of GNPs in 3D and 2D images, respectively (scale bar: 500 nm), and (3, 4) In the same way, the obtained images from GNP–VGB3 were dedicated within a scan area of 0.5 μm × 0.5 μm. f TEM images to verify the size, shape and the size distributions of GNPs and GNP–VGB3, respectively (scale bar: 50 nm) and g EDS analysis of GNPs-VGB3 and its mapping to identify each element in the sample (Au: red; C: green, and O: Violet)

Next, the FAAS and ICP-MS analysis was performed to achieve the gold concentration. Accordingly, the Au concentration was 48 mg L−1 for free GNP and 19 mg L−1 for GNP–VGB3 (data not shown).

The FT-IR spectra enable us to address the functional groups in the structure of synthesized peptide, MU/MUA modified GNP, and the peptide immobilized on modified GNPs. As indicated in Fig. 2d, VGB3 was bonded from its NH2 group to the –COOH functional groups of MU/MUA modified GNP. In MUA modified GNP, the absorption bands in 3414, 1730 and 1279 cm−1 can be related to the stretching vibrations of hydroxyl group of COOH, –C=O and –C–O bands in the MU/MUA linker, respectively. The band at 2927 cm−1 was assigned to the stretching vibrations of –C–H in –CH2– groups of MU/MUA. The broad intense absorption band in the wavenumbers of 3367 cm−1 (in VGB3 spectrum) and 3412 cm−1 (in GNP–VGB3 spectrum) was related to the vibrations of the hydroxyl group of –COOH and –NH groups in the peptide [24]. The stretching vibrations of amidic –C=O band in the peptide were observed at 1676 cm−1 (for unconjugated peptide) that overlapped by the vibrations of the carbonyl group of COOH. In both VGB3 and GNP–VGB3 spectra, the stretching vibrations of amidic –NH was occurred in 3150–3350 cm−1 that overlapped with –COOH band and the vibrations of –C–S band was appeared in 600–700 cm−1. The bands in 1207 and 2982 cm−1 can be assigned to the vibrations of –C–N of amid group and –C–H in CH2, respectively. Also, the bending vibrations of NH observed in 1548 cm−1 represented the amide bands. In the GNP–VGB3 spectra, reduction in the band intensity of the carbonyl group of COOH in 1730 cm−1 and increase in the intensity of the band at 1641 cm−1 is related to amidic –C=O vibrations, indicating the formation of amid band by the covalent linkage between the NH2 group of peptide with –COOH group of MU/MUA conjugated GNP [23, 33].

Figure 2e(1–4) shows atomic force microscopy (AFM) images of naked GNP and GNP-peptide within a scan area of 0.50 μm × 0.50 μm for both samples. In three dimensional images, the average size of GNPs and GNP-peptide were estimated about 15 ± 4 nm and 39 ± 4 nm, respectively [25].

TEM and FESEM were utilized to specify the morphology of NPs. According to the results obtained from TEM, the shape of free GNPs and GNP-peptide were spherical with a monodispersed size from 5–12 nm (Fig. 2f). The differences in the mean diameters of nanoparticles obtained by TEM and DLS could be due to the methods of size measurement. FESEM images of GNP-peptide is shown in Additional file 1: Fig S2 that displayed surface morphology of this sample. Compared to GNP–VGB3 displaying the mean size of 25 nm, free GNP displayed much larger size, reflecting the susceptibility of free GNPs for aggregation. Based on EDS analysis, the fundamental elements of the nanoparticles were the Au element (92.2%), C and O elements (6.8 and 1.1%, respectively). The presence of C and O elements is attributable to the MU/MUA linkers, and the peptide molecules on the surface of GNPs.

GNP-VGB3 recognizes and neutralizes VEGFR1 and VEGFR2

VEGFR2 and VEGFR1 are highly expressed on the surface of endotheliral cells (ECs) [34]. The specific cell binding of GNP–VGB3 to VEGFR1 and VEGFR2 were investigated by immunocytochemical assay using human umbilical vein endothelial cells (HUVECs). Pre-incubation of HUVECs by increasing concentrations of free VGB3 (500, 700 and 1200 ng mL−1) and GNP–VGB3 (250, 500, 1000 ng mL−1) reduced binding of fluorescently labeled anti-VEGFR1 or anti-VEGFR2 (20 ng mL−1), whereas fluorescence intensities was not affected when HUVECs treated with GNP (Additional file 1: Fig. S3). These results indicate that VGB3 retained its ability to recognize VEGFR1 and VEGFR2 after conjugation to the gold nanoparticles (Fig. 3a, b). In addition, GNP–VGB3 treatment inhibited VEGF-induced phosphorylation of VEGFR2 and VGEFR1. As indicated in Fig. 3c, d, incubation with GNP–VGB3 (1000 ng mL−1) as well as VGB3 (1200 ng mL−1) reduced fluorescent signals of anti-phospho VEGFR1 (anti-pVEGFR1) or anti-pVEGFR2 in a dose-dependent manner compared to controls, whereas binding of the antibodies was not affected by GNP (Additional file 1: Fig S4). These results indicate that GNP–VGB3 abrogated the VEGF-induced activation (phosphorylation) of VEGFR1 and VEGFR2.

Fig. 3

GNP-VGB3 recognizes VEGFR1 and VEGFR2 and suppresses their VEGF-induced phosphorylation in endothelial cells. a Immunocytochemical images of HUVE cells treated with PBS, GNP, VGB3 and GNP–VGB3 using FITC-secondary anti-mouse antibody (green) to bind to VEGFR1 (left) and VEGFR2 (right) (scale bar: 20 μm). b Statistical analysis of VEGFR1/2 fluorescence intensity under various concentrations were performed by prism software 8; Oneway ANOVA and all data displayed mean ± SEM (n = 3). c Immunoflourescent staining images (PE-secondary anti-mouse antibody (red)) of phosphorylated-VGEFR1 (p-VEGFR1) and p-VEGFR2 with various treatments. d Quantitative analysis of the fluorescence intensity of p-VEGFR1/2 by prism software 8 analyzed by One-way ANOVA for different treatments. (****P < 0.0001, ***P < 0.001 and NS: not significant in comparison with control)

Inhibition of proliferation and migration of endothelial and tumor cells

VEGFR1 and VEGFR2 undergo dimerization and VEGFR ligand-dependent phosphorylation, which trigger mitogenic, chemotactic, and prosurvival signals, along with stimulation of tumor vessel formation [35]. We have previously indicated that VGB3 can inhibit proliferation, migration and tube formation of HUVECs, and proliferation of 4T1 mammary carcinoma tumor cells that expresses both VEGFR1 and VEGFR2 [13]. Here, to confirm whether VGB3 retained its effects in nanoformulation, the antiproliferative and antimigrative properties of GNP–VGB3 were determined in comparison to GNP and VGB3 in HUVE and 4T1 cells when stimulated by VEGF (20 ng mL−1). Notably, blank-GNP had no significant effects on the cell viability and showed a similar result to the non-treated cells, inferring that the blank-GNP composition is biocompatible (Fig. 4a). In contrast, VGB3 and GNP–VGB3 exhibited time and dose-dependent cytotoxicity; the half-maximal inhibition (IC50) values of GNP–VGB3 against HUVECs were 554 ng mL−1 (24 h) and 440 ng mL−1 (48 h) and for free VGB3 were 710 ng mL−1 (24 h) and 561 ng mL−1 (48 h). Similarly, the data of 4T1 cells demonstrate that the IC50 values of GNP–VGB3 were 1238 ng mL−1 (24 h) and 423 ng mL−1 (48 h) and for free VGB3 were 1971 ng mL−1 (24 h) and 771 ng mL−1 (48 h). These results in agreement with previous studies indicated that the cytotoxicity of peptide-conjugated GNP was higher than that of free peptides [36, 37].

Fig. 4

Inhibition of HUVE and 4T1 cells proliferation, migration and cell cycle progression. a The investigation of cell viability in the HUVECs and 4T1 cells after treatment with different concentrations (0–1000 ng mL−1) of GNPs, VGB3, and GNP–VGB3 in the presence of VEGF (20 ng mL−1). Also, the cell viability rates of 4T1 cells after treatments with different concentrations (0–2000 ng mL−1) of GNPs, VGB3 and GNP–VGB3 after 24 and 48 h incubation. The cell viability rates of treated cells were evaluated by MTT assay for 24 and 48 h incubation using prism software 8 (One-way ANOVA method, based on mean ± SEM of six independent observations). b HUVE and 4T1 cells wound closure cell migration assay after different treatment with PBS, GNPs, VGB3, or GNP–VGB3 in the presence of VEGF (20 ng mL−1) after incubation for 0, 24 or 48 h. By Wimasis image analysis, wound areas in the images were shown with a gray color. c Statistical analysis of wound area of HUVE and 4T1 cells after treatments by One-way ANOVA, mean ± SD, n = 3). d Cell cycle analysis of HUVE and 4T1 cells after exposure to PBS, GNPs, VGB3, and GNP–VGB3 and VI stain followed by flow cytometry analysis. e Cell cycle distribution of HUVE and 4T1 treated cells were analyzed statistically by One-way ANOVA pathway (mean ± SEM, and n = 3, ****P < 0.0001, **P < 0.01, *P < 0.05, or ns: not significant)

Cell migration is essential for angiogenesis of endothelial cells and invasion of tumor cells. We carried out wound healing assay to investigate the antimigrative effect of GNP–VGB3 in HUVE and 4T1 cells. Based on the above mentioned IC50 values, endothelial cells were incubated with VGB3 (710 ng mL−1), GNP–VGB3 (554 ng mL−1) for 24 and 48 h, and filling the scratch area with cells was evaluated compared to controls and GNP-treated  cells. When stimulated by VEGF (20 ng mL−1), HUVE and 4T1 cells were able to fill the wound area after 24 h. However, the VEGF-induced migrations were inhibited by VGB3 and GNP–VGB3, and to a lesser extent by GNP after 24 h and suppression was maximal in GNP–VGB3-treated groups (P < 0.0001). After 48 h, whereas GNP-treated HUVE and 4T1 cells completely migrated to the wound area, VGB3 and GNP–VGB3 potently inhibited VEGF-induced migration of cells compared to controls and GNP-treated cells (P < 0.0001). Results of wound healing assay indicated that GNP–VGB3 could suppress endothelial and tumor cells locomotion in response to growth factor-attractive surroundings (Fig. 4b, c).

The induction of cell cycle arrest is a strategy to control aberrant cancer cell proliferation [38]. Hence, to investigate the mechanism of antiproliferative effects, we evaluated the cell cycle distribution using flow cytometry. Cell cycle analysis of HUVECs and 4T1 cells exposed to GNP, VGB3 and GNP–VGB3 is shown in Fig. 4d. Consistent with the results of proliferation and migration, treatment with VGB3 and GNP–VGB3 resulted in an arrest in the G2/M phase, with a significant decrease in G0/G1 phase versus control cells, whereas GNP was ineffective in 4T1 cells or moderately (P < 0.05) effective in HUVECs. Importantly, the accumulation in G2/M phase was significantly higher in HUVE and 4T1 cells treated GNP–VGB3 (40.3 and 97.8%, respectively) than in cells treated with free peptide (34.4 and 30.3%, respectively)(Fig. 4e), suggesting that binding to GNP induced the inhibitory effects of VGB3. These results are consistent with the data obtained by MTT and scratch analyses and provide further evidence for enhanced potency of VGB3 after as a result of ligation to GNPs.

Induction of ROS production and apoptosis in endothelial and tumor cells

Inhibition of VEGF binding to VEGFRs on the endothelial and 4T1 cells results in apoptosis induction [39]. On the other hand, ROS overproduction can activate the apoptotic signaling pathways and cell death [40]. We therefore evaluated the potential of GNP–VGB3 to induce ROS overproduction followed and apoptosis induction in VEGF-induced endothelial and tumor cells. First, the intracellular ROS production was measured in HUVECs and 4T1 cells after treatment with free GNPs, free VGB3, and GNP–VGB3. When treated with concentrations equal to IC50 values and in the presence of VEGF (20 ng mL−1), the fluorescence intensity of 2-7-dichlorofluorescin diacetate (DCFDA) as ROS production probe in response to GNP–VGB3 treatment were 40.98 and 19.03 in HUVECs and 4T1 cells, respectively, which were significantly higher than that of GNP (16.11 and 8.00, respectively), free peptide (27.28 and 15.69, respectively) and untreated cells (13.8 and 5.24, respectively) (Fig. 5a, b), indicating that neutralization of VEGF receptors led to the overproduction of ROS.

Fig. 5

ROS overproduction and apoptosis induction in HUVE and 4T1 cells. a Intracellular ROS induction of HUVE and 4T1 cells treated with PBS, GNPs, VGB3 and GNP–VGB3 in the presence of VEGF (20 ng mL−1) specified after staining with DCFH-DA by flow cytometry (scale bar: 20 μm). b Detection of ROS based on fluorescence intensity using prism 8.0 software in HUVE and 4T1 cells after treatments. c Flow cytograms of cell apoptosis in HUVE and 4T1 cells induced by PBS, GNPs, VGB3, and GNP–VGB3 in the presence of VEGF (20 ng mL−1) using Annexin V/PI staining. d The total cell apoptosis in HUVE and 4T1 treated cells were obtained from the sum of early and late apoptosis, which placed at the corner of each panel (in lower-right (Annexin V-FITC+, PI−), and upper-right (Annexin V-FITC+, PI+) quadrants, respectively) and represented in the diagram using prism software. (All data analyzed based on mean ± SEM; One-way ANOVA; n = 3; number sign symbol (#) was used for comparing the treatments with control, and asterisk symbol (*) was used for comparison between treatments, ****P < 0.0001, ***P < 0.001, **P < 0.01, *P < 0.05, or ns: not significant)

To test GNP–VGB3-mediated induction of apoptosis, we conducted Annexin V and propidium iodide (PI) stainings [41] in VEGF (20 ng mL−1)-induced HUVE and 4T1 cells treated with GNP, VGB3 and GNP–VGB3. As indicated in Fig. 5c and d, the percentage of apoptotic cells remarkably increased from 30.98 and 50.83% in GNP- and VGB3-treated cells, respectively, to 74.50% in GNP–VGB3-treated HUVECs. More strikingly, the proportion of apoptotic 4T1 cells were increased in GNP–VGB3-treated group (82.00%) compared to the cells treated by VGB3 (12.35%) and GNP (1.92%) (Fig. 5c, d).

Inhibition of breast tumor growth in mice

To determine whether the superior in vitro potency of GNP–VGB3 over free peptide and GNP is recapitulated in vivo, mice harboring 4T1 mammary carcinoma tumors, which is known as a VEGF-dependent model, were treated with GNP–VGB3, GNP, free peptide or phosphate buffer saline (PBS) as control. When tumor size reached an average volume of ~ 100 mm3, different treatment groups intravascularly (i.v.) injected once a week for 3 weeks (Fig. 6a), and during this period, tumor volume, body weight and survival curve were measured. On day 32, the average tumor volume in the GNP–VGB3-treated group (1062 mm3) was significantly lower than in PBS (1828 mm3), GNP (1605 mm3), and free VGB3 (1485 mm3). These results indicate that tumor regression occurred in both VGB3 and GNP–VGB3-treated groups, but GNP–VGB3 was significantly more effective. Notably, GNP group showed no tumor growth inhibition (Fig. 6b).

Fig. 6

In vivo antitumor activity of PBS, GNP, VGB3, and GNP–VGB3 in murine 4T1 mammary carcinoma tumor model. a Schedule for animal experiments. b Tumor growth inhibition; lines, mean tumor volume for each group of six animals per group;  error bars signify ± SEM. n = 6; **P < 0.01, ***P < 0.001; two-way ANOVA (c) The survival rates of 4T1-bearing mice after treatment were illustrated by Kaplan–Meier curves., and d Body weight measurements taken every 5 days until day 32, presented as mean ± SEM. All groups compared with control and analyzed by prism by two-way ANOVA statistical analysis, mean ± SEM and n = 6 (***P < 0.001, **P < 0.01, ns: not significant). e [18F]-fluorodeoxyglucose (18F-FDG) PET imaging of mice treated with GNP, VGB3, GNP–VGB3 or PBS at day 32. Representative PET images are shown with arrows indicating 4T1 mammary carcinoma tumors

To further assess the in vivo efficacy of treatments, animal survivals was followed up in the treatment groups (n = 6) and the results were compared with PBS-treated controls. The survival curve deduced from Kaplan–Meier analysis until the day 32 after implantation indicated that GNP–VGB3 (one mouse dead; 83.2% survival at day 32) prolonged the survival rate more than VGB3 (two mice dead; 64% survival at the day 32) and GNP (four mice dead; 16% survival at day 32) (Fig. 6c). All members of the control group were lost before day 32. In addition, the body weight of all animals was increased during the treatment period (Fig. 6d), suggesting that the treatments are nontoxic at the dosages used in this work.

To evaluate the effect of treatments on the tumor progression in Balb/c mice, we performed 18F-FDG-PET imaging at the end of treatments (day 32). The FDG-PET images of Balb/c mice is presented in Fig. 6e. After treatment for 4 weeks (day 32), the mean uptake values of 18F-FDG was more significantly decreased in GNP–VGB3 group than VGB3 group compared within the group that received PBS, whereas the uptake value did not change in GNP group. These results provided strong suggestive evidence that the VGB3-mediated inhibition of tumor progression is improved by gold nanoformulation.

Suppression of VEGFR-1/-2-mediated signaling in 4T1 mammary carcinoma tumors

Our recent study demonstrated that VGB3/48 h is an effective treatment for metastatic murine 4T1 mammary carcinoma tumors [6] through the inhibition of tumor cells proliferation (decreased Ki-67 expression), angiogenesis (decreased expression of CD31 and CD34), and the induction of apoptosis in tumors (increased TUNEL staining and p53 expression, and decreased Bcl-2 expression) [13]. Results of the current study revealed that weekly treatments attenuate the peptide efficacy, but conjugation to GNP significantly improved its antitumor properties. To explore the molecular mechanisms underlying the superior antitumor effects of GNP–VGB3 against 4T1 mammary carcinoma tumors compared with free peptide, tumors were harvested at the end of the treatment period (day 32 after implantation) and the VEGFR1/R2 signaling pathways were assessed by western blot.

First, tumor lysates were analyzed for total and phosphorylated VEGFR-1 and VEGFR-2. VEGFR1 and VEGFR2 expressions were abundant in 4T1 tumors, and their expression levels were roughly equivalent in untreated tumors (Fig. 7a). This result, in accordance to our previous investigations [6, 8, 9, 13], confirms that 4T1 model is appropriate for investigation of responses to VEGFR1 and VEGFR2 inhibition. Obviously, VEGFR1 and VEGFR2 expression levels were much more effectively inhibited when weekly treated by GNP–VGB3 (P < 0.0001) than VGB3 (P < 0.01) and GNP compared to controls.

Fig. 7

Inhibition of VEGFR1/2-mediated signaling after treatment with PBS, GNP, VGB3 or GNP–VGB3 in 4T1 mammary carcinoma tumor-bearing mice. All of the measurements were done at the end of the treatment period (day 32). a Tumor lysates were probed and quantitatively analyzed for levels of total and phosphorylated VEGFR1 and VEGFR2. b Representative images and quantitative analysis of CD31 as microvessel formation index (Scale bar = 20 µm). c Tumor lysates were probed and quantitatively analyzed with the indicated antibodies. d Representative images and quantitative analysis of Ki67 as tumor proliferation index (Scale bar = 100 µm). e Tumor lysates were probed and quantitatively analyzed with the indicated antibodies. f Representative images of TUNEL and the statistical graph based on the percentage of apoptosis cells (Scale bar = 100 µm). g Tumor lysates were probed and quantitatively analyzed with the indicated antibodies. All data were analyzed by prism software (One-way ANOVA method, mean ± SD, n = 3, number sign symbol (#) was used for comparing the treatments with control, and asterisk symbol (*) was used for comparison between treatments ****P < 0.0001, ***P < 0.001, **P < 0.01,*P < 0.1, ns: not significant compared to untreated control)

Obviously, blockade of VEGFR1- and especially VEGFR2-mediated signaling attenuates angiogenesis. To assess whether the antitumor effect of GNP–VGB3 is associated with the inhibition of angiogenesis, the microvascular density (MVD) was quantified by immunohistochemical staining of CD31, as an index of angiogenesis. Compared to PBS-treated control group, MVD was significantly reduced in VGB3-treated tumors (P < 0.01) (Fig.  7b); however, the greatest reduction was observed in the GNP–VGB3 group (51% reduction, P < 0.001) compared to controls. In GNP group, there was no significant reduction in MVD (P > 0.9999).

The VEGFR-2 and, to a lesser extent, VEGFR1 has been proved to mediate various cellular signal transduction, including endothelial and tumoral cell survival, proliferation, migration, and induction of permeability [3]. Therefore, the consequences of VEGFR-1/-2 blockade on the constitutive and phosphorylated forms of common downstream proteins were assayed by immunoblotting of the tumor lysates.

The RAS/RAF/MEK/ERK signaling pathway is crucial for the regulation of different cellular processes, including cell proliferation, differentiation and migration. Through cyclin D1 and cyclin-dependent kinases (Cdk)-2/-4, RAS/RAF/MEK/ERK pathway is involved in the regulation of cell cycle progression. We investigated the effects of different treatments on this pathway by measurement of the expression levels of RAF, MEK, CyclinD1, CDK and constitutive and phosphorylated forms of ERK1/2. The results showed that GNP is mostly ineffective on these signaling mediators. Furthermore, GNP–VGB3 resulted in downregulation of RAF, MEK, cyclinD1, CDK-4 and phosphorylated form of ERK1/2 more potently than VGB3, which indicate that the superior antitumor effect of GNP–VGB3 than free peptide is associated with more effective inhibition of proliferation signaling. In agreement with these results, immunohistochemical staining of Ki-67, an index tumor cell proliferation, was decreased by GNP–VGB3 (P < 0.001) more potently than VGB3 (P < 0.01) compared to controls (Fig. 7d). Furthermore, suppression of proliferation signaling supported by decreased expression of glycogen synthase kinase-3 (GSK-3), an inhibitor of cyclin D1, in VGB3- and GNP–VGB3-treated tumors (P < 0.0001) compared to controls (Fig. 7c).

Supporting cell survival and inhibition of apoptosis is another consequence of RAS/RAF/MEK/ERK signaling. Moreover, VEGF promotes cell survival, cancer development as well as metastasis via the PI3K/Akt/mTOR signaling pathway [42]. Accordingly, we sought to further investigate the effects of treatments on the survival, apoptosis and metastasis by analysis of PI3K, AKT, p-AKT, mTOR and p-mTOR, NF-κB, p-NF-κB, P53, Bcl2 and Bax. Compared to controls, the expression level of PI3K increased after treatment with GNP and VGB3 (P < 0.05) but markedly decreased after GNP–VGB3 treatment (P < 0.0001) (Fig. 7e). In agreement with these results, GNP–VGB3 potently reduced p-AKT formation in 4T1 tumors (P < 0.0001), whereas VGB3 was less effective (P < 0.001) and GNP had no effect. mTOR known as promoter of tumor cell migration and invasion [43]. GNP could not change the expression level of mTOR and p-mTOR. In contrast, both VGB3 and GNP–VGB3 treatments drastically suppressed total mTOR expression (P < 0.0001). Furthermore, phosphorylation of mTOR was inhibited more effectively by GNP–VGB3 (P < 0.0001) than by free VGB3 (P < 0.001) compared to PBS-treated tumors. A major target of Akt is the NF-κB pathway [44]. Analysis of tumors revealed highly significant decrease in the expression of NF-κB in VGB3- and GNP–VGB3-treated tumors (P < 0.001 and P < 0.0001, respectively). More importantly, GNP–VGB3 resulted in strong suppression of NF-κB phosphorylation, whereas GNP had no effect and VGB3 group modestly showed NF-κB phosphorylation. Given that activation of NF-κB leads to blockade of apoptosis and promotion of cell proliferation [45], downregulation of NF-κB and p-NF-κB is expected to induce apoptosis in tumors. Accordingly, VGB3-treated tumors present with much higher P53 levels than control 4T1 tumor lysates (P < 0.0001); however, P53 even more increased upon administration of GNP–VGB3 (Fig. 7e). In parallel, GNP–VGB3 treatment resulted in a decrease Bcl2 expression with a concomitant increase in the protein level of Bax (Fig. 7e). These data, consistent with ROS overproduction and annexin V staining in HUVE and 4T1 cells, suggest that GNP–VGB3 enhanced the VGB3-driven inhibition of survival signaling, leading to apoptosis induction in 4T1 mammary carcinoma tumors. To corroborate these results, tumors were analyzed with the TUNEL apoptosis assay. Notably, whereas VGB3 treatment alone had a low effect on the TUNEL-positive cells (P < 0.01), the conjugation of VGB3 to GNP appeared to reinforce the apoptosis induction property of the VEGFR1/2-blocking peptide, as evidenced by the marked increasing of the TUNEL-positive tumor cells in GNP–VGB3-treated tumors (P < 0.0001) (Fig. 7f).

FAK/Paxillin signaling axis, in downstream of PI3K/AKT and MAPK/ERK signaling pathways, is involved in cell adhesion, migration, proliferation, and survival. Analysis of tumor lysates revealed a moderate reduction of total FAK for both VGB3 and GNP–VGB3-treated tumors (P < 0.05) while Paxillin expression was considerably decreased in VGB3 and GNP–VGB3-treated tumors compared to control (P < 0.001 and P < 0.0001, respectively). More strikingly, p-FAK as well as p-paxillin formation were strongly inhibited by all treatments (P < 0.0001) so that GNP–VGB3 and GNP were the most and less effective groups, respectively. These results are indicative of more efficient suppression of cell detachment, as an initial step in the metastatic transformation, by GNP–VGB3 than by other treatments (Fig. 7e).

The process of cell invasion is a combination of cell migration with concurrent degradation of the surrounding extracellular matrix (ECM) by matrix metalloproteases [46]. Decreased expression of MMP-9 has been observed in 4T1 mammary carcinoma tumors treated with VEGF blockading peptides [9]. Importantly, GNP–VGB3 potently inhibited MMP-9 expression in tumor tissue (P < 0.0001), whereas free peptide was moderately effective (P < 0.05) and GNP had no effect (Fig. 7e).

Inhibition of VEGFR2 phosphorylation was shown to inhibit metastasis and cancer progression via eNOS/Akt signaling [47], underlined by the fact that endothelial nitric oxide synthase (eNOS) induces nitric oxide (NO) production, which plays important role in vascular protection, focal adhesion formation and cell migration. Our results indicated that GNP-treatment has no effect on eNOS expression in tumor tissues, whereas both VGB3 and GNP–VGB3 treatments markedly suppressed eNOS expression compared to controls (P < 0.0001) (Fig. 7e).

p38 MAP kinase has been implicated in a variety of cellular processes, including cell proliferation, cell differentiation, apoptosis, cell migration, and invasion [48, 49]. The total expressions of p38 MAPK were unaffected by GNP treatment but increased by VGB3 (P<0.05) and GNP–VGB3 (P < 0.001) compared to controls. More strikingly, p38 MAP kinase activation, i.e. p-p38 MAP kinase formation, strongly suppressed by GNP–VGB3 (P < 0.0001) but not by the other treatments compared to controls (Fig. 7g).

Epithelial–mesenchymal transition (EMT) promotes metastasis by enhancing mobility, invasion, and resistance to apoptotic stimuli [50]. Importantly, EMT is characterized by decreased expression of cell adhesion molecules such as E-cadherin and increased expression of vimentin and N-cadherin. We therefore compared the potential of treatments to affect the expression of E-cadherin, vimentin and N-cadherin. GNP–VGB3 treatment was more effective than VGB3 in decreasing the expression of vimentin and N-cadherin as well as in increasing the expression of E-cadherin in 4T1 tumors (P < 0.0001), whereas the expression levels of both proteins in GNP-treated tumors were comparable with controls (Fig. 7g). These results indicate that GNP–VGB3 attenuated EMT in 4T1-bearing Balb/c mice.

Although most of known responses to VEGFA are mediated by VEGFR2, endothelial cell functions can be stimulated by VEGFR1 especially through PI3K/Akt pathway [51]. In agreement with the in vitro results, analysis of signaling pathways in tumor tissues showed that even greater inhibitory effects than those resulted from VEGFR1/VEGFR2 blockading peptide can be obtained from its combination with the positive effects of GNP. Figure 8 represents the wide ranges of signaling mediators targeted by GNP–VGB3.

Fig. 8

Signaling transduction and biological processes mediated by inhibition of VEGFR1 and VEGFR2 using GNP–VGB3. a Schematic illustration of the effect of GNP–VGB3 on the downstream signaling pathways of VEGFR1/2. Binding of GNP–VGB3 to VEGFR1/R2 inhibits the activity of receptors that were activated by VEGFA/B and the downstream signaling pathways is subsequently prevented. VGB3 after conjugation to GNPs suppresses the signaling pathways in via preventing cell proliferation, migration, apoptosis, permeability and metastasis. b  The comparison of in vitro and in vivo effects exerted free GNP, free VGB3 peptide and GNP-VGB3

MicroCT imaging

To identify the accumulation of nanoparticles in the tumors, we examined mice treated with the GNP, VGB3 and GNP-VGB3  using MicroCT imaging. Initially, one group of 4T1 tumor-bearing Balb/c mice (n = 3) were intravenously injected with free GNPs. After injection for 3 h, the strongest CT signals were observed in kidneys (Fig. 9a). In addition, CT signals were observed in tumors and liver (Fig. 9a).  However, the signals in the tumor regions were augmented in mice treated with peptide bound GNP (GNP–VGB3) (Fig. 9a). Furthermore, target specificity of GNP–VGB3 elicited by a blocking experiment. As shown in Fig. 9a, accumulations were suppressed effectively by coadministration of competing free VGB3 peptide. These observations suggest that GNP–VGB3 can specifically target mammary carcinoma tumors.

Fig. 9

Biodistribution study by MicroCT imaging and ICP MS analysis. a Representative 3D-reconstructed whole-body CT images of mice bearing 4T1 tumors at 3 h following intravenous (i.v.) injection of GNP, GNP–VGB3, block and PBS (untreated control). The red circles and arrows indicate tumor locations. b The Au concentrations in major organs, including kidney, spleen, tumor, liver and heart, which received GNP, GNP–VGB3 and block samples quantified after 24 h using inductively coupled plasma mass spectrometry (ICP-MS) analysis. The quantified results were defined based on the percentage of injection dose per tissue (%ID/tissue) and analyzed statistically by two-way ANOVA method (mean ± SEM, and n = 3, ****P < 0.0001, **P < 0.01, *P < 0.05, or ns: not significant). (Asterisk symbol (*) was used for comparison between GNP with GNP–VGB3 and block and number sign symbol (#) was used for comparing GNP–VGB3 and block)

Biodistribution study of nanoparticles

Quantitative biodistribution analysis was performed by examining the gold content (% ID/tissue). To this end, the mice were sacrificed, a series of dissected organs (kidney, spleen, liver and heart) and tumor tissues of mice (n = 3) were freshly collected 24 h post-injection, and inductively coupled plasma mass spectrometry (ICP-MS) measurements were immediately taken. As shown in Fig. 9b, a very high amount of gold was found in liver (88.8% ID/tissue) and spleen (80.3% ID/tissue) for GNP and blocking groups (n = 3), respectively, suggesting that these nanoparticles are cleared mainly through the reticuloendothelial system. In these groups, the tumor accumulations were 1.3 and 8% ID/tissue, respectively. The tumor accumulation of GNP–VGB3, however, increased to 15.2% ID/tissue, which was significantly more than those of GNP (1.4%) and blocking (9.4%). Importantly, the highest accumulation of GNP–VGB3 was observed in kidney. The observation that free GNP is accumulated mostly in the liver was also reported in previous investigations [52, 53]. Nanosystems with renal clearance are more desirable than those cleared through the reticuloendothelial system [54]. Thus, higher accumulation of GNP–VGB3 kidney may induce less damage than free GNP in the normal tissues. These data suggest that the tumor accumulation of GNP–VGB3 is more efficient than free GNP. In addition, a decreased amount of GNP–VGB3 in the presence of free peptide (block) further confirms its specific binding to tumors.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit The Creative Commons Public Domain Dedication waiver ( applies to the data made available in this article, unless otherwise stated in a credit line to the data.


This article is autogenerated using RSS feeds and has not been created or edited by OA JF.

Click here for Source link (