Synergistic Suppression of Tumor Angiogenesis by the Co-delivering

Research Article www.acsami.org

Synergistic Suppression of Tumor Angiogenesis by the Co-delivering of Vascular Endothelial Growth Factor Targeted siRNA and Candesartan Mediated by Functionalized Carbon Nanovectors Xuefang Ding,†,‡ Yujie Su,† Cheng Wang,† Fangrong Zhang,† Kerong Chen,† Yu Wang,*,§ Min Li,† and Wei Wang*,† †

State Key Laboratory of Natural Medicines, Department of Pharmaceutics, China Pharmaceutical University, 24 Tongjiaxiang, Nanjing 210009, China ‡ School of Pharmaceutical Science, Nanjing Tech University, 30 South Puzhu Road, Nanjing 211816, China § Department of Pharmacology, Collaborative Innovation Center for Cardiovascular Disease Translational Medicine, Nanjing Medical University, 140 Hanzhong Road, Nanjing 210029, China

ABSTRACT: Single-walled carbon nanotubes (SWNTs) with unique physicochemical properties have exhibited promising biomedical applications as drug and gene carriers. In this study, polyethylenimine (PEI)-modified SWNT conjugates linked with candesartan (CD) were developed to deliver vascular endothelial growth factor (VEGF)-targeted siRNA (siVEGF) for the synergistic and targeted treatment of tumor angiogenesis. The characterization results revealed that SWNT−PEI−CD conjugates were successfully synthesized and exhibited desirable dispersibility and superior stability. Confocal laser scanning microscopy (CLSM) and flow cytometry (FCM) results showed that SWNT−PEI−CD/siVEGF complexes could achieve high cellular uptake and specific intracellular distribution of siRNA in AT1R overexpressed PANC-1 cells. Strong down-regulation of VEGF was also verified by qualitative real-time polymerase chain reaction, enzyme-linked immunosorbent assay, and Western blot in complex-treated PANC-1 cells. The in vitro angiogenesis assay showed that SWNT−PEI−CD/siVEGF complexes highly inhibited tube formation of human umbilical vein endothelial cells. Furthermore, in vivo observation in PANC-1 xenografted nude mice demonstrated that SWNT−PEI−CD/siVEGF complexes exhibited significant distribution at tumor sites and caused obvious inhibition of tumor growth and tumor-associated angiogenesis repression induced by the drug combination of CD and siVEGF. Finally, a WST-1 assay indicated that SWNT−PEI−CD possessed low cytotoxicity, and a hemolysis test showed good biocompatibility of SWNT−PEI−CD. Hematological and histological analyses confirmed that SWNT−PEI−CD/siVEGF complexes did not cause any obvious toxic effects to blood and major organs. These findings suggested that the SWNT−PEI− CD/siVEGF co-delivery system with tumor-targeting specificity, improved endosomal escaping properties, and collaboration of angiogenesis inhibition could be a prospective method for efficient tumor antiangiogenic therapy. KEYWORDS: single-walled carbon nanotubes, candesartan, VEGF-targeted siRNA, synergistic inhibition, tumor angiogenesis

1. INTRODUCTION

tumors such as colon, gastric, breast, liver, and prostate cancers. Therapies for targeting the VEGF pathway have been developed for cancer therapy.3 Recently, small interfering RNAs (siRNAs) have been considered as potential therapeutic

Angiogenesis is a physiological process of new vascular generation from pre-existing blood vessels or endothelial progenitor cells. It plays an important role in tumor growth and metastatic progression and is regulated by a very-sensitive interplay of several growth factors and inhibitors.1,2 Therein, vascular endothelial growth factor (VEGF), a cytokine, has been shown to be the main cause of angiogenesis that occurs in © 2017 American Chemical Society

Received: April 8, 2017 Accepted: June 15, 2017 Published: June 15, 2017 23353

DOI: 10.1021/acsami.7b04971 ACS Appl. Mater. Interfaces 2017, 9, 23353−23369

Research Article

ACS Applied Materials & Interfaces

Scheme 1. Schematic Illustration of the Assembly Process of the Targeted SWNT−PEI−CD/siRNA Co-delivery System and the Intracellular Trafficking Pathwaya

a Briefly, the pristine SWNT was shortened and covalently conjugated with PEI and CD, and then the siRNA was condensed with the SWNT−PEI− CD by electrostatic attraction to form the targeted SWNT−PEI−CD/siRNA complexes. After intravenous administration, the co-delivery system would result in improved tumor therapy in vivo through the cell penetration and the AT1R-mediated endocytosis, membrane depletion, and protonsponge-induced endosome escape and the synergistic anti-angiogenesis by VEGF down-regulation related to CD and siRNA silencing.

to enhancing the dispersibility and biocompatibility while mitigating the toxicity of SWNTs. Positively charged SWNTs are needed for the establishment of new gene-delivery carriers because of their ability to bind negatively charged siRNAs by electrostatic interaction, thus enabling the efficient delivery of the siRNAs into cells and tissues. Herein, we employed branched polyethylenimine (PEI), one of the most efficient cationic polymers, to modify SWNTs for improving the dispersibility and siRNA-binding capacity of SWNTs and enhancing the ability of the siRNAs to escape from endosomes and relocate in cytoplasm. SWNTbased siRNA delivery systems could be equipped with tumortargeting ligands so that internalization efficiency of the systems would be improved and potential side effects would be minimized. Candesartan (CD), an Ang-II receptor blocker, shows a strong affinity for the angiotensin II type-1 (AT1) receptor expressed on the surface of neovascular endothelial cells and tumor cells. Numerous researches have confirmed higher expression of AT1 receptor (AT1R) in various tumor cells, such as pancreatic carcinoma,36,37 breast carcinoma,38 laryngeal cancer, and choriocarcinoma,39 than in normal cells. Thus, CD could be a promising targeting ligand to AT1R in anticancer therapy. More importantly, encouraging results have demonstrated that CD can decrease VEGF expression in prostate cancer40 and evidently down-regulate the production of VEGF in MCF-7 tumor-bearing mice, leading to significant inhibition of tumor growth and angiogenesis in vivo.37,41 In our research, SWNTs were shortened, covalently connected with PEI and CD and subsequently complexed with siRNA. The formed SWNT−PEI−CD/siRNA complexes possessed combined capabilities of AT1R-specific targeting, cell penetration, gene silencing, and drug delivery (Scheme 1). Such a SWNT-based co-delivery system of VEGF-targeted siRNA and CD is favorable for improving tumor therapy efficacy through antiangiogenesis. Flow cytometry and confocal imaging were employed to examine the specific targeting and intracellular distribution of the SWNT−PEI−CD/siVEGF com-

agents of many diseases, including cancer; viral infections; cardiovascular, neurological, and metabolic diseases; and other gene-related diseases.4−7 siRNAs can be incorporated into RNA-induced silencing complex (RISC) and lead to the degradation of targeted mRNAs in a sequence specific posttranscription gene-silencing manner.8−10 It is well-known that the action mechanism of small molecule antivascular drugs is comprehensive, and these drugs can cause unavoidable systemic toxicity and myelosuppression. Compared to the defects of antivascular drugs, the strategy of VEGF gene silence triggered by siRNA is more specific and more powerful and would be much more preferred in the inhibition of VEGF expression.11 Studies have shown that siRNAs can interfere with the process of neovascularization12,13 by selectively and successfully downregulating the expression of VEGF, which was an important signaling protein involved in tumor growth, progression, and metastasis.14−17The therapeutic application, however, of this technology is still limited as siRNAs have certain disadvantages, such as their relatively high molecular weights, negatively charged characteristics, and low stability and bioavailability, both in vitro and in vivo. To overcome these obstacles, safer and more-efficient gene vectors are required for the stabilization of siRNAs and effective uptake by targeted cells.18 Recently, researchers revealed single-walled carbon nanotubes (SWNTs) as increasingly prospective candidates for the delivery of genes.19−21 SWNTs possess numerous unique physicochemical and biological properties, including ultrahigh surface area;22 excellent optical,23 electrical, and thermal properties;24 and the strong ability to easily penetrate through the cell membrane.25−27 SWNTs have already shown to be promisingly useful in biomedical applications including biosensors,28 biological imaging, biomedical delivery systems (including drugs,29 genes,30−32 and proteins),33 and photothermal therapy.34,35 However, the highly hydrophobic surface and cytotoxicity of SWNTs severely hinder their applications. Therefore, surface modification could be a remarkable approach 23354


Research Article


COOH suspension and sonicated for 30 min and reacted with stirring for 24 h at room temperature. Subsequently, the aqueous suspension was centrifugated for 30 min at 13 000 rpm, and the precipitate was collected and washed to remove unreacted polymer. Finally, the SWNT−PEI−CD hybrids were obtained by drying under vacuum. SWNT−PEI was also synthesized in the same experiment procedures. 2.4. Characterization of SWNT−PEI−CD. 2.4.1. Transmission Electron Microscopy Investigation. A small amount of pristine SWNTs, SWNT−COOH, SWNT−PEI−CD, and SWNT−PEI−CD/ siVEGF complexes were dispersed in water, separately, and ultrasonicated for 30 min. A small amount of suspension was taken and put on grids. Transmission electron microscopy (TEM) images were observed using a JEM-2100 HR-TEM microscopy (JEOL Ltd., Tokyo, Japan). 2.4.2. Raman Spectrum analysis. The pristine SWNTs, SWNT− COOH, and SWNT−PEI−CD were covered on glass slides, respectively. The samples were irradiated by an exciting light of 585.25 nm generated by confocal micro-Raman spectroscopy (LabRam HR800, France). The Raman spectra of the samples were recorded from 200 to 3000 nm. 2.4.3. FTIR Spectrum Analysis. The SWNT−COOH and SWNT− PEI−CD were ground, respectively, with potassium bromide (KBr) in an agate mortar and pestle to form very fine powders. They were compressed into thin pellets and scanned using a Bruker MPA Fourier transform infrared (FTIR) spectrometer (Bruker Corporation, Billerica, MA). 2.4.4. NMR Spectrum Analysis. CD and SWNT−PEI−CD were dissolved in DMSO-d6 and D2O (pH 7.0), respectively. They were then analyzed by 1H nuclear magnetic resonance (NMR; Avance AV500, Bruker Corporation). 2.4.5. Thermogravimetric Analysis. The thermogravimetric analysis (TGA) was performed to pristine SWNTs, SWNT−COOH, SWNT−PEI−CD, and PEI−CD by scanning from 25 to 700 °C using a TG209C thermogravimetric analyzer (Netzsch Gmbh and Co. KG, Selb, Germany). 2.4.6. Dispersion Stability Study. Pristine SWNT, SWNT−COOH, and SWNT−PEI−CD were dispersed in distilled water by ultrasonic for 30 min, respectively. Afterward, the samples were sealed and stored at room temperature for observation. 2.4.7. Measurement of Size and Potential. The size distribution of SWNT−PEI−CD and the ζ potential of pristine SWNT and SWNT− PEI−CD were revealed by Zetasizer 3000HS (Malvern, UK). In addition, SWNT−PEI−CD/siRNA complexes were obtained by mingling SWNT−PEI−CD with same amount of siRNA at different w/w ratios in pure water. After incubation for 30 min, the ζ potentials of the complexes were measured. 2.5. Preparation of SWNT−PEI−CD/siRNA Complexes. Briefly, aqueous suspension of SWNT−PEI−CD was dropped into siRNA solution, eddied gently for 30 s, and incubated for 30 min at room temperature. The prepared SWNT−PEI−CD/siRNA complexes were then loaded on 2% agarose gels containing 5 μL/100 mL GoldView (Viswagen Biotech Pvt. Ltd., Kerala, India) and electrophoresed at 80 mV for 20 min. Next, gels were measured by UV analysis, and the siRNA retardation was visualized using Gel-Pro Analyzer software (Media Cybernetics, MD). 2.6. In Vitro Cytotoxicity of SWNT−PEI−CD. The in vitro cytotoxicity of SWNT−PEI−CD was evaluated by WST-1 assay. PANC-1, MCF-7, and HEK293 cells were seeded in 96-well plates at a density of 1 × 104 cells per well, and then incubated for 12 h. Various concentrations of PEI−CD, SWNT−PEI−CD, PEI−CD + SWNTs, and PEI 25 KDa were added to wells and incubated for 24 h. Subsequently, 100 μL of WST-1 (5 mg/mL, Roche, Mannheim, Germany) was added and incubated for 4 h. The absorbance of each well was monitored using a BioTek Instruments, Inc., EL800 universal microplate reader (Winooski, VT) at a wavelength of 450 nm. The cytotoxicity was expressed as the percentage of cell viability compared to that of untreated control cells. 2.7. In Vitro Cellular Uptake Studies. The cellular uptake of the complexes was studied in HEK293, MCF-7, and PANC-1 cells, which was chosen as AT1R nonexpressing, normal-expressing, and over-

plexes to AT1R over-expressed PANC-1 cells. The VEGFsilencing effect of SWNT−PEI−CD/siVEGF complexes was also investigated, and the inhibitory effect of angiogenesis by SWNT−PEI−CD/siVEGF complexes was then observed. Finally, the tumor inhibition of SWNT−PEI−CD/siVEGF complexes in PANC-1 xenograft nude mice was investigated, and the toxicity of complexes was evaluated to further indicate the prospective clinical application of the SWNT−PEI−CD/ siVEGF complexes in tumor antiangiogenesis therapy.

2. MATERIALS AND METHODS 2.1. Materials. SWNTs were purchased from Chengdu Organic Chemicals Co. Ltd. (Chengdu, China). Anti-VEGF siRNA (siVEGF), negative-control siRNA (siRNA), and fluorescein labeled siRNA (FAM-siRNA) were synthesized and purified by GenePharma Co. Ltd. (Shanghai, China). Branched polyethylenimine (PEI, MWCO = 1.8 and 25 KDa) was purchased from Sigma-Aldrich (St. Louis, MO). Candesartan was purchased from AstraZeneca (Hamburg, Germany). N-Hydroxysuccinimide (NHS) and 1-(3-(dimethylamino)propyl)-3ethylcarbodiimide hydrochloride (EDC) were purchased from Aladdin Reagent Inc. (Shanghai, China). LysoTracker Red, TRIzol reagent, Lipofectamine 2000 reagent, and Gibco DMEM RIPA lysis and extraction buffer were obtained from Life Technologies Ltd. (Grand Island, New York). RIPA Lysis, extraction buffer, and PMSF and BCA protein assay kits were provided by Beyotime Biotechnology (Haimen, China). A real-time polymerase chain reaction (RT-PCR) Kit was obtained from Thermo Fisher Scientific Inc. (Shanghai, China). Cy7 was purchased from Beijing Fanbo Biochemical Co., Ltd. (Beijing, China). ECMatrix was purchased from BD Biosciences Inc. (Heidelberg, Germany). 2.2. Cell Culture. Human umbilical vein endothelial cells (HUVECs), human embryonic kidney epithelial cell line HEK293, human breast adenocarcinoma cell line MCF-7 and human pancreatic adenocarcinoma cell line PANC-1 were obtained from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China) and cultured in Gibco Dulbecco’s modified Eagle medium (DMEM) culture medium supplemented with 10% fetal bovine serum (FBS, HyClone), in a humidified incubator containing 5% CO2 at 37 °C. All experiments were performed on cells in the logarithmic phase of growth. 2.3. Synthetic Procedure for SWNT−PEI−CD. 2.3.1. Puri®cation of SWNTs. Pristine SWNTs were dispersed in concentrated HNO3 and stirred for 24 h at room temperature. Next, filtration was performed on a Millipore membrane followed by rinsing with ultrapure water thoroughly until the filtrate was neutral. The purified SWNTs were then collected and dried under a vacuum. 2.3.2. Oxidation of SWNTs. Approximately 50 mg of purified SWNTs was suspended in an ultrasonic bath in 250 mL of mixed acids (H2SO4/HNO3, v/v = 3:1) for 5 h and subsequently refluxed at 80 °C for 24 h. Followed by cooling and dilution, the suspension was then filtered on a Millipore membrane. The filter was further rinsed with abundant of ultrapure water. Consequently, the carboxylated SWNTs (SWNT−COOH) were obtained under a vacuum. 2.3.3. Synthesis of PEI−CD. The conjugate PEI−CD was synthesized as previously described by our laboratory37 with an amide bond between the primary amines of PEI and the carboxy termini of CD molecules. Briefly, CD was dissolved in DMF and activated by EDC and NHS with agitation for 2 h. The solution was dropped into PEI 1.8 KDa solution to react for 24 h at room temperature. Subsequently, the conjugate was collected by precipitate with excess acetone. The precipitate was dissolved in pure water followed by dialyzation using a dialysis membrane (Spectra/Por, MWCO = 1000). Finally, the product was obtained by lyophilization of the resulting solution and stored at −20 °C for further use. 2.3.4. Synthesis of SWNT−PEI−CD. SWNTs were covalently functionalized with PEI−CD by amidation reaction between SWNT−COOH and the amine group of PEI−CD using EDC and NHS as coupling reactants. Briefly, EDC and NHS were added into ultrasonicated SWNT−COOH aqueous suspension (1 mg/mL). Then, PEI−CD was added slowly into the activated SWNT− 23355


Research Article

ACS Applied Materials & Interfaces expressing models, respectively. Cells were plated onto confocal dishes at a density of 1 × 105 cells per dish and grown for 12 h to reach about 70% confluence at the time of cellular uptake. Subsequently, the primary growth medium was replaced with serum-free media and 100 μL of SWNT−PEI−CD/FAM−siRNA complexes were added and incubated for 4 h. After being washed three times with PBS, the cells were fixed with 4% paraformaldehyde and rinsed with ultrapure water to eliminate the paraformaldehyde. Cellular uptake with PEI−CD/ FAM−siRNA, PEI 1.8 KDa/FAM−siRNA, and SWNT−PEI/FAM− siRNA complexes were conducted in the same way. Finally, cells were imaged by confocal laser scanning microscopy (CLSM) (Leica Microsystems, Wetzlar, Germany), and the cell uptake efficiency was quantified by flow cytometry (FACSCalibur, Beckton, Dickinson and Company). 2.8. Intracellular Distribution Studies. To reveal the intracellular trafficking of SWNT−PEI−CD/siRNA on AT1R highexpressing PANC-1 cells, the intracellular localization of SWNT− PEI−CD/FAM−siRNA complexes was investigated by CLSM. Cells were seeded into confocal dishes at a density of 2 × 104 cells per well and incubated for 24 h at room temperature. SWNT−PEI−CD/ FAM−siRNA (100 nM) complexes were added to the dishes. At designated intervals, the cells were washed and stained with LysoTracker Red (50 nM) for 30 min. After rinsing and fixation, the cells were redyed using Hoechst 33258 (10 μg/mL, Invitrogen, Carlsbad, CO) and investigated by CLSM. 2.9. In Vitro Gene Silencing Effects. 2.9.1. qRT-PCR. The effect of SWNT−PEI−CD/siVEGF complexes on VEGF mRNA expression was studied by qRT-PCR. PANC-1 cells were transfected with naked siVEGF, SWNT−PEI−CD/siRNA complexes, SWNT−PEI−CD/ siVEGF complexes, SWNT−PEI/siVEGF complexes + CD, SWNT−PEI/siVEGF complexes, PEI−CD/siVEGF, and Lipofectamine 2000/siVEGF complexes, respectively. After 48 hof transfection, extraction of total RNA and synthesis of first strand cDNA were performed according to the manufacturer’s manual. Thereafter, the VEGF mRNA expression in the samples was quantified by the reverse transcription system (Promega, Madison, WI) by the performance of 40 cycles of PCR. β-Actin was used as normalizing reference control. The primers for VEGF and β-actin were: VEGF forward primer: 5′CGAAGTGGTGAAGTTCATGGATG-3′ and reverse primer: 5′TTCTGTATCAGTCTTTCCTGGTGAG-3′; β-actin forward primer: 5′-TCATGTTTGAGACCTTCAA-3′ and reverse primer: 5′GTCTTTGCGGATGTCCACG-3′. 2.9.2. ELISA for VEGF. PANC-1 cells were cultured and transfected, as described in section 2.9.1. At 48 h post-transfection, each plate was harvested and centrifuged. The expression level of VEGF in the medium was evaluated by a R&D human VEGF enzyme-linked immunosorbent assay (ELISA) kit. 2.9.3. Western Blotting. After 48 h of transfection, the PANC-1 cells were washed and lysed in lysis buffer optionally with PMSF. The protein concentration was evaluated in the supernatant after the centrifugation of the lysates by a BCA protein assay kit. Equivalent lysates containing 50 μg of protein of the samples were loaded and size-fractionated using 15% sodium dodecyl sulfate polyacrylamide gel electrophoresis. Proteins were transferred onto PVDF membranes and blocked with 5% nonfat milk. Subsequently, the primary antibody (anti-VEGF at a 1:1000 dilution and anti-β-actin at a 1:5000 dilution) was added and incubated overnight at 4 °C, followed by rinsing and incubation with secondary antibody (1:3000 diluted horseradish peroxidase conjugated antibody) for 1 h. The target proteins were measured with a Western blotting kit (Life Technologies, Grand Island, NY). The bands were visualized using an Odyssey image system (LI-COR Biotechnology, Lincoln, NE) and normalized to βactin expression. 2.10. In Vitro Angiogenesis Assay. In vitro angiogenesis assay was conducted to evaluate the ability of tubular formation of endothelial cells after treatment. Liquefied matrigel was placed in precooled 96 well plates and then polymerized for 30 min at 37 °C. HUVECs were transfected with SWNT−PEI−CD/siRNA complexes, SWNT−PEI/siVEGF complexes and SWNT−PEI/siVEGF + CD, PEI−CD/siVEGF complexes, and SWNT−PEI−CD/siVEGF, respec-

tively. Cells were then collected and seeded onto the precoated plates and cultured for 24 h for tubular formation. The results were visualized using an ECLIPSE TE2000-U microscope (Nikon, Tokyo, Japan). Untreated cells were used as a control. The inhibition rate of the formation of tubular structure was calculated as follow:

inhibition rate (%) = (TLcontrol − TLsample)/TLcontrol × 100% where TL stands for tube length. 2.11. In Vivo Antitumor Studies. 2.11.1. PANC-1 Tumor Xenograft Models. BALB/c male nude mice at 4−6 weeks of age were provided by the Shanghai Laboratory Animal Center (Chinese Academy of Sciences [SLACCAS], Shanghai, China). The mice were fed ad libitum in a pathogen-free environment and allowed to access food and water. All animal experiments complied with the guidelines of the National Institutes of Health Guide for the Care and Use of Laboratory Animals. A PANC-1 nude mouse tumor-xenograft model was established by inoculating subcutaneously 2 × 106 PANC-1 cells into the left flank. The treatment was initiated when the tumor reached a median of 100 mm3 in volume (3−4 weeks after tumor inoculation). Tumor volume was detected with calipers and calculated as V = a2 × b/2 mm3 (a: width; b: length). 2.11.2. In Vivo Imaging and Quantitative Analysis. To study biodistribution and tumor targeting efficiency of the complexes, in vivo imaging analysis was performed using the near-infrared fluorophore Cy7. PANC-1 tumor xenograft mice were administered with Cy7− SWNT−PEI−CD/siVEGF, Cy7−PEI−CD/siVEGF or Cy7− SWNT−PEI/siVEGF complexes intravenously at doses of 50 μg Cy7/kg. The fluorescence images were taken and analyzed at 2, 6, and 12 h after injection. 2.11.3. In Vivo Antitumor Efficacy. The mice were randomly divided into 6 groups (n = 5 for each): (1) saline control, (2) SWNT− PEI−CD/siRNA complexes, (3) SWNT−PEI/siVEGF complexes + CD, (4) SWNT−PEI/siVEGF complexes, (5) PEI−CD/siVEGF complexes, and (6) SWNT−PEI−CD/siVEGF complexes. The amount of siRNA was set at a dose of 1 mg/kg. The first day of injection was regarded as Day 0. The treatment was conducted at a two day interval until Day 14. Tumor volume and body weight of each group were detected every second day. At the end of treatment, mice were sacrificed and tumor tissues were harvested and divided into three pieces: one for immunohistochemical staining, one for ELISA, and one for hematoxylin and eosin (H&E) staining. For the observation of microvessel density in tumor tissues, immunohistochemical analysis was performed as the standard operation procedure. Briefly, tumors resected from mice were fixed using formalin and embedded in paraffin for cryosectioning to 5 mm sections. Then, sections were sequentially treated in xylene, 90% ethanol, 80% ethanol, 75% ethanol, and PBS, followed by staining with a monoclonal rat antimouse anti-CD31 antibody and secondary antibody. The resulting stains were redyed with hematoxylin and dehydrated, and tissue images were obtained using an Olympus optical microscope (Tokyo, Japan) at a magnification of 400. The microvessel density (MVD, number per mm2) was calculated to evaluate the mechanism of antitumor effects in vivo. The VEGF level of tumor tissue was measured with ELISA to test the down-regulation of VEGF in vivo. The tumor tissues were homogenized in PBS and centrifuged at 15 000 rpm for 15 min. The obtained supernatant was analyzed using an R&D Systems human VEGF ELISA kit. The amount of VEGF was represented as the ratio of VEGF protein (ng) to total protein (mg). For histological staining, tumors were harvested and fixed in 10% formalin, embedded in paraffin, sliced into approximately 5 mm thick sections, and stained with H&E. The tissue morphology was observed using an ECLIPSE TE2000-U optical microscope. 2.12. In Vitro and in Vivo Safety Analysis. 2.12.1. Hemolysis Assay. Hemolysis assay was required for the safety evaluation of the SWNT−PEI−CD conjugate for use in vitro. Fresh rabbit red blood cells (RBCs) were obtained after centrifugation and suction. Next, the cells were washed and diluted with PBS buffer to a fixed concentration of 2% (v/v). Then, 2.5 mL of diluted RBC solution was gently mixed 23356


Research Article


Figure 1. Synthetic scheme of SWNT−PEI−CD conjugates.

Figure 2. TEM images of (a, b) pristine SWNTs, (c) purified SWNTs, (d) SWNT−COOH, (e) SWNT−PEI−CD, and (f) SWNT−PEI−CD/ siVEGF complexes. with equal SWNT−PEI−CD to a final concentration of 2, 4, 6, 8, and 10 mg/mL of the conjugate, respectively. A total of 2.5 mL of pure water and an equal amount of PBS was conducted as the positive and negative controls, respectively. A mixture containing PEI 1.8 KDa was also prepared for comparison. The mixtures were stand for 2 h and centrifuged to obtain the supernatants. Absorbance at 540 nm was detected to calculate the degree of hemolysis as follows:

hemolysis (%) =

A sample − A 0% A100% − A 0%

× 100%

2.12.2. Hematology Analysis. For hematology studies, blood was collected at day 14 post-injection and sent to the Medical Examination Center of Zhongda Hospital, Nanjing, China. The hematologic 23357


Research Article

ACS Applied Materials & Interfaces parameters and serum biochemistry indicators for hepatic function were determined. 2.12.3. Histological Analysis. Mice were sacrificed at day 14. The kidneys, heart, spleen, lungs, and liver were removed and fixed in 10% buffered formalin. They were embedded in paraffin and sectioned to 5 mm thicknesses and finally stained with H&E for histological analysis. 2.13. Statistical Analysis. Quantitative results were represented as means ± SD. Statistical analysis was carried out by Student’s unpaired t-test between two groups or analysis of variance (ANOVA) among three or more groups. Significance of difference was indicated with comparison to corresponding controls as *, P < 0.05; **, P < 0.01; or ***, P < 0.001.

3. RESULTS AND DISCUSSION 3.1. Synthesis and Characterization of SWNT−PEI−CD Conjugates. The quasi-one-dimensional nanoscale structure

Figure 4. Raman spectra (585.25 nm) of (A) raw SWNT, (B) SWNT−COOH, and (C) SWNT−PEI−CD.

intended to a further research to exploit comprehensive utilization with thermotherapy or biosensors. The pristine SWNTs are highly hydrophobic, which severely restricts their biological application. The surface modification of SWNTs with hydrophilic conjugates has been widely explored.42−44 Various modifications have been effectively employed for dispersing SWNTs in aqueous solutions. After chemical modification with a cationic polymer, positively charged SWNTs could electrostatically interact with the negatively charged siRNA to facilitate the process of cellular uptake, and this can be employed to establish a tumor-targeting gene and drug co-delivery system for potential cancer therapy.32,45 It has been reported that mixed-acid treatment, together with sonication, causes surfaces corrosion of SWNTs, resulting in the production of shorter and open-end SWNTs.46,47 SWNT−PEI−CD was synthesized in this study, as shown in Figure 1. The SWNTs were first oxidized and shortened for the introduction of carboxylic groups. PEI and CD were conjugated by an amide reaction.

Figure 3. (A) Particle size and (B) ζ potential of SWNT−PEI−CD and (C) ζ potential of SWNT−PEI−CD/siRNA complexes at various w/w ratios.

provides SWNTs with a better-defined diameter, smaller dimension, and more-stable nanostructure toward MWNT, which may have a better penetrability and drug delivery performance. Besides, compared to MWNT, SWNT tend to exhibit richer electrical and optical properties and are morepreferred for near-infrared (NIR) fluorescence. Our choosing of SWNT rather than MWNT considered the extensibility and 23358


Research Article


Figure 5. Characterization of SWNT−PEI−CD. (A) FTIR spectra of SWNT−COOH and SWNT−PEI−CD; (B) 1H NMR spectra of CD and SWNT−PEI−CD; (C) TGA curves of (a) pristine SWNTs, (b) SWNT−COOH, (c) SWNT−PEI−CD, and (d) PEI−CD; (D) photographs showing dispersibility of (a) pristine SWNTs, (b) SWNT−COOH, and (c) SWNT−PEI−CD suspended in water as prepared (left) or stored at room temperature for 6 months (right).

of raw SWNT showed a strong G mode at 1594 cm−1 accompanied by a weak D mode at 1356 cm−1, which was, respectively, associated with sp2-hybridized carbons and sp3hybridized carbons from the sidewalls. The IG-to-ID ratio of raw SWNT was high, up to 9.26, indicating an integral structure of SWNT with low degree of defect sites. The presence of radial breathing modes appearing at 249 cm−1 suggested that the nanotubes were truly single-walled. As for SWNT−COOH, the IG-to-ID ratio decreased to 1.95. This change indicated that oxidation occurring on graphitic sidewalls increased in the number of sp3-hybridized carbons and partially destroyed the SWNT structure. When covalently modifying the sidewalls of SWNT with PEI−CD, no significant change was observed in Raman spectrum with an IG-to-ID ratio of 1.94. This result showed that the mild amidation does not further break the graphitic sidewalls of SWNT. Figure 5A showed the FTIR spectra of SWNT−COOH and SWNT−PEI−CD. The SWNT−COOH presented an absorption peak at 1710 cm−1, which can be regarded as the carbonyl stretch of the carboxyl group.51 After grafting with PEI−CD, new absorption peaks appeared at 1650 and 1580 cm−1, which corresponded to the amide carbonyl vibration and the N−H bending vibration, respectively. The intensity of the absorption band at 1710 cm−1 attributed to the carboxyl group of −COOH was significantly reduced by the PEI−CD grafting, agreeing with the formation of amide structure. The successful conjugation of PEI−CD to SWNTs was further confirmed using 1H NMR. As depicted in Figure 5B, the SWNT−PEI−CD showed the same chemical shift as the alkyl

The resulting PEI−CD was then covalently grafted onto the oxidized SWNTs via the formation of amide bonds between the carboxylic groups of SWNT−COOH and the amine groups of PEI. SWNT−PEI was also synthesized as a negative control. As indicated by the TEM images (Figure 2), the original SWNTs were long, twisted, and bundled or aggregated with some visible impurities, such as black amorphous carbon particles or metal catalyst (Figure 2a,b). However, after purification and oxidation, the smooth surfaces (Figure 2c) and the hollow structures (Figure 2d) of SWNT−COOH could be clearly identified. After functionalization with PEI−CD, the surfaces of nanotube became rougher (Figure 2e). The increased wall thickness of SWNT−PEI−CD compared with pristine SWNTs suggested that the surfaces of the SWNTs were successfully functionalized with PEI−CD. After the SWNT−PEI−CD complexing with siVEGF, the TEM image showed some particles attached to the side wall of SWNTs (Figure 2f), which is presumably that the siVEGF complexation lead to these morphology changes. The SWNTs were shortened to a range of 100−450 nm with a mean length of ∼200 nm, as shown in Figure 3A. The ζ potential was also measured. As depicted in Figure 3B, the pristine SWNTs were found to be negative, agreeing with that in the literature.48−50 After SWNTs were functionalized with positively charged PEI−CD, the ζ potential of the resulting SWNT−PEI−CD conjugate was positive, which verified the successful synthesis of SWNT−PEI−CD. Figure 4 showed the Raman spectrum of raw SWNT, SWNT−COOH, and SWNT−PEI−CD. The Raman spectrum 23359


Research Article


Figure 7. Viabilities of PANC-1 (A), MCF-7 (B), and HEK293 (C) cells incubated with PEI−CD, SWNT−PEI−CD, PEI−CD + SWNT, and PEI 25 KDa at different concentrations for 24 h, respectively. Results were presented as mean ± SD (n = 3). *, P < 0.05; **, P < 0.01; and ***, P < 0.001.

Figure 6. Agarose electrophoresis of complexes at different w/w ratios of PEI−CD (A), SWNT−PEI−CD (B), and PEI 25 KDa (C) to siRNA. The arrows point to the w/w ratios under which the siRNA movability was entirely blocked by the carriers.

obvious precipitates were observed in the SWNT−COOH suspension. However, the homogeneity of SWNT−PEI−CD remained, even up to 6 months. The good dispersibility of the SWNT−PEI−CD can be explained in a number of ways. First, the HNO3 purification and mixed-acid treatment made the SWNT−COOH rather hydrophilic. In addition, the peripheral amine groups of SWNT−PEI−CD were protonated in water; thus, the electrostatic repulsion between positively charged SWNT−PEI−CD conjugates forced them apart and improved the dispersibility. The highly dispersed SWNT−PEI−CD hybrids favored the efficient binding of biomolecules. 3.2. Preparation and ζ Potential of SWNT−PEI−CD/ siRNA Complexes. The electrostatic interaction of negatively charged siRNA with the cationic nanotubes provided ideal conditions for the application of the SWNT−PEI−CD as a gene carrier. To confirm the condensation capability of SWNT−PEI−CD with siRNA, agarose gel electrophoresis was performed. PEI 25 KDa, a common gene carrier, was employed as positive control. As represented in Figure 6, full retardation of siRNA by PEI−CD, SWNT−PEI−CD, and PEI 25 KDa was, respectively, obtained at the w/w ratio of 1.5, 1 and 0.3. This suggested that SWNT−PEI−CD was sufficiently potent in condensing siRNA and was an effective gene-delivery agent.

peaks (−CH3) of CD at δ1.17−1.39 ppm and heterocyclic ring peaks (Ar−H) of CD at 6.76−7.63 ppm, as well as the typical peak of PEI (−NHCH2CH2−) at δ 2.25−3.42 ppm, indicating the successful coupling of PEI−CD on the surface of SWNTs. The amount of PEI−CD attached to the SWNTs was measured by TGA analysis. The SWNTs remained steady under 600 °C (Figure 5Ca), whereas the PEI−CD was definitely degraded at about 450 °C (Figure 5Cd). At 450 °C, the weight losses of SWNTs, SWNT−COOH, and SWNT−PEI−CD were 0%, 23.1%, and 35.7%, respectively; thus, the weight of PEI−CD on SWNT surface was about 12.6% (Figure 5Cb,c). As mentioned above, the development of SWNTs in biomedical applications was seriously hampered on account of the highly hydrophobic surfaces and poor biocompatibility of SWNTs. Therefore, it is of extensive scientific significance to improve the biocompatibility of SWNTs.52 In this study, the dispersive properties of pristine SWNT, SWNT−COOH, and SWNT−PEI−CD in pure water were investigated. As shown in Figure 5D, the dispersibility of pristine SWNTs was poor, while SWNT−COOH and SWNT−PEI−CD showed much-better dispersibility. After standing for 15 days (up to 6 months), 23360


Research Article


Figure 9. Down-regulation of VEGF mRNA and protein levels in PANC-1 cells after treatment with various formulations. (A) qRT-PCR analysis of VEGF suppression on mRNA levels in PANC-1 cells. βActin mRNA was utilized as internal control. Results were represented as mean ± SD (n = 3). *, P < 0.05; **, P < 0.01; and ***, P < 0.001. (B) ELISA analysis of the secretion of VEGF protein. Results were expressed as mean ± SD (n = 3). *, P < 0.05 and **, P < 0.01. (C) Western blot analysis of VEGF suppression in PANC-1 cells.

Figure 8. Cellular uptake and intracellular distribution of siRNAloaded complexes. CLSM images (A) and FACS analysis (C) of HEK293, MCF-7, and PANC-1cells transfected with different nanoparticles (SWNT−PEI−CD, PEI−CD, SWNT−PEI, and PEI− 1.8 KDa) containing FAM-siRNA (green). (B) Intracellular distribution of SWNT−PEI−CD/siRNA complexes in PANC-1 cells. The green fluorescence represented FAM-siRNA, while endosomes were labeled by Lyso Tracker Red (red) and nuclei were labeled by Hoechst 33258 (blue). (D) FACS analysis of cell uptake efficiency in PANC-1 cell. The final concentration of siRNA was 100 nM. Results were shown as mean ± SD (n = 3). *, P < 0.05; **, P < 0.01; and ***, P < 0.001.

intensively explored. Previous results have shown that the toxicity of SWNTs is correlated with multiple factors, including the nanotube length,53 degree of impurities,54 and dispersion status,55 as well as the type of functionalization.56,57 Untreated SWNTs showed obvious toxicity in vitro and in vivo. As shown in Figure 7, for the PEI−CD + SWNT group, a low concentration of SWNT (5 μg/mL) led to an almost 20% decrease in cellular viability. Only less than 40% of the cells were alive after being incubated with 20 μg/mL SWNTs for 24 h, which was even less than that of the PEI 25 KDa group. The PEI−CD + SWNT group showed the highest cytotoxicity, which was possibly attributed to the combination of positively charged PEI and non-functionalized SWNTs. For the PEI−CD alone group, its cytotoxicity was low, as shown in Figure 7. On the contrary, for SWNT−PEI−CD group, a significant increase in cellular viability was observed. Cell viabilities remained above 80% at a SWNT−PEI−CD dose of 20 μg/mL, which was higher than that adopted in the subsequent studies. No

Furthermore, the ζ potential of the SWNT−PEI−CD/ siRNA complexes was investigated. As depicted in Figure 3C, the SWNT−PEI−CD/siRNA complexes showed a negative potential at low w/w ratios. However, when the w/w ratio exceeded 1.5, the ζ potential changed to positive and remained stable (28−30 mV) at w/w ratios beyond 10, indicating the complete complexation of siRNA molecules by the SWNT− PEI−CD. 3.3. In Vitro Cytotoxicity Studies. As a new-type of gene delivery carriers, the potential toxicity of SWNTs has been 23361


Research Article


Figure 10. In vitro angiogenesis assay. (A) Microscopic observation of the inhibition of tubule formation in HUVECs on ECMatrix after treatment of different formulations for 24 h at siRNA or siVEGF concentration equivalent to 100 nM. (B) The quantitative analysis of the inhibition of tubule formation. Results were represented as mean ± SD (n = 3). *, P < 0.05; **, P < 0.01; and ***, P < 0.001.

conjugate without CD modification. In addition, the cell specific fluorescence intensity of the PEI−CD/FAM−siRNA complexes decreased in the order: PANC-1 cells > MCF-7 cells > HEK293 cells. These observations indicated that cellular uptake of nanotubes could be promoted after CD modification due to the specificity of CD to AT1 receptor that was overexpressed in PANC-1 cells. More importantly, it was also noticed that fluorescence intensity was significantly stronger in cells treated with the SWNT−PEI−CD/FAM−siRNA complexes than those treated with the PEI−CD/FAM−siRNA complexes in each cell line. Furthermore, it was found that the cellular uptake efficiency of the SWNT−PEI/FAM−siRNA complexes was lower than that of the PEI−CD/FAM−siRNA complexes in AT1R over-expressed PANC-1 cells. Interestingly, with the AT1R normally expressed MCF-7 cells and negatively expressed HEK293 cells, the SWNT−PEI/FAM−siRNA complexes uptake was more efficient than that of the PEI− CD/FAM−siRNA complexes. Taking these findings into consideration, it was clear that the SWNT-based nanocomplexes could enter into cells more effectively than nanocomplexes without SWNTs. The enhanced cellular uptake of FAM−siRNA probably resulted from the cell penetrating ability of the SWNTs. It has been reported that SWNTs could lead to formation of transient nanoscaled holes on cells membrane, which was similarly to the nanosyringe model, thus obviously breaking down the integrity of cell membranes. Thus, the entrance of FAM−siRNA became easier.58,59 To further confirm the SWNT−PEI−CD enhanced FAM− siRNA transfection efficiency, the intracellular distribution was studied by CLSM. As seen in Figure 8B, analysis of the green (FAM−siRNA) and red (endosomes) fluorescence dots and the merged yellow stains suggested that siRNA was successfully delivered into cells via the endocytosis mechanism. Notably, more green dots could be investigated in the cells with time, indicating the separation of red and green dots and the release of green dots into cytoplasm. These results demonstrated that SWNT−PEI−CD/FAM−siRNA complexes could escape from endosomes to cytoplasm for effective gene transfection. They showed that SWNTs had the ability to be internalized efficiently into different cell types. Also, the results showed that the gene delivery activity of SWNTs covalently bound PEI

significant difference was observed between the different cell lines for all groups, suggesting that PEI−CD functionalization greatly reduced cytotoxicity and improved biocompatibility of SWNTs. We believe there were three main factors that might account for the findings. First, the SWNTs in our study were initially purified to remove the impurities. Second, the SWNTs were oxidized and shortened to a uniform particle size about 200 nm. Moreover, the functionalization of SWNTs with PEI− CD prevented the aggregation of SWNTs in water and, therefore, greatly improved dispersibility. In view of these results, we concluded that the SWNT−PEI−CD with low toxicity was feasible for drug delivery system at the test concentration. 3.4. In Vitro Cell Uptake and Intracellular Distribution Studies. It is suggested that the tumor-targeting properties of CD are based on its ability to bind to AT1R expressed on cells.37 Research has revealed that AT1R is highly expressed in PANC-1 cells, so these cells were used in this study. AT1R normally expressed MCF-7 cells and negatively expressed HEK293 cells were used as controls. For siRNA transfection, it is important to achieve high cellular uptake and specific intracellular distribution of siRNA via an effective delivery system. In this study, FAM-labeled siRNA was used as a fluorescent indicator to investigate the cellular uptake capability of each FAM−siRNA-loaded complex at optimal w/w ratios in PANC-1, MCF-7, and HEK293 cells. The fluorescence distribution and intensity of different complexes were detected by CLSM and flow cytometry. As shown in Figure 8A, all of the prepared complexes could deliver siRNA into cells to a certain extent. The fluorescence intensity representing FAM−siRNA uptake by PANC-1 cells decreased gradually in the order: SWNT−PEI−CD/FAM− siRNA > PEI−CD/FAM−siRNA > SWNT−PEI/FAM− siRNA > PE -1.8 KDa/FAM−siRNA. In contrast, the cellular uptake of SWNT−PEI/FAM−siRNA was more efficient than PEI−CD/FAM−siRNA in the MCF-7 and HEK293 cells. Based on the quantitative results from fluorescence-activated cell sorting (FACS) analysis shown in Figure 8C,D, the CDcontained conjugates SWNT−PEI−CD and PEI−CD without SWNTs exhibited significantly higher FAM−siRNA fluorescence intensities in PANC-1 cells than the SWNT−PEI 23362


Research Article


dynamic structures that continuously loose and gain phospholipid molecules released by membrane sculpting processes.60 In the presence of a sink such as that provided by SWNTs, depletion of the endosomal membrane would be expected to occur, and this is believed to be the mechanism by which cationic liposomes enable endosomal release when they act as transfection agents. Thus, functionalized SWNTs with PEIs might in principle enable both osmotic (proton sponge) and membrane depletion mechanisms to act cooperatively in endosomal membrane disruption. Also, the large, rigid, nonbiodegradable structure of SWNTs may contribute a physical disruption mechanism to damage endosomal membranes. Endosomes undergo contraction due to bud formation and membrane sculpting before fusion with lysosomes. The large, rigid structure of endocytosed SWNT may enhance transfection efficiencies by delaying the contraction step and thus delaying the destruction of siRNA by lysosomal enzymes. When the above results are taken together, it is clear that an SWNT-based delivery system could enhance cellular uptake efficiency via its cell-penetrating ability and improve gene transfection efficiency via an endosomal membrane-disruption mechanism. 3.5. In Vitro VEGF Gene Silencing. VEGF, secreted by most solid tumors, plays important roles in the process of angiogenesis, involving endothelial cell proliferation, invasion, and migration.37 Therefore, the inhibition of VEGF expression is considered as a powerful strategy for inhibiting tumor growth and metastasis. Hence, the effectiveness of SWNT−PEI−CD for siVEGF delivery in PANC-1 cells was detected by qRTPCR, Western blot, and ELISA. Lipofectamine 2000 was used as the positive control. VEGF silence was first tested by measuring VEGF mRNA levels by qRT-PCR. As shown in Figure 9A, the reduction of VEGF mRNA expression was ∼20% in the group treated with SWNT−PEI−CD/siRNA (scrambled VEGF). This could possibly be exclusively attributed to the inhibition caused by CD. In comparison, the expression of VEGF mRNA in the SWNT−PEI/siVEGF + CD, SWNT−PEI/siVEGF, PEI−CD/ siVEGF, and SWNT−PEI−CD/siVEGF treatment groups was reduced by ∼50%, ∼30%, ∼35%, and ∼70%, respectively. The trend of this result was consistent with the cellular uptake experiment in PANC-1 cells, indicating that the performance of SWNT−PEI−CD was superior to that of SWNT−PEI and PEI−CD conjugates in inhibiting VEGF mRNA transcription. From the above results, it could be suggested that an AT1R targeting SWNT-based delivery system would be beneficial for enhancing cellular uptake of siRNA and thus promoting gene silencing efficacy. More importantly, compared with the SWNT−PEI/siVEGF and CD mixed-delivery system, SWNT−PEI−CD/siVEGF complexes co-delivered CD and siVEGF into PANC-1 cells and significantly enhanced the inhibition of VEGF mRNA level. This result suggested a combined function of CD and siVEGF, and the synergistic effect of the two therapeutic agents was demonstrated by the reduction of VEGF protein expression after treating with SWNT−PEI−CD/siVEGF complexes, which was in accordance with the Western blot data (Figure 9C). To further verify the VEGF silence effect of SWNT−PEI− CD/siVEGF complexes, the level of VEGF protein was determined using a human ELISA kit at 48 h post-transfection. As depicted in Figure 9B, the results were in agreement with the qRT-PCR assay. It could be found that SWNT−PEI−CD/ siVEGF complexes show more dramatic downregulation of

Figure 11. In vivo tumor-targeting analysis. In vivo tumor-targeted observation after the intravenous injection of (A) Cy7-labled SWNT− PEI/siVEGF, (B) Cy7-labled PEI−CD/siVEGF, and (C) Cy7-labled SWNT−PEI−CD/siVEGF complexes in PANC-1 tumor xenograft mice. (D) The quantification of average flourescence intensity in the tumor site of different groups. Results were expressed as mean ± SD (n = 5). **, P < 0.01 and ***, P < 0.001.

was more efficient in comparison to PEI 1.8 KDa. There were several probable mechanisms by which covalent attachment of PEI−CD to SWNTs could result in the enhanced transfection efficiencies observed in the study. First, the characterization studies above demonstrated that PEI−CD-functionalized SWNTs were able to fully condense siRNA (Figures 3C and 6B). Most importantly, particle-like SWNTs might result in improvements in the endocytosis step of the transfection. The electrostatic interaction between the negatively charged cell membrane and the positively charged SWNT−PEI−CD/ siRNA complexes and the presence of the physical property of SWNTs might facilitate endosomal escape by providing a sink for lipid components released from the membrane enclosing the endocytotic vesicle. Biological membranes are 23363


Research Article


Figure 12. In vivo antitumor effects of different formulations on BALB/c nude mice bearing of PANC-1 tumor. (A) Tumor-growth curve after treatment with different formulations. (B) Changes of body weight. (C) The VEGF expression levels in tumor tissues evaluated by a human-specific ELISA kit. (D) Ex vivo images of tumor tissues and tumor weight. (E) H&E analysis of tumor tissues in mice (400×) with the treatment of Control, SWNT−PEI−CD/siRNA, SWNT−PEI/siVEGF + CD, SWNT−PEI/siVEGF, PEI−CD/siVEGF, and SWNT−PEI−CD/siVEGF. Results were expressed as mean ± SD (n = 5 in panels A, B, D, and E and n = 3 in C). *, P < 0.05; **, P < 0.01; and ***, P < 0.001.

VEGF secretion than any of the other mono- or mixed-delivery systems. These investigations markedly indicated the synergistic effect of CD and siVEGF gene in VEGF gene silence, which could ultimately restrain tumor associated angiogenesis. Such effects could be a consequence of the following factors. First, VEGF plays an essential role in tumor growth and is the most important factor regulating neovascularization, an effect it exerts via specific receptors. When siRNA silences the VEGF

gene, the expression of VEGF decreases, resulting in inhibition of tumor growth, invasion, and metastasis. Second, there is evidence that CD inhibits VEGF expression in tumor cells throught down-regulating the VEGF transcriptional factors Ets1 and HIF-1, thereby significantly inhibiting tumor angiogenesis.61 3.6. In Vitro Angiogenesis Assay. The development of endothelial cell capillary structure is a multistep process 23364


Research Article


Figure 13. (A) Immunohistochemistry investigations of representative tumor tissues after staining with CD31 antibody. (B) Quantitative analysis of CD31-positive microvessels in PANC-1 tumor-bearing nude mice with treatment of different formulations at Day 14. Results were represented as mean ± SD (n = 3). *, P < 0.05; **, P < 0.01; and ***, P < 0.001.

Figure 14. In vitro and in vivo safety evaluation. (A) In vitro hemolysis of PEI 1.8 KDa and SWNT−PEI−CD at various concentrations. (B) In vivo toxicity examined of SWNT−PEI−CD/siVEGF complexes by hemotologic analysis. The hematologic parameters (red blood cell count (RBC), white blood cell count (WBC), and hemoglobin (Hb)) and serum biochemistry indicators for hepatic functions (alanine aminotransferase (ALT) and aspartate aminotransferase (AST)) and for kidney functions (creatinine (Cr)) were determined. Results were expressed as mean ± SD (n = 3).

involving cell adhesion, migration, differentiation, and growth and is a vital step in neovascularization. To assess the angiogenic inhibition of the various formulations in vitro, HUVECs were co-cultured with the formulations and plated onto Matrigel for 24 h. As shown in Figure 10, in the investigated time, the cells aligned themselves end-to-end and formed interconnecting networks of capillary tubes, a sign of the facilitation of angiogenesis. All of the formulations exerted a suppressive effect on the tubular structure formation to some extent. SWNT−PEI−CD/siRNA complexes showed low suppressive effect, while PEI−CD/siVEGF and SWNT−PEI/ siVEGF complexes also exerted an inhibition effect toward the control group. In comparison, treatment with the SWNT−PEI/ siVEGF complexes plus free CD groups and treatment with the SWNT−PEI−CD/siVEGF group resulted in the nearly unchanged morphology of HUVECs and almost no tube formation. In particular, SWNT−PEI−CD/siVEGF complexes exerted the highest tube formation inhibition rate of 88.36% in all groups. These observations suggest that SWNT−PEI−CD/ siVEGF complexes as a co-delivery system were more effective

in inhibiting morphological differentiation of the HUVECs and angiogenesis. The effects were derived from both CDdependent angiogenesis repression and the silence of the most-important angiogenesis cytokine VEGF, instigated by siVEGF. 3.7. In Vivo Imaging Analysis. SWNT−PEI−CD/ siVEGF complexes have exhibited preferable AT1R targeting ability and considerable VEGF down-regulation efficiency in vitro. However, it is still necessary to investigate the drug accumulation in tumor tissue and main organs in vivo to evaluate the efficacy and safety of the nanodelivery system. In this study, Cy-7 was used as near-infrared fluorescence probe to monitor the tumor-targeting efficiency of the complexes by employing the Kodak in vivo FX PRO imaging system. The mice were treated with Cy7-labeled SWNT−PEI/siVEGF, Cy7-labeled PEI−CD/siVEGF, or Cy7-labeled SWNT−PEI− CD/siVEGF complexes, respectively. As shown in Figure 11, at 2 h post-injection, the Cy7 fluorescence signal was obviously observed in the tumors of the nude mice treated with SWNT−PEI−CD/siVEGF complexes, 23365


Research Article


Figure 15. Histologic assessments of major organs in mice (400×) treated with (A) nothing (control), (B) SWNT−PEI−CD/siRNA, (C) SWNT− PEI/siVEGF + CD, (D) SWNT−PEI/siVEGF, (E) PEI−CD/siVEGF, and (F) SWNT−PEI−CD/siVEGF.

while a weak fluorescent signal was detected in the PEI−CD/ siVEGF group and almost no fluorescent signal was detected in the SWNT−PEI/siVEGF group. This result showed a preferential accumulation of SWNT−PEI−CD/siVEGF complexes at the tumor site. Throughout the observation time, an enhancement of fluorescence signal was monitored at the tumor sites for the SWNT−PEI−CD/siVEGF group as compared to the SWNT−PEI/siVEGF and PEI−CD/siVEGF groups. The highest fluorescence intensity at the tumor sites was found at 6 h after injection for each formulation, after which time the fluorescence intensity decreased gradually. The quantitative analysis of average fluorescence intensity in Figure 11(D) showed the significant difference between each group. It was easy to find that the fluorescence intensity of SWNT− PEI−CD/siVEGF at 6 h post-injection was almost twice as strong as SWNT−PEI/siVEGF and PEI−CD/siVEGF groups. In addition, for SWNT−PEI/siVEGF and PEI−CD/siVEGF complexes, strong fluorescence signals were detected in the liver tissue after intravenous injection, suggesting that a strong entrapment of nanotubes by the reticuloendothelial system (RES) had occurred. These findings provided decisive evidence that SWNT−PEI−CD/siVEGF complexes can significantly enhance tumor accumulation for AT1R targeted delivery. In general, it is well-known that the enhanced permeability and retention (EPR) effect is an important mechanism influencing the accumulation of nanoscale particles in tumors. The powerful tumor targeting ability of SWNT−PEI−CD/siVEGF complexes could be attributed to a combined effect of preferable EPR effect and the strong affinity of CD to AT1R over-expressed in tumors. Above all, these in vivo biodis-

tribution studies indicated that SWNT−PEI−CD/siVEGF complexes were a highly specific co-delivery system to AT1R over-expressed tumors. 3.8. In Vivo Antitumor Efficacy and Angiogenesis Suppression. To evaluate the in vivo tumor growth inhibition and antiangiogenesis effects of the SWNT−PEI−CD/siVEGF co-delivery system, the PANC-1 xenografted tumor model was constructed. As shown in Figure 12A, compared with the 5% glucose treatment group, tumor-growth inhibition was clearly seen in the SWNT−PEI−CD/siRNA group. However, neither the mono- nor the mixed-delivery systems with siVEGF showed significant inhibition of tumor growth. These results suggested that the decrease in tumor volume was mainly attributed to the treatment of siVEGF delivered by the complexes. Compared to the nontargeted SWNT−PEI/siVEGF group (Figure 12D), a noticeable reduction in tumor volume was found in the targeted SWNT−PEI−CD/siVEGF group, indicating that the AT1R targeted delivery method was valuable for tumor therapy. Moreover, the SWNT−PEI−CD/siVEGF group exhibited significantly greater tumor growth inhibition than the SWNT−PEI/siVEGF with free CD group, demonstrating the advantage of the co-delivery system. In addition, it is noteworthy that the inhibition of tumor growth was remarkable stronger for the SWNT−PEI−CD/siVEGF group than for the PEI−CD/siVEGF group without SWNT (Figure 12E). This result could be attributed to the distinctive performance of the SWNT-based delivery system with the excellent enhancement of cellular uptake and gene transfection efficiency as mentioned above. Overall, the superior co-delivery system might provide a promising approach for effective anticancer treatment. 23366


Research Article


SWNT−PEI−CD/siVEGF complexes did not cause mortality at the injected doses. The behaviors of mice were normal, indicating that the SWNT−PEI−CD/siVEGF complexes did not cause acute toxicity. Moreover, no significant change of body weight was found in the mice between treatment groups and the control group (Figure 12B). Hematological and biochemical analyses were carried out to quantify the potential in vivo toxicity of SWNT−PEI−CD after administration. Quantitative evaluation of hematologic parameters (RBC; white blood cells, WBC; and hemoglobin, Hb) and serum biochemistry indicators for hepatic function (alanine aminotransferase, ALT; and aspartate aminotransferase, AST) and for kidney function (creatinine, Cr) was performed. As seen in Figure 14B, these indicators were at normal levels, and no difference was detected between the control and test groups. Furthermore, we conducted a histological analysis for the safety of the complexes in some major organs, including liver, heart, lungs, and kidneys. The representative photographs of H&E-stained samples were shown in Figure 15. No obvious tissue damage or inflammatory cell infiltration occurred in any of the test groups, and no significant toxicity was found in any of the organs compared to the control group. Taken together, we concluded that the administration of SWNT−PEI−CD/ siVEGF complexes does not cause any obvious toxic effects to blood and major organs.

To further confirm the tumor inhibition by SWNT−PEI− CD/siVEGF complexes, which was concerned with the downregulation of VEGF level, the tumors were isolated for the analysis of the VEGF expression levels by ELISA. As seen in Figure 12C, VEGF expression levels in tumor tissues, from highest to lowest, were in the following order: 5% glucose ≈ naked siVEGF ≈ SWNT−PEI−CD/siRNA > PEI−CD/ siVEGF ≈ SWNT−PEI/siVEGF ≈ SWNT−PEI/siVEGF + CD > SWNT−PEI−CD/siVEGF. A significant down-regulation of VEGF expression in tumor was detected in the SWNT−PEI−CD/siVEGF group compared to the control group. It is revealed that the VEGF-targeted siRNA delivery system could successfully down-regulate the expression of VEGF protein by gene silencing and therefore induce the inhibition of neovasculature. The MVD, a surrogate marker of tumoral angiogenesis, was highlighted by means of an immunohistochemistry assay. As shown in Figure 13, microvessel density was significantly higher in the control group (VEGF-positive) than that in the SWNT− PEI−CD/siVEGF group (VEGF-negative). With VEGF expression increasing, microvessel density became higher. These results have demonstrated that constructed SWNT− PEI−CD/siVEGF complexes exert a notable angiogenesis inhibition in tumors. The reasons for such powerful tumor angiogenesis inhibition are as follows: (1) preferential accumulation in tumor and superior AT1R targeting specificity toward the tumor cells, (2) higher transfection efficiency of internalized SWNT−PEI−CD/ siVEGF complexes owing to stronger buffering capacity of PEI−CD conjugates and cell-penetrating ability of SWNTs, and (3) synergistic efficacy of potential anticancer agents (CD) and powerful VEGF gene-silencing siRNA in antiangiogenic therapy. This powerful tumor angiogenesis repression caused by SWNT−PEI−CD/siVEGF complexes ultimately led to an anticipated enhanced antitumor efficacy in vivo. In summary, the smartly designed SWNT−PEI−CD/siVEGF nanosystem could be used as an effective tumor-targeted and antiangiogenesis co-delivery system, holding great potential for achieving more-effective anticancer activity. 3.9. Safety Evaluation in Vitro and in Vivo. The potential toxicity of SWNTs cause close attention for in vivo applications. Studies have shown that PL−polyethylene glycolfunctionalized SWNTs are excreted from mice via the biliary and renal pathways after intravenous injection and, therefore, are safe for use as a drug carrier.62 Other studies have shown that well-functionalized SWNTs are innocuous to cells and mice in the range of doses tested.29,52 In our study, the in vitro and in vivo toxicity of SWNT−PEI−CD/siVEGF complexes was extensively investigated. For in vivo applications, nanomaterials must possess excellent blood compatibility, such as low hemolytic effect. Herein, the hemolysis experiments were performed for evaluating the blood compatibility of SWNT−PEI−CD. As presented in Figure 14A, SWNT−PEI−CD and CD did not display significant hemolytic activity at all tested concentrations (

Synergistic Suppression of Tumor Angiogenesis by the Co-delivering

Recommend Documents