Bufalin

Increasing the Anticancer Performance of Bufalin (BUF) by Introducing an Endosome-Escaping Polymer and Tumor-Targeting Peptide in the Design of a Polymeric Prodrug

Xiao-jing Shi, Yan-yan Qiu, Hui Yu, Cheng Liu, Yu-xia Yuan, Pei-hao Yin, Tao Liu

Abstract

A well-defined multifunctional brush-type polymeric prodrug covalently linked with an anticancer drug (bufalin, BUF), a tumor-targeting peptide (RGD), and an endosome-escaping polymer, poly(N,N-diethylaminoethyl methacrylate-co-butyl methacrylate (P(DEA-co-BMA)), was developed. Its anticancer performance against colon cancer was investigated in vitro and in vivo. Reversible addition-fragmentation transfer (RAFT) polymerization of oligo(ethylene glycol) monomethyl ether methacrylate (OEGMA), 2-((3-(tert-butoxy)-3-oxopropyl)thio)ethyl methacrylate (BSTMA), and 2-(2-bromoisobutyryloxy)ethylmethacrylate (BIEM) afforded the multifunctional random copolymer, P(OEGMA-co-BSTMA-co-BIEM), in which hydrophilic POEGMA can stabilize nanoparticles in water, PBSTMA can be converted into carboxyl groups, and PBIEM can be employed as a macromolecular atom radical transfer polymerization (ATRP) initiator. The ATRP of DEA and BMA using P(OEGMA-co-BSTMA-co-BIEM) as a macromolecular ATRP initiator led to the formation of the pH-responsive brush-type copolymer, P(OEGMA-co-BSTMA)-g-P(DEA-co- BMA). After hydrolysis by trifluoroacetic acid and post-functionalization the final polymeric prodrug, P(OEGMA-co-BUF-co-RGD)-g-P(DEA-co-BMA), was obtained with a drug content of ~7.8 wt%. P(OEGMA-co-BUF-co-RGD)-g-P(DEA-co-BMA) can be assembled into nanoparticles (BUF- NP-RGD) in aqueous solution with a diameter of 148.4 ± 0.7 nm and a zeta potential of -7.6 ± 0.4 mV. BUF-NP-RGD exhibited controlled drug release in the presence of esterase. Additionally, P(OEGMA-co- BSMA)-g-P(DEA-co-BMA) showed a significant hemolysis effect at a pH comparable to that of endosomes/lysosomes. Cell viability and a tumor-bearing nude mouse model were employed to evaluate the anticancer efficacy of BUF-NP-RGD. It was revealed that BUF-NP-RGD showed improved anticancer performance compared with that of free BUF both in vitro and in vivo. Histological and immunochemical analysis further demonstrated that BUF-NP-RGD exhibited improved cell apoptosis, angiogenesis inhibition, and an anti-proliferation effect.

Keywords: Colon cancer, Targeting, P(DEA-co-BMA), Brush copolymers

1. Introduction

Bufalin (BUF), the main active component extracted from the traditional Chinese medicine Ch’an Su (toad venom), exhibits promising anticancer activity against a broad range of cancers by inhibiting cell proliferation and inducing apoptosis of tumor cells [1-8]. However, its poor water solubility and severe adverse effects (e.g. high cardiac toxicity, allergic shock, acute fever, and sinus bradycardia) limit its further clinical applications [9, 10]. Additionally, organic solvents such as dimethylsulfoxide (DMSO) must be employed to dissolve BUF, leading to potential risks of vasoconstriction and neurological damage [11, 12].
In this context, BUF was physically encapsulated into the hydrophobic domains of nanoparticles to improve its anticancer performance and expand its application areas. Nanoparticles such as wheat germ agglutinin-grafted lipid, mPEG-PLGA-PLL-cRGD, pluronic polyetherimide nanoparticles, and biotinylated chitosan were developed to encapsulate BUF into nanoparticles and improved anticancer activity was achieved [10, 13, 14]. However, premature/burst drug release after intravenous administration constrains their further application. Covalently linking pharmacologically active drugs with polymers seems to be a feasible strategy. By carefully designing the linkage between polymer and drug, triggered drug release under specific stimuli can be obtained, which can effectively circumvent premature/burst drug release as well as maintain the high anticancer activity [15]. Unfortunately, chemical modification of BUF often leads to dramatically reduced anticancer activity because the 3-hydroxyl group (3C moiety) is the most chemically active group as well as anticancer active moiety [16]. This challenges the design of applicable prodrugs of BUF.
Recently, we employed the esterase-sensitive -thioester bond as a linker to fabricate a novel poly(ethylene glycol) (PEG)-based polymeric prodrug of BUF, PEGS-BUF [15]. It was revealed that PEGS-BUF showed dramatically improved water solubility and stability at the prerequisite of maintaining its original anticancer activity both in vitro and in vivo. Next, we developed another multifunctional polymeric prodrug covalently linked with a tumor-targeting peptide (RGD) [17]. The obtained tumor targeting polymeric prodrug exhibited improved anticancer performance in addition to improved water solubility and stability.
Nevertheless, it was still possible to further optimize the design of the polymeric prodrug of BUF. It was repeatedly documented that nanoparticles enter cells mainly via endocytosis and the predominant fate of these nanoparticles is enzymatic degradation in the lysosomes or recycling and extracellular clearance [18, 19]. Designs that can help nanoparticles escape from endo/lysosomes would significantly improve the effective utilization of drug molecules and thus ultimately improve their anticancer activities. Various active components including pH-responsive lipids, viral fusogenic proteins, peptides, and synthetic polymers were developed to improve the endosome escaping capabilities of nanoparticles via the mechanism of triggering membrane destabilization at acidic pH[20-22]. In this context, poly(N,N-diethylaminoethyl methacrylate-co-butyl methacrylate, P(DEA-co- BMA), was employed to enhance the pDNA delivery efficiency of the cationic polymer, poly(N,N-dimethylaminoethyl methacrylate) by improved endosome escaping effects [22]. It was shown that the reduction of the solution pH to values representative of endosomal/lysosomal compartments can effectively destabilize endo/lysosomal membranes and ultimately improve the gene expression efficiency.
In the current work, we developed a novel type of multifunctional brush-type polymeric prodrug, in which BUF was attached onto the polymer by an esterase-sensitive -thioester bond, a tumor-targeting peptide (RGD) was anchored onto the polymer by amide bond, and P(DEA-co-BMA) was introduced by atom radical transfer polymerization (ATRP) to act as an endosome escaping segment (Scheme 1). Reversible addition-fragmentation transfer (RAFT), ATRP, and post-functionalization were employed for the fabrication of the polymeric prodrug. Nuclear magnetic resonance (NMR), gel permeation chromatography (GPC), and dynamic laser light scattering (LLS) were employed to characterize the polymers and their assemblies. Typical malignant colon cancer cell line HCT116 cells and tumor-bearing mice were used to evaluate the anticancer performance of the obtained polymeric prodrug in vitro and in vivo, respectively. Flow cytometry, histological and immunochemical analysis were utilized to examine the cell apoptosis, angiogenesis inhibition, and anti-proliferation effect caused by the polymeric prodrug.

2. Materials and methods

2.1 Materials

Oligo(ethylene glycol) monomethyl ether methacrylate (OEGMA, Mn = 500 g/mol, mean degree of polymerization, DP, is 8-9) purchased from Aladdin (Shanghai, China) was passed through a neutral alumina column to remove the inhibitor and then was stored at -20 °C prior to use. 2,2′-Azobis(2-methylpropionitrile) (AIBN; Aladdin) was recrystallized from 95% ethanol. N,N-Diethylaminoethyl methacrylate (DEA; Sigma-Aldrich, St Louis, MO, USA) and butyl methacrylate (BMA, 98%; Sigma-Aldrich) were vacuum-distilled over calcium hydride (CaH2) and were stored at -20 °C prior to use. N,N,N′,N′′,N′′-Pentamethyldiethylenetriamine (PMDETA, 98%) and copper(I) bromide (CuBr, 98%) were purchased from Sigma-Aldrich and were used as received.
Bufalin (BUF, 99%; Chengdu PuRuiFa Technology Development Co. Ltd., Chengdu, China), cyclic RGD (cRGD, shortened as RGD in the subsequent sections; Chinese Peptide Company, Hangzhou, China), and esterase solution (porcine liver, 5 KU; Sigma-Aldrich) were used as received. Fetal bovine serum (FBS), penicillin, streptomycin, Dulbecco’s modified Eagle’s medium (DMEM), and hematoxylin and eosin (H&E) were purchased from Thermo Fisher Scientific (Waltham, MA, USA) and were used as received. Neutral buffered paraformaldehyde (4%), the cholecystokinin (CCK-8) assay kit, bovine serum albumin (BSA, 5%), and 4,6-diamidino-2-phenylindole (DAPI) were purchased from Beyotime Biotech (Shanghai, China) and were used as received. Annexin V-FITC/PI apoptosis detection kit (BD Biosciences, San Jose, CA, USA), TdT-dependent dUTP-biotin nick end labeling (TUNEL) assay kit (Promega, Madison, WI, USA) and 3,3′-diamino-benzidine substrate kit (Vector Laboratory, Burlingame, CA, USA) were used as received. Triton X-100, N,N′-dicyclohexylcarbodiimide (DCC), 4-dimethylaminopyridine (DMAP), trifluoroacetic acid (TFA) and all other reagents were purchased from Sinopharm Chemical Reagent Co. Ltd. (Shanghai, China) and were used as received. Triethylamine (TEA) and dichloromethane (CH2Cl2) were dried over CaH2 and distilled just before use. Water was deionized with a Milli-Q SP reagent water system (Millipore) to a specific resistivity of 18.4 MΩ cm. Benzyl 2-methyl-2-(((propylthio)carbonothioyl) thio)propanoate (BPTPA) [23], 2-(2-bromoisobutyryloxy)ethylmethacrylate (BIEM) [24], and 2-((3-(tert-butoxy)-3-oxopropyl)thio)ethyl methacrylate (BSTMA) [17] were synthesized according to literature procedures.

2.2 Sample synthesis

The synthetic routes employed for the preparation of P(OEGMA-co-BUF-co-RGD)-g-P(DEA- co-BMA) are shown in Scheme 2. All 1H NMR spectra were recorded using a Bruker AV300 NMR spectrometer (resonance frequency 300 MHz) operated in the Fourier transform mode. CDCl3 was used as the solvent. The molecular weight (Mn) and dispersity (ÐM = Mw/Mn, where Mw represents weight-average molecular weight) were determined by GPC equipped with a Waters 1515 pump and Waters 2414 differential refractive index detector (set at 30 °C), employing a series of two linear Styragel columns (HR2 and HR4) at an oven temperature of 45 °C. The eluent was DMF at a flow rate of 1.0 mL/min. A series of low ÐM polystyrene standards was employed for calibration.
Synthesis of P(OEGMA-co-BSTMA-co-BIEM). BPTPA (33 mg, 0.1 mmol), OEGMA (3.0 g, 6 mmol), BSTMA (686 mg, 2.5 mmol), BIEM (140 mg, 0.5 mmol), and AIBN (1.6 mg, 0.01 mmol) were charged into a glass ampoule equipped with a magnetic stirring bar. The ampoule was degassed by three freeze-pump-thaw cycles and was sealed under vacuum. After thermostatting at 70 °C in an oil bath and stirring for 5 h, the reaction was terminated by quenching into liquid N2, exposure to air, and dilution with methanol. The mixture was then purified by dialysis (cellulose membrane; molecular weight cutoff, MWCO, 3,500 Da) against deionized water for 48 h to afford P(OEGMA-co- BSTMA-co-BIEM) (2.5 g, yield: 64.8%). The obtained polymer was then treated at 70 °C for 2 h with an excess of AIBN (164 mg, 10 mmol) in 1,4-dioxane (5 mL) under nitrogen to remove the trithiocarbonate end-groups. After treatment the yellow product became almost colorless, implying the complete removal of trithiocarbonate end-groups. The Mn and ÐM were determined to be 31,000 and 1.03 by GPC using DMF as the eluent (Table 1). 1H NMR analysis revealed the contents of OEGMA, BSTMA, and BEMA in the polymer were ~60 mol%, ~35 mol%, and ~5 mol%, respectively, and the DP of the polymer was determined to be ~74 (Fig. S1). Thus the polymer was denoted as P(OEGMA0.60-co-BSTMA0.35-co-BIEM0.05)74 and shortened as P(OEGMA-co-BSTMA- co-BIEM) in the subsequent sections.
Synthesis of P(OEGMA-co-BSMA)-g-P(DEA-co-BMA). P(OEGMA-co-BSTMA-co-BIEM) (606 mg, 74 mol Br moieties), DEA (822 mg, 4.44 mmol), BMA (420 mg, 2.96 mmol), DMF (3 mL), and PMDETA (13 mg, 74 mol) were charged into a glass ampoule equipped with a magnetic stirring bar. The ampoule was degassed by three freeze-pump-thaw cycles, and then CuBr (11 mg, 74 mol) was introduced under the protection of N2 before freezing and sealing under vacuum. After thermostatting at 60 °C in an oil bath and stirring for 3 h, the reaction was terminated by quenching into liquid N2, exposure to air, and dilution with tetrahydrofuran. The mixture was then purified by dialysis (cellulose membrane; MWCO: 3,500 Da) against deionized water for 24 h before lyophilization. The obtained polymer was added in anhydrous CH2Cl2 (10 mL). Next, TFA (10 mL) was added and the reaction mixture was stirred at room temperature for 5 h. The mixture was evaporated to dryness on a rotary evaporator, dissolved in 10 mL of DMF, and then purified by dialysis (cellulose membrane; MWCO: 3,500 Da) against deionized water to afford P(OEGMA-co- BSMA)-g-P(DEA-co-BMA) (1.3 g, yield: 70.3%). The Mn and ÐM of the polymer were determined to be 68,600 and 1.19, respectively, by GPC using DMF as the eluent (Table 1). The conversion rates of DEA and BMA were calculated to be ~66 mol% and ~63 mol%, respectively, based on 1H NMR analysis of the crude product of P(OEGMA-co-BSMA)-g-P(DEA-co-BMA). The DP of P(DEA-co-BMA) block was determined to be ~65 by 1H NMR analysis in CDCl3 (Fig. S2). Thus, the polymer was denoted as P(OEGMA0.60-co-BSMA0.35-co-BIEM0.05)74-g-P(DEA0.61-co-BMA0.39)65 and shortened as P(OEGMA-co-BSMA)-g-P(DEA-co-BMA) in the subsequent sections.
Synthesis of P(OEGMA-co-BUF)-g-P(DEA-co-BMA). P(OEGMA-co-BSMA)-g-P(DEA-co- BMA) (709 mg, 259 mol -COOH moieties), DCC (53 mg, 259 mol), and DMAP (3 mg) were dissolved in anhydrous CH2Cl2 (15 mL) and was cooled to 0 °C in an ice-water bath. BUF (70 mg, 181 mol) in anhydrous CH2Cl2 (5 mL) was then added dropwise over 10 min. The reaction mixture was stirred at 0 °C for 2 h and then for 48 h at room temperature. After filtration, the mixture was purified by dialysis (cellulose membrane; MWCO, 3,500 Da) against water for 8 h to afford P(OEGMA-co-BUF)-g-P(DEA-co-BMA) (500 mg, yield: 64.2%). The Mn and ÐM of P(OEGMA-co- BUF)-g-P(DEA-co-BMA) were determined to be 71,600 and 1.20, respectively (Table 1). The BUF content in the polymer was determined to be ~7.8wt% by UV-vis spectroscopy in ethanol using BUF as the calibration standard. Thus, the polymer was denoted as P(OEGMA0.60-co-BUF0.21-co- BSMA0.14-co-BIEM0.05)74-g-P(DEA-co-BMA)65 and was shortened as P(OEGMA-co-BUF)-g- P(DEA-co-BMA) in the subsequent sections.
Synthesis of P(OEGMA-co-BUF-co-RGD)-g-P(DEA-co-BMA). P(OEGMA-co-BUF)-g-P(DEA- co-BMA) (383 mg, 52 mol COOH moieties), DCC (3 mg, 15 mol), and DMAP (0.5 mg) were dissolved in anhydrous CH2Cl2 (15 mL) and were cooled to 0 °C in an ice-water bath. NHS (1.7 mg, 15 mol) in anhydrous CH2Cl2 (5 mL) was then added. The reaction mixture was stirred at 0 °C for 2 h and then for 48 h at room temperature. The reaction mixture was evaporated to dryness, and the crude product was used for the next reaction step directly. The obtained crude product was dissolved in DMF (5 mL). RGD (5 mg, 8 mol) and TEA (2.3 L, 16 mol) were then added, and the reaction mixture was stirred at ambient temperature overnight. The mixture was then purified by dialysis (cellulose membrane; MWCO: 3,500 Da) against water for 10 h to afford P(OEGMA-co-BUF-co- RGD)-g-(DEA-co-BMA) (304 mg, yield: 78.6%). GPC characterization revealed an Mn of 72,500 and ÐM of 1.23 (Table 1). HPLC analysis (LC-20AD Series, Shimadzu Corporation, Kyoto, Japan) results indicated ~1.48 RGD moieties per polymer chain (Fig. S4). Thus, the polymer was denoted as P(OEGMA0.60-co-BUF0.21-co-RGD0.02-co-BSMA0.12-co-BEMA0.05)74-g-P(DEA-co-BMA)65 and shortened as P(OEGMA-co-BUF-co-RGD)-g-P(DEA-co-BMA) in the subsequent sections.

2.3 Preparation of micelles

Micelles assembled from polymeric prodrug were prepared via the co-solvent approach. In a typical example, P(OEGMA-co-BSMA)-g-P(DEA-co-BMA) (10 mg, shortened as NP in the subsequent sections) was dissolved in DMF (1 mL). The polymer solution was added into 9 mL deionized water under vigorous stirring. Next, the solution was dialyzed against deionized water to remove organic solvent. Finally, the colloidal dispersion was diluted to the desired concentrations for further experiments. Using similar procedures, nanoparticles assembled from P(OEGMA-co-BUF)- g-P(DEA-co-BMA) (BUF-NP) and P(OEGMA-co-BUF-co-RGD)-g-P(DEA-co-BMA) (BUF-NP- RGD) were fabricated.

2.4 Determination of the hydrodynamic diameter (Dh) and zeta potential ()

Malvern Zetasizer Nano ZS (Malvern Instruments, Malvern, UK) was used to characterize the  and hydrodynamic dimensions of micellar nanoparticles with 632 nm set at a scattering angle of 173°. The solution was first sonicated for 30 s and then the measurements were conducted in disposable sizing cuvettes or zeta-potential measurement cells. Each measurement was performed in triplicate. The micelles were characterized in PBS buffer (10 mM, pH 7.4) at a concentration of 0.5 g/L.

2.5 In vitro drug release measurements

An aqueous solution of BUF-NP-RGD (1.0 g/L, 100 μL) was transferred to a dialysis cell with molecular weight cuto of 2.0 kDa and then was dialyzed against 3.4 mL of PBS bu er (pH 7.4 or 5.0) in the absence and presence of 10 U of esterase at 37 °C. The BUF concentration in the dialysate was quantified by HPLC against a standard calibration curve.

2.6 Cell culture

Human colorectal cancer cell line HCT116 cells were purchased from the Cell Bank of Type Culture Collection of Chinese Academy of Sciences (Shanghai, China) and were cultured in high-glucose DMEM supplemented with 10% fetal bovine serum (FBS), penicillin (100 U/mL) and streptomycin (100 g/mL). Cells were incubated in a humidified atmosphere of 95% air and 5% CO2 at 37 °C.

2.7 In vitro cytotoxicity evaluation

The CCK-8 assay was used to detect cell viability. Briefly, cells were seeded in a 96-well plate at an initial density of ca. 10,000 cells/well in 200 L of complete medium. After incubating for 24 h, the old medium was replaced with fresh one, and the cells were treated with NP, free BUF, BUF-NP, and BUF-NP-RGD at varying concentrations. After incubation for 24 h and 48 h further, 10 µL of CCK-8 assay agents were added into the culture medium, and the cells were incubated for 1 h. The absorbance was measured at 450 nm using a microplate reader (Thermo Fisher Scientific). Each experimental condition was performed in quadruple, and the data were shown as mean values ± standard deviation.

2.8 Hemolysis experiment

The hemolysis experiment of P(OEGMA-co-BSMA)-g-P(DEA-co-BMA) was carried out as described in the literature [22]. Briefly, P(OEGMA-co-BSMA)-g-P(DEA-co-BMA) at different concentrations were incubated for 1 h at 37 °C in the presence of human red blood cells (RBC) in PBS (100 mM supplemented with 150 mM NaCl) of varying pH (7.4, 7.0, 6.0, 5.0 and 4.0) mimicking the acidifying pH gradient of endosomes at di erent phases. PBS and Triton X-100 (1% v/v) were employed as negative and positive hemolysis controls, respectively. After centrifugation (1000 rpm for 10 min) the supernatant absorbance at 405 nm was recorded using a microplate reader (Thermo Fisher). The hemolytic activity was defined as H = (A − A0)/(ATX − A0) where A, A0 and ATX denote the absorbance reading of the sample well, negative control, and positive control, respectively.

2.9 Establishment of a tumor model

Subcutaneous tumors were established in nude mice (BALB/c, female, 4-6 weeks old) by injecting HCT116 cells (5 × 106 cells in 0.2 mL) into their left armpits. All experiments were carried out according to the Guidelines of the Laboratory Protocol of Animal Handling (affiliated Putuo Hospital of Shanghai University of Traditional Chinese Medicine) and were approved by the Institutional Animal Care and Use Committee of Shanghai University of Traditional Chinese Medicine. Tumor sizes were measured using a caliper and the tumor-bearing mice were randomized and classified into five groups (n = 6): Saline (blank), free BUF, NP, BUF-NP, and BUF-NP-RGD.

2.10 In vivo study of the therapeutic efficacy in mice

The mice were classified into five groups based on the solutions they were administered: saline group (13.5 mL/kg), BUF group (1 mg BUF/kg), NP group (24.5 mg/kg), BUF-NP group (1 mg BUF/kg), and BUF-NP-RGD group (1 mg BUF/kg). The samples were intravenously injected through the vena caudalis of mice every two days. On day 12, the animals were sacrificed, and the tumors were collected for further examination.

2.11 Histological examination of tumor tissues

The removed xenograft tumors were fixed with 4% neutral buffered paraformaldehyde and embedded in paraffin before making tissue sections. Four micrometer tissue sections were prepared and stained with H&E for histological examination.

2.12 TUNEL assay

Slides were deparaffinized through xylene and a graded alcohol series and prefixed with 4% formaldehyde. Apoptosis in situ was then detected using the TUNEL assay and the apoptosis detection kit according to the manufacturer’s instructions. The slides were then rinsed with PBS and mounted to a cover slip using Vectashield. All stained slides were evaluated and digital images were acquired using an Eclipse Ti-U inverted microscope (Nikon Corp., Tokyo, Japan) at 20× magnification. Five randomly selected microscopic fields were quantitatively analyzed by Image J.

2.13 Immunohistochemistry detection

After deparaffinization and dehydration, microwave antigen retrieval was performed for 5 min before peroxidase quenching with 3% H2O2 in PBS for 15 min. Subsequently, the sections were blocked with 5% BSA for 30 min and then were incubated, respectively, with the primary antibodies (anti-MVD and anti-Ki-67 antibody) overnight at 4 °C at a dilution ratio of 1:100 in PBS. After washing with PBS, the sections were incubated with biotinylated secondary antibodies for 30 min and then were stained with 3,3′-diamino-benzidine for 2-5 min. The slides were counterstained with hematoxylin for 2-3 min, mounted, and examined.

2.14 In vivo study of the circulation of prodrugs in rats

Ten male Sprague-Dawley rats (250 ± 10 g) purchased from Shanghai SLAC Laboratory Animal Co. Ltd. (Shanghai, China) were kept in an environmentally controlled breeding room (temperature: 22 ± 2 °C, humidity: 60 ± 5%, 12 h dark-light cycle) with free access to standard laboratory food and water for 5 days before starting the experiment. They were fasted overnight before dosing. All experiments were carried out according to the Guidelines of the Laboratory Protocol of Animal Handling (affiliated Putuo Hospital of Shanghai University of Traditional Chinese Medicine) and were approved by the Institutional Animal Care and Use Committee of Shanghai University of Traditional Chinese Medicine.
Ten rats were randomly separated into two groups. Group 1 received free BUF dissolved in normal saline containing 3.75% ethanol (BUF dosage: 0.6 mg/kg) intravenously via the tail vein and group 2 received BUF-NP-RGD dissolved in normal saline containing 3.75% ethanol (BUF dosage: 0.6 mg/kg) intravenously via the tail vein. Animals had free access to water and food 4 h after drug administration. The blood samples (0.5 mL) were collected into heparinized tubes at 30, 45, 60, and 90 min after intravenous injection via the retinal vein plexus. The plasma was separated from heparinized blood by centrifugation at 3,500 rpm for 15 min and was stored at -20 °C prior to analysis.
To determine the concentration of total BUF in each plasma sample for BUF-NP-RGD, 200 μL of plasma was mixed with 10 μL of NaOH (1.0 M) and stored at room temperature for 1 h, leading to the release of free BUF from the polymer. Next, 15 μL of HCl (1.0 M) was added. The proteins were precipitated by the addition of methanol (40 μL) to the plasma samples and mixed by vortexing for 1 min, and then 500 μL of ethyl acetate/petroleum ether (1:1, v/v) was added and mixed by vortexing for 5 min before centrifuging at 4,000 rpm (10 min) twice. The supernatant was transferred to another tube, and then the organic phase was evaporated by bubbling N2. The residue was dissolved in 50 μL of acetonitrile, and the solution was centrifuged at 12,000 rpm for 10 min. Next, 2 μL of the clear supernatant was injected into the HPLC system and the concentrations of BUF in all samples were determined based on the peak area by reference to a calibration curve. Additionally, the BUF concentration was calculated using free BUF as the calibration standard with an absorption detection wavelength at 296 nm. Chromatographic separation was performed using a Waters C18 column (2.1 × 50 mm, 1.7 μm particle size) with the mobile phase composed of water-acetonitrile (65:35, v/v) at a flow rate of 0.5 mL/min. The column temperature was 30 °C. The injection volume was 2 μL and the analysis time was 3 min per sample.

3. Results and Discussion

The main purpose of this study was to develop a novel type of polymeric prodrug of BUF to improve its anticancer performance. The typical colon cancer cell line HCT116 was employed to explore its cellular uptake and anticancer performance both in vitro and in vivo.

3.1 Synthesis and characterization of polymeric prodrug of BUF

The synthetic routes employed for the fabrication of the multifunctional polymeric prodrug, P(OEGMA-co-BUF-co-RGD)-g-P(DEA-co-BMA), covalently linked with BUF, RGD, and the endosome escaping polymer, P(DEA-co-BMA), are shown in Scheme 2. Random copolymers, P(OEGMA-co-BSTMA-co-BIEM), were fabricated via RAFT polymerization in the presence of OEGMA, BSTMA, and BIEM. The Mn and ÐM of the obtained polymer were determined to be 31,000 and 1.03, respectively, by GPC characterization (Table 1). As shown in Fig. S1, 1H NMR analysis revealed the DP to be ~74, and the contents of OEGMA, BSTMA, and BEMA in the polymer were ~60 mol%, ~35 mol%, and ~5 mol%, respectively. Typical peaks ascribed to the RAFT initiator moieties (benzyl methylene, peak b), OEGMA (peaks g and h), BIEM (peak k), and BSTMA (peak o) were observed. P(OEGMA-co-BSMA)-g-P(DEA- co-BMA) was obtained by the ATRP of DEA and BMA followed by hydrolysis with TFA. GPC characterization showed the Mn and ÐM to be 68,600 and 1.19, respectively (Table 1). The dramatically improved Mn indicated successful polymerization. Typical peaks of the DEA moieties (peaks o and p) of the obtained polymer in 1H NMR spectrum confirmed this polymerization (Fig. S2). The conversion rates of DEA and BMA were calculated to be ~66 mol% and ~63 mol%, respectively, based on 1H NMR analysis of the crude product of P(OEGMA-co-BSMA)-g- P(DEA-co-BMA). The DP of P(DEA-co-BMA) block was determined to be ~65 by 1H NMR analysis. Next, the esterification reaction of BUF was performed, affording P(OEGMA-co-BUF)-g- P(DEA-co-BMA). Typical peaks of BUF (peaks w, x, y) were observed in the 1H NMR spectrum (Fig. S3). The BUF content in the polymer was determined to be ~7.8wt% by UV-vis spectroscopy in ethanol using BUF as the calibration standard. GPC examination demonstrated the Mn and ÐM to be 71,600 and 1.20, respectively (Table 1). The tumor-targeting polymeric prodrug, P(OEGMA-co- BUF-co-RGD)-g-P(DEA-co-BMA), was then obtained via the reaction of P(OEGMA-co-BUF)-g- P(DEA-co-BMA) with RGD. GPC analysis revealed an Mn of 72,500 and ÐM of 1.20 (Table 1). The typical peak of RGD (peak u) was observed in the 1H NMR spectrum, demonstrating the successful attachment of RGD and BUF molecules onto the polymer (Fig. S4).
In aqueous media at pH 7.4, P(OEGMA-co-BSMA)-g-P(DEA-co-BMA) tended to assemble into micellar nanoparticles (NP) with Dh 165.8 ± 4.3 nm and 2/Γ2 0.13 (Table 2). The  was determined to be 7.7 ± 0.4 mV. On the other hand, by introducing hydrophobic BUF onto the polymer backbone, nanoparticles (BUF-NP) with Dh 153.4 ± 0.4 nm, 2/Γ2 0.14, and -9.2 ± 0.3 mV were obtained. Further modification of the nanoparticles with RGD (BUF-NP-RGD) showed little change and the Dh, 2/Γ2, and were determined to be 148.4 ± 0.7 nm, 0.22, and-7.6 ± 0.4 mV, respectively.

3.2 In vitro drug release

As shown in Fig. 1a, only ~12% BUF was released over 24 h under simulated physiological conditions, confirming the reported result that -thioester bonds are relatively stable under neutral conditions [15]. The relatively stable -thioester bond can effectively circumvent drug burst release during the circulation period. However, dramatically improved BUF release was observed when esterase was added. As shown in Fig. 1b, as much as ~51% BUF was released in the first 9 h. Moreover, the release rate was almost constant. The BUF release rate gradually decreased thereafter and ultimately ~83% BUF was released over 24 h. This endows the nanoparticles with triggered release capability after entering cancer cells, considering the presence of abundant esterase in cancer cells.
In addition to esterase, the effect of the acidic condition on the hydrolysis of the prodrug was also explored. As shown in Fig. 1c, BUF was slowly released from nanoparticles at pH 5.0, and approximately 38% of BUF was released over the period of 24 h. Compared with the neutral condition, a lower pH accelerated the hydrolysis of the prodrug caused by the increased hydrolysis rate of the -thioester bond under acidic conditions. This means that, after the nanoparticles enter the endosomes/lysosomes some of the BUF molecules will be released. By adding esterase into the acidic solution the BUF release rate further increased although the release rate was still lower than that at pH 7.4 with 10 U of esterase addition (Fig. 1d). This might be due to the reduced esterase activity at a lower pH condition. This demonstrated that both acid and esterase play roles in drug release in tumor cells.

3.3 In vitro cytotoxicity

The typical malignant colon cancer cell line HCT116 was employed to investigate the in vitro anticancer performance of the obtained polymeric prodrug. As shown in Fig. 2, when treated with as high as ~100 g/mL of the polymeric carrier, P(OEGMA-co-BSMA)-g-P(DEA-co-BMA), HCT116 cells still showed over 80% viability. By extending the incubation time from 24 h to 48 h, the cell viability showed little change, implying the good biocompatibility of the polymer backbone. However, treatment of HCT116 cells with BUF, BUF-NP or BUF-NP-RGD resulted in significant cell death (Fig. 3a). At a BUF concentration of 20 nM, ~94.5% cell viability was observed when HCT116 cells were treated with free BUF. However, the cell viability was reduced to ~81.5% when the same BUF dosage was used for BUF-NP. Moreover, the cell viability further declined to ~53.4% for BUF-NP-RGD. If the BUF concentration leading to 50% cell death is regarded as the IC50, the IC50 values of BUF, BUF-NP, and BUF-NP-RGD can be determined to be ~318 nM, ~131 nM, and ~26 nM, respectively. By extending the incubation time from 24 h to 48 h, the IC50 values of BUF, BUF-NP, and BUF-NP-RGD further declined to ~133 nM, ~69 nM, and ~10 nM, respectively (Fig. 3b).
It was already proven that copolymers P(DEAEMA-co-BMA) could effectively help the active molecules escape from endosomes/lysosomes after entering cells via endocytosis [22]. In the current work, it was reasonable to assume that the improved anticancer performance of polymeric prodrugs of BUF (BUF-NP and BUF-NP-RGD) was ascribed to the introduction of the P(DEAEMA-co-BMA) block. Additionally, RBC hemolysis experiments at different pH values mimicking the endosomal/lysosomal trafficking pathway were conducted to evaluate the endosome escaping abilities of P(OEGMA-co-BSMA)-g-P(DEA-co-BMA). As shown in Fig. 4, human RBCs incubated with P(OEGMA-co-BSMA)-g-P(DEA-co-BMA) at pH 7.4, corresponding to extracellular pH values showed little hemolysis effect. However, by decreasing the pH to 7.0, which corresponds to the beginning pH of the endocytosis stage, the hemolysis started to increase. Further decreasing the pH to 6.0, 5.0, and 4.0 dramatically increased the hemolysis effect (approximately 90%). This hemolysis transition coincides with the literature results [22]. In addition to the endosome escaping effect, the introduction of the targeting peptide RGD might be another positive factor to further improve the anticancer performance of BUF-NP-RGD compared with that of its non-targeting counterpart, BUF-NP [17].
Cumulative reports have shown that BUF caused cell growth inhibition via the apoptosis pathway [15]. In the current work, flow cytometry was employed to explore whether the polymeric prodrug of BUF induced growth inhibition because of apoptosis. As shown in Fig. S5, the apoptotic population in the untreated cells was ~15.2%. At 10 nM BUF, BUF-NP-RGD exhibited ~36.6% cell apoptosis, much higher than that of free BUF (~18%) and BUF-NP (~17.7%). This result agreed with the cell cytotoxicity assay result. By increasing the BUF concentration from 10 nM to 160 nM, the apoptotic population increased to ~20.3% and finally to ~30.6% for free BUF treated cells. BUF-NP led to a comparable apoptotic population, and there was no significance difference between the samples with free BUF and BUF-NP in the concentration range of 10 nM to 160 nM. However, BUF-NP-RGD caused a significantly higher apoptotic population than both free BUF and BUF-NP (P < 0.05). When the BUF concentration was increased from 10 nM to 160 nM, the apoptotic population of BUF-NP-RGD treated cells increased to ~53.1% and finally to ~67.9%. This might explain the enhanced anticancer performance of BUF-NP-RGD. 3.4 In vivo anticancer activity in human colon xenograft tumors BUF-NP-RGD showed significantly improved antitumor activity in the tumor-bearing mouse model. As shown in Fig. 5, NP showed no cytotoxic effect in mice, exhibiting a comparable tumor growth rate with that of the saline group. The almost unchanged body weight among different groups implied the low cytotoxicity of BUF and its polymeric prodrugs. However, when treated with BUF the tumor growth was effectively inhibited as vividly revealed by the decreased tumor volume growth. Moreover, when the same dose of BUF was administered to the mice, the tumor inhibiting effect was much more obvious in the BUF-NP and BUF-NP-RGD groups than in the BUF group. The tumor size was significantly smaller in the BUF-NP-RGD group than in the other groups after 12 days of treatment. It was revealed that the anticancer efficacy of BUF was effectively improved via polymer-BUF conjugation. Additionally, the therapeutic efficacy can be further improved through RGD introduction. It was repeatedly documented that polymeric modification of small-molecule drugs often changes its pharmacokinetic characteristics. Thus, the pharmacokinetic profile of BUF-NP-RGD was then conducted. It was shown that BUF-NP-RGD exhibited dramatic increase in the concentrations of BUF in plasma compared with that of free BUF (Fig. S6). The significantly improved circulation performance dramatically improves its effective utilization in vivo. In combination with the tumor-targeting role of RGD this might partially explain the improved anticancer performance of BUF-NP-RGD in vivo. 3.5 Histological analysis of xenograft tumors HE staining, MVD and Ki67 analysis were conducted to examine the efficacy of BUF-NP-RGD. As shown in Fig. 6, the tumor tissues treated with saline and NP showed almost no necrosis. However, free BUF and BUF-NP led to relatively large necrotic areas of tumor tissues. Additionally, the BUF-NP-RGD treated group exhibited much larger necrotic areas than the BUF and BUF-NP groups. Furthermore, H&E staining results indicated no significant change in normal organs (Fig. S7). Cell proliferation-related protein Ki-67 was then examined via an immunohistochemistry technique. As shown in Fig. 6, Ki-67 was positive in tumor tissues with strong staining in the groups treated with saline and NP groups in the cell nucleus but was inhibited significantly in groups treated with free BUF, BUF-NP, and BUF-NP-RGD. This further confirmed the anticancer activity of the polymeric prodrugs. Finally, the expression of MVD, an indicator of blood vessel growth, in groups treated with saline and NP was observed. The decreased expression of MVD in groups treated with BUF, BUF-NP, and BUF-NP-RGD implied that the growth of new blood vessels was effectively inhibited by BUF and BUF-containing nanoparticles. By carefully comparing the MVD staining of groups treated with free BUF, BUF-NP, and BUF-NP-RGD, we can observe that BUF-NP-RGD exhibited much more effective therapeutic efficiency than that of free BUF and BUF-NP, which is in accordance with the results of in vivo tumor inhibition experiments. To further investigate the cellular damage caused by drugs tumor cell apoptosis was quantified via the TUNEL assay after drug treatment (Fig. 7). The TUNEL assay by labeling fragmented DNA is a widely accepted method to detect apoptotic cells. Saline and NP caused little cell apoptosis with TUNEL positive cells merely approximately 5%. However, BUF and BUF-NP led to significant cell apoptosis with TUNEL-positive cell rates of ~35.4% and ~43.3%, respectively. Although BUF-NP led to more TUNEL-positive cells, there was no significant difference between BUF and BUF-NP with P > 0.05. Furthermore, the BUF-NP-RGD treated group showed as high as ~93.1% TUNEL-positive cells, much higher than that of both free BUF and its non-targeting counterpart, BUF-NP. It was revealed that the P values between both BUF-NP-RGD/BUF-NP and BUF-NP- RGD/free BUF were smaller than 0.05, indicating that the introduction of RGD onto the polymer effectively enhanced this anticancer performance. This finding correlates with the tumor inhibition data shown in Fig. 6.

4. Conclusions

In summary, a novel type of polymeric BUF prodrug (BUF-NP-RGD) with endosome-escaping and tumor-targeting capabilities was successfully developed and used to treat colon cancer in mice. It was demonstrated that BUF-NP-RGD can more effectively enhance cell death in vitro and reduce tumor growth in vivo than free BUF. Histological and immunochemical analysis demonstrated that BUF-NP-RGD yielded improved cell apoptosis, angiogenesis inhibition, and anti-proliferation effect compared with free BUF. The reported BUF-NP-RGD with synergistically integrated cancer targeting, controlled release, and endosome-escaping abilities shows promise for clinical applications in nanomedicine systems.

References

[1] Y.T. Jiang, Y. Zhang, J.L. Luan, H.Y. Duan, F. Zhang, K. Yagasaki, G.Y. Zhang, Effects of bufalin on the proliferation of human lung cancer cells and its molecular mechanisms of action, Cytotechnology, 62 (2010) 573-583.
[2] C.M. Xie, W.Y. Chan, S. Yu, J. Zhao, C.H.K. Cheng, Bufalin induces autophagy-mediated cell death in human colon cancer cells through reactive oxygen species generation and JNK activation, Free Radic. Biol. Med., 51 (2011) 1365-1375.
[3] X.B. Xie, J.Q. Yin, L.L. Wen, Z.H. Gao, C.Y. Zou, J. Wang, G. Huang, Q.L. Tang, C. Colombo, W.L. He, Q. Jia, J.N. Shen, Critical role of heat shock protein 27 in bufalin-induced apoptosis in human osteosarcomas: A proteomic-based research, Plos One, 7 (2012) e47375.
[4] P.H. Yin, X. Liu, Y.Y. Qiu, J.F. Cai, J.M. Qin, H.R. Zhu, Q. Li, Anti-tumor activity and apoptosis-regulation mechanisms of bufalin in various cancers: New hope for cancer patients, Asian Pac. J. Cancer Prev., 13 (2012) 5339-5343.
[5] D.Z. Qiu, Z.J. Zhang, W.Z. Wu, Y.K. Yang, Bufalin, a component in Chansu, inhibits proliferation and invasion of hepatocellular carcinoma cells, BMC Complement Altern. Med., 13 (2013) 185-195.
[6] M.Y. Li, X.J. Yu, H. Guo, L.M. Sun, A.J. Wang, Q.J. Liu, X.W. Wang, J.S. Li, Bufalin exerts antitumor effects by inducing cell cycle arrest and triggering apoptosis in pancreatic cancer cells, Tumor Biol., 35 (2014) 2461-2471.
[7] Y.Y. Qiu, Q. Hu, Q.F. Tang, W. Feng, S.J. Hu, B. Liang, W. Peng, P.H. Yin, MicroRNA-497 and bufalin act synergistically to inhibit colorectal cancer metastasis, Tumor Biol., 35 (2014) 2599-2606.
[8] A. Zou, X. Zhao, U.A. Handge, V.M. Garamus, R. Willumeit-Romer, P. Yin, Folate receptor targeted bufalin/beta-cyclodextrin supramolecular inclusion complex for enhanced solubility and anti-tumor efficiency of bufalin, Mater. Sci. Eng. C, 78 (2017) 609-618.
[9] Z. Yang, Y. Teng, H. Wang, H. Hou, Enhancement of skin permeation of bufalin by limonene via reservoir type transdermal patch: Formulation design and biopharmaceutical evaluation, Int. J. Pharm. 447 (2013) 231-240.
[10] Q. Hu, B. Liang, Y. Sun, X.L. Guo, Y.J. Bao, D.H. Xie, M. Zhou, Y.R. Duan, P.H. Yin, Z.H. Peng, Preparation of bufalin-loaded pluronic polyetherimide nanoparticles, cellular uptake, distribution, and effect on colorectal cancer, Int. J. Nanomed., 9 (2014) 4035-4041.
[11] Z.P. Yuan, L.J. Chen, L.Y. Fan, M.H. Tang, G.L. Yang, H.S. Yang, X.B. Du, G.Q. Wang, W.X. Yao, Q.M. Zhao, B. Ye, R. Wang, P. Diao, W. Zhang, H.B. Wu, X. Zhao, Y.Q. Wei, Liposomal quercetin efficiently suppresses growth of solid tumors in murine models, Clin. Cancer Res. 12 (2006) 3193-3199.
[12] T.H. Wu, F.L. Yen, L.T. Lin, T.R. Tsai, C.C. Lin, T.M. Cham, Preparation, physicochemical characterization, and antioxidant effects of quercetin nanoparticles, Int. J. Pharm. 346 (2008) 160-168.
[13] Y. Liu, P.F. Wang, C. Sun, J.H. Zhao, Y. Du, F. Shi, N.P. Feng, Bioadhesion and enhanced bioavailability by wheat germ agglutinin-grafted lipid nanoparticles for oral delivery of poorly water-soluble drug bufalin, Int. J. Pharm. 419 (2011) 260-265.
[14] X. Tian, H.Z. Yin, S.C. Zhang, Y. Luo, K. Xu, P. Ma, C.G. Sui, F.D. Meng, Y.P. Liu, Y.H. Jiang, J. Fang, Bufalin loaded biotinylated chitosan nanoparticles: An efficient drug delivery system for targeted chemotherapy against breast carcinoma, Eur. J. Pharm. Biopharm., 87 (2014) 445-453.
[15] T. Liu, X. Yuan, T. Jia, C. Liu, Z. Ni, Z. Qin, Y. Yuan, Polymeric prodrug of bufalin for increasing solubility and stability: Synthesis and anticancer study in vitro and in vivo, Int. J. Pharm., 506 (2016) 382-393.
[16] J. Zhang, Y. Sun, J.H. Liu, B.Y. Yu, Q. Xu, Microbial transformation of three bufadienolides by Nocardia sp. and some insight for the cytotoxic structure-activity relationship (SAR), Bioorg. Med. Chem. Lett., 17 (2007) 6062-6065.
[17] G. Pan, Y. Bao, J. Xu, T. Liu, C. Liu, Y. Qiu, X. Shi, H. Yu, T. Jia, X. Yuan, Z. Yuan, P. Yin, Y. Cao, Esterase-responsive polymeric prodrug-based tumor targeting nanoparticles for improved antitumor performance against colon cancer, RSC Adv., 6 (2016) 42109-42119.
[18] L.K. Bogart, G. Pourroy, C.J. Murphy, V. Puntes, T. Pellegrino, D. Rosenblum, D. Peer, R. Levy, Nanoparticles for imaging, sensing, and therapeutic intervention, ACS Nano, 8 (2014) 3107-3122.
[19] L.K. Medina-Kauwe, J. Xie, S. Hamm-Alvarez, Intracellular trafficking of nonviral vectors, Gene Ther., 12 (2005) 1734-1751.
[20] J. Ren, J.C. Sharpe, R.J. Collier, E. London, Membrane translocation of charged residues at the tips of hydrophobic helices in the T domain of diphtheria toxin, Biochemistry, 38 (1999) 976-984.
[21] D.M. Copolovici, K. Langel, E. Eriste, U. Langel, Cell-penetrating peptides: Design, synthesis, and applications,