Research Article www.acsami.org
Surface Engineered Protein Nanoparticles With Hyaluronic Acid Based Multilayers For Targeted Delivery Of Anticancer Agents Sreeranjini Pulakkat,† Sai A Balaji,‡,§ Annapoorni Rangarajan,‡ and Ashok M. Raichur*,†,∥ †
Department of Materials Engineering, Indian Institute of Science, Bangalore, 560012, India Molecular Reproduction, Development and Genetics, Indian Institute of Science, Bangalore, 560012, India § Department of Pharmacology, Manipal College of Pharmaceutical Sciences, Manipal University, Manipal, 576104, India ∥ Nanotechnology and Water Sustainability Research Unit, University of South Africa, The Science Campus, Florida Park, 1710 Roodepoort, Johannesburg, South Africa ‡
S Supporting Information *
ABSTRACT: Layer-by-layer (LbL) technique was employed to modify the surface of doxorubicin (Dox)-loaded bovine serum albumin (BSA) nanoparticles using hyaluronic acid (HA) to enable targeted delivery to overexpressed CD44 receptors in metastatic breast cancer cells. LbL technique offers a versatile approach to modify the surface of colloidal nanoparticles without any covalent modification. Dox-loaded BSA (Dox Ab) nanoparticles optimized for their size, zeta potential, and drug encapsulation efficiency were prepared by modified desolvation technique. The cellular uptake and cytotoxicity of the LbL coated Dox Ab nanoparticles were analyzed in CD44 overexpressing breast cancer cell line MDA-MB-231. Nanoparticles with HA as the final layer (Dox Ab HA) showed maximum cellular uptake in MDA-MB-231 cells owing to the CD44 receptor-mediated endocytosis and hence, exhibited more cytotoxicity as compared to free Dox. Further, luciferase-transfected MDA-MB-231 cells were used to induce tumor in BALB/c female nude mice to enable whole body tumor imaging. The mice were imaged before and after Dox treatment to visualize the tumor growth. The in vivo biodistribution of Dox Ab HA nanoparticles in nude mice showed maximum accumulation in tumor, and importantly, better tumor reduction in comparison with free Dox, thus paving the way for improved drug delivery into tumors. KEYWORDS: LbL technique, hyaluronic acid, Cd44, in vivo targeted delivery, albumin nanoparticles
■
charged entities to form multilayer architectures with nanometer precision.5 The driving forces involved in multilayer assembly are primarily electrostatic in nature but can also involve hydrogen bonding,6 hydrophobic interactions,7 van der Waals forces,8
INTRODUCTION The modular layer-by-layer (LbL) technique offers a versatile approach to functionalize the surface of charged drug carriers independent of their internal design. Such a strategy can facilitate targeted or enhanced cellular uptake, stimuli-responsive release, and protection from opsonization, as well as early degradation during carrier transport.1−4 The LbL technique involves the deposition of oppositely charged polyelectrolytes or other © XXXX American Chemical Society
Received: April 8, 2016 Accepted: August 25, 2016
A
DOI: 10.1021/acsami.6b04179 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Research Article
ACS Applied Materials & Interfaces covalent bonding, or click chemistry modifications.9,10 The LbL assembly has been utilized to manipulate the surface characteristics of drug-loaded polymeric nanoparticles,11 liposomes,12 mesoporous silica nanoparticles (MSN),13 quantum dots,14 and carbon nanotubes.15 In most of these systems, LbL modification enables stimuli-responsive gating or controlling the release of pre-encapsulated molecules by minimizing the initial burstrelease, maintaining a sustained release, or extending the duration of release. For instance, Luo et al. characterized the effects of LbL modification on submicrometer-sized poly(lactide-co-glycolide) (PLGA) particles and demonstrated that poly(allylamine hydrochloride) (PAH)/poly(styrenesulfonate) (PSS) or poly(L-lysine hydrobromide) (PLL)/dextran sulfate (DS) coatings could suppress the burst release of bioactive molecules as well as extend release over 4 days under physiological conditions in vitro.16 Similarly, Haider et al. demonstrated that the drug encapsulation efficiency of cationic liposomes is enhanced with the deposition of polyelectrolyte multilayers of natural polymers: alginate and chitosan.17 Another example is that of doxorubicin (Dox)-loaded liposomes modified with PLL/poly(L-aspartic acid)-b-polyethylene glycol (PEG), which exhibit a pH dependent release profile.18 At slightly acidic pH, the protonation of aspartic acid’s carboxylic acid side chain initiates film breakdown and release of cargo, whereas at pH 7.4, the films are intact and significantly improve drug retention relative to bare liposomes.12 LbL-coated MSNs have also been designed and crafted to release encapsulated anticancer drugs by changing the pH or by adding competitive agents. Bis-aminated poly(glycerol methacrylate) and cucurbit[7]uril-coated LbL-MSN hybrids possessing excellent biostability, negligible premature drug leakage at pH 7.4, and exceptional stimuli-responsive drug release performance were reported by Li et al.13 The LbL modification also facilitates the deposition of highly negatively charged polyelectrolytes as the outer layer of drug carriers designed for systemic delivery and greatly improves their steric stability and circulatory half-life. This could also result in decreased serum protein adsorption and opsonization and hence, reduced clearance by the mononuclear phagocyte system. Polyanionic biopolymers, such as hyaluronic acid (HA) and alginate, have been employed in LbL architectures to attain significantly longer residence times for drug carriers as compared to that of their unmodified counterparts. For example, alginate and HA-terminated PLGA formulations developed by Morton and colleagues demonstrated a 3-fold increase in drug retention and sustained delivery of the model drug indocyanine green dye in vivo as opposed to bare PLGA cores.19 Increased circulation half-life also guarantees passive targeting and better accumulation of the drug carriers in solid tumors due to enhanced permeability and retention effect (EPR).20 Further, active targeting of the LbL-modified drug carriers can be facilitated by careful selection of the outer surface layer that allows for the attachment of specific ligands or targeting moieties. For instance, LbL modification using PSS/PAH polyelectrolytes and further attachment of aptamers imparted suspension stability and targeting ability to bovine serum albumin (BSA) nanoparticles, which were then loaded with Dox.21 LbL in combination with covalent chemistry has been used to attach PEG, anti-TNF-α, and folic acid to PLGA nanoparticles to impart tailored recognition, enhanced cellular uptake, protection, and release of encapsulated cargo.22 In similar lines, the possibility of targeted delivery of drugloaded BSA nanoparticles to overexpressed CD44 receptors in cancer cells, by modifying the surface of the nanoparticles via LbL
deposition of HA, has been investigated in this work. HA is a biodegradable and biocompatible, linear, negatively charged polysaccharide having alternating units of N-acetyl D-glucosamine and D-glucuronic acid. Interaction of HA with receptors such as CD44 and receptor for hyaluronic acid-mediated motility is responsible for various functions within the extracellular matrix, such as cell growth, differentiation, and migration.23−26 Different types of HA-based drug carriers have been reported to target CD44 receptors that are overexpressed in various tumors, for example, ovarian, breast, colon, stomach, and acute leukemia. HA has multiple functional groups like the carboxylate on the glucuronic acid, the N-acetylglucosamine hydroxyl, and the reducing end, which can be successfully utilized in conjugation reactions with drugs to form a nontoxic prodrug. Direct conjugations of a low molecular weight HA to cytotoxic drugs, such as butyric acid, cisplatin, PTX, and Dox have been reported.27−29 Alternatively, self-assembled nanoparticles and polymersomes of amphiphilic HA derivatives encapsulating various anticancer drugs have also been reported.30−32 Drug-loaded colloidal nanocarriers like liposomes and polymer nanoparticles have been surface-modified with HA to impart targeting characteristics. HA-bound liposomes containing anticancer agents such as Dox and mitomycin C exhibited enhanced targeting ability to the cancer cells and higher therapeutic efficacy compared to free drugs.33−36 However, covalent modification of HA to conjugate drugs or drug carriers have been found to affect the binding sites or result in masking of receptor recognition sites in HA, limiting the targeting efficiency.37 In such a scenario, incorporation of HA, which is also a polyanionic polysaccharide, in a multilayer assembly that does not require any covalent modification can be beneficial in imparting stealth properties and improving the targeting efficacy of CD44-targeted delivery systems. Further, the CD44 targeting ability of HA incorporated in multilayer assemblies has not been explored in detail so far, although a few HA based multilayer films and capsules have been reported earlier.38,39 Dox-loaded BSA nanoparticles optimized for drug encapsulation efficiency were prepared by modified desolvation technique. Biopolyelectrolytes HA and PLL were utilized as polyanion and polycation, respectively. After deposition of 2 bilayers, the LbL deposition was terminated with HA as the final layer. The influence of this surface modification in the cellular uptake and cytotoxicity of the LbL coated BSA nanoparticles were analyzed in CD44 overexpressing breast cancer cell line MDAMB-231. Further, luciferase transfected MDA-MB-231 cells were used to induce tumor in BALB/c nude mice to enable whole body tumor imaging. The mice were imaged before and after the Dox treatment to visualize the tumor growth. The in vivo tumor reduction, as well as biodistribution of LbL-coated BSA nanoparticles, was also studied in comparison with free Dox.
2. MATERIALS AND METHODS 2.1. Materials. HA (Mw = 200 kDa), PLL, BSA (Mw = 66 kDa), doxorubicin hydrochloride (C27H29NO11HCl, Fw = 580 Da), glutaraldehyde, trypsin, phosphate-buffered saline (PBS), Dulbecco’s modified eagle medium (DMEM), Dulbecco’s phosphate buffered saline (DPBS), and fetal calf serum were purchased from Sigma-Aldrich (Bangalore, India). Sodium dodecyl sulfate (SDS), chloroform, isopropyl alcohol, and sodium hydroxide (NaOH) were obtained from Rankem, RFLC Limited (Bangalore, India). Annexin V-FITC apoptosis detection kit was purchased from BioVision Incorporated (USA). MDA-MB-231 cell line was from ATCC. Double autoclaved Milli-Q water (Millipore, Billerica, MA, USA) was used for all the experiments. B
DOI: 10.1021/acsami.6b04179 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Research Article
ACS Applied Materials & Interfaces 2.2. Preparation of Dox-Loaded BSA Nanoparticles. Doxloaded BSA (Dox Ab) nanoparticles were prepared using a modified desolvation method. The concentration of Dox (100−750 μM) and BSA (10−50 mg/mL) was varied to obtain Dox Ab nanoparticles optimized for their stability and loading efficiency. Briefly, different amounts of 10 mM Dox solution were added to 2 mL aqueous solution of BSA at pH 7 and stirred for 30 min. Then ethanol was added dropwise to make the final volume of the reactant solution to 10 mL. At that point, a turbid solution was obtained to which 25 μL of 5% glutaraldehyde solution was further added. The nanoparticle suspension was then stirred overnight to yield cross-linked Dox Ab nanoparticles. The nanoparticles were then washed first with ethanol and then thrice with water by centrifuging at 14,000 rpm for 5 min, redispersed in Milli-Q water, and stored until further use. Bare BSA nanoparticles (Ab Np) were prepared in a similar process except that Dox was not added to aqueous solution of BSA. 2.3. Polyelectrolyte Multilayer Modification of Dox Ab Nanoparticles. Dox Ab nanoparticles were first incubated with PLL (0.5 mg/mL in 0.15 M NaCl, pH 6) for 15 min. The sample was then centrifuged at 3000 rpm for 5 min and washed thrice with 0.15 M NaCl to remove the unadsorbed polyelectrolyte. Subsequently, the sample was incubated with HA (0.5 mg/mL in 0.15 M NaCl, pH 6) for 15 min and the same procedure for washing was followed. After depositing 2 bilayers, the Dox Ab nanoparticles with HA as the final layer (Dox Ab HA) was redispersed in Milli-Q water. To obtain Dox Ab nanoparticles with PLL as the final layer (Dox Ab PLL), 3 layers of polyelectrolytes were deposited, washed, and redispersed in Milli-Q water. 2.4. Dynamic Light Scattering (DLS) and Zeta Potential Measurements. DLS and zeta potential measurements were done using Zetasizer Nano ZS (Malvern, South borough MA). Zeta potential measurements were taken after every adsorption step to understand and ensure the multilayer growth of polyelectrolytes on the surface of Dox Ab nanoparticles. Three parallel measurements were taken for all samples. The size of Dox Ab and Dox Ab HA nanoparticles were determined by DLS measurements. 2.5. Morphological Characterization. Morphological characteristics of Dox Ab and Dox Ab HA nanoparticles were analyzed using TEM (Tecnai F30 FEI-Eindhoven, Netherlands). For TEM imaging, a drop of various samples was put on a 300 mesh carbon coated copper grid (Toshniwal Bros SR Pvt. Ltd., Bangalore, India) and dried at ambient conditions overnight before analysis. Dox Ab nanoparticles were also analyzed by SEM equipped with field emission gun (FEISIRION, Eindhoven, Netherland) at operating voltages of 10 kV/5 kV and working distance of 6 mm. A drop of sample suspension was put on a silicon wafer and dried under ambient conditions overnight. Since all the samples were nonconductive in nature, they were subjected to gold sputtering (JEOL JFC 1100E Ion sputtering device; JEOL, Tokyo, Japan). Confocal images of Dox Ab nanoparticles were also obtained to confirm the presence of Dox in the nanoparticles. Dox is a highly fluorescent molecule and can be visualized under confocal microscopy using emission filter of 570 nm upon exciting at 488 nm. 2.6. Drug Loading Studies. UV−visible spectroscopy was used to determine the concentration of Dox in the loading and release studies.21 The free Dox remaining in the supernatant obtained after centrifugation of Dox Ab nanoparticles was estimated by measuring the absorbance at 480 nm. The amount of Dox Ab nanoparticles was calculated as the difference in initial BSA amount and the amount of BSA remaining in the supernatant. The encapsulation efficiency (EE) and the loading capacity (LC) of Dox Ab nanoparticles was then calculated using the equations given below:
2.7. Drug Release Studies. To monitor the effect of pH on the drug release profile, Dox Ab and Dox Ab HA nanoparticles were resuspended in 2 mL of pH 4 acetate buffer or in PBS buffer (physiological pH 7.4) at 37 °C. The enzyme-dependent release of Dox from the nanoparticles was similarly obtained in the presence of 250 μg/mL trypsin enzyme in PBS buffer solution. The nanoparticle suspension was then centrifuged at 2000 rpm at stipulated periods (1, 3, 6, 12, 18, 24, 48 h) and the amount of Dox in the supernatant was estimated spectrophotometrically. An equivalent volume of prewarmed buffer solution was then added to compensate for the supernatant taken out for analysis. 2.8. In Vitro Cellular Uptake Studies. Breast cancer cell line MDA-MB-231 (CD44+) was used for in vitro cellular uptake studies. About 1 × 105 cells were grown in 6-well plates with DMEM supplemented with 10% FBS at 37 °C and 5% CO2 supply. All experiments were carried out at 70% confluence and repeated in triplicates for statistical significance. Different concentrations (1 and 2 μM) of free Dox, Dox Ab, Dox Ab PLL, and Dox Ab HA nanoparticles were incubated with MDA-MB-231 cells for 1 h followed by DPBS wash and trypsinization. To study the effect of HA pretreatment and to confirm the internalization of Dox Ab HA nanoparticles by receptor-mediated endocytosis, the cells were first incubated with 1 mg/mL HA solution for 2 h before treating with different samples for 1 h. All the above samples were further centrifuged and the pelleted cells were resuspended in 200 μL of DPBS before subjecting to FACS analysis (BD FACS Canto-II). The cell-associated fluorescence was monitored at an excitation and emission wavelength of 488 and 550 nm, respectively, and the data was analyzed with Summit 5.2 software. The cellular uptake of drug-loaded nanoparticles was also visualized using a fluorescence microscope (Carl Zeiss, Germany) using an oil immersion lens at 63 X magnification. The different samples were treated with 5 × 104 MDA-MB-231 cells for 1 h, washed with DPBS, and fixed with 4% formaldehyde before obtaining images. 2.9. Cytotoxicity Assay. The in vitro cytotoxicity of bare Ab Np and Dox Ab HA nanoparticles in comparison with free Dox was assessed using MTT assay. Eight ×103 MDA-MB-231 cells were seeded in 96-well plates for 12 h followed by addition of different concentrations (0.25−2 μM) of the various samples. After incubating for 48 h at 37 °C, 20 μL of 5 mg/mL of MTT reagent was added to each well and again incubated for 4 h at 37 °C. Later, the media was removed completely and 100 μL of DMSO was added in each well to solubilize formazan. The percentage of cell viability was then determined at 570 nm relative to untreated cells by measuring the absorbance after solubilization of the formazan crystals. 2.10. Apoptosis Assay. MDA-MB-231 cells (1 × 105) were seeded in 12-well plates and treated with 1 μM of different samples for 20 h. The cells were then washed twice with DPBS, trypsinized, centrifuged, and the cell pellet was resuspended in 1× binding buffer provided by the Annexin V-FITC apoptosis detection kit. The cells were then treated with 5 μL of Annexin V-FITC conjugate provided in the kit and incubated in dark for 10 min. The fluorescence of the cells was then immediately determined using flow cytometer (Ex/Em = 488 nm/530 nm). 2.11. In Vivo Biodistribution Studies. The in vivo biodistribution studies were conducted in BALB/c nude mice, bred, and housed at the Central Animal Facility, Indian Institute of Science, Bangalore, India. All animals were maintained as per institutional guidelines and all procedures were conducted in agreement with the institutional ethics committee guidelines. Three ×106 MDA-MB-231 cells were subcutaneously injected in both flanks of 4−5 weeks old athymic nude mice (n = 12). After 4 weeks of injection, the mice were assigned into three groups, one control and two groups injected with either 5 mg/kg of free Dox or Dox Ab HA nanoparticles by intravenous injection in the tail vein. After 24 h, the mice were sacrificed and Dox content in the plasma, heart, liver, lung, kidney, and tumor was analyzed. The whole blood was collected via cardiac puncture, immediately placed in EDTA containing eppendorfs and the plasma was obtained by centrifugation at 1500 rpm for 10 min. The organs were collected, weighed, and homogenized to a 10% homogenate (w/v in water) with a hand-held tissue homogenizer. The Dox content was then quantified by a solvent extraction and fluorescence detection assay reported previously.40 50 μL of plasma or 200 μL of 10% tissue homogenate were adjusted to 0.8 mL with water,
EE = [(C i − Cs)/C i] × 100
LC = [(C i − Cs)/amount of Dox Ab Np] × 100 where Ci = initial amount of Dox and Cs = amount of Dox in the supernatant. The loss of drug from Dox Ab nanoparticles during the polyelectrolyte multilayer modification, although negligible, was accounted for while determining the amount of drug loaded. C
DOI: 10.1021/acsami.6b04179 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX
Research Article
ACS Applied Materials & Interfaces to which 100 μL each of 10% SDS and 10 mM H2SO4 were then added. Subsequently, Dox extraction was achieved by adding 2 mL of chloroform/isopropyl alcohol (1:1, v/v) followed by vigorously vortexing and freezing at −20 °C overnight. After thawing and further vortexing, the samples were centrifuged at 1500 rpm for 10 min and the fluorescence of organic phase (lower phase) collected was determined fluorometrically (λex = 470 nm, λem = 590 nm). The samples were compared with a standard curve made from the fluorescence emission of known amounts of Dox added to acidified isopropanol extracts of homogenized tissue from untreated mice. 2.12. In Vivo Bioluminescence Imaging. To enable in vivo bioluminescence imaging, luciferase-transfected MDA-MB-231 cells were utilized. MDA-MB-231 cells were transfected with a plasmid pLenti CMV LUC Puro using the transfection reagent lipofectamine 2000. After 48 h, the cells were treated with 0.5 μg of puromycin to select transfected cells and make stable cell lines. These cells were subjected to luciferase assay (data not shown) to confirm luciferase expression in the cells. 3 ×106 luciferase-transfected MDA-MB-231 cells were subcutaneously injected in both flanks of BALB/c nude mice (4−5 weeks, n = 18). After 4 weeks of injection, the mice were intraperitoneally injected with luciferin substrate and immediately imaged using a whole body animal imaging system (Xenogen IVIS 200 in vivo Imaging, PerkinElmer, USA) before starting the Dox treatment. The mice were later imaged when the drug treatment was terminated. 2.13. In Vivo Tumor Suppression Efficacy. The in vivo tumor suppression efficiency of Dox Ab HA nanoparticles in comparison with free Dox was studied using 4−5 weeks old nude mice (n = 18) divided into three groups, namely control, free Dox treated, and Dox Ab HA treated groups. Three ×106 luciferase-transfected MDA-MB-231 cells were subcutaneously injected in both flanks of all 3 groups of mice. After 4 weeks of injection, 2 groups of mice were treated with 2 mg/kg of free Dox or Dox-encapsulated nanoparticles every fifth day for a period of 25 days. The tumor size was measured using a digital Vernier caliper every week during the treatment and the tumor volume was calculated using the following formula:
3. RESULTS AND DISCUSSION 3.1. Preparation and Characterization of Dox Ab Nanoparticles. Dox Ab nanoparticles were prepared by desolvation method as reported previously41 with slight modifications. The method provides the opportunity of controlling the size of nanoparticles by varying a number of process parameters such as desolvating agent, rate of addition of desolvating agent, stirring speed, temperature, and pH of the media. The initial BSA concentration, as well as the concentration of Dox, was varied to prepare Dox Ab nanoparticles with maximum encapsulation efficiency. It has been established that the desolvation process when carried out at pH values near the isoelectric point leads to faster precipitation and formation of larger aggregates.42 Hence, the synthesis of BSA has been carried out at the neutral pH of 7, which is well above the isoelectric point of BSA (pH 4.4). The stirring speed and the rate of ethanol addition were fixed at 500 rpm and 1 mL/min, respectively. Cross-linking of the nanoparticles was required to make the nanoparticles stable and to prevent them from redissolving in water. Glutaraldehyde is a commonly used cross-linker for proteins and the aldehyde groups in it react with amine groups and lysine residues on the protein chain leading to the formation of Schiff’s bases.43 Although the amount of cross-linker does not have an effect on particle size, it is known to have a marked effect on the zeta potential of the nanoparticles and their drug loading and release behavior. The concentration of glutaraldehyde and the reaction time required to cross-link all the amino groups in HSA was reported to be about 40% and 24 h, respectively.44,45 A much lesser glutaraldehyde concentration (5%) and time of cross-linking (15 h), just enough to maintain particle stability, was used in the synthesis process so that the drug release properties are not affected. A schematic diagram depicting the synthesis of Dox Ab nanoparticles and further modifiacation using LbL technique is given in Scheme 1. First, the Dox concentration was fixed at 250 μM and the variation of zeta potential and size of the Dox Ab nanoparticles as a function of BSA concentration was observed (Table 1). Since the pI of BSA is around 4.4, negatively charged nanoparticles will be formed at pH 7. As expected, all the samples had highly negative zeta potential. However, at a low concentration of BSA
volume (mm 3) = (A × B2 )/2 where A is the largest diameter (mm) and B is the smallest diameter (mm). 2.14. Statistical Analysis. The experimental data were subjected to statistical analysis by applying the Student’s t-test and One-way ANOVA using commercially available GraphPad Prism 5 software. A P value
