Article pubs.acs.org/Biomac
Polyphenolic Polymersomes of Temperature-Sensitive Poly(N‑vinylcaprolactam)-block-Poly(N‑vinylpyrrolidone) for Anticancer Therapy Veronika Kozlovskaya,†,∥ Fei Liu,†,∥ Bing Xue,† Fahim Ahmad,‡ Aaron Alford,† Mohammad Saeed,‡ and Eugenia Kharlampieva*,†,§ †
Department of Chemistry, University of Alabama at Birmingham, Birmingham, Alabama 35294, United States Department of Infectious Disease, Drug Discovery Division, Southern Research, Birmingham, Alabama 35205, United States § Center for Nanoscale Materials and Biointegration, University of Alabama at Birmingham, Birmingham, Alabama 35294, United States ‡
S Supporting Information *
ABSTRACT: We report a versatile synthesis for polyphenolic polymersomes of controlled submicron ( 36 °C are not stable and dissolve at T < LCST and therefore cannot be reliably used at physiological temperature.11,13 To “lock” the vesicle morphology in the physiologically relevant range of 36− 42 °C, the polymersomes formed at elevated temperatures are often chemically or physically cross-linked. For instance, the Received: May 15, 2017 Revised: July 10, 2017 Published: July 12, 2017 2552
DOI: 10.1021/acs.biomac.7b00687 Biomacromolecules 2017, 18, 2552−2563
Article
Biomacromolecules
hydrophilic drugs.50 Despite the remarkable biological properties of the layer-by-layer TA-containing microcapsules, decreasing their size to the submicron level is challenging due to the aggregation of sacrificial particles under 1 μm in size during the multistep assembly process as well as the timeconsuming fabrication process. To overcome this challenge, we integrate characteristics of PVCLn−PVPONm polymersomes, such as monodispersity, onestep temperature-induced self-assembly, and submicron ( LCST of the PVCLn−PVPONm diblock copolymer via the hydrogen bonding between TA molecules and PVPON blocks that stabilizes them against low temperature-induced dissociation. The PVCLn−PVPONm diblock copolymers with n = 179 and m = 107, 166, 205, and 234 are attained by controlled RAFT polymerization of PVCL179 used as a macro chain transfer agent. The initial size of the TA(PVCL179−PVPONm) polymersomes obtained at room temperature and their temperature-induced size changes are controlled by varying PVPON chain length as well as the TA:PVPON molar ratio. In situ study of polymersome membrane permeability is achieved using the small-molecular weight anticancer drug doxorubicin (DOX) and higher molecular weight fluorescein isothiocyanate (FITC)−dextran (Mw = 4000 Da), while particle stability against aggregation is explored in bovine serum in the physiologically relevant pH and temperature range. Intracellular degradation of the polymersomes and the relationship between drug concentration and incubation time on cancer cell viability are investigated in cocultures of DOX-loaded polymersomes with human alveolar adenocarcinoma A549 cells. Considering their high loading capacity, biocompatibility, biodegradability, monodispersity, and facile two-step assembly based on the reversible temperature-responsiveness of PVCL, these polymeric vesicles have substantial potential as a novel type of drug nanocarrier as well as offer a new perspective for fundamental studies of thermo-triggered block copolymer assemblies in solutions.
PNIPAM-based diblock copolymer poly(N-(3-aminopropyl)methacrylamide hydrochloride)-b-PNIPAM could be assembled into polymersomes in aqueous solutions at 45 °C and were stabilized via formation of an ionically paired polyelectrolyte complex with an anionic homopolymer.13 Thermoresponsive poly(N-vinylpyrrolidone) (PVPON)-b-PNIPAM block copolymer micelles were assembled into multilayers at 37 °C and stabilized using tannic acid (TA) as a physical cross-linker sandwiched between the layers of the micelles.20 Importantly, unlike the previously reported chemical crosslinking of micelles21,22 or polymersomes,23,24 noncovalent cross-links are advantageous, as they allow much faster “lockin” and also the reversibility of the process. Conversely, the ionic cross-links very often can be removed only by adding high salt concentrations to shield the electrostatic binding between strong polyelectrolyte components or by changing pH to extreme values to lower ionization degrees of weak polyelectrolytes.13,25 In this regard, cross-linking via hydrogen bonds avoids ionic interactions and therefore can increase biocompatibility of the resultant nanocarriers. H-bonded crosslinking additionally increases their biodegradability in the presence of intracellular enzymes, which promotes easy breakdown and removal by the body. The current lack of such a system motivates our development of biocompatible nanocarriers that can be easily assembled and loaded with therapeutic cargo in an aqueous environment followed by release of the therapeutics inside target cells. We recently demonstrated controlled synthesis of poly(N-vinylcaprolactam)-b-PVPON (PVCL−PVPON) diblock copolymers using reversible addition−fragmentation chain transfer polymerization (MADIX/RAFT).26 These block copolymers were noncytotoxic and could self-assemble into polymersomes above the copolymer’s LCST due to dehydration of the PVCL block.26 However, those polymeric vesicles were stable only at elevated temperatures. We therefore explored a versatile synthesis of polyphenolic polymersomes via stabilization of the vesicular morphology of thermally responsive PVCLn−PVPONm diblock copolymers with TA through noncovalent hydrogen bonding at a temperature above the copolymer’s LCST. PVCL has excellent biocompatibility, stability against hydrolysis, and complexation ability.27,28 In contrast to PNIPAM, PVCL possesses a classical Flory−Huggins thermoresponsive phase diagram with a continuous coil-to-globule phase transition from 36 to 50 °C, depending on molar mass and concentration.29,30 This feature allows for convenient control of the temperature sensitivity of PVCL polymers via varying the molecular weight or through functionalization of the vinylcaprolactam monomer before copolymerization.31−33 One the other hand, TA is a hydrolyzable tannin containing a glucose core with connected digalloyl ester groups used for many biomedical applications.34−39 The TA phenolic groups can promote ionic pairing,40,41 hydrogen bonding,42,43 and metal coordination.35,44 We have previously shown that hydrogen-bonded TA/ PVPON coatings made via multilayer assembly prolonged the viability and function of pancreatic islets in vitro45 and in vivo,46 while (TA/PVPON) multilayer microcapsules showed antioxidant and immunomodulatory properties by suppressing synthesis of proinflammatory cytokines and chemokines from immune cells.46,47 The TA/PVPON microcapsules were also found to be noncytotoxic to other types of cells48,49 and demonstrated excellent retention of small molecular weight
■
EXPERIMENTAL SECTION
Materials. TA (Mw = 1700 Da), N-vinylcaprolactam (VCL), Nvinylpyrrolidone (VPON), 1,4-dioxane, 2,2′-azobis(2-methylpropionitrile) (AIBN), tetrahydrofuran, and tetra-n-butylammonium fluoride trihydrate (TBAF) were purchased from Sigma-Aldrich. AIBN was recrystallized from methanol before use. VCL and VPON were purified by distillation under vacuum. 1,4-Dioxane was freshly distilled before use. Hexane, methyl-2-bromopropionate, methanol, and potassium ethyl xanthate were purchased from Fisher Scientific and used as received. Dextran of 4000 Da fluorescently labeled with fluorescein isothiocyanate (FITC-dextran) was purchased from SigmaAldrich. Doxorubicin hydrochloride was from LC laboratories. Ultrapure deionized water (18.2 Ω cm) was used for aqueous solutions (Evoqua). Synthesis of PVCL Macroinitiator. The PVCL macroinitiator was synthesized using reversible addition−fragmentation chain transfer (RAFT) controlled radical polymerization per our previous report.26 Briefly, the RAFT chain transfer agent (CTA) was prepared first by reaction of methyl-2-bromopropionate (30 mmol) with potassium ethyl xanthate (35 mmol) in methanol for 24 h as reported previously.51 The product was extracted with a mixture of hexane:ethyl ether (1:1) and collected after solvent evaporation. Freshly distilled 1,4-dioxane (5 mL), VCL (15 g, 107 mmol), and AIBN (20 mg, 0.12 2553
DOI: 10.1021/acs.biomac.7b00687 Biomacromolecules 2017, 18, 2552−2563
Article
Biomacromolecules mmol) were mixed with CTA (55 mg, 0.26 mmol) in a Schlenk flask under constant stirring, and the solution was degassed by three cycles of freeze−pump−thaw under argon (Airgas) and heated at 60 °C in an oil bath for 6 h. The reaction was quenched by placing the flask in a dry ice bath for 20 min. The product was purified by precipitation in hexane and characterized with gel permeation chromatography (GPC, Waters) using linear polystyrene standards (Polymer Standards Service). PVCL with Mn = 21250 g mol −1, Đ = 1.17 was obtained with 93% yield (14 g). Synthesis of PVCL−PVPON Diblock Copolymers Using RAFT. The PVCL−PVPON diblock copolymers were synthesized with a similar protocol to the PVCL homopolymer but substituting the CTA with a prepared macro-transfer agent and using VPON as a monomer according to our earlier protocol.26 For that, the PVCL macroinitiator (0.5 g, 0.014 mmol) in 2.5 mL of 1,4-dioxane was mixed with VPON (2 g, 18 mmol) and AIBN (2 mg, 0.012 mmol) in a Schlenk flask; the solution was degassed with three freeze−pump−thaw cycles and heated at 60 °C for 2 h in an oil bath. The synthesized block copolymer was purified as described above, and the molar ratio of PVCL to PVPON was obtained using 1H NMR spectroscopy. Self-Assembly of TA/(PVCL−PVPON) Polymersomes. PVCL− PVPON polymersomes were assembled by gradually increasing the temperature of PVCL−PVPON aqueous solutions (1 mg mL−1) to 48 °C at 2 °C min−1. To lock the polymersome morphology, after reaching 48 °C, 0.5 mg mL−1 TA solution (300 μL) was added dropwise to 0.8 mL of the PVCL179−PVPON166 polymer solution. The mixture was cooled to room temperature at 2 °C min−1 followed by dialysis in water for 48 h using a Float-A-Lyzer (MWCO 10 kDa, Spectrum Laboratories). Loading TA(PVCL−PVPON) Polymersomes with Doxorubicin (DOX) or FITC−Dextran. The DOX- or FITC−dextran-loaded polymersomes were prepared by dropwise addition of 300 μL TA solution (0.5 mg mL−1) into the solution containing 0.2 mL DOX (1.5 mg mL−1) or FITC-dextran (1.5 mg mL−1) and 0.8 mL PVCL179− PVPON166 (1 mg mL−1) solutions at 48 °C. The mixture was cooled to room temperature at 2 °C min−1, and the solution was dialyzed in water for 48 h using Float-A-Lyzer with 10 kDa MWCO (Spectrum Laboratories). The DOX loading (wt %) was quantified by measuring the DOX weight content in the polymersomes. For that, the dialyzed DOX-loaded polymersomes were lyophilized (Labconco) and weighed. The DOX was extracted from polymer mass with ethanol and quantified using UV−visible spectroscopy (Varian Cary 50; Agilent Technology) at λ = 490 nm using a DOX calibration curve. The FITC-dextran loading (wt %) was calculated similarly to that of DOX. Interaction of Serum with DOX-Loaded TA(PVCL−PVPON) Polymersomes. 300 μL of DOX-loaded TA(PVCL179−PVPON166) (1 mg mL−1) were incubated in 1 mL of 10% FBS-containing PBS solution at 25 °C for 2 h. The FBS-treated polymersome solution (300 μL) was added to a Lab-Tek chambered coverglass and confocal microscopy images of the DOX-polymersomes were obtained with Nikon A1R+ confocal laser microscope equipped with a 60× oil immersion objective using TRITC channel (Ex/Em = 557/576 nm). The hydrodynamic size of the serum-treated polymersomes was determined using DLS. Release of Cargo from TA/(PVCL−PVPON) Polymersomes. The release experiments were carried out using dialysis method. For that, 1 mL of DOX-loaded TA(PVCL179−PVPON166) polymersome solution in Float-A-Lyzer dialysis tube (MWCO 3 kDa, Spectrum Laboratories) was placed in 50 mL PBS solution (pH = 7.4, HyClone) and kept under stirring (80 rpm) at 37 °C in an incubator (Precision Scientific Company). The PBS solution (3 mL) was withdrawn periodically and analyzed using UV−visible spectroscopy. The amount of DOX was quantified using a DOX calibration curve. Turbidity Measurements. To determine the turbidity of the diblock copolymer solutions as a function of temperature, 3 mL of aqueous copolymer solution (1 mg mL−1) was measured using a fluorescence spectrophotometer (Varian, Cary Eclipse). The scattering intensity of the polymer solution was obtained at λ = 700 nm from 30 to 50 °C with the temperature increase rate of 0.2 °C min −1. The
measurements were carried out in 0.01 M phosphate buffer at pH = 7.4. The increase in optical density with increasing temperature resulted from the copolymer phase transition. The LCST of the thermoresponsive copolymers was calculated as the inflection point of the curve of optical density as a function of temperature. Cell Studies. Human alveolar adenocarcinoma A549 cells were used for polymersome cytotoxicity and cell viability studies. For polymersome internalization studies, cells were plated at a density of 1 × 105 cells per well in a 96-well plate. Cells were allowed to adhere for 24 h followed by culturing with DOX-free, DOX-loaded polymersomes, and DOX solution for 1, 6, and 24 h at 0.1, 0.5, 1, 5 μg mL−1. After that, liquid medium was aspirated, and the cells were washed once with PBS and covered to a depth of 2−3 mm with 4% paraformaldehyde diluted in PBS for 20 min. Cells were rinsed three times with PBS for 5 min and cell nuclei were stained with 2-(4amidinophenyl)-1H-indole-6-carboxamidine (DAPI, Sigma-Aldrich) to stain the cell nuclei. Fluorescent optical sections of cells were collected with a Nikon A1R+ confocal laser microscope equipped with a 60× oil immersion objective using a TRITC/DAPI filter set. Cell viability was assessed with MTT viability assay (Sigma-Aldrich). Cells were seeded in 96-well plates at a density of 2 × 104 cells per well in Dulbecco’s Modification of Eagle’s Medium (DMEM; high glucose), supplemented with 10% heat-inactivated fetal bovine serum (FBS; Mediatech Inc.) and allowed to grow to near confluency. Subsequently, the cells were incubated with DOX-free or DOX-loaded polymersomes. Solution of DOX-free polymersomes (1 mg mL−1) was prepared as a stock and used to prepare 100, 50, 0.5, and 0.1 μg mL−1 polymersome solutions by dilution with PBS buffer. For DOX-loaded polymersomes, the stock polymersome solution containing 150 μg mL−1 DOX was diluted with PBS buffer to result in final DOX concentrations of 5, 1, 0.5, and 0.1 μg mL−1. DOX solutions with the same concentrations were used as positive cytotoxicity controls, while nontreated cells were used as a negative cytotoxicity control. The cell plates with samples were incubated at 37 °C (humidity ≥85%; an air/ CO2 ratio = 95/5%). After 24 h, liquid medium was aspirated, and the cells were rinsed three times with PBS for 5 min each, and cell viability was quantified in each well by adding 100 μL of Cell-Titer Glo reagent (Promega) and measuring the luminescence using an Envision plate reader (PerkinElmer). Prior to cell viability measurements, images of cells in each well were captured for visual control. Transmission Electron Microscopy (TEM). TEM images of the polymersomes were obtained with a FEI Tecnai T12 Spirit TWIN TEM microscope operated at 80 kV. Polymersome solution (1 mg mL−1, 7 μL) was dropped onto an argon plasma-treated Formvar/ Carbon-coated copper grid (200 mesh, TED Pella) for 30 s followed by blotting off the excess solution with Kimwipe paper. The TEM grid was stained with 1 wt % uranyl acetate (Electron Microscopy Sciences) for 30 s followed by removing excess solution and air-drying. For cryoelectron microcopy (cryo-TEM), 3 μL of sample was applied to glowdischarged 200 mesh Quantifoil R 2/1 grids (Electron Microscopy Sciences, Hatfield, PA). The grid was loaded into the FEI Vitrobot Mark IV (FEI, Eindhoven, Netherlands), where it was blotted briefly and plunged into liquid ethane.52 Frozen grids were transferred to a Gatan 622 cryo-holder and observed in an FEI Tecnai F20 electron microscope (Eindhoven, Netherlands) operated at 200 kV with magnification at 65 500× and defocus settings of −2.5 to −3.0 μm. Images were collected under low-dose conditions on a Gatan Ultrascan 4000 CCD camera. Gel Permeation Chromatography (GPC). Waters GPC supplied with a Waters 1515 pump, 2414 differential refractive index detector and two connected Styragel columns was used to determine molecular weights of polymers. Tetrahydrofuran was used as eluent with 0.02 M TBAF at the flow of 1 mL min−1 at room temperature. The calibration was performed with monodisperse linear polystyrene standards (Waters). Nuclear Magnetic Resonance (NMR). 1H NMR spectra of the copolymers were recorded on a Bruker 400 MHz NMR spectrometer. The copolymer solutions (1 mg mL−1) were prepared in CDCl3 and measured at 25 °C. 2554
DOI: 10.1021/acs.biomac.7b00687 Biomacromolecules 2017, 18, 2552−2563
Article
Biomacromolecules Dynamic Light Scattering (DLS). DLS was conducted on a Nano-ZS Zetasizer (Malvern) with He−Ne laser at 663 nm at 173°. Average hydrodynamic sizes of polymersomes were obtained from three independent runs from of 1 mg mL−1 aqueous polymer solutions. Polydispersity index was used to evaluate the size distribution.
■
RESULTS AND DISCUSSION Synthesis of PVCLn−PVPONm Diblock Copolymer Polymersomes and Their Interaction with TA. Temper-
Figure 1. (a) Synthesis of the PVCL−PVPON diblock copolymers and (b) optical density of various PVCL179−PVPONm diblock copolymers as a function of temperature.
Figure 2. PVCLn-PVPONm diblock copolymers (a) assembled into polymersomal nanocapsules at T > LCST (b) can be locked with TA (c) via hydrogen-bonds with PVPON, which results in PVCLn− PVPONm nanocapsules stable at T < LCST (d). (e,f) TEM images of TA(PVCL179−PVPON107) polymersomes stained with uranyl acetate. (g) UV−vis spectra of 0.2 mg mL−1 TA solution (red circles), PVCL179−PVPON107 aqueous solution (blue triangles), and TA(PVCL179−PVPON107) vesicles in deionized (DI) water (green squares) at 25 °C.
Table 1. Number-Average Molecular Weights (Mn), Polydispersity (Đ), and Chemical Composition of the Copolymers sample PVCL179 PVCL179− PVPON107 PVCL179− PVPON166 PVCL179− PVPON205 PVCL179− PVPON234
Mna GPC
Đ (GPC)
Mnb NMR
PVCL, wt %
PVPON, wt %
21250 21400
1.17 1.29
24881 36758
100 68
0 32
22688
1.20
43307
57
43
23674
1.28
47636
52
48
23722
1.18
50855
49
51
(Figure 1b). The data agrees with the earlier reported LCST decrease from 46 to 38 and to 33 °C for linear PVCL chains with increased Mw from 18 000 to 31 500, and to 150 000 g mol−1, respectively.51,53 The collapse of the hydrophobic PVCL block shifted to a higher temperature of ∼37.8 °C when a PVPON hydrophilic block with polymerization degree of 107 was extended from the PVCL179 (Figure 1b). The dependence of the LCST of the PVCL block copolymer on the hydrophilicity of the following block has been demonstrated by us26,53 and others31,32 and allows tuning of the copolymer LCST to the useful biological range. Indeed, increasing the polymerization degree of PVPON to 166, 205, and to 234 resulted in a temperature shift of the optical density onset of the PVCL179−PVPON diblock copolymers to 38.6, 39, and 39.8 °C, respectively (Figure 1b). The coil-to-globule transitions of the copolymers were completely reversible, and the solutions became clear after the temperature was decreased to room temperature as we reported previously.26 Heating solutions of these diblock copolymers above the copolymer LCST leads to their self-assembly into macromolecular structures such as micelles or vesicles due to the dehydration of the PVCL block.26,33 Vesicular morphology for PVCL179−PVPONm assemblies is expected when the ratio of the hydrophilic fraction (PVPON) to the total copolymer mass, f, is 0.25 < f < 0.5.9,54 The hydrophilic ratio of the synthesized PVCL179−PVPONm was varied from 0.32 to 0.43, to 0.48, and to 0.51 for m = 107, 166, 205, 234, respectively (Table 1). However, these vesicles disassemble into free polymer chains when temperature is decreased back to room temperature (Figure 2a,b). We found that by adding TA into the polymersome solution at T > LCST, the polymersome vesicles
a
The GPC number-average molecular weights were obtained using polystyrene linear standards. bCalculated from 1H NMR, the molecular weight of the second block was calculated from the molar ratio of the first and the second blocks.
ature-sensitive PVCLn−PVPONm diblock copolymers were synthesized by RAFT polymerization using O-ethyl-S-(1methoxycarbony) ethyl dithiocarbonate as a chain transfer agent as we previously reported (Figure 1a).26 The PVCL179 macroinitiator with Mn = 21 250 g mol −1, (Đ = 1.17) was used for synthesis of PVCL179−PVPONm diblock copolymers in dioxane at 60 °C with the reaction time of 2, 3, 4, and 6 h to obtain the PVPONm block with m = 107, 166, 205, 234, respectively (Table 1). The PVPON chain extension leading to the PVCL179−PVPONm diblock copolymers was confirmed by size-exclusion chromatography using linear polystyrene standards for molecular weight calibration (Figure S1), while the PVCL:PVPON composition of the diblock copolymers was analyzed using H1 NMR analysis (Figure S2). The turbidity of the PVCL179 and its diblock copolymers performed by heating their aqueous solutions from 32 to 42 °C and measuring their optical density at 700 nm using fluorescence spectroscopy revealed that PVCL179 had a coilto-globule transition at ∼34.5 °C when its aggregation occurred 2555
DOI: 10.1021/acs.biomac.7b00687 Biomacromolecules 2017, 18, 2552−2563
Article
Biomacromolecules
Figure 3. Average hydrodynamic sizes of PVCL179−PVPONm diblock copolymer solutions (1 mg mL−1) with n = 107 (a), 166 (b), 205 (c), and 234 (d) as a function of temperature before (squares) and after (circles) interaction with TA.
Figure 4. Hydrodynamic size of PVCL179−PVPONn with n = 107, 166, 205, and 234 before and after interaction with TA at 50 °C and cooling to 25 °C.
molecules to PVPON blocks via multiple hydrogen bonds,43,50 which resulted in the formation of a thin TA(PVPON−PVCL) shell stable at room temperature (Figure 2d). Figure 2g demonstrates that the UV absorbance from the solution of “locked” TA(PVCL179−PVPON107) vesicles is shifted to lower energy (282 nm) compared to the peak at 276 nm observed for free TA in water. This shift confirms hydrogen bonding between TA hydroxyl groups and the carbonyls of PVPON (Figure 2g). This result agrees with our previous UV−vis studies on the sensitivity of this region to TA−PVPON hydrogen-bonded complex formation50 and with the reported
can be stabilized at T < LCST, and polyphenolic TA(PVCL179−PVPONm) nanocapsules can be obtained (Figure 2b−d). Figure 2 shows TEM images of vesicles made from PVCL179−PVPON107 by adding 0.5 mg mL−1 TA solution into the aqueous copolymer solution at 50 °C, cooling to room temperature, and imaging at 25 °C (Figure 2e,f). The vesicle TEM images demonstrate light interiors implying their impermeability to the TEM staining dye. The “freezing” of the vesicular morphology at temperatures lower than the copolymers LCST was possible due to the binding of TA 2556
DOI: 10.1021/acs.biomac.7b00687 Biomacromolecules 2017, 18, 2552−2563
Article
Biomacromolecules
tional entropy of macromolecular chains in a crowded environment. The addition of TA molecules to the self-assembled structures at 50 °C caused substantial shrinkage for all studied copolymer self-assemblies, which did not dissolve into molecular chains when the temperature was lowered to 25 °C (Figure 3a−d). The decrease in average hydrodynamic sizes was from 1000 to 710 nm (PDI = 0.185) for PVCL179− PVPON107 (Figure 4a), from 950 to 460 nm (PDI = 0.167) for PVCL179−PVPON166 (Figure 4b), from 770 to 290 nm (PDI = 0.143) for PVCL179−PVPON205 (Figure 4c), and from 250 to 190 nm (PDI = 0.026) for PVCL179−PVPON234 (Figure 4d) upon interaction with TA at 50 °C. This remarkable shrinkage after interaction of the polymersomes with TA can be attributed to PVPON chain collapse due to hydrogen bonding with TA. These observations are consistent with the decreased size of hydrogel microshells where anionic hydrogel hollow capsules shrank upon complexation with macromolecular cations,25 and with the decreased diameter of poly(N-(3aminopropyl)methacrylamide hydrochloride)-b-(PNIPAM) cationic diblock copolymer vesicles assembled at 45 °C and stabilized at 25 °C via ionic cross-links of the shell with anionic poly(sodium 2-acrylamido-2-methylpropanesulfonate) added to the vesicle solution at 45 °C.13 After lowering the solution temperature to 25 °C, a slight increase in the average hydrodynamic sizes of the selfassembled structures was observed for PVCL179−PVPON166 from 460 to 530 nm (PDI = 0.247) (Figure 4b), for PVCL179− PVPON205 from 290 to 310 nm (PDI = 0.188) (Figure 4c), and for PVCL179−PVPON234 from 190 to 210 nm (PDI = 0.029) (Figure 4d). The slightly elevated values of the average hydrodynamic size observed at 25 °C also correlate with the slight increase in the polydispersity indexes of the samples. These results may imply that when the temperature becomes lower that the LCST of a diblock copolymer, the PVCL chains tend to relax and transition back from the globule to coil conformations. In this case, the elastic force of the PVCL chain extensions may cause an increase in the final size of the polymersomes at room temperature. No measurable change in the average size of the PVCL179−PVPON107 vesicles upon cooling may suggest that the collapsing PVCL chains are not completely shielded by the short PVPON107 blocks from TA molecules at 50 °C, and the hydrogen bonding interactions may occur between the PVCL and TA along with TA and exterior PVPON chains. The TEM analysis of the “locked” TA(PVCL179−PVPON) self-assemblies prepared from their solutions at 25 °C confirmed the presence of polymersomes with spherical morphology and the absence of aggregation in all samples (Figure 5). In the TEM image of TA(PVCL179−PVPON205) (top panel, Figure S3), the self-assembled structures are stained with the dye and clearly demonstrate hollow vesicles with an unstained cavity inside the vesicles and a darker vesicle membrane, which is in agreement with previous reports on visualizing vesicular structures, in contrast to micelles or particles, which would appear with dark cores (Figure S4b). No additional compartments can be seen in the cryo-TEM image (Figure S3, bottom panel), and the thickness of the TA coating was estimated to be 15 ± 3 nm by ImageJ analysis. As seen from the TEM images (Figure 5) and DLS data in Figure 4, the use of shorter PVPON107 block results in the formation of the largest PVCL179−PVPON polymersomes, which is consistent with reports where decreasing hydrophilic block lengths led to
Figure 5. TEM images of TA/(PVCL179−PVPONm) polymersomes with m = 107 (a), 166 (b), 205 (c), and 234 (d) dried from solutions at 25 °C.
Figure 6. Temperature-dependent hydrodynamic size of PVCL179− PVPON166 before (open squares) and after interaction with 0.5 mg mL−1 TA solution with (VPON:TA) molar ratio of 53:1, filled spheres; of 35:1, filled squares; and of 26:1, filled triangles.
correlation between the electronic absorption frequencies of phenolic groups and their hydrogen-bonded states.55,56 Control of TA(PVCLn−PVPONm) Polymersome Size and Their Stability. Figure 3 demonstrates the evolution of hydrodynamic sizes of the diblock copolymers with temperature ranging from 25 to 60 °C as measured with DLS. The hydrodynamic diameter of the copolymer assemblies was below 15 nm at 30 °C. The size increased at the temperature close to the copolymer LCSTs followed by decrease at ∼45 °C and reached the stabilized values of 1000 nm (PDI = 0.002), 950 nm (PDI = 0.018), 770 nm (0.081), and 250 nm (PDI = 0.099) at ∼50−60 °C for PVCL179−PVPONm with m = 107, 166, 205, 234, respectively (Figure 3a−d). The observed decrease of the hydrodynamic size of the assemblies after 10−15 degrees past their LCSTs has been reported earlier for PVCL copolymers and is explained by formation of aggregates due to gradual dehydration of PVCL blocks after heating above the LCST unlike PNIPAM which exhibits a sharp phase transition.57,58 At higher temperatures, larger aggregates dissociate into smaller self-assembled structures due to the decrease in the configura2557
DOI: 10.1021/acs.biomac.7b00687 Biomacromolecules 2017, 18, 2552−2563
Article
Biomacromolecules
Figure 7. (a) UV−visible spectroscopy from solutions of PVCL179−PVPON166 diblock copolymer (black), empty TA(PVCL179−PVPON166) polymersomes (blue), and DOX-loaded (red) and FITC-dextran-loaded (green) TA(PVCL179−PVPON166) polymersomes at 25 °C. (b) The intensity and (c) number size distributions of DOX-loaded TA(PVCL179−PVPON166) polymersomes as prepared (Day 1) and 3 and 5 days after preparation as measured with dynamic light scattering. (d) Release (%) of DOX or FITC-dextran from DOX (544 Da)-loaded and FITC-dextran (4 kDa)-loaded TA(PVCL179−PVPON166) polymersomes.
were obtained when PVCL179−PVPON166 diblock copolymer solution was mixed with TA at the molar unit ratio of VPON:TA = 35:1 at 50 °C (Figure 6, filled squares). At the lower TA concentration, with the VPON:TA ratio of 53:1, the TA(PVCL179−PVPON166) vesicles first shrank upon addition of TA at 50 °C from 950 nm (PDI = 0.018) to 620 nm (PDI = 0.165), and then dissociated into solution at T < 45 °C with the final average hydrodynamic size of ∼18 nm at 25 °C (Figure 6, filled circles). Interestingly, when the TA amount was increased by 30% with the resulting VPON:TA ratio of 26:1, the average hydrodynamic size increased from 950 to 1300 nm implying cross-linking between TA(PVCL179−PVPON166) polymersome vesicles with, on average, three nanovesicles per aggregate (Figure 6, filled triangles). Loading and In Vitro Release of Molecules by TA(PVCL−PVPON) Polymersomes. We explored the ability of the locked polymersomes to load and carry a therapeutic cargo by loading the TA(PVCL179−PVPON166) with either high or low molecular weight molecules of FITC−dextran (Mw = 4000 g mol−1) or DOX (Mw = 543.5 g mol−1), respectively. Figure 7 demonstrates that, after loading of the polymersomes with DOX and FITC−dextran by assembling the vesicles at 48 °C, adding TA, and dialyzing in DI water, both loaded vesicle solutions exhibited absorbance at 282 nm due to TA present in the polymersome shell (Figure 7a). Also, the absorbance bands at 490 and 480 nm were present for DOX- and FITC-dextranloaded TA(PVCL179−PVPON166) polymersomes, respectively, unlike cargo-free TA(PVCL179−PVPON166) nanoshells and (PVCL179−PVPON166) copolymer solution. The intensity and number distributions of size of the loaded polymersomes measured using DLS revealed the stability of the loaded polymersome size during 5 days of measurement with no sign of aggregation (Figure 7b,c).
Figure 8. Viability of human alveolar adenocarcinoma A549 cells (%) after incubation with various concentrations of (a) PVCL179− PVPON166 vesicles for 24 h, and (b) DOX-loaded PVCL179− PVPON166 vesicles (DOX-vesicles) and DOX solution (DOX) for 24 h at 37 °C. Each data point represents an average of three replicates ± SD.
an increased number of self-assembling diblock chains owing to a lowered steric repulsion in the hydrophilic layer.59 We investigated the amount of TA required for a reliable cross-linking of PVPON chains and formation of the TA(PVCL179−PVPON) nanoshells that were stable at room temperature. Figure 6 demonstrates that these nanocapsules 2558
DOI: 10.1021/acs.biomac.7b00687 Biomacromolecules 2017, 18, 2552−2563
Article
Biomacromolecules
Figure 9. Confocal microscopy images of human alveolar adenocarcinoma A549 cells after incubation with DOX (a−c and g−i) and DOX-loaded (5 μg mL−1) TA(PVCL179−PVPON166) polymersomes (d−f and j−l) for 1 (a−f) and 6 (g−l) hours at 37 °C, pH = 7.4. DAPI-labeled cell nuclei emit a blue fluorescence signal, and DOX emits a red fluorescence signal. The scale bar is 5 μm in all images.
with human alveolar adenocarcinoma A549 cells, and vesicle cytotoxicity to the cells was quantified using Cell-Titer Glo assay (see Experimental Section). The viability of the cells did not decrease after 24 h and was above 90% after incubation with the TA(PVCL179−PVPON166) polymersomes at concentrations ranging from 0.1 to 100 μg mL−1 (Figure 8a). This result agrees with our previous reports on the noncytotoxicity of PVCL−PDMS−PVCL triblock copolymer polymersomes54 and PVCL−PVPON diblock copolymers, 26 and (TA/ PVPON), (TA/PVCL), (PVCL), and (PVPON) multilayer microcapsules.39,46−49 By contrast, when the cells were incubated with DOX-loaded TA(PVCL179−PVPON166) polymersomes, the cell viability decreased from 100 to 57 ± 8% for 0.1 μg mL−1 DOX-loaded vesicles after 24 h (Figure 8b). This result indicates successful intracellular delivery of DOX from the TA-locked polymersomes potentially due to biodegradation of TA by intracellular enzymes. We also observed a concentration-dependent cytotoxicity for the DOX-loaded TA(PVCL179−PVPON166) vesicles when the
The dialysis of DOX- and FITC−dextran-loaded TA(PVCL179−PVPON166) vesicles at pH = 7.4 in PBS solution at 37 °C revealed a negligible release of DOX or FITC−dextran (less than 2%) under these conditions over 74 h (Figure 7d). Unlike previously reported DOX-loaded PEO-b-PHEMA micelles, which, when assembled with TA into multilayers, released ∼20% of DOX at pH = 7.4 during 24 h,60 the TA(PVCL−PVPON) vesicles demonstrate an outstanding ability to store a small molecular weight drug like DOX for long time periods. Our data is in excellent agreement with our previous reports on 5 μm (PVPON/TA) multilayer microcapsules, which showed an insignificant release of DOX from the capsule interiors after 24 h with ∼1 and ∼2% of the released drug, at pH = 7.4 and pH = 5, respectively.50 Intracellular Release of DOX from TA(PVCL179− PVPON166) Polymersomes. The therapeutic properties of TA(PVCL179−PVPON166) vesicles were evaluated by incubating DOX-free or DOX-loaded vesicles (DOX-vesicles) with various concentrations of DOX at pH = 7.4 and 37 °C for 24 h 2559
DOI: 10.1021/acs.biomac.7b00687 Biomacromolecules 2017, 18, 2552−2563
Article
Biomacromolecules
mediated drug uptake by tumor cells in contrast to free drug at the same dose as reported previously.54,61,62 Figure 9 demonstrates the representative confocal microscopy images of the cells incubated for 1 and 6 h with free DOX solution (Figure 9a−c and g−i), and DOX-loaded TA(PVCL179−PVPON166) polymersomes (Figure 9d−f and j−l). The blue fluorescence in the DAPI channel indicates cell nuclei, while the red fluorescence from the DOX channel observed within the A549 cytosolic space after 1 h of incubation indicates effective DOX transport into the cancer cells (Figure 9a,b). The overlap of the two channels with the transmitted light view shows the resultant pink color from the nuclei within the brightfield contours of cells (Figure 9c). Unlike free DOX, no DOX fluorescence was present in the cell nuclei after 1 h when DOX-vesicles were used (Figure 9e,f). This result suggests that the DOX−TA(PVCL179−PVPON166) vesicles require a longer time for cell internalization and could be removed from the cell surfaces during the PBS rinsing before cells were fixed with formaldehyde (see Experimental Section). Indeed, after a 6 h incubation, red DOX fluorescence was observed within the cell nuclei for both DOX- and DOX-vesicle-treated cells (Figure 9h,i and k,l, respectively), confirming the successful release of DOX from the vesicles and its delivery to the cytoplasm and to the cell nuclei. Intriguingly, along with the homogeneous DOX fluorescence from the nuclei, red-fluorescent dot-like species were also observed within the cytosol after 6 h when the cells were treated with DOX-vesicles (Figure 9k,l). Presumably, the DOXvesicles could increase in size upon their uptake by the cells either due to swelling or interaction with intracellular proteins and become visible with the confocal microscopy. In the control experiment, DOX-loaded TA(PVCL179−PVPON166) vesicles were exposed to 10% FBS solution in PBS for 2 h at 25 °C, and their size was determined using DLS. Figure 10 shows that the serum treatment resulted in the increased average size of the vesicles from 400 to 1000 nm (Figure 10a), and they were visible via the confocal fluorescence microscopy as bright red dots (Figure 10b), which is agreeable with the confocal
Figure 10. (a) Hydrodynamic size of DOX-loaded TA(PVCL179− PVPON166) before (circles) and after interaction with BSA at 20 °C (squares). (b) Confocal microscopy image of DOX-loaded (5 μg mL−1) TA(PVCL179−PVPON166) polymersomes after being incubated with BSA at 25 °C for 2h. DOX emits a red fluorescence signal. The scale bar is 5 μm.
cell viability decreased from 57 ± 8 to 49 ± 5, to 36 ± 3, and to 8 ± 1% after 24-h incubation with the vesicles loaded with 0.1, 0.5, 1, and 5 μg mL−1 DOX, respectively (Figure 8b). DOX solutions with the matching concentrations of the drug used as a positive cytotoxic control resulted in a similar cytotoxicity trend after 24 h incubation. In this case, the cell viability decreased from 50 ± 1% for 0.1 μg mL−1 DOX to 33 ± 8% for 0.5 μg mL−1 DOX, to 31 ± 1% for 1 μg mL−1 DOX, and to 30 ± 1% for 5 μg mL−1 DOX (Figure 8b). Importantly, the TA(PVCL179−PVPON166) vesicles loaded with 5 μg mL−1 DOX were 3.8-fold more cytotoxic to the cells than free DOX (Figure 8b). This result agrees with the enhanced carrier-
Figure 11. Confocal microscopy images of human alveolar adenocarcinoma A549 cells after incubation with DOX (a−c) and DOX-loaded (5 μg mL−1) TA(PVCL179−PVPON166) polymersomes (d−f) for 24 h at 37 °C, pH = 7.4. DAPI-labeled cell nuclei emit a blue fluorescence signal, and DOX emits a red fluorescence signal. The scale bar is 5 μm in all images. 2560
DOI: 10.1021/acs.biomac.7b00687 Biomacromolecules 2017, 18, 2552−2563
Biomacromolecules
■
microscopy imaging of the cells incubated with DOX-loaded TA(PVCL179−PVPON166) vesicles for 6 h (Figure 9j−l). Note that the hydrodynamic size of free FBS solution was less than 10 nm as determined with DLS (Figure S5, Supporting Information). Therefore, this data suggests the presence of BSA protein (as the main protein component of FBS) around the TA-locked vesicles after the FBS treatment, as TA can interact with proteins through hydrogen bonding and form TA−protein complexes.63,64 Despite the observed size increase of the DOXnanocapsules upon interaction with BSA, we believe this should not be the issue for future in vivo applications, since in contrast to solid particles, soft multilayer capsules and hydrogels of micrometer size have been demonstrated to enter the cancer cells and deliver a therapeutic cargo.65,66 Remarkably, after 24 h of incubation with the DOX-vesicles, the DOX fluorescence in the cell nuclei increased, while the visibility of the polymersomes significantly decreased, seemingly due to their dissolution (Figure 11). Apparently, the acidic endosomal/lysosomal environment can facilitate the escape of the TA-locked polymersomes via dissolution of TA-BSA complexes63 and subsequent osmotic rupture of the lysosomal vesicles. The resultant access of the intracellular enzymes to TA(PVCL179−PVPON166) causes TA degradation and release of DOX into the cytosolic space. The degradation of complex hydrolyzable tannins by intracellular enzymes has been shown to proceed to lower molecular weight phenolic compounds. Thus, for instance, the biochemical pathway for the degradation of TA by cell-free extracts from Lactobacillus plantarum was the hydrolysis of TA to gallic acid and glucose after 6 h incubation of TA with the intracellular enzymes, followed by decarboxylation of the gallic acid to pyrogallol after 24 h incubation.67
Article
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.biomac.7b00687. GPC traces of temperature responsive diblock copolymers of PVCL179−PVPONn with n = 107, 130, 166, 205, 234; 1H NMR spectra of temperature responsive PVCL179 macroinitiator and PVCL179−PVPONn diblock copolymers with n = 107, 166, 205, and 234; TEM images of self-assembled structures obtained from aqueous solutions of PVCL 155 −PVPON 164 and PVCL155−PVPON404; TEM and cryo-TEM images of TA(PVCL179−PVPON205); hydrodynamic size measured from 8% solution of FBS in PBS at 25 °C (PDF)
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. ORCID
Eugenia Kharlampieva: 0000-0003-0227-0920 Author Contributions ∥
V.K. and F.L. contributed equally to this work.
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was supported by NSF-DMR 1608728 (E.K.), and by funds from Southern Research (M.S.). UAB transmission electron microscopy facility and UAB High Resolution Imaging facility (Confocal laser scanning microscopy) are also acknowledged. The Cryo-EM work was carried out by Terje Dokland, Ph.D. in the Cryo-EM facility (UAB Department of Microbiology). We also thank Cynthia Rodenburg (UAB cryo-EM Facility) for technical assistance.
■
CONCLUSIONS We report on the synthesis of polyphenolic polymersomes using temperature-responsive PVCLn−PVPONm diblock copolymers with n = 179, and m = 107, 166, 205, and 234 synthesized by controlled RAFT polymerization of PVPON using PVCL as a macro-chain transfer agent. We show that the assembly of PVCL−PVPON into submicrometer polymersomes at T > LCST of the copolymers can be stabilized at room temperature by interaction with TA at T > LCST of the diblock copolymers. We demonstrate that the size of the (PVCLn−PVPONm) polymersomes ranging from 1000 to 950, 700, and 250 nm is controlled by the PVPON chain length with n = 107, 166, 205, 234, and decreases significantly to the corresponding range from 710 to 460, 290, and 190 nm upon hydrogen bonding of the polymersome shell with TA. We also show that the TA-locked polymersomes can encapsulate and store the small-molecular weight anticancer drug doxorubicin (DOX) and higher molecular weight FITC−dextran (Mw = 4000 Da) in the physiologically relevant pH and temperature range and that these particles are stable in serum solutions. In addition, we show that encapsulated DOX is released intracellularly in the human alveolar adenocarcinoma tumor cells after 6 h incubation, and the cytotoxicity of DOX-loaded polymersomes is concentration-dependent. This approach allows for obtaining biocompatible nanovesicles of a controlled submicron (
