Article pubs.acs.org/molecularpharmaceutics
Mechanisms of Drug Resistance Reversal in Dox-Resistant MCF-7 Cells by pH-Responsive Amphiphilic Polyphosphazene Containing Diisopropylamino Side Groups Liyan Qiu,* Cheng Zheng, and Qinghe Zhao College of Pharmaceutical Sciences, Zhejiang University, Hangzhou 310058, China ABSTRACT: pH-responsive drug carriers derived from polymers containing weak base groups have been shown to improve the antitumor effect of chemotherapeutics. The common interpretation is that a “proton sponge effect” caused by pH-responsive polymers facilitates endosomal membrane destruction and accelerates cytoplasmic drug release in tumor cells. However, the mechanisms by which pH-responsive weak base polymers disrupt membranes have not been expatiated clearly. Herein, we synthesized a series of pH-responsive amphiphilic polyphosphazenes containing diisopropylamino (DPA) side groups with various contents and investigated the effect of DPA content on the actions of polymers with cell membranes. In a certain pH range, the polymers with elevated DPA content showed enhanced membrane disruptive activity. Electrical interactions between the protonated DPA groups of polymers and the cell lipid bilayer are critical for pH-dependent membrane disruption, which can be competitively prevented by serum proteins. On the other hand, the hydrophobic unprotonated DPA moieties can insert into lipophilic regions of cell membrane. These synergic actions caused the alteration of biomembrane permeability consequently. More interestingly, it was also found that DPA-rich polymers exhibit higher P-glycoprotein (P-gp) inhibition activity as compared with the polymer containing only low levels of DPA by efficiently blocking the internal epitope of P-gp. These findings strongly provide rational support for pHresponsive amphiphilic polyphosphazenes containing DPA side groups to be quite promising drug carriers for intracellular drug delivery applications, especially the treatment of P-gp overexpressing, drug-resistant tumors. KEYWORDS: polyphosphazene, pH-response, P-glycoprotein, drug resistance, nanoparticle
■
INTRODUCTION Nanosized biomaterial-based drug delivery systems are an evolving technique whose development has been accelerated by the emerging field of nanotechnology in recent years. They are extensively proposed as a promising strategy for tumor targeting, although some obstacles exist that currently preclude them from being considered as an optimal tumor therapy. One of the most serious problems encountered in clinical treatment of cancer is multidrug resistance (MDR). MDR tumor cells can quickly decrease intracellular drug concentration through drug efflux and/or drug metabolism.1 Therefore, optimal drug carriers are expected to possess certain anti-MDR characteristics. First, drug carriers should facilitate drug escape from endosomes. Because many drugs or therapeutic biomacromolecules (e.g., nucleic acids and proteins) are endosomal membrane-impermeable, nanosized drug carriers and cargo drugs would be quickly eliminated by tumor cells through endosome recycling if they could not escape from these intracellular compartments after endocytic internalization.2−6 Second, drug carriers should instantly release cargo drugs after internalization into the tumor cells. Stimulus-coupled drug release can facilitate rapid intracellular accumulation of drug to reach therapeutic levels as soon as possible, namely, improving the effectiveness of these drugs in MDR cells.7 In addition, © 2012 American Chemical Society
optimal drug delivery systems should evade P-glycoprotein (Pgp) efflux and preferably inhibit P-gp activity since the major form of drug resistance in tumor cells has been correlated with the increased expression of P-gp.3−5,8,9 Some cationic polymers such as polyethylenimines with high molecular weight cause destruction of endosomal membranes via a “proton sponge” effect and/or interaction with the endosomal membrane,10,11 but high cytotoxicity prohibits their widespread application. Polyanions containing carboxylic acid groups also have membrane-destabilizing activity12−14 but must usually be complexed with other drug carriers such as liposomes to achieve stimulus-responsive drug release.15 In recent years, pH-responsive weak base polymers have excited wide interest as drug carriers.16,17 These polymers are nonionic and are nontoxic in physiological environments but can be protonated within the range of endosomal pH, thus destabilizing the endosomal lipid bilayer.18 Drug carriers derived from these polymers will destabilize rapidly and release cargo drugs upon protonation.16,19 Hence, pH-responsive weak Received: Revised: Accepted: Published: 1109
July 16, 2011 March 23, 2012 April 11, 2012 April 11, 2012 dx.doi.org/10.1021/mp200356w | Mol. Pharmaceutics 2012, 9, 1109−1117
Molecular Pharmaceutics
Article
Figure 1. 1H NMR spectra of pH-sensitive polyphosphazenes in CDCl3. Monomethoxy PEG (mPEG2000, Mn = 2000) was purchased from Fluka, and amino-terminal PEG2000 (PEG2000-NH2) was prepared via a two-step protocol described previously.21 AlexaFluor 546 was purchased from Invitrogen and used without further purification. Rhodamine-123 (R-123) and cyclosporine A (CsA) were obtained from Sigma Aldrich. The micro-BCA protein assay kit was obtained from Biyuntian Co. Ltd. (Jiang Su, China). Doxorubicin hydrochloride (Dox·HCl) was kindly supplied by Juhua Group Pharmaceutical Factory (Zhejiang, China). All other reagents were commercially available and used without further purification. Polymer Synthesis and Composition Determination. The synthesis of DPA- and mPEG-grafted polyphosphazene involves sequential grafting of amino-terminal PEG2000 and DPA side groups onto the poly(dichlorophosphazene) backbone through nucleophilic substitution as reported previously.23 The product was abbreviated as PPD-n [n refers to the molar ratio of DPA to amino-terminal PEG2000 grafted on poly(dichlorophosphazene) backbone]. The 1H NMR spectra of these copolymers were recorded on an Avance DMX500 spectrometer using tetramethylsilane (TMS) as an internal reference at 25 °C. The molecular weight of the resultant polyphosphazenes was determined using size exclusion chromatography with a Waters 515 HPLC Pump and a Waters 2410 refractive index detector. Tetrahydrofuran (THF) was used as a solvent (flow rate of 1.5 mL/ min at 40 °C), and narrow-disperse polystyrene standards were employed for calibration. The chemical structure of polymers is illustrated in Figure 1, and the detailed compositions of polymers are shown in Table 1. Polymers were purified by dissolving precipitation twice in ether and dialysis against pure water for 2 days. Potentiometric Titration Measurements. Potentiometric titration measurements were performed using a Mettler Toledo pH meter coupled to an InLab 423 combined pH electrode. The polymer solutions (0.5 mg/mL, pH 3.0) were prepared, and potentiometric titration curves were obtained by monitoring the pH increase in the range of 3.0−11.0 as a function of 0.1 mol/L NaOH (increments of 0.05 mL in a 10 mL starting volume of polymer-containing solution). Hemolysis Assay. Membrane-lytic activity of the polymers was examined by hemolysis assay.24,25 Isosmotic phosphate buffers (100 mM) in the pH range of 5.0−7.4 were prepared. Polymer solutions
base polymers represent a promising new strategy in the design of optimal tumor targeting, anti-MDR drug carriers. However, while the phenomena of endosomal disruption and MDR inhibition by these polymers are well-documented, much less is known about the mechanisms that underlie these properties. Diisopropylamino (DPA) is a weakly basic, pH-sensitive chemical moiety. DPA can be protonated in slightly acidic solutions, so DPA-bearing polymers can imitate cationic polymers to rupture biomembranes within endosomal pH ranges. Researchers previously constructed DPA-bearing block copolymers based on methacrylate monomers, and drug carriers derived from these polymers exhibit pH-responsive drug release profiles.20 However, vinyl polymers are not biodegradable, and thus, the fate of polymeric carriers in vivo is unclear. Similarly, the capability of DPA-based pH-responsive polymers to disrupt biomembranes and inhibit P-gp is also not well characterized. Recently, we have conjugated DPA moieties onto amphiphilic polyphosphazene, which notably displays biodegradability and biocompatibility characteristics. Furthermore, drug carriers derived from this polymer series are capable of releasing endocytosed drug into the cytoplasm of tumor cells21 and significantly suppress the drug resistance of tumor cells.21,22 Here, we synthesized a series of graft polyphosphazenes containing DPA and PEG segments with different graft ratios and investigate the role of DPA played in the cell membrane disruption and P-gp inhibition behaviors of DPAgrafted polyphosphazenes. The mechanisms of these effects are discussed.
■
MATERIALS AND METHODS
Materials. Hexachlorocyclotriphosphazene (Acros Organics) was purified by sublimation at 80−90 °C. Poly(dichlorophosphazene) was synthesized by ring-opening polymerization at 250 °C. DPA was purchased from Alfa Aesar and was dried using molecular sieve 4A. 1110
dx.doi.org/10.1021/mp200356w | Mol. Pharmaceutics 2012, 9, 1109−1117
Molecular Pharmaceutics
Article
for 4 h. The intracellular metabolized product, MTT formazan, was retrieved by adding dimethyl sulfoxide (DMSO) and incubating at room temperature for 10 min. The plates were read at 550 nm, and the cell viability was calculated. The Dox resistance of adriamycin-resistant MCF-7 cells was validated by cytotoxicity assay. The IC50 of Dox in these cells after 48 h of incubation is 16-fold lower than that in adriamycin-sensitive MCF-7 cell lines. Monitoring Cell Membrane Penetrability. For confocal microscopy observation, the MCF-7/adr cells were seeded into sixwell culture plates (1.6 × 105 cells/well) with a cover glass on the bottom of each well and incubated for 24 h. After the media were discarded, AlexaFluor 546 solutions with or without polymer at different pH values and AlexaFluor 546 loaded polymer nanoparticles were added to the appropriate cell plates. At the predetermined time, the media were discarded, and the cells were washed three times with PBS (pH 7.4) to remove noninternalized samples. Microscope slides were then imaged with a Zeiss LSM 510 confocal microscope. AlexaFluor 546 loaded nanoparticles were prepared as following: AlexaFluor 546 was first combined with the DPA-contained polymer in 0.5 mL of DMSO and stirred for 1 h. Then, the solutions were transferred into dialysis bag and dialyzed against Milli-Q distilled water for 24 h. A 0.45 μm microporous filter was used to eliminate the dust, and the content of AlexaFluor 546 in the micelles was determined by spectrofluorometry at excitation/emission ∼543/570 nm. Rhodamine-123 Cellular Accumulation Studies. Solutions of 2 μg/mL R-123 in assay buffer (122 mM sodium chloride, 25 mM sodium bicarbonate, 10 mM glucose, 10 mM HEPES, 3 mM potassium chloride, 1.2 mM magnesium sulfate, 1.4 mM calcium chloride, and 0.4 mM potassium phosphate dibasic, pH 7.4) with various concentrations of copolymers or 6 μg/mL CsA were equilibrated for a minimum of 30 min before use. MCF-7/adr cells (1 × 105/well) were seeded in 96-well culture plates and incubated in RPMI 1640 medium for 24 h. The cell culture medium was then replaced with assay buffer and incubated for 30 min at 37 °C. After preincubation, the assay buffer was removed, and the R123/polymer or R123/CsA solution was added to the wells. After 60 min of incubation, the supernatant was removed, and cells were washed with ice-cold PBS and lysed with PBS containing 1.0% Triton X-100. The R-123 concentration in the supernatants was measured using a Flexstation II Fluoresce microplate reader [excitation wavelength (λex), 485 nm; emission wavelength (λem), 530 nm] and normalized for cellular protein levels as determined by BCA assay (Pierce). All experiments were carried out in quadruplicate. P-gp Labeling. A sandwich technique was used for labeling P-gp. For labeling internal antibody, 1 × 106 MCF-7/adr cells per 100 μL were fixed in methanol at −20 °C for 15 min. After they were washed three times with pH 7.4 PBS containing 3% BSA, cells were permeabilized with Triton X-100 (0.1%, w/v) for 30 min and washed again three times. Preparations were then incubated for 40 min with the primary antibody (specific anti-P-gp JSB-1, diluted to 1:50) in the
Table 1. Molecular Characterizations of PEG/DPA-PPPs molar fraction label
xa
ya
pKb
molecular weightb (g/mol)
PPD-10 PPD-6.7 PPD-3.0
0.18 0.26 0.50
1.82 1.74 1.50
6.20 6.45 6.71
10100 10400 17200
a
Calculated from 1H NMR analysis. bThe GPC molecular weights were calculated as Mn, PDI = 1.5−2.8. were prepared with different concentrations and pH values and were allowed to equilibrate for 24 h before use. Mouse red blood cells (RBCs) were washed twice with 150 mM NaCl in aqueous solution and once more with PBS solutions of specific pH values. RBCs were then resuspended in 150 μL of polymer solution and plated into a 96well culture plate (1 × 108 RBCs/well). Two samples of RBCs in buffer alone and in deionized water were used as negative control and positive control, respectively. Each test was performed in triplicate. The samples were incubated at 37 °C for 0.5 h and then centrifuged at 4000 rpm for 10 min. The absorbance of the supernatant from each sample was measured at 550 nm using a microplate reader (Thermo MK 3), and the percentage of hemolysis was determined for each well. Preparation of Dox-Loaded Nanoparticles. To encapsulate the Dox, the weighted Dox·HCl was dissolved in dimethylformamide along with the copolymer in the presence of triethylamine.21 Then, the solution was transferred into a dialysis bag and dialyzed against Milli-Q distilled water for 12 h. A 0.45 μm microporous filter was used to eliminate the dust or Dox precipitates. The obtained solution was lyophilized and stored in desiccator before used. The content of Dox in the nanoparticles was determined by spectrofluorometry using UV− vis spectrometry at 483 nm by the premeasured calibration curve. The drug content was calculated using eq 1, where the “mass of Dox-loaded nanoparticles” refers to total weight of Dox-loaded nanoparticles (Dox + polymer). Dox loading content (LC) mass of Dox encapsulated in nanoparticles = × 100% mass of Dox‐loaded nanoparticles
(1)
Cell Line and Cytotoxicity Evaluation. Adriamycin-resistant MCF-7 human breast cancer cell lines (MCF-7/adr) were seeded in a 96-well culture plate (10000 cells/well) and incubated for 24 h. After preincubation, the cell culture medium was replaced with fresh medium. Serial dilutions of free Dox, Dox-loaded nanoparticles, polymers with different amounts of calf serum, and polymers with different concentrations at different pH values were added to the prepared cell plate. At the predetermined time, media were discarded, and 3-(4,5-dimethylthiazol-2-yl)-2,5- diphenyltetrazolium bromide (MTT) solution was added prior to a further incubation at 37 °C
Figure 2. Potentiometric acid−base titration curves of different pH-responsive polymers at 0.5 mg/mL polymer solutions: (a) PPD-10, (b) PPD-6.7, and (c) PPD-3.0. 1111
dx.doi.org/10.1021/mp200356w | Mol. Pharmaceutics 2012, 9, 1109−1117
Molecular Pharmaceutics
Article
presence of 3% BSA and 100 μg/mL polymer. After it was washed three times, a secondary antibody, FITC-labeled goat antimouse IgG (diluted to 1:1000), was applied for 40 min. For labeling external antibody, 1 × 106 cells/100 μL of PBS (containing 3% BSA) were incubated for 40 min with 1:50 diluted anti-P-gp MRK 16. After they were washed three times, the cells were incubated with FITC-labeled goat antimouse IgG (diluted to 1:1000) for another 40 min. After they were washed again three times with PBS (pH 7.4), the cells were fixed with 4% paraformaldehyde and resuspended in 0.5 mL of PBS (pH 7.4). Measurement of FITC labeling was by flow cytometry (FC500MCL flow cytometer, Beckman-Coulter), with ∼10000 cells counted per sample. Excitation was at 488 nm, and emission was at 520 nm. To evaluate the background fluorescence intensity of residual secondary antibody, MCF-7/adr cells treated by the same procedures used for P-gp labeling but in the absence of JSB-1 and MRK-16 were employed as a background control group. Statistical Analysis. Statistical analysis was performed using oneway analysis of variance (ANOVA). Differences between groups were considered statistically significant at P < 0.05 (*).
■
Figure 3. Relative hemolysis of mouse RBCs by polymers at different pH values. The polymer concentration is fixed at 1 mg/mL. Error bars represent standard deviation (n = 3).
RESULTS AND DISCUSSION Synthesis of pH-Responsive Polyphosphazene. The structure and 1H NMR spectra of resultant pH-responsive polyphosephazene are given in Figure 1, and the 31P NMR spectra exhibit a similar broad main peak at ∼3.0 ppm. The molar ratio of PEG/DPA was calculated from 1H NMR data by comparing the peak intensities of the methylene protons of the ethylene oxide units of PEG at 3.4 ppm to the methyl protons of DPA at 1.0 ppm, which was in a good agreement with the feed ratio. The detailed compositions of polymers are shown in Table 1. pH-Responsive Characterization of Polymers. The potentiometric measurements were performed to investigate the pH sensitivity of polymers. As shown in Figure 2, each of the DPA-based polymers went through a sharp protonation transition in the pH range of 6.0−7.0. The pKb values of these polymers decreased slightly with increasing DPA content. pH-Triggered and Concentration-Dependent Hemolysis. As the previous study has shown, a correlation between the hemolytic efficiency and the endosomal disruption,26−28 the standard hemolysis assay was applied to assess the potential of DPA-grafted polyphosphazenes polymers to disrupt endosomal membrane at the specific pH in this study. First, we examined the relative hemolytic activity of DPA-grafted polyphosphazenes as a function of pH at a polymer concentration of 1.0 mg/mL. As shown in Figure 3, PPD-10 and PPD-6.7 induced significant degrees of hemolysis (>20%) over the pH range of 6.60−6.82. PPD-10 and PPD-6.7 reached their maximum degrees of hemolysis near to pH 6.6 and 6.7, respectively, which seems to be influenced by the pKb value of each polymer. In contrast, PPD-3.0 (which has the lowest DPA content) did not display any hemolytic activity over the test pH range. Next, we investigated the degree of hemolysis of RBCs incubated in solutions with varying polymer concentrations at pH values fixed at pH 6.6 and 6.7 for PPD-10 and PPD-6.7, respectively (corresponding to the maximum hemolytic pH of each polymer). As shown in Figure 4, the relative hemolytic activity of DPA-grafted polyphosphazenes is determined by the DPA content of polymer. The minimal concentration of PPD-10 for the maximum degree of hemolysis (>80%) is significantly lower than that of PPD-6.7. Several polymers were reported as being disruptive to biomembranes or possessing hemolytic activity,10−14 and these polymers were observed to interact with lipid bilayers of biomembranes in different ways. For example, cationic
Figure 4. Relative hemolysis of mouse RBCs as a function of polymer concentration for PPD-10 (at pH 6.6) and PPD-6.7 (at pH 6.7), respectively. Error bars represent standard deviation (n = 3).
polymers (e.g., poly(ethylene imine) (PEI) 25K) produce holes or pores in biomembranes in a wide pH range (Figure 3), either by solubilizing phospholipids or by inducing bilayer rearrangement (known as “flip-flop”). Different from polycations, DPA-grafted polyphosphazenes display hemolytic activity only in a certain pH range. At hemolytic pH, PPD-10 and PPD-6.7 polymers are partially protonated, with the protonated DPA groups anchoring polymer chains on the negatively charged lipid bilayer and the unprotonated DPA moieties creating multiple hydrophobic segments on the polymer chains. These hydrophobic segments can insert into lipophilic regions of lipid bilayers, causing membrane solubilization or altering the permeability of the membrane, hence, inducing hemolysis.13,14 However, if the pH value is so low that the majority of DPA groups are protonated, the polymer will instead be absorbed on the external leaflet of the lipid membrane with no adverse impact on membrane strength. This explains why the PPD-3.0 polymer (which has low DPA content) is devoid of hemolytic activity. However, it should be noted that PPD-3.0 has high PEG grafting density, which may also mask hemolytic activity. To explore the mechanism by which DPA-grafted polyphosphazenes rupture biomembranes, we investigated the hemolytic activity of PPD-6.7 in the presence of different 1112
dx.doi.org/10.1021/mp200356w | Mol. Pharmaceutics 2012, 9, 1109−1117
Molecular Pharmaceutics
Article
additive agents at a fixed concentration of 10 mM. As shown in Figure 5, polyethylene glycol of molecular mass greater than 5
Furthermore, polyethylene glycols with molecular masses ranging from 1 to 10 kDa have already been found to protect RBCs from mechanical damage,31 which implies that polyethylene glycols may enhance the mechanical strength of the lipid bilayer. The inhibition of hemolysis by polyethylene glycols may result from a concerted action via these two mechanisms, but defining this process is beyond the scope of the current study. The changes in RBC shape were studied using an optical microscope, and representative images of RBCs are shown in Figure 6. The lipid membranes of erythrocytes treated with 1.0 mg/mL PPD-6.7 at pH 6.7 were completely destroyed. Interestingly, RBCs treated by the same way but in the presence of 10 mM PEG8000 were found to aggregate and fuse into flowerlike, hemoglobin-containing structures (Figure 6c). This phenomenon suggests that PPD-6.7 changes the fluidity of RBC membranes, while the PEG8000 protects RBCs from hemolysis. Figure 6d shows a diagrammatic representation of how this combination of changes in membrane fluidity and protection from hemolysis may affect RBC morphology. pH-Induced Disruption of Plasma Membrane. The plasma membrane of living cells is much more complex than the erythrocyte membrane model used here. To investigate the disruptive effects of PPD-6.7 on cell membranes, we employed the plasma membrane-impermeable dye, AlexaFluor 546. As shown in Figure 7, considerable intracellular fluorescence can be detected in MCF-7/adr cells incubated with 1000 μg/mL of PPD-6.7 at pH 6.7. In contrast, cells incubated with 1000 μg/ mL PPD-6.7 at pH 7.4 or with blank PBS at pH 6.7 exhibit only minimal intracellular fluorescence. Interestingly, once AlexaFluor 546 was encapsulated in PPD-6.7 nanoparticles, a large amount of AlexaFluor 546 stained the whole cells (Figure 7d), which indicates that this membrane-impermeable dye succeeded in the phagocytosis by MCF-7/adr cells via polymeric nanoparticles and then fled from acidic later endosomes/ lysosomes with the help of PPD-6.7 owning membrane disruptive ability at weak acid pH. These results are accordance in our previous study of doxorubicin intracellular distribution in MCF-7/adr cells via DPA-contained polymeric nanoparticles.21 Considering that such alteration in cell membrane permeability would likely cause cytotoxicity, the cytotoxic
Figure 5. Effect of additives on the relative hemolysis of mouse RBCs induced by PPD-6.7 solution at pH 6.7. Error bars represent standard deviation (n = 3).
kDa can efficiently inhibit RBC hemolysis induced by PPD-6.7 at pH 6.7. This result is very similar to a previously reported study by Chen et al.,13 indicating that protonated DPA-grafted polyphosphazenes induce hemolysis by puncturing the RBC membrane and thus altering its permeability. These small holes allow small molecular weight salts to freely diffuse through the RBC membrane,29,30 while high molecular weight proteins inside RBCs (e.g., hemoglobin) remain trapped. Hence, the isotonic balance between the intracellular and the extracellular fluids of RBC is disturbed. Additionally, the polymer− membrane interaction also reduces the mechanical strength of the RBC membrane, weakening it sufficiently that the osmotic pressure derived from hemoglobin can rupture it, despite the miniscule pressure difference induced by the proteins (
