Polyethyleneimine-associated polycaprolactone—Superparamagnetic iron oxide nanoparticles as a gene delivery vector Min-Cheol Kim,1 Meng Meng Lin,2 Youngjoo Sohn,3 Jwa-Jin Kim,4 Bo Sun Kang,5 Do Kyung Kim6 1
Department Department 3 Department 4 Department 5 Department 6 Department 2
of of of of of of
Mechanical Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139 Chemical Engineering, Tsinghua University, Beijing 100084, People’s Republic of China Anatomy, College of Korean Medicine, Kyung Hee University, Seoul 130-701, South Korea Anatomy, College of Medicine, Konyang University, Daejeon 302-718, South Korea Radiological Science, Konyang University, Daejeon 302-718, South Korea Pharmacology, College of Medicine, Konyang University, Daejeon 302-718, South Korea
Received 8 April 2015; revised 17 August 2015; accepted 23 August 2015 Published online 00 Month 2015 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/jbm.b.33519 Abstract: This study describes the synthesis of novel gene delivery vector with low toxicity and high transfection efficiency for magnetofection. The rational design of magnetofection vector called PPMag (PEI-associated polycaprolactone (PCL)-SPIONs) composed of oleic acid (OA) stabilized superparamagnetic iron oxide nanoparticles (SPPIONs) prepared by thermolysis of iron oleate with a combination of hydrophobic PCL and proton absorbing polymer polyethyleneimine (PEI) (PEI-PCL-SPIONs) is described. Encapsulation of amphiphilic PEI with SPIONs not only improves water dispersity of SPIONs, but also allows nucleic acid (NA) condensation and endosomal/lysosomal escape via proton sponge effect after internalization in cells. MTT cytotoxicity assay showed that cell viability was improved compared to conventional PEISPIONs. The luciferase activity of magneto-polyplexes treated
cells significantly improved compared to both controls revealed that transfection efficiency of PPMag- pCIKlux polyplexes group was improved compared to naked pCIKlux group. The application underneath of a rare earth magnet significantly improve the transfection efficiency (i.e., the luciferase activity doubles) compared to cells without magnet, indicating that sedimentation induced by magnetic field plays important role in accumulation of magneto-polyplexes on cell surfaces. The results demonstrate that PPMag can be used as a novel gene transfection vector to improve transfecC 2015 Wiley Periodicals, Inc. J Biomed Mater Res tion efficiency. V Part B: Appl Biomater 00B: 000–000, 2015.
Key Words: superparamagnetic, iron oxide, polyethyleneimine, polycaprolactone, gene delivery
How to cite this article: Kim M-C, Lin MM, Sohn Y, Kim J-J, Kang BS, Kim DK. 2015. Polyethyleneimine-associated polycaprolactone—Superparamagnetic iron oxide nanoparticles as a gene delivery vector. J Biomed Mater Res Part B 2015:00B:000–000.
INTRODUCTION
The goal of gene transfection is to deliver nucleic acids (NA) into the cells using nonviral vectors to achieve overexpression of an introduced therapeutic gene and/or down regulation of an endogenous mutated gene. Nonviral vectors for NA delivery should not only protect the heterogeneous genetic materials from enzymatic degradation, but also mimic the infection mechanism of viral vectors through the cellular barriers to the nucleus during the transfection process.1 The criteria of an ideal transfection vector are low cytotoxicity, low immunogenicity, high transfection efficiency, and capability to be monitored in in vivo gene therapy. Cationic lipid (e.g., LipofectamineTM 2000) and cationic polymers (e.g., polyethyleneimine, PEI)2,3 are the most commonly used vectors for transfection. PEI is capable of DNA condensation and protection from enzymatic attack in the cytosol, by forming polyplexes with
DNA prior to transfection.2 The key feature of PEI is the proton sponge effect for escape of endosomal/lysosomal pathway and increased transfection efficiency. Despite the low immunogenicity and low cost of a PEI based gene carrier, the major problem is balancing gene delivery efficiency and cytotoxicity to patients.4,5 New strategies have been suggested and shown improved overall transfection efficiency, such as modifying the chemical structures of PEI6 or combination with other targeting/diagnostic moieties to construct multifunctional nanodevices.7,8 NA vectors associated with magnetic particles have been developed as a novel strategy where a physical parameter such as the magnetic force is utilized to increase molecular contact of vectors with target cells and thereby transduce DNA for biomedical applications.9 Mykhaylyk et al. developed “Magnetofection” protocols for viral and nonviral gene
Correspondence to: Do Kyung Kim; e-mail:
[email protected] Contract grant sponsor: National Research Foundation of Korea (NRF), Ministry of Education, Science and Technology; contract grant number: NRF-2014R1A1A4A03005726
C 2015 WILEY PERIODICALS, INC. V
1
delivery. These magnetic complexes composed of magnetic particles, nucleic acids and cationic lipids or polymers are sedimented onto the surface of the cells to be transfected within minutes by the application of a magnetic gradient field. The transfection using magnetic particles can be improved using an external magnetic field. The transfection efficiency is enhanced several thousand times compared with conventional transfections carried out with nonmagnetic gene vectors.10 Magnetofection takes advantages of both biochemical (cationic lipids or polymers) and physical forces (magnetic forces) to accelerate NA delivery into the target cells.11 DNA complexed cationic SPIONs were first bind with high affinity to lipid groups on the surface membrane and effectively internalized via receptor-mediated endocytosis or nonspecific endocytosis. Once they enter the acidic endosomal environment, the unsaturated amino groups (e.g., PEI coating) sequester the pumped-in-protons by the ATPase on the endosomal membrane and lead to retention of Cl- ions and water molecules in the endosome. This process keeps the pump functioning and leads to the retention of one Cl2 ion and one H2O molecule per proton. As the water retention increases, the endosome swells and bursts resulting in release of genetic materials and magnetic particles into the cytoplasm.12 There are two methods to associate NA with the magnetic vectors: formation of lipoplexes or polyplexes followed by association with the SPIONs13 and direct association of genetic substances with cationic SPIONs via electrostatic interaction.1 Li et al. demonstrated that transfection efficiency of PEI/DNA polyplexes was 35- to 85-fold higher when covalently linked with a commercial available magnetic nanobead.13 Similarly, Xenariou et al. reported a 300fold increase in reporter gene expression of Lipofectamine 2000/plasmid DNA (pDNA) lipoplexes, when complexed with TransMAGPEI (consisting of SPION and PEI coating) under a magnetic field at suboptimal DNA concentrations.14 However, the potentially toxic cationic lipid or polymers are not eliminated in these approaches, and translation into in vivo applications is still difficult. The second magnetofection strategy involves construction of cationic SPIONs that directly associate with naked genetic substances via electrostatic interaction. Plank’s group, who invented magnetofection, has recently described the advanced design by self-assembly of DNA with cationic lipid or polymers-associated SPIONs for in vitro transfection of both adherent and suspension cells with an external magnetic force.1 Namiki et al. reported small interference RNA (siRNA) delivery with LipoMag, which consists of OASPIONs and cationic lipid layer, in mice gastric tumor models.15 Cohorny et al. reported that formation of small clusters of OA-SPIONs is induced by coating of polylactide, subsequently PEI/oleate ion pair was formed on the surface of the nanospheres by an emulsion-evaporation method.16 Currently, PEI based SPIONs are directly prepared by insitu coprecipitation within the PEI matrix or step-wise modification of presynthesized SPIONs with PEI.17 The SPIONs prepared by coprecipitation method are usually 56 nm in
2
KIM ET AL.
diameter, poorly crystallized and have low magnetization that result in weak and heterogeneous response to the available laboratory magnetic arrays. Several authors attempted to form small clusters of SPIONs by back-filling them into mesoporous silica structures18 or bridging them together with hydrophobic polymers,16 followed by electrostatic deposition of PEI onto surfaces of the SPION clusters for DNA condensation and protection, but no further progress has been reported. Similar to this work, Sersa et al. developed surface modified SPIONs with a combination of polyacrylic acid (PAA) and polyethylenimine (PEI) (SPIONsPAA-PEI) proved to be safe and effective for magnetofection of cells and tumors in mice. The hierarchical structure of the particles shows effective for in vitro magnetofection of different cells with pDNA encoding either GFP or cytokine interleukin 12 (IL-12), and even superior in transfection efficacy than some other nonviral transfection approaches.19 To improve the cytotoxicity and transfection efficiency, here we suggested a novel construction of gene delivery vector with PEI-associated polycaprolactone (PCL)-SPIONs, referred to as PPMag. Moreover, there is no report on preparation of magnetic gene transfection vector with SPIONs prepared by thermal decomposition to our knowledge. EXPERIMENTAL
Materials Iron (III) nitrate nonahydrate (Fe(NO3)3 9H2O, 97.0%), cis9-Octadecenoic acid (CH3(CH2)7CH@CH(CH2)7COOH, OA, 90.0%), hexane (99.0%), ethanol (95%), 1-octadecene (CH3(CH2)15CH@CH2, ODE, 90.0%), sodium hydroxide pellet (NaOH, 99.9%) and branched polyethyleneimine 25K (PEI, 99%), polycaprolactone (PCL, average Mw 14,000, average Mn 10,000) were purchased from Sigma-Aldrich Korea, Seoul, South Korea. All the chemicals were used without further purification except ODE. ODE was purified by heating at 2008C for 3 h prior to use to remove absorbed water and organic impurities with a low boiling point. Double distilled and deionized water (ddH2O) was used throughout chemical synthesis and sample preparation for characterization. Sterilized ddH2O was used for cell-related experiments. Dulbecco’s Modified Eagles Medium (DMEM), RPMI medium, fetal bovine serum (FBS), 200 mM L-glutamine, penicillin/streptomycin and amphotericin B were purchased from Biosera, Nuaille, France. 3-[4, 5-dimethylthiazol-2-yl]-2, 5-diphenyl tetrazolium bromide (MTT) assay kit and luciferase assay kit were purchased from Promega Korea, Seoul, South Korea. LipofectamineTM 2000 was purchased from Life Technology Korea, Seoul, Korea. One-pot synthesis of SPIONs Hydrated FeOl complexes were prepared with FeCl36H2O and sodium oleate (NaOl) as starting materials in the mixture of hexane, ethanol, and water as solvent, which was published elsewhere.20 One mmole (0.404 g) of Fe(NO3)39H2O and a desired amount (1–5 mmol) of OA and 5 mL of ODE were dissolved in a three-neck round-bottom flask. For ease of reference sample temperature and collected for analysis. By increasing the amount of OA the
POLYETHYLENEIMINE-ASSOCIATED POLYCAPROLACTONE
ORIGINAL RESEARCH REPORT
and a similar milky brownish emulsion was obtained. The organic solvent was removed by vacuum for 4 h at room temperature. The brownish PEI-PCL magnetic particles (denoted as PPMag) were finally dispersed in water by sonication. The NaOH-treated particles oleate stabilized PCL magnetic particles (denoted as Ol-PMag) were also dispersed in water. All samples were stored at 148C for further characterization and experiments.
FIGURE 1. The schematic illustration of the proposed structure of PPMag. The blue spherical represent individual core SPION that are simultaneously coated with a layer of OA prepared in thermolysis of nonhydrated FeOl complexes; the pink part represent the hydrophobic PCL shell that bridging several core particles; finally phase transfer agent PEI was associated to carboxyl head groups of outer layer of OA via formation of PEI/oleate iron pair. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
nonhydrated FeOl complexes tend to be more tacky solids; The mixture was heated to 1208C for 1 h to evaporate the physically absorbed water and HNO3. After 30 min an air condenser was connected for refluxing. The reaction mixture was then heated to 3208C for 30 min with a heating rate of 3.38C/min under magnetic stirring, resulting in the reddish brown solution turning into black. The lipid-coated SPIONs were dissolved in 5 mL hexane, precipitated by 10 mL ethanol and centrifuged at 14,000 rpm for 10 min, after which the supernatant was carefully decanted. The washing process was repeated five times and the lipophilic magnetic nanoparticles were redispersed in 5 mL hexane, forming a stable magnetic fluid and kept at 48C for characterization and further experimentation. Phase transfer from organic solvent to aqueous system with PEI The structure of PPMag nanosphere is illustrated and explained in Figure 1. Freshly prepared OA-SPIONs (Fe content 0.1 mmol), 25 mg OA and 50 mg polycaprolactone (PCL) were re-dispersed in 4 mL chloroform/tetrahydrofuran (THF) mixture (1 : 1), followed by incubation and occasional shaking of the reaction mixture for 2 h at room temperature. 100 mL 25K PEI was added to the mixture and shaken for another 2 h at room temperature. 10 mL ddH2O was added to the solution and the mixture was sonicated for 15 min to form a milky brownish emulsion. For comparison of the effect of different bases (PEI and NaOH), 10 mL 0.1N NaOH solution was added to the mixture and sonicated for 15 min in parallel
Characterization The size and morphology of nanoparticles was examined on two transmission electron microscopes (TEM, JEOL 2100F (200 kV) and 1230 (120 kV), Tokyo, Japan) for example, the physical sizes of iron oxide cores were measured on JEOL 2100F (200 kV); and both core and shell structures were observed under JEOL 1230 (120 kV). TEM samples were prepared by placing a few drops of nanoparticle suspension onto a carbon-coated grid and air drying under ambient conditions. For physical size measurement, the diameter of >100 nanoparticles were measured on digital TEM images using image analysis software ImageJ. Fourier transform infrared (FTIR) spectra were recorded at 208C using an Alpha FTIR Spectrometer (Bruker Optics, ALPHA FTIR Spectrometer, Karlsruhe, Germany) equipped with Platinum ATR (single reflection diamond attenuated total reflectance). The samples were dried in oven at 508C overnight and grinded into fine powder with a motor and pestle prior to measurement. Spectra were measured with a resolution of 1 cm21 and the wavenumber range was 500–4000 cm21. Thermal gravimetric analysis (TGA), differential scanning calorimetry (DSC) and differential thermal analysis (DTA) analysis were carried out using Q600 simultaneous DSCDTA-TGA system (TA Instruments, Q600, New Castle, USA). The Q600 system was controlled by Thermal Advantage and Universal Analysis software for data acquisition and analysis. In data plots the weight loss was expressed as a percentage of the initial sample weight and plotted vs. temperature. Nitrogen purge gas (100 mL/min) were used to as protection gas during heating of samples (5–10 mg) in platinum pans with pierced lids (TA Instruments). Samples were measured from room temperature to 6008C at a nominal heating rate of 108C/min. Samples were dried in an oven at 508C overnight and grinded into fine powder with a motor and pestle prior to measurement. The hydrodynamic size of nanoparticles was determined by photon correlation spectroscopy (PCS), (Malvern, Zeta-Sizer HA300, Worcestershire, UK). Samples were dispersed in ddH2O with iron concentration of 10 lg/mL. For zeta-potential measurement, all samples were diluted with ddH2O into 10 lg/mL iron concentration, and the pH was adjusted with 0.1N NaOH or 0.1N HCl to the range of pH3 to pH12, respectively. Acute cytotoxicity determined by MTT assay NCI-H292 (human lung mucoepidermoid carcinoma) cells were cultured in DMEM supplemented with 10% fetal bovine serum (FBS), 2 mM L-glutamine and 1% penicillin/ streptomycin at 378C in a humidified atmosphere at 5% CO2
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | MONTH 2015 VOL 00B, ISSUE 00
3
FIGURE 2. Schematic illustration of magnetofection using PPMag as gene delivery vector. (a) Luciferase reporter gene encoding plasmid DNA (insertion: the backbone design of pGL4 vector, incorporated a variety of features: firefly luciferase gene (Luc 2) and selectable mammalian markers) was quickly mixed with Magneto-PEI to form polyplexes under physiological conditions; (b) Magnetofection using a static magnetic field underneath the cell culture plate to enhance the transfection efficiency; (c) expression of luciferase gene; (d) chemical reaction to yield light which can be quantified by an fluorescence spectrophotometer upon addition of luciferase assay substrate (luciferin). (Figure was modified based on information from Ref. 39).
in 25 cm2 cell culture flasks. Medium was changed every 2 or 3 days until 90% confluence was achieved. The cytotoxic effect of samples at various concentrations for up to 5-day incubation time were assessed by 3-[4, 5-dimethylthiazol-2-yl]-2, 5-diphenyl tetrazolium bromide (MTT) assay. MTT assay was a cell proliferation assay based on the ability of a mitochondrial dehydrogenase enzyme in viable cells to cleave the tetrazolium rings of the pale yellow MTT and form a dark blue formazan crystal. The number of surviving cells was directly proportional to the level of the formazan product created, which can then be quantified by reading absorbance at 570 nm. Cells were incubated with PPMag at a serise particle concentration from 20 lg/mL to 500 lg/mL. Totally, 10 lL MTT solution was added to the medium and the mixtures were incubated for 4 h at 378C, followed by addition of 100 lL DMSO to dissolve the purple crystals with gentle pipetting. Absorbance was measured at 570 nm. Control cells were cultured with complete medium only. Each experiment was carried out in triplet and repeated twice. The results were analyzed and plotted in Origin 7.0; one-way analysis of variance (ANOVA) was performed to compare the means of each sample groups, in which Tukey test was used as means comparison test, and Levenell test was used as equal variance test. Transfection efficiency determined by Luciferase assay For transfection studies, NCI-H292 cells were cultured in RPMI medium supplemented with 10% FBS, 100 U/mL pen-
4
KIM ET AL.
icillin, 0.1 mg/mL streptomycin, 0.25 mg/mL amphotericin B, and 2 mM L-glutamine. Cells were grown in 25 cm2 tissue culture flasks (T 25) with 15 mL of complete medium that was changed every two or three days until 90% confluence was achieved. Cells were seeded at 2 3 104 cells per well in 96-well tissue culture plates and incubated overnight at 378C with 5% CO2. The transfection protocol was illustrated in Figure 2. Briefly, PPMag particles were suspended in serum free RPMI tissue culture medium at 70 mg mL21, and combined with pCIKlux plasmid DNA at a final concentration of 14 mg mL21. The samples were left at room temperature for 20 min to allow the DNA to bind to the particle surface. The media was removed from the cells and replaced with 0.1 mL serum free RPMI containing 0.5 mg PPMag particles associated with pCIKlux pDNA. At 2 h post transfection, the media was removed from each well and replaced with 0.1 mL RPMI medium supplemented with 10% FBS, 100 U/mL penicillin, 0.1 mg/mL streptomycin, 0.25 mg/ mL amphotericin B, and 2 mM L-glutamine. Lipofectamine 2000 (Invitrogen) transfection was performed using 0.1 mg DNA per well following the manufacturers’ recommended protocols. Magnetofection was performed for 10 min by applying the static magnetic field with NeFeB magnets (sintered NdFe-B magnets (NeoDelta; remanence Br, 1080–1150 mT) directly under the 96-well plates. Dimensions for magnetofection for 96-well plate format: cylindrical; d 5 6 mm,
POLYETHYLENEIMINE-ASSOCIATED POLYCAPROLACTONE
ORIGINAL RESEARCH REPORT
FIGURE 3. (a) TEM and (b) HR-TEM images of OA-SPIONs (dispersible in hexane) with insertion of diffraction pattern and (c) TEM and (d) HRTEM images of PPMag (dispersible in water).
h 5 5 mm, inserted in an acrylic glass template in 96-well plate format with strictly alternating polarization. Untransfected (media only) and naked DNA (no transfection agents) conditions were included as controls. At 48 h post transfection the media was removed from each well. The cells were washed once with phosphate buffered saline (PBS) and then lysed in 0.03 mL of reporter lysis buffer (Roche). Each sample was assayed for luciferase activity using a luciferase assay system (Promega) following the manufacturer’s recommended protocol. The total protein concentration was determined using a BCA protein assay (Pierce, USA). RESULTS
Morphology and particle size TEM and High-Resolution Transmission Electron Microscopy (HR-TEM) images of SPIONs prepared via thermolysis of non-hydrate iron oleate complexes before and after PEI assisted phase transfer are shown in Figure 3. Before phase transfer, the OA-SPIONs were readily dispersed in organic solvents (i.e. hexane and chloroform). Monodispersed OASPIONs with readily spherical shapes with an average diameter of 10 nm were obtained. The insertion showed the diffraction pattern of OA-SPIONs, indicating the existence of gFe2O3 phase. PPMag particles were completely dispersed in water without agglomeration and the particle size of PPMag
was unchanged during the phase transfer process compared to the OA-SPIONs. Both HR-TEM images before and after phase transfer showed the lattice was determined to be 0.25 nm, indicating the crystal structures of SPIONs are not influenced during the post-synthetic processing. FTIR analysis Figure 4 shows the FTIR spectra of freshly prepared samples of OA-SPIONs, Ol-PMag, and PPMag. Iron oxide nanoparticles prepared by thermolysis of nonhydrated FeOl are denoted as OA-SPIONs. Subsequently, hydrophobic particles which are transferred into water via NaOH titration are denoted as Ol-PMag; and particles transferred via PEI are denoted as PPMag. In the FTIR spectrum of OA-SPIONs, the band appearing at 1709 cm21 is assigned to C@O stretching bands of surface bound OA. A very weak shoulder appeared at 1656 cm21, possibly due to the C@C stretching of OA.21 Three peaks at 1588 cm21, 1554 cm21, and 1523 cm21 are assigned to asymmetrical carboxylate vibration; whereas the 1456 cm21, 1434 cm21, and 1406 cm21 are assigned to symmetrical carboxylate vibration.22,23 The difference of asymmetrical and symmetrical carboxylate position is an indicator of the binding mode of oleate ligand on the iron oxide surface, according to Bronstein et al. these oleate ligand bind to SPIONs via both bidentate and bridging modes.23 In Figure 1, a hydrophobic polymer PCL chain was
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | MONTH 2015 VOL 00B, ISSUE 00
5
organic coating is 2408C and all the organic substance is burned out at 4508C leaving the remaining iron oxide cores attributed to 30.0% of total mass. The DSC profile was analyzed together with TGA curve. Four shallow insignificant endothermic peaks can be distinguished: the first one occurs at 1408C21458C, possibility due to the loss of both physically or chemically bounded water; the second endothermic event takes place at 2118C; the third endothermic peak appears at 3528C which can be assigned to the decomposition temperature of the PEI coating, in close agreement to Td (3578C) of PEI reported elsewhere;5 the fourth peak occurs at 4568C which can be explained as the OA decomposition temperature, Td of OA in PPMag slightly right-shifted, compared to pure OA (Td 5 4498C).
FIGURE 4. FTIR spectra of PPMag, Ol-PMag, and OA-SPIONs in the wavenumber range of 4000 cm21 to 600 cm21. (The FTIR spectra wavenumber range from 3000 cm21 to 2000 cm21 was omitted because no peaks were detected in this range, and ma and ms means asymmetric and symmetric stretching respectively). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
interdigitated with the hydrocarbon tail of OA whereas carboxylic functional group of OA is attached to the SPIONs surfaces. From the region of 2000 cm21 to 600 cm21 of the spectrum of Ol-PMag, there is an intensive band that appears at 1709 cm21, which is assigned to the C@O stretching mode of dimeric OA.21 A small shoulder also appears at 1,736 cm21, corresponding to the C@O vibration mode of carbonyl groups in crystalline PCL.24 There is no particular step for PCL crystallization on SPION surface in this work. It is possible that the amorphous PCL carbonyl band at 1726 cm21 is difficult to distinguish due to the strong interference of 1709 cm21 carboxylic band.24 In the region of 1600 cm21 to 1400 cm21, both 1588 cm21, and 1543 cm21 are assigned to symmetrical carboxylate stretching, whereas the three bands appearing at 1456 cm21, 1436 cm21, and 1412 cm21 are assigned to asymmetrical stretching of oleate ligand. The bend at 1240 cm21 can be assigned to the CAO stretching of the oleate ligand.23 The spectrum of PPMag shows enormous differences with OlPMag. Since the carboxylic functional group of the second layer of OA interacts with the amine groups of PEI electrostatically, the characteristic PCL and OA bands may not be easily distinguished. The intensity of 1736 cm21 carbonyl band decreases dramatically possibly due to the full coverage of PEI on the particle surfaces.5,7 The new band at 1656 cm21 can be assigned to stretching vibration of ammonium cation (d(NH1)).7 Thermal properties Figure 5 shows both TGA and DSC profiles of PPMag. PPMag does not show significant step-wise degradation. It can only be claimed that the onset degradation temperature of the
6
KIM ET AL.
Colloidal properties Figure 6 shows the zeta-potential of Ol-PMag and PPMag from pH 5 3–12. For dispersion of Ol-PMag in aqueous system a NaOH solution was used to create the electrostatic emulsion during phase transfer. The zeta-potential of OlPMag shows a clear reverse pH-dependency of surface charges in the acidic pH range, the isoelectric point (pHiep) was determined as pHiep 5 4.3. At low pH range of pH 5 3– 4, the surface charges are almost neutral, because the carboxylic head groups exist in the form of carboxylic acid, thus these particles are not practically soluble in water. In the range of pH 5 5–8, zeta-potential value decreases as pH increases, indicating the outer layer of OA react with OH- in the system and yield carboxylate group to display a negatively charged surface. The higher the pH is, the more OH2 exists in the system for conversion of neutral OA to negatively charged oleate ligand on the surface. When the pH is above 8, zeta-potential starts to become constant within experimental errors, in the range of 235 mV to 240 mV, regardless of the pH change. The data is consistent with the observation that Ol-PMag precipitate very quickly when pH < 8.
FIGURE 5. TGA and DSC profiles of PPMag as a function of temperature form room temperature to 6008C, with a heating rate of 108C and N2 protection gas flow with a rate of 100 mL min21. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
POLYETHYLENEIMINE-ASSOCIATED POLYCAPROLACTONE
ORIGINAL RESEARCH REPORT
FIGURE 6. Zeta-potential of PPMag (no pHiep) and Ol-PMag (pHiep 5 4.3) against pH from pH3 to pH12. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
The zeta-potential of PPMag before and after binding the DNA at various ratios of particles to DNA are given in Figure 7. Before DNA binding, the zeta-potential of PPMag is found to be highly positive 1 50 6 0.7 mV at pH 7.4, indicating that PEI coating provides both electrostatic and steric repulsion. Zeta-potential before and after complexing with DNA was shown in Figure 7. When the ratio of the particle to DNA is 2/1, the zeta-potential of polyplexes are 1 27.2 6 2.1 mV, which may be not charged enough to achieve maximum transfection. When the ratio of particle to DNA is increased to 5/1, 10/1, and 20/1, the zeta-potential of polyplexes is increased to 1 36.1 6 0.9 mV, 1 37.4 6 0.9 mV, and 1 39.4 6 1.8 mV, respectively, which are suitable for magnetofection studies. Cytotoxicity and transfection efficiency assays Figure 8(a) shows the cell survival rate after 24-h incubation with PPMag at a series of different concentration up to 500 mg mL21. Clearly, PPMag exhibits the dose-dependent pattern of cytotoxicity of NCI-H292 cells. At a low concentration (20 mg mL21) of PPMag, 104% 6 0.5% of cells survived. The cell survival rate decreased by increasing the amount of PPMag particles. When the concentration of PPMag exceeds 200 mg/mL, the cell survival rate decreased to less than 55.5% 6 0.3% which is not suitable for magnetofection. Figure 8(b) shows cytotoxicity of NCI-H292 cells by applying the static magnetic field of 0.3T NeFeB Magnets. From the results, external magnetic field will not induce the cytotoxicity at the series of different times up to 120 min.
small clusters and aggregation hence required further size selection process.25 In our rational step-by-step design of PPMag, SPIONs cores are firstly produced by thermolysis, with good dispersion and arrangement in organic solvent with high magnetization. Table I shows the comparison between PEI-SPIONs25 and PPMag. The major drawback of thermal decomposition is the simultaneous coating of hydrophobic OA or oleylamine which hinders the nanomedicinal applications.26–28 Biocompatible and biodegradable PCL16,29 was selected for further coating on OA-SPIONs. The second layer of OA binds to the PCL layer via hydrophobic interaction between PCL chains and the carbon tail of OA and leaves the carboxylic groups to interact with the environment solution. Finally, branched PEI is associated to the second layer of OA by formation of a PEI/oleate ion pair, hence some of the amines are occupied to reduce the electrostatic binding with NA to facilitate the NA unpacking and release in the cytoplasm to achieve maximum gene transfection. One of the advantages of PPMag is that PEI is associated to the particle surface via electronic interactions with negatively charged oleate as a balancing strategy for attenuation of primary amines of PEI. Generally primary amine coated particles can protect DNA from enzymatic attack efficiently and have the strongest DNA binding capacity. However, high concentration of primary amines often results in high cytotoxicity and slow DNA dissociation upon entering the acidic organelles (i.e., endosome and lysosomes). Gabrielson et al. had demonstrated a 20- to 60-fold increase in gene delivery efficiency when DNA binding capacity is significantly reduced by the introduction of carboxylic groups to PEI.30 Zintchenko et al. have been able to induce remarkable gene knockdown by succinylation and acetylation of primary amines in branched PEI.31 Yezhelyer et al. reported synthesis of proton sponge coated quantum dots (QDs) with balanced composition of carboxylic and tertiary amine on the surface of QDs to achieved high gene silencing efficiency
DISSCUSSION
Steitz et al. reported the preparation of PEI-SPIONs by a two-step process: SPIONs fabricated by water-based coprecipitation and subsequent PEI physical absorption, and demonstrated some promising results in gene transfection.25 Unfortunately, the size of SPION cores are limited to 5– 10 nm with relatively low magnetization, and often result in
FIGURE 7. Zeta-potential of PPMag before and after DNA complexion at pH 5 7.4.
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | MONTH 2015 VOL 00B, ISSUE 00
7
FIGURE 8. Histogram of (a) NCI-H292 cells viability after 24-h incubation of PPMag at a series particle concentration from 20 mg mL21 to 200 mg mL21 (measured by iron content) (b) magnetic cytotoxicity of NCI-H292 cells by applying the static magnetic field of 0.3T NeFeB Magnets at the series of different times up to 120 min. The data are represented as means 6 SD from four experiments. **p < 0.01 and ***p < 0.001 vs. control. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
with minimum cytotoxicity.8 It is possible to optimize the proton sponge coating with carboxylic groups to reduce cytotoxicity and accelerate DNA unpacking and endosomal release to achieve maximum delivery. As discussed previously, PEI binds to OA strongly via electrostatic interaction; therefore any molecules that contained primary or secondary amines can assist phase transfer. In this work, PEI was selected as a phase transferring agent due to its proton sponge effect for gene delivery and the fact that PEI provides high electrostatic and steric repulsion for the separation of nanoparticles in water. The pHdependency of zeta-potential is a supplementary evidence for full oleate coverage of Ol-PMag. Even though it is possible to dissolve Ol-PMag in water, the particle solubility and stability at physiological pH is still needed to improve by introducing a proton-sponge coating layer for both steric and electrostatic repulsion. Upon mixing PPMag and plasmid DNA (pDNA) at room temperature, the cationic amines of PEI bind to the negatively charged plasmid DNA backbone via electrostatic interactions driving the formation of magneto-polyplexes. The zeta-potential is dramatically reduced [mt]10 mV at the entire range of the ratio of particles to DNA, suggesting that DNA was successfully bound to the PPMag. It is clearly indicated that the reduction of the surface charges of magnetopolyplexes increases as the DNA amount is increased. The cellular accumulation of the polyplexes strongly rely on the surface charges, the magneto-polyplexes with a zetapotential of more than 1 30 mV is not only a display of good colloidal stability in water, but also show noticeable increased cellular uptake attributed to the electrostatic interaction with the negatively charged cell membrane.32 Therefore, zeta-potential values shed some light for optimization of ratio of particles to DNA for optimum transfection. Grayson et al. reported cells showed significant gene knockout when treated with polyplexes with 1 35 mV surface, while gene knockout was undetectable when polyplexes with negative surface charge were used.33 No maximum DNA binding (saturation) was observed for the ratios of particles to DNA used in this experiment;
8
KIM ET AL.
however when the ratio of particle to DNA reaches 20/1, aggregations started to form in the solution, indicating such high ratio of particles to DNA is not suitable for cellular transfection experiments. The above mentioned particle aggregations are possibly attributed to the free PPMag (without DNA binding) and negatively charged glucose. Similarly, Funhoff et al. also reported surface charges of polyplexes become steady when the ratio of polymer to DNA reaches a certain value.34 It has to be noted that this is nonspecific cellular uptake, because PPMag lacks of a biological targeting moiety. For gene delivery, it worth more investigation in engineering cell-type specific or nucleus specific delivery vectors by conjugation of targeting agents, such as the Arg-Gly-Asp (RGD) peptide or nucleus location signal (NLS) peptide for a more sophisticated method of gene delivery.35 Yezhelyer et al. reduced cytotoxicity of transfection agents by balancing composition of carboxylic and tertiary amine of PEI on CdSe/ZnS QD surface.30 Kievit et al. reported that the cell survival rate of PEI-SPIONs treated C6 rat gliomas cells for 24 h was 12.4%; however, the cell viability of PEI attached PEG-chitosan copolymer functionalized SPIONs was 90%, indicating very little cytotoxicity effect.36 In contrast with artificial SPIONs, bacterial magnetic particles BMPs were found to attenuate the cytotoxicity of linear PEI significantly in Chinese hamster cells (CHO), and human cervical adenocarcinomas cells.37 Mannosylation of PEI was also found to mask the high density of primary amine of PEI to significantly reduce the cytotoxicity of PEI stabilized silica nanoparticles.38 It is possible to optimize PEI with carboxylic groups to reduce cytotoxicity and accelerated NA unpacking and endosome release to achieve maximum delivery. To compare the transfection efficiency of PPMag with commercially available ones, NCI-H292 cells were transfected by PPMag and Lipofectamine 2000 with or without static magnetic field. In Figure 9, the negative control cells without any treatment showed no luciferase activity as expected, and the control cells treated with PCIKlux DNA only showed trace luciferase activity, indicating plasmid
POLYETHYLENEIMINE-ASSOCIATED POLYCAPROLACTONE
ORIGINAL RESEARCH REPORT
TABLE I. A List of Different Properties of PEI-SPIONs25 and PPMag
Core particle morphologies Core particle sizes Surface chemistry Water dispersity Proton-sponge coating method Enzymatic protection DNA/RNA binding capacity DNA/RNA unpacking Cytotoxicity
Steitz (PEI-SPIONs)
In this work (PPMag)
Often clustered aggregation 6 nm Limited amount of –OH groups Water-based preparation, can be used directly in aqueous system Physical absorption Weak association and easy to dissociate
Perfect dispersion and arrangement Up to 15 nm, can be tuned Simultaneous coating of OA Oil-based preparation, need surface engineering and phase transfer Electrostatic interaction with OA
Very strong Very efficient Slow High
Less Less efficient Increased Reduced
DNA itself cannot penetrate the cell plasma membrane, since both plasmid DNA backbone and cell plasma membrane are negatively charged and repelled each other. LipofectamineTM 2000 is cationic lipid-based transfecting agent capable of transfecting with very high efficiency in a broad range of mammalian cells that has been widely used for the past decade.18 Transfection efficiency by LipofectamineTM 2000 was unchanged with or without an external magnetic field, as expected. The luciferase activity of magnetopolyplexes treated cells significantly improved compared to both controls, suggesting that it is possible to use PPMag as a gene carrier for magnetofection. The application underneath of a rare earth magnet significantly improve the transfection efficiency (i.e., the luciferase activity doubles) compared to cells without magnet, indicating that sedimentation induced by magnetic field plays important role in accumulation of magneto-polyplexes on cell surfaces. However, only 883.4 6 11.2 RLU mg21 of transfection efficiency was achieved by our PPMag compared to
LipofectamineTM 2000 (2711.5 6 87.5 RLU mg21). The chemical compositions, the vector-to-DNA ratio and the experimental conditions (i.e., incubation time and cell number) had to be optimized to improve transfection efficiency. CONCLUSIONS
The purpose of this study was to develop a magnetic gene transfection agent with high transfection efficiency with a minimal cytotoxicity. Oleic acid (OA) stabilized SPIONs prepared by thermolysis were used as the magnetic cores of PPMag. PEI coating was associated with OA-SPIONs via electrostatic interaction with carboxylic groups of oleate ligands. MTT cytotoxicity results showed that the cell viability of PPMag group (75.3% 6 0.9%) was higher than PEI-SPIONs prepared by conventional method (53.6% 6 1.2%) at the fixed iron content 20 mg mL21, possibly due to the primary amine attenuation of PEI coating; hence, the minimization of the cell membrane damages during transfection. Preliminary luciferase assay results showed limited gene delivery efficiency (883.4 6 11.2 RLU/mg) of PPMag, compared to LipofectamineTM 2000 (2,711.5 6 87.5 RLU/mg) as positive control. However, the experimental conditions need to be optimized, including the incubation time and the concentration of PPMag. REFERENCES
FIGURE 9. Histogram of luciferase activity in NCI-H292 cells transfected with pCIKlux DNA using PPMag and LipofectamineTM 2000. All transfection were performed in 96-well cell culture plates using 0.1 mg DNA/ well for 10 min transfection times. (Data were normalized to protein content.) The data are represented as means 6 SD from four experiments. ***p < 0.001 vs. tatic magnet. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
1. Mykhaylyk O, Antequera YS, Vlaskou D, Plank C. Generation of magnetic nonviral gene transfer agents and magnetofection in vitro. Nat Prot 2007;2:2391–2411. 2. Kichler A. Gene transfer with modified polyethylenimines. J Gene Med 2004;6:S3–S10. 3. Yang J, Lee C-H, Ko H-J, Suh J-S, Yoon H-G, Lee K, Huh Y-M, Haam S. Multifunctional magneto-polymeric nanohybrids for targeted detection and synergistic therapeutic effects on breast cancer. Angew Chem Int Ed 2007;46:8836–8839. 4. Fisher D, Li D, Ahlemeyer B, Krieglstein J, Kissel T. In vitro cytotoxicity testing of polycations: influence of polymer structure on cell viability and hemolysis. Biomaterials 2003;24:1121–1131. 5. Shuai X, Merdan T, Unger F, Kissel T. Supermolecular gene delivery vectors showing enhanced transgene expression and good biocompatibility. Bioconjugate Chem 2005;16:322–329. 6. Huang H, Tang G, Wang Q, Li D, Shen F, Zhou J, Yu H. Two novel non-viral gene delivery vectors: Low molecular weight polyethylenimine cross-linked by (2-hydroxypropyl)-beta-cyclodextrin or (2hydroxylpropyl)-alpha-cyclodextrin. ChemComm 2006:2382–2384. 7. Peterson H, Fechner PM, Fisher D, Kissel T. Synthesis, characterization and biocompatibility of polyethylenimine-graft-
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | MONTH 2015 VOL 00B, ISSUE 00
9
8.
9.
10.
11. 12.
13.
14.
15.
16.
17.
18.
19.
20.
21. 22.
23.
24.
10
poly(ethylene glycol) block copolymers. Macromolecules 2002;35: 6867–6874. Yezhelyev MV, Qi L, O’Regan RM, Nie S, Gao X. Proton-spongecoated quantum dots for siRNA delivery and intracellular imaging. J Am Chem Soc 2008;130:9006–9012. Govindarajan S, Kitaura K, Takafuji M, Ihara H, Varadarajan KS, Patel AB, Gopal V. Gene delivery into human cancer cells by cationic lipid-mediated magnetofection. Int J Pharma 2013;446(1–2): 87–99. Mykhaylyk O, Antequera YS, Vlaskou D, Plank C. Generation of magnetic nonviral gene transfer agents and magnetofection in vitro. Nat Protocols 2007;2:2391–2411. Dobson J. Magnetic nanoparticle-based gene delivery. Gene Thera 2006;13:283–287. Nel AE, Madler L, Velegol D, Xia T, Hoek EM, Somasundaran P, Klaessig F, Castranova V, Thompson M. Understanding biophysicochemical interactions at the nano-bio interface. Nat Mater 2009; 8:543–557. Li W, Ma N, Ong LL, Kaminski A, Skrabal C, Ugurlucan M, Lorenz P, Gatzen H-H, L€ utzow K, Lendlein A, P€ utzer BM, Li RK, Steinhoff G. Enhanced throacic gene delivery by magnteic nanobeadmediated vector. J Gene Med 2008;10:897–909. Xenariou S, Griesenbach U, Ferrari S, Dean D, Scheule RK, Cheng SH, Geddes DM, Plank C, Alton EWFW. Using magnetic forces to enhance non-viral gene transfer to airway epithelium in vivo. Gene Thera 2006;13(1545–1552). Namiki Y, Namiki T, Yoshida H, Ishii Y, Tsubota A, Koido S, Nariai K, Mistunaga M, Satoru Y, Hideyuki K, et al. A novel magnetic crystal-lipid nanostructure for magnetically guided in vivo gene delivery. Nat Nanotechnol 2009;4:598–606. Cohorny M, Polyak B, Aferiev IS, Walsh K, Friedman G, Levy RJ. Magnetically driven plasmid DNA delivery with biodegradable polymeric nanoparticles. FASEB J 2007;21:2510–2519. rez-Yag€ Mulens-Arias V, Rojas JM, Pe ue S, Morales MP, Barber DF. Polyethylenimine-coated SPIONs trigger macrophage activation through TLR-4 signaling and ROS production and modulate podosome dynamics. Biomaterials 2015;52:494–506. Yiu HHP, McBain SC, El Haj AL, Dobson J. A triple-layer design for polyethylenimine-coated magnetic particles and their use in DNA binding and transfection. Nanotechnology 2007;18:25–29. Prijic S, Prosen L, Cemazar M, Scancar J, Romih R, Lavrencak J, Bregar VB, Coer A, Krzan M, Znidarsic A, et al. Surface modified magnetic nanoparticles for immuno-gene therapy of murine mammary adenocarcinoma. Biomaterials 2012;33:4379–4391. Lin M, Kim D. In situ thermolysis of magnetic nanoparticles using non-hydrated iron oleate complex. J Nanoparticle Res 2012;14:1– 13. Shukla N, Liu C, Jones PM, Weller D. Oxidation of FePt Nanoparticles. J Magn Magn Mater 2003;266(1-2):178–184. Mahajan SV, Dikerson JH. Synthesis of monodispersed Sub-3 nm RE2O3 and Gd2O3:RE31 nanocrystals. Nanotechnology 2007;18: 325605–325611. Bronstein LM, Huang X, Retrum J, Schmucker A, Pink M, Stein BD, Dragnea B. Influence of iron oleate complex structure on iron oxide nanoparticle formation. Chem Mater 2007;19:3624–3632. Huang L, Allen E, Tonelli AE. Study of the inclusion compounds formed between alpha-cyclodextrin and high molecular weight
KIM ET AL.
25.
26. 27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
poly(ethylene oxide) and polycaprolactone). Polymer 1997;39: 4657–4865. Steitz B, Hofmann H, Kamau SW, Hassa PO, Hottiger MO, von Rechenberg B, Hofmann-Amtenbrink M, Petri-Fink A. Characterization of PEI-coated superparamagnetic iron oxide nanoparticles for transfection: Size distribution, colloidal properties and DNA interaction. J Magn Magn Mater 2006;311:300–305. Sun S, Zeng H. Size-Controlled Synthesis of Magnetite Nanoparticles. J Am Chem Soc 2002;124:8204–8205. Hyeon T, Lee SS, Park J, Chung Y, Na HB. Synthesis of highly crystalline and monodispersed maghemite nanocrystallites without a size-selection process. J Am Chem Soc 2001;123:12798– 12801. Park J, An K, Hwang Y, Park J-G, Noh H-J, J-YK, Park J-H, Hwang N-M, Hyeon T. Ultra-large-scale synthesis of monodispersed nanocrystals. Nat Mater 2004;3:891–895. Beaune G, Dubertret B, Clement O, Vayssettes C, Cabuil V, Menager C. Giant vesicles containing magnetic nanoparticles and quantum dots: Feasibility and tracking by fiber confocal fluorescence microscopy. Angew Chem Int Ed 2007;46:5421–5424. Gabrielson NP, Pack DW. Acetylation of polyethylenimine enhances gene delivery via weakened polymer/DNA interactions. Bioconjugate Chem 2006;7:2427–2435. Zintchenko A, Philipp A, Dehshahri A, Wagner E. Simple modification of branched PEI lead to highly efficient siRNA carriers with low toxicity. Bioconjugate Chem 2008;19:1448–1455. Funhoff AM, Nostrum CF, Gerben AK, SchuurmansNieuwenbroek NME, Crommelin DJA, Hennink WE. Endosomal escape of polymeric gene delivery complexes is not always enhanced by polymers buffering at low pH. Biomacromolecules 2004;5:32–39. Grayson ACR, Doody AM, Putnam D. Biophysical and strctural characterization of polyethylenimine-mediated siRNA delivery in vitro. Pharm Res 2006;23:1868–1875. Funhoff AM, van Nostrum CF, Janseen APCA, Fens MHAM, Crommelin DJA, Hennink WE. Polymer side-chain degradation as a tool to control the destablization of polyplexes. Pharm Res 2004;21:170–176. Josephson L, Tung C-H, Moore A, Weissleder P. High-efficiency intracellular magnetic labeling with novel superparamagnetic-tat peptide conjugates. Bioconjugate Chem 1999;10:186–191. Kievit FM, Veiseh O, Bhattarai R, Fang C, Gunn JW, Lee D, Ellenbogen RG, Olson JM, Zhang MQ. PEI-PEG-chitosan-copolymer-coated iron oxide nanoparticles for safe gene delivery: Synthesis, complexation, and tranfection. Adv Funct Mater 2009;19: 2244–2251. Li X, Wang B, Jin H, Jiang W, Tian J, Guan S, Li Y. Baterial magentic particles (BMPs)-PEI as a novel and efficient non-viral gene delivery system. J Gene Med 2007;9:679–690. Park IY, Kim IY, Yoo MK, Choi YJ, Cho M-H, CHo CS. Mannsylated polyethylenimine coupled mesoporous silica nanoparticles for receptor-mediated gene delivery. Pharm Nanotechnol 2008;359:280–287. Invitrogen. Luciferase reporter assay kit manual and application notes. https://www.thermofisher.com/kr/ko/home/life-science/ protein-biology/protein-biology-learning-center/protein-biologyresource-library/protein-biology-application-notes/versatile-luciferases-tools-reporter-assays.html
POLYETHYLENEIMINE-ASSOCIATED POLYCAPROLACTONE
