as efficient reagent for siRNA transfection

0 downloads 0 Views 2MB Size Report
Dec 14, 2015 - Polyimine . Cationic polymer . Transfection . siRNA. Introduction. Gene therapy, as a potential method in medical diagnostics and treatment of ...
J Polym Res (2016) 23:10 DOI 10.1007/s10965-015-0898-9

ORIGINAL PAPER

Poly(ethylene glycol) modified poly(2-hydroxypropylene imine) as efficient reagent for siRNA transfection Alma Bockuviene 1 & Juste Balciunaite 1 & Kristina Slavuckyte 2 & Lolita Zaliauskiene 2 & Ausvydas Vareikis 1 & Ricardas Makuska 1

Received: 11 July 2015 / Accepted: 14 December 2015 # Springer Science+Business Media Dordrecht 2015

Abstract MPEG’ylated poly(2-hydroxypropylene imines) were synthesised using 1,3-diamino-2-propanol, 1,3dibromo-2-propanol and methoxy poly(ethylene glycol) (MPEG, Mn 1000, 2000, 5000). Trying to find conditions favourable for attachment of MPEG chains, MPEG iodide was either added to the initial mixture of the monomers (method A), or was introduced later, at the end of the reaction (method B). MPEG’ylated derivatives of PHPI were characterized by FTIR and NMR spectroscopy, SEC, DLS, DSC and potentiometric titration and were tested for DNA and siRNA delivery in vitro. Efficiency of high MW / low MPEG content PHPI derivatives in DNA transfection was similar to that of PHPI. MPEG grafting had a positive effect on siRNA delivery: high MW polymers with low content of MPEG performed significantly better than non-MPEG’ylated PHPI and low-molecular-weight polymers with high content of MPEG. The length of MPEG chains attached to the polymers had no effect over the transfection efficiency.

Keywords PEG’ylation . Polyimine . Cationic polymer . Transfection . siRNA

* Alma Bockuviene [email protected] * Lolita Zaliauskiene [email protected] * Ricardas Makuska [email protected] 1

Department of Polymer Chemistry, Vilnius University, Naugarduko 24, LT-03225 Vilnius, Lithuania

2

Thermo Fisher Scientific Baltics, Graiciuno 8, LT-02241 Vilnius, Lithuania

Introduction Gene therapy, as a potential method in medical diagnostics and treatment of genetic disorders such as immunodeficiency, cystic fibrosis, neurodegenerating diseases or cancer, has gained essential attention over the past decade [1–7]. Over the last three decades, both viral and non-viral gene delivery techniques have been investigated intensively. Viral vectors have been proven to generate high transfection efficiency in most cell lines. However, the limitations associated with viral vectors in terms of cytotoxicity, low transgene size and high cost, have limited their use [8, 9]. These limitations are serious hurdle if thinking about in vivo applications of viral vectors. Non-viral vectors lack the pathogenicity of viruses, are cheaper and can carry large inserts, but are limited by low transfection efficiency. Importantly, the ideal non-viral vector should not only exhibit low cytotoxicity and immune response, but also enable to achieve excellent gene expression [10, 11]. Various non-viral vectors have been described as gene delivery systems, including polymeric vectors such as poly[2-(dimethylamino)ethyl methacrylate)] [12, 13], polyamidoamine [14, 15], poly(ethylene imine) (PEI) [16], poly(L-lysine) [17], poly(amino methacrylate) [18] and carbohydrate-based [19, 20] polymers. Among these polymeric materials, PEI is one of the most successful and efficient non-viral vector used both for in vitro and in vivo gene delivery [21–26]. Favourable properties of PEI allow protection of the genetic material in the extracellular environment and promote disruption of endosomes and subsequent release of nucleic acids into the cell interior. PEI has constitutional repeating unit comprising two carbons followed by one nitrogen atom, and containing primary, secondary and, in the case of branched PEI, tertiary amino groups, all of which have potential to be protonated [26]. Despite these advantages, the use of

10

J Polym Res (2016) 23:10

Page 2 of 12

PEI experiences several limitations the main being rather high toxicity of PEI-polyplexes [21]. Although transfection efficiency of PEI-based complexes increases with molecular weight of PEI, high-molecular-weight PEI (25 kDa) exhibits significant cytotoxicity [24]. Low-molecular-weight PEI (1.8 kDa) is less cytotoxic, but its efficiency in gene transfection is lower [24]. PEI-induced cytotoxicity is mainly associated with strong positive charge of its macromolecules leading to strong interactions of PEI with cell surfaces and their damage [26]. To deal with the dilemma of choosing between high transfection efficiency and low cytotoxicity, PEI was modified by poly(ethylene glycol) (PEG) [27], poly(caprolactone) [28], polyester [29], and poly(L-lysine) [30]. PEG is one of the most suitable graft-forming polymers because of its unique physicochemical and biological properties, including hydrophilicity, solubility in water and organic solvents, lack of toxicity, ease of chemical modification and absence of antigenicity and immunogenicity [31]. Since PEG has been accepted as non-toxic, non-immunogenic and generally biocompatible polymer, it has been widely investigated for drug and gene delivery [32–35]. Several types of PEI-PEG copolymers have been synthesized with block, alternating or statistical composition, arranged in linear, branched, comb or star topology, and with varying molecular weight of PEG segments [36–39]. A number of synthetic routes for preparation of PEI-PEG copolymers have been reported in the literature. Most research groups employed commercially available activated PEG, such as methoxy-PEG-succinimidyl propionate [38], PEG diacrylate [40], sulfone-ω-N-hydroxysuccinimide ester of poly(ethylene glycol) [41], PEG isocyanate [42], etc. These PEG derivatives were directly used for coupling reactions with PEI. PEG length and molecular weight has a great influence on gene delivery systems. Regarding PEG‘ylation, both PEG chain length and graft density clearly affect properties of polyplexes. Some authors suggest that low graft density of high-molecular-weight PEG is more effective in reducing protein adsorption than higher surface density graft of lowmolecular-weight PEG [37]. Contrarily, others authors found that density of PEG grafting had more pronounced effect on protein adsorption than PEG chain length on steric repulsion and van der Waals attraction [41]. Notably, the stability, resistance to nuclease and in vitro knockdown efficiency of PEG’ylated polyplexes has been shown to depend greatly on the PEG chain length. In general, PEG provides polyplexes with improved solubility, lower surface charge, diminished aggregation, and of lower cytotoxicity [43, 44]. At first sight, siRNA and DNA have many common properties. They are both double-stranded nucleic acids, have anionic phosphodiester backbones with the same negative charge to nucleotide ratio and both can interact electrostatically with cationic polymers. However, siRNA and DNA have significant differences in molecular topography and complex size which play a role in complex uptake and transfection

efficiency. For instance, upon PEG’ylation of 25 kDa PEI, DNA transfection remained at the same level or was increasing, and complex uptake remained unchanged [44]. Meanwhile, other groups reported a decrease in the transfection efficacy for PEG-grafted PEI [22, 45–47], which likely was due to the reduction of binding sites on the complexes essential for the interaction with the negatively charged cell membrane [45]. Some studies on siRNA delivery showed that PEG grafting exerted negative effects on the intracellular release of nucleic acids [39], but other researchers hypothesized that PEG’ylation enhanced siRNA release in the cytoplasm, leading to an increased gene-suppression activities [41]. Actually, the parallel effects of PEG’ylation on size, charge, stability and cytotoxicity of the complexes may show different and opposing consequences regarding their bioactivity [44]. Furthermore, the effects of PEI PEG’ylation may be different for the complexes of (PEI-PEG)/DNA and (PEI-PEG)/siRNA [22, 41]. The differences may be determined by the size of DNA and siRNA molecules the former being more than 200fold larger, which has a significant impact on the molar ratios of PEI nitrogen to DNA or siRNA phosphate (N/P). Moreover, siRNA and plasmid DNA have major structural and functional differences, and need to be delivered to different cellular compartments, and these differences may be influenced by polymer/DNA or siRNA complex stability [48]. Taken together, this indicates that the effect of PEI PEG’ylation on efficiency of plasmid DNA transfection and siRNA delivery need to be studied separately. Poly(2-hydroxyalkylene imines) possessing both imine and hydroxyl groups in their repeating units have been synthesized recently and shown to be efficient gene transfection reagents [49, 50]. More detailed examination of poly (2hydroxypropylene imine) revealed that this polymer had no significant effect on tested cell viability and was very efficient gene delivery reagent, comparable or even better than commercially available transfection reagents Exgen500, Lipofectamine, and others. Unfortunately, at higher concentrations PHPI showed substantial cytotoxicity [50]. In the present paper, we report synthesis and study of MPEG’ylated derivatives of poly(2-hydroxypropylene imine) using MPEG with Mn 1000, 2000 or 5000. Novel branched MPEG’ylated PHPI containing terminal segments of methoxy poly(ethylene glycol) are expected to be non-cytotoxic and efficient reagents for DNA and siRNA transfection.

Experimental Materials Methoxy poly(ethylene glycols) (MPEG-1000, MPEG2000, and MPEG-5000), 1,3-diamino-2-propanol (DAP, 98 %), 1,3-dibromo-2-propanol (DBP, 95 %), N,N-

J Polym Res (2016) 23:10

dimethylacetamide (DMAC, 99 %) and dichloroethane (CH2Cl2, 99 %) were purchased from Sigma-Aldrich. Prior to use, DMAC was purified according to standard procedures and stored over molecular sieves (3 Å). Other reagents – Dubelcco’s Modified Eagle’s Medium (DMEM), RPMI-1640 medium, Trypsin-EDTA, L-glutamine, gentamicin sulfate and doxycycline were purchased from Sigma Aldrich, Fetal bovine serum (FBS) – from GE Healthcare Life Sciences, HeLa cells – from ATCC (LGS Standards, UK), HEK293iGFP cells were kindly provided by colleagues from Thermo Fisher Scientific, Lafayette, US. GFP-specific and control siRNAs were purchased from Ambion (Life Technologies), eGFP encoding vector was synthesized at GenScript USA Inc. All other reagents were of analytical reagent grade. Iodination of poly(ethylene glycol) monomethyl ether Methoxy poly(ethylene glycol) iodide (MPEGI) was synthesised according to the modified method of Rydon [51] with yields ranging from 81.3 % (MPEGI 1000) to 98.4 % (MPEGI 5000). The degree of iodination of MPEG was determined by 1H-NMR spectroscopy and was in the range of 0.86–1.03.

Page 3 of 12 10

slowly preheated to 60 °C and kept at that temperature for 2 h. Then, under vigorous stirring, DBP (2.06 g, 9.5 mmol) was added dropwise to the reaction mixture. Subsequently, temperature of the reaction mixture was raised up to 80 °C, and the condensation reaction was continued for 168 h. Method B DAP (0.9 g, 10 mmol), DBP (2.06 g, 9.5 mmol) and 8.5 ml of dry DMAC were placed in a 25 ml capacity round-bottom reaction flask submerged in an oil bath. Temperature of the reaction mixture was raised up to 80 °C, and the condensation reaction was continued for 144 h. Then MPEG iodide (1 mmol, according to degree of iodination) was added to the reaction mixture, and the reaction was continued for additional 24 h. In both cases the reaction products were dissolved in 100 ml aqueous 0.1 M NaCl and extracted with 100 ml CH 2 Cl 2 to remove unreacted MPEGI and MPEG (as impurity of MPEGI). Aqueous phase was separated and ultrafiltrated through 10 kDa membrane against water. Solid MPEG’ylated PHPI was isolated by freeze-drying.

Synthesis of PHPI Spectroscopic measurements PHPI was synthesized by the method described before [52]. DAP 1.49 g (16.6 mmol) and dry DMAC 8.1 mL were placed in a 25 ml capacity round-bottom reaction flask submerged in an oil bath. Under moderate stirring, the reaction mixture was slowly preheated to 80 °C. Then, under vigorous stirring, DBP (3.62 g, 16.6 mmol) was added dropwise to the reaction mixture, and the polycondensation was carried out for 168 h. The resulting solution was diluted with 0.15 M NaCl solution (100 mL) and ultrafiltrated through a Pellicon membrane (Millipore) with nominal cut-off 10 kDa against water. PHPI as a solid polymer was isolated by freeze-drying. Yield 0.99 g (19.5 %). 1H NMR (400 MHz, D2O) δ ppm: 2.36 (6H, NH-CH2-NH-CH 2-CH(OH)-CH2-), 3.45 and 3.62 (H, CH2-CH(OH)-CH2-). MPEG’ylation of PHPI MPEG’ylated derivatives of PHPI were synthesized using two different methods: Method A DAP (0.9 g, 10 mmol), MPEGI (1 mmol, according to degree of iodination) and 8.5 ml of dry DMAC were placed in a 25 ml capacity roundbottom reaction flask submerged in an oil bath. Under moderate stirring, the reaction mixture was

FT-IR spectra were recorded on a PERKIN-ELMER Frontier spectrometer using Universal ATR Sampling Accessory. 1H and 13C-NMR spectra were recorded on a Brucker 400 Ascend™ 400 MHz spectrometer at 29 °C in D2O. Content of oxyethylene units (OE, mol %) and MPEG content in MPEG’ylated PHPI was calculated from 1HNMR spectra comparing intensities of the signals of oxyethylene protons of MPEG (at 3.7 ppm, 4H) and methyne protons of PHPI (at 2.7–3.2 ppm, 1H). Degree of branching (N2/N3) of PHPI and PHPI segments in MPEG’ylated PHPI were calculated from 13C-NMR spectra [52]. Determination of molecular weight Molecular characteristics of the polymers were determined by size-exclusion chromatography (SEC). SEC analysis was performed on a Viscotek TDAmax system equipped with a triple-detection array TDA305 consisting of refractive index detector, right-angle and low-angle light scattering detectors, and four-capillary bridge viscosity detector. The universal column was A6000M (Viscotek). The system was eluted with 250 mM acetate buffer of pH 4 at a 0.5 ml/ min flow rate and temperature 30 °C. Sample concentration was 0.5 mg/ml. The molecular weight were calibrated

10

J Polym Res (2016) 23:10

Page 4 of 12

with poly(ethylene glycol) standards. OmniSEC software from Viscotek was used to collect and analyse the data. Refractive index increment dn/dc of MPEG’ylated PHPI in 250 mM acetate buffer (pH 4) at 30 °C was determined using differential refractometer Waters 410 and was 0.145 (for polymers of method A) and 0.184 (for polymers of method B).

Potentiometric titration Potentiometric titration was carried out at 25 °C using METLER TOLEDO MP220 pH-meter equiped with HANNA HI 1230 combined glass electrode. 0.1 g of ultrafiltrated and freeze-dryed MPEG-PHPI or PHPI was dissolved in 10 mL of 0.15 M aqueous NaCl and was titrated with 0.1 M HCl using a microburet of 2 mL cappacity. The same procedure was repeated using 0.1 M NaOH. pK a values were calculated by the method of Henderson-Hasselbalch . It was assumed that the degree of ionization of protonated amino groups at the acidic and basic equivalence points were equal to zerro and one, respectively. Degree of unprotonation of amino groups (β) at pH of cytosol of HeLA and HEK293iGFP was calculated from experimental data fitted to the HendersonHasselbalch equation. Buffering cappacity of MPEG’ylated PHPI was calculated as pH = pKa ± 1.

Calculation of number of MPEG segments in MPEG’ylated derivatives of PHPI Number of MPEG segments (k) per one average macromolecule was calculated solving the system of linear equations with two variables: 8 m⋅73 <   k⋅ M MPEG ¼M n   þ M PEG . M PEG þ m ⋅100 ¼ X k⋅ k⋅ : 44 44 Where: m k X MMPEG Mn 73 44

number of constitutional repeating units in PHPI segment, number of MPEG segments per one molecule of MPEG’ylated PHPI, fraction of oxyethylene (OE) units in MPEG’ylated PHPI calculated from 1H-NMR data, mol. %, molecular weight of MPEG, number average molecular weight of MPEG’ylated PHPI calculated from SEC data, molecular weight of constitutional repeating unit of PHPI, molecular weight of constitutional repeating unit of MPEG.

Cell culture and transfection Differential scanning calorimetry DSC measurements were performed on a Perkin Elmer DSC 8000. Scans were run in nitrogen atmosphere at a heating and cooling rate of 10 °C/min (temperature range −50 to 120 °C). Glass transition temperature (Tg) and melting temperature (Tm) of the polymers were derived from the second heating curve.

Dynamic light scattering Size of PHPI and MPEG’ylated PHPI in aqueous solutions was measured using Zetasizer Nano ZS (Malvern Instruments) with a 4 mW laser at wavelength 633 nm. Sample concentration was 1 mg/ml, temperature 25 °C, and the scattered light was registered at the angle 173°. The size distribution data were analysed by Malvern Zetasizer software 7.03. Solutions for DLS experiments were prepared as described in the section “Cell culture and transfection”, but 0.15 M aqueous NaCl was used instead of the serum-free medium; the measurements were performed at the time intervals 0, 20 and 60 min.

For DNA transfections, 1 day before the experiment HeLa cells were seeded in a 24-well tissue culture plate at the density of 6 · 104 cells per well in the total volume of 1 mL DMEM culture medium supplemented with 10 % FBS. The cells were incubated at 37 °C in a CO2 incubator until they reached 70–80 % confluency (usually within 24 h). Transfection complexes were prepared as follows: eGFP encoding plasmid DNA (0,5 μg) was diluted in 100 μL of serum-free medium; cationic polymers were deposited on the wall of the same Eppendorf tube and vortexed immediately for few seconds to ensure an even distribution of the material. DNApolymer mixtures were incubated for 15 min at room temperature and added to the cells in a dropwise manner. The final concentration of the cationic polymers in the cell culture (based on N) was 7,5/15 μM for PHPI, 60/90/120 μM for polymer no.1, 7.5/10/15 μM for polymers no.2 and no.4, 30/ 45/60 μM for polymer no.3, 1/2/3 μM for polymer no.5 and 2/3/4 μM for polymer no.6. For siRNA delivery, HEK293iGFP cells were seeded in a 48 well tissue culture plate at the density of 2·104 cells per well in the total volume of 250 μl of RPMI culture medium supplemented with 10 % FBS. GFP expression was induced with doxycycline (final concentration in cell culture - 1 μg/ml). The cells were incubated in a CO2 incubator for 24 h at 37 °C.

J Polym Res (2016) 23:10

Polymer-siRNA complexes were prepared as follows: nontargeting control or eGFP-specific siRNA were diluted in 50 μL of serum-free medium, cationic polymers were deposited on the wall of the tube and vortexed immediately; the mixtures were incubated for 15 min at room temperature and added to the cells in a dropwise manner. The final concentration of siRNA in cell culture was 25 nM, the final concentration of cationic polymers was 15 / 140 / 15 / 45 / 15 / 2 / 3 μM for PHPI and polymers no. 1 through 6, respectively. The transfection efficiency was evaluated 48 h later: cells were trypsinized, collected and mixed at 1:1 ratio with 2 % paraformaldehyde/PBS solution. Fixed cells were analysed by flow cytometry (acquiring 5000 events for each sample) using Guava EasyCyte8HT system and Guava CytoSoft 2.2.3 cell acquisition/analysis software (Millipore). Toxicity (the percent of dead cells) within the total population of analyzed cells was estimated based on forward and side scatter results smaller and/or more granular cells were gated out. The percent of GFP+ cells and fluorescence intensity was evaluated within the population of viable cells.

Results and discussion Synthesis and characterization of MPEG’ylated derivatives of PHPI MPEG’ylated derivatives of PHPI were synthesized by polycondensation of DAP and DBP in the presence of MPEGI [51] (Scheme 1). Trying to find conditions suitable for the synthesis of PHPI derivatives with different content of MPEG chains, MPEGI was added to the initial mixture of the monomers (method A), or it was introduced later, at the end of the reaction (method B). Moreover, MPEGI with different molecular weight (Mn 1000, 2000, and 5000) was used, which expanded possibilities to synthesize PHPI-MPEG derivatives with different structure, i.e. length and density of MPEG chains and molecular weight. PHPI derivatives were purified by extracting unreacted MPEGI along with residual MPEG with CH2Cl2, and low-molecular-weight fractions of the polymer were eliminated by ultrafiltration through 10 kDa membrane. Important parameters for delivery of DNA and siRNA are molecular weight and structure of cationic polymer [44, 53]. Molecular characteristics of MPEG’ylated derivatives of PHPI are summarized in Table 1. Yield of MPEG’ylated PHPI was low, usually less than 30 %. Especially low yield was characteristic for PHPI derivatives containing segments of MPEG 5000. Possibly, this is predetermined by sterically hindered reaction between high-molecular-weight MPEGI-5000 and amino groups of DAP. Molecular characteristics of MPEG’ylated PHPI depend also on the synthesis method. High content of MPEG segments exceeding 60 mol % was

Page 5 of 12 10

characteristic for PHPI derivatives synthesized by the method A. In this case, 1 to 1.25 MPEG chains were attached per one macromolecule of PHPI. Contrarily, the content of MPEG and the number of MPEG segments per one molecule of MPEG’ylated PHPI (k) obtained by the method B was considerably lower. Figure 1 presents molecular weight distributions (MWD) of MPEG’ylated PHPI. Obviously, molecular weight and MWD of the polymers depend on molecular weight of MPEG and, especially, on the method of the synthesis. Weightaverage molecular weight of MPEG’ylated derivatives of PHPI synthesized by the method A was rather low, from 8 to 53 kDa. The polymers synthesized by the method B (MPEGI was introduced at the end of the reaction) were characterized by considerably higher weight-average molecular weight (138–162 kDa) and very high dispersity (Mw/Mn up to 13.6) (Table 1). MWD curves of these polymers were bimodal showing the presence of the fractions with molecular weight up to several millions. Why characteristics of PHPI derivatives synthesized by the methods A and B differ so substantially? Using the method A, MPEGI was introduced at an early stage of the process. The polycondensation started when DBP was loaded, and proceeded at equimolar amounts of amino and halo- groups. Thus, MPEGI had a possibility to react with DAP from the early stage of the polycondensation. Moreover, MPEG is monofunctional compound and acts as a chain terminating agent. Therefore, polycondensation of DAP and DBP in the presence of MPEGI, as expected, generated the products with lower molecular weight and containing considerable amount of MPEG. Using the method B, poly(2-hydroxypropylene imine) was synthesized in the first step, and then it was modified by various MPEG iodides. During the first step, the reacting amino and bromo groups were not at equimolar ratio (the stoichiometric ratio 1:0.95). Under conditions of equilibrium polycondensation, such stoichiometric imbalance allows reaching the degree of polymerization at around 40 only (Mn about 6200). Polycondensation between alkyl diamines and alkyl dibromides is nonequilibrium (irreversible) because of the absence of interchange reactions between imine groups. Moreover, PHPI is only slightly soluble in DMAC, which can distort stoichiometry of the reacting groups even more and add the factor of the diffusion control. Under these conditions, formation of the polycondensation products with high molecular weight is possible and even expected [54]. In such a case, the polycondensation products should contain two fractions, with DP higher and lower than expected. At the end of the first step, the reaction mixture contained the solid and liquid phases which could be attributed to the fractions with high and low molecular weights, respectively. When these polycondensation products were treated with MPEGI, likely, only marginal amount of MPEG was attached to PHPI of high molecular

10

J Polym Res (2016) 23:10

Page 6 of 12

Scheme 1 Synthesis of MPEG‘ylated derivatives of PHPI

weight because of low content / bad accessibility of terminal aminogroups in this fraction and/or low mobility of terminal segments of PHPI in viscous media. The vast amount of MPEGI reacted with low-molecular-weight PHPI or even left unreacted. Since low-molecular-weight compounds were removed by ultrafiltration, PHPI derivatives obtained by the method B were characterized by high molecular weight, very high polydispersity and low content of MPEG. This explanation is supported also by low yield of the polycondensation products. Structure of MPEG’ylated derivatives of PHPI was established using 1H and 13C-NMR spectra. 1H NMR spectrum of PHPI modified by MPEG 2000 (Fig. 2) contains the signals attributed to -CH2CH2O- (3.63 ppm), -CH2-CH(OH)-CH2(4.09–4.22 ppm), and -NH-CH2-NH-CH2-CH(OH)-CH2(2.49–3.31 ppm) [51, 52]. 13C NMR spectra of MPEG’gylated PHPI indicate the presence of MPEG chains since they have the signals attributed to CH3O- (57.4 ppm) and -O-CH2-CH2(70 ppm) groups (Fig. 2). The signals attributed to PHPI segments evidence branched structure of the polymers: the signal at 42.28 ppm belongs to the carbon associated with the primary amino group (NH2-CH2-), at 50.8 ppm is related to the carbon attached to the secondary amine (-CH2-NH-CH2-CH(OH)CH2-), at 59.1 ppm is related to the carbon next to the tertiary

Table 1 Molecular characteristics of MPEG’ylated PHPI

Method

A

B

amino group (N(CH2-CH(OH)-CH2-)3), and that at 64.2– 63.8 ppm belongs to the secondary carbon associated with the hydroxyl group -CH2-CH(OH)-CH2-. Since tertiary amino group represents the branching point on PHPI, the degree of branching was calculated by the ratio of secondary to tertiary amino groups. It was determined that every second or third repeating unit had a branching point. The polymers synthesized by the method B appeared to be slightly more branched than those synthesized by the method A (Table 1). SEC analysis confirmed branched structure of PHPI segments: the values of the constant α in Mark-Houwink equation were very low varying from 0.29 to 0.38 (Table 1). Such values are typical for branched polymers [55]. The presence of MPEG segments in PHPI derivatives was confirmed also by FT-IR spectra and DSC measurements. FTIR spectra showed the absorption bands at 1100 cm−1 corresponding to the stretching vibrations of C-O-C in (-O-CH2CH2-), and 1470 cm−1 associated to deformation vibrations of -OCH3. DSC thermograms of PHPI derivatives with high content of MPEG showed melting at 31, 48 and 51 °C for the polymers no 1, 2 and 3, respectively. Partial crystallinity is a characteristic for MPEG [42] but not for PHPI [52]. DSC thermograms of PHPI derivatives with low content of MPEG didn’t show melting behaviour. Glass transition temperature

Sample

MPEG

q, %

OE, mol %

k

Mw, kDa

Mw/Mn

α

N2/N3

1 2 3 4 5 6

1000 2000 5000 1000 2000 5000

28.1 23.2 5.97 20.9 17.1 6.04

59 67 72 3.3 2.4 9.3

1.25 1.06 1.20 0.22 0.09 0.15

7.7 15.6 52.6 138 145 162

2.84 4.03 5.29 11.7 12.1 13.6

0.38 0.33 0.29 0.35 0.33 0.32

2.5 2.4 2.3 2.7 2.8 2.6

J Polym Res (2016) 23:10

Fig. 1 MWD of fractionated MPEG’ylated PHPI. Numbers on the curves correspond to the number of the samples in Table 1

(Tg) of MPEG’ylated derivatives of PHPI was at negative temperatures and ranged from −2 to −7 °C. Ionization of MPEG’ylated derivatives of PHPI The protonization ability of cationic non-viral systems is an important property for its success as a gene delivery vehicle.

Page 7 of 12 10

Cationic polymers are asumed to induce facilitated escape of polyplexes from endosomes due to the uptake of protons from lyosomal environment. Acid-base potentiometric titrations of MPEG’ylated PHPI gave titration curves (Fig. 3) very similar to those of PHPI [49, 50] and similar polymers [56]. For cationic gene delivery vehicle, its buffering capacity and degree of unprotonation β at physiological pH (pH 7.3) are very important characteristics with respect to endosomal escape of polyplexes and high transfection efficiency. PHPI and MPEG’ylated derivatives of PHPI are characterized by high slope values (n) in the coordinates of the HendersonHasselbalch equation (Table 2). This suggests that ionization of amino groups of these polymers is seriously affected by neighbouring already protonized imino groups. All the polymers studied exhibited very close pKa values (Table 2) with the only difference that the polymers synthesized by the method A appeared to be slightly more basic than those obtained by the method B. Probably, this is due to the differences in molecular weight of the polymers which influences their ionization capability. According to pKa values, PHPI derivatives synthesized by the method A are more similar to PHPI than those obtained by the method B. Proximity of pKa values of all MPEG’ylated PHPI derivatives indicates that MPEG segments have no appreciable influence on the basicity of PHPI

Fig. 2 1H-NMR (top) and 13C-NMR (bottom) spectra in D2O of PHPI modified with MPEG 2000 (sample 5 in Table 1)

10

J Polym Res (2016) 23:10

Page 8 of 12

Fig. 3 Acid-base titration curves of MPEG’ylated PHPI. Numbers on curves correspond to the sample No in Table 1

segments. Overall buffering capacity of PHPI derivatives covers pH range from 6.4 to 8.8, and the degree of unprotonation of amino groups at the pH of the cytosol (pH 7.3) is at about 39–48 %. These numbers indicate that MPEG’ylated derivatives of PHPI, if used as siRNA/DNA delivery vehicles, should be capable to accumulate protons from lysosomal environment and facilitate release of endosomal content into the cytosol [57, 58]. For instance, PEI is only partly protonated at physiological pH and therefore may accumulate protons from more acidic surroundings of lysosomes [58] (i.e., may act as a “proton sponge”). Dynamic light scattering is very useful tool for determination of particle size, size distribution and aggregation phenomena. These parameters are very important for cationic polymers since condensation of polycations with DNA and siRNA into small particles is prerequisite for gene delivery [44]. It is known that diameter of macromolecules of PHPI is at about 10 nm [49]. Unfortunately, there is no information on how dimensions of these polymers depend on pH. DLS investigation revealed that macromolecules of MPEG’ylated PHPI are coiled to a smaller diameter both in neutral and acidic environment, compared to nonMPEG’ylated counterpart (Fig. 4). This phenomenon might Table 2 Method

A

B

Results of potentiometric titration of MPEG’ylated PHPI Sample

pKa

n

β

Buffering capacity

1 2 3 4 5 6

7.8 7.8 7.6 7.5 7.4 7.5

3.1 2.8 2.8 3.0 2.9 2.8

41.2 39.4 43.4 45.6 48.3 45.6

6.8–8.8 6.8–8.8 6.6–8.6 6.5–8.5 6.4–8.4 6.5–8.5

be due to additional branching and/or screening imparted by MPEG segments. The size of the solvated macromolecules of PHPI and MPEG’ylated PHPI are not susceptible to pH, at least, in acidic and neutral solutions (Fig. 4). Such behaviour is unexpected, since at low pH amino groups are expected to be substantially protonated which leads to repulsions between chains and expansion of polymer coils. Likely, stability in dimensions of PHPI and MPEG’ylated PHPI in respect to pH is related to highly branched structure of these polymers: even under protonation polymer coils have no possibility to expand since the distance between branching points is very short. Stability in dimensions of MPEG’ylated PHPI in respect to pH is a valuable feature considering their use in gene delivery. DNA and siRNA complexation and delivery by MPEG’ylated PHPI derivatives PEG’ylation has been shown to influence the size, morphology and stability of DNA and siRNA polyplexes [43]. The characteristics of DNA or siRNA complexes with PHPI and its MPEG’ylated derivatives were examined using DLS technique (Fig. 5). Considering structural and size differences of DNA and siRNA, it was not surprising to see that plasmid DNA formed larger complexes (100–170 nm) compared to siRNA (40–70 nm). However, no obvious differences were observed in relation to the density of MPEG chains or molecular weight (MW) of tested polymers. The overall DNA polyplex size for polymer no. 3 (method A, longer PEG chains) was higher compared to other polymers (1 and 2) from the same group. The stability of particles was not significantly affected as well. Slight shift in size (DNA polyplexes) within 60 min time frame was observed for polymers no. 4 and 5, but not for polymer no 6 (synthesis by the method B).

J Polym Res (2016) 23:10

Page 9 of 12 10

Fig. 4 Volume-based DLS size distribution curves of PHPI (dashed line) and MPEG’ylate PHPI (solid line) (a - sample 2 and b - sample 5 in Table 1) recorded at different pH

Transfection experiments were carried out to assess the ability of MPEG’ylated polymers to deliver DNA or siRNA into the cell. For DNA delivery, enhanced green fluorescent protein (eGFP) encoding plasmid was transfected into HeLa cells. DNA-polymer complexes were formed and transfections carried out using a range of polymer concentrations (1–150 mM) in order to delineate each polymer activity limits and identify an optimal concentration yielding the highest transgene expression and generating the lowest cellular toxicity (Fig. 6). The percent of eGFP+ cells and transgene expression level (mean fluorescense intensity) varied between 50 and 90 % and 1000–4000 units, respectively. High MW polymers (no. 4-6) with low content of MPEG chains performed

better than low MW / high MPEG content polymers. The difference was especially apparent (3–4 fold) at the level of GFP fluorescense (1000 vs. 4000 units). Comparing to PHPI, low MW / high MPEG content polymers demonstrated reduced transfection efficiency, while low MPEG content had no effect on polymer activity. Toxicity in all cell samples was only moderately increased compared to the negative control and could be attributed to the overexpression of foreign protein. To extend the studies on the ability of MPEG’ylated PHPI to deliver genetic material into the cells, HEK293iGFP – GFP inducible cell line was used for siRNA transfections. GFP expression was induced by doxycycline 1 day before the

Fig. 5 Size of polyplexes formed by DNA and siRNA with PHPI and its MPEG’ylated derivatives. The time of incubation is shown on X axis, the numbers below correspond to the samples in Table 1. Data shows particle size mean values ± SD (n = 3)

10

J Polym Res (2016) 23:10

Page 10 of 12

Fig. 6 DNA transfer analysis of polyplexes with PHPI and its MPEG’ylated derivatives. Concentration of the polymers is shown on X axis, the numbers below correspond to the samples in Table 1. Graph shows mean values ± SD (n = 3)

transfection. Non-targeting control and GFP-specific siRNA were transfected using the amount of polymers determined to give the best DNA transfection result (Fig. 6). GFP fluorescense was measured 48 h later using flow cytometer, the GFP knockdown efficiency was calculated based on the reduction of mean fluorescense intensity comparing samples transfected with non-targeting control vs. GFP-specific siRNA (Fig. 7). Similarly as with DNA transfection, high MW/ low MPEG content polymers performed better, for which GFP suppression levels reached 70 % (black histograms), while for low MW/ high MPEG content polymers – 40 % (grey histograms). In both cases, MPEG’ylated derivatives performed better than non-PEG’ylated PHPI. Thus, MPEG’ylation of PHPI facilitates improved siRNA mediated gene suppression effect. Comparison of the results of DNA and siRNA transfection experiments lead to the conclusion, that high MW polymers (no. 4-6) are clearly superior over the low MW polymers (no. 1-3). High MW polymers were of much higher dispersity (Table 1) which could have a certain effect on efficiency of transfection. Some published data indicate [59], however, that molecular weight of a polymer has much stronger effect on delivery efficiency of a vehicle than its molecular weight distribution. For instance, PEI with a weight average molecular weight of 216,000 and a number average molecular weight of

17,000 (dispersity index 12.93) was much more effective than PEI with a weight average molecular weight of 7600 and a number average molecular weight of 5600 (dispersity index 1.36) [59]. The length of MPEG chains attached to the polymers had no effect on the transfection efficiency (little variation was seen between the polymers 1–3 or 4–6) for both, DNA and siRNA experiments. The content of MPEG chains in PHPI derivatives, on the other hand, had a significant effect. It was reported before, that an attachment of PEG chains reduced the net charge of polyplexes and, subsequently, their ability to bind to the negatively charged cell membranes [43, 56, 60]. Similar results were obtained in the present study where the polymers no. 1-3 were less capable of DNA and siRNA delivery than the polymers no. 4-6. PEG’ylation has been stated to have different effect on nucleic acid release inside the cell: negative for DNA [61] and positive for siRNA [39]. In our studies, the reduction in DNA transfection efficiency, when using MPEG’ylated PHPI instead of non-MPEG’ylated counterparts, could also be attributed to poor DNA release into the cytosol. An increase in gene knockdown level with all MPEG‘ylated derivatives of PHPI in siRNA experiments suggests that MPEG grafting improves siRNA release and fosters its biological activity.

Conclusions

Fig. 7 siRNA-mediated gene knockdown using PHPI and its MPEG’ylated derivatives. The numbers on X axis correspond to the samples in Table 1. Graph shows mean values ± SD (n = 3)

MPEG‘ylated derivatives of PHPI were synthesized by polycondensation of 1,3-diamino-2-propanol and 1,3-dibromo-2propanol in the presence of MPEG iodide (MPEGI) of different molecular weight (Mn 1000, 2000, and 5000). Addition of MPEGI to the initial monomer feed resulted in polymers with relatively low molecular weight (Mw 8–53 kDa) and high MPEG content (over 60 mol %, 1 to 1.25 MPEG chains per macromolecule). Introduction of MPEGI during the final stage of the polycondensation gave high-molecular-weight products (Mw 140–160 kDa) with high dispersity and low

J Polym Res (2016) 23:10

content of MPEG (less than 9 mol %, 0.09 to 0.22 MPEG chain per macromolecule). Yield of MPEG’ylated PHPI was low, less than 30 %; these polymers were highly branched, irrespective of the method of the synthesis. Buffering capacity of PHPI covers pH range from 6.4 to 8.8, and the degree of unprotonation of amino groups at physiological pH 7.3 is at about 40–45 %. Hydrodynamic diameter of MPEG’ylated derivatives is considerably lower compared to non-MPEG’ylated PHPI and is almost independent of pH. Complexes of MPEG’ylated derivatives of PHPI with DNA are larger (100–170 nm) compared to those with siRNA (40– 70 nm). Comparing to PHPI, low MW / high MPEG content PHPI derivatives demonstrated reduced DNA transfection efficiency. Efficiency of high MW / low MPEG content PHPI derivatives in DNA transfection was similar to that of PHPI. The length of MPEG chains attached to the polymers had no effect over the transfection efficiency. MPEG grafting had a positive effect on siRNA delivery: high MW polymers with low content of MPEG performed significantly better than nonMPEG’ylated PHPI and low-molecular-weight polymers with high content of MPEG.

Page 11 of 12 10 18. 19. 20. 21. 22. 23. 24. 25. 26. 27.

28. 29. 30. 31. 32.

References 33. 1. 2.

3. 4. 5.

6. 7. 8. 9. 10. 11. 12. 13. 14.

15. 16. 17.

Boucher RC (1999) Clin Investig 103:441–445 Cavazzana-Calvo M, Hacein-Bey S, de Saint BG, Gross F, Yvon E, Nusbaum P, Selz F, Hue C, Certain S, Casanova JL, Bousso P, Deist FL, Fischer A (2000) Science 669:669–672 Merdan T, Kopeček J, Kissel TJ (2002) Adv Drug Deliv Rev 54: 715–758 Li W, Szoka FC (2007) Pharm Res 3:438–449 LeWitt PA, Rezai AR, Leehey MA, Ojemann SG, Flaherty AW, Eskandar EN, Kostyk SK, Thomas K, Sarkar A, Siddiqui MS, Tatter SB, Schwalb JM, Poston KL, Henderson JM, Kurlan RM, Richard IH, Van Meter L, Sapan CV, During MJ, Kaplitt MG, Feigin AJ (2011) Lancet Neurol 10:309–319 Resnier P, Montier T, Mathieu V, Benoit J, Passirani C (2013) Biomaterials 34:6429–6443 Gould DJ, Favorov P (2003) Gene Ther 10:912–927 Stone D, David A, Bolognani F, Lowenstein PR, Castro MG (2000) J Endocrinol 164:103–118 Lundstrom K (2003) Trends Biotechnol 21:117–122 Yang J, Liu H, Zhang X (2014) Biotechnol Adv 32:804–817 Mintzer MA, Simanek EE (2009) Chem Rev 109:259–302 Agarwal S, Zhang Y, Maji S, Greiner A (2012) Mater Today 15: 388–393 Synatschke CV, Schallon A, Jerome V, Freitag R, Muller AHE (2011) Biomacromolecules 12:4247–4255 Kurosaki T, Yamashita Y, Aki K, Harasawa H, Nakagawa H, Kodama Y, Higuchi N, Nakamura T, Kitahara T, Sasaki HJ (2011) Pharm Sci 100:4855–4863 Wu Y, Liu C, Zhao X, Xiang J (2008) J Polym Res 15:181–185 Baker A, Saltik M, Lehrmann H, Killisch I, Mautner V, Lamm G, Christofori G, Cotten M (1997) Gene Ther 4:773–782 Szabo R, Mezo G, Hudecz F (2002) J Bioact Compat Polym 17: 399–415

34. 35. 36. 37. 38. 39. 40.

41. 42. 43. 44. 45. 46. 47. 48. 49.

Dubruel P, Christiaens B, Vanloo B, Bracke K, Rosseneu M, Vandekerckhove J, Schacht (2003) Eur J Pharm Sci 18:211–220 Ahmed M, Jawanda M, Ishihara K, Narain R (2012) Biomaterials 33:7858–7870 Wu C, Liu C, Hao J, Deng X (2010) J Polym Res 17:869–876 Wightman L, Kircheis R, Rossler V, Carotta S, Ruzicka R, Kursa M, Wagner E (2001) J Gene Med 3:362–372 Lungwitz U, Breunig M, Blunk T, Gopferich A (2005) Eur J Pharm Biopharm 60:247–266 Von Harpe A, Petersen H, Li Y, Kissel T (2000) J Control Release 69:309–322 Jager M, Schubert S, Ochrimenko S, Fischer D, Schubert US (2012) Chem Soc Rev 41:4755–4767 Zhang S, Zhao B, Jiang H, Wang B, Ma B (2007) J Control Release 123:1–10 Hu Y, Tang G, Liu J, Cheng W, Yue Y, Li J, Zhang P (2012) J Biomed Biotechnol 2012:1–7 Ngyugen HK, Lemieux P, Vinogradov SV, Gebhart CL, Guerin N, Paradis G, Bronich TK, Alakhov VY, Kabanov AV (2000) Gene Ther 7:126–138 Liu Y, Nguyen J, Steele T, Merkel O, Kissel T (2009) Polymer 50: 3895–3904 Xun M, Liu Y, Guo Q, Zhang J, Zhang Q, Wu W, Yu X (2014) Eur J Med Chem 78:118–125 Rimann M, Luhmann T, Textor M, Guerino B, Ogier J, Hall H (2008) Bioconjug Chem 19:548–557 Kolate A, Baradia D, Patil S, Vhora I, Kore G, Misra A (2014) J Control Release 192:67–81 Zintchenko A, Philipp A, Dehshahri A, Wagner E (2008) Bioconjug Chem 19:1448–1455 Roberts MJ, Bentley MD, Harris JM (2002) Adv Drug Deliv Rev 54:459–476 Petersen H, Fechner PM, Fischer D, Kissel T (2002) Macromolecules 35:6867–6874 Zhang X, Pan S, Hu H, Wu G, Feng M, Zhang W, Luo X (2008) J Biomed Mater Res 84A:795–804 Banerjee P, Weissleder R, Bogdanov AJ (2006) Chem Bioconjug 17:125–131 Petersen H, Kunath K, Martin AL, Stolnik S, Roberts CJ, Davies MC, Kissel T (2002) Biomacromolecules 3:926–936 Parka MR, Hanb KO, Hanb IK, Cho MH, Nahd JW, Choia YJ, Choa CS (2005) J Control Release 105:367–380 Kursa M, Walker GF, Roessler V, Ogris M, Roedl W, Kircheis R, Wagner E (2003) Bioconjug Chem 14:222–231 Zhong Z, Feijen J, Lok MC, Hennink WE, Christensen LV, Yockman JW, Kim YH, Kim SW (2005) Biomacromolecules 6: 3440–3448 Mao S, Neu M, Germershaus O, Merkel O, Sitterberg J, Bakowsky U, Kissel T (2006) Bioconjug Chem 17:1209–1218 Shen Y, Deng J, Luo X, Zhang X, Zeng X, Feng M, Pan S (2009) J Biomater Sci 20:1217–1233 Malek A, Czubayko F, Aigner A (2008) J Drug Target 16:124–139 Gary DJ, Puri N, Won YJ (2007) Control Release 121:64–73 Mady MM, Mohammed WA, El-Guendy NM, Elsayed AA (2011) Rom J Biophys 2:151–165 Boussif O, Lezoualcht F, Zanta MA, Mergnyt MD, Schermant D, Demeneixt B, Behr (1995) J Proc Natl Acad Sci 92:7297–7301 Ohlfest JR, Perkinson S, Lobitz P, Hasz D, Largaespada DA (2004) Mol Ther 10:260–268 Aravindana L, Bicknella KA, Brooks G, Khutoryanskiy VV, Williams AC (2009) Int J Pharm 378:201–210 Zaliauskiene L, Bernadisiute U, Vareikis A, Makuska R, Volungeviciene I, Petuskaite A, Riauba L, Lagunavicius A, Zigmantas S (2010) Bioconjug Chem 21:1602–1611

10 50.

J Polym Res (2016) 23:10

Page 12 of 12

Lagunavicius A, Zigmantas S, Riauba L, Zaliauskiene L, Makuska R, Vareikis A (2010) Bernadisiute U US Patent Application 12/333, 453 51. Bernadisiute U, Antanelis T, Vareikis A, Makuska R (2008) Chemija 19:43–49 52. Bockuviene A, Vareikis A, Makuska R (2015) Chemija 26:51–59 53. Wang J, Lu Z, Wientjes MG, Au JLS (2010) AAPS J 12:492–503 54. Odian G (2004) Principles of polymerization, Fourth ed., John Wiley & Sons, Inc. Chapter 2, pp 39–195 55. Malmström E, Hult A (1997) J Macromol Sci C Polym Rev 37: 555–579

56.

Benjaminsen RV, Mattebjerg MA, Henriksen JR, Moghimi SM, Andresen TL (2013) Mol Ther 21:149–157 57. Sonawane ND, Szoka FC, Verkman AS (2003) J Biol Chem 278: 44826–44831 58. Boussif O, Lezoualc’h F, Zanta MA (1995) Proc Natl Acad Sci U S A 92:7297–7301 59. Godbey WT, Kenneth KW, Mikos AG (1999) J Biomed Mater Res 45:268–275 60. Erbacher P, Rem J-S, Behr J-P (1999) Gene Ther 6:138–145 61. Knorr V, Allmendinger L, Walker GF, Paintner FF, Wagner E (2007) Bioconjug Chem 18:1218–1225