Synthesis of a novel PEGDGA-coated hPAMAM complex as an efficient and biocompatible gene delivery vector: an in vitro and in vivo study

Abstract

hPAMAM/DNA polyplexes, compared to viral vectors, display unique characteristics including more safety, less immune response outcomes, a simpler synthesis and an easier process. Given the importance of these polymers, hPAMAM coated with the PEGDGA copolymer was developed as a promising non-viral gene carrier. In the present study, a new complex of hPAMAM, PEGDGA-modified hyperbranched polyamidoamine (hPAMAM), was established as a versatile non-viral gene vector. The hPAMAM polymer was synthesized by using a modified one-pot method. The resulting hPAMAM–PEGDGA polymer was able to efficiently protect encapsulated-DNA against degradation for over 2 h. In addition to low cytotoxicity, the transfection efficiency of hPAMAM–PEGDGA represented much higher (p < 0.05) than that of Lipofectamine 2000 in both MCF7 and MDA-MB231 cells (an approximately 4.5-fold increase). Cellular uptake of hPAMAM–PEGDGA in MDA-MB231 cells, 254.79 ± 2.1, was significantly higher than that in MCF7 cells, 51.61 ± 6.1 (p < 0.05). EMA-labeled DNA can be clearly observed in the tumor tissue of mice receiving hPAMAM-PEGDGA/EMA-labeled DNA. However, a significant number of fluorescent spots can be found in the tumor tissue of mice receiving hPAMAM/DNA, when compared to those treated with naked hPAMAM/DNA. It has been observed that GFP is expressed more highly in hPAMAM-PEGDGA/EMA-labeled/DNA than the one in PAMAM/DNA. The results indicated that hPAMAM-PEGDGA-mediated gene delivery to breast cancer cells is a feasible and effective strategy that may offer a new therapeutic avenue as a non-viral gene delivery carrier. Notably, According to these findings, this newly-introduced copolymer, the hPAMAM–PEGDGA complex, has proved to be a promising strategy for drug or gene delivery to tissues or cell types of interest, particularly to triple-negative breast cancer.

Keywords:

• Breast cancer
• copolymer PEG–glutamic acid
• gene delivery
• glutamic acid
• hPAMAM
• hPAMAM–Glu

Introduction

Synthetic cationic polymers as non-viral gene delivery carriers, compared to viral gene delivery carriers, have demonstrated a higher safety profile without undesirable immune responses and malicious transformation, proving to be a favorable candidate for clinical gene therapy (Takakura et al., 2002; Pack et al., 2005; Park et al., 2006; Han et al., 2014). On the other hand, the currently-available cationic polymers have not yet been shown to be a suitable vector for clinical application. The major limitations of these cationic polyamidoamines are the relatively low gene transfer efficiency as well as high cytotoxicity, which hinder their clinical applications for gene therapy (Mazda, 2002; Wang & Chang, 2003).

This emphasizes the need for the development and application of effective, biocompatible and affordable non-viral vectors for gene therapy (Gao et al., 2008). Among several cationic polyamidoamine vectors studied, dendrimers provide significant benefits over conventional polymeric systems due to their distinctive molecular architecture and individual properties (Xiong et al., 2007). There is no doubt that dendrimers have a precisely-defined globular structure and well-controlled size, assembled by step-by-step synthesis (Dufès et al., 2005; Tomalia, 2005; Bi et al., 2015). However, a PAMAM dendrimer molecule has a well-defined spherical shape, representing a high positive surface charge. PAMAM dendrimer molecules have demonstrated considerable efficacy in both gene therapy and drug delivery (Bielinska et al., 1999; Fant et al., 2008; Wang et al., 2010). These molecules combine to form a DNA/dendrimer complex with a relatively low polydispersity index (PDI) which inhibits aggregations. Generally, this is a critical factor for the successful in vivo application of cationic polyamidoamines as a gene delivery carrier. In addition, a high density of amine groups on the surface of and inside the polyamidoamines provides various functions to improve gene delivery carriers (Haensler & Szoka, 1993; Boussif et al., 1995). However, the presence of extremely compact amino groups at the periphery of hPAMAM limits high-dose administration and long-term application due to its high cytotoxicity (Zinselmeyer et al., 2002).

Some investigations have shown that these compounds not only are able to serve as protonic sponges, but also provide endosome escape (Kumar et al., 2010). Giving more attention to these hyperbranched dendrimers is closely associated with high-efficiency gene transfer and minimal toxicity (Yang et al., 2011). Moreover, the inflexible structure of dendrimers, although not favorable for DNA binding, prevents DNA degradation during gene delivery (Tang & Szoka, 1997). In order to improve their properties as highly efficient gene delivery carriers, numerous approaches have been widely investigated and developed to modify dendrimers, including PEGylation and coupling of amino acids or ligands. Jevprasesphant et al. showed that the surface of PAMAM modified with six lauroyl or four polyethylene glycol (PEG) chains could successfully lead to decreased cytotoxicity, ascribing to the shielding effect (Jevprasesphant et al., 2003). In addition, Choi et al. reported that l-arginine-grafted-PAMAM showed more enhanced transfection efficiency by changing rigid amino structures on the surface of PAMAM (Choi et al., 2004a; Zeng et al., 2011).

Some recent studies have optimized dendrimer surface groups with various agents, such as PEG, cyclodextrin and l-arginine, indicating that significant advances had been achieved over recent years (Roessler et al., 2001; Choi et al., 2004b; Kim et al., 2004). Using direct administration into the body (in vivo) or transformation (ex vivo), gene therapy strategies can be successfully applied to a variety of solid tumors for various purposes, including cell growth control, allergy medication, tumor immunity induction and angiogenesis inhibition (Crystal, 1995; Kommareddy & Amiji, 2007). The major problem in systemic gene delivery is to develop a safe and effective vector system. However, one potential approach to circumvent this issue is to transfer genes into target cells and tissues, and to express the protein of interest (Thomas et al., 2003).

To confront the problems of hPAMAM as a safe and efficient in vivo gene delivery vector, we have recently developed a new method to synthesize a biodegradable, flexible and pompon-like dendrimer, hPAMAM attached to PEGDGA [di-glutamic acid (DGA) grafted with PEG 400]. The newly-synthetized copolymer was composed of hPAMAM-G2.0 as the inner core and DGA grafted to PEG 400 as the surrounding multiple arms. The pompon-like structure of hPAMAM–PEGDGA was proposed to combine the advantages of hPAMAM, and the low-toxicity and flexible chains of DGA-PEG. The reasons for using hPAMAM–PEGDGA as a novel gene delivery vector are that: (i) hPAMAM used as a core possesses a well-defined globular structure, allowing the formation of a well-defined size and shape of PEGDGA; (ii) biodegradable PEGDGA, as grafted chains, decreases the cytotoxicity of native hPAMAM through a shielding effect like PEG chains (Tian et al., 2007; Zeng et al., 2011); (iii) the DGA-PEG grafted chains also change the rigid structure of hPAMAM to the relatively flexible structure of hPAMAM–PEGDGA. High structural flexibility of hPAMAM was reported to increase the transfection efficiency because of weak base polycations, not only involving in the inhibition of DNA degradation by lysosomal enzymes, but also leading to an increased rate of DNA escape from endocytic vesicles (11) and (iv) finally, DGA-grafted PEG with a molecular weight of 400 Da, attached on the hPAMAM surface, is able to successfully compact and protect DNA from degradation (Wen et al., 2009; Deng et al., 2011).

In the present study, hPAMAM, synthesized by the one-pot method (Zheng et al., 2015), was modified by PEG-400 and poly-glutamic acid (hPAMAM–PEGDGA). Following synthesis, we attempted to investigate gene transfer efficiency and toxicity in the breast cancer cell lines, MCF7, MDA-MB231. It should be noted that fourth-generation PAMAM nanoparticles and Lipofectamine were also studied as controls.

Materials and methods

Materials

pEGFP-N1 (Clontech, Palo Alto, CA) was extracted using a QIAGEN Plasmid Mega Kit (Qiagen GmbH, Hildden, Germany). Breast cancer cell lines, Mcf7 and MDA-MB231, were bought from National Cell Bank of Iran (Pasture Institute, Iran). Polyamidoanimine (PAMAM-G4) (M_w = 14 000 kDa), glutamic acid, Lipofectamine 2000 and PEG 400 were purchased from Sigma Aldrich (St Louis, MO). A plasmid of enhanced green fluorescent protein (pEGFP) was obtained from Invitrogen (Carlsbad, CA).

Animals

Four- to five-week-old (20–25 g) female Balb/c mice were bought from Department of Laboratory Animals, Razi Vaccine and Serum Research Institute, Karaj, Iran. Mice were housed for 1 week before the experiment, given free access to food and water, and maintained in a light/dark cycle with lights on from 6:00 to 18:00 h. All animal experiments were carried out according to assessment guide of the ethic committee of Shahid Beheshti University of Medical Sciences. The mice were subcutaneously inoculated on day 0 with 1 × 10⁶ of MDA-MB-231 cells (Piao et al., 2013). MDA-MB-231 cells were orthotopically injected into the mammary fat pad of 6-week-old BALB/c female mice. Twenty days after cell inoculation, the mice were intravenously injected by ethidium monoazide (EMA)-labeled hPAMAM–PEGDGA/DNA via mouse tail vein, once tumors reached a palpable size. The in vivo fluorescence images were prepared 120 minutes after the injection. GFP gene expression assessed in breast tumor sections, 48 hours after the injection of 50 μg/mouse of hPAMAM-PEGDGA/EMA-labeled/DNA or PAMAM/DNA. Profile of gene expression were assessed in breast tumor tissue of Balb/c mice treated with Naked plasmid, hPAMAM/DNA NPs and PAMAM-PEGDGA/DNA complexes 48 h after i.v injection. Frozen sections (thickness of 20 mm) of breast tumor tissue of all treated group were examined by fluorescent microscopy.

Synthesis of hPAMAM derivatives and characterization of nanoparticles

hPAMAM was synthesized by a one-pot method using DETA and MA, as raw materials. Briefly, Michael addition, DETA in flask (20.6 g) was dissolved in 25 mL methanol, and MA (20.6 g) was added dropwise to the reaction medium. The mixture was placed in the mixer at room temperature for 48 h. Because the Amidation step most certainly takes place between molecules, the Michael and Amidation steps were continuously repeated until the desired generation of dendrimers was obtained (Schmidt et al., 2008). Following each step, Michael and Amidation, the mixture was put in a rotary dryer at room temperature to remove additional MA in the mixture. The product was precipitated three times in ethylene oxide, and kept in a sealed container (Cao et al., 2007; Zeng et al., 2011).

Synthesis of the PEGDGA copolymer

The PEGDGA copolymer was synthesized by grafting DGA onto PEG. 29.43 g (0.2 mol) of glutamic acid was placed in a flask, and 40 g of PEG 400 was then added to the flask. Under nitrogen atmosphere, the reaction mixture temperature was raised from 25 to 175 °C within a period of 1 h, and stirred at this temperature for 1 h. The contents of the reservoir were connected to a vacuum system (10 mm Hg), and stirred at 175 °C for 30 min. For purification, the product was dissolved in chloroform, and then precipitated with ether at 4 °C. Afterwards, the pellet was recovered at low temperatures, which was rapidly transformed into a viscous liquid at 25 °C. Schematic presentation of DGA-grafted-PEG and hPAMAM–PEGDGA syntheses is shown in scheme 1.

An amount equivalent to 0.24 g of the PEGDGA polymer was added to hPAMAM (core DETA G = 2, 4 ml hPAMAM 20% solved in methanol), and the mixture was then stirred for 1 h in a shaker. In general, about half of the NH2 groups in hPAMAM are involved in PGA-g-PEG copolymer ionic bonds, which may lead to a decrease in toxicity and increase in loading efficiency. The schematic diagram in Scheme 1 shows how to graft the PEGDGA polymer to hPAMAM.

A molar ratio of 1:20 (hPAMAM–PEGDGA) (mol/mol) was prepared in ethanol. It should be noted that, at this stage, the primary amine groups present on the surface of hPAMAM reacted with the terminal carboxyl (–CO) groups of DGA-PEG copolymers. The resulting conjugates were dissolved in PBS, pH 7.0. To prepare BODIPY labeled-conjugates, hPAMAM was initially dissolved in 100 mM NaHCO₃ solution, incubated with BODIPY for 12 h at 4 °C, and purified by using the standard method (Twyman et al., 2004). The resulting hPAMAM–PEGDGA conjugates were also labeled with BODIPY using the above-mentioned method.

Characterization of the synthesized nanoparticles

¹H NMR and Zeta seizer were used to characterize the newly-synthesized nanoparticles. To analyze the nanoparticles by ¹H NMR (400 MHz), a dry powder was prepared by lyophilization from the initial solution of hPAMAM–PEGDGA. The particle size and zeta potential of the nanoparticles were evaluated by dynamic light scattering and zeta plus analyzer Zetasizer Nano-ZS (Malvern Instruments, Worcestershire, UK). Atomic force microscopy (AFM) was used in the tapping mode to survey the shape and surface morphology of the nanoparticles (Digital Instrument, Santa Barbara, CA).

Gel retardation assay

When hPAMAM, hPAMAM–Glu, hPAMAM–PLG–PEG and PAMAM-G4 were prepared, the complexes were diluted in distilled water at various concentrations. To obtain an appropriate N/P ratio, certain amounts of the nanoparticles were added to DNA solution (1 μg, containing 3 nM phosphate). To this end, DNA (100 mg/ml) was dissolved in sodium sulfate solution (50 mM) at specified weight ratios, and then added to the nanoparticle solution (0.2:5, 1:1, 5:1 and 10:1 hPAMAM/DNA w/w). The mixture was gently vortexed for 30 s (37 °C), and the complex was then incubated at room temperature for 30 min to allow self-assembly formation. Agarose gel electrophoresis was performed to demonstrate both the formation of DNA nanoparticles and the stability of the complex at a voltage of 100 mV/cm. The FT-IR spectra and GPC data of hPAMAM and PEGDGA nanoparticles are shown in the Supplementary data.

Cell lines

MCF7 and MDA-MB231 cell lines were provided from the National Cell Bank of Iran (Pasture Institute) (NCBI), and cultured according to standard procedures. Briefly, the cells were cultured in RPMI 1640 culture medium, supplemented with 10% FBS at 37 °C in a humidified atmosphere containing 5% CO₂ (Zhu et al., 2012).

Cellular uptake of dendrimers

The cell lines, MCF7and MDA-MB231, were incubated at a density of 2 × 10⁴ cells/well in 24-well plates (Corning-Coaster, Sigma) for the first 48 h, and examined for cell morphology and confluency. Afterwards, the cells were exposed to a 1.0 μM concentration of BODIPY-labeled PAMAMG4, as well as the concentrations of 1.0, 2, 4 and 8 μM of BODIPY-labeled hPAMAM–PLG–PEG at room temperature for 30 min. The culture medium was removed, and washed three times by cold PBS at pH 7.4. The cells were then observed by a fluorescent microscope (Olympus, Osaka, Japan). In our recently published experience, we have shown that the most efficient cellular uptake of hPAMAM–Glu (monomer amino acid) occurs at a ratio of 1/20 (Hemmati et al., 2016). To quantitatively analyze the cellular uptake of hPAMAM–PEGDGA, PAMAM-GLU, hPAMAM and PAMAM at various concentrations, the cells were cultured at a density of 8 × 10⁴ cells/well in six-well plates (Corning-Coaster, Sigma) for 72 h, and then examined for cell morphology and confluency under a microscope. Subsequently, the cells were incubated at concentrations of 1.0, 2, 4 and 8 μM of hPAMAM–PEGDGA or 1 μM of PAMAM-GLU, hPAMAM and PAMAM nanoparticles labeled with BODIPY for 30 min. Similarly, the cells were washed three times with PBS (pH 7.4), and the treated cells were trypsinized and centrifuged at 1600 rpm for 8 min. Afterwards, the cell pellet was resuspended in PBS, and fluorescence was measured by a flow cytometer (FACSCalibur, BD Biosciences, Bedford, MA) equipped with an argon ion laser (488 nm). Data were analyzed using WinMDI 2.9 (Microsoft, Redmond, WA) (Ke et al., 2009). The cells cultured under normal conditions and live cells were considered as a negative control and gate, respectively; this means that the only cells at the gate were analyzed. The mean density of fluorescent cells was calculated as a histogram plot.

Gene transfer efficiency into MCF7 cells

The plasmid DNA was covalently labeled with fluorescent dye ethidiummonoazide bromide (EMA). Briefly, plasmid DNA solution (1 mg/ml in TE buffer, pH 7.0) was diluted with EMA solution (0.1 mg/ml), and subsequently incubated for 30 min in a dark room. Finally, the solution was precipitated by the addition of ethanol to a final concentration of 30% (v/v). Following centrifugation, the pellet was resuspended in 50 mM sodium sulfate solution. The solution was used to prepare the EMA-DNA/nanoparticle complex by the same method. The cells were seeded at a density of 2.0 × 10⁴ cells/well in a 24-well plate, and then incubated for 48 h. At various ratios, the hPAMAM–PEGDGA, PAMAM-GLU, hPAMAM and PAMAM complexes containing EMA-DNA were subsequently added to the cells (Pardridge, 2005). After 1 h, the supernatant was removed, and the cells were washed three times with PBS, pH 7.4, and observed under a fluorescent microscope.

In vitro cell viability assay

Toxicity evaluation and cell viability were determined by the MTT assay. Briefly, the cells were cultured at a density of 10 000 cells (MCF7 and MDA-MB231) per well in 96-well plates, and then exposed to the newly-synthetized nanoparticles, hPAMAM–PEGDGA, PAMAM-GLU and hPAMAM, at various concentrations (10, 50 100, 150 and 200 μg/ml).

After cell transfection, the in vitro cytotoxicity of hPAMAM–PEGDGA/pDNA was also evaluated by the MTT-based assay. After 24 h, the hPAMAM–pDNA complexes were added to the cells at various weight ratios (1, 3, 5 and 10) as well as DNA dosages (1 μg), and incubated for 4 h in serum free medium. Culture medium was replaced with a medium containing 10% FBS, and incubated for 24 h. For the MTT assay, 100 μl of MTT solution (0.5 mg/ml) was added to each well, and incubated for another 4 h. In the next step, the medium was carefully removed, and 200 μl of dimethyl sulfide (DMSO) was added to each well. Afterwards, the plate was incubated for 15 min at 37 °C. The absorbance of formazan was read on a spectrophotometric microplate reader (MOBEL 550, Bio-Rad Laboratories, Hercules, CA), using a wavelength of 570 nm (Huang et al., 2010). The experiment was repeated three times, and relative cell viability was expressed as the percentage of live cells compared to untreated control cells.

Green fluorescence protein expression

The cell lines, MCF7 and MDA-MB231, were cultured at a density of 1.0 × 10⁶ cells/well in six-well plates in 2 ml of culture medium for 18 h to reach 50% confluency for gene therapy. Prior to transfection, the medium was replaced with a serum-free medium, and the cells were then incubated by the complexes hPAMAM–PEGDGA/DNA (8 μg/well) for 4 h at a specific weight ratio. The medium was replaced with a medium containing 10% FBS, and then incubated again for 24 h. GFP expression was detected under a fluorescent microscope (Olympus) (Huang et al., 2011).

Luciferase activity assay

MCF7 and MDA-MB231 cell lines were cultured in a 24-well plate with appropriate confluences for 24 h to reach their confluences to 50% at the transfection time. Prior to transfection, the medium was replaced with serum-free medium, and the cells were then incubated with hPAMAM–PEGDGA/pDNA complexes (4 μg/well) at a specific weight ratio under standard conditions for 4 h. Subsequently, the medium was replaced with a medium containing 10% FBS, and incubated again for 24 h.

To measure the luciferase activity, cells were washed two times with PBS, and 100–200 μl of cell lysis buffer was added to each well for 5 min at 37 °C. Cell lysis solution was centrifuged at 1200 rpm for 5 min, 10 μl of the supernatant was mixed with 25 μl of both luciferase substrate and ATP solution, and luciferase activity was measured with Sirius luminometer (Autolumat LB953, EG and G, Berthold, Germany). Small amounts of the protein were determined by the bicinchoninic acid assay (Bio-Rad Laboratories). Relative light units (RLUs) were normalized with the concentration of the proteins extracted from the cells. All transfections were performed in triplicate. The transfection activity was reported by RLUs (Ma et al., 2009). Lipofectamine was used as a control to compare each sample.

Confocal microscopy

Intracellular distribution of the nanoparticles was investigated by a confocal microscopy. The cell line MDA-MB231 was cultured at a density of 10 000 cells per well in the 35 mm glass-bottom culture dish at 37° C, 5% Co₂, for 48 h. Afterwards, the cells were incubated with the hPAMAM–PEGDGA complex (5 μg pDNA/well), washed with cold PBS, and finally incubated with 50 nM Lyso Tracker Green and DAPI for 15 and 60 min, respectively. Henceforth, the cells were washed three times with cold PBS, fixed with 3.7% paraformaldehyde (Ke et al., 2009) and observed under the confocal laser scanning microscopy (BioRad MRC1024 MP, CA), with a scan head on a Nikon E800 microscope (60x objective).

Results

Nanoparticle characterization

Figure 1 shows the chemical structure of the DGA-PEG copolymer and hPAMAM. The chemical structure of hPAMAMs is similar to that of PAMAM dendrimers. The resulting copolymer was confirmed by GPC spectra (Figure 2A; Table 1). The number of Glu residues attached to PEG was estimated using the GPC Chromatogram. The synthesized PEGDGA contains two glutamic acid monomers coupled with one PEG 400. The resulting copolymer–hPAMAM complex was confirmed by NMR spectra (Figure 2B). ¹H NMR (400 MHz, D₂O) branches and hPAMAM protons represented several peaks between 2.0 and 3.2 ppm, respectively. The Glu units and PEG 400 showed a sharp peak at 3.4–3.6 ppm (Figure 2B). The PEGylation degree and the number of Glu attached to hPAMAM were estimated using the proton integration method, by taking into account the characteristic peaks of PEG and hPAMAM (Zhu et al., 2010). The percentage of PEGDGA copolymers attached to hPAMAM was calculated from the integral ratio of –NHCH2CH2–(C/c, 2.2–3.2 ppm) to PEG (i, 3.6 ppm), by the following formula, PAMAM–PEG= 2(C/c)/3(i). Based on the integral proportion (area under curve) of protons (2.2–3.2) to glutamic acid (3.4–3.6), the binding of amine groups in the hPAMAM–PEGDGA and hPAMAM–Glu20 (without PEG) samples were 30% and 20%, respectively. Approximately 15 and 30 molecules of the DGA-PEG copolymer and monomer glutamic acid residues were attached to one molecule of the hPAMAM–PEGDGA complex, respectively.

Figure 1. Presentation of PEGDGA copolymer and PAMAM–PEGDGA synthesis.

Figure 2. GPC chromatogram of PEGDGA copolymer (A) ¹H NMR spectrum of hPAMAM–PEGDGA (B). The percentage of PEGDGA copolymer attachment to hPAMAM was calculated from the integral ratio of –NHCH2CH2– (C/c, 2.0–3.2 ppm) to PEG (i, 3.7 ppm), by the following formula, hPAMAM–PEG= 2(C/c)/3(i)).

Table 1. GPC parameters of PEGDGA copolymer.

PEAK 1	MP	MN	MW	MZ	MZ 1	MV	PD
PEGDGA	672	649	704	761	820	696	1.084

AFM and DLS were performed to study the size and morphology of nanoparticles, as shown in Figure 3(A) and Table 2, respectively. The mean size and zeta potential of hPAMAM–pDNA complexes are shown at a weight ratio of 10:1 in Table 2. As illustrated in Figure 3(A), two- and three-dimensional images of AFM suggested that the dendrimer-to-plasmid ratio was 10:1, representing that, depending on the properties of their structure, the size is between 90 ± 30 nm. As depicted in Figure 3(B), while the non-capsulated pDNA was completely destroyed during the first half hour, pDNA encapsulated with dendrimers was protected from enzymatic degradation for 2 h.

Figure 3. (A) AFM images with 3D view of the nanoparticle complexes, image of hPAMAM–PEGDGA/pDNA (120 nm) (complex at a weight ratio of 10). (B) hPAMAM–PEGDGA encapsulated DNA showed stability against DNase I for up to 120 min, while the naked DNA was fully degraded in 30 min.

Table 2. Particle size and zeta potential of hPAMAM and PAMAM-PEGDGA with or without DNA (n = 3).

Nanoparticle	Mean size (mean ± SD, nm)	Zeta potential (mV)	PDI
PAMAM-G4	3.8 ± 0.6	3.8 ± 0.8	0.55
hPAMAM	1.94 ± 0.5	3.11 ± 0.87	0.1
hPAMAM–PEGDGA	4.41 ± 2.8	1.5 ± 1.1	0.7
hPAMAM–PEGDGA/DNA  (N/P 10)	120 ± 25	−34 ± 13.9

Cell viability assay

In this study, we sought to investigate the potential toxicity of hPAMAM–PEGDGA, Lipofectamine and PAMAM-G4 to tumor cells. In this light, the toxicity of hPAMAM–PEGDGA, PAMAM-G4 and Lipofectamine derivatives to MCF7 and MDA-MB231 cells was first determined by the MTT assay. As shown in Figure 4(TQ1)(A), at varying concentrations (10, 50, 100, 150, 200 μ/ml) of hPAMAM, hPAMAM–PEGDGA NPs on the human cell lines, no significant toxicity was observed after 4 h, demonstrating that both nanoparticles hPAMAM-G2 and PAMAM-G4 lead to the decreased cell viability with increased concentration. The modified MTT assay not only showed no toxicity up to 200 mg/ml for hPAMAM–PEGDGA, but also displayed no significant toxicity, when compared to PAMAM-G4 and Lipofectamine, as illustrated in Figure 4(A) (P < 0.05).

Figure 4. Cell viability of hPAMAM and hPAMAM–PEGDGA. (A) Cell viability on 4 h after treatment by nanoparticles (Lipofectamine, PAMAM-G4, hPAMAM, hPAMAM–PEGDGA) in MCF7 (A) and MDA-MB231 (B) cell lines at various polymer concentration.

Interestingly, hPAMAM–PEGDGA exhibited lower cytotoxicity effects on the cell viability, as compared with hPAMAM. A sharp drop was observed in the number of PAMAM-treated G4 viable cells when the polymer concentration was slightly increased to 50 μg/ml. hPAMAM and its derivatives, hPAMAM/Glu₂₀ and hPAMAM–PEGDGA, displayed a better performance in the cell viability assay than PAMAM-G4 and Lipofectamine. However, hPAMAM was observed to display significantly greater cytotoxicity than either hPAMAM–Glu20, hPAMAM–PEGDGA. In contrast to this finding, hPAMAM was able to support high cell viability (over 81%) in both cell lines after 4 h, when compared to Lipofectamine and PAMAM-G4. In addition, hPAMAM/Glu20 and hPAMAM–PEGDGA exhibited a lower cytotoxicity at the longer transfection times compared with PAMAM-G4.0. It must be noted that significant differences in cytotoxicity were found between hPAMAM–PEGDGA and hPAMAM in Mcf7 and MDA-MB231 cells (data not shown). In this study, we showed a flexible and biocompatible coating method that is able to cover amino groups present on the surface of hPAMAM dendrimers, successfully reducing the toxicity, as has been shown by other researchers and (Huang et al., 2010, Hemmati et al., 2016). According to the above findings, the MTT results proposed that the biocompatible hPAMAM–PEGDGA is able to reduce the toxicity of hPAMAM compared to naked hPAMAM and PAMAMG4.

Figure 7. Quantification of luciferase expression in MCF7 and MDA-MB231 cells. Luciferase activity was measured 48 h post-transfection with different complexes (Lipofectamine, hPAMAM/DNA and hPAMAM–PEGDGA/DNA) for 4 h as described before, and was expressed as light units per mg protein. ∗, p < 0.05 between group; †, p < 0.05 versus Lipofectamine (as control). Data represent the mean ± SD (n = 3).

In- vitro uptake characteristics of dendrimers by MCF7 and MDA-MB231

To assess the concentration-dependent entry of nanoparticles into MCF7 and MDA-MB231 cells, BODIPY-labeled hPAMAM–PEGDGA was evaluated in different concentrations (1, 2, 4, 8 μM) following 1-h cell exposure, as shown in Figure 5(A–D MCF7 and E–H in MDA-MB231 cell lines).

Figure 5. Fluorescent microscopy and flow cytometry analysis data. Cellular uptake of BODIPY-labeled hPAMAM–PEGDGA in the concentration of 1, 2, 4 and 8 μm in MCF7 (A–E) and MDA-MB231 (E–H) cells and the data of both cells showed in figure (J). Cellular uptake of BODIPY-labeled hPAMAM–PEGDGA complex and PAMAM-G4 in the concentration of 1 μm (I). The number of BODIPY-positive cells was analyzed by setting a gate according to the control. Green: BODIPY.

As illustrated in Figure 5(J), It was well-demonstrated that while the hPAMAM–PEGDGA concentration was increased from 1.0 to 4.0 μM, the percentage of positive BODIPY cells treated with hPAMAM–PEGDGA and hPAMAM–Glu20 was elevated from 42.07 and 41.84 to 244.61 and 114.64 in MDA-MB231 cells, respectively. In addition, the cellular uptake rate of all BODIY-labeled complexes was determined at the concentration of 1.0 μM, shown in Figure 5(I).

At a concentration of 8 μM, the transfection efficiency of hPAMAM–PEGDGA was increased up to 254.79, demonstrating that its efficiency was 4-fold more than that of hPAMAM/Glu20 (63.87) (p < 0.01) in MDA-MB231 cells. We further investigated the transfection efficiency in another cell line, MCF7. However, it turned out that the transfection efficiency of hPAMAM–PEGDGA increased from 41.74 to 172.4 at the concentration of 8 μM, shown in Figure 5. Under optimal condition, a significantly higher transfection efficiency was achieved by hPAMAM–PEGDGA (p < 0.01), compared with hPAMAM/Glu20.

Gene transfer efficiency into MCF7

The optimal transfection condition was W_hPAMAM/W_DNA ratio of 10 using 2 μM DNA per well. The EMA-labeled pDNA was used to assess the quality of gene introduction by hPAMAM/pDNA-EMA nanoparticles into the cells. The qualitative results are shown by fluorescent images in Figure 6. The elevated entry level of nanoparticles into the MCF7 cells was found to correlate with the increased weight ratio. W_hPAMAM/W_DNA ratios of hPAMAM (A ratio of 10) and hPAMAM–PEGDGA (B, C and D at ratios of 2, 4 and 10, respectively) were shown at different ratios, illustrated in Figure 6 (A)–(D). hPAMAM–PEGDGA pDNA (D) at a ratio of 10:1 was found to result in higher gene transfer efficiencies than those at A, B and C ratios.

Figure 6. Cellular uptake of hPAMAM/DNA in the concentration of 8 μm (A), hPAMAM–PEGDGA/DNA NPs in the concentration of 1, 4 and 8 μm on MCF7 cells after a 30-min incubation (B–D). The weight ratio of PAMAM to DNA was 10:1. Results were performed as fluorescent microscopy images. Red: EMA-labeled DNA. Original magnification: 100×.

Quantification of gene-transfer efficiency in Mcf7 and MDA-MB231 cells

Gene transfer efficiency in cancer cells was carried out by using luciferase gene expression, and the results were reported as the lighting Lux unit per mg protein (RLU/mg). As can be seen in Figure 7, there are two factors required for gene transfer efficiency that led us to develop a carrier system, including (I) the proportion of the coated amino acid and (II) W_hPAMAM/W_DNA ratios of nanoparticles and DNA. A ratio of 10:1 was investigated in hPAMAM, hPAMAM–PEGDGA and Lipofectamine nanoparticles. The optical expression of the HPAMAM–PEGDGA nanoparticle (at the polymer/DNA ratio of 10:1) and Lipofectamine reached 8.75 × 10⁴ and 1.8 × 10⁴ RLU/mg protein in the MDA-MB231 cell, respectively.

Using GFP gene expression, it was demonstrated that a significant increase in hPAMAM–PEGDGA gene transfer efficiency was found at an N/P ratio of 10:1, as shown in Figure 8(A)–(D). The fluorescent microscopy images showed that a 10:1 ratio (D) presents a significantly higher efficiency level, when compared to the non-treated group as well as other ratios, such as 1:1 and 5:1 ones, as depicted in Figure 8(A)–(D).

Figure 8. Gene expression on MDA-MB231 cells. The fluorescence images of GFP expression were taken 48 h post-infection with hPAMAM–PEGDGA/DNA in the different concentrations: (A) No treated, (B) 2 μm, (C) 4 μm, (D) 8 μm. Magnification: 100×. Results were performed as fluorescent microscopy images.

Cellular localization of PAMAM–hPAMAM–PEGDGA/pDNA

To study the intracellular distribution of hPAMAM–PEGDGA/pDNA nanoparticles, the confocal microscopy was used to monitor MDA-MB231, shown in Figure 9. As can be seen in Figure 10, the red (hPAMAM–PEGDGA/EMA labeled DNA) and green dots (Lyso tracker Green) were found outside the cells, and concentrated on the cell membrane, while the nanoparticles were able to enter the cells 4 min after transfection. As depicted in Figure 9 (A)–(D), the green and red dots were localized within the lysosomes and endosomes after 15 min, indicating the entry of DNA into these compartments. As shown in Figure 9(E)–(H), DNA was almost entirely distributed in the cells, some of which have penetrated into the nucleus after 60 min.

Figure 9. Intracellular tracking hPAMAM–PEGDGA/DNA NPs in MDA-MB231 cells. Images were taken after incubated with NPs for 15 min (A–D) or 60 min (E–H). Red: EMA-labeled DNA; Green: LysoTracker Green.

Figure 10. In vivo imaging of mice treated by hPAMAM/DNA NPs (A) or PAMAM–PEGDGA/DNA NPs (B). Images were taken 120 min after NPs injection. (C) Semi-quantitative fluorescence intensity of brain and different organs. p value < 0.05.

In vivo imaging analysis

hPAMAM/EMA-labeled DNA (A as a control group) and hPAMAM–PEGDGA/EMA-labeled DNA nanoparticles (B as a treated group) were injected into the tail veins of mice (Figure 10A and B). The in vivo fluorescence images were prepared 120 min after the injection. As shown in Figure 10(B), EMA-labeled DNA can be clearly observed in the tumor tissue of mice receiving hPAMAM–PEGDGA/EMA-labeled DNA. However, a significant number of fluorescent spots can be found in the tumor tissue of mice receiving hPAMAM/DNA, when compared to those treated with naked hPAMAM/DNA (Figure 10C).

Qualitative analysis of gene expression distribution in mouse tumor tissues

Figure 11 shows GFP gene expression in breast tumor sections, 48 h after the injection of 50 μg/mouse of hPAMAM–PEGDGA/EMA-labeled/DNA or PAMAM/DNA. The PAMAM/DNA nanoparticle was able to induce relatively pEGFP gene expression in a tumor tissue section (Figure 11B). Meanwhile, it has been observed that GFP is expressed more highly in hPAMAM–PEGDGA/EMA-labeled/DNA (C) than the one in PAMAM/DNA (B) (Figure 11). Naked plasmid used as a control (Figure 11A).

Figure 11. The qualitative evaluation of gene expression in vivo. Profile of gene expression in breast tumor tissue of Balb/c mice treated with naked plasmid (A), hPAMAM/DNA NPs (B) and PAMAM–PEGDGA/DNA (C) complexes 48 h after i.v. injection. Frozen sections (thickness of 20 mm) of breast tumor tissue (A–C) were examined by fluorescent microscopy. The sections were stained with 300 nM DAPI for 10 min at room temperature. Green: GFP. Blue: cell nuclei. Original magnification: ×100.

Discussion

hPAMAM/DNA polyplexes, compared with viral vectors, display unique characteristics, including more safety, a simpler synthesis, less immune response outcomes and an easier process, whereas safe and effective for clinical application, a non-viral gene carrier system should provide a cost-effective benefit, high-efficiency gene transfer and biological compatibility. Given the importance of these polymers, hPAMAM coated with glutamic amino acids was developed as a promising non-viral gene carrier in our recent publication. The manufacturing procedure to synthesize hPAMAM has already been optimized by Choi et al. (2010), Zhu et al. (2012) and Zheng et al. (2015). In the present study, we attempted to optimize the PEGDGA copolymer present on the surface of these polymeric nanoparticles (hPAMAM). In this work, the PEGDGA copolymer was first added to amine groups present on the surface of hPAMAM, and then the gene transfer efficiency and toxicity of the resulting polymer were assessed in vitro and in vivo.

In general, a hPAMAM-synthesis technique is known to be simple and cost-effective (Zhu et al., 2012; Zheng et al., 2015). While non-viral carriers (e.g. PAMAM) are considered to be a useful method, several issues impede their extensive usage, such as the high cost and complexity (Kumar et al., 2010). The raw materials used to manufacture hPAMAM are DETA and MA, which are synthesized through the one-pot method (Zheng et al., 2015), a time consuming and consistent strategy (Zhu et al., 2012). Therefore, there is no doubt that hPAMAM can be used as a reasonable non-viral gene vector. Because of the necessity of hPAMAM application as a gene carrier, we sought to study the effect of modified hPAMAM amine groups on reducing toxicities and increasing gene transfer efficiencies in the breast cancer cells, MCF7 and MDA-MB231.

To modify the hPAMAM surface, we used PEGDGA as a copolymer (hPAMAM–PEGDGA), along with PEGDGA attached to each other through the ionic bond, carboxyl and amino groups (scheme 1). Due to more flexibility, hyperbranched dendrimers were found to be more effective nanoparticles for gene transfer and DNA compaction, when compared to PAMAMs (Wang et al., 2010). Moreover, a study by Tang and Szoka showed that partial degradation of the dendrimer not only results in improved gene-transfer efficiencies into cells, but also plays a highly important role in the flexibility of the chain (Zinselmeyer et al., 2002; Wang et al., 2010; Zheng et al., 2015). The findings in our recent study, similar to other studies (Wang et al., 2010; Zhu et al., 2012), reflect that the synthesized hPAMAM has shown to be exquisitely effective with low-toxicity in the gene delivery, compared to PAMAMG4. Therefore, it is recommended that the use of these polymeric nanoparticles may be a more cost effective and perfect alternative to polyamidoamine nanoparticles.

In this study, the surface amine groups of hPAMAM were also conjugated with the PEGDGA copolymer, resulting in higher biocompatibility of these nanoparticles. Some studies have reported the same methods to optimize hPAMAM and PAMAM (Tang et al., 1996; Kono et al., 2005; Wang et al., 2010; Hemmati et al., 2016). Some studies, as our recent research findings, reported that after being conjugated, the toxicity and gene-transfer efficiency of nanoparticles have been improved, while attempting to modify the surface of hPAMAM and PAMAM dendrimers by using phenylalanine, glutamic acid and arginine. In this study, the improved gene-transfer efficiency was attributed to the synergistic effect of the proton sponge, resulting from the induction of the inter-polymer secondary amines as well as hydrophobic interactions between hydrophobic groups of amino acids present on the surface of the dendrimers (Tang et al., 1996; Han et al., 2014). Wang et al. (2010) showed that the phenylalanine-modified hPAMAM polymer leads to a decrease and increase in the toxicity and gene-transfer efficiency, respectively. In 2011, when investigating the PAMAM polymerized by glutamic acid, Zeng et al. (2011) were encouraged to study the flexibility and reduced toxicity potentials of PLGE-chain-modified PAMAM combined with glutamic acid residues.

In the present study, hPAMAM conjugated with PEGDGA was investigated as hPAMAM–PEGDGA. The modification ratio of the PEGDGA copolymer residue was estimated through the molar value of PEGDGA copolymer residues added to the surface of amino groups. According to GPC analysis data, the ratio of glutamic acid actually grafted to PEG was two glutamic acid residues. Based on NMR analysis data, the replacement ratio actually performed in hPAMAM–PEGDGA was 30%. Because its side carboxyl group has a PK with approximately 4.1, glutamic acid residues are negatively-charged at physiological pH. A higher level of gene transfer was shown by further modification of hPAMAM by PEGDGA in higher ratios of N/P. However, gene transfer efficiency in hPAMAM–Glu20 was lower than that in hPAMAM–PEGDGA. Therefore, an increase in the complex concentration, from 1.0 to 4.0 μM, resulted in the elevated percentage of BODIPY-positive cells treated with hPAMAM–PEGDGA and hPAMAM–Glu20, from 42.07 and 41.84 to 244.61 and 114.64 in MDA-MB231 cells, respectively. In addition, the cellular uptake of a 1.0 μM concentration of all BODIY-labeled complexes was also compared, shown in Figure 5(I). In the concentration of 8 μM, the transfection efficiency of hPAMAM–PEGDGA was increased up to 254.79, demonstrating that its efficiency was 4-fold more than that of hPAMAM/Glu20 (63.87) (p < 0.01) in MDA-MB231 cells. We further investigated the transfection efficiency in another cell line, MCF7. However, it turned out that the transfection efficiency of hPAMAM–PEGDGA increased from 11.43 to 251.61 at the concentration of 8 μM, shown in Figure 5.

Cellular uptake of nanoparticles into the MDA-MB231 cells, 254.79 ± 2.15, was significantly higher than that into the MCF7 cells, 51.61 ± 1.73. It is also quite likely that the excessive accumulation of negatively-charged groups around the polymer was one of the reasons for the reduction in transfection value in the monomer glutamic acid modification (hPAMAM–Glu20). This effect appears to result from two ways that reduce gene transfer efficiency into the cells. It can be deduced that either negative charge or reduced DNA compaction inhibit nanoparticle binding to cell surface receptors. Although exhibiting a lower loading capacity than hPAMAM–Glu20, hPAMAM–PEGDGA showed a significantly higher efficiency. The similar studies showed that a further modification of nanoparticles by phenylalanine and higher N/P ratios was able to increase the transfer-efficiency level (Carrabino et al., 2005; Wang et al., 2010). In a study done by Wang et al., it has been noted that phenylalanine hydrophobicity can be considered as an important factor governing transfer efficiency. However, similar to other studies and our previous work, the present study demonstrated that the DNA compaction rate in hPAMAM–PEGDGA/DNA was less than the one in hPAMAM/DNA (Carrabino et al., 2005; Wang et al., 2010; Bai et al., 2013). In addition, this study strongly demonstrated significantly higher efficiency and cell uptake rates in MDA-MB231 cells than those in MCF7 cells. This effect is attributed to the glutamate receptor expressed on MDA-MB231 cells, Figures 5 and 7 (Seidlitz et al., 2009).

As shown in Figure 7, the PEGDGA copolymer can greatly increase luciferase gene expression in hPAMAM–PEGDGA treated cells compared to hPAMAM and Lipofectamine. Gene transfer efficiency of hPAMAM–PEGDGA was approximately 4.6 times higher than that of Lipofectamine in MDA-MB231 cells. Another advantage of these nanoparticles is their low toxicity. In this study, the synthesized nanoparticles exhibited no obvious toxicity to both cells (Figure 4A and B).

While Lipofectamine revealed high toxicity, the hPAMAM–Glus complexes showed a very low toxicity profile during gene transfer (Thomas et al., 2005; Wang et al., 2010; Hemmati et al., 2016). Size is one of the major factors in gene transfer efficiency into cells (Kukowska-Latallo et al., 1996; Zhu et al., 2012). As shown in Figure 3(A), the size was 120 ± 15 nm at the weight ratio of 10:1. In addition, the size of the hPAMAM–PEGDGA nanoparticle at a ratio of 10:1, confirmed by AFM, is illustrated in Figure 3(A). The smaller size of the hPAMAM–hPAMAM–PEGDGA/DNA nanoparticle than hPAMAM/DNA is probably due to the presence of negatively-charged glutamic acid residues on the surface of the nanoparticle, resulting in decreased DNA binding to the dendrimer. This is probably (appears to be) one of the reasons why coated nanoparticles tend to be smaller. Endocytosis, Clathrin and/or Caveolae are thought to be other ways to uptake nanoparticles into cells (Zhu et al., 2012). Spherical nanoparticles smaller than 200 nm are usually endocytosed by the Clathrin pathway (Kukowska-Latallo et al., 2005). Because its size is 120 ± 15 nm, hPAMAM/PEGDGA/DNA can be easily absorbed through Clathrin. However, further investigation is required to determine the exact route of entry. Both Results of confocal microscopy, demonstrating lysosomes containing nanoparticles labeled with fluorescent, and these findings could be the reasons for our claim. However, trends for increased internalization of nanoparticles in 15 and 60 min are depicted in Figure 9 (Choi et al., 2004b; Fant et al., 2008).

In this study, it was well-demonstrated that the higher zeta potential (−34 mV) is able to improve gene transfer efficiency, achieving more stability of nanoparticles in in vivo and in vitro application. However, it was found that zeta potentials higher than 26 mV had no effect on the binding level of nanoparticles to the cell membrane and gene transfer efficiency (Rejman et al., 2004; Ye et al., 2007). The surface potential of hPAMAM–PEGDGA/pDNA complexes significantly increased from −0.4 to −34, when the N/P ratio did from 1:1 to 10:1, shown in Table 1.

Moreover, the modified hPAMAMs have a completely significant effect on the transfection efficiency, so hPAMAM–PEGDGA/DNA that showed the highest transfer efficiency, compared to PAMAM-G4 as a control (Figures 5A–J). In a study performed by Zhu et al., the efficiency of hPAMAM was greater than Lipofectamine and PEI (Zhu et al., 2012), similar to the findings in our recent experiment (Hemmati et al., 2016). In addition, we confirmed that gene transfection efficiency is higher in hPAMAM–PEGDGA compared to naked hPAMAM, determined by in vitro GFP gene expression, Figure 8.

MDA-MB231 is well-known as a triple-negative breast cancer (TNBC). Triple- TNBC is categorized by the lack of estrogen receptor, progesterone receptor and human epidermal growth factor receptor 2 on malignant cells. Interestingly, our findings showed that this newly-synthesized copolymer (hPAMAM–PEGDGA) is able to efficiently deliver genes or drugs more specific into the TNBC cells, which may be resulting in highly expressed glutamate receptor in this type of breast cancer.

Speyer et al. (2012) showed that the mGluR1 receptors exist in TNBC cells and confirmed that the inhibition of mGluR1 activity by BAY36-7620 or Riluzole, at the same doses that are currently being used clinically in treatment of ALS, negatively affected the growth of MDA-MB-231 xenografts in mice. In another study, Speyer et al. (2014) also observed that the reduced activity of mGluR1 by the same agent significantly inhibited the growth of 4T1 tumors. Therefore, it is reasonable that the mGluR1 plays a critical role in TNBC, both in the tumor compartment, where it directly stimulates tumor cell growth, and tumor microenvironment, which stimulates angiogenesis. It has been shown that the γ-glutamyl transpeptidase is present in some tissues such as brush border of the proximal convoluted tubules of the kidney (Curto et al., 1988), lactating mammary glands (Meister, 1983; Meister & Anderson, 1983; Vina et al., 1989), the apical portion of the (Speyer et al., 2012,2014) and the BBB (Orlowski et al., 1974) and were also found to have a crucial role in actively transportation of amino acids. Nevertheless, it has been shown that the expression rate of GluR1 in TNBC cells was significantly higher than other tissues even brain (more than 3.6-fold) (Speyer et al., 2014).

Taking together, the rate of gene expression in MCF7 cells was relatively lower than that in MDA-MB231 cells. The results from nanoparticle perfusion showed that hPAMAM–PEGDGA/DNA uptake was higher in breast tumor tissue than naked hPAMAM/DNA nanoparticles (Figure 10). Finally, gene expression profiles of hPAMAM–PEGDGA/DNA nanoparticles in breast tumor tissue showed that the PEGDGA copolymer not only serves as a ligand, but also has a high potential for drug and gene delivery to these types of tumors.

Conclusion

In this study, we showed that PEGDGA-modified hPAMAM has the potential to be a safe and highly effective gene carrier, and to elicit a weaker immune response, as compared to naked hPAMAM. The advantages of this modified nanoparticle are low cytotoxicity and high gene transfection efficiency in MCF7 and MDA-MB231 cells, when compared with PAMAM-G4 and Lipofectamine. In addition, it was found that gene transfection efficiency is approximately 4.6-fold higher in MDA-MB231 compared to MCF7 cells. According to these findings, this newly-introduced copolymer, the hPAMAM–PEGDGA complex, has proved to be a promising strategy for drug or gene delivery to tissues or cell types of interest, particularly to TNBC.

Supplementary material available online.

Declaration of interest

This work was financially supported by the Cancer research Center, Cancer Institute of Iran, Tehran University of Medical Sciences, Cellular and Molecular Biology Research Center, Shahid Beheshti University of Medical Sciences.