Amphoteric Agmatine Containing Polyamidoamines as Carriers for

Sep 3, 2010 - Cancer Research and Treatment, Strada Provinciale 142, Km 3.95, 10060 Candiolo, Torino, Italy, and. Dipartimento di Chimica Organica e ...
0 downloads 0 Views 3MB Size
Biomacromolecules 2010, 11, 2667–2674

2667

Amphoteric Agmatine Containing Polyamidoamines as Carriers for Plasmid DNA In Vitro and In Vivo Delivery Roberta Cavalli,† Agnese Bisazza,† Roberto Sessa,‡ Luca Primo,‡ Fabio Fenili,§ Amedea Manfredi,§ Elisabetta Ranucci,§ and Paolo Ferruti*,§ Dipartimento di Scienza e Tecnologia del Farmaco, Universita` degli Studi di Torino, Via P. Giuria 9, 10125 Torino, Italy, Dipartimento di Scienze Cliniche e Biologiche, Universita` di Torino and Institute for Cancer Research and Treatment, Strada Provinciale 142, Km 3.95, 10060 Candiolo, Torino, Italy, and Dipartimento di Chimica Organica e Industriale, Universita` degli Studi di Milano, via Venezian 21, 20133 Milano, Italy Received June 18, 2010; Revised Manuscript Received August 13, 2010

In this paper we report on the investigation, as DNA nonviral carriers, of three samples of an amphoteric polyamidoamine bearing 4-aminobutylguanidine deriving units, AGMA5, AGMA10, and AGMA20, characterized j w 5100, 10100, and 20500, respectively). All samples condensed DNA in spherical, by different molecular weights (M positively charged nanoparticles and protected it against enzymatic degradation. AGMA10 and AGMA20 polyplexes had average diameters lower than 100 nm. AGMA5 polyplexes were larger. All polyplexes showed negligible cytotoxicity and were internalized in cells. AGMA10 and AGMA20 performed differently from AGMA5 as nucleic acid carriers in vitro. AGMA10 and AGMA20 effectively promoted transfection, whereas AGMA5 was ineffective. FITC-labeled AGMA10 was prepared and the intracellular trafficking of its DNA polyplex was studied. DNA/ AGMA10 polyplex was largely localized inside the nucleus, while AGMA10 concentrated in the perinuclear region. DNA/AGMA10 polyplex intravenously administered to mice promoted gene expression in liver but not in other organs without detectable toxic side effects.

Introduction In the last decades, several systems, including viral and nonviral carriers, have been developed to transfer genetic materials into cells with the aim of enhancing gene transfer. However, the scope of gene therapy is presently still limited by the scarcity of nucleic acid delivery systems acting safely and effectively not only in vitro, but also in vivo. The use of viral vectors involves recognized toxicity and immunogenicity risks, and therefore, nonviral vectors are becoming increasingly attractive. Among these, several cationic polymers have been considered, showing large loading capacity and remarkable transfection ability in vitro.1–4 However, some of them, such as for instance polylysine (PLL), polyethylenimine (PEI), and polyamidoamine dendrimers (PAMAM),5,6 are too toxic to be considered for in vivo applications. Other cationic polymers, such as for instance poly(β-aminoester)s,7 R,β-poly(asparthylhydrazide)- glycidyltrimethylammoniumchloride copolymers,8 and polyethylenimine-alt-poly(ethyleneglycol) copolymers,9 exhibit adequate safety profiles, but their transfection efficiency in vivo is disappointingly low due to rapid capture by the reticulo-endothelial system (RES) cells. Novel biodegradable poly(disulfideamine)s were also synthesized and tested as nonviral gene delivery carriers.10–13 Poly(amidoamine)s (PAAs) are synthetic biodegradable polymers that can be designed to be highly biocompatible. They are obtained by stepwise Michael-type polyaddition of primary or secondary amines to bisacrylamides and contain tert-amine and amide groups regularly arranged along the polymer chain. * To whom correspondence should be addressed. Phone: +39-0250314128. Fax: +39-02-50314129. E-mail: [email protected]. † Universita` degli Studi di Torino. ‡ Universita` di Torino and Institute for Cancer Research and Treatment. § Universita` degli Studi di Milano.

Many functionalized amines and bisacrylamides can be used as monomers resulting in an almost endless variety of polymer structures. PAAs were first described in 1970.14 Subsequently, their physicochemical and biological properties were reviewed at intervals.15–17 As a rule, PAAs are water soluble and all of them show different charge distribution profiles as a function of pH.18–20 In regard to biocompatibility, most PAAs exhibit LD50 values in vitro higher by 2 orders of magnitude than PLL, PEI, or PAMAM dendrimers.17 Amphoteric PAAs, carrying carboxyl groups as side substituents, can be as biocompatible as dextran.19 The ability of linear PAAs to promote the transfection in vitro of HEPG2 cells by pSV-β-galactosidase was first demonstrated for some PAAs.18 Subsequently, other PAAs have been considered as transfection promoters.21–25 More recently, PAAs carrying primary amine groups as side substituents26 and bioreducible PAAs containing disulfide linkages in the polymer chain were synthesized and evaluated as nonviral vectors.27 Up to now, however, no PAAs seem to have been reported as in vivo DNA carriers. Recently, we have considered a linear amphoteric, but prevailingly cationic PAA nicknamed AGMA1, obtained by polyaddition of 4-aminobutylguanidine (Agmatine) to 2,2bisacrylamidoacetic acid (Figure 1), as DNA complexing agent and transfection promoter in vitro.19,20 jw j n 4800, M The AGMA1 sample previously studied had M 7200, and PD 1.50. As transfection promoter, it gave positive results when tested in vitro with Hela cell line. AGMA1 was found to combine three key features essential as DNA nonviral vector: negligible toxicity, stealthlike behavior when injected in animals, and high DNA complexing ability.20 This peculiar combination of properties led us to investigate in detail the potential of AGMA1 as nonviral vector not only in vitro, but

10.1021/bm100685t  2010 American Chemical Society Published on Web 09/03/2010

2668

Biomacromolecules, Vol. 11, No. 10, 2010

Figure 1. Chemical structure of AGMA1.

also in vivo. First of all, we determined the molecular weight dependence of AGMA1 effectiveness as transfection promoter by investigating in vitro with Hela cell line three polymer samples, henceforth named AGMA5, AGMA10 and AGMA20 of the same structure as AGMA1 but with lower (AGMA5) and higher (AGMA10 and AGMA20) molecular weights. Then we tested the most convenient sample, namely AGMA10, as DNA carrier in vivo. The aim of this paper is to report on these issues.

Experimental Section Instruments and Methods. 1H and 13C NMR spectra were run on a Bru¨ker Advance 400 spectrometer operating at 400.132 (1H) and 100.623 (13C) MHz. Size exclusion chromatography (SEC) traces were obtained with a Knauer Pump 1000 equipped with a Knauer Autosampler 3800, TSKgel G4000 PW and G3000 PW TosoHaas columns connected in series, Light Scattering (LS) Viscotek 270 Dual Detector, UV detector Waters model 486, operating at 230 nm, and a refractive index detector Waters model 2410. The mobile phase was a 0.1 M Tris buffer pH 8.00 ( 0.05 with 0.2 M sodium chloride. The flow rate was 1 mL/min and sample concentration 1% w/w. Fluorimetric measurements were performed with an RF 551 Shimadzu fluorimeter. Materials. Ultrapure water was obtained using a 1-800 Milli-Q (Millipore, F) system. Fluorescein 5-isothiocyanate (FITC), ethidium bromide, and Etoposide were purchased from Fluka (CH). DNase I was supplied by Sigma (U.K.). Plasmidic DNA was purified with Qiagen Midiprep Kit. Agarose was purchased from BIO-RAD Laboratories S.r.l (Milan) and JetPEI from Polyplus-Transfection (Strasbourg, France). All other reagents (ACS grade) were from Sigma and were used as received. High-performance liquid chromatography (HPLC) solvents were from Carlo Erba (Italy). Synthesis of AGMA5, AGMA10, and AGMA20. The three linear PAA samples were prepared by a procedure similar to that previously reported for AGMA1.19 Briefly, Agmatine sulfate (2.000 g, 8.5 mmol) and lithium hydroxide monohydrate (0.360, 8.5 mmol) were added to a solution of BAC (1.689 g, 8.5 mmol) and lithium hydroxide monohydrate (0.360 g, 8.5 mmol) in distilled water (2.8 mL). This mixture was maintained under a nitrogen atmosphere and occasionally stirred for 48 (AGMA5), 78 (AGMA10), or 240 (AGMA20) hrs. After this time, it was diluted with water (2.8 mL), acidified with hydrochloric acid to pH 4-4.5, and then ultrafiltered through membranes with nominal cutoff 3000 (AGMA5), 5000 (AGMA10), and 10000 (AGMA20), respectively. The fractions retained in each case were freeze-dried and the product was obtained as a white powder. Yields: 2.1, 1.9, and 1.75 g. 1H NMR (D2O): δ (ppm) 1.61 (br, NHCH2CH2CH2), 1.76 (br, NHCH2CH2), 2.79 (br, NHCOCH2CH2), 3.19 (m, NHCH2, NCH2), 3.44 (br, NHCOCH2), 5.55 (s, CHCOOH). 13 C NMR (D2O): δµ (ppm) 22.3 (NHCH2CH2CH2), 25.0 (NHCH2CH2), 28.9 (NHCOCH2CH2), 40.4 (CH2NCH2), 49.1 (NHCOCH2), 52.5

Cavalli et al. (NHCH2), 56.0 (COOHCH), 155.1 (NH2CNNH), 171.3 (NHCO), 173.5 j n ) 3300, M jw ) (CHCOOH). Molecular weight values: AGMA5, M j w ) 10100, and PD ) j n ) 7800, M 4500, PD ) 1.38; AGMA10, M j n ) 16400, M j w ) 20500, and PD ) 1.25. 1.29; AGMA20, M Preparation of Fluorescent PAAs (FITC-AGMA10). Labeled AGMA10 (FITC-AGMA10) was prepared by treating with a FITC solution in methanol (0.2 mg/mL) a 10 mg/mL solution in buffer pH 7.4 of an AGMA10 sample (AGMA10-NH2) carrying amine groups as side substituents. In turn, AGMA10-NH2 was obtained as previously described20 by a general procedure established for preparing aminated PAAs. The resultant mixture was stirred overnight at room temperature and then centrifuged to eliminate insoluble impurities. The resultant clear solution was then dialyzed and the fluorescein-labeled polymer isolated by freeze-drying the retained portion. The recovery was practically quantitative. The conjugation of AGMA10-NH2 with FITC was confirmed by NMR and fluorescence microscopy. Characterization of the Polymers. ζ-Potential Determination. The ζ-potential values of AGMA5, AGMA10, and AGMA20 were determined in aqueous solutions at increasing pH values, ranging from 4.0 to 7.4, to verify the polymer charge distribution as function of the pH. A 90 Plus instrument (Brookhaven, NY) was used to determine the electrophoretic mobility and the ζ-potential of the three polymers. For the determinations, the aqueous solutions of the polymers were placed in the electrophoretic cell, where an electric field of about 15 V/cm was applied. Each value reported is the average of 10 measurements. The electrophoretic mobility measured was converted into ζ-potential using the Smoluchowsky equation.28 Biocompatibility Assessment. Hemolytic activity of AGMA5, AGMA10, and AGMA20 was studied on human blood. Increasing amounts of the polymers up to 15 mg/mL were added to a suspension of erythrocytes (30% v/v) in phosphate buffer pH 7.4 and then incubated for 90 min at 37 °C. A suspension containing only a 30% v/v of erythrocytes in phosphate buffer pH 7.4 was used as blank. Another suspension added with an excess of ammonium chloride was used to obtain complete hemolysis as 100% hemolytic control. After centrifugation at 2000 rpm for 5 min, the supernatants were analyzed using a Lambda 2 Perkin-Elmer spectrophotometer at a wavelength of 543 nm. The percentage of hemolysis was calculated versus the 100% hemolysis control. The cytoxicity of AGMA5, AGMA10, and AGMA20 was assessed on Hela cells seeded in 24-well plates. The cytotoxicity of the polymer samples was evaluated by MTT assay [MTT ) (3-(4,5dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide)]. The measurement was performed at 24 and 48 h postincubation. The viability of the treated cells was compared to untreated cells and to cells treated with Etoposide (0.1 µg/µL) as negative compound. Preparation of the DNA Polyplexes. The DNA/AGMA polyplexes were prepared at different pH values. PAA aqueous solutions (1.6 mg/ mL) were added under stirring to plasmidic DNA solutions (30 µg/ mL) in 20 mM HEPES buffer pH 7.4 or lower, adjusted by adding dilute hydrochloric acid. Different w/w DNA/polymer ratios, namely, 1:5, 1:10, 1:15, 1:20, and 1:30 were used for AGMA5, AGMA10, and AGMA20. In the case of AGMA5, two further ratios, that is, 1:90 and 1:100, were tested. In all cases, complexes were incubated for 30 min at room temperature before characterization. Size, ζ-Potential Determination, and Morphology of the DNA Polyplexes. The average diameters and polydispersity indices of the DNA/AGMA polyplexes were determined by photocorrelation spectroscopy (PCS) using a 90 Plus instrument (Brookhaven, NY) at a fixed angle of 90° and a temperature of 25 °C. Each reported value is the average of five measurements of three different samples. The electrophoretic mobility and ζ-potential were determined, as previously described in the case of polymers. The morphology of DNA polyplexes was determined by transmission electron microscopy (TEM). TEM analyses were carried out using a Philips CM10 instrument (Eindoven, NL). DNA/AGMA polyplexes in solution were dropped onto a Formwar-coated copper grid and dried before observation.

Polyamidoamines as Carriers for Plasmid DNA Stability of the DNA Polyplexes. The stability of the DNA polyplexes was evaluated over time and after freezing at -20 °C. At fixed time the sizes and morphology of DNA complexes were determined. Gel Retardation Assay. Gel electrophoresis assay was used to evaluate the formation of DNA/AGMA complexes. The complexes were subjected to electrophoresis on agarose gel (0.7% w/v) with ethidium bromide (0.25 µg/mL) for 1 h at 100 V to confirm the DNA complexation. The banding pattern was obtained using a UV transilluminator and photographed with a Polaroid camera. DNase Degradation Assay. To determine the complex ability to protect DNA from enzymatic degradation, DNA (10 µg/mL) and the DNA complexes (1 mg/mL) were incubated at 37 °C with DNase I (1000 units/mL) in phosphate buffer pH 7.4. At fixed times (up to 1 h), samples were withdrawn and centrifuged, and the supernatants were analyzed at 260 nm using a Perkin-Elmer Lambda II UV spectrophotometer. Results were expressed as a percentage of the control degradation (naked DNA). In Vitro DNA Transfection Studies. Transfection experiments were performed on Hela cell lines. Hela cells line was grown in DME (Cambrex) supplemented with 10% FCS, 2 mM L-glutamine (Cambrex), and antibiotics. Cells were transfected in 6-well plates at 4.5 × 105 cells per well. The DNA (pEGFP) and the polymers were preincubated in a 150 nM NaCl solution pH 5.5 for 30 min at room temperature. The DNA quantity used for transfections in 6- and 24-well plates were 3.5 and 0.7 µg, respectively. Before transfection, cell medium was replaced with serum-free medium and the complex of transfection was added to the culture medium for 3 h. Nuclear Localization of the DNA/AGMA Polyplexes. After transfection, the cells were washed, fixed with 4% PAF for 10 min, and stained with DAPI and then were mounted and analyzed with a confocal laser-scanning microscope (TCS SP2 with DM IRE2; Leica) equipped with 63X/1.40 HCX Plan-Apochromat oil-immersion objective. Confocal images are maximun projections of a z section of ∼3 µm. Nuclear localization was quantified with high resolution confocal images stacks reconstructed by isosurface rendering using Imaris software (version 6.2.0, Bitplane, AG). Isosurface is a computer generated representation of specified range of fluorescence intensities that allows the creation of an artificial solid object of a specific area. Cytotoxicity Studies. The cytotoxicity of the DNA/polymer complexes was evaluated by MTT assay. Cells were cultured for 24 and 48 h after transfection then 0.2 mg/mL of MTT in DMEM medium without Phenol Red (Lonza) was added. After 3 h of incubation with MTT, the supernatant was removed, and 200 µL of DMSO was added to dissolve the formazan crystal. Optical density value of each sample was measured at a wavelength of 595 nm. All experiments were performed in triplicate. In ViVo Administration of DNA/AGMA10 Polyplexes. Preliminary in vivo experiments were performed on 6-10 weeks old Balb/c mice. To four groups of six mice, 10 µg DNA (pEGFP) as 100 µL solution in 5% w/w glucose of 1:30 DNA/AGMA10 polyplex prepared at pH 6.0 was injected in the tail vein. Free DNA and AGMA10 solutions at the same concentrations were used as controls. Three mice of each group were sacrificed at 24 h and three at 48 h after the intravenous administration, and samples of liver, heart, kidney, and lungs were collected, washed with saline, and frozen. Then they were homogenized in PBS solution pH 7.4 using a Politron homogenizer before the evaluation of the in vivo gene expression. The animal experiments complied with the rules set forth in the NIH Guide for the Care and Use of Laboratory Animals. Western Blot Experiments. Total proteins were extracted in Laemmli buffer (62.5 mM Tris-HCl pH 6.8, 2% SDS, 10% glycerol) and quantified, and equal amounts of each sample were resolved by SDSPAGE and transferred to PVDF membrane. After blocking with TBS/ 0.1% Tween 20/5% BSA, membranes were incubated with primary antibody overnight at 4 °C. Primary antibodies used are: R-Akt (Cell Signaling), R-GFP (Molecular Probes), R-tubulin, and R-actin (Santa

Biomacromolecules, Vol. 11, No. 10, 2010

2669

Table 1. ζ-Potential Values of AGMA10 at Different pH Values ζ-Potential (mV) pH 4.0

pH 5.0

pH 6.0

pH 7.4

21.00 ( 1.16

15.32 ( 0.74

11.10 ( 0.52

2.10 ( 0.37

Cruz Biotechnology). Immunoreactive proteins were identified with secondary antibody coupled to horseradish peroxidase (HRP) antibody and visualized by ECL.

Results Polymers Preparation. Three samples of linear Agmatinecontaining PAAs of different molecular weights, AGMA5, AGMA10, and AGMA20, were synthesized and studied as intracellular DNA carriers on Hela cell line in vitro with the aim of determining the molecular weight dependence of their efficiency as transfection promoters. The most convenient sample, namely, AGMA10, was then tested as nonviral DNA carrier in vivo by intravenously administering to mice a nanosuspension of its polyplex nanoparticles. All samples were prepared by the same preparation recipe, that is, the Michael polyaddition of 4-aminobutylguanidine (Agmatine) to 2,2-bisacrylamidoacetic acid in water. The molecular weights of the samples were tuned by adopting different reaction times and different cutoff membranes for the final ultrafiltration step during isolation. In particular, AGMA5 j w 5100, PD 1.38, AGMA10 had M j n 7800, M jw j n 3700, M had M j n 16400, M j w 20500, PD 10100, PD 1.29, and AGMA20 had M 1.29. These values should be compared with those of the previously studied AGMA1 sample (see Introduction). Besides covering a wider range of molecular weights, the new samples studied in the present work have a significantly lower PD value, that is, are less polydisperse. A sample of AGMA10 containing a small percentage (4-5%) of units carrying a primary amine group as side substituent, named AGMA10-NH2, was prepared according to a general method for side-aminated PAAs, namely, by substituting a small percentage of monoprotected 1,2-diaminoethane for the same amount of Agmatine in the polymerization recipe19 and subsequently removing the protecting group. Labeling with FITC was then performed by a standard procedure.20 The conjugation of AGMA10 with FITC to give FITC-AGMA10 was confirmed by NMR and the efficiency of the labeling procedure determined by measuring the fluorescence intensity at λex ) 480 nm and λem ) 520 nm of a solution of FITC-AGMA10 of known concentration versus a standard FITC solution. The molecular weight values found for FITC-AGMA10 were very similar to those previously determined for AGMA10, showing that the labeling procedure did not induce any significant alteration of polymer properties. ζ-Potential Measurements. All samples were positively charged in aqueous solution at pH 7.4 and the positive charge increased by lowering the pH to 4.0. The ζ-potential values relative to AGMA10, which will be more deeply investigated both in vitro and in vivo, are reported in Table 1. The values obtained with AGMA5 and AGMA 20 were similar. This is in agreement with the expected charge distribution due to the pKa values of the ionic groups (carboxyl, amine, and guanidine) present in their structure, as previously determined.19 Cytotoxicity and Hemolytic Properties. No significant hemolytic activity was observed for all AGMA samples after 90 min incubation in blood at pH 7.4 up to a concentration of 15 mg/mL. Moreover, the three polymers showed no cytotoxic

2670

Biomacromolecules, Vol. 11, No. 10, 2010

Cavalli et al.

Table 2. ζ-Potential Values of DNA/AGMA Polyplexes at Different DNA/AGMA Ratios and pH Values ζ-potential (mV) code

DNA/AGMA ratio (w/w)

pH 4.0

pH 6.0

pH 7.4

DNA/AGMA5

1:5 1:15 1:30 1:5 1:15 1:30 1:5 1:15 1:30

+12.34 ( 0.77 +18.23 ( 0.81 +23,81 ( 1.25 +17.62 ( 0.63 +19.53 ( 0.81 +27.61 ( 1.25 +24.20 ( 1.46 +29.31 ( 0.92 +31.60 ( 1.07

+8.55 ( 0.65 +17.29 ( 0.60 +20.16 ( 1.11 +11.44 ( 0.77 +17.95 ( 0.60 +10.5 ( 1.91 +10.5 ( 1.91 +18.71 ( 0.41 +21.04 ( 1.74

+3.23 ( 1.0 +5.23 ( 0.7 +13.29 ( 0.88 +5.23 ( 0.70 +6.11 ( 0.46 +5.62 ( 0.86 +5.62 ( 0.86 +9.34 ( 0.54 +13.54 ( 0.74

DNA/AGMA10 DNA/AGMA20

effects on Hela cells after 48 h incubation up to a concentration of 7 mg/mL. Nucleic Acid Complexation. All the AGMA samples formed stable complexes with DNA in the entire pH range considered (4.0-7.4). The ζ-potential values of the DNA/AGMA polyplexes at different weight ratios are reported in Table 2. A moderate excess of AGMA (5:1) shifts the negative ζ-potential of DNA to positive values already at pH 7.4. The ability to complex DNA were confirmed by the strong electrophoretic retardation (Figure 2), showing the disappearance of the free DNA band for AGMA10 and AGMA20 and a strong reduction but not complete disappearance for AGMA5. Direct observation by TEM revealed that the DNA/AGMA polyplexes were obtained in the form of discrete nanospheres (Figure 3A-C), whose average size and polydispersity depended on the DNA/AGMA ratio, the molecular weight of the AGMA polymer (Table 3), and also the pH of the medium, as confirmed by PCS. The stability of the complex was confirmed also after freezing the complexes showing no changes in sizes and morphology. Compared with AGMA5, the higher molecular weight AGMA10 and AGMA20 exhibit a superior capability to condense DNA in small nanoparticles. At a 1:15 w/w DNA/ AGMA ratio, they formed nanoparticles with average diameters lower than 100 nm with a narrow size distribution at pH 5.0, whereas AGMA5 formed larger nanoparticles. AGMA5, AGMA10, and AGMA20 significantly protected DNA from degradation by DNase I. For instance, the protection afforded by 1:20 DNA/AGMA polyplexes is shown in Figure

Figure 2. Electrophoresis assay of DNA/AGMA polyplexes at different w/w ratios. From left to right: (1) DNA marker (white spots), (2) DNA (white spot), (3) DNA/AGMA10 1:20, (4) DNA/AGMA10 1:15, (5) DNA/ AGMA5 1:20, (6) DNA/AGMA5 1:15, (7) DNA/AGMA20 1:20, (8) DNA/AGMA20 1:15.

4. It may be noticed that, after 1 h, less than 10% of the DNA in the DNA/AGMA polyplexes was degraded, compared with 100% of free DNA. Gene Transfer Efficacy and Cytotoxicity of the Polyplexes. The ability of the AGMA5, AGMA10, and AGMA20 to act as a DNA carrier in vitro was assessed using a plasmidic DNA carrying GFP protein and analyzing the percentage of fluorescent cells by flow cytometric analysis. To establish the relationship between the molecular weight of AGMA5, AGMA10, and AGMA20 and their gene transfer ability, we carried out transfection experiments in Hela cells, a well-characterized cell line reputedly difficult to transfect. The gene delivery efficacy was first examined by fluorescence microscopy. Only a very limited number of fluorescent cells were obtained by transfecting Hela cells with DNA/AGMA5 polyplexes even at a DNA/ polymer ratio as of 1:100 (data not shown). By contrast, AGMA10 transfected cells showed a higher fluorescent cells number and a well spread cell morphology compared to cell transfected with commercial PEI. Cell transfected with PEI are shown in Figure 5B. It may be observed that PEI transfected cells were round and partially floating, suggesting a toxic effect exerted by the DNA/PEI complex. To confirm this, the viability of the transfected cells was evaluated by measuring the activity of mitochondrial dehydrogenase with MTT substrate (Figure 5C). Compared to untreated cells, the DNA/AGMA10 complex reduced cell viability by about 5-10% and DNA/PEI by about 40% (Figure 5C). AGMA20 gave approximately the same results as AGMA10 (data not shown). Because, however, its preparation involves much longer reaction times and reduced yields compared with that of AGMA10, we decided to pursue our investigations only on the latter. The transfection efficiency of AGMA10 was quantified using two different w/w DNA/polymer ratios (1:10 and 1:30) and determining the fluorescent cell number 24 or 48 h after transfection. A good transfection efficiency was already obtained at 24 h. The percentage of GFP positive cells was of 35 and 71% with DNA/AGMA10 ratios of 1:10 and 1:30, respectively (Figure 5A). Intracellular Trafficking and Nuclear Localization. The intracellular trafficking of FITC-AGMA10 and FITC-labeled DNA/AGMA10 was analyzed by means of confocal microscopy. The results are shown in Figure 6. AGMA10 and DNA/ AGMA10 polyplex, both internalized inside cells after 30 min treatment, differently localized in cell compartments. In particular, it is apparent from pictures shown in Figure 6B that FITC-AGMA10 predominantly concentrated in the perinuclear region, while FITC-labeled DNA/AGMA10 was largely localized inside the nucleus (clearly visualized by blue staining in Figure 6A). The different localization of labeled AGMA10 and DNA/AGMA10 is quantified in Figure 6C. In Vivo Transfection Experiment. AGMA10 was selected as a nonviral DNA vector in vivo. In particular, a dose of 1:30

Polyamidoamines as Carriers for Plasmid DNA

Biomacromolecules, Vol. 11, No. 10, 2010

2671

Figure 3. TEM microphotograph of (A) 1:15 w/w DNA/AGMA5 polyplex at pH 6.0, (B) 1:15 w/w DNA/AGMA10 polyplex, (C) 1:15 w/w DNA/ AGMA20 polyplex, and (D) 1:15 w/w DNA/AGMA5 polyplex at pH 6.0 after freeze-drying. Table 3. Average Diameter and Polydispersity Index of DNA/AGMA Polyplexes at Different pH Values pH 5.0 code DNA/AGMA5 DNA/AGMA10 DNA/AGMA20

pH 7.4

DNA/AGMA ratio (w/w)

avg diameter (nm)

polydispersity index

avg diameter (nm)

polydispersity index

1:5 1:15 1:30 1:5 1:15 1:30 1:5 1:15 1:30

108.2 ( 3.2 189.3 ( 4.7 197.5 ( 4.3 85.2 ( 5.6 78.9 ( 1.2 72.0 ( 2.3 107.3 ( 5.9 95.0 ( 3.4 89.5 ( 7.9

0.26 0.30 0.30 0.10 0.12 0.13 0.33 0.18 0.18

450 ( 10.0 310 ( 5.0 333 ( 21.3 161 ( 15.4 158 ( 9.4 148 ( 0.52 271.9 ( 20.8 115.0 ( 10.8 94.6 ( 22.3

0.23 0.22 0.15 0.20 0.18 0.21 0.25 0.16 0.21

DNA/AGMA10 polyplex nanoparticles corresponding to 10 µg DNA (pEGFP) was injected in the tail vein of mice, as reported

in the Experimental Section. The Western blot results of the in vivo gene expression in animals sacrificed 24 and 48 h after treatment are reported in Figure 7. The results showed a remarkable gene expression in the liver, suggesting liver localization of the DNA/AGMA10 polyplex, whereas in the lung, as well as in the heart and kidneys, no gene expression was detected.

Discussion

Figure 4. Enzymatic degradation of DNA and DNA/AGMA polyplexes with time.

Three samples of an amphoteric Agmatine-containing PAA of the same structure, but different molecular weights AGMA5, AGMA10, and AGMA20, were tested as nucleic acid carriers. A previous study of the acid-base properties of this particular PAA with the consequent calculation of its charge distribution profiles as a function of pH had shown

2672

Biomacromolecules, Vol. 11, No. 10, 2010

Cavalli et al.

Figure 7. Western blot analysis of total lysates from livers and lungs of mice i.v. injected with the DNA/AGMA10 polyplex. Animals were sacrificed 24 and 48 h post-treatment and the organs processed for protein extraction. Each point represents a sample from a single animal.

Figure 5. Evaluation of AGMA10 and AGMA5 efficiency as transfection promoters. (A) Cytofluorimetric analysis of Hela cells transfected with DNA/AGMA10 complexes with different w/w ratios. The transfection efficiency is expressed by the GFP positive cell number. (B) Fluorescent micrographs of Hela cells transfected with 1:30 w/w DNA/AGMA10, and DNA/PEI polyplexes. (C) Viability of transfected cells evaluated by MTT assay. Viability of 100% is referred to untreated cells.

Figure 6. Intracellular trafficking and nuclear localization of FITCAGMA10 and FITC-labeled DNA/AGMA10. (A) Nucleus is stained by DAPI (blue). Labeled AGMA10 and DNA/AGMA10 polyplex are visualized (green). (B) The confocal images are rendered to enhance the nuclear localization and the mean pixel intensity of different cell regions is quantified and reported in the graph. Experiments were performed in triplicate. At least 50 cells were analyzed for each experiment. (C) Quantification of the different localizations of labeled AGMA10 and DNA/AGMA10.

that it is prevailingly cationic at all physiological pH values, having, for instance, an average of 0.55 positive charges per unit at pH 7.4.19,20 This feature is hardly influenced by the molecular weight of the sample, as it depends primarily from the chemical groups present in the repeating units, and moreover, it is typical of PAAs having practical independence of each unit in regard to acid-base properties.15,17 The samples considered in the present study were first characterized in terms of ζ-potential, cytotoxicity, and hemolytic activity. They combined positive charge with negligible cytotoxicity and hemolytic activity up to a concentration of 7 and 15 mg/mL, respectively. This combination of properties makes these polymers rather exceptional among polycations, which are most often toxic,29-31 and lead us to think of them as valuable candidates as nucleic acid carriers for gene

delivery, not only in vitro, but also in vivo. All the samples were able to complex DNA at different weight ratios and pH values forming nanoparticles with spherical shape, but the lower molecular weight sample, AGMA5, under the conditions adopted formed larger polyplex nanoparticles than AGMA10 and AGMA20, suggesting a weaker or incomplete binding ability. In particular, AGMA10 and AGMA20 polyplexes had sizes lower than 100 nm, which varied according to the DNA/polymer ratio but did not increase over time nor after freezing and did not change morphology. This size is considered suitable for cell internalization both in vitro and in vivo. In fact, it is generally recognized that polyplexes with dimensions in the range 200-300 nm are gene transfer agents effective in vitro, but unsuitable for in vivo use, whereas 150 nm represents the highest threshold size value for nonspecific cell uptake in vivo. Upon complexation, all samples were able to protect DNA against enzymatic degradation. The protection capacity confirmed the encapsulation of the nucleic acids by the polymer chains. To establish the relationship between the molecular weight of the samples and their gene transfer ability we first carried out transfection experiments in vitro on Hela cells, a well characterized and difficult to transfect cell line that represents a model for molecular medicine studies. The ability of AGMA5, AGMA10, and AGMA20 to form polyplexes and act as a DNA carrier was assessed using a plasmidic DNA carrying GFP protein and analyzing the percentage of fluorescent cells by flow cytometric analysis. The results clearly highlighted a remarkable difference in transfection capacity. AGMA10 and AGMA20 showed good transfection ability with little difference between the two and were already effective at a 1:5 DNA/polymer w/w ratio. By contrast, AGMA5 did not show a good transfection activity even at a 1:100 w/w DNA/polymer ratio. Indeed, the transfection efficiency of AGMA10 in vitro was remarkably good, as revealed by the high number of fluorescent cells observed by fluorescence microscopy and flow cytometric analysis, which compares favorably with that obtained using Jet PEI, a well-known commercial transfection promoter, whose molecular weight is not disclosed. The DNA/AGMA10 polyplexes exhibited over Jet PEI the advantage of a much lower cytotoxicity. It may be mentioned that linear low molecular weight PEI (10 kDa) has been reported as an efficient gene transfer agent with a lower toxicity compared with high molecular weight PEI.32 However, it has also been reported that direct intravenous dosing of linear PEI can be lethal.33 The poor efficiency of AGMA5 as transfection promoter was related to incomplete DNA encapsulation, leading to large and

Polyamidoamines as Carriers for Plasmid DNA

unstable polyplexes. A similar behavior was reported for PLL. Short PLL peptides were able to weakly bind and condense plasmid DNA forming particles greater than 300-800 nm. Increasing the PLL chain length leads to DNA condensation in nanoparticles with diameter lower than 100 nm.34 Comparing the in vitro data obtained with the three AGMA samples studied in the present work with the data previously reported for AGMA1,20 we noticed a lack of significant improvement in transfection efficiency passing from AGMA10 to AGMA20, a little improvement passing from AGMA1 to AGMA10, but a dramatic improvement passing from AGMA5 to AGMA10. Comparing with the molecular weights of the same j w 4500; AGMA1: M j n 4800, M jw j n 3300, M samples (AGMA5: M j n 7800, M j w 10100; AGMA20: M j n 16400, 7200; AGMA10: M j w 20500), there is little doubt that the molecular weight plays M a paramount role in determining the transfection efficiency but with qualifications. We did not explore the behavior of AGMA samples with molecular weight higher than 20000, but our data strongly suggest that a rather sharp molecular weight threshold exists for the AGMA polyamidoamines to act as transfection promoters. With low to moderate PD values, this threshold can be placed between molecular weight 7000 and 10000. Amphoteric PAAs other than AGMA have been reported to undergo in aqueous solution some conformational changes by increasing the molecular weight, with consequent variations in the Mark-Houwink-Sakurada constants.35 This might occur also in the case of AGMA polyamidoamines and explain the existence of a molecular weight threshold for a stable polyplex formation. Owing to the above considerations, we selected AGM10 for further studies and in vivo transfection experiments. The success of nonviral gene therapy has been largely limited by inefficient gene delivery in vivo due to the presence of multiple extracellular and intracellular barriers, of which the major, not yet successfully overcome, is the nuclear envelope. Following endocytosis, to enter the nucleus, DNA must escape the endosomal vesicle and become “freed” into the cytoplasm. Our experiments suggested that AGMA10 remained complexed, with the DNA transporting it to the nucleus without “sinking” in the endosomal or lysosomal vesicles. Preliminary intracellular trafficking studies on FITClabeled AGMA10, and its DNA polyplex showed, in fact, that FITC-labeled DNA/AGMA10 was largely localized inside the nucleus, while FITC-AGMA10 predominantly concentrated in the perinuclear region. AGMA10 appeared to have all the qualifications needed by a carrier for effective gene delivery in vivo. A preliminary experiment was designed to substantiate this assumption. A single dose of 1:30 DNA/ AGMA10 polyplex nanoparticles in buffer pH 6.0, corresponding to 10 µg DNA (pEGFP), was intravenously administered to mice that were subsequently sacrificed 24 and 48 h after treatment. GFP was unequivocally expressed in the liver cells, whereas under the same conditions naked DNA was totally ineffective. No transfection evidence was found in all the other tissues examined, including lungs, suggesting that liver is the primary localization site of DNA/ AGMA10 polyplexes. This might imply that the Kupffer cells present in the liver capture DNA/AGMA10 polyplexes. In addition, liver has a fenestrated epithelium, which might permit the extravasation and accumulation of the nanoparticles. We have no data, at present, to ascertain which mechanism of capture prevails. The absence of gene expression in the lung might indicate that the DNA/AGMA10 polyplexes circulating in blood were not entrapped in the

Biomacromolecules, Vol. 11, No. 10, 2010

2673

lung capillaries notwithstanding their positive surface charge. This was rather unexpected. Broadly speaking, in fact, electropositive nanoparticles are supposed to distribute mainly in the lung after i.v. administration to mice36,37 because they attract and bind electronegative proteins such as serum albumin, quickly grow in size, and, consequently, become liable to be physically entrapped in the lung capillary beds. In previous animal experiments, a soluble Agmatine-containing PAA, though prevailingly cationic, showed a prolonged circulation time and a stealth-like behavior after intravenous injection to rats.16 It would seem that these properties are maintained to some extent by the DNA/AGMA10 nanoparticles, because the in vivo data obtained in this work would be difficult to explain unless considering that DNA/AGMA10 polyplexes do not extensively bind negative proteins and are sufficiently stable in the bloodstream to avoid DNA dissociation and degradation by DNase. The intravenous administration of DNA/AGMA10 polyplexes to mice did not cause any detectable toxic effect and, not unexpectedly, the same was true for free AGMA10. The acute toxicity of free AGMA10 in mice is being determined by Intox Laboratories (Urawade, India), although this study is not yet complete, the first results indicate that the maximum tolerated dose after intravenous administration is higher than 500 mg/kg. This is an outstandingly low value for an effective DNA carrier and warrants further investigations on AGMA10 for gene transfection in vivo.

Conclusions Three samples of an amphoteric Agmatine-containing PAA, namely, AGMA5, AGMA10, and AGMA20, having the same chemical structure but different the molecular weights, were studied as DNA carriers in vitro and, limited to AGMA10, in vivo. They were water-soluble and prevailingly positive at all physiological pH values. Notwithstanding, they lacked significant toxicity on Hela cells and hemolytic activity on human red blood cells. All of them proved able to complex DNA forming stable nanoparticles with positively charged surfaces, but the size of DNA/AGMA10 and DNA/AGMA20 polyplexes were smaller than those of DNA/AGMA5. The polyplex nanoparticles proved highly biocompatible when tested on Hela cells. As well as plain AGMA10, they were easily cell internalized and largely localized inside the nucleus. The molecular weight of the polymer samples had a paramount influence on their efficiency as DNA carriers. The lowest molecular weight sample, AGMA5, was ineffective. Oppositely, the higher molecular weight samples, AGMA10 and AGMA20, were highly effective in vitro. AGMA10 was chosen for animal experiments and proved effective also in vivo. In fact, the 1:30 DNA/AGMA10 polyplex after intravenous administration to mice induced remarkable gene expression in the liver but not in other organs, including lungs. Toxicity determinations pointed to a remarkable biocompatibility of AGMA10 and its polyplexes. It may be concluded that AGMA10, and probably also other AGMA polyamidoamines with a molecular weight above a threshold estimated between 7000 and 10000, apparently satisfies the requirements of a nonviral gene delivery carrier and holds a definite potential for in vivo application.

References and Notes (1) De Smedt, S. C.; Demeester, J.; Hennink, W. E. Pharm. Res. 2000, 17, 113–126. (2) Luten, J.; Van Nostrum, C. F.; De Smedt, S. C.; Hennink, W. E. J. Controlled Release 2008, 126, 97–110.

2674

Biomacromolecules, Vol. 11, No. 10, 2010

(3) El-Aneed, A. J. Controlled Release 2004, 94, 1–14. (4) Hung, W. C.; Shau, M. D.; Kao, H. C.; Shih, M. F.; Cherng, J. Y. J. Controlled Release 2009, 133, 68–76. (5) Patil, S. D.; Rhodes, D. G.; Burgess, D. J. AAPS J. 2005, 7, E61E77. (6) Lynn, D. M.; Anderson, D. G.; Putnam, D.; Langer, R. J. Am. Chem. Soc. 2001, 123, 8155–8156. (7) Pedone, E.; Cavallaro, G.; Richardson, S.; Duncan, R.; Giammona, G. J. Controlled Release 2001, 77, 139–153. (8) Parck, M. R.; Han, K. O.; Han, I. K.; Cho, M. H.; Nah, J. W.; Choi, Y. J.; Cho, C. S. J. Controlled Release 2005, 105, 367–380. (9) Shi, S.; Guo, Q. F.; Kann, B.; Fu, S. Z.; Wang, X. H.; Gong, C. Y.; Deng, H. X.; Luo, F.; Zhao, X.; Wei, Y. Q.; Qian, Z. Y. BMC Biotechnol. 2009, 9, 65. (10) Ou, M.; Wang, X. L.; Xu, R.; Chang, C. W.; Bull, D. A.; Kim, S. W. Bioconjugate Chem. 2008, 19, 626–633. (11) Gao, K.; Huang, L. Mol. Pharm. 2008, 6, 651–658. (12) Convertine, A. J.; Benoit, D. S.; Duvall, C. L.; Hoffman, A. S.; Stayton, P. S. J. Controlled Release 2009, 133, 221–229. (13) Jeong, J.; Christensen, V.; Lane, V:; Yockman, J. W.; Zhong, Z.; Engbersen, J. F. J.; Kim, W. J.; Feijen, J.; Kim, S. W. Biomaterials 2007, 28, 1912–1917. (14) Danusso, F.; Ferruti, P. Polymer 1970, 11, 88–113. (15) Ferruti, P.; Marchisio, M. A.; Barbucci, R. Polymer 1985, 26, 1336– 1348. (16) Ferruti, P. In Polymeric Materials Encyclopedia; Salamone, J. C., Ed.; CRC Press, Inc.: Boca Raton, FL, 1996; Vol. 5, pp 33343359. (17) Ferruti, P.; Marchisio, M. A.; Duncan, R. Macromol. Rapid Commun. 2002, 23, 332–355. (18) Richardson, S. C. W.; Pattrick, N. G.; Stella Man, Y. K.; Ferruti, P.; Duncan, R. Biomacromolecules 2001, 2, 1023–1028. (19) Franchini, J.; Ranucci, E.; Ferruti, P.; Rossi, M.; Cavalli, R. Biomacromolecules 2006, 7, 1215–1222. (20) Ferruti, P.; Franchini, J.; Bencini, M.; Ranucci, E.; Zara, G. P.; Serpe, L.; Primo, L.; Cavalli, R. Biomacromolecules 2007, 8, 1498– 1504.

Cavalli et al. (21) Hill, I. R.; Garnett, M. C.; Bignotti, F.; Davis, S. S. Biochim. Biophys. Acta 1999, 1427, 161–174. (22) Jones, N. A.; Hill, I. R.; Stolnik, S.; Bignotti, F.; Davis, S. S.; Garnett, M. C. Biochim. Biophys. Acta 2000, 1517, 1–18. (23) Hill, I. R.; Garnett, M. C.; Bignotti, F.; Davis, S. S. Anal. Biochem. 2001, 291, 62–68. (24) Rackstraw, B. J.; Stolnik, S.; Bignotti, F.; Garnett, M. C. Biochim. Biophys. Acta 2002, 1576, 269–286. (25) Garnett, M. C.; Ferruti, P.; Ranucci, E.; Suardi, M.; Heyde, M.; Sleat, R. Biochem. Soc. Trans. 2009, 37, 713–716. (26) Min, L.; Chen, J.; Cheng, Y. P.; Xue, Y. N.; Zhuo, R. X.; Huang, S. W. Macromol. Biosci. 2010, 10, 384–392. (27) Christensen, V.; Chang, C. W.; Kim, W. J.; Kim, S. W.; Zhong, Z.; Lin, C.; Engbersen, J.; Feijen, J. Bioconjugate Chem. 2006, 17, 1233– 1240. (28) Sze, A.; Erickson, D.; Ren, L.; Li, D. J. Colloid Interface Sci. 2003, 261, 402–410. (29) Hongtao, L.; Shubiao, Z.; Bing, W.; Shaohui, C.; Jie, Y. J. Controlled Release 2006, 114, 100–109. (30) Bos, G. W.; Kanellos, T.; Crommelin, D. J.; Hennink, W. E.; Howard, C. R. Vaccine 2004, 23, 460–469. (31) Gary, D. J.; Puri, N.; Won, Y. Y. J. Controlled Release 2007, 121, 64–73. (32) Fisher, D.; Bieber, T.; Li, X.; Elasser, H.; Kissel, T. Pharm. Res. 1999, 16, 1273–1279. (33) Chollet, P.; Favrot, M. C.; Hurbin, A.; Coll, J. L. J. Gene Med. 2002, 4, 84–91. (34) Wadhwa, M. S.; Collard, W. T.; Adami, R. C.; McKenzie, D. L.; Rice, K. G. Bioconjugate Chem. 1997, 8, 81–88. (35) Ferruti, P.; Manzoni, S.; Richardson, S. C. W.; Duncan, R.; Pattrick, N. G.; Mendichi, R.; Casolaro, M. Macromolecules 2000, 33, 7793– 7800. (36) Li, S.; Rizzo, M. A.; Bhattacharya, S.; Huang, L. Gene Ther. 1998, 5, 930–997. (37) Collard, W. T.; Yang, Y.; Kwok, K. Y.; Park, Y.; Rice, K. G. J. Pharm. Sci. 2000, 89, 499–512.

BM100685T