Poly-l-glutamic Acid Derivatives as Multifunctional Vectors for Gene

Biomacromolecules 2003, 4, 1168-1176

1168

Poly-L-glutamic Acid Derivatives as Multifunctional Vectors for Gene Delivery. Part A. Synthesis and Physicochemical Evaluation Peter Dubruel, Luc Dekie, and Etienne Schacht* Polymer Materials Research Group, Department of Organic Chemistry, Ghent University, Ghent 9000 Belgium Received January 13, 2003; Revised Manuscript Received May 2, 2003

This paper describes the synthesis and evaluation of a series of multifunctional poly-L-glutamic acid derivatives that can be used as vectors for gene delivery. They readily form polyelectrolyte complexes with DNA, resulting in a reduced surface charge and size of the DNA. The formation of a polymer-DNA complex and the stability toward serum albumin was analyzed by ethidium bromide fluorescence measurements and agarose gel retardation studies. Most polymers, except those with more than 80% imidazoles, are able to condense calf thymus DNA, thus forming complexes with sizes varying between 105 and 172 nm. The surface charge of the complexes was determined at different charge ratios by zeta potential measurements. The buffering properties of the polymers were determined via titration studies. The results show that the polymers are able to buffer the endosomal environment, although to a smaller extent than polyethyleneimine. The first part of this study is devoted to the synthesis and the physicochemical evaluation of the multifunctional polymers and their use as carriers for genetic information. The second part, to be published subsequently, discusses the biological evaluation of the polymers and their complexes with DNA. 1. Introduction When a mutation occurs in a specific gene, it can stop the production of a vital protein, start the synthesis of a malfunctioning protein, or start the production of lifethreatening proteins.1 In each case, this will affect the function of a tissue or organ, which in turn can cause disease or death. At present, most therapies for genetic diseases are based on the treatment of the symptoms of the disease, rather than its cause. Unfortunately in many cases, these nongenetic manipulations usually have a modest effect on the disease. Gene therapy has been proposed as an alternative for these therapies. The first scientific publications discussing the use of gene therapy for experiments in humans date from 1966.2 Nevertheless, one had to wait until 1990 before the first rather successful gene therapy experiment was performed on a four year old girl suffering from adenosine deaminase deficiency (ADA).3,4 Other diseases that are currently under investigation are cystic fibrosis,5 muscular dystrophy,6 and hypercholesterolemia. However, gene therapy is not limited to genetic diseases. This approach can also be used for a number of acquired diseases having a genetic component. Examples are cancer7,8 and acquired immunodeficiency syndrome (AIDS).9-11 In theory, gene therapy is easy: the right gene has to be delivered into the right type of cell. However, finding a way that allows efficient gene transfer is not straightforward. * To whom correspondence should be addressed. Prof. Dr. E. H. Schacht, Krijgslaan 281 (S4 Bis), B-9000 Ghent, Belgium. Phone: 0032-(09)2644497. Fax: 0032-(0)9-2644998. E-mail: [email protected].

During evolution, the human body has developed all kinds of defense mechanisms to protect itself from surrounding dangers, including the introduction of foreign genetic material into its genome. Nucleases are present that are able to digest DNA.12 In case the genetic material can circumvent this hurdle and thus reach the target cell, it is usually incompetent to enter the cell. Therefore, it is necessary to combine the genetic material with a carrier, a vector. The vector has to protect the DNA from nucleases and target it to specific cells or tissues. Once the vector system is inside the cell, the vector assists in the introduction of the DNA in the nucleus and the subsequent expression of the gene.13 The use of viruses would be natural because they are capable of penetrating cells. In addition, they are also able to use the hosts’ transcription mechanism to synthesize their viral proteins.14 The limitations of viral vectors, specifically their immunogenicity and small capacity for therapeutic DNA, has led to the development of synthetic nonviral vectors. Nonviral vectors have the advantage over viral ones that they can be constructed and modified to carry a broad range of properties. In addition, they can be adopted for mass production. Poly(L-lysine) (pLL),15 polyethyleneimine (pEI),16 and poly(2-(dimethylamino)ethyl-methacrylate) (pDMAEMA)17-19 have been widely investigated. The results have shown that it is possible to use these polyamines as gene delivery vectors. They readily form polyelectrolyte complexes with DNA,20 resulting in a reduced size and surface charge and an enhanced stability toward nucleases. pEI has, under physiological conditions, only 20% of its amine functions protonated. A further 25% is protonated in the endosomal

10.1021/bm034014j CCC: $25.00 © 2003 American Chemical Society Published on Web 07/10/2003

Part A. Synthesis and Physicochemical Evaluation

Figure 1. Chemical structure of the multifunctional vectors by aminolysis of poly(γ-benzyl-L-glutamate) or poly(γ-trichloroethyl-Lglutamate) by different amine containing molecules.

compartment where the pH drops to about 5.5 going from the early endosomes to the late endosomes. This results in some kind of endosomal buffering effect that enables the complexes to be released into the cytoplasm.21 The main disadvantages of these polymers are their relatively high toxicity and, in case of pEI and pDMAEMA, the fact that they are not biodegradable, two properties that might be important for repeated in vivo applications. The aim of this study is the synthesis of multifunctional biodegradable poly-L-glutamic acid derivatives that can be used as vectors for gene therapy. Their DNA condensation properties were determined by ethidium bromide (EtBr) exclusion tests and agarose gel electrophoresis. The size and charge of the complexes were determined via photon correlation spectroscopy and zeta potential measurements, respectively. The buffering properties of the polymers were determined by titration studies. Finally, the stability of the complexes in the presence of serum albumin was also studied. 2. Materials and Methods All cationic polymers were synthesized by aminolysis of poly-γ-benzyl-L-glutamate or poly-γ-trichloroethyl-L-glutamate (Figure 1). In this way, polymers were prepared having only tertiary amines or a mixture of tertiary amines with either primary amines, guanidines, or imidazole functions. Furthermore, polymers were prepared containing only imidazole functions or copolymers containing imidazole functions and guanidine groups. The molecular weight of the polymers, prepared by NCA polymerization, was controlled by the monomer-to-initiator ratio. 2.1. Chemicals. Glutamic acid (Bachem), 2-dimethylaminoethylamine (Fluka), histamine (Fluka), agmatine sulfate (Acros), benzylchloroformiate (Aldrich), p-methoxybenzenesulfonyl chloride (Alkemi), bovine serum albumin (Sigma), branched polyethyleneimine (Aldrich) Mw ) 25 000, and 2-hydroxypyridine (Acros), highly polymerized calf thymus DNA sodium salt (Sigma) were used as such. Dimethylformamide, chloroform, and dichloromethane were dried over calciumhydride and distilled. Ethyl acetate was dried over potassium carbonate and distilled after filtration. Tetrahydrofuran was dried over sodium and distilled. Methanol was

Biomacromolecules, Vol. 4, No. 5, 2003 1169

dried over calcium oxide and distilled. Cyclohexane was dried over magnesium sulfate. The chemical characterization of the different prducts was carried out by 1H NMR spectroscopy and IR spectroscopy. 2.2. Preparation of the Monomers. 2.2.1. Synthesis of the N-Carboxyanhydride DeriVatiVe of γ-Benzyl-L-glutamate. γ-Benzylglutamate was prepared by the method of Gutmann and Boissonnas.22 The corresponding N-carboxyanyhydride (NCA) was prepared by reaction of the ester with diphosgene in ethyl acetate and recrystallized from ethyl acetate as described in the literature.23 In a typical experiment, γ-benzyl-L-glutamate (5 g, 0,0211 mol) was suspended in dry ethyl acetate (50 mL). The suspension was heated in an oil bath at 60 °C, and a solution of diphosgene in EtOAc (5% v/v) was added in portions of 5-10 mL. After approximately 15 min of reaction, nitrogen was flushed through the suspension for about 2 min to remove the HCl formed. This procedure was repeated until only a small amount of γ-benzyl glutamate was left. The oil bath was removed, and the solution was flushed with nitrogen for 30 min. The suspension was filtered off and the filtrate was concentrated under reduced pressure to about 75 mL. Dry hexane (15 mL) was added slowly until crystallization started, and the mixture was then put in the freezer for 1 h. The white crystalline product was filtered off and washed with cyclohexane. After drying, the product was recrystallized from EtOAc until the product was sufficiently chloride-free. 2.2.2. Synthesis of the N-Carboxyanhydride DeriVatiVe of γ-Trichloroethyl-L-glutamate. γ-Trichloroethyl-L-glutamate was prepared by the method described by Ishikawa and Endo.24 The N-carboxyanhydride derivative was synthesized via a similar method as for the N-carboxyanhydride derivative of γ-benzyl-L-glutamate. Tetrahydrofuran was used as solvent during this reaction. 2.3. Polymer Synthesis. 2.3.1. Synthesis of Poly(γ-benzylL-glutamate). Poly(γ-benzyl-L-glutamate) (pBG) was synthesized via polymerization of the corresponding NCA in an ethyl acetate/dichloromethane mixture (1/6), initiated by tributylamine at room temperature. The reaction was followed by IR spectroscopy. The polymer was obtained after precipitation in a methanol/diethyl ether (2/1) mixture, filtered off, washed with diethyl ether (3×), and dried under vacuum.25 The molecular weight of the polymer was determined by viscosimetry measurements in dichloroacetic acid, using the following equation:26,27 η ) 2.78 × 10-5Mv0.87. Polymers with a molecular weight of 91 × 103 and 102 × 103 Da were prepared. 2.3.2. Synthesis of Poly(γ-trichloroethyl-L-glutamate). Poly(γ-trichloroethyl-L-glutamate) (pTCEG) was synthesized via polymerization of the corresponding NCA in chloroform, initiated by 2-triphenylmethylaminoethylamine at room temperature. The reaction was followed by IR spectroscopy. The polymer was obtained after precipitation in pentane and dried under vacuum. The molecular weight of the polymer was determined by 1 H NMR in deuterated trifluoroacetic acid taking into account the trityl headgroup protons. Polymers with a molecular weight of 26 × 103 and 36 × 103 Da were prepared.

1170

Biomacromolecules, Vol. 4, No. 5, 2003

Dubruel et al.

Figure 2. Master curve for the aminolysis of pTCEG with histamine and 2-dimethylaminoethylamine. This curve represents the ratio of the aminolysis reagents (tertiary amine/imidazole) versus the ratio of both compounds in the copolymers obtained as determined by 1H NMR spectroscopy.

2.3.3. Synthesis of Poly(dimethylaminoethyl-L-glutamine). Poly(dimethylaminoethyl-L-glutamine) was prepared by aminolysis of pBG (Mv ) 91 × 103 Da). In a typical experiment, pBG (1 g, 4.57 mmol esters) was dissolved in 10 mL of dimethylformamide (DMF) and heated in an oil bath at 40 °C. 2-Dimethylaminoethylamine (10.02 mL, 91.3 mmol) and 2-hydroxypyridine (2.17 g, 22.8 mmol) were added and the solution was stirred for 48 h. The reaction was followed by 1H NMR and IR spectroscopy. The product was obtained after precipitation in cold diethyl ether. Finally, the product was dissolved in 0.1 N HCl, dialyzed against water (2 days), and freeze-dried. The molecular weight, as determined by MALLS, was 59 × 103 Da, d (Mw/Mn) was 1.05. 2.3.4. Synthesis of Poly(dimethylaminoethyl-L-glutamine)85%co-poly(aminoethyl-L-glutamine)15%. Poly(dimethylaminoethyl-L-glutamine)85%-co-poly(aminoethyl-L-glutamine)15% was prepared by aminolysis of pBG (Mv ) 102 × 103 Da). In a typical experiment, pBG (1 g, 4.57 mmol esters) was dissolved in 10 mL of DMF and heated in an oil bath at 50 °C. 2-Dimethylaminoethylamine (7 mL, 63.9 mmol), 2-triphenylmethylaminoethylamine (6.9 g, 22.8 mmol), and 2-hydroxypyridine (2.17 g, 22.8 mmol) were added, and the solution was stirred for 72 h. The reaction was followed by 1 H NMR and IR spectroscopy. The product was obtained after precipitation in cold diethyl ether. The trityl group was removed via acidolysis in trifluoroacetic acid (20 mL) during 1.5 h at room temperature. The final product was obtained after precipitation in cold diethyl ether. Finally, the product was dissolved in 0.1 N HCl, dialyzed against water (2 days), and freeze-dried. The molecular weight as determined by MALLS was 77 × 103 Da, d was 1.11. The chemical composition, as determined by 1H NMR, was obtained by comparing the signals from the methyl protons (2.8 ppm) with the R-CH proton (4.2 ppm).

2.3.5. Synthesis of Poly(histamino-L-glutamine). Histamine (512 mg, 4.6 mmol) was dissolved in 4 mL DMF. pTCEG (Mn ) 26 × 103 Da, 300 mg, 1.15 mmol esters) and 2-hydroxypyridine (546 mg, 5.75 mmol) were added and the solution was stirred overnight at 10°C. The product was obtained after precipitation in cold diethyl ether. Finally, the product was dissolved in 0.1 N HCl, dialyzed against water (2 days), and freeze-dried. The molecular weight as determined by 1H NMR was 22 × 103 Da. 2.3.6. Synthesis of Poly(dimethylaminoethyl-L-glutamine)co-poly(histamino-L-glutamine). The composition of the polymers was controlled by the 2-dimethylaminoethyl-amine/ histamine ratio (Figure 2). The synthesis of poly(dimethylaminoethyl-L-glutamine)82%co-poly(histamino-L-gluta-mine)18% is given as an example. pTCEG (210 mg, 0.807 mmol esters) was dissolved in 3 mL of DMF. 2-Dimethylaminoethylamine (265 µl, 1.6 mmol), histamine (90 mg, 0.807 . mmol), and 2-hydroxypyridine (384 mg, 4.04 mmol) were added and the solution was stirred overnight at 10 °C. The product was obtained after precipitation in cold diethyl ether. Finally, the product was dissolved in 0.1 N HCl, dialyzed against water (2 days), and freeze-dried.The chemical composition, as determined by 1H NMR, was obtained by comparing the signals from the R-CH protons from the imidazole group (4.05 ppm) and the tertiary amine function (4.1 ppm). The molecular weight and the composition of the different copolymers is shown in Table 1. 2.3.7. Synthesis of Guanidine Containing Polymers. 2.3.7.1. Synthesis of Benzyloxycarbonyl Agmatine. Agmatine sulfate (3 g, 13.14 mmol) was dissolved in 10 mL of cold 2 N NaOH. The solution was kept at 0 °C while alternately benzylchloroformiate (3.75 mL, 26.3 mmol) and 2 N NaOH were added. During the addition, the pH of the solution was kept between 9 and 10. The solution was stirred for another


Part A. Synthesis and Physicochemical Evaluation Table 1. Composition and Molecular Weight of the Poly(dimethylaminoethyl-L-glutamine)-co-poly(histamino-L-glutamine) Derivatives molar composition polymer

Mw (Da)

% tertiary amines

pDMAEG82%-pHisG18% pDMAEG64%-pHisG36% pDMAEG35%-pHisG65% pDMAEG24%-pHisG76% pDMAEG12%-pHisG88%

23 × 23 × 103 32 × 103 32 × 103 31 × 103

82 64 35 24 12

103

% imidazole 18 36 65 76 88

2 h, during which the pH dropped to 7-7.5. The resulting precipitate was filtered off, washed with distilled water and diethyl ether, and dried under vacuum. 2.3.7.2. Protection of the Guanidine Function of Benzyloxycarbonyl Agmatine. Benzyloxycarbonyl agmatine (0.5 g, 1.9 mmol) was dissolved in a mixture of 2 mL of 4 N NaOH and 16 mL of aceton and cooled to 0 °C. A solution of p-methoxybenzenesulfonyl chloride (800 mg, 3.8 mmol) in 4 mL aceton was added dropwise while the solution was intensively stirred.28 The solution was stirred for another 2 h at 0 °C and overnight at room temperature. The mixture was then acidified with citric acid. The solvent was removed under reduced pressure. The residue was redissolved in EtOAc and extracted with distilled water. EtOAc was removed under reduced pressure, and the product was further purified via column chromatography (eluent ) ethyl acetate, Rf ) 0.32). 2.3.7.3. Deprotection of the Benzyloxycarbonyl Group. The product synthesized in paragraph 2.3.7.2 (0.5 g) was dissolved in 20 mL methanol. Palladium on activated charcoal (250 mg) was added, and the mixture was placed under a hydrogen atmosphere. The deprotection was followed by TLC. The catalyst was filtered off over Celite, and the solvent was removed under reduced pressure. 2.3.8. Synthesis of Poly(dimethylaminoethyl-L-glutamine)83%co-poly(agmatino-L-glutamine)17%. 2-Dimethylaminoethylamine (126 µl, 0.76 mmol), the agmatine derivative (196 mg, 0.644 mmol), and 2-hydroxypyridine (109.6 mg, 1.15 mmol) were dissolved in 4 mL of DMF. pTCEG (Mn ) 36 × 103 Da, 300 mg, 1.15 mmol esters) was added, and the solution was stirred overnight at 10 °C. The product was obtained after precipitation in cold diethyl ether. Finally, the product was dissolved in 0.1 N HCl, dialyzed against water (2 days), and freeze-dried. The p-methoxybenzenesulfonyl group was removed in a methanesulfonic acid/anisol mixture.28 The product (130 mg) was dissolved in 4.5 mL of methanesulfonic acid/anisol (20/ 1) and stirred for 5 h at room temperature. The product was obtained after precipitation in cold diethyl ether. Finally, the product was dissolved in 0.1 N HCl, dialyzed against water (2 days), and freeze-dried. The chemical composition, as determined by 1H NMR, was obtained by comparing the signals from the methyl protons (2.8 ppm) with the R-CH proton (4.2 ppm). The molecular weight as determined by 1H NMR (see § 2.3.2.) was 24 × 103 Da. 2.3.9. Synthesis of Poly(histamino-L-glutamine)83%-co-poly(agmatino-L-glutamine)17%. Histamine (128 mg, 1.15 mmol),

Table 2. Mass Per Charge and IC50 of the Cationic Polymers Studied polymer

mass/charge

IC50a

pDMAEG pDMAEG85%-pAEG15% pDMAEG82%-pHisG18% pDMAEG64%-pHisG36% pDMAEG35%-pHisG65% pDMAEG24%-pHisG76% pDMAEG12%-pHisG88% pHisG pDMAEG84%-pAgmG16% pHisG73%-pAgmG27%

235 231 233 230 226 225 224 222 241 237

0.77:1 0.8:1:1 0.8:1 0.82:1 1.3:1

0.92:1

a The IC 50 is the charge ratio (+:-) causing 50% reduction of EtBr fluorescence.

the agmatine derivative (115 mg, 0.383 mmol), and 2-hydroxypyridine (109.6 mg, 1.15 mmol) were dissolved in 4 mL of DMF. pTCEG (Mn ) 36 × 103 Da, 300 mg, 1.15 mmol esters) was added, and the solution was stirred overnight at 10 °C. The product was obtained after precipitation in cold diethyl ether. Finally, the product was dissolved in 0.1 N HCl, dialyzed against water (2 days), and freezedried. The p-methoxybenzenesulfonyl group was removed by the method described in paragraph 2.9.4. The molecular weight as determined by 1H NMR (see section 2.3.2) was 24 × 103 Da. 2.4. Assembly of Complexes at Different Charge Ratios. For calculation of the charge ratio, a mass per charge of 325 was used for DNA. The mass per charge of the polymer was calculated assuming 100% protonation of all amines (Table 2). These assumptions are not fully correct, but they allow us to estimate the mass per charge in a consistent way. Each sample contained 20 µg/mL of CT DNA and the appropriate amount of polymer to obtain the desired charge ratio. The total sample volume was 2 mL. The complexes were allowed to self-assemble in water and were left at room temperature for at least half an hour before use. 2.4.1. Measurement of the Complex Formation. 2.4.1.1. EtBr Exclusion Tests. The ability of the new polymers to form complexes with DNA was assessed by measuring the change in EtBr-DNA fluorescence. Complex formation is known to be accompanied with a loss of fluorescence (λex ) 510 nm, λem ) 590 nm).29 A sample containing 20 µg/mL CT DNA and 400 ng/mL EtBr was used to calibrate the fluorimeter to 100% fluorescence against a background of 400 ng/mL EtBr in a Perkin-Elmer fluorescence luminescence spectrometer LS 50B. The polymers were added stepwise so that 10 additions were needed to obtain a final charge ratio of 2:1 (+:-). The samples were allowed to stabilize before the fluorescence was measured. All tests were performed on the same day and in 3-fold. 2.4.1.2. Agarose Gel Retardation Studies. Condensation of DNA by cationic polymers results in a reduced mobility of the DNA in an agarose gel. Complete condensation leads to retention of the complexes in the wells of the gel.30 The polymer-DNA complexes were prepared at different charge ratios and analyzed on a 1% (w/v) agarose gel (tris-

1172


borate/EDTA buffer) for 1 h at 90 V. A 0.5 µg/mL EtBr solution was used to visualize the DNA (1 h incubation) and the gels were photographed using a transilluminator (Gibco BRL UV Transluminator TFX-20M, Life Technologies, 312 nm) and a digital camera (Kodak DC120). 2.4.2. Stability of the Complexes in the Presence of BoVine Serum Albumin. The stability of the complexes was measured by agarose gel retardation studies. The complexes were prepared at different charge ratios, and bovine serum albumin was added to a final concentration of 5, 25, and 50 mg/mL. The DNA release was determined after 1 and 24 h, both at room temperature and at 37 °C. 2.4.3. Zeta Potential Measurements. Zeta potential measurements were performed on a Zetamaster (Malvern Instruments Ltd.) equipped with a 10 mW laser. Complexes were prepared half an hour before performing the tests at a 0.5:1, 1:1, 2:1, and 4:1 (+:-) charge ratio. Each sample contained 20 µg/mL CT DNA. The samples were measured 5 times at 1000 Hz and 220 V. The apparatus was calibrated with a -55 mV standard (carboxylated polystyrene latex, Malvern). 2.4.4. Photon Correlation Spectroscopy. PCS analysis was performed on a Zetamaster (Malvern Instruments Ltd.) equipped with a 10 mW laser. The sampling time was set to 5 min and each run consisted of 15 subruns. The measurements, carried out in 3-fold, were done at 25 °C at an angle of 90°. The complexes used in the experiments were made at least half an hour before performing the tests. Each sample contained 20 µg/mL of DNA and polymer in a 2:1 (+:-) charge ratio. 2.4.5. Determination of the Buffering Properties of the Multifunctional Polymers. The pH area in which the polymers have a buffering effect was determined via titration studies. The amount of polymer that is present in complexes prepared at a 60:1 (+:-) charge ratio was dissolved in 10 mL of CO2-free water. The pH of the solution was brought to 10, using a CO2-free NaOH solution. The resulting solution was titrated with a CO2-free 0.01 N HCl solution. The titration studies were performed under nitrogen atmosphere. 3. Results and Discussion 3.1. Complex Formation Monitored by the Loss of EtBr Fluorescence and Agarose Gel Retardation Studies. Ethidium bromide is a planar fluorescent dye that is used to render nucleic acids fluorescent. The dye is able to intercalate between the base pairs of the DNA, more specifically in the minor groove of the DNA helix.31-33 Polycations, including those synthesized in this work, are able to interact with the negatively charged phosphate groups of DNA, in this way excluding EtBr from the DNA minor groove. The loss of EtBr fluorescence was used to monitor DNA condensation by interaction with cationic polymers, resulting in the formation of polyelectrolyte complexes. The condensation profiles are shown in Figure 3. The results show that all polymers, except those with more than 70% imidazole functions, are able to condense DNA. Condensation occurs between a 0.8:1 and a 1.3:1 (+:-) charge ratio. In the case of the polymers containing more than 70% imidazole groups, it can be seen that the fluorescence value at a 2:1 (+:-)

Dubruel et al.

Figure 3. Inhibition profiles of DNA/EtBr fluorescence for the different cationic polymers. The tests were performed on the same day, n ) 3. Measurements were perfomed at λex ) 510 nm and λem ) 590 nm. A sample containing 20 µg/mL of CT DNA and 400 ng/mL of EtBr was used to calibrate the fluorimeter to 100% fluorescence.

charge ratio is more than 50. This clearly indicates that these polymers do not form tight complexes with DNA at a 2:1 (+:-) charge ratio. The charge ratio needed to obtain 50% inhibition of fluorescence (IC50) is shown in Table 2. pDMAEG, pDMAEG85%-pAEG15%, pDMAEG82%-pHisG18%, and pDMAEG64%-pHisG36% have a similar IC50 value (( 0.8:1). An increase in the imidazole content results in an increase of the IC50 value of the corresponding polymer. The fluorescence of the complexes based on the polymers with more than 65% imidazole groups does not drop below 50% at a 2:1 (+:-) charge ratio. This is due to the fact that the imidazole groups are partially protonated under the experimental conditions in comparison with tertiary amines, primary amines, or guanidine functions. The determination of the real fraction of protonated amines is difficult. It is known that pKa values change when neighboring amine functions are protonated34 or when they interact with the phosphate groups of DNA. pDMAEG84%-pAgmG16% shows a higher IC50 than pDMAEG85%-pAEG15%. This might be due to the longer distance between the charged guanidine group and the main polymer chain in comparison with the polymers containing tertiary and primary amines, resulting in a weaker electrostatic interaction between the polymer and the DNA. Agarose gel electrophoresis was used to investigate if the polymers were able to form polyelectrolyte complexes with DNA at higher charge ratios (>2:1 (+:-)). pHisG73%pAgmG27% is able to condense DNA at a 4:1 (+:-) charge


Figure 4. Agarose gel electrophoresis of imidazole containing polymers at different charge ratios (pH ) 7.5). The polymer-DNA complexes were prepared at different charge ratios and analyzed on a 1% (w/v) agarose gel for 1 h at 90 V. A 0.5 µg/mL EtBr solution was used to stain the DNA. (A) pHisG, (B) pDMAEG12%-pHisG88%, (C) pDMAEG24%-pHisG76%, (D) pDMAEG35%-pHisG65%, (E) pDMAEG64%-pHisG36%, (F) pHisG73%-pAgmG27%, (G) pDMAEG.

Figure 5. Agarose gel electrophoresis of imidazole containing polymers at different charge ratios and pH. The polymer-DNA complexes were prepared at different charge ratios and analyzed on a 1% (w/v) agarose gel for 1 h at 90 V. A 0.5 µg/mL EtBr solution was used to stain the DNA. (A) pDMAEG35%-pHisG65%, (B) pDMAEG24%-pHisG76%, (C) pDMAEG12%-pHisG88%, (D) pHisG.

ratio, whereas the other polymers are not able to do so (Figure 4). In Figure 5, the results of the DNA condensation at higher charge ratio (5:1 until 8:1 (+:-)) are shown. In the same figure, the effect of the pH on the condensation is also shown. From the results, it can be concluded that pDMAEG35%-pHisG65% and pDMAEG24%-pHisG76% condense DNA at a 5:1 and 6:1 (+:-) charge ratio, respectively. Polymers with a higher imidazole content do not seem to be able to condense the DNA or the formed complexes are so weak that they dissociate in the applied electric field.35


Figure 6. Stability of complexes based on pDMAEG (A) and pDMAEG35%-pHisG65% (B) after 1 h incubation with serum albumin (calbumin solution ) 100 mg/mL). The polymer-DNA complexes were prepared at different charge ratios, incubated with serum albumin, and analyzed on a 1% (w/v) agarose gel for 1 h at 90 V. A 0.5 µg/mL EtBr solution was used to stain the DNA.

Complexes that are formed at pH ) 6.5 condense the DNA at a lower charge ratio and to a higher extent than those formed at pH ) 7.5. This can be explained by the fact that, at lower pH, more groups are charged. These are able to interact with the DNA to a larger extent. 3.2. Stability of the Complexes in the Presence of Bovine Serum Albumin. In in ViVo gene therapy, vectors are brought into contact with the blood of patients or animals. Therefore, it is very important that the complexes do not interact with the blood components. Albumin is the most abundant protein in the blood (30-50 g/L) and was therefore used as a model protein to investigate the stability of the complexes toward serum proteins. The interaction of albumin with the complexes was investigated with agarose gel electrophoresis. The complexes were prepared at different charge ratios and bovine serum albumin was added to a final concentration of 5, 25, and 50 mg/mL. The DNA release was assessed after 1 and 24 h. Results are shown in Figures 6 and 7. It can be concluded that albumin is unable to release any DNA from the complexes. 3.3. Zeta Potential Measurements. The membrane of most cells is negatively charged. This implies that the interaction between complexes and cell membranes will depend on the surface charge (zeta potential) of the complexes and thus their charge ratio. The charge of the polymer-DNA complexes was determined by zeta potential measurements. The technique is based on the measurement of the electrophoretic mobility of a particle in an electric field.36 This can be determined by laser doppler velocimetry. The results are shown in Table 3. The complexes, based on polymers with 65 molar % imidazole groups and more, show a negative zeta potential at a 2:1 (+:-) charge ratio. These results are in accordance with the DNA condensation studies where it was shown that polymers with a high imidazole content are unable to properly condense DNA, even at high charge ratio. As already mentioned, this is due to the low protonation degree of the imidazole groups under the experimental conditions. The complexes based on pD-

1174


Dubruel et al.

Table 3. Zeta Potential of the Various Polymer-DNA Complexes at Different Charge Ratiosa

a

polymer

0.5:1 (+:-) (mV)

1:1 (+:-) (mV)

2:1 (+:-) (mV)

4:1 (+:-) (mV)

pDMAEG pDMAEG82%-pHisG18% pDMAEG64%-pHisG36% pDMAEG35%-pHisG65% pDMAEG24%-pHisG76% pDMAEG12%-pHisG88% pHisG pDMAEG84%-pAgmG16% pHisG73%-pAgmG27% pDMAEG85%-pAEG15%

-14.7 ( 2.3 -14.7 ( 1.8 -15.6 ( 0.9 -20.6 ( 0.4 -19.3 ( 1.8 -14.9 ( 2.6 -22.9 ( 1.1 -14.3 ( 0.8 -22.1 ( 2.3 -16.6 ( 1.8

23.8 ( 0.8 12.2 ( 0.4 -0.9 ( 0.1 -16.5 ( 1.2 -15.6 ( 3.7 -17.8 ( 1.6 -23.2 ( 0.7 19.6 ( 0.7 -20.2 ( 2.3 22.6 ( 4.1

24.8 ( 2.5 22.5 ( 0.7 19.9 ( 1.4 -7.2 ( 0.5 -13.3 ( 1.6 -16.6 ( 1.9 -20.7 ( 1.3 31.9 ( 1.1 -3.2 ( 0.3 23.7 ( 3.3

27.1 ( 1.8 26.3 ( 2.9 23.4 ( 3.2 11.6 ( 0.5 -3.1 ( 0.3 -11.3 ( 0.4 -18.4 ( 0.5 32.7 ( 1.9 -1 ( 0.3 31.4 ( 2.4

The samples were measured 5 times at 1000 Hz and 220 V.

Table 4. Particle Size of Polymer-DNA Complexes as Studied by Photon Correlation Spectroscopya polymer pDMAEG pDMAEG82%-pHisG18% pDMAEG64%-pHisG36% pDMAEG35%-pHisG65% pDMAEG24%-pHisG76% pDMAEG12%-pHisG88% pHisG pDMAEG84%-pAgmG16% pHisG73%-pAgmG27% pDMAEG85%-pAEG15% a

Figure 7. Stability of complexes after 24 h incubation with serum albumin at room temperature and 37°C (calbumin solution ) 100 mg/mL). The polymer-DNA complexes were prepared at different charge ratios, incubated with serum albumin, and analyzed on a 1% (w/v) agarose gel for 1 h at 90 V. A 0.5 µg/mL EtBr solution was used to stain the DNA. A ) pDMAEG-DNA complexes; B ) pDMAEG65%pHisG35%-DNA complexes.

MAEG84%-pAgmG16% show a higher zeta potential than those composed of pDMAEG. This can be explained by the more pronounced positive charge of the guanidine functions (pKa ) 12.5) in comparison with the tertiary amines (pKa ) 10). 3.4. Particle Size Determination via Photon Correlation Spectroscopy. For adequate receptor mediated endocytosis, it is necessary that the particles are smaller than 150-200 nm.37 Photon correlation spectroscopy was used to determine

size (nm) 120 ( 2.0 119 ( 4.0 137 ( 2 165 ( 4 172 ( 1 219 ( 2 193 ( 0 (92%) 896 ( 10 (8%) 117 ( 3 1725 ( 6 105 ( 1

The complexes were formed at a 2:1 (+:-) charge ratio.

the size of the complexes (Table 4). From the results, it can be seen that the introduction of imidazole groups leads to an increase in the size of the complexes. The smallest complexes are formed with pDMAEG85%-pAEG15% (105.1 ( 1.1 nm). This can be explained by the fact that the primary amines that are present in this polymer allow a better interaction with the DNA in comparison with the tertiary amines (less steric hindrance, higher protonation degree). pHisG and pHis73%-pAgmG27% formed the largest complexes. The former polymer forms complexes with a bimodal size distribution of 193 ( 0 nm (92%) and 896 ( 10 nm (8%). The large size is due to the fact that only a few of the imidazole functions are protonated, so that there is only a weak interaction between the polycation and the DNA. The complexes based on pHis73%-pAgmG27% have a size of 1725 ( 63 nm. This is probably due to the fact that the zeta potential of the complexes is close to neutrality what results in an aggregation of the complexes. 3.5. Buffering Capacity of the Various Polymers. Up to now, the best synthetic vector is pEI. This is probably due to the high amount of amine functions present in the polymeric chain. Only a small portion is protonated under physiological conditions.38 The remaining amines can help in the buffering of the endosomal pH. The early endosomes have a pH of 7.2. Just before they fuse with the lysosomes, the pH drops to around 5. Because of the buffering of the endosomes, an osmotic effect occurs that causes the endosomes to swell and ultimately burst (proton sponge theory).16,39 By incorporating functional groups with different pKa, we tried to prepare polymers that are, like pEI, capable of



Figure 8. Titration curves of the imidazole containing polymers. An amount of polymer that is present in complexes prepared at a 60:1 (+:-) charge ratio was dissolved in 10 mL of CO2-free water. The solution was titrated with a CO2-free 0.01 N HCl solution. The titration studies were performed under nitrogen atmosphere. As reference, the titration curves of pEI and pDMAEG are given in the same graph. Table 5. Buffering Properties of the Various Multifunctional Polymers polymer

buffering area

pDMAEG pDMAEG82%-co-pHisG18% pDMAEG64%-co-pHisG36% pDMAEG35%-co-pHisG65% pDMAEG24%-co-pHisG76% pDMAEG12%-co-pHisG88% pHisG pDMAEG84%-pAgmG16% pHisG73%-pAgmG27% pDMAEG85%-pAEG15%

7.7-5.7 7.6-5.6 7.5-5.6 7.3-5.4 7.3-5.4 7.2-5.3 7.1-5.2 7.7-5.8 7.2-5.2 7.7-5.7

buffering the endosomal pH and increase the endosomal release of the polymer-DNA complexes. Titration studies were performed to determine the pH area where the polymers buffer.40 During the analyses, an equal amount of amines was dissolved in distilled and CO2-free water. The solutions were titrated with a 0.01 N CO2-free HCl solution (Figure 8). The results show that none of the synthesized polymers buffer as well as pEI (buffers between pH ) 11 and 4). pDMAEG buffers between pH 7.8 and 5.8. Incorporation of pyridine groups did not improve the buffering properties of the polymers (pKa is too low, data not shown). The incorporation of imidazole functions changed the buffering region of the polymers. For instance, pDMAEG35%pHisG65% buffers the pH between 7.3 and 5.4, whereas pHisG buffers between 7.1 and 5.2 (Table 5). One can conclude that the lower pKa value of the imidazole function compared to that of the tertiary amines causes the buffering properties of the polymers to shift to a lower pH area. pDMAEG84%pAgmG16% and pDMAEG85%-pAEG15% have similar buffering properties as pDMAEG. 4. Conclusions A number of multifunctional polymers were prepared via aminolysis of poly-γ-benzyl-L-glutamate and poly-γ-trichlo-

roethyl-L-glutamate. Under physiological conditions, these polymers are protonated to a certain extent depending on their chemical structure. When mixed with a DNA solution, they spontaneously form polyelectrolyte complexes. The size of these complexes depended on the polymer used. The complexes were stable toward bovine serum albumin. Titration studies were performed to determine the buffering capacities of the various polymers. It was shown that the polymers buffer the medium to a certain extent between pH 7.5 and 5, although to a smaller extent than pEI. These results indicate that the multifunctional polymers are suitable carriers for DNA. The biological evaluation of the polymers and polymer-DNA complexes will be discussed in a subsequent paper. Acknowledgment. The authors thank the Flemish institute for the promotion of Scientific-Technological Research in Industry (IWT), the Fund for Scientific Research-Flanders (FWO), the Belgian Ministry of Scientific Programming, IUAP/PAI-V, and the European Union’s Biotechnology Program Contract Number 97 2334 with support from INCO Contract Number IC 20 CT 970005. Abbreviations pLL: poly(L-lysine) pEI: polyethyleneimine pBG: poly(γ-benzyl-L-glutamate) pTCEG: poly(γ-trichloroethyl-L-glutamate) NCA: N-carboxyanhydride pDMAEG: poly(dimethyl-aminoethyl-L-glutamine) pDMAEG-pAEG: poly(dimethylamino-ethyl-L-glutamine)co-poly(aminoethyl-L-glutamine) pDMAEG-pHisG: poly(dimethyl-aminoethyl-L-glutamine)co-poly(histamino-L-glutamine) pHisG: poly(histamino-L-glutamine) pDMAEG-pAgmG: poly(dimethylaminoethyl-L-glutamine)co-poly(agmatino-L-glutamine) pHisG-pAgmG: poly(histamino-L-glutamine)-co-poly(agmatino-L-glutamine) DMF: dimethylformamide PCS: photon correlation spectroscopy

1176


MALLS: multi angle laser light scattering GPC: gel permeation chromatography EtBr: ethidium bromide CT DNA: calf thymus DNA Mw: weight average molecular weight Mn: number average molecular weight Mv: molecular weight as determined by viscosimetry TLC: thin-layer chromatography.

References and Notes (1) Clark, W. R. In The New Healers: the Promise and Problems of Molecular Medicine in the Twenty-First Century; Oxford University Press: New York, 1997. (2) Lederberg, J. Am. Nat. 1966, 100, 519-531. (3) Anderson, W. F. Science 1992, 256, 5058, 808-813. (4) Blease, R. M.; Culver, K. W.; Miller, A. D.; Carter, C. S.; Fleisher, T.; Clerici, M.; Shearer, G.; Chang, L.; Chiang, Y.; Tolstoshev, P.; Greenblatt, J. J.; Rosenberg, S. A.; Klein, H.; Berger, M.; Mullen, C. A.; Ramsey, W. J.; Muul, L.; Morgan, A.; Anderson, W. F. Science 1995, 270, 475-480. (5) Dorin, J. R.; Dickinson, P.; Alton, E. W. F. W.; Smith, S. N.; Geddes, D. M.; Stevenson, B. J.; Kimber, W. L.; Fleming, S.; Clarke, A. R.; Hooper, M. L.; Anderson, L.; Beddington, R. S. P.; Porteous, D. J. Nature 1992, 359 (6392), 211-215. (6) Martin, P. In The development of gene therapy in Europe and the United States: a comparative analysis (Brussels: European Union), 1996. (7) Boulikas, T. Int. J. Oncol. 1996, 9 (6), 1239-1251. (8) Sugaya, S.; Fujita, K.; Kikuchi, A.; Ueda, H.; Takakuwa, K.; Kodama, S.; Tanaka, K. Human Gene Therapy 1996, 7 (2), 223-230. (9) http://www.aidsmeds.com. (10) Yu, M.; Leavitt, M. C.; Maruyama, M.; Yamada, O.; Young, D.; Ho, A. D.; Wongstaal, F. Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 699-703. (11) Riddell, S. R.; Greenberg, P. D.; Overell, R. W.; Loughran, T. P.; Gilbert, M. J.; Lupton, S. D.; Agosti, J.; Scheeler, S.; Coombs, R. W.; Corey, L. Human Gene Therapy 1992, 3 (3), 319-338. (12) Chiou, H. C.; Tangco, M. V.; Levine, S. M.; Robertson, D.; Kormis, K.; Wu, C. H.; Wu, G. Y. Nucleic Acids Res. 1994, 22 (24), 54395446. (13) De Ron, A.; Vandendriessche, T.; Khim, M. C. Nat. Tech. 1998, 66 (20), 21-29. (14) Smith A. E. Ann. ReV. Microbiol. 1995, 49, 807-838. (15) Dash, P. R.; Toncheva, V.; Schacht, E.; Seymour, L. W. J. Controlled Release 1997, 48 (2-3), 269-276. (16) Abdallah, B.; Hassan, A.; Benoist, C.; Goula, D.; Behr, J. P.; Demeneix, B. A. Human Gene Therapy 1996, 7, 1947-1954.

Dubruel et al. (17) van de Wetering, P.; Cherng, J. Y.; Talsma, H.; Crommelin, D. J. A.; Hennink, W. E. J. Controlled Release 1998, 53 (1-3), 145-153. (18) Dubruel, P.; Toncheva, V.; Schacht, E. H. J. Bioact. Compat. Polym. 2000, 15 (3) 191-213. (19) Dubruel, P.; Schacht, E. H. J. Bioact. Compat. Polym. 2000, 15 (4) 279-296. (20) Kabanov, A. V., Kabanov, V. A. Bioconjugate Chem. 1995, 6 (1), 7-20. (21) Behr, J. P. Chimia 1997, 51 (1-2), 34-36. (22) Boissonnas, R. A.; Guttmann, St.; Huguenin, R. L.; Jaquenoud, P. A.; Sandrin, E. HelV. Chim. Acta 1958, 41, 1867-1882. (23) De Marre, A.; Soyez, H.; Schacht, E.; Pytela, J. Polymer 1994, 35 (11), 2443-2446. (24) Ishikawa, K.; Endo, T. J. Am. Chem. Soc. 1988, 110 (6), 20162017. (25) Dekie, L.; Toncheva, V.; Dubruel, P.; Schacht, E. H.; Barrett, L.; Seymour, L. W. J. Controlled Release 2000, 65 (1-2), 187-202. (26) Doty, P.; Bradbury, J. H.; Holtzer, A. M. J. Am. Chem. Soc. 1956, 78, 947-954. (27) Spach, G.; Freund, L.; Daune, M.; Benoit, I. J. Mol. Biol. 1963, 7, 468-451. (28) Nishimura, O.; Fujino, M. Chem. Pharm. Bull. 1976, 24 (7), 15681575. (29) Wolfert, M. A.; Seymour, L. W. Gene Therapy 1996, 3 (3), 269273. (30) Wolfert, M. A.; Schacht, E. H.; Toncheva, V.; Ulbrich, K.; Nazarova, O.; Seymour, L. W. Human Gene Therapy 1996, 7 (17), 2123-2133. (31) Haugland, R. P. In Handbook of Fluorescent Probes and Research Chemicals; Molecular Probes Inc.: Eugene, OR, 1996. (32) Waring, M. J. J. Mol. Biol. 1965, 13, 269-282. (33) LePecq, J. B.; Paoletti, C. J. Mol. Biol. 1967, 27, 87-106. (34) Nagaya, J.; Homma, M.; Tanioka, A.; Minakata, A. Biophys. Chem. 1996, 60, 45-51 (35) Ruponen, M.; Yla-Herttuala.; S, Urtti, A. Biochim. Biophys. ActaBiomembr. 1999, 1415 (2), 331-341. (36) Schurz, J.; Erk, G.; Schempp, W.; Ribitsch, V. J. Macromol. Sci.Chem. 1990, A27 (13-14), 1673-1692. (37) Perales, J. C.; Ferkol, T.; Molas, M.; Hanson, R. W. Eur. J. Biochem. 1994, 226 (2), 255-266. (38) Boussif, O.; Lezoualc’h, F.; Zanta, M. A.; Mergny, M. D.; Scherman, D.; Demeneix, B.; Behr, J. P. Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 7297-7301. (39) Remy, J. S.; Behr, J. P. J. Liposome Res. 1996, 6:3, 535-544. (40) Tang, M. X.; Szoka, F. C. Gene Therapy 1997, 4:8, 823-832.

BM034014J

Poly-l-glutamic Acid Derivatives as Multifunctional Vectors for Gene

Recommend Documents